Progress in Particle and Nuclear Physics PERGAMON
Progress
in Particle
and Nuclear
Physics 40 (1998)
113-121
Neutral-Current Detection via 3He(n,p)3H in the Sudbury Neutrino Observatory for the Sudbury
R. G. H. ROBERTSON”
Neutrino
Observatory
Collaboration
“U!rIl’r!Yir~ 0/ Uirs/lin,qloU. Smfrlr. W./lYXIYS.1 .s..‘l. The rate of neutral-current comparison
neutrino
of the rate for electron
interactions
in SNO can be measured
elastic scattering
rate on deuterium,
b) neutron
c) neutron
in an array of 3He-filled proportional
capture
capture
on Cl ions added to the heavy water, and,
proportional
counters,
radioactivity
are essential to minimize photodiiintegration
for producing
the technique
ultra-pure
by: a)
to the pure charged-current
counters.
In the case of
described in this paper, extremely
nickel counter
low levels of
backgrounds.
A method
bodies by chemical vapor deposition
from
Ni(CO)k has been shown to yield purities with respect to U and Th of order 10-r’ by weight. between
Pulse-shape
3He(n,p)3H
digitization
affords a powerful
events and background
have been developed
for deriving
means for diitinguishing
alphas or electrons.
Novel methods
the position of events with singleended
readout
from counter strings up to 12 m in length.
1
The Sudbury Neutrino Observatory
The Sudbury
Neutrino
Observatory
to resolve the “solar neutrino pure heavy water Creighton
(SNO) [l] is a laboratory
problem.”
Deuterium’s
modes by which low-energy
d + u,
+
p+p+e--
d + uz
+
p+n+v,-2.22
d+p=
excavated
neutrinos (CC)
MeV
(NC)
-)
n + n + e+ - 4.03 MeV
(CA)
-+
e-+v,
(ES)
PlI: SOl46-6410(98)00015-5
0
1998
and
is 1000 tonnes of 99.92% isotopically 2070 m below ground in the INCO
make it ideal for the study of neutrino
1.44 MeV
0146~6410~98~$19.00+0.00
chamber
of neutrinos
Ontario.
unique nuclear properties
are four principal
e-+v,
The heart of the detector
(DsO) placed in a specially
nickel mine near Sudbury,
to study the properties
can interact
interactions.
with the heavy water:
Elsevier Science BV. All rights reserved. Printed In Great Britain
There
114
R. G. H. Roherrsort 1 Prog. Part. Nucl. Ph_w. 40 (1998)
The first of these reactions
proceeds
only, the second is the neutral-current probability
charged-current
interaction
(ES). While neutrinos
W-boson
SNO therefore
events produce
The resulting
diameter
in coincidence
are produced
One is to dissolve
Cerenkov
the third is the
from electrons,
the cross section
of neutrino
radiation
contribution
of
oscillations
largely indepen-
of the CC rate to the NC rate.
photomultipliers
that is detected surrounding
is the signal that
would identify strategies
on 75% abundant can be detected
in SNO by an array of 9,500
the acrylic sphere that holds the
a NC event has occurred a pe interaction).
for detecting
in the heavy water several tonnes
captures
of for
by the Sun as it burns its nuclear fuel, but
to reveal the presence
of possible
with equal
may reach the Earth as other flavors.
with a positron
and a number
MgC12. When a neutron
and can be initiated
(CA), and the fourth is the elastic scattering elastically
Cerenkov radiation
of a free neutron
moderator,
be devised.
neutrinos
via the direct comparison
R1408 204mm
or two neutrons
of deuterium
of electron neutrinos
6 times that for other flavors as a result of the additional
has the capability
Charged-current
DzO. The detection
antineutrinos
occur, then the neutrinos
dent of solar properties
(CC) interaction
(vZ = I+, v,, or Y,) and their antiparticles,
flavors can scatter
Only electron
oscillations
Hamamatsu
of electron
all active
is about
exchange.
if neutrino
excellent
(NC) disintegration
by any of the active neutrinos
neutrinos electron
by the charged-current
113-121
(detection
Heavy water is an
the thermalized
of a chloride
of one
neutrons
can
salt, such as NaCl or
35C1, it emits 8.6 MeV in gammas,
which shower.
by the PMT array in the same way CC events are
detected. The method
described
in the heavy water.
