MARIA CURIE-SKŁODOWSKA UNIVERSITY in

LUBLIN
Faculty of Mathematics, Physics and Computer
Science

Bartłomiej Kiczek
Register No.: 260923

Structures and traces of dark matter in

different physical systems – from condensed
matter to black hole physics

A doctoral thesis
prepared in the Department of Theoretical Physics
under supervision of prof. Marek Rogatko

LUBLIN 2023
UNIWERSYTET MARII CURIE-SKŁODOWSKIEJ
W LUBLINIE
Wydział Matematyki, Fizyki i Informatyki

Bartłomiej Kiczek
Nr albumu: 260923

Struktury i ślady ciemnej materii w różnych

układach fizycznych – od materii
skondensowanej do fizyki czarnych dziur

Praca doktorska
napisana w Katedrze Fizyki Teoretycznej
pod kierunkiem prof. dr hab. Marka Rogatko

LUBLIN ROK 2023

Abstract

English
This thesis is concerned with several scenarios of non-gravitational interactions between dark
matter (represented by dark photons and axions) and visible matter, which were investigated
using a similar theoretical framework. Using the general relativity with different matter
contributions we modelled, i. a. , the structures around black holes – both visible and dark
– and showed possible implications on their physics generated by interaction terms between
two kinds of matter. The thesis consists of two parts, first provides an introductory material
to the conducted research. The second one collects the articles, published in peer-reviewed
scientific journals, being the substantive scientific contribution of this dissertation.

Polski
Niniejsza praca dotyczy kilku scenariuszy związanych z niegrawitacyjnymi oddziaływaniami
pomiędzy ciemną materią (reprezentowaną przez ciemne fotony oraz aksjony), a widzialną
materią, które zostały zbadane z wykorzystaniem podobnych technik teoretycznych. Bazu-
jąc na ogólnej teorii względności z dodatkowymi wkładami materii zamodelowano m. in.
struktury wokół czarnych dziur - zarówno widzialne, jak i ciemne - i pokazano ich możliwy
wpływ na fizykę, który wynika z oddziaływania pomiędzy dwoma rodzajami materii. Praca
składa się z dwóch części: pierwsza stanowi materiał wprowadzający do podjętej tem-
atyki badawczej, w drugiej zebrano artykuły opublikowane w recenzowanych czasopismach
naukowych, stanowiące merytoryczny wkład naukowy tej rozprawy.

i
Acknowledgements

My gratitude is directed twofold. First, I would like to thank my advisor, professor Marek
Rogatko for excellent mentoring during my doctoral studies. I truly appreciate our
discussions, engagement in scientific projects and enormous dose of freedom. On the other
hand, I would like to thank my wife Anna, for gigantic patience and endless support. Her
efforts allowed me to work on this thesis and complete it in the best possible way I could.

ii
Contents

Abstract i

Acknowledgements ii

1 Introduction 1

2 Theoretical background 3
2.1 Dark matter in a nutshell . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1.1 Axions and axionlike particles . . . . . . . . . . . . . . . . . . . . . . 4
2.1.2 Dark photons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1.3 Alternative theories of gravity . . . . . . . . . . . . . . . . . . . . . . 13
2.1.4 Macroscopic objects . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2 Hairy black holes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.3 AdS/CFT correspondence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

3 Methods 20
3.1 Theoretical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.1 Equation of motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.2 Boundaries and asymptotics . . . . . . . . . . . . . . . . . . . . . . . 21
3.1.3 Other quantities and benchmarks . . . . . . . . . . . . . . . . . . . . 23
3.2 Numerical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.1 Pseudo-spectral collocation method . . . . . . . . . . . . . . . . . . . 24
3.2.2 Shooting method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.3 Spectral differentiation . . . . . . . . . . . . . . . . . . . . . . . . . . 27

4 Original contributions 29
4.1 Summary of articles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
4.1.1 Ultra-compact spherically symmetric dark matter charged star objects 29
4.1.2 Influence of dark matter on black hole scalar hair . . . . . . . . . . . 31
4.1.3 Holographic DC SQUID in the presence of dark matter . . . . . . . . 34
4.1.4 Axionlike dark matter clouds around rotating black holes . . . . . . . 38
4.1.5 Static axionlike dark matter clouds around magnetized rotating worm-
holes - probe limit case . . . . . . . . . . . . . . . . . . . . . . . . . . 40

iii
CONTENTS iv

4.2 Attached articles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.3 Other activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
4.3.1 Conference presentations . . . . . . . . . . . . . . . . . . . . . . . . . 113
4.3.2 Other papers in physics . . . . . . . . . . . . . . . . . . . . . . . . . 113
4.3.3 Papers in computer science . . . . . . . . . . . . . . . . . . . . . . . . 113
4.4 Future investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
4.5 Co-author statements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
Chapter 1

Introduction

Space research constitutes a specific part of science. Except its solid, methodologically
rigorous part it also contains a pinch of mystery and poetry. The view of a starlit sky
has not been an inspiration only to generations of researches and philosophers, but also to
ordinary people. This phenomenon was extensively used in culture, mostly in science-fiction
genre, where thousands of authors stimulated their reader’s curiosity and need to explore
the unknown.
The issue of dark matter (DM) fits perfectly into this theme of galactic secrets. Despite
the solid observational foundations it also has an element of mystery. The mystery that
has been bothering minds of many scholars for almost one hundred years, yet still remains
without definite answers, and even constantly brings more questions.
It seems quite complicated to start a brief historical review from nobody else but Fritz
Zwicky [1]. Although the premises for the existence of an additional, invisible matter in
galaxies were given by researches such as Jan Oort, the Swiss astronomer is considered the
father of dark matter. In 1930s, he collaborated with Edwin Hubble, who provided him
the observational data, related to galaxy clusters and their dynamics. Zwicky built a model
based on the viral theorem to describe the motion of the galaxies in a cluster. The viral
theorem gives the relation between mean kinetic energy and mean potential energy of the
system in question.
Zwicky concentrated his efforts on the Coma Cluster data, especially on the velocities of
the galaxies that built this cluster. In the next step, he calculated the difference between
the slowest and the fastest galaxy. Then he compared observationally measured velocities
to these determined by his model. The result was stunning as the difference was to the
order of several hundred percent. It lead him to the conclusion that there is a large amount
of dark, unobserved matter in the system, which significantly affects the dynamics of the
galaxy motion.
Another clue for dark matter existence has been brought by spectroscopic observations
of spiral galaxies with 21 cm wavelength. The rotation curve was the key characteristic -
the velocity of a galaxy element as a function of its distance from the centre. Although
such measurements were carried out since the 1920s, a significant breakthrough in this field

1
CHAPTER 1. INTRODUCTION 2

took place in 1970s. The milestone was the device built by Vera Rubin and Kent Ford [2].
The authors studied the neighbouring Andromeda Galaxy. The new spectrograph allowed
the collection of the rotation curve far from the centre of the galaxy, much farther than its
predecessors.
It turned out that the measured rotation curve behaved as predicted up to a certain
point. However the farther from the centre, where it should have decreased, it reached a
plateau. Similar behaviour was also observed in other spiral galaxies such as NGC2403,
NGC6946 and M101. This observation shows that the spatial distribution of mass does not
correspond to that of visible matter. Moreover, the galaxies extend much further beyond
the range of glowing visible matter.
In those days, it was a truly shocking discovery that many researches of that time did
not believe in. Even though, new evidence was emerging in favour of this hypothesis in
the following years. More accurate measurements of rotation curves or gravitational lensing
caused by clouds of dark matter are among many examples.
Nowadays, however, few dare to doubt the existence of dark matter. The number of
astrophysical and cosmological implications makes the explanation of this mystery one of
the burning problems of contemporary physics. However, it is not known how and when
the experiment will yield a robust answer. Despite the significantly greater abundance of
27% of the mass-energy in the Universe [3], dark matter remains elusive to the most sublime
apparatus.
However it is not so certain that any dark matter particle will be discovered in the Earth-
based experiments. Particle colliders seem to lack power to achieve the high enough energy
levels. On the other hand, there is a severe problem with the detection itself. As dark matter
is defined to surely interact gravitationally, the non-gravitational interactions are predicted
to be ultra-weak at best. Different particle probes located on our planet have not brought
unquestionable results.
It seems, however, that the traces of dark matter may appear in the distant stellar
systems. From the theoretical point of view, it is extremely fascinating to study known
physical systems enriched with additional dark sector fields and novel interactions. This
may be of particular importance in some astrophysical systems, where a high density of
dark matter is expected. The research performed for the sake of this thesis fits in this
newly emerging tide in the dark matter community. The lessons learned in the past three
decades show that it is imminent to look for indirect methods for testing different dark
matter candidates. These scenarios developed on paper and in computer simulations might
be confronted with better and better astronomical surveys.
Therefore, the incoming decades will be exceptionally interesting in the study of dark
matter. The variety of theoretical models might eventually obtain some feedback from the
experiment, thanks to the continuous development of empirical methods. Since the 2010s
managed to ”catch” a gravitational wave, it is possible that the following years will bring a
coveted discovery of an exotic phenomena related to a particular dark matter candidate.
Chapter 2

Theoretical background

The dark matter community used to suffer from some kind of stagnation. Multiple proposed
solutions seemed not to meet requirements brought by cosmological surveys and the data
from experiments aimed at detecting them remained meaningless. Recently Bertone and Tait
in their short Nature paper [4] raised a call to the community, that our research methods
should be diversified and new ideas are dramatically needed. The community reacted suitably
to the manifest and by browsing scientific journals devoted to astronomy, gravitation and
cosmology we can notice the appearance of the new trend. Not only new ideas for ground-
based experiments are emerging, but also, which personally are far more interesting to me,
ones related to astrophysical observations.
In this chapter I would like to discuss theoretical foundations and sources of inspiration
for the research of this thesis. Because the dark matter community is very diverse, containing
particle physicists, cosmologists, astronomers and relativists, the points of view are different
and ideas versatile. I will briefly review the state-of-the-art, aforementioned new ideas (some
of them are rather new-old) and where do I fit between them.
In whole work I use the geometrized unit system, with c = G = kB = 1 and naturally
the Einstein summation convention.

2.1 Dark matter in a nutshell

Since classical investigations by Zwicky through the Rubin-Ford breakthrough the contem-
porary dark matter market has split into three most popular branches, that used to be
seemingly mutually exclusive. First branch says that dark matter is made of some new par-
ticles, living beyond the Standard Model (SM). This idea emerged from particle physicists
studying cosmology, where on particular stages of the Universe evolution different ingredients
were needed to match the observational data. Although it is not always the case, particle
physics theory creates new beings that occasionally seem to be a nice dark matter candidate.
However, what are the requirements that particle ought to meet to be promoted with a
candidate title? Large astronomical and cosmological evidence requires four basic features
from such particles. They should be massive, stable for at least several billions of years, non-

3
CHAPTER 2. THEORETICAL BACKGROUND 4

relativistic, and very weakly interacting. If one takes all these requirements and compare
them to all particles known by particle physics there is only conclusion - dark matter most
likely cannot be a part of the Standard Model. Therefore there is large pressure on proposing
and testing various scenarios involving hypothetical particles, where some of them solve both
existing problems in SM and make a great DM candidate.
WIMPs (Weakly Interacting Massive Particles) used to be one of the most prominent
candidates for dark matter. They constitute a thermal relic of the early stages of the evolu-
tion of the Universe. At larger energy scales they used to be coupled to the baryonic matter,
being in thermal equilibrium with it. As the Universe cooled down the WIMP dark sector
decoupled from the other ingredients and created a medium which interacted mostly gravi-
tationally (hopefully not only) with the visible matter. WIMPs’ masses are supposed to be
heavier than 100keV, moreover they are supposed to have self-annihilation cross section on
the order of 10−26 cm3 /s.
We owe WIMPs the burst of interest in direct detection of dark matter. The so-called
WIMP paradigm fuelled a lot of ideas for stand alone experiments, but also for studying
Large Hadron Collider (LHC) reactions for traces of dark matter. Unfortunately, so far all
in vain, and it looks like the WIMP paradigm is falling apart. This lack of direct evidence
turns the attention of the scientific community to new methods and new candidates for DM
particles.
Under these circumstances, more and more different particles are considered in the plan-
ing of dark matter experiments. Modern literature contains a plethora of ideas based on
various assumptions - from supersymmetry and string theories to fixes in SM and their
by-products. Picking one of them makes a first challenge for a researcher starting one’s
investigations in the area of dark matter.

2.1.1 Axions and axionlike particles

Axions
From the big family of potential candidates I was specifically interested in axions and dark
photons. Axions originate from the so-called charge-parity (CP) problem of quantum chro-
modynamics (QCD). This vital part of SM describes the strong interaction of quarks and
their gauge bosons gluons. It is widely known that symmetries play key role in modern
physics. They govern they existence of conserved quantities, what we know from Noether
theorem, and the interaction among elementary particles. We expect a physical quantum
field theory to be invariant under charge-parity-time (CPT) symmetry, that stands for charge,
parity and time. In other words CPT symmetry corresponds to Lorentz invariance, which is
of the key importance in Minkowski spacetime, where the quantum field theory (QFT) lives.
Naturally, we know that some physical phenomena may break the time invariance. For
instance the presence of magnetic field in the system can break time symmetry (T). However,
as we stated before, the whole CPT symmetry must stay intact, for the sake of physical-
ity of the theory. Therefore the remaining CP symmetry should break in such way, that
compensates the breaking of T symmetry, so the total CPT symmetry is not violated.
CHAPTER 2. THEORETICAL BACKGROUND 5

In Standard Model, breaking of CP symmetry occurs in weak interactions. This can

be (and actually was) seen experimentally in different counts of some process and its CP
transformed counterpart. In the realm of strong interactions no CP symmetry breaking was
observed. This appeared to be a very surprising fact to the community as the QCD theory
expects CP violation to occur. The so-called θ term is the suspect here
g 2 θ µν
G G̃µν , (2.1)
32π 2
where Gµν is the gluon field strength tensor and g is a coupling constant and is currently
constrained to g > 6 − 34 × 10−14 GeV −1 . Theta itself is an angle, so it naturally should
take some value from 0 to 2π. On the other hand, the presence of this term induces an
electric dipole moment of the neutron. This dipole moment is proportional to theta |d| ∼
3.6 × 10−16 θe cm. From precise measurements we know that |d| < 1.8 × 10−26 e cm, which
means that θ < 5 × 10−11 . This is extremely tiny, and it tends to be called unnatural by
several researchers, especially if this parameter is theoretically allowed to be θ ∈ [0, 2π]. This
diminishing of θ might be either a coincidence or a sign of new physics.
One of solutions to this fascinating problem was proposed by Roberto Peccei and Helen
Quinn in their paper from 1973 [5]. In their idea θ was promoted to a dynamical field instead
of being merely a SM constant. In this way a new fundamental field and a particle - axion -
was born, with properties that keep θ close to zero, so the electric dipole moment of neutron
vanishes.
The emergence of axion is connected with introduction of a new symmetry, called the PQ
symmetry from the names of the authors. In the early stage of the Universe PQ symmetry
was present in particle interactions. As the universe cooled down a spontaneous symmetry
breaking took place, leaving the axion as the Nambu-Goldstone boson of PQ symmetry. In
consequence a new fundamental boson, next to Higgs boson, is added to the set of particles
of SM. Its mass is estimated to be
 12 
10 GeV
ma = 5.70 ± 0.007µeV , (2.2)
fa
where fa is the energy scale of spontaneous breaking of PQ symmetry. However, this value
remains unknown so far. It is speculated that it might be grand unification scale or even
Planck scale. From current estimations of possible PQ symmetry breaking energy scale the
QCD axion mass bounds range from 1 × 10−13 to 2 × 10−11 eV.
Axions are pseudo-scalar particles, therefore there are certain possible interactions al-
lowed for them with other SM sectors. The nonrelativistic Hamiltonian density schematically
expresses them
r p
ϵ0
H= gaγγ aE · B + gaf f ℏc∇a · Ŝ + ϵ0 (ℏc)3 gEDM aŜ · E, (2.3)
µ0
where a is axion field; gaγγ is the axion coupling to photons; gaf f is coupling to fermionic
matter and gEDM is the strength of axion induced neutron electric dipole moment. Vectors
CHAPTER 2. THEORETICAL BACKGROUND 6

E and B are the electric and magnetic fields, Ŝ is the spin direction of the particle that an
axion interacts with.
However, on the macroscopic level in a gravitational setup we are the most interested in
the first term of this Hamiltonian, which enables interactions with electromagnetism. The
covariant Lagrangian of the axion is following:
1 1 1
Laxion = − ∇µ a∇µ a − m2a a2 − gaγγ a ∗ F µν Fµν , (2.4)
2 2 2
where Fµν = 2∇[µ Aν] is electromagnetic field strength tensor and ∗ denotes Hodge dual
operator. The first term of the Lagrangian is a dynamical one, second is the mass term,
while the third one enables interactions of axions with photons in the magnetic field.
Soon after the appearance of the axion it was found that it might be a very good dark
matter candidate. The idea was proposed independently by three groups in their papers
published in Physics Letters B [6, 7, 8], and recently reviewed in [9]. The properties of
axions, such as their bosonic nature, little couplings to SM and small masses, allow them to
form structures from stellar up to galactic scale. First of all, if we consider the fundamental
principle of quantum mechanics - the particle-wave duality - we can think of DM structures
as waves. These waves naturally have some wavelength, using de Broglie’s formula in the
non-relativistic limit we have
h
λ= . (2.5)
mv
Relatively, heavy particles such as electrons have wavelengths of the order of nanometers.
Because of that, in most of day-to-day life phenomena we experience electrons as particles,
with their wave nature revealing on the appropriate length scale, such as diffraction on a
crystalline lattice. However, for light DM we expect these waves to be very long, up to even
hundreds of light years, with very low frequencies. Such waves can be seen as DM forming
the web-like structures on the large scale.
Another argument for usefulness of axion as DM is its bosonic nature. As it is widely
known bosons do not follow the Pauli exclusion principle. This means that many bosonic
particles can occupy the same quantum state. If this number is large, and so macroscopic,
the particles can be described with classical field theory, rather than quantum. Moreover,
under specific circumstances, bosons can form Bose-Einstein condensate, which grants them
new collective properties. As an illustration, Helium-4 atoms in superfluid state occupy the
ground state and grant ability of collective movement without viscosity.
Bosonic condensates are especially interesting, as they might appear in gravitational
setups. There are two types of them. First are hairy black holes, which are black hole
solutions with additional external structures of matter fields, such as scalar condensates.
This case is not only reserved to a black hole, as other objects such as neutron stars or
regular stars might possess such hairy configurations too. I will return to this topic in the
following sections of the thesis. Second setting are the so-called boson stars, or more precisely
axion stars in this case. They constitute a self-gravitating droplet of a (pseudo-)scalar field,
which is cold and macroscopic. Having in mind these ideas, axions might not only form large
scale DM structures but also local aggregates of DM on sub-galactic distances.
CHAPTER 2. THEORETICAL BACKGROUND 7

Axionlike particles
Dark matter theory also considers axion-like particles (ALPs), which share similar physical
foundations and the theoretical construction as QCD axions, however they are free from the
PQ theory related constrains [10, 11].
First of all, mass range of ALPs is much broader, as it does not depend on the energy
scale of spontaneous symmetry breaking of PQ symmetry. Therefore we talk about very
broad range from 10−22 up to 10−2 eV. This naturally adds a lot of nuances to the large
picture. For instance the ultralight, so called fuzzy dark matter made of 10−22 eV ALPs
cannot be the only DM we need, as it would give only few percent to the total mass of
DM the Universe needs to look like it looks. On the other hand, heavy ALPs could be
spotted in nuclear interactions, especially in fierce environments, like inside supernovas. The
production of ALPs adds an additional cooling channel to the system, which gives some
observational possibilities.
The couplings of ALPs to the SM particles are also more untied. While QCD axions
have a bound, originating from effective field theory, saying that the axion couplings are
proportional to its mass, ALPs generally do not have such constrains.
Finally there are other methods allowing to introduce axion-like particles to physics,
different than solution to QCD of CP violation problem. The top-down approach of string
theories allows to derive ALPs in the compactification procedure. In fact ALPs are quite
common in string compactifications as they rise naturally as zero modes of antisymmetric
tensor fields. These, on the other hand, are generally present in most string theories. For
illustration, let us consider a compactification of a weakly coupled heterotic string theory.
Here the axion emerges as the Poincare dual of the antisymmetric tensor field Bµν [12]. The
E8 ⊗ E8 theory in 9 + 1 dimensions is given by the action [13]
Z Z Z
2πMs8 √ Ms6 1 ∗ 2πMs4 1
SH = 2
10
d x −gR − 2
F∧ F− 2
H ∧∗ H + . . . , (2.6)
gs 2πgs 4 gs 2
where R is the Ricci scalar, F the gauge field strength and H is the field strength of the
aforementioned Bµν field. This action describes the dynamics of the massless bosonic exci-
tations of the heterotic string. Compactification of the above theory on a six-dimensional
manifold with volume V6 leads to the effective action in 3 + 1 dimensions
Z Z Z
MP2 4
√ 1 √ 1 √
S3+1 = d x −gR− 2 d x −gFµν F − 2 d4 x −gHµνρ H µνρ +. . . , (2.7)
4 µν
2 4gY M 4fa
with parameters provided by
4π 4πgs2 gs2
MP2 = 2 Ms8 V6 ; gY2 M = 6 ; fa2 = . (2.8)
gs Ms V6 2πMs4 V6
Using the above dependencies one can rule out the volume V6 and the string scale Ms and
get the axion energy scale
√
fa = αY M MP /(2π 2) ≃ 1.1 × 1016 GeV, (2.9)
CHAPTER 2. THEORETICAL BACKGROUND 8

with αY M = gY2 M /(4π) ∼ 1/25, we get the energy scale of PQ symmetry breaking close to
GUT scale.
On the other hand, in string theories of type II we can reach different values of fa and a
pretty large range
MP
fa ∼ p ∼ 104−17 GeV, (2.10)
V6 Ms 6

spreading over 13 orders of magnitude. This example shows how diverse ALPs can be,
with large ranges of possible values of parameters describing their physics, depending on the
creating mechanism. The ongoing experiments shall place constrains on these ranges and
consequently narrow them in upcoming iterations.

Experiments
Axions and ALPs have been widely theoretically discussed in the past years. Recent works
can be split into two different categories. First one (the mainstream one) is focused on direct
detection of these new particles in Earth-based experiments. The other, which seems minor
so far, focuses on space-based indirect effects, with axions playing important part in.
Most of the ALP-oriented experiments utilize the hypothetical particle reaction, which is
typical for ALPs. The annihilation of an axion and creation of two photons in magnetic field
or vice versa is a natural consequence of the last term from equation (2.4). The hypothetical
existence of such process yields a basis for experiments, which can be colloquially named as
shining through a wall. In such setups [14] the laser beam, placed in the strong magnetic
field, simply strikes towards a wall. The goal is to turn these photons into axions, which
can travel freely through the obstacle. As they pass it, they once more enter the region with
strong magnetic field and hopefully a pair of photons can be recovered, so they might be
easily observed. In practice, it is obviously not that easy and such phenomenon has not been
observed yet. Microwave cavities such as ADMX [15] aim at ALPs of sub-eV mass range of
the cosmic origin. Once again it uses the resonant conversion of axions into microwave range
photons. On the other hand, it was predicted that axions may cause a time-varying nucleon
electric dipole moment, which results in precession of nuclear spins in the background electric
field (the counterpart of the magnetic nuclear resonance).
There are also some space-oriented experiments that aim into detecting the axions, which
were created by stars (such as the Sun). The so-called Primarkoff effect allows for the
production of axions in the solar core. Analogically to the devices mentioned in the previous
paragraph, helioscopes use strong magnetic field to convert hypothetical solar axions into
photons. The probability of such conversion in homogeneous magnetic field B over the
distance L is  2
sin(qL/2)
Pa→γ = gaγ B , (2.11)
q
with q = m2a /2E being the axion-photon momentum transfer in vacuum. The examples
of such devices are International Axion Observatory (IAXO) [16], CERN Axion Solar Tele-
CHAPTER 2. THEORETICAL BACKGROUND 9

scope (CAST) [17] and the newest BabyIAXO [18, 19]. However, as these ideas are totally
interesting they have not brought any substantial results yet.
Despite already realised experimental projects there are various theoretical propositions
that can be put into flesh in near future. Last twenty years were generous for condensed
matter physics, both in experimental and theoretical progress. The fruits of these scientific
advances became our everyday comfort. Fast CPUs, large memories, this abundance of com-
puter resources was merely a dream in the previous century. Apart from practical aspects
of condensed matter physics, plenty of interesting ideas emerged, such as novel states of
matter or topological materials. The dark matter community did not stay indifferent to new
physical mechanisms and opportunities offered by novel states of matter. Plenty of theoret-
ical proposals have been presented. The usage of superfluid helium and its interaction with
hypothetical ultra-light bosons were considered in [20]. Absorption of ultra-light particles
in superconductors was studied in [21]. Authors calculated the outcomes of hypothetical
scattering of particles such as ALPs and dark photons off conduction electrons in Cooper
pairs. In [22] polar materials were considered. Authors used density functional theory for
modelling the condensed matter setup and calculating adequate matrix elements. Scattering
of ultra-light bosons on electrons in different crystals such as GaAs or sapphire was studied.
On the other hand, novel materials such as Weyl semi-metals could host phenomena, like ax-
ion originated charge density waves [23]. Although these ideas seem theoretically interesting,
their experimental utilization looks rather challenging.
In the recent years, researchers were looking for axion-involved space-based phenomena.
Adding ALP to different astrophysical and gravitational theories gave a plethora of inter-
esting outcomes. Let me briefly mention some of them. Laser-like blasts of light caused
by axionic superradiant instability were predicted around rotating black holes [24, 25]. It
was shown that an unstable axionic cloud can emit electromagnetic radiation, exponentially
in time. In [26] authors showed that the polarization of light might be twisted by axionic
fields. This result was obtained by modelling light rays emitted from protoplanetary neb-
ula, which emits polarized light (with polarization perpendicular to the plane of the nebular
disc). Then as the light passes the axionic cloud somewhere in the space, its polarization
angle appears to be different than before the passage. Authors also propose the measure-
ments of this angle as an indirect method of ALP detection. Similar effect concerning cosmic
microwave background (CMB) was predicted [27]. It appears that fuzzy axionic dark matter
can influence the polarization of CMB both in the early stage of the Universe and in the
late time. Surprisingly these effects can be distinguishable, as the early-time oscillations of
the axion field wash out the polarization of the CMB produced at last scattering. Neutron
stars [28, 29, 30] host a suitable environment for axionic configurations. It was shown that
depending on pulsar frequency axions can form a stable condensate or radiate away. More-
over pulsars can also be used in polarimetry measurements. As their light is polarized, any
fluctuations for the polarization angle might indicate some new physics originated from the
hidden sector interactions.
CHAPTER 2. THEORETICAL BACKGROUND 10

2.1.2 Dark photons

Another kind of particle that I considered in my research is the dark photon. It is a particle
originating from a U (1) symmetry gauge group, that is kinetically mixed with the Maxwell
field strength tensor. The hidden photon shares similar properties with its visible cousin.
However, it might be massive and adds a lot of new physics into the consideration. The
idea was initially introduced by Bob Holdom [31] and it was widely studied since that time
[32, 13].
The motivation for dark photon is based on several clues. First is the so-called natural-
ness, based purely on particle physics arguments. Dark matter is thought to be, most likely,
not a single kind of particle, but the whole new set of particles. This new sector can be as
various as the SM itself. Given the extremely low interaction cross-section between SM and
DM, most of the dark sector particles can be uncharged under SM gauge groups.
Most of experiments assume that some kind of interaction between SM and DM exist. In
fact processes such as scattering of a heavy nuclei require the adequate non-zero interaction
Hamiltonian terms

Hint ̸= 0 =⇒ ⟨f |Hint |i⟩ > 0, (2.12)

where the transition amplitude from initial |i⟩ to final |f ⟩ state is non-zero. In such picture
the DM cannot be completely dark, as there is some weak coupling with SM particles.
However this image can be totally false. The dark sector might exist as a parallel world,
containing different particles - scalars, fermions - that are fully decoupled from the standard
model.
This sad perspective, especially in its strong variant, can exclude any direct detection
of a DM particle from our experimental planning. However there is a shade of hope if we
assume that there might a portal to the dark world. A particle that is both coupled to the
visible and the dark side. Theoretically such portal can take various forms: vector, fermionic
(neutrino), scalar or previously introduced pseudo-scalar - axions.
From all these possibilities the vector solution looks most promising, as this kind of
particles constitutes gauge bosons, which mediate the fundamental interactions between
particles. By analogy, the dark sector might have its own interactions, with its own gauge
bosons. Therefore the kinetic mixing between the dark photon and the visible photon (which
is Abelian gauge boson - that is allowed for such mixing) can take place.
This addition opens a new spectrum of possibilities, which are two-fold. First, the dark
photon itself can be investigated with its reaction with visible matter. On the other hand,
it may allow to indirectly probe dark particles that are decoupled from SM. One way or
another the consequences are quite serious.
The case of supernovae is an interesting example and simultaneously a second, phe-
nomenological clue for dark photon. Surprisingly it appears that supernovae are colder than
they should be. It means there occurs some kind of excessive cooling mechanism, which
allow them to reach lower temperatures. This new mechanism can be related to the dark
sector particles, as new channels for thermal energy evaporation.
CHAPTER 2. THEORETICAL BACKGROUND 11

Dark photons are thought to be produced in the core of a supernova. From there they
can escape from the star area and decay into an electron and positron pair. Later on, these
pairs can produce 511keV photons, which are naturally measurable and can be registered as
an excess of annihilation radiation. The bounds on this phenomenon were placed using the
Raffelt criterion in the measurements of SN 1987a [33, 34]. The new particle cannot carry
more energy than 3 × 1052 erg/s out of the system. On the other hand, the over abundance of
galactic positrons can not occur. This conditions place specific constraints on the coupling
parameter and the dark photon mass.
These γ-ray flashes are also predicted in different astrophysical systems. For instance
binary neutron star mergers are also high energy physics labs, which can host phenomena
related to physics beyond SM [35]. When the merging occurs a metastable remnant, similar
to a protoneutron star after the supernova, is formed. As this extremely hot body cools
down it can produce new weakly coupled particles.
Similarly to ALPs dark photon also can be derived from strings. A compactification of
E8 ⊗ E8 heterotic string supergravity serves as an example. In this case the first exceptional
Lie group E8 contains Standard Model gauge groups, while the second E8 group includes
a hidden gauge group, which interacts with the first one only gravitationally. During com-
pactification the second factor can be broken into non-Abelian and U (1) gauge groups. For
instance in the orbifold compactification of heterotic string, one obtains SM groups and two
hidden sector symmetries, one non-Abelian and another U (1) [36]

E8 ⊗ E8 → SU (3) ⊗ SU (2) ⊗ SU (1) ⊗ [SU (6) ⊗ U (1)]. (2.13)

In low energy scales this results in an effective Lagrangian where the non-gravitational
coupling occurs through kinetic mixing. This Lagrangian has the form
1 α 1
LEM +DP = − Fµν F µν − Fµν B µν − Bµν B µν − m2 Bµ B µ , (2.14)
4 4 4
where the Maxwell field strength tensor, is defined as Fµν = 2∇[µ Aν] . The other field Bµ is
the vector potential of dark photon field, with corresponding field strength tensor, defined
analogically Bµν = 2∇[µ Bν] . Finally m denotes the mass of the dark photon and α is the
coupling parameter between two sectors.
In general case the mass of the dark photon might be non-zero, as it arises either from
standard Higgs mechanism or the Stueckelberg mechanism. The mass range is quite large
as values from meV up to GeV scales are expected. For the second parameter of the theory,
the kinetic mixing, different sources (observational constraints and string predictions) expect
the value of α to be in range
10−12 < α < 10−2 . (2.15)
Similarly to axions, dark photons have been intensively probed throughout two past
decades. Most of the experiments focus on the detection of space-originated dark photons,
however some aim at capturing particles created in hypothetical processes in accelerators. It
is expected that dark matter particles can interact with nuclei or electrons in the detecting
CHAPTER 2. THEORETICAL BACKGROUND 12

materials. So far it was only DAMA/Libra collaboration that claimed to capture an excess
modulation of the signal [37, 38]. However, their results have not been replicated for a long
period of time [39]. The gamma rays incoming from dwarf galaxies were carefully studied by
Dark Energy Survey [40] and other projects [41, 42, 43, 44]. Astronomers aim at registering
excessive cosmic rays that can be related to non-gravitational dark matter-ordinary matter
interactions. For instance regular photons can contain traces of past interactions with dark
photons or dilaton-like particles. Moreover these interactions may cause oscillations of fine
structure constant [45]. Recently authors in [46] revised the data from 1987A supernova.
Many new effects were brought into consideration, such as finite temperature and density
of coupling parameter. The final result brought a significant decrease of the previous upper
limit on dark photon mass. It has been reduced below 15MeV, while previously it was over
100MeV.
Another experiment focused on direct detection (or even production) is BABAR, where
researchers try to produce dark matter particles in accelerators. Precisely it is thought
that positron-electron e+ e− collisions can be a source of dark photons [47]. SENSEI (Sub-
Electron-Noise Skipper CCD Experimental Instrument) collaboration created a device based
on the idea of a CCD matrix. The electrons in the valence band of a semiconductor can be
excited to the conductance band where the device can detect them. Possible dark photon-
electron interaction can cause such process, leading to the measurement of the interaction.
The detector is placed in an underground cave, which helps to eradicate different noises.
Recent report on SENSEI’s activity [48] shown no registered dark matter related events and
provided constraints on the probed particles.
One of the most prominent experiments of our times is run by XENON collaboration.
The Italy-based laboratory, deep under Gran Sasso d’Italia massif, is one of the best isolated
from the outer world labs. Such level of isolation allows for conducting precise measurements
of subtle phenomena. The heart of the laboratory’s apparatus is the large tank filled with two
tones of ultra-pure liquid xenon. Whole tank is surrounded by hundreds of photomultipliers.
Such set up can remind the great Super-Kamiokande experiment, which successfully found
the evidence for neutrinos oscillation. However XENON collaboration aims at detecting
different beings. The WIMPs were the first target, as a photon emission was expected after
a collision between a WIMP and a xenon nucelus. Although as the attention of the dark
matter hunters turned towards lighter particles such as axions and dark photons, researchers
realised that XENON1T might measure these particles too, but via absorption. It is expected
that light dark matter particles can cause an analogue of photoelectric effect in a xenon atom.
The measured energy would be then proportional to the dark particle mass.
Recently XENON collaboration reported an excess of events [49] in a possible mass range
of dark matter particles. This paper caused an avalanche of papers by particle-oriented dark
matter theorists in the autumn of 2020 in Physical Review Letters. So far it is not known
what caused the excess. Plenty of ideas have appeared, with axions and dark photons as
most probable culprits. However, sceptics indicate that it might have been a radioactive
background issue. Upcoming experiments - LUX-ZEPLIN and PandaX, based in the US
and China respectively, might hopefully replicate the result or falsify it.
CHAPTER 2. THEORETICAL BACKGROUND 13

The remaining two branches of dark matter are alternative theories of gravity and macro-
scopic objects.

2.1.3 Alternative theories of gravity

Alternative theories of gravity assume that the gravitational interaction behaves differently
on different length scales. They do not deny the predictions of General Relativity (GR) in
the Solar system, as there is too much experimental evidence that it is hard to overthrow.
Starting from the famous Mercury perihelion, going through Sir Arthur Eddington’s obser-
vation of light bending during the 1919 solar eclipse and ending with the newest gravitational
waves detection. First try of modified physics for solving dark matter problem is the so-called
MOND (MOdified Newtonian Dynamics). It was aimed to describe the flattening rotation
curve of spiral galaxies. In fact the hypothesis added a slight modification to the Newton’s
second law of dynamics  
a
F⃗ = m⃗a, (2.16)
a + a0
where a0 is an arbitrary parameter of the acceleration dimension and is predicted to be very
small a0 ∼ 10−10 m/s2 . The model seemed to describe well the rotational curves of galaxies
given only the baryonic mass of a galaxy. However it miserably failed in different applications.
First of all it does not eradicate the need for dark matter, as there is still the problem of
missing mass in different galactic systems. Secondly, it does not describe correctly other
phenomena, such as collisions of galaxies. Moreover, the idea can be considered inelegant
and full of ad hoc assumptions.
Another, more general approaches are generalizations of the General Relativity. They
assume that standard GR is valid in limited cases and the Einstein-Hilbert action
Z
√
SEH = d4 x −gR, (2.17)

where R is Ricci scalar, can include another gravity-related terms. As there are many ideas
for extension of this action, let me shortly mention some of the most popular. Introducing
higher order curvature terms into the Einstein-Hilbert action is a natural treatment. These
terms are topological invariants which can consist not only of Ricci scalar, but different
contractions of Ricci and Riemann tensors - the Gauss-Bonnet theory is an example here,
with its term
G = Rµνρσ Rµνρσ − 4Rµν Rµν + R2 . (2.18)
More general extensions are named as f (R) theories, with action
Z
√
Sf (R) = d4 x −gf (R), (2.19)

where f (R) is some generic function of the Ricci scalar. There is also even more general
family called scalar-tensor theories, where an additional scalar field is introduced. The
CHAPTER 2. THEORETICAL BACKGROUND 14

example of such theory is the Brans-Dicke theory of gravitation, with action

Z  
4
√ ω µ
SBD = d x −g ϕR − ∇µ ϕ∇ ϕ , (2.20)
ϕ

where ϕ is the scalar field and ω is a dimensionless coupling constant. In this theory ϕ plays
the role of a spatially varying gravitational constant G.
To summarize, alternative theories of gravity bring a plethora of new physical phenomena.
It considers various aspects of research - new black hole solutions, cosmological modifications
and even changes of stellar evolution. However they do not fully remove the need for excessive
mass in galaxies, therefore they cannot be considered a final solution to the dark matter
problem.

2.1.4 Macroscopic objects

At last but not least let me briefly discuss the third branch of dark matter research - the
macroscopic objects. This idea suggests the dark matter merely consists of non-shining
objects, no new particles or new gravity are needed. The so-called MACHOs (Massive
Astrophysical Compact Halo Objects) are the bodies that build the galactic halos and are
hardly detectable. These can be black holes, neutron stars, brown dwarfs, jupiters (aka
gas giants) or another hypothetical exotic objects such as boson stars, gravastars or quark
stars. Moreover invisible structures formed on the visible objects can also be classified as
macroscopic objects (I shall discuss this topic in the next subsection).
Finally, to summarize this subsection. I believe that the Nature does not really care
about human’s segmentation of ideas. Therefore it will rather not listen to our quarrels
and prayers - the answer to the dark matter quest might be the golden mean between these
seemingly separate branches. The macroscopic objects might be built from new particles
that undergo modified laws of gravity.

2.2 Hairy black holes

Everyone who studies physics realize at some point that physicists truly adore the language
of metaphors and hilarious names. Terms such as black hole, wormhole or hair sound rather
funny to a stranger. However their symbolic meaning is enormous as they capture the depth
and complication of the physical structure of these objects.
During the so-called golden age of General Relativity (roughly 1960s and 1970s) many
fascinating and inspiring ideas were discovered and introduced. Among them was the famous
no-hair conjecture, named by Jacob Bekenstein and advertised to broad auditorium by John
Archibald Wheeler.
It states that solutions of Einstein-Maxwell equations (General Relativity with electro-
magnetism) can be fully described by merely three physical parameters - mass, angular
momentum and electric charge. General Relativity predicts that any physical process of
CHAPTER 2. THEORETICAL BACKGROUND 15

gravitational collapse should lead to the emergence of a black hole. Apparently the end
state of this process should be independent of the initial conditions. Naturally the values of
these three parameters would depend on the initial conditions, but the functions describing
the solution will belong to the same family and have the same form. This is the reason
behind the statement black holes have no hair, as they seem to not have any structures that
would alter the three-parameter picture, such as the outer configurations of some matter
fields.
The conjecture inspired the truly rigorous formulation of the problem in question, the so
called uniqueness theorem. First mathematically rigorous proof of the uniqueness theorem of
the black hole solution was given by Werner Israel in his pioneering paper from 1967 [50]. He
showed that the Schwarzschild metric is the unique static asymptotically flat solution of the
Einstein vacuum field equations. Among all, in proving uniqueness theorems for black holes,
the key role is played by Killing vectors describing symmetries of the underlying spacetimes.
Moreover, the black hole event horizon is identified with a null hypersurface to which the
aforementioned Killing vectors are tangent (orthogonal). Soon after, Israel generalized his
concepts on the Reissner-Nordstrom solution in the realm of Einstein-Maxwell theory. Later
on, the uniqueness was found for the rotating solutions - the stationary and axisymmetric
Kerr and Kerr-Newman metrics - by Brandon Carter [51] and Stephen Hawking and George
Ellis [52].
However it is not always the case, as going beyond the standard General Relativity
brings the possibility of obtaining the hairy solutions. Such solutions are possible to obtain
considering scalar fields, vector fields or non-abelian Yang-Mills fields (I will briefly review
recent advances and examples below). Including new degrees of freedom to the theory
of gravity can vastly change the structure of the obtained spacetimes. Because of that
there are many examples that involve different stable matter field configurations emerging
outside black hole event horizons. Non-trivial hair naturally appears in standard GR but in
asymptotically anti-de Sitter (AdS) spacetimes. In asymptotically flat spacetimes obtaining
hair requires a little bit more effort.
One of the first fields studied in gravitation was the scalar field, which is relatively simple
to model and can phenomenologically mimic more complex matter contributions. The scalar
fields gained on physical significance in 2012, when the Large Hadron Collider at CERN found
the evidence for the Standard Model Higgs boson. This discovery has shown that scalar fields
can exist fundamentally in nature and their meaning seems bigger than just toy models.
Depending on the specific construction of the Langrangian, field composition, couplings
and symmetries the analysis of the solution could provide three distinct outcomes [53].

