Measurement of the rate of nu(e) + d --> p + p + e(-) interactions produced by (8)B solar neutrinos at the Sudbury Neutrino Observatory.

Abstract

Solar neutrinos from (8)B decay have been detected at the Sudbury Neutrino Observatory via the charged current (CC) reaction on deuterium and the elastic scattering (ES) of electrons. The flux of nu(e)'s is measured by the CC reaction rate to be straight phi(CC)(nu(e)) = 1.75 +/- 0.07(stat)(+0.12)(-0.11)(syst) +/- 0.05(theor) x 10(6) cm(-2) s(-1). Comparison of straight phi(CC)(nu(e)) to the Super-Kamiokande Collaboration's precision value of the flux inferred from the ES reaction yields a 3.3 sigma difference, assuming the systematic uncertainties are normally distributed, providing evidence of an active non- nu(e) component in the solar flux. The total flux of active 8B neutrinos is determined to be 5.44+/-0.99 x 10(6) cm(-2) s(-1).

Full text

VOLUME 87, N
UMBER 7 PHYSICAL REVIEW LETTERS 13A
UGUST 2001
Measurement of the Rate of n
e
1 d ! p 1 p 1 e
2
Interactions Produced
by
8
B Solar Neutrinos at the Sudbury Neutrino Observatory
Q. R. Ahmad,
15
R. C. Allen,
11
T. C. Andersen,
12
J. D. Anglin,
7
G. Bühler,
11
J. C. Barton,
13,
* E. W. Beier,
14
M. Bercovitch,
7
J. Bigu,
4
S. Biller,
13
R. A. Black,
13
I. Blevis,
3
R. J. Boardman,
13
J. Boger,
2
E. Bonvin,
9
M. G. Boulay,
9
M. G. Bowler,
13
T. J. Bowles,
6
S. J. Brice,
6,13
M. C. Browne,
15
T. V. Bullard,
15
T. H. Burritt,
15,6
K. Cameron,
12
J. Cameron,
13
Y. D. Chan,
5
M. Chen,
9
H. H. Chen,
11,
†
X. Chen,
5,13
M. C. Chon,
12
B. T. Cleveland,
13
E. T. H. Clifford,
9,1
J. H. M. Cowan,
4
D. F. Cowen,
14
G. A. Cox,
15
Y. Dai,
9
X. Dai,
13
F. Dalnoki-Veress,
3
W. F. Davidson,
7
P. J. Doe,
15,11,6
G. Doucas,
13
M. R. Dragowsky,
6,5
C. A. Duba,
15
F. A. Duncan,
9
J. Dunmore,
13
E. D. Earle,
9,1
S. R. Elliott,
15,6
H. C. Evans,
9
G. T. Ewan,
9
J. Farine,
3
H. Fergani,
13
A. P. Ferraris,
13
R. J. Ford,
9
M. M. Fowler,
6
K. Frame,
13
E. D. Frank,
14
W. Frati,
14
J. V. Germani,
15,6
S. Gil,
10
A. Goldschmidt,
6
D. R. Grant,
3
R. L. Hahn,
2
A. L. Hallin,
9
E. D. Hallman,
4
A. Hamer,
6,9
A. A. Hamian,
15
R. U. Haq,
4
C. K. Hargrove,
3
P. J. Harvey,
9
R. Hazama,
15
R. Heaton,
9
K. M. Heeger,
15
W. J. Heintzelman,
14
J. Heise,
10
R. L. Helmer,
10,
‡
J. D. Hepburn,
9,1
H. Heron,
13
J. Hewett,
4
A. Hime,
6
M. Howe,
15
J. G. Hykawy,
4
M. C. P. Isaac,
5
P. Jagam,
12
N. A. Jelley,
13
C. Jillings,
9
G. Jonkmans,
4,1
J. Karn,
12
P. T. Keener,
14
K. Kirch,
6
J. R. Klein,
14
A. B. Knox,
13
R. J. Komar,
10,9
R. Kouzes,
8
T. Kutter,
10
C. C. M. Kyba,
14
J. Law,
12
I. T. Lawson,
12
M. Lay,
13
H. W. Lee,
9
K. T. Lesko,
5
J. R. Leslie,
9
I. Levine,
3
W. Locke,
13
M. M. Lowry,
8
S. Luoma,
4
J. Lyon,
13
S. Majerus,
13
H. B. Mak,
9
A. D. Marino,
5
N. McCauley,
13
A. B. McDonald,
9,8
D. S. McDonald,
14
K. McFarlane,
3
G. McGregor,
13
W. McLatchie,
9
R. Meijer Drees,
15
H. Mes,
3
C. Mifflin,
3
G. G. Miller,
6
G. Milton,
1
B. A. Moffat,
9
M. Moorhead,
13,5
C. W. Nally,
10
M. S. Neubauer,
14
F. M. Newcomer,
14
H. S. Ng,
