Você está na página 1de 10

PHYSICAL REVIEW D VOLUME 23, NUMBER 2 15 JAN UAR Y 1981

Infiationary universe: A possible solution to the horizon and fiatness problems


Alan H. Guth*
Stanford Linear Accelerator Center, Stanford University, Stanford, California 94305
(Received 11 August 1980)
The standard model of hot big-bang cosmology requires initial conditions which are problematic in two ways: (1)
The early universe is assumed to be highly homogeneous, in spite of the fact that separated regions were causally
disconnected (horizon problem); and (2) the initial value of the Hubble constant must be fine tuned to extraordinary
accuracy to produce a universe as flat (i.e., near critical mass density) as the one we see today (flatness problem).
These problems would disappear if, in its early history, the universe supercooled to temperatures 28 or more orders
of magnitude below the critical temperature for some phase transition. A huge expansion factor would then result
from a period of exponential growth, and the entropy of the universe would be multiplied by a huge factor when the
latent heat is released. Such a scenario is completely natural in the context of grand unified models of elementary-
particle interactions. In such models, the supercooling is also relevant to the problem of monopole suppression.
Unfortunately, the scenario seems to lead to some unacceptable consequences, so modifications must be sought.

I. INTRODUCTION: THE HORIZON AND FLATNESS completely described.


PROBLEMS Now I can explain the puzzles. The first is the
The standard model of hot big-bang cosmology well-known horizon problem. 2 The initial uni-
relies on the assumption of initial conditions which verse is assumed to be homogeneous, yet it con-
are very puzzling in two ways which I will explain sists of at least -10" separate regions which are
below. The purpose of this paper is to suggest a causally disconnected (i. e. , these regions have
modified scenario which avoids both of these puz- not yet had time to communicate with each other
zles. via light signals). ' (The precise assumptions
By "standard model, " I refer to an adiabatically which lead to these numbers will be spelled out in
expanding radiation- dominated universe described Sec. II. ) Thus, one must assume that the forces
by a Robertson-%alker metric. Details will be which created these initial conditions were capable
given in Sec. II. of violating causality.
Before explaining the puzzles, I would first like The second puzzle is the flatness problem. This
to clarify my notion of "initial conditions. The " puzzle seems to be much less celebrated than the
standard model has a singularity which is conven- first, but it has been stressed by Dicke and Pee-
tionally taken to be at time t =0. As t the-0, bles. I feel that it is of comparable importance
temperature T ~. Thus, no initial-value prob- to the first. It is known that the energy density p
lem can be defined at t=0. However, when T is of the universe today is near the critical value p
of the order of the Planck mass (Mz, = I/~6=1. 22 (corresponding to the borderline between an open
&&10~~ GeV)' or greater, the equations of the stan- and closed universe). One can safely assume that~
dard model are undoubtedly meaningless, since
quantum gravitational effects are expected to be-
0. 01 & Q& 10, (
come essential. Thus, within the scope of our where
knowl, edge, it is sensible to begin the hot big-bang
0
= p/p= (8w/3)Gp/H2,
scenario at some temperature To which is com-
fortably below Mp, let us say To 10" GeV. At and the subscript p denotes the value at the present
this time one can take the description of the uni- time. Although these bounds do not appear at first
verse as a set of initial conditions, and the equa- sight to be remarkably stringent, they, in fact,
tions of motion then describe the subsequent evolu- have powerful implications. The key point is that
tion. Of course, the equation of state for matter the condition 0=1 is unstable. Furthermore, the
at these temperatures is not really known, but one only time scale which appears in the equations for
can make various hypotheses and pursue the con- a radiation-dominated universe is the Planck time,
sequences. 1/I= 5. 4 10 sec. A typical closed universe
&&

In the standard model, the initial universe is will reach its maximum size on the order of this
taken to be homogeneous and isotropic, and filled time scale, while a typical open universe will
with a gas of effectively massless particles in dwindle to a value of p much less than p. A uni-
thermal equilibrium at temperature To. The ini- verse can survive -10' years only by extreme fine
tial value of the Hubble expansion "constant" H is tuning of the initial values of p and H, so that p is
taken to be Ho, and the model universe is then very near p. For the initial conditions taken at
ALAN H. GUTH

To 10 GeV, the value of Ho must be fine tuned H


k
+ R2 = 8m Gp (2. 2b)
to an accuracy of one part in 10". In the standard 3
model this incredibly precise initial relationship where H = R/R is the Hubble "constant" (the dot de-
must be assumed without explanation. (For any notes the derivative with respect to f). Conserva-
reader who is not convinced that there is a real tion of energy is expressed by
problem here, variations of this argument are giv-
en in the Appendix. ) (2. 3)
The reader should not assume that these incredi-
ble numbers are due merely to the rather large where p denotes the pressure. In the standard
value I have taken for 7.'0. If I had chosen a modest model one also assumes that the expansion is adi-
value such as To 1 MeV, I mould still have con- abatic, in which case
cluded that the "initial" universe consisted of at
least -10" causally disconnected regions, and that (sR )=0 (2. 4)
the initial value of Ho was fine tuned to one part in dt
10 . These numbers are much smaller than the where s is the entropy density.
previous set, but they are still very impressive. To determine the evolution of the universe, the
Of course, any problem involving the initial above equations must be supplemented by an equa-
conditions can always be put off until we under- tion of state for matter. It is now standard to de-
stand the physics of T ~M&. However, it is the scribe matter by means of a field theory, and at
purpose of this paper to show that these puzzles high temperatures this means that the equation of
might be obviated by a scenario for the behavior state is to a good approximation that of an ideal
-
of the universe at temperatures mell below M&. quantum gas of massless particles. Let N, (T) de-
The paper is organized as follows. The assump- note the number of bosonic spin degrees of free-
tions and basic equations of the standard model are dom which are effectively massless at temperature
summarized in Sec. II. In Sec. ID, I describe the T (e. g. , the photon contributes two units to N, );
inflationary universe scenario, showing how it can and let N&(T) denote the corresponding number for
eliminate the horizon and flatness problems. The fermions (e. g. , electrons and positrons together
scenario is discussed in the context of grand mod- contribute four units). Provided that T is not near
els in Sec. IV, and comments are made concerning any mass thresholds, the thermodynamic functions
magnetic monopole suppression. In Sec. V I dis- are given by
cuss briefly the key undesirable feature of the
scenario: the inhomogeneities produced by the = 3p =
7TR

