The cosmological constant and Quintessence

Daniele Scalabrini

Abstract
We first review the conditions that lead to an accelerated expansion of the uni-
verse from a theoretical point of view. We then move on to a description of Dark
Energy in term of the cosmological constant and briefly discuss the problems
that come with this choice. We will end by discussing Quintessence and tracking
solutions, and how they might offer a solution to the problems of the cosmological
constant.

1 Introduction
The observational discovery of late time cosmic acceleration from the Supernovae type
Ia [7, 6] opened up a new area of research in modern cosmology. This has been also
confirmed by other observations, such as Cosmic Microwave Background (CMB) [8, 1]
and Baryon Acoustic Oscillations (BAO)[5]. The source of this accelerated expansion
is an unknown negative pressure fluid, called Dark Energy, that makes up about 70%
of the energy budget of the Universe. Understanding the nature of this energy form
is a serious problem in cosmology. The simplest candidate is Einstein’s cosmological
constant ( which gives rise to the ΛCDM model ), but there are more complicated
dynamical models that can be divided into two classes:
⋄ Modified matter models: models in this class are based on a specific form of
matter, such as quintessence or k-essence;
⋄ Modified gravity models: models in this class are based on the modification of
gravity at large distances. Examples includes f(R) gravity, scalar-tensor theories
and so on;
In both classes the dark energy equation of state changes in time, which is the feature
that distinguishes them from the ΛCDM model.

1.1 Conditions for accelerated expansion

Since there is increasing observational evidence that the universe is nearly flat, from
now on we will restrict ourselves to the FLRW metric for a flat universe (k = 0):

ds2 = −dt2 + a(t)2 (δij dxi dxj )

The Einstein equations for this metric, in the presence of a mixture of perfect fluids
with equation of state Pa = wa ρa (wa =cost) are given by:

ρ ä ρ + 3P
H2 = =−
3Mpl2 a 6Mpl2

1
 P
where ρ = a ρa . From the second equation it’s easy to see thatP an accelerated
P
expansion is possible only if (ρ + 3P ) < 0. Using the fact that P = a Pa = a wa Pa
we can write accelerated expansion condition as:
X
(1 + 3wa )ρa < 0
a

This condition is satisfied if at least one of the energy components has wa < − 31 and
dominates over the others.

2 The Cosmological Constant

For explaining the observed accelerating expansion of the universe the simplest idea is
to borrow Einstein’s cosmological constant. Let us consider Einstein’s equation, with
a cosmological constant term:

Gµν + Λgµν = 8πGTµν

If we take the cosmological constant term on the left side we can define:
Λ µ
Tνµ = − δ
8πG ν
Where we have:
T00 = −ρΛ Tji = −ρΛ δji = PΛ δji
From these two equations we can read of the equation of state for the cosmological
constant, which is PΛ = −ρΛ , i.e. wΛ = −1. By solving the continuity equation for
ρΛ we get ρΛ = const. The cosmological constant then satisfies the conditions to be
a source for accelerated expansion since wΛ < − 13 and ρΛ =cost. which means that,
depending on the value of the constant, it will eventually take over radiation ρr ∝ a−4 ,
and dust ρd ∝ a−3 . What comes out of this picture is the so called ΛCDM model,
which is the simplest model in accordance with observations.

2.1 The problems of the Cosmological constant

Despite the success of the ΛCDM model, there are two problems that plague the
cosmological constant: one is a fine tuning problem tied to the interpretation of Λ as
vacuum energy, and the other is a problem known as the cosmic coincidence problem.

2.1.1 Cosmological Constant as Vacuum energy

The origin of the cosmological constant can be explained by a particle physics point of
view as vacuum energy. However this explanation brings with it a serious problem of
energy scale. From observations we know that Λ is of the order of the Hubble parameter
H0 , that is:
Λ ≈ H02 = (2.13h × 10−42 GeV )2
This corresponds to a critical density:

ΛMpl2
ρΛ = ≈ 10−47 GeV 4
8π

2
The vacuum energy can also be computed on the QFT side [4, 11], by summing zero
point energies of quantum field of mass m, momentum k and frequency ω:

1 ∞√ 2
Z ∞ √
d3 k
Z
1
ρvac = k +m 2 = 2 k 2 k 2 + m2 dk
2 0 (2π)3 4π 0

This exhibits an ultraviolet divergence: ρvac ∝ k 4 . However we expect that quantum

field theory is valid up to some cut-off scale kmax in which case the integral is finite:
4
kmax
ρvac ≈
16π 2
Taking the cutoff scale to be the Planck mass Mpl , the vacuum energy density can
be estimated as ρvac ≈ 1074 GeV 4 . This is about 10121 times larger than the observed
value.