10
A. J. Noble,
3,
‡
E. B. Norman,
5
V. M. Novikov,
3
M. O’Neill,
3
C. E. Okada,
5
R. W. Ollerhead,
12
M. Omori,
13
J. L. Orrell,
15
S. M. Oser,
14
A. W. P. Poon,
5,6,10,15
T. J. Radcliffe,
9
A. Roberge,
4
B. C. Robertson,
9
R. G. H. Robertson,
15,6
J. K. Rowley,
2
V. L. Rusu,
14
E. Saettler,
4
K. K. Schaffer,
15
A. Schuelke,
5
M. H. Schwendener,
4
H. Seifert,
4,6,15
M. Shatkay,
3
J. J. Simpson,
12
D. Sinclair,
3
P. Skensved,
9
A. R. Smith,
5
M. W. E. Smith,
15
N. Starinsky,
3
T. D. Steiger,
15
R. G. Stokstad,
5
R. S. Storey,
7,†
B. Sur,
1,9
R. Tafirout,
4
N. Tagg,
12
N. W. Tanner,
13
R. K. Taplin,
13
M. Thorman,
13
P. Thornewell,
6,13,15
P. T. Trent,
13,
* Y. I. Tserkovnyak,
10
R. Van Berg,
14
R. G. Van de Water,
14,6
C. J. Virtue,
4
C. E. Waltham,
10
J.-X. Wang,
12
D. L. Wark,
13,6,
§
N. West,
13
J. B. Wilhelmy,
6
J. F. Wilkerson,
15,6
J. Wilson,
13
P. Wittich,
14
J. M. Wouters,
6
and M. Yeh
2
(SNO Collaboration)
1
Atomic Energy of Canada Limited, Chalk River Laboratories, Chalk River, Ontario K0J 1J0 Canada
2
Chemistry Department, Brookhaven National Laboratory, Upton, New York 11973-5000
3
Carleton University, Ottawa, Ontario K1S 5B6 Canada
4
Department of Physics and Astronomy, Laurentian University, Sudbury, Ontario P3E 2C6 Canada
5
Institute for Nuclear and Particle Astrophysics and Nuclear Science Division, Lawrence Berkeley National Laboratory,
Berkeley, California 94720
6
Los Alamos National Laboratory, Los Alamos, New Mexico 87545
7
National Research Council of Canada, Ottawa, Ontario K1A 0R6 Canada
8
Department of Physics, Princeton University, Princeton, New Jersey 08544
9
Department of Physics, Queen’s University, Kingston, Ontario K7L 3N6 Canada
10
Department of Physics and Astronomy, University of British Columbia, Vancouver, BC V6T 1Z1 Canada
11
Department of Physics, University of California, Irvine, California 92717
12
Physics Department, University of Guelph, Guelph, Ontario N1G 2W1 Canada
13
Nuclear and Astrophysics Laboratory, University of Oxford, Keble Road, Oxford, OX1 3RH, United Kingdom
14
Department of Physics and Astronomy, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6396
15
Center for Experimental Nuclear Physics and Astrophysics, and Department of Physics, University of Washington,
Seattle, Washington 98195
(Received 18 June 2001; published 25 July 2001)
Solar neutrinos from
8
B decay have been detected at the Sudbury Neutrino Observatory via the
charged current (CC) reaction on deuterium and the elastic scattering (ES) of electrons. The flux of
n
e
’s is measured by the CC reaction rate to be f
CC
共n
e
兲 苷 1.75 6 0.07共stat兲
10.12
20.11
共syst兲 6 0.05共theor兲 3
10
6
cm
22
s
21
. Comparison of f
CC
共n
e
兲 to the Super-Kamiokande Collaboration’s precision value of the
flux inferred from the ES reaction yields a 3.3s difference, assuming the systematic uncertainties are
normally distributed, providing evidence of an active non-n
e
component in the solar flux. The total flux
of active
8
B neutrinos is determined to be 5.44 6 0.99 3 10
6
cm
22
s
21
.