p 3t(T)T', (2. 6)
random nucleation of bubbles. Some vague ideas 30
which might alleviate these difficulties are men-
tioned in Sec. VI. ' (2. 6)

II. THE STANDARD MODEL OF THE VERY EARI Y


UNIVERSE n =, 3V(T)T' (2. 7)

In this section I will summarize the basic equa- where


tions of the standard model, and I will spell out the -
assumptions which lead to the statements made in m{T) =N, (T)+ '.N, (T}, (2. 6)
the Introduc tion. 3t'(T) =N, (T) + ', N, (T) . - (2. 9}
The universe is assumed to be homogeneous and
isotropic, and is therefore described by the Rob- Here n denotes the particle number density, and
ertson-Walker metric: t(3) = 1. 202 06. . . is the Riemann zeta function.
dJ' The evolution of the universe is then found by
dv =dF R~(t) 1 0
+r (dB +sin 9dg ) rewriting (2. 2b) solely in terms of the tempera-
ture. Againing assuming that T is not near any
(2. 1)
mass thresholds, one finds
where 4=+1, 1, or 0 for a closed, open, or
flat universe, respectively. It should be empha- + e(T)T' = 4~' Got(T)T4, (2. 10)
sized that any value of k is possible, but by con-
~

vention r and R(t) are rescaled so that k takes on where


one of the three discrete values. The evolution of
R(t) is governed by the Einstein equations u ~
2v'~&T} "' (2. 11)
RT i45 S
R =
4n
G(p+ 3p)R, (2. 2a) where S
= A s denotes the total entropy in a volume
INFLATIONARY UNIVERSE: A POSSIBLE SOLUTION TO. . .

specified by the radius of curvature R. ratio of volumes, so


Since S is conserved, its value in the early uni- 's~
l 11 45 (gs Mp
verse can be determined (or at least bounded) by I
43 4~') L, TT )
current observations. Taking p & 10p today' it
follows that today = 4x10-ss&-t s(M~/T}s (2. 20)
k
9H2. Taking 'X- 10s and Ts 10" GeV, one finds ls'/Lo
R2 & (2. 12) =10 . This is the horizon problem.
From now on I will take k=+ 1; the special case k
=0 is still included as the limit R ~. Then today III. THE INFLATIONARY UNIVERSE
B + 3 g - 3 x 2 0 years. Taking the present photon In this section I will describe a scenario which
temperature Tas 2. 7'K, one then finds that the is capable of avoiding the horizon and flatness
photon contribution to S is bounded by problems.
From Sec. II one can see that both problems
S& 3 &&1085 (2. ia)
could disappear if the assumption of adiabaticity
Assuming that there are three species of massless were grossly incorrect. Suppose instead that
'
neutrinos (e, p, and r), all of which decouple at a Z
time mhen the other effectively massless particles Sp So, (3. i)
are the electrons and photons, then S=21/22S. where S~ and So denote the present and initial val-
Thus, ues of R s, and Z is some large factor.
S& 1086 Let us look first at the flatness problem. Given
(2. 14
(3. 1), the right-hand side (RHS) of (2. 16) is multi-
and plied by a factor of Z . The "initial" value (at Ts
0&s3ts' =10'~ GeV) of Ip p I/p could be of order unity,
s
. (2. iS)
and the flatness problem would be obviated, if
But then
Z& 3 X102'. (3. 2)

s~ &3x 10 ssX (M /T) Now consider the horizon problem. The RHS of
%2
~

4m'
P I

(2. 19) is multiplied by Z,


which means that the
(2. ie) length scale of the early universe, at any given
Taking T= 10' GeV and%-10 (typical of grand temperature, was smaller by a factor of Z than
unified models), one finds Ip p I/p & 10+s. This had been previously thought. If Z is sufficiently
is the flatness problem. large, then the initial region which evolved into
The sT term can now be deleted from (2. 10), our observed region of the universe would have
which is then solved (for temperatures higher than been smaller than the horizon distance at that time.
all particle masses) to give To see how large Z must be, note that the RHS of
(2. 20) is multiplied by Z . Thus, if
P
(2. 17) 5&20",
2yt ' Z& (3. 3}
where ys = (4v /45)X. (For the minimal SUs grand then the horizon problem disappears. (It should be
unified model, N, =82, N& 90, and y=21. 05. ) noted that the horizon will still exist; it will simply
Conservation of entropy implies RT= constant, so be moved out to distances which have not-been ob-
A o-t" 2. A light pulse beginning at t=0 will have served. )
traveled by time t a physical distance It is not surprising that the RHS's of (3. 2) and
t (3. 3) are approximately equal, since they both
l(t) =R(t) dt'R (t') = 2t, (2. iS) correspond roughly to Sp of order unity.
0
I will now describe a scenario, which I call the
and this gives the physical horizon distance. This inflationary universe, which is capable of such a
horizon distance is to be compared with the radius large entropy production.
L(t) of the region at time t which will evolve into Suppose the equation of state for matter (with all
our currently observed region of the universe. A- chemical potentials set equal to zero} exhibits a
gain using conservation of entropy, first-order phase transition at some critical tem-
perature T, . Then as the universe cools through
(2. i9)
the temperature Tone would expect bubbles of
where s~ is the present entropy density and L~ the low-temperature phase to nucleate and grow.
-10' years is the radius of the currently observed However, suppose the nucleation rate for this
region of the universe. One is interested in the phase transition is rather low. The universe will
ALA% H. GUTH

