Reich-Moore Analysis of the 127I and 129I
Resolved Resonance Range
G. Noguère1, A. Brusegan2, A. Leprêtre3, N. Herault3, O. Bouland1,
and G. Rudolf4
1
CEA Cadarache, F-13108 Saint Paul Lez Durance, France
2
EC-JRC-IRMM,B-2440 Geel, Belgium
3
CEA Saclay, F-91191 Gif sur Yvette, France
4
IReS, F-67037 Strasbourg, France
Abstract. High-resolution transmission and capture Time-Of-Flight (TOF) measurements of 127I and 129I have been
carried out at the 150-MeV pulsed neutron source GELINA of the Institute for Reference Materials and Measurements
(IRMM). Below 10 keV, the resonance parameters were extracted from the resolved resonance energy range of the TOF
data with the REFIT and SAMMY shape analysis codes. In a last step, valuable s-wave mean radiation width <Γγ >,
strength function S0, and level density D0 were deduced from the statistical analysis of the new resonance parameters.
radioactive iodine, extensive measurements of 127I
have been carried out under the same experimental
conditions as for the 129I. The data-reduction procedure
was performed with the AGS system [1], and the
resonance parameters were extracted with the
SAMMY [2] and REFIT [3] shape analysis codes.
INTRODUCTION
Iodine 129 is a long-lived fission product. With a
very long β- half-life of 1.57×10-7 years, iodine
requires disposal strategies that will isolate this isotope
from the environment for very long periods of time.
However, iodine is considered difficult to isolate as it
easily moves throughout the ecosystem. Thereby, 129I
is potentially a key long-lived fission product for P&T
applications: after one single neutron capture, 129I
transmutes in 130I and decays to the noble gas 130Xe
with a 12.36-h half-life.
The lack of reliable data on neutron cross sections
is one of the major problems faced in evaluating the
transmutation potential of 129I in appropriate reactor
configurations. Improvements of the capture cross
section within the thermal and epithermal energy
ranges would help to reduce uncertainties in waste
management concepts. For that purpose, Time-OfFlight measurements covering the [0.5 eV-100 keV]
energy range have been carried out at the 150-MeV
pulsed neutron source GELINA of the Institute for
Reference Materials and Measurements (IRMM). Two
types of experiments have been performed, namely
capture and transmission, respectively related to the
neutron capture and total cross sections. Since the PbI2
samples used in this work contain natural and
SAMPLE PREPARATION
The ideal target material would have been
elemental iodine, but after a few tries this was
considered not to be practical in view of the
radiological hazard, volatility, and reactivity of I2. The
PbI2 compound was considered a reasonable
compromise, with no lead resonances in the resolved
resonance energy range of the iodine isotopes. The
thicknesses in atom per barn of the samples prepared
at the IRMM are given in Table 1.
Four samples containing commercial Pb127I2
powder were prepared together with a 1-mm-thick LiI
liquid sample. For the samples containing 129I, the
iodine was extracted at the IRMM from 210 l of waste
solution provided by the French fuel reprocessing
facility of La Hague [4]. The final radioactive PbI2
samples exhibited a complex composition involving
CP769, International Conference on Nuclear Data for Science and Technology,
edited by R. C. Haight, M. B. Chadwick, T. Kawano, and P. Talou
© 2005 American Institute of Physics 0-7354-0254-X/05/$22.50
1462
TABLE 1. 127I and 129I thicknesses in atoms per barn of the samples.