DOI: 10.1103/PhysRevLett.87.071301 PACS numbers: 26.65.+t, 14.60.Pq, 95.85.Ry
071301-1 0031-9007兾01兾 87(7)兾071301(6)$15.00 © 2001 The American Physical Society 071301-1

VOLUME 87, N
UMBER 7 PHYSICAL REVIEW LETTERS 13A
UGUST 2001
Solar neutrino experiments over the past 30 years [1–6]
have measured fewer neutrinos than are predicted by mod-
els of the Sun [7,8]. One explanation for the deficit is
the transformation of the Sun’s electron-type neutrinos
into other active flavors. The Sudbury Neutrino Obser-
vatory (SNO) measures the
8
B solar neutrinos through the
reactions
n
e
1 d ! p 1 p 1 e
2
共CC兲 ,
n
x
1 d ! p 1 n 1n
x
共NC兲 ,
n
x
1 e
2
! n
x
1 e
2
共ES兲 .
The charged current (CC) reaction is sensitive exclusively
to electron-type neutrinos, while the neutral current (NC)
is sensitive to all active neutrino flavors (x
苷 e, m, t). The
elastic scattering (ES) reaction is sensitive to all flavors as
well, but with reduced sensitivity to n
m
and n
t
. By it-
self, the ES reaction cannot provide a measure of the total
8
B flux or its flavor content. Comparison of the
8
B flux
deduced from the ES reaction, assuming no neutrino os-
cillations [f
ES
共n
x
兲], to that measured by the CC reaction
[f
CC
共n
e
兲] can provide clear evidence of flavor transfor-
mation without reference to solar model flux calculations.
If neutrinos from the Sun change into other active flavors,
then f
CC
共n
e
兲 ,f
ES
共n
x
兲.
This Letter presents the first results from SNO on the ES
and CC reactions. SNO’s measurement of f
ES
共n
x
兲 is con-
sistent with previous measurements described in Ref. [5].
The measurement of f
CC
共n
e
兲, however, is significantly
smaller and is therefore inconsistent with the null hypothe-
sis that all observed solar neutrinos are n
e
. A measurement
using the NC reaction, which has equal sensitivity to all
neutrino flavors, will be reported in a future publication.
SNO [9] is an imaging water
ˇ
Cerenkov detector lo-
cated at a depth of 6010 m of water equivalent in the
INCO, Ltd. Creighton mine near Sudbury, Ontario. It
features 1000 metric tons of ultrapure D
2
O contained in
a 12-m diameter spherical acrylic vessel. This sphere
is surrounded by a shield of ultrapure H
2
O contained in
a 34-m-high barrel-shaped cavity of maximum diameter
22 m. A stainless steel structure 17.8 m in diameter sup-
ports 9456 20-cm photomultiplier tubes (PMTs) with light
concentrators. Approximately 55% of the light produced
within 7 m of the center of the detector will strike a PMT
if it is not absorbed by intervening media.
The data reported here were recorded between Novem-
ber 2, 1999 and January 15, 2001 and correspond to a live
time of 240.95 days. Events are defined by a multiplic-
ity trigger of 18 or more PMTs exceeding a threshold of
⬃0.25 photoelectrons within a time window of 93 ns. The
trigger reaches 100% efficiency at 23 PMTs. The total
instantaneous trigger rate is 15–18 Hz, of which 6–8Hz
is the data trigger. For every event trigger, the time and
charge responses of each participating PMT are recorded.
The data were partitioned into two sets, with approxi-
mately 70% used to establish the data analysis procedures
and 30% reserved for a blind test of statistical bias in the
analysis. The analysis procedures were frozen before the
blind data set was analyzed, and no statistically significant
differences in the data sets were found. We present here
the analysis of the combined data sets.