2 2.1
in this paper is to detect
the neutron
The NC and CC signals are completely
Neutral-Current
in 3He proportional
separate
in this approach
counters
placed
[2, 3, 41.
Detector Array
The Signal
SNO is sensitive reactions
to only the sB neutrinos
on deuterium,
sB are constrained
and the existence
by current
are reported
the corresponding
of a Cerenkov threshold.
observations
rate RES in terms of the effective
The results
from the Sun, as a result of the Q-values
as 2.44f0.06
?i$
The fluxes of active neutrinos
with the SuperKamiokande
CC and NC cross sections
detector
x lo6 v, cm- 2 s-l . If conversion
and Pinsomreault
from
[5]. Writing the total
a, and a,,,, and the fluxes & and 4,,,,,
flux is 17.2 x lo6 + + u, cme2 s-l, which provides
in SNO. The solar model of Bahcall
for CC and NC
[6] predicts
to uG + v, is complete,
an upper bound to the NC rate
a v, flux of 6.6 x lo6 v, cmd2 s-l,
R. G. H. Robertson / Prog. PUI.I. Nucl. Phy.
Table 1: Rates for neutral-current MeV induced
interactions
40 /199X/ 113-I-71
in SNO and charged-current
115
above T, = 5
interactions
by sB neutrinos.
FhlX lo6 cm-’ s-l
Detected
Species
CC rate
Total NC rate
Detected
NC rate
Y-1
Y-1
Y-1
NC/CC
2.44
“.Z
4170
2210
550
1.78
ve
3040
1610
400
4.82
VP + VT
0
4370
1090
6.6
v, + up + r+
3040
5980
1500
0.493
17.2
“fi + VT
0
15600
3900
00
yielding an intermediate those scenarios
scenario.
are summarized
The expected
0.132
rates in SNO for CC and NC neutrino
in Table 1. For simplicity,
interactions
the shape of the 8B spectrum
in
is taken to
be undistorted. In deriving
the detected
rates from the total rates in Table 1, the efficiency for neutron
(after cuts) is taken to be 25%, and the threshold energy.
Bahcall
and Lisi [7] integrate
for detecting
the Kubodera-Nozawa
electrons
is taken to be 5 MeV kinetic
cross sections
[8] over the sB spectrum
and find for the NC cross section 0.478 x 10e4’ cm2, and for the ratio of the CC interaction 5 MeV electron
energy to the total NC interaction
of the effect of instrumental sre estimated
2.2
resolution.
rate 1.882. The latter figure includes
Uncertainties
detection
in the NC and CC cross sections
rate above an estimate individually
to be S%, but only 1% in the ratio.
Physics Issues in Design of Array
Detection
of thermal
takes advantage applications counters
neutrons
of an enormous
in neutrino
physics
for SNO poses unusual
by the reaction
cross section,
[9, 10, 111. Nonetheless,
from naturally
be allowed to leach significantly permeates
readily
water at absolute
through pressures
art that has included
the design of an array of 3He proportional
problems.
