Improved Constraints on Sterile Neutrinos in the MeV to GeV Mass Range

D. A. Brymana,b and R. Shrockc

(a) Department of Physics and Astronomy, University of British Columbia,
Vancouver, British Columbia, V6T 1Z1, Canada
(b) TRIUMF, 4004 Wesbrook Mall, Vancouver, British Columbia, V6T 2A3, Canada and
(c) C. N. Yang Institute for Theoretical Physics and Department of Physics and Astronomy,
Stony Brook University, Stony Brook, NY 11794, USA

Improved upper bounds are presented on the coupling |Ue4 |2 of an electron to a sterile neutrino
ν4 from analyses of data on nuclear and particle decays, including superallowed nuclear beta decays,
(π) (K) (D ) +
the ratios Re/µ = BR(π + → e+ νe )/BR(π + → µ+ νµ ), Re/µ , Re/τs , and Be2 decay, covering the
mass range from MeV to GeV.
arXiv:1904.06787v1 [hep-ph] 15 Apr 2019

Neutrino oscillations and hence neutrino masses and
lepton mixing have been established and are of great im-
portance as physics beyond the original Standard Model
(SM). Most oscillation experiments with solar, atmo-
spheric, accelerator, and reactor (anti)neutrinos can be
explained within the minimal framework of three neu-
trino mass eigenstates with values of ∆m2ij = m2νi − m2νj
given approximately by ∆m221 = 0.74 × 10−4 eV2 and
|∆m232 | = 2.5 × 10−3 eV2 , with normal mass ordering
mν3 > mν2 favored; furthermore, the lepton mixing an-
gles θ23 , θ12 , and θ13 have been measured, with a ten-
tative indication of a nonzero value of the CP-violating
quantity sin(δCP ) [1]-[6].
In addition to the three known neutrino mass eigen-
states, there could be others, which would necessarily be
primarily electroweak-singlets (sterile) [7]. Indeed, ster-
ile neutrinos are present in many ultraviolet (UV) exten-
sions of the SM. Whether sterile neutrinos exist in na-
ture is one of the most outstanding questions in particle
physics, and therefore, improved constraints on their cou-
plings are of fundamental and far-reaching importance.
Taking account of the possibility of sterile neutrinos, the
neutrino interaction eigenstates νℓ would be given by
3+n
Xs
νℓ = Uℓi νi , (1)
i=1
where ℓ = e, µ, τ ; ns denotes the number of sterile
neutrinos; and U is the lepton mixing matrix [8].
Here we obtain improved upper limits on |Uei |2 for a
sterile neutrino νi in a wide range of masses from the MeV
to GeV scale and point out new experiments that would
be worthwhile and could yield further improvements. For
simplicity, we assume one heavy neutrino, ns = 1, with
i = 4; it is straightforward to generalize to ns ≥ 2. Since
a ν4 in this mass range decays, it is not excluded by the
cosmological Pupper limit on the sum of effectively stable
neutrinos, i mνi < ∼ 0.12 eV [9]. Such a ν4 is subject
to a number of constraints from cosmology (e.g., [10]);
however, since these depend on assumptions about the
early universe, we choose here to focus on direct labo-
ratory bounds. Constraints from the non-observation of
neutrinoless double beta decay are satisfied by assum-
ing that ν4 is a Dirac neutrino [11]. Since sterile neu-
trinos violate the conditions for the diagonality of the
weak neutral current [12, 13], ν4 has invisible tree-level
decays of the form ν4 → νj ν̄i νi where 1 ≤ i, j ≤ 3 with
model-dependent invisible branching ratios. Because our
bounds are purely kinematic, they are complementary to
bounds from searches for neutrino decays, which involve
model-dependent assumptions on branching ratios into
visible versus invisible final states.
We first obtain improved upper bounds on |Ue4 |2 from
nuclear beta decay data. The emission of a ν4 via lepton
mixing in nuclear beta decay has several effects, including
producing a kink in the Kurie plot and reducing the decay
rate [14]. For the nuclear beta decays (Z, A) → (Z +
1, A) + e− + ν̄e or (Z, A) → (Z − 1, A) + e+ + νe into
a set of neutrino mass eigenstates νi ∈ νe , i = 1, 2, 3
of negligibly small masses, plus a ν4 of of non-negligible
mass, the differential decay rate is

 i1/2 
dN h
= C (1 − |Ue4 |2 )pE(E0 − E)2 + |Ue4 |2 pE(E0 − E) (E0 − E)2 − m2ν4 θ(E0 − E − mν4 ) , (2)
dE

where p ≡ |p| and E denote the 3-momentum and (to- and C = G2F |Vud |2 FF |M|2 /(2π 3 ), where M denotes the
tal) energy of the outgoing e± , E0 denotes its maximum nuclear transition matrix element, V is the Cabibbo-
energy for the SM case, the Heaviside θ function is de- Kobayashi-Maskawa (CKM) quark mixing matrix, and
fined as θ(x) = 1 for x > 0 and θ(x) = 0 for x ≤ 0, FF is the Fermi function. Early bounds on |Ue4 |2 were
 2

set from searches for kinks in Kurie plots in [14] and for ν4 masses in the range 0.5 MeV <∼ mν4 < 9.4 MeV,
analyses of particle decays [15]-[17], and from dedicated indicated in Fig. 1 as BD2. (These and other limits
experiments. For example, a search for kinks in the Kurie presented are at the 90 % confidence level.)
