Nuclear Fission Research at IRMM Franz-Josef Hambsch EC-JRC-Institute for Reference Materials and Measurements, Retieseweg 111, B-2440 Geel, Belgium Abstract. The Institute for Reference Materials and Measurements (IRMM) will celebrate its 45th anniversary in 2005. With its 150-MeV Geel Electron Linear Accelerator (GELINA) and 7-MV Van de Graaff accelerator as multi-purpose neutron sources, it served the nuclear physics community for this period. The research in the field of nuclear fission was focused in recent years on both the measurement and calculation of fission cross sections, and the measurement of fission fragment properties. Fission cross sections were determined for 233Pa and 234U; the fission process was studied in the resolved resonance region of 239Pu(n,f) and for 251Cf(nth,f). These measurements derive their interest from accelerator driven systems, the thorium fuel cycle, high temperature reactors, safety issues of current reactors, and basic physics. The measurements are supported by several modeling efforts that aim at improving model codes and nuclear data evaluation.
Of importance for the success of the nuclear data measurements was the fact that the IRMM (CBNM) had a very well equipped Sample Preparation Group. Due to the difficult personnel situation, the supply of high-quality sample material is presently very much reduced and efforts towards improvement of this situation are under way
INTRODUCTION The Institute for Reference Materials and Measurements (IRMM) was founded in 1960 as the Central Bureau for Nuclear Measurements (CBNM) mentioned in the Euratom treaty. Since its creation, activities were dedicated to nuclear data measurements and, in particular, to the investigation of neutroninduced reactions.
This paper concentrates on nuclear fission research. Total and capture cross section measurements and activities in the field of neutron data standards are presented in several other contributions to this conference (see e.g., [1-3]).
The IRMM is well known for its very highresolution cross-section measurements. This is possible with the unique Geel Electron Linear Accelerator (GELINA) facility, delivering neutron beams with repetition rates up to 800 Hz and burst widths of less than 1 ns. The neutron energy is determined by the time-of-flight (TOF) technique with flight path lengths of up to 400 m.
In the following, recent examples of successful neutron-induced fission measurement campaigns and modeling efforts are given.
IRMM’s 7-MV Van de Graaff, solely used for neutron production via charged particle reactions, complements GELINA in the higher neutron energy region above 1 MeV and is primarily used for activation cross section, fission cross section, standards, and fission fragment property measurements.
FISSION CROSS SECTION MEASUREMENTS In the past, fission cross-section measurements were successfully performed for even very active isotopes, hence very difficult to be measured [4,5]. A special detector system had been developed to cope with intrinsic radioactivity rates of up to 107 s-1 [4].
With its two accelerators, the IRMM is very well equipped to cover today’s still most important neutron energy range up to 20 MeV.
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
902
In recent years the focus was more on fission cross sections of isotopes with very short half-lives (233Pa, T1/2 = 27 d) [6] and where the purity of the sample plays a very crucial role (234U, usually with contaminants of the fissile isotope 235U) [7].
explanation for the observed differences of the direct fission cross-section determination compared to the so-called surrogate method is given in [11]. As a side product of the 233Pa(n,f) cross section measurements, the 233U(n,f) cross section has also been measured in the same energy region. This was necessary for correcting the 233Pa fission cross section due to the buildup of 233U in the 233Pa sample and because shielding was used. Also, the evaluated libraries showed a 10% variation of this cross section at MeV incident neutron energies. In Fig. 2, the results are given, showing between 3 MeV and 6 MeV higher cross-section values as compared to the evaluations.
The 233Pa fission cross section was, for the first time, directly measured at our Van de Graaff driven neutron source [6]. The direct measurement of the fission cross section was not considered feasible in a recent IAEA report [8]. The isotope 233Pa is an important intermediate nuclide in a Th-U fuel-based reactor where the reactions involving 233Pa have a large impact on the balance of nuclei as well as the average number of prompt fission neutrons. The presence of 233Pa in a 232Th-fueled reactor also gives rise to what is known as the protactinium effect [9]. In the reactor, an equilibrated fraction of 233Pa in the fuel is reached after some time and, after a reactor stop, this fraction continues to decay into the fissile 233U. This increases the reactivity, potentially causing criticality. Hence, knowledge of the capture and fission cross section is important to be able to calculate the severity of such an effect. The resulting fission cross section measured in the incident neutron energy range from 1 to 8.5 MeV is shown in Fig. 1, together with the available evaluations, and compared to a recent indirect measurement of this cross section by Petit et al. [10].
FIGURE 2. The measured 233U(n,f) cross section2. The experimental data are compared to the ENDF/B-VI, JEFF-3.0, and JENDL-3.3 evaluations.
