Comenius University in Bratislava

Faculty of Mathematics, Physics and

Informatics

Universe expansion and dark matter

freeze-out in 1+1 dimension

Bachelor thesis

Alºbeta Miklá²ová

Bratislava 2020
 Comenius University in Bratislava

Faculty of Mathematics, Physics and

Informatics

Universe expansion and dark matter

freeze-out in 1+1 dimension

Bachelor thesis

Study program: Physics

Field of study: 4.1.1 Physics

Supervisor: Mgr. Peter Maták, PhD.

Department: Theoretical Physics

Alºbeta Miklá²ová

Bratislava 2020
 Comenius University in Bratislava
Faculty of Mathematics, Physics and Informatics

THESIS ASSIGNMENT

Name and Surname: Alžbeta Miklášová

Study programme: Physics (Single degree study, bachelor I. deg., full time
form)
Field of Study: Physics
Type of Thesis: Bachelor´s thesis
Language of Thesis: English
Secondary language: Slovak

Title: Universe expansion and dark matter freeze-out in 1+1 dimension

Annotation: The main goal of the thesis is the study of the expansion of the homogeneous
and isotropic universe with one spatial and one temporal dimension. In the
case of 1+1 dimensional space-time the Einstein tensor vanishes identicaly
and gravity gets trivial. Therefore, instead of the straightforward reduction of
general relativity we use the Nordström theory as the alternative theory of
gravity in the lower dimensional case. This allows us to derive the Friedmann
equations for the homogeneous and isotropic universe. In the second part of the
thesis the freeze-out of massive particle annihilations will be studied. We will
compare the results with the standard 1+3 dimensional case.
Literature: Baumann, D., Primordial Cosmology, Proceedings, Theoretical Advanced
Study Institute in Elementary Particle Physics: Physics at the Fundamental
Frontier (TASI 2017): Boulder, CO, USA, June 5-30, 2017, arXiv:1807.03098
Riotto, A., Particle cosmology, High-energy physics. Proceedings, 5th CERN-
Latin-American School, Recinto Quirama, Colombia, 2010, arXiv:1010.2642
Bernstein, J., Kinetic theory in the expanding universe, Cambridge University
Press, 1988
Boozer, A.D., General relativity in (1+1) dimensions, European Journal of
Physics 29 (2008), 319-333, 10.1088/0143-0807/29/2/013

Supervisor: Mgr. Peter Maták, PhD.

Department: FMFI.KTF - Department of Theoretical Physics
Head of doc. RNDr. Tomáš Blažek, PhD.
department:
Assigned: 24.10.2019

Approved: 29.10.2019 prof. RNDr. Jozef Masarik, DrSc.

Guarantor of Study Programme

Student Supervisor
 Univerzita Komenského v Bratislave
Fakulta matematiky, fyziky a informatiky

ZADANIE ZÁVEREČNEJ PRÁCE

Meno a priezvisko študenta: Alžbeta Miklášová

Študijný program: fyzika (Jednoodborové štúdium, bakalársky I. st., denná
forma)
Študijný odbor: fyzika
Typ záverečnej práce: bakalárska
Jazyk záverečnej práce: anglický
Sekundárny jazyk: slovenský

Názov: Universe expansion and dark matter freeze-out in 1+1 dimension

Rozpínanie vesmíru a hustota tmavej hmoty v 1+1 rozmeroch
Anotácia: Hlavným cieľom tejto záverečnej práce je štúdium rozpínajúceho sa vesmíru
s jedným priestorovým a jedným časovým rozmerom. V 1+1 rozmernom
časopriestore je Einsteinov tensor nulový a gravitácia triviálna. Preto,
namiesto jednoduchej redukcie všeobecnej teórie relativity budeme uvažovať
Nordströmovu teóriu ako jej alternatívu v 1+1 rozmernom prípade. To nám
umožní odvodiť netriviálne Friedmannove rovnice. V druhej časti práce
sa zameriame na časový priebeh hustoty častíc tmavej hmoty. Budeme
pri tom numericky riešiť Boltzmannovu rovnicu pre anihiláciu takýchto častíc.
Nakoniec výsledky porovnáme so známym 1+3 rozmerným prípadom.
Literatúra: Baumann, D., Primordial Cosmology, Proceedings, Theoretical Advanced
Study Institute in Elementary Particle Physics: Physics at the Fundamental
Frontier (TASI 2017): Boulder, CO, USA, June 5-30, 2017, arXiv:1807.03098
Riotto, A., Particle cosmology, High-energy physics. Proceedings, 5th CERN-
Latin-American School, Recinto Quirama, Colombia, 2010, arXiv:1010.2642
Bernstein, J., Kinetic theory in the expanding universe, Cambridge University
Press, 1988
Boozer, A.D., General relativity in (1+1) dimensions, European Journal of
Physics 29 (2008), 319-333, 10.1088/0143-0807/29/2/013

Vedúci: Mgr. Peter Maták, PhD.

Katedra: FMFI.KTF - Katedra teoretickej fyziky
Vedúci katedry: doc. RNDr. Tomáš Blažek, PhD.
Spôsob sprístupnenia elektronickej verzie práce:
bez obmedzenia
Dátum zadania: 24.10.2019

Dátum schválenia: 29.10.2019 prof. RNDr. Jozef Masarik, DrSc.

garant študijného programu

študent vedúci práce

Adavit
I hereby arm, that this Bachelor's Thesis represents my own written work
and that I have used no sources and aids other, than those indicated.

