..

·-
1 2

August 1991 LBL-31169 1 Introduction

UCB-PTH-9141
The Electroweak Phase Transition In the minimal standard model, electroweak symmetry breaking is induced
and Baryogenesis • by the ground state of a single doublet scalar field. We can write the potential for
the real scalar component of the doublet which acquires a vacuum expectation
value as
Greg W. Anderson t U(</>) = ~,\ (¢2- u2)2. (1.1)
and
In a cold and empty (or relatively empty) universe, the Higgs field </>can mini-
Lawrence J. Hall
mize its energy and hence the free energy of the system by choosing a vacuum
expectation value (¢) = u. However, the early universe was neither cold nor
Department of Physics
empty, and the presence of an ambient, thermal distribution of particles changes
University of California, Berkeley
this picture. Although the vacuum energy of the system is still reduoed by shift-
and
Lawrence Berkeley Laboratory
ing the classical value of the Higgs field away from (¢) =0, we now pay the
price of adding free energy to the particles in the surrounding plasma as they
1 Cyclotron Road, Berkeley, CA 94720.
acquire a mass (see Figure 2). When the temperature is high enough, the free
energy required to give mass to a thermal distribution of particles exceeds the
Abstract vacuum energy liberated by displacing the Higgs field vacuum expectation value
from the origin. This occurs for temperatures above a critical temperature we
We give an analytic treatment of the one Higgs doublet, electroweak phase tran- call T,, where the configuration of the Higgs field that minimizes the free energy
sition which demonstrates that the phase transition is first order. The phase of the system is (<I>) = 0. Thus, at temperatures large compared to the scale
transition occurs by the nucleation of thin walled bubbles and completes at a of electroweak physics, the minimum of the effective Higgs potential is at the
temperature where the order parameter, (¢)T is significantly smaller than it is origin. This is the origin of symmetry restoration at high temperature[I, 2, 3].
when the origin becomes absolutely unstable. The rate of anomalous baryon As the universe cools, when the temperature drops to the critical temperature
number violation is an exponentially sensitive function of (<I>)T· In very minimal T., a new minimum appears, separated from the origin by a hump. When the
extensions of the standard model it is quite easy to increase (<Ph so that anoma- free energy barrier r.;parating the two extrema is small enough, bubbles of true
lous baryon number violation is suppressed after the completion of the phase vacuum are nucleated and grow. The thermal decay rate from the unstable state
transition. Hence, baryogenesis at the electroweak phase transition is tenable in carries a suppression [4, 5],
minimal extensions of the standard model with one Higgs doublet.
r jV <X e-Fo(T)/T (1.2)
'
•This work was supported by the Director, Office of Energy Research, Office of High Energy
and Nuclear Physics, Djvision of High Energy Physics of the U.S. Department of Energy nnder where Fo(T) is the surplus free energy needed to create a bubble of true vacuum
contract DE-AC03-76SFOOD98 and in part by the National Science Foundation under grant large enough to grow indefinitely. Bubbles of true vacuum smaller than this crit-
PHY!J0.21139.
IAddress after September 27, 1991; Physics Department, Ohio State University, 174 W.
ical size collapse under surface tension. For a bubble larger than this critical size,
18th Street, Columbus, Ohio 43210. EJectronic mail: [email protected] or as the radius of the bubble increases the free energy liberated by the expanding
[email protected] volume of true vacuum exceeds the free energy required to increase the bubble's
 •' ..

Disclaimer

Thia docwnent was prepand M an accO\.mt of work spoll!IOn'!d by the United States Gov-
enunent. Neither the United StateA Government nor any agency thereof, nor The Regenl:.8
of the University of California, nor any of their employee!!!l, makes any warranty, exprese or
implied, or assumes any legal liability or nosponsibility for the accuracy, completeness, or use-
fulness of any infonnation, apparatUA, product, or process disclosed, or represents that its use
would not infringe privately owned righ1:.8. Reference herein to any specific conunen:ial prod-
ucts proceLV~, or service by its trade name, trademark, manufacturer, or otherwise, does not
necessarily constitute or imply its endoBement, reconunendation, or favoring by the United
States Government or any agf!ncy thereof, or The Regents of the University of California. The
views and opinioruJ of authors exp~ herein do not necessarily state or reflect thoee of the
United States Gove.nunent or any agency thereof of The Regents of the University of Califomia
and shall not be U8ed for advertising or product endonement p1.l.TJXI8e!!l.

Lawrence Berkeley Laborotory is an equal opportunity employer.

3 4

surface area. Such a bubble will grow and convert space to true vacuum. A of creating the baryon asymmetry at the electroweak phase transition, (EWPT),
static bubble which is exactly the critical size is in unstable equilibrium, it is a [11, 12, 13, 14, 15, 16, 17]. A successful scenario must explain why the baryon
saddle point solution of the free energy functional. excess created at the electroweak phase transition is not washed out after the
At temperature T,, where the second derivative of the potential at the origin phase transition completes. We show that extremely simple additions to the
vanishes, fluctuations can classically roll towards the global minimum without standard model avoid washout for any Higgs mass up to 150 GeV.
surmounting an energy barrier. If the phase transition has not yet completed by
the time the temperature drops to T,, the transition no longer occurs through 2 Evolution of the Potential
bubble nucleation. We call the transition first-order if it proceeds by bubble
nucleation. So a necessary condition for a first order phase transition is that The tree level potential for the physical Higgs scalar is
bubbles occupy most of space before the temperature drops to T,. We will
satisfy a slightly more stringent condition. At temperatures very close toT, the
loop expansion parameter becomes large[2]. So in addition, a reliable analytical
U(¢) = 7(¢2- u')'' (2.1)

determination of the phase transition requires that it complete while the effective where -\0 is related to the Higgs boson mass by m~ = 2-\ 0 u 2 . To reliably analyze
loop expansion parameter is smalL the dynamics of this field, we need to include the interactions of the Higgs field
with virtual particles and with the heat bath.
In the next section we write the finite temperature effective potential for
the one Higgs doublet standard model in a form which is useful for the analytic The one loop, zero temperature potential, V(¢) can be written as the sum
understanding of the nature of the phase transition. We take the Higgs boson =
of the classical potential and a one loop correction V ( ¢) U (¢) + V 1 ( ¢). If we
to be lighter than about 150 GeV, and use the high temperature approximation, adopt the renormalization prescriptions i) V"(u) = m~, and ii) V'(u) = 0, for
which we will show is highly accurate for all aspects of the phase transition. each degree offreedom to which the Higgs boson is coupled, the zero temperature
In section three we discuss the nature of bubbles which could be nucleated to one loop correction to the effective potential is (see Appendix II)
trigger the phase transition. The scaling arguments of section three suggest 1
that an examination of whether the phase transition occurs v"1a nudeation of
1--\(¢) = ± -
64rr
,{m (¢)ln (m (¢)/m (u))
4 2 2

