Professional Documents
Culture Documents
J. S. Nico
National Institute of Standards and Technology, Stop 8461, Gaithersburg, Maryland 20899, USA
arXiv:0901.2200v3 [nucl-ex] 10 Aug 2009
which agrees well with the pp flux predicted by the standard solar model. Finally, we make several tests and
show that the data are consistent with the assumption that the solar neutrino production rate is constant in time.
I. INTRODUCTION was presented in Ref. [2]. Part II of this series, although not
called by this name, described the changes to the experiment
and gave the results for the period January 1998 to December
The SAGE experiment was built to measure the capture rate 2001 [3]. In Secs. II and III of the present article we do the
of solar neutrinos by the reaction 71 Ga + νe → 71 Ge + e− and same for the six-year period January 2002 to December 2007.
thus to provide information to aid in understanding the deficit We then discuss the four neutrino source experiments with Ga
of neutrinos observed in the 37 Cl experiment [1], in which in Sec. IV, give the present interpretation of the SAGE results
only about one-third of the solar neutrino capture rate pre- in Sec. V, derive the contemporary value of the neutrino flux
dicted by the standard solar model was detected. The feature produced by the proton-proton fusion reaction in Sec. VI, and
that distinguishes the Ga experiment from all other past or present a brief consideration of the question of possible time
present solar neutrino detectors is its sensitivity to the proton- variation in the data in Sec. VII.
proton fusion reaction, p + p → d + e+ + νe , which generates
most of the Sun’s energy. Ga experiments have provided the
only direct measurement of the current rate of this reaction.
A full description of the SAGE experiment and the results In addition to SAGE, there also existed a second Ga solar
of each measurement from its inception to December 1997 neutrino experiment called Gallex. It contained 30 tons of
gallium in a solution of GaCl3 and measured the solar neutrino
capture rate from 1991 to 1997. In 1998 this experiment was
reconstituted under the name of GNO and it took data until
∗ Corresponding
author. Present address: SNOLAB, PO Box 159, Lively, 2003. We give the results of these experiments and combine
Ontario P3Y 1M3, Canada; bclevela@snolab.ca. them with the SAGE data in Sec. III.
SAGE results, Part III 2
II. EXPERIMENTAL PROCEDURES for all but two extractions since that time. In contrast to the
usual counters with a solid cathode, the cathode of the new
A. Overview counters is made from vapor-deposited carbon, thus eliminat-
ing the usual dead volume behind the cathode. The dead vol-
The SAGE experiment is in a dedicated deep-underground ume is further reduced and end effects are nearly eliminated
laboratory excavated into the side of Mt. Andyrchi in the by curving inwards the regions of the counter where the cath-
northern Caucasus mountains of Russia. The rock overburden ode ends. The cathode and anode leads are sealed into the
is equivalent to 4700 m of water and the measured muon flux counter body with Mo ribbon that makes the counter leak free
at the location of the experiment is (3.03±0.10)×10−9 /(cm2 s). and ensures excellent gain stability. The cathode is so thin that
The mass of gallium used in SAGE at the present time is the counter body is transparent, making it possible to visually
50 tonnes. It is contained in seven chemical reactors which are inspect all the internal counter parts.
heated to 30 ◦ C so the gallium metal remains molten. A mea- During 2004–2005 an extensive series of measurements of
surement of the solar neutrino capture rate begins by adding the efficiency of these new counters was made. The methods
to the gallium a stable Ge carrier. The carrier is a Ga-Ge alloy of measurement were described in Ref. [2] and counter fill-
with a known Ge content of approximately 350 µg and is dis- ings of 69 Ge, 71 Ge, and 37 Ar were used. The measured volume
tributed equally among all reactors. The reactor contents are efficiency of the new counters was 96% with a spread in effi-
stirred thoroughly to disperse the Ge throughout the Ga mass. ciency of only ±1% for all counters of this type. This should
After a typical exposure interval of one month, the Ge carrier be compared with an average volume efficiency of 89% for
and 71 Ge atoms produced by solar neutrinos and background our original counter design. Further, the fraction of events that
sources are chemically extracted from the Ga. The final step is degraded in energy was found to be significantly less than
of the chemical procedure is the synthesis of germane (GeH4 ), in the old design. These decreases in degraded fraction com-
which is used as a proportional counter fill gas with an admix- bined with the increase in volume efficiency lead to a quite
ture of (90–95)% Xe. The total efficiency of extraction is the dramatic increase in efficiency for these new counters com-
ratio of mass of Ge in the germane to the mass of initial Ge pared to the old type, approximately 25% in the K peak and
carrier and is typically (95 ± 3)%. 10% in the L peak.
Another innovation in the new counter design is that the
Suprasil counter body is etched in hydrofluoric acid to a thick-
B. Extraction of Ge from Ga
ness of ∼0.2 mm. This permits calibration of the counter with
our standard 55 Fe source over nearly its entire volume. As
The extraction procedures from 1990 to 1997 are described an undesired side effect, however, the thin body, combined
in Ref. [2]. At the beginning of 1998 some minor modifica- with the very thin cathode, makes these counters sensitive to
tions were made as described in Ref. [4]. low-energy x rays from local radioactivity. To eliminate this
Beginning with the December 2005 extraction, the carrier response, a graded shield consisting of an outer layer of 1 mm
used to measure the extraction efficiency was isotopically en- of Cu and an inner layer of 3 mm of low-background acrylic
riched in either 72 Ge or 76 Ge. At the end of each extrac- (to absorb Cu x rays) is placed over the counter body during
tion a sample was taken from the final extraction solution and measurement with 71 GeH4 .