Bratislava, June 2020

iv
Acknowledgements
I would like to express my gratitude to my supervisor Mgr. Peter Maták,
PhD. for a great patience and tons of useful advices. I am also grateful to
my friends and family for their kindness and support, and to all physicists
for the amount of beautiful knowledge I could immerse myself in.

v
Abstrakt
V tejto práci sa zaoberáme modelom vesmíru s jedným £asovým a jedným
priestorovým rozmerom. Takýto zjednodu²ený vesmír si ponecháva mnoho
atribútov a zaujímavých vlastností z 3 + 1 rozmerného vesmíru a môºe do-
bre slúºit na testovanie nových teórií ako aj na pedagogické ú£ely. V prvej
£asti práce ²tudujeme geometrické a dynamické vlastnosti takéhoto vesmíru
a odvodíme analóg Friedmannovej rovnice. V druhej £asti sa zaoberáme
stabilizáciou hustoty tmavej hmoty a porovnáme numerické rie²enia Boltz-
mannovej rovnice v 1 + 1 a 1 + 3 rozmeroch.

vi
Abstract
In this work we deal with the model of the universe with one temporal and one
spatial dimension. Such a simplied universe retains many of the attributes
and interesting properties of the 3 + 1 dimensional universe and can serve
well for testing new theories as well as for pedagogical purposes. In the rst
part of the work we study the geometric and dynamic properties of such a
universe and derive an analog of the Friedmann equation. In the second part
we study the freeze-out of dark matter and compare the numerical solutions
of the Boltzmann equation in 1 + 1 and 1 + 3 dimensions.

vii
Contents
Adavit iv
Acknowledgements v
Abstrakt vi
Abstract vii
Introduction 1
Theory of relativity . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Motivation for Einstein eld equations . . . . . . . . . . . . . . . . 6

Geometry of the universe 10

Friedmann equation in 1 + 1 dimension 13
Dark matter freeze-out in 1 + 1 dimension 18
Contents of the universe . . . . . . . . . . . . . . . . . . . . . . . . 18
Boltzmann equation . . . . . . . . . . . . . . . . . . . . . . . . . . 21

viii
Introduction
Theory of relativity
In special relativity with a space-time considered a four-dimensional mani-
fold, the events are represented by four coordinates [14]. The coordinates of
inertial observer (the observer moving along a free particle trajectory) are
conventionally x0 = ct, x1 = x, x2 = y , x3 = z so that all coordinates xa ,
(a = 0, 1, 2, 3) have the same dimension. The inertial observer exists in in-
ertial frame, the frame without acceleration, where the rst Newton's law of
motion is obeyed.
Objects in space-time, particles and observers, measure their own proper
time τ with standard clocks, the clocks of some standard construction. World-
line is a curve in space-time tracing the history of particle or observer. It is
dened by four coordinate functions parametrized in terms of proper time τ

xa = xa (τ ). (1)

The special relativity is based on two postulates:

1. The same physical laws are valid in all inertial frames. That means, the
absolute motion is not detectable by any experiment. Only a relative
motion can be detected.

1
Introduction 2

2. In an empty space, light has the same constant velocity c in all reference
frames.

As a consequence, the simultaneity of events is not the same for dierent

observers. Based on the second postulate, we can use the photons to measure
distance. Let observer O send two light signals in dierent directions. They
reect at the events E and E' and come back to his worldline at the equal
time t2 . That means the events E and E 0 ' occurred simultaneously according
to the observer O. Moreover, the observer assigns the time t = 12 (t1 + t2 ) and
spatial distance d = 2c (t2 − t1 ) to this events. It is important to note that
other observer O0 moving by relative velocity v with respect to the observer
O may not consider these events simultaneous.

Figure 1: The events E and E 0 are simultaneous for observer O but not for
O0

For two inertial observers O and O0 moving constant velocity v relative

to each other, the time ows at dierent speed. The proper time τ of O is
Introduction 3

related to the proper time τ 0 of O0 as dτ /dτ 0 = γ , where the γ factor is

1
γ=q . (2)
2 2
1 − v /c

Let E1 and E2 be two events with coordinates xµ1 and xµ2 in some coordi-
nate system. Then xµ1 − xµ2 = X µ are components of four-vector and set
of four numbers X = (X 0 , X 1 , X 2 , X 3 ) are subject to Lorentz transforma-
tions. Four-vectors make a four-dimensional vector space, the Minkowski
space. Minkowski space is a vector space with additional structure, metric
g = diag(1, −1, −1, −1) called Minkowski tensor. If X and Y are two four-
vectors, their scalar product is dened as g(X, Y ) = gµν X µ Y ν . The scalar
product is invariant, it is the same in all coordinate systems.
Let us denote W : τ → X µ (τ ) the worldline of some particle. Then one
can dene the four-velocity as

dX µ (τ ) dX µ dt dX µ
Vµ = = =γ , (3)
dτ dt dτ dt

where τ is the proper time measured along the worldline, and t is the coor-
dinate time. The scalar product of V is always g(V, V ) = c2 .
Now, we will consider a particle moving with constant acceleration. The
velocity is changing along the particle worldline, but we can assign a comov-
ing observer with corresponding velocity v and proper time τ to any innites-
imal element of the worldine. The displacement four-vector on worldline is
X = (ct, xi ). Using a dot notation for derivative with respect to proper time
i i
τ , we obtain ṫ = γ , ẋi = dx
dτ
= dt dx
dτ dt
= γv i . In the original coordinate
system, in which the accelerating particle is observed, v i and ai = d2 xi /dt2
are velocity and acceleration in the usual sense. The four-acceleration may
be dened as
d2 X µ a · ~v
4~
Aµ = 2 = γ v ) + γ 2 (0, ~a).
2 (c, ~ (4)
dτ c
Introduction 4

Now, let us consider the notoriously known example of an accelerating ele-

vator. With the x-axis pointing upward and the elevator moving by accel-
eration g , we obtain V µ = (cṫ, ẋ, 0, 0) and Aµ = (cẗ, ẍ, 0, 0) which implies
Vµ V µ = c2 = c2 ṫ2 − ẋ2 and Aµ Aµ = −g 2 = c2 ẗ2 − ẍ2 . Therefore, the motion
of the elevator can be described by two dierential equations
p
cẗ = g ẗ2 − 1,

ẍ = g ṫ.