(22)
thin walled bubbles is warranted. In section four, the free energy of these thin 3 4 (¢) + 2m2 (¢)m2 (u)- -m
I 4 (u) } ,
-:)m
walled bubbles is calculated, and in section five it it shown that the thin walled - 2
bubble free energy is small enough that the phase transition does complete by where the± is for bosons (fermions) and m(¢) is the mass of the particle in
the rapid nucleation of exclusively thin walled bubbles. Our analysis is purely the presence of a background field ¢. Equation (2.2) is valid for particles which
analytic, and we obtain formulae for such quantities as the temperature of the have a mass of the form m 2 =a+ b¢ 2 in the mass eigenstate basis. In addition
universe at the completion of the phase transition and the number of bubbles to these quantum corrections, we must also include the interaction between the
nucleated per horizon volume. Higgs field and the hot electroweak plasma. Taking the Higgs boson sufficiently
Section six provides an important application of our results to the ques- light that we can ignore the contribution of scalar loops, it is shown in Appendix
tion of depletion of the baryon asymmetry after the phase transition. The II that, using the high temperature expansion, the effective potential for the
standard model contains an anomaly which is baryon number violating [6]. At standard model can be reliably written
high temperatures the rate of anomalous baryon number violation can be quite
large[5, 7, 8, 9, 10]. This has stimulated a great deal of interest in the possibility V(¢, T) = D(T2 - T?)¢ 2 - ET¢ 3 + ~,\T¢\ (2.3)
 5 6

where D =f., {6(mw/17) 2 + 3(mz/17) 2 + 6(m,jo-) 2 ), and the coefficient of the For Higgs Boson masses above the current experimental limit, the difference in
term linear in temperature is E = 1 ~" [6(mw/17) 3 + 3(mz/17) 3 ] co: 10-2 The temperature between T1 and T2 is small compared to the temperature. Writing
temperature dependent ¢>4 coupling is: T, = T, + r, we find r << T, provided mH ;':; 10 GeV, where

AT= A- Lg
B
8 m:
1617
4

1T
2
2 ln(m!/c8 T ) + F
4
LgF 16m; 2 ln(m~/cFT ),
17 1T
2
(2.4) T = (2~2D) T,. (2.10)

where the B(F) denotes bosons (fermions), gB(F) is the number of degrees of From equations (2.8) and (2.9) we see that
freedom, Cp and c8 are constants which can be found in appendix II, and the
" (T,0 ) = 3ET, ± ET,
masses in (2.4) are evaluated at (¢>) = 17. The physical Higgs mass is related to 'I'± 2AT 2AT
A by (2.11)
" (T ) _ 3ET, 3ET,
m~ = (2A + 128)<7 2 , (2.5) 'I'± ,-2AT±2AT.

where B = ; a
64 2 4
( 6m~ + 3mi - 12m~). We define Tl as the temperature It will prove convenient to write the potential in terms of the scaled field ¢>',
where V"(¢> = 0) = 0. From Appendix II, where ¢> = ( P,:)
¢>', for which

= m~ -
2
BBo-
V(,P)- A (ET)• (T";(T2-T,2),p'2-¢>'"+ ~¢>,.)
2 2( ) 2
T~ An. =X mo mH mw (2.6) (2.12)
- T AT T2(T,2 - 7{) 4 .
Because this result was obtained at the origin, it is valid to all orders in m/T.
If T, - T, < < T, + T,, then for any T such that IT, - Tl << T,, the potential
The quantity x is plotted against mH for various values of m, in Figure 1.
reduces to the simple form
At temperatures well above T,, the only minimum of the potential is (¢>) = 0.
As the early universe cools down from this high temperature, a second local min-
imum of the potential first appears (as an inflection point) when the temperature
V(¢>) =AT ( ~~r ((1- c)¢>' 2
- ¢>'" + ~¢>,.), (2.13)

reRChes where c = (T, - T)/(T,- T,). In terms of the scaled field

2 1
T_' = T, 1 _ 9£2 j8ATD (2.7)
1>± = 21 (3± Jl+Bc) (2.14)
at a value of the field <f>, = 3ET./2AT. At lower temperatures, this point splits
into a barrier 1>- and a local minimum ¢>+ which subsequently evolve as The critical temperatures T., T., and T, correspond to the following values of c:

<f>± = -3ET 1
± ---;- .j9£ 2T2 -
£, = -3/8, c, = 0, £2 = 1. (2.15)
8ATD (T2 - 7{) · (2.8)
2AT 2AT
The largest values of mfT which are of importance in this paper correspond
The evolution of¢>± is shown in Figure 2. We define the temperature T, to be the
toT co: T, and¢> co: ¢>+(T,). Is the high temperature expansion a good approxi-
temperature at which the second minimum becomes degenerate with the origin,
mation in this case? If the high temperature approximation is valid for the top
V(¢>+(T,)) = 0. Hence, if we divide equation (2.3) by ¢> 2, T, occurs where the
quark, it will be valid for all other particles as well. In the temperature region
resulting quadratic equations has two real equal roots. This gives the relation
of interest, for the top quark we have
2 1 2
T~= £lT'J. (2.9) m,(T) ~ .x,<f>+(T,) ~ .X,2E
1- - T- T, - .xT· (2.16)
\TD
7 8

In Appendix II we show that the high temperature approximation is valid to the height and the width of the barrier near ¢_ are not small compared to the
better than 5 percent provided this quantity is less than 1.6 . An inspection depth and width of the well at¢+, for the optimal solution, ¢will change quickly
of Figures three and four demonstrates that our use of the high temperature between 0 and ¢', and 6R will be small. This is the situation for temperatures
approximation is well justified. just below T,. Hence, the first bubbles which could be formed are thin wall
bubbles.

3 A Heuristic Discussion of the Saddle Point As tbe temperature subsequently drops towards T,, tbe barrier in ¢ space
tends to zero, and the difference in free energy density between the states (¢) = 0
After the universe cools down to a temperature below T,, the previously and (¢) = ¢+ increases. When the size of the hump in the potential at q,_
global minimum (¢) = 0 becomes metastable. The subsequent conversion of the becomes small compared to the depth of the well at ¢+, it is favorable to make
universe to the true vacuum state (¢) = ¢+(T) takes place by the nucleation of 6R as large as possible so as to minimize the 'surface' term Fs. Hence, 6R ~ R,
true vacuum bubbles. Accordingly, we need to determine the free energy barrier and we should work in a thick wall approximation. So, whether the EWPT
such bubbles must surmount in order to grow. Consider a true vacuum bubble proceeds by the nucleation of thick walled or thin walled bubbles depends on
in a sea of false vacuum (¢) = 0. Let (¢) = ¢'at the center of the bubble (see how large the rate of bubble nucleation becomes before thick walled bubbles
Figure 5). are energetically preferred. Our purpose, throughout this section, is to gain
a qualitative understanding of the dependence of bubble free energies on the
By convention we choose the state (¢) = 0 to have free energy zero, V(O) = shape of the effective potential. Accordingly, we shall not be too concerned
0. Then the surplus free energy of a true vacuum bubble is
with the precise value of the numerical prefactors which occur in estimates of
bubble parameters. Simple qualitative and semi-quantitative estimates of these
F= jc?xg(v¢)'+V(¢,T)}. (3.1)
two cases will tell us what kind of bubble to examine with closer scrutiny. Let's
first consider the simpler case of thick walled bubbles.
The free energy of a true vacuum bubble has two contributions: a surface
• Thick Walled Bubbles
free energy Fs, coming from the derivative terms in (3.1), and a volume term
Fv, which arises from the difference in free energy density inside and outside For thick walled bubbles 6R ~ R, and the surface energy of the bubble
the bubble. These two contributions scale like grows like R. In contrast, the negative volume term increases in magnitude
like R". Thus, a thermal fluctuation producing a bubble of true vacuum, which
2
Fs ~ 211-R2 (6¢)
R 6R, starts from a radius of zero and expands in radius to envelope the system must
1
6 have a free energy greater than or equal to some critical value. The critical
(3.2)
4rr n3- radius of the bubble, Rc, is the bubble radius where the total free energy of the
Fv~- -nV
3 ' bubble reaches a maximum. Differentiating (3.2), Rc ~ 6¢jV2V, and the thick
where R is the radius of the bubble, 6R is the thickness of the bubble wall, wall bubble free energy is
6¢ = ¢', l and V is minus the average value of the potential inside the bubble. 4rr R'.-
When V(¢-) is large compared to -V(¢'), it is important to minimize the
Fcce2rrRc(6¢) 2 -3 cV
(3.3)
contribution to Fv coming from regions near ¢ ce ¢_. More precisely, when (6¢)'
~JV
t For thin walled bubbles ¢'1 will lie at. the absolute minimum, while for thick walled bubbles
we must allow for the possibility that 4>' is somewhat Jess than¢+· Note that for thick walled bubbles, the magnitude ofF" depends on the relative
9 10