this sample was analyzed with an inductively-coupled plasma
mass spectrometer to determine the fractional content of the The pulses from the proportional counter are sent to a fast
various Ge isotopes. The efficiency of Ge extraction from the transient analyzer where they are digitized for 800 ns after
Ga metal was then calculated using the method outlined in pulse onset at two different gains, one chosen for the L peak
Appendix A. This procedure for determining the extraction and the other for the K peak. The transient digitizer serves
efficiency has the advantage that it gives a direct measure of to differentiate fast-rising 71 Ge pulses from generally slower-
any Ge that may enter the sample from unknown sources. rising background pulses. This can be seen by comparing
the upper and lower panels of Fig. 1, which show the pulses
from the 77 extractions that have been measured in the new
C. Counting of 71 Ge proportional counters. The upper panel is for all events that
pass the time cuts for Rn (see Sec. II D), are not high-voltage
71
Ge decays to 71 Ga by pure electron capture with a half breakdown, do not have a NaI coincidence, and occur during
life of 11.4 days. Two peaks are observed in the proportional the first 30 days of counting. The total live time is 1999.8
counter–the K peak at 10.4 keV and the L peak at 1.2 keV. The days and there are 2063 events. The lower panel of this fig-
counter containing the GeH4 from the extraction is placed in ure shows the 1545 events that occurred between days 100.0–
the well of a NaI detector that is within a large passive shield 130.1 (the same live time duration as in the upper panel). The
and is counted for a typical period of 6 months. To reduce the fast-rising 71 Ge events in the L and K peaks are evident in the
influence of 222 Rn, the volume inside the shield around the upper panel but missing in the lower panel because the 71 Ge
counters is purged with boil-off gas from a dewar filled with has decayed away.
liquid nitrogen. Aside from replacing some modules that failed, no changes
A completely redesigned proportional counter [5] began to were made to the counting system electronics since their de-
be used with the extraction of April 2001 and has been used scription in Ref. [2].
SAGE results, Part III 3
πσ
√
r !
2 E −C
F(E) = he−[(E−C)/( 2σ)] + hd erfc √ , (1)
2C 2σ
where h, C, and σ are the peak height, center, and width
and d is a parameter related to the fraction of degraded
FIG. 1: Upper panel: Count rate vs energy and rise time for events events. The error function term here is the integral of
during the first 30 days of counting. Regions where the L and K the Gaussian from energy E to ∞; it is essentially flat
peaks are predicted to occur based on 55 Fe calibrations are shown below the peak, monotonically decreases in the peak re-
darkened. There are 427 counts in the L-peak region and 287 counts gion, and is zero above the peak. This new line shape
in the K-peak region. The counts in both regions are a combination only makes a very small change to the counting effi-
of events from 71 Ge decay and background. Lower panel: Equivalent ciency in the L peak for a few runs whose energy win-
graph for all events that occurred during an equal live time interval dow width is obliged to be less than 2 full widths at half
beginning at day 100 after extraction. There are 226 counts in the
maximum.
L-peak region and 94 counts in the K-peak region.
• For all runs after August 1992 the likelihood function
was modified to include a factor that weights each event
D. Data analysis according to its measured energy. This requires knowl-
edge of the energy distribution for 71 Ge pulses and for
Based on criteria described in [2], a group of events is se- background events, both of which can be determined
lected from each extraction that are candidate 71 Ge decays. from the long duration of counting data that we have
These events are fit to a maximum likelihood function [6], accumulated. When this method is applied, it is found
assuming that they originate from an unknown but constant- that the overall statistical uncertainty decreases by (0.1-
rate background and the exponentially-decaying rate of 71 Ge. 0.2) SNU, but the systematic uncertainty increases by
Because only a few 71 Ge counts are detected from each ex- ∼0.1 SNU.
traction, a single run result has a large statistical uncertainty
and thus little significance. These changes in analysis methods have been applied to all
Several minor changes in the methods of analysis have oc- data.
curred since Part II of this series. These include the following:
140
120
Capture rate (SNU) 100
80
60
40
20
0
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Year
FIG. 2: Combined SAGE results for each year. Shaded band is the combined best fit and its uncertainty for all years. Vertical error bars are
statistical with 68% confidence.
1.0
ability is small, but still quite possible.
R i = φ♁ ♁
Z ∞
R= σ(E)Φ♁ (E)dE, (3) i hσi i, (10)
Ethreshold
where
where σ(E) is the cross section of the neutrino-capture reac-
tion and Φ♁ (E) is the total flux of electron neutrinos at the hσ
i i
hσ♁
i i= ee . (11)
Earth, which can be expressed as hPi i
φ♁ ♁
X
Φ♁ (E) = i S i (E). (4) In Table III we give values of hPee ♁
i i and hσi i for
i each neutrino component. These were calculated assum-
In this expression the index i refers to the various nuclear re- ing three-neutrino mixing to active neutrinos with parame-
actions in the Sun that produce neutrinos (pp, 7 Be, pep, 13 N, ters from Ref. [25]: ∆m212 = (7.65+0.23 −0.20 ) × 10
−5
eV2 , θ12 =
+1.36 +3.5
15
O, 17 F, 8 B, and hep), φ♁ 33.46−1.00 degrees, and θ13 = 5.7−5.7 degrees. The approxi-
i is the amplitude of flux component
♁ mate formulae given in Ref. [26] were used for the survival
i at the Earth, and S (E) is the spectrum of the ith neutrino probability Pee
i i (E). As we show in Sec. VII there is no ap-
component at the Earth, each of which is normalized such that preciable difference between the day and night capture rates
R∞
0
S i♁ (E)dE = 1. The neutrino spectrum at the Earth is re- in Ga and thus regeneration in the Earth was neglected. The
lated to the spectrum produced in the Sun S i (E) by cross sections σ(E) were taken from Appendix C for Ga
and Ref. [27] for Cl. The neutrino spectra φ i (E) are from
S i♁ (E) = Ai S i (E)Pee
i (E), (5) Refs. [19] (pp, 13 N, 15 O, 17 F), [27] (8 B), and [28] (hep).