The signal, from 4 to 43 neutrons by backgrounds
5330 b, and is a well developed
per day produced
occurring
radioactivity
in the heavy water, must not be overwhelmed in construction
into the heavy water and interfere
many otherwise
acceptable
materials.
materials,
nor can substances
with water purification. The counters
up to 3.5 atm for ten years or more, placing stringent
Helium
are to survive under demands on detector
longevity and stability. A counter array with high neutron efficiency is needed; but not at the cost of significantly obscuring the cerenkov light from CC events. These conditions severely constrain detector
R. G. H. Robertson/
116
design, the materials
employed
Prog. Pmt. Nucl. Phys. 40 (1998)
in constructing
the detector
113-121
array, and the methods
of handling
and
deployment. Backgrounds by gammas
above 2.22 MeV, as the neutrons
odisintegration energetic
of several kinds can be identified.
neutrons.
gammas.
At the bottom
In principle,
capture
of neutrons
cannot
of the natural
emerging
activities
events,
be distinguished
from neutrin-
also have high energy gammas,
78 days, to be relevant.
Energetic
particularly
from the wall of the proportional
in the gas as a 3He(n,p)T
Even very low-energy
directly
of deuterium
Th and U decay chains are two sufficiently
and by (cr,py) and (cu,nr) reactions,
Alpha particles of ionization
produced
many cosmogenic
one, 56Co, has a long enough halflife,
Most serious is photodisintegration
event, as can electrons
such as the decay of tritium,
photons
but only
can be produced
by
outside the SNO detector.
counter
can leave the same amount
from /3 decay and Compton can compromise
scattering.
the signal via random
summing. The principal materials
defense
and to assure
pulse-shape
discrimination
ion-current
profiles
against
these
a high degree
of cleanliness
and position
event by event.
backgrounds
encoding
during
breakdown
Radioactivity
and electromagnetic
must not only be minimized,
constraints
dictate
requiring
low-noise
interference
but in addition
assembly.
radioactivity Beyond
in construction
that,
techniques
are used, both of which require digitizing
the vessel at the ends of cables 8 - 15 m in length, high-voltage
is to minimize
that
design.
must be controlled.
it is essential
preamplifiers
for
the pulse are outside
Spurious
pulses from
Backgrounds
of all types
to have robust techniques
to messure
them
in situ.
2.3
Neutron-Capture
Because
of the effectiveness
neutron
capture
lattice with 1 m spacing
optical
on lattice
constant,
photons
calculations
gives a neutron
capture
via the charged are absorbed
subject
to systematic
For a given array geometry 3He used in the counters. thermal
neutrons.
(NCD) can be employed. on a square
efficiency
interference
current
(121 of 45 f5%
interaction.
and 3He partial
uncertainties
the neutron
for
with a total length of N 800 m arranged
approximations
Even at modest
The efficiency
detectors
pressure
capture
pressures
is highly sensitive
and different and differences
with tolerable
Approximately
12% of uniformly
of neutron are shown.
source is also shown ss a function
[12, 131 make use of certain
and are therefore
current
and the high cross-section
[?]. In Figure 1 the dependence
enrichment,
from a point
moderator
array of neutral counters
deuterium
of neutrons
sparse
as a neutron
proportional
light produced
generated
detection
of heavy water
on 3He a rather
An array of 5cm diameter
to Cerenkov
Efficiency
of radius. reference
capture
efficiency
The efficiency The Monte
for
Carlo
designs for detectors,
of order 10%.
efficiency is governed
by the partial
of l-3 atm, the counters
are essentially
to isotopic
As can be seen, neutrons
enrichment.
pressure
of
“black” to
R. G. H. Robertson / Prog. Purr. Nucl. Phw.
117
40 i 199X I I13~117l
60
.E 30 g w 0 20 ; E Q loo OA
1
I
1
I
I
0
1
2
3
4
5
01
6
’
’
98.4
1.0 8 0.8
i=/
.f 0.6 g W Q
Ll2
1 il
0.4
J
’
99.6
100.0
I ’
I
I
I
600
Radial Position of Source, cm
40
80
K
I
I I (a)
\
$0.2
400
’
’
r=5cm
_’
i?
