Searching for deviations from the General Relativity Theory with X-ray surface

brightness observations and complementary probes

R. F. L. Holanda1,2,3 , S. H. Pereira4, and S. Santos-da-Costa5

arXiv:1612.09365v1 [astro-ph.CO] 30 Dec 2016

Departamento de Fsica,
Universidade Estadual da Paraba,
58429-500, Campina Grande - PB, Brasil,
2
Departamento de Fsica,
Universidade Federal de Campina Grande,
58429-900, Campina Grande - PB, Brasil,
3
Departamento de Fsica,
Universidade Federal do Rio Grande do Norte,
59300-000, Natal - RN, Brasil.
4
Universidade Estadual Paulista (Unesp)
Faculdade de Engenharia, Guaratinguet
a
Departamento de Fsica e Qumica
Av. Dr. Ariberto Pereira da Cunha 333
12516-410 Guaratinguet
a, SP, Brazil
5
Observat
orio Nacional, 20921-400,
Rio de Janeiro - RJ, Brasil.
Nowadays, thanks to the improved precision of cosmological data, it has been possible to search
for deviation from general relativity theory with tests on large cosmic scales. Particularly, there
is a class of modified gravity theories that breaks the Einstein equivalence principle (EEP) in the
electromagnetic sector, generating variations of fundamental constants and violations of the cosmic
distance duality relation and the evolution law of the cosmic microwave background (CMB) radiation. In recent papers, this class of theories has been tested with angular diameter distance of
galaxy clusters, type Ia supernovae and CMB temperature.
In this work, we propose a new test by considering the most recent X-ray surface brightness
observations of galaxy clusters jointly with type Ia supernovae and CMB temperature. Particularly,
we show how luminosity distances can be obtained from current X-ray gas mass fractions if the EEP
fails. Our basic result is that this new approach is competitive with the previous one and also does
not show significant deviations from general relativity.

I.

INTRODUCTION

Recently, several models of modified gravity theories have been appeared in order to deal with some
problems that General Relativity (GR) can not solve
directly. Although being the best theory of gravity we
know, GR fail when one tries to understand some local observations concerning galactic velocities in galaxy
clusters or the rotational curve of spiral galaxies. The
addition of a new kind of attractive matter that does
not interact electromagnetically, the so called Dark
Matter (DM), is the standard solution in order to
maintain the GR as the background gravity theory.

Electronic

address: holanda@uepb.edu.br
address: shpereira@feg.unesp.br
Electronic address: simony@cct.uepb.edu.br
Electronic

Nevertheless, the nature, origin and dynamics of such

new kind of matter is still a mystery. At large scales
the GR has also been tested, and recent observations
of Supernovae type Ia (see [1] for a recent compilation
and references therein) shows that the universe is currently in an accelerating phase, which can be explained
in the GR context with the addition of a Cosmological
Constant (CC) or with the addition of a new kind of
repulsive energy, the so called Dark Energy (DE) [2],
whose nature is also undetermined.
However, if we take off such put by hand ingredients,
namely DM, CC and DE, GR alone is not enough to explain the large amount of observational data, in several
scales. For this reason, alternative models have been
proposed recently in order to accommodate the observations, as massive gravity theories, modifications of
Newtonian dynamic, known as MOND, f (R, T ) theories that generalizes the Lagrangian of GR, models