Now that all the terms have been calculated, we can use
where Ai is a constant of normalization and Pee i (E) is the prob- the fluxes in Table II, combined with Eqs. (8) and (11), to
ability that an electron neutrino produced in the Sun by reac- predict the capture rate in Ga from each of the solar neutrino
tion i with energy E will reach the Earth without a change of components. The individual rates and the total rate are given
flavor, commonly called the survival factor. The physical ori- in Table IV for two recent solar models from Table II. For
gin for the reduction of the electron component of the solar both models there is good agreement between the calculated
neutrino flux is the now well-established mechanism of MSW total rate and the observed capture rate of 66.1 ± 3.1 SNU.
neutrino oscillations [24]. Pee i (E) is different for each flux The major contribution to the uncertainty in the predicted total
SAGE results, Part III 7
TABLE III: Factors needed to compute the solar neutrino capture rate in 71 Ga and 37 Cl solar neutrino experiments. The uncertainty values are
at 68% confidence. The parameter hσ♁
i i is defined in Eq. (11).
TABLE IV: Capture rates Ri for Ga experiments calculated with fluxes from Ref. [21].
With GS98 composition With AGS05 composition
Spect. Cap. rate Percent uncertainty in rate due to Total unc. Cap. rate Percent uncertainty in rate due to Total unc.
comp. (SNU) φ σ ∆m212 θ12 θ13 in rate (%) (SNU) φ σ ∆m212 θ12 θ13 in rate (%)
pp 39.35 + 0.6,- 0.6 + 2.4,- 2.3 + 0.0,- 0.0 + 2.2,- 2.8 + 2.0,- 3.1 + 3.9,- 4.8 39.81 + 0.5,- 0.5 + 2.4,- 2.3 + 0.0,- 0.0 + 2.2,- 2.8 + 2.0,- 3.1 + 3.9,- 4.8
pep 1.43 + 1.1,- 1.1 +17.0,- 2.4 + 0.3,- 0.3 + 1.9,- 2.3 + 1.9,- 2.9 +17.2,- 4.6 1.47 + 1.0,- 1.0 +17.0,- 2.4 + 0.3,- 0.3 + 1.9,- 2.3 + 1.9,- 2.9 +17.2,- 4.5
7
Be 18.73 + 6.0,- 6.0 + 7.0,- 2.3 + 0.1,- 0.1 + 2.1,- 2.6 + 2.0,- 3.0 + 9.7,- 7.5 16.81 + 6.0,- 6.0 + 7.0,- 2.3 + 0.1,- 0.1 + 2.1,- 2.6 + 2.0,- 3.0 + 9.7,- 7.5
13
N 0.89 +15.0,-15.0 + 9.8,- 2.3 + 0.1,- 0.1 + 2.1,- 2.6 + 2.0,- 3.0 +18.1,-15.7 0.58 +14.0,-13.0 + 9.8,- 2.3 + 0.1,- 0.1 + 2.1,- 2.6 + 2.0,- 3.0 +17.3,-13.8
15
O 1.23 +17.0,-16.0 +12.9,- 2.3 + 0.2,- 0.2 + 2.0,- 2.4 + 1.9,- 3.0 +21.5,-16.6 0.77 +16.0,-15.0 +12.9,- 2.3 + 0.2,- 0.2 + 2.0,- 2.4 + 1.9,- 3.0 +20.7,-15.6
17
F 0.03 +19.0,-17.0 +12.9,- 2.3 + 0.2,- 0.2 + 2.0,- 2.4 + 1.9,- 3.0 +23.1,-17.6 0.02 +16.0,-15.0 +12.9,- 2.3 + 0.2,- 0.2 + 2.0,- 2.4 + 1.9,- 3.0 +20.8,-15.6
8
B 4.64 +11.0,-11.0 +31.8,-14.4 + 0.5,- 0.4 + 5.4,- 3.9 + 1.8,- 2.8 +34.1,-18.7 3.68 +11.0,-11.0 +31.8,-14.4 + 0.5,- 0.4 + 5.4,- 3.9 + 1.8,- 2.8 +34.1,-18.7
hep 0.02 +15.0,-15.0 +32.7,-15.4 + 0.3,- 0.3 + 6.2,- 4.5 + 1.9,- 2.9 +36.5,-22.2 0.02 +15.0,-15.0 +32.7,-15.4 + 0.3,- 0.3 + 6.2,- 4.5 + 1.9,- 2.9 +36.5,-22.2
Total 66.31 + 1.9,- 1.9 + 3.3,- 1.8 + 0.1,- 0.1 + 1.5,- 1.8 + 1.3,- 2.0 + 4.3,- 3.8 63.16 + 1.8,- 1.8 + 3.1,- 1.8 + 0.1,- 0.0 + 1.5,- 1.9 + 1.4,- 2.1 + 4.1,- 3.8
37
rate is from the neutrino capture cross section, with smaller Cl solar neutrino experiment, the total calculated rate
contributions from the solar model flux, θ12 , and θ13 . is 3.09(1+0.094
−0.091 ) SNU using fluxes based on GS98 and
+0.091
In this analysis we have used the cross sections in Ap- 2.53(1−0.089 ) SNU using fluxes based on AGS05. These
pendix C, in which the contribution of the two lowest-lying should be compared with the experimental rate of 2.56 ±
excited states in 71 Ge has been set to zero. If instead we use 0.16 (stat) ± 0.16 (syst) SNU [1].