plot in 20 F beta decay reported in Ref. [18] yielded an
upper bound on |Ue4 |2 decreasing from 5.9 × 10−3 for
mν4 = 0.4 MeV to 1.8 × 10−3 for mν4 = 2.8 MeV. (Some
recent reviews of searches for sterile neutrinos include
[19]-[24].)
In addition to kink searches, a powerful method to set
constraints on massive neutrino emission, via lepton mix-
ing, in nuclear beta decays is to analyze the decay rates.
Since, in general, the heavy neutrino would also be emit-
ted in µ decay, the measurement of the µ lifetime per-
formed assuming the SM would yield an apparent (app)
value of the Fermi constant, denoted GF,app , that would
be smaller
√ than the true value, GF , given at tree level by
GF / 2 = g 2 /(8m2W ), where g is the SU(2) gauge cou-
pling [15–17]. To avoid this complication, the ratios of
rates of different nuclear beta decays are compared.
The integration of dN/dE over E gives the kinematic
rate factor f . The combination of this with the half-life FIG. 1: 90 % C.L. upper limits on |Ue4 |2 vs. mν4 from var-
for the nuclear beta decay, t ≡ t1/2 , yields the product f t. ious sources: PIBETA, pion beta decay[64]; BD1, previous
Incorporation of nuclear and radiative corrections yields limits from beta decay [18]; BD2, beta decay with the two
the corrected f t value for a given decay, denoted F t. dashed horizonal lines based on our analysis using [31] and
BR(π + →e+ νe )
Conventionally, analyses of the F t values for the most [32]; PIENU and PIENU-H, the ratio BR(π + →µ+ ν ) in the
µ
precisely measured superallowed 0+ → 0+ nuclear beta kinematically allowed and forbidden regions for ν4 emission;
+ +
decays have been used, in conjunction with the value of P ie2, π → e νe4 peak search [47]; KENU and KENU-H, the
BR(K + →e+ νe )
GF,app from µ decay, to infer a value of the weak mixing ratio BR(K + →µ+ ν ) in the kinematically allowed and forbid-
µ
matrix element, |Vud | [25]-[34]. A first step in these anal- den regions for ν4 emission; Ke2, K + → e+ νe4 peak search
yses has been to establish the mutual consistency of the [48]; Dse2, Ds+ → e+ νe4 ; and, Be2, B + → e+ νe4 .
F t values for these superallowed 0+ → 0+ decays. Since
the emission of a ν4 with mass of a few MeV would have a
different effect on the kinematic functions and integrated We next discuss upper bounds from two-body leptonic
rates for nuclear beta decays with different Q (energy decays of charged pseudoscalar mesons (generically de-
release) values, it would upset this mutual consistency. noted as M + ) [14, 15]. This method is quite powerful,
Hence, from this mutual agreement of F t values, an because the signal is a monochromatic peak in dN/dpℓ
+
upper limit on |Ue4 |2 can be derived for values of mν4 and for Me2 decays, the strong helicity suppression in
such that a ν4 could be emitted in some of these superal- the SM case is removed when a heavy neutrino is emit-
lowed decays. Ref. [26] obtained upper bounds on |Ue4 |2 ted. The presence of a massive ν4 also changes the ratio
ranging from 10−2 down to 2×10−3 for mν4 from 0.5 to 2 BR(M + → e+ νe )/BR(M + → µ+ νµ ) from its SM value,,
MeV, while Ref. [18] obtained the limits |Ue4 |2 < 1×10−3 and this was used to set further bounds [14, 15, 38]. A
to |Ue4 |2 < 2 × 10−3 for mν4 from 1 to 7 MeV. The number of dedicated experiments have been performed
maximum Q value in the current set of 14 superallowed to search for a peak due to heavy neutrino emission and
0+ → 0+ beta decays used for the F t fit in [30, 31] is also to measure BR(M + → e+ νe )/BR(M + → µ+ νµ )
+ + +
9.4 MeV (for 74 Rb). A measure of the mutual agree- with πℓ2 , Kℓ2 , and Bℓ2 , where ℓ = e, µ [36]-[49].
ment is the precision with which |Vud |2 is determined, In the SM with only the three known neutrinos with
so a reduction in the fractional uncertainty of the value negligibly small masses, the ratio
of |Vud |2 results in an improved upper limit on |Ue4 |2 .