The case of the 234U(n,f) cross section is also connected with the Th-U fuel cycle. 234U is the capture product of the main fissile isotope 233U. Its existence leads to unwanted 232U production through the (n,3n) reaction. Moreover, it will contribute significantly to the long-lived radioactive waste from used fuel elements. Due to the lack of accurate experimental data, the commonly used evaluated data files for 234U show large discrepancies. In order to improve this situation, the fission cross section of 234U was measured at the thermal point at the Institute LaueLangevin (ILL) in Grenoble, France. FIGURE 1. The measured 233Pa(n,f) cross section1. The directly measured data from IRMM are compared to the ENDF/B-VI and JENDL-3.3 evaluations and the indirect measurement from [10].
The ILL measurements resulted in a thermal fission cross-section value of (67 ± 14) mb, as already reported [7]. This value has been adopted in a recent evaluation of this cross section by Maslov et al. [12].
It is obvious from Fig. 1 that the directly measured fission cross section is lower than both evaluations, as well as the indirect measurement of [10]. An
A high-resolution measurement of the fission cross section of 234U in the resonance region is being performed at GELINA. The aim is to cover the incident neutron energy region from 10 meV to
1
2
Figure 1 is reprinted from Nucl. Phys A733, No 1-2, F. Tovesson et al., “233Pa(n,f) cross section up to 8.5 MeV,” pages 3-19, (2004), with permission from Elsevier.
Figure 2 is reprinted from Nucl. Phys A733, No 1-2, F. Tovesson et al., “233Pa(n,f) cross section up to 8.5 MeV,” pages 3-19, (2004), with permission from Elsevier.
903
1 MeV. As detection system a Frisch-gridded ionisation chamber with very pure methane as detector gas is used.
FISSION FRAGMENT PROPERTY MEASUREMENTS
Figure 3 shows the resolved resonances in the incident neutron energy range from 100 eV to 2000 eV, nicely illustrating the intermediate structure in the fission cross section due to the double-humped fission barrier of 234U. For the measurement, a highly enriched (99.868 at %) 234U sample was used, with only 0.076 at % 235U. The purity is of crucial importance since the fission cross section of 235U is orders of magnitude higher than the one of 234U.
Fission fragment property measurements are another domain, where the IRMM has a renowned reputation. In the past couple of years fission fragment mass and total kinetic energy distributions have been measured for resonance-neutron induced fission of 239 Pu, without and with coincidence of prompt neutron emission, for spontaneous fission of 252Cf and for neutron induced fission of 251Cf at thermal incident neutron energy. Resonance neutron induced fission of 235U has revealed fluctuations of the fission fragment mass yield and total kinetic energy (TKE) from resonance to resonance [14]. The TKE was found to be anticorrelated with the number of prompt neutrons emitted in fission. In addition, a fission mode analysis in the frame of the multi-modal random neck-rupture model [15] showed fluctuations in the modal weights from resonance to resonance. In the case of 239Pu(n,f), a cleaner situation was expected due to the fact that well separated 0+ and 1+ transition states above the barrier are present.
10 234
U(n,f)
σf [b]
8 6 4 2 0 100
En[eV]
At resolved resonances below 200 eV the variation of the average TKE and mass yield distribution for 239 Pu(n,f) as a function of resonance energy has been investigated. Figure 5 shows the fluctuation of the TKE with resonance energy. The deviation from the thermal value of (177.83 ± 0.02) MeV is small (about 300 keV) but visible. Error bars are rather large, and only statistical errors are given. Nevertheless, it is observed that the deviation of the average TKE in resonances is only about half the deviation as being
1000
FIGURE 3. The first cluster of intermediate resonances in the 234U(n,f) cross section.
Figure 4 shows a comparison of the new measurement results with those of James et al. [13] for a part of the first resonance cluster. Obvious is the superior energy resolution of the GELINA measurement. 8
178.8
IRMM 2003 ORNL 1977
6
π
[MeV]
178.4
σf [b]
4
2
0 550
+
J =1 π + J =0 th
178.6
178.2 178.0 177.8 177.6 177.4
560
570
580
177.2 -20
590
En [eV]
0
20
40
60
80 100 120 140 160 180 200 220
En [eV]
FIGURE 4. Comparison of the new measurements performed at IRMM (histogram) with data of James et al. (full line) [13].
FIGURE 5. Mean TKE as a function of resolved resonance energy in 239Pu(n,f) compared to the thermal value.
904
observed in resonance fission of 235U [14]. The shape is very similar, except that the maximum is at about twice the incident neutron energy (see [14]).
0.07 0.06
relative mass yield
Figure 6 shows the changes in the fission fragment mass yield distribution relative to the thermal value for the sum of all Jπ = 1+ resonances. Again, a much smaller effect as compared to 235U is observed. A possible explanation could be the following: in case of a single transition state leading to fission with a given modal structure, no fluctuations in the fission fragment properties should be observable in contrast to several transition states with individual modal structure present in 236U. The observed fluctuations point to the fact that more than one transition state leads to fission for some of the resonances.