continue to cool as it expands, and it will then su- theory, the form of T~ is determined by the con-
percool in the high- temperature phase. Suppose servation requirement
~gatv up to the possible modifica-
that this supercooling continues down to some tem- tion
perature Tmany orders of magnitude below T, . +
Ttt, v Ttt v s (s. 6)
When the phase transition finally takes place at
temperature T the latent heat is released. How- for any constant X. (X cannot depend on the values
ever, this latent heat is characteristic of the ener- of the fields, nor can it depend on the temperature
gy scale Twhich is huge relative to T, . The uni- or the phase. ) The freedom to introduce the mod-
verse is then reheated to some temperature T ification (3. 6) is identical to the freedom to intro-
which is compa, rable to T, . The entropy density is duce a cosmological constant into Einstein's equa-
then increased by a factor of roughly (T/T, ) (as- tions. One can always choose to write Einstein's
suming that the number X of degrees of freedom equations without an explicit cosmological term;
for the two phases are comparable), while the val- the cosmological constant A is then defined by
ue of R remains unchanged. Thus,
(0 T, I 0) =Ag. ,
[ (3. 7)
Z= T/T, . (s. 4) where Io) denotes the true vacuum. A is identified
If the universe supercools by 28 or more orders of as the energy density of the vacuum, and, in prin-
magnitude below some critical temperature, the ciple, there is no reason for it to vanish. Empir-
horizon and flatness problems disappear. ically A is known to be very small ( IA 10+6 I ("
In order for this scenario to work, it is neces- GeV ) so I will take its value to be zero. The
sary for the universe to be essentially devoid of value of p0 is then necessarily positive and is de-
any strictly conserved quantities. Let n denote the termined by the particle theory. '2 It is typically
density of some strictly conserved quantity, and of 0(r, ').
let y= n/s denote the ratio of this conserved quan- Using (3. 5), Eq. (2. 10) becomes
tity to entropy. Then x~ = Z 3x0 + 10 +0, Thus,
only an absurdly large value for the initial ratio
T2
T
4n
45
Gx(r)r -e(r)r + 8w Gp 0'. (3. 8)
would lead to a measurable value for the present
ratio. Thus, if baryon number were exactly con- This equation has two types of solutions, depending
served, the inflationary model would be untenable. on the parameters. If e& e0, where
However, in the context of grand unified models, 8w2v' 30
baryon number is not exactly conserved. The net 0 GvPO (s. 9)
45 ~

baryon number of the universe is believed to be


created by CP-violating interactions at a tempera- then the expansion of the universe is halted at a
ture of 10 -10 GeV. Thus, provided that T, lies temperature T given by
in this range or higher, there is no problem. The
baryon production would take place after the re- r '=, 3Pp '-
~ ~-(~ 2
+ ~
g(60
q
1/2 2
(s. io)
heating. (However, strong constraints are im- 0

posed on the entropy which ca.n be generated in any and then the universe contracts again. Note that
phase transition with T, 10 GeV, in particular,
the Weinberg-Salam phase transition. )
T
io.
is of this is not the desired scenar-
0(T,), so
The case of interest is e& e0, in which case the
Let us examine the properties of the supercooling expansion of the universe is unchecked. [Note that
universe in more detail. Note that the energy den- co-v9LT, '/M~' ispresumably avery small number.
sity p(T), given in the standard model by (2. 5), Thus 0( e & eo (a closed universe) seems unlikely,
must now be modified. As T-O, the system is but e & 0 (an open universe) is quite plausible. ]
cooling not toward the true vacuum, but rather to- Once the temperature is low enough for the p0 term
ward some metastable false vacuum with an energy to dominate over the other two terms on the HHS
density p0 which is necessarily higher than that of of (3. 8), one has
the true vacuum. Thus, to a good approximation T(t) = const&&e "', (3. 11)
(ignoring mass thresholds)
where
p(T) = Ot(r)T' + po. (s. 5)

X' = Sm' G~0 ~ (3. 12)
3
Perhaps a few words should be said concerning
the zero point of energy. Classical general rela- Since RT = const, one has"
tivity couples to an energy-momentum tensor of
matter, T~, which is necessarily (covariantly)
R(t) = const&& e" '.
conserved. When matter is described by a field The universe is expanding exponentially, in a false
IN F LATIONAR 7 UNIVERSE: A POSSIBLE SOLUTION TO. . . 351

vacuum state of energy density pp. The Hubble formed at time t&.
constant is given by H=R/R=y. (More precisely, I will now' assume that the nucleation rate is suf-
H approaches X monotonically. from above. This ficiently slow so that no significant nucleation takes
behavior differs markedly from the standard rnod- place until TTwhen exponential growth has set
el, in which H falls as t"'. ) in. I will further assume that by this time &(t) is
The false vacuum state is Lorentz invariant, so given approximately by the zero-temperature nu-
T& = p'pg+ . I't follows tha't p = pp the pressure cleation rate Xp. One then has
is negative. This negative pressure allows for the
conservation of energy, Eq. (2. 3). From the sec- P(t) = exp +0(1) (3. Is)
ond-order Einstein equation (2. 2a), it can be seen
that the negative pressure is also the driving force where
behind the exponential expansion.
The Lorentz invariance of the false vacuum has T =
3X
(s. iv)
4nzp'
one other consequence: The metric described by
(3. 13) (with @=0) does not single out a comoving and O(1) refers
to terms which approach a constant
frame. The metric is invariant under an O(4, 1) as Xt ~.
During one of these time constants, the
group of transformations, in contrast to the usual universe will expand by a factor
Robertson-Walker invariance of O(4). It is known 3X4 l
as the de Sitter metric, and it is discussed in the Z, = exp(}t~) = exp ' (s. is}
s tandard literature. ' 4mXp)