2.1.2 The Coincidence problem

The question at the heart of the cosmic coincidence problem is very simple: why is the
observed density of Dark Energy of the same order of magnitude of the observed energy
density of Dark Matter? Or in other words, why does cosmic acceleration happen to
begin at roughly the present time, or not at some time in the past or the future? If we
were to try to explain this using the ΛCDM model we’d run into a fine tuning issue,
not dissimilar from the one encountered when dealing with the flatness problem. Since
for the cosmological constant ρΛ = cost and for Dark Matter ρDM ∝ a−3 , it appears
that their ratio must be set to a very specific infinitesimal value in the early universe,
in order for the two densities to nearly coincide today.

3 Quintessence
One way to address the coincidence problem is to turn our attention towards dynamical
models of dark energy, such as Quintessence. Quintessence is described by a scalar
field ϕ, minimally coupled to gravity, that slowly varies along a potential V (ϕ) leading
to late time accelerated expansion. This mechanism is similar to the one employed
to describe slow roll inflation in the early Universe, but the difference is that non
relativistic matter and radiation cannot be ignored in order to describe the dynamics
of Dark Energy correctly.

3.1 Equations of motion for the scalar field

We consider Quintessence in the presence of a conventional energy background( such as
NR matter or radiation, or both ) described as a perfect fluid with a constant equation
of state parameter wB and an action SB [10]. The total action is given by:

√
Z  
4 1 2 1 µν
S= d x −g Mpl R + g ∂µ ϕ∂ν ϕ − V (ϕ) + SB
2 2

We study the dynamics of quintessence on the flat FLRW background. By using the
explicit form for the stress energy tensor for ϕ we get the energy density and pressure
of quintessence:
1 1
ρϕ = ϕ˙2 + V (ϕ) Pϕ = ϕ˙2 − V (ϕ)
2 2

3
and its equation of state parameter:
1 ˙2
Pϕ ϕ − V (ϕ)
wϕ = = 21 (1)
ρϕ
2
ϕ˙2 + V (ϕ)

Since ϕ = ϕ(t), wϕ varies with time and has a range −1 < wϕ < 1. In this range, we are
interested in the negative pressure: −1 < wϕ ≤ 0. We see that whenever V dominates
over the kinetic energy then wϕ ≈ −1, recovering in this way a cosmological constant
EoS capable of accelerating the universe. The equation of motion for the field can be
derived by using the continuity equation:
ρ̇ϕ = −3H(ρϕ + Pϕ )

and inserting the explicit expressions of Pϕ and ρϕ :

ϕ̈ + 3H ϕ̇ + V,ϕ = 0 (2)
This equation is complemented by the Friedmann and acceleration equations and by
the continuity equation for the background:

ϕ̇2
3MP2 H 2 = + V (ϕ) + ρB
2
(3)
2MP2 Ḣ = −[ϕ̇2 + (1 + wB )ρB ]
ρ̇B = −3H(ρB + PB )

In order to better deal with the cosmological dynamics of the field, it’s useful to intro-
duce the following variables:
p
ϕ̇ V (ϕ)
x= √ y=√ (4)
6Mpl H 3Mpl H

In terms of these variables the Friedmann equation and the equation of state become:

x2 − y 2
Ωϕ + ΩB = 1 wϕ = (5)
x2 + y 2
ρ
with Ωϕ = 3M 2ϕH 2 = x2 +y 2 and ΩB = 3Mρ2BH 2 . Equation (2) is replaced by an equivalent
pl pl
system of ODEs for x and y:
√
′ 6 2 3
x = −3x + λy + x[(1 − wB )x2 + (1 + wB )(1 − y 2 )]
√ 2 2 (6)
′ 6 3 2 2
y =− λxy + y[(1 − wB )x + (1 + wB )(1 − y )]
2 2
V
where ′ = d/dN with N = log (a) and λ = −Mpl V,ϕ . The parameter λ is just the slow
roll parameter used for inflation and it obeys the equation:
V V,ϕϕ
q
′
λ = − 3(1 + w)Ωϕ (Γ − 1)λ2 Γ= (7)
V,ϕ2