2
in higher dimensions, as brane world models, string
theories at low energies, Kaluza-Klein theories,among
others. Nevertheless several of these theories explicitly breaks the Einstein equivalence principle (EEP),
which leads to explicit modifications of some fundamental constants of nature. This allows us to test,
from a statistical point of view, the degree of confidence of such modified theories, as some constants of
nature are known to several orders of magnitude.
A powerful mechanism to test the signatures of a
class of modified gravity theories has been developed
by A. Hees et al. [3, 4]. These authors showed that theories that explicitly breaks the EEP by introducing an
additional coupling of the usual electromagnetic part of
matter fields to a new scalar field (motivated by scalartensor theories of gravity, for instance [5, 6, 7, 8, 9, 10]),
lead to modifications of the fine structure constant of
the quantum electrodynamics [11]. Thus, all the electromagnetic sector of the theory is also affected, which
leads to a non-conservation of the photon number and
consequently a modification to the expression of the
luminosity distance, which is the basis for various cosmological evaluation.
After that, based on the results by A. Hees et al.,
some recent papers [12, 13] have also searched for deviations from the GR by considering the same class of
modified gravity theories which explicitly breaks the
EEP. These studies used angular diameter distances
(ADD) of galaxy clusters obtained via their X-ray surface brightness + Sunyaev-Zeldovich effect (SZE) observations [14, 15], SNe Ia samples [1] and the Cosmic
Microwave Background (CMB) temperature in different redshifts, TCMB (z), [16, 17]. No deviation from
GR was verified, although the results do not completely rule out the models under question.
In this paper, we follow searching for deviations from
the GR by considering cosmological observations and
the class of models that explicitly breaks the EEP in
the electromagnetic sector. We show that measurements of the X-ray gas mass fraction (GMF) in galaxy
clusters jointly with SNe Ia distance moduli and CMB
temperature at different redshift also furnish an interesting test for the GR. In our analyses, we use the most
recent GMF sample from the Ref. [18] with galaxy
clusters in the redshift range 0.078 z 1.063. This
sample contains 40 massive and morphologically relaxed systems obtained from the Chandra observations,
with high-quality weak gravitational lensing data, fundamental to X-ray mass calibration. We show that our

analyses present competitive results with those ones

found in the Ref.[13] and no significant deviation from
GR is verified.
This paper is organized as follows: Section II we
briefly revise the changes on important results of the
standard cosmology when it is assumed a time varying
fine structure constant. In Section III we present as
GMF measurements can be used to test these kind of
modifications. The samples and analyses are in Section
IV. The Section V shows the results and in the Section
VI we conclude.

II.
A.

BREAK OF EEP

Deviations from the standard relations

Some kinds of modified gravity theories that explicitly breaks the EEP can be represented by a matter
action that has the following form:
XZ

(1)
Sm =
d4 x ghi ()Li (g , i ) ,
i

where Li are the Lagrangians for different matter fields

i and hi () represents a non-minimal couplings between the extra field (motivated by several alternative models [5, 6, 7, 8, 9, 10]) and i . When hi () = 1
we recover the standard GR scenery. When the coupling is with the electromagnetic sector of the Lagrangian, the fine structure constant of the quantum
electrodynamics turns to be time dependent [7, 11],
h1 ((t)). The main implication of that time dependence is that the photon number is not conserved
along geodesics, leading to a modification to the expression of the luminosity distance, DL , and also to
the violation of the so-called cosmic distance-duality
relation (CDDR), which happens to be written as [4]:
s
h(0 )
DL (z)
= (z) =
,
(2)
DA (z)(1 + z)2
h((z))
where DA is the angular diameter distance [19] and
(z) parametrizes the deviation from standard GR,
which corresponds to (z) = 1 and can be recovered
for the present time (0 (z = 0)). Some usual
parametrization for (z) are [13, 20]:
P1: (z) = 1 + 0 z
P2: (z) = 1 + 0 z/(1 + z)
P3: (z) = (1 + z)0

3
P4: (z) = 1 + 0 ln(1 + z)
where 0 is the parameter to be constrained. The limit
0 = 0 corresponds to the standard GR results. In
terms of the parametrization (z), the change on the
fine structure constant is [4]:

= (z) 1 = 2 (z) 1.
(3)

Moreover, also due to the non-conservation of the

photon number, a variation of the evolution of the
CMB radiation is also expected, which affect its temperature distribution as [4]:
T (z) = T0 (1 + z)[0.88 + 0.12 2 (z)].