Calibration of the PMT time and charge pedestals,
slopes, offsets, charge vs time dependencies, and second
order rate dependencies are performed using electronic
pulsers and pulsed light sources. Optical calibration is
obtained by using a diffuse source of pulsed laser light at
337, 365, 386, 420, 500, and 620 nm. The absolute energy
scale and uncertainties are established with a triggered
16
N source (predominantly 6.13-MeV g’s) deployed over
two planar grids within the D
2
O and a linear grid in the
H
2
O. The resulting Monte Carlo predictions of detector
response are tested using a
252
Cf neutron source, which
provides an extended distribution of 6.25-MeV g rays
from neutron capture, and a
3
H共p, g兲
4
He [10] source
providing 19.8-MeV g rays. The volume-weighted mean
response is approximately nine PMT hits per MeV of
electron energy.
Table I details the steps in data reduction. The first of
these is the elimination of instrumental backgrounds. Elec-
trical pickup may produce false PMT hits, while electrical
discharges in the PMTs or insulating detector materials
produce light. These backgrounds have characteristics
very different from
ˇ
Cerenkov light, and are eliminated by
using cuts based only on the PMT positions, the PMT time
and charge data, event-to-event time correlations, and veto
PMTs. This step in the data reduction is verified by com-
paring results from two independent background rejection
analyses.
For events passing the first stage, the calibrated times
and positions of the hit PMTs are used to reconstruct the
vertex position and the direction of the particle. The re-
construction accuracy and resolution are measured using
Compton electrons from the
16
N source, and the energy
and source variation of reconstruction are checked with a
8
Li b source. Angular resolution is measured using Comp-
ton electrons produced more than 150 cm from the
16
N
source. At these energies, the vertex resolution is 16 cm
and the angular resolution is 26.7
±
.
An effective kinetic energy, T
eff
, is assigned to each
event passing the reconstruction stage. T
eff
is calculated
TABLE I. Data reduction steps.
Analysis step Number of events
Total event triggers 355 320 964
Neutrino data triggers 143 756 178
N
hit
$ 30 6 372 899
Instrumental background cuts 1 842 491
Muon followers 1 809 979
High level cuts
a
923 717
Fiducial volume cut 17 884
Threshold cut 1169
Total events 1169
a
Reconstruction figures of merit, prompt light, and 具u
ij
典.
071301-2 071301-2

VOLUME 87, N
UMBER 7 PHYSICAL REVIEW LETTERS 13A
UGUST 2001
using prompt (unscattered)
ˇ
Cerenkov photons and the po-
sition and direction of the event. The derived energy re-
sponse of the detector can be characterized by a Gaussian:
R共E
eff
, E
e
兲 苷
1
p
2ps
E
共E
e
兲
exp
∑
2
1
2
µ
E
eff
2 E
e
s
E
共E
e
兲
∂
2
∏
,
where E
e
is the total electron energy, E
eff
苷
T
eff
1 m
e
, and s
E
共E
e
兲 苷 共20.4620 1 0.5470
p
E
e
1
0.008722E
e
兲 MeV is the energy resolution. The uncer-
tainty on the energy scale is found to be 61.4%, which
results in a flux uncertainty nearly 4 times larger. For
validation, a second energy estimator counts all PMTs
hit in each event, N
hit
, without position and direction
corrections.
Further instrumental background rejection is obtained
by using reconstruction figures of merit, PMT time residu-
als, and the average angle between hit PMTs (具u
ij
典), mea-
sured from the reconstructed vertex. These cuts test the
hypothesis that each event has the characteristics of single
electron
ˇ
Cerenkov light. The effects of these and the rest of
the instrumental background removal cuts on neutrino sig-
nals are quantified using the
8
Li and
16
N sources deployed
throughout the detector. The volume-weighted neutrino
signal loss is measured to be 1.4
10.7
20.6
% and the residual in-
strumental contamination for the data set within the D
2
O
is ,0.2%. Lastly, cosmic ray induced neutrons and spal-
lation products are removed using a 20 s coincidence win-
dow with the parent muon.