4
Equations (6) together with (7) form a 3D dynamical system. Their solutions will
describe curves in the (x, y, λ) space. Note that x and y are bounded 0 ≤ x2 + y 2 ≤ 1,
for ρB > 0 and so the evolution of this system is completely described by trajectories
within the unit disc in the (x, y) plane. The system (6) is also symmetric under
the reflection (x, y) → (x, −y) and under time reversal t → −t. The last important
ingredient is the effective EoS parameter wef f of the universe:
Ptot PB + Pϕ
wef f = = = wB ΩB + wϕ Ωϕ = x2 − y 2 + wB (1 − x2 − y 2 ) (8)
ρtot ρB + ρϕ
The parameter wef f is of fundamental importance in our analysis because it tells us,
along the time evolution, in which epoch the Universe is in and whether it undergoes
an accelerating ( wef f < − 31 ) or a decelerating ( wef f > − 13 ) expansion.

3.2 Exponential potential

In order to study the dynamics of the field we need to choose a potential. The simplest
case arises in models with an exponential potential, where λ is a constant:
 
λϕ
V (ϕ) = V0 exp −
Mpl
The set of ODE (6) is sufficient to determine the evolution of the system. For what
follows we’ll consider only NR-matter i.e. wB = wm = 0. The first thing to do is to
find the critical points of the dynamical system and perform their stability analysis.
These can be found by setting x′ = y ′ = 0. There can be up to five critical points
depending on the numerical value of λ:
⋄ O = (x, y) = (0, 0), Ωϕ = 0, wef f = wm ;
⋄ A± = (x, y) = (±1, 0), Ωϕ = 1, wef f = wϕ = 1;
√ p
⋄ B = (x, y) = (λ/ 6, 1 − λ2 /6), Ωϕ = 1, wef f = w = −1 + λ2 /3;
p p
⋄ C = (x, y) = ( 3/2(1 + wm )/λ, [3(1 − wm 2 )/2λ2 ), Ω = 3(1 + w )/λ2 , w
ϕ m ef f =
wϕ = wm ;
The stability of the fixed points (x, y) = (xc , yc ) can be studied considering linear
perturbations about them, x = xc + δx and y = yc + δy. Inserting these expressions in
(6) and keeping only the first order in perturbation we get:
   
d δx δx
= M(xc , yc )
dN δy δy
If both the eigenvalues µ1 and µ2 of the matrix M are negative, the corresponding fixed
point is stable, i.e. an attractor. If either µ1 or µ2 is negative, the point corresponds
to a saddle. If both µ1 and µ2 are positive, the fixed point is unstable. The complete
analysis has been performed in various papers ([2, 3]) and here we only report the
results.
⋄ Point O, corresponding to a dust dominated or radiation dominated universe, it’s
a critical point for all values of λ. This point is always a saddle point attracting
trajectories along the x-axis and repelling them towards the y-axis. The effective
EoS parameter matches the matter EoS, wef f = wm and thus for physically
admissible values of wm there is no acceleration;

5
 ⋄ Points A± correspond to a universe dominated by the scalar field kinetic energy
(x2 = Ωϕ = 1) and thus the effective EoS parameter reduces to a fluid with
wef f = w = 1 and no acceleration. Their existence is always guaranteed and
they never represent stable points. They are unstable
√ or saddle points depending
on the value of λ being smaller or greater than 6;
⋄ Point B represents a scaling solution. By using Ωϕ = x2 + y 2 and the definition
(5) for wϕ it’s easy to show that, at this critical point, w = wϕ and wef f = w.
This means that the universe evolves under the influence of both the matter and
the scalar field, but it expands as if it were completely matter dominated. This
solution is of great physical interest since according to it, a scalar field could be
present in the universe hiding its effect on cosmological scales. However, since
for point B wef f = wm there cannot be accelerated expansion. When this point
exists, i.e. for λ2 ≥ 3(1 + wm ), it always represents a stable point attracting the
trajectories in the physical phase space;
⋄ Point C stands for the cosmological solution where the universe is completely
scalar field dominated. This implies Ωm = 0 and Ωϕ = x2 + y 2 = 1, meaning that
point C lies on the unit circle in the phase space. It exists for λ2 < 6 and is a
stable attractor for λ2 < 3(1+wm ) and a saddle point for 3(1+wm ) ≤ λ2 < 6. The
effective EoS parameter is wef f = λ2 − 1 which implies an accelerating universe
for λ2 < 2. In the limit λ → 0 we recover a universe dominated by a cosmological
constant.
We see that, for λ2 < 2, a dark matter to dark energy transition can be achieved by
an orbit connecting point O and point C. However this is not sufficient to solve the
coincidence problem. Since λ is constant, in order to obtain a late time transition
we would need to resort once again to a fine tuning of the initial conditions, which is
exactly what we want to avoid.

