RAPID COMMUNICATIONS

PHYSICAL REVIEW D 70, 031901共R兲 共2004兲

A 4 symmetry and neutrinos with very different masses

Ernest Ma
Physics Department, University of California, Riverside, California 92521, USA
and Phenomenology Institute, Physics Department, University of Wisconsin, Madison, Wisconsin 53706, USA
共Received 30 April 2004; published 18 August 2004兲
The A 4 model of leptons was proposed originally for nearly degenerate neutrino masses. With minimal
modification, it is shown to accommodate as well the cases of normal hierarchy and inverted hierarchy. In all
cases, this new model predicts sin22␪atm⫽1 and tan2 ␪ sol ⫽0.5 for U e3 ⫽0. It also predicts their deviations from
1 and 0.5 respectively for U e3 ⫽0.

DOI: 10.1103/PhysRevD.70.031901 PACS number共s兲: 11.30.Hv, 14.60.Pq

Neutrino oscillations are now firmly established 关1兴. The ␪ 23⫽ ␲ /4, ␪ 12⫽ ␪ , ␪ 13⫽0, 共6兲
mixing matrix U ␣ i linking (e, ␮ , ␶ ) to the neutrino mass
eigenstates ( ␯ 1 , ␯ 2 , ␯ 3 ) is also determined to a large extent.
This information is, however, not sufficient to fix all the which is of course consistent with present experimental data,
elements of the neutrino mass matrix M␯ 共assumed here to i.e.,
be Majorana at the outset兲, but is indicative of its possible
approximate form in terms of a small number of parameters.
For example, if the Z 2 symmetry ␯ ␮ ↔ ␯ ␶ is imposed on sin2 2 ␪ 23⯝1.0, tan2 ␪ 12⯝0.4, sin2 ␪ 13⬍0.067 共 3 ␴ 兲 .
M␯ , then 关2兴 共7兲

M␯ ⫽ d
d
冉 冊 e d
a
b
d
b ,
a
共1兲 Furthermore, it has been shown 关2兴 that M␯ of Eq. 共1兲 is
flexible enough to allow for all possible hierarchies of the
neutrino mass eigenvalues. However, since ␪ 13 is required to
which is diagonalized by be zero, there can be no CP violation in U.

冉 冊
Versatile as it is, the proposed M␯ has a very serious flaw.
m1 0 0 Since each neutrino has its corresponding charged lepton to-
gether in a weak SU(2) L doublet, any symmetry imposed on
U T M␯ U⫽ 0 m2 0 , 共2兲 M␯ must affect the charged-lepton Dirac mass matrix Ml
0 0 m3 linking (e, ␮ , ␶ ) to (e c , ␮ c , ␶ c ). In particular, the Z 2 symme-
try ␯ ␮ ↔ ␯ ␶ implies that there is also a Z 2 symmetry ␮ ↔ ␶ in
where Ml . This requirement is not consistent with the implicit as-

冉 冊
cos ␪ ⫺sin ␪ 0 sumption of Eq. 共1兲 that Ml is diagonal with different values
for m ␮ and m ␶ . Attempts to fix this problem by choosing
U⫽ sin ␪ / 冑2 cos ␪ / 冑2 ⫺1/冑2 , 共3兲 other simple symmetries such Z 3 关3兴 and Z 4 关4兴 necessarily
sin ␪ / 冑2 cos ␪ / 冑2 1/冑2 entail additional arbitrary assumptions regarding Ml . One
exception is the recent application 关5兴 of the non-Abelian
with discrete symmetry S 3 . Another is the A 4 model 关6 –9兴, based
on the symmetry of the tetrahedron 共Plato’s fire兲. This group
tan 2 ␪ ⫽2 冑 2d
e⫺a⫺b
. 共4兲
has three inequivalent one-dimensional representations and
one three-dimensional representation, allowing Ml to have
three independent eigenvalues and in the basis where it is
In the convention diagonal, M␯ has the form