sizes of the shift in ¢ and the potential difference between the center and the make a careful study of bubble free energies in the thin wall approximation the
outside of the bubble and not the height of the barrier in ¢ -space. § AB a instant after the temperature drops below T,.
quick estimate for the potential (2.3), defining r = 1 - <, as the temperature
approaches T, the critical free energy scales like
4 Saddle Point Free Energies in the Thin Wall
F - r 3/2
ET (3.4) Approximation
" ,...., A;/2

• Thin Walled Bubbles We can adapt Coleman's thin wall approximation[18] to derive an analytic
For thin walled bubbles, in addition to the contribution of the derivative formula for the critical free energy valid in the limit that the temperature ap-
term to the bubble wall free energy, inside the bubble walls there is also a proaches T, from below.~ The thin wall analysis has been applied to the general
positive contribution to the free energy from the barrier in ¢-space, vb ~ v_. case of thermal vacuum stability by Linde [19]. However we find his conclusions
AB a function of the bubble radius and thickness, the thin wall bubble free energy regarding the applicability of the thin wall approximation to realistic gauge the-
IS ories overly pessimistic. For the potential of equation (2.3), if we shift our field
6¢ 47T - ¢ _, <!>' = ¢ + (ET/ Ar), we can cast the potential in a form where the validity
F(R) ~ 2rrR2 6R { ( R
6 )
2
+ 2Vb } - 3R"V. (3.5)
of Coleman's thin wall approximations will be transparent. Recall6 E 2 JArD, =
< = (T, - T)j(T, - T,), and define v = ET/Ar. In terms of the shifted field the
The saddle point corresponds to 6R "' 6¢/,fi'V;, and Rc ~ ~-!Jf;· So the
potential becomes
critical free energy is
Fo ~ (6¢)3 (Vb)3/2 V(<J>) = { ~ [¢"- v2(1 + 2<)] 2 - < (2Arv3) (¢ + v)} {1 + 0(6<)} (4.1)
(3.6)
Vv v .
So the thin wall bubble radius is found by scaling the thick wall radius by plus terms independent of¢. For a universe filled with false vacuum (<I>)= -v,
312
~ jV:/V and the thin wall free energy increases by a factor of ~ ( ~) the true vacuum bubble of minimum free energy, which is just large enough to
relative to the thick wall case. For the potential of (2.3), V"' -V(¢+), V,"' grow, is a static 0(3) invariant solution to the equations of motion. Hence it
V(¢-), and 6¢ = ¢+· Since the first bubbles to form as the universe cools to a satisfies
temperature below T, will be thin walled bubbles, substituting ¢± into equation d2¢ + ~ d¢ = V'(¢, T), (4.2)
dr2 r dr
(3.6), recalling< = (T,- T)j(T,- T,), and expanding for small<, the first bubbles
where the ' denotes differentiation with respect to <J>. Integrating the appropriate
to form must surmount a free energy barrier which scales like
solution to (4.2) gives the critical free energy
ET,
Fo(<) ~ <2(2Ar)3/2 (3.7)

As we will discuss in section 5, a free energy barrier 0(100T) is small enough

- J {12
Fc-41T r 2 dr (d¢)
dr
2
+V(</>,T) } . (4.3)

to allow bubbles to nucleate. Accordingly, the estimate (3.7) tells us we should

In the limit of small <, the bubble wall thickness is negligible when compared
~Whileit is tme that a large barrier in V(¢) vs. ¢ space makes it hard to fluctuate a to the bubble radius. So for small <, inside the bubble walls we can neglect
bubble of true vacuum, and a smaller barrier makes it easier, we should remember that the
real barrier is in configuration space, and we should be wary of intuition based on the one 'tiThe result we obtain here will be exact in the limit € goes to zero, for the potential of
dimensional mechanics. (2.3).
11 12

the term in (4.2) linear in spatial derivatives. This, together with the boundary cooled from temperature T 1 to temperature T'J, a point was reached where clas-
conditions that the derivative of¢ and the free energy density vanish outside of sical thermal fluctuations were large enough to nucleate bubbles of true vacuum.
the bubble, implies These thermal fluctuations produced bubbles of true vacuum at a rate per unit
d¢ = ,fiV (4.4) volume
dr r IV= A.(T)e-F.(T)/T' (5.1)
inside the bubble walls. We have defined the zero of V(¢) in (4.1) so that
V(-v) = 0. From (4.3) and (4.4) we canwrite the resulting bubble free energy where F.(T) is the free energy of a fluctuation large enough to pass over the

as
¢(R+6R) lR energy barrier separating the two vacua, and A is a characteriStic scale in the
theory. For definiteness we take i\. 4 = w'I"'. As we shall see, the temperatures we
Fo::4rrR 2
1¢(R-6R)
)2V(¢)d¢+4rr
0
drr 2 V(¢+,T). (4.5)
are interested in are on the order of the particle masses. Moreover, because the
nucleation rate is dominated by the exponential, the exact value of the prefactor
As a function of the bubble radius, the free energy is given byll
is not very important, so the effect of w # I will be negligible.
8rr In> 16rr ~'
F(R) = 3v2.XTv3 R 2 - <3ATv•n-. (4.6) • The Onset of Nucleation
Let's begin by determining when the onset of bubble nucleation occurs. In
Varying with respect toR, we find that Rc = 2/ (~(3<v)). Hence, in the the radiation dominated era, the time-temperature relationship is t = f,Mp,fr>,
limit of small <, the critical free energy to temperature ratio is where t is the age of the universe, MP, = 1.22 x 10 19 GeV is the Planck mass,
and f.>:::: 1/34 near the electroweak phase transition. " Because the horizon size
64rr) E
F.jT, = ( Sl <2(2,\T )3/2 scales like dH = 2t, the size of a causal volume at a temperature T is
(4.7)
3
50GeV) -2 VH(T) ~ 8f, 3 Mf,,;r<. (5.2)
>:::: (2.85) ( mH f .