3.3 Tracker solution

The concepts of tracker field and tracker solution were first introduced in [9, 12] and
they offer a way to avoid the coincidence problem. Tracker fields have an equation of
motion with attractor-like solutions in which a wide range of initial conditions rapidly
converge to a common evolutionary track, solving the need for fine tuning. The condi-
tions for tracking Quintessence are:
(1) (Γ − 1) > 0;
(2) (Γ − 1) ≈ const over range of plausible initial conditions;
Condition (1), together with (7), implies a decreasing λ, meaning that eventually λ →
0. Using the definition of λ = −Mpl V,ϕ /V we see that this leads to an epoch of
accelerated expansion, i.e. V ≫ ϕ̇2 and wϕ ≈ −1. If condition (1) is satisfied then
Condition (2) makes sure that independently of the initial conditions, the rate at which
λ decreases remains more or less the same. This can be seen once again thanks to (7),
which implies λ′ ∝ (Γ − 1). If (Γ − 1) varies too abruptly for different initial conditions,
then different trajectories will lead to an epoch of accelerated expansion, but at different
times. Tracker solutions are also defined by the fact that wϕ is nearly constant and
−1 < wϕ ≤ wB . This means that when the universe is dominated by the background
( radiation or matter ), ρϕ decreases less rapidly than ρB . Eventually, ρϕ becomes
the dominant component leading, as λ approaches zero, the universe in a phase of

6
accelerated expansion. In [9], Steinhardt et.al derive an expression that relates wϕ to
w. Combining the equation of motion (2) with the definition (1) of wϕ we get:
s  
V,ϕ 1 p 1 d log x
± =3 1 + wϕ 1 + (9)
V 3Mpl2 Ωϕ 6 d log a

where x = (1 + wϕ )(1 − wϕ ) = 21 ϕ̇2 /V (ϕ). The ± signs depend on whether V,ϕ > 0 or
V,ϕ < 0 respectively. If we take the derivative with respect to ϕ of (9) and we combine
it with (9) we get an equation for Γ:

wB − wϕ 1 + wB − 2wϕ x′ 2 x′′
Γ=1+ − − (10)
2(1 + wϕ ) 2(1 + wϕ ) (6 + x′ ) (1 + wϕ ) (6 + x′ )2

where ′ = d/d log a. For a tracker solution Γ ≈ const. and wϕ ≈ const. which means
that x′ and x′′ are negligible. In this case we can use (10) to derive a relation between
wϕ and wB :
wB − 2(Γ − 1)
wϕ ≈ (11)
1 + 2(Γ − 1)
Since for tracking solutions Γ > 1 then wϕ < wB .

3.3.1 Inverse power law potential

The simplest potential that leads to tracker solutions is the inverse power law potential:

V (ϕ) = M 4+α ϕ−α (12)

The parameter Γ is given by:
α+1
Γ= (13)
α
and it is both greater than one and constant. Thus the inverse power law potential
satisfies the tracking conditions. Given Γ we can compute wϕ using equation (11):

αwB − 2
wϕ = (14)
2+α
As expected wϕ ≤ wB . For α ≫ 1 we get wϕ ≈ wB . Once we fix α, for every value of
M we have a different tracking solution. In this regard we can give a rough estimate
of the value M should have in order for ρϕ to match the current energy density:
(0)
ρϕ ≈ Mpl2 H02 ≈ 10−47 GeV4