Ref. [26] obtained |Vud | = 0.9740 ± 0.001. The recent (M) BR(M + → ℓ+ νℓ )
Rℓ/ℓ′ ≡ (5)
analyses in [31] and [32] obtain |Vud | = 0.97420(21) and BR(M + → ℓ′+ νℓ′ )
|Vud | = 0.97370(14), respectively [35]. Applying these
factors of improvement from [31] and [32] to the previous is given by
bounds in [26], improved upper bounds are obtained as (M) 2
m2ℓ

(M) 1 − δℓ
2
|Ue4 | < 4 × 10 −4
(3) Rℓ/ℓ′ ,SM = (1 + δRC ) , (6)
m2ℓ′ (M)
1 − δℓ ′
and (M)
where δℓ = m2ℓ /m2M and δRC is the radiative correc-
2 −4
|Ue4 | < 2.7 × 10 (4) tion (RC) [50]-[55].
 3

We denote the ratio of the experimental measurement so

(M)
of Rℓ/ℓ′ to the SM prediction as (M) (M)
(M) 1 − |Uℓ4 |2 + |Uℓ4 |2 ρ̄(δℓ , δν4 )
R̄ℓ/ℓ′ = (M) (M)
. (16)
(M) 1 − |Uℓ′ 4 |2 + |Uℓ′ 4 |2 ρ̄(δℓ′ , δν4 )
(M)
Rℓ/ℓ′
R̄ℓ/ℓ′ ≡ (M)
. (7)
Rℓ/ℓ′ ,SM With no loss of generality, we order ℓ and ℓ′ such that
(M)
mℓ′ > mℓ and define the mass intervals (i) I1 : mν4 <
(π) (M)
The most precise measurement of Re/µ is from the mM − mℓ′ ; (ii) I2 : mM − mℓ′ < mν4 < mM − mℓ ;
(π) (M)
PIENU experiment at TRIUMF, with the result Re/µ = and (iii) I3 : mν4 > mM − mℓ . Thus, a ν4 with
(M) +
(1.2344 ± 0.0023stat ± 0.0019syst ) × 10−4 [45]. The resul- mν4 ∈ I1 contributes to both Mℓ2 and Mℓ+′ 2 decays,
(π) (M) +
tant PDG world average is Re/µ = (1.2327 ± 0.0023) × while if mν4 ∈ I2 , then ν4 contributes to Mℓ2 , but not
+ (M)
10−4 [1], in agreement with the SM prediction with RC, to Mℓ′ 2 decay, and if mν4 ∈ I3 , then ν4 cannot be
+
(π)
Re/µ = (1.2352 ± 0.0002) × 10−4 [51, 52, 54]. Using the emitted in either Mℓ2 or Mℓ+′ 2 decay.
(π) If for a given mν4 , one knows, e.g., from peak-search
PDG value of Re/µ , one finds
experiments, that |Uℓ′ 4 |2 is sufficiently small that the de-
(π)
nominator of (16) can be approximated well by 1, then
R̄e/µ = 0.9980 ± 0.0019 . (8) (M)
an upper bound on the deviation of R̄ℓ/ℓ′ from 1 yields
(K) an upper bound on |Uℓ4 |2 . Thus, one has the bound
The ratio Re/µ has recently been measured by the NA62
experiment at CERN [43], dominating the world average (M)
R̄ℓ/ℓ′ − 1 (M)
[1] |Uℓ4 |2 < (M) (M)
for mν4 ∈ I2 . (17)
ρ̄(δℓ , δν4 ) −1
(K)
Re/µ = (2.488 ± 0.009) × 10−5 . (9)
This gives very stringent upper limits on |Uℓ4 |2 because
(M (M) (M)
The SM prediction with RC [52, 55] is ρ̄(δe , δν4 ) >> 1 over much of the interval I2 (see
(M)
Figs. 3-5 in [15]). If mν4 ∈ I3 , then (16) reduces to
(K)
Re/µ,SM = (2.477 ± 0.001) × 10−5 , (10) (M)
R̄ℓ/ℓ′ = (1 − |Uℓ4 |2 )/(1 − |Uℓ′ 4 |2 ), so if |Uℓ′ 4 |2 << 1 in
this interval, then the upper limit is
resulting in
(M) (M)
(K) |Uℓ4 |2 < 1 − R̄ℓ/ℓ′ for mν4 ∈ I3 . (18)
R̄e/µ = 1.0044 ± 0.0037 . (11)
(π)
With emission of a heavy neutrino ν4 , the ratio We now apply this analysis to Re/µ , using (17) and (18)
+
(M)
Rℓ/ℓ′ ,SM for general ℓ, ℓ′ changes to with M + = π + , ℓ = e, and ℓ′ = µ. From previous πµ2
peak search experiments [36]-[49] and the calculation of
(π) (π)
ρ̄(δµ , δν4 ), it follows that |Uµ4 |2 is sufficiently small for
" #
2 (M) 2 (M) (M)
(M) [(1 − |Uℓ4 | )ρ(δℓ , 0) + |Uℓ4 | ρ(δℓ , δν4 )
Rℓ/ℓ′ = (M) (M) (M)
× m ∈ I (π) that we can approximate the denominator of