U⫽ 0
0
冉 1 0
c 23
s 23
0
⫺s 23
c 23
冊冉 c 13

⫺s 13e
0
i␦
0
1
0
s 13e ⫺i ␦
0
c 13
冊 M␯ ⫽m 0 0 冉 冊
1 0
0
0
1 . 共8兲

冉 冊
0 1 0
c 12 ⫺s 12 0
⫻ s 12 c 12 0 , 共5兲
0 0 1 This results in three degenerate neutrino masses, but the ad-
dition of the most general flavor-changing one-loop radiative
this means that corrections 关8,9兴 automatically changes it to

1550-7998/2004/70共3兲/031901共5兲/$22.50 70 031901-1 ©2004 The American Physical Society

 RAPID COMMUNICATIONS

ERNEST MA PHYSICAL REVIEW D 70, 031901共R兲 共2004兲

M␯ ⫽m 0 冉 1⫹ ␦ 0 ⫹ ␦ ⫹ ␦ * ⫹2 ␦ ⬘
␦⬙
␦ ⬙*
␦⬙
␦
1⫹ ␦ 0 ⫹ 共 ␦ ⫹ ␦ * 兲 /2
␦ ⬙*
1⫹ ␦ 0 ⫹ 共 ␦ ⫹ ␦ * 兲 /2 .
␦*
冊 共9兲

Without loss of generality, ␦ may be chosen real by absorb- L i ⫽ 共 ␯ i ,l i 兲 ⬃3 共 i⫽1,2,3 兲 , 共11兲

ing its phase into ␯ ␮ and ␯ ␶ and ␦ 0 set equal to zero by
redefining m 0 and the other ␦ ’s. Remarkably, this M␯ has
l c1 ⬃1, l c2 ⬃1 ⬘ , l c3 ⬃1 ⬙ . 共12兲
the same form as Eq. 共1兲, if ␦ ⬙ is real. However ␦ ⬙ is in
general complex, in which case ␪ 13 becomes nonzero, and
the CP violation in U is predicted to be maximal 关2,10兴. Both Using the Higgs doublets
results are very interesting and sure to be tested in future
neutrino experiments. ⌽ i ⫽ 共 ␾ 0i , ␾ ⫺
i 兲 ⬃3 共 i⫽1,2,3 兲 , 共13兲
The neutrino mass matrix of Eq. 共9兲 predicts three nearly
degenerate neutrino masses. This is naturally expected from the 3⫻3 mass matrix linking l i to l cj is given by

冉 冊
the assumed A 4 symmetry. Since all the mass splitting comes
from radiative corrections, m 0 should be greater than the h 1v 1 h 2v 1 h 3v 1
square root of the observed ⌬m 2 for atmospheric neutrino h 2␻ v 2 h 3␻ 2v 2 ,
Ml ⫽ h 1 v 2 共14兲
oscillations by at least an order of magnitude. This puts a
lower bound of about 0.3 eV on m 0 , which is very relevant h 1v 3 h 2␻ v 32
h 3␻ v 3
for neutrinoless double beta decay 关11兴 and the large-scale
structure of the Universe 关12兴. In fact, the experimental up- where ␻ ⫽e 2 ␲ i/3 and v i ⫽ 具 ␾ 0i 典 . As shown in Ref. 关6兴, for a
per bound on m 0 in either case is also about 0.3 eV. continuous range of parameter values, the minimum of the
In the following, the A 4 model is modified to show that it Higgs potential is given by v 1 ⫽ v 2 ⫽ v 3 ⫽ v . In that case, Ml
can also accommodate three neutrinos of very different is diagonalized by