(4)

In [13] the four parametrizations above were analysed and the constraints in the parameter 0 were obtainded by using 29 data of ADD of galaxy clusters
[15] obtained via their SZE + X-ray surface brightness
observations, 29 data of luminosity distances of SNe
Ia [1] at same redshifts of galaxy clusters and 38 data
of CMB temperature. The inclusion of CMB temperature has diminished about one order of magnitude on
the parameter 0 related to previous analysis without
such data [12] (see table I).
In this work, we search for the EEP breaking by
testing jointly the CDDR (2) by using GMF + SNe Ia
observations and the CMB temperature evolution law
(4) by using the most recent TCMB (z) data. Following,
we show how it is possible to obtain luminosity distance
from the X-ray GMF measurements of galaxy clusters
bt tanking into account the dependence of the GMF
with and .

B.

In the subsection, we discuss the consequences of the

EEP breaking on GMF measurements and explained
the basic equations used in our analyses.
The baryonic matter content of galaxy clusters is
dominated by the X-ray emitting intracluster gas via
predominantly thermal bremsstrahlung (see [21] for
more details). On the other hand, the total mass
within a given radius R is obtained by assuming
that the intracluster gas is in hydrostatic equilibrium.
Thus, the gas mass fraction is defined as [22]:
Mgas
,
Mtot

where the symbol * denotes quantities that were used

obs
in the observations, fXray
, from a fiducial cosmological model (usually a flat CDM model where = 1)
and the parameter N defines the astrophysical model
of the cluster, such as stellar mass fraction, nonthermal pressure and the depletion parameter, which
indicates the amount of cosmic baryons that are thermalized within the cluster potential [22]. However, our
method is completely based on the recent results from
the Refs.[20, 25]. In Ref.[25], the authors showed that
the gas mass fraction measurements extracted from Xray data are affected by a possible departure of = 1
and the Eq. (6) must be rewritten as
#
"
3/2
1/2 DL
obs
.
(7)
fXray (z) = N
3/2
DL
The parameter appears after using (2). On the other
hand, in the Ref.[20] was also showed that the gas mass
fraction measurements extracted from X-ray data are
affected by a possible departure of (z) = 1, such as
fXray [(z)]3/2 ,

(8)

or, by considering the class of theories explored by Hees

et al., from the Eq. (6):

Consequences for Gas mass fraction

measurements

fgas =

where Mtot is the total mass and Mgas is the gas mass
obtained by integrating the gas density model. As it is
largely known, assuming that the hot gas mass fraction
does not evolve with redshift [23, 24], X-ray observations of galaxy clusters can also be used to constrain
cosmological parameters from the following expression
[22]:
"
#
1/2
DL
DA
obs
fXray (z) = N
,
(6)
1/2
DL DA

(5)

fXray 3 .

(9)

obs
As one may see, the quantity fXray
may still be deviated from its true value by a factor 3 , which does
not have a counterpart on the right side in the Eq. (7).
In this way, this expression has to be modified to
"
#
3/2
7/2 DL
obs
fXray (z) = N
.
(10)
3/2
DL

Thus, it is possible to obtain the luminosity distance

of a galaxy cluster from its GMF by:

obs
DL = 7/3 DL
[N/fXray
(z)]2/3 .

(11)

4
16000

14000

0.22

from SNe Ia Union 2.1 data

0.20

0.18

12000

0.16

10000

0.14

8000

fgas

DL(Mpc)

Mantz et al. 2014

6000

0.12

0.10

4000
0.08

2000

0.06

b)

a)
0
0.0

0.04

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

z
16

14

12

TCMB(z)

10

c)
0
0.0

0.5

1.0

1.5

2.0

2.5

3.0

FIG. 1: In Fig.(a) We plot the 580 distance moduli of SNe Ia [1]. In Fig.(b) we plot the 40 GMF data [18]. In Fig.(c) we
plot the 36 TCM B (z) data [16, 17].

at radii near r2500 1 , rather than integrated at all

radii (< r2500 ). This significantly reduces the
corresponding theoretical uncertainty in gas depletion from hydrodynamic simulations (see Fig.
8 in their paper). Moreover, the bias in the mass
measurements from X-ray data arising by assuming hydrostatic equilibrium was calibrated by robust mass estimates for the target clusters from
weak gravitational lensing (see also [26]).

