Radiat Environ Biophys (2007) 46:327–338
DOI 10.1007/s00411-007-0126-z
ORIGINAL PAPER
Neutron-induced 63Ni in copper samples from Hiroshima
and Nagasaki: a comprehensive presentation of results
obtained at the Munich Maier-Leibnitz Laboratory
W. Rühm · K. L. Carroll · S. D. Egbert · T. Faestermann · K. Knie · G. Korschinek ·
R. E. Martinelli · A. A. Marchetti · J. E. McAninch · G. Rugel · T. Straume · A. Wallner ·
C. Wallner · S. Fujita · H. Hasai · M. Hoshi · K. Shizuma
Received: 18 May 2007 / Accepted: 22 July 2007 / Published online: 8 September 2007
© Springer-Verlag 2007
Abstract Those inhabitants of Hiroshima and Nagasaki
who were aVected by the A-bomb explosions, were
exposed to a mixed neutron and gamma radiation Weld.
Few years later about 120,000 survivors of both cities were
selected, and since then radiation-induced late eVects such
as leukemia and solid tumors are being investigated in this
cohort. When the present study was initiated, the fast neutron Xuences that caused the neutron doses of these survivors had never been determined experimentally. In
principle, this would have been possible if radioisotopes
produced by fast neutrons from the A-bomb explosions
had been detected in samples from Hiroshima and Nagasaki at distances where the inhabitants survived. However,
no suitable radioisotope had so far been identiWed. As a
contribution to a large international eVort to re-evaluate
the A-bomb dosimetry, the concentration of the radionuclide 63Ni (half-life 100.1 years) has been measured in
copper samples from Hiroshima and Nagasaki. These measurements were mainly performed at the Maier-LeibnitzLaboratory in Munich, Germany, by means of accelerator
mass spectrometry. Because the 63Ni had been produced in
these samples by fast A-bomb neutrons via the reaction
63
Cu(n,p)63Ni, these measurements allow direct experimental validation of calculated neutron doses to the members of the LSS cohort, for the Wrst time. The results of
these eVorts have already been published in a compact
form. A more detailed discussion of the methodical aspects
of these measurements and their results are given in the
present paper. Eight copper samples that had been signiWcantly exposed to fast neutrons from the Hiroshima
W. Rühm (&)
Institute for Radiation Protection,
GSF National Research Center for Environment
and Health, Ingolstädter Landstr. 1,
85764 Neuherberg, Germany
e-mail: [email protected]
W. Rühm · G. Rugel · A. Wallner
Radiobiological Institute, Universität München,
80336 Munich, Germany
K. L. Carroll · R. E. Martinelli · A. A. Marchetti · J. E. McAninch
Lawrence Livermore National Laboratory,
Livermore, CA 94551, USA
S. D. Egbert
Science Applications International Corporation,
10260 Campus Point Drive, San Diego, CA 92121, USA
T. Faestermann · K. Knie · G. Korschinek · G. Rugel ·
A. Wallner · C. Wallner
Fakultät für Physik, Technische Universität München,
85747 Garching, Germany
T. Straume
University of Utah, 729 Arapeen Drive,
Salt Lake City, UT 84108, USA
S. Fujita
Department of Statistics,
Radiation EVects Research Foundation,
5-2 Hijiyama Park, Minami-ku, Hiroshima 732-0815, Japan
H. Hasai
Hiroshima Kokusai Gakuin University,
6-20-1 Nakano, Aki-ku, Hiroshima 739-0321, Japan
M. Hoshi
International Radiation Information Center,
Research Institute for Radiation, Biology and Medicine,
Hiroshima University, 1-2-3 Kasumi, Minami-ku,
Hiroshima 734-8553, Japan
K. Shizuma
Quantum Energy Applications,
Graduate School of Engineering,
Hiroshima University, Higashi-Hiroshima 739-8527, Japan
123
328
A-bomb explosion were investigated. In general, measured
63
Ni concentrations decreased in these samples with
increasing distance to the hypocenter, from 4 £ 106 63Ni
nuclei per gram copper at 391 m, to about 1 £ 105 63Ni
nuclei per gram copper at about 1,400 m. Additional measurements performed on three large-distant copper samples
from Hiroshima (distance to the hypocenter 1,880–
7,500 m) and on three large-distant copper samples from
Nagasaki (distance to the hypocenter 3,931–4,428 m) that
were not exposed signiWcantly to A-bomb neutrons, suggest a typical background concentration of about 8 £ 104
63
Ni nuclei per gram copper. If the observed background is
accounted for, the results are consistent with state-of-theart neutron transport calculations for Hiroshima, in particular for those distances where the victims survived and
were included in the life span study cohort.
Introduction
It was only a couple of days after the A-bombs had
exploded when Japanese scientists entered Hiroshima and
Nagasaki, to document the eVects caused by the explosions.
One of their goals was to verify speculations that the tremendous demolition of both cities was due to a nuclear
chain reaction that had produced neutrons. More speciWcally, they intended to detect the radionuclide 32P (half-life
14.26 days) which they expected to have been produced on
sulfur by the fast neutrons from the bombs, via the
32
S(n,p)32P reaction. While suitable samples could not be
found in Nagasaki, electric power poles that contained sulfur as an adhesive could be identiWed in Hiroshima. The
measured 32P activities were signiWcantly above background in samples that were found up to a distance of about
700 m from the hypocenter1 [1–3]. For many decades, these
results remained the only direct information on the fast neutrons emitted by the Hiroshima bomb, and they played an
important role, e.g. when parameters such as bomb yield
and height-of-burst were Wxed for the previous dosimetry
system DS86 [4]. Other radionuclides that had been produced by the neutrons were detected later in samples from
Hiroshima and Nagasaki, such as 60Co [5], 152Eu [6], 154Eu
[6], 36Cl [7–10] and 41Ca [11, 12]. Although these radionuclides had been predominantly produced by thermal neutrons that were not responsible for the neutron doses to the
survivors, they became important in the 1990s because they
seemed to suggest much more thermal neutrons than the
DS86 calculations had implied, at those distances where the
inhabitants of Hiroshima had survived the blast (i.e. at
ground ranges larger than about 1,000 m) [9, 13–16]. See
1
Hypocenter denotes the vertical projection of the point of explosion
(epicenter) to the ground.
123
Radiat Environ Biophys (2007) 46:327–338
[17] for more details on these measurements. The thermal
neutron measurements also questioned the predicted fast
neutron Xuences and accordingly, the neutron doses estimated for the survivors. Unfortunately, the early 32P data
could not be used to clarify the situation since they are reliable (although slightly lower than the DS86 values would
imply) only for distances to the hypocenter smaller than
700 m where only few inhabitants survived.
For this reason, eVorts have been intensiWed since the
mid 1990s to Wnd alternatives that would allow the detection of the fast neutrons at large distances, even 50 years
after the explosions. Among several options the radionuclides 63Ni and 39Ar were identiWed to be most promising
[17–19]. While Wrst results on 39Ar have only recently been
published [20], the 63Ni measurements already became part
of an international eVort to re-evaluate all major aspects of
the DS86 dosimetry.
A major feature of the 63Ni beta decay (half-life
100.1 § 2.0 years [21]) is the fact that no -radiation is
emitted. Therefore, the 63Ni nuclei produced in copper samples from Hiroshima cannot be detected by means of spectroscopy and, accordingly, eVorts have been made to
detect the emitted -particles by means of liquid scintillation counting [22] or detectors based on solid scintillators
[23]. While this method is feasible in principal, much more
sample material is needed than for accelerator mass spectrometry (AMS). This is so because the half-life of 63Ni is
long and, therefore, the activities induced in the samples are
low.