冉 冊 冉 冊
masses, either in the normal hierarchy, i.e., 兩 m 1 兩 ⬍ 兩 m 2 兩
⬍ 兩 m 3 兩 , or the inverted hierarchy, i.e., 兩 m 3 兩 ⬍ 兩 m 1 兩 ⬍ 兩 m 2 兩 , h1 0 0 me 0 0
with solar neutrino oscillations coming from ␯ 1 -␯ 2 splitting 冑3 v ⫽
U L† Ml U R ⫽ 0 h2 0 0 m␮ 0 ,
and atmospheric neutrino oscillations from the larger split-
ting of ␯ 3 with ␯ 1,2 . It is also shown that the resulting neu- 0 0 h3 0 0 m␶
trino mixing matrix reduces to that of Eq. 共3兲 in a certain 共15兲
symmetry limit, but with tan2 ␪ also fixed at 1/2. This specific
form of U has long been advocated as a desirable ansatz 关13兴 where

冉 冊
and here is the first example of how it can be derived from a
complete theory, without arbitrary assumptions regarding its 1 1 1
1
charged-lepton sector. Away from this symmetry limit as pa- U L⫽ 1 ␻ ␻2 , 共16兲
rametrized by a single 共complex兲 parameter, U e3 becomes 冑3
1 ␻2 ␻
nonzero, ␪ 23 becomes greater or less than ␲ /4, and tan2 ␪ 12
moves away from 0.5. The present experimental bound on
and U R is the unit matrix.
U e3 means that this model has specific predictions for ␪ 12
Instead of using the canonical seesaw mechanism for Ma-
and ␪ 23 , which can be tested in future neutrino experiments.
jorana neutrino masses as in previous versions of the A 4
The group theory and representations of A 4 are already
model, the neutrino mass matrix M␯ is assumed here to
discussed in Ref. 关6兴. Briefly, it is the group of even permu-
come from the naturally small vacuum expectation values of
tations of four objects, which is also the symmetry group of
heavy Higgs triplets 关15兴. Let
the tetrahedron, one of five perfect geometric solids known
to the ancient Greeks, and associated by Plato with the ele-
ment ‘‘fire.’’ It has 12 elements and 4 irreducible representa- ␰ 1 ⬃1, ␰ 2 ⬃1 ⬘ , ␰ 3 ⬃1 ⬙ , ␰ i ⬃3 共 i⫽4,5,6 兲 ,
tions: 1,1 ⬘ ,1 ⬙ , and 3, with the multiplication rule 共17兲

3⫻3⫽1⫹1 ⬘ ⫹1 ⬙ ⫹3⫹3. 共10兲 where ␰ i ⫽( ␰ ⫹⫹

i ,␰⫹
i , ␰ i ). Then M␯ in the original basis is
0

given by

冉 冊
As such, it is ideal for describing three families of leptons
共and quarks兲. „Note that A 4 is isomorphic to the dihedral a⫹b⫹c 0 0
group ⌬(12), and is a discrete subgroup 关14兴 of SO(3) as
well as SU(3).… In the convention that all fermions are left- M␯ ⫽ 0 a⫹ ␻ b⫹ ␻ c 2
d ,
handed, the leptons are assumed to transform under A 4 as 0 d a⫹ ␻ b⫹ ␻ c
2

follows: 共18兲

031901-2
 RAPID COMMUNICATIONS

A 4 SYMMETRY AND NEUTRINOS WITH VERY . . . PHYSICAL REVIEW D 70, 031901共R兲 共2004兲

where a comes from 具 ␰ 01 典 , b from 具 ␰ 02 典 , c from 具 ␰ 03 典 , and d If m 4 ⫽0, ␯ 1 mixes with ␯ 3 , but ␯ 2 remains the same. Let
from 具 ␰ 04 典 , assuming that 具 ␰ 05 典 ⫽ 具 ␰ 06 典 ⫽0, which is also a the new mass eigenstates be
natural minimum of the Higgs potential for a continuous
range of parameter values. ␯ 1⬘ ⫽ ␯ 1 cos ␪ ⫹ ␯ 3 e i ␦ sin ␪ , ␯ 3⬘ ⫽⫺ ␯ 1 e ⫺i ␦ sin ␪ ⫹ ␯ 3 cos ␪ ,
In the basis where the charged-lepton mass matrix is di- 共26兲
agonal, the neutrino mass matrix becomes then the new mixing matrix U has elements
␮,␶)