Let c = 1 and set t0 = τ0 = 0 in the middle of the elevators oor at

the beginning of the motion. By solving the above equations of motion
we can show that the worldline of the bottom point of the elevator can be
parametrized in terms of the proper time tau as
1
t= sinh gτ ,
g
1
x = (cosh gτ − 1).
g
The person inside the elevator can assign his own coordinates Ξ = (τ, ξ 1 , ξ 2 , ξ 3 )
to any event he observes. Although these are behind the scope of special rel-
ativity (the frame of such an observer is non-inertial) we can always consider
the comoving inertial observer (at the same place with the same velocity)
using exactly the same cartesian coordinates Ξ. These are related to X by
the standard Lorentz transformation
 
γ vγ 0 0
 
 
 vγ γ 0 0 
Λ= 


 0 0 1 0 
 
0 0 0 1

where γ = dt
dτ
= cosh gτ and vγ = dx
dτ
= sinh gτ . With conveniently chosen
reference point we get t = (ξ + g1 ) sinh gτ , x = (ξ + g1 ) cosh gτ − g1 .
Introduction 5

Figure 2: The coordinates in the elevator are the so-called Rindler coordi-
nates in the Rindler space

The bottom point of the elevator is moving with the constant acceler-
ation g . Any other point with non-zero coordinate ξ can be assigned the
acceleration g 0 = g
1+gξ
resulting from 1
g
0 = ( g1 + ξ). The metric in eleva-
tor is no more Minkowski metric. In terms of the coordinates Ξ we have
ds2 = dt2 + dx2 = (1 + gξ)2 dτ 2 − dξ 2 . To move forward from special to gen-
eral relativity (the move to be accomplished in the next section) we need to
add the third postulate to the two mentioned previously. This corresponds to
the principle of equivalence, the notion that for the person inside the elevator
there is no way to distinguish between the eect of inertial forces, related to
Introduction 6

the acceleration of the frame, and the eect of gravitational eld.

Einstein eld equations

The main motivation leading to the Einstein eld equations comes from the
idea that the geometry of space-time may be related to its matter content. In
Euclidean space, particles move along geodesics, the straight lines. The mo-
tion of particle aected by gravitational eld can be interpreted as a geodesic
motion in a space-time curved by matter. In general, such motion can be
described by geodesic equation

Ẍ α + Γαµν Ẋ µ Ẋ ν = 0, (5)

where X α are the coordinate functions of the geodesic γ(t) and Γcab = 12 g ad (
gcd,b + gbd,c − gbc,d ) are the Christoel symbols. The notation gbd,c means
derivative with respect to X c . The geometry behind the geodesics, as well
as other tools from dierential geometry used in this thesis, are explained
beautifully in [5].
Following the t'Hooft papers [7], let us consider a particle moving slowly
in a weak and static gravitational eld. The four-velocity of such particle can
be approximated as Ẋ ≈ (1, ~0). In consequence, we can reduce the geodesic
equation to
Ẍ i ≈ −Γi00 . (6)

The weakness of the gravitational eld lead us to another approximation. We

will write the space-time metric as ηµν = diag(1, −1, −1, −1) with a small
perturbation h
gµν = ηµν + hµν , (7)

This is in agreement with our experience that a trajectory of a particle should

be aected only a little if the gravitational eld is weak. We can see easily
Introduction 7

from elevator's metric that h00 should be equal to 2gξ = 2φg .

1
Γcab ' η ad (hcd,b + hbd,c − hbc,d ),
2
1 ij 1
Γ00 = − η h00,j = − ∂ i h00 = −∂ i φg ,
i
2 2
i i
Ẍ = ∂ φg ⇒ ~g = −∇φ. ~

The matter content of the universe can be described by the energy-momentum

tensor. This symmetric tensor has two coordinate indices. Therefore, we can
look for symmetric tensors describing the geometry of space-time. There is
the curvature tensor dened on any riemannian manifold.
α
Rβγδ = heα , R(eβ , eγ )eδ i = Γαβδ,γ − Γαβγ,δ + Γρβδ Γαργ − Γρβγ Γαρδ , (8)

where R((eβ , eγ ) = [∇β , ∇γ ] − ∇[β,γ] and ∇β ∂γ = Γαγβ ∂α is a covariant deriva-

tive, a derivative along the vector eld eβ . After the contraction we obtain
γ
the Ricci tensor Rαβ ≡ Rαγβ . Like the energy-momentum tensor, Ricci tensor
carries symmetric indices too. Therefore, it appears to be a good candidate
to be proportianal to the energy-momentum tensor to obtain the equation
relating the matter and geometry of space-time. The proportionality comes
from the Lovelock theorem stating that higher derivative terms may lead to
causality violations [9]. However, the simple linear dependence Rαβ = κTαβ
is not sucient. Unlike the Ricci tensor, the four- divergence of the energy-
momentum tensor vanishes. Therefore we use the Bianchi identity to prove
that Gαβ
;β = 0, where
1
Gαβ = Rαβ − Rgαβ , (9)
2
is the Einstein tensor and the Einstein equations can be written as