Without being overly precise about the numerical prefactor, we take (5.2) as
As we will discuss in section 5, the phase transition completes when the free our definition of a causal volume. For probabilities small compared to one,
energy to temperature ratio is on the order of 100. This is achieved for < o:: the probability that a bubble was nucleated inside a causal volume during a
(1/6)(50 GeV/mH)"/ 2. For completeness, we note that to lowest order in<, inside temperature interval efT is given by
the bubble walls ¢ behaves like a domain wall and the well known solution is
dP ~ 16 w<• M~,
'T"' e
-F.(T)/T dT (5.3)
'
". -T.
¢(r) = -v tanh ( fjv(r- R)) . (4.8) In this section, it will be convenient to make use of the fact that the critical
bubble free energy to temperature ratio, (4.7), has a Taylor expansion about a
temperature To given by
5 Bubble Production, Evolution, and Number
F.(T)fT = F.(To)/To (1 + 2x + 3x 2 + ... ), (5.4)
Having determined the free energy of a bubble large enough to grow in-
where x = :J.=~o. Between the temperatures T 1 and T'l the free energy is a
definity, we examine the rate for producing such bubbles. As the early universe
quickly changing function of temperature but the change in temperature itself
liThe reader is cautioned that formulas (4.12) and (4.21) of reference [18] are off by a factor
of two.
··e /( oo )
32'11"3
,
LaYa+(7/B)Lp9F
 13 14

between T1 and T'l is insignificant. Define TN as the temperature reached when During the eledroweak phase transition a bubble nucleated at a temperature
the first bubble is nucleated in a typical horizon. Using the expansion (5.4), we T'l and expanding with a velocity (3l has a radius
integrate equation (5.3) and note that the integrand is sharply peaked about
x = 0. The first bubble nucleation in a typical horizon when P ~ I giving

F<(TN )/TN ~ 134. (5.5) Ra(T, T') = RN(T') T

T' MP, (
+ 2fJe T 1 1)
T- T' ' (5.9)
This ratio varies by about 2 percent as the Higgs Boson mass increases from 50
GeV to 100 GeV.
• The Phase Transition where RN(T') is the radius of a true vacuum bubble when it is produced at a
In order to claim that the phase transition really proceeds by bubble nucle- temperature T'.
ation, it is necessary but not sufficient to require that T'l is small enough relative We can imagine two qualitatively different types of first order phase tran-
to TN to ensure that nucleated bubbles grow enough to convert the universe to sitions. In a bubble expansion dominated first order phase transition, a typical
true vacuum before the temperature drops to T2 , where the free energy barrier observer in the fluid sees the surrounding volume converted to true vacuum as
disappears. A reliable determination of the phase transition requires that it com- a nearby bubble of true vacuum expands and a bubble wall passes over the vol-
pletes before the loop expansion parameter in the SU(2) x U(1) becomes large[2]. ume. A second type of first order phase transition occurs if the nucleation rate
Moreover, because the anomalous baryon number violating p_·ocesses which per- turns on very suddenly and is so rapid that a typical volume element in the fluid
sist after the EWPT completes are very sensitive to the value of 4>+(T0 ) tt, a is converted to true vacuum by a nucleation event. We call this second type of
careful determination of the temperature at which the phase transition completes first order phase transition nucleation dominated. The nucleation rate increases
is required. on a time scale set by the expansion rate of the universe. At a time when the
Define dN(T) as the number of bubbles per unit volume nucleated between bubble nucleation rate is large enough to produce bubbles copiously, but small
temperatures T and T + dT. From (5.1), enough that the conversion of space to true vacuum by the nucleation event it-
dN(T) = 2weMp1Te-Fo(T)/T dT. self is still negligible, previously nucleated bubbles will grow appreciably in size
(5.6)
before there is a substantial increase in the nucleation rate. This is a result of
Consider a true vacuum bubble expanding with a constant terminal velocity the smallness of the temperature relative to the Planck mass. As a result, the
(J with respect to the plasma. There are two contributions to the bubble wall vast majority of space is converted to true vacuum by the expansion of bubbles
expansion. One comes from the propagation of the bubble wall through the fluid as opposed to being on the site of a nucleation event. So, the volume of a bubble
and the other is from the expansion of the universe. Hence, produced at temperature T' can be safely written
dT
dR = (Jdt- a(T)T, (5.7)
where a(T) is the Robertson- Walker scale factor. In the radiation dominated
era, this leads to the differential equation
327r 3 M3
V(T, T') = 3fJ"~ 'J'3
PI (1 1)3
__ -
T T (5.10)

dR = _ fJeMe 1 _ a(T).
2 (5.8)
dT T" T
--------------~
ft tP+ increases by a factor of 3/2 as the temperatures drops &om T 1 to T2 . Integrating over the duration of bubble production, the fraction of a typical
15 16

horizon that is true vacuum at time t is '" of the phase transition satisfies the simple relation:

f ~ J dt'V(t',t)f/V n ~ _J_{J_ 3
16rr
( dFo)
3

dT 11 .
(5.11) (5.14)
3 3
~ 64rr wpe•(M /T)4 {T' TdT' (1- T)3 -Fo(T')/T' = ...!:._fJ-3 ( To ) (Fo(To))
3 Pt JT Tl2 T' e . 2rr T,- To To

• End of the Phase Transition It is easy to understand why the number of bubbles per horizon increases as
the derivative of the saddle point free energy increases. At the end of the phase
Define T0 as the temperature reached when the phase transition has com-
transition, the faster the nucleation rate is changing, the shorter the duration
pleted. In the region of interest, the temperature change is insignificant com-
of bubble nucleation. A larger number of bubbles needs to be produced for the
pared to the change in Fo(T). Defining x = (T- T 0 )/T0 , using the Taylor
phase transition to complete in a shorter time interval.
expansion (5.4) about T0 , we see that the integrand of (5.11) is sharply peaked
at temperatures very near T0 . The phase transition completes when the fraction
of true vacuum approaches unity, giving 6 Avoiding Washout of the Baryon Asymmetry
2
2-XTo- ) (100GeV)
Fo(T0 )/T0 ~96-7ln ( ( 0GeV)2 -4ln T, -lnP. (5.12)
Can baryogenesis occur at the one Higgs doublet EWPT? The picture of
5
the phase transition which we have given, valid for Higgs masses from the ex-
The expanding bubbles are not extremely non-relativistic, giving Fo(To)/To ~
perimental limit of 46 GeV up to about 150 GeV, certainly shows that the
100. Hence <n ~ (1/7)(50GeV/mH) 3 12 , and the phase transition completes the universe is far from thermal equilibrium after the temperature drops below TN.
instant after the universe cools to temperature T,. The precise value of <n is At high temperature, anomalous baryon number violation can be quite rapid.
shown as a function of the Higgs boson mass for several top quark masses in
[5, 7, 8, 9, 10] Hence the possibility of successful baryogenesis rests on two issues:
Figure 6.
• sufficient CP violation
• The Number of Bubbles
• avoiding washout of the B +L asymmetry, after T 0 .