the original Bahcall cross sections, then the total rate increases
by 1.2 SNU with the GS98 composition and by 1.1 SNU with
the AGS05 composition. Whatever cross sections are assumed VI. THE pp NEUTRINO FLUX FROM THE SUN
is thus not a significant factor in the interpretration of the total
rate in the Ga experiment. The cross sections are, however, of In this section we will use the Ga measurement given in
vital importance in understanding the origin of the unexpect- Eq. (2) and the results of other solar neutrino experiments to
edly low result in the source experiments. determine the pp flux from the Sun. The conventional way to
The attentive reader may be concerned that there is a log- make this calculation is by a combined fit to all experiments,
ical inconsistency in the argument presented here: the pre- as for example is presented in Ref. [29] and [30]. Here we
dicted capture rates we derive for the Ga experiment depend give an alternate approach that successively decomposes the
on the neutrino oscillation parameters, but the measured total total measured rate into the components from each neutrino
rate of the Ga experiment is itself one of the inputs used to de- source. The final result is identical to what one obtains in a
termine the oscillation parameters. Although this is true in a combined fit and has the advantage that the argument is simple
strict sense, the neutrino oscillation parameters derived from and transparent.
a global fit of all experiments are for all practical purposes The rate in Eq. (2) is the sum of the rates from all the com-
independent of the rate in the Ga experiment. Rather, the pa- ponents of the solar neutrino flux, which we denote by
rameter θ12 is principally determined by the SNO experiment,
[pp+7 Be+CNO+pep+8 B|Ga] = 66.1(1 ± 0.047) SNU. (12)
θ13 by the CHOOZ experiment, and ∆m212 by the KamLAND
experiment. Although it was not true in the past, the result of We ignore the tiny hep contribution and combine the 13 N, 15 O,
the Ga experiment is at present only a very minor input to the and 17 F components into a single value, called here “CNO”.
determination of these parameters. In an experiment of great technical difficulty, the 7 Be flux
Incidentally, if we carry out the same analysis for the has been directly measured by Borexino and they report the
SAGE results, Part III 8
result as φ
7 Be = 5.18(1 ± 0.098) × 10 neutrinos/(cm s) [30].
9 2
years [3, 32, 33], with the major change being a gradual reduc-
Using Eqs. (8) and (11), we multiply this flux by the electron tion of the uncertainty. In the future, as Borexino continues to
neutrino survival factor for 7 Be and by the cross section of collect data, and as direct measurements are made of the CNO
7
Be on Ga (the values of these factors and their uncertainties and pep fluxes, the uncertainty in this flux should be further
are given in Table III) and obtain the rate of 7 Be in Ga of reduced and eventually may be dominated by the uncertainty
in the Ga rate itself. By that time, however, there will hope-
[7 Be|Ga] = 19.1(1+0.12
−0.11 ) SNU. (13) fully be direct experiments that measure the pp flux in real
time.
The 8 B flux at the Earth has been directly measured by SNO
For comparison, we see from Table II that the predicted pp
to be φ♁
8 B = (1.67±0.05)×10 electron neutrinos/(cm s) [31].
6 2
flux from the two recent solar models with different compo-
7
In a similar way to Be, we multiply this flux by the spectrum- sition is φ
pp = 5.97 ± 0.04 and 6.04 ± 0.03, both in units of
integrated cross section for 8 B neutrinos on Ga and obtain the 10 νe /(cm2 s). There is good numerical agreement between
10
8
B contribution to the Ga experiment of these flux values and the result in Eq. (19), but, as made clear
by Bahcall and Peña-Garay [34], there is a large difference in
[8 B|Ga] = 3.6(1+0.32
−0.16 ) SNU. (14)
interpretation: the result in Eq. (19) was derived from contem-
Subtracting these measured rates of 7 Be and 8 B from the porary solar neutrino experiments and is the pp neutrino flux
total Ga rate in Eq. (12) gives at the present time. In contrast, energy generation in the solar
model is highly constrained by the measured solar luminosity
[pp+CNO+pep|Ga] = 43.3(1+0.087
−0.094 ) SNU. (15) and thus, when the luminosity constraint is imposed, as is the
case for the models given in Table II, the calculated pp flux
We can obtain an approximate value for the contribution is what the Sun was producing some 40 000 years ago. The
of CNO and pep to the Ga experiment from the measured agreement between the present pp flux, as measured by the Ga
capture rate in the Cl experiment [7 Be+CNO+pep+8 B|Cl] = experiment, and the past flux, as inferred from the solar model
2.56(1 ± 0.088) SNU [1]. As in the case of Ga, we use the with the luminosity constraint, implies that the pp flux from
7
Be flux measured by Borexino, the 8 B flux measured by the Sun has not altered (within our 14% uncertainty) during
SNO, and the cross sections in Table III to determine [7 Be|Cl] the last 40 000 years.
= 0.67(1+0.105 +0.068
−0.108 ) SNU and [ B|Cl] = 1.73(1−0.067 ) SNU. We
8
subtract these values from the total Cl rate and are left with
[CNO+pep|Cl] = 0.19(1+1.36−1.00 ) SNU. VII. CONSIDERATION OF TIME VARIATION
If we attribute this entire rate to the neutrinos from pep
then, using the cross sections for pep on Cl and Ga, we cal-
In a plot of the SAGE results as a function of time there is
culate a rate of [pep|Ga]test = 2.35(1+1.37
−1.00 ) SNU. On the other a slight visual hint of a long-term decrease, as illustrated in
hand, if we attribute this entire rate to CNO, we obtain in the
Fig. 2. The average rate prior to 1996 is somewhat higher
same manner a rate of [CNO|Ga]test = 3.11(1+1.37 −1.00 ) SNU. The than after 1996. A plot of the Gallex-GNO data shows a
upper extreme of these two test rates is 3.11 × (1 + 1.37) =
similar behavior [10]. When examined quantitatively, how-
7.37 SNU. As a reasonable estimate we can thus set the sum
ever, the evidence for a long-term decrease in the capture rate
of CNO and pep rates at half this upper limit with an uncer-
is unconvincing. A χ2 test applied to these yearly SAGE
tainty of 100%:
data points assuming the rate is constant at 65.4 SNU gives
[CNO+pep|Ga] = 3.68(1+1.00 χ2 /DOF = 12.0/17, which has a probability of 80%. The fit to
−1.00 ) SNU. (16)
a constant rate is thus quite good.