As an alternative it was emphasized that AMS is also
suitable to detect 63Ni in copper samples [24–28], and it
was pointed out that the Munich AMS facility might be the
only laboratory where 63Ni could be measured in copper
samples from ground ranges larger than 1,000 m [24]. AMS
is a method that allows identiWcation of radioisotopes
directly prior to their decay. For this reason, the 63Ni in
copper samples from Hiroshima is measurable by means of
AMS even at large distances from the hypocenter, although
the activity induced in these samples is small. Thus, the low
activity associated with a long half-life, which is essential if
an A-bomb-produced radioisotope is to be measured nowadays, is no obstacle. Application of AMS to detect 63Ni in
large-distant copper samples from Hiroshima required,
however, a suYcient suppression of molecular background
(i.e. molecules with the same mass as 63Ni), isobaric background (i.e. nuclides with the same mass as 63Ni), and isotopic background (i.e. neighboring stable nickel isotopes
62
Ni and 64Ni). Isobaric background turned out to be most
critical for the 63Ni measurements, and special emphasis
was placed on methods to suppress this type of background
as eVective as possible.
In a joint eVort involving teams from the United States
and Germany, the method to measure 63Ni in copper
Radiat Environ Biophys (2007) 46:327–338
samples by means of AMS was successfully developed.
The method includes a dedicated chemistry to extract small
quantities of nickel from a large copper matrix [25, 26], and
detection of the 63Ni extracted by means of AMS [27–31].
The results of these eVorts have recently been summarized
[32]. In the present paper, the long-lasting eVorts made at
the Munich tandem laboratory—now known as MaierLeibnitz Laboratory (MLL)—are highlighted. The results
brieXy reported in [32] are presented and discussed in detail
here, and new results based on additional measurements are
given.
Materials and methods
Samples
The RERF (Radiation EVects Research Foundation) at
Hiroshima and the Hiroshima University routinely keep
track of samples that might be useful for retrospective dose
construction. As a part of this eVort, some of the copper
samples used here had been collected even before the 63Ni
was suggested as a fast neutron Xuence monitor. Additionally, a systematic search for copper samples was carried out
in Hiroshima and Nagasaki in the year 2000, within ground
ranges of about 5,000 m. Most promising to Wnd suitable
samples were ferro-concrete buildings that were not completely destroyed by the explosions, that were refurbished
afterwards, and that were still in use when a potential copper sample was identiWed. Additionally, other structures
such as monuments were also investigated. The search was
based on detailed lists of structures that had been exposed
to A-bomb radiation [33, 34]. A sample was identiWed as
suitable if (a) it was made of pure copper, (b) it was sure
that it had been exposed to A-bomb neutrons, (c) it
remained at its location until it was dismantled, (d) it was in
line-of-sight to the explosions with no or only minor intervening shielding, and (e) stable nickel content in the sample
was suYciently low to assure that the 63Ni production was
not dominated by the thermal-neutron induced
62
Ni(n,)63Ni reaction and that the 63Ni to Ni ratio was not
diluted by too much stable nickel. Samples made of coppercontaining alloys such as bronze or brass were not suitable
because the chemical procedure needed to extract the nickel
from the copper was not applicable for those materials.
Once a copper sample was rated suitable, negotiations with
the owner were conducted to obtain permission to dismantle the sample. If permission was granted, the geometry of
sample and local environment was carefully documented,
and compared with historical photographs of the building
or the sample from before the war. Sample coordinates
were reconstructed based on aerial photographs, and the
distance to the hypocenter was determined based on a
329
recent re-evaluation of the hypocenter coordinates in both
cities [35].
Exposed samples, Hiroshima
The sample closest to the hypocenter in Hiroshima was
found in 2000, on the roof of the former building of the
Bank of Japan, Hiroshima Branch (BoJ). The copper sample of that building is located 391 m to the southeast from
the hypocenter. From a wire hanging down from a lightning
rod in front of a wall that faced the epicenter, several Wlaments were taken.
A copper wire mounted inside an iron pipe on the roof of
the Sanin Godoh Bank (SGB) (ground range 621 m) was
taken in 1983 when the building was demolished. This
sample was thought to have been in line-of-sight to the epicenter, but doubts arose later and the sample was not used
for further analyses (see below).
In the Hiroshima Peace Museum, the upper part of a
concrete chimney that belonged to a soy sauce brewery
(SSB, ground range 964 m) is exhibited. This chimney—it
survived the blast wave and was dismantled in 1969—was
found to include copper material. The sample was probably
part of a lightning rod and consisted of seven copper Wlaments to form an electric wire. For the purpose of this
study, permission was granted to remove those parts of the
wire that were led through an iron pipe and thus were not
visible to the visitors of the museum.
Another lightning rod was found on the roof of the old
City Hall of Hiroshima (CH). It had already been taken in
1985 when the building was dismantled and the new city
hall was built. The sample (ground range 1,018 m) consisted of ten copper Wlaments that formed a wire led
through a vertical iron pipe.
On the former campus of the Hiroshima University,
copper samples were found on three diVerent buildings—
on the Elementary School (ES), the former Faculty of Science (FOS), and the Radioisotope Building (RIB). The ES
sample was nearest to the hypocenter (ground range
1,308 m). It was part of a rain gutter mounted line-of-sight
to the epicenter at the edge of the roof of the building. The
sample was taken in 1996 when the building was knocked
down.
The building of the former FOS (“E-building”) is still
existing. On its facade, several vertical rain gutter pipes are
mounted. Material was taken from two adjacent pipes that
had directly faced the epicenter (ground ranges 1,386 and
1,388 m), and from one additional pipe that had been
mounted opposite to the epicenter on the back side of the
building (ground range 1,469 m). This sample was shielded
against the fast neutrons by the building itself, and served
as a control sample. The samples from the FOS were taken
in 2002.
123
330
Radiat Environ Biophys (2007) 46:327–338
2
1.5
1
0.5
0
0.5
1
1.5
Distant control samples, Hiroshima and Nagasaki
2
2
2
N
1.5
1.5
Distance [km]
1
1
0.5
0.5
Hypocenter
0
0
SGB
SSB
BoJ
0.5
1
CH
FOS
1.5
0.5
1
ES
RIB
1.5
2
2
2
1.5
1
0.5
0
0.5
Distance [km]
1
1.5
2
Fig. 1 Location of the exposed copper samples relative to the hypocenter in Hiroshima
Copper rain gutters were also found on the RIB that was
built in 1934 (ground range 1,470 m). The sample used was
a sprout protruding from the wall that faced the epicenter at
a height of 6 m above ground. Figure 1 shows the location
of these copper samples relative to the hypocenter in
Hiroshima.
A systematic search for copper samples was performed in
2000, and emphasis was also put on those copper samples
that were not exposed signiWcantly to A-bomb neutrons.