Now consider the process of bubble formation in If the phase transition is associated with the ex-
a Robertson-Walker universe. The bubbles form pectation value of a Higgs field, then Xp can be cal-
randomly, so there is a certain nucleation rate culated using the method of Coleman and Callan. "
X(t), which is the probability per (physical) volume The key point is that nucleation is a tunneling pro-
per time that a bubble will form in any region cess, so that &p is typically very small. The
which is still in the high-temperature phase. I Coleman-Callan method gives an answer of the
will idealize the situation slightly and assume that form
the bubbles start at a point and expand at the speed
&0 =A po exp(-B), (s. 19}
of light. Furthermore, I neglect k in the metric,
so d7 =dt R (t)dx . where B is a barrier penetration term and A is a,
I want to calculate p(t), the probability that any dimensionless coefficient of order unity. Since Z,
given point remains in the high-temperature phase
at time t. Note that the distribution of bubbles is
totally uncorrelated except for the exclusion prin-
very easily' '
is then an exponential of an exponential, one can
obtain values as large as log&pZ
=28, or even log&pZ=10
ciple that bubbles do not form inside of bubbles. Thus, if the universe reaches a state of exponen-
This exclusion principle causes no problem be- tial growth, it is quite plausible for it to expand
cause one can imagine fictitious bubbles which and supercool by a huge number of orders of mag-
form inside the real bubbles with the same nuclea- nitude before a significant fraction of the universe
tion rate X(t). With all bubbles expanding at the undergoes the phase transition.
speed of light, the fictitious bubbles mill be forever So far I have assumed that the ea, rly universe can
inside the real bubbles and will have no effect on be described from the beginning by a Robertson-
p(t). The distribution of all bubbles, real and fic- Walker metric. If this assumption were really
titious, is then totally uncorrelated. necessary, then it would be senseless to talk about
P(t) is the probability that there are no bubbles "solving" the horizon problem; perfect homogeneity
which engulf a given point in space. But the num- was assumed at the outset. Thus, I must now ar-
ber of bubbles mhich engulf a given point is a Pois- gue that the assumption can probably be dropped.
son-distributed variable, so P(t) =exp[-1V(t)], Suppose instead that the initial metric, and the
where Z(t) is the expectation value of the number distribution of particles, mas rather chaotic. One
of bubbles engulfing the point. Thus would then expect that statistical effects mould tend
to thermalize the particle distribution on a local
p(t) = exp dtqX(t~)R (tq) V(t, t~) (3. 14) scale. 2 It has also been shown (in idealized cir-
p
cumstances} that anisotropies in the metric are
where damped out on the time scale of -10' Planck

v(~, t, ) =-s, ' R(t,


4m dt
)
(s. 15)
times. The damping of inhornogeneities in the
metric has also been studied, 22 and it is reasonable
to expect such damping to occur. Thus, assuming
is the coordinate volume at time t of a bubble which that at least some region of the universe started at
ALAR H.

temperatures high compared to Tone would ex- order Mx/n-10~6 GeV, where n =g /4m is the
pect that, by the time the temperature in one of grand unified fine structure constant. Since the
these regions falls to T it will be locally homoge- monopoles are really topologically stable knots in
neous, isotropic, and in thermal equilibrium. By the Higgs field expectation value, they do not exist
locally, I am talking about a length scale $ which in the high-temperature phase of the theory. They
is of course less than the horizon distance. It will therefore come into existence during the course of
then be possible to describe this local region of the a phase transition, and the dynamics of the phase
universe by a Robertson-Walker metric, which will transition is then intimately related to the mono-
be accurate at distance scales small compared to pole production rate.
When the temperature of such a region falls be- The problem of monopole production and the sub-
low T the inflationary scenario will take place. sequent annihilation of monopoles, in the context of
The end result will be a, huge region of space which a second-order or weakly first-order phase transi-
is homogeneous, isotropie, and of nearly critical tion, was analyzed by Zeldovieh and Khlopov30 and
mass density. If Z is sufficiently large, this re- by Preskill.
' In Preskill's analysis, which was
gion can be bigger than (or much bigger than) our more specifically geared toward grand unified
observed region of the universe. models, it was found that relic monopoles would
exceed present bounds by roughly 14 orders of
IV. GRAND UNIFIED MODELS AND MAGNETIC magnitude. Since it gems difficult to modify the
MONOPOLE PRODUCTION estimated annihilation rate, one must find a scen-
In this section I will discuss the inflationary ario which suppresses the production of these
model in the context of grand unified models of monopole s.
elementary-par ticle interac tions. 2 ~ Kibble 2 has pointed out that monopoles are pro-
A grand unified model begins with a, simple gauge duced in the course of the phase transition by the
group G which is a valid symmetry at the highest process of bubble coalescence. The orientation of
energies. As the energy is lowered, the theory the Higgs field inside one bubble will have no cor-
undergoes a hierarchy of spontaneous symmetry relation with that of another bubble not in contact.
breaking into successive subgroups: G -0 When the bubbles coalesce to fill the space, it will
' ' '-Ho, where Hi SU3&&SU2&&Uq [@CD (quantum be impossible for the uncorrela, ted Higgs fields to
chromodynamics) & Weinberg-Salam] and Ho SU3 align uniformly. One expects to find topological
x U&". In the Georgi-Glashow model, which is knots, and these knots are the monopoles. The
the simplest model of this type, G = SU, and n = 1. number of monopoles so produced is then compar-
The symmetry breaking of SU& SU3 x SU2 x U$ oc- able to the number of bubbles, to within a few or-
curs at an energy scale M~- 10 GeV. ders of magnitude.
At high temperatures, it was suggested by Kirzh- Kibble's production mechanism can be used to
nits and Linde25 that the Higgs fields of any spon- set a "horizon bound" on monopole production
taneously broken gauge theory would lose their ex- which is valid if the phase transition does not sig-
pectation values, resulting in a high-temperature nificantly disturb the evolution of the universe.
phase in which the full gauge symmetry is re- At the time of bubble coalescence t the size l of
stored. A formalism for treating such problems the bubbles cannot exceed the horizon distance at
was developed by Weinberg and by Dolan and that time. So
Jackiw. In the range of parameters for which the
tree potential is valid, the phase structure of the l & 2tcoal = (4. 1)
' ' We 7 T coal 2
SU5 model was analyzed by Tye and me. ~