127
Experiment
Sample
I Thickness (at/b)
127
I transmission
Thin PbI2 sample
(7.089 ± 0.998) ×10-4
LiI target
(9.451 ± 0.106) ×10-4
Thick PbI2 sample
(2.380 ± 0.040) ×10-2
127
I capture
Thin PbI2 sample
(9.476 ± 0.223) ×10-4
Thick PbI2 sample
(4.764 ± 0.086) ×10-3
129
I transmission
Thin PbI2 sample
(1.468 ± 0.044) ×10-4
Thick PbI2 sample
(1.465 ± 0.044) ×10-3
129
I capture
Thin PbI2 sample
(1.564 ± 0.047) ×10-4
Thick PbI2 sample
(7.897 ± 0.235) ×10-4
129
I Thickness (at/b)
(7.091 ± 0.174) ×10-4
(7.076 ± 0.173) ×10-3
(7.556 ± 0.184) ×10-4
(3.814 ± 0.091) ×10-4
natural and radioactive iodine − (3.36±0.08)% and
(16.49±0.40)%, respectively − together with a nonnegligible amount of natural lead (53.5±3.0)%, sulphur
(6.2±0.4)%, sodium (0.75±0.04)%, hydrogen ~0.09%,
oxygen (14.5±1.5)%, and nitrogen (1.2±0.4)%. The
global accuracies on the 127I and 129I mass fractions
were close to 2.4%.
monitored by the acquisition system. As the
experiments took several weeks, the sequence had to
be short enough to reduce the systematic uncertainties
due to a possible instability of the accelerator and to
any variations of the sample temperature.
EXPERIMENTAL TECHNIQUES
A transmission experiment consists of measuring
the attenuation of the incident neutron beam impinging
on a sample of thickness n atoms per barn. The
experimental transmission T(E) is then expressed as
the ratio of the transmitted neutron flux φin(E) to the
incident flux φout(E):
φ in ( E )
a in ( E ) N in ( E ) − B in ( E )
(1)
=NT
T (E )= N T
φ out ( E )
a out ( E ) N out ( E ) − B out ( E )
DATA REDUCTION
The transmission and capture experiments on
iodine were performed at the Geel Electron LInear
Accelerator (GELINA) of the Institute for Reference
Materials and Measurements. The Measurements were
founded on the Time-Of-Flight (TOF) method.
The neutron source of the GELINA facility
consists of a mercury-cooled uranium target. A short
burst of fast neutrons is produced via (γ,n) and (γ,f)
reactions. To enhance the neutron spectrum intensity
below a few hundred kilo-electron Volts, two 36-mm
water slabs in beryllium canning are placed as close as
possible above and below the uranium target. The
neutrons emitted from the target-moderator assembly
are led to the experimental areas along neutron flight
paths.
In our transmission measurement, the neutrons
crossed a filter setup and the sample at a distance of
23.7 m. Neutrons were detected at 49.34 m with a
1/4″-thick Li-glass (NE912) viewed by a 5″ EMI 9823
KQB photomultiplier. The normalisation of the
transmission was provided by four BF3 tubes located
in the concrete wall above the uranium target.
For our capture experiment, the sample was located
at 28.67 m from the neutron source. The neutron flux
was measured with a 10B ionisation chamber, while
capture γ-rays emitted from the sample were detected
by two C6D6 liquid scintillators placed at the right
angle to the beam axis.
The capture and transmission experiments
consisted of a sequence of sample changer positions
in which NT stands for the normalisation factor, ain(E)
and aout(E) are the dead time corrections, Nin(E) and
Nout(E) represent the measured count rates, and Bin(E)
and Bout(E) are the estimated background.
Neutron capture cross sections are measured by
detecting γ-rays emitted when the compound nucleus
de-excites to its ground state. In practice we have
measured the capture yield Y(E) defined as the number
of capture events Nc(E) per incident neutron φ(E). The
major requirement for a neutron radiative capture
detector is that its efficiency be independent of the
capture γ-ray spectrum (εc=cst). This result can be
achieved by using a total energy detector that has the
property of making the probability of detecting a
capture event proportional to the excitation energy E*
of the compound nucleus. This condition can be
satisfied by weighing the pulse amplitude response of
the C6D6 scintillators with an appropriate weighting
function W(I) expressed as a function of the γ-ray
energy I. The experimental capture yield is defined as:
1463
Y (E )= N Y
N c(E )
φ (E )
a ( E )∑ N γ ( I , E ) W ( I ) − B γ ( E ) B
= NY
σ n ,α ( E )
E*
a (E )N F (E )− B F (E )
D(E)⊗σγ,i(E) stand for the total and capture Doppler
broadened cross sections. Examples of data and
theoretical curves are shown in Fig. 1.