These samples were considered important because the 63Ni/
Cu concentration measurable in these samples deWnes the
background that has to be subtracted from the 63Ni signals
found in the exposed samples. It is assumed that distant
copper samples from Hiroshima and Nagasaki that were
exposed to cosmic radiation provide the most realistic way
to deWne the 63Ni background concentration for those copper samples exposed to A-bomb neutrons. Altogether, three
distant copper samples from Hiroshima were included in
the present study: part of a rain gutter sprout protruding
from a wall of the Sumitomo Bank (SuB) building that
faced the epicenter (ground range 1,880 m), and parts from
the copper roof of the Kusatsu Hachiman Shrine (KHS,
ground range 5,062 m) and from the Kamesaki Shrine (KS,
ground range 7,500 m). In addition, three copper samples
from Nagasaki were also used: part of a lightning rod from
the Minori En building (ME, ground range 3,931 m), a
sample from the Hongkong Shanghai Bank (HSB, ground
range 4,187 m), and a part from a rain gutter found at the
Oura Church (OC, ground range 4,428 m). Table 1
shows details of all the copper samples used for the present
investigations.
Table 1 Copper samples from Hiroshima and Nagasaki used for the present study
Name of building
ID
gr (m)
Height
sr (m)
Year of
construction [33, 34]
Year of
sampling
Shielding
against epicenter
Bank of Japan
BoJ
391
18
701
1936
2000
No
Sanin Godoh Banka
SGB
621
11
856
1929
1983
Yes
Soy Sauce Brewery
SSB
964
15.5
1,127
1933
1998
0.3 cm iron
City Hall
CH
1,018
18
1,173
1928
1985
0.4 cm iron
Elementary School
ES
1,308
13
1,434
1938
1996
No
E-building #1
FOS1
1,386
6
1,508
1931
2002
No
E-building #2
FOS2
1,388
6
1,510
1931
2002
No
E-building #3
FOS4
1,388
6
1,510
1931
2002
No
E-building #4
FOS3
1,469
6
1,585
1931
2002
Whole building
Radioisotope Building
RIB
1,470
6
1,585
1934
1996
No
Sumitomo Bank, Hiroshima
SuB
1,880
9.4
1,971
1921
2000
No
Kusatsu Hachiman Shrine, Hiroshima
KHS
5,062
–b
»5,097
1931
2000
–c
Kamesaki Shrine, Hiroshima
KS
7,500
–b
»7,524
»1900
2000
–c
3,931
–
b
»3,963
1927
2000
–c
–
b
»4,217
1896
2000
–c
b
»4,456
1864
2000
–c
Minori En, Nagasaki
Hongkong Shanghai Bank, Nagasaki
Oura Church, Nagasaki
ME
HSB
OC
4,187
4,428
–
Ground range (gr) and slant range (sr) are based on DS02 coordinates of the epicenter [35]
a
The SGB sample was not used for further analyses because it turned out to have been signiWcantly shielded
b
No data available
c
Not shielded against cosmic radiation
123
Radiat Environ Biophys (2007) 46:327–338
Sample preparation
When it became clear that the 63Cu(n,p)63Ni reaction would
oVer a unique possibility to reconstruct fast neutron
Xuences in Hiroshima and Nagasaki, eVorts were Wrst made
to extract minor amounts of nickel from a large matrix of
natural copper. It was found that a two-step procedure was
most suitable, involving an initial electrolysis step to
remove the bulk of copper. The nickel (both, stable nickel
isotopes and 63Ni) left in solution was then converted to
nickel tetracarbonyl, (Ni(CO)4), by means of carbon monoxide, CO, that bubbled through the solution. Finally, the
Ni(CO)4 was carried with helium gas to a heated graphite
target holder used for the AMS measurements, where it was
thermally decomposed at about 200°C [25, 26]. The stable
nickel content in the initial copper samples was usually
determined after electrolysis in the solution, by means of
graphite furnace atomic absorption spectrometry. Both the
electrolysis and the carbonyl chemistry are highly eVective
to separate the nickel from the copper and thus, the 63Ni
from its stable isobar 63Cu. If the natural stable nickel content of the copper sample was too low to produce a nickel
amount suYcient for the AMS measurement, stable nickel
was added as a carrier. Typically, between 10 and 50 g of a
copper sample were necessary to produce a measurable Nisample.
Experimental setup
The AMS target holder which contained the decomposed
Ni(CO)4 material was introduced into a cesium sputter
source. This ion source was speciWcally designed to extract
high currents of negatively charged Ni¡ ions from the sample material, and at the same time to reduce production of
any Cu¡ ions from stable copper that might exist as an
impurity in the sample, the sample holder, or in the structure material of the ion source [29]. The extracted ions are
accelerated by means of an electrostatic tandem accelerator
set typically to a terminal voltage of about 13.7 MV. In the
center of the accelerator (“terminal”) a foil removes electrons from the 63Ni¡ ions and the positively charged 63Ni
ions are further accelerated towards the end of the tandem.
Any molecules extracted from the ion source are destroyed
in the foil and molecular background does therefore not
play any further role. After the accelerator a 90° analyzing
magnet is used to select the 63Ni12+ that had been accelerated to an energy of 170 MeV. Isotopic background is further reduced by means of a Wien velocity Wlter, and
isobaric background by means of a gas-Wlled 135° magnet.
Both types of background are further reduced by means of a
multi-E ionization chamber that is installed after the gasWlled magnet.
331
All AMS results are basically obtained in terms of 63Ni/
Ni ratios. Typically, the combination of the chemical preparation method developed at LLNL and the AMS method
developed at MLL allows detection of 63Ni/Ni ratios as low
as 2 £ 10¡14. Once the stable nickel content in a certain
sample was known, the 63Ni/Ni ratio could be converted to
the corresponding 63Ni/Cu ratio. Taken together, the
applied method allows to measure 63Ni concentrations as
low as about 104 63Ni nuclei per gram copper. The AMS
setup is described in detail in [27–30].
Measurement procedure and data evaluation
The procedure involved comparative measurements of standard samples (i.e. samples with known 63Ni/Ni ratios), blank
samples (i.e. samples without 63Ni), and real samples (i.e.
samples with unknown 63Ni/Ni ratios). Typically, the set of
standards included samples with 63Ni/Ni ratios of 10¡10 and
10¡11. These were used to identify and optimize the 63Ni
signals in the detector (“63Ni cuts”), and to quantify the
transmission of the 63Ni ions from the entrance of the accelerator to the detector. The blank samples were used to quantify any isobaric background, i.e. those events that entered
the 63Ni regions in the spectra without being 63Ni. The real
samples included those from Hiroshima, but also those produced from a nickel standard with a nominal 63Ni/Ni ratio of
3.67 £ 10¡13. Although the 63Ni/Ni ratio of this sample was
known, it was treated as the Hiroshima samples, and its
63
Ni/Ni ratio was determined and compared to the nominal
ratio as another means of quality control. This was to assure
that 63Ni/Ni ratios as low as those expected for the distant
Hiroshima samples could be measured.
For a real sample, the contribution from the isobaric
background was calculated based on results obtained on the
blank samples, and on the number of 63Cu events in the
actual spectrum [31]. ConWdence levels for the resulting
diVerence were calculated by means of an approach including Poisson statistics developed for experiments with small
signals and known background [36].
Once an exposed sample was measured successfully, it
was measured again several months later at another run, if
possible. For this purpose, an aliquot from the exposed copper sample was used to produce a new AMS target. Most of
the exposed samples were even measured a third time. All
AMS measurements performed were blind in the sense that
the stable nickel content determined at the LLNL and the
amount of stable nickel carrier added at the LLNL was not
disclosed to the Munich group before the Wnal result for the
63
Ni/Ni ratio determined at Munich was opened to the
LLNL group. This procedure minimized any potential bias
resulting from expected 63Ni atoms/g Cu concentrations.