Finally, we define the distance module of a galaxy

cluster as
obs
cluster (, z) = 5 log[ 7/3 DL [N/fXray
(z)]2/3 ] + 25.
(12)
As one may see, if we have SNe Ia distance module
measurements, (z), in the identical redshifts of galaxy
clusters, we can put observational constraints on the
parameter.

III.

The full SNe Ia sample is formed by 580 SNe

Ia data compiled by Suzuki et al. [1], the socalled Union2.1 compilation, with redshift range
0.015 z 1.4 (see Fig. 1a). The Union2.1 SNe
Ia compilation is an update of the Union2 compilation and all SNe Ia were fitted using SALT2
[27]. All analyses and cuts were developed in

DATA

In our analyses, we use the following data set:

40 GMF measurements in redshift range 0.078
z 1.063 from Mantz et al. [18]. This is the
most recent GMF sample (see Fig. 1c). Differently from previous GMF measurements, these
authors measured the GMFs in spherical shells

This radii is that one within which the mean cluster density is
2500 times the critical density of the Universe at the clusters
redshift.

5
1.0

1.0

0.9

0.9

GMF

0.8

P1

TCMB

0.7

0.6
0.5
0.4
0.3
0.2

P2

TCMB

0.7

Likelihood

Likelihood

GMF

0.8

1
0.6
0.5
0.4
0.3
0.2

0.1
0.0
-0.3

-0.2

-0.1

0.0

0.1

0.2

0.1

a)
0.3

b)

0.0
-0.4

-0.3

-0.2

-0.1

0.0

0.1

0.2

0.3

FIG. 2: In both figures, the solid blue and black lines correspond to analyses by using SNe Ia + GMF and TCM B (z),
respectively. The dashed area corresponds to the joint analysis (SNe Ia + GMF + TCM B (z)). In Fig.(a) we plot the
results by using the parametrization P1 and in Fig.(b) by using P2.

a blind manner, i.e., with the cosmology hidden. In order to perform our test we need SNe
Ia and galaxy clusters in the identical redshifts.
Thus, for each GC in the GFM sample, we select SNe Ia with redshifts obeying the criteria
|zGC zSN e | 0.006 and calculate the following
weighted average for the SNe Ia data:

IV.

ANALISES AND RESULTS

We evaluate our statistical analysis by defining the

2
likelihood distribution function, L e /2 , where
2 =

40
X
(
(zi ) cluster (, zi ))2
i=1
38
X

i=1

P
(i /2 i )
P
,
2
1/
i

2 =

P 1 2
1/i

(13)

Then, we finalize with 40

i and 2i measurements. Following [1] we added a 0.15 systematic
error to SNe Ia data.

The TCMB (z) sample is composed by 36 points

(see Fig. 1c). The data in low redshifts are from
SZE observations [16] and from observations of
spectral lines we have the data at high redshifts
[17]. In total, this represents 36 observations of
the CMB temperature at redshift between 0 and
2.5. We also use the estimation of the current
CMB temperature T0 = 2.7250.002 K [28] from
the CMB spectrum as estimated from the COBE
collaboration.