(1 − |Uℓ′ 4 |2 )ρ(δℓ′ , 0) + |Uℓ′ 4 |2 ρ(δℓ′ , δν4 ) ν4 2
(π)
Eq. (16) by 1. From R̄e/µ in Eq. (8), using the procedure
× (1 + δRC ) , (12)
(π)
from [56], we obtain the limit R̄e/µ < 1.0014. Then, for
(M)
where δν4 = m2ν4 /m2M , and the kinematic function (π)
ν4 ∈ I2 , we find
ρ(x, y) is [14, 15]
(π)
R̄e/µ′ − 1 0.0014
ρ(x, y) = [x + y − (x − y)2 ][λ(1, x, y)]1/2 , (13) 2
|Uℓ4 | < < . (19)
(π) (π) (π) (π)
ρ̄(δe , δν4 ) −1 ρ̄(δe , δν4 ) −1
with
This bound is labelled as PIENU in Fig. 1. If mν4 ∈
λ(z, x, y) = x2 + y 2 + z 2 − 2(xy + yz + zx) . (14) (π)
I3 , i.e., mν4 > 139 MeV, then, using (18), we obtain
the upper bound on |Ue4 |2 given by the flat line labelled
Thus, in the SM case, ρ(x, 0) = x(1 − x)2 . Here and PIENU-H in Fig. 1.
(M) (M)
below, it is implicitly understood that ρ(δℓ , δν4 ) = 0 We next obtain a bound on |Ue4 |2 by applying the same
if mν4 ≥ mM − mℓ , where the decay M + → ℓ+ ν4 is (K)
type of analysis to Re/µ . From Kµ2 peak search exper-
kinematically forbidden. We define (K) (K)
iments [37, 44, 48] and the calculation of ρ̄(δµ , δν4 ),
ρ(x, y) ρ(x, y) |Uµ4 |2 is sufficiently small that we can approximate the
ρ̄(x, y) = = (15)
ρ(x, 0) x(1 − x)2 denominator of Eq. (16) well by 1. Using Eq. (11) for

(K)
ν4 ∈ I2 , we find
(K)
R̄e/µ′ − 1 0.010
|Uℓ4 |2 < (K) (K)
< (K) (K)
. (20)
ρ̄(δe , δν4 ) −1 ρ̄(δe , δν4 ) −1
This upper limit on |Ue4 |2 is labelled KENU in Fig. 1.
(K)
For mν4 ∈ I3 , i.e., mν4 > 493 MeV, using (18), we
obtain the flat upper bound labelled KENU-H in Fig. 1.
One can also apply these methods to two-body leptonic
decays of heavy-quark hadrons. We first consider Ds+ →
ℓ+ νℓ decays [57], using (17) and (18) with M + = Ds+ , ℓ =
e, and ℓ′ = τ . Experimental data from CLEO, BABAR,
Belle, and BES have determined BR(Ds+ → µ+ νµ ) =
(5.49 ± 0.17) × 10−3 and BR(Ds+ → τ + ντ ) = (5.48 ±
0.23) × 10−2 [58]-[62]. Furthermore, searches by CLEO
[58], BABAR [59], and Belle [60] have yielded the limit
(D )
BR(Ds+ → e+ νe ) < 0.83 × 10−4 . Hence, Re/τs < 1.6 ×
(D )
10−3 . For Re/τs , using the results of [51], we calculate
1 + δRC = 0.948. Substituting this in Eq. (6) with M =
Ds , ℓ = e, ℓ′ = τ , we find
(D )
s
Re/τ,SM = 2.29 × 10−6 .
(D ) (D )
Therefore, R̄e/τs < 7.0 × 102 . For Re/τs , the interval
(D )
I2 s is 191 MeV < mν4 < 1.457 GeV. Actually, we re-
strict mν4 to a lower-mass subset of this interval, because
for sufficiently great mν4 , even though the Ds+ → e+ ν4
decay is kinematically allowed to occur, the momentum
pe (in the Ds rest frame) would be below the minimal
value set by experimental cuts in the BES III event re-
construction. With pe,cut ≃ 0.8 GeV [63], this means
that mν4 must be less than 0.85 GeV for the event to be
accepted. Thus, we consider 0.191 GeV < mν4 < 0.85
(D )
GeV. Substituting the experimental limit on R̄e/τs in
the special case of (16) with M = Ds , ℓ = e, ℓ′ = τ
and using the fact that |Uτ 4 |2 << 1 for this mν4 mass
range [1], we obtain a resultant limit from (17). For
(D ) (D )
mν4 = 0.191 GeV, ρ̄(δe s , δν4 s ) = 1.37 × 105 , increas-
(Ds ) (Ds )
ing to ρ̄(δe , δν4 ) = 1.83 × 106 for mν4 = 0.85 GeV.