found that the SU5 symmetry is restored at T


By Kibble's argument, the density nof monopoles
& -10' GeV and that for most values of the param-
then obeys
eters there is an intermediate-temperature phase
with gauge symmetry SU4xU&, which disappears at 3T co816
~ l'~3 + Y (4. 2)
T-10 GeV. Thus, grand unified models tend to
-
N
P
provide phase transitions which could lead to an in-
flationary scenaxio of the universe. By considering the contribution to the mass density
of the present universe which could come from 10~
Grand unified models have another feature with
GeV monopoles, Preskill ' concludes that
important cosmological consequences: They con-
tain very heavy magnetic monopoles in their parti- n/n& 10 ~4, (4. 3)
cle spectrum. These monopoles are of the type
discovered by 't Hooft and Polyakov, 2 and will be where n, is the density of photons. This ratio
present in any model satisfying the above descrip- changes very little from the time of the phase
tion. These monopoles typically have masses of transition, so with (2. 7) one concludes
28
'
INFLATIONARY UNIVERSE: A POSSIBLE SOLUTION TO. . .

] 0-24/2 qi/3 bubble. This energy can be thermalized only


(4. 4) when the bubble walls undergo many collisions.
2$ (3)
(ii} The de Sitter metric does not single out a co-
If T, -10' GeV, this bound implies that the uni- moving frame. The O(4, 1) invariance of the de
verse must supercool by at least about four orders Sitter metric is maintained even after the forma-
of magnitude before the phase transition is com- tion of one bubble. The memory of the original
pleted. Robertson-Walker comoving frame is maintained
The problem of monopole production in a strongly.
by the probability distribution of bubbles, but the
first-order phase transition with supercooling was local comoving frame can be reestablished only af-
treated in more detail by Tye and me. 6 We ter enough bubbles have collided.
showed how to explicitly calculate the bubble den- (iii) The size of the largest bubbles will exceed
sity in terms of the nucleation rate, and we con- that of the smallest bubbles by roughly a factor of
sidered the effects of the latent heat released in Z; the range of bubble sizes is immense. The
the phase transition. Our conclusion was that surface energy density grows with the size of the
(4. 4) should be replaced by bubble, so the energy in the walls of the largest
T, .&2&10" GeV, (4. 6) bubbles can be thermalized only by colliding with
other large bubbles.
where Trefers to the temperature just before (iv) As time goes on, an arbitrarily large frac-
the release of the latent heat. tion of the space will be in the new phase [see
Tye and I omitted the crucial effects of the mass (3. 16)). However, one can ask a more subtle
density po of the false vacuum. However, our work question about the region of space which is in the
has one clear implication: If the nucleation rate is new phase: Is the region composed of finite sepa-
sufficiently large to avoid exponential growth, then rated clusters, or do these clusters join together
far too many monopoles would be produced. Thus, to form an infinite region& The latter possibility
the monopole problem seems to also force one into is called "percolation. " It can be shown. that the
the inflationary scenario. 5 system percolates for large values of &0/Z4, but
In the simplest SU5 model, the nucleation rates that for sufficiently small values it does not. The
have been calculated (approximately) by Weinberg critical value of Ao/}t has not been determined,
and me. ' The model contains unknown parame- but presumably an inflationary universe would have
ters, so no definitive answer is possible. We do a value of Xo/y be'low critical. Thus, no matter
find, however, that there is a sizable range of pa- how long one waits, the region of space in the new
rameters which lead to the inflationary scenario. phase will consist of finite clusters, each totally
surrounded by a region in the old phase.
(v) Each cluster will contain only a few of the
V. PROBLEMS OF THE INFLATIONARY SCENARIO
largest bubbles. Thus, the collisions discussed in
As I mentioned earlier, the inflationary scenario (iii) cannot occur.
seems to lead to some unacceptable consequences. The above statements do not quite prove that the
It is hoped that some variation can be found which scenario is impossible, but these consequences
avoids these undesirable features but maintains the are at best very unattractive. Thus, it seems that
desirable ones. The problems of the model will be the scenario will become viable only if some mod-
discussed in more detail elsewhere, ' but for com- ification can be found which avoids these inhomog-
pleteness I will give a brief description here. eneities. Some vague possibilities will be men-'
The central problem is the difficulty in finding a tioned in the next section.
smooth ending to the period of exponential expan- Note that the above arguments seem to rule out
the possibility that the universe was ever trapped
stant as t - -
sion. Let us assume that X(t) approaches a con-
and T 0. To achieve the desired in a, false vacuum state, unless Xo/y4 ~ 1. Such a
expansion factor Z & 10'8, one needs Xo/y4 & 10 2 large value of Xo/y does not seem likely, but it
[see (3. 18)], which means that the nucleation rate is possible. '
is slow' compared to the expansion rate of the uni-
verse. (Explicit calculations show that &0/}t4 is VI. CONCLUSION
typically much smaller than this value. '8 '9 ~6) The
randomness of the bubble formation process then I have tried to convince the reader that the stan-
leads to gross inhomogeneities. dard model of the very early. universe requires the
To understand the effects of this randomness, assumption of initial conditions which are very im-
the reader should bear in mind the following facts. plausible for two reasons:
(i} All of the latent heat released as a bubble ex- (i) The horizon problem. Causally disconnected
pands is transferred initially to the walls of the regions are assumed to be near'ly identical; in par-
ALAN H.