200
’ 99.2
D, 0 isotopic Enrichment, %
3He Absolute Pressure, atm
0
c
98.8
i
I 1
120
Lattice Constant, cm
Figure 1: Neutron capture efficiency as a function of (a) 3He pressure, (b) isotopic enrichment of heavy water, (c) position of a point neutron source, and, (d) lattice constant of the square grid, with detector radius as a parameter (2 atm 3He). From ref. [12].
created near the wall of the vessel (at 600 cm), have a high probability of escaping rather than being captured, which reduces the detector’s sensitivity to backgrounds generated at or outside the vessel wall. For neutrons generated uniformly in the heavy water, Table 2 indicates their fates. For 99.85% enriched heavy water, the mean distance from the point of generation to the point of capture for thermal neutrons is 110 cm and the mean time to capture is 16 ms with the NCD array, compared with 48 cm and 4 ms, respectively, for 0.25% NaCl.
2.4
Gas Fill
The detectors are filled with a gas mixture of 85% 3He and 15% CFd at a total pressure of 2.5 atm to provide a good compromise between gas gain and stopping power (to mitigate “wall effect” wherein either the proton or the triton strikes the wall before the end of its range). A lower pressure, and
118
R. G. H. Robertson/
Prog. Part. Nucl. Phys. 40 (1998)
113-121
Table 2: Capture fraction by isotope, in %, uniform source: MCNP 4.2 calculation, 99.85% D20, 2.5 atm 3He, 0.5 atm CFI, O&mm thick Ni wall, 112 strings of detector with total length 808 m. “Acrylic” neutrons are generated at the vessel wall by (r,n) at E., = 2.6 MeV. Also shown are the fractions for 0.25% NaCl. 3He Detectors
0.25% NaCl
1Uniform
Acrylic
4.5
2.9
1.5
11.6
4.4
2.9
1.5
1.9
0.6 82.9
43.4
1.3
0.7
Material
Uniform
4crylic
3He
46.8
14.8
2H
11.8
‘H ~-6ZNi
r
33,37(-J
23Na IS-160
2.1
0.8
0.5
0.3
Acrylic
19.2
55.0
7.9
38.2
HsO (external)
5.2
19.7
2.5
15.8
1
therefore operating voltage, would simplify microdischarge management and increase drift speeds, but would require thick-walled counters to resist collapse. At the bottom of the SNO vessel, the absolute pressure is that due to a column of DzO of height 18 m plus 1.3 atmospheres of air pressure, for a total of 3.2 atmospheres. The copper anode wire of diameter 50 pm is low in radioactivity and ohmic losses. At 1800 V the gas gain is approximately 100. For higher gas gains positiveion space charge at the wire [4] causes the avalanche multiplication to depend increasingly on track orientation, ‘Ike
amounts of electronegative contaminants such as oxygen and water degrade severely the
performance of the gae in a sealed counter. Consequently, counter surfaces are etched in a solution of 4-M nitric acid, 0.17-M hydrofluoric acid for 50 minutes at room temperature, b&d
under vacuum,
and purged with boiloff N2 prior to fill. (The etch also removes surface debris stripped off the mandrel during the fabrication process.) The 3He gas from the Department of Energy facility in Savannah River contains a small amount of tritium, about 0.5 mCi/l, which is reduced to 5 nCi/l or lees by passage through a charcoal-loaded cold trap and by recirculation through a SAES St101 getter. At that level, random coincidences and pileup of tritium decay pulses are no longer a concern in the energy regime of interest. Some detectors, about 5% by length, are filled with 4He to provide a check on backgrounds.