We thus obtain the upper bound on |Ue4 |2 labelled Dse2
in Fig. 1.
A dedicated peak-search experiment to search for the
heavy-neutrino decay Ds+ → e+ ν4 would be worthwhile
and could improve the upper bound on |Ue4 |2 . Similarly,
a search for leptonic D decays like D+ → e+ ν4 would
be valuable and will be discussed elsewhere [64]. The
(D ) (D ) (D) (D)
very large values of ρ̄(δe s , δν4 s ) and ρ̄(δe , δν4 ) over
a large portion of the kinematically allowed ranges of mν4
in Ds+ → e+ ν4 and D → e+ ν4 mean that there would be
quite strong kinematic enhancement of the heavy neu-
+
trino decay relative to the corresponding (Ds+ )e2 and De2
decays. In particular, these searches could be performed
by the BES III experiment, which recently reported re-
sults from a data sample of 3.19 fb−1 and expects to
collect considerably higher statistics.
4
Finally, we consider B + → ℓ+ νℓ decays. There is
an upper limit BR(B + → e+ νe ) < 0.98 × 10−6 from
Belle [65] and BABAR [66]. For the other two leptonic
decay modes, BR(B + → µ+ νµ ) = (6.46 ± 2.22stat ±
1.60syst ) × 10−7 from Belle [67], with a recent update
BR(B + → µ+ νµ ) = (5.3 ± 2.0stat ± 0.9syst ) × 10−7
[68, 69], and BR(B + → τ + ντ ) = (1.09±0.24)×10−4 from
BABAR [70] and Belle [71, 72]. The measured values of
BR(B + → µ+ νµ ) are in agreement with the SM predic-
tion BR(B + → µ+ νµ )SM = (3.80 ± 0.31) × 10−7 [67].
The measured value of BR(B + → τ + ντ ) is also in agree-
ment with the SM prediction BR(B + → τ + ντ )SM =
(0.75+0.10
−0.05 ) × 10
−4
[72, 73].
We focus on data from a Bℓ2 +
peak search experiment
by Belle [46]. In general [14],
(M) (M)
BR(M + → ℓ+ ν4 ) |Uℓ4 |2 ρ̄(δℓ , δν4 )
= . (22)
+ +
BR(M → ℓ νℓ )SM 1 − |Uℓ4 |2
For mν4 in the range from 0.1 GeV to 1.4 GeV, the Belle
experiment obtained an upper limit on BR(B + → e+ ν4 )
of 2.5 × 10−6 , while in the interval of mν4 from 1.4
GeV to 1.8 GeV, this upper limit increased to 7 × 10−6 .
In the range of mν4 from 0.1 to 1.3 GeV, the Belle
(21)
experiment obtained (non-monotonic) upper limits on
BR(B + → µ+ ν4 ) of approximately 2 − 4 × 10−6 , and
in the interval of mν4 from 1.3 GeV to 1.8 GeV, it ob-
tained upper limits varying from 2 × 10−6 to 1.1 × 10−5 .
Substituting the BR(B + → e+ ν4 ) limits in Eq. (22)
with M = B and ℓ = e, we obtain the upper limits
on |Ue4 |2 shown as the curve Be2 in Fig. 1 [74]. From
the BR(B + → e+ ν4 ) limits we infer upper limits on
|Uµ4 |2 that decrease from 0.83 to 3.4 × 10−2 as mν4 in-
creases from 0.1 GeV to 1.2 GeV. Further peak searches
for B + → ℓ+ ν4 with ℓ = e, µ at Belle II would be worth-
while as a higher-statistics extension of [46].
We briefly remark on other constraints on a Dirac ν4
in the mass range considered here. From the results
of [12, 75], it follows that there is a negligibly small
contribution to decays such as µ → eγ and µ → eeē.
Similarly, there is no conflict with bounds on neutrino
magnetic moments [1, 76], and contributions to invisible
Higgs decays [77] are well below the current upper limit
of BR(H → invis.) < 19% [78].