The 127I and 129I resonance parameters obtained in
this work are reported in the JEFDOC-986 available
from the OECD/NEA databank. The respective
assignments of 719 and 400 resonances up to 10 keV
significantly improves the knowledge of the Resolved
Resonance Range of the iodine isotopes. The present
set of 129I resonance parameters is the first to be
extracted over a wide energy range with modern shape
analysis techniques. Concerning the resolved
resonance parameters, the sources of systematic errors
are mainly the sample thickness, the incomplete
description of the resolution function, and the timedependent background. From the combination of the
above-mentioned effects, we may assign on the 127I
and 129I neutron width values an average systematic
error of about 3.3% and 3.6%, respectively.
(2)
in which NY represents the normalisation factor, a(E) is
the dead time correction, Nγ(I,E) is the count rate
observed at a given pulse height energy, NF(E) stands
for the count rate delivered by the boron chamber,
σBn,α(E) is the 10B(n,α) cross section, and Bγ(E) and
BF(E) are the estimated background contributions. In
this work, we have used the experimental weighting
function measured at the IRMM by Corvi et al. [5].
The data-reduction procedure was performed using
a code called AGS [1]. The AGS system deals with
basic operations such as spectrum addition or division,
dead time correction, nonlinear fitting, and several
subsidiary operations, preserving all the steps in a
single file together with full uncertainty propagation.
A large number of data sets covering the energy range
[0.5 eV-100 keV] were obtained, together with their
covariance matrices.
RESONANCE SHAPE ANALYSIS
The SAMMY and REFIT codes adjust nuclear
parameters so that the theoretical curve agrees with the
observed data within the limit of the uncertainties. In
the Resolved Resonance Range, the nuclear
parameters are the energies (Eo), the partial widths
(Γn,Γγ), the total angular momentum (J), the relative
orbital momentum (l), and the effective radii (R'). The
cross sections are generated using the Reich-Moore
approximation of the R-Matrix formalism [6] and are
Doppler broadened using the ideal Free Gas Model
with a given effective temperature Teff [7]. Point-wise
distributions provided by Monte-Carlo simulations
were introduced in the resonance shape analysis to
described the experimental resolution R(E) of the
GELINA facility [8]. Theoretical curves are given by:
∞
T ( E ) = ∫ R ( E , E ') T th ( E ') dE '
(3)
0
Y (E ) =
∞
∫ R ( E , E ' )Y
th
( E ' ) dE '
(4)
0


T th ( E ) = exp  − ∑ n i D ( E ) ⊗ σ t , i ( E ) 
i


n
D
(
E
)
⊗
σ
(
E)
γ
i
,
i
∑i
Y th ( E ) = µ ( E ) (1 − T th ( E ) )
n
i
D
(
E
)
⊗
σ
t
,
i
(
E
)
∑i
(5)
FIGURE 1. Examples of 127I and 129I resonance peaks
measured with the thin capture sample and the thick
transmission sample. The solid lines are the corresponding
adjusted theoretical curves. The energies of the 129I
resonances are displayed in the figure.
(6)
in which ni defines the sample thicknesses, µ(E) is the
multiple scattering correction, D(E)⊗σt,i(E) and
1464
provided a confident s-waves sample together with
relevant prior average parameters. Mean level
spacings, neutron strength functions, and mean
radiation widths are crucial prior parameters in
analysing the unresolved energy range of the iodine
neutron cross sections, to determine consistent pointwise cross sections above 100 keV, and to improve the
Reference Input Parameter Library (RIPL-2) around
the compound nucleus mass numbers 128 and 130.