All results are given with 68.3% conWdence levels.
123
332
Results and discussion
63
Ni/Ni ratios and 63Ni/g copper concentrations obtained
for the exposed samples, Hiroshima
All AMS results obtained on the exposed Hiroshima samples are given in Table 2, in terms of 63Ni/Ni ratios (column
6). Based on these values, on the concentration of stable
nickel in the copper samples and the amount of stable
nickel carrier added to some of the samples during sample
preparation, the corresponding 63Ni/Cu ratios were also calculated (last column). At this point, corrections for radioactive decay, contribution from thermal neutrons, shielding
by surrounding material or the sample itself, and for the
background observed in the distant samples were not yet
applied. The only correction applied to the measured 63Ni/
Ni ratios was due to isobaric background (column 5) that
was erroneously identiWed as 63Ni in the AMS measurement. It should be noted that the 63Ni/Ni ratios obtained for
diVerent AMS samples prepared from the copper sample of
a certain building need not necessarily be similar as the
amount of stable nickel present in the copper or added as a
carrier during chemical processing might have been diVerent.
Bank of Japan
Less than 10 g copper was enough to perform a reliable
AMS measurement. Although this sample was located
close to the hypocenter, the 63Ni/Ni ratios measured were
relatively low, due to the high level of stable nickel present
in the sample. The 3/2001 run was characterized by high
background levels, and diVerent procedures had to be
applied to identify 63Ni ions resulting in comparatively
large uncertainties. In terms of 63Ni/Cu, this sample showed
the highest values, and all of the three independent measurements gave results consistent with their weighted average (Fig. 2).
The high nickel concentration in the sample indicates
that thermal activation by the 62Ni(n,)63Ni reaction might
also be an important source of 63Ni. Indeed, estimates
based on DS02 neutron spectra at that distance (1 m above
ground) [35] suggest that about 24% of the measured
63
Ni was produced by thermal neutrons. Thus, our data
must be corrected for this contribution if they are used to
verify fast neutron calculations and, thus, neutron dose
estimates.
Radiat Environ Biophys (2007) 46:327–338
was induced by thermal neutrons in the iron pipe in which
the copper wire had been mounted. Thus, it is believed that
this sample was signiWcantly shielded, and the 63Ni measurement given in Table 2 was not used for further analysis.
Soy Sauce Brewery
The high level of stable nickel present made it extremely
diYcult to measure this sample, since it resulted in a very
low 63Ni/Ni ratio. In principle, the measured ratios of 1–
2 £ 10¡13 were still well above our detection limit of
2 £ 10¡14 [30], but this limit could not be reached during
the 4/2000 run, and major corrections had to be made for
machine background (see Table 2). As a result, the mean
value given in Fig. 3 is dominated by the results obtained
during the 7/2001 run. Another consequence of the high
stable nickel concentration is that—similar to the BoJ sample—activation due to thermal neutrons through the
62
Ni(n,)63Ni reaction is not negligible. Based on DS02
[35] we would estimate for the SSB sample that about 6%
of the measured 63Ni signal was due to thermal neutrons.
Although the stable nickel concentration is almost the same
as that in the BoJ sample, the resulting relative thermal activation is a factor of about 4 less because the fraction of
thermal/fast neutrons decreases with increasing ground
range.
City Hall
Two samples were measured during the 9/1998 run, and
another sample during the 7/1999 run. Although the samples were located much more distant from the hypocenter
than those from the BoJ or the SSB, the measured 63Ni/Ni
ratios were similar or even larger (Table 2) since the level
of stable nickel was relatively low (4.1–9.1 ppm). Because
of this variation in the stable nickel concentration, the
determined individual 63Ni/Cu ratios are quite consistent
and agree with their weighted mean, although the corresponding individual 63Ni/Ni ratios vary considerably.
Samples from the City Hall were also independently
measured at LLNL in May and September 1998, and a
weighted mean of (2.9 § 0.4) £ 105 63Ni/g Cu was
obtained, in agreement with the weighted mean value of
(2.4 § 0.4) £ 105 63Ni/g presented here. For further analyses, both data sets were combined and a weighted mean of
(2.65 § 0.28) £ 105 63Ni/g was used (Fig. 4).
Elementary School
Sanin Godoh Bank
The 63Ni signal both in terms of the 63Ni/Ni and the 63Ni/Cu
ratios was signiWcantly lower than expected. Later it turned
out that this had also been the case for the 60Co signal that
123
The results obtained for the Elementary School samples are
consistent both in terms of 63Ni/Ni and 63Ni/Cu (Fig. 5).
Again reasonable agreement between the individual 63Ni/
Cu values and their weighted mean is observed.
Radiat Environ Biophys (2007) 46:327–338
333
Table 2 Summary of results obtained on copper samples from Hiroshima and Nagasaki, during various beam times
Ground
range (m)
[35]
Mass of
copper (g)
Events in
the cut
Subtracted
background
events
63
Ni/Ni
(10¡12)
Nickel
(ppm)
Nickel
carrier
(g)
63
CH
1,018
36.89
17
2.2
1.9 § 0.6
4.1 § 0.7
192.8
1.9 § 0.6
CH
1,018
29.63
19
0.2
1.5 § 0.4
9.1 § 2.5
192.9
2.4 § 0.7
ES
1,308
24.16
38
0.6
0.63 § 0.11
16.1 § 3.2b
103.5
1.32 § 0.31
CH
1,018
39.78
145
5.7
3.1 § 0.4
6.9 § 2.1
193.6
3.7 § 0.8
ES
1,308
13.71
33
7.3
0.42 § 0.11
18.4 § 1.7
104.2
1.11 § 0.30
ES
1,308
24.72
57
2.8
0.55 § 0.10
13.8 § 0.4
105.9
1.02 § 0.18
Blind10¡13
–