ticular, they are simultaneously at the same tem- man and Leonard Susskind, and for the invaluable
perature. help I received from my collaborators Henry Tye
(ii) The flatness problem. For a fixed initial and Erick Weinberg. I would also like to acknowl-
temperature, the initial value of the Hubble "con- edge very useful conversations with Michael Aizen-
stant" must be fine tuned to extraordinary accura- man, Beilok Hu, Harry Kesten, Paul Langacker,
cy to produce a universe which is as flat as the one Gordon LasherSo- Young Pi, John Preskill, and
we observe. EdwardWitten. This work was supported by the
Both of these problems would disappear if the Department of Energy under Contract No. DE-
universe supercooled by 28 or more orders of AC03-76SF00515.
magnitude below the critical temperature for some
phase transition. (Under such circumstances, the
universe would be growing exponentially in time. )
However, the random formation of bubbles of the
new phase seems to lead to a much too inhomoge-
neous universe. APPENDIX: REMARKS ON THE FLATNESS
The inhomogeneity problem would be solved if PROBLEM
one couM avoid the assumption that the nucleation
rate X(t) approaches a small constant Xp as the
-0. This appendix is added in the hope that some
temperature T If, instead, the nucleation
skeptics can be convinced that the flatness problem
rate rose sharply at some T&, then bubbles of an
is real. Some physicists would rebut the argument
approximately uniform size would suddenly fill
given in Sec. I by insisting that the equations might
space as T fell to T&. Of course, the full advant-
make sense all the way back to t=0. Then if one
age of the inflationary scenario is achieved only if
fixes the value of II corresponding to some arbi-
T, &10 "T,. trary temperature Tone always finds that when
Recently Witten has suggested that the above
chain of events may in fact occur if the parameters
of the SU5 Higgs field potential are chosen to obey
Q -
the equations are extrapolated backboard in time,
-1 as t 0. Thus, they would argue, it is na-
tural for 0 to be very nearly equal to 1 at early
the Coleman-Weinberg condition4P (i. e. , that O'V/
times. For physicists who take this point of view,
8&fP=O at /=0). Witten has studied this possi-
the flatness problem must be restated in other
bility in detail for the case of the Weinberg-Salam
terms. Since Hz and Tz have no significance, the
ph3se transition. Here he finds that thermal tun-
model universe must be specified by its conserved
neling is totally ineffective, but instead the phase
transition is driven when the temperature of the quantities. In fact, the model universe is com-
@CD chiral-symmetry-breaking phase transition pletely specified by the dimensionless constant &
=
Ip/R2T2, where k and R are parameters of the
is reached. For the SU, case, one can hope that a
much larger amount of supercooling will be found;
Robertson-Walker metric, Eq. (2. 1). For our
however, it is difficult to see how 28 orders of universe, one must take lel &3&10~ . The prob-
magnitude could arise.
lem then is the to explain why le should have such
l

Another physical effect which has so far been left a startlingly small value.
out of the analysis is the production of particles Some physicists also take the point of view that
e=0 is plausible enough, so to them there is no
due to the changing gravitational metric. 2 This
effect may become important in an exponentially problem. To these physicists I point out that the
universe is certainly not described exactly by a
expanding universe at low temperatures.
In conclusion, the inflationary scenario seems
Robertson-Walker metric. Thus it is difficult to
like a natural and simple way to eliminate both the imagine any physical principle which would require
horizon and the flatness problems. I am publishing a parameter of that metric to be exactly equal to
this paper in the hope that it will highlight the ex- zero.
istence of these problems and encourage others to In the end, I must admit that questions of plausi-
find- some way to avoid the undesirable features of bility are not logically determinable and depend
the inflationary scenario.
somewhat on intuition. Thus I am sure that some
physicists will remain unconvinced that there real-
ly is a flatness problem. However, I am also sure
ACKNOWLEDGMENTS
that many physicists agree with me that the flatness
I would like to express my thanks for the advice of the universe is a peculiar situation which at
and encouragement I received from Sidney Cole- some point will admit a physical explanation.
INFLATIONARY UNIVERSE: A POSSIBLE SOLUTION TO. . .

~Present address: Center for Theoretical Physics, 38, 255 (1977).