In this work, improved upper limits on |Ue4 |2 have been
presented covering most of the range from mν4 = 0.5
MeV to mν4 ≃ 1 GeV, representing the best available
laboratory bounds for a Dirac neutrino ν4 that do not
make model-dependent assumptions concerning visible
neutrino decay modes. Over parts of this range, the
bounds obtained are competitive with those that assume
specific visible ν4 decays. For example, for mν4 = 30
MeV, our upper bound is |Ue4 |2 < 0.8 × 10−6 , while the
best bound for this value of mν4 from experiments search-
ing for neutrino decays is |Ue4 |2 < 1 × 10−6 [79]. New
peak search experiments to search for Ds+ → e+ ν4 and
D+ → e+ ν4 as well as a continued search for B + → e+ ν4
would be valuable; these could improve the bounds fur-
ther. Other constraints on sterile neutrinos such as from

π + → π 0 e+ νe decay,and a detailed report of the results
presented here will be published elsewhere [64].
We thank J. Benitez, J. Hardy, V. Luth, W. Marciano,
M. Ramsey-Musolf, C. Yuan, and G. Zhao for useful dis-
cussions. This work was supported in part by the Nat-
[1] M. Tanabashi et al. (Particle Data Group, PDG), Phys.
Rev. D 98, 030001 (2018); online at http://pdg.lbl.gov.
[2] See http://www.nu-fit.org.
[3] See talks at Neutrino 2018, Heidelberg,
https://www.mpi-hd.mpg.de/nu2018 (June, 2018).
[4] S. Gariazzo, C. Giunti, M. Laveder, and Y. F. Li, Phys.
Lett B 782, 13 (2018).
[5] M. Dentler, A. Hernández-Cabezudo, J. Kopp, P.
Machado, I. Martinez-Soler and T. Schwetz, JHEP 08
(2018) 010.
[6] I. Esteban, M. C. Gonzalez-Garcia, A. Hernández-
Cabezudo, M. Maltoni, and T. Schwetz, JHEP 01 (2019)
106.
[7] We will use the term “sterile neutrino” both in its pre-
cise sense as an electroweak-singlet interaction eigenstate
and in a commonly used approximate sense as the cor-
responding mainly sterile mass eigenstate(s) in this neu-
trino interaction eigenstate.
[8] See, e.g., [4, 5] for global fits to neutrino data that allow
for a possible sterile neutrino with mass ∼ O(1) eV2 .
[9] N. Aghanim et al. (Planck Collab.), arXiv:1807.06209;
P. A. R. Ade et al. (Planck Collab.), Astron. Astrophys.
594, A13 (2016).
[10] There is an extensive literature on these constraints and
on models that are designed to evade one or more of them.
See, e.g., the following and references therein: S. Hannes-
tad and G. G. Raffelt, Phys. Rev. D 72, 103514 (2005);
G. Gelmini, E. Osoba, S. Palomares-Ruiz, and S. Pas-
coli, JCAP 10 (2008) 029; G. Steigman, Adv. HEP 2012
(2012) 268321; J. Lesgourgues and S. Pastor, Adv. High
Energy Phys. 2012, 608515 (2012); H. Ishida, M. Kusak-
abe, and H. Okada, Phys. Rev. D 90, 083519 (2014);
M. Archidiacono, S. Hannestad, R. S. Hansen, and T.
Tram, JCAP 1608 (2016) 067; K. N. Abazajian and M.
Kaplinghat, Ann. Rev. Nucl. Part. Sci. 66, 401 (2016).
[11] Although Majorana neutrino masses have often been re-
garded in the past as more generic, many ultraviolet ex-
tensions of the SM lead to Dirac neutrinos instead. See,
e.g., K. Dick, M. Lindner, M. Ratz, and D.Wright, Phys.
Rev. Lett. 84, 4039 (2000); D. A. Demir, L. L. Everett,
and P. Langacker, Phys. Rev. Lett. 100, 091804 (2008);
Y. Grossman and D. J. Robinson, JHEP 1101 (2011)
132; N. Memenga, W. Rodejohann, and H. Zhang, Phys.
Rev. D 87, 053021 (2013); A. Aranda et al., Phys. Rev.
D 89, 033001 (2014); E. Ma and O. Popov, Phys. Lett.
B 764, 142 (2017); A. Dasgupta, S. K. Kang, and O.
Popov, arXiv:1903.12558 and references therein.
[12] B. W. Lee and R. E. Shrock, Phys. Rev. D 16, 1444
(1977). The theorem proved in this work is that the nec-
essary and sufficient condition for the weak leptonic neu-
tral current to be diagonal in mass eigenstates is that
leptons of the same chirality and charge must have the
same weak T and T3 .
[13] J. Schechter and J. W. F. Valle, Phys. Rev. D 22, 2227
5
ural Sciences and Engineering Research Council and the
National Research Council of Canada (D.B.) and by the
U.S. National Science Foundation Grant NSF-PHY-16-
1620628 (R.S.).
(1980).
[14] R. E. Shrock, Phys. Lett. B 96, 159 (1980).
[15] R. E. Shrock, Phys. Rev. D 24, 1232 (1981).
[16] R. E. Shrock, Phys. Rev. D 24, 1275 (1981).