We subtract this estimate for the CNO plus pep rate from In previous articles we have demonstrated the agreement
the rate in Eq. (15) and obtain the result for the measured pp between the assumption of a constant production rate and the
rate in the Ga experiment SAGE measurements by use of the cumulative distribution
of the capture rate C(p), defined as the fraction of data sets
[pp|Ga] = 39.7(1+0.13
−0,14 ) SNU. (17) whose capture rate is less than p. Figure 4 shows this distri-
bution for the data and the expected distribution derived from
Dividing this capture rate by the cross section for capture of
100 simulations of all 168 runs, where it is assumed in the sim-
pp neutrinos from Table III gives the measured electron neu-
ulations that the production rate is constant and has a value of
trino pp flux at Earth of
65.4 SNU. For each run the rates from the separate L and K
peaks are used in this figure, not the rate from the L+K combi-
φ♁pp = 3.38(1+0.14
−0.14 ) × 10 /(cm s).
10 2
(18)
nation. To ensure that the simulations parallel the real data as
+0.030 closely as possible, all parameters of the simulation, such as
i i = 0.561(1−0.042 ) from
If we use Eq. (7) and the value of hPee
Table III then the pp flux produced in the Sun is background rates, efficiencies, exposure times, and counting
times, were chosen to be the same as for the real data. Only
φ
pp = 6.0(1 ± 0.14) × 10 /(cm s).
10 2
(19) the number of counts in each run and the times when these
counts occurred were allowed to vary.
Our present result for the pp flux is in good agreement with The data spectrum and the simulated spectrum are very
the previous estimates that we have made during the past six similar to each other, indicating that the distribution of cap-
SAGE results, Part III 9
1.0
Cumulative distribution function
10 2
0.9
0.8
0.7
0.6
10 1
0.4
0.3
0.2
0.1
10 0
0.0
0 50 100 150 200 250
Capture rate (SNU)
0.0 2.0 4.0 6.0
FIG. 4: Measured capture rate for all 310 SAGE data sets (jagged Power
curve) and the expected distribution derived by 100 Monte Carlo sim-
ulations of each set (smooth curve). FIG. 6: Histogram of powers in spectrum of Fig. 5. The bin size is
0.07 power units. The solid line is the expected distribution if there
is no time variation, i.e., the number of frequencies × exp(−power),
6.0
integrated over the limits of each bin.
dent of the Russian Federation under grants 00-1596632, NS- are to be determined. For N extractions there are thus 2N
1782.2003, NS-5573.2006.2 and NS-959.2008.2, by the Pro- variables (the extraction efficiency and mass of extra natu-
gram of Basic Research “Neutrino Physics” of the Presidium ral Ge for each extraction) and 5N equations that relate these
of the Russian Academy of Sciences, by the International variables, one for each of the naturally-occurring Ge isotopes,
70
Science and Technology Center under grant 1431, and by Ge, 72 Ge, 73 Ge, 74 Ge, and 76 Ge. Because there are more re-
the US Civilian Research and Development Foundation under lationships than unknowns, the problem is solved by finding
grants CGP RP2-159 and CGP RP2-2360-MO-02. the set of variables that minimizes the function
5
N X
X Mnp (i) − M e (i) 2
APPENDIX A: CALCULATION OF EXTRACTION χ2 = n
, (A2)
n=1 i=1
σn (i)
EFFICIENCY FROM ISOTOPIC ANALYSIS
where Mne (i) is the measured mass of isotope i in extraction
It is assumed the extracted Ge consists of a combination of n and σn (i) is the total uncertainty in the knowledge of the
• Ge from the carrier added for the current extraction, predicted and measured masses in extraction n.
[1] B. T. Cleveland, T. J. Daily, R. Davis, Jr., J. R. Distel, K. Lande, Jr., K. Lande, P. S. Wildenhain, S. R. Elliott, and J. F. Wilk-
C. K. Lee, and P. S. Wildenhain, Astrophys. J. 496, 505 (1998). erson, Phys. Rev. C 60, 055801 (1999) [arXiv.org/abs/astro-
[2] J. N. Abdurashitov, V. N. Gavrin, S. V. Girin, V. V. Gorbachev, ph/9907113].
T. V. Ibragimova, A. V. Kalikhov, N. G. Khairnasov, T. V. Kn- [3] J. N. Abdurashitov, V. N. Gavrin, S. V. Girin, V. V. Gor-
odel, I. N. Mirmov, A. A. Shikhin, E. P. Veretenkin, V. M. Ver- bachev, P. P. Gurkina, T. V. Ibragimova, A. V. Kalikhov, N.
mul, V. E. Yants, G. T. Zatsepin, T. J. Bowles, W. A. Teasdale, G. Khairnasov, T. V. Knodel, I. N. Mirmov, A. A. Shikhin,
D. L. Wark, M. L. Cherry, J. S. Nico, B. T. Cleveland, R. Davis, E. P. Veretenkin, V. M. Vermul, V. E. Yants, and G. T. Zat-
SAGE results, Part III 15
TABLE VI: Approximate cross section for neutrino capture by 71 Ga if the contribution of the first two excited states is set to zero.