Massachusetts Institute of Technology, Cambridge, ~3The effects of a false vacuum energy density on the
Massachusetts 02139. evolution of the early universe have also been con-
I use units for which @ = c =k (Boltzmann cons tant) sidered by E. W. Kolb and S. Wolfram, CAL TECH
= 1. Then 1 m= 5.068x 10~~ GeV, 1 kg= 5.610x 10 6 Report No. 79-0984 (unpublished), and by S. A. Blud-
GeV, 1 sec = 1.519 x 10+ GeV ~, and 1 'K = 8.617x 10 ~4 rnan, University of Pennsylvania Report No. UPR-
GeV. 0143T, 1979 (unpublished).
W. Rindler, Mon. No~. . B. Astron. Soc. 116, 663 {1956). More precisely, the usual invariance is O(4) if' =1,
See also Ref. 3, pp. 489-490, 525-526; and Ref. 4, pp. O(3, 1) ifk = -1, and the group of rotations and trans-
740 and 815. 1.ations in three dimensions ifk = 0.
S. Weinberg, Gravitation and Cosmology (Wiley, New See for example, Bef. 3, pp. 385-392.
York, 1972). 6A. H. Guth and S. -H. Tye, Phys. Rev. Lett. 44, 631
C. W. Misner, K. S. Thorne, and J. A. Wheeler, Grav- (1980); 44, 963 {1980).
itation (Freeman, San Francisco, 1973). S. Coleman, Phys. Rev. D 15, 2929 (197?); C. G. Cal-
5In order to calculate the horizon distance, one must of lan and S. Coleman, ibid. 16, 1762 (1977); see also
course follow the light trajectories back to t = 0. This S. Coleman, in The Whys of Subnuclear Physics,
violates my contention that the equations are to be proceedings of the International School of Subnuclear
trusted only for T & T p Thus, the horizon problem Physics, Ettore Majorana, Erice, 1977, edited by
could be obviated if the full quantum gravitational A, Zichichi plenum. New York. 19?9).
theory had a radically different behavior from the ~
J.
A. H. Guth and E. Weinberg, Phys. Rev. Lett. 45,
naive extrapolation. Indeed, solutions of this sort 1131 (1980).
have been proposed by A. Zee, Phys. Bev. Lett. 44, ~~E. J.
Weinberg and I are preparing a manuscript on
703 (1980) and by F. W. Stecker, Astrophys. J. 235, the possible cosmological implications of the phase
Ll (1980). However, it is the point of this paper to transitions in the SU& grand unified model.
show that the horizon problem can also be obviated by 2PJ. Ellis and G. Steigman, Phys. Lett. 89B, 186 (1980);
mechanisms which are more within our grasp, occur- J.
Ellis, M. K. Gaillard, and D. V. Nanopoulos, ibid.
ring at temperatures below Tp. SOB, 253 {1980).
R. H. Dicke and P. J. E. Peebles, General Relativity: 2~B. L. Hu and L. Parker, Phys. Bev. D 17, 933 (1978).
An Einstein Centenary Survey, edited by S. W. Hawk- See Ref. 4, Chap. 30.
ing and W. Israel {Cambridge University Press, K,on- 23The simplest grand unified model is the SU(5) model
don, 1979). of H. Georgi and S. L. Glashow, Phys. Rev. Lett. 32,
See Ref. 3, pp. 475-481; and Bef. 4, pp. 796-797. 438 (1974). See also H. Georgi, H. R. Quinn, and
For example, see Bef. 3, Chap. 14. S. Weihberg, ibid. 33, 451 {1974); and A. J. Buras,
9M. Yoshimura, Phys. Rev. Lett. 41, 281 {1978); 42, J. Ellis, M. K. Gaillard, and D. V. Nanopoulos, Nucl.
746 (E} (1979); Phys. Lett. 88B, 294 (1 979); S. Dimopou- . Phys. B135, 66 {1978).
los and L. Susskind, Phys. Bev. D 18, 4500 {1978); Other grand unified models include the SO(10) model:
Phys. Lett. 81B, 416 (1979); A. Yu Ignatiev, N. V.
H. Georgi, in Particles and Eields 1975, proceedings
Krashikov, V. A. Kuzmin, and A. N. Tavkhelidze, of the 1975 meeting of the Division of Particles and
ibid. 76B, 436 (1978); D. Toussaint, S. B. Treiman, Fields of the American Physical Society, edited by
F. Wilczek, and A. Zee, Phys. Bev. D 19, 1036 (1979); Carl Carlson (AIP, New- York, 1S75); H. Fritzsch and
S. Weinberg, Phys. Rev. Lett. 42, 850 (1979); D. V. P; Minkowski, Ann. Phys. (N. Y.) 93, 193 (1975);
,

Nanopoulos and S. Weinberg, Phys. Bev. D 20, 2484 H. Georgi and D. V. Nanopoulos, Phys. Lett. 82B,
{1979);J. Ellis, M. K. Gaillard, and D. V. Nanopoulos, 392 (1979) and Nucl. Phys. B155, 52 (1979). The E(6)
Phys. Lett. 80B, 360 {1979); 82B, 464 (1979); M. Hon- model: F. Gursey, P. Ramond, and P. Sikivie, Phys.
da and M. Yoshimura, Prog. Theor. Phys. 62, 1704 Lett. 60B, 177 (1975); F. Gursey and M. Serdaroglu,
(1979); D. Toussaint and F. Wilczek, Phys. Lett. 81B, Lett. Nuovo Cirnento 21, 28 (1978). The K(7) model:
238 {1979); S. Barr, G. Segre, and H. A. Weldon, F. Giirsey and P. Sikivie, Phys. Rev. Lett. 36, 775
Phys. Rev. D 20, 2494 {1979); A. D. Sakharov, Zh. (1976), and Phys. Bev. D 16, 816 (1977); P. Ramond,