If we assume that, at present, ϕ0 ≈ Mpl and ρ0ϕ ≈ V (ϕ0 ), using (12) we obtain the
mass scale: 1
M = (ρ0ϕ Mplα ) 4+α

7
3.3.2 Stability of tracker solutions
For potentials that allow tracking solutions λ is not constant and it becomes much
more difficult to perform an analysis just like the one done for the exponential case.
Nonetheless it’s possible to show is that the tracker solution is stable. We follow
[9]. Consider a solution in which wϕ differs from the tracker solution value w0 by an
amount δ. Then, equation (10) can be expanded to lowest order in δ to obtain after
some algebra:
1 9
δ ′′ + 3[ (wB + 1) − w0 ]δ ′ + (1 + wB )(1 − w0 )δ = 0 (15)
2 2
where as before ′ = d/d log a. The solution of this equation is δ ∝ aγ where:
r
3 1 i 1
γ = − [ (wB + 1) − w0 ] ± 18(1 + wB )(1 − w0 ) − 9[ (wB + 1) − w0 ]2 (16)
2 2 2 2
The real part of the exponent γ is negative for −1 < w0 ≤ 12 (1 + wB ), which includes
our range of interest. So, without imposing any further conditions, this means that δ
decays exponentially and the solution approaches the tracker solution. As δ decays,
it also oscillates with a frequency described by the second term. In deriving equation
(15) we assumed that Γ is strictly constant, which is true for pure inverse power law. It
is possible possible to prove that the tracker solution is stable also in the more general
case of (Γ − 1) approximately constant over the range of possible initial conditions.

4 Conclusions
As we’ve seen Quintessence and tracker solutions in particular, offer a viable way of
solving of solving the need for a fine tuning of the value of energy density at early time.
This however doesn’t mean that either of the two cosmological constant problems are
completely solved in this way. In order for the scalar field component to overtake the
matter density at the present epoch, we still need to adjust the scale of the potential,
as we have seen for (12), introducing another fine tuning problem.

References
[1] P. A. R. et. al Ade. Planck2013 results. xvi. cosmological parameters. Astronomy
amp; Astrophysics, 571:A16, October 2014.
[2] Sebastian Bahamonde, Christian G. Böhmer, Sante Carloni, Edmund J. Copeland,
Wei Fang, and Nicola Tamanini. Dynamical systems applied to cosmology: Dark
energy and modified gravity. Physics Reports, 775–777:1–122, November 2018.
[3] Edmund J. Copeland, Andrew R. Liddle, and David Wands. Exponential po-
tentials and cosmological scaling solutions. Physical Review D, 57(8):4686–4690,
April 1998.
[4] EDMUND J. COPELAND, M. SAMI, and SHINJI TSUJIKAWA. Dynamics
of dark energy. International Journal of Modern Physics D, 15(11):1753–1935,
November 2006.
[5] Daniel J. et. al Eisenstein. Detection of the baryon acoustic peak in the large-
scale correlation function of sdss luminous red galaxies. The Astrophysical Journal,
633(2):560–574, November 2005.

8
 [6] S. et. al Perlmutter. Measurements of and from 42 high-redshift supernovae. The
Astrophysical Journal, 517(2):565–586, June 1999.
[7] Adam G. et. al Riess. Observational evidence from supernovae for an ac-
celerating universe and a cosmological constant. The Astronomical Journal,
116(3):1009–1038, September 1998.
[8] D. N. et. al Spergel. First-year wilkinson microwave anisotropy probe ( wmap ) ob-
servations: Determination of cosmological parameters. The Astrophysical Journal
Supplement Series, 148(1):175–194, September 2003.
[9] Paul J. Steinhardt, Li-Min Wang, and Ivaylo Zlatev. Cosmological tracking solu-
tions. Phys. Rev. D, 59:123504, 1999.
[10] Shinji Tsujikawa. Quintessence: a review. Classical and Quantum Gravity,
30(21):214003, October 2013.
[11] JAEWON YOO and YUKI WATANABE. Theoretical models of dark energy.
International Journal of Modern Physics D, 21(12):1230002, November 2012.
[12] Ivaylo Zlatev, Li-Min Wang, and Paul J. Steinhardt. Quintessence, cosmic coin-
cidence, and the cosmological constant. Phys. Rev. Lett., 82:896–899, 1999.