–
84
3.7
0.34 § 0.05
–
–
–
1,470
29.75
–
–
0.55 § 0.25
8.8 § 0.2b
103.9
0.69 § 0.31
–
–
–
–
0.40 § 0.16
Sample/location
Ni/Cu
(105 at/g)
Beamtime 9/1998a
Beamtime 7/1999
Beamtime 12/1999
RIB
¡13
Blind10
Beamtime 4/2000
SSB
964
12.31
7
2.3
0.13 + 0.10 ¡ 0.09
258.3 § 4.8
–
3.5 + 2.7 ¡ 2.4
RIB
1,470
26.92
46
6.5
1.30 + 0.34 ¡0.28
11.1 § 0.8b
–
1.5 + 0.4 ¡ 0.3
Blind10¡13
–
–
9
0.5
0.18 + 0.08 ¡ 0.06
–
–
–
BoJ
391
9.52
200
0.8
2.05 § 0.21
212.8 § 14.5
–
45 § 5
SGB
621
9.73
18
1.7
0.26 + 0.09 ¡ 0.07
21.1 § 0.6
RIB
1,470
24.41
27
1.8
1.1 § 0.3
9.3 § 0.8b
–
1.04 § 0.26
BoJ
391
3.37
–
–
2.7 + 0.7 ¡ 0.9
199.1 § 7.5
–
55 + 14 ¡ 19
SuB, Hiro
1,880
12.95
–
–
<0.11
18.5 § 1.1
–
<0.21
KHS, Hiro
5,062
3.48
–
–
0.09 + 0.20 ¡ 0.09
124 § 3
–
1.14 + 2.54 ¡ 1.14
Blind10¡13
–
–
–
–
0.59 + 0.76 ¡ 0.26
–
–
–
BoJ
391
5.94
77
5.4
1.62 § 0.24
207.8 § 1.2
–
35 § 5
SSB
964
12.31
25
8.3
0.18 § 0.06
258.3 § 4.8
–
4.9 + 1.7 ¡ 1.6
SuB, Hiro
1,880
19.96
16
1.5
0.41 + 0.14 ¡ 0.11
21.3 § 3.9
–
0.93 + 0.34 ¡ 0.27
Beamtime 8/2000a
0.54 + 0.20 ¡ 0.15
Beamtime 3/2001
Beamtime 7/2001
KHS, Hiro
5,062
5.76
4
1.7
0.05 + 0.06 ¡ 0.04
136 § 10
–
0.7 + 0.8 ¡ 0.5
Blind10¡13
–
–
62
14
0.31 § 0.06
–
–
–
Beamtime 1/2002
KS, Hiro
7,500
20.19
–
–
<0.036
251.9 § 19.6
–
<0.9
ME, Naga
3,931
94.96
4
0.6
0.58 + 0.48 ¡ 0.30
3.0 § 0.9
–
0.18 + 0.15 ¡ 0.09
HSB, Naga
4,187
78.94
19
0.7
2.7 + 0.8 ¡ 0.7
7§1
–
1.8 § 0.5
HSB, Naga
4,187
78.94
36
1.2
3.3 § 0.6
7§1
–
2.2 § 0.5
OC, Naga
4,428
21.65
5
1.6
0.17 + 0.31 ¡ 0.06
12.2 § 1.2
–
0.21 + 0.38 ¡ 0.08
Blind10¡13
–
–
15
2.8
0.22 + 0.09 ¡ 0.07
–
–
–
FOS1
1,386
16.85
30
3.1
0.45 + 0.13 ¡ 0.10
16.6 § 0.6
–
0.77 + 0.22 ¡ 0.17
FOS2
1,388
29.67
14
0.6
0.29 + 0.10 ¡ 0.08
33.8 § 0.8
–
1.0 + 0.4 ¡ 0.3
FOS3
1,469
30.00
26
1.7
0.33 + 0.08 ¡ 0.07
12.1 § 0.1
–
0.4 § 0.1
FOS4
1,388
29.67
12
0.3
1.9 + 0.5 ¡ 0.7
33.8 § 0.8
–
6.6 + 2.4 ¡ 1.7
SuB, Hiro
1,880
13.63
6
0.3
1.2 + 0.7 ¡ 0.5
21.3 § 3.9
–
2.4 + 1.4 ¡ 1.0
HSB, Naga
4,187
45.75
1
0.2
0.41 + 0.89 ¡ 0.32
2.0 § 0.1
–
0.09 + 0.19 ¡ 0.07
Beamtime 7/2002
Beamtime 7/2003c
123
334
Radiat Environ Biophys (2007) 46:327–338
Table 2 continued
Sample/location
Ground
range (m)
[35]
Mass of
copper (g)
Events in
the cut
Subtracted
background
events
63
Ni/Ni
(10¡12)
Nickel
(ppm)
Nickel
carrier
(g)
63
Blind10¡13
–
–
3
0.12
0.49 + 0.39 ¡ 0.33
–
–
–
Blind10
¡13
–
–
6
0.3
0.66 + 0.39 ¡ 0.26
–
–
–
Blind10¡13
–
–
1
0.02
0.31 + 0.55 ¡ 0.20
–
–
–
Ni/Cu
(105 at/g)
BoJ Bank of Japan, SSB Soy Sauce Brewery, CH City Hall, ES Elementary School, FOS Faculty of Science, RIB Radioisotope Building, SuB Sumitomo Bank, KHS Kusatsu Hachiman Shrine, KS Kamesaki Shrine, ME Minori En Building, HSB Hongkong Shanghai Bank, OU Oura Church; the
“Blind 10¡13” sample had a nominal 63Ni/Ni ratio of 0.367 £ 10¡12
a
Measurement of the blind sample was not possible
b
ppm value deduced from mean of the other samples from the same copper piece
c
Samples from the A-bomb dome had also been measured in this beam time, but will be discussed in a separate paper
80.0
4
63
Ni/Ni [10
-12
63
]
4
5.0
63
5
Ni/Cu [10 at/g]
3.5
70.0
3.5
3
60.0
3
2.5
50.0
2.5
2
40.0
2
1.5
30.0
1.5
20.0
1
Ni/Ni [10
-12
]
63
5
Ni/Cu [10 at/g]
4.0
3.0
2.0
1
8/2000 3/2001 7/2001
8/2000 3/2001 7/2001
0.0
9/1998 9/1998 7/1999
Fig. 2 63Ni/Ni ratios measured for the Bank of Japan samples (left
panel); the corresponding 63Ni/Cu ratios (right panel) are compared to
their weighted mean (dashed line) and its uncertainty (dotted lines)
0.5
1.0
9/1998 9/1998 7/1999
Fig. 4 63Ni/Ni ratios measured for the City Hall samples (left panel);
the corresponding 63Ni/Cu ratios (right panel) are compared to their
weighted mean (dashed line) and its uncertainty (dotted lines), as obtained from the Munich measurements
10.0
63
Ni/Ni [10
-12
]
63
5
Ni/Cu [10 at/g]
0.4
8.0
0.3
6.0
0.2
4.0
2.5
1
63
0.1
2.0
0
4/2000 7/2001
Ni/Ni [10
-12
]
63
5
Ni/Cu [10 at/g]
0.8
2.0
0.6
1.5
0.4
1.0
0.0
4/2000 7/2001
63
Fig. 3 Ni/Ni ratios measured for the Soy Sauce Brewery samples
(left panel); the corresponding 63Ni/Cu ratios (right panel) are compared to their weighted mean (dashed line) and its uncertainty (dotted
lines)
Faculty of Science
Although the two exposed copper samples (FOS1 and
FOS2) were from somewhat diVerent locations, the results
are combined in Fig. 6 to calculate their weighted mean,
123
0.2
0.5
9/1998 7/1999 7/1999
9/1998 7/1999 7/1999
Fig. 5 63Ni/Ni ratios measured for the Elementary School samples
(left panel); the corresponding 63Ni/Cu ratios (right panel) are compared to their weighted mean (dashed line) and its uncertainty (dotted
lines)
Radiat Environ Biophys (2007) 46:327–338
335
2.0
1
63
Ni/Ni [10 -12]
63
Ni/Cu [10 5 at/g]
0.8
1.6
0.6
1.2
0.4
0.8
0.2
0.4
0.0
0
1/2002 1/2002 1/2002
1/2002 1/2002 1/2002
Fig. 6 63Ni/Ni ratios measured for the Faculty of Science samples
FOS1 and FOS2 (black symbols, left panel); the corresponding 63Ni/Cu
ratios (open symbols, right panel) are compared to their weighted mean
(dashed line) and its uncertainty (dotted lines); for comparison, the results obtained on the shielded sample FOS3 are also shown (open symbol, left panel; black symbol, right panel) (see text for details)
because their ground ranges are similar (1,386 m for sample FOS1 and 1,388 m for sample FOS2). The signiWcantly
higher result of the FOS4 sample (7/2003), an aliquot of
FOS2, is not included in the Wgure, because it has to be remeasured before any Wnal conclusion on the reliability of
its result can be drawn. The weighted mean given in the
Wgure [(0.84 + 0.19 ¡ 0.15) £ 105 63Ni/g Cu] is also based
only on the FOS1 and FOS2 samples. It should therefore be
seen as preliminary, until the measurements of the FOS4
sample are completed.