ν energy Cross section (10−46 cm2 ) ν energy Cross section (10−46 cm2 ) ν energy Cross section (10−46 cm2 )
(MeV) Best −1σ +1σ (MeV) Best −1σ +1σ (MeV) Best −1σ +1σ
0.240 1.310 × 101 1.280 × 101 1.340 × 101 1.445 1.944 × 102 1.897 × 102 2.274 × 102 9.500 4.749 × 104 4.053 × 104 6.275 × 104
0.250 1.357 × 101 1.326 × 101 1.388 × 101 1.500 2.153 × 102 2.104 × 102 2.495 × 102 10.000 5.653 × 104 4.820 × 104 7.474 × 104
0.275 1.499 × 101 1.465 × 101 1.533 × 101 1.600 2.451 × 102 2.395 × 102 2.859 × 102 10.500 6.638 × 104 5.650 × 104 8.785 × 104
0.300 1.662 × 101 1.624 × 101 1.700 × 101 1.700 2.771 × 102 2.707 × 102 3.252 × 102 11.000 7.703 × 104 6.548 × 104 1.020 × 105
0.325 1.836 × 101 1.794 × 101 1.878 × 101 1.750 2.939 × 102 2.871 × 102 3.458 × 102 11.500 8.848 × 104 7.510 × 104 1.173 × 105
0.350 2.018 × 101 1.972 × 101 2.064 × 101 2.000 3.932 × 102 3.702 × 102 4.712 × 102 12.000 1.007 × 105 8.535 × 104 1.336 × 105
0.375 2.208 × 101 2.157 × 101 2.259 × 101 2.500 6.428 × 102 5.986 × 102 7.826 × 102 12.500 1.137 × 105 9.621 × 104 1.508 × 105
0.400 2.406 × 101 2.351 × 101 2.461 × 101 3.000 9.806 × 102 9.043 × 102 1.211 × 103 13.000 1.274 × 105 1.077 × 105 1.692 × 105
0.425 2.606 × 101 2.546 × 101 2.799 × 101 3.500 1.449 × 103 1.323 × 103 1.811 × 103 13.500 1.418 × 105 1.197 × 105 1.884 × 105
0.450 2.810 × 101 2.746 × 101 3.029 × 101 4.000 2.108 × 103 1.905 × 103 2.662 × 103 14.000 1.569 × 105 1.322 × 105 2.086 × 105
0.500 3.231 × 101 3.157 × 101 3.516 × 101 4.500 3.043 × 103 2.722 × 103 3.880 × 103 14.500 1.728 × 105 1.454 × 105 2.298 × 105
0.600 4.109 × 101 4.015 × 101 4.585 × 101 5.000 4.336 × 103 3.842 × 103 5.573 × 103 15.000 1.893 × 105 1.590 × 105 2.520 × 105
0.700 5.027 × 101 4.912 × 101 5.776 × 101 5.500 6.072 × 103 5.337 × 103 7.853 × 103 15.500 2.064 × 105 1.731 × 105 2.749 × 105
0.800 6.478 × 101 6.329 × 101 6.924 × 101 6.000 8.350 × 103 7.290 × 103 1.086 × 104 16.000 2.241 × 105 1.875 × 105 2.988 × 105
0.900 7.829 × 101 7.649 × 101 8.470 × 101 6.500 1.133 × 104 9.832 × 103 1.479 × 104 18.000 3.010 × 105 2.495 × 105 4.025 × 105
1.000 9.299 × 101 9.085 × 101 1.017 × 102 7.000 1.515 × 104 1.309 × 104 1.985 × 104 20.000 3.860 × 105 3.162 × 105 5.184 × 105
1.100 1.168 × 102 1.142 × 102 1.306 × 102 7.500 1.989 × 104 1.712 × 104 2.613 × 104 22.500 5.013 × 105 4.028 × 105 6.778 × 105
1.200 1.372 × 102 1.341 × 102 1.549 × 102 8.000 2.550 × 104 2.188 × 104 3.357 × 104 25.000 6.233 × 105 4.883 × 105 8.501 × 105
1.300 1.593 × 102 1.557 × 102 1.814 × 102 8.500 3.198 × 104 2.737 × 104 4.216 × 104 30.000 8.701 × 105 6.354 × 105 1.217 × 106
1.400 1.831 × 102 1.789 × 102 2.100 × 102 9.000 3.928 × 104 3.357 × 104 5.186 × 104
sepin, T. J. Bowles and W. A. Teasdale, J. S. Nico, B. T. Lett. B 616, 174 (2005) [arXiv.org/abs/hep-ex/0504037].
Cleveland, S. R. Elliott, and J. F. Wilkerson, Zh. Eksp. Teor. [11] J. N. Abdurashitov, V. N. Gavrin, S. V. Girin, V. V. Gorbachev,
Fiz. 122, 211 (2002) [J. Exp. Theor. Phys. 95, 181 (2002)] T. V. Ibragimova, A. V. Kalikhov, N. G. Khairnasov, T. V. Kn-
[arXiv.org/abs/astro-ph/0204245]. odel, V. N. Kornoukhov, I. N. Mirmov, A. A. Shikhin, E. P.
[4] J. N. Abdurashitov, V. N. Gavrin, S. V. Girin, V. V. Gor- Veretenkin, V. M. Vermul, V. E. Yants, G. T. Zatsepin, Yu. S.
bachev, P. P. Gurkina, T. V. Ibragimova, A. V. Kalikhov, N. Khomyakov, A. V. Zvonarev, T. J. Bowles, J. S. Nico, W. A.
G. Khairnasov, T. V. Knodel, V. A. Matveev, I. N. Mirmov, Teasdale, D. L. Wark, M. L. Cherry, V. N. Karaulov, V. L. Lev-
A. A. Shikhin, E. P. Veretenkin, V. M. Vermul, V. E. Yants, itin, V. I. Maev, P. I. Nazarenko, V. S. Shkol’nik, N. V. Skorikov,
G. T. Zatsepin, T. J. Bowles, S. R. Elliott, W. A. Teasdale, B. B. T. Cleveland, T. Daily, R. Davis, Jr., K. Lande, C. K. Lee, P.