1. A strict no-hair theorem - giving the lack of possibility for the formation of a non-trivial
scalar.

2. A no-hair theorem under some set of assumptions, but hairy solutions if any of these
assumptions are broken.

3. A hairy solution, where the scalar profile is always non-trivial.

CHAPTER 2. THEORETICAL BACKGROUND 16

Let me now list a couple of examples of theories belonging to each group. The Lagrangian
1 1 1
L = R − ∇µ Φ∇µ Φ − µ2 Φ2 , (2.21)
4 2 2
corresponds to scalar-vacuum (µ = 0) and massive-scalar-vacuum (µ ̸= 0) theories. Both
of them belong to the first group, as they exhibit clean no-hair behaviours, which were
proved by Chase [54] and Bekenstein [55], respectively. Second class contains theories such
as V-scalar-vacuum
1 1
L = R − ∇µ Φ∇µ Φ − V (Φ), (2.22)
4 2
where except gravitational and kinetic term we have a generic potential. Massive-complex
scalar-vacuum theory (with complex scalar field and V (Φ) = µ2 Φ† Φ) exhibits similar proper-
ties. These theories show both hairy and no-hair behaviours, depending on certain assump-
tions and parameter ranges. They were studied by Heusler and Straumann [56], Bekenstein
[57], Pena and Sudarsky [58]. Finally the example of the third class is Einstein-Skyrme
theory
1 1
L = R − ∇µ Φa ∇µ Φa − κ|∇[µ Φa ∇ν] Φb |2 , (2.23)
4 2
where only hairy solutions are present, such as Droz-Heusler-Straumann [59] and its gener-
alizations.
The pure scalar hair is not the only option that is being considered by researchers. The
simplest generalization is the addition of electromagnetic field and granting the scalar field
electric charge by including it in the covariant derivative
1 1 1
L = R − Fµν F µν − (∇µ Φ − iqAµ Φ)† (∇µ Φ − iqAµ Φ) − µ2 Φ† Φ, (2.24)
4 4 2
with q being the charge of the scalar field. In this theory, hairy solutions emerge naturally in
asymptotically AdS spacetimes. This feature has been widely used in the realm of holography
aka AdS/CFT correspondence., which I will describe in the following section. If one wishes
to obtain a hairy solution within this theory in an asymptotically flat spacetime, the usage
of the so-called box boundary conditions is imminent.
Recently it was shown that rotating black holes can undergo a process of spontaneous
scalarization in the Gauss-Bonnet gravity
1 1
L = R − ∇µ Φ∇µ Φ + ηΦ2 G, (2.25)
2 8
where η is a coupling constant and G is the Gauss-Bonnet invariant defined before in equation
(2.18). Moreover the scalarization appears to be induced by the black hole spin, which was
investigated by several authors [60, 61].
One may notice this abundance of theories can serve as a motivation to test various
scenarios of matter composition. The same situation takes place in searches for dark matter
candidates. One can wonder how many hypothetical terms and couplings may be attached
CHAPTER 2. THEORETICAL BACKGROUND 17

to standard GR gravity (or extended) and elaborated under various circumstances. This is
one of the main targets of the research conducted for this thesis.
The particle oriented dark matter theory gives a birth to various candidates and moreover
there are some experimental clues for the correctness of these models. However, it is the space
where dark matter was originally observed and serves as the main theatre of its evolution.
Seeking for new space-oriented dark matter physics is truly necessary. The exoticism of
these new fields can be a source of interesting phenomena, such as new stellar objects and
black hole hair like clouds and condensates. Hairy black holes are usually considered as
theoretical concepts, but can dark matter fields form condensates on the compact objects
in astrophysical environment? Moreover the black holes can serve as a high energy physics
laboratories, which can host subtle dark matter related effects. Building this theoretical
literature will allow astronomers and other experimentalists to conduct observations and
precise measurements of space based phenomena, that can carry the traces of dark matter
in imaginable future.

2.3 AdS/CFT correspondence

One of contributions collected in this thesis utilizes the AdS/CFT correspondence in mod-
elling of the dark matter related physics. Therefore I consider it justified to briefly describe
fundamental concepts of the gauge/gravity duality.
The anti-de Sitter/Conformal Field Theory (AdS/CFT) correspondence is a duality orig-
inating from the string theory, which links a n-dimensional gravitational theory with a
(n − 1)-dimensional quantum field theory. On the first side one has a theory of (quantum)
gravity, which apart from gravitational components can contain different matter fields. Such
Lagrangian possesses a set of symmetries, which are of the key importance, because they
persist on both sides of the correspondence. This gravitational theory lives in the bulk of
the n-dimensional AdS spacetime. It should be noted that this family of spacetimes has
the asymptotic boundary, which naturally has the dimensionality reduced by one. On the
second side, there is a quantum field theory, living in the Minkowski spacetime, on the afore-
mentioned boundary of the AdS spacetime. Depending on the specific set of fields, their
symmetries and couplings the weakly coupled gravitational theory may refer to boundary
field theory that can describe different strongly coupled many-body systems.
This incredible idea was proved by Juan Maldacena in his famous 1998 paper [62], which
is one of the most cited papers in high energy theoretical physics. The proof itself is com-
plicated, but simply speaking it relies on the thermodynamical properties of both theories
presented in the language of path integrals. One ought to consider the partition function,
which gives a complete description of the field theory, as it is possible to produce different
relevant quantities from it. For both sides of the duality we can compute separate partition
functions, one for the bulk theory and another for the boundary. Finally it appears, if the
bulk partition function is taken to the limit of the AdS boundary it is equal to the boundary
CHAPTER 2. THEORETICAL BACKGROUND 18

theory partition function, in other words

ZCF T [ηµν ] = ZAdS [gµν |∂ ] ≃ eSg [gµν |∂ ] , (2.26)

where SG is the action of the gravitational bulk theory and ∂ represents the boundary of the
AdS spacetime. This fact leads to the conclusion that both theories describe systems which
share the same physical properties and behaviours.
Such observation can be utilized in various ways. The most important is the overcoming
the problem of solving a strongly coupled theory. Instead of struggling with a correlated
many-body system, one can find the gravitational weakly coupled counterpart, which can
be solved even in the perturbative regime. The gauge/gravity correspondence guarantees
that the easily computed quantities on the gravitational side can be translated to relevant
parameters in the boundary field theory. The adequate correlation functions can be obtained
via standard path integral relation
δ n log Z
⟨O1 (x1 )...On (xn )⟩ = , (2.27)
δJ1 (x1 )...δJn (xn )

where functional derivatives are taken with respect to specific sources.

The gauge/gravity duality found many applications, just to mention a few disciplines -
in modelling condensed matter systems, neutron star matter and even in collider nuclear
physics.
As an illustrative example let us take a simple holographic superconductor [64]. On the
gravitational side of the theory we consider a charged complex scalar field
Z  
4
√ 1 µν 1 † µ µ 2 2
S= d x −g R − 2Λ − Fµν F − (∇µ ψ − iqAµ ψ) (∇ ψ − iqA ψ) − µ |ψ| ,
M 4 2
(2.28)
where Λ is the cosmological constant and is naturally negative, which guarantees that the
resulting spacetime will be asymptotically AdS. If one varies the above action with respect
to the fields, the equations of motion will be the result. For a certain set of conjectures and
parameters a Schwarzschild-AdS hairy or hairless black holes are the solutions.
Now if one wishes to interpret such system with the AdS/CFT toolbox, the whole prob-
lem should be taken to the asymptotic boundary of the AdS spacetime. In terms of a radial
coordinate, that is r → ∞. After such operation the equations of motion reach their asymp-
totic forms and the asymptotic solutions can be computed. By virtue of the gauge/gravity
duality the behaviours of the bulk fields are related to the operators and currents which are
present in dual field theory. In this example, the scalar field asymptotic form corresponds
to the condensation operator (order parameter) of the superconducting condensate.
Detailed relations between physical quantities of the bulk gravitational theory and the
boundary CFT are collected in the holographic dictionary. For example the metric tensor
in the bulk corresponds to the stress-energy tensor of the boundary theory. Similarly gauge
field potential corresponds to some currents.
CHAPTER 2. THEORETICAL BACKGROUND 19

By using this machinery one can compute relevant quantities, that describe the boundary
strongly coupled field theory using only the computations from the gravitational bulk. In
this way one can note that the scalar condensation operator undergoes a phase transition of
the second kind and the conductivity reveals characteristic patterns being the result of the
superconducting gap.
Chapter 3

Methods

In the previous section I described the physical foundations of problems I was dealing with
during my doctoral research. Now I shall focus on the technical side of the research. I will
briefly explain the theoretical methods I used to model considered systems and then the
numerical ones that helped me to extract viable physical information from the theory.

3.1 Theoretical methods

Below I list several theoretical methods which were used for establishing the problems con-
sidered in all papers included in this thesis. I give a basic description of these methods, that
allowed me to set up the model, derive the equations of motion, introduce the boundary
conditions and analyse the solutions.

3.1.1 Equation of motion

The research collected in this thesis utilizes rather standard methods of contemporary theo-
retical physics in the field of gravitation. I start with defining a theoretical set up, by writing
down the action Z
√
S = dn x −g(R + LM ), (3.1)

where R is the Ricci scalar and L is the matter Lagrangian density. At this point one
defines the composition of the physical space, namely the fields that take part in the theory.
Moreover the dimensionality n of the space is defined, too. Here the gravitational part of
the action comes from standard General Relativity, no other curvature terms were used in
this research. At this point some general properties of the theory might be investigated such
as symmetries.
Using the widely known variational principle,

δS = 0, (3.2)

20
CHAPTER 3. METHODS 21

we calculate adequate functional derivatives, and we derive equations of motion, that define
the behaviour of the matter fields.
Obtaining the gravitational part requires varying the action with respect to the metric
tensor gµν . With few transformations we obtain the Einstein equation
1
Rµν − gµν R = Tµν , (3.3)
2
which gives us a complete set describing both the spacetime and the fields dynamics.
At this point the problem needs to be precisely specified. Depending on the physical
problem and other theoretical justifications one should make a series of assumptions on
what the final solution should look like. This requires us to select a form of the metric and
fields, which should be put into the generic equations of motion in order to obtain a solveable
system of differential equations.
For example an ansatz for static and spherically symmetric spacetime is given by

ds2 = gµν dxµ dxν = f (r)dt2 + g(r)dr2 + r2 dθ2 + r2 sin2 θdϕ2 (3.4)

and a sample dynamical scalar field would take a form

Ψ = ψ(r)Ylm (θ, ϕ)e−iωt , (3.5)

where Ylm are spherical harmonics. Substituting this ansatz into the equations of motion
gives us a set of differential equations which sometimes can be solved analytically, but more
often numerically.

3.1.2 Boundaries and asymptotics

Solving a differential equation requires providing boundary conditions on both ends of the
domain. Usually, some of these boundary conditions are straightforward, like gtt = 0 on the
event horizon. On the other hand, they can originate from the properties of the spacetime,
for example axial symmetry of the spacetime forces the axial symmetry of the fields inside.
However there are situations where the boundary conditions are not given by physics
right away. Instead one has to find an analytical behaviour of the solution in some certain
limits and then impose it on the numerical solution.
Such cases take place in many black hole physics scenarios. Let us consider a few exam-
ples.
First when we have several fields in the vicinity of a black hole. If one of the fields
is precisely bound, but the remaining are not, one ought to take a limit of the equations
of motion, introduce the known boundary conditions and then after several mathematical
transformations obtain some conditions on the remaining functions.
In the second case, we have a similar situation to the first one. However one of the
equations of motion has a singularity in a spatial point, for instance the event horizon. If
adjust to zero value, a field on a horizon brings regularity of the system and eradicates the
singularity from the equation then this is the boundary condition we were looking for.
CHAPTER 3. METHODS 22

Third example is a bit more complex and for an illustration we will take the case of
the holographic superconductor from one of the contributions contained in this thesis. The
equation will be stripped from physical constants that are irrelevant in this explanation. In
the asymptotically AdS spacetime the metric behaves like

gtt ∼ r2 , (3.6)

when r → ∞. Therefore by taking a simple limit of an equation of motion, we are guaranteed

to obtain infinite terms. This is not something that we are looking for. However the Frobenius
method comes with a rescue. The equation of motion for the scalar field in this particular
case yields  ′   2 
′′ f 2 ′ ϕ 2 ψ
ψ + + ψ + −m = 0, (3.7)
f r f f
where gtt = f (r) = r2 (1 − (rh /r)3 ), ϕ is the electric potential, m is the mass of the scalar
and prime denotes a derivative with respect to r.
If one takes the limit r → ∞ of the above equation to the leading order terms, we get
4 m2
ψ ′′ + ψ ′ − 2 ψ = 0. (3.8)
r r
Having an equation in such form allows us to use the Frobenius method and using the
expansion X
ψ(r) = rρ cn r−n . (3.9)
n

By differentiating the series and substituting it into the equation we obtain the characteristic
equation for leading order exponents ρ,

(ρ − n)(ρ − n − 1) + 4(ρ − n) − m2 = 0, (3.10)

with final solutions

1 √ 
ρ± = −3 ± 9 + 4m2 , (3.11)
2
which provides us with asymptotic solution
 
ψ1 ψ2 1
ψ = −ρ+ + −ρ− + O −ρ− +1 . (3.12)
r r r
At the end of the day such analysis delivers us the boundary conditions for the numerical
attempts. Moreover it gives the constrains on the physical parameters, such as scalar mass.
In this particular case in order to obtain a stable scalar condensate one requires it to vanish
in infinity, therefore ρ± must be negative. On the other hand, the exponents also must be
real, so the scalar does not have oscillating modes. This requires the term under the square
root in (3.11) to be positive and gives the bounds on the mass parameter
9
m2 ⩾ − . (3.13)
4
CHAPTER 3. METHODS 23

Such limitation in these theories is called Breitenlohner-Freedman limit for tachyonic fields
in AdS spacetimes.
Summarizing this subsection. The analysis of boundary behaviour of the equation in
question brings not only the clues for selecting the boundary conditions, but also useful
information about the role of some physical parameters in the theory.

3.1.3 Other quantities and benchmarks

Finding the precise solution of the equations of motion is definitely not the final goal of most
of research projects. It is necessary to study the behaviour of relevant physical quantities.
In this short paragraph I will discuss the calculation of free energy of the physical state of
the system, as it is the most used benchmark in all research conducted for this thesis.
Helmholtz free energy, or simply free energy, is a thermodynamical potential that was
introduced for gas related processes, where the temperature and volume can be controlled.
Free energy also sets the equilibrium conditions and plays a key role in phase transitions.
The preferable physical state is the one where free energy is minimal, which is an obvious
analogy to the least action principle in the realm of field theory.
Classically free energy is defined as
F = U − T SE , (3.14)
where U is the internal energy, T is the temperature and SE is the entropy.
Surprisingly this definition is still valid in black hole physics. However the right-hand side
components have different definitions and meanings. Let us start one by one, the internal
energy counterpart for a gravitational system is the quasilocal energy introduced by Brown
and York in their famous work on conserved quantities derived from the gravitational action
[63]. The definition of the energy in the gravity is a vast problem itself, here however the
authors define it relatively with respect to the flat Minkowski spacetime. It takes the form
of I
1 √
U =− lim (k − k0 ) σd2 x, (3.15)
8π r→rb S 2
where k is the external curvature of the studied metric, while k0 is the external curvature
of the flat spacetime. Next, σ is the determinant of a metric on a time-like hypersurface S 2
with radius rb , which is present in the limit.
Following quantities such as temperature and entropy are well defined in black hole ther-
modynamics. The definition of gravitational analogue for temperature requires the surface
gravity evaluated on the event horizon of the black hole [65]
1
κ2 = − (∇a χb )(∇a χb ) , (3.16)
2 Horizon

where χa is the Killing field of the considered spacetime. Then the Hawking temperature is
simply defined as
κ
TH = . (3.17)
2π
CHAPTER 3. METHODS 24

On the other hand, the black hole entropy is simply proportional to the area of the even
horizon I √
1
SE = hd2 x, (3.18)
4 SH2
with h being the metric determinant on the event horizon.
By using these quantities one can obtain the free energy for different black hole solutions.
However it is not a fully general method. Unfortunately it is unable to capture non-trivial
configurations of fields present in the vicinity of a black hole, such as scalar (or more compli-
cated) hair. Therefore for such research it is recommended to compute free energy straight
from the basis of the theoretical set up, using the classical action

F = T Scl . (3.19)

The term classical refers to the signature of the metric. In relativistic physics the time
component of the spacetime always has the different sign than the other, spatial, components.
Depending on the convention, it might be negative or positive. Nevertheless the goal is to
impose the Euclidean signature to the metric, so all coordinates have the same sign. To
achieve such goal one has to perform the Wick rotation on the time coordinate

t → iτ, (3.20)

with τ being the imaginary time.

3.2 Numerical methods

Most of the field theory research I dealt with can be brought down to boundary value
problems (BVPs). These however are not easily solvable, especially if one considers a more
sophisticated gravitational set up, such as Kerr metric, or if the problem is non-linear.
Some of them, can be solved using analytic methods such as Sturm-Liouville approach.
Nevertheless in general, it requires making some assumptions that are not always possible
and we have to use numerical approximations to the problem.
The core philosophy of various numerical methods is the same in most cases – translating
a differential problem into an algebraic one. Below, I briefly discuss the methods I used in
works being the part of this thesis and illustrate them with simple examples.

3.2.1 Pseudo-spectral collocation method

The key idea of (pseudo-)spectral methods is to approximate the sought function in a dif-
ferential equation with a combination of spectral functions [66]. We require such functions
ϕi to be orthogonal Z
ϕi (x)ϕj (x)dx = δij . (3.21)
M
CHAPTER 3. METHODS 25

The most popular choices are sines and cosines, where we talk about Fourier spectral meth-
ods, or Chebyshev polynomials. Chebyshev polynomials of the first kind can be defined in
the following, recursive way
T0 (x) = 1,
T1 (x) = x,
Tn+1 (x) = 2xTn (x) − Tn−1 (x),
or with a more general formula
Tn (x) = cos(n arccos(x)). (3.22)
Second formulation reveals that Chebyshev polynomials are in fact cosines in a disguise.
However they are used in problems that do not require the solution to be periodic. The
domain of Chebyshev polynomials is the interval [−1, 1]. Therefore if one wants to use
this technique, appropriate transformations of the independent variable are required. After
preparing the equation to a Chebyshev-eligible form, we can solve it. For illustrative purpose,
we use a standard second order ordinary differential equation as an example
y ′′ + g(x)y ′ + h(x)y = s(x), (3.23)
where g(x), h(x) and s(x) are generic functions of x. To solve a BVP with the Chebyshev
pseudo-spectral method one has to substitute a series of Chebyshev polynomials
N
X
y(x) ≈ ai Ti (x), (3.24)
i

into the given equation, where ai are the series coefficients (and become the values, which we
want to compute as a solution to the problem). Mind that N is the number of polynomials
used and it also denotes its highest order. The great advantage of this method is that
the derivatives can be computed analytically, as it is naturally straightforward to compute
a derivative of a polynomial. Then comes a crucial point, where one has to collocate the
equation on a grid. For the sake of convergence one should not use any uniform grids. Instead
natural grids for this method are Chebyshev-related ones – either zeros (aka Gauss-Lobatto
grid) or extremes of the highest N -order polynomial. The former one is defined as
xn = cos(nπ/N ), (3.25)
and it gives a high density of points near the edges of the numerical domain, while the
central part remains significantly less dense. After substituting these points into the master
equation we achieve the goal of translating a differential problem into an algebraic one. We
obtain a set of N algebraic equation with N unknowns


f1 (a1 , a2 , · · · , an ) = s1 ,


f2 (a1 , a2 , · · · , an ) = s2 ,
(3.26)
...



f (a , a , · · · , a ) = s ,
n 1 2 n n
CHAPTER 3. METHODS 26

where the ai coefficients ought to be found.

The next problem is to solve such system. In case of a linear differential equation it
is only a matter of computing an inverse of the matrix. However in physical applications
we usually deal with non-linear differential equation. In such case one has to use iterative
schemes for minimizing the equations, for instance Newton-Raphson method.
One of the greatest advantages of the pseudo-spectral method is its convergence. Al-
though it requires only a few points in a spatial discretization in order to achieve high
accuracy. Moreover a properly constructed algorithm should demonstrate exponential con-
vergence with exponentially decreasing error with an increase of N . This method was used
in the contribution [72].

3.2.2 Shooting method

A BVP is usually defined as a second order differential equation with its values (Dirichlet
boundary conditions) specified on the boundaries or the values of the derivatives (Neumann
boundary conditions). The idea of the shooting method can be brought down to translating
a BVP into an initial value problem (IVP).
Let us proceed with an exemplar ODE, exactly the same as in the previous paragraph

y ′′ + g(x)y ′ + h(x)y = s(x), (3.27)

with Dirichlet boundary conditions

y(xL ) = a, y(xR ) = b, (3.28)

denoting values of the solution on the left and right-hand side of the numerical domain.
We can reconsider equation (3.27) and instead of using boundary conditions (3.28), utilize
an alternative pair
y(xL ) = a, y ′ (xL ) = η, (3.29)
which describes the value of the solution and its derivative on the left-hand side of the
numerical domain. The parameter η is arbitrary at the moment. Our second order ODE
can be rewritten into a set of two first order ODEs

y ′ = u,
u′ + g(x)u + h(x)y = s(x),

which allows us to utilize a variety of solvers for IVPs and integrate these equations from
xL to xR . We can use standard Runge-Kutta schemes, Adams-Bashford method or more
sophisticated adaptative stepsize algorithms such as Dormand-Prince or Fehlberg methods.
After the integration we check if the solution fulfills the right-hand side boundary con-
dition 3.28. The odds that our initial guess of η will bring the desired behaviour of our
function are rather small. Therefore our goal is to minimize the objective function

F (η) = |b − y(xR ; η)|, (3.30)

CHAPTER 3. METHODS 27

which is the absolute error between computed and target value of the solution. Alternatively
one can use different error, such as squared error. Having the objective function defined we
have to minimize it with a chosen algorithm. It can be a simple bisection search, gradient
descent or Newton-Raphson procedure. After several iterations we should obtain the value
η, for which the solution fulfills the BVP boundary conditions. This numerical method was
used in contributions [70, 71].

3.2.3 Spectral differentiation

Spectral differentiation relies on constructing a differentiation matrix, which is applied to
a function discretized on a grid. This method is thoroughly described in [67], here I only
briefly describe its usage.
It sounds similar to techniques such as finite differences method (FDM) and from the
outside it really is, especially in the simple usage of this method. At the end of the day
a differential equation can be translated to a matrix equation, with specific differentiation
matrices as the components.
One of the greatest advantages of this method is the accuracy of the computed deriva-
tive. In case of FDM, usually the differentiation error is proportional to the stepsize of the
grid. Because of that it is necessary to exploit relatively large grids, sometimes containing
thousands of points, in order to achieve acceptable results. However this is not the case
of spectral differentiation, where only few grid points are required for obtaining satisfying
results.
Suppose we can interpolate any function in our domain x ∈ [−1, 1] with a polynomial,
e.g. Chebyshev polynomial. Let p(x) be an interpolant of vi values of some function, defined
on the Chebyshev points (3.25). On a small grid, with N = 2 (three points in the grid
x0 = 1, x1 = 0 and x2 = −1), we have a quadratic polynomial
1 1
p(x) = x(1 + x)v0 + (1 − x)(1 + x)v1 + x(x − 1)v2 . (3.31)
2 2
The derivative of the interpolant yields
1 1
p′ (x) = (x + )v0 − 2xv1 + (x − )v2 . (3.32)
2 2
The differentiation operation can be rewritten as a matrix equation

w = D2 v, (3.33)

where w is a vector of values of the derivatives of v, defined on the grid and the differentiation
matrix is defined as  3 
2
−2 12
D2 =  12 0 − 12  . (3.34)
1 3
−2 2 −2
CHAPTER 3. METHODS 28

This result can be generalized to N number of grid points and used in discretization of
differential equations.
Taking the equation from previous paragraphs (3.27) as an example, we write
2
DN y + Ig DN y + Ih y = Is , (3.35)

where y is a solution vector and I are diagonal matrices containing the values of the functions
g, h and s. These matrices can be collected into a single operator
2
L = DN + Ig DN + Ih . (3.36)

At this point one should introduce the boundary conditions to the system by substituting
adequate rows and columns of the L matrix and right-hand side of the equation. Finally,
this allows us to straightforwardly solve the equation

y = L−1 Is , (3.37)

where we invert the L matrix with a preferred method. In case of a non-linear equation
the algorithm gets only little more complicated. One can assume the form of the solution
and substitute it into the non-linearity as the initial guess. Then one ought to solve the
problem (3.37) iteratively until stopping conditions are fulfilled. This method was used in
contributions [73, 74].
Chapter 4

Original contributions

In this chapter I include the copies of the articles that were co-authored by me during my
doctoral studies. They do not present a strict cycle, but they are oriented around the issue
of interactions between visible and dark sectors. For every article I present a short summary
listing the theoretical basis of the article and its most important results.

4.1 Summary of articles

4.1.1 Ultra-compact spherically symmetric dark matter charged
star objects
Reference
B. Kiczek and M. Rogatko, Ultra-compact spherically symmetric dark matter charged star
objects, JCAP 09 (2019) 049.

Motivation
As it was mentioned in the previous chapters, the addition of dark sector fields to the known
systems may change interestingly their behaviour. In this work, the goal was to study the
influence of the additional dark sector U(1) gauge group on the solutions describing ultra-
compact starlike objects. Naturally the dark photon is mixed with the Maxwell field via
kinetic mixing, therefore we expect new dependencies on some kind of effective charge, being
a mixture of the electric charge and a hypothetical charge originating form the dark sector
group.
Therefore we wish to obtain the solution of the spacetime around a compact spherically
symmetric object charged, in general, with electric charge, the dark charge and some mixture
of both. After obtaining such solution it is interesting to see what properties it has.

29
CHAPTER 4. ORIGINAL CONTRIBUTIONS 30

Methods
We start with the definition of action
Z  
4
√ 1 µν α µν 1 µν
S = d x −g R − Fµν F − Bµν F − Bµν B , (4.1)
4 4 4

where Fµν = 2∇[µ Aν] is the ordinary Maxwell field while Bµν = 2∇[µ Bν] is the U(1) gauge
field related to the dark sector.
By varying the action over all fields we get the equations of motion,

Gµν = Tµν , (4.2)

α
∇µ F µν + ∇µ B µν = 0, (4.3)
2
α
∇µ B µν + ∇µ F µν = 0. (4.4)
2
In the first approximation we pick the spherically symmetric ansatz for the line element
given by

ds2 = −e2ϕ(r) dt2 + e2λ(r) dr2 + r2 (dθ2 + sin2 θdϕ2 ), (4.5)

with gauge fields in form

Q Qd
At = − , Bt = − . (4.6)
r r
By plugging these quantities into Einstein equations and solving them we get

2M Q2
e−2λ(r) = 1 − + c2 , λ(r) = −ϕ(r), (4.7)
r 4r
with the effective charge Q2c = Q2 + Q2d + αQQd , which constitutes the solution of the
metric outside spherically symmetric ultra-compact starlike object.
Using this derived spacetime we considered a fluid sphere, with Tolman-Oppenheimer-
Volkoff (TOV) equations derived in the end
 
′
Qc (r) Qc (r) ρ + p(r) h Fc (r) r3 p(r) Qc (r)2 ih 2m(r) Fc (r) i−1
p′ (r) = − m(r) + + − 1 − − .
2r4 r2 2 2 8r r r
(4.8)
In the fifth section of the paper we numerically investigated the emergence of scalar hair
and the influence of the dark charge on their distribution.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 31

Results
First of all, we derived the gravitational metric of the outside of spherically symmetric ultra-
compact starlike object. The metric can be seen as a generalised Reisner-Nordstrom solution,
with an effective charge being a composition of electric charge, the dark sector related charge
and a mixing term.
Then we considered a dark matter charged fluid sphere, where TOV equations have been
derived. Solving them was beyond of the scope of this article, however, we hope to return to
these calculations elsewhere, as the dark sector impact on the stellar evolution constitutes a
fascinating problem.
Next, the stability bounds on mass and charge for ultra-compact object were found. It
was revealed that the compact object with hidden sector charge can reach smaller size than
the object made from the visible components only.
Finally we investigated the possibility of emergence of the visible scalar hair. We checked
what are the conditions and how these solutions behave. It was found that the large quantity
of dark sector charge in the system results with shrunk and smaller scalar condensate.

4.1.2 Influence of dark matter on black hole scalar hair

Reference
B. Kiczek and M. Rogatko, Influence of dark matter on black hole scalar hair, Phys. Rev.
D 101, 084035 (2020).

Motivation
In the previous paper we have studied scalar hair around an ultra-compact object influenced
by U(1) dark matter sector. This work is a natural continuation of it. We have already seen
that in the probe limit, where the gravitational background and both gauge fields are fixed,
the visible scalar hair was shrinking with the growth of dark charge in the system. In this
work we quit the probe limit scenario and solve the fully backreacting system.

Methods
We start with the action
Z  Z
4
√  1 µν α µν 1 µν 2 2 2
√
S= d x −g R − Fµν F − Bµν F − Bµν B − |Dψ| − m ψ − d3 x −γK,
M 4 4 4 ∂M
(4.9)
as in the previous case Fµν is a Maxwell field strength tensor, Bµν is a strength tensor
of a hidden sector vector boson. The complex scalar field Ψ = ψeiθ , where θ denotes
the phase, is coupled only to the ordinary electromagnetic field by the covariant derivative
Dµ = ∇µ − iqAµ . The last term of the action is the Gibbons-Hawking boundary term of our
box, where the considered objects are contained. The box is a 3-dimensional hypersurface
with γ metric and the extrinsic curvature K.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 32

Varying the action gives us the equations of motion, for our fields we choose following
ansatz
dr2
ds2 = −g(r)h(r)dt2 + + r2 (dθ2 + sin θ2 dϕ2 ), (4.10)
g(r)
Aµ dxµ = ϕ(r)dt, (4.11)
Bµ dxµ = χ(r)dt, (4.12)
Ψ = ψ(r), (4.13)

where the metric is spherically symmetric and matter fields depend only on the radial coor-
dinate. Finally we get the equations of motion
q 2 rϕ2 ψ 2
h′ − rhψ ′2 − = 0, (4.14)
g2
 
′ 1 1 ′2 q 2 rϕ2 ψ 2 1 r
g +g + rψ + − + (ϕ′2 + αχ′ ϕ′ + χ′2 + m2 hψ 2 ) = 0, (4.15)
r 2 2gh r 2h
 
′′ 2 h′ 2q 2 ϕψ 2
ϕ + − ϕ′ − = 0, (4.16)
r 2h α̃g
   2 2 
′′ 2 h′ g′ ′ q ϕ 2 ψ
ψ + + + ψ + −m = 0, (4.17)
r 2h g gh g
 
′′ 2 h′ αq 2 χψ 2
χ + − χ′ + = 0. (4.18)
r 2h α̃g
We solve these equations simultaneously using the shooting method. In simple words,
we integrate the fields from the inner boundary to the outer one using the standard Runge-
Kutta 4 method. At the outer boundary we check if the boundary conditions are fulfilled.
If not we pick new initial conditions using the bisectional search and repeat the procedure
until reaching the convergence.
After obtaining adequate solutions we analyse the system using the thermodynamical
framework. Apparently the theory has three phases:

• Generalised Reisner-Nordstrom (RN) black hole (the analytic result of the previous
work),

• Hairy black hole (HBH),

• Boson star (BS).

We adopt thermodynamical treatment by calculating the free energy of the system

F = TH Scl , (4.19)

with Scl being the classical action, and checking what are physical implications of the presence
of the dark charge in the system.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 33

Results
We have investigated the influence of the dark sector U(1) gauge group related charge on the
possible configurations in the Einstein-Maxwell-dark photon-scalar theory. The main result
of this calculation is the phenomenon of shifting phase boundaries caused by the presence of
the dark charge. This process can be seen on the schematic phase diagram presented below.

μ
(A)
Boson
star
(B)

(D)

Hairy
BH RN-like
(C)
BH

T
Figure 4.1: The schematic phase diagram of the Einstein-Maxwell-dark photon-scalar theory
wit spherically symmetric asymptotically flat solutions.

The diagram is basically a plot on the µ-T plane, which stand for the chemical poten-
tial (which is a boundary value of the Maxwell field on the box boundary, the name is
a loose analogy to the AdS/CFT rigorously defined chemical potential) and the Hawking
temperature.
We found that adding more dark charge to the system significantly shrinks the parameter
space available for hairy black hole solutions to emerge. It happens both on the HBH-RN
boundary (second kind phase transition) and on the BS-HBH boundary (first kind phase
transition).
We also noted a very interesting thing. Namely there are points in the parameter space
that are completely immune to the presence of dark sector, as the system with these particular
parameters remains invariant under the change of dark charge amount. These are so-called
isosbestic points, which appeared to be well-known in the realm of condensed matter physics.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 34

4.1.3 Holographic DC SQUID in the presence of dark matter

Reference
B. Kiczek, M. Rogatko and K. I. Wysokiński, Holographic DC SQUID in the presence of
dark matter, JCAP 01 (2021) 063.

Motivation
Superconducting quantum interference devices or simply SQUIDs constitute one of the most
sensitive sensors constructed by humans. They are able to detect a slightest change of the
magnetic field, a.k.a. magnetic flux quantum.
A SQUID is a loop made of two Josephson junctions. A Josephson junction (JJ) con-
sists of two superconductors divided by a thin layer of non-superconducting material, i.e.
a metal or an insulator. Its characteristic feature is the flow of a direct current, which
magnitude depends on the phase difference between the Cooper pair wave functions in both
superconductors

I = Imax sin ϕ. (4.20)

Where ϕ = ϕR − ϕL , with ϕL(R) being the phases of the superconductor order parameter
on the left (right) side of the junction. If one applies external magnetic field to a SQUID,
the maximal current which can flow through it is given by
πΦ
Imax = Ic cos , (4.21)
Φ0
h
where Φ is the magnetic flux and Φ0 = 2e is the magnetic flux quantum.
The sensitivity of SQUIDs and interesting progress in the usage of superconductors in
the dark matter search [20, 21, 22] motivates us to pursue in that direction. We discuss the
usage of SQUID in possible dark sector searches.
In this work we used the AdS/CFT correspondence to build a holographic SQUID, where
we modelled an injection of a DM particle into one of Josephson junctions. To show the
effect of this interaction we compare the DM case with a DM-free scenario.

Methods
The first ingredient of the AdS/CFT based model is the adequate gravitational background,
which naturally must be asymptotically anti de Sitter. For this purpose we use the gravita-
tional action
Z  
√ 4
Sg = −g d x R − Λ , (4.22)

with a negative cosmological constant Λ = −6/L2 . Later on we use the AdS-Schwarzschild

black hole line element as a gravitational background
CHAPTER 4. ORIGINAL CONTRIBUTIONS 35

ds2 = −f (r)dt2 + f −1 (r)dr2 + r2 (dξ 2 + dy 2 ). (4.23)

2
With f (r) = Lr 2 (1 − r03 /r3 ) and the corresponding Hawking temperature of the BH is
3r0
TH = 4πL 2 , which is a very important quantity for the model in question.

The matter part of the action is the Abelian-Higgs sector with additional U(1) gauge
field belonging to the hidden sector
Z 
√ 1
Sdm+EM +H = −g d x − Fµν F µν − [∇µ ψ − i Aµ ψ]† [∇µ ψ − i Aµ ψ] − m2 |ψ|2
4
4

1 µν ζ µν
− Bµν B − Fµν B . (4.24)
4 4
We model the presence of the dark sector in one of Josephson junctions by using a
spatially dependent kinetic mixing parameter
2 /λ2
ζ(ξ) = α0 e−(ξ−ξ0 ) . (4.25)

With ξ0 being the centre of the affected JJ and λ a Gaussian broadening, which can be
loosely interpreted as a de Broglie’s wavelength of a DM particle. By varying the action and
substituting all relevant quantities into the obtained formulas we get our system of equations
of motion
   
2 1 2 2 ∂r f 1 2 1 2 2 m2
∂r |ψ| + 2 ∂ξ |ψ| + + ∂r |ψ| + 2 Mt − 2 Mξ − Mr − |ψ| = 0,
r f r f f r f f
(4.26)
h 2 1 i 2 1 ζ 1
ζ̃ ∂r2 Mt + ∂r Mt + 2 ∂ξ2 Mt − Mt |ψ|2 + ∂ξ ζ 2 ∂ξ Mt = 0,
r r f f 2 2r f
(4.27)
h i 1 ζ  
ζ̃ ∂ξ2 Mr − ∂ξ ∂r Mξ − 2 r2 Mr |ψ|2 − ∂ξ ζ ∂ξ Mr − ∂r Mξ = 0,
2 2
(4.28)
h  i
∂r f 2 Mξ
ζ̃ ∂r2 Mξ − ∂r ∂ξ Mr + ∂r Mξ − ∂ξ Mr − |ψ|2 = 0,
f f
(4.29)

which is a set of non-linear second order partial differential equations. By using the Frobenius
method we obtain asymptotic forms of the fields on the AdS boundary

|ψ (1) (ξ)| |ψ (2) (ξ)|

|ψ| = ∆(1)
+ ∆(2)
+ O(r−3 ), (4.30)
r r
ρ(ξ)
Mt = µ(ξ) − + O(r−2 ), (4.31)
r
Mr = O(r−3 ), (4.32)
CHAPTER 4. ORIGINAL CONTRIBUTIONS 36

J(ξ)
Mξ = ν(ξ) + + O(r−2 ). (4.33)
r
These expansions carry parameters such as µ, ρ, ν, J and ψ (i) , which are of key importance
as they constitute relevant physical quantities of the CFT living on the boundary of the bulk
AdS spacetime. They denote respectively the chemical potential, charge density, velocity
of the super current carriers, electric current and the condensation parameter (with i = 1
or i = 2, depending on the particular choice of the model). By solving the EOMs and
using these expansions we can obtain the information about the strongly coupled boundary
quantum field theory.
To solve these complicated equations we use a numerical method based on spectral meth-
ods. Precisely it is a pseudo-spectral Chebyshev collocation method. In brief, we assume
that our fields are combinations of Chebyshev polynomials with some factors. We substitute
this assumption into the EOMs and collocate the equations on a Gauss-Lobato grid. In
such way we translate a calculus problem into an algebraic one. Then we solve a nonlinear
system of algebraic equations using the iterative Newton-Raphson method, until we reach
convergence.

Results
As a warm up we check the simplified case of a homogeneous holographic superconductor
under the influence of the dark sector field. We find that its critical quantities such a critical
temperature and critical chemical potential are shifted due to the presence of the coupling.
This gives us a clue for further investigation in the full problem.
When we move to solving the equation of the holographic SQUID we start with the
DM-free scenario. We suppose that both Josephson junctions are identical, considering
every parameter from shape to chemical potential. In such case the SQUID is trivial, as no
interference is observed.
As we add the hidden sector into the consideration we observe an interference pattern,
which is generated due to the current imbalance between junctions, that finally causes an
effective magnetic flux. This gives us a criterion for a possible detection of such particle.
We also tested the dependence of the Gaussian width λ on the signal strength - the
maximal current. It can be seen on a figure below.
The strongest signal is seen when λ is very close to the width of the gap in the Josephson
junction. It is also an interesting clue for some possible experiments, as the gap width might
be engineered for a particular wavelength, corresponding to the mass, of a sought particle.
Naturally our considerations cannot be treated as some blueprint for an experiment. A
relevant calculations in a realm of conventional condensed matter and high energy physics
should be conducted. However they might be a bit challenging, as considering such phe-
nomenon on the holographic ground is much easier than in quantum field theory based
approach.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 37

×10−5
0.0

−0.2

−0.4

−0.6
δJ

−0.8
α
0.5
−1.0
0.05

−1.2
−1.00 −0.75 −0.50 −0.25 0.00 0.25 0.50 0.75 1.00
−2
Φ [10 ]

Figure 4.2: A dependence of a current difference as a function of magnetic flux. Two curves
represent different couplings of the hidden sector.

6.60

6.58
Jc [10−2]

6.56

6.54

6.52

6.50
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
λ

Figure 4.3: Critical current vs. width of the Gaussian incursion.

Erratum
During the collection of materials for this thesis an essential typos have been spotted in the
original article. After the equation (4.11) on page 18 following definitions shall be corrected:
F̃µν = 2∇(µ Ãν) → F̃µν = 2∇[µ Ãν] ,
B̃µν = 2∇(µ B̃ν) → B̃µν = 2∇[µ B̃ν] .
CHAPTER 4. ORIGINAL CONTRIBUTIONS 38

4.1.4 Axionlike dark matter clouds around rotating black holes

Reference
B. Kiczek and M. Rogatko, Axionlike dark matter clouds around rotating black holes, Phys.
Rev. D103, 124021 (2021).

Motivation
There is no doubt that axions constitute one of the most prominent candidates for dark
matter. Their physical properties such as stability and their ability to form large scale
structures, due to they possible low masses, have granted them this status. However, it is
interesting to evaluate if there are any chances on the emergence of axionlike DM structures
on the stellar scale.
The studies of dark matter at the stellar scale can be theoretically performed in the black
hole, as well as other compact object, surroundings. Conclusions from these investigations
can be used to apply similar calculations for regular stars, such as the Sun. Lessons learnt
from these considerations can be very beneficial for planning future experiments, as it is not
excluded that dark matter may form some kind of structures even in the Solar System. The
questions is what kind of structures are these and what do they look like?
In a recent paper Boskovic with collaborators [25] have shown, using an analytical approx-
imation, that axions may form stationary configurations around slowly rotating black holes.
In our paper we generalise this result on different masses of axions and angular momentum
of the hosting black holes.