Figure 1 shows the radial distribution of all remaining
events above a threshold of T
eff
$ 6.75 MeV. The distri-
bution is expressed as a function of the volume-weighted
radial variable 共R兾R
AV
兲
3
, where R
AV
苷 6.00 m is the ra-
dius of the acrylic vessel. Above this energy threshold,
there are contributions from CC events in the D
2
O, ES
events in the D
2
O and H
2
O, a residual tail of neutron cap-
ture events, and high energy g rays from radioactivity in
the outer detector. The data show a clear signal within the
D
2
O volume. For 共R兾R
AV
兲
3
. 1.0 the distribution rises
into the H
2
O region until it is cut off by the acceptance
of the PMT light collectors at R ⬃ 7.0 m. A fiducial vol-
ume cut is applied at R 苷 5.50 m to reduce backgrounds
from regions exterior to the D
2
O, and to minimize system-
atic uncertainties associated with optics and reconstruction
near the acrylic vessel.
Possible backgrounds from radioactivity in the D
2
O and
H
2
O are measured by regular low level radio assays of
U and Th decay chain products in these regions. The
ˇ
Cerenkov light character of D
2
O and H
2
O radioactivity
backgrounds is used
in situ to monitor backgrounds be-
tween radio assays. Low energy radioactivity backgrounds
are removed by the high threshold imposed, as are most
neutron capture events. Monte Carlo calculations predict
that the H
2
O shield effectively reduces contributions of low
energy (,4 MeV) g rays from the PMT array, and these
predictions are verified by deploying an encapsulated Th
source in the vicinity of the PMT support sphere. High
energy g rays from the cavity are also attenuated by the
Acrylic vessel
PMT array
(R/R
AV
)
3
Events/(0.07 Cubic AV Radii)
0
20
40
60
80
100
120
140
0 0.5 1 1.5 2 2.5 3 3.5
FIG. 1. Distribution of event candidates with T
eff
$ 6.75 MeV
as a function of the volume-weighted radial variable 共R兾R
AV
兲
3
.
The Monte Carlo simulation of the signals, weighted by the
results from the signal extraction, is shown as a histogram. The
dotted line indicates the fiducial volume cut used in this analysis.
H
2
O shield. A limit on their leakage into the fiducial vol-
ume is estimated by deploying the
16
N source near the
edge of the detector’s active volume. The total contribution
from all radioactivity in the detector is found to be ,0.2%
for low energy backgrounds and ,0.8% for high energy
backgrounds.
The final data set contains 1169 events after the fidu-
cial volume and kinetic energy threshold cuts. Figure 2(a)
displays the distribution of cosu
Ø
, the angle between the
reconstructed direction of the event and the instantaneous
direction from the Sun to the Earth. The forward peak
in this distribution arises from the kinematics of the ES
reaction, while CC electrons are expected to have a distri-
bution which is
共1 2 0.340 cosu
Ø
兲 [11], before accounting
for detector response.
The data are resolved into contributions from CC, ES,
and neutron events above threshold using probability den-
sity functions (pdf’s) in T
eff
, cosu
Ø
, and 共R兾R
AV
兲
3
, gen-
erated from Monte Carlo simulations assuming no flavor
transformation and the shape of the standard
8
B spec-
trum [12] (hep neutrinos are not included in the fit). The
extended maximum likelihood method used in the signal
extraction yields 975.4 6 39.7 CC events, 106.1 6 15.2
ES events, and 87.5 6 24.7 neutron events for the fiducial
volume and the threshold chosen, where the uncertainties
given are statistical only. The dominant sources of sys-
tematic uncertainty in this signal extraction are the energy
scale uncertainty and reconstruction accuracy, as shown in
Table II. The CC and ES signal decomposition gives con-
sistent results when used with the N
hit
energy estimator,
as well as with different choices of the analysis threshold
071301-3 071301-3

VOLUME 87, N
UMBER 7 PHYSICAL REVIEW LETTERS 13A
UGUST 2001
cos θ
O
Events/(0.1 wide bin)
.
.
...
..