Define Tn as the temperature reached by the time n bubbles have been

The numerical size of CP violation in the standard model is insufficient
nucleated inside the comoving volume coincident with the horizon volume at the
for baryogenesis. This is true even if the strong CP parameter 0 were of order
end of the phase transition.
unity at high tempe•atures, because the physical effects of 0 are repressed by
light quark Yukawa couplings. Nevertheless, it is simple to add new physics to
n ~ I6we• (i:') 4 £~' e-f!Fo(T)/T d:. (5.13) the standard model which yields sufficient CP violation without changing the
behavior of the EWPT. One possibility is that this new physics yields operators
Equating Tn with T 0 , the number of bubbles produced per horizon by the end of the form ..j, q,• </JF F in the low energy theory, where F is the electroweak field
strength [11]. Hence it may be possible to create a significant baryon asymmetry
*This expression for the fraction of space containing true vacuwn over counts the volwne
. - JrtVV(t',r)dt'
at the one Higgs EWPT.
where bubbles overlap. The exact [20] formula 1s 1 - f = e . Hence, our
definition of the end of the phase transition corresponds to the era where only 1/e of space is We are left with the problem of how to avoid the B + L asymmetry being
still false vacuum. depleted just after T 0 as soon as thermal equilibrium is re-established. Sev-
 17 18

era! authors [12, 16] have argued that anomalous baryon number violation will o All particle masses are such that at temperatures near T, the high T
washout any baryon asymmetry for Higgs masses larger than some critical value, approximation, (A2.6), for V(</>, T) is valid, with the exception of particles so
mHo· Although there are uncertainties in the calculation of ffiH it is in the
0
, massive that their thermal contributions to the effective potential are Boltzmann
vicinity of 50 GeV, perilously close to the experimental lower bound. Hence, the suppressed to the point that they can be ignored.
one Higgs EWPT of the minimal standard model, extended only to include CP • The EWPT proceeds by the thermal fluctuation of thin wall bubbles. This
violating operators of the form if,</>'</>FF, does not yield an acceptable baryon means that t
asymmetry, with the possible exception that the Higgs boson mass is very close
to 46 GeV. (There is also the possibility that ffiH is much la:ger than 100 GeV,
, ~ 1 El/2
in which case we do not know the nature of the phase transition.). n -12
- >.3/4 (6.1)
<< 1.
T
In this section we show that the problem of baryon washout in one Higgs
doublet models can be solved in a way similar to the solution of the problem of These one Higgs models all have the EWPT proceed as we have described,
sufficient CP violation. Particles can be added to the standard model such that except now the parameters B, D, E, >.T can differ significantly from their stan-
our analysis of the one Higgs EWPT persists, but baryon washout is avoided for dard model values. We now calculate how the baryon washout rate depends on
Higgs masses all the way up to 150 GeV. We find that these additional particles these parameters. We will not attempt to calculate the numerical value for the
affect the EWPT indirectly, by changing the numerical values of the parameters washout rate, however.
of the effective potential (B, D, E, .Xr ). However, our formulae for quantities of At T < To the rate for anomalous baryon number violation is proportional
interest (T1,2, </>±(T!,2), <n, etc.) are still correct when written in terms of these to exp(-F,p(T)/T), where F,p(T) is the sphaleron free energy at temperature
parameters B, D, E, Ar. T. In the region of interest to us T c= T0 (which is just below T,), and we do
As we go beyond the minimal standard model, why not go to the two not know what F,p(T) is. This is because the usual approximation of keeping
Higgs doublet model? We do not do this because this greatly complicates the only the T 2¢ 2 terms in the high Texpansion of ~Vi(</>,T) [21] is not good at
EWPT. In general, one cannot just define a single linear combination of the temperatures near T0 . In particular it is clear that the term -ET¢3 cannot
Higgs as the one which gets a vev, because this combination is T dependent. be neglected. Although the ¢ 2T 2 terms are the largest </> dependent terms in
No complete analysis of the phase transition exists. For example, we do not the mfT expansion, they combine with the zero temperature ¢ 2 terms to cancel
know the quantities ¢J+(T0 ) and ¢ 2+(T0 ) which are relevant for baryogenesis. when the temperature is T,. One possibility is to try to find some lower temper-
The advantage of the one Higgs doublet EWPT is that we know essentially ature where the ¢ 3 term can be dropped, but where the high T expansion is still
everything about the phase transition, so that we can use the requirement of good. However, since the baryon washout rate decreases as T is lowered, this
avoiding baryon washout as a guide to what new physics should exist. will only yield a lower bound on the amount of depletion. Alternatively one can
The analysis of this section will be valid for a whole class of models. This do a numerical analysis for the sphaleron energy at temperatures just below T,
class of models has a single Higgs doublet and has the EWPT proceed by the [12]. We will assume that the sphaleron free energy is linear in </>(T)/T, a form
nucleation of thin wall bubbles. More precisely, the class is defined by three
criteria: I In order to use this formula for ~:n it is actually necessary to also require that T1 - T2 <<
T1 + T2. This is because our thin wall analysis for t:n: was based on the effective potential
o The EWPT is induced by a single Higgs doublet, and the coefficient of
of (2.13) which assumes T1- T2 << T1 + T2. However if th.is constraint on T1- T2 is not
the tree level </> 4 term is sufficiently small that the Higgs contribution to V(</>, T) satisfied, equation 6.1 still holds provided t: is replaced by Fro, where F :::= R T~~")-. ::::; 1. In
2
can be neglected. the majority of models where F is not close to 1, we find that F is not small either.
19 20

motivated by the inclusion of the T 2 </> 2 terms. Hence we take for low top quark masses.

dnB = _!_dT We can now see how simple it is to avoid baryon washout at the one Higgs
(6.2a)
nB HT EWPT. Particles should be added to the standard model so that Ef >.r > ~ for
where any desired value of mH in the region of 45-150 GeV. This can be accomplished
r in two ways:
HT = C, exp { - ( c, <i>+JT)) } {6.2b)
• add bosons with small SU(2) x U{1) preserving masses to increase E
and H = H(T) is the Hubble parameter: C, is a large dimensionless constant
• add bosons so that D.>. + 6B is increased. For a given Higgs mass this
which we assume has a more mild dependence on (B, D, E, >.r) than does rf>+(T)
decreases >.r as can be seen from (6.5).
in the region ofT= T 0 • We take C, to be a constant, although our results are
It is interesting that both possibilities involve additional bosons. Although
unchanged if it has a large power dependence on T.
fermions never contribute toE, they do contribute to D.>.+ 6B. However, they
The most likely T region for baryon washout is immediately below T 0 . If
tend to increase baryon depletion. This can be seen from the fact that a heavy
the washout is to be limited, the washout rate must have frozen out well before
top quark mass in the standard model decreases the critical Higgs mass:
T2 : (fj.)T, << 1. Taking rf>+(T) = f,.' 1>'r(T) where rf>'r is given in equation
(2.14), and integrating (6.2a) gives a logarithmic depletion of
3
"'(m')
4
m1Jm,) = mHo(O)- (4ln(~)+3).
In (nB(To)) ~ ~
nB(O) - 2C,E
(T,- T,) (_!:_)
T, H To '
(6.3)
2 4rr2 o- ,jCiT, (6.7)