T. Cleveland, W. C. Haxton, J. F. Wilkerson, J. S. Nico, A. S. Wildenhain, S. R. Elliott, and J. F. Wilkerson, Phys. Rev. C.
Suzuki, K. Lande, Yu. S. Khomyakov, V. M. Poplavsky, V. 59, 2246 (1999) [arXiv.org/abs/hep-ph/9803418].
V. Popov, O. V. Mishin, A. N. Petrov, B. A. Vasiliev, S. A. [12] W. C. Haxton, Phys. Rev. C 38, 2474 (1988).
Voronov, A. I. Karpenko, V. V. Maltsev, N. N. Oshkanov, A. [13] V. N. Gavrin, A. L. Kochetkov, V. N. Kornoukhov, A. A.
M. Tuchkov, V. I. Barsanov, A. A. Janelidze, A. V. Korenkova, Kosarev, and V. E. Yants, Institute for Nuclear Research of the
N. A. Kotelnikov, S. Yu. Markov, V. V. Selin, Z. N. Shakirov, A. Russian Academy of Sciences Report No. P-777 (1992).
A. Zamyatina, S. B. Zlokazov, Phys. Rev. C 73, 045805 (2006) [14] W. Hampel, G. Heusser, J. Kiko, T. Kirsten, M. Laubenstein, E.
[arXiv.org/abs/nucl-ex/0512041]. Pernicka, W. Rau, U. Rönn, C. Schlosser R. v. Ammon, K. H.
[5] S. Danshin, A, Kopylov, and V. Yants, Nucl. Instrum. Methods Ebert, T. Fritsch, D. Heidt, E. Henrich, L. Stieglitz, F. Weirich,
Phys. Res. A 349, 466 (1994). M. Balata, F. X. Hartmann, M. Sann, E. Bellotti, C. Cattadori,
[6] B. T. Cleveland, Nucl. Instrum. Methods Phys. Res. 214, 451 O. Cremonesi, N. Ferrari, E. Fiorini, L. Zanotti, M. Altmann,
(1983). F. v. Feilitzsch, R. Mößbauer, G. Berthomieu, E. Schatzmann,
[7] J. N. Abdurashitov, T. J. Bowles, C. Cattadori, B. T. Cleveland, I. Carmi, I. Dostrovsky, C. Bacci, P. Belli, R. Bernabei, S.
S. R. Elliott, N. Ferrari, V. N. Gavrin, S. V. Girin, V. V. Gor- d’Angelo, L. Paoluzi, A. Bevilacqua, M. Cribier, I. Gosset, J.
bachev, P. P. Gurkina, W. Hampel, T. V. Ibragimova, F. Kaether, Rich, M. Spiro, C. Tao, D. Vignaud, J. Boger, R. L.Hahn, J. K.
A. V. Kalikhov, N. G. Khairnasov, T. V. Knodel, I. N. Mirmov, Rowley, R. W. Stoenner, and J. Weneser, Phys. Lett. B420, 114
L. Pandola, H. Richter, A. A. Shikhin, W. A. Teasdale, E. P. (1998).
Veretenkin, V. M. Vermul, J. F. Wilkerson, V. E. Yants, and G. T. [15] W. Hampel, J. Handt, G. Heusser, D. Jaether, J. Kiko, T.
Zatsepin, Astropart. Phys. 25, 349 (2006) [arXiv.org/abs/nucl- Kirsten, M. Laubenstein, E. Neder, E. Pernicka, W. Rau, H.
ex/0509031]. Richter, U. Rönn, U. Schwan, M. Wojcik, Y. Zakharov, R.
[8] W. Hampel and L. Remsberg, Phys. Rev. C 31, 666 (1985). v. Ammon, K. H. Ebert, T. Fritsch, D. Heidt, E. Henrich, L.
[9] F. Kaether, Ph. D. thesis, Datenanalyse der Sonnenneu- Stieglitz, F. Weirich, M. Balata, F. X. Hartmann, E. Bellotti,
trinoexperiments Gallex, Heidelberg, 2007 [www.ub.uni- C. Cattadori, O. Cremonesi, N. Ferrari, E. Fiorini, L. Zanotti,
heidelberg.de/archiv/7501/]. M. Altmann, F. v. Feilitzsch, R. Mößbauer, G. Berthomieu,
[10] M. Altmann, M. Balata, P. Belli, E. Bellotti, R. Bernabei, E. E. Schatzmann, I. Carmi, I. Dostrovsky, C. Bacci, P. Belli,
Burkert, C. Cattadori, R. Cerulli, M. Chiarini, M. Cribier, S. R. Bernabei, S. d’Angelo, L. Paoluzi, M. Cribier, J. Rich, M.
d’Angelo, G. Del Re, K. H. Ebert, F. von Feilitzsch, N. Fer- Spiro, C. Tao, D. Vignaud, J. Boger, R. L.Hahn, J. K. Rowley,
rari, W. Hampel, F. X. Hartmann, E. Henrich, G. Heusser, F. R. W. Stoenner, and J. Weneser, Phys. Lett. B436, 158 (1998).
Kaether, J. Kiko, T. Kirsten, T. Lachenmaier, J. Lanfranchi, M. [16] M. A. Acero, C. Giunti, and M. Laveder, Phys. Rev. D 78,
Laubenstein, K. Lützenkirchen, K. Mayer, P. Moegel, D. Motta, 073009 (2008) [arXiv.org/abs/0711.4222v3].