Gαβ = κTαβ . (10)

To nd the value of the proportionality constant kappa, we employ the per-
turbation theory to the rst order in the metric perturbation h to reproduce
Introduction 8

the newtonian limit of the relativistic theory. This corresponds to the as-
sumption that all the elds are weak. Moreover, we assume that the move-
ments of all the bodies are slow enough. In consequence, time derivatives of
the components of h can be neglected completely. The Ricci tensor and its
contraction, the Ricci scalar, are equal to

1
Rab ' (−∂µ ∂ µ hαβ + ∂α ∂µ hµβ + ∂β ∂µ hµα − ∂α ∂β hµµ ),
2
1 1 1
R00 ' − ∂µ ∂ µ h00 = − 2h00 = 4h00 = 4φg ,
2 2 2
R = Rαα = g αβ Rαβ ' −2hµm u + ∂ α ∂ β hαβ

to the rst order in h. Here 2 = ∂µ ∂ν = g µν ∂µ ∂ν is the d'Alambert operator.

To avoid the unknown spatial components of h, which cannot be neglected,
we perform a little trick [7]. Making the contraction of the both sides of
equation (9) we obtain

Gαα = −R = κTαα = κT, (11)

where T is the trace of energy-momentum tensor. Putting this back to the

Einstein equations we get

1 1
Rαβ = Rgαβ + κTαβ = κ(Tαβ − T gαβ ),
2 2
1 α
R00 = 4φg = κ(T00 − Tα g00 ).
2

In the limit of static weak eld the dominant component of energy-momentum

tensor is the mass density T00 = ρ ' T . Using the Gauss law from Newtonian
gravity results to
1
4φg = 4πGρ = κρ. (12)
2
Finally, we obtain the Einstein equations in the form

Gαβ = 8πGTαβ . (13)

Introduction 9

These equations has been proded by A.Eddington in the experiment during

the Solar eclipse in 1917 and many others since then.
It is interesting to note, that there had been previous attempts to describe
the gravitation in relativistic way. The Nordström's theory described in
Misner's Gravitation [10] was one of the most signicant. In Nordström's
theory the gravitational eld is considered to be a scalar function φ generating
the spacetime from the Minkowski tensor as g = φ2 η . In non-relativistic limit,
the spatial components of the energy-momentum tensor are negligible and
therefore Nordström suggested that ρ can be approximated by the trace of
energy-momentum tensor ρ ' Tµµ = T (unlike the h00 component as Einstein
did). This leads to a eld equation

η αβ φ,αβ = 4πGη αβ , Tαβ . (14)

This can be also expressed in terms of the scalar curvature as

R = 24πGT. (15)

Interestingly enough, the information about geometry in Ricci scalar R is not

reduced. In Nordström's theory the metric has only one degree of freedom in
the scalar eld φ. Nordström's theory required a at-spacetime background,
a geometry that is independent of matter content. The rejection of such
"prior geometry" was needed to develop Einstein's theories.
Geometry of the universe
The geometry of the universe is determined by its matter content and de-
scribed by the metric tensor. We will study the Friedmann-Lemaître-Robert-
son-Walker (FLRW) model of universe [1]. It is based on the assumption that
on large scales our universe is homogenous and isotropic. That means it has
the same properties and looks the same at every position and any direction.
Thanks to this symmetry we can treat the universe as constantly curved or
at. At present, the space seems to be nearly at from the measurements,
but positively or negatively curved space is possible as well within the ex-
perimental error. In all three cases the line element in spacetime takes the
form
ds2 = dt2 − dl2 = dt2 − a2 (t)γij dxi dxj , (16)

with variant γij . Here we follow the lecture notes on cosmology by Daniel
Baumann [1]

1. Flat space is the Euclidean 3-dimensional space E 3 with spatial distance

element of the form

dl2 = γij dxi dxj = δij dxi dxj . (17)

2. Positively curved space is a 3-sphere S 3 embedded in 4-dimensional

Euclidean space. In analogy with S 1 it is dened as ~x2 + u2 = a2 ,

10
Geometry of the universe 11

where a is a radius of 3-sphere. The line element is d~l2 = d~x2 + du2

p
with constraint ~xd~x = −udu, where u = a2 − ~x2 . It follows that
2 2
d~l2 = d~x2 + ~
x d~
2
x
a −~
x
2 and with scaling x → ax we obtain

~
h ~x2 d~x2 i
2 2 2
dl = a d~x + . (18)
1 − ~x2

3. Negatively curved 3-space is a hyperboloid H 3 dened as ~x2 −u2 = −a2 .

Similarly to previous case we derive that the line element is

~
h ~x2 d~x2 i
2 2 2
dl = a d~x − . (19)
1 + ~x2

We can parametrize all the three cases as

h ~x2 d~x2 i
ds2 = dt2 − a2 (t) d~x2 + k , (20)
1 − k~x2
where k = 0 is for at space, k = 1 for positively curved space and k = −1
for negatively curved space. In expanding universe a(t) is the scale factor.
Free particles in spacetime are moving along geodesics that can be derived
from the metric tensor.
The matter content and energy, which are the source for spacetime curva-
ture, are represented in energy-momentum tensor T . In FLRW model, it is a
energy-momentum tensor for an isotropic uid T µν = (ρ + P )U µ U ν − P η µν .
Component T 00 is the energy density (or the mass density) if c = 0, T 0i = T i0
represents a density of energy ux, which is zero for isotropic uid. Com-
ponent T ij is j th component of ux density of ith momentum component.
For perfect uid only the diagonal components are non-zero and represent
its pressure.
The Einstein equations deal with the relation between the energymomen-
tum tensor and Ricci tensor representing the curvature. Due to isotropy we
expect both of the tensors to have vanishing components with one spatial and
Geometry of the universe 12

one temporal index. From Einstein equations we can derive the Friedmann
equation, which describes the evolution of the scale factor a(t)