We now give a specific simple extension of the standard model which avoids
where baryon washout even for Higgs boson masses up to 150 GeV. We add a spin 0
(~)To"' C,T,e-C,,f;:(3+/I+B<o). {6.4) multiplet S which is a singlet, so that the Lagrangian of the standard model is
augmented by
When <n < < 1, it can be dropped from equation 6.4. Whether depletion is
significant is largely a question of whether the baryon washout rate freezes out
before or after Tn, i.e. of whether (rjH)To is greater or less than unity. This £s= o" s·a,.s- M's• s- >.s(s· 5) 2 - 2( 2 s· sn· H, (6.8)
entire class of one Higgs doublet models has a baryon washout rate which is
where H is the Higgs doublet field: fH 0 f = -7,. We take M 2 and ('
to be
exponentially sensitive to the ratio E/>.r, but is relatively insensitive to Band
positive so that < S >= 0 at all temperatures, and the EWPT is that of the
D, and therefore to T1 and T,. There is significant sensitivity to mH because
single Higgs.
>.r is related to ffiH by
m'
>.r = ;, -(D.>.+ 6B) (6.5)
A simple possibility is to take M 2 sufficiently small that where ¢>"' ¢>+ (T,)
2 the scalar mass m1 = M 2 + ( 2 ¢>~ "' ( 2 </>~. In this case the S field contributes
where we have defined the logarithmic terms of equation 2.4 to be D.>. : >.r = to E. If it gives the dominant contribution to E, then
>.- D.>..
2
Others have found a critical Higgs mass of about 50 GeV [12, 16] in the .!!._ = g, ( 3 ( 100GeV) (6.9)
minimal standard model. This corresponds to Ar 1T mH

where g, is the degeneracy of the S multiplet, and we have chosen parameters

(!)c"'~ {6.6)
such that D.>.+ 6B can be dropped in equation {6.5). The size of Ef>.r is
21 22

restricted by requiring that the high T expansion be valid at T,: of M = 0.56( TeV. Baryon depletion can be avoided even if the additions to the
E 1 standard model have masses in the TeV region.
-<- (6.10)
Ar- ( We conclude that successful baryogenesis at the one Higgs doublet EWPT
Hence we must take ( ::; 1. Note that the criterion (6.1) for the thin wall can occur provided the Higgs doublet is given two new interactions: one to
approximation is easily satisfied for E/-'r much bigger than we need. There are violate CP and the other to enhance E/>-r to avoid baryon workout.
a wide range of parameters that avoid baryon washout for all Higgs masses up
to 150 GeV.
7 Conclusion
As an example, consider ( = 1,gs = 1, and mH = 100 GeV. In this case
E/>-r ~ 2/rr, and all constraints and approximations are satisfied. Taking m, = We have presented a completely analytic treatment of the electroweak phase
125 GeV gives T2 ~ 130 GeV and T, ~ 148 GeV. How small must the SU(2) transition (EWPT), valid for all Higgs boson masses from the experimental limit
invariantS mass be in order that ms(T,) ~ (<i>+(T,) is a valid approximation? of 46 GeV up to about 150 GeV. The electroweak phase transition is first order
We find that M ::; (f.,. T, "' 94 GeV, which does not involve any more fine and proceeds by the nucleation of thin walled bubbles. We give the precise value
tuning than for the Higgs doublet. of the temperature at which the phase transition completes as a function of the
The simple extension of the standard model described by equation (6.8) can top quark and Higgs boson masses. In addition to characterizing the electroweak
also avoid baryon washout, even if the S particle has a mass in the TeV range. phase transition, we determine the value of the Higgs field vev after the phase
At first sight this is surprising since S does not contribute to E. This is because transition completes, the number of bubbles nucleated per horizon, and many
M >> (<i>+(T,), and because the S particles have an exponentially suppressed other quantities. Our formulae also apply to many extensions of the standard
number density at T,. The contribution of the S field to the effective potential, model which have the EWPT occur through just one field acquiring a vev.
Vs(t/>), can then be obtained by expanding Al.6 in a power series in !)$: Additional particles can significantly alter quantities such as the temperature at
which the phase transition completes, through their virtual effects on parameters
V. (</>) = g,(• (-<1>2(</>2 - 2) + ~ q,s) (2u2 +0 ((•u•) (6.11) in the Higgs potential. An important result of our analysis is determination of
' 64rr2 "" 3 u2 M2 M• .
(.P)r at the end of the phase transition. The temperature dependent vev, (,P)r,
The decoupling behavior as M--> oo is manifest. Since v;(u) = v;'(u) = 0, this
increases by a factor of 3/2 as the temperature drops from T, to T,. Accordingly,
wntribution does not alter the minimum of the potential or the relation for the
determination of when the phase transition completes is essential because the
Higgs mass: m"k = 2(>-TsM + 6BsM)u 2 For our purposes the most important
rate of anomalous baryon number violation is an exponentially sensitive function
consequence of V, is to correct the coefficient of ,P4 :
of (,P}r. If the rate of anomalous baryon number violation is large after thermal
g,(• (20"2
Ar = ATsM - - 2 M 2 . (6.12) equilibrium is reestablished, any B + L asymmetry generated during the EWPT
161T will be at best ephemeral.
This is a crucial correction since it changes the mH / ,\T relation, increasing the
Much attention has recently been paid to the exciting possibility that the
critical Higgs mass:
cosmological baryon excess may be produced at the EWPT. The possibility that
2 - g,(• 2 (20"2 this occurs in a model with a single Higgs doublet has largely been ignored. This
m"kJC) = mHJ(- 0) + 8rr2"" M2 · (6.13)
is because it has been shown that even if sufficient baryon asymmetry could be
Hence if (g,( 4 )( ~
~ ) = 3, the critical Higgs mass squared is increased by (50 generated, immediately after the EWPT it would be destroyed by anomalous
2
k
GeV) For example taking g,( 4 = 15 and c;;,' = would give a bareS mass baryon number violation, at least for Higgs boson masses above about 50 GeV.
23 24

As a demonstration of the utility of the analysis of the EWPT presented m Appendix 1: The Zero Temperature, One-Loop Effective
this paper, we have shown that this baryon washout is very easily avoided in Potential
simple extensions of the standard model. Perhaps the simplest is the addition
of a gauge singlet scalar boson that receives a contribution to its mass from the
Higgs boson vev. In a subsequent paper we will show that the baryon asymmetry Consider an ensemble of particles i, which receive a contribution to their mass
of the universe could be generated at the one Higgs doublet EWPT. from the vacuum expectation value of a scalar field ¢. In the mass eigenstate
basis, the unrenormalized one-loop contribution to the effective. potential is

We would like to thank David Bralbm and Steven Hsu for useful discussions.
We acknowledge support from the Director, Office of Energy Research, Office of
High Energy and Nuclear Physics, Division of High Energy Physics of the U.S. il.V,=± { --i
2
J -a'k
- l4 n ( -k 2 +m(¢)-t£
(27r)
2 . )} (Al.l)
Department of Energy under Contract No. DE-AC03-76SF00098, and from the
National Science Foundation under research grant No. PHY-9Q--21139. L.H.
acknowledges support from a Presidential Young Investigator Award. where the ± is for bosons (fermions) respectively. Going to Euclidean space,
introducing a cutoff A, and integrating we find

[~m4 (¢) ln ( m:~¢)) - ~m4 (¢) + m2 (¢)A 2]

1
il. V, = ± 32" 2 (A1.2)

We write the renormalized potential V1 as the sum of the tree level potential
U(¢), the one loop correction il.V1 , and a counter term potential Vet.