(a)
Kinetic energy (MeV)
Events/(0.51 MeV bin)
(b)
Super-K ES flux
Kinetic energy (MeV)
Data/BPB01 ( B only)
(c)
8
0
25
50
75
100
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1
0
100
200
7 8 9 10111220
0
0.2
0.4
0.6
7 8 9 10111220
FIG. 2. Distributions of (a) cosu
Ø
and (b) extracted kinetic
energy spectrum for CC events with R # 5.50 m and T
eff
$
6.75 MeV. The Monte Carlo simulations for an undistorted
8
B
spectrum are shown as histograms. The ratio of the data to the
expected kinetic energy distribution with correlated systematic
errors is shown in (c). The uncertainties in the
8
B spectrum [12]
have not been included.
and the fiducial volume up to 6.20 m with backgrounds
characterized by pdf’s.
The CC spectrum can be extracted from the data by
removing the constraint on the shape of the CC pdf and
repeating the signal extraction.
Figure 2(b) shows the kinetic energy spectrum with sta-
tistical error bars, with the
8
B spectrum of Ortiz et al. [12]
scaled to the data. The ratio of the data to the predic-
tion [7] is shown in Fig. 2(c). The bands represent the 1s
uncertainties derived from the most significant energy-
dependent systematic errors. There is no evidence for a
deviation of the spectral shape from the predicted shape
under the nonoscillation hypothesis.
Normalized to the integrated rates above the kinetic en-
ergy threshold of T
eff
苷 6.75 MeV, the measured
8
B neu-
trino fluxes assuming the standard spectrum shape [12] are
f
CC
SNO
共n
e
兲 苷 1.75 6 0.07共stat兲
10.12
20.11
共syst兲 6 0.05共theor兲
3 10
6
cm
22
s
21
f
ES
SNO
共n
x
兲 苷 2.39 6 0.34共stat兲
10.16
20.14
共syst兲
3 10
6
cm
22
s
21
,
TABLE II. Systematic error on fluxes.
Error source CC error ES error
(percent) (percent)
Energy scale 25.2, 16.1 23.5, 15.4
Energy resolution 60.5 60.3
Energy scale nonlinearity 60.5 60.4
Vertex accuracy 63.1 63.3
Vertex resolution 60.7 60.4
Angular resolution 60.5 62.2
High energy g’s 20.8, 10.0 21.9, 10.0
Low energy background 20.2, 10.0 20.2, 10.0
Instrumental background 20.2, 10.0 20.6, 10.0
Trigger efficiency 0.0 0.0
Live time 60.1 60.1
Cut acceptance 20.6, 10.7 20.6, 10.7
Earth orbit eccentricity 60.1 60.1
17
O,
18
O 0.0 0.0
Experimental uncertainty 26.2, 17.0 25.7, 16.8
Cross section 3.0 0.5
Solar Model 216, 120 216, 120
where the theoretical uncertainty is the CC cross section
uncertainty [13]. Radiative corrections have not been ap-
plied to the CC cross section, but they are expected to
decrease the measured f
CC
共n
e
兲 flux [14] by up to a few
percent. The difference between the
8
B flux deduced from
the ES rate and that deduced from the CC rate in SNO
is 0.64 6 0.40 3 10
6
cm
22
s
21
,or1.6s. The SNO’sES
rate measurement is consistent with the precision measure-
ment by Super-Kamiokande Collaboration of the
8
B flux
using the same ES reaction [5]:
f
ES
SK
共n
x
兲 苷 2.32 6 0.03共stat兲
10.08
20.07
共syst兲 3 10
6
cm
22
s
21
.
The difference between the flux f
ES
共n
x
兲 measured by
Super-Kamiokande via the ES reaction and the f
CC
共n
e
兲
flux measured by SNO via the CC reaction is 0.57 6
0.17 3 10
6
cm
22
s
21
,or3.3s [15], assuming that the sys-
tematic errors are normally distributed. The probability
that a downward fluctuation of the Super-Kamiokande re-
sult would produce a SNO result $3.3s is 0.04%. For
reference, the ratio of the SNO CC
8
B flux to that of the
BPB01 solar model [7] is 0.347 6 0.029, where all uncer-
tainties are added in quadrature.