Eksp. Teor. Fiz. 76, 1172 (1979) fSov. Phys. JETP Nucl. Phys. B110, 214 {1976). For some general
49, 594 {1979)];A. Yu Ignatiev, N. V. Krashikov, properties of grand unified models, see M. Gell-Mann,
V. A. Kuzmin, and M. E. Shaposhnikov, Phys. Lett. P. Ramond, and R. Slansky, Rev. Mod. Phys. 50, 721
87B, 114 (1979); E. W. Kolb and S. Wolfram, ibid. (1978). For a review, see P. Langacker, Report No.
91B, 217 (1980); Noel. Phys. B172, 224 (1980); J. N. SLAC-PUB-2544, 1980 (unpublished).
Fry, K. A. Olive, and M. S. Turner, Phys. Rev. D 22, D. A. Kirzhnits and A. D. Linde, Phys. Lett. 42B,
2953 (1980); 22, 2977 (1980); S. B. Treiman and 471 (1972).
F. Wilczek, Phys. Lett. 95B, 222 (1980); G. Senjanovic S. Weinberg, Phys. Bev. D 9, 3357 (1S74); L. Dolan
and F. W. Stecker, Phys. Lett. B (to be published). and R. Jackiw, ibid. 9, 3320 (1974); see also D. A.
~~The reason A is so small is of course one of the deep Kirzhnits and A. D. Linde, Ann. Phys. {Y.
) 101, 195
mysteries of physics. The value of A is not determined (1S76); A. D. Linde, Bep. Prog. Phys. 42, 389 (19?9).
,

by the particle theory alone, but must be fixed by &-expansion techniques are employed by P. Ginsparg,
whatever theory couples particles to quantum gravity. Nucl. Phys. B (to be published).
27
This appears to be a separate problem from the ones In the case that the Higgs quartic couplings are corn-
discussed in this paper, and I merely use the empiri- parable to g or smaller (g = gauge coupling), the
cal fact that A=O. phase structure has been studied by M. Daniel and
S. A. Bludman and M. A. Buderman, Phys. Rev. Lett. C. E. Vayonakis, CERN Report No. TH. 2860 1980
ALAN H. GUTH

(unpublished); and by P. Suranyi, University of Cin- been proposed by P. Langacker and S.-Y. Pi, Phys.
cinnati Report No. SO-0506 (unpublished). Rev. Lett. 45, 1 {1980). By modifying the Higgs
8G. 't Hooft, Nucl. Phys. 879, 276 {1974); A. M. Poly- structure, they have constructed a model in which the
akov, Pis'ma Zh. Eksp. Teor. Fiz. 20, 430 (1974) high-temperature SU5 phase undergoes a phase transi-
[JETP Lett. 20, 194 (1974)]. For a review, see tion to an SU3 phase at T 10~4GeV. Another phase
P. Goddard and D. I. Olive, Rep. Prog. Phys. 41, 1357 transition occurs at T -10 GeV, and below this
(1978). temperature the symmetry is SU3x U~ + . Monopoles
If II&(|") and 02(C) are both trivial, then II2(G/Kp) cannot exist until T + 10 GeV, but their production is
= IIf (0'p). In our case IIf +p) is the group of integers. negligible at these low temperatures. The suppression
For a general review of topology written for physicists, of monopoles due to the breaking of U~ symmetry at
see N. D. Mermin, Rev. Mod. Phys. 51, 591 (1979). high temperatures was also suggested by S. -H. Tye,
Y. B. Zeldovich and M. Y. Khlopov, Phys. Lett. 79B, talk given at the 1980 Guangzhou Conference on Theo-
239 (1978). retical Particle Physics, Canton, 1980 (unpublished).
@J. P. Preskill, Phys. Rev. Lett. 43, 1365 (1979). +The Weinberg- Salam phase transition has also been
T. W. B. Kibble, J. Phys. A 9, 1387 (1976). investigated by a number of authors: E. Witten, Ref.
This argument was first shown to me by John Pres- 41; M. A. Sher, Phys. Rev. D 22, 2989 (1980); P. J.
kill. It is also described by Einhorn et al. , Ref. 34, Steinhardt, Harvard report, 1980 (unpublished); and
except that they make no distinction between T~~ and A. H. Guth and E. J. Weinberg, Ref. 1S.
Tc' 37This section represents the work of E. J. Weinberg,
+The problem of monopole production was also exam- H. Kesten, and myself. Weinberg and I are preparing
ined by M. B. Einhorn, D. L. Stein, and D. Toussaint, a manuscript on this subject.
Phys. Rev. D 21, 3295 (1980), who focused on second- The proof of this statement was outlined by H. Kesten
order transitions. The structure of SU(5) monopoles (Dept. of Mathematics, Cornell University), with de-
has been studied by C. P. Dokos and T. N. Tomaras, tails completed by me.
Phys. Rev. D 21, 2940 (1980); and by M. Daniel, 39E. Witten, private communication.
G. Lazarides, and Q. Shafi, Nucl. Phys. B170, 156 4
J.
S. Coleman and E. Weinberg, Phys. Rev. . D 7, 1888
(1980). The problem of suppression of the cosmologi- (1973); see also, J.
Ellis, M. K. Gaillard, D. Nano-
cal production of monopoles is discussed by G. Laza- poulos, and C. Sachrajda, Phys. Lett. 83B, 339 (1979),
rides and Q. Shafi, Phys. Lett. 948, 149 {19SO), and and J.Ellis, M. K. GaiQard, A. Peterman, and
G. Lazarides, M. Magg, and Q. Shafi, CERN Report C. Sachrajda, Nucl. Phys. B164, 253 (1980),
No. TH. 2856, 1980 (unpublished); the suppression dis- 4~E. Witten, Nucl. Phys. B (to be published).
cussed here relies on a novel confinement mechanism, L. Parker, in Asymptotic Structure of Spacetime,
and also on the same kind of supercooling as in Ref. editedby F. Esposito and L. Witten (Plenum, New York,
16. See also J. N. Fry and D. N. Schramm, Phys. Rev. 1977); V. N. Lukash, I. D. Novikov, A. A. Starobin-
Lett. 44, 1361 (1980). sky, and Ya. B. Zeldovich, Nuovo Cimento 35B, 293
35An alternative solution to the monopole problem has (1976).

Você também pode gostar