For comparison, the results obtained on the shielded
sample are also shown in Fig. 6. Although the 63Ni/Ni ratio
for the FOS3 sample is similar to those obtained for the
FOS1 and FOS 2 samples, the corresponding 63Ni/Cu ratio
for the shielded FOS3 sample is lower than those for the
line-of-sight samples, because the amount of nickel in the
FOS3 sample was smaller than that in the exposed samples.
It appears that the numerical result for the shielded FOS3
sample (63Ni/Cu = 0.4 § 0.1 £ 105 63Ni nuclei per gram
copper, see Table 2) is somewhat lower than the results
obtained on the rest of the unexposed samples that gave a
mean value of (0.79 + 0.24 ¡ 0.19) £ 105 63Ni nuclei per
gram copper (see discussion below and Table 2). It is noted
that, contrary to all other exposed samples, those from the
FOS originate from diVerent copper pieces and each of the
four samples has only been measured once so far. Additional measurements are therefore required to conWrm these
results.
Radioisotope Building
Three samples from the RIB were measured in three successive runs. Although the result obtained in December
1999 in terms of the 63Ni/Ni ratio was signiWcantly lower
than those obtained later, all three measurements are in
agreement in terms of the 63Ni/Cu ratio (Fig. 7). This is
partly due to the fact that for preparation of the 12/1999
sample stable nickel carrier was added which lowered the
63
Ni/Ni ratio, whereas the preparation of the remaining
samples was done without adding any carrier.
63
Ni/Ni ratios and 63Ni/g copper concentrations obtained
for the control samples not exposed to A-bomb neutrons,
Hiroshima and Nagasaki
The results obtained for those copper samples from Hiroshima and Nagasaki that were not signiWcantly exposed to
A-bomb neutrons are also given in Table 2. So far two samples have already been measured more than once—the SuB
sample and the KHS sample. Because only an upper limit
for the 63Ni/Cu ratio was obtained for the SuB sample in the
3/2001 run, the measurements from 2001 for this sample
were combined and treated as a single run. In combination
with the measurement from 2003 that suggested a higher
ratio, the weighted mean is (0.79 + 0.26 ¡ 0.19) £ 105 63Ni
nuclei per gram copper. The weighted average for the concentration of the KHS sample was (0.7 + 0.8 ¡ 0.5) £ 105
63
Ni nuclei per gram copper. The weighted average for the
SuB and the KHS samples combined—(0.79 + 0.24 ¡
0.19) £ 105 63Ni nuclei per gram copper—is at present seen
as representative for the background concentration in the
Hiroshima copper samples. It should be noted that, based
on an earlier evaluation, the numerical value for the background used in the earlier publication [32] was 0.73 £ 105
63
Ni nuclei per gram copper.
Although the results obtained for the remaining unexposed samples must be interpreted with caution since independent second measurements have not yet been
performed, they appear in most cases to be consistent with
this background value of (0.79 + 0.24 ¡ 0.19) £ 105 63Ni
2
2.3
63
Ni/Ni [10
-12
]
63
5
Ni/Cu [10 at/g]
1.6
1.9
1.2
1.5
0.8
1.1
0.4
0.7
0
0.3
12/1999 4/2000 8/2000
12/1999 4/2000 8/2000
Fig. 7 63Ni/Ni ratios measured for the Radioisotope Building samples
(left panel); the corresponding 63Ni/Cu ratios (right panel) are compared to their weighted mean (dashed line) and its uncertainty (dotted
lines)
123
336
Radiat Environ Biophys (2007) 46:327–338
1E+01
5
N i ( 10 at / gr C u)
1E+02
63
nuclei per gram copper. This is the case for the KS sample
from Hiroshima and for the ME and OC samples from
Nagasaki (see Table 2). As for the HSB sample from Nagasaki, two measured 63Ni concentrations are clearly higher
(1/2002 beam time), one is also low (3/2003 beam time).
Additional measurements must be performed to clarify this
result before it is possible to draw any conclusions. Similarly, the concentration of (0.4 § 0.1) £ 105 63Ni nuclei per
gram copper in the shielded copper sample from the FOS
building that is lower than the background value deduced
above, has to be veriWed (see Fig. 6, Table 2).
1E+00
Comparison with DS02 free-in-air calculations
1E-01
The mean background concentration of (0.79 + 0.24 ¡
0.19) £ 105 63Ni nuclei per gram copper observed in the
copper samples not exposed to A-bomb neutrons, is interpreted here as an overall background level present in any
copper sample from Hiroshima. Most probable, it is due to
the applied chemical procedure and the instrumental background of the AMS measurements. Production of 63Ni in
the copper samples by cosmic rays is discussed in detail in
[37], and it was concluded that this source is not suYcient
to explain the observed background level. As is shown
below, a similar level of 63Ni in the copper samples is
reached at a ground range of about 1,500 m, due to neutrons from the A-bomb.
For comparison with the averaged experimental results
shown in Table 3, 63Ni concentrations are used here that are
based on DS02 neutron Xuences calculated free-in-air (Wa)
and one meter above ground [35], folded with the ENDF/BVI cross section of the 63Cu(n,p)63Ni reaction [38]. They
are corrected for radioactive decay for the year 2000, and
with a constant background concentration of (0.79 +
0.24 ¡ 0.19) £ 105 63Ni nuclei per gram copper added
Table 3 Weighted means for the 63Ni/Cu concentrations obtained in
the exposed copper samples from Hiroshima (column 3); all data are
given for the date of measurement and no corrections for production
due to thermal neutrons or for shielding were applied
Sample
gr (m)
Measured 63Ni/Cu (105/g Cu)
BoJ
391
40.0 § 4.0
SSB
964
4.4 § 1.4
CH
1,018
2.65 § 0.27a
ES
1,308
1.10 § 0.14
FOS
1,386–1,388
0.84 + 0.19 ¡ 0.15b
RIB
1,470
1.03 § 0.17
gr Ground range
Weighted mean of the results obtained at the MLL and the LLNL (for
details see text)
a
b
Weighted mean of the results obtained on the FOS1 and FOS2 samples [if the result obtained on the FOS4 sample is included, the weighted mean changes only marginally (for details see text)]
123
0
500
1000
1500
2000
ground range (m)
Fig. 8 63Ni/Cu concentrations measured by means of AMS (black
symbols; see also Table 2) versus distance from the hypocenter
(ground range); thick solid line: calculated DS02 free-in-air (1 m above
ground) concentration of 63Ni in copper due to fast neutrons, decaycorrected for 2000 and with a constant background level of 0.79 £ 105
63
Ni nuclei per gram copper added; thin solid lines: same calculations
but with upper and lower limit of the background level (i.e. 1.03 £ 105
and 0.60 £ 105 63Ni nuclei per gram copper) added; note: the experimental values have not yet been corrected for shielding eVects (e.g.
self-shielding of the sample) and height above ground; compared to
Table 3, the experimental values for the BoJ and the SSB samples were
corrected for an assumed 24 and 5.7% contribution from thermal neutrons through the 62Ni(n,)63Ni reaction [32, 35]
(Fig. 8). It is important to note that—at this stage—the calculated 63Ni concentrations do not include any contribution
from the thermal-neutron induced 62Ni(n,)63Ni reaction
which is important for the BoJ and the SSB samples (see
above), nor any correction for the actual sample height and
for shielding or self-shielding eVects. Calculation of these
correction factors requires extensive computational eVorts
that have been made in the framework of the new dosimetry
system DS02 [35], but which is beyond the scope of this
paper. However, corrections due to height above ground
and shielding are not expected to alter the estimated Wa activation by more than 10% [35]. In order to facilitate comparison between measured and calculated results for the
BoJ and the SSB samples, the corresponding measured 63Ni
concentrations were corrected in Fig. 8 for an estimated
contribution from thermal neutrons of 24.5 and 5.7%,
respectively [32, 35].