2
obs

[T (zi ) Ti,obs ]
,
T2 i ,obs

(14)

2
2
with obs
= 2 + cluster
and T (z) given by Eq.(8).
In our analyses, the normalization factor N is taken
as a nuisance parameter so that we marginalize over
it. The EEP breaking is sought for allowing deviations
from = 1, such as: (P1) (z) = 1 + 0 z and (P2)
(z) = 1 + 0 z/(1 + z). As one may see, if 0 = 0 from
the data the GR is verified.
Our results are plotted in Figs. (2a) and (2b) for
each parametrization. Note that in each case the solid
blue and black lines correspond to analyses by using
separately CMB temperature and GMF + SNe Ia data
in Eq.(14), respectively. The dashed area are the results from the joint analysis, i.e., the complete Eq.(17)
with CMB temperature + GMF + SNe Ia. In Table I
we put our 1 results from the joint analyses for each
parametrization and several 0 values already present
in literature which consider correctly possible variations of and in their analyses. As one may see, our
results are in full agreement each other and with previous one regardless the galaxy cluster observations and
(z) functions used. Moreover, our analyses present
competitive results with those found in Ref.[13] and
no significant deviation from GR is verified.

6
TABLE I: A summary of the current constraints on the parameters 0 for P1, P2, P3 and P4, from angular diameter
distance from galaxy clusters and different SNe Ia samples. The symbol * corresponds to angular diameter distance (ADD)
from Ref. [14] and ** angular diameter distance from Ref. [15].
Reference

Data Sample

[12]
ADD + SNe Ia
[13]
ADD + SNe Ia + TCM B
[13]
ADD + SNe Ia + TCM B
This paper GMF + SNe Ia + TCM B

V.

0 (P1)

0 (P2)

0.069 0.106
0.005 0.025
0.005 0.032
0.020 0.027

CONCLUSION

In this work we continue the search for analysis in

order to test some cosmological signatures of modified
gravity theories that explicitly breaks the EEP. As already pointed by Hees et al. [3, 4], alternative theories
described by the action (1) leads naturally to modifications on standard constants of physics, on the cosmic
distance duality relation and also on the CMB temperature distribution. The possibility to test the validity
of such new theories is of great interest.
The main contribution of this work is use the dependence of current measurements of X-ray gas mass
fraction in galaxy clusters on the EEP validity to test
the class of theories pointed by [3, 4]. Following the
analysis of previous works [12, 13], where data of ADD

[1] N. Suzuki et al., Astrophys J., 85, (2012) 746.

[2] P. J. E. Peebles and B. Ratra, Rev. Mod. Phys. 75,
(2003) 559;
[3] A. Hees, O. Minazzoli, J. Larena, Phys. Rev. D 90,
(2014) 124064, [arXiv:1406.6187];
[4] A. Hees, O. Minazzoli, J. Larena, Gen. Rel. and Grav.
47, (2015) 2, [arXiv:1409.7273]; O. Minazzoli, A. Hees,
Phys. Rev. D 90, (2014) 023017, [arXiv:1404.4266].
[5] T. Damour and A. M. Polyakov, Nucl. Phys. B423
(1994) 532; [hep-th/9401069]; Gen.Rel.Grav. 26
(1994) 1171; [gr-qc/9411069].
[6] J. M. Overduin and P. S. Wesson, Phys. Rept. 283
(1997) 303; [gr-qc/9805018].
[7] J. D. Bekenstein, Phys. Rev. D 25 (1982) 1527; H. B.
Sandvik, J. D. Barrow and M. Magueijo, Phys. Rev.
Lett. 88 (2002) 031302; [astro-ph/0107512]; J. D. Barrow and S. Z. W. Lip, Phys. Rev. D 85 (2012) 023514,
[arXiv:1110.3120]; J. D. Barrow and A. A. H. Grahan,
Phys.Rev. D 88 (2013) 103513; [arXiv:1307.6816];
[8] M. Dine, W. Fischler and M. Srednicki, Phys. Lett.

0 (P3)

0 (P4)

0.000 0.135
0.048 0.053 0.005 0.04 0.005 0.045
0.007 0.036 0.015 0.045 0.015 0.047
0.041 0.042
-

+ SNe Ia + TCMB were used to constraint possible

deviations from GR, here we consider data of GMF
+ SNe Ia + TCMB in order to obtain constraints on
the 0 parameter for the two parametrizations P1 and
P2 of CDDR. As presented in Table I, no significant
deviation from GR was found at 1.