If oscillation solely to a sterile neutrino is occurring, the
SNO CC-derived
8
B flux above a threshold of 6.75 MeV
will be consistent with the integrated Super-Kamiokande
ES-derived
8
B flux above a threshold of 8.5 MeV [16]. By
adjusting the ES threshold [5], this derived flux difference
is 0.53 6 0.17 3 10
6
cm
22
s
21
,or3.1s . The probabil-
ity of a downward fluctuation $3.1s is 0.13%. These
data are therefore evidence of a nonelectron active flavor
component in the solar neutrino flux. These data are also
inconsistent with the “Just-So
2
” parameters for neutrino
oscillation [17].
071301-4 071301-4

VOLUME 87, N
UMBER 7 PHYSICAL REVIEW LETTERS 13A
UGUST 2001
φ(ν
e
) (10
6
cm
-2
s
-1
)
φ(ν
µτ
) (10
6
cm
-2
s
-1
)
φ(ν
e
) (relative to BPB01)
φ(ν
µτ
) (relative to BPB01)
φ
ES
= φ(ν
e
) + 0.154 φ(ν
µτ
)φ
SK
φ
CC
φ
ES
φ
SNO
φ
SK
φ
x
φ
x
φ
SK+SNO
φ
SSM
0123456
0
2
4
6
8
0 0.2 0.4 0.6 0.8 1 1.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
FIG. 3. Flux of
8
B solar neutrinos which are m or t flavor
vs the flux of electron neutrinos as deduced from the SNO and
Super-Kamiokande data. The diagonal bands show the total
8
B flux f
共n
x
兲 as predicted by BPB01 (dashed lines) and that
derived from the SNO and Super-Kamiokande measurements
(solid lines). The intercepts of these bands with the axes repre-
sent the 6
1s errors.
Figure 3 displays the inferred flux of nonelectron fla-
vor active neutrinos [f共n
mt
兲] against the flux of electron
neutrinos. The two data bands represent the one stan-
dard deviation measurements of the SNO CC rate and the
Super-Kamiokande ES rate. The error ellipses represent
the 68%, 95%, and 99% joint probability contours for
f共n
e
兲 and f共n
mt
兲. The best fittof共n
mt
兲 is
f共n
mt
兲 苷 3.69 6 1.13 3 10
6
cm
22
s
21
.
The total flux of active
8
B neutrinos is determined to be
f共n
x
兲 苷 5.44 6 0.99 3 10
6
cm
22
s
21
.
This result is displayed as a diagonal band in Fig. 3, and is
in excellent agreement with predictions of standard solar
models [7,8].
Assuming that the oscillation of massive neutrinos ex-
plains both the evidence for the electron neutrino flavor
change presented here and the atmospheric neutrino data
of the Super-Kamiokande collaboration [18], two separate
splittings of the squares of the neutrino mass eigenval-
ues are indicated: ,10
23
eV
2
for the solar sector [19,17]
and ⯝ 3.5 3 10
23
eV
2
for atmospheric neutrinos. These
results, together with the beta spectrum of tritium [20],
limit the sum of mass eigenvalues of active neutrinos to
be between 0.05 and 8.4 eV, corresponding to a constraint
of 0.001 ,V
n
, 0.18 for the contribution to the critical
density of the Universe [21,22].
In summary, the results presented here are the first direct
indication of a nonelectron flavor component in the solar
neutrino flux, and enable the first determination of the total
flux of
8
B neutrinos generated by the Sun.
This research was supported by the Natural Sciences
and Engineering Research Council of Canada, Industry
Canada, National Research Council of Canada, North-
ern Ontario Heritage Fund Corporation, the Province of
Ontario, the United States Department of Energy, and in
the United Kingdom by the Science and Engineering Re-
search Council and the Particle Physics and Astronomy
Research Council. Further support was provided by INCO,
Ltd., Atomic Energy of Canada Limited (AECL), Agra-
Monenco, Canatom, Canadian Microelectronics Corpo-
ration, AT&T Microelectronics, Northern Telecom, and
British Nuclear Fuels, Ltd. The heavy water was loaned by
AECL with the cooperation of Ontario Power Generation.
*Permanent address: Birkbeck College, University of Lon-
don, Malet Road, London WC1E 7HX, UK.