冑 冑
M (e, ⫽U L† M␯ U L*

冉 冊
␯ 2 1 2 i␦
U e1 ⫽ cos ␪ , U e2 ⫽ , U e3 ⫽⫺ e sin ␪ ,
a⫹ 共 2d/3兲 b⫺ 共 d/3兲 c⫺ 共 d/3兲 3 冑3 3
⫽ b⫺ 共 d/3兲 c⫹ 共 2d/3兲 a⫺ 共 d/3兲 , 共27兲
c⫺ 共 d/3兲 a⫺ 共 d/3兲 b⫹ 共 2d/3兲 1 1
共19兲 U ␮ 3 ⫽⫺ cos ␪ ⫹ e i ␦ sin ␪
冑2 冑6
which is the main result of this paper. For b⫽c⫽d⫽0, it
reduces to the model of Ref. 关8兴 before radiative corrections.
For b⫽c, it reduces to the form of Eq. 共1兲 with the concomi-
⫽⫺
冑
冑
1
2
3 1
1⫺ 兩 U e3 兩 2 ⫺ U e3 .
2 2
共28兲

tant ␯ ␮ ↔ ␯ ␶ symmetry and thus the results of Eq. 共6兲, but

with an additional constraint, i.e., e⫽a⫹b⫺d, or tan 2 ␪ ⫽ Therefore, the experimental constraint 关16兴
⫺2 冑2 in Eq. 共4兲. The mixing matrix U of Eq. 共3兲 then be- 兩 U e3 兩 ⬍0.16 共29兲
comes

冉 冊
implies
冑2/3 1/冑3 0
0.61⬍ 兩 U ␮ 3 兩 ⬍0.77, 共30兲
U⫽ ⫺1/冑6 1/冑3 ⫺1/冑2 , 共20兲
⫺1/冑6 1/冑3 1/冑2 or, using sin22␪atm⫽4兩U␮3兩2(1⫺兩U␮3兩2),

which is a well-known ansatz 关13兴 with the prediction 0.94⬍sin2 2 ␪ atm ⬍1. 共31兲
tan2 ␪ 12⫽1/2, but has never been derived from the symmetry Similarly, using tan2 ␪ sol ⫽ 兩 U e2 兩 2 / 兩 U e1 兩 2 ,
of a complete theory, without arbitrary assumptions regard-
ing its charged-lepton sector, until now. 0.5⬍tan2 ␪ sol ⬍0.52 共32兲
In the basis defined by U of Eq. 共20兲, i.e.,
is obtained. Whereas Eq. 共31兲 is well satisfied by the current

␯ 1⫽ 冑 2
3
␯ e⫺
1
共 ␯ ⫹␯ 兲,
冑6 ␮ ␶
共21兲
data, Eq. 共32兲 is at the high end of the 90% confidence limit–
allowed range centered at tan2 ␪ sol ⯝0.4 关17兴.
If future experimental measurements persist in getting a
value of tan2 ␪ sol outside the range predicted by Eq. 共32兲, one
1 possible explanation within the context of this model is that
␯ 2⫽ 共 ␯ e⫹ ␯ ␮⫹ ␯ ␶ 兲, 共22兲
冑3 具 ␰ 5,6
0
典 are not zero as assumed, but rather 具 ␰ 06 典 ⫽⫺ 具 ␰ 05 典 ,
which can be maintained in the Higgs potential by postulat-
ing the interchange symmetry ␰ 5 ↔⫺ ␰ 6 in addition to
1
␯ 3⫽ 共 ⫺ ␯ ␮⫹ ␯ ␶ 兲, 共23兲 ␾ 2 ↔ ␾ 3 for the soft trilinear ␰ ␾␾ terms which break the A 4
冑2 symmetry. In that case, the neutrino mass matrix of Eq. 共18兲
has the additional piece