Acknowledgments

RFLH acknowledges financial support from CNPq

and UEPB (No. 478524/2013-7, 303734/2014-0). SHP
is grateful to CNPq - Conselho Nacional de Desenvolvimento Cientfico e Tecnologico, Brazilian research agency, for financial support, grants number
304297/2015-1.

[9]

[10]
[11]

[12]
[13]
[14]

B104 (1981) 199; D. B. Kaplan, Nucl. Phys. B260

(1985) 215.
P. Brax et. al., Phys.Rev. D70 (2004) 123518;
[astro-ph/0408415]; P. Brax, C. van de Bruck and
A. C. Davies, Phys. Rev. Lett. 99 (2007) 121103;
[hep-ph/0703243]; M. Ahlers et. al., Phys. Rev.
D 77 (2008) 015018; [arXiv:0710.1555]; J. Khoury
and A. Weltman, Phys. Rev. D 69 (2004) 044026;
[astro-ph/0309411].
T. Hargo, F. S. N. Lobo and O. Minazzoli, Phys. Rev.
D 87 (2013) 047501; [arXiv:1210.4218].
T. Damour and F. Dyson, Nucl. Phys. B480 (1996)
37, [hep-ph/9606486]; M. T. Murphy et al., Lect. Notes
Phys. 648 (2004) 131, [astro-ph/0310318].
R. F. L. Holanda and K. N. N. O. Barros, Phys. Rev.
D 94, (2016) 023524, [arXiv:1606.07923].
R. F. L. Holanda and S. H. Pereira, Phys. Rev. D 94,
(2016) 104037, [arXiv:1610.01512].
E. De Filippis, M. Sereno, M. W. Bautz and G. Longo,
Astrophys J. 625, (2005).

7
[15] M. Bonamente et al., Astrophys J. 647, (2006) 25.
[16] G. Luzzi et al., Astrophys. J. 705, (2009) 1122.
[17] G. Hurier, N. Aghanim, M. Douspis and E. Pointecouteau, Astron. Astrophys. 561, (2014) A143;
[arXiv:1311.4694 [astro-ph.CO]].
[18] A. B. Mantz et al., MNRAS, 440, (2014) 2077.
[19] I. M. H. Etherington, Phil. Mag 15, (1933) 761; B. A.
Bassett and M. Kunz, Phys. Rev. D 69 (2004) 101305;
R. F. L. Holanda, R. S. Goncalves and J. S. Alcaniz,
JCAP 1206 (2012) 022, [arXiv:1201.2378]; Z. Li, P.
Wu and W. Yu, Astrophys. J. 729 (2011) L14; X.
Yang et. al., Astrophys. J. Lett. 777 (2013) L24; R. F.
L. Holanda and V. C. Busti, Phys. Rev. D 89 (2014)
103517, [arXiv:1402.2161].
[20] R. F. L. Holanda, J. A. S. Lima, M. B. Ribeiro, As-

tronomy & Astrophysics 528, (2011) L14.

[21] C. L. Sarazin, X-Ray Emission from Cluster of Galaxies, (Cambridge University Press: Cambridge) 1988.
[22] S. W. Allen et al., MNRAS 383, (2008) 879.
[23] S. Planelles et al., MNRAS 431, (2013) 1487.
[24] N. Battaglia, J. R. Bond, C. Pfrommer and J. L. Sievers, ApJ, 777 (2013) 123.
[25] R. S. Goncalves, R. F. L. Holanda, J. S. Alcaniz, MNRAS 420, (2012) L43.
[26] D. E. Applegate et al., MNRAS 457, (2016) 1522.
[27] J. Guy et al., Astronomy & Astrophysics 466, (2007)
11.
[28] J. C. Mather, D. J. Fixsen, R. A. Shafer, C. Mosier,
and D. T. Wilkinson, Astrophys. J. 512, (1999) 511.