R. G. H. Robertson / Prog. Parr. Nucl. Phj,s. 40 (1998)
2.5
119
113-121
Construction
The bodies of the proportional counters are made of ultra-pure nickel tubes fabricated at Mirotech (Toronto) Inc., by thermolysis of Ni(CO)d vapor at the surface of an anodized aluminum mandrel heated to 215 C. A limited number of elements (Pb, Ra, Th, and U not among them) react with CO to form carbonyls, and the metals formed by chemical vapor deposition (CVD) from this precursor can be expected to be free of the most troublesome radioactivities. Cobalt does form a carbonyl, and may be conveyed through this chemistry at a level that is not at present known. Analysis of the nickel deposit, which has properties very similar to conventional met8llurgic8l products, by radiochemical neutron activation analysis has shown [14] Th levels in the bulk material of order lo-i2
(1 ppt) by
weight or less. Nickel is a relatively inert metal and has been commonly used to provide corrosion protection for other metals. The exposed macroscopic surface 8rea in SNO is 125 m2. In pure (17.3MR)
water
the measured leach rate is 122 pg rnb2 d-‘, which is not expected to compromise the operation of the water purification plant. Endcaps are also made by CVD, in this case on stainless steel mandrels, and sre welded into the tubes with 8 Lumonics 50-W Nd-YAG laser welder. Insulators are Heraeus-Amersil Suprssil T-21 synthetic fused silica tubes. The insulators are internally coated with a layer of pyrolytic graphite at anode potential to eliminate electric fields inside them. They extend 2.5 cm into the gas volume to act 8s field tubes and prevent multiplication of electrons from regions where the electric field is distorted. Silica-nickel is 8 highly mismatched seal, so the design places the silica under compressive stress at working temperature8 to take advantage of the high compressive strength of that material. Techniques were developed in collaboration with IJ Research (Santa Ana), Inc. to metallize and solder the seals with a 965~3.5 eutectic Sn:Ag alloy. Counters are filled through copper tubes, which are then pinched off. All assembly is carried out in a Class 10-1000 cleanroom. A cross section of a counter is given in Hime [4]. The detectors have a wall thickness of 0.36 - 0.48 mm and 8re fabricated in unit lengths of 200, 227, and 272 cm in order to fill the sphere efficiently. ( A firm upper limit is set by the dimensions of the cage in the INCO Number 9 shaft, which permits 8n object 3.7 m long to be brought down without the need to suspend it beneath the cage.) The active length is 13 cm less than the mechanical length. A total mech8nical length of 750 m is to be deployed in the heavy water in the form of 96 “strings” up to 11 m long welded during deployment. Spaces of 35 to 50 cm are allowed between the string ends and the vessel. The bottom of each string is terminated with a 3Cknsopen-ended 415ohm delay line in aid of position reedout by pukreflection
timing. A single Ql-ohm cosxial cable, made
of copper and polyethylene by South Bay Cable (Idyllwyld, CA), Inc., carries signals from the top of each string up the neck of the vessel to preamplifiers. The cable is positively buoyant in heavy water.
R. G. H. Robertson/
120
Prog. Pm.
Nucl. Phys. 40 (1998)
Table 3: Photodisintegration
are installed
2.6
beta emitters.
y branch, %
(r,n) Probability
2.615
36%
i/470
(per Y)
3.26 (18%)
2.445
1.5%
l/750
(per 7)
ssco
1.8
var.
s2sAC
2.06 (11%)
Var.
234mPa
2.28 (100%)
4eK
1.3 (89%)
Isotope
Ep(,,), MeV E,, MeV
212Bi
1.8 (18%)
214Bi
Each string floats upward,
and high-energy
113-l.?/
is anchored
restrained
l/1125 0
89%
0 0
1.46 (EC)
to an attachment
point afhxed to the bottom
by a flexible braided
by a remotely
(per decay)
anchor attachment
of the acrylic vessel and
made of Vectran
fiber. The strings
vehicle (ROV) once the vessel is filled with heavy water.
operated
Backgrounds
In Table 3 are listed some of the decay properties light is the principal
means of quantifying
of light by other activities
components,
counting,
predictions
that
the photodisintegration
techniques
alpha counting)
quantitative
other backgrounds
isotopes.
Since detection
background
of Cerenkov
in situ, the production
is also important.