†
Deceased.
‡
Permanent address: TRIUMF, 4004 Wesbrook Mall, Van-
couver, BC V6T 2A3, Canada.
§
Permanent address: Rutherford Appleton Laboratory,
Chilton, Didcot, Oxon, OX11 0QX, and University of
Sussex, Physics and Astronomy Department, Brighton
BN1 9QH, United Kingdom.
[1] B. T. Cleveland
et al., Astrophys. J. 496
, 505 (1998).
[2] K. S. Hirata et al., Phys. Rev. Lett. 65
,
1297 (1990); K. S.
Hirata et al., Phys. Rev. D 44
, 2241 (1991); 45
, 2170(E)
(1992); Y. Fukuda et al., Phys. Rev. Lett. 77
, 1683 (1996).
[3] J. N. Abdurashitov et al., Phys. Rev. C 60
,
055801
(1999).
[4] W. Hampel et al., Phys. Lett. B 447
, 127 (1999).
[5] S. Fukuda et al., Phys. Rev. Lett. 86
, 5651 (2001).
[6] M. Altmann et al., Phys. Lett. B 490
,
16 (2000).
[7] J. N. Bahcall, M. H. Pinsonneault, and S. Basu, Astro-
phys. J. 555
, 990 (2001). The reference
8
B neutrino flux
is 5.05 3 10
6
cm
22
s
21
.
[8] A. S. Brun, S. Turck-Chièze, and J. P. Zahn, Astrophys. J.
525
,
1032 (1999); S. Turck-Chièze et al., Astrophys. J.
Lett. 555
,
L69 (2001).
[9] The SNO Collaboration, J. Boger et al., Nucl. Instrum.
Methods Phys. Res., Sect. A 449
, 172 (2000).
[10] A. W. P. Poon et al., Nucl. Instrum. Methods Phys. Res.,
Sect. A 452
, 115 (2000).
[11] P. Vogel and J. F. Beacom, Phys. Rev. D 60
, 053003
(1999).
[12] C. E. Ortiz et al., Phys. Rev. Lett. 85
, 2909 (2000).
[13] S. Nakamura, T. Sato, V. Gudkov, and K. Kubodera, Phys.
Rev. C 63
, 034617 (2001); M. Butler, J.-W. Chen, and
X. Kong, Phys. Rev. C 63
, 035501 (2001); G. ’t Hooft,
Phys. Lett. 37B
, 195 (1971). The Butler et al. cross section
with
L
I,A
苷 5.6 fm
3
is used.
[14] I. S. Towner, J. Beacom, and S. Parke (private commu-
nication); I. S. Towner, Phys. Rev. C 58
, 1288 (1998);
J. Beacom and S. Parke, hep-ph/0106128; J. N. Bahcall,
M. Kamionkowski, and A. Sirlin, Phys. Rev. D 51
, 6146
(1995).
[15] Given the limit set for the hep flux by Ref. [5], the effects
of the hep contribution may increase this difference by a
few percent.
071301-5 071301-5

VOLUME 87, N
UMBER 7 PHYSICAL REVIEW LETTERS 13A
UGUST 2001
[16] G. L. Fogli, E. Lisi, A. Palazzo, and F. L. Villante, Phys.
Rev. D 63
,
113016 (2001); F. L. Villante, G. Fiorentini, and
E. Lisi, Phys. Rev. D 59
, 013006 (1999).
[17] J. N. Bahcall, P. I. Krastev, and A. Yu. Smirnov, J. High
Energy Phys. 05
, 015 (2001).
[18] T. Toshito
et al., hep-ex/0105023.
[19] M. Apollonio et al., Phys. Lett. B 466
,
415 (1999).
[20] J. Bonn et al., Nucl. Phys. (Proc. B Suppl.) 91
,
273 (2001).
[21] Allen’s Astrophysical Quantities, edited by Arthur Cox
(Springer-Verlag, New York, 2000), 4th ed.; D. E. Groom
et al., Eur. Phys. J. C 15
, 1 (2000).
[22] H. Pas and T. J. Weiler, Phys. Rev. D 63
,
113015 (2001).
071301-6 071301-6