V, = U(¢) + Ll V, (¢) + Vct(rP) (Al.3)

We denote the renormalized one loop correction by V 1 = il. V, (¢) + Vet (¢). At
=
¢ u, we choose to impose renormalization conditions that preserves the tree
level values of mH, and u:

d-
i) d¢ v, = 0
(A1.4)
d2
ii) d¢2 v, = 0.
25 26

For a collection of particles with masses of the form m 2(¢) = ,.. 2+ g¢ 2 , imposing between mH and Ao wiH be preserved. In this case, for each degree of freedom,
the renormalization conditions above, we find t the one loop corrections to the effective potential is given by (A1.5): §

IIi(¢)= I> 6 !~2 [mt(¢)1n :~i~i- ~mt(<P) + 2m; 2(¢)m:(u)] (A1.5) V,(¢) = ±
1 4 2 2
rr {m (¢) In (m (¢)/m (u))
64 2
' (A2.2)
3 4 I
plus terms independent of ¢. We will have occasion to use two special cases of -2m (¢) + 2m2 (¢)m2 (u)- m 4(u)},
2
this formula. For a particle with mass m(¢) = >.¢, we recover the well known
result [22]: where the± is for bosons (fermions) and m(¢) is the mass of the particle in the
presence of a background field. Neglecting the Higgs doublet contribution to
V\(¢) = ±6~:2 { ¢•ln(¢2/u')- ~¢• + 2¢2u2}. (A1.6)
V,. The one-loop, zero temperature, effective potential for the physical Higgs
scalar is given by the well known expression [22]:
Finally, for a particle with a large SU(2) conserving mass m 2(¢) = M 2
+ g¢ 2
where M 2 >> gu 2 , the one loop contribution decou pies as I I
V(¢) = --(>. + 2B)u2 ¢ 2 + -,\¢ 4 + B¢4 ln(¢ 2fu 2 ) (A2.3)
2 4
V(¢) = ±_t__ (u•¢2 _ u2¢• + !¢6) M-2 + O(M-•) (A1.7)
M~ 3 . plus terms independent of ¢. Here ¢, the physical Higgs scalar, has mass
m~ = 2>.0 u = (2>. + 12B)u , (¢) = u, and B = 64
2 2
;,a•
(6mt +3m~- 12mn.
Appendix II: The Effective Potential of the Standard When the system is in contact with a hot thermal reservoir, such as in the early
Model universe, the effective potential for the Higgs boson must be modified to include
the interactions between the Higgs field and the hot ambient plasma. The ther-
For the tree level potential mal one-loop corrections to the effective potential for the Higgs boson is just
the free energy of the Bose- Einstein and Fermi-Dirac distributions of particles
0 2
U(¢) = : (¢2- u 2 ) , (A2.1) getting a mass from ¢,

>.a is related to the Higgs mass by m~ = 2>.ou 2 . The one loop potential is the tll/i (¢, T) = - L 9;:; l"' dx x 2
In (I + e-Jx'+!Pm})
sum of the classical potential and a one loop correction V = U + V 1 . If we adopt F
(A2.4)
the renormalization prescription i) V"(u) = m~, and ii) V'(u) = 0, the relation
+L
B
g; T" 1"" dxx 2 In (1- e-Jx'+f3'm~)
1r
2
0 ,
I Strictly speakin'g, the tree leve11-Uggs boson mass preserved by (AlA) is the Higgs boson
mass at zero euclidesn momentwn. The formulas we obtain here will have corrections coming
from running the mass from zero euclidean momentwn up to mH [23] The clivergence in where mB(F) is the mass of a boson (fermion) in the presence of a background
(AI.5) for Goldstone bosons, is an artifact of this running. This is because the Goldstone field ¢, YB(F) is the number of degrees of freedom, f3 = 1/T and (F)B denotes
boson contribution gives an infinite rwm.ing of the Higgs mass between p 2 = m~ and zero
euclidean momentum. Altemath•ely one could impose renonna)jzation conditions at different
a sum over (fermions) bosons respectively. Expanding the argument of the
value of cP. In that case, a singularity in V" would still exist for the reasons mention above, logarithm and integrating, the integral equation for tlV1 (¢, T), can be written
and parameters or effective potential would be related to measurable quantities by a running
in both ¢ space and momentum space. In this paper we work with Higgs Boson self couplings §We have added a constant so that the one loop contribut.ion to cosmological constant
small enough that we can neglect scalar loops which include the Goldstone Bosons. vanishes at (¢) = q,
27 28

in terms of a sum. As well as providing good qualitative behavior, the standard m/T expansion
(A2.6) gives reliable quantitative results up to surprisingly large values of mjT.
L'.V,(¢, T) = "'gFr<
~ 11" 2
r~(-1)"
~ --;-;2 ({3mF) K2({3mFn)
2 J For what values of m/T is the high temperature approximation accurate? From
2
Figure A1 we see that the high temperature approximation agrees with the exact
(A2.5)
potential to better that 5 percent for m/T < 1.6 (2.2) for fermions (bosons). It is
- "\"'Yay<
~ n12 (f3ma) 2 K2({3m 8 n) ] ,
~ 211"2 [ ~ interesting to note that unless great accuracy is required, for any value of m/T,
either the high temperature or a low temperature approximation can be used to
Since the modified Bessel function K2 falls off exponentially for large values of its obtain a simple formula for the free energy. For large values of x, the modified
argument, the expression (Al.5) is well suited to numerical computation when Bessel function K2 has the asymptotic behavior:
m/T is large. In the high temperature limit, when m(¢)/T is small, equation
(2.6) can be expanded in powers of m(¢)/T [3]. Excluding the terms independent
of¢,
K2(x) = {fe-x (1 + ~! + ..) (A2.9)

So for large values of m/T the temperature dependent contribution to the free
L..J9a [ 1 m 82T 2 - 11l"m38 T- m!
L'.Vi(¢,T) = "\"' 2]
rr ln(m28 /c8 T) energy is given by
24 12 64 2
B

m
4
(A2.6) L'.V,(¢) = (m(¢))3/2 e-m(¢)/T (1 + 15T + ···). (A210)
+ ~gF 1
[ 48m;T2 + 64;2ln(m;/cFT2)
]
2rrT Sm(¢)

For any value of m/T, the better of (A2.6) or (A2.10) will give a value for L'. Vi (¢)
+0{ M 6 /T2 } + O{M 6 /T 21n(m/T) },
which is good to better than ten percent.
where ln(ca) = ~ + 2ln4rr- 2-y c=: 5.41 and ln(cF) = ~ + 2ln 11"- 2-y c=: 2.64. 'II
Expanding to order T" in m/T and neglecting all one loop Higgs self interac-
tions, we add (A2.3) and (A2.6) to obtain the one-loop, temperature dependent
potential:

V(¢, T) = [ DT2- G,\ + Bu2) .,.2] ¢2- E¢3T + ~,\T¢4

(A27)
=D(T2 - T,2)¢2 - ET¢3 + ~ ,\T¢4
4

where the constants D, E, and AT are given in section 2. The absolute instability
of the origin occurs when the temperature reaches T2 .