The experimental results shown in Table 3 and Fig. 8
suggest a signiWcant decrease of the measured 63Ni/Cu
ratios with increasing distance, for ground ranges lower
than about 1,300 m. This is interpreted as strong indication
that the 63Ni in the exposed samples was indeed produced
by neutrons from the A-bomb. Except for the BoJ sample
whose 63Ni signal shows also a signiWcant contribution
Radiat Environ Biophys (2007) 46:327–338
from thermal neutrons (see above), all experimental results
agree with the Wa DS02 calculation for fast neutrons. At
ground ranges beyond 1,300 m, the measured 63Ni concentrations level oV and approach concentrations that are consistent with those measured in the distant copper samples.
This overall interpretation also holds if the required modiWcation factors were applied [32, 35].
Conclusion
This paper shows the results of a long-lasting eVort to
determine concentrations of the radioisotope 63Ni in copper
samples from Hiroshima and Nagasaki. This was done in
an attempt to reconstruct the neutron doses to the survivors
of the A-bomb explosions. Emphasis is placed here on
work done at the MLL in Munich, Germany. In particular,
it is demonstrated that the method of AMS applied to detect
the 63Ni atoms produces reliable results for 63Ni/Ni ratios in
real samples as low as 10¡13. This was shown in a series of
measurements using samples of a known nominal 63Ni/Ni
ratio of 3.67 £ 10¡13. The results obtained on material
from the same copper samples from Hiroshima were shown
to be consistent. This is remarkable in so far as the experimental setup used at the MLL by far exceeds conventional
AMS standards. In most cases, AMS is used to detect light
radionuclides such as 10Be and 14C, by means of dedicated
machines which include small electrostatic accelerators. At
the MLL, a large tandem accelerator (terminal voltage
14 MV) is available to accelerate 63Ni ions to an energy
high enough that a gas-Wlled magnet can be used for isobaric suppression, in combination with a multi-anode ionization chamber. It was this unique combination that made
the Munich AMS setup suitable for the detection of 63Ni at
large distances of the hypocenter in Hiroshima.
The 63Ni/Cu concentrations obtained for the copper samples that were exposed to A-bomb neutrons are close to calculations that are based on DS02 fast neutron Xuences freein-air one meter above the ground [35]. This might already
indicate that the DS02 neutron dose estimates should not be
too far from the real values. It is evident, however, that
detailed transport calculations are required to take into
account (a) the actual height of the investigated copper
samples above the ground, (b) the modiWcation of the fast
neutron spectrum by the local environment around the samples and the sample itself, and (c) the contribution of thermal neutrons to the 63Ni concentration in the samples. The
contribution of thermal neutrons does not only include that
from the 62Ni(n,)63Ni reaction, but possibly also that from
the exothermic 63Cu(n,p)63Ni reaction. A detailed description of those transport calculations that have already been
performed, is beyond the scope of the present paper, but is
presented elsewhere [35].
337
The 32P data provided, already in the very early phase of
neutron dose reconstruction, reliable data on the fast neutron Xuence in Hiroshima for ground ranges smaller than
about 700 m. Almost 60 years later the 63Ni data summarized in [32] and presented here in much more detail,
extend the ground ranges for which the detection of fast
neutrons is possible to almost 1,500 m. Thus, more than
half a century after the tragic explosions over Hiroshima
and Nagasaki the neutron doses to the survivors can now be
calculated, based on sound experimental grounds.
Acknowledgment The authors would like to thank A.M. Kellerer
for his essential support of this project. This work was supported by
the German Federal Ministry of Environment, Nature Conservation
and Nuclear Safety under contracts StSch 4235 and StSch 4267, and
by the European Commission under contract FIGD-CT2000-0079.
One of us (G.R.) would like to thank the Bavarian Government for a
grant. Part of this work was performed under the auspices of the US
Department of Energy by the Lawrence Livermore National Laboratory under contract W-7405-Eng-48, and by a grant to the University
of Utah (T. Straume) from the US Department of Energy (#DE-FG0300ER62963).
References
1. Yamasaki F, Sugimoto A (1953) Radioactive 32P produced in sulfur in Hiroshima. Collection of investigative reports on atomic
bomb disaster. Science Council of Japan, Tokyo, pp 18–19
2. Sugimoto A (1953) Determination of number of fast neutron particles emitted at the time of the Hiroshima A-bomb explosion. Collection of investigative reports on atomic bomb disaster. Science
Council of Japan, Tokyo, p 19
3. Arakatsu B (1953) Report on survey of radioactivity in Hiroshima
several days after the atomic bomb explosion. Collection of investigative reports on atomic bomb disaster. Science Council of Japan, Tokyo, pp 5–10
4. Roesch WC (ed) (1987) US–Japan joint reassessment of atomic
bomb radiation dosimetry in Hiroshima and Nagasaki—Wnal report, vol 1,2. Radiation EVects Research Foundation, Hiroshima,
Japan
5. Hashizume T, Maruyama T, Shiragai A, Tanaka E, Izawa M, Kawamura S, Nagaoka S (1967) Estimation of the air dose from the
atomic bombs in Hiroshima and Nagasaki. Health Phys 13:149–
161
6. Nakanishi T, Imura T, Komura K, Sakanoue M (1983) 152Eu in
samples exposed to the nuclear explosions at Hiroshima and Nagasaki. Nature 302:132–134
7. Haberstock G, Heinzl J, Korschinek G, Morinaga H, Nolte E,
Ratzinger U, Kato K, Wolf M (1986) Accelerator mass spectrometry with fully stripped 36Cl ions. Radiocarbon 28:204–210
8. Kato K, Habara M, Yoshizawa Y, Biebel U, Haberstock G, Heinzl
J, Korschinek G, Morinaga H, Nolte E (1990) Accelerator mass
spectrometry of 36Cl produced by neutrons from the Hiroshima
bomb. Int J Radiat Biol 58:661–672
9. Straume T, Egbert SD, Woolson WA, Finkel RC, Kubik PW,
Gove HE, Sharma P, Hoshi M (1992) Neutron discrepancies in the
DS86 Hiroshima Dosimetry System. Health Phys 63:412–426
10. Huber T, Rühm W, Kato K, Egbert S, Kubo F, Lazarev V, Nolte E
(2005) The Hiroshima thermal neutron discrepancy for 36Cl at
large distances; part I: New 36Cl measurements in granite samples
exposed to A-bomb neutrons. Radiat Environ Biophys 44:75–86
123
338
11. Korschinek G, Morinaga H, Nolte E, Preisenberger E, Ratzinger
U, Urban A, Dragovitsch P, Vogt S (1987) Accelerator mass spectrometry with completely stripped 41Ca and 53Mn ions at the Munich tandem accelerator. Nucl Instrum Methods B29:67–71
12. Rühm W, Kato K, Korschinek G, Morinaga H, Nolte E (1992) 36Cl
and 41Ca depth proWles in a Hiroshima granite stone and the
Dosimetry System 1986. Z Phys A Hadrons Nuclei 341:235–238
13. Hoshi M, Yokoro K, Sawada S, Shizuma K, Iwatani K, Hasai H,
Oka T, Morishima H, Brenner DJ (1989) Europium-152 activity
induced by Hiroshima atomic bomb neutrons: comparison with the
32
P, 60Co, and 152Eu activities in the dosimetry system 1986
(DS86). Health Phys 57:831–837
14. Nakanishi T, Ohtani H, Mizuochi R, Miyaji K, Yamamoto T, Kobayashi K, Imanaka T (1991) Residual neutron-induced radionuclides in samples exposed to the nuclear explosion over Hiroshima:
comparison of the measured values with the calculated values. J
Radiat Res 32(suppl):69–82
15. Shizuma K, Iwatani K, Hasai H, Hoshi M, Oka T, Morishima H
(1993) Residual 152Eu and 60Co activities induced by neutrons
from the Hiroshima atomic bomb. Health Phys 65:272–282
16. Shizuma K, Iwatani K, Hasai H, Oka T, Endo S, Takada J, Hoshi
M, Fujita S, Watanabe T, Imanaka T (1998) Residual 60Co activity
in steel samples exposed to the Hiroshima atomic-bomb neutrons.