S. Nisi, J. Oehm, L. Pandola, F. Petricca, W. Potzel, H. Richter, [17] Y. Farzan, T. Schwetz, and A. Yu Smirnov, Jour. High Energy
S. Schoenert, M. Wallenius, M. Wojcik, and L. Zanotti, Phys. Phys. 0807, 067 (2008) [arXiv.org/abs/0805.2098].
SAGE results, Part III 16
[18] W. C. Haxton, Phys. Lett. B 431, 110 (1998) [33] V. N. Gavrin and B. T. Cleveland, talk at XXII Int. Conf. on
[arXiv.org/abs/nucl-th/9804011]. Neutrino Physics and Astrophysics, Santa Fe, NM, USA, 13–
[19] J. N. Bahcall, Phys. Rev. C 56, 3391 (1997) [arXiv.org/abs/hep- 19 June 2006 [arXiv.org/abs/nucl-ex/0703012].
ph/9710491]. [34] J. N. Bahcall and C. Peña-Garay, Jour. High Energy Phys. 11,
[20] J. N. Bahcall, A. M. Serenelli, and S. Basu, Astrophys. J. Suppl. 004 (2003) [arXiv.org/abs/hep-ph/0305159].
Ser. 165, 400 (2006) [arXiv.org/abs/astro-ph/0511337]. [35] B. T. Cleveland, Nucl. Instrum. Methods Phys. Res. A 416, 405
[21] C. Peña-Garay and A. M. Serenelli, [arXiv.org/abs/0811.2424]. (1998).
[22] N. Grevesse and A. J. Sauval, Space. Sci. Rev. 85, 161 (1998). [36] W. H. Press, S. A. Teukolsky, W. T. Vetterling, and B.
[23] M. Asplund, N. Grevesse, and A. J. Sauval, in ASP Conf. Ser. P. Flannery, Numerical Recipes in Fortran, Cambridge Uni-
336, Cosmic Abundances as Records of Stellar Evolution and versity Press, Cambridge, England, 1992, Section 13.8
Nucleosynthesis, ed. by T. G. Barnes III and F. N. Bash, Astro. [www.nrbook.com/a/bookfpdf/f13-8.pdf].
Soc. Pac., San Francisco, 2005, pp. 25. [arXiv.org/abs/astro- [37] L. Pandola, Astroparticle Physics 22, 219 (2004)
ph/0410214]. [arXiv.org/abs/hep-ph/0406248].
[24] S. P. Mikheev and A. Yu. Smirnov, Yad. Fiz. 42, 1441 (1985), [38] G. L. Fogli, E. Lisi, D. Montanino, and A. Palazzo, Phys. Rev.
[Sov. J. Nucl. Phys. 42, 913 (1985)]; A. Yu. Smirnov, 10th Int. D 61, 073009 (2000) [arxiv.org/abs/hep-ph/9910387].
Workshop on Neutrino Telescopes, Venice, 11-14 March 2003 [39] D. Krofcheck, E. Sugarbaker, J. Rapaport, D. Wang, J. N. Bah-
[arXiv.org/abs/hep-ph/0305106]. call, R. C. Byrd, C. C. Foster, C. D. Goodman, I. J. Van Heer-
[25] T. Schwetz, M. Tórtola, and J. W. F. Valle, New J. Phys. 10, den, C. Gaarde, J. S. Larsen, D. J. Horen, and T. N. Taddeucci,
113011 (2008) [arXiv.org/abs/0808.2016v2]. Phys. Rev. Lett. 55, 1051 (1985).
[26] V. Barger, D. Marfatia, and K. Whisnant, Physics Letters B 617, [40] H. Ejiri, H. Akimune, Y. Arimoto, I. Daito, H. Fujimura, Y. Fu-
78 (2005) [arXiv.org/abs/hep-ph/0501247]. jita, M. Fujiwara, K. Fushimi, M. B. Greenfield, M. N. Harakeh,
[27] J. N. Bahcall, E. Lisi, D. E. Alburger, L. de Braeckeleer, S. F. Ihara, T. Inomata, K. Ishibashi, J. Jänecke, H. Kohri, S.
J. Freedman, and J. Napolitano, Phys. Rev. C 54, 411 (1996) Nakayama, C. Samanta, A. Tamii, M. Tanaka, H. Toyokawa,
[arXiv.org/abs/nucl-th/9601044]. and M. Yosoi, Phys. Lett. B 433, 257 (1998).
[28] J. N. Bahcall web site www.sns.ias.edu/∼jnb. [41] H. Ejiri. Physics Reports 338, 265 (2000).
[29] M. C. Gonzalez-Garcia and M. Maltoni, Phys. Rept. 469, 1 [42] R. G. T. Zegers, T. Adachi, H. Akimune, S. M. Austin, A.
(2008) [arXiv.org/abs/0704.1800v2]. M. van den Berg, B. A. Brown, Y. Fujita, M. Fujiwara, S.
[30] C. Arpesella, et al. (Borexino collaboration), Phys. Rev. Lett. Gales, C.J. Guess, M. N. Harakeh, H. Hashimoto, K. Hatanaka,
101, 091302 (2008) [arXiv.org/abs/0805.3843]. R. Hayami, G. W. Hitt, M. E. Howard, M. Itoh, T. Kawa-
[31] B. Aharmim, et al. (SNO collaboration), Phys. Rev. Lett. 101, bata, K. Kawase, M. Kinoshita, M. Matsubara, K. Nakanishi,
111301 (2008) [arXiv.org/abs/0806.0989]. S. Nakayama, S. Okumura, T. Ohta, Y. Sakemi, Y. Shimbara,
[32] V. N. Gavrin (SAGE collaboration), Proc. of Eighth Int. Work- Y. Shimizu, C. Scholl, C. Simenel, Y. Tameshige, A. Tamii,
shop on Topics in Astroparticle and Underground Physics, Seat- M. Uchida, T. Yamagata, and M. Yosoi, Phys. Rev. Lett. 99,
tle, WA, USA, 5–9 September 2003, ed. by F. Avignone and W. 202501 (2007) [arXiv.org/abs/0707.2840].
Haxton, Nucl. Phys. B (Proc. Suppl.) 138, 87 (2005).