Methods
We start with the action of the theory
Z  
4
√ 1 µν 1 µ µ2 2 k µν
S = d x −g R − Fµν F − ∇µ Ψ∇ Ψ − Ψ − Ψ ∗ F Fµν ,
4 2 2 2
where R is the Ricci scalar, Fµν = 2∇[µ Aν] is the Maxwell field strength tensor and Ψ is
the scalar field (axion) with mass µ. The respective equation of motion for the axion field is
following
k
∇µ ∇µ Ψ − µ2 Ψ − ∗ F µν Fµν = 0.
2
The obtained equation of motion (EoM) is a generalised Klein-Gordon equation with a
source term, being the invariant of the Maxwell field. This particular source term plays a
very important role in this scheme, because if it vanishes (when the magnetic field is zero or
when the spacetime is spherically symmetric) the equation then undergoes a no-hair theorem
and no axionic configurations can be established. Thus, in this situation we can speak of
magnetically induced axionlike hair.
We consider the above EoM in two gravitational backgrounds in the probe limit scenario.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 39

1. Kerr black hole immersed in uniform magnetic field using Wald’s technique,
 
2 2M r 4M ra sin2 θ Σ Ξ sin2 θ 2
ds = − 1 − dt2 − dtdϕ + dr2 + Σdθ2 + dϕ ,
Σ Σ ∆ Σ
2. Kerr-Newman black hole,

 
2 2M r Q2 2a(2M r − Q2 ) sin2 θ Σ Ξ sin2 θ 2
ds = − 1 − + dt2 − dtdϕ + dr2 + Σdθ2 + dϕ ,
Σ Σ Σ ∆ Σ

with auxiliary functions given by

Σ(r, θ) = r2 + a2 cos2 θ,

∆(r) = r2 − 2M r + a2 ,
Ξ(r, θ) = (r2 + a2 )2 − a2 ∆ sin2 θ,
where all parameters have their standard textbook definitions. Naturally, in KN case ∆(r) =
r2 −2M r +a2 +Q2 is enriched with the electric charge of the black hole. For every considered
point in the parameter space we also compute the free energy of the system as a benchmark
of thermodynamical stability of the cloud.
To solve these problems we use a numerical method based on the spectral differentiation
using Chebyshev polynomials. We build up a differentiation matrix for a two dimensional
Gauss-Lobatto grid. From this basic matrix, we are able to construct other differentiation
matrices - the second and mixed derivatives, and the whole left-hand side operator in the
end. In this way we are able to translate a differential problem into an algebraic one and
solve it by calculating the inverse of the system matrix.

Results
We have found that axions, indeed, can have stationary solutions around Kerr black holes
immersed in magnetic field and Kerr-Newman black holes. Moreover these solutions exist in
the whole range of parameter space, as they are magnetically induced.
The mass of the axions significantly influences the spatial distribution of the cloud. For
ultralight axions the cloud spreads from polar areas down almost to θ ≃ 80◦ . As one increases
the mass of the axionic field we find out that their angular localisation shifts to θ ≃ 45◦ ,
leaving polar areas depleted. It should also be noted that larger mass decreases the spatial
extent of the cloud, which is a quite expected result.
From the thermodynamical point of view, the free energy of axionic cloud happens to be
the smallest (the strongest binding) as the hosting black hole is nearly extreme (a → 1) and
the mass is tiny (the ultralight case). This theoretical clue allows to think that black holes
in the galaxies (which are naturally rotating and have some magnetic field around them)
may host axionlike dark matter structures that can spread over a large distance from them.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 40

4.1.5 Static axionlike dark matter clouds around magnetized ro-

tating wormholes - probe limit case
Reference
B. Kiczek and M. Rogatko, Static axionlike dark matter clouds around magnetized rotating
wormholes - probe limit case, Eur. Phys. J. C 82, 586 (2022).

Motivation
In this work we return to the issue of axionlike clouds, but this time in the vicinity of
wormholes. Previously we studied axionlike clouds around rotating black holes and found
their stationary solutions.
Wormholes are tunnel-like structures that can link two separate regions of the Universe,
or even different universes. Their origins are in early works of Einstein and Rosen [68] on
their bridge. This discovery resulted in a blossom of interest in these objects, from their
mathematical structure up to their possible usage in space travels. Later on Morris and
Thorne [69] gave the contemporary understanding of wormholes using the survivability of
human travellers as a criterion for calling a wormhole traversable.
In modern physics, wormholes can be seen as black hole mimickers in many cases and
from large distances it is truly difficult to distinguish one from another. So far there is no a
reliable criterion that could surely exclude a possibility if the observed compact object is not
a wormhole. Therefore the primary objective of this paper was to study the possible axionlike
dark matter halos around a wormhole and compare the achieved results with aforementioned
black hole solutions.

Methods
For the matter part we use exactly the same Einstein-Maxwell-axion theory as in the previous
work on black holes, therefore I do not write it here again. What is essential for these
computations are the gravitational background we used. Let me briefly comment on them.
The Kerr-like wormhole is the wormhole extension of the standard Kerr black hole solu-
tion. Its line element is following
 
2 2M r 2 4M ar sin2 θ Σ
ds = − 1 − dt − dtdϕ + dr2 + Σdθ2 (4.34)
Σ Σ ˜
∆
 2M a2 r sin2 θ  2
+ r 2 + a2 + sin θdϕ2 ,
Σ
˜
where ∆(r) = r2 + a2 − 2M (1 + λ2 )r, with λ being the deviation parameter. If one puts
λ = 0 we return to a plain Kerr black hole. However, using this gravitational background
is bittersweet. The sweet part is that it allows us to test the behaviour of axionlike clouds
both in wormhole and black hole regimes. The bitter part is that it is not an exact vacuum
CHAPTER 4. ORIGINAL CONTRIBUTIONS 41

solution of Einstein field equations, thus we should treat it as a toy model. In the end we
immerse the Kerr-like wormhole in uniform magnetic field using the Wald method.
During the work on this project we came to the conclusion that the comparison between
a black hole and one type of a wormhole is not enough, and an additional wormhole should
be included into the consideration. The axionlike cloud requires the gravitational object to
be axially symmetric, and therefore rotating. We decided to include a rotating counterpart
of the well-known (and typical solution) Morris-Thorne wormhole, which is Teo wormhole.
It serves as a benchmark for observed phenomena, if they are tied to only one solution or
they might be rather more generally wormhole related. The used background is described
by a line element

dr2  
ds2 = −N 2 dt2 + b
+ K 2 r2 dθ2 + sin2 θ(dϕ − ωdt)2 , (4.35)
1− r

with functions provided by

h r i  r γ 2ar+
+ +
N = exp − , b(r) = r+ , ω= , K = 1, (4.36)
r r r3
which are quite common choices in the literature.
To solve obtained field equations we used the same method as in the Axionlike dark
matter clouds around rotating black holes paper.

Results
The main results of this work are the spatial distributions of axionic clouds. Once again,
we found that they are magnetically induced and stationary. However their distribution
significantly differs from the black hole scenario. Specifically, there appears a gap between
the surface of wormhole’s throat and the axionic cloud itself. For a black hole the axionic
field had a non-zero finite value on the horizon. In case of a Kerr-like wormhole the axionic
field vanishes, then grows to some maximal value and fades away with the growth of radius.
We checked this also for the Teo wormhole and found a very similar result. In that case
the field did not vanish totally, as there was a little value on the throat. However the overall
tendency was similar, and probably if we used ψ(rth ) = 0 boundary condition, we could have
reached a similar picture.
Nevertheless, a significant difference, with respect to black holes, in the shape of axionlike
dark matter structures has been found. Naturally this finding is only a tip of the iceberg
as it can become now a starting point in finding some observable imprints caused by the
axionic condensate, like some possible influence on the trajectory of light. We hope to return
to these problems in future investigations.

4.2 Attached articles

J ournal of Cosmology and Astroparticle Physics
An IOP and SISSA journal

Ultra-compact spherically symmetric

JCAP09(2019)049
dark matter charged star objects

Bartlomiej Kiczek and Marek Rogatko

Institute of Physics, Maria Curie-Sklodowska University.
pl. Marii Curie-Sklodowskiej 1, 20-031 Lublin, Poland
E-mail: [email protected], [email protected]

Received April 17, 2019

Accepted August 12, 2019
Published September 24, 2019

Abstract. We study the properties of ultra-compact spherically symmetric dark matter sec-
tor star objects, being the solution of Einstein equations with two U(1)-gauge fields. One
of them is the ordinary Maxwell field, while the auxiliary gauge field pertains to the hidden
sector, and mimics the properties of dark matter. The visible and hidden sectors are coupled
by a kinetic mixing term with a coupling constant α. We also investigate the possibility of
condensation of charged scalar field around the reflecting dark matter star object. It happens
that dark matter sector, both dark matter charge and coupling constant, cause shrinking of
the star radius for which the condensation may occur.

Keywords: gravity, dark matter theory

ArXiv ePrint: 1904.07232

c 2019 IOP Publishing Ltd and Sissa Medialab https://doi.org/10.1088/1475-7516/2019/09/049

Contents

1 Introduction 1
1.1 Dark matter model 2

2 Ultra-compact charged star-like object in hidden sector 3

3 Dark matter charged fluid sphere 4

3.1 TOV equation 6

JCAP09(2019)049
4 Bound on the gravitational mass and charge of stable dark matter sector
compact object 6

5 Scalar hair on dark matter compact star 9

5.1 Model of scalar field configuration 9
5.2 Numerical results 10

6 Discussion and conclusions 12

1 Introduction

The pursuit for an illusive ingredient of our Universe constituting over 23 percent of its mass,
non-baryonic dark matter, authorizes of the most important topic of the astronomy and
physics [1]. Dark matter builds a thread-like structure of the so-called cosmic web being a
scaffolding for the ordinary matter to accumulate [2, 3]. For the first time Hubble Space Tele-
scope, revealed images of the structure in question, a giant filament of dark matter, being the
part of the comic web extending from one of the most massive galaxy clusters MACS J071 [3].
The contemporary understanding of the creation of structures in the Universe, the
so-called ΛCDM model, envisages that galaxies are embedded in extended massive halos
composed of dark matter. They are surrounded by smaller dark matter sub-halos, which are
large enough to accumulate gas and dust and form satellite galaxies, orbiting around the
host ones. However, smaller galaxies can be circled by much smaller sub-halo dark matter
satellites, almost invisible to telescopes [4]. It can be concluded that in the proximity of the
Milky Way we can suppose that such kind of structures may also exist.
On the other hand, dark matter interaction with the Standard Model particles is one
of the main theoretical investigations of the particle physics in the early stages of our
Universe [5, 6]. Several new types of fundamental particle have been claimed as candidates
for dark matter sector, which are expected to interact with nuclei in suitable detecting ma-
terials on Earth. It was claimed only by the DAMA collaboration [7, 8] that they observed
modulation in the rate of interaction events in their detectors which might be the trace of
dark matter sector. Several groups want to reproduce the DAMA results but in vain [9].
This situation triggers the discussion concerning the composition, interaction with ordinary
matter, the self-interaction and the possible ways to discriminate between various models of
dark matter [10].
One also tries to implement physics beyond the Standard Model for the explanation of
dark matter non-gravitational interactions. This fact resurgences interests in gamma rays

–1–
emissions coming from dwarf galaxies, possible dilaton-like coupling to photons caused by
ultra-light dark matter, and oscillations of the fine structure constant [11]–[13]. The Earth ex-
periments are also exploited for the detection of possible low-energy mass of dark matter sec-
tor, especially in e+ e− colliders [14]. BABAR detector set some energy range for dark photon
production, i.e., 0.02 < m < 10.2 GeV, but in vain. No significant signal has been observed.
However the new experiments are planned to cover the energy region 15 ≤ m ≤ 30 MeV.
Recently, the revision of the constraints on dark photon with masses below 100 MeV was pro-
posed based on the observation of supernova 1987A event [15]. Collisions among galaxy clus-
ters can also provide new tools for testing non-gravitational forces acting on dark matter [16].
Collapse of neutron stars and emergence of the first star generations [17]–[19], alternative

JCAP09(2019)049
theories [20]–[22], or compact objects studies [23]–[26], can deliver some other hints for these
researches in question. The existence of dark matter can affect black hole growth during the
early stages of our Universe. The numerical studies of dark matter and dark energy collapse
and their interactions with black holes and wormholes were investigated in [27, 28].
Because of the absence of evidences for the most popular particle candidates for dark
matter, it was claimed the growing sense of ‘crisis’ in the dark matter particle community,
and necessity of diversifying the experimental efforts. The efforts in question should accom-
plish upcoming astronomical surveys, gravitational wave observatories, to deliver us some
complimentary information about the dark matter sector [1].

1.1 Dark matter model

Having all the above arguments in mind, we shall examine the problem of the influence of
dark matter on the properties of spherically symmetric static charged star-like solution. In
our research we shall consider the model of dark matter sector in which the additional U(1)-
gauge field is coupled to the ordinary Maxwell one. The action describing Einstein-Maxwell
dark matter gravity yields [29, 30]
Z  
4 √ 1 µν 1 µν α µν
S = d x −g R − Fµν F − Bµν B − Fµν B , (1.1)
4 4 4
where Fµν = 2∇[µ Aν] is the ordinary Maxwell field while Bµν = 2∇[µ Bν] stands for the
auxiliary U(1)-gauge field mimicking the dark matter sector coupled to Maxwell one. The
coupling constant is denoted by α. Predicted values of α-coupling constant, being the kinetic
mixing parameter between the two U(1)-gauge fields, for realistic string compactifications
range between 10−2 and 10−16 [31]–[34]. It turns out that astrophysical observations of 511 eV
gamma rays [35], experiments detecting the electron positron excess in galaxies [36, 37], as
well as, possible explanation of muon anomalous magnetic moment [38], strongly advocate
the idea of dark matter sector coupled to the Maxwell one. Moreover, the kinetic mixing
term between ordinary boson and relatively light one (the dark one) arising from U(1)-gauge
symmetry connected with a hidden sector, may cause a low energy parity violation [39]. The
low energy gauge interaction in the hidden sector may manifest itself by the Higgs boson
decays, and a relatively light vector boson with mass m ≥ 10 GeV can be produced [40].
The presented model has also its justification in string/M-theory, where the mixing
portal (term which couples Maxwell and the additional U(1)-gauge field) arises in open string
theory. Both gauge states are supported by D-branes separated in extra dimensions [41], in
supersymmetric Type I, Type II A, Type II B models, where the massive open strings stretch
between two D-branes. The massive string/brane states existence connect the different gauge
sectors.

–2–
 The other cognizance of the above scenario can be achieved by M2-branes wrapped on
surfaces which intersect two distinct codimension four orbifolds singularities. It happened
that the natural generalization can be carried out in M, F-theory and heterotic string models.
In the considered action (1.1) the auxiliary gauge field is connected with some hidden
sector [41]. For the first time such a model was introduced [42], in order to describe the
existence and subsequent integrating out of heavy bi-fundamental fields charged under the
U(1)-gauge groups. In general, such kind of terms emerge in the theories having in addition
to some visible gauge group, the other one in the hidden sector. Such scenario is realized e.g.,
in compactified string or M-theory solutions generically possess hidden sectors which contain
the gauge fields and gauginos, due to the various group factors included in the gauge group

JCAP09(2019)049
symmetry of the hidden sector.
The hidden sector in the low-energy effective theory contains states which are uncharged
under the Standard Model gauge symmetry groups. They are charged under their own groups
and interact with the visible ones via gravitational interaction. We can also think out other
portals to our visible sector [43, 44]. For the consistency and supersymmetry breaking [32],
the realistic embeddings of the Standard Model in E8 × E8 string theory, as well as, in type
I, IIA, or IIB open string theory with branes, require the existence of the hidden sectors.
The organization of the paper is as follows. In section 1 we give some introduction to the
presented model of dark matter sector. In section 2 we present the direct derivation of static
spherically symmetric solution of Einstein — dark matter equations of motion. Section 3
is devoted to the analysis of dark matter charged fluid sphere, while in section 4 we derive
the bound on the gravitational mass and charges for a stable dark matter compact star-like
object. In section 5 we considered the problem of hair growing of charged under visible sector
scalar field, on the ultra-compact reflecting star. In section 6 we concluded our investigations.

2 Ultra-compact charged star-like object in hidden sector

In this section we consider the spherically symmetric static solution of Einstein dark matter
sector equations of motion, being a model of ultra-compact charged dark matter star-like
object.
√
The components of energy momentum tensor defined as Tµν = −δS/ −gδg µν , devoted
to Maxwell, dark matter U(1)-gauge field and the mixture of them, may be written as

Tµν = Tµν (F ) + Tµν (B) + αTµν (F, B), (2.1)

where the adequate energy momentum tensors are given by

1 1
Tµν (F ) = Fµγ Fν γ − gµν Fαβ F αβ , (2.2)
2 8
1 1
Tµν (B) = Bµγ Bν γ − gµν Bαβ B αβ , (2.3)
2 8
1 1
Tµν (F, B) = Fµγ Bν − gµν Fαβ B αβ .
γ
(2.4)
2 8
In what follows we suppose that the existence of t-component of the gauge field potentials
Q Qd
At = − , Bt = − , (2.5)
r r
where Q is the ordinary Maxwell charge while Qd is connected with dark matter sector.

–3–
 The line element of spherically symmetric spacetime implies
ds2 = −e2φ(r) dt2 + e2λ(r) dr2 + r2 (sin2 θdφ2 + dθ2 ). (2.6)
Having in mind (2.5) and (2.6), the (t t) and (r r) components of Einstein equations Gµν =
Tµν , are provided by
" #
2φ(r) 1 e2λ(r) 2λ0 (r) Q2c
−e − − = , (2.7)
r2 r2 r 4r4
1 e2λ(r) 2φ0 (r) Q2c
− + = − , (2.8)
r2 r2 r 4r4 e2φ

JCAP09(2019)049
where the total charge Q2c = Q2 + Q2d + αQQd , the prime denotes taking derivative with
respect to r-coordinate. Consequently, analyzing the above equations, we draw a conclusion
that λ(r) = −φ(r), and the metric tensor component yields
2M Q2
e−2λ(r) = 1 − + c2 , (2.9)
r 4r
where M is the total gravitational mass and Qc the total charge, measured by the observer
at future timelike infinity.

3 Dark matter charged fluid sphere

This section will be devoted to the examination of the charged dark sector fluid sphere and
the Tolmann-Openheimer-Volkov (TOV) equation for the object under inspection. In order
to derive the relativistic hydro and electro-dark matter dynamic equation for the charged
sphere let us consider the action for visible and hidden sector U(1)-gauge fields with current
sources. It is provide by the relation of the form as
Z  
√ 1 1 α
SEM +dm+cur (F, B) = d4 x −g − Fµν F µν − Bµν B µν − Fµν B µν −2Aµ jµ (F )−2B µ j̃µ (B) .
4 4 4
(3.1)
Varying the underlying action with respect to the adequate gauge fields we arrive at the
following equations of motion
α
∇µ F µν + ∇µ B µν = 2j ν (F ), (3.2)
2
α
∇µ B µν + ∇µ F µν = 2j̃ ν (B). (3.3)
2
It can be shown that for the Maxwell field it implies
α̃∇µ F µν = 2j ν (F ) − αj̃ ν (B), (3.4)
where α̃ = 1 − α2 /4 and on the right-hand side we get currents responsible both for visible
and dark matter sector. The adequate components of the energy momentum tensor imply
1 1
Tµν (F ) = Fµγ Fν γ − gµν Fαβ F αβ − gµν Aδ j δ (F ) + 2A(µ jν) (F ), (3.5)
2 8
1 1
Tµν (B) = Bµγ Bν − gµν Bαβ B αβ − gµν Bδ j̃ δ (B) + 2B(µ j̃ν) (B),
γ
(3.6)
2 8
1 1
Tµν (F, B) = Fµγ Bν γ − gµν Fαβ B αβ . (3.7)
2 8

–4–
Moreover, for the charged sphere we shall consider perfect fluid component of the energy
momentum tensor. It yields

Tµν (p, ρ) = (ρ + p)uµ uν + pgµν , (3.8)

where p(r) is the fluid pressure while ρ stands for the total mass density of the sphere in
question.
Having in mind that the considered ultra compact object has the spherical symmetry,
the elaborated electric Maxwell field ought to posses the radial symmetry. Just the electro-
magnetic strength tensor will be of the form as Frt = E(r). Then using equation (3.4) we
get the following relation:

JCAP09(2019)049
1 φ(r)+λ(r)
E(r) = e Qc (F, B)(r), (3.9)
r2
where Qc (F, B)(r) yields
Z
Qc (F, B)(ζ) = dζ ζ 2 eφ(ζ)+λ(ζ) j t (F, B)(ζ), (3.10)

while the currents for the mixture of Maxwell and dark matter fields is written in the form
as follows:
j t (F, B) = α1 j t (F ) − α2 j̃ t (B), (3.11)
with the adequate coefficients α1 = 2/α̃ and α2 = α/α̃.
Using the Einstein equation (tt)-component provided by the equation

e−2λ(r) 1 2λ(r)0 e−2λ(r) Q2c (r)

− + + = ρ + , (3.12)
r2 r2 r 4r4
we arrive at the final expression for the metric tensor component describing dark sector ultra
compact charged sphere of a radius r
2m(r) Fc (r)
e−2λ(r) = 1 − − , (3.13)
r r
where we set Z Z
ρ ζ2 Qc (ζ)
m(ζ) = dζ , Fc (ζ) = dζ . (3.14)
2 4ζ 2
m(r) stands for the interior mass accumulated in a sphere of r-radius. It will be also necessary
to introduce the notion of gravitational mass. To commence with, let us match the exterior
and interior solutions of the considered field equations. Namely at the distance R = r, we
have equality Z R Z R
2M Q2 1 1 Qc (ζ)2
1− + c2 = 1 − dζρζ 2 − dζ , (3.15)
R 4R R 0 R 0 4ζ 2
which provides the explicit relation for gravitational mass M at the distance equal to R
Q2c Fc (R)
M= + m(R) + . (3.16)
8R 2
Just the definition of gravitational mass at any radius r implies
Fc (r) Qc (r)2
mg (r) = m(r) + + . (3.17)
2 8r

–5–
All the above lead to the conclusion that the gravitational mass will modify the metric
function e−2λ(r) . Now, consequently with the help of the aforementioned notion, we obtain
that the following relation:

2mg (r) Qc (r)2

e−2λ(r) = 1 − + , (3.18)
r 4r2
where the metric component depends on the gravitational mass and the total (Maxwell and
dark matter charges) of the examined sphere.

3.1 TOV equation

JCAP09(2019)049
In order to find the Tolmann-Oppenheimer-Volkov equation for the dark sector ultra-compact
charged star we have to use the conservation of the energy momentum tensor, ∇µ T µν = 0.
Namely having in mind this relation
h i Qc (r)0 Qc (r)  
0 0
∇r Tr r (F, B) + Tr r ρ, p) = − + p(r) + φ(r) ρ + p(r) = 0, (3.19)
2r4
and the Grr component of the Einstein equation, as well as, relation (3.13), we arrive at
the following form of the equation binding pressure, density and charge of the dark sector
star-like object
 
ρ+p(r)   
0
0
Qc (r) Qc (r) Fc (r) r3 p(r) Qc (r)2 2m(r) Fc (r) −1
p (r) = − m(r)+ + − 1− − .
2r4 r2 2 2 8r r r
(3.20)

4 Bound on the gravitational mass and charge of stable dark matter sector
compact object

In this section we look for the upper bound on gravitational mass for given radius and total
charge, for the ultra-compact dark matter sector star objects. It turns out that following
the procedure presented in [45, 46], after some algebra, one can get almost the same result.
Namely for sphere of radius r the inequality implies
q √ r
r r Q2c (r)
mg (r) ≤ + + . (4.1)
3 9 12r
On the other hand, in order to derive the inequality for the case when r ≥ R, let us suppose
that ψk = mg (r) + r3 p(r) − Q2c (r)/4r is a sequence of regular solutions with support in the
range [Rk , R] of the Tolmann-Oppenheimer-Volkov equation (3.20). Moreover, we suppose
that limk→∞ Rk /R = 1. Then, following the steps presented in [46], one arrives at the
inequality among M, R, Qc valid for the exterior solution of the Einstein equations with two
U(1)-gauge fields. Namely, it implies
√ r
√ R R Q2
M≤ + + c . (4.2)
3 9 12R
However, an interesting problem was suggested in [47], i.e., it was posed a question if the
addition of some characteristics of horizonless object could improve the upper bound on its
gravitational mass. The refinement of the bounds derived in [46] was presented in [47].

–6–
 On the other hand, geodesic motion determines important features of the spacetime
and objects in it. Namely, null unstable geodesics are closely connected with the appearance
of a compact object visible by the external observer. They are also bounded with the char-
acteristic modes of black holes, i.e., computing the Lyapunov exponent one can prove that,
in the eikonal limit, quasi-normal modes of black holes are determined by parameters of the
circular null geodesics [48]. Circular geodesics are of special interests. For example in Kerr
spacetime the binding energy of the last stable circular timelike geodesics is related to grav-
itational binding energy that can be radiated to infinity. This fact can be implemented to
the estimation of a spin of astrophysical black objects through observations of the accretion
discs surrounded the compact object in question [49].

JCAP09(2019)049
Circular null orbits (photonspheres) were also analyzed in the background of a regu-
lar horizonless ultra-compact stars [47, 50], where it was revealed that photonspheres at-
tributed to the aforementioned objects possessed an upper bound expressed in terms of mass
and charge.
To proceed further, let us discuss the upper bound of the gravitational mass allowed for
the ultra-compact spherical dark matter sector star objects. To begin with, one finds that the
Lagrangian describing the geodesics in the spacetime of dark matter ultra-compact star yields

2L = −e2φ(r) ṫ2 + e2λ(r) ṙ2 + r2 θ̇2 + r2 sin θφ̇2 , (4.3)

where the dot means taking derivate with respect to the proper time. We shall restrict our
considerations to the equatorial orbits. Then, following the methods of classical mechanics,
the generalized momenta with respect to the coordinates (t, r, φ)

pt = −e2φ(r) ṫ = −E, pr = e2λ(r) ṙ = L, pφ = r2 φ̇, (4.4)

where the constants E and L are attributed to the total energy and orbital momentum of
the system, respectively. The Legendre transformation leads to the Hamiltonian of the form

2H = −E ṫ + Lφ̇ + e2λ(r) ṙ2 =  = const. (4.5)

If  = 0, one obtains null geodesics, whereas  = 1 we have to do with the timelike ones.
Consequently from equation (4.5) we get
 
2 1 2 −2φ(r) L2
(ṙ) = 2λ(r) E e − 2 − . (4.6)
e r
The right-hand side of the above relation authorizes the effective potential which will be
denoted by Vr (r).
The condition for the circular null orbits, Vr = Vr0 = 0, implies

E2 e−2λ(r)
= , 1 + r λ0 (r) = 0, (4.7)
L2 r2
which gives us the characteristic relation N (r = rc ) = 0, with the definition
 
−2λ(r) 2 Qc (r)2
N (r) = 3e − 1 − r p(r) − . (4.8)
8r4
One can see that the light rings of the spherically symmetric ultra-compact dark matter sector
object, which line element is provided by (2.6), are characterized by the functional (4.8). The

–7–
dimensionless function N (r) determines the discrete radii of the null circular geodesics of the
spacetime in question. It has the following boundary conditions

N (r → 0) = 2, N (r → ∞) = 2. (4.9)

The above feature implies that spherically regular horizonless compact objects are charac-
terized by an even number of null circular geodesics. Let us remark that in [51, 52] it is
derived that horizonless matter configurations having a light ring must have pairs of them.
For such regular configurations one of the closed light ring is stable. However, in the spherical
symmetric case being subject to the equation ∼ rc (ρ + pT ) = 1, where rc is the radius of the
light ring and pT tangential pressure, we have odd number of the light rings [53].

JCAP09(2019)049
For the exterior region of the ultra-compact object in question, one has that

r ≥ R, p(r) = ρ = 0. (4.10)

Further suppose that the spatially regular charged dark matter sector compact object, has
the external light ring for rc > R. Inspection of the equations (2.9) and (4.8) reveals that
p
6 + 36M 2 − 7Q2c
rc (out) = , (4.11)
4
which gives us the stronger bound for mass and the total charge. Namely, we arrive at the
condition q
Q2 + Q2d + αQQd 6
≤√ . (4.12)
M 7
Let us analyze the derived condition for the specific choice of the charges from visible and
hidden sectors. Namely, if Q = Qd we get

Q 6
≤p . (4.13)
M 7(2 + α)

The larger α one takes into account, the smaller value of the ratio we get. On the other
hand, when Qd = βQ and α = const, we obtain

Q 6
≤p . (4.14)
M 7(1 + β 2 + αβ 2 )

Just the bigger multiplying factor β one takes, the smaller value of Q/M we have.
In principle the presence of the light ring outside the surface of a spherically regular
horizonless compact star object, may cause the appearance of the another light ring which
is included inside the outer one. It was revealed [54] that the existence of the inner stable
null circular geodesic, might show that the ultra-compact object under consideration was
nonlinearly unstable to perturbations massless fields. Therefore we draw a conclusion that
spherically symmetric horizonless spacetimes describing a compact object within the theory
with dark matter sector has to have no light rings. It implies that the lower on the radii of
the stable dark matter star-like object yields
p
6 + 36M 2 − 7Q2c
R> . (4.15)
4

–8–
Having in mind the exact form of the total charge, from (4.15), one can remark that the
radius of ultra-compact dark matter sector star object will be smaller that this without
hidden sector.
One has that the larger dark charge Qd (M, Q, α = const) we consider, the smaller
radius one obtains. Consequently, if we assume that mass and the charges of the object are
constant, and the bigger value of α one takes, the smaller radius of ultra-compact dark matter
sector we receive.

5 Scalar hair on dark matter compact star

The no-hair theorem for asymptotically flat horizonless neutral compact star being subject

JCAP09(2019)049
to the reflecting boundary conditions, with the influence of scalar field with potential, was
derived in [55]. Then the problem was elaborated in [56], implementing the spacetime with a
positive cosmological constant. The case of the massless scalar field non-minimally coupled
to gravity was studied in [57].
On the other hand, the charged reflecting sphere influenced by scalar field was examined
in [58], while the analytical formulae for the discrete spectrum of star radii were revealed
in [59]. It turns out that the low frequency scalar perturbations in the spacetime of a black
hole in a box can lead to the superradiant instability, as well as, formation of a quasi-local
hair [61]. It is believed that box boundary may enforce fields to bounce back and trigger
the condensation around the black object. This kind of a problem was analyzed in the case
of scalar field configurations near charged compact reflecting star [62], where the lower and
upper bounds for the radii of the object were provided.

5.1 Model of scalar field configuration

To find out the possibility of formation of the scalar hair we start our considerations with
the following action:
Z  
4 √ 1 µν α µν 1 µν α 2 2
SEM +dm+sc = d x −g − Fµν F − Bµν F − Bµν B −Dα ψD ψ−µ ψ , (5.1)
4 4 4
where the covariant derivative is given by D = ∇µ − qAµ . q and µ are respectively charge
and mass of the scalar field ψ(r). The scalar field in question is charged with respect to
visible sector.
For the brevity of further notation we rename the left-hand side of the relation (2.9),
i.e., e−2λ(r) = f (r). Moreover, for the convenience of numerical calculations we also rescale
charges in the line element describing dark matter charged static line element
Q(i)
→ Q(i) , (5.2)
2
where the subscript (i) denotes charges bounded with visible and hidden sectors.
Having in mind the form of the action (5.1), we get equation of motion for the scalar
field, δS/δψ = 0, which is provided by
 
∇µ ∇µ ψ − q 2 Aµ Aµ + µ2 ψ = 0. (5.3)

Its explicit for the static background implies the second order ordinary differential equation
of the form as follows:
 0   
f 2 q 2 Q2 µ2
ψ 00 + + ψ0 + − ψ = 0, (5.4)
f r r2 f 2 f

–9–
where prime denotes derivative with respect to r-coordinate. It can be noticed that the
relation (5.4) reveals the scaling symmetry of its parameters. Namely, one has that

r → sr, M → sM, Q(i) → sQ(i) , q → q/s, µ → µ/s. (5.5)

In order to solve the differential equation we have to establish the adequate boundary con-
ditions imposed on ψ(r). In what follows we implemet the reflecting boundary conditions,
which means that at the surface of the dark matter star the scalar field will vanish, ψ(rs ) = 0.
Moreover, one supposes that the timelike box boundary conditions should to be satisfied in
the considered spacetime. It means that at r = rb the scalar field in question is reflected
back, ψ(rb ) = 0.

JCAP09(2019)049
The function describing the scalar field has to have at least one extremum point in the
range from rs to rb . At the aforementioned point the function ψ(rext ) should satisfy the
following conditions:
ψ 0 (rext ) = 0, ψ(rext )ψ 00 (rext ) ≤ 0. (5.6)
√
Following the procedure presented in [62], we apply the transformation ψ̃ = rψ in
the equation (5.4) and then collecting terms with ψ̃ derivatives, we use the extremum condi-
tion (5.6) in order to achieve the inequality valid for r = rextr
1 1
µ2 r2 f ≤ − f 2 − f 0 f r + q 2 Q2 . (5.7)
4 2
In the next step we check the monotonicity and the sign of each term on the right-hand side
of the inequality, because of the fact that both f (r) and rf 0 (r) are descending and greater
than zero, and moreover left-hand side is a descending function for the distance greater than
the radius of the star. Consequently, one concludes that

µ2 rs2 f (rs ) ≤ µ2 r2 f (r) ≤ q 2 Q2 . (5.8)

Neglecting the middle term we solve the inequality

 
2M Q2
µ2 rs2 1 − + 2c 6 q 2 Q2 (5.9)
rs rs
treating µrs as the variable in the quadratic inequality. Finally it leads to the possible range
where rs lies within. It is provided by
p p
µM + µ2 (M 2 − Q2c ) 6 µrs 6 µM + µ2 (M 2 − Q2c ) + q 2 Q2 . (5.10)

The lower bound is a Schwarzschild radius for charged dark matter sector black hole and the
upper one is connected with the larger solution of the inequality (5.9).

5.2 Numerical results

Integrating from rs to rb , we numerically solve the ordinary differential equation (5.4) gov-
erning the motion of the scalar field ψ in the spacetime of ultra-compact dark matter star-like
object. By virtue of the shooting method, we have found the adequate radius of the star
surface rs , with previously set rb fulfilling the auxiliary condition ψ(rb ) = 0 on the second
boundary. We perform our shooting by bisection method in the range derived from the maxi-
mum point analysis (5.10). The algorithm may find more than one value of rs , corresponding
to the boundary condition in a given interval. If such situation takes place, we choose the
largest solution.

– 10 –
 JCAP09(2019)049
Figure 1. The distribution of the condensed scalar field ψ around a compact charged star-like object.
The calculation was carried out for parameters M = 8, Q = 4, Qd = 4, q = 1 (normalized to the
mass µ).

After obtaining solution of the equation of motion satisfying the aforementioned require-
ments, we shall investigate the influence of the dark matter charge and the coupling strength
on the possibility of the scalar hair formation in the vicinity of ultra-compact star-like object.
In figure 1 one depicts the value of ψ scalar field versus µr, for the fixed values of mass of
the star and charges of visible and hidden sectors equal to 4. The physical parameters of
the system have been normalized to µ due to the scaling symmetry (5.5). One can observe
that the larger α-coupling constant we examine, the smaller radius of the hairy compact star
we arrive at. The influence of dark matter causes that the condensation takes place for the
smaller objects. On the other hand, the maximal value of ψ also depends on the α-coupling
constant. If α increases the maximal value of the scalar field that form hair also grows.
To proceed further, let us take a closer look on the scenario when one of the charges
dominates the other. We set one of them to Q(i) = 3 and plot the star radius for condensing
scalar, for different values of the other kind of charge. In figure 2 the dependence of the star
radius on the charges, the electric and the dark matter one, is envsaged respectively. One
can see that increasing the charge forces the star to be smaller in order to hold the condensed
scalar. Every curve separately shows similar parabolic behavior, but when we are aware of
the coupling presence there are some differences to be spotted. The effect of the coupling
is similar in the change of both charges. The splitting distance between line of interest and
baseline scales more or less linear with α. Both plots tend to the same final value of rs when
the charges are equal (Q = 3, Qd = 3) but the curves differ quantitatively. The initial value
of rs from the left panel of figure 2 is smaller than the one from the right panel. In such a
case the increase of Qd causes more steep drop of rs than in case of increasing Q.
The larger value of α-coupling constant we consider, the smaller µrs we get. This leads
to the conclusion that the both charges result in shrinking the star radius so the scalar hair
may occur.

– 11 –
 15.6
15.4

15.5

15.3
15.4

15.2
15.3

15.1 15.2
rs

rs
15.1

15.0 

15.0
14.9

14.9

14.8
14.8
 =0  =0

JCAP09(2019)049
14.7  = 0.1  = 0.1
14.7
 = 0.25  = 0.25
 = 0.5  = 0.5
14.6 14.6
0.5 1.0 1.5 2.0 2.5 3.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0

Q Qd

Figure 2. Star radius dependence on an electric charge (left panel), dark matter charge (right panel)
and the dark matter coupling constant. For this plot M = 8, q = 1. The unvaried charge is equal
Q(i) = 3, in both panels.
17.0
=0
16.0
= 0.1
16.0 = 0.25

15.0 15.5

14.0
15.0

13.0
rs

rs

14.5
12.0

11.0
14.0

10.0

13.5
=0
9.0
= 0.1
= 0.25
8.0 13.0
6.0 6.5 7.0 7.5 8.0 8.5 9.0 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

M q/

Figure 3. Star radius dependence on the star mass with different dark matter coupling (left panel)
and on the scalar field charge (right panel). In both plots Q = 4, Qd = 4, in the left panel q = 1 and
in the right panel M = 8.

The left panel of the figure 3 depicts the hairy star radius as a function of its mass.
Adding the dark matter coupling to the system shrinks the obtained radius. A similar
behavior is also shown on in the right panel of the figure 3, where the radius versus scalar
field charge is plotted. Both of these plots are presented for equal charges Q = Qd = 4.

6 Discussion and conclusions

In our paper we have studied the charged spherically horizonless solution of Einstein dark
matter sector equation of motion, being the two U(1)-gauge theory in which the ordinary
Maxwell field pertain to the visible sector while the auxiliary U 1)-gauge field is responsible
for the hidden one. The solution in question models the ultra-compact star-like object in the
underlying theory.

– 12 –
 We have found the lower bound on the radius of the charged dark matter sphere, which
takes place for any solutions satisfying the condition of pressure being less than the density
of the object. However, it is possible to derive some refinement on the bound in question.
Namely, the investigations of photonsphere in the background of the spherically symmetric
ultra-compact dark matter sector star objects reveals the fact that it has an upper bound
expressed in terms of the ADM mass and total charge, which in turn gives us some additional
condition binding the total charge and mass.
Due to the fact that spherically symmetric horizonless ultra-compact star-like object
may be unstable to perturbations of massless fields, we obtain another restriction on its
radius. It turns out that the larger dark charge is considered, the smaller radius of the

JCAP09(2019)049
compact object we get. When the charge and mass of the dark matter object are constant,
the increase of the α-coupling constant value, binding the hidden and visible sectors, causes
the decrease of the radius of dark star.
We have also investigated the possibility of condensation of scalar field around compact
reflecting stars in a box-boundary regime. The presented model includes the dark matter
charge and its influence on the condensing scalar through the background metric. From every
plot presented in this article we can conclude that the dark matter -electromagnetic field cou-
pling decreases the largest star radius which is capable of holding hair. Moreover this effect is
stronger when both charges are present in the system and when they are sufficiently large, so
the mixing term makes significant contribution in the metric function. On the other hand the
effective charge may not be too large so the square root in (5.10) remains real. We are aware of
the restraints of the probe limit and that the problem requires a full study with back reaction
and solving the Maxwell equations. Also it is interesting to see the stability of the scalar con-
densate when the charged dark matter would be dropping into the system in a time-dependent
approach. We take it as an inspiration to further research to be presented elsewhere.

References
[1] G. Bertone and M.P. Tait, Tim, A new era in the search for dark matter, Nature 562 (2018) 51
[arXiv:1810.01668] [INSPIRE].
[2] R. Massey et al., Dark matter maps reveal cosmic scaffolding, Nature 445 (2007) 286
[astro-ph/0701594] [INSPIRE].
[3] J.P. Dietrich et al., A filament of dark matter between two clusters of galaxies, Nature (2012)
[arXiv:1207.0809] [INSPIRE].
[4] T.K. Starkenburg, A. Helmi and L.V. Sales, Dark influences II: gas and star formation in
minor mergers of dwarf galaxies with dark satellites, Astron. Astrophys. 587 (2016) A24
[arXiv:1508.06281] [INSPIRE].
[5] M. Regis, J.-Q. Xia, A. Cuoco, E. Branchini, N. Fornengo and M. Viel, Particle dark matter
searches outside the Local Group, Phys. Rev. Lett. 114 (2015) 241301 [arXiv:1503.05922]
[INSPIRE].
[6] Y. Ali-Haı̈moud, J. Chluba and M. Kamionkowski, Constraints on Dark Matter Interactions
with Standard Model Particles from Cosmic Microwave Background Spectral Distortions, Phys.
Rev. Lett. 115 (2015) 071304 [arXiv:1506.04745] [INSPIRE].
[7] R. Bernabei et al., Searching for WIMPs by the annual modulation signature, Phys. Lett. B
424 (1998) 195 [INSPIRE].
[8] R. Bernabei et al., Final model independent result of DAMA/LIBRA-phase1, Eur. Phys. J. C
73 (2013) 2648 [arXiv:1308.5109] [INSPIRE].

– 13 –
 [9] COSINE-100 collaboration, An experiment to search for dark-matter interactions using
sodium iodide detectors, Nature 564 (2018) 83 [arXiv:1906.01791] [INSPIRE].
[10] N. Masi, Dark matter: TeV-ish rather than miraculous, collisionless rather than dark, Eur.
Phys. J. Plus 130 (2015) 69 [INSPIRE].
[11] A. Geringer-Sameth et al., Indication of Gamma-ray Emission from the Newly Discovered
Dwarf Galaxy Reticulum II, Phys. Rev. Lett. 115 (2015) 081101 [arXiv:1503.02320] [INSPIRE].
[12] K.K. Boddy and J. Kumar, Indirect Detection of Dark Matter Using MeV-Range Gamma-Ray
Telescopes, Phys. Rev. D 92 (2015) 023533 [arXiv:1504.04024] [INSPIRE].
[13] K. Van Tilburg, N. Leefer, L. Bougas and D. Budker, Search for ultralight scalar dark matter
with atomic spectroscopy, Phys. Rev. Lett. 115 (2015) 011802 [arXiv:1503.06886] [INSPIRE].