冉 冊
the neutrino mass matrix of Eq. 共19兲 rotates to

冉 冊
0 e ⫺e
m1 0 m4
⌬M␯ ⫽ e 0 0 , 共33兲
␮,␶)
M ␯(1,2,3) ⫽U † M (e,
␯ U *⫽ 0 m2 0 , 共24兲 ⫺e 0 0
m4 0 m3
resulting in

冉 冊
where

冉 冊冉 冊冉 冊
0 0 0
m1 1 ⫺1/2 ⫺1/2 1 a ⌬M ␯(1,2,3) ⫽U † 共 U L† M␯ U L* 兲 U * ⫽ 0 0 m5 ,
m2 1 1 1 0 b 0 m5 0
⫽ . 共25兲 共34兲
m3 ⫺1 1/2 1/2 1 c
m4 0 ⫺ 冑3/2 冑3/2 0 d where m 5 ⫽i 冑2e. Thus Eq. 共24兲 becomes

031901-3
 RAPID COMMUNICATIONS

ERNEST MA PHYSICAL REVIEW D 70, 031901共R兲 共2004兲

M ␯(1,2,3) ⫽ 冉 m1
0
m4
0
m2
m5
m4
m5 .
m3
冊 共35兲
共 m 1 ⫺␭ 兲共 m 2 ⫺␭ 兲共 m 3 ⫺␭ 兲 ⫺m 24 共 m 2 ⫺␭ 兲 ⫺m 25 共 m 1 ⫺␭ 兲 ⫽0.

For ␭ i ⫽m i ⫹ ␦ i , the eigenstates are approximately given by

共41兲

If m 4 ⫽0 but m 5 ⫽0, then ␯ 2 mixes with ␯ 3 while ␯ 1 re-

mains the same. The analogs of Eqs. 共27兲 and 共28兲 are then
␯ i⫹ 兺
j⫽i
⑀i j␯ j
共42兲
冑
,
U e1 ⫽ 冑 2
3
, U e2 ⫽
1
冑3
cos ␪ , U e3 ⫽⫺
冑3
1
e i ␦ sin ␪ , 1⫹ 兺
j⫽i
⑀ 2i j

共36兲
where each ⑀ i j is much smaller than 1. Thus

U ␮ 3 ⫽⫺
1
cos ␪ ⫺
1
e i ␦ sin ␪ 冑2 ⑀ 31⫹ ⑀ 32
冑2 冑3 U e3 ⫽ 具 ␯ e 兩 ␯ 3 典 ⯝ , 共43兲
冑3
1
⫽⫺ 冑1⫺3 兩 U e3 兩 2 ⫹U e3 . 共37兲 冑3⫹ ⑀ 31⫺ 冑2 ⑀ 32
冑2 U ␮ 3 ⫽ 具 ␯ ␮ 兩 ␯ 3 典 ⯝⫺ , 共44兲
冑6
Therefore, the experimental constraint 关18兴
冑2⫹ ⑀ 12
0.91⬍sin2 2 ␪ atm ⬍1 共38兲 U e1 ⫽ 具 ␯ e 兩 ␯ 1 典 ⯝ . 共45兲
冑3
implies Consider first ␯ 3 . The solutions of Eq. 共41兲 are
兩 U e3 兩 ⬍0.11, 共39兲 m 24 m 25
␦ 3⯝ ⫹ , 共46兲
and thus m 3 ⫺m 1 m 3 ⫺m 2

0.48⬍tan2 ␪ sol ⬍0.5. 共40兲 m4 m5

⑀ 31⯝ , ⑀ 32⯝ . 共47兲
m 3 ⫺m 1 m 3 ⫺m 2
If both m 4 and m 5 are nonzero and all m i ’s are real, the
eigenvalues of Eq. 共35兲 obey the cubic equation Consider next ␯ 1 . The solutions of Eq. 共41兲 are