[17] R. E. Shrock, in Proc. of the 1980 VPI Conference on
Weak Interactions as Probes of Unification, A.I.P. Conf.
Proc. 72, Particles and Fields Subseries No. 23 (A.I.P.,
New York, 1981) p. 368.
[18] J. Deutsch, M. LeBrun, and R. Prieels, Nucl. Phys. A
518, 149 (1990).
[19] A. Kusenko, Phys. Rept. 481, 1 (2009); A. Boyarsky, O.
Ruchayskiy, and M. Shaposhnikov, Ann. Rev. Nucl. Part.
Sci. 59, 191 (2009).
[20] J. C. Helo, S. Kovalenko, and I. Schmidt, Nucl. Phys. B
853, 80 (2011); Phys. Rev. D 84, 053008 (2011).
[21] A. Abada et al., JHEP 1302 (2013) 048; A. Abada et al.,
JHEP 1402 (2014) 091.
[22] A. de Gouvêa and A. Kobach, Phys. Rev. D 93, 033005
(2015).
[23] S. Alekhin et al., Rep. Prog. Phys. , 124201 (2016); M.
Drewes et al., JCAP 1701 (2017) 025; K. N. Abazajian,
Phys. Rept. 711-712, 1 (2017); A. Boyarsky, M. Drewes,
T. Lasserre, S. Mertens, and O. Ruchayskiy, Prog. Part.
Nucl. Phys. 104, 1 (2019).
[24] B. Batell, T. Han, D. McKeen, and B. Es Haghi, Phys.
Rev. D 97, 075016 (2018).
[25] J. C. Hardy and I. S. Towner, Nucl. Phys. A 254, 221
(1975).
[26] J. C. Hardy, I. S. Towner, V. T. Kozlowsky, E. Hagberg,
and H. Schmeing, Nucl. Phys. A 509, 429 (1990).
[27] A. Czarnecki, W. J. Marciano, and A. Sirlin, Phys. Rev.
D 70, 093006 (2004).
[28] J. C. Hardy and I. S. Towner, Phys. Rev. C 71, 055501
(2005); J. C. Hardy and I. S. Towner, Phys. Rev. C 79,
055502 (2009).
[29] W. J. Marciano, and A. Sirlin, Phys. Rev. Lett. 96,
032002 (2006).
[30] J. C. Hardy and I. S. Towner, Phys. Rev. C 91, 025501
(2015).
[31] J. C. Hardy and I. S. Towner, arXiv:1807.01146 C18-05-
29 (CIPANP-2018).
[32] C. Y. Seng, M. Gorchtein, H. H. Patel, and M. Ramsey-
Musolf, Phys. Rev. Lett. 121, 241804 (2018); C. Y. Seng,
M. Gorchtein, and M. Ramsey-Musolf, arXiv:1812.03352.
[33] See talks by J. Hardy, W. Marciano, and M.
Ramsey-Musolf in the Workshop on First-Row
CKM Unitarity (Texas A&M Univ., Jan. 2019),
https://cyclotron.tamu.edu/ckmuw2019.
[34] Since in the presence of massive neutrinos, GF,app would
differ from GF , the value of |Vud |, denoted |Vud,app |, de-
rived from these fits to Ft values would also differ from
the true value [15, 17]. However, all that we use here is the
relative agreement between the Ft values, as measured
by the quoted precision in the determination of |Vud,app |,

so we keep the subscript app implicit.
[35] Using their value of |Vud | in [31], together with |Vus | =
0.2243(5) from [1] (the |Vub |2 term makes a negligible
contribution), Hardy and Towner obtain |Vud |2 +|Vus |2 +
|Vub |2 = 0.99939(64), in agreement with CKM unitarity.
Ref. [32] notes that its value of |Vud | leads to tension with
CKM unitarity. See also E. Blucher and W. J. Marciano,
in [1].
[36] R. Abela et al., Phys. Lett. 105B, 263 (1981); R. C.
Minehart et al., Phys. Rev. Lett. 52, 804 (1984).
[37] Y. Asano et al., Phys. Lett. B 104, 84 (1981); R. S.
Hayano et al., Phys. Rev. Lett. 49, 1305 (1982).
[38] D. Bryman et al., Phys. Rev. Lett. 50, 1546 (1983).
[39] T. Yamazaki, in Proc. of Neutrino-84 Conf., Nord-
kirchen, Germany (World Scientific, Singapore, 1984).
[40] D. Bryman and T. Numao, Phys. Rev. D 53, 558 (1986).
[41] M. Daum et al., Phys. Rev. D 36, 2624 (1987).
[42] D. I. Britton et al., Phys. Rev. D 46, R885 (1992).
[43] C. Lazzeroni et al. (NA62 Collab.), Phys. Lett. B 719,
326 (2013).