JCAP09(2019)049
[14] BaBar collaboration, Search for a Dark Photon in e+ e− Collisions at BaBar, Phys. Rev. Lett.
113 (2014) 201801 [arXiv:1406.2980] [INSPIRE].
[15] J.H. Chang, R. Essig and S.D. McDermott, Revisiting Supernova 1987A Constraints on Dark
Photons, JHEP 01 (2017) 107 [arXiv:1611.03864] [INSPIRE].
[16] D. Harvey, R. Massey, T. Kitching, A. Taylor and E. Tittley, The non-gravitational
interactions of dark matter in colliding galaxy clusters, Science 347 (2015) 1462
[arXiv:1503.07675] [INSPIRE].
[17] J. Bramante and T. Linden, Detecting Dark Matter with Imploding Pulsars in the Galactic
Center, Phys. Rev. Lett. 113 (2014) 191301 [arXiv:1405.1031] [INSPIRE].
[18] J. Fuller and C. Ott, Dark Matter-induced Collapse of Neutron Stars: A Possible Link Between
Fast Radio Bursts and the Missing Pulsar Problem, Mon. Not. Roy. Astron. Soc. 450 (2015)
L71 [arXiv:1412.6119] [INSPIRE].
[19] I. Lopes and J. Silk, A particle dark matter footprint on the first generation of stars, Astrophys.
J. 786 (2014) 25 [arXiv:1404.3909] [INSPIRE].
[20] S.I. Blinnikov and M.Y. Khlopov, On possible effects of ‘mirror’ particles, Sov. J. Nucl. Phys.
36 (1982) 472 [INSPIRE].
[21] S.I. Blinnikov and M.Y. Khlopov, Possible astronomical effects of mirror particles, Sov. Astron.
27 (1983) 371 [INSPIRE].
[22] M.Y. Khlopov, G.M. Beskin, N.E. Bochkarev, L.A. Pustylnik and S.A. Pustylnik,
Observational Physics of Mirror World, Sov. Astron. 35 (1991) 21 [INSPIRE].
[23] J.P.S. Lemos and E.J. Weinberg, Quasiblack holes from extremal charged dust, Phys. Rev. D
69 (2004) 104004 [gr-qc/0311051] [INSPIRE].
[24] N. Dadhich, On compactness limit of static charged object, arXiv:1903.03436 [INSPIRE].
[25] S. Chakraborty and S. Sengupta, Packing extra mass in compact stellar structures: An
interplay between Kalb-Ramond field and extra dimensions, JCAP 05 (2018) 032
[arXiv:1708.08315] [INSPIRE].
[26] N. Dadhich and S. Chakraborty, Buchdahl compactness limit for a pure Lovelock static fluid
star, Phys. Rev. D 95 (2017) 064059 [arXiv:1606.01330] [INSPIRE].
[27] A. Nakonieczna, M. Rogatko and R. Moderski, Dynamical Collapse of Charged Scalar Field in
Phantom Gravity, Phys. Rev. D 86 (2012) 044043 [arXiv:1209.1203] [INSPIRE].
[28] A. Nakonieczna, M. Rogatko and L. Nakonieczny, Dark sector impact on gravitational collapse
of an electrically charged scalar field, JHEP 11 (2015) 012 [arXiv:1508.02657] [INSPIRE].
[29] T. Vachaspati and A. Achucarro, Semilocal cosmic strings, Phys. Rev. D 44 (1991) 3067
[INSPIRE].

– 14 –
[30] A. Achucarro and T. Vachaspati, Semilocal and electroweak strings, Phys. Rept. 327 (2000) 347
[hep-ph/9904229] [INSPIRE].
[31] S.A. Abel and B.W. Schofield, Brane anti-brane kinetic mixing, millicharged particles and
SUSY breaking, Nucl. Phys. B 685 (2004) 150 [hep-th/0311051] [INSPIRE].
[32] S.A. Abel, J. Jaeckel, V.V. Khoze and A. Ringwald, Illuminating the Hidden Sector of String
Theory by Shining Light through a Magnetic Field, Phys. Lett. B 666 (2008) 66
[hep-ph/0608248] [INSPIRE].
[33] S.A. Abel, M.D. Goodsell, J. Jaeckel, V.V. Khoze and A. Ringwald, Kinetic Mixing of the
Photon with Hidden U(1)s in String Phenomenology, JHEP 07 (2008) 124 [arXiv:0803.1449]
[INSPIRE].

JCAP09(2019)049
[34] NA64 collaboration, Search for invisible decays of sub-GeV dark photons in missing-energy
events at the CERN SPS, Phys. Rev. Lett. 118 (2017) 011802 [arXiv:1610.02988] [INSPIRE].
[35] P. Jean et al., Early SPI/INTEGRAL measurements of 511 keV line emission from the 4th
quadrant of the Galaxy, Astron. Astrophys. 407 (2003) L55 [astro-ph/0309484] [INSPIRE].
[36] J. Chang et al., An excess of cosmic ray electrons at energies of 300–800 GeV, Nature 456
(2008) 362 [INSPIRE].
[37] PAMELA collaboration, An anomalous positron abundance in cosmic rays with energies
1.5–100 GeV, Nature 458 (2009) 607 [arXiv:0810.4995] [INSPIRE].
[38] Muon g-2 collaboration, Final Report of the Muon E821 Anomalous Magnetic Moment
Measurement at BNL, Phys. Rev. D 73 (2006) 072003 [hep-ex/0602035] [INSPIRE].
[39] H. Davoudiasl, H.-S. Lee and W.J. Marciano, ‘Dark’ Z implications for Parity Violation, Rare
Meson Decays and Higgs Physics, Phys. Rev. D 85 (2012) 115019 [arXiv:1203.2947]
[INSPIRE].
[40] H. Davoudiasl, H.-S. Lee, I. Lewis and W.J. Marciano, Higgs Decays as a Window into the
Dark Sector, Phys. Rev. D 88 (2013) 015022 [arXiv:1304.4935] [INSPIRE].
[41] B.S. Acharya, S.A.R. Ellis, G.L. Kane, B.D. Nelson and M.J. Perry, The lightest visible-sector
supersymmetric particle is likely to be unstable, Phys. Rev. Lett. 117 (2016) 181802
[arXiv:1604.05320] [INSPIRE].
[42] B. Holdom, Two U(1)’s and  Charge Shifts, Phys. Lett. 166B (1986) 196 [INSPIRE].
[43] D. Lüst, Intersecting brane worlds: A Path to the standard model?, Class. Quant. Grav. 21
(2004) S1399 [hep-th/0401156] [INSPIRE].
[44] S. Abel and J. Santiago, Constraining the string scale: From Planck to weak and back again, J.
Phys. G 30 (2004) R83 [hep-ph/0404237] [INSPIRE].
[45] P. Karageorgis and J.G. Stalker, Sharp bounds on 2m/r for static spherical objects, Class.
Quant. Grav. 25 (2008) 195021 [arXiv:0707.3632] [INSPIRE].
[46] H. Andreasson, Sharp bounds on the critical stability radius for relativistic charged spheres,
Commun. Math. Phys. 288 (2009) 715 [arXiv:0804.1882] [INSPIRE].
[47] S. Hod, Upper bound on the gravitational masses of stable spatially regular charged compact
objects, Phys. Rev. D 98 (2018) 064014 [arXiv:1903.10530] [INSPIRE].
[48] V. Cardoso, A.S. Miranda, E. Berti, H. Witek and V.T. Zanchin, Geodesic stability, Lyapunov
exponents and quasinormal modes, Phys. Rev. D 79 (2009) 064016 [arXiv:0812.1806]
[INSPIRE].
[49] R. Narayan, Black holes in astrophysics, New J. Phys. 7 (2005) 199 [gr-qc/0506078] [INSPIRE].
[50] Y. Peng, Upper bound on the radii of regular ultra-compact star photonspheres, Phys. Lett. B
790 (2019) 396 [arXiv:1812.04257] [INSPIRE].

– 15 –
[51] V. Cardoso, L.C.B. Crispino, C.F.B. Macedo, H. Okawa and P. Pani, Light rings as
observational evidence for event horizons: long-lived modes, ergoregions and nonlinear
instabilities of ultracompact objects, Phys. Rev. D 90 (2014) 044069 [arXiv:1406.5510]
[INSPIRE].
[52] P.V.P. Cunha, E. Berti and C.A.R. Herdeiro, Light-Ring Stability for Ultracompact Objects,
Phys. Rev. Lett. 119 (2017) 251102 [arXiv:1708.04211] [INSPIRE].
[53] S. Hod, On the number of light rings in curved spacetimes of ultra-compact objects, Phys. Lett.
B 776 (2018) 1 [arXiv:1710.00836] [INSPIRE].
[54] J. Keir, Slowly decaying waves on spherically symmetric spacetimes and ultracompact neutron
stars, Class. Quant. Grav. 33 (2016) 135009 [arXiv:1404.7036] [INSPIRE].

JCAP09(2019)049
[55] S. Hod, No-scalar-hair theorem for spherically symmetric reflecting stars, Phys. Rev. D 94
(2016) 104073 [arXiv:1612.04823] [INSPIRE].
[56] S. Bhattacharjee and S. Sarkar, No-hair theorems for a static and stationary reflecting star,
Phys. Rev. D 95 (2017) 084027 [arXiv:1704.02873] [INSPIRE].
[57] S. Hod, No nonminimally coupled massless scalar hair for spherically symmetric neutral
reflecting stars, Phys. Rev. D 96 (2017) 024019 [arXiv:1709.01933] [INSPIRE].
[58] S. Hod, Charged massive scalar field configurations supported by a spherically symmetric
charged reflecting shell, Phys. Lett. B 763 (2016) 275 [arXiv:1703.05333] [INSPIRE].
[59] S. Hod, Marginally bound resonances of charged massive scalar fields in the background of a
charged reflecting shell, Phys. Lett. B 768 (2017) 97 [arXiv:1806.06831] [INSPIRE].
[60] C.A.R. Herdeiro, J.C. Degollado and H.F. Rúnarsson, Rapid growth of superradiant instabilities
for charged black holes in a cavity, Phys. Rev. D 88 (2013) 063003 [arXiv:1305.5513]
[INSPIRE].
[61] N. Sanchis-Gual, J.C. Degollado, P.J. Montero, J.A. Font and C. Herdeiro, Explosion and Final
State of an Unstable Reissner-Nordström Black Hole, Phys. Rev. Lett. 116 (2016) 141101
[arXiv:1512.05358] [INSPIRE].
[62] Y. Peng, Scalar field configurations supported by charged compact reflecting stars in a curved
spacetime, Phys. Lett. B 780 (2018) 144 [arXiv:1801.02495] [INSPIRE].

– 16 –
 PHYSICAL REVIEW D 101, 084035 (2020)

Influence of dark matter on black hole scalar hair

†
Bartlomiej Kiczek* and Marek Rogatko
Institute of Physics, Maria Curie-Sklodowska University,
20-031 Lublin, pl. Marii Curie-Sklodowskiej 1, Poland

(Received 13 February 2020; accepted 3 April 2020; published 14 April 2020)

Searches for dark matter sector field imprints on the astrophysical phenomena are one of the most active
branches of the current researches. Using numerical methods, we elaborate the influence of dark matter on
the emergence of black hole hair and formation of boson stars. We explore thermodynamics of different
states of the system in Einstein-Maxwell-scalar dark matter theory with box boundary conditions. Finally,
we find that the presence of dark sector within the system diminishes a chance of formation of scalar hair
around a black hole.

DOI: 10.1103/PhysRevD.101.084035

I. INTRODUCTION fields and gravity [11] are widely studied in literature [12–16]
for a quite long period of time.
The astrophysical evidence of the illusive ingredient of
The purpose of our paper is to examine thermodynamical
our Universe, dark matter, is overwhelming and authorizes
properties and stability of the black holes and horizonless
the galaxy rotation curves, gravitational lensing, a thread-
objects-boson stars in Einstein-Maxwell-scalar system
like structure (cosmic web) on which ordinary matter
influenced by dark matter sector and envisage the role
accumulates [1,2]. On the contrary, the absence of the
of the dark matter in the elaborated problems.
evidence of the most popular particle candidates for
The organization of the paper is as follows. In Sec. II, we
baryonic dark matter stipulates the necessity of diversifying describe the basic features of the hidden sector model and
experimental efforts [3]. Black holes and ultracompact derived the basic equations needed in what follows.
horizonless objects being the ideal laboratories for dark Section III is devoted to the description of the obtained
matter studies may help us to answer the tantalizing numerical results. In Sec. IV, we concluded our researches.
question of how dark matter sector leaves its imprint in
the physics of these objects. However, it happens that II. MODEL
Schwarzschild black hole has a negative specific heat and it
cannot be in equilibrium with thermal radiation. To over- We consider the spacetime manifold with timelike boun-
come this difficulty, the idea of enclosing the black hole dary ∂M, which will be referred as a box. The action for
within a box was proposed [4,5]. Einstein-Maxwell sys- Einstein-Maxwell scalar dark matter gravity is provided by
tems with box boundary conditions were elaborated in [6], Z 
where it was established that the phase structure of the pffiffiffiffiffiffi 1 α
S¼ d x −g R − Fμν Fμν − Bμν Fμν
4
models was similar to AdS gravity. Inclusion of the M 4 4
 Z
additional scalar field to the theory in question envisages 1 pffiffiffiffiffiffi
the correspondence of phase transitions in gravity in a box − Bμν Bμν − jDΨj2 − m2 jΨj2 − d3 x −γ K; ð1Þ
4 ∂M
with s-wave holographic superconductor [7–9]. The ther-
modynamical studies of Einstein-Maxwell scalar systems where Fμν is a Maxwell field strength tensor, Bμν is a strength
in the asymptotically flat spacetime with reflecting boun- tensor of a hidden sector vector boson. The complex scalar
dary conditions were conducted in [10]. A certain range of field Ψ ¼ ψeiθ , where θ denotes the phase, is coupled only to
parameters allows to obtain stable black hole solution, the ordinary electromagnetic field by the covariant derivative
giving a way to circumvent no-hair theorem. Dμ ¼ ∇μ − iqAμ . The theoretical justifications of the model
The next compact objects studied in our paper, from the
in question originate from M/string theories, where such
point of view of the influence of dark matter on their physics, mixing portals coupling Maxwell and auxiliary gauge fields
are boson stars. Boson stars being a self-gravitating solution can be encountered [17]. The hidden sectors states are
of massive scalar field with a potential coupled to gauge charged under their own groups and interact with the visible
sector via gravitational interactions. The realistic string
*
[email protected] compactifications establish the range of values for α between
†
[email protected] 10−2 and 10−10 [18–21]. It seems that astrophysical

2470-0010=2020=101(8)=084035(8) 084035-1 © 2020 American Physical Society

BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 101, 084035 (2020)

observations of gamma rays of energy 511 keV [22], positron α

excess in galaxies [23], and muon anomalous magnetic ∇μ Fμν þ ∇μ Bμν − 2q2 Aν ψ 2 ¼ 0; ð9Þ
2
moment [24] argue for the aforementioned idea of coupling
α
Maxwell field with dark matter sector. Recent experiments ∇μ Bμν þ ∇μ Fμν ¼ 0: ð10Þ
aimed at gamma ray emissions from dwarf galaxies [25], 2
dilatonlike coupling to photons caused by ultralight dark
As in the case of Eq. (3), the last two equations can be
matter [26], oscillations of the fine structure constant [27],
rewritten as
revisions of the constraints on dark photon 1987A supernova
emission [28], measurements of excitation of electrons in
α̃∇μ Fμν − 2q2 Aν ψ 2 ¼ 0; ð11Þ
CCD-like detector [29], as well as the examinations in eþ e−
Earth colliders [30], give us some hints for the correctness of α
the proposed model. They and the future planned ballon ∇μ Bμν þ q2 Aν ψ 2 ¼ 0: ð12Þ
α̃
d’essai will ameliorate the mass constraints on the hidden
sector particles, especially for dark photons. Consequently, the explicit forms of the equations of
The second integral denotes the Gibbons-Hawking motion yield
boundary term of our box with γ metric on the three-
dimensional hypersurface ðr ¼ rb Þ, with the extrinsic q2 rϕ2 ψ 2
curvature K. h0 − rhψ 02 − ¼ 0; ð13Þ
g2
Varying the action (1), we get the equations of motion of
 
the forms 1 1 02 q2 rϕ2 ψ 2 1
g0 þ g þ rψ þ −
r 2 2gh r
ð∇μ − iqAμ Þð∇μ − iqAμ ÞΨ − m2 Ψ ¼ 0; ð2Þ
r 02
þ ðϕ þ αχ 0 ϕ0 þ χ 02 þ m2 hψ 2 Þ ¼ 0; ð14Þ
2h
α̃∇μ Fμν ¼ jν ; ð3Þ
 
00 2 h0 2q2 ϕψ 2
2
where α̃ ¼ 1 − α4 and the current jν is provided by the ϕ þ − ϕ0 − ¼ 0; ð15Þ
r 2h α̃g
relation
   2 2 
ν † ν ν ν ν † 00 2 h0 g0 0 qϕ 2 ψ
j ¼ iq½Ψ ð∇ − iqA ÞΨ − Ψð∇ þ iqA ÞΨ : ð4Þ ψ þ þ þ ψ þ −m ¼ 0; ð16Þ
r 2h g gh g
In what follows, we use a time-independent spherically  
2 h0 0 αq2 χψ 2
symmetric line element, with the metric coefficients being χ 00 þ − χ þ ¼ 0: ð17Þ
functions of r coordinate, r 2h α̃g

To solve the equations of the theory in question, one has to

dr2
ds2 ¼ −gðrÞhðrÞdt2 þ þ r2 ðdθ2 þ sin θ2 dϕ2 Þ; ð5Þ provide adequate boundary conditions. Namely, we can
gðrÞ pick either a horizonless or a black hole solution. In case of
a black hole, we expand the underlying functions in a
and the adequate components of the fields in the theory will Taylor series around the horizon of radius rh ,
constitute radial functions of the forms
ψ ¼ ψ 0 þ ψ 1 ðr − rh Þ þ ψ 2 ðr − rh Þ2 þ Oðr3 Þ; ð18Þ
Aμ dxμ ¼ ϕðrÞdt; Bμ dxμ ¼ χðrÞdt; Ψ ¼ ΨðrÞ: ð6Þ
ϕ ¼ ϕ1 ðr − rh Þ þ ϕ2 ðr − rh Þ2 þ Oðr3 Þ; ð19Þ
In general, the scalar field can have harmonic time
dependence which can be absorbed by a redefinition of
g ¼ g1 ðr − rh Þ þ g2 ðr − rh Þ2 þ Oðr3 Þ; ð20Þ
the gauge field function. Having this in mind it can be seen
that the r component of the equations of motion for the
h ¼ 1 þ h1 ðr − rh Þ þ Oðr2 Þ; ð21Þ
gauge and scalar fields leads to the conclusion that
ΨðrÞ ¼ ψðrÞ. By virtue of this, the following equations
of motion are provided: χ ¼ χ 1 ðr − rh Þ þ χ 2 ðr − rh Þ2 þ Oðr3 Þ: ð22Þ

1 We set g0 ¼ 0 due to occurrence of the black hole

Rμν − gμν R ¼ T μν ; ð7Þ event horizon. For the regularity of the Uð1Þ-gauge
2 fields on the event horizon, one also puts ϕ0 and χ 0 equal
to zero (in order to keep the terms with division by gðrh Þ in
∇μ ∇μ ψ − q2 Aμ Aμ ψ − m2 ψ ¼ 0; ð8Þ equations of motion finite). By implementing the expansions

084035-2
INFLUENCE OF DARK MATTER ON BLACK HOLE SCALAR … PHYS. REV. D 101, 084035 (2020)

(18)–(22) into the equations of motion, we find out that constrains in our shooting procedure for ϕ1 and χ 1 . We set a
frh ; ψ 0 ; ϕ1 ; χ 1 ; αg comprise free parameters of the theory in domain of shooting parameters from the series expansion of
question, while the remaining ones can be expressed the solutions near the horizon, then by using the iterative
by them. bisections one finds a solution that meets constrains, with a
As far as the boson star scenario is concerned, we desired tolerance. Therefore, parameters fμd ; αg, which are
perform a similar expansion. However, since the configu- controlling, respectively, amount of dark charge and the
ration in question is horizonless, the expansion accom- coupling strength remain free; thus, they can be varied to
plishes around the origin of the reference frame. At r ¼ 0, see their impact on the system in question. For conven-
we require that the derivatives of all the functions are set ience, let us refer to the parameter ψ ð1Þ, as a condensation,
equal to zero, which ensures that there is no kink at this which serves as a handy analogy to holographic theory. As
point. At r ¼ rb , we establish the Dirichlet boundary mentioned above, in our numerical scheme, we treat μd as
condition for the scalar field ψðrb Þ ¼ 0 (the reflecting an input parameter in our code; however, one might not be
mirrorlike boundary conditions). interested in expressing these relations in a language of
Asymptotic analysis of matter fields, at the box boun- chemical potentials. Therefore, one might compute the total
dary, enables us to write dark charge of the system
Z
ψ ∼ ψ ð0Þ þ ψ ð1Þ ðrb − rÞ þ Oðr2 Þ; ð23Þ 1 pffiffiffiffiffiffi
Qd ¼ lim Bμν tμ nν −gd2 θ; ð26Þ
r→rb 4π S 2

ϕ ∼ ϕð0Þ þ ϕð1Þ ðrb − rÞ þ Oðr2 Þ; ð24Þ where tμ is a unit timelike vector and nμ is a normal vector
to the boundary. In the similar manner, we compute
χ ∼ χ ð0Þ þ χ ð1Þ ðrb − rÞ þ Oðr2 Þ: ð25Þ electrical charge for Fμν.
As was proposed in Refs. [10,16], because of the fact
that the scalar field satisfies the reflecting mirrorlike III. RESULTS
boundary conditions ψðrb Þ ¼ 0, one can fix ψ ð0Þ ¼ 0 We commence with the hairy black hole solution (HBH),
and the other parameter ψ ð1Þ can be used for the phase i.e., a system with an event horizon and nontrivial scalar
transition description. This approach to the problem in field profile. The parameter space of HBH can be illustrated
question is widely exploited in holographic studies of on a plane of chemical potential and Hawking temperature
superconductors and superfluids. (μ-T) as a triangular shape. That region is bound between
For the gauge fields, one has that ϕð0Þ ¼ μ and χ ð0Þ ¼ μd boson star phase from the left-hand side and generalized
as chemical potentials for visible and hidden sector fields, RN solution from the right-hand side. A schematic phase
treating the system as a grand canonical ensemble. In order diagram has been presented in the Fig. 1, where both
to conduct the thermodynamical analysis, we calculate the mentioned lines are marked. Moreover, the influence of the
free energy of the system, to see which phase is thermo- dark sector on phase boundaries is visualized by arrows,
dynamically preferable, for a fixed temperature. In the case showing the trend of the flow by increasing the hidden
of a hairless solution, we take into account the classical sector chemical potential.
formula F ¼ E − TS − μQ − μd Qd , where E is Brown- The hairy configuration can be achieved for a specific
York quasilocal energy [4,5]. Nevertheless, this approach value of the chemical potential. Below the value μRN scalar
may cause problems in hairy solution analysis, being cannot condensate and we get RN-dark matter black hole.
insufficient to capture the mass of the scalar field. On the other hand, for the value greater than the critical one,
Therefore, we treat this problem evaluating the on shell μc , the system becomes unstable. By stable hairy solution,
action in Euclidean signature F ¼ TScl , which enables to we mean a constrained solution of the equations of motion
take into considerations the nontrivial profile of scalar field (13)–(17) that fulfils the boundary conditions with desired
constituting the solution of the underlying system of tolerance and its free energy is lower than the free energy of
differential equations. RN and BS, making it the ground state of the system. We
Solution of Eqs. (13)–(17) with ψ ¼ 0 can be achieved can define μc as the chemical potential for which the phase
analytically, giving the Reissner-Nordstrom (RN) dark transition driven by temperature is no longer of second
matter black object [31]. To proceed further and accomplish order and the condensate collapses. In the range between
the complete numerical analysis of the underlying equa- μRN and μc , we contend a typical second order phase
tions, we implement the shooting method, integrating the transition, depicted in Fig. 2. In the vicinity of critical
aforementioned relations from rh to rb , using the fourth temperature, condensation can be described by a function
order Runge-Kutta method. From the set of free parameters, ψ ð1Þ ∼ ðT c − TÞ1=2 . It should also be noted that establishing
we fix the scalar magnitude on the event horizon ψ 0 and an HBH solution requires relatively a large value of the
pick rh , ϕ1 , and χ 1 to be shooting parameters. Moreover, scalar charge. In our calculation, we used q ¼ 100 and a
we impose values on both chemical potentials that serve as small mass of m ¼ 10−6 .

084035-3
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 101, 084035 (2020)

FIG. 3. Double valued profiles of condensation as functions of

Hawking temperature caused by an increasing amount of dark
matter in the system with α ¼ 10−3 and μ ¼ 0.14. While the first
FIG. 1. A scheme of phase diagram of the described system. transition for μd ¼ 0.08 might still be considered as a regular, the
Blue-yellow line indicates the border between boson star and another strictly not—the value of condensation becomes double
hairy black hole parameter space, while the red line depicts hairy valued for some range of temperatures. Moreover, these solutions
BH—generalized RN BH phase boundary. The arrows on the obey the boundary conditions but their free energy is larger than
scheme show us the flow of phase boundaries driven by the both BSs and RNs; therefore, they cannot be considered as
chemical potential of dark matter. Lines have been split and thermodynamically preferred.
labeled from A to C, with a point D being the center of rotation of
left-hand side boundary.
numerical method finds constrained solutions, but due to
free energy leap, they are not thermodynamically preferred.
The exotic phase effect occurs in case when scalar mass is
close to or equal zero, for a mass away from this limit we do
not obtain that phase. Instead, we have a sharp crossing,
from stable solutions below μc to the situation when the
equations of motion do not provide solutions with con-
densed scalar above the μc threshold at all. It is worth
mentioning that a similar condensation-temperature profile
has been shown in the so-called vector p-wave holographic
superfluids [32], but a first order phase transition was
hidden behind it. However, it was revealed that for the real
value of the vector field, the model in question gave us the
same description as holographic s-wave model with dark
matter sector [33–34].
Dark matter gauge field plays an interesting role in this
transition, as it accelerates the appearance, let us say, the
FIG. 2. Condensation ψ ð1Þ as a function of temperature for the exotic phase. For a larger value of dark sector chemical
different values of μd and α ¼ 10−3 . For μ ¼ 0.1, a typical second potential, ψ ð1Þ becomes double valued for the lower
order phase transition takes place, the dark matter presence chemical potential, which is depicted in Fig. 3, where μd
influences the transition point and the condensation. has gradually increasing value. Moreover, when system
enters the exotic phase, its free energy rises repeatedly and
exceeds the free energy of RN black hole, so the hairy
Let us now discuss physical mechanisms behind the phase is no longer a preferred option. In this way, that effect
phase boundaries flow from Fig. 1. When we cross the line restricts the range of chemical potential where the second
of the critical chemical potential value μc , one encounters order phase transition may occur, μc becomes a descending
the exotic phase, where for one value of temperature we function of the dark charge [see curve (A) in the Fig. 1].
have two values of the condensation parameter ψ ð1Þ. However, it cannot be increased without a limit. For every
Moreover, by evaluating its free energy, we can find it is value of the electric charge, there exists a certain limit of
so high that the hairy state is no longer stable—our dark charge, below which a formation of scalar hair is

084035-4
INFLUENCE OF DARK MATTER ON BLACK HOLE SCALAR … PHYS. REV. D 101, 084035 (2020)

FIG. 4. Relative change of the critical temperature of hairy FIG. 5. Quasirotation of the phase transition boundary between
black hole—generalized RN black hole as a function of a ratio of boson star and hairy black hole caused by dark sector charge with
chemical potentials. The temperature ratio has been normalized to α coupling equal to 10−4 . It can be observed that the μ threshold
the critical temperature of dark matter free solution, for hairy BH solution is significantly lowered and some param-
where μd ¼ 0. eter space of boson star is taken for the advantage of hairy BH for
lower values of the chemical potential.

possible. Above it, such condensation cannot take

place and no stable solutions are found. This phenomenon potential [line (C) in Fig. 1]. This process is depicted in
adds up both gravitational influence of the charge on the Fig. 5, which asserts a phase diagram at the boundary
metric and nongravitational coupling between both gauge between hairy black hole and boson star. While the boson
fields. star is a horizonless object, its Hawking temperature
Now, let us draw our attention to the HBH-RN BH remains undefined. However, it is possible to calculate
(B) border. An interesting effect that hidden sector exerts on its characteristic—condensation and free energy as a
the hairy black hole system is the shifting of the critical function of chemical potential. Then to obtain the phase
temperature of the phase transition. The larger growth of boundary curve, we start in the hairy black hole regime,
dark matter charge (and also μd ) we observe, the lower then one moves toward lower Hawking temperatures and
value of the transition temperature one achieves. Such study the value of the free energy of the hairy black hole on
effect has been depicted in the Fig. 4, where the critical the way. When it exceeds the free energy of a boson star, for
temperature ratio described by the relation the corresponding value of the chemical potential, the
transition point is found. Both phases of the system are
influenced by the hidden sector; nonetheless, the free
T c ðμd Þ − T c ð0Þ
δT c ¼ energy of a boson star is affected much less than that of
T c ð0Þ the black hole. Dark sector causes a significant drop of free
energy of a hairy black hole. It means that the stability of a
is shown as a function of the chemical potential hairy solution is preserved for lower temperatures given the
of hidden sector normalized to the visible sector chemi- presence of the dark matter in the system. Such effect
cal potential. One can notice that the shift of the critical causes that hairy black hole solution is thermodynamically
temperature is proportional to the square of μd . Obvi- preferable for the lower Hawking temperatures and limits
ously, it cannot decrease as low as one wishes and a the emergence of boson star. The presence of α-coupling
certain limit exists which has been discussed in analytic constant slightly diminishes the space of parameters for
solution of dark matter charged RN-like black hole [31]. which boson star can emerge.
The descent of the critical temperature becomes steeper At last, it is sensible to mention some points that seem to
for larger value of the chemical potential of the visible be dark sector resistant. One of them appears on the phase
sector. It can be explained by the nongravitational boundary, labeled with (D) on the phase diagram scheme in
interaction between fields via kinetic mixing term, which Fig. 1. This point or rather its neighborhood does not seem
plays a significant role when both fields are sufficiently to be susceptible on the dark charge presence in the system.
strong. For a particular numerical example, like in Fig. 5, it is
To proceed further, we shed some light on the influence placed around μ ≈ 0.1160942 and T ≈ 0.2929537. Another
of dark matter sector on the black hole-boson star phase one can be noticed in the condensation-Hawking temper-
transition in the stable area of small values of the chemical ature dependence presented in Fig. 2. All the curves

084035-5
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 101, 084035 (2020)

The second order term takes the approximated form

(a)
ð1Þ ψ ð1Þ ðT; μd1 Þ − ψ ð1Þ ðT; μd2 Þ
ψ 1 ðTÞ ¼ ; ð28Þ
μ2d1 − μ2d2

where in a certain example of curves from Fig. 2 we took

μd1 ¼ 0.12 and μd2 ¼ 0.08. The zero of this function refers
to the isosbestic point, where the contribution of μd is by
definition none. By calculating the above function with
help, the leading order of the influence of the dark sector
may be subtracted from the main function
ð1Þ
ψ̃ ð1Þ ðT; μd Þ ¼ ψ ð1Þ ðT; μd Þ − μ2d ψ 1 ðTÞ: ð29Þ
In the similar manner, we can expand and analyze the
chemical potential as a function of Hawking temperature,
(b) parametrized by μd from boson star-hairy black hole phase
boundary,
μðT; μd Þ ¼ μðT; 0Þ þ μ2d μ1 ðTÞ þ Oðμ3d Þ: ð30Þ
We define μ1 ðTÞ analogically to (28) with μd1 ¼ 0.2 and
μd2 ¼ 0.08 and perform the same transformation for
μðT; μd Þ curve as for ψðT; μd Þ in (29). The effect of these
transformations is depicted in Fig. 6, where all curves tend
to be much closer to each other than before. Obviously, the
total effect of μd is not ruled out completely, since it is much
more complex than in the considered expansion.

IV. CONCLUSION
FIG. 6. Panel (a) presents condensation profile after the trans-
formation performed in Eq. (29). The plot refers to the same data In our paper, based on Einstein-Maxwell scalar dark
as in Fig. 2; however, it can be seen that the separation between matter theory, where the hidden sector is mimicked by the
curves is significantly smaller. In case of panel (b), which refers auxiliary Uð1Þ-gauge field coupled to the ordinary
to Fig. 5, all curves appear to overlap with the dark matter free Maxwell one by the kinetic mixing term with a coupling
solution. The aforementioned transformation had removed the
constant α, we elaborate two scenarios of emergence of a
leading term of the dark sector influence.
hairy black hole or a boson star. The main motivation
standing behind our research was to shed some light on the
influence of dark matter sector on the physics and thermo-
certainly cross each other in one point, located at ψ ð1Þ ≈ dynamics of these systems.
0.345 and T ≈ 0.3325. This interesting phenomenon shows The obtained results reveal that the coupling between
that while the dark sector may modify the phase structure of visible and hidden sectors plays a complex role in the
the system and has an imprint on its critical quantities, there behavior of scalar hair. The parameter space (μ-T), where
exists a specific configuration of the system, that remains these solutions constitute a thermodynamically favorable
completely untouched. phase, is being narrowed on two boundaries and extended
The points in question constitute the so-called isosbestic to another one. The dark sector’s presence strongly reduces
ones [35], where the curves dependent on temperature T the value of critical chemical potential, above which the
and parametrized by values of dark matter chemical hairy solution becomes unstable. Moreover, the critical
potential, intersect. They illustrate the influence of temper- temperature of HBH-RN-like solution is shifted toward the
lower value of Hawking temperature. However, the boun-
ature on condensation ψ ð1Þ and chemical potential of visible
dary between HBH and boson stars is shifted toward the
sector. At this point, we may perform a short analysis,
latter. The presence of the dark sector lowers the free
which would reveal the leading order term of the dark
energy of HBH system, which broadens the parameter
sector influence. We take a following expansion of the
space available for the emergence of the object in question
condensate:
by a noticeable extent, i.e., leaving boson star as an adverse
ð1Þ phase in the low μ regime. It appears that the free energy of
ψ ð1Þ ðT; μd Þ ¼ ψ ð1Þ ðT; 0Þ þ μ2d ψ 1 ðTÞ þ Oðμ3d Þ: ð27Þ boson stars in the considered configuration reacts faintly to

084035-6
INFLUENCE OF DARK MATTER ON BLACK HOLE SCALAR … PHYS. REV. D 101, 084035 (2020)

the presence of Uð1Þ-gauge dark matter field. While the different obstacles a significant abundance of dark sector
response of the system is visible, it is much smaller in may prevent hairy solutions from emerging.
magnitude than of the condensate around a black hole. The curves ψ ð1Þ ðTÞ and μðTÞ, parametrized by the values
However, we suppose that interesting results may be of dark matter chemical potential, reveal the isosbestic
achieved for more robust model of a scalar field, e.g., points, where they all intersect. One has the specific
containing self-interacting terms. configurations of the considered system which is unaf-
In the view of presented results, it seems that the hairy fected by the influence of hidden sector. At the points in
solutions are not only battled by no-hair theorems origi- question, we perform analysis revealing that the leading
nating from the theory of black holes, but also by a factor order influence of dark matter on the condensation ψ ð1Þ and
that is commonly present in our Universe—the dark matter. chemical potential of ordinary matter is quadratic in μd .
Even if such formation would be possible despite different As a concluding remark, we present promising future
obstacles a significant abundance of dark matter may research directions. We have elaborated the simple box-
boundary models of a hairy black hole and a boson star (the
prevent hairy solutions from emerging.
so-called small boson star). The tantalizing question can be
To visualize the impact of dark sector, we compute the
asked about the different boson star configurations with
area of HBH parameter space between both phase R boun-
additional fields and potentials. Further investigations in
daries. One can consider simple integration ðμBS ðTÞ −
this direction will be published elsewhere.
μRN ðTÞÞdT of the curves from Fig. 5, which reveals that the
dark sector with μd ¼ 0.14 takes away approximately 27%
ACKNOWLEDGMENTS
of the hairy black hole’s parameter space, compared to dark
matter free solution. It is indeed a significant difference, We acknowledge K. I. Wysokinski and N. Sedlmayr for
because even if such formation would be possible despite fruitful discussions on various occasions.

[1] R. Massey et al., Dark matter maps reveal cosmic scaffold- [12] B. Kleihaus, J. Kunz, C. Lammerzahl, and M.List, Charged
ing, Nature (London) 445, 286 (2007). boson stars and black holes, Phys. Lett. B 675, 102 (2009).
[2] J. P. Dietrich, N. Werner, D. Clowe, A. Finoguenov, T. [13] B. Hartmann, B. Kleihaus, J. Kunz, and I. Schaffer,
Kitching, L. Miller, and A. Simionescu, A filament of dark Compact boson stars, Phys. Lett. B 714, 120 (2012).
matter between two cluster of galaxies, Nature (London) [14] D. Pugliese, H. Quevedo, J. A. Rueda, and R. Ruffini,
487, 202 (2012). Charged boson stars, Phys. Rev. D 88, 024053 (2013).
[3] G. Bertone and T. M. P. Tait, A new era in the search for dark [15] S. Kumar, U. Kulshreshtha, D. S. Kulsreshtha, S. Kahlen,
matter, Nature (London) 562, 51 (2018). and J. Kunz, Some new results on charged compact boson
[4] J. W. York, Black-hole thermodynamics and the Euclidean stars, Phys. Lett. B 772, 615 (2017).
Einstein action, Phys. Rev. D 33, 2092 (1986). [16] Y. Peng, Large regular reflecting stars have no scalar field
[5] H. W. Braden, J. D. Brown, F. B. Whiting, and J. W. York, hair, Eur. Phys. J. C 79, 309 (2019).
Charged black hole in a grand canonical ensemble, Phys. [17] B. S. Acharya, S. A. R. Ellis, G. L. Kane, B. D. Nelson, and
Rev. D 42, 3376 (1990). M. J. Perry, Lightest Visible-Sector Supersymmetric Particle
[6] G. W. Gibbons and M. J. Perry, Black Holes in Thermal is Likely Unstable, Phys. Rev. Lett. 117, 181802 (2016).
Equilibrium, Phys. Rev. Lett. 36, 985 (1976). [18] S. A. Abel and B. W. Schofield, Brane-antibrane kinetic
[7] S. A. Hartnoll, C. P. Herzog, and G. T. Horowitz, Building a mixing, millicharged particles and SUSY breaking, Nucl.
Holographic Superconductor, Phys. Rev. Lett. 101, 031601 Phys. B685, 150 (2004).
(2008). [19] S. A. Abel, J. Jaeckel, V. V. Khoze, and A. Ringwald,
[8] Y. Peng, B. Wang, and Y. Liu, On the thermodynamics of Illuminating the hidden sector of string theory by shining
the black hole and hairy black hole transitions in the light through a magnetic field, Phys. Lett. B 666, 66 (2008).
asymptotically flat spacetime with a box, Eur. Phys. J. C [20] S. A. Abel, M. D. Goodsell, J. Jaeckel, V. V. Khoze, and A.
78, 176 (2018). Ringwald, Kinetic mixing of the photon with hidden U(1)s
[9] Y. Peng, Studies of a general flat space/boson star transition in string phenomenology, J. High Energy Phys. 07 (2008)
model in a box through a language similar to holographic 124.
superconductors, J. High Energy Phys. 07 (2017) 042. [21] D. Banerjee et al., Search for Invisible Decays of Sub-GeV
[10] P. Basu, C. Krishnan, and P. N. Bala Subramanian, Hairy Dark Photons in Missing-Energy Events at the CERN SPS,
black holes in a box, J. High Energy Phys. 11 (2016) 041. Phys. Rev. Lett. 118, 011802 (2017).
[11] P. Jetzer, Boson stars, Phys. Rep. 220, 163 (1992); S. L. [22] P. Jean et al., Early SPI/INTEGRAL measurements of
Libling and C. Palenzuela, Dynamical boson stars, Living 511 keV line emission from the 4th quadrant of the Galaxy,
Rev. Relativity 15, 6 (2012). Astron. Astrophys. 407, L55 (2003).