Cf(SF)
0.05 0.04 0.03 0.02
0.00 80
90
100
110
120
A [amu]
FIGURE 7. Comparison of the mass distribution for the light fragments for the system 251Cf(nth,f) and 252Cf(SF). An evaluation for this mass distribution is also given by the full line and shaded area [16].
0.5
multiplicities and spectra were determined for several actinides [17,18]. The most important improvement to the statistical model code used to calculate the fission cross section was the implementation of the multi modality of the fission process. The multi-modal fission process was also included in an improved version [18] of the Los Alamos code [19] used in the calculation of prompt neutron multiplicities and spectra.
0.4 0.3 0.2 0.1 0.0 -0.1
π
+
252
0.01
0.6
ΣiYi(J =1 ) − Y(thermal) [%]
England evaluation Cf(nth,f)
251
-0.2 -0.3 -0.4 70
80
90
In the recent calculation of the prompt fission neutron spectrum of 252Cf(SF) [20], the fission fragment residual nuclear temperature distribution had also been modified and an anisotropy in the prompt neutron emission has been assumed. In Fig. 8 the superposition of the experimental mass-yield and TKE distribution is shown, which entered into the prompt fission neutron spectrum (PFNS) calculation. Since the mass distribution of 252Cf(SF) is rather broad, the normally sufficient number of three fission modes for the superposition did not work out here. Two additional theoretically predicted modes, the Standard 3 and Standard X modes, had to be introduced. With a total of five fission modes, the mass-yield and TKE distributions as well as the prompt neutron evaporation could be described consistently (as seen in Figs. 8 and 9).
100 110 120 130 140 150 160 170
[amu] FIGURE 6. Difference of the mass yield distribution for the sum of resolved 1+ resonances compared to thermal mass yield distribution.
The thermal neutron induced fission of 252Cf* has been investigated at the LOHENGRIN spectrometer of ILL, Grenoble, France, via the reaction 251Cf(nth,f) at E* = 6.2 MeV. This is the heaviest system, in which a comparison is possible between a thermal neutron induced and a spontaneously fissioning system. Figure 7 shows the mass distribution deduced for the 251Cf(nth,f) reaction in comparison with 252Cf(SF) and an evaluation of England et al. [16]. A definite difference is obvious and is probably related to a change in the modal weights of the so-called standard modes S1 and S2. A more elaborated analysis is presently ongoing.
Figure 9 gives the calculated PFNS relative to a Maxwellian distribution with a fixed temperature, TM = 1.42 MeV. Additionally, the present calculation is compared to the evaluation by Mannhart [21]. The full and dash-dotted lines show the improvement in the agreement of the presently calculated spectrum with the Mannhart evaluation if the improved temperature dependence of the neutron emission and
905
1
252
IRMM exp.data
Cf(SF) SX
0
10
S2
Y (%)
Ratio to Maxwellian TM = 1.42 MeV
10
S1
-1
10
SL
S3
-2
10
-3
10
80
90
100
110
120
130
140
150
160
170
1.4 Manhart eval.
(1987)
1.0
0.8
Madland (Nucl.Sci.Eng.'82) C=9.15 multi-modal calculations: b=0.10, s=1.3 b=0.15, s=1.3
0.6 0.01
0.1
1
10
E (MeV)
FIGURE 9. The calculated PFNS relative to a Maxwellian distribution with TM = 1.42 MeV (full and dash-dotted line) compared to the evaluation of Mannhart [21]. The calculation of Madland et al. [19] (dashed line) is also shown.
200 252
Cf(SF)
190
TKE (MeV)
Cf (SF)
1.2
A
180
252
(1989)
Maerten et al. (1990) EXFOR sub.acc.num.: , 22219003,005 , 30969002,003
S1
SX
S2
170
In the meantime, fission cross-section calculations have also been performed for higher fission chances up to several tens of MeV, e.g., for 235U and 231,233Pa.
SL S3
160
In Fig. 10, as one example, the result of the fission cross-section calculation for 235U(n,f) up to 50-MeV incident neutron energy is given. The present calculation is compared to the available evaluations existing only up to 20 MeV. Also the contributions of the different fission chances involved are given, again compared to the results from previous evaluations. Although the new fission cross-section calculation is in good agreement with the evaluations, the first- and second-chance fission cross-section contributions appear to be quite different.