[44] A. V. Artamonov et al. (BNL E949), Phys. Rev. D 91,
052001 (2015).
[45] A. Aguilar-Arevalo et al. (PIENU Collab.), Phys. Rev.
Lett. 115, 071801 (2015).
[46] C.-S. Park et al. (Belle Collab.), Phys. Rev. D 94, 012003
(2016).
[47] A. Aguilar-Arevalo et al. (PIENU Collab.), Phys. Rev. D
97, 072012 (2018).
[48] C. Lazzeroni et al. (NA62 Collab.), Phys. Lett. B 772,
712 (2017); E. Cortina Gill et al., (NA62 Collab.), Phys.
Lett. B 778, 137 (2018).
[49] A. Aguilar-Arevalo et al. arXiv:1904.03269.
[50] S. Berman, Phys. Rev. Lett. 1, 468 (1959); T. Kinoshita,
Phys. Rev. Lett. 2, 477 (1959); T. Kinoshita and A. Sir-
lin, Phys. Rev. 113, 1652 (1959).
[51] W. J. Marciano and A. Sirlin, Phys. Rev. Lett. 71, 3629
(1993).
[52] V. Cirigliano and I. Rosell, Phys. Rev. Lett. 99, 231801
2007); V. Cirigliano and I. Rosell, JHEP 10 (2007) 005.
[53] Sufficiently soft photons are understood to be included in
the experimentally measured M + → ℓ+ νℓ decays, which
are thus often labelled as M + → ℓ+ νℓ (γ).
[54] D. Bryman, W. J. Marciano, R. Tschirhart, and T. Ya-
manaka, Ann. Rev. Nucl. Part. Sci. 61, 331 (2011).
[55] Hadronic structure-dependent bremsstrahlung (SDB)
contributions to the radiative corrections are estimated
(π)
to be negligibly small for Re/µ and hence are not included
in the calculations of [51, 52]. Bremsstrahlung contribu-
(K)
tions to Re/µ are discussed in [48, 52].
6
[56] G. Feldman and R. Cousins, Phys. Rev. D57, 3873
(1998).
[57] Charge-conjugate decays are understood to be included
here and below.
[58] J. P. Alexander et al. (CLEO Collab.), Phys. Rev. D 79,
052001 (2009).
[59] P. del Amo Sanchez et al. (BABAR Collab.), Phys. Rev.
D 82, 091103(R) (2010).
[60] A. Zupanc et al. (Belle Collab.), JHEP 09 (2013) 139.
[61] M. Ablikim et al. (BES III Collab.), Phys. Rev. D 94,
072004 (2016).
[62] M. Ablikim et al. (BES III Collab.), Phys. Rev. Lett.
122, 071802 (2019) [arXiv:1811.10890].
[63] Private communication from Changzheng Yuan and
Guang Zhao (BES III).
[64] D. Bryman and R. Shrock, to be published.
[65] N. Satoyama et al. (Belle Collab.), Phys. Lett. B 647, 67
(2007).
[66] B. Aubert et al. (BABAR Collab.), Phys. Rev. D 79,
091101 (2009).
[67] A. Sibidanov et al. (Belle Collab.), Phys. Rev. Lett. 121,
031801 (2018).
[68] M. Prim (Belle Collab.), talk at Moriond-2019 [69].
[69] Moriond-2019 Workshop,
http://moriond.in2p3.fr/2019/EW.
[70] J. P. Lees et al. (BABAR Collab.), Phys. Rev. D 88,
031102 (2013).
[71] K. Hara et al. (Belle Collab.), Phys. Rev. Lett. 110,
131801 (2013).
[72] B. Kronenbitter et al. (Belle Collab.), Phys. Rev. D 92,
051102 (2015).
[73] See CKMfitter website, http://ckmfitter.in2p3,fr.
[74] Here, we focus on the effect of ν4 in these B decays. For
recent discussions of possible indications of other new
physics in the B system, see, e.g., [69].
[75] W. J. Marciano and A. I. Sanda, Phys. Lett. B 67, 303
(1977); S. M. Bilenky, S. T. Petcov, and B. Pontecorvo,
Phys. Lett. 67, 309 (1977); B. W. Lee, S. Pakvasa, R.
E. Shrock, and H. Sugawara, Phys. Rev. Lett. 38, 1230
(1977).
[76] K. Fujikawa and R. E. Shrock, Phys. Rev. Lett. 45, 963
(1980).
[77] R. E. Shrock and M. Suzuki, Phys. Lett. B 110, 250
(1982).
[78] S. Cooperstein (CMS Collab.), in [69]; see also V.
Khachatryan et al. (CMS Collab.), JHEP 02, 135
(2017); arXiv:1809.05937; G. Aad (ATLAS Collab.),
arXiv:1809.06682.
[79] B. Bernardi et al., Phys. Lett. B 166, 479 (1986).