1 1
␦ 1 ⫽ 关 m 2 ⫺m 1 ⫺ 共 m 24 ⫹m 25 兲 / 共 m 3 ⫺m 1 兲兴 ⫺ 冑关 m 2 ⫺m 1 ⫺ 共 m 24 ⫹m 25 兲 / 共 m 3 ⫺m 1 兲兴 2 ⫹4m 24 共 m 2 ⫺m 1 兲 / 共 m 3 ⫺m 1 兲 , 共48兲
2 2

m 5␦ 1 ␦1
⑀ 12⫽ , ⑀ 13⫽ . 共49兲
m 4 共 m 1 ⫺m 2 ⫹ ␦ 1 兲 m4

As a numerical example, consider are obtained. Accounting also for the mass shifts ␦ i , this
example has
m 1 ⫽0.003 eV, m 2 ⫽0.009 eV, m 3 ⫽0.047 eV,

m 4 ⫽0.003 eV, m 5 ⫽0.003 eV, 共50兲

⌬m atm
2
⫽2.2⫻10⫺3 eV2 , ⌬m sol
2
⫽6.9⫻10⫺5 eV2 .
then 共54兲

⑀ 31⫽0.068, ⑀ 32⫽0.079, ⑀ 12⫽0.034, ⑀ 13⫽⫺0.071,

共51兲 As for CP violation, it is allowed in both Eq. 共24兲 and Eq.
共35兲, but is not otherwise constrained in this model.
resulting in Another possibility of obtaining a value of tan2 ␪ sol ⬍0.5
U e3 ⫽0.10, U ␮ 3 ⫽⫺0.686, U e1 ⫽0.834, 共52兲 in Eq. 共24兲 is through radiative corrections. Just as Eq. 共8兲 is
radiatively corrected to become Eq. 共9兲, Eq. 共19兲 may also
from which get corrected so that ␯ 1 mixes with ␯ 2 in Eq. 共24兲. For ex-
ample, if b,c,d⬍a, then combining Eq. 共9兲 and Eq. 共19兲
sin2 2 ␪ atm ⫽0.996, tan2 ␪ sol ⫽0.42 共53兲 with b⫽c, the analog of Eq. 共4兲 is given by