150 IRMM experimental data
140 130
140
150
160
170
A
FIGURE 8. Superposition of fission modes and experimental distributions Y(A) and TKE(A) for 252Cf(SF).
an anisotropy of 10% to 15% in the angular distribution of the prompt emitted neutrons is assumed. This anisotropy would point to the fact that the source of prompt neutron emission is not purely from fully accelerated fragments, but other sources such as scission neutron emission could exist. The search for scission neutrons was recently again in the focus, since a reanalysis of the measured prompt neutron emission in 252Cf(SF) [22], showed some evidence that this process is really occurring. At present an INTAS project has been approved for funding, which is partly related to the search for scission neutrons.
4.0
EXFOR eval.based on exp.data: Konshin Liu Ji-Cai EXFOR: Kari(KFK) Smith(USALAS) Grundl(USALAS) Czirr(USALRL) Poenitz(USAANL)
Fission cross-section (b)
3.5
Last but not least, the calculation of the fission cross section and other reaction cross sections using a statistical model code [17] is being described.
3.0
ENDF/B-VI rev.5 (nov.89) and JEF3.0 JENDL3.3 rev.4 (feb.02) only MT=18 is given
2.5
new evaluation
n+
235
U
2.0 1.5 1.0 0.5 0.0 0
5
10
15
20
25
30
35
40
45
50
En (MeV)
For the first-chance fission, the statistical model code has been improved taking into account the modality of the fission process. Based on modal weights determined from the experimental fission yield and TKE distributions for several actinides, the fission cross section and the other reaction cross sections involved have been calculated for the same actinides investigated experimentally (see, e.g., [17]).
FIGURE 10 The fission cross section up to 50 MeV for 235 U compared to EXFOR data (symbols) and ENDF/B-VI (dotted lines). The dashed and dotted lines show the contributions of the different fission chances involved.
906
7. C. Wagemans, J. Wagemans, O. Serot, Nucl. Sci. Eng. 141 (2002) 171. 8. B.D. Kuzminov, V.N. Manokhin, Nuclear Constants, Issue No. 3-4,41 (1997), IAEA INDC(CCP)-416 (1998). 9. C. Rubbia, J.A. Rubbio, S. Buono et al., CERN Report No. CERN/AT/95-44(ET) (1995). 10. M. Petit, M. Aiche, G. Barreau et al., Nucl. Phys. A735 (2004) 345-371. 11. G. Vladuca, F.-J. Hambsch et al., Nucl. Phys. A740 (2004) 3-19. 12. V. Maslov, Yu.V. Porodzinskij et al., INDC(BLR)-017 (2003). 13. G.D. James, J.W.T. Dabbs, J.A. Harvey, N.W. Hill, Phys. Rev. C15 (1977) 2083. 14. F.-J. Hambsch, H.-H. Knitter, C. Budtz-Jorgensen, J.P. Theobald, Nucl. Phys. A491 (1989) 56-90. 15. U. Brosa, S. Grossmann, A. Müller, Phys. Rep. 197 (1990) 167-262. 16. T.R. England, B.F. Rider, „Evaluation and compilation of fission product yields“, ENDF-349 (1992). 17. G. Vladuca, A. Tudora, F.-J. Hambsch, S. Oberstedt, Nucl. Phys. A707 (2002) 32-46. 18. F.-J. Hambsch, S. Oberstedt, G. Vladuca, A. Tudora, Nucl. Phys. A709 (2002) 85-102. 19. D. G. Madland, J. R. Nix, Nucl. Sci. Eng. 81 (1982) 213. 20. F.-J. Hambsch, A. Tudora, G. Vladuca, S. Oberstedt, to be published. 21. W. Mannhart, “Status of the 252Cf fission neutron spectrum evaluation with regard to recent experiments”, IAEA INDF 220 (1989) 305-336. 22. N. Kornilov, A. B. Kagalenko et al., Nucl. Phys. A686 (2003) 187-203.
CONCLUSIONS Nuclear fission research is still a major part of the nuclear data work program at IRMM. The experimental results are used in improvements of model codes for calculations of total reaction cross sections, including the fission channel, as well as for the calculation of the prompt fission neutron multiplicities and spectra. The work is carried out in collaboration with European and non-European laboratories and with participation of fellows at graduate and undergraduate level. Many visiting scientists from New Member States and Candidate Countries to the European Union have participated in the work program.
REFERENCES 1. A. J. P. Plompen, contribution to this conference. 2. F.-J. Hambsch, A. D. Carlson, H. Vonach, contribution to this conference. 3. G. Giroginis, V. Khriatchkov, contribution to this conference. 4. H.-H. Knitter, C. Budtz-Jorgensen, Atomkernenergie 33 (1979) 205-212 5. C. Budtz-Jorgensen, H.-H. Knitter, Nucl. Sci. Eng. 79 (1981) 380-392. 6. F. Tovesson, E. Birgersson, M. Fleneus, et al., Nucl. Phys. A733 (2004) 3-19.
907