ȧ2 8πG k
2 = ρ − 2. (21)
a 3 a

The four-divergence of energy-momentum tensor is conserved

∇µ T µν = T;µµν = T,µµν + Γµρµ T ρν + Γνρµ T µρ = 0, (22)

which leads to the local conservation law for isotropic uid

ȧ
ρ̇ + 3 (ρ + P ) = 0 (23)
a

where H ≡ ȧ
a
is a Hubble parameter.
We will study a 1+1 dimensional model of universe, which is much simpler
but still possesses a many features analogous to 3 + 1 dimensional universe.
Friedmann equation in 1 + 1
dimension
In 1 + 3 dimensional universe, the theory of gravity is described by the Ein-
stein eld equations which determine the dynamics of the universe.

Gαβ = 8πGTαβ , (24)

where
1
Gαβ = Rαβ − Rgαβ . (25)
2
Calculating both the energy-momentum tensor and the Ricci tensor, we get
the evolution of the scale factor a(t), depending on the matter content and
the curvature of the FLRW spacetime. In 3 + 1 dimension we obtained the
Friedmann equation, so we expect to have some similar equation in 1+1
dimension. First, to get the components of the Ricci tensor, we need to cal-
culate the Christoel symbols Γαβγ = 21 g αδ (gδγ,β + gβδ,γ − gβγ,δ ). The FLWR
metric takes the form
xx
ds2 = dt2 − a2 (t)γij dxi dxj where γij = δij + k 1−kx
i j
2 . The individual compo-
2
nents of FLWR metric g are g00 = 1, g01 = g10 = 0, g11 = −a2 (t)(1 + k x
2 ).
1−kx
The inverse metric is dened by g αβ gβγ = δαγ . The spatial derivative of the
component g00,i vanishes, so all the Christoel symbols with two time indices

13
Friedmann equation in 1+1 dimension 14

are zero.

1
Γ110 = g 11 (g10,1 + g11,0 − g10,1 ),
2
ȧ 1
g11,0 = 2 g11 , g 11 = ,
a g11
ȧ
Γ110 = .
a

Due to the symmetry of metric, Γ110 and Γ101 are equal. The other Christoel
symbols are derived in a similar way

1  x2 
Γ011 = g 00 (g01,1 + g10,1 − g11,0 ) = aȧ 1 + k ,
2 1 − kx2
1 kx
Γ111 = g 11 g11,1 = .
2 1 − kx2

The derivatives of Christoel symbols needed in Ricci tensor are

Γ101,1 = Γ110,1 = 0 = Γ111,0 ,

ä  ȧ 2
Γ101,0 = Γ110,0 = − .
a a

The contraction of the Riemann tensor (8) gives Rµν = Γρµν,ρ −Γρµρ,ν +Γσµν Γρσρ −
Γσµρ Γρσν . Both R10 and R01 components are zero R10 = Γ110,1 − Γ111,0 + Γ110 Γ111 −
Γ111 Γ110 = 0. The diagonal components are

ä
R00 = − ,
a
 x2  ä
R11 = äa 1 + k 2 = − g11 .
1 − kx a

The Ricci scalar is a contraction of the Ricci tensor

ä
R = Rαα = g αα Rαα = −2 . (26)
a

Since Gαβ = Rαβ − 21 Rgαβ = 0, we can see, that the Einstein tensor vanishes
and the Einstein equations have no dynamical content. Thus, we need an
Friedmann equation in 1+1 dimension 15

alternative theory of gravity in 1 + 1 dimension. Due to obvious lack of ex-

perimental data from two-dimensional universe, we cannot be certain how to
describe a relativistic gravitation in two dimensions. We will follow Sikkema
and Mann [13] using a theory of gravity based on string theory and Liouville
quantum gravity
R = 8πGT, (27)

where R is the Ricci scalar, T is the trace of the energy-momentum tensor

and G is a 1 + 1 dimensional gravitational constant. In two dimensions,
the Ricci scalar completely characterizes the curvature of a two-dimensional
surface.
Otherwise we could transform a Nordström's theory as in [3]. We men-
tioned this theory in 3 + 1 dimensions, however the 1 + 1 dimensional model
has a dierent constant. The dierences are in the scalar curvature and the
gravitational eld equation ∂ 2 φ = 2Gρ. The 1 + 1 dimensional analog of
Nordström's theory leads to R = 4GT .
Unlike in the 3 + 1 dimensional space, in 1 + 1 dimensions the evolution
of the scale factor a(t) is the same for at and positively or negatively curved
space. So we can use coordinates independent of k and the metric takes a
simpler form ds2 = dt2 − a2 (t)dx2 .
Due to the isotropy and homogeneity of the universe the energymomen-
tum tensor takes the form

Tµν = (ρ + P )Uµ Uν − P gµν . (28)

With 2-velocity of comoving observer U = (1, 0), we have the components

of energymomentum tensor T00 = ρ, T10 = T01 = 0, T11 = P so with
rised indices they are T00 = ρ, T11 = −P and the trace of the tensor equals
T = T00 + T11 = (ρ − P ). Thus we obtain the 1 + 1 dimensional eld equation,
Friedmann equation in 1+1 dimension 16

the analogue of the Friedmann equation

ä = 4πG(P − ρ)a. (29)