PRIOR AVERAGE PARAMETERS
Information on average parameters can be obtained
from statistical analyses of resolved resonance
parameters. A careful determination of prior 129I
average parameters has never been performed since
the resonance parameters available in the ENDF files
were only deduced from the capture areas
Aγ=gΓnΓγ/(Γγ+Γn) reported by Macklin [9]. Herein, the
prior average parameters for 127,129I have been deduced
from a confident s-wave resonances sample suggested
by the Bayesian treatment implemented in the
ESTIMA code [10].
The <Γγ> mean radiation widths given in Table 2
are the weighted average values of the Γγ values
determined from our simultaneous analysis of the
transmission and capture data. We have obtained the
127
I and 129I radiation widths for, respectively, 155 and
55 resonances. The quoted uncertainties are the
standard deviations of the distributions of the fitted
values. For determining D0 and S0, the estimation
procedure in the ESTIMA code takes into account the
missing weak levels and uses a method based on the
Porter-Thomas integral distribution properties. The
uncertainties reported in Table 2 include fitting and
sampling errors as explained in the ESTIMA user’s
manual.
ACKNOWLEDGMENTS
The authors wish to express their appreciation for
the work of C. Ingelbrecht and also to his sample
preparation staff. Special thanks go to the IRMM
Linac operators for providing the neutron beam for the
TOF measurements. We also express our gratitude to
N.M. Larson and M. Moxon for valuable discussions
and their relevant advice during this work.
REFERENCES
1. Bastian C., “AGS, a set of UNIX commands for neutron
data reduction,” Proc. Int. Conf. on Neutrons in Research
and Industry, Crête, Greece (1979).
2. Larson N.M., RISC peripheral shielding routine
collection, SAMMY-M6, Oak Ridge National
Laboratory, ORNL/TM-9179/R6 (2003).
3. Moxon M.C., Brisland J.B., “REFIT, a least square
fitting program for resonance analysis on neutron
transmission and capture data,” Technical Report
Harwell laboratory, CBNM/ST/90-131/1 (1990).
4. Ingelbrecht C., Lupo J., Raptis K., Altizouglou T.,
Noguere G., “Iodine-129 targets for nuclear waste
transmutation studies,” Nucl. Instrum. Methods Phys.
Res. A 480, 204 (2002).
5. Corvi F., Fioni G., Gasperini F., Smith P.B., “The
weighting function of a neutron capture detection
system,” Nucl. Sci. Eng. 102, 272 (1991).
6. Lane A.M., Thomas R.G., “R-Matrix theory of nuclear
reactions,” Rev. Mod. Phys. 17, 350 (1960).
7. Lamb W.E., “Capture of neutrons by atoms in a crystal,”
Phys. Rev. 55, 190 (1939).
8. Coceca C., Magnani M., “Resolution rotary target”
Internal Report IRMM, GE/R/ND/06/96 (1996).
9. Macklin R.L., “Neutron capture cross sections and
resonances of iodine-127 and iodine-129,” Nucl. Sci.
Eng. 85, 350 (1983).
10. Fort E., Doat J.P., “ESTIMA: a code to calculate average
parameters from sets of resolved resonance parameters,”
NEANDC-161(1983).
TABLE 2. 127I and 129I average resonance parameters
deduced from the statistical analysis of the resonance
parameters. The mean level spacing D0 is in eV, the mean
radiation width <Γγ> is in meV, and the dimensionless
parameter S0 is the neutron strength function.
Parameters
Iodine 127
Iodine 129
104S0
0.72 ± 0.06
0.54 ± 0.07
D0
12.5 ± 0.3
27.3 ± 0.9
100.0 ± 22.6
106.0 ± 15.2
<Γγ>
CONCLUSION
Brand new experimental 127I(n,γ), 127I(n,tot),
I(n,γ), and 129I(n,tot) cross sections measured at the
IRMM have been analysed within the 0.5 eV to
10 keV energy range. The systematic errors affecting
the accuracy of the present results have been
determined using sensitivity studies. For the key
isotope of the present work, the error on the 129I
capture cross section ranges from 3% to 14%.
The statistical analysis of the resolved resonances
with a method relying on the Bayes’ theorem has
129
1465