Health Phys 75(3):1–7
17. Rühm W, Kellerer AM, Korschinek G, Faestermann T, Knie K,
Rugel G, Kato K, Nolte E (1998) The dosimetry system DS86
and the neutron discrepancy in Hiroshima—historical review,
present status, and future options. Radiat Environ Biophys
37:293–310
18. Shibata T, Imamura M, Shibata S, Uwamino Y, Ohkubo T, Satoh
S, Nogawa N, Hasai H, Shizuma K, Iwatani K, Hoshi M, Oka T
(1994) A method to estimate the fast-neutron Xuence for the Hiroshima atomic bomb. J Phys Soc Jpn 63(10):3546–3547
19. Marchetti AA, Straume T (1996) A Search for neutron reactions
that may be useful for Hiroshima dose reconstruction. Appl Radiat
Isot 47(1):97–103
20. Nolte E, Rühm W, Loosli HH, Tolstikhin I, Kato K, Huber TC, Egbert SD (2006) Measurements of fast neutrons in Hiroshima by use
of 39Ar. Radiat Environ Biophys 44:261–271
21. Barnes IL, GarWnkel SB, Mann WB (1971) Standardization of
63
Ni by eYciency tracing. Int J Appl Radiat Isot 22:781–783
22. Shizuma K, Iwatani K, Hasai H, Oka T, Hoshi M, Shibata S,
Imamura M, Shibata T (1997) IdentiWcation of 63Ni and 60Co produced in a steel sample by thermal neutrons from the Hiroshima
atomic bomb. Nucl Instrum Methods Phys Res A384:375–379
23. Ito Y, Shibata T, Imamura M, Shibata S, Nogawa N, Uwamino Y,
Shizuma K (1999) A method to detect low-level 63Ni activity for
estimating fast neutron Xuence from the Hiroshima atomic bomb.
Health Phys 76(6):635–638
24. Straume T, Marchetti AA, McAninch JE (1996) New analytical
capability may provide solution to the neutron dosimetry problem
in Hiroshima. Radiat Prot Dosimetry 67(1):5–8
25. Marchetti AA, Hainsworth LJ, McAninch JE, Leivers MR, Jones
PR, Proctor ID, Straume T (1997) Ultra-separation of nickel from
123
Radiat Environ Biophys (2007) 46:327–338
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
copper metal for the measurement of 63Ni by AMS. Nucl Instrum
Methods Phys Res B 123:230–234
McAninch JE, Hainsworth LJ, Marchetti AA, Leivers MR, Jones
PR, Dunlop AE, Mauthe R, Vogel JS, Proctor ID, Straume T
(1997) Measurement of 63Ni and 59Ni by accelerator mass spectrometry using characteristic projectile X-rays. Nucl Instrum
Methods Phys Res B 123:137–143
Knie K, Faestermann T, Korschinek G (1997) AMS at the Munich
Gas-Wlled Analyzing Magnet System GAMS. Nucl Instrum Methods Phys Res B 123:128–131
Knie K, Faestermann T, Korschinek G, Rugel G, Rühm W, Wallner C (2000) High-sensitivity AMS for heavy nuclides at the Munich Tandem accelerator. Nucl Instrum Methods Phys Res B
172:717
Rugel G, Faestermann T, Knie K, Korschinek G, Marchetti AA,
McAninch JE, Rühm W, Straume T, Wallner C (2000) Accelerator
mass spectrometry of 63Ni using a gas-Wlled magnet at the Munich
tandem laboratory. Nuclear Instrum Methods B 172:934
Rühm W, Knie K, Rugel G, Faestermann T, Marchetti AA, McAninch JE, Straume T, Wallner C, Korschinek G (2000) Detection of
63
Ni by means of accelerator mass spectrometry for estimating the
Xuence of fast neutrons from the Hiroshima atomic bomb. Health
Phys 79:358
Rugel G, Arazi A, Carroll KL, Faestermann T, Knie K, Korschinek G, Marchetti AA, Martinelli RE, McAninch JE, Rühm W,
Straume T, Wallner A, Wallner C (2004) Low level measurement
of 63Ni by means of accelerator mass spectrometry. Nucl Instrum
Methods Phys Res B223–223:776–781
Straume T, Rugel G, Marchetti AA, Rühm W, Korschinek G,
McAninch JE, Carroll K, Egbert S, Faestermann T, Knie K, Martinelli R, Wallner A, Wallner C, Fujita S, Shizuma K, Hoshi M,
Hasai H (2003) Measuring fast neutrons in Hiroshima at distances
relevant to atomic-bomb survivors. Nature 424, 539–542, and
(2004) Nature 430, 483
Hiroshima Peace Memorial Museum (1996) Architectural witnesses to the atomic bombing—a record for the future. City of
Hiroshima (in Japanese)
City of Nagasaki (1996) Record of exposed buildings and other
structures—Nagasaki 50 years A-bomb anniversary. City of
Nagasaki (in Japanese)
Young RW, Kerr GD (eds) (2005) Report of the Joint US–Japan
Working Group. Reassessment of the atomic bomb radiation
dosimetry for Hiroshima and Nagasaki: Dosimetry System DS02.
Radiation EVects Research Foundation, Hiroshima
Feldman GJ, Cousins RD (1998) UniWed approach to the classical
statistical analysis of small signals. Phys Rev D 57:3873
Rühm W, Rugel G, Faestermann T, Knie K, Wallner A, Heisinger
B, Nolte E, Marchetti AA, Martinelli RE, Carroll KL, Korschinek
G (2003) Cosmic-ray induced 63Ni—a potential confounder of
fast-neutron induced 63Ni in copper samples from Hiroshima. Eur
Phys J A17:633–639
Rose PF (ed) (1991) ENDF/B-VI summary documentation, crosssection evaluation group. Report BNL-NCS-17541, Nuclear data
Center, Brookhaven National Laboratory, Upton, NY