The local conservation of energy can be written as T;µµν = T,µµν + Γµρµ T ρν +

Γνρµ T µρ = 0. Referring to [5] it can be rewritten as
1 √
T;µµν = g µν ∂µ P + p ∂µ [ g(P + ρ)U µ U ν ] + Γνµρ (P + ρ)U µ U ρ . (30)
|g|
~ = 0,
Since the pressure P for perfect uid is isotropic and therefore ∇P
the conservation law is always satised for spatial component. For a time
component ν = 0 we get
1
T;000 = g 00 ∂t P + ∂t (a(P + ρ)) ⇒ a∂t P = ∂t (a(P + ρ)).
a
This can be rewritten as
ρ̇ + H(ρ + P ), (31)

which is similar to local conservation law in 3 + 1D (23). The dierence is in

the factor of 3, which arises from expanding in one direction instead of three
directions.
We can combine the conservation law and the eld equation (29), which
is a dierential equation of the second order, to get a dierential equation of
the rst order.
ȧ
ä = 4πG(−ρ̇ − 2ρ)a,
a
2ȧä = −8πG(ρ̇a2 + 2ρȧa),
d 2 d
ȧ = −8πG (ρa2 ),
dt dt
ȧ2 = −8πGρa2 + C.

This leads to the equation of the similar form as the Friedmann equation in
3 + 1D
ȧ2 C
2 = −8πGρ + 2 . (32)
a a
Friedmann equation in 1+1 dimension 17

From the three equations (29), (31) and (32) there are just two linearly
independent equations. This is the same as in 3 + 1 dimensions, where we
have two Friedmann equations and the local conservation law.
Dark matter freeze-out in 1 + 1
dimension
Contents of the universe
There are three main components of universe: radiation, matter and dark
energy. As matter we refer to baryons and cold dark matter. These are non-
relativistic particles. The pressure is proportional to kinetic energy, so it is
negligible compared to energy density. The energy density scales with the
expansion parameter as ρm ∼ a−3 . Baryons are ordinary matter consisted
of protons and neutrons. Dark matter is still a subject of research, we don't
know what it is. We will deal with WIMPs, weakly interacting massive
particles, which are hypothetical dark matter particles with masses and the
interaction strength given in terms of the weak scale.
Radiation are relativistic particles, such as photons and neutrinos, which
have the pressure around a one third of the energy density. Here, the energy
density scales as ρr ∼ a−4 . This comes from the fact that volume itself scales
as V ∼ a3 while the wavelength of the particles stretches as λ ∼ a1 .
Dark energy is a form of energy predicted by measurements but still of
unknown nature. Its density remains constant despite the expansion and
so the scaling of the energy density is ρΛ ∼ a0 . It is responsible for the

18
Dark matter freeze-out in 1+1 dimension 19

accelerated expansion of the universe.

In 1 + 1 dimensional universe these relations will be dierent, since the
universe expands in one direction only. For matter and radiation we have
ρm ∼ a−1 and ρr ∼ a−2 respectively.
Since the densities of matter, radiation and dark energy are diluting at
dierent rates, there is always another dominant component. We will study
the rst stage of the universe lasting only a few fractions of a second. During
the period dominated by radiation the freeze-out of dark matter occurred.
We will describe it in 1 + 1 dimensional universe following the third chapter
in Baumann's Cosmology lectures [1], which deals with 3 + 1 dimensional
case.
We describe the thermodynamics of the universe in terms of the expan-
sion rate H and the interaction rate Γ. The early universe was hot and dense,
and massive particles were interacting at high rate. Antiparticles were anni-
hilating with particles and creating the light particles in this process. Until
the interaction rate was higher than the expansion rate, the universe was in
thermal equilibrium. Massive particles would be eventually suppressed by
the factor of e−m/T but as the universe expands and cools down the interac-
tions decrease which results in freeze-out of the massive particles and their
relic density.
In quantum mechanics, we have discrete momentum values for particles
on a segment of the length L: pn = 2π~
L
n, where we set ~ ≡ 1. Therefore, the
number of states with momentum within some interval (p, p + ∆p) equals to
∆n = L
2π
∆p.
In thermal equilibrium we can describe the distribution of bosons with
Bose-Einstein statistics (− sign) and fermions with Fermi-Dirac statistics (+
Dark matter freeze-out in 1+1 dimension 20

sign)
1
f (E) = . E−µ (33)
e ±1 kB T

The chemical potential µ for photons and electrons (and probably for neu-
trinos as well) is zero. Also, we will use the natural units ~ = kB = c = 1.
p
The energy of a relativistic particle, Ep~ = m2 + p~2 , can be approximated
by Ep~ = |~
p|. Therefore we obtain the distribution

1
f (p) = p . (34)
e ±1
T

For the particle with g degrees of freedom, the number density is given by
the product of the density of states and the distribution of the particle in
momentum space
gL
∆nf (p) = f (p)∆p. (35)
2π
Integrating over the momentum space we get a number density of relativistic
particles Z Z
1 g 1
nr (T ) = g dnf (p) = p dp, (36)
L 2π e ±1
T

and the energy density

Z
g p
ρr (T ) = p dp. (37)
2π e ±1
T

For fermions we have

∞
gπT 2
Z
g p
ρf (T ) = p dp = . (38)
2π 0 eT + 1 24

For bosons we will use the following Riemann zeta-function

∞
∞
π2
Z
xdx X 1
ζ(2) = = = , (39)
0 ex − 1 n=1 n2 6

so we obtain
∞
gπT 2
Z
g p
ρb (T ) = p dp = . (40)
2π 0 eT − 1 12
Boltzmann equation 21