031901-4

A 4 SYMMETRY AND NEUTRINOS WITH VERY . . .
tan 2 ␪ sol ⯝⫺2 冑2 冋 b⫺ 共 d/3兲 ⫹ ␦ ⬙ a
b⫺ 共 d/3兲 ⫺2 ␦ ⬘ a
册 , 共55兲
where
1 charged-lepton mass matrix is diagonal with three indepen-
␦ ⬙⫽ ␦ e␮⫹ ␦ e␶ , ␦ ⬘ ⫽ ␦ ee ⫺ 共 ␦ ␮␮ ⫹ ␦ ␶␶ 兲 ⫺ ␦ ␮␶ , 共56兲
2
and tan2 ␪ sol ⯝0.4 is obtained if 关 b⫺(d/3)⫹ ␦ ⬙ a 兴 / 关 b
⫺(d/3)⫺2 ␦ ⬘ a 兴 ⯝0.75. Note that this may occur even if
␦ ␣␤ ⫽0 for ␣ ⫽ ␤ , i.e., in the absence of flavor changing
radiative corrections, in contrast to the requirement of Refs.
关8兴 and 关9兴.
In the Higgs sector, the triplets are very heavy and may all
be integrated away 关15兴. Their only imprint on the low-
energy theory is the neutrino mass matrix of Eq. 共19兲. On the
other hand, the three Higgs doublets are at the electroweak
mass scale and their Yukawa couplings to the leptons are
关1兴 For a brief review, see for example A.Yu. Smirnov,
hep-ph/0402264.
关2兴 E. Ma, Phys. Rev. D 66, 117301 共2002兲.
关3兴 See, for example, E. Ma, Phys. Lett. B 583, 157 共2004兲; Mod.
Phys. Lett. A 19, 577 共2004兲.
关4兴 See, for example, E. Ma, Phys. Rev. Lett. 90, 221802 共2003兲;
E. Ma and G. Rajasekaran, Phys. Rev. D 68, 071302共R兲
共2003兲.
关5兴 S.-L. Chen, M. Frigerio, and E. Ma, hep-ph/0404084, and ref-
erences therein.
关6兴 E. Ma and G. Rajasekaran, Phys. Rev. D 64, 113012 共2001兲.
关7兴 E. Ma, Mod. Phys. Lett. A 17, 289 共2002兲; 17, 627 共2002兲.
关8兴 K.S. Babu, E. Ma, and J.W.F. Valle, Phys. Lett. B 552, 207
共2003兲; see also E. Ma, Mod. Phys. Lett. A 17, 2361 共2002兲.
关9兴 M. Hirsch, J.C. Romao, S. Skadhauge, J.W.F. Valle, and A.
Villanova del Moral, Phys. Rev. D 69, 093006 共2004兲.
关10兴 W. Grimus and L. Lavoura, Phys. Lett. B 579, 113 共2004兲.
关11兴 For a review, see, for example, S.R. Elliott and P. Vogel, Annu.
Rev. Nucl. Part. Sci. 52, 115 共2002兲. There is also a reported
031901-5
RAPID COMMUNICATIONS
PHYSICAL REVIEW D 70, 031901共R兲 共2004兲
completely determined by the charged-lepton masses as
shown in Ref. 关6兴. Phenomenological consequences are dis-
cussed in detail there.
In conclusion, a much more predictive version of the A 4
model of leptons has been proposed. In the basis where the
dent eigenvalues 共i.e. m e , m ␮ , m ␶ ), the neutrino mass ma-
trix is fixed to have the form of Eq. 共19兲. Using the mixing
matrix of Eq. 共20兲, it becomes Eq. 共24兲, with 4 parameters
m 1,2,3,4 . If m 4 ⫽0, then U e3 ⫽0, sin22␪atm⫽1, and tan2 ␪ sol
⫽0.5, independent of the neutrino mass eigenvalues m 1,2,3 .
To obtain the experimentally more favorable result of
tan2 ␪ sol ⯝0.4, one possibility is to add m 5 according to Eq.
共35兲, another is to include radiative corrections as in Ref. 关8兴.
I thank Vernon Barger for discussions. This work was
supported in part by the U.S. Department of Energy under
Grant No. DE-FG03-94ER40837.
first evidence of its observation: H.V. Klapdor-Kleingrothaus
et al., Mod. Phys. Lett. A 17, 1475 共2002兲; Phys. Lett. B 586,
198 共2004兲.
关12兴 D.N. Spergel et al., Astrophys. J., Suppl. 148, 175 共2003兲; S.
Hannestad, J. Cosmol. Astropart. Phys. 05, 004 共2003兲.
关13兴 P.F. Harrison, D.H. Perkins, and W.G. Scott, Phys. Lett. B 530,
167 共2002兲. For an updated review, see P.F. Harrison and W.G.
Scott, hep-ph/0402006.
关14兴 A. Hanany and Y.-H. He, J. High Energy Phys. 02, 013 共1999兲.
关15兴 E. Ma and U. Sarkar, Phys. Rev. Lett. 80, 5716 共1998兲.
关16兴 M. Apollonio et al., Phys. Lett. B 466, 415 共1999兲; F. Boehm
et al., Phys. Rev. D 64, 112001 共2001兲.
关17兴 SNO Collaboration, Q.R. Ahmad et al., Phys. Rev. Lett. 89,
011301 共2002兲; 89, 011302 共2002兲; 92, 181302 共2004兲; Kam-
LAND Collaboration, K. Eguchi et al., ibid. 90, 021802
共2003兲.
关18兴 C.K. Jung, C. McGrew, T. Kajita, and T. Mamm, Annu. Rev.
Nucl. Part. Sci. 51, 451 共2001兲.