084035-7
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 101, 084035 (2020)

[23] J. Chang et al., An excess of cosmic ray electrons at energies [30] J. P. Lees et al., Search for a Dark Photon in eþ e− Collisions
of 300–800 GeV, Nature (London) 456, 362 (2008). at BABAR, Phys. Rev. Lett. 113, 201801 (2014).
[24] O. Adriani et al. (PAMELA Collaboration), An anomalous [31] B. Kiczek and M. Rogatko, Ultra-compact spherically
positron abundance in cosmic rays with energies 1.5– symmetric dark matter charged star objects, J. Cosmol.
100 Gev, Nature (London) 458, 607 (2009). Astropart. Phys. 09 (2019) 049.
[25] A. Geringer-Sameth and M. G. Walker, Indication of [32] R.-G. Cai, L. Li, and L.-F. Li, A holographic p-wave
Gamma-Ray Emission from the Newly Discovered Dwarf superconductor model, J. High Energy Phys. 01 (2014) 032.
Galaxy Reticulum II, Phys. Rev. Lett. 115, 081101 (2015). [33] M. Rogatko and K. I. Wysokinski, P-wave holographic
[26] K. K. Boddy and J. Kumar, Indirect detection of dark matter superconductor/insulator phase transitions affected by dark
using MeV-range gamma-rays telescopes, Phys. Rev. D 92, matter sector, J. High Energy Phys. 03 (2016) 215.
023533 (2015). [34] M. Rogatko and K. I. Wysokinski, Condensate flow in
[27] K. Van Tilburg, N. Leefer, L. Bougas, and D. Budker, holographic models in the presence of dark matter, J. High
Search for Ultralight Scalar Dark Matter with Atomic Energy Phys. 10 (2016) 152.
Spectroscopy, Phys. Rev. Lett. 115, 011802 (2015). [35] D. Vollhardt, Characteristic Crossing Points in Specific Heat
[28] J. H. Chang, R. Essig, and S. D. McDermott, Revisiting Curves of Correlated Systems, Phys. Rev. Lett. 78, 1307
Supernova 1987A constraints on dark photons, J. High (1997); S. Uchida, T. Ido, H. Takagi, T. Arima, Y. Tokura,
Energy Phys. 01 (2017) 107. and S. Tajima, Optical spectra of La2−x Srx CuO4 : Effect of
[29] M. Crisler, R. Essig, J. Estrada, G. Fernandez, J. Tiffenberg, carrier doping on the electronic structure of the CuO2 plane,
M. S. Haro, T. Volansky, and T.-T. Yu (SENSEI Collabo- Phys. Rev. B 43, 7942 (1991); J. K. Freericks, T. P.
ration), SENSEI: First Direct-Detection Constraints on Sub- Devereaux, and R. Bulla, Exact theory for electronic Raman
GeV Dark Matter from a Surface Run, Phys. Rev. Lett. 121, scattering of correlated materials in infinite dimensions,
061803 (2018). Phys. Rev. B 64, 233114 (2001).

084035-8
J ournal of Cosmology and Astroparticle Physics
An IOP and SISSA journal

Holographic DC SQUID in the

presence of dark matter

JCAP01(2021)063
Bartłomiej Kiczek, Marek Rogatko and Karol I. Wysokiński
Institute of Physics, Maria Curie-Sklodowska University,
pl. Marii Curie-Skłodowskiej 1, Lublin 20-031, Poland
E-mail: [email protected], [email protected],
[email protected]

Received October 21, 2020

Accepted December 2, 2020
Published January 28, 2021

Abstract. The gauge-gravity duality has been applied to examine the properties of holo-
graphic superconducting quantum device (SQUID), composed of two S-N-S Josephson junc-
tions, influenced by dark sector modelled by the additional U(1)-gauge field coupled to the
ordinary Maxwell one. The dark matter sector is known to affect the properties of supercon-
ductors and is expected to enter the current-phase relation. The kinetic mixing between two
gauge fields provides a mechanism allowing for the conceivable observation of the effect. We
find small but visible effect of the dark matter particle traversing the device, which shows up
as a change of its maximal current.

Keywords: dark matter theory, gravity, string theory and cosmology

ArXiv ePrint: 1904.00653

c 2021 IOP Publishing Ltd and Sissa Medialab https://doi.org/10.1088/1475-7516/2021/01/063

Contents

1 Introduction 1

2 The holographic model and equations of motion 5

2.1 How to build holographic SQUID influenced by dark matter sector? 5
2.2 Equations of motion for the model 6

3 Numerical results 11

JCAP01(2021)063
3.1 Homogeneous superconductor — warm up 11
3.2 Analysis of the holographic SQUID 12

4 Holographic versus condensed matter SQUID 16

5 Discussion and conclusion 20

1 Introduction

The quest for the dark matter in the Universe has been one of the most important topics of
the current research in cosmology and physics [1]. Contemporary astronomical observations
of galaxies and primordial radiation endorse the fact that our Universe is made mostly (over
23 percent of its mass) of non-luminous dark matter. Several new types of fundamental
particles have been claimed as candidates for dark matter sector. They are expected to
interact with nuclei in suitable detector materials on Earth. It was claimed only by the
DAMA collaboration [2, 3] that they observed modulation in the rate of interaction events
which might be the trace of dark matter sector. Several groups want to reproduce the DAMA
results but in vain [4]. Recently the problem of dark matter identity and its detection has
gained new impetus after the XENON1T experiment has reported the signal which source
could not be explained [5]. This discovery has sparked intensive theoretical investigations
of various scenarios which elucidate the observed effect [6–10]. This situation triggers the
discussion concerning the composition, interaction with ordinary matter, the self-interaction
and the possible ways to discriminate between various models of dark matter [11].
It was argued that due to the growing sense of ‘crisis’ in the dark matter particle com-
munity, arising from the absence of evidences for the most popular candidates for dark matter
particles like WIMPs, axions and sterile neutrinos, diversifying the experimental effort should
be paid attention to. These efforts ought to accomplish upcoming astronomical surveys, grav-
itational wave observatories which can provide us some complimentary information about the
dark matter sector. They constitute our best hope for making progress in this direction. Un-
conventional experiments and techniques are also being looked for. One of the prospect
directions is related to clever usage of molecular or condensed matter systems including su-
perconductors or superconducting devices. Recent proposals authorise search for bosonic
dark matter via absorption in superconductors [12], using superfluid helium [13] or optical
phonons in polar materials [14] to detect light dark matter. More exotic proposals are based
on the observations of color centres production in crystals [15], or the usage of bulk three-
dimensional Dirac semi-metals [16] and other topological semiconducting compounds [17], as

–1–
well as, multilayered optical devices [18], for the detection of possible candidates for dark
matter sector.
The theoretical approach we have been using in our research relies on the AdS/CFT
correspondence [19, 20], which has been found to provide strong coupling description [21, 22]
of many condensed matter models. In particular this gauge-gravity duality has been shown to
describe U(1) and SU(2) symmetry breaking superconducting transitions. The applications
of the holography to study various models of superconductivity have been recently reviewed
in [23].
Our previous AdS/CFT studies [24]–[34] on holographic superconductors and super-
fluids reveal the influence of dark matter sector on various properties of them, e.g., super-
conducting transition temperature, Ginzburg-Landau ratio, vortices and condensation flow,

JCAP01(2021)063
viscosity bound for anisotropic superfluids and interacting currents in holographic Dirac fluid.
These findings may in principle help to design future condensed matter experiments oriented
on detection of dark matter. However, more sensitive probes are still searched for. One of
them is proposed in the current paper, i.e., examination of the properties of DC SQUID
device built of two Josephson junctions and influenced by dark sector.
The Josephson effect [35] is the well known quantum phenomenon observed in devices
consisting of two superconductors separated by a thin layer of normal metal, insulator or a
superconductor with much lower transition temperature. Such junctions, known as Josephson
junctions, even without external bias sustain the direct current I which magnitude is related
to the phase difference φ = φL − φR
I = Imax sin φ, (1.1)
where φL and φR are the phases of the superconducting order parameter on the left and
right hand side of the junction. When the external magnetic flux Φ penetrates the interior
of SQUID its maximal current depends on the flux as Imax (Φ) = Ic cos πΦ φ0 , where φ0 = 2e is
h

the magnetic flux quantum. This relation is at the heart of many devices able to detect tiny
magnetic fields [36]. As it has already alluded to, our aim is to study the effect of the dark
sector particle on the holographic SQUID. Recently the so-called second order Josephson
effect has been proposed to detect axionic dark matter [37, 38].
The holographic model of Josephson junction was also paid attention to. In [39] a
gravitational dual of S-N-S junction was constructed and the calculations on the gravity side
reproduced the standard relation between the current across the junction and the phase dif-
ference of the considered condensate. In the probe limit, Maxwell field coupled to complex
scalar one, was examined in the background of AdS-Schwarzschild black brane. The Joseph-
son junction array was considered in [40], where a model was proposed in the background of
multi-supergravity theories on products of distinct asymptotically AdS spacetimes coupled
by mixed boundary conditions. Among all it was found that the Cooper-pair condensates
were described by a discretised Schrödinger-like equation. In a continuum limit the relation
in question became a generalised Gross-Pitajevski equation, known from the long-wavelength
description of superfluids.
A holographic model of S-I-S junction constructed by examining a complex scalar field
coupled to Maxwell one, in the background of four-dimensional AdS soliton was examined
in [41]. The dependence of the maximal current on the dimension of the condensate operator
and the width of the junction were presented. On the other hand, a holographic configuration
including a chiral time-reversal breaking superconductors in (2 + 1)-dimensions was discussed
in [42], where the ansatz for px + ipy superconductors [43] was implemented. Such kind of

–2–
 IA
I

IB

JCAP01(2021)063
ξ
Figure 1. The schematic view of the condensed matter superconducting quantum interference device
consisting of two Josephson junctions (dark-blue) embedded into the superconducting loop (yellow-
bright). The current I flowing across the system is split into IA in the upper and IB in the lower
branch of the device. The presence of the magnetic flux Φ induces the interference in the system. For
the holographic setup see text for details.

superconductor is believed to support topologically protected gapless Majorana-Weyl edge

modes. On the other hand, the model with two coupled vector fields, was also implemented
in a generalisation of p-wave superconductivity, for the holographic model of ferromagnetic
superconductivity [44].
The construction of a Josephson junction in non-relativistic case with a Lifshitz ge-
ometry as the dual gravity was discussed in [45], where the effect of the Lifshitz scaling
was elaborated. Among all the standard sinusoidal relation between the current and phase
difference was revealed for various values of scaling. The relation featuring exponential de-
creasing between condensate operator and the width of the weak link, as well as, the relation
connected the critical current with the width were found.
Holographic s-wave and p-wave Josephson junctions with backreaction were studied
in [46]. It turns out that the critical temperature of the considered junctions decreases with
the increase of backreaction, while the tunnelling current and condensation also decrease
with the growth of backreaction. However, the relation between current and phase difference
remains in the form of a sine-function. On the other hand, the problems of holographic
models of hybrid and coexisting both types of junctions, i.e., s- and p-wave were analysed
in [47].
The studies of S-N-S junction in massive gravity theory unveil that the graviton mass pa-
rameter make it difficult for the normal metal-superconductor phase transition to occur [48].
Moreover the mass graviton parameter increase will cause the decrease of the maximal tun-
nelling current.
In recent years there has been also an interest in building holographic model of super-
conducting quantum interference device (SQUID) consisting of two appropriately connected
Josephson junctions (see figure 1 for the condensed matter model). Consequently, in order
to construct the holographic analog of SQUID one considers Einstein-Maxwell gravity with
complex scalar field in (3 + 1)-dimensional AdS Schwarzschild black brane background. One
of the spatial dimensions is compactified into a circle and with the properly chosen profiles of
the chemical potential constitute the main building blocks of the holographic SQUID [49, 50].
Taking our main goal into account we interpret the relation I = IA + IB for the total
current in the SQUID, with the currents flowing across each of the junctions given by the

–3–
relation (1.1), in the slightly different manner. Namely suppose, that the phase difference,
∆φA (or ∆φB ), in one of the branches of the considered device has changed as a result of
external factors. In the model under inspection we understand this phenomenon as a hidden
sector particle appearing in one of the Josephson junctions of the SQUID. This means that
the current phase relation changes in the same way as if the effective magnetic flux would
appear. The large sensitivity of the device, rises the hope of detecting such transition of the
dark matter particle, under the assumption of the validity of the coupling between hidden
and visible sectors. In this paper we shall suppose that a single dark matter particle appears
in one SQUID branch at a time, leaving aside the issue of possible fluctuations of the SQUID
signal due to consecutive events. This problem will be studied in future publications.
The main motivation standing behind our consideration is to reinvestigate the holo-

JCAP01(2021)063
graphic model of the Josephson junctions and the SQUID [39, 49, 50], in order to include
the hidden sector particles responsible for the dark matter fields. In conventional physics
the hidden sector should be added to the Standard Model. On the contrary, in our attitude
we shall build holographic model in which dark sector naturally emerges from the string/M
theory. In our research we shall consider the model in which the additional U(1)-gauge field
is coupled to the ordinary Maxwell one. The action describing Maxwell dark matter sector
is provided by [53, 54].
Z  
√ 4 1 1 α
Sdm+EM = −g d x − Fµν F µν − Bµν B µν − Fµν B µν , (1.2)
4 4 4
where Fµν = 2∇[µ Aν] stands for the ordinary Maxwell strength tensor, while the hidden
U(1)-gauge field Bµν is given by Bµν = 2∇[µ Bν] . α is a coupling constant between two
gauge fields. Predicted values of α-coupling constant, being the kinetic mixing parameter
between the two U(1)-gauge fields, for realistic string compactifications range between 10−2
and 10−16 [55]–[58].
The compatibility with the current observations authorises its order up to 10−3 . It
happens that some astrophysical phenomena like observations of 511 eV gamma rays [59],
experiments connected with the detection of the electron positron excess in galaxies [60,
61], as well as, possible explanation of muon anomalous magnetic moment [62], strongly
argue for the idea of dark sector coupled to the Maxwell one. Moreover, the kinetic mixing
term between ordinary boson and relatively light one (the dark one) arising from U(1)-
gauge symmetry bounded with a hidden sector, may be responsible for a low energy parity
violation [63]. On the other hand, the low energy gauge interaction in the hidden sector
may be envisaged by the Higgs boson decays, and a relatively light vector boson with mass
m ≥ 10 GeV emergence [64].
As was mentioned the justifications of such kind of models might be obtained from
the top-down perspective [65], starting from the string/M-theory. It is very important from
the AdS/CFT correspondence point of view, because string/M-theory constitutes a fully
quantum description and this guarantees that any phenomena envisaged by the top-down
reduction are physical.
The second term in the above action is connected with some hidden sector [65], while
the interaction of the visible and hidden sectors are described by the so-called kinetic mixing
term, for the first time introduced in [66], in order to describe the existence and subsequent
integrating out of heavy bi-fundamental fields charged under the U(1)-gauge groups. The
term in question is characteristic for theories having in addition to some visible gauge group
an additional one, in the hidden sector. The hidden sector describes states in the low-energy

–4–
effective theory, uncharged under the Standard Model gauge symmetry groups. On the
contrary they are charged under their own groups.
It is gravity which enables interaction between those two sectors [67, 68]. It turns out
that the realistic embeddings of the Standard Model in E8 × E8 string theory, as well as, in
type I, IIA, or IIB open string theory with branes, require the existence of the hidden sectors
for the consistency and supersymmetry breaking [56, 69]. On the other hand, in string
phenomenology [56] the dimensionless kinetic mixing term parameter can be produced at an
arbitrary high energy scale and it does not deteriorate from any kind of mass suppression
from the messenger introducing it. This fact is of a great importance from the experimental
point of view, due to the fact that its measurement can provide some interesting features of
high energy physics beyond the range of the contemporary colliders.

JCAP01(2021)063
Moreover the mixing term of two gauge sectors are typical for states for open string
theories, where both U(1)-gauge groups are advocated by D-branes that are separated in
extra dimensions. It happens in supersymmetric Type I, Type IIA, Type IIB models and
results in the existence of massive open strings which stretch between two D-branes [55].
The organization of the paper is as follows. In the next section we pay attention to the
holographic model of DC SQUID composed of two Josephson junctions in Einstein-Maxwell
gravity with complex scalar field affected by the dark sector particles modelled by the hidden
U(1)-gauge field interacting with the ordinary Maxwell one. One considers the model of
the junction, where a central part constitutes a normal metal (S-N-S). We also mentioned
the case with an insulator (S-I-S). Section 3 is devoted to the numerical solutions of the
adequate equations of motion, for the homogeneous case and the case of a dark matter beam.
We propose an ‘experiment’ which helps to distinguish the presence of dark matter sector.
In section 4 we concluded our investigations.

2 The holographic model and equations of motion

In this section we present the examined holographic model and the resulting equations of
motion. We shall examine the holographic DC SQUID, composed of the two S-N-S Josephson
junctions. We give some comments concerning the construction of holographic model of such
a device.

2.1 How to build holographic SQUID influenced by dark matter sector?

As the SQUID is composed of two Josephson junctions we shall concentrate on the problem
how to build holographic model of the S-N-S Josephson junction with the phase difference
across the normal region, say γ. In order to simulate it one has to implement the holographic
description of the superconductor, i.e., we should have the AdS background because of the
fact that we describe an approximately conformal field theory.
In the next step we want to describe the conserved U(1) currents Jµ in the field theory.
According to the AdS/CFT dictionary [22] in the gravity dual they correspond to gauge
fields. In our case these gauge fields will be bounded with ordinary Maxwell field and the
auxiliary one, pertaining to the hidden sector.
The superconducting state requires spontaneous symmetry breaking,1 which on the
gravity side corresponds to the condensation of the (complex) scalar field. The operator for
which vacuum expectation value is non-zero will correspond to the scalar field in the bulk.
1
The global symmetry of the boundary field theory is the conservation of the charge, which is precisely
broken below the superconducting transition temperature. On the gravity side this is marked as an instability
of the black hole which becomes hairy.

–5–
 In order to incorporate temperature in the theory we take into account the black brane
with Hawking temperature. On the other hand, to have TC 6= 0, one incorporates the scale to
generate it, which can be done by either introducing chemical potential µ or a non-zero charge
density. In the AdS/CFT dictionary µ, on the field theory side, is correlated with the vacuum
expectation value of the time component of gauge field. Thus if we take a non-zero chemical
potential to generate superconductivity, one achieves a critical temperature proportional to
it. All these points lead the holographic model of Josephson junction in which one should
also have varying chemical potential µ(~x) at fixed temperature (we have black brane at fixed
Hawking temperature) such that in most of the space temperature T is below TC connected
with µ (the system in question is in a superconducting state) but in a narrow gap is above
it, and the system is in a normal state.

JCAP01(2021)063
For technical reasons we introduce gauge invariant fields described by Mµ = Aµ − ∂µ φ,
which have the adequate asymptotics at the boundary of AdS spacetime, i.e. for r → ∞
proportional to chemical potential, charge density, superfluid velocity and current. Because
of the fact that the ansatz for the scalar field has a quantum phase φ and Mx = −∂x ψ + Ax ,
the gauge invariant phase difference across the normal gap is given by
Z Z
γ=− dxMx = ∆φ − dxAx . (2.1)
gap gap

It concludes, that from the boundary at infinity one obtains the following relation:
Z +∞
γ=− dx [ν(x) − ν(∞)] . (2.2)
−∞

Moreover, at the black brane event horizon we need to have the regularity conditions
fulfilled, i.e., Mt (r = r0 ) = 0. One also imposes that the junction is unbiased so µ(∞) =
µ(−∞) meaning that near the boundary µ approaches x-independent value.
Finally, in order to construct the holographic model of Josephson junction one demands
the adequate profile for the chemical potential, for which we ought to obtain that TC ∼ µ(∞)
with THawking < TC , and also we demand that µ is much smaller in the junction leading to
TC < THawking there. All these conditions ensure that we arrive at superconductor for most
of the range of ~x, only in the thin region we shall get a normal state.
To conclude, the philosophy underlying our considerations is to incorporate visible and
hidden sectors into AdS spacetime, apply the AdS/CFT machinery, and look if there are some
implications of the auxiliary U(1)-gauge field at r → ∞, at the boundary of the spacetime.

2.2 Equations of motion for the model

The holographic model of the 2+1 dimensional system in question relies on the 3+1 dimen-
sional gravitation background of AdS static Schwarzschild black brane spacetime. One will
work in the probe limit, which means that matter field do not cause the fluctuations of the
gravitational one, we have no backreaction effects. In the probe limit approximation the
dynamics of the gravitational part of the considered model will decouple from the rest, there-
fore one can elaborate gauge field perturbations on the fixed black brane background which
is therefore uncharged.
The gravitational action in (3+1) dimensions is taken in the form
Z
√
Sg = −g d4 x (R − Λ) , (2.3)

–6–
where Λ = −6/L2 stands for the negative cosmological constant, while L is the radius of the
AdS spacetime. In what follows we set L = 1.
We shall examine the Abelian-Higgs sector coupled to the hidden U(1)-gauge field,
coupled to the visible one by the kinetic mixing term, with a coupling α
Z 
√ 1
Sdm+EM+H = −g d4 x − Fµν F µν − [∇µ ψ − i Aµ ψ]† [∇µ ψ − i Aµ ψ] − m2 |ψ|2
4

1 α
µν
− Bµν B − Fµν B µν
. (2.4)
4 4

In many analysis of the hidden sector in holographic approach [24]–[34], α is taken as a

JCAP01(2021)063
constant. In [44], where the generalisation of the considered model to SU(2) gauge fields
was studied in the context of holographic model of ferromagnetic superconductivity, α was
incorporated in the redefined gauge fields being the mixture of visible and hidden sectors
components. In our attitude we are interested in the influence of the dark sector on the
physics and therefore we shall treat both sectors and kinetic mixing term between them.
However, as it will be revealed in section 3.1 of our work, such assumption about the
homogeneity of α-coupling constant gives no response on the considered holographic model
of SQUID.
In order to make the investigations more realistic, and find the influence of dark sector
on the physics of the holographic device, we suppose that instead of the coupling in the kinetic
mixing term we shall take into account the strength parameter function ζ, which controls
the strength of visible and hidden sectors interaction. Having in mind that our line element
is static, i.e., that one has time-like Killing vector field orthogonal to the three-dimensional
space-like hypersurface, our strength parameter function will depend on spatial coordinates.
In our considerations we shall elaborate the problem of the response of a holographic
superconducting junction/SQUID to the appearance of dark matter particle. Due to the
geometric constraints (see for the details [49, 50]), the nontrivial response of the holographic
(2+1) dimensional SQUID is expected to appear only if one of its junctions is affected by a
dark matter particle.
As it was already mentioned, to mimic the passage of the dark particle via one of the
junctions of the holographic SQUID we assume that the coupling ζ depends on space-like
coordinates. This is taken into account in the derivation of the underlying equations of
motion. Later on we specify the particular dependence ζ(xa ) used to solve the model.
The equations of motion are provided by

(∇µ − i Aµ ) (∇µ − i Aµ ) ψ − m2 ψ = 0, (2.5)

1 h i
∇µ F µν + ∇µ (ζ B µν ) − i ψ † (∇ν − i Aν ) ψ − ψ (∇ν + i Aν ) ψ † = 0, (2.6)
2
1
∇µ B µν + ∇µ (ζ F µν ) = 0. (2.7)
2
Consequently, using the last two equations one has that
 
1 1 h i
ζ̃ ∇µ F µν
+ ∇µ ζ B µν − αF µν − i ψ † (∇ν − i Aν ) ψ − ψ (∇ν + i Aν ) ψ † = 0, (2.8)
2 2

where we set ζ̃ = 1 − ζ 2 /4.

–7–
 As we have mentioned above, as a gravitational background we shall consider static
four-dimensional AdS-Schwarzschild black brane line element, given by
1
ds2 = −f (r)dt2 + dr2 + r2 (dξ 2 + dx2 ), (2.9)
f (r)
where f (r) = r2 − r03 /r and r0 is the horizon radius of the black brane, while the direction ξ
is compactified with the periodicity −π R ≤ ξ ≤ π R, where R is the radius of the ξ-loop.
For further analysis it is convenient to assume that r0 = 1 and π R = 10. According
to the claim in [39], the coherence length for the considered model is estimated to be about
∼ 1.20 within these units. We argue that the length of the SQUID loop should be much
larger than the coherence length, in order to decrease the influence of the proximity effect

JCAP01(2021)063
and consequently the mixing of the phases. Fulfilling this condition, we obtain a homogeneous
superconducting phase at the both ends of the ξ-direction. It results in a well defined phase
difference for the junction under inspection. On the other hand, one wants to establish our
weak link of the junction to be narrow enough, that its critical current receives a sufficiently
high value. This is important from the point of view of the numerics, i.e., the precision of the
solutions. Having all the above arguments in mind, we establish R = 10/π as a reasonable
choice.
In what follows one chooses the following components of the fields in the underlying
theory:
ψ = |ψ|eiφ , Aµ = (At , Ar , Aξ , 0), Bµ = (Bt , Br , Bξ , 0), (2.10)
where we assume that all the field components and the phase are real functions dependent
on (r, ξ)-coordinates. Moreover, one defines the gauge invariant quantities Mµ = Aµ − ∂µ φ.
Consequently, the equations of motion yield
  
1 2 ∂r f 1 1
∂r2 |ψ| + ∂ξ2 |ψ| + + ∂r |ψ| + M 2 − 2 Mξ (2.11)
r2 f r f f2 t r f
#
m2
− Mr2 − |ψ| = 0,
f
   
1 2 Mξ 2 ∂r f
∂r M r + 2 ∂ξ M ξ + Mr ∂r |ψ| + 2 ∂ξ |ψ| + + Mr = 0, (2.12)
r f |ψ| r f r f
 
2 1 2
ζ̃ ∂r2 Mt + ∂r Mt + 2 ∂ξ2 Mt − Mt |ψ|2 (2.13)
r r f f
 
1 ζ
+ ∂r ζ (∂t Br − ∂r Bt ) − (∂t Mr − ∂r Mt )
2 2
 
1 1 ζ 1
+ ∂ξ ζ − 2 (∂ξ Bt − ∂t Bξ ) + (∂ M − ∂ M ) = 0,
2 2 r2 f
ξ t t ξ
r f

h i 1
ζ̃ ∂ξ2 Mr − ∂ξ ∂r Mξ − 2 r2 Mr |ψ|2 + ∂ξ ζ (∂ξ Br − ∂r Bξ ) (2.14)
2

ζ
− (∂ξ Mr − ∂r Mξ ) = 0,
2
 
2 ∂r f 2 Mξ
ζ̃ ∂r Mξ − ∂r ∂ξ Mr + (∂r Mξ − ∂ξ Mr ) − |ψ|2 (2.15)
f f
 
1 f ζ f
+ ∂r ζ 2 (∂r Bξ − ∂ξ Br ) − (∂r Mξ − ∂ξ Mr ) = 0.
2 r 2 r2

–8–
The components of the relation (2.7) imply
2 1 1 1 1
∂r Bt + ∂r2 Bt + 2 ∂ξ2 Bt + ∂r ζ ∂r Mt + ∂ξ ζ 2 ∂ξ Mt (2.16)
r r f 2 2 r f
 
ζ 2 1
+ ∂r Mt + ∂r2 Mt + 2 ∂ξ2 Mt = 0,
2 r r f
∂r f 1
∂r2 Bξ − ∂r ∂ξ Br + (∂r Bξ − ∂ξ Br ) + ∂r ζ (∂r Mξ − ∂ξ Mr ) (2.17)
f 2
 
ζ ∂r f
+ ∂r2 Mξ − ∂r ∂ξ Mr + (∂r Mξ − ∂ξ Mr ) = 0,
2 f

JCAP01(2021)063
1 ζ
∂ξ (∂ξ Br − ∂r Bξ ) + ∂ξ ζ (∂ξ Mr − ∂r Mξ ) + ∂ξ (∂ξ Mr − ∂r Mξ ) = 0. (2.18)
2 2
To proceed further, we ought to choose the form of ζ(xi ). One of the possible represen-
tations of it is a Gaussian function form
2 /λ2
ζ(ξ) = α0 e−(ξ−ξ0 ) , (2.19)

where α0 is the peak value of the coupling and λ its decay length (half of the Gaussian width).
Our choice of the strength function controlling interaction of both sectors in question ζ(ξ),
was motivated by the features of static line element chosen for the description of holographic
device, given by (2.9), as well as, the astrophysical observations revealing the space-like
dependence of dark matter distribution in the observable Universe.
The half of the Gaussian width λ is an important parameter of the model as its ratio to
the holographic SQUID characteristic length-scale πR and the shape of the weak link affect
the holographic SQUID’s response.
In principle, there are two possible attitudes for modelling the local presence of the
dark matter sector fields in this theory. We can either pick α to be constant and impose
specific boundary conditions for dark matter fields or as it was mentioned before, promote
the strength function ζ with a spatial dependence and keep the fields in the homogeneous
scenario. In this approach we selected the latter one, so we can choose the x-independent
and source free solution of Bµ fields, which simply yields
 
r0
Bα dx = µD α
1− dt. (2.20)
r
By putting this assumption into our system of differential equations we receive a significant
simplification, with only four functions to be obtained numerically. In particular the resulting
equations depend on the dark sector only via the effective coupling ζ̃ and not µD .
  " #
1 2 ∂r f 1 1 m2
∂r2 |ψ| + ∂ξ2 |ψ| + + ∂r |ψ| + 2 Mt2 − 2 Mξ2 − Mr2 − |ψ| = 0, (2.21)
r2 f r f f r f f
 
2 1 2 1 ζ 1
ζ̃ ∂r2 Mt + ∂r Mt + 2 ∂ξ2 Mt − Mt |ψ|2 + ∂ξ ζ 2 ∂ξ Mt = 0, (2.22)
r r f f 2 2r f
h i 1 ζ
ζ̃ ∂ξ2 Mr − ∂ξ ∂r Mξ − 2 r2 Mr |ψ|2 − ∂ξ ζ (∂ξ Mr − ∂r Mξ ) = 0, (2.23)
2 2
 
∂ f 2 Mξ
ζ̃ ∂r2 Mξ − ∂r ∂ξ Mr + (∂r Mξ − ∂ξ Mr ) − |ψ|2 = 0. (2.24)
r
f f

–9–
The above forms of the equations envisage that comparing to the case when ζ = 0, the
modification is significant. Some terms are multiplied by ζ̃ and moreover we have additional
terms with ζ∂ξ ζ factor. Numerical solutions of these equations will be presented in the
subsequent part of the paper.
As we have remarked at the beginning of the section, the presented model describes
SQUID with S-N-S type Josephson junctions. Nevertheless, for the completeness of the
results, we shall mention the modeling of S-I-S Josephson junction. In the holographic model
of the insulator the AdS soliton line element will play the crucial role. Namely, performing
the double Wick rotation on the metric (2.9), we arrive at the AdS soliton line element [70].
It is provided by
dr2

JCAP01(2021)063
ds2 = −r2 dt2 + + r2 dx2 + f (r)dχ2 , (2.25)
f (r)
with the same function f (r) as before. However, the coordinate χ has the period β = 4πl/3r0 ,
in order to avoid the conical singularity at r = r0 .
As in the previous case we assume that the following components of the gauge field
are non-zero Aµ = (At , Ar , Ax , 0), and that all the field components and the phase φ are
real functions dependent on (r, x)-coordinates. The equations of motion are very similar to
the ones obtained in the previous case. In fact only the factor in front of ∂r Mt instead of
2
r reads f . The results will differ only quantitatively, therefore they will be not analysed
∂r f

numerically.
Having discussed the model set up we shall describe appropriate boundary conditions
for our model. Starting with AdS boundary where r → ∞, the asymptotic of the fields in
question are provided by

|ψ (1) (ξ)| |ψ (2) (ξ)|

|ψ| = + + O(r−3 ), (2.26)
r∆(1) r∆(2)
ρ(ξ)
Mt = µ(ξ) − + O(r−2 ), (2.27)
r
Mr = O(r−3 ), (2.28)
J(ξ)
Mξ = ν(ξ) + + O(r−2 ), (2.29)
r
where
1 p 1 p
∆(1) = (3 − 9 + 4m2 ), ∆(2) = (3 + 9 + 4m2 ). (2.30)
2 2
In the boundary field theory the quantities µ(ξ), ρ(ξ), ν(ξ), and J(ξ) are connected with the
chemical potential, charge density, superfluid velocity and current, respectively. On the other
hand, |ψ (1) (ξ)| and |ψ (2) (ξ)| may be considered as the source and the vacuum expectation
value of the dual operator of the scalar field |ψ(ξ)(2) | = hOi. In the numerical calculation we
set |ψ (1) (ξ)| = 0 (as turning off the source), because one has that the U(1)-gauge symmetry
be broken. At the black brane event horizon we demand that g tt Mt ought to be regular,
but because of the fact that g tt is divergent on the horizon, Mt should vanish at r = r0 .
The remaining fields should be finite at the event horizon and the simplified equations of
motion place the boundary conditions therein. In our numerical calculations for simplicity
we set m2 = −2 which is above Breitenlohner-Freedman limit m2 ≥ − 49 , therefore ∆(1) = 1
and ∆(2) = 2.

– 10 –
3 Numerical results

This section will be devoted to the results of our numerical solutions of the equations of
motion for the holographic Josephson junction and holographic DC SQUID device. At first
we pay attention to the case of homogeneous holographic superconductor influenced by the
dark matter sector. Then we proceed to analyse the model of SQUID with a beam of dark
matter, i.e., we take into account the strength parameter controlling the strength of visible
and hidden sectors interaction and allow for its ξ dependence.

3.1 Homogeneous superconductor — warm up

To commence with, let us consider the properties of homogeneous holographic superconduc-

JCAP01(2021)063
tor, i.e., we take α as a constant value, influenced by the dark matter sector, and look for
its critical temperature, critical chemical potential and value of the condensation. For this
calculation we straight out the ξ dimension renaming it to x. Then by neglecting the spatial
dependence of x-coordinate in the equations of motion, we receive the simplified system of
differential equations
  " #
2 ∂r f 1 1 m2
∂r2 |ψ| + + ∂r |ψ| + 2 Mt2 − 2 Mx − |ψ| = 0, (3.1)
r f f r f f
2 2
∂r2 Mt + ∂r Mt − Mt |ψ|2 = 0, (3.2)
r α̃ f
∂r f 2 Mx
∂r2 Mx + ∂r Mx − |ψ|2 = 0. (3.3)
f α̃ f
2
where α̃ = 1 − α4 and all the above three functions, i.e., |ψ|, Mt , Mx , depend only on r.
These equations establish the boundary conditions for a homogeneous region of su-
perconductor in x-dependent approach, but for the analysis of phase transitions we set the
current to zero, which implies that Mx = 0. Now we can move towards numerical solu-
tion of our equations. In order to proceed further, we perform coordinate transformation
z = 1 − 1r in order to work on a (0, 1) grid, where z = 0 is the black hole horizon and
z = 1 is the holographic boundary. We solve the boundary value problem using the standard
relaxation method for different values of temperature and chemical potential, to investigate
the superconductor-normal phase transition. In figures 2 and 3 we plot the condensation
operator hOi as a function of temperature and the chemical potential for the different values
of dark matter α-coupling constants.
One concludes that the bigger T /Tc is the smaller value of hOi/Tc we obtain. In the
case when we fix the value of T /Tc , then the bigger α-coupling constant is taken into account
the smaller hOi/Tc we get. Just, the condensation diminishes when dark matter coupling
constant grows. On the other hand, the bigger µ one considers, the larger value of hOi we
receive, for the fixed value of α-coupling constant. In the case when α grows, we deduce that,
the larger α is, the smaller value of the condensation operator one perceives.
In the both types of the analysed phase transitions, the presence of dark matter does
not shift the phase transition point (critical temperature, critical chemical potential), yet it
changes the value of condensation.
Figure 4 depicts the relationship between the condensation operator and dark matter
coupling constant, for the fixed temperature and chemical potential. The differences are not
too big, but they are visible and the lines are distinguishable. However there might be a

– 11 –
 16 3.0

14
2.5
12

10 2.0
hOi/Tc

hOi
8
1.5
6 α
0.00 1.0
4 0.10
0.25
2 0.5
0.50
0

JCAP01(2021)063
0.5 0.6 0.7 0.8 0.9 1.0 0.0
0.340 0.345 0.350 0.355 0.360 0.365 0.370 0.375 0.380
T /Tc T

Figure 2. The superconductor-normal phase transition driven by the temperature, affected by dark
matter. The right panel shows the zoom of the area of left plot which is close to the phase transition
and without temperature normalisation. The coupling with the dark sector causes the increase of
critical temperature.

4.0

3.5

3.0

2.5
hOi

2.0 α
0.00
1.5 0.10
0.25
1.0
0.50
0.5

0.0
3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2
µ

Figure 3. The normal to superconducting phase transition driven by the change of the chemical
potential (which can be interpreted as doping) for different values of dark matter α-coupling constants.
The presence of the dark sector coupling causes the phase transition for smaller value of the chemical
potential.

problem with measuring such a deviation. Our idea to overcome this obstacle is presented
in the next section — by using superconducting devices.
The calculations of this section served as a test of our numerical procedure. The proce-
dure passed the test as the closer inspection of the equations of motion (3.1)–(3.3) shows that
the parameter α̃ can be eliminated by rescaling the ψ field, as ψ̃ = √ψα̃ , with ψ̃ being indepen-
dent on α. This rescaling leads to the similar rescaling of the condensation parameter √ hOi,
leading to the solution with α independent value hOi0 . The expected behavior hOi = α̃hOi0
is clearly visible in figure 4 as are the concomitant changes of other parameters depending
on hOi.

3.2 Analysis of the holographic SQUID

In this subsection we shall treat the case when one has the strength function ζ controlling
the interaction between visible and hidden sectors.

– 12 –
 6.35

6.30

hOi
6.25

6.20

0.0 0.1 0.2 0.3 0.4 0.5

JCAP01(2021)063
α

Figure 4. Value of the condensation operator for the constant value of the chemical potential µ = 6
and temperature T = 3/4π versus the dark matter coupling constant.

5.0

4.8

4.6

4.4
µ

4.2

4.0

3.8

3.6
−10.0 −7.5 −5.0 −2.5 0.0 2.5 5.0 7.5 10.0
ξ

Figure 5. The schematic profile of the chemical potential in the superconducting quantum interfer-
ence device consisting of two holographic Josephson junctions (lower values of µ) embedded into the
superconducting loop with µ = 5.

To imitate the two insulating regions embedded into the superconducting ring we take
the chemical potential as given in [49]
    
X ki (ξ − pi + wi ) ki (ξ − pi − wi )
µ(ξ) = h − di tanh − tanh , (3.4)
i=1, 2
π π

where i = 1, 2 stand for two junctions in the SQUID ring, and h, di , ki , pi , and wi are
related to the highest value of the chemical potential (inside the superconductor), depth,
slope, position, and width of the junction i, respectively. The idea behind this choice is the
following: from the previous studies we know that with and/or without dark matter it is
the value of the chemical potential which tunes the transition between insulator and metal
and the temperature decrease induces metal — superconductor transition. To construct a
junction it is thus enough to make the chemical potential in the normal part of the junction
lower than its critical value for superconducting transition, while leaving it above this value
in the superconducting parts of the junction. This is just embodied in the equation (3.4).
And shown in figure 5, for the weak links located at p1 = 5 and p2 = −5 with h = 5 and
identical other parameters d1 = d2 = 0.7 widths w1 = w2 = 1.6 and slopes k1 = k2 = 7.

– 13 –
 In the discussion of the holographic SQUID we change γL and γR , defined in what
follows. In our model we assume that the dark matter is present only in one of the junctions
of SQUID. Such approach gives us a possibility to calculate the difference of the phases
between junctions caused by the dark matter coupling. In the trivial case (without the
presence of dark matter sector) the effective magnetic flux defined as Φ = γL − γR vanishes,
therefore the critical current
q
Jc = JL2 + JR2 + 2JL JR cos(Φ), (3.5)
becomes constant and equals to |JL + JR | On the contrary, when dark sector particle is
present in one of the junctions, we receive a “yes or no” type criteria for the dark matter
occurrence.

JCAP01(2021)063
In adverse to the previously obtained results [49], we do not impose the continuity
conditions on our fields in the node points ξ = 0 and ξ = ±10. Moreover we argue that these
points can be singular due to the presence of the additional source terms. Especially the
supercurrent might not be well behaved when it inflows and outflows from our system. If we
wish to model the SQUID with two currents flowing parallel, we have to take J ∼ sgn(ξ),
because current flows into the system at ξ = 0, then in one branch it flows to ξ = −10 and
in the another to ξ = 10, against and with the ξ axis, respectively.
As far as the boundary value problem is concerned, for the AdS/CFT boundary one
has the asymptotic expansions, for the black hole horizon we require that Mt = 0, while the
conditions for the remaining functions are given by the adequate equation of motion. On the
other hand, for the ξ boundary, in both cases, we impose that the functions |ψ|, Mt and Mx
have to be even and Mr is an odd function with respect to the central point of the junction.
In our numerical computations we implement the pseudo-spectral method with Cheby-
shev points, which constitutes an extremely efficient tool for smooth functions we deal with in
the problem. For convenience we compactify the r-coordinate using following transformation
z = 1 − 1/r, after which z = 0 corresponds to the black hole horizon and z = 1 to the AdS
boundary. Also we squeeze ξ dimension into (−1, 1) range which is the domain of the Cheby-
shev polynomials. Generally pseudo-spectral methods require only few points in the spatial
discretisation to achieve satisfying results, therefore we use the grid with 37 points along ξ
direction and 20 points along z axis. By expanding the functions into Chebyshev series we
translate a differential problem stated by equations of motion into an algebraic one. Then we
deal with a system of nonlinear algebraic equations by standard Newton-Raphson method. In
the explained way we solve numerically our set of partial differential equations (2.21)–(2.24)
for different values of J0 , varying from −0.06 to 0.06. In figure 6 we plot the exemplary
solutions of the underlying equations of motion for J = 0.03 and α0 = 0.2. We can see the
imposed shape of the chemical potential at the boundary z = 1 of the Mt function. The
remaining functions react accordingly to the presence of the normal phase. The shapes of
the solution are similar to the ones obtained in [39, 41, 49, 50]. However it contains the
traces of the influence of α coupling, but they are too small to be visible on such 3D plots.
Nevertheless they do have the influence on some of the properties, which will be elaborated
below.
After solving our differential equations we obtain the phase differences for the left and
right Josephson junctions separately, using the following formula:
Z 0
γL = − [ν(ξ) − ν(0)] dξ, (3.6)
−10
Z 10
γR = − [ν(ξ) − ν(0)] dξ. (3.7)
0

– 14 –
 JCAP01(2021)063
Figure 6. Plots of numerical solutions of equations of motion (2.21)–(2.24) for one of holographic
SQUID’s Josephson junction with parameters J = 0.03 and α0 = 0.2. We can see the imposed
chemical potential boundary condition for Mt at the AdS boundary and the reaction of the remaining
functions to it.

Next, for each of the Josephson junction we fit in a sine relation between JL/R and γL/R , to
obtain the value of the maximum current.
Having obtained the values of the critical currents for each junction we can proceed to
calculation of the dark matter particle induced effective magnetic flux Φ, as defined previously.
We wish to see how the critical current of holographic SQUID, defined by the relation (3.5),
changes with the presence of dark matter in the system, hence one elaborates the case when
ζ 6= 0. To proceed further let us define the critical current ratio, given by
Jc (Φ) − Jc (0)
δJ = . (3.8)
Jc (0)
In figure 7 the critical current ratio as a function of the magnetic flux, for different values of
α0 , is depicted. In our model such behaviour is only possible when dark particle appears in
one of the junctions. Otherwise, if Jc is independent of Φ the interference does not occur.
This gives us the possibility to establish the criterion for detection of the discussed coupling
which is correlated with the dark sector.
While the residuals of numerical solution of our equations (2.21)–(2.24) and the con-
straint relation (2.12) are relaxed to the values less than 10−12 , the relation from the figure 7
carries a burden of inaccuracy of the least squares method. Namely the sine function is not
a perfectly fitting one and the best result we can obtain for the norm of residuals is to the
order of 10−6 . It results that the uncertainty of the critical current is ∼ 10−4 . Having this in
mind, for little and yet realistic values of coupling our computed values of J(Φ) are below the
level of accuracy, so we cannot state its direct value. Although when working on differences,

– 15 –
 ×10−5
0.0

−0.2

−0.4

−0.6
δJ

−0.8
α
0.5

JCAP01(2021)063
−1.0
0.05

−1.2
−1.00 −0.75 −0.50 −0.25 0.00 0.25 0.50 0.75 1.00
Φ [10−2]

Figure 7. Relative change of critical current ratio from the relation (3.8) versus the effective magnetic
flux, which appears to be non-zero in the presence of dark matter. The yellow curve represents solution
when α0 = 0.5, while the blue one α0 = 0.05. In the case with α = 0 such a dependence would not
occur, resulting a trivial SQUID with two identical Josephson junctions and δJ ≡ 0.

like in our ratio given by the relation (3.8), these uncertainties are subtracted, so using this
quantity seems sensible.
We also investigated the influence of increasing the Gaussian packet decay length λ
on the current-phase relation of one of the holographic SQUID’s Josephson junctions. By
implementing to our code various values of λ (from 1 to 5), we have found out an interesting
behaviour which is depicted in figure 8. It envisages the critical Josephson current as a
function of the wavelength of the decay length λ. This plot might be interpreted as the
sensitivity of the holographic SQUID on detecting dark matter sector. On the other hand,
it is a very interesting result and may constitute a guidance for future experiments. Namely,
by adjusting the size of the normal/insulating part of the holographic Josephson junctions
one can tune up the system to react most sensitively on possible detection of dark particle.
The obtained value of the optimal λ ≈ 1.5 is correlated, as it is expected to be, with the
width w of the normal region of the junction affected by the hidden particles.
Having a possible range of the Gaussian packet decay length λ, one receives a possibility
to prepare an adequate detector for measurements. In the present approach we are assuming
that the dark matter particle passes the center of the junction. However, one expects that
dark particles move slightly off the center. In such situation the sensitivity can be roughly
read of from the curve shown in figure 8, as the decrease/increase of λ, mimicking the distance
from the center of the holographic junction.

4 Holographic versus condensed matter SQUID

Here we compare the holographic results with the expectations based on the standard theory
of laboratory SQUID as shown in figure 1. Each branch of the SQUID, in figure 1, contains

– 16 –
 6.60

6.58

Jc [10−2]
6.56

6.54

6.52

JCAP01(2021)063
6.50
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
λ
Figure 8. The critical current of dark matter influenced Josephson junction as a function of Gaussian
broadening λ from the relation (2.19). It is clearly seen that there is a maximal sensitivity for some
value of λ. In this plot we used α0 = 0.5, the size of the junctions is 1.6, while the separation between
the junctions is 10, in the units used in the plot.

one Josephson junction. In general, if the device is biased the current phase relation reads
 
2 eV
I = Imax sin ∆φ + t , (4.1)
~
where V = (µ(∞)−µ(−∞))/e is the voltage drop across the junction, ∆φ is earlier introduced
phase difference between two superconductors. Equation (4.1) shows that the current of
the biased junction oscillates with frequency 2 ~eV . In the following, in full analogy to the
holographic treatment we shall consider only direct current (dc) SQUID i.e. with V = 0.
The presence of the magnetic flux perpendicular to the plane of the device induces the
screening currents in the superconductor and changes the phase relations in both branches.
By the gauge invariance principle one finds that the flux Φ penetrating the SQUID loop
modifies the relation between the phases of two junctions A and B and one gets [51]
Φ = 2πφ0 (∆φA − ∆φB ). (4.2)
The total current I entering the device from the left splits into IA and IB . At the right hand
side of the device both IA and IB fulfilling the relation (1.1) add to give again the current I.
One thus gets (for IA = IB and with the phase differences ∆φA and ∆φB in the respective
branches)    
∆φA + ∆φB ∆φA − ∆φB
I = IA + IB = 2Imax sin cos . (4.3)
2 2
As was shown in [52] the phase difference is related to the total flux φtotal , which for the
negligible inductance of the loop reduces to the external flux Φ as in the equation (4.2). With
the symmetric distribution of the phase modifications
πΦ
∆φA = ∆φ + (4.4)
φ0
πΦ
∆φB = ∆φ − (4.5)
φ0

– 17 –
one gets      
πΦ πΦ πΦ
I = 2Imax cos sin ∆φB + = 2Imax sin(∆φ) cos , (4.6)
φ0 φ0 φ0
where φ0 ≡ hc/2e ≈ 2.07 · 10−7 Gs · cm2 is the quantum flux.
 One sees that the amplitude of
the current in the SQUID is modulated by the factor cos πΦ φ0 and the ratio of the maximal
currents flowing in the device subject to the external flux to that without it is given by this
factor  
Im (Φ) πΦ
= cos . (4.7)
Im (0) φ0
For the completeness it has to be noted that if the magnetic flux Φ is piercing one of

JCAP01(2021)063
the weak links (junctions), then another “diffraction pattern” is observed [36]. It may be
characterised by the ratio of the maximum current in the device with (Im (Φ)) and without
(Im (0)) extra flux
 
Im (Φ) sin πΦ
φ0
= . (4.8)
Im (0) πΦ
φ0

The above two possibilities to change the maximum current in the device formulated by
equations (4.7) and (4.8) by applying to it the external magnetic field are plotted in the
upper panel of the figure 9. The blue curve on this panel shows the current as a function of
the magnetic flux ratio Φ/φ0 for the flux penetrating the extended junction and the magnetic
field parallel to the plane of the SQUID, while the other curve shows standard geometry with
the flux penetrating the loop.
Now we discuss the possible effect of ıdark matter particle on the current through the
condensed matter SQUID. For this we imagine the dark matter particle traversing one of its
junctions. In order to find the changes of the flux Φ induced by the passage of the dark
particle, let us consider the action (1.2), and define new gauge fields which enable us to get
rid of the kinetic mixing term. Namely one has
√
2−α
õ = (Aµ − Bµ ) , (4.9)
2
√
2+α
B̃µ = (Aµ + Bµ ) . (4.10)
2
It leads to
1 1 α 1 1
Fµν F µν + Bµν B µν + Fµν B µν =⇒ F̃µν F̃ µν + B̃µν B̃ µν , (4.11)
4 4 4 4 4
where F̃µν = 2∇(µ Ãν) and respectively B̃µν = 2∇(µ B̃ν) . The corresponding magnetic fluxes
are provided by
√  
2 − α 2π a 2π b
Φõ = − , (4.12)
2 e ed
√  
2 + α 2π a 2π b
ΦB̃µ = + , (4.13)
2 e ed

where e stands for the charge of Maxwell field while ed is connected with the charge of dark
matter sector gauge field, while a, b ∈ Z.

– 18 –
 1.0

0.8

Im(Φ)/Im(0) 0.6

0.4

| sin(Φ/Φ 0)
Φ/Φ0 |

JCAP01(2021)063
0.2
| cos(Φ/Φ0 )|

0.0
−3 −2 −1 0 1 2 3
Φ/Φ0
0.030 ×10−9 + 1

0
0.025
−1
Without DM
0.020 With DM
Im(Φ)/Im(0)

Im(Φ)/Im(0)

−2

0.015
−3

0.010
−4
Without DM
0.005 With DM
−5

0.000
−1.60 −1.59 −1.58 −1.57 −1.56 −1.55 −1.54 −1.00 −0.75 −0.50 −0.25 0.00 0.25 0.50 0.75 1.00
Φ/Φ0 Φ/Φ0[10−4]

Figure 9. The interference patterns in SQUID is shown in the upper panel. Left-lower panel shows
the detail SQUID responses for magnetic field tuned close to the minimum. The passage of dark
particle shifts the minimum and the SQUID responds with a well visible signal (lower left panel
α = 0.004). The detection of the signal is much more complicated if the SQUID is tuned close to the
maximum of the current (lower right panel). The differences are small even for very large values of
the coupling, which in this panel is assumed to be α = 0.2.

Thus the total magnetic flux encompassed by the SQUID is given as Φõ +ΦB̃µ . Having
in mind that the coupling constant α  1, one can rewrite the sum as follows:
√    
2 2π α e
Φõ + ΦB̃µ ' a+ b . (4.14)
e 4 ed
If we assume that a = b it leads to the relation
  
α e
Φõ + ΦB̃µ ' ΦA 1 + , (4.15)
4 ed
where ΦA is the magnetic flux bounded with Maxwell field.
On the basis of the equation (4.15) and assuming that eed = 1 we argue that due to the
kinetic mixing (with the coupling α) the effective magnetic field and thus the flux through
the device would change from the value Φ to the new value (1 + α/4)Φ when the dark particle
is traversing the SQUID, so obviously the current in the device would also change. This
change is illustrated in the lower panels of figure 9.

– 19 –
 The lower left panel of figure 9 compares the detected signals close to its minimum. The
blue curve is the signal without dark particle. The zero current is expected at Φ/φ0 = π/2. If
the dark particle appears in the SQUID it changes the flux across the device. This change of
the flux obtained for α = 0.004 is visible as a nonzero current at the same external magnetic
field value. In the figure both curves have been plotted as a function of the external flux
Φ (no dark particle in the SQUID), which value is changed (locally in space and time) to
(1 + α/4)Φ (when dark particle traverses the device).
On the contrary if one looks at the SQUID signal close to the maximum current (lower
right panel) the effect is hardly visible for the same value of α (not shown). In the panel
we are comparing two signals using α = 0.2. This large value of the coupling makes the
signal visible. However, even the tiny change of the effective flux causes the shift of the

JCAP01(2021)063
minimum in the diffraction curves and makes the tiny effect visible. Thus the response of
the SQUID which is tuned by the external B field to the minimum of the current seem
easier to be detected. This is due to the fact that in the latter case one expects the sudden
increase of the current from zero to some finite (albeit small) value. Unless the changes of
the current are below the thermal or quantum fluctuations in the system the effect should be
observable. This depends on the value of the kinetic mixing α and thus on the mass(es) of
the dark particle(s). To define real experimental conditions one should also take into account
fluctuations of the external flux. The field theory analysis of the possibility of detecting dark
matter particles is an important and complicated subject [12], which we shall not pursue here
any further.

5 Discussion and conclusion

Using the holographic approach we have studied the influence of the dark sector particle
on the holographic model of the SQUID. To this end we have assumed gravity background
consisting of the four-dimensional static AdS Schwarzschild black brane. The holographic
modelling of the device is achieved by use of the special form of the chemical potential
mimicking the two insulating/normal regions embedded into superconducting ring. Moreover
one assumes that dark matter sector is present only in one of the holographic Josephson
junctions of the SQUID under inspection.
This supposition enables us to calculate the difference of the phases between the junc-
tions, caused by dark matter particle. We solve numerically equations of motion and receive
values of Mi and thus the phase, for the specific choice of the current J and ζ strength
parameter function, controlling the interaction of visible and hidden sectors. This procedure
leads to the critical current and the phase for each of the examined holographic junctions
and allows establishing the current phase relation which has a typical sinusoidal character.
The passage of the dark sector particle by one of the holographic SQUID junctions
modifies the phase difference in that junction and leads to the modification of the critical
current. This shows up as the appearance of the dark sector induced effective flux Φ. The
difference of the critical currents of the SQUID measured for ζ 6= 0 and ζ = 0, signals the
dark sector field detection.
We have assumed that the presence in the junction of the dark matter sector field can
be described by the Gaussian type of dependence of the coupling ζ on the ξ coordinate.
It turns out that the increase of the Gaussian packet decay length λ has an effect on the
current-phase relation of one of the holographic SQUID’s Josephson junctions. Moreover, the
critical Josephson current is a function of the decay length λ. This fact supplies us the tool

– 20 –
for changing the sensitivity of the holographic SQUID in the dependence on the expected
properties dark matter sector. Interestingly the simplest possible assumption that due to the
kinetic mixing between the fields the dark particle traversing the condensed matter SQUID
increases the effective magnetic field leads to the results similar to those obtained by means
of holographic analogy.
To conclude, we remark that the presented results have been obtained for the holo-
graphic s-wave Josephson junction and SQUID composed of such junctions. They envisage
some features which can be utilised in future experiments aimed at detecting dark matter
sector particles. Of course, we are aware that examination of more complicated models like
p-wave or px + ipy might lead to even better detection methods. We shall investigate these
problems elsewhere.

JCAP01(2021)063
Acknowledgments
We thank J.E. Santos and R. Moderski for valuable comments concerning the numerical
methods. This work has been partially supported by the M. Curie-Sklodowska University
and the National Science Center grant DEC-2017/27/B/ST3/01911 (Poland).

References
[1] G. Bertone and T.M.P. Tait, A new era in the search for dark matter, Nature 562 (2018) 51
[arXiv:1810.01668] [INSPIRE].
[2] R. Bernabei et al., Searching for WIMPs by the annual modulation signature, Phys. Lett. B
424 (1998) 195 [INSPIRE].
[3] R. Bernabei et al., Final model independent result of DAMA/LIBRA-phase1, Eur. Phys. J. C
73 (2013) 2648 [arXiv:1308.5109] [INSPIRE].
[4] COSINE-100 collaboration, An experiment to search for dark-matter interactions using
sodium iodide detectors, Nature 564 (2018) 83 [Erratum ibid. 566 (2019) E2]
[arXiv:1906.01791] [INSPIRE].
[5] XENON collaboration, Excess electronic recoil events in XENON1T, Phys. Rev. D 102 (2020)
072004 [arXiv:2006.09721] [INSPIRE].
[6] B. Fornal, P. Sandick, J. Shu, M. Su and Y. Zhao, Boosted Dark Matter Interpretation of the
XENON1T Excess, Phys. Rev. Lett. 125 (2020) 161804 [arXiv:2006.11264] [INSPIRE].
[7] J. Bramante and N. Song, Electric But Not Eclectic: Thermal Relic Dark Matter for the
XENON1T Excess, Phys. Rev. Lett. 125 (2020) 161805 [arXiv:2006.14089] [INSPIRE].
[8] N.F. Bell, J.B. Dent, B. Dutta, S. Ghosh, J. Kumar and J.L. Newstead, Explaining the
XENON1T excess with Luminous Dark Matter, Phys. Rev. Lett. 125 (2020) 161803
[arXiv:2006.12461] [INSPIRE].
[9] A. Bally, S. Jana and A. Trautner, Neutrino self-interactions and XENON1T electron recoil
excess, Phys. Rev. Lett. 125 (2020) 161802 [arXiv:2006.11919] [INSPIRE].
[10] F. Takahashi, M. Yamada and W. Yin, XENON1T Excess from Anomaly-Free Axionlike Dark
Matter and Its Implications for Stellar Cooling Anomaly, Phys. Rev. Lett. 125 (2020) 161801
[arXiv:2006.10035] [INSPIRE].
[11] N. Masi, Dark matter: TeV-ish rather than miraculous, collisionless rather than dark, Eur.
Phys. J. Plus 130 (2015) 69 [INSPIRE].
[12] Y. Hochberg, T. Lin and K.M. Zurek, Detecting Ultralight Bosonic Dark Matter via Absorption
in Superconductors, Phys. Rev. D 94 (2016) 015019 [arXiv:1604.06800] [INSPIRE].

– 21 –
[13] S. Knapen, T. Lin and K.M. Zurek, Light Dark Matter in Superfluid Helium: Detection with
Multi-excitation Production, Phys. Rev. D 95 (2017) 056019 [arXiv:1611.06228] [INSPIRE].
[14] S. Knapen, T. Lin, M. Pyle and K.M. Zurek, Detection of Light Dark Matter With Optical
Phonons in Polar Materials, Phys. Lett. B 785 (2018) 386 [arXiv:1712.06598] [INSPIRE].
[15] R. Budnik, O. Chesnovsky, O. Slone and T. Volansky, Direct Detection of Light Dark Matter
and Solar Neutrinos via Color Center Production in Crystals, Phys. Lett. B 782 (2018) 242
[arXiv:1705.03016] [INSPIRE].
[16] Y. Hochberg et al., Detection of sub-MeV Dark Matter with Three-Dimensional Dirac
Materials, Phys. Rev. D 97 (2018) 015004 [arXiv:1708.08929] [INSPIRE].
[17] T. Liang, B. Zhu, R. Ding and T. Li, Direct Detection of Axion-Like Particles in

JCAP01(2021)063
Bismuth-Based Topological Insulators, Int. J. Mod. Phys. A 33 (2018) 1850135
[arXiv:1807.11757] [INSPIRE].
[18] M. Baryakhtar, J. Huang and R. Lasenby, Axion and hidden photon dark matter detection with
multilayer optical haloscopes, Phys. Rev. D 98 (2018) 035006 [arXiv:1803.11455] [INSPIRE].
[19] J.M. Maldacena, The Large N limit of superconformal field theories and supergravity, Int. J.
Theor. Phys. 38 (1999) 1113 [Adv. Theor. Math. Phys. 2 (1998) 231] [hep-th/9711200]
[INSPIRE].
[20] E. Witten, Anti-de Sitter space and holography, Adv. Theor. Math. Phys. 2 (1998) 253
[hep-th/9802150] [INSPIRE].
[21] S. Sachdev, What can gauge-gravity duality teach us about condensed matter physics?, Ann.
Rev. Condensed Matter Phys. 3 (2012) 9 [arXiv:1108.1197] [INSPIRE].
[22] J. Zaanen, Y.-W. Sun, Y. Liu and K. Schalm, Holographic Duality in Condensed Matter
Physics, Cambridge University Press, Cambridge U.K. (2015) [INSPIRE].
[23] R.-G. Cai, L. Li, L.-F. Li and R.-Q. Yang, Introduction to Holographic Superconductor Models,
Sci. China Phys. Mech. Astron. 58 (2015) 060401 [arXiv:1502.00437] [INSPIRE].
[24] L. Nakonieczny and M. Rogatko, Analytic study on backreacting holographic superconductors
with dark matter sector, Phys. Rev. D 90 (2014) 106004 [arXiv:1411.0798] [INSPIRE].
[25] L. Nakonieczny, M. Rogatko and K.I. Wysokiński, Magnetic field in holographic superconductor
with dark matter sector, Phys. Rev. D 91 (2015) 046007 [arXiv:1502.02550] [INSPIRE].
[26] L. Nakonieczny, M. Rogatko and K.I. Wysokiński, Analytic investigation of holographic phase
transitions influenced by dark matter sector, Phys. Rev. D 92 (2015) 066008
[arXiv:1509.01769] [INSPIRE].
[27] M. Rogatko and K.I. Wysokiński, P-wave holographic superconductor/insulator phase
transitions affected by dark matter sector, JHEP 03 (2016) 215 [arXiv:1508.02869] [INSPIRE].
[28] M. Rogatko and K.I. Wysokiński, Holographic vortices in the presence of dark matter sector,
JHEP 12 (2015) 041 [arXiv:1510.06137] [INSPIRE].
[29] M. Rogatko and K.I. Wysokiński, Viscosity of holographic fluid in the presence of dark matter
sector, JHEP 08 (2016) 124 [arXiv:1603.07858] [INSPIRE].
[30] M. Rogatko and K.I. Wysokiński, Condensate flow in holographic models in the presence of
dark matter, JHEP 10 (2016) 152 [arXiv:1608.00343] [INSPIRE].
[31] Y. Peng, Holographic entanglement entropy in superconductor phase transition with dark
matter sector, Phys. Lett. B 750 (2015) 420 [arXiv:1507.07399] [INSPIRE].
[32] Y. Peng, Q. Pan and Y. Liu, A general holographic insulator/superconductor model with dark
matter sector away from the probe limit, Nucl. Phys. B 915 (2017) 69 [arXiv:1512.08950]
[INSPIRE].

– 22 –
[33] M. Rogatko and K.I. Wysokiński, Viscosity bound for anisotropic superfluids with dark matter
sector, Phys. Rev. D 96 (2017) 026015 [arXiv:1612.02593] [INSPIRE].
[34] M. Rogatko and K.I. Wysokiński, Two interacting current model of holographic Dirac fluid in
graphene, Phys. Rev. D 97 (2018) 024053 [arXiv:1708.08051] [INSPIRE].
[35] B.D. Josephson, Possible new effects in superconductive tunnelling, Phys. Lett. 1 (1962) 251
[INSPIRE].
[36] M. Tinkham, Introduction to Superconductivity, Dover Publications, New York NY U.S.A.
(2004).
[37] C. Beck, Possible resonance effect of axionic dark matter in Josephson junctions, Phys. Rev.
Lett. 111 (2013) 231801 [arXiv:1309.3790] [INSPIRE].

JCAP01(2021)063
[38] F. Wilczek, Emergent Majorana Mass and Axion Couplings in Superfluids, New J. Phys. 16
(2014) 082003 [arXiv:1401.4379] [INSPIRE].
[39] G.T. Horowitz, J.E. Santos and B. Way, A Holographic Josephson Junction, Phys. Rev. Lett.
106 (2011) 221601 [arXiv:1101.3326] [INSPIRE].
[40] E. Kiritsis and V. Niarchos, Josephson Junctions and AdS/CFT Networks, JHEP 07 (2011)
112 [Erratum ibid. 10 (2011) 095] [arXiv:1105.6100] [INSPIRE].
[41] Y.-Q. Wang, Y.-X. Liu, R.-G. Cai, S. Takeuchi and H.-Q. Zhang, Holographic SIS Josephson
Junction, JHEP 09 (2012) 058 [arXiv:1205.4406] [INSPIRE].
[42] M. Rozali and A. Vincart-Emard, Chiral Edge Currents in a Holographic Josephson Junction,
JHEP 01 (2014) 003 [arXiv:1310.4510] [INSPIRE].
[43] E. Chang-Young, H. Kim and H. Nakajima, Noncommutative Supertori in Two Dimensions,
JHEP 08 (2008) 058 [arXiv:0807.0710] [INSPIRE].
[44] A. Amoretti, A. Braggio, N. Maggiore, N. Magnoli and D. Musso, Coexistence of two vector
order parameters: a holographic model for ferromagnetic superconductivity, JHEP 01 (2014)
054 [arXiv:1309.5093] [INSPIRE].
[45] H.-F. Li, L. Li, Y.-Q. Wang and H.-Q. Zhang, Non-relativistic Josephson Junction from
Holography, JHEP 12 (2014) 099 [arXiv:1410.5578] [INSPIRE].
[46] Y.-Q. Wang and S. Liu, Holographic s-wave and p-wave Josephson junction with backreaction,
JHEP 11 (2016) 127 [arXiv:1608.06364] [INSPIRE].
[47] S. Liu and Y.-Q. Wang, Holographic model of hybrid and coexisting s-wave and p-wave
Josephson junction, Eur. Phys. J. C 75 (2015) 493 [arXiv:1504.06918] [INSPIRE].
[48] Y.-P. Hu, H.-F. Li, H.-B. Zeng and H.-Q. Zhang, Holographic Josephson Junction from Massive
Gravity, Phys. Rev. D 93 (2016) 104009 [arXiv:1512.07035] [INSPIRE].
[49] R.-G. Cai, Y.-Q. Wang and H.-Q. Zhang, A holographic model of SQUID, JHEP 01 (2014) 039
[arXiv:1308.5088] [INSPIRE].
[50] S. Takeuchi, Holographic Superconducting Quantum Interference Device, Int. J. Mod. Phys. A
30 (2015) 1550040 [arXiv:1309.5641] [INSPIRE].
[51] J.F. Annett, Superconductivity, superfluids and condensates, Oxford University Press, Oxford
U.K. (2004).
[52] J.B. Ketterson and S.N. Song, Superconductivity, Cambridge University Press, Cambridge U.K.
(1999).
[53] T. Vachaspati and A. Achúcarro, Semilocal cosmic strings, Phys. Rev. D 44 (1991) 3067
[INSPIRE].
[54] A. Achúcarro and T. Vachaspati, Semilocal and electroweak strings, Phys. Rept. 327 (2000) 347
[hep-ph/9904229] [INSPIRE].

– 23 –
[55] S.A. Abel and B.W. Schofield, Brane anti-brane kinetic mixing, millicharged particles and
SUSY breaking, Nucl. Phys. B 685 (2004) 150 [hep-th/0311051] [INSPIRE].
[56] S.A. Abel, J. Jaeckel, V.V. Khoze and A. Ringwald, Illuminating the Hidden Sector of String
Theory by Shining Light through a Magnetic Field, Phys. Lett. B 666 (2008) 66
[hep-ph/0608248] [INSPIRE].
[57] S.A. Abel, M.D. Goodsell, J. Jaeckel, V.V. Khoze and A. Ringwald, Kinetic Mixing of the
Photon with Hidden U(1)s in String Phenomenology, JHEP 07 (2008) 124 [arXiv:0803.1449]
[INSPIRE].
[58] NA64 collaboration, Search for invisible decays of sub-GeV dark photons in missing-energy
events at the CERN SPS, Phys. Rev. Lett. 118 (2017) 011802 [arXiv:1610.02988] [INSPIRE].

JCAP01(2021)063
[59] P. Jean et al., Early SPI/INTEGRAL measurements of 511 keV line emission from the 4th
quadrant of the Galaxy, Astron. Astrophys. 407 (2003) L55 [astro-ph/0309484] [INSPIRE].
[60] J. Chang et al., An excess of cosmic ray electrons at energies of 300–800 GeV, Nature 456
(2008) 362 [INSPIRE].
[61] PAMELA collaboration, An anomalous positron abundance in cosmic rays with energies
1.5–100 GeV, Nature 458 (2009) 607 [arXiv:0810.4995] [INSPIRE].
[62] Muon g-2 collaboration, Final Report of the Muon E821 Anomalous Magnetic Moment
Measurement at BNL, Phys. Rev. D 73 (2006) 072003 [hep-ex/0602035] [INSPIRE].
[63] H. Davoudiasl, H.-S. Lee and W.J. Marciano, ‘Dark’ Z implications for Parity Violation, Rare
Meson Decays, and Higgs Physics, Phys. Rev. D 85 (2012) 115019 [arXiv:1203.2947]
[INSPIRE].
[64] H. Davoudiasl, H.-S. Lee, I. Lewis and W.J. Marciano, Higgs Decays as a Window into the
Dark Sector, Phys. Rev. D 88 (2013) 015022 [arXiv:1304.4935] [INSPIRE].
[65] B.S. Acharya, S.A.R. Ellis, G.L. Kane, B.D. Nelson and M.J. Perry, The lightest visible-sector
supersymmetric particle is likely to be unstable, Phys. Rev. Lett. 117 (2016) 181802
[arXiv:1604.05320] [INSPIRE].
[66] B. Holdom, Two U(1)’s and  charge shifts, Phys. Lett. B 166 (1986) 196 [INSPIRE].
[67] D. Lüst, Intersecting brane worlds: A Path to the standard model?, Class. Quant. Grav. 21
(2004) S1399 [hep-th/0401156] [INSPIRE].
[68] S. Abel and J. Santiago, Constraining the string scale: From Planck to weak and back again, J.
Phys. G 30 (2004) R83 [hep-ph/0404237] [INSPIRE].
[69] K.R. Dienes, C.F. Kolda and J. March-Russell, Kinetic mixing and the supersymmetric gauge
hierarchy, Nucl. Phys. B 492 (1997) 104 [hep-ph/9610479] [INSPIRE].
[70] G.T. Horowitz and R.C. Myers, The AdS/CFT correspondence and a new positive energy
conjecture for general relativity, Phys. Rev. D 59 (1998) 026005 [hep-th/9808079] [INSPIRE].

– 24 –
 PHYSICAL REVIEW D 103, 124021 (2021)

Axionlike dark matter clouds around rotating black holes

* †
Bartlomiej Kiczek and Marek Rogatko
Institute of Physics, Maria Curie-Sklodowska University, pl. Marii Curie-Sklodowskiej 1,
20-031 Lublin, Poland

(Received 5 March 2021; revised 19 April 2021; accepted 17 May 2021; published 9 June 2021)

Numerical analysis of a dark matter axionlike cloud in the vicinity of a rotating black hole has been
performed. The model where an axionlike scalar field is nontrivially coupled to the Maxwell field is studied
in the spacetime of a Kerr black hole in a uniform magnetic field and in the Kerr-Newman one. The
dependence of scalar mass and black hole angular momentum on accumulation of the axion dark matter
cloud was given. It was revealed that condensation of the dark matter clouds is preferable for a very small
mass of axion.

DOI: 10.1103/PhysRevD.103.124021

I. INTRODUCTION It was established [10] that the superradiant instability can

The nature of the elusive ingredient of our Universe, lead to the generation of extremely dense axion clouds in the
dark matter, is one of the most intriguing mysteries of the nearby rotating black holes. Moreover, the stimulated decay
contemporary physics and astrophysics. Ultralight bosons may lead to extremely bright lasers. A possible connection
like axion, axionlike particles, and dark photons could be with the observed fast radio bursts was proposed.
the answer for these tantalizing questions. From the point of Neglecting the rotational effects, axion configuration
view of UV theory, the QCD axions are well motivated as around pulsars was studied in Ref. [11]. Among all it was
the solution of the CP problem [1–3]. Recently, axionlike found that the axions form a localized condensate or radiate
particles widely emerging in the realm of string theory [4] as outgoing waves, depending on if the pulsar frequency is
also attract much attention. smaller or greater than that of the axion mass.
Both axion and axionlike particles are regarded as On the other hand, the analysis of broad-band radio
constituting the possible hidden sector. This fact triggers telescope observations of magnetar PSR J1745-2900,
the motivation to search for them in various kinds of enables to establish with the confidence of 95 percent
experiments and theoretical researches. Namely, it turns out limits, the resonant axion-photon conversion emission line
that axion dark matter has novel effects in polarization of flux density. These data were translated into limits on
the cosmic microwave background [5] and can be detected axion-photon coupling constant gaγγ versus axion mass. If
in the future terrestrial or astrophysical observations. In there is a dark matter cusp, then the limits reduce to
Ref. [6] the new mechanism where a coherently oscillating gaγγ > 6–34 × 10−14 GeV−1 , overlapping the axion models
axionlike particle field can transfer its energy to a dark with mass range over 33 eV [12,13]. It is argued [14,15]
photon has been elucidated. Recently, it has been argued that the axion coupling to photon depends on the specific
that radio telescope observations of neutron stars will model and is related to the values ∼10−11 –10−15 GeV−1 for
enable the possible detection of axion dark matter, through intermediate, ∼10−19 GeV−1 for Grand Unification Theory,
the axion resonant conversion into radio-frequency pho- and ∼10−21 GeV−1 for Planck energy scales.
tons. The conversion probabilities are proportional to the Studies of light rays passing through an axion and
strength of the magnetic field surrounding the neutron axionlike clouds surrounded by a stationary axisymmetric
star [7,8]. black hole, focusing on the experimental setup that is
The process of lasing of an ultralight axion condensate required for the detection of such an effect, and paying
into photons, relevant for a superradiant axion condensate attention to the radio observations of linearly polarized
around a stellar mass black hole, was elaborated in [9]. The astrophysical sources, like active galactic nuclei, have been
influence of plasma properties placed around the black hole
performed in [16].
in question, on the lasing of the axion condensate, was also
In [17] it was proposed to detect axionlike dark matter by
revealed.
using linearly polarized pulsar light. A pulsar linear
polarization angle may vary with time, due to the birefrin-
*
[email protected] gence effect which is caused by an oscillating galactic
†
[email protected] aforementioned hidden sector component.

2470-0010=2021=103(12)=124021(10) 124021-1 © 2021 American Physical Society

BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 103, 124021 (2021)

The numerical solution of the laser emission problem from where R is the Ricci scalar, Fμν ¼ 2∇½μ Aν is the Maxwell
an axion dense cloud around a spinning black hole was field strength tensor, and Ψ is the scalar field (axion) with
presented in [18,19], where it was envisaged that the laser mass μ. The last term of the action describes the coupling of
emission existed at classical level and the presence of electric axion field Ψ to one of the electromagnetic field invariants,
charge or rotation leads to the appearance of the black hole composed of Maxwell and dual Maxwell filed strengths,
with nontrivial axionic hair. Moreover, the coupling constant where by  we have denoted the Hodge dual operator. Note
of the hidden sector triggers the strong instabilities affecting that k constitutes the axionic coupling constant to the Uð1Þ-
superradiant clouds around black holes. On the other hand, in gauge field.
[20] the entire spectrum of the most unstable superradiant Varying the action with respect to the scalar field Ψ, we
modes of the Proca field around a Kerr black hole was obtain the equation
obtained, as well as constraints on dark photon and axionlike
particles were given. k
In our paper we elaborate the subject of the possible ∇μ ∇μ Ψ − μ2 Ψ −  Fμν Fμν ¼ 0: ð2Þ
2
existence of axionic dark matter clouds in the spacetime of
stationary axisymmetric black holes. Numerical simula- On the other hand, the Uð1Þ-gauge field is subject to the
tions based on the axion dark matter model, where axions relation
are coupled to the Maxwell field invariant composed of
dual and ordinary Uð1Þ-gauge field strengths, enable us to
reveal the basic characteristics of the system in question. ∇μ Fνμ þ 2k  Fνμ ∇μ Ψ ¼ 0: ð3Þ
We shall pay attention to two cases of black holes, i.e., a
Kerr black hole in a uniform magnetic field and Kerr- The resulting Klein-Gordon-like equation (2) contains,
Newman spacetime. despite the standard dynamical and mass terms, an addi-
The rest of the paper is organized as follows. In Sec. II tional source term, being independent of axionlike field Ψ.
we give a short overview of the axionlike dark matter portal The presence of the nonzero source term, containing the
and provide information about studied black hole back- dual invariant, explicitly defined as
grounds. In the subsections we discuss underlying equa-
tions of motion and the problem of free energy for dark 1
matter axionic clouds around rotating black holes in I ¼ Fμν Fμν ¼ ϵμνρλ Fρλ Fμν ; ð4Þ
2
question. Section III is devoted to the description of the
achieved results. Namely, we examine the possibilities of where ϵαβγδ stands for the totally antisymmetric Levi-Civita
condensations of dark matter in the vicinity of Kerr black symbol, is crucial for the scalarization of a black hole.
holes in a uniform magnetic field and around stationary Namely, if it is equal to zero, the axionlike scalar field
axisymmetric Kerr-Newman black holes. In Sec. IV we equation of motion reduces to the simple massive Klein-
conclude our investigations. Finally, Appendix contains the Gordon case, without any self-interaction potential. Then
relevant technical details concerning the numerical method. the no-hair theorem plays its role and prevents any scalar
hair configuration on the black hole from emerging.
II. AXIONLIKE DARK MATTER SECTOR On the other hand, it is easy to check that the invariant in
question, Fμν Fμν , is equal to zero in the case when
In this section we shall present the basic equations standing
Fμν ¼ 0, or for spherically symmetric spacetime. In order
behind the axion dark matter sector model, viewed as the
axionlike scalar field coupled to the Maxwell Uð1Þ-gauge to be nontrivial, Fμν Fμν ≠ 0 has to ensure both rotational
field. The basic idea lies in the nontrivial axionic coupling to and magnetic Uð1Þ-gauge field components.
the Maxwell strength field invariant constructed from dual In what follows the main objective of our paper will be to
and ordinary Maxwell field strengths. In what follows one elaborate on the behavior of axionlike field dark matter
investigates the behavior of axionlike dark matter clouds sector in the vicinity of a black hole. As it was remarked the
surrounded spinning black hole in a uniform, say galactic survivability of the I term in Eq. (2) would be crucial for
magnetic field, as well as besieged the Kerr-Newman black our studies. Therefore, we implement a magnetic field in
hole. For convenience, we also refer to them as axions. To the considered stationary axisymmetric black hole space-
commence with, we start with the Einstein-Maxwell-axion time in two ways: internally, as a consequence of the Kerr-
dark matter theory described by the following action: Newman black hole solution and externally, as, e.g., a
 galactic magnetic field surrounding a Kerr black hole. The
Z
4 pffiffiffiffiffiffi 1 1 latter idea was originally proposed by Wald in [21], where
S ¼ d x −g R − Fμν Fμν − ∇μ Ψ∇μ Ψ the uniform magnetic field around a black hole was studied.
4 2
 We shall consider both of these background line ele-
μ2 k
− Ψ2 − Ψ  Fμν Fμν ; ð1Þ ments and investigate properties of axionic dark matter
2 2 clouds around the black holes in question.

124021-2
AXIONLIKE DARK MATTER CLOUDS AROUND ROTATING … PHYS. REV. D 103, 124021 (2021)

A. Kerr black hole in a uniform magnetic field 1

Aμ ¼ Bðmμ þ 2akμ Þ: ð6Þ
In this section we recall, for the reader’s convenience, the 2
basic idea concerning the Wald’s introduction of the uni-
form magnetic field in the spacetime of a Kerr black hole In this way we can introduce a static magnetic field to the
[21]. The line element of a Kerr black hole in Boyer- system, which is oriented along the black hole rotation axis.
Lindquist coordinates is provided by the following: From an astrophysical perspective, such a case may seem
  quite idealized, however, it is an interesting starting point
2Mr 4Mra sin2 θ
ds2 ¼ − 1 − dt2 − for including magnetic fields into field theories around
Σ Σ black holes. One way or another, any external (galactic)
Σ 2 Ξ sin2 θ 2 magnetic field can be cast on the parallel and perpendicular
× dtdϕ þ dr þ Σdθ2 þ dϕ ; ð5Þ (to the rotation axis) components, and the perpendicular
Δ Σ
component can be neglected. Using this set up allows us to
with the auxiliary functions defined as utilize all the mathematical properties of the Kerr geometry,
such as axial symmetry, in constructing the numerical
Σðr; θÞ ¼ r2 þ a2 cos2 θ; solution for the scalar. One has to remember, however, that
ΔðrÞ ¼ r2 − 2Mr þ a2 ; a nonzero magnetic field breaks the reflection symmetry
with respect to the equatorial plane.
Ξðr; θÞ ¼ ðr2 þ a2 Þ2 − a2 Δsin2 θ: As it has been already mentioned, we are interested in a
static magnetic field, parallel to the rotation axis, so we can
The solution naturally describes a rotating black hole and drop the timelike Killing vector from the general form of
is parametrized by two physical quantities, black hole the gauge potential and write it in the form as follows:
mass M and angular momentum parameter a ¼ MJ . The
stationary axisymmetric line element (5) possesses two
Killing vector fields, the timelike kμ ¼ð∂=∂tÞμ and axial 1 B sin2 θ
Aμ dxμ ¼ Bgμν mν dxμ ¼ ð−2Mardt þ ΞdϕÞ: ð7Þ
one mμ ¼ ð∂=∂ϕÞμ . 2 2Σ
If we consider the electromagnetic field equations in the
spacetime of a Kerr black hole, neglecting the metric back In order to proceed to the analysis of the axion dark matter
reaction, it is possible to derive a general analytical form of equation of motion, we should find the invariant I in the
the vector potential, being a combination of Killing vectors spacetime under consideration. Its explicit form is as
of the underlying spacetime, such as follows:

aB2 M sin2 θ cos θ 6

I ¼− ½3a þ 2a4 Mr − 5a4 r2 − 8a2 Mr3 − 32a2 r4 − 24r6
2Σ4
þ 4a2 ða4 − a2 r2 þ 2ðM − rÞr3 Þ cos 2θ þ a4 ða2 − 2Mr þ r2 Þ cos 4θ: ð8Þ

Because of the fact that the obtained formula is a bit long and complicated and it might not be easy to imagine its shape, for
the convenience of the reader, we visualize it in Fig. 1.

B. Kerr-Newman black hole spacetime

As far as the Kerr-Newman spacetime is concerned, it generalizes a Kerr solution and represents a black hole that does
not only rotate but also is electrically charged. The line element implies
 
2 2Mr Q2 2að2Mr − Q2 Þ sin2 θ Σ Ξ sin2 θ 2
ds ¼ − 1 − þ dt2 − dtdϕ þ dr2 þ Σdθ2 þ dϕ : ð9Þ
Σ Σ Σ Δ Σ

The auxiliary functions Σ and Ξ are defined in the same way as in the previous case, however, ΔðrÞ has the form given by

ΔðrÞ ¼ r2 − 2Mr þ a2 þ Q2 ; ð10Þ

where Q is the electric charge of the black hole. The solution naturally possesses a nonzero electromagnetic vector potential
of the form

124021-3
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 103, 124021 (2021)

which allows us to fix one quantity to unity. The scaling

concerns the quantities from both backgrounds. For con-
venience, in our numerical simulations, we use the above
scaling and fix the black hole mass to unity M ¼ 1.

D. Free energy of dark matter axionic clouds

in the spacetime of rotating black holes
The existence of the solution to the field equation, i.e.,
some state of the system, does not guarantee that it is the
physically preferable configuration. To verify this, one
ought to consider the thermodynamics of the system and
look for the relevant quantities [22]. As we consider the
gravitational system without backreaction, the thermody-
FIG. 1. Angular and radial dependence of the Maxwell field namical quantities of the black hole, such as entropy and
invariant in Kerr spacetime evaluated at the event horizon. Colors of Hawking temperature, remain unaffected by the axionic
the lines indicate subsequent surfaces of constant r. The solid lines dark matter condensate. Thus we wish to examine the free
correspond to a ¼ 0.5, while the dashed lines to a ¼ 0.99. The rise energy difference generated by the nontrivial profile of the
of angular momentum pumps up the value of the invariant
significantly and shifts its peak towards the black hole equator.
scalar ψ with respect to the hairless solution. To proceed
further, we consider the ψ dependent part of the underlying
action,
rQ arQ sin2 θ Z 
Aμ dxμ ¼ dt − dϕ: ð11Þ pffiffiffiffiffiffi 1
Σ Σ S axion ¼ d4 x −g − ∇μ Ψ∇μ Ψ
On the other hand, the corresponding electromagnetic 2
2

invariant, needed in the axion dark matter equations of μ 2 k μν
− Ψ − Ψ  F Fμν : ð14Þ
motion, acquires a new simpler form, described by 2 2

4aQ2 rða2 − 2r2 þ a2 cos 2θÞ cos θ In order to find the free energy contribution of the axionic
I ¼− : ð12Þ
Σ4 cloud, we evaluate the Euclidean on-shell action related
with (14).
C. Equation of motion Firstly, we use the equation of motion for the axionic
Let us suppose that the axion dark matter field will be a field ψ (2) and substitute it into the action. This allows us to
function depending on two coordinates, radial and azimu- remove the term with the Maxwell field strength tensor.
thal ones, i.e., Ψ ¼ ψðr; θÞ. It leads to the axion equation of After few transformations we make use of the Gauss
motion provided by the relation theorem and split the action into volume and surface terms.
Because of the boundary conditions (see the next section
kΣ for details), the surface integral vanishes. In last step we
Δ∂ 2r ψ þ 2ðr − MÞ∂ r ψ þ ∂ 2θ ψ þ cot θ∂ θ ψ − μ2 Σψ ¼ I;
2 perform Wick’s rotation of the time coordinate and get the
ð13Þ explicit formula provided by
Z
∂ m stands for the derivative with respect to the m- Σ sin θ
F ¼ −2π drdθ ½ð∂ r ψÞ2 grr þ ð∂ θ ψÞ2 gθθ þ μ2 ψ 2 :
coordinate. The obtained equation is an elliptic partial M 2
differential equation, which is linear in ψ. The general form ð15Þ
of this equation remains the same in both backgrounds.
They differ by the shape of Δ function, which can be
enriched with the electric charge, and by the source term on As the integrand is positive, in the whole domain, the
the right-hand side. It turns out that Eq. (13) follows a free energy shift is negative for every configuration of ψ
scaling transformation of the form field being the solution to the considered equations of
motion. Although there is a caveat. It can be supposed that
r → cr a → ca any nontrivial ψ will be preferred by nature. However, this
is not really the case. For the given ansatz, the system has
M → cM k → kc2 only trivial zero solution when the source term I is zero.
Q → cQ B → B=c This allows us to state that the considered axionic dark
matter clouds are magnetically induced and are only
μ2 → μ2 =c2 ; present in the system when I is nontrivial.

124021-4
AXIONLIKE DARK MATTER CLOUDS AROUND ROTATING … PHYS. REV. D 103, 124021 (2021)

Equation (15) will be extensively exploited, in the 2

∂ 2r ψ þ ∂ r ψ − μ2 ψ ¼ 0; ð19Þ
following section, to achieve the free energy plots. The r
aforementioned integral will be computed numerically.
which implies that
III. NUMERICAL RESULTS e−μr eμr
ψðrÞ ∼ þ : ð20Þ
This section will be devoted to the obtained numerical r r
solutions of both axion dark matter clouds surrounding
Kerr black holes immersed in a uniform magnetic field and The asymptotic flatness and regularity cause that the
a Kerr-Newman one. As we have already mentioned, they second term in the relation (20) vanishes, for a finite mass
differ by the shape of the source term originating from solution vanishes in the infinity. Thus, the requirement that
axionic coupling and the metric function Δ. We deal with ψðr → ∞Þ ¼ 0 comprises the second boundary condition.
the partial differential equation (13) by virtue of Chebyshev In the case of the boundaries imposed on the polar angle
spectral methods. As the equation is fully linear in ψ, the θ, we use the argumentation based on the symmetries of the
acquired solution is unique and well defined. For technical spacetime. For the north pole θ ¼ 0, the axial symmetry of
details of the numerical method see Appendix. the rotating black hole obliges the solution to be invariant
In the above set up one considers the field only above the under the transformation ϕ → ϕ  π. In other words the
black hole event horizon (including the horizon itself). The solution ought to be even along a meridian, with respect to
bounds for the radial coordinate are from the event horizon the pole. Therefore, ∂ θ ψ ¼ 0 is the reasonable choice.
to spatial infinity, precisely r ∈ ½rþ ; ∞Þ. The symmetry of However, for the equatorial plane θ ¼ π=2 the situation is
the spacetime allows us to narrow the domain to one quarter different, as it constitutes the place where both source terms
of the ðr; θÞ plane. For convenience, we pick the first change signs, and so does the solution. For that reason we
quarter, with θ ∈ ½0; π=2. Values of the solution for demand that ψ ¼ 0, there.
remaining quarters can be achieved by the negative
reflection with respect to the equatorial plane A. Axionic dark matter clouds around Kerr black holes
in a uniform magnetic field
ψðr; θ − π=2Þ ¼ −ψðr; π=2 − θÞ; ð16Þ
Now we proceed with the conclusions achieved from
the analysis of the numerical solutions of Eq. (13) in the
and the remaining part of the solution can be obtained by
adequate spacetimes of the rotating black holes. To
the rotation. Having our numerical domain defined, we can
commence with one considers the results obtained for a
move to the necessary transformations.
Kerr black hole in a uniform magnetic field.
To implement spectral methods based on Chebyshev
In Fig. 2 we present the first series of spatial distribution
polynomials, we have to map the coordinates of the
plots of the axionic dark matter field. For convenience, we
manifold onto ½−1; 1 intervals. In order to do this we
present the squared distribution ψ 2 , where the ðr; θÞ plane
use the following transformations for r and θ:
has been cast into Cartesian coordinates ðx; yÞ. The black
2rþ circle corresponds to the area hidden under the black hole’s
z¼1− ; ð17Þ event horizon. In the first pair of plots, depicted in panels
r
(a) and (b), one can see the distribution of the ultralight
4θ axionic dark matter field. It can be noticed that it is loosely
u¼ − 1; ð18Þ concentrated around polar regions. The equatorial plane
π
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi remains free from the scalar ψ, which results from the
where rþ ¼ M þ M2 − a2 − Q2 , which is the standard imposed boundary conditions and the symmetry of the
definition of the outer event horizon of rotating black holes. problem in question. The increase of the black hole angular
After the transformation, z ¼ −1 represents the inner momentum does not drastically change the field distribu-
boundary—the event horizon and z ¼ 1 spatial infinity. tion. However, it significantly influences its magnitude (see
Similarly, for u ¼ −1 one thinks about a north pole of a the values on colorbars). Furthermore, the field slowly
black hole, while u ¼ 1 represents the equator of the object decays with distance and spreads up to r ¼ 10rþ outside
in question. the event horizon. This is not the case for the massive
Let us now discuss the boundary condition imposed on axionic dark matter field, which is presented in panels (c)
the solution of the axionic dark matter field equation. and (d) of Fig. 2. The mass μ2 is quite large, in terms of
Namely, for the black hole event horizon r ¼ rþ , we geometrical units, however, it serves the purpose of
demand that the derivative with respect to the r-coordinate visualizing the contrast. In the case under inspection the
is given by ∂ r ψ ¼ 0. This fact ensures the regularity of the increase of the angular momentum causes dragging the
solution. For r → ∞, the field equation takes the simple axionic clouds towards the equatorial plane. This effect
angle independent form provided by the relation seems to be quite intuitive, by the analogy to the centrifugal

124021-5
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 103, 124021 (2021)

FIG. 2. The distribution of the axionic dark matter clouds around a black hole with an external uniform magnetic field. Every panel
corresponds to different values of field mass and black hole angular momentum. For ultralight mass the cloud concentrates in polar
areas, while the large mass field is dragged towards the equator.

forces in classical mechanics. Similarly to the former case, curves have in common is a significant decrease of free
the angular momentum increase causes a boost in the field energy for small values of the axion mass. It means that
ψ magnitude, by several orders. axionic dark matter clouds are the most stable and the most
It should be noted that the approximated analytical strongly bound for very small, almost zero, field masses.
solution for this background has been derived in One might notice an interesting correlation of this result
Ref. [19]. Our numerical solution perfectly matches the with theoretical predictions for dark matter axion mass, to
results obtained before, in the considered limit—a slow be ultralight in sub eV region. Namely, the recent con-
rotation of the black hole and zero mass of the axion field. straints on bosonic dark matter for ultralow field nuclear
To proceed further with the studies of hairy configura- magnetic resonance were proposed in [23]. The new
tions in Kerr-Wald spacetime, let us take a look at the experimental bounds for axionlike dark matter particles
behavior of the free energy, with respect to the change of are ranging from 1.8 × 10−16 to 7.8 × 10−14 eV. Recently,
the other parameters of the theory. In Fig. 3 the free energy the direct implications on the mass of ultralight dark matter
shift as a function of the axion field mass can be observed. particles by studies on mass and spin of accreting and
Here we present several curves for different values of the jetted astrophysical black holes have been established [24].
black hole angular momentum. All curves share the similar It was revealed that axionlike particles with the mass
behavior, which is scaled differently. Moreover, what all range 10−21 < μðeVÞ < 10−19 could contribute at almost

124021-6
AXIONLIKE DARK MATTER CLOUDS AROUND ROTATING … PHYS. REV. D 103, 124021 (2021)

FIG. 4. Free energy shift vs the black hole angular momentum.

FIG. 3. Free energy shift as a function of axionlike dark matter The extremal black holes (with very high angular momentum)
mass. Ultralight particles are the most preferred ones as they constitute a perfect environment for dark matter axionic hair since
cause the most significant free energy decrease. the free energy fall off is the biggest.

(a) (b)

(c) (d)

FIG. 5. Axionic dark matter clouds around Kerr-Newman black holes with Q ¼ 0.1. High angular momentum of the black hole reveals
rich geometrical structure of the clouds.

124021-7
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 103, 124021 (2021)

10 percent of the dark matter mass. On the contrary, for the

mass range 10−19 > μðeVÞ > 10−17 , they constitute only
0.01 to 1 percent of the dark sector mass.
On the other hand, Fig. 4 presents the free energy shift as
a function of black hole angular momentum for different
masses of the field ψ. The course of the curves is also very
similar and surely they follow some kind of μ2 dependent
scaling. The shift is slight for moderate values of the
angular momentum and becomes stronger for quickly
rotating black holes. Extremal black holes, with a
approaching to 1, bring the biggest fall off of free energy.
Once again the decrease is the most drastic for the ultralight
field (dark blue curve in the plot). All these observations
allow us to conclude that extremal black holes constitute
good environments for the emergence of the ultralight FIG. 6. Free energy shift vs black hole angular momentum in
axionic dark matter clouds. Kerr-Newman background. The presented function shows the
slight drop of free energy for moderate values of angular
momentum. However, the lowest free energy value is found
B. Axion dark matter clouds in the for the extreme black hole regime.
vicinity of Kerr-Newman black holes
Let us now discuss the characteristic features of axionic
dark matter clouds nearby Kerr-Newman black holes. This also shrinks the spread of the clouds, as they quickly decay
background has a distinct electromagnetic vector potential, with distance from the horizon. Besides the rise of angular
thus the behavior of axionic hair differs significantly from momentum, it significantly increases the magnitude of the
the former case. It can be seen in Fig. 5, where we plot axionic field (see colorbars).
analogical spatial distributions of ψ 2 , just like in the case of The Kerr-Newman background can be analyzed thermo-
a Kerr black hole dipped in a uniform magnetic back- dynamically in the similar manner as it has been performed
ground. However, these solutions are essentially various for in the Kerr in a uniform magnetic field system. However,
two main reasons. Firstly, the shape of the I expression is for the elaborated case, the free energy dependence on the
different, hence the source term envisages the other kind of black hole angular momentum reveals a slightly different
solution. Secondly, the electromagnetic component (the behavior. These curves are portrayed in Fig. 6. Energy
electrical charge Q) enters the geometric relations, such as characteristics are monotonic and decreasing with growth
the radius of the event horizon. Because of this interplay of the angular momentum. Nevertheless, in the extreme
between a and Q, the black hole angular momentum is black hole regime, the dynamics of the free energy rises and
limited to the value below one, because of the fact that we the curves are steeper as a → 1. Once again, the curve
require rþ to be a real number. In this illustrative example corresponding to the zero mass has the lowest free energy.
one sets Q ¼ 0.1. The panels (a) and (b) illustrate ultralight
axionic dark matter clouds. In the case under inspection, IV. CONCLUSION
their distribution is slightly affected by the black hole
angular momentum. When its value is moderate, the cloud In our paper we have elaborated on the axioniclike dark
aggregates around polar regions of the Kerr-Newman black matter model, where the scalar field (axion) is nontrivially
hole. As the angular momentum rises, axions flow towards coupled to the electromagnetic Uð1Þ-gauge field, via
the equator and are spread over whole hemispheres. The coupling to the Fαβ Fαβ invariant. We considered the
clouds of dark matter are distributed on the majority of the possibility of accumulating axion dark matter clouds in
slice’s area, except the equatorial region, which is naturally the vicinity of a rotating black hole. Namely, we have
the result of the imposed boundary conditions. However, studied the Kerr black hole immersed in a uniform
contrary to previous background, axionic clouds are magnetic field and Kerr-Newman black hole spacetime.
strongly localized in the vicinity of the black hole event In both cases axion dark matter clouds tend to accumulate
horizon. Their distribution quickly vanishes with the in polar regions of the black hole in question.
growth of the distance. The large mass case, presented As far as the Kerr black hole is concerned, it turns out
on panels (c) and (d), reveals a strong concentration of the that the increase of a black hole angular momentum does
field for both values of the angular momentum. In the not change the distribution of ultralight dark matter clouds
former case the cloud is visible over the majority of black but influences its magnitude. For the increase of the axion
hole hemispheres, while in the latter case it is mostly mass, the cloud gathers in the equatorial area of the object.
present in the equatorial area. Moreover, the axion dark For the Kerr-Newman black hole the ultralight axion
matter field blurs further into the space. The rise of mass dark matter cloud distribution depends on the black hole

124021-8
AXIONLIKE DARK MATTER CLOUDS AROUND ROTATING … PHYS. REV. D 103, 124021 (2021)

angular momentum. The increase of its value spreads the

cloud over the space surrounding the Kerr-Newman black
hole. In the case of a large mass axion, the field is strongly
dragged towards the equatorial area. Moreover, it was
revealed that axionlike dark matter clouds are preferable for
very small, almost zero mass, axion fields.
Considering that axionic dark matter clouds do not
emerge spontaneously, but are rather magnetically induced,
this mechanism naturally requires a magnetic component,
such as a galactic magnetic field or a charged rotating black
hole. Nevertheless, if such dark matter clouds constitute
reality, it will be a complicated observational challenge to
reveal their existence. This task will require advanced
numerical relativity simulations, which should take into FIG. 8. Convergence of the free energy of the system. Left y-
account additional astrophysical and particle-related sce- scale corresponds to the Wald background (purple stars), while
narios, such as, e.g., plasma-axion dark matter interactions. the right y-scale to the Kerr-Newman solution (red triangles). The
We hope to return to these problems elsewhere. value of the free energy quickly converges to the limit.

The numerical code has undergone two convergence

APPENDIX: TECHNICAL DETAILS ON THE trials on the N × N grid. The first relies on evaluating the
NUMERICAL METHOD mean of the residuals
The implemented numerical method relies on the
Chebyshev differentiation matrices, which are used to dis- ζ ¼ hjLψ − Bji; ðA1Þ
cretize the partial differential equation, on a Chebyshev grid.
Similarly to finite difference methods (FDMs), the usage of on a set of random points, which do not belong to the
differentiation matrix allows one to translate a differential spectral grid. However, we calculate this metric using
equation into a system of linear equations Lψ ¼ B. Unlike another differentiation scheme, in this case, standard central
the FDM, spectral differentiation requires only a few grid finite difference derivative. This allows us to verify if the
points in order to achieve high accuracy. Having our discrete spectral solutions are relevant [26]. The result of this test is
differential operator constructed, we impose the boundary presented in Fig. 7. The increase of the number of grid
conditions by substituting particular rows in the matrix and points lowers the error of the solution in both gravitational
solve the system with standard linear algebra tools. In result backgrounds.
we obtain the vector of ψ values on the grid points. We have The second numerical test uses the free energy from
implemented the numerical scheme in PYTHON, based on the Eq. (15). For the same set of physical parameters we
MATLAB counterpart [25], using open source libraries. evaluate the free energy, increasing the number of grid
points in each step. This test is visualized in Fig. 8. One can
see that the free energy quickly converges to its limit. For

FIG. 7. Convergence of the mean value of residuals calculated FIG. 9. Dependence of the value of the solution at a distant
at the set of random points. The error smoothly decays with the point in the numerical domain on the location of the inner
growth of the grid. boundary. The event horizon radius is rþ ≈ 1.86 for this case.

124021-9
BARTLOMIEJ KICZEK and MAREK ROGATKO PHYS. REV. D 103, 124021 (2021)

convenience, we refer to the Kerr black hole in a uniform However, to make sure that our results are reliable, one can
magnetic field as the Wald solution. shift the inner boundary arbitrarily and check the behavior
By the analysis of the convergence of the algorithm we of the solution in the deep interior of the numerical domain.
picked N ¼ 50 in each direction, as it constitutes a This is being tested by calculating the value of ψ at the
reasonable compromise between the accuracy and the point ηðr ¼ 10; θ ¼ π=4Þ, systematically for subsequent
length of computation. Both presented tests were executed inner boundary locations. We define a ratio
for μ ¼ 1, a ¼ 0.5, and Q ¼ 0.1 in case of the Kerr-
Newman background. This is one of the considered cases in ψðηÞrb − ψðηÞrb ¼rþ
Figs. 4 and 6. For different physical parameters the δψ ¼ ; ðA2Þ
numerical scheme revealed similar behaviors. All solutions ψðηÞrb ¼rþ
shown in the plots in this work meet the require-
ment ζ < 10−5 . and compute it for several values of the position of inner
Finally, we check if the numerical solution is invariant boundary rb. The result of this test is presented in Fig. 9.
for any change of the location of the inner boundary. In the The value of the function very weakly depends on the
whole work we always use the black hole event horizon as position of the inner boundary. Relative error is to the order
the inner boundary rb ¼ rþ of our numerical domain. of 10−4 within huge changes of rb —up to 3rþ .

[1] R. D. Peccei and H. R. Quinn, CP Conservation in the [14] P. Arias, D. Cadamuro, M. Goodsell, J. Jaeckel, J. Redondo,
Presence of Pseudoparticles, Phys. Rev. Lett. 38, 1440 (1977). and A. Ringwald, WISPy cold dark matter, J. Cosmol.
[2] S. Weinberg, A New Light Boson?, Phys. Rev. Lett. 40, 223 Astropart. Phys. 06 (2012) 013.
(1978). [15] P. W. Graham, I. G. Irastorza, S. K. Lamoreaux, A. Lindner,
[3] F. Wilczek, Problem of Strong P and T Invariance in the and K. A. van Bibber, Experimental searches for axion and
Presence of Instantons, Phys. Rev. Lett. 40, 279 (1978). axion-like particles, Annu. Rev. Nucl. Part. Sci. 65, 485
[4] P. Svrcek and E. Witten, Axions in string theory, J. High (2015).
Energy Phys. 06 (2006) 051. [16] A. D. Plascencia and A. Urbano, Black hole supperadiance
[5] M. A. Fedderke, P. W. Graham, and S. Rajendram, Axion and polarization-dependent bending of light, J. Cosmol.
dark matter detection with CMB polarization, Phys. Rev. D Astropart. Phys. 04 (2018) 059.
100, 015040 (2019). [17] X. Bi, Y. Gao, J. Guo, N. Houston, T. Li, F. Xu, and X.
[6] R. T. Co, A. Pierce, Z. Zhang, and Y. Zhao, Dark photon Zhang, Axion and dark photon limits from Crab Nebula
dark matter produced by axion oscillations, Phys. Rev. D 99, high energy gamma-rays, arXiv:2002.01796.
075002 (2019). [18] T. Ikeda, R. Brito, and V. Cardoso, Blasts of Light From
[7] F. P. Huang, K. Kadota, T. Sekiguchi, and H. Tashiro, Radio Axions, Phys. Rev. Lett. 122, 081101 (2019).
telescope search for the resonant conversion of cold dark [19] M. Boskovic, R. Brito, V. Cardoso, T. Ikeda, and H. Witek,
matter axions from magnetized astrophysical sources, Phys. Axionic instabilities and new black hole solutions, Phys.
Rev. D 97, 123001 (2018). Rev. D 99, 035006 (2019).
[8] J. W. Foster, Y. Kahn, O. Macias, Z. Sun, R. P. Eatough, V. I. [20] V. Cardoso, O. J. C. Dias, G. S. Harnett, M. Middleton, P.
Kondratiev, W. M. Peters, C. Weniger, and B. R. Safdi, Pani, and J. E. Santos, Constraining the mass of dark
Green Bank and Effelsberg Radio Telescope Searches for photons and axion-like particles through black-hole super-
Axion Dark Matter Conversion in Neutron Star Magneto- radiance, J. Cosmol. Astropart. Phys. 03 (2018) 043.
spheres, Phys. Rev. Lett. 125, 171301 (2020). [21] R. M. Wald, Black hole in a uniform magnetic field, Phys.
[9] S. Sen, Plasma effects on lasing of a uniform ultralight axion Rev. D 10, 1680 (1974).
condensate, Phys. Rev. D 98, 103012 (2018). [22] B. Kiczek and M. Rogatko, Influence of dark matter on
[10] J. G. Rosa and T. W. Kephart, Stimulated Axion Decay in black hole scalar hair, Phys. Rev. D 101, 084035 (2020).
Superradiant Clouds Around Primordial Black Holes, Phys. [23] A. Gracon et al., Constraints on bosonic dark matter for
Rev. Lett. 120, 231102 (2018). ultralow-field nuclear magnetic resonance, Sci. Adv. 5,
[11] B. Garbrecht and J. I. Mc Donald, Axion configurations eaax4539 (2019).
around pulsars, J. Cosmol. Astropart. Phys. 07 (2018) 044. [24] C. Ünal, F. Pacucci, and A. Loeb, Properties of ultralight
[12] J. Darling, New limits on axionic dark matter using the bosons from heavy quasar spins via superradiance, arXiv:
magnetar PSR J1745-2900, Astrophys. J. Lett. 900, L28 2012.12790.
(2020). [25] L. N. Trefethen, Spectral Methods in MATLAB (SIAM,
[13] J. Darling, Search for Axion Dark Matter Using the Philadelphia, 2000).
Magnetar PSR J1745-2900, Phys. Rev. Lett. 125, 121103 [26] J. P. Boyd, Chebyshev and Fourier Spectral Methods
(2020). (DOVER Publications Inc, Mineola, New York, 2000).

124021-10
Eur. Phys. J. C (2022) 82:586
https://doi.org/10.1140/epjc/s10052-022-10545-0

Regular Article - Theoretical Physics

Static axionlike dark matter clouds around magnetized rotating

wormholes–probe limit case
Bartlomiej Kiczeka , Marek Rogatkob
Institute of Physics, Maria Curie-Sklodowska University, pl. Marii Curie-Sklodowskiej 1, 20-031 Lublin, Poland

Received: 22 April 2022 / Accepted: 22 June 2022

© The Author(s) 2022

Abstract The problem of the distribution of axionlike par- The current understanding of wormholes was revealed in
ticle, being the model of dark matter, in the nearby of rotating [5], where the conditions for traversability for Lorentzian
wormholes has been investigated numerically. In the model wormholes were defined by the survivability of human trav-
in question the axion scalar is non-trivially coupled to the ellers. This redefinition was not only of great importance to
Maxwell gauge field. We consider two toy models of rotat- physics, but also to futurology and is still seen as a main way
ing wormholes embedded in magnetic field, Kerr-like and to travel at large distances in space by humans. On the other
Teo rotating wormholes. Moreover one assumes that the mat- hand, models of a wormhole, possessing no event horizon
ter fields will not backreact on the wormhole spacetimes, i.e., and physical singularities, were elaborated in [6–8].
we shall study the problem in the probe limit case. We point In order to obtain such kind of wormhole solutions one
out the differences in the distribution of dark matter compar- should invoke phantom field (exotic matter), whose energy
ing to the location of it in the vicinity of rotating magnetized momentum tensor violates the null, weak and strong energy
black holes. conditions, as well as, its kinetic energy term is of the reversed
sign.
However, traversability requires also stability of the
1 Introduction wormhole solution, except small acceleration and tidal
forces. To achieve this goal we may consider a generalized
Recently there has been observed a big resurgence of interests Einstein gravity theories, like Gauss-Bonnet-dilaton theory.
in a special class of Einstein field equation solutions repre- Moreover in this theory wormholes can be built with no use
senting tunnel-like structures connecting spatially separated of such exotic kind of matter [9,10]. On the other hand, the
regions or even more different Universes, nowadays called method of constructing traversable wormholes by applying
wormholes. These fascinating objects are not only important duality rotation and complex transformations was proposed
for popular culture, but also gain a lot of scientific attention [11,12]. By assuming that the dilaton field constitutes a phan-
as their properties allow them to be black hole mimickers. tom one, an electrically charged traversable wormhole solu-
From historical point of view, the first description of such tion in Einstein–Maxwell-phantom dilaton gravity, has been
kind of objects begins with the issue of [1], devoted to spa- revealed [13].
tial part of Schwarzschild solution studies. The prototype Soon after the rotating wormhole solutions were paid
of wormhole emerged from the studies devoted to parti- attention to [14–16]. There were also conceived perturbative
cle model, where the mathematical construction which tried and numerical attempts to construct spinning generalization
to eliminate coordinate or curvature singularities, dubbed of static wormhole solutions [17–19]. It was claimed that the
as Einstein–Rosen bridge, was proposed in [2]. Later on, rotating wormholes would be with a higher possibility stable
the Kruskal-Szekeres coordinates were implemented for [21] and therefore traversable.
the description of Schwarzschild wormhole [3], while the The other interesting problem in wormhole physics is their
Euclidean form of wormhole solution was obtained in [4]. classification. Having in mind classification delivered by the
One should remark that all these concepts were postulated at black hole uniqueness theorem, the first work in this direc-
quantum scale. tion was provided in [22], delivering the uniqueness theorem
for wormhole spaces with vanishing Ricci scalar. Further, the
a e-mail: [email protected] (corresponding author) uniqueness of Ellis–Bronikov wormhole with phantom field
b e-mail: [email protected]

0123456789().: V,-vol 123

 586 Page 2 of 11 Eur. Phys. J. C (2022) 82:586

was found in [23], while the uniqueness for four-dimensional in binary coalescences, in the case of static and rotating
case of the Einstein–Maxwell-dilaton wormholes with the toy models of traversable wormholes, has been elucidated
dilaton coupling constant equal to one, was presented in [24]. in [29].
The case of higher-dimensional generalization of wormhole The other subject acquiring much attention in contempo-
solution, valid from the point of view of the unification theo- rary astrophysics and physics is the unrelenting search for
ries like string/M-theory attracts also attention. The unique- finding dark matter sector particles. The nature of this elu-
ness theorem for higher-dimensional case of the static spher- sive ingredient of our Universe is a mystery and several mod-
ically symmetric phantom wormholes was treated in [25], els try to explain it and constitute the possible guidance for
while the case of of static spherically symmetric traversable the future experiments. The main aim of our work will be
wormholes with two asymptotically flat ends, subject to the to investigate the behavior of axion-like particle dark matter
higher-dimensional solutions of Einstein–Maxwell-phantom model clouds, around the mimickers of rotating black holes,
dilaton field equations with an arbitrary dilaton coupling con- stationary axially symmetric wormholes. The work will pro-
stant, was elaborated in [26]. vide some continuity with our previous studies [32], where
Various other aspects of physics of these objects were we have paid attention to the main features of axionic clouds
under intensive studies (for a detailed review of the blossom- dark matter in the vicinity of magnetized rotating black holes.
ing subject the reader may consult [27]). The principal goal of the investigations will be to find
Wormholes being a fascinating subject of their possible the possible differences in characteristic features of the
impact on space and time travels, may also be regarded as axion-like condensate, between those two classes of com-
potential astrophysical objects, that can be observationally pact objects, i.e., rotating black holes and black hole mim-
search for. From the astrophysical point of view, it is per- ickers. Our studies will constitute the first glimpse at the
suasive to consider rotating wormholes. The problem that problem in question. Namely, we restrict our consideration
arises is how to observationally distinguish rotating worm- to the probe limit case, when one has the complete separation
holes from stationary axisymmetric black holes of Kerr-type. of the degrees of freedom, i.e., matter fields do not backreact
Remarkable attention to the aforementioned problem was on wormhole spacetime.
paid to after the Even Horizon Telescope observed the black The organization of the paper is as follows. In Sect. 2 we
hole shadow in the center of the galaxy M87. deliver the basic facts about the axion-like dark matter model.
The first studies to what extent wormholes can imitate the Section 3 will be devoted to the description of the rotating
observational characteristics of black holes were conducted wormholes models surrounded by dark matter clouds, in the
in [28], where the simple generalization of Schwarzschild- considered model of axion-like dark matter. In Sect. 4 we
like line element was revealed. The considered metric dif- describe the numerical results of the studies, while in Sect. 5
fers from the static general relativity one by introducing the we conclude our investigations and aim the possible problems
dimensionless parameter λ. The value of the parameter equal for the future investigations.
to zero is responsible for the ordinary Schwarzschild black
hole solution.
Of course one should be aware that for non-zero values 2 Model of axion-like dark matter sector
of the parameter the presented line element is no longer the
static solution of Einstein equations and changes the structure The explanation of astronomical and cosmological observa-
of the manifold. Therefore the matter with almost vanishing tions require dark matter existence, whose nature is one of the
energy density ought to be required to maintain the aforemen- most tantalizing questions confronting contemporary physics
tioned gravitational configuration (for the discussion of the and cosmology. A large number of ongoing or planned exper-
influence of the parameter λ on the static manifold structure imental searches for its detection and understanding of the
see, e.g., [29]). dark sector role in a fundamental description of the Universe.
Further generalization of the idea given in [28] to describe Axions are among the strongest candidates for the possible
Kerr-like wormhole spacetime as a toy model, was achieved explanation of the existence of hidden sector [33–35]. Their
by applying a modification on the Kerr metric similar to existence has been postulated to explain the apparent lack
the procedure performed in [28]. The embedding diagrams, of violation of charge conjugate parity [36–38] and in the
geodesic structure, as well as, shadow characteristics of the strong interaction motivated the absence of observable elec-
obtained Kerr-like wormhole were given in [30]. On the other tric dipole moment of the neutron [39]. Axionlike particles
hand, the throat-like effects on the shadow of Kerr-like worm- are also widely spotted in the realm of string theories [40].
holes were elaborated in [31]. In what follows, we shall study axionlike scalar particles
However, the problem of the structure at the horizon scale coupled to the Maxwell U (1)-gauge field. The non-trivial
of black hole which gives rise to echoes of the gravitational coupling of axion field to the Maxwell field strength invariant
wave signal bounded with the postmeger ring-down phase plays the crucial role in the model in question. The field

123
Eur. Phys. J. C (2022) 82:586 Page 3 of 11 586

equations of motion are provided by the variation procedure To introduce the magnetic field we use the method pro-
with respect to the action given by posed by Wald [55], where the vector potential is sourced by
  Killing vectors of the rotating spacetime. In general it has a
4 √ 1 1
S = d x −g R − Fμν F μν − ∇μ ∇ μ  form
4 2
 1
μ2 2 k μν Aμ = B(m μ + 2akμ ), (5)
−  −  ∗ F Fμν , (1) 2
2 2
where kμ and m μ are the Killing vectors connected with
where we set R for the Ricci scalar, Fμν = 2∇[μ Aν] , while temporal invariance and φ rotation respectively.
 stands for the scalar field (axion) with mass μ. ∗F μν = As in [32], where we have studied rotating magnetized
1/2μναβ F αβ is the dual to Maxwell field strength. black holes submerged into axionic dark matter cloud, one
The equation of motion for the scalar field , which con- can introduce a static magnetic field to the system, which
stitutes a covariant Klein–Gordon equation with a source will be oriented along the rotation axis. It seems to be plau-
term of the dual Maxwell field invariant, implies sible from the point of astrophysical perspective and can
k be regarded as a starting point for studies of the magnetic
∇μ ∇ μ  − μ2  − ∗ F μν Fμν = 0, (2)
2 field influence of the system in question. Because of the fact
while the U (1)-gauge field is subject to the relation as fol- that our investigations focus on static magnetic field, parallel
lows: to the wormhole rotation axis, the gauge potential may be
rewritten in the form as Aμ d x μ = B/2 gμν m ν d x μ .
∇μ F νμ + 2k ∗ F νμ ∇μ  = 0. (3) For our considerations we choose a static, time indepen-
We refer to the  field as axionlike, because the axions (orig- dent ansatz. The symmetry of the problem enables us to elab-
inating from QCD) have adequate constrains on both mass orate the axion field in the form provided by
and coupling parameter. Here however we consider particles  = ψ(r, θ ), (6)
with physics given by an analogical Lagrangian yet with arbi-
trary values of physical parameters. However for simplicity which will be plugged into the Eq. (2), for the considered
we might refer to the studied axionlike particles as simply line element.
axions.
The dark matter model in question was widely elaborated
in studies of black hole superradiance and light polarization 3 Rotating wormhole metrics
effects, possible experimental signals of dark sector around
these objects [32,41–45], and neutron stars [46–49], as well The simplicity of the static line element describing a worm-
as, the influence of axionic dark matter on the physics on hole may suggest that the spinning generalization can be
early Universe and primordial black holes [50–53]. achieved analytically and ought to be globally regular. But in
The form of the relation (2) envisages the fact that the vain, it happens that finding the stationary solution with an
presence of the non-zero source term, containing the dual extended source is far more complicated (see for the recent
invariant, given by aspects of this problem [20]). However, the rotating worm-
hole solutions are widely discussed in literature [14,19], but
I = ∗ F μν Fμν = 0, (4)
one should be aware that they do not constitute the exact solu-
is crucial. In the opposite case, when the invariant is equal to tions of the equations of motion but rather comprise some
zero, the axion-like scalar field equation of motion reduces model of geometries.
to the simple massive Klein–Gordon case, without any self- In this section, we shall study two kinds of rotating worm-
interaction potential. It means that no scalar hair configura- hole model metrics. First one accounts for the extension of
tion on the studied line element can emerge. Although it has the regular black hole Kerr metric [29,30]. The other is the
been shown that in Kerr spacetime scalar hair may emerge Teo class wormhole [14], a rotating generalization of Morris-
in certain situations [54], here we pick a different ansatz Thorne wormhole, which serves us as comparison to a bit
(see below) as we focus on stationary configurations, which more realistic Kerr-like wormhole.
appear to be magnetically induced in this approach. On the
other hand, it can be noticed that the discussed invariant, 3.1 Kerr-like wormhole
∗Fμν F μν , is equal to zero in the case when Fμν = 0, or
for spherically symmetric spacetime. However, it has a non- To begin with, we consider the metric of Kerr-like rotating
trivial form, ∗Fμν F μν = 0, when both rotation and magnetic wormhole. It is constructed by a slight modification of sta-
U (1)-gauge field components are present in the spacetime tionary axisymmetric line element with a parameter λ. For
under consideration. the first time, such construction was proposed in [28], where

123
 586 Page 4 of 11 Eur. Phys. J. C (2022) 82:586

the static Schwarzschild black hole was considered. Then, it μ2 → μ2 /η2 , r+ → ηr+ , (13)
was generalized to the case of stationary axisymmetric line
which allows us to fix one of model parameters to unity. For
element [29,30]. The Kerr-like wormhole line element yields
this we pick M = 1.
 
2Mr 4Mar sin2 θ
ds 2 = − 1 − dt 2 − dtdφ 3.2 Teo rotating wormhole

+ dr 2 + dθ 2 The well-known Morris–Thorne metric, introduced in Ref.

˜
  [5], describes a traversable wormhole spacetime, which is
2Ma 2 r sin2 θ
+ r +a +
2 2
sin2 θ dφ 2 , (7) stabilised by exotic matter in the area of its throat. That solu-
tion was achieved by using reverse engineering of general
where we set relativity, namely the metric was postulated first and with a
help of Einstein equations the suitable matter components
(r, θ ) = r 2 + a 2 cos 2 θ, (8) were found. Generalization of the aforementioned solution,
by including the rotation into the consideration, was per-
˜ (r ) = r 2 + a 2 − 2M(1 + λ2 )r. (9)
formed in [14]. The resulting metric of the rotating wormhole
has a following form:
The parameters M and a M correspond to mass and angular
momentum of a wormhole. For a small deviation parameter dr 2
ds 2 = −N 2 dt 2 +
λ, one achieves almost indistinguishable from of Kerr black 1− b
r
hole line element. These three parameters describe the sys-
tem as seen from the outside. Moreover its Arnowitt-Deser- +K 2 r 2 dθ 2 + sin2 θ (dφ − ωdt)2 , (14)
Misner (ADM) mass, as seen by the observer at asymptotic where, as in the Morris-Thorne case, one has a lot of freedom
spatial infinity, is given by M AD M = M(1 + λ2 ). in choosing the shape of N , b, K and ω functions, as long as
The largest root of ˜ (r ) = 0, establishes the surface pro- they meet specific requirements. Firstly, all the functions can
vided by be functions of r and θ and should be regular on the sym-

r+ = M(1 + λ2 ) + M 2 (1 + λ2 )2 − a 2 . (10) metry axis θ = 0, π . Secondly, N , the gravitational redshift
function, ought to be finite and nonzero, b as the shape func-
For the model in question it does not constitute a radius of tion determining the shape of the wormhole throat, should
the event horizon, but describes the radius of the throat of satisfy b  r . K accounts for the radial distance with respect
the rotating wormhole, which connects two asymptotically to the coordinate origin and ω stands for the angular velocity
flat regions of the spacetime. It can be explicitly seen by the of the wormhole.
adequate changes of variables [29,30]. The points with the The embedding of constant t and θ -cross sections in
condition r < r+ do not exist. the three-dimensional Euclidean space reveals the well-
Consequently the axion field equation written in the Kerr- recognizable form of the wormhole spacetime. The con-
like wormhole spacetime implies the following: structed geometry describes two regions, where the radial
coordinates are given by r ∈ [r+ , ∞), which are joined
2(r − M) ˜ − Mλ2 (r 2 + a 2 ) together at the wormhole throat r = r+ . At spatial infinity,
˜ ∂r2 ψ + ∂r ψ + ∂θ2 ψ
the requirement of asymptotic flatness regions provides that
k the metric coefficients ought to satisfy the following expan-
+ cot θ ∂θ ψ − μ2 ψ = IK W H , (11)
2 sions:

where the electromagnetic field invariant is provided by M 1 1 b 1

N =1− +O 2 , K =1+O , =O ,
r r r r r
2J 1
a B 2 M ˜ sin2 θ cos θ ω = 3 +O 4 , (15)
IK W H = − 4 r r
2
 6
× 3a + 2a Mr − 5a 4 r 2 − 8a 2 Mr 3
4 where we have denoted by M the mass of the wormhole and
− 32a 2 r 4 − 24r 6 + 4a 2 (a 4 − a 2 r 2 by J its angular momentum. In general, one encounters the
whole range of functions, which fulfil the aforementioned
+ 2(M − r )r 3 ) cos 2θ + a 4 (a 2 − 2Mr + r 2 ) cos 4θ .
conditions and constitute a regular rotating wormhole solu-
(12)
tion.
For the numerical calculations, we pick a set of functions
The Eq. (11) undergoes a following scaling transformation
which appear to be quite popular in the literature of the sub-
r → ηr, a → ηa, M → ηM, B → B/η, ject, and were previously used by different authors [56–60]

123
Eur. Phys. J. C (2022) 82:586 Page 5 of 11 586

The straightforward integration of the Eq. (21) appears to

γ be problematic. It is because both considered backgrounds
r+ r+ 2ar+
N = exp − , b(r ) = r+ , ω= have singular metric determinant at the throat, which makes
r r r3
simple integration from throat to infinity impossible in these
K = 1, (16)
coordinates. It should be noted that this singularity is merely
where we use the r+ symbol, for denoting the wormhole a coordinate singularity, as the curvature of both wormholes
throat radius. The angular momentum parameter is defined is regular and finite at the throat.
in the standard way a = J/M. Using the asymptotic relations In the case of Kerr-like wormhole metric, we have
(15) we find that for the picked set of functions (16) M = r+ . 
√
Thus, the family of the above solutions is described by three −g = sin2 θ, (22)
parameters, i.e., the throat radius r+ , angular momentum ˜
parameter a and the shape parameter γ . where for the case of λ equal to zero we obtain that =
After putting the ansatz (6) and the metric into the field ˜ . This fact naturally eradicates the singularity problem in
equation (20) we arrive at the equation of motion the black hole scenario. Here, however, as we radially fall
 toward the wormhole, the root of ˜ comes first and creates
r+ γ 2
r 2 − r+ r ∂r ψ + 2r + r+ the singularity. On the other hand, for the Teo rotating line
r  element we get
γ 2
r+ 1 r+ 3  
+ r+ γ − − r+ ∂r ψ √ exp − rr+ r 2 sin θ
r 2 r 2 −g =   r+ γ +1 , (23)
1 2 1− r
+ ∂θ2 ψ + cot θ ∂θ ψ − μ2 r 2 ψ = kr IT W H , (17)
2
with the denominator naturally generating the infinity.
which radial part is strongly dependent on γ . The Maxwell To deal with the integration in such spacetimes we use
field invariant related to uniform magnetic field in this space- energy differences instead. Also we introduce a cutoff to the
time implies lower integration bound, so we start from r+ +  rather than
  r γ simply r+ . In this way we ensure the finiteness of energy dif-

12a B 2 r+ cos θ sin2 θ  r − r+ r+ ferences and give them straightforward physical interpreta-
IT W H =  . (18)
r 5/2 tion. With the change of the background parameter the solu-
exp − 2rr+
tion becomes more or less thermodynamically stable with
The Eq. (17) follows a scaling transformation respect to some ground solution.

r → ηr, r+ → ηr+ , a → ηa,

B → B/η, μ2 → μ2 /η2 . (19) 4 Results
Using this transformation we fix r+ = 1.
In this section we pay attention to the solutions of the equa-
tions of motion for the previously described two toy models
3.3 Free energy of rotating wormholes. Due to the complications of the rela-
tions (11) and (17), we solve them numerically by virtue of
As a benchmark for the thermodynamical preference of the spectral methods. Firstly the adequate equation is discretized
obtained states we use free energy by evaluating the on-shell on Gauss-Lobato grid [61] and next translated into a system
action of the axion dependent part of the theory of algebraic equations with spectral differentiation matrices.
  The method in question has already been implemented in
4 √ 1 μ2 2
Saxion = d x −g − ∇μ ∇ μ  −  Python and tested for the numerical stability. The technical
2 2
 details, especially convergence tests of the numerical method
k
−  ∗ F μν Fμν . (20) are described in the Appendix of [32], where we studied the
2 problem of axionlike particle clouds in the spacetimes of
By substituting the equations of motion into the action and rotating magnetized black holes.
imposing the ansatz of the field we arrive to the formula for The spectral nature of the numerical scheme requires
the free energy remapping the coordinates onto the [−1, 1] intervals. It can
   be achieved by the coordinate transformation provided by
√
F = −2π dr dθ −g (∂r ψ)2 grr + (∂θ ψ)2 g θθ + μ2 ψ 2 .
M 2r+
(21) z =1− , (24)
r

123
 586 Page 6 of 11 Eur. Phys. J. C (2022) 82:586

4θ depicted in Fig. 1. In the following panels we see the increas-

u= − 1, (25)
π ing mass of the axionic field. In the panel (a) the field is
ultralight, subsequently in (b) μ2 = 0.01, (c) μ2 = 0.1 and
where r+ is the wormhole throat radius. After such operation,
finally in (d) μ2 = 1. In every panel we have a = 0.99 and
our numerical domain may be written in the form [−1, 1] ×
λ = 0.5. We can clearly see how the mass of the axionlike
[−1, 1]. For z-coordinate, the boundaries are the wormhole
field changes the angular distribution of it around the worm-
throat (z = −1) and spatial infinity (z = 1), while for u =
hole. For little masses the clouds are concentrated around
−1, one talks about north pole of a wormhole and the equator
the poles of the wormhole and spread in the space for sev-
with u = 1.
eral throat radii. As we increase the axion mass we see that
Consequently after the coordinate transformation in the
the polar regions of the wormhole become depleted and the
underlying equations, one shall impose the adequate bound-
field drifts towards the equator. The largest concentration is
ary conditions. Namely, on the throat surface we demand that
visible on the latitude θ π/4.
the axion field should be regular, therefore ∂r ψ = 0 provides
Second important effect is the influence of the field mass
a desirable conduct of the field. Alternatively, setting the field
on the magnitude of the field. Inspection of the colorbars
to a constant value, such as zero in a wormhole scenario, is
reveals that the larger the mass the smaller the field. The
also a possible choice. However we wish to explore the Kerr-
spatial tail of the field is also much shorter, when the mass of
like solution for different values of λ parameter, including its
the field is larger. Intuitively, in the asymptotic solution (29)
zeroing when it simplifies to the Kerr black hole. Given that
μ enters the suppressing exponential term. The field decays
for the consistency between these two kinds of solutions we
faster for larger masses, which means the massive fields are
use the Neumann boundary condition. At the spatial infinity,
localized in the vicinity of the throat surface.
we take a look on the asymptotic behaviour of the equation
Another important thing that stands out in relation to the
itself and the source term I. It appears that the Maxwell field
black hole solutions is the repulsion of the axion cloud from
invariants in both backgrounds are vanishing functions. As
the wormhole throat surface. While in the case of the black
r → ∞, we have
  hole, the field had non-zero values on the surface of the event
1 horizon, and its radial character was monotonically decreas-
IK W H = O 4 , (26)
r ing, here we have a completely different situation. For the
 
1 wormhole, the field vanishes or at least has a significantly
IT W H = O 2 . (27) smaller value on the throat. Then it grows with the radius as
r
it reaches the maximum and finally decreases. This effect is
Which means that both Eqs. (11) and (17) reach a simple, particularly visible for the high values of the angular momen-
asymptotic form, to the leading order tum.
2 The radial behaviour of the axionic field can be seen more
∂r2 ψ + ∂r ψ − μ2 ψ = 0. (28)
r precisely in Fig. 2. We present there a slice of ψ as a function
This simple equation has a solution of r in throat radius units, for constant θ = π/4 and few
different values of the λ parameter. In contrast, we also plot
exp(μr ) exp(−μr )
ψ=A +B , (29) the behaviour of axions in Kerr black hole metric (that is
r r λ = 0).
where A and B are constants. Naturally the field ought to What we can see is the increasing λ consequently extin-
decay for the sake of asymptotic flatness of the spacetime. guishes the axionic hair. In the foreground a structural change
Given that we are allowed to choose A = 0, with arbitrary in the field profile is visible as we compare it to the black
B. This means that a boundary condition ψ(r → ∞) = 0 is hole scenario. An axionic field over a black hole has a max-
an adequate and mathematically motivated choice. imum value on the event horizon. The opposite is true for a
On the other hand, the boundary conditions for the angular wormhole, on the throat the field vanishes, then grows to its
dependency are built on the basis of the spacetime symme- maximum and fades away with the radius. Then, the bigger
try. Both considered spacetimes are rotating, therefore we is λ the smaller are the maxima and overall magnitude of the
demand ∂θ ψ = 0 on the north pole. On the equator, the axionic hair.
presence of magnetic field combined with the spacetime sym- In the next step we investigate the free energy of the
metry implies that ψ = 0. obtained axion cloud configurations. It is interesting to
see how the parameters describing the spacetime geometry
4.1 Kerr-like wormhole around the wormhole influence the thermodynamics of the
axion clouds. Due to the previously mentioned difficulties in
To commence with, we solve the Eq. (11) for the Kerr-like computing the free energy in these metrics, we rather talk
background metric. A portion of obtained distributions is about energy differences, than the exact values. In Fig. 3

123
Eur. Phys. J. C (2022) 82:586 Page 7 of 11 586

(a) μ2 → 0+ , λ = 0.5 (b) μ2 = 0.01, λ = 0.5

(c) μ2 = 0.1, λ = 0.5 (d) μ2 = 1, λ = 0.5

Fig. 1 Axion field distribution around Kerr-like wormholes for given field appears to be non-zero on the event horizon. Subsequent panels
sets of parameters. The blank space in the vicinity of wormhole throat for each mass parameter show how the angular distribution of the field
distinguishes the solution from the black hole counterpart, where the is affected

we present the differences of the free energy versus angu-

0.8
λ lar momentum a, with respect to the λ = 0 level, which
0.0 constitutes a plain Kerr black hole. It is clearly visible that
0.6 0.33
the larger value of the distortion parameter λ one takes into
ψ(θ = π/4)

0.66
1.0 account, the higher value of the free energy of the cloud we
0.4 1.33 achieve. It turns out, that the more the gravitational back-
1.66 ground deviates from from the black hole metric, the less
0.2 thermodynamically desirable axion clouds are. This effect
works together with the diminishing magnitude of the field
0.0 on the previously discussed Fig. 2. Additionally the increas-
0 10 20 30 40 50 ing angular momentum of the wormhole also increases the
r/r+ free energy difference. This means that for extreme Kerr-like
wormhole axion hair is the least favourable.
Fig. 2 A closer look on the axion cloud gap near the wormhole throat.
Here we show slices of ψ for constant θ = π/4, with parameters
a = 0.99, μ2 = 0+ . Increasing of λ decreases the magnitude of the
axions and cuts off its tail

123
 586 Page 8 of 11 Eur. Phys. J. C (2022) 82:586

3.0 4.2 Teo rotating wormhole

λ
2.5 0.25
0.5 Teo class wormhole has a different set of parameters and it
2.0 0.75 does not simply transform into a black hole solution, just like
ΔF

1.0
1.5 a Kerr-like wormhole does. Here the throat radius is indepen-
1.0
dent of the other parameters and is imposed manually in gtt
and grr . With the particular choice of functions (16), we can
0.5
only steer with the shape of metric components via γ param-
0.0 eter. Therefore, let us consider values of γ in the interval
0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
(−1, 1], where for γ = −1 the function grr is singular, so
a
we can only approach this value.
Fig. 3 Free energy differences as a function of angular momentum As it was mentioned, the Kerr-like wormhole can be
(with λ = 0 as the ground curve) for different values of λ. With axion reduced to a black hole solution by setting λ = 0. Teo solu-
mass μ2 = 0+ we see that the cloud thermodynamical favourability
tion does not share this feature, but is a well-known wormhole
decreases with growth of both angular momentum and λ parameter
metric, just like its non-rotating counterpart the archetypi-

(a) μ2 → 0+ , γ = −0.99 (b) μ2 = 0.1, γ = −0.99

(c) μ2 = 0.5, γ = −0.99 (d) μ2 = 1, γ = −0.99

Fig. 4 Axion cloud distribution around the magnetized Teo type worm- again a big mass results with a localized field, concentrated closely to
hole. The negative value of γ resembles the results for Kerr-like worm- the throat with depleted polar region. For bigger values of γ we observe
hole, with large λ value, i.e., the decreasing of the field magnitude. Each similar influence of μ on the spatial distribution
panel depicts the field distribution for different axionlike masses. Once

123
Eur. Phys. J. C (2022) 82:586 Page 9 of 11 586

cal Morris-Thorne wormhole. While it can not serve as a 1.2

testing field for differences between axion clouds around
1.0 γ
wormholes and black holes, one can treat it as a benchmark −0.99

ψ(θ = π/4)
for behaviours of the axion hair in another wormhole envi- 0.8 −0.53
−0.03
ronment. Using this background might help us to see if the 0.6 0.99
obtained axion solutions share similar features. 0.4
2.0

In Fig. 4 one can see the distribution of the axionic cloud

0.2
around a Teo wormhole for different axionlike field masses.
In the panel (a) we have an ultralight field, then it takes values 0.0
0 5 10 15 20 25 30
0.1, 0.5 and 1.0 for (b), (c) and (d) respectively. In all panels
r/r+
we use γ = −0.99, which gives us tiny value of the axionic
field (see the colorbars). Fig. 5 Radial slices of the axion field for constant angle θ = π/4. The
The angular distribution has similar features to the Kerr- growth of γ increases the maximum value of the field. However in this
like metric. For ultralight axions the field is localized in metric the significant growth of the blank space near the throat is not
present. Also the growth of γ does not seem to greatly affect the tail
the majority of the wormhole surroundings. As the mass of the field, away from the throat, which is different from the Kerr-like
increases the hair tightens spatially and disappears from the wormhole results
polar regions. For the large mass case the axionic clouds are
drifting toward the equator with the polar caps left almost
empty. Moreover the radial reach is very short - around one 25 γ
throat radius. −0.99
20 −0.5
One can clearly notice that for the negative gamma value 1.0
the analogous effect to the Kerr-like scenario is observed. The 15 2.0
ΔF

axionic cloud is also pushed away from the throat surface - in 10

its vicinity the field acquires small values, reaches the max-
5
imum and the descends monotonically to zero. However in
this gravitational background this effect is not as dramatic as 0

in case of Kerr-like wormhole. The weakening of the axionic 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
field in the vicinity of the throat is easily visible in the distri- a
butions, although it is not that large. Increasing γ up to zero
Fig. 6 Free energy differences vs. angular momentum for different
and beyond causes the rise of ψ field. It grants bigger values, values of γ . The curve of γ = 0 is the reference level. We see that free
but the spatial qualitative characteristic remains intact. energy increases for negative values of gamma and slightly drops for
The next figure brings a closer look on the field drop near positive ones, as angular momentum grows
the throat. In Fig. 5 we present radial slices of the field distri-
bution with θ = π/4. In this particular figure we depict the
behaviour for ultralight field, however a similar tendencies Free energy difference curves for γ < 0 are positive,
are shown by more massive fields. First of all, we observe especially the γ = −0.99 curve reaches relatively big values.
a significant amplification of the field with the growth of γ . Therefore thermodynamically speaking wormholes with γ
The field does not acquire new features however, but it seems close to −1 have least chances to hold axionic hair. With
that the curves follow some kind of scaling related to gamma. the growth of the parameter, the free energy of the cloud
Additionally the profiles resemble the results obtained for decreases, which makes the axions more thermodynamically
Kerr-like wormholes. As we have previously mentioned, Teo favourable. However this fall is rather moderate comparing
class wormhole cannot be simply transformed into a black to the rise of the top curve.
hole by a simple choice of parameter value. However the In both cases the increase of angular momentum ampli-
features like a drop near the throat surface, then maximum fies the tendencies of the curves. Curves for negative gamma
and monotonic fall show that these might be more general grow, while the positive fall. A consequent growth of γ
wormhole related behaviours of axionic hair. parameter leads the hair to some limit characteristics which
If we consider the free energy, it appears that the axion can be seen in Figs. 5 and 6.
clouds for the background with negative γ , are definitely Finally let us conduct qualitative comparison of the
less thermodynamically favourable. In Fig. 6 we plot the free axionic clouds in considered metrics. The solutions have
energy difference as a function of angular momentum for undoubtedly similar features, especially when one takes a
several values of the gamma parameter. We use the curve look on the ψ slices. We also observe the separation of the
with γ = 0 as a baseline for calculating energy differences. cloud from the surface of the throat in both cases. This allows
us to notice some general wormhole related phenomena,

123
 586 Page 10 of 11 Eur. Phys. J. C (2022) 82:586

which are not present around the black holes [32]. Naturally Open Access This article is licensed under a Creative Commons Attri-
we cannot speak in a fully general manner, as we only con- bution 4.0 International License, which permits use, sharing, adaptation,
distribution and reproduction in any medium or format, as long as you
sidered here merely two distinct gravitational backgrounds, give appropriate credit to the original author(s) and the source, pro-
which on the other hand should be treated as toy models. vide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article
are included in the article’s Creative Commons licence, unless indi-
cated otherwise in a credit line to the material. If material is not
5 Conclusion included in the article’s Creative Commons licence and your intended
use is not permitted by statutory regulation or exceeds the permit-
In our paper, we have considered the problem of the distri- ted use, you will need to obtain permission directly from the copy-
bution of axionlike particles, being regarded as dark matter right holder. To view a copy of this licence, visit http://creativecomm
ons.org/licenses/by/4.0/.
sector model, around the toy models of rotating wormholes. Funded by SCOAP3 . SCOAP3 supports the goals of the International
We have investigated the Kerr-like wormhole line element Year of Basic Sciences for Sustainable Development.
with the distortion parameter λ and Teo model of a rotating
axisymmetric wormhole. The models under inspection were
characterized by the mass, angular momentum, distortion
parameter (for Kerr-like wormhole) and the shape parameter References
γ (for the Teo model). We numerically solve the equations
of motion for the underlying cases, using spectral methods. 1. L. Flamm, Beiträge zur Einsteinischen Gravitationstheorie. Phys.
Z. 17, 448 (1916)
Among all we have found that the axion clouds are pushed 2. A. Einstein, N. Rosen, The particle problem in the general theory
forward from the wormhole throat, especially for the case of relativity. Phys. Rev. 48, 73 (1935)
of large value of the rotation parameter a. The voids in the 3. J.A. Wheeler, Geons. Phys. Rev. 97, 511 (1955)
vicinity of the wormhole throat appear for the larger value of 4. S.W. Hawking, Wormholes in space-time. Phys. Rev. D 37, 904
(1988)
the distortion parameter λ. This phenomenon distinguishes 5. M.S. Morris, K. Thorne, Wormholes in space-time and their use
the studied system from the previously elaborated black holes for interstellar travel: A tool for teaching general relativity. Am. J.
[32]. On the other hand, for the larger λ, one achieves higher Phys. 56, 395 (1988)
value of the free energy, and therefore this solution is less 6. H.G. Ellis, Ether ow through a drainhole: A particle model in gen-
eral relativity. J. Math. Phys. 14, 104 (1973)
thermodynamically favoured. 7. K. Bronikov, Scalar-Tensor Theory and Scalar Charge. Acta Phys.
As far as the Teo class of rotating wormholes is concerned, Polonica B 4, 251 (1973)
we have for the negative value of γ the analogous effect as 8. H.G. Ellis, The evolving, owless drainhole: A nongravitating-
in the latter case is obtained. However for the positive value, particle model in general relativity theory. Gen. Rel. Grav. 10, 105
(1979)
the behavior of axionic clouds resembles features of the dark 9. P. Kanti, B. Kleinhaus, J. Kunz, Wormholes in dilatonic Einstein–
matter clouds around Kerr black hole in a uniform magnetic Gauss–Bonnet Theory. Phys. Rev. Lett. 107, 271101 (2011)
field. The solution with negative γ is not thermodynamically 10. T. Harko, F.S.N. Lobo, M.K. Mak, S.V. Sushkov, Modifíed-gravity
favourable, as it has been revealed in free energy studies. wormholes without exotic matter. Phys. Rev. D 87, 067504 (2013)
11. G.W. Gibbons, M.S. Volkov, Ring wormholes via duality rotations.
However when γ increases, the free energy of the axionic Phys. Lett. B 760, 324 (2016)
cloud decreases. 12. G. W. Gibbons and M. S. Volkov, Weyl metrics and wormholes,
We have found that the behavior of the axionic clouds JCAP 05 (039) 2017
significantly differs from the black hole scenario, which we 13. P. Goulart, Phantom wormholes in Einstein-Maxwell-dilaton the-
ory. Class. Quant. Grav. 35, 025012 (2018)
discussed in our previous work [32]. This fact will account for 14. E. Teo, Rotating tranversable wormholes. Phys. Rev. D 58, 024014
the possible guidance, enabling one to distinguish between (1998)
these two classes of compact objects. Nevertheless, the search 15. G. Clement, A class of wormhole solutions to higher-dimensional
for astronomically observable criteria require far more com- general relativity. Gen. Rel. Grav. 16, 131 (1984)
16. K.A. Bronikov, V.G. Krechet, J.P.S. Lemos, Rotating cylindrical
plex approach. A more realistic dynamical gravitational wormholes. Phys. Rev. D 87, 084060 (2013)
model is needed, when the time dependence of the studied 17. P.E. Kashargin, S.V. Sushkov, Slowly rotating scalar fíeld worm-
fields is taken into account, as well as, the direct mathemati- holes: The second order approximation. Phys. Rev. D 78, 064071
cal proofs of the stabilities of rotating wormhole spacetimes (2008)
18. B. Kleinhaus, J. Kunz, Rotating Ellis wormholes in four dimen-
ought to be found. These subjects impose a real mathematical sions. Phys. Rev. D 90, 121503 (2014)
challenge and require also solid numerical relativity machin- 19. X.Y. Chew, B. Kleinhaus, J. Kunz, Geometry of spinning Ellis
ery. These problems shall be investigated elsewhere. wormholes. Phys. Rev. D 94, 104031 (2016)
20. M.S. Volkov, Stationary generalization for Bronnikov-Ellis worm-
Data Availability Statement This manuscript has no associated data hole and for the vacuum ring wormhole. Phys. Rev. D 104, 124064
or the data will not be deposited. [Authors’ comment: The numerical (2021)
code is not ready yet to be published. However, it is available upon 21. T. Matos, D. Nunez, Rotating scalar fíeld wormhole. Class. Quant.
Reader’s request.] Grav. 23, 4485 (2006)

123
Eur. Phys. J. C (2022) 82:586 Page 11 of 11 586

22. P. J. Ruback, A uniqueness theorem for wormholes in quantum 43. T. Ikeda, R. Brito, V. Cardoso, Blasts of light from axions. Phys.
gravity. Class. Quant. Grav. 6, L21 (1989) Rev. Lett. 122, 081101 (2019)
23. S. Yazadjiev, Uniqueness theorem for static wormholes in Einstein 44. M. Boskovic, R. Brito, V. Cardoso, T. Ikeda, H. Witek, Axionic
phantom scalar fíeld theory. Phys. Rev. D 96, 044045 (2017) instabilities and new black hole solutions. Phys. Rev. D 99, 035006
24. B. Lazov, P. Nedkova, S. Yazadjiev, Uniqueness theorem for static (2019)
phantom wormholes in Einstein-Maxwell-dilaton theory. Phys. 45. V. Cardoso, O.J.C. Dias, G.S. Harnett, M. Middleton, P. Pani, J.E.
Lett. B 778, 408 (2018) Santos, Constraining the mass of dark photons and axion-like par-
25. M. Rogatko, Uniqueness of higher-dimensional phantom fíeld ticles through black-hole superradiance. JCAP 03, 043 (2018)
wormholes. Phys. Rev. D 97, 024001 (2018) 46. B. Garbrecht and J. I. Mc Donald, Axion confígurations around
26. M. Rogatko, Uniqueness of higher-dimensional Einstein– pulsars. JCAP 07, 044 (2018)
Maxwell–phantom dilaton fíeld wormholes. Phys. Rev. D 97, 47. J. Darling, New limits on axionic dark matter using the magnetar
064023 (2018) PSR J1745–2900. Astrophys. J. Lett. 900, L28 (2020)
27. M. Visser, Lorentzian Wormholes, New York, American Institute 48. J. Darling, Search for axion dark matter using the magnetar PSR
of Physics 1995, F. S. N. Lobo ed., Wormholes, warp drives and J1745–2900. Phys. Rev. Lett. 125, 121103 (2020)
energy conditions, New York, Springer Fundamentals in Theor. 49. P. W. Graham, I. G. Irastorza, S. K. Lamoreaux, A. Lindner, K. A.
Phys., 2017 van Bibber, Experimental searches for axion and axion-like parti-
28. T. Damour, S.N. Solodukhin, Wormholes as black holes foils. Phys. cles. Annu. Rev. Nucl. Part. Sci. 65, 485 (2015)
Rev. D 76, 024016 (2007) 50. M. A. Fedderke, P. W. Graham, S. Rajendram, Axion dark matter
29. P. Bueno, P.A. Cano, F. Goelen, T. Hertog, B. Vercnocke, Echoes detection with CMB polarization. Phys. Rev. D 100, 015040 (2019)
of Kerr-like wormholes. Phys. Rev. D 97, 024040 (2018) 51. R. T. Co, A. Pierce, Z. Zhang, Y. Zhao, Dark photon dark matter
30. M. Amir, K. Jusufí, A. Banerjee, S. Hansraj, Shadow images of produced by axion oscillations. Phys. Rev. D 99, 075002 (2019)
Kerr-like wormholes. Class. Quant. Grav. 36, 215007 (2019) 52. S. Sen, Plasma efíects on lasing of a uniform ultralight axion con-
31. S. Kasuya, M. Kobayashi, Throat efíects on shadows of Kerr-like densate. Phys. Rev. D 98, 103012 (2018)
wormholes. Phys. Rev. D 103, 104050 (2021) 53. J. G. Rosa, T. W. Kephart, Stimulated axion decay in superradiant
32. B. Kiczek, M. Rogatko, Axionlike dark matter clouds around rotat- clouds around primordial black holes. Phys. Rev. Lett. 120, 231102
ing black holes. Phys. Rev. D 103, 124021 (2021) (2018)
33. J. Preskill, M.B. Wise, F. Wilczek, Cosmology of the invisible 54. C. A. R. Herdeiro, E. Radu, Kerr Black Holes with Scalar Hair.
axion. Phys. Lett. B 120, 127 (1983) Phys. Rev. Lett. 112, 221101 (2014)
34. L.F. Abbott, P. Sikivie, Cosmological bound on the invisible axion. 55. R. M. Wald, Black hole in a uniform magnetic fíeld. Phys. Rev. D
Phys. Lett. B 120, 133 (1983) 10, 1680 (1974)
35. M. Dine, W. Fischler, The not so harmless axion. Phys. Lett. B 120, 56. R. Shaikh, Shadows of rotating wormholes. Phys. Rev. D 98,
137 (1983) 024044 (2018)
36. R. D. Peccei, H. R. Quinn, CP conservation in the presence of 57. P. G. Nedkova, V. Tinchev, S. S. Yazadjiev, Shadow of a rotating
pseudoparticles. Phys. Rev. Lett. 38, 1440 (1977) traversable wormhole. Phys. Rev. D 88, 124019 (2013)
37. S. Weinberg, A new light boson? Phys. Rev. Lett. 40, 223 (1978) 58. A. Abdujabbarov, B. Juraev, B. Ahmedov, Z. Stuchlik, Shadow of
38. F. Wilczek, Problem of strong P and T invariance in the presence rotating wormhole in plasma environment. Astrophys. Space Sci.
of instantons. Phys. Rev. Lett. 40, 279 (1978) 361, 226 (2016)
39. J. M. Pendlebury et al., Revised experimental upper limit on the 59. T. Harko, Z. Kovacs, F. Lobo, Thin accretion disks in station-
electric dipole moment of the neutron. Phys. Rev. D 92, 092003 ary axisymmetric wormhole spacetimes. Phys. Rev. D 79, 064001
(2015) (2009)
40. P. Svrcek, E. Witten, Axions in string theory. JHEP 06, 051 (2006) 60. C. Bambi, Broad Kfí iron line from accretion disks around
41. A. D. Plascencia, A. Urbano, Black hole supperadiance and traversable wormholes. Phys. Rev. D 87, 084039 (2013)
polarization-dependent bending of light. JCAP 04, 059 (2018) 61. N. Lloyd, Trefethen (Spectral methods in MATLAB, SIAM,
42. X. Bi, Y. Gao, J. Guo, N. Houston, T. Li, F. Xu, and X. Zhang, Axion Philadelphia, 2000)
and dark photon limits from Crab Nebula high energy gamma-rays.
arXiv:2002.01796 (2020)

123
CHAPTER 4. ORIGINAL CONTRIBUTIONS 113

4.3 Other activities

During my doctoral studies I also participated in different research, engineering and R&D
projects, which might not be fully connected with the main topic of this thesis. However,
for completeness I shall list them here.

4.3.1 Conference presentations

1. The 6th Conference of the Polish Society on Relativity (POTOR-6), Szczecin, Poland,
23-26.09.2019 (talk).

2. New Trends in Physics, Paris, France, 28-29.11.2019 (poster).

3. Virtual Conference of the Polish Society on Relativity 2020 (Virtual POTOR), online,
24-26.09.2020 (talk).

4. Alternative Gravities and Fundamental Cosmology AlteCosmoFun ’21, online, 6-10.09.2021

(talk).

5. The 28th Annual International Conference On Mobile Computing And Networking

(MobiCom 2022), Sydney, Australia, 17–21 Oct 2022 (poster).

4.3.2 Other papers in physics

1. B. Kiczek, M. Rogatko and K. I. Wysokinski, Anomalous Hall conductivity of the
holographic Z2 Dirac semimetals, Phys. Rev. D 104, 086022 (2021).

4.3.3 Papers in computer science

1. D. Wójcik, B. Stefaniak, M. Woś, B. Kiczek, T. Rymarczyk, Image reconstruction
for lung monitoring in wearable electrical impedance tomography, Przegląd elektrotech-
niczny, ISSN 0033-2097, r. 98 nr 3/2022.

2. B. Kiczek, M. Gołąbek, D. Wójcik, K. Kania, E. Kozłowski, T. Rymarczyk, and J.

Sikora, A wearable ultrasonic bladder monitoring device, In Proceedings of the 28th
Annual International Conference on Mobile Computing And Networking (MobiCom
’22). Association for Computing Machinery, New York, NY, USA, 886–888.

3. B. Kiczek, D. Wójcik, M. Oleszek, T. Rymarczyk, J. Sikora, B. Baran, B. Przysucha,

LETS - a wearable heart and lung monitoring device for the diagnosis of cardiac and
respiratory diseases, To be published in proceedings of Ubicomp 2022.

4. B. Kiczek, D. Wójcik, M. Oleszek, T. Rymarczyk, K. Król, E. Kozłowski, Electrical

impedance tomography bladder monitoring wearable device, To be published in proceed-
ings of ISWC 2022.
CHAPTER 4. ORIGINAL CONTRIBUTIONS 114

5. T. Rymarczyk, G. Kłosowski, P. Adamkiewicz, M. Styła, B. Kiczek, Use of a Long

Short-Term Memory Network in Radio Tomography to Track People Indoors, To be
published in proceedings of Sensys 2022.

6. M. Styła, B. Kiczek, et al., Machine Learning-Enhanced Radio Tomographic Device

for Energy Optimization in Smart Buildings, Energies 2023, 16(1), 275

4.4 Future investigations

The results obtained within this thesis can be treated as a starting point for some future
investigations. There are multiple directions that can be taken, let me list few examples.

Study of the backreaction Most of the dark matter structures in this thesis were ob-
tained in the probe limit. It means only equations of matter were solved with the gravi-
tational background being fixed. However, this should be extended on the analysis of the
impact of dark sector clouds on the metric. It could be a quite challenging task, especially
if one considers a rotating spacetime. Having this, one can obtain several quantities which
can be observed with telescopes.
For instance, one can study light rays bending by a compact object hosting an axion cloud.
Similarly, the geodesics of charged particles forming the accretion disk can be examined.
Finally, a timely topic, one can verify how a dark matter structure impacts the shape of the
black hole shadow. These quantities can be compared to dark matter free scenarios, where
the key differences can be pointed.

Time dependent simulations In works [73, 74] we have shown the existence of stationary
axion cloud solutions around magnetised black holes and wormholes. Nevertheless, this is
merely a beginning, as there are lots of things that should be tested concerning that setup.
The time dependence ought to be included at first. Then one can look for physical pro-
cesses that would lead to formation of such axion clouds. On the other hand, the stationary
clouds can be tested under various perturbations, such as changes of magnetic field.

Stellar evolution It is particularly interesting to see if an addition of some dark matter

candidates into the stellar plasma can impact the evolution of a star. By using the equation
of state and TOV equations one could model these phenomena. Following questions may
arise: How would it affect the ageing process? How the addition of DM would distort star’s
path on Hertzsprung–Russell diagram?
There are, naturally, more examples that could be brought here. However this thesis
shall end somewhere, and this is a good place.

4.5 Co-author statements

Bibliography

[1] F. Zwicky, The Redshift of Extragalactic Nebulae, Helvetica Physica Acta, Vol. 6, p.
110-127, 1933.
[2] V. C. Rubin and W. K. Ford, Radial Velocities and Line Strengths of Emission Line
Across the Nuclear Disk of M31, Astrophys. J. 170, 25-52 (1971).
[3] Planck Collaboration (2016), Planck 2015 results. XIII. Cosmological parameters, As-
tron. Astrophys. 594, 13:A13 (2016).
[4] G. Bertone and T. M. P. Tait, A New Era in the Quest for Dark Matter, Nature 562,
51 (2018).
[5] R. D. Peccei and H. R. Quinn, CP Conservation in the Presence of Pseudoparticles,
Phys. Rev. Lett. 38, 1440 (1977).
[6] J. Preskill, M. B. Wise, F. Wilczek, Cosmology of the Invisible Axion, Phys. Lett. B
120, 127 (1983).
[7] L. F. Abbott, P. Sikivie, A Cosmological Bound on the Invisible Axion, Phys. Lett. B
120, 133 (1983).
[8] M. Dine, W. Fischler, The Not So Harmless Axion, Phys. Lett. B 120, 137 (1983).
[9] F. Chadha-Day, J. Ellis, D. J. E. Marsh, Axion dark matter: What is it and why now?,
Sci. Adv. 8, eabj3618 (2022).
[10] S. Weinberg, A New Light Boson?, Phys. Rev. Lett. 40, 223 (1978).
[11] F. Wilczek, Problem of Strong P and T Invariance in the Presence of Instantons, Phys.
Rev. Lett. 40, 279 (1978).
[12] E. Witten, Some Properties of O(32) Superstrings, Phys. Lett. B 149, 351 (1984).
[13] J. Jaeckel and A. Ringwald, The Low-Energy Frontier of Particle Physics, Annu. Rev.
Nucl. Part. Sci. 2010. 60:405-37.
[14] K. Ehret et al., New ALPS results on hidden-sector lightweights, Phys. Lett. B 689,
149-155 (2010).

120
BIBLIOGRAPHY 121

[15] T. Braine et al. (ADMX Collaboration), Extended Search for the Invisible Axion with
the Axion Dark Matter Experiment, Phys. Rev. Lett. 124, 101303 (2020).

[16] E. Armengaud et al, Conceptual design of the International Axion Observatory (IAXO),
JINST 9 (2014) T05002.

[17] CAST Collaboration, New CAST limit on the axion–photon interaction, Nature Phys.
13, 584–590 (2017).

[18] A. Abeln et al, Conceptual Design of BabyIAXO, the intermediate stage towards the
International Axion Observatory, arXiv: physics.ins-det 2010.12076 (2020).

[19] J. Galan, Axion search with BabyIAXO in view of IAXO, arXiv: physics.ins-det
2012.06634 (2020).

[20] K. Schutz and K. M. Zurek, Detectability of Light Dark Matter with Superfluid Helium,
Phys. Rev. Lett. 117, 121302 (2016).

[21] Y. Hochberg et al., Detecting ultralight bosonic dark matter via absorption in supercon-
ductors, Phys. Rev. D 94, 015019 (2016).

[22] S. Griffin et al., Directional detection of light dark matter with polar materials, Phys.
Rev. D 98, 115034 (2018).

[23] J. Gooth et al., Axionic charge-density wave in the Weyl semimetal (T aSe4 )2I, Nature
575, 315 (2019).

[24] T. Ikeda et al., Blasts of Light from Axions, Phys. Rev. Lett. 122, 081101 (2019).

[25] M. Boskovic et al., Axionic instabilities and new black hole solutions, Phys. Rev. D 99,
035006 (2019).

[26] T. Fujita et al., Hunting Axion Dark Matter with Protoplanetary Disk Polarimetry,
Phys. Rev. Lett. 122, 191101 (2019).

[27] M. A. Fedderke, P. W. Graham, and S. Rajendram, Axion dark matter detection with
CMB polarization, Phys. Rev. D 100, 015040 (2019).

[28] B. Garbrecht and J. I. Mc Donald, Axion configurations around pulsars, J. Cosmol.

Astropart. Phys. 07 (2018) 044.

[29] J. Darling, New limits on axionic dark matter using the magnetar PSR J1745-2900,
Astrophys. J. Lett. 900 (2020) L28.

[30] J. Darling, Search for axion dark matter using the magnetar PSR J1745-2900, Phys.
Rev. Lett. 125, 121103 (2020).

[31] B. Holdom, Two U (1)’s and ϵ charge shifts, Phys. Lett. B 166, 196-198 (1986).
BIBLIOGRAPHY 122

[32] M. Fabbrichesi, E. Gabrielli, G. Lanfranchi, The Physics of the Dark Photon: A Primer,
SpringerBriefs in Physics, Springer 2021.

[33] W. DeRocco, P. W. Graham, D. Kasen, G. Marques-Tavares and S. Rajendran, Ob-

servable signatures of dark photons from supernovae, J. High Energy Phys. 02 (2019)
171.

[34] G. G. Raffelt, Stars as laboratories for fundamental physics, Chicago Univ. Pr., Chicago,
IL, USA (1996).

[35] M. D. Diamond and G. Marques-Tavares, γ-Ray Flashes from Dark Photons in Neutron
Star Mergers, Phys. Rev. Lett. 128, 211101 (2022).

[36] B. S. Acharya, S. A. R. Ellis, G. L. Kane, B. D. Nelson, and M. J. Perry, Lightest

Visible-Sector Supersymmetric Particle is Likely to be Unstable, Phys. Rev. Lett. 117,
181802 (2016).

[37] R. Bernabei et al., Searching for WIMPs by the annual modulation signature, Phys.
Lett. B 424, 195 (1998).

[38] R. Bernabei et al., Final model independent result of DAMA/LIBRA-phase 1, Eur.

Phys. J. C 73, 2648 (2013).

[39] COSINE-100 collaboration, An experiment to search for dark-matter interactions using

sodium iodide detectors, Nature 564, 83 (2018).

[40] A. Geringer-Sameth et al., Indication of Gamma-Ray Emission from the Newly Discov-
ered Dwarf Galaxy Reticulum II, Phys. Rev. Lett. 115, 081101 (2015).

[41] P. Jean et al., Early SPI/INTEGRAL measurements of 511 keV line emission from the
4th quadrant of the Galaxy, Astron. Astrophys. 407, L55 (2003).

[42] J. Chang et al., An excess of cosmic ray electrons at energies of 300-800 GeV, Nature
456, 362 (2008).

[43] O. Adriani et al. (PAMELA Collaboration), An anomalous positron abundance in cosmic

rays with energies 1.5-100 GeV, Nature 458, 607 (2009).

[44] K. K. Boddy and J. Kumar, Indirect detection of dark matter using MeV-range gamma-
ray telescopes, Phys. Rev. D 92, 023533 (2015).

[45] K. Van Tilburg et al., Search for Ultralight Scalar Dark Matter with Atomic Spec-
troscopy, Phys. Rev. Lett. 115, 011802 (2015).

[46] J. H. Chang, R. Essig, and S. D. McDermott, Revisiting Supernova 1987A constraints

on dark photons, J. High Energy Phys. 01 (107) 2017.
BIBLIOGRAPHY 123

[47] J. P. Lees et al., Search for a Dark Photon in e+ e− Collisions at BaBar, Phys. Rev.
Lett. 113, 201801 (2014).

[48] L. Barak et al. (SENSEI Collaboration), SENSEI: Direct-Detection Results on sub-GeV

Dark Matter from a New Skipper CCD, Phys. Rev. Lett. 125, 171802 (2020).

[49] E. Aprile et al. (XENON Collaboration), Excess electronic recoil events in XENON1T,
Phys. Rev. D 102, 072004 (2020).

[50] W. Israel, Event Horizons in Static Vacuum Space-Times, Phys. Rev. 164, 1776 (1967).

[51] B. Carter, Axisymmetric Black Hole Has Only Two Degrees of Freedom, Phys. Rev.
Lett. 26, 331 (1971).

[52] S. W. Hawking and G. F. R. Ellis, The Large Scale Structure of Space-Time, Cambridge
University Press, Cambridge, England, 1973.

[53] C. A. R. Herdeiro and E. Radu, Asymptotically flat black holes with scalar hair: A
review, Int. J. Mod. Phys. D 24, 1542014 (2015).

[54] J. E. Chase, Event horizons in static scalar-vacuum space-times, Commun. Math. Phys.
19, 276-288 (1970).

[55] J. D. Bekenstein, Transcendence of the Law of Baryon-Number Conservation in Black-

Hole Physics, Phys. Rev. Lett. 28, 452 (1972).

[56] M. Heusler and N. Strauman, Scaling arguments for the existence of static, spherically
symmetric solutions of self-gravitating systems, Class. Quantum Grav. 9, 2177 (1992).

[57] J. D. Bekenstein, Novel “no-scalar-hair” theorem for black holes, Phys. Rev. D 51,
R6608(R) (1995).

[58] I. Peña and D. Sudarsky, Do collapsed boson stars result in new types of black holes?,
Class. Quantum Grav. 14, 3131 (1997).

[59] M. Heusler, S. Droz, N. Straumann, Stability analysis of self-gravitating skyrmions,

Phys. Lett. B 271, 61-67 (1991).

[60] E. Berti et al., Spin-Induced Black Hole Scalarization in Einstein-Scalar-Gauss-Bonnet

Theory, Phys. Rev. Lett. 126, 011104 (2021).

[61] C. A. R. Herdeiro et al., Spin-Induced Scalarized Black Holes, Phys. Rev. Lett. 126,
011103 (2021).

[62] J. M. Maldacena, The Large N Limit of Superconformal Field Theories and Supergravity,
Adv. Theor. Math. Phys. 2, 231-252 (1998).
BIBLIOGRAPHY 124

[63] J. D. Brown and J. W. York, Quasilocal energy and conserved charges derived from the
gravitational action, Phys. Rev. D 47, 1407-1419 (1993).

[64] S. A. Hartnoll et al., Building a Holographic Superconductor, Phys. Rev. Lett. 101,
031601 (2008).

[65] R. M. Wald, General Relativity, The University of Chicago Press, Chicago and London,
1984.

[66] John P. Boyd, Chebyshev & Fourier Spectral Methods, DOVER Publications, Inc., New
York, 2000.

[67] Lloyd N. Trefethen, Spectral Methods in MATLAB, SIAM, Philadelphia, 2000.

[68] A. Einstein and N. Rosen, The particle problem in the general theory of relativity, Phys.
Rev. 48, 73 (1935).

[69] M. S. Morris and K. Thorne, Wormholes in space-time and their use for interstellar
travel: A tool for teaching general relativity, Am. J. Phys. 56 (1988) 395.

[70] B. Kiczek and M. Rogatko, Ultra-compact spherically symmetric dark matter charged
star objects, J. Cosmol. Astropart. Phys. 09 (2019) 049.

[71] B. Kiczek and M. Rogatko, Influence of dark matter on black hole scalar hair, Phys.
Rev. D 101, 084035 (2020).

[72] B. Kiczek, M. Rogatko and K. I. Wysokiński, Holographic DC SQUID in the presence

of dark matter, J. Cosmol. Astropart. Phys. 01 (2021) 063.

[73] B. Kiczek and M. Rogatko, Axionlike dark matter clouds around rotating black holes,
Phys. Rev. D 103, 124021 (2021).

[74] B. Kiczek and M. Rogatko, Static axionlike dark matter clouds around magnetized
rotating wormholes - probe limit case. Eur. Phys. J. C 82, 586 (2022).

[75] B. Kiczek, M. Rogatko and K. I. Wysokiński, Anomalous Hall conductivity of the

holographic Z2 Dirac semimetals, Phys. Rev. D 104, 086022 (2021).