The energy density for radiation then equals to

π 2
ρ(T ) = T g∗ (T ), (41)
12

where  2  2
X Ti 1 X T
g∗ (T ) = gi + gi i (42)
bosons
T 2 f ermions T

is the number of eective degrees of freedom. In the case of cold or non-

relativistic particles in thermal equilibrium we have the Maxwell-Boltzmann
distribution
Ep
f (p) = e− T (43)
p
Since p  m for matter, the energy of the particle Ep = m2 + p2 can be
2
expanded in p as Ep ≈ m + p
2m
. The number density is
  12
mT m
nm = g e− T (44)
2π

Energy of the matter can be approximated as a mass of the particle Ep ≈ m

at the lowest order so the energy density is ρm ≈ nm and with the correction
of the second order
1
ρm = nm + nT (45)
2

Boltzmann equation
Without interactions, the Boltzmann equation is just a conservation of num-
ber density of particles

1 d(ni a)
= ṅi + Hni = 0. (46)
a dt

The interactions could be included in an interaction term on the right-hand

side of the equation. In a simplied model with four species of particles
Boltzmann equation 22

which can be annihilated and created in a reversible process φ1 φ2 ↔ φ3 φ4 ,

the Boltzmann equation takes the form

1 d(n1 a)
= −αn1 n2 + βn3 n4 , (47)
a dt

where the term −αn1 n2 stands for annihilation and βn3 n4 for creation of
the type 1 particles. The rate of interactions is proportional to the cross
section, so the parameter α is equal to hσvi, the cross section thermally
averaged with velocity v . The r.h.s. of the Boltzmann equation vanishes in
equilibrium, which implies
 
nn
β=α 1 2 (48)
n3 n4 eq

where nieq are the number densities in thermal equilibrium. The Boltzmann
equation becomes
 
1 d(n1 a) n1 n2
= −hσvi(n1 n2 + n3 n4 ) (49)
a dt n3 n4 eq

Approximately 95% of all matter in our universe corresponds to a dark

matter [8]. One of the observations that proves the existence of the dark
matter is the measurement of the spiral galaxies rotation curves. They would
be dierent from what was observed if there was only the luminous matter.
The actual dark matter particle was not observed yet, but there are a few
theoretical candidates with dierent properties. We will study the weakly-
interacting massive particles. We know that the relative abundance of dark
matter is of order one, which can be obtained if the annihilation cross section
is similar to that of the electroweak interactions. The rst candidates, a heavy
neutrinos, were excluded by experiments. The current perspective candidate
is the lightest supersymmetric particle (LSP), the lightest neutralino with
mass of the order omχ ∼ 100 GeV.
Boltzmann equation 23

We want to describe the freeze-out in 1 + 1 dimensional universe, which

is not supported by any experimental data. For this purpose we have to
make some assumptions about the model of the universe. We have two
particle species only. The WIMP particles and antiparticles χ, χ̄, which can
annihilate and produce the light particles and antiparticles l, ¯l. This process
is reversible χ + χ̄ ↔ l + ¯l. In addition, we consider both of the particles as
scalar particles with g = 1. The cross section of WIMP in 3 + 1D can be
obtained by means of dimensional analysis
2
αef f
σχ ' 2 ∼ 10−9 GeV−2 . (50)
mχ
In 1 + 1D the cross section is dimensionless σχ ' αef
2
f and we will take the

value of αef f the same as in 3 + 1D.

The light particles keep their equilibrium densities throughout the pro-
cess. The Boltzmann equation (49) for dark matter particles can be written
as
1 d(na)
= −hσvi(n2 − n2eq ). (51)
a dt
It is convenient to introduce the entropy density
ρ+P
s= . (52)
T
Following the Bernstein's Kinetic Theory [2] we assume the expansion of the
universe to be slow enough for entropy to be conserved. For the entropy
density in comoving volume we write
1 d(sa)
= ṡ + Hs = 0. (53)
a dt
Now we can rewrite the Boltzmann equation for a new quantity, a ratio of
the number density and the entropy density N = n/s, which is proportional
to number of particles in comoving volume

Ṅ s + N ṡ + HN s = −hσvis2 (N 2 − Neq
2
), (54)
Boltzmann equation 24

which simplies using the entropy conservation (53) to

dN
= −hσvis(N 2 − Neq
2
). (55)
dt

Moreover, it is useful to write the Boltzmann equation in terms of variable

M
x= . (56)
T

Therefore we need to nd the relation between x and t. For relativistic

particles in 1 + 1 dimension we have P = ρ. From the eld equation (32)
we will get the evolution of the scale factor during the radiation dominated
period. The energy density of radiation evolves as ρ ∼ a−2 , so the equation
(32) can be reduced to
ȧ2 = −8πGρ0 + C. (57)

The solution for this dierential equation is a linear function of t. With the
initial condition a(t0 ) = a0 we get a(t) = a0 tt . We usually set the scale factor
0

at the present time to one, a0 = 1. Therefore

t
a(t) = . (58)
t0

The energy density scales as ρ ∼ a−2 . Hence the analogue of the Friedmann
equation in 1 + 1 dimension becomes