2 (2,\ + 4B)u2 = m~- 8Bu2

T,= 4D
= x2(m,, mH)m~. (A2.8)
4D
1fThis comes from Dolan and Jackiw [3} where the 2.64 was misprinted as 2.84, we thank
David Brahm for bringing tills to our attention.
29 30

References [17] L. McLerran, Phys. Rev. Lett. 62, 1989 (191075)

[1] D. A. Kirzhnitz and A. D. Linde, Ann. Phys.10119576 [18] S. Coleman Phys. Rev. 015, 2929 (1977); S. Coleman and C. Callan, Phys.
Rev. 016, 1762 (1977)
[2] S. Weinberg, Phys. Rev. 09, 3357 (1974)
[19] A. D. Linde, Nucl. Phys. B216, 421 (1983), Phys. Lett. B100, 137 (1981)
[3] L. Dolan and R. Jackiw, Phys. Rev. 09, 3320 (1974)
[20] A. H. Guth and S.-H. Tye, Phys. Rev. Lett. 44, 631 (1980), A. H. Guth and
[4] I. Affleck, Phys. Rev. Lett. 46, 388 (1981)
E. J. Weinberg, Phys. Rev. 023, 876 (1981)
[5] A. D. Linde, Phys. Lett. 70B, 306 (1977)
[21] L. Carson, X. Li, L. McLerran, and R.-T. Wang, Phys. Rev. 042, 2127
[6] G. 't Hooft, Phys. Rev. Lett. 37, (1976) Phys. Rev. 014, 3432 (1976) (1990), L. Carson, L. McLerran, Phys. Rev. 041, 647 (1990)
[7] S. Dimopoulos and L. Susskind, Phys. Rev. 018, 4500 (1978) [22] M. Sher, Phys. Rep. 179, 275 (1989)
[8] F.R. Klinkhamer and N.S. Manton, Phys. Rev. 030, 2212 (1985) [23] J. Iliopoulos, C. Itzykson, and A. Martin, Rev. Mod. Phys. 47, 165 (1975)
[9] V.A. Kuzmin, V.A. Rubakov, and M. E. Shaposhnikov, Phys. Lett. B155,
36 (1985)

[10] P. Arnold and L. McLerran, Phys. Rev. 036, 581 (1987), Phys. Lett.
155B, 36 (1985)

[11] M. Dine, P. Huet, R.S. Singleton Jr., and L.Susskind, Phys. Lett. B257,
351 (1991)

[12] M. Dine, P. Huet, and R.S. Singleton Jr., Santa Cruz preprint SCIPP-91-08

[13] L. McLerran, M. Shaposhnikov, N. Throk, and M. Voloshin, Phys. Lett.

B256, 451 (1991)

[14] N. Throk, J. Zadrozny, Phys. Rev. Lett. 65, 2331 (1990), Princeton
preprints PUTP-90-1216,PUTP-90-1224, PUTP-91-1225.

[15] A. Cohen, D. Kaplan, and A. Nelson, Phys. Lett. B245, 561 (1990), Nucl.
Phys. B349, 727 (1991), Phys. Lett. B263, 86 (1991)

[16] M.E. Shaposhnikov Nucl. Phys. B287, 757 (1987); Nucl. Phys. B299,
797 (1988) ;A.!. Bochkarav, S. Yu. Khlebnikov, and M. E. Shaposhnikov,
Nucl. Phys. B329, 490 (1990) ; A.!. Bochkarav, S.V. Kuzmin, and M. E.
Shaposhnikov, Phys. Lett. 244B, 275 (1990) ; A.!. Bochkarav, and M. E.
Shaposhnikov, Mod. Phys. Lett. A2 417 (1987)
31 32

Figure 1: The ratio x = T2/mH plotted as against the Higgs Boson mass for three
different top quark masses.

Figure 2: A schematic picture of the effective potential for temperatures T. ,TJ ,T1 >
T>T'J and T2,

Figure 3: The temperature dependent vacuum expectation value of the Higgs field
verses the Higgs Boson mass for five di~erent top quark masses. The dashed line
is the exact result, while the solid curve represents the approximation of equation
2.3.

Figure 4: The tree level, one loop, and temperature dependent quartic scalar couplings
verses the Higgs boson mass for three different top quark masses.

Figure 5: The effective potential

Figure 6: Epsilon at the end of the phase transition verses the Higgs boson mass for
three different top quark masses.

Figure 7: A fermion's contribution to the free energy as a function of m/T. The solid
curve is the exact contribution to one loop. The dashed (dotted) curve represents
the approximation A2.6 (A2.10).

Figure 8: A boson's contribution to the free energy as a function of m/T. The solid
curve is the exact contribution to one loop. The dashed (dotted) curve represents
the approximation A2.6 (A2.10).
 X = T2/MH
2.0.-~~~~~~~~~~~~~~~~~~----~~~--

/ M1 ~ 100 GeV
1.8

\ 130

><
1.6f-"... \ -- ~~------------
·:---,·., /
.,. ...._. ... 160
--
--

,·...
" ........ . /
1.4 190
"' ........................... .
' .......... ..._ 0 0 ••• 0 • •• 0. 0 •• 0 •• ••••

..._
- - -- -
1.2
-- ------

La~~~~~~~~~~~~~~~~~~~~~~~~~

50 75 100 125 150 175

MH (GeV)

Figure 1
V(¢)

\
\cP._(Tl

Figure 2
 Tem per atu re dep end ent VEV
1.2~~~~--r-~~~--~~~--
~~~~--.-~-.--.-.-~

:r- --- --g o

- - - - - exac t
1.0

130 I
expa nswn
0,8 I '' \_'\. I (2.3)
T"i

~
~
,.--....
T"i 0.6
~
'-'
f9i

0.4~ \
\
170

190 150
0.2

0. 0 L-----~--l___l~~__L___L__J__j___L_..L_..L__L_L----L-L__J____
L__]
40 60 80 100 120
MH (GeV)

Figure 3
 Three Quartic Couplings
0.25~-.-.-.-.-.-.-,-,~~~~~-.-.~~~~~~~~~~

. ... .... . . . . . ~

0.20 I- I I 190
~T 160
130
~0
.· . //
0.15 .. ·· .. ··· ///
.'}, .·. //'i
f-<
. .·· \J . .·· ~/
/ / </ /

~/_;:x
r<

0.10 . / ::;;?_.,;# . \ \
190
.· ~..---~ 160
.·~~
.. ··
•.
.· .. . . . -;::?"
130
.. ·· .· .. ~~~~
···· ~~
0.05 - . -·~ ~ .:;::;...-~
---.· .·. ~
Mt = 130,160,190 GeV

o.oo~~~~~~~~~~~~~~~~~~~~L-L-L-L-~

60 80 100 120 140

MH (GeV)

Figure 4
v(~)

y_

v.

¢
~\ 4>
+

Figure 5
 0
"<!'
.......
> (j)
I:
0 I :
0 0 0
Q) co 0 I :
....... ....... .......
II
....
s 0
C\2
.......

~
0
0
0
.......
>
Q)
......,
.---I
0 <D
r/l ~
(j)
1...
~ :::J
ILl ::r:: ·-
Cl
u...

~
0
co

...l 0
·. '; co
.. I

0 1.0 0 1.0 0
C\2 ....... ....... 0 0
0 0 0 0 0
 -- -- --
--

....---..
0
----It>·

0
0 N. ""<!' co co 0
0 0 0 0 . 0 ......
0 0 0 0 0 0
I I I I I

f'.L/ (.L/W) d
 - - -
-- -- --
/
/ -- --
/
/
I
I .........
co
I N
rn I ~
.........
~ \ 00
0 E-t
.......... OJ
rn \
:::?1 '-
::::l
0
P=1
\
~
~
·-
01
u..

"'
..-...
-.;!<
N
~
.........

..-...
0
..-<
-{>-
N
~
.........

0
0 N -.;!< co o:J 0
0 0 0 0 0 ..-<
0 0 0 0 0 0
I I I I I

t-.L/ (.L/W).!I: