Development of photonuclear cross section library for CINDER’90

GIACRI Marie-Laure Theoretical division (T-16), Los Alamos National Laboratory, University of California, Los Alamos, NM 87545, USA DSM/DAPNIA/SPhN, CEA Saclay, F-91191 Gif/Yvette, FRANCE [email protected]

ABSTRACT The major goal of this study is to develop a photonuclear cross section library for the CINDER’90 evolution code using the existing evaluations from the IAEA library and calculations with HMS-ALICE and GNASH. Detailed calculations with GNASH of photonuclear cross sections have been also done for 235 U, 237 Np and 239 Pu. Emission of delayed and prompt neutrons have also been studied.

1

Contents 1 Introduction

6

2 Projects on photonuclear reactions 2.1 Detecting delayed neutrons . . . . . . . . . . . . . . . . 2.2 Burning nuclear waste . . . . . . . . . . . . . . . . . .

7 7 7

3 HMS-ALICE 3.1 Description . . . . . . . . . 3.2 Results . . . . . . . . . . . . 3.2.1 Heavy non-fissionable 3.2.2 Light nuclei . . . . . 3.2.3 Actinides . . . . . . 3.3 Conclusion . . . . . . . . . . 4 GNASH 4.1 235 U . . . . . . . . 4.2 237 Np . . . . . . . . 4.2.1 (γ,abs) cross 4.2.2 Results . . . 4.3 239 Pu . . . . . . . . 4.3.1 (γ,abs) cross 4.3.2 Results . . . 4.4 Conclusion . . . . .

. . . . . . . . nuclei . . . . . . . . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

9 9 9 9 11 11 14

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

16 16 17 17 19 20 20 20 21

5 CINDER’90 5.1 Results . . . . . . . . . . . . . 5.1.1 Transmutation of 90 Sr 5.1.2 Transmutation of 93 Zr 5.1.3 Transmutation of 137 Cs 5.2 Work left . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

22 22 22 24 26 26

6 Photofission neutron emission 6.1 Simulation of prompt neutron emission . . . . . . . . . 6.2 Delayed neutrons sources . . . . . . . . . . . . . . . . .

28 28 28

7 Conclusion

31

8 Acknowledgement

32

. . . . . . . . . . section . . . . . . . . . . section . . . . . . . . . .

2

A Transmutation of

90

B Transmutation of

137

C Transmutation of

93

Sr

35

Cs

36

Zr

38

3

List of Figures 1 2

3 4

5

6 7 8 9 10 11

12 13 14 15 16 17

Preliminary results of 237 Np incineration by fission. . . Photoneutron production cross sections (γ,1n) and (γ,2n) obtained with HMS-ALICE for 208 Pb and compared with IAEA evaluation (solid lines). . . . . . . . . . . . . . . Photoneutron production cross sections (γ,1n) and (γ,2n) obtained with HMS-ALICE for 181 Ta and compared with IAEA evaluation (solid lines). . . . . . . . . . . . . . . Particle production cross sections (γ,1n) and (γ,1p) obtained with HMS-ALICE for 12 C and compared with IAEA evaluations (solid lines). . . . . . . . . . . . . . . Particle production cross sections (γ,1n) and (γ,1p) obtained with HMS-ALICE for 27 Al and compared with IAEA evaluations (solid lines). . . . . . . . . . . . . . . Photonuclear cross sections calculated with HMS-ALICE for 235 U compared with IAEA evaluations (solid lines). Photonuclear cross sections calculated with HMS-ALICE for 238 U compared with IAEA evaluations (solid lines). Results of HMS-ALICE for photoabsorption cross section compared with IAEA evaluation. . . . . . . . . . . Results of GNASH calculation for 235 U compared with the experimental data (see legend for details). . . . . . Integrated photoabsorption cross sections for different actinides as a function of mass A. . . . . . . . . . . . . Photoabsorption cross section for 237 Np : both data and different parametrizations1 are plotted. Intermedaite curve giving integral of 3.54 . . . . . . . . . . . . . . . Results of GNASH for 237 Np in the case of partial photonuclear cross sections . . . . . . . . . . . . . . . . . . Photoabsorption cross section for 239 Pu . . . . . . . . . Results of GNASH for 239 Pu in the case of partial photonuclear cross sections . . . . . . . . . . . . . . . . . . Transmutation of 90 Sr. Left part is the irradiation at 1017 γ/cm2 /s and right part is natural decay. . . . . . . Flux dependence for transmutation of 90 Sr with CINDER’90 . . . . . . . . . . . . . . . . . . . . . . . . . . Transmutation of 93 Zr. Left part is the irradiation at 1017 γ/cm2 /s and right part is natural decay. . . . . . . 4

8 10

10

12

12 13 13 15 17 18

18 19 20 21 23 23 25

18 19 20 21 22 23 24 25 26 27 28 29 30

Flux dependence for transmutation of 93 Zr with CINDER’90 . . . . . . . . . . . . . . . . . . . . . . . . . . Transmutation of 137 Cs. Left part is the irradiation at 1017 γ/cm2 /s and right part is natural decay. . . . . . . Flux dependence for transmutation of 137 Cs with CINDER’90 . . . . . . . . . . . . . . . . . . . . . . . . . . νp (Eγ) for 235 U, 238 U, 237 Np and 239 Pu. . . . . . . . . . 90 Sr : Flux 1017 γ/cm2 /s . . . . . . . . . . . . . . . . . 90 Sr : Flux 1018 γ/cm2 /s . . . . . . . . . . . . . . . . . 90 Sr : Flux 1019 γ/cm2 /s . . . . . . . . . . . . . . . . . 137 Cs : Flux 1017 γ/cm2 /s . . . . . . . . . . . . . . . . . 137 Cs : Flux 1018 γ/cm2 /s . . . . . . . . . . . . . . . . . 137 Cs : Flux 1019 γ/cm2 /s . . . . . . . . . . . . . . . . . 93 Zr : Flux 1017 γ/cm2 /s . . . . . . . . . . . . . . . . . 93 Zr : Flux 1018 γ/cm2 /s . . . . . . . . . . . . . . . . . 93 Zr : Flux 1019 γ/cm2 /s . . . . . . . . . . . . . . . . .

5

25 27 27 29 35 35 36 36 37 37 38 38 39

1

Introduction

The major problem in studying photonuclear reactions is the lack of experimental data and corresponding evaluations. The only evaluations available come from the IAEA photonuclear library [IAE00] where cross sections are given only for 164 isotopes. A lot of projects encounter that problem. Some of these projects are presented in section 2. The solution we choose is to use simulation tools to calculate needed photonuclear cross sections. For this purpose we use HMS-ALICE and GNASH codes. The goal of this work is to develop a photonuclear cross sections library for CINDER’90 (section 5) material evolution code, which will allow to perform material activation analysis based on photonuclear reactions. For example, we will be able to simulate transmutation of nuclear waste in a photon flux or to evaluate the background level in typical photonuclear experiments.

6

2 2.1

Projects on photonuclear reactions Detecting delayed neutrons

Detecting 235 U and 239 Pu at the borders is an important goal of the US homeland defense program. A study has been launched at LANL [Lit03] to develop a device able to detect these materials using delayed neutrons. This device should detect special nuclear material (SNM) and be easily used by non-specialist persons. The photons will be created by Bremsstrahlung using a 8 MeV electron accelerator. They will induced photofission on actinides. The delayed neutron flux will be measured and analyzed to detect SNM. For the above project one should need an extended photonuclear cross section library for actinides including photofission and fission yields. In addition, some of the photonuclear reactions might result in delayed neutrons emitters, which will contribute to the unwanted background. All the reactions should be indentified providing corresponding photonuclear cross sections. Another project using delayed neutrons has been launched at the CEA [Saf03]. Delayed neutrons will be used for a non-destructive characterization of nuclear waste barrels. With this method the proper end cycle for each nuclear waste barrel could be chosen. The LINAC SAPHIR, presently available at CEA Saclay, will be used for feasibility experiments. It can deliver electrons from 10 to 30 MeV with a step of 5 MeV. Within this project, some experiments will be done to define the accelerator parameters. In addition, a number of fundamental physics experiments will improve knowledge about delayed neutron data including their yields, energy and time characteristic. Finally this investigation is supposed to develop into a large scale industrial facility to examine all barrels containing nuclear waste in the CEA laboratory. 2.2

Burning nuclear waste

Using photonuclear reactions for transmutation of actinides and long life fission products might be an option which should not be neglected as discussed in more detail in Refs [Mat88, Ben02, Gia03].

7

1.0

Quantity of remaining

237

Np

0.8

0.6

0.4

0.2

0.0 12.0

Thermal neutrons Fast neutrons Photons

14.0

16.0 Log(flux)

Figure 1: Preliminary results of

237 Np

18.0

20.0

incineration by fission.

A preliminary simulation of 3 month-irradiation of 237 Np in different neutron and photon flux has been done. The photon energy was equally distributed between 10 and 20 MeV. The uncertainty on cross sections (about 50%) lead to uncertainty on transmutation rate. That’s why the value for transmutation rate is between the two dashed lines is presented on figure 1. The maximal flux of thermal neutrons is available at Grenoble (France) and is of the order of 1015 n/cm2 /s. This would allow a transmutation of 11% of the 237 Np, which can be obtained with a flux of photons at 1017 γ/cm2 /s. With a flux of 1018 γ/cm2 /s we can reasonably hope to transmute at least 75% within 3 months. However such a flux is not easy to obtain with today’s technology. fast neutrons show even worse performance than thermal ones. Calculation has been done with ORIGEN2 by adding manually average cross sections for each nuclide belonging to the transmutation chain (˜ 20 actinides). After completion of the full library for CINDER’90 similar simulations will be done more easily and we will be able to improve these data by using a more physical flux energy spectrum and hopefully more accurate cross section evaluations.

8

3

HMS-ALICE

HMS-ALICE is a powerful multiparticle reaction code to calculate quickly cross sections for every nuclide at any energy from a few MeV up to the pion mass threshold. If the results are precise enough we will use them to build the photonuclear cross sections library we need. 3.1

Description

HMS-ALICE [Bla96] is a nuclear reaction code which was designed for the use in the bombarding energy range of a few MeV to several hundred MeV (under the pion mass threshold). The minimum input required to run ALICE is the target and projectile charge and mass, projectile energy and a title card. Other parameters can also be changed by the user. The output contains the cross sections for each reaction product. This code uses a hybrid Monte-Carlo method to calculate precompound nucleus decay. It permits multiple precompound emissions for each interaction, and it may also be used to calculate exclusive spectra and yields. 3.2

Results

At the very beginning we needed to verify the accuracy of the results of HMS-ALICE by comparing them with evaluation from the IAEA library. In the following sections we will present comparison between HMS-ALICE results and evaluated data from IAEA [IAE00]. As light nuclei have always been difficult to simulate and photofission was not included into previous HMS-ALICE version we will begin the study with heavy non-fissionable nuclei. 3.2.1

Heavy non-fissionable nuclei

The first calculation has been done for 208 Pb because this material is often used in nuclear technology for shielding purposes and is a potential candidate for photoneutron production target. In addition, the existing data are of good quality. Figure 2 is an example of the results we may obtain. In brief our calculations are satisfactory. 208 Pb has a very regular shape so it is 9

γ+

208

Pb reactions

700 σ(γ,1n) IAEA σ(γ,1n) HMS-ALICE σ(γ,2n) IAEA σ(γ,2n) HMS-ALICE

Cross section (mbarns)

600 500 400 300 200 100 0

5

10

15

20

25

Energy (MeV)

Figure 2: Photoneutron production cross sections (γ,1n) and (γ,2n) obtained with HMS-ALICE for 208 Pb and compared with IAEA evaluation (solid lines).

γ+

181

Ta reactions

600 σ(γ,1n) IAEA σ(γ,1n) HMS-ALICE σ(γ,2n) IAEA σ(γ,2n) HMS-ALICE

Cross section (mbarns)

500 400 300 200 100 0

5

10

15

20

25

Energy (MeV)

Figure 3: Photoneutron production cross sections (γ,1n) and (γ,2n) obtained with HMS-ALICE for 181 Ta and compared with IAEA evaluation (solid lines).

10

easy to simulate. If we look at a more exotic nucleus such as 181 Ta the results may differ from what we expect. For 181 Ta the results presented in figure 3 do not agree so well as for Pb. The major reason may be that HMS-ALICE calculates the photoabsorption cross sections with a single lorentzian. But as 181 Ta is a deformed nucleus (β=0.269) [Mol95] the absorption cross section should be modeled with a sum of two lorentzians. The above example prove that HMS-ALICE should be used with a certain care for deformed nuclei unless an improved absorption cross section is provided manually. A new routine for photoabsorption cross section will be included in the next version. 3.2.2

Light nuclei

We choose to study 12 C and 27 Al, and the results are presented in figures 4 and 5. Carbon and aluminium are some of the most used among the light nuclides and the absorption cross section can be modeled by a single lorentzian in the GDR region. For 12 C the cross sections are underestimated because some other reaction channels are open in the present simulation. For example the probability to emit an α particle is much higher than experimental obtained values. For 27 Al the situation is opposite. The peak value is overestimated while for high energy photons we underestimate the evaluation. There is a lot of work left in order to improve the prediction for light nuclei, but whatever the code used they are the most difficult to model because of the quick change of characteristics. Simulation for nuclides with strong shell closure such as magic nuclei and nuclei close to them have to be improved. This will be done in the next version of HMS-ALICE. For very light nuclides (below Be) the code can not give any results. We plan to use experimental data, when available. 3.2.3

Actinides

We choose to study 235 U and 238 U, being the dominant in the case of nuclear fuel cycle. Here a lot of experimental efforts have been made and data should be accurate. The results are given figures 6 and 7. The ability to use HMS-ALICE for photonuclear reactions on actinides is a new feature of the code. Improvement can be done in the 11

12

γ + C reactions

Cross section (mbarns)

20 σ(γ,1n) IAEA σ(γ,1n) HMS-ALICE σ(γ,p) IAEA σ(γ,p) HMS-ALICE

15

10

5

0 10

20

15

30

25

35

Energy (MeV)

Figure 4: Particle production cross sections (γ,1n) and (γ,1p) obtained with HMS-ALICE for 12 C and compared with IAEA evaluations (solid lines).

27

γ + Al reactions 30 σ(γ,1n) IAEA σ(γ,1n) HMS-ALICE σ(γ,p) IAEA σ(γ,p) HMS-ALICE

Cross section (mbarns)

25 20 15 10 5 0

5

10

20

15

25

30

35

Energy (MeV)

Figure 5: Particle production cross sections (γ,1n) and (γ,1p) obtained with HMS-ALICE for 27 Al and compared with IAEA evaluations (solid lines).

12

γ+

235

U reactions

Cross section (mbarns)

500

400

300

σ(γ,1n) IAEA σ(γ,1n) HMS-ALICE σ(γ,2n) IAEA σ(γ,2n) HMS-ALICE σ(γ,F) IAEA σ(γ,F) HMS-ALICE

200

100

0

10

5

15

20

Energy (MeV)

Figure 6: Photonuclear cross sections calculated with HMS-ALICE for compared with IAEA evaluations (solid lines).

γ+

238

235 U

U reactions

Cross section (mbarns)

500

400

300

σ(γ,1n) IAEA σ(γ,1n) HMS-ALICE σ(γ,2n) IAEA σ(γ,2n) HMS-ALICE σ(γ,F) IAEA σ(γ,F) HMS-ALICE

200

100

0

5

10

15

20

Energy (MeV)

Figure 7: Photonuclear cross sections calculated with HMS-ALICE for compared with IAEA evaluations (solid lines).

13

238 U

future versions, where we hope fission yields will also be calculated. Results are not really accurate. It seems that the error comes from the method to calculate the total absorption cross section due to the use of a single Lorentzian. These cross section are not consistent with the IAEA data as we can see from figure 8. As we mentioned previously the new calculation of total absorption cross section will be implemented in the next version shortly. 3.3

Conclusion

This version of HMS-ALICE does not include fission yield but work is in progress. For the actinides and deformed nuclei we are looking for a better method to calculate the absorption cross section. Marshall Blann (the author of the HMS-ALICE code) is working on the absorption and the shell closure effect problems, so in the next version we expect improved results. We should not forget that this is the first time HMS-ALICE is used to calculate photonuclear cross sections. The results are really encouraging. But to date they are not precise enough to build the library using this code exclusively. So we decided that for the nuclei which are in the IAEA library, we will use these data and the other cross sections will be provided from HMS-ALICE. For very light nuclides we will use experimental data since no code is able to reproduce them with desired precision. If we study the evolution of nuclei included in the IAEA library, the data from HMS-ALICE will be used only for second chance reactions. The error induced by the uncertainties in cross sections may be somewhat decreased. If we want to study a precise nucleus, which is not in the IAEA library, we will try to have the more accurate cross sections by calculations done with GNASH. The reason why we don’t use GNASH instead of HMS-ALICE is that GNASH requires a lot of preparation work and computation time. So using it for systematic calculations for hundreds of nuclei is less practical.

14

Photoabsorption cross sections 235

U

800 HMS-ALICE IAEA

600 400 200 0

0

5

10

15

Cross section (mbarns)

238

20

25

30

U

800 HMS-ALICE IAEA

600 400 200 0

0

5

10

15 239

20

25

30

Pu

800 HMS-ALICE IAEA

600 400 200 0

0

5

10

20 15 Energy (MeV)

25

30

Figure 8: Results of HMS-ALICE for photoabsorption cross section compared with IAEA evaluation. 15

4

GNASH

GNASH [You96] is another nuclear reaction code, which is based on the Hauser-Feshbach statistical approach including a preequilibrium stage. It requires as input ground state spin, parity, masses, nuclear structure data, particle transmission coefficients, etc. The absorption cross section is not necessary but is strongly recommended to be more confident in the output results. Our computational procedure was to fit experimental photonuclear cross section by adjusting physical parameters. The parameters we can play with are the fission barrier parameters as energy and width, but also the level-density parameter and the pairing energy. GNASH has never been used at Los Alamos National Laboratory for photonuclear cross section calculations on actinides. The only evaluations available come from Russia however they don’t include delayed or prompt neutrons spectra, which is one of our major interests for various applications. We choose to begin with nuclei which have already been studied at LANL for neutron induced-reaction cross sections : 235 U. 4.1

235

U

Lot of work has been done with GNASH on 235 U for evaluation of neutron cross sections. We began with these data. As we study photonuclear reactions and not neutron induced reactions we have removed artificially the (n,γ) channel which is responsible for a creation of 236 U as composite nucleus. The first point is to verify that we can obtain the same values as in the case of the IAEA evaluations by using input parameters from neutron calculation. To be on a safe side, we use the existing data for photoabsorption, so we take the data from IAEA even if they do not come from the same experiment as (γ,fission) and (γ, xn) cross sections. The results are acceptable as it can be seen on figure 9. At energies higher than 15 MeV we do not reproduce (γ,1n) and (γ,2n) cross sections. On the other hand we don’t know if we can trust these data due to the opening of the (γ,3n) channel. In the experiment this particular channel could not be separated. That’s maybe why our calculation are

16

400.0

Cross section (mbarn)

300.0

(γ,1n) calc. (γ,2n) calc. (γ,fiss) calc. Caldwell data (1980) Varlamov data (1987)

200.0

100.0

0.0 0.0

10.0 Energy(MeV)

Figure 9: Results of GNASH calculation for imental data (see legend for details).

20.0

235 U

compared with the exper-

underestimated. As a conclusion we have confidence to use this approach to calculate more rapidly the cross sections for other actinides. 4.2 4.2.1

237

Np

(γ,abs) cross section

As we mentioned, preferably GNASH needs the (γ,abs) cross section as input. Data [Die88] from two different experiments done by Berman and Caldwell at Livermore and by Veyssiere at Saclay are available for 237 Np but they are not consistent. As photoabsorption cross sections have small variations for close nuclei we decide to impose the value of the integrated photoabsorption cross sections between 0 and 20 MeV. Photonuclear data had also been measured for 235 U, 236 U, 238 U, 239 Pu by Berman and Caldwell at Livermore, 238 U had been also measured by Veyssiere at Saclay. We compare the experimental values of the integrated photoabsorption cross sections with the values from IAEA in figure 10. 17

4.4

B Berman, Caldwell values VVeyssiere value IAEA evaluations 237 XIntegral of Np

Integral (barn.MeV)

4.2

B

4.0

B

3.8 B 3.6

X B

3.4 V 3.2

V

B

3.0 234

235

236

237 A

238

239

240

Figure 10: Integrated photoabsorption cross sections for different actinides as a function of mass A. 237

Np (γ,abs)

600

Cross section (mbarn)

Berman data (1) Berman fit Veyssiere data (2) Veyssiere fit 0.356*(1)+0.664*(2)

400

200

0

5

10

15

20

Energy (MeV)

Figure 11: Photoabsorption cross section for 237 Np : both data and different parametrizations1 are plotted. Intermedaite curve giving integral of 3.54 1

The parametrization is a weighted average of the two fits from Berman and Veyssiere. Berman fit is weighted by 0.356 and Veyssiere fit by 0.644. The sum of the two weighted is our absorption cross section

18

According to figure 10 the value for integrated photoabsorption cross sections for 237 Np has been fixed at 3.54. This figure shows us that the data never agree and that it might be reasonable to make a compromise between the two experiments. Data from Berman are higher for all actinides, if we don’t pay attention to 236 U, and data from Veyssiere are systematically lower. The resulting absorption cross section is presented in figure 11. 4.2.2

Results 237

γ+

Np reactions

400 1n, GNASH calc. Bermann, Caldwell (1986) Veyssiere (1973) 2n, GNASH calc. Bermann, Caldwell (1986) Veyssiere (1973) F, GNASH calc. Bermann, Caldwell (1986) Veyssiere (1973)

Cross sections (mbarns)

300

200

100

0 5

10

15

20

Energy (MeV)

Figure 12: Results of GNASH for cross sections

237 Np

in the case of partial photonuclear

Using the previous photoabsorption cross section figure 11, we can obtain with GNASH the partial cross sections presented on figure 12. The compromise we have chosen is respected for the fission cross sections, where the two experiments disagree. For photoneutron production they agree within the error bars and we tried to follow their values.

19

4.3 4.3.1

239

Pu

(γ,abs) cross section

239

Pu (γ, abs)

600 IAEA data Berman (1986) Berman fit Gurevich (1976) Moraes (1993)

Cross section (mbarn)

500

400

300

200

100

0

5

10

15

20

Energy (MeV)

Figure 13: Photoabsorption cross section for

239 Pu

Finding an absorption cross section for 239 Pu was easier than for Np because data are more consistent as we can see figure 13. The major difference concerns the first fission hump. We choose Berman fit as input for GNASH because data from this experiment for (γ,1n), (γ,2n) and (γ,F) reactions are also available, but almost the same results can be obtain by using photoabsorption cross section from the IAEA library. 237

4.3.2

Results

By using the absorption cross section given in figure 13 we obtain the results for separate reaction channels on the figure 14. The difference we have with (γ,2n) and (γ,F) cross section over 15 MeV is due to the fact that we do not have data for (γ,3n) reaction and other reactions in competition. It is possible that the experimental data are overestimated or the absorption cross section underestimated. So we can

20

γ+

239

Pu reactions

400 1n, GNASH calc Berman (1986) Moraes 1n 2n, GNASH calc Berman (1986) F, GNASH calc. Berman (1986) Moraes (1993)

Cross section (mbarn)

300

200

100

0

5

10

15

20

Energy (MeV)

Figure 14: Results of GNASH for cross sections

239 Pu

in the case of partial photonuclear

not adjust the values which simulate the competition between these reaction channels. The difference around the first fission hump can be linked to the uncertainty we have in the absorption cross section in the same region. 4.4

Conclusion

We have shown that GNASH is a very powerful device to calculate photonuclear cross sections if experimental data, at least for photoabsorption cross section, are available. As we have seen for 237 Np it is not always easy to find experimental data. In addition the use of GNASH requires a lot of preparation work for each nucleus, that’s why we had preferred to use HMS-ALICE to build the library and reserve the use of GNASH for particular cases where better precision on evaluation is required.

21

5

CINDER’90

CINDER’90 neutron library contains 3400 isotopes while the IAEA can provide only 164 of them in the case of photonuclear reactions. It is clear that we need a theoretical model to complete the library. Unfortunately HMS-ALICE does not give accurate results in some cases as we discussed above. To improve the precision of the library we choose to insert the existing data from the IAEA library, when they are available and to complete them with HMS-ALICE calculations. For particular nuclei, not included in the IAEA library and for which HMS-ALICE failed, such as 237 Np we use GNASH results. To date the isotopes from the IAEA and data calculated with HMSALICE for more than 400 isotopes have be added. Data from IAEA for actinides go only to 20 MeV. We put those data in the library. For energies higher than 20 MeV cross sections were taken equal to the 20 MeV cross section for the same reaction. 5.1

Results

In the past some work had been done to calculate transmutation of long live nuclear waste in a photon flux with ORIGEN2 [RSI91]. This work was divided into two parts. The first one consisted into validating the method by comparing the results with the article from T. Matsumoto [Mat88] for 90 Sr. After the approach was validated [Ben02] the second part consisted of adding (γ,2n) reactions and using this method for 93 Zr. The flux used has a flat distribution between 10 and 20 MeV. We will compare the results of this work with the results of CINDER’90 and its new library. Our results are quite different from Matsumoto because he did not take into account the (γ,2n) reactions. Only radioactive nuclei will be presented on the following figures in this section. 5.1.1

Transmutation of

90 Sr

The differences we can find between two calculations come from the difference in cross sections. ORIGEN2 requires the average value of

22

90

Variation of radioactive nuclei for Sr 17

2

Flux : 10 γ/cm /s 1 90

Sr Sr Y 88 Y 89 Zr 87 Y 88 Zr 85 Sr 90 Sr [Ben02] 89 90

Log (N) (arb unit)

0.001

1e-06

89

Sr [Ben02]

90

Y [Ben02]

88

Y [Ben02]

89

Zr [Ben02]

1e-09

87

Y [Ben02]

88

Zr [Ben02]

85

Sr [Ben02]

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (days)

Figure 15: Transmutation of 90 Sr. Left part is the irradiation at 10 17 γ/cm2 /s and right part is natural decay.

1 90

Sr Y

90

88Y 81

Log (N) (arb. unit)

Kr

1e-10

1e-20

1e-30 16

17

18 2 Log(φ) (γ/cm /s)

19

Figure 16: Flux dependence for transmutation of

23

20

90 Sr

with CINDER’90

the cross section. For an irradiation of 72 days in a flux of 1018 γ/cm2 /s the difference is less than 7%. This difference increases with the flux. If we compare figure 15 with Benomard work [Ben02], our results are in a good agreement. Differences may come from the value of the integral cross sections used by ORIGEN, where they were extrapolated from existing data. Only (γ,1n) and (γ,2n) reactions have been taken into account in ORIGEN2 simulation whereas all possible reactions are included into CINDER’90 even if they have small cross section. This can explain the small differences. The major difference concerns the production of 88 Zr. This nuclide comes only from the (γ,1n) reaction on 89 Zr. 89 Zr is not in the IAEA library so the value of the cross section in ORIGEN2 has been extrapolated according to cross section from the IAEA for other isotopes of Zr. The value chosen was 77.9 mbarns as HMS-ALICE gives a value of 93.5 mbarns. It was the same for the reaction cross section on 88 Zr but here extrapolated and HMS-ALICE values are similar (72.9 vs 72.5). Results for other flux intensities can be found in appendix A. If we compare figure 16 with Matsumoto article [Mat88] we can see that values for 1020 γ/cm2 /s differ. At this flux (γ,2n) cross section have an important role. The quantity of 81 Kr is greater also because of those reactions. 5.1.2

Transmutation of

93 Zr

For the quantity of 93 Zr after irradiation and cooling we find an error of 8% compared to Bernomard work [Ben02]. But for the other radioactive nuclides the error is much bigger as we can see figure 17. The closest radioactive nuclide is 90 Y, where errors on cross sections are accumulated. No data for any Y isotopes are available in the IAEA library so they were taken equal to neighbor nuclei for the ORIGEN2 simulation. As the amount of produced daughter radioactive nuclides are so small the least error in cross section leads to big differences in the quantity of nuclides left in the sample after irradiation. The efficiency of the transmutation of 93 Zr is 45% with a flux of 1018 γ/cm2 /s. But we have no point of comparison for the flux dependence of the transmutation. Results for other values of the flux can be found in appendix B. 24

93

Variation of radioactive nuclei for Zr 17

2

Flux : 10 γ/cm /s 1 93

Zr Sr Y 89 Zr 89 Sr 88 Y 87 Y 85 Sr 93 Zr [Ben02] 90 90

Log (N) (arb unit)

0.001

1e-06

90

Sr [Ben02]

90

Y [Ben02]

89

Sr [Ben02]

88

Y [Ben02]

1e-09

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (days)

Figure 17: Transmutation of 93 Zr. Left part is the irradiation at 1