ρ0 C
H 2 = −8πG 2 + 2 (59)
a a

For Hubble constant we can write

4π 2 2
H02 =− T + C. (60)
3 0

Because of the unknown parameter C , and unlike the 3 + 1 dimensional case,

the Hubble constant and the present temperature are independent. Therefore
Boltzmann equation 25

we will need both T0 and H0 as as an input for our calculations. Moreover,

from the previous equations the following can be derived

ȧ
H 2 = H02 a−2 , = H0 a−1 ,
a
da = H0 dt, a = H0 t,
1
H=
t

Now, we obtain
 
dx d M x dT 1 1
= =− ' xt 2 = x = Hx. (61)
dt dt T T dt t t

Therefore
dN 1
=− hσvis(N 2 − Neq
2
). (62)
dx Hx
Surprisingly, this Riccati-type equation takes the same form (62) in both
dimensions, although there are important dierences in temperature depen-
dence of the Hubble parameter, entropy density and the cross section. From
relations (41) and (52) follows that entropy density for radiation dominated
universe equals to
π
s= T g (T ) (63)
6 ∗
The equation (62) cannot be solved analytically but we will nd a nu-
merical solution. The numerical value of the gravitational constant is found
experimentally and is expected to be higher in the universe with more dimen-
sions. However, in our model, we will take the value of the four-dimensional
constant except for the dierent units [16, 11]. Consequently, the reduced
Planck mass remains
r
1
Mpl = = 2.435x1018 GeV
8πG
Boltzmann equation 26

Further we use the present values of the Hubble parameter and the temper-
ature of the thermal medium

H0 = 1.42 × 10−42 GeV,

T0 = 23.5 × 10−14 GeV.

The numerical solutions with dierent cross sections can be seen in gures
3 and 4.

Figure 3: Dark matter freeze-out in 1 + 1 dimensions

Figure 4: Dark matter freeze-out in 3 + 1 dimensions

Boltzmann equation 27

The dashed line is for the exponetially supressed equilibrium abundance

and the color lines represent the density of relics. For x → ∞ the equilibrium
density Neq drops exponentially. Therefore, it is negligible compared to the
freeze-out density N∞ . Thus we can nd an approximate solution for the
Riccati equation (62)
2
dN αef fs 2
=− N . (64)
dx Hx
The entropy density in terms of x equals to
π
s= mx−1 . (65)
6
The Hubble parameter is proportional to temperature so it can be expressed
as
H02 2 H02 2 −2
H2 = T = 2m x . (66)
T02 T0
We get the dierential equation
dN 2
π αeff mx−1 2
=− N . (67)
dx 6 HT 0 m
0

Apparently, the dark matter abundance does not depend on the mass of the
WIMP particles. Let us introduce the constant
π T0 2
λ= α . (68)
6 H0 eff
To obtain the approximate solution we write
dN λ
= − N 2.
dx x
Integrating this equation from x = xf to x = ∞ we obtain
1 1 xf
− = −λ ln , (69)
N∞ Nf x∞
where Nf is typically much larger than N∞ , so the approximate relic density
long after freeze-out can be written as
xf −1
 
N∞ = − λ ln . (70)
x∞
Boltzmann equation 28

In gure 3 and gure 4 we can see that the dark matter abundance de-
creases for bigger values of annihilation cross section. That was expected
since the particles with smaller cross section stop interacting earlier. Com-
paring the 1 + 1D and 3 + 1D freeze-out, the deviation from the equilibrium
density occurred around the same xf but the abundance density in 1 + 1D is
a few orders of magnitude lower.
Bibliography
[1] Daniel Baumann. Cosmology. lecture notes.

[2] Jeremy Bernstein. Kinetic Theory in the Expanding Universe. Cam-

bridge Monographs on Mathematical Physics. Cambridge University
Press, 1988.

[3] A D Boozer. General relativity in (1 + 1) dimensions. European Journal

of Physics, 29(2):319333, feb 2008.

[4] J.D. Brown. Lower Dimensional Gravity. World Scientic, 1988.

[5] Marián Fecko. Dierential Geometry and Lie Groups for Physicists.
Cambridge University Press, 2006.

[6] Samuel C. Fletcher, J.B. Manchak, Mike D. Schneider, and James Owen
Weatherall. Would two dimensions be world enough for spacetime?
Studies in History and Philosophy of Science Part B: Studies in His-
tory and Philosophy of Modern Physics, 63:100–113, Aug 2018.

[7] G. Hooft. Introduction to General Relativity. Rinton Press, 2001.

[8] Marc Kamionkowski. Wimp and axion dark matter, 1997.

[9] David Lovelock. The Einstein Tensor and Its Generalizations. Journal
of Mathematical Physics, 12(3):498501, March 1971.

29
BIBLIOGRAPHY 30

[10] C.W. Misner, K.S. Thorne, J.A. Wheeler, W.H. Freeman, and Company.
Gravitation. Gravitation. W. H. Freeman, 1973.

[11] Mario Rabinowitz. n-dimensional gravity: Little black holes, dark mat-
ter, and ball lightning, 2001.

[12] Antonio Riotto. Lecture notes on cosmology. lecture notes, 2013.

[13] A E Sikkema and R B Mann. Gravitation and cosmology in (1+1)

dimensions. Classical and Quantum Gravity, 8(1):219235, jan 1991.

[14] N.M.J. Woodhouse. Special Relativity. Springer Undergraduate Mathe-

matics Series. Springer London, 2012.

[15] A. Zee. Einstein Gravity in a Nutshell. In a Nutshell. Princeton Uni-

versity Press, 2013.

[16] Barton Zwiebach. A First Course in String Theory. Cambridge Univer-

sity Press, 2004.