By mesns of a suite of radioassay direct gamma
of relevant
activation
applied to samples
identified
analysis, radiochemical
and to the complete
can be made concerning
will accompany
events can be unambiguously
(neutron
methods,
inventory
of smaller
the levels of photodiiintegration
the NCD array in SNO. About
half of all neutron
via track length ZISenergy as being distinct
and capture
from alpha particles
(for this reason the array efficiency is taken to be 25% rather than the nominal 45% capture efficiency). Electron
and Compton
backgrounds,
from neutron
events.
determination
and subtraction.
3
As a result,
and microdischarge photodisintegration
events, have topologies is the only background
separated
still further
requiring
a separate
Conclusions
In 1997, radioassay
indicates
that
approximately
130 neutrons
volume of the vessel by the NCD array, and a further components.
Initial
photodisintegration
56Co activity
produces
rate once the array
about
is deployed
per year will be produced
in the main
300 near the wall by cables and termination
200 per year. depends
The determination
primarily
on observing
of the actual the Cerenkov
Prog. Purr. Nd. Phys.40 (IYYN)113-121
R. G. H. Rohrrtsotll
spectrum water.
of the gammas
responsible,
and is expected
If the levels are as anticipated controlled
production
rate will be less than 8% at 6000 neutrons
flavor content
and independently
Neutrino
assayed,
Observatory
of the 8B neutrino
oscillations
after one year’s operation
the uncertainty
suited to make a determination
flux from the Sun, essentially data.
in the water can be in the NC
per year.
is uniquely
that flux within the range allowed by current of neutrino
by Th and U in the heavy
(5 5 fg/g of Th, 5 10 fg/g U), and “‘Rn
sdequately
The Sudbury
to be dominated
121
independent
of the neutrino
of the actual magnitude
By this means a model-independent
can be made, if they are the explanation
for the solar neutrino
of
determination problem.
References PI G. Aardsma
et al., “A Heavy Water Detector
to Resolve the Solar Neutrino
Problem”,
Phys. Lett.
194 (1987) 321.
PIT.J.
Bowles,
P.J. Doe, A. Hime, R.G.H.
in Neutral-Current
helmy and J.F. Wilkerson, Los Alamos National
131 R.G.H. Robertson Moriond Workshop,
Laboratory,
P.M. Thornewell,
Detection in the Sudbuy
FIN-94ER-E324,
Villars-sur-Ollon,
Solar Neutrino
Switzerland,
Detectors
Conf., Heidelberg,
[51 Y. Suzuki, these proceedings J.N. 161
T.C. Spencer,
January
J.B. Wil-
Neutrino Observatory,
31 (1992).
et al., in Perspectives in Neutrinos, Atomic Physics, and Gravitation, XIIIth
141 A. Hime et al., Neutral-Current ternational
Robertson,
January
for the Sudbury Germany,
30 - February Neutrino
6 (1993).
Observatory,
Proc. 4th In-
April 8-11, 1997 (to be published).
(1997).
Bahcall and M.H. Pinsonneault,
Revs. Mod. Phys. 67 (1995) 781.
171 J.N. Bahcall and E. L&i, Phys. Rev. D54 (1996) 5417. PI K. Kubodera
and S. Nozawa, Irk. J. Mod. Phys. E 3 (1994) 101.
PI E.L. Pasierb,
Ph.D. Dissertation,
WI 1.1. Gurevich
WI G.S.
Vidyakin
WI J.B. Wilhelmy,
University
et al., I.V. Kurchatov et al., JETP
1141G.G. Miller, unpublished
Irvine, 1979.
of Atomic Energy Preprint
IAE4986/2
Lett. 59 (1994) 364.
Los Alamos
P31 M.M. Lowry, unpublished
Institute
of California,
National (1992).
(1995).
Laboratory
preprint
LA-UR-92-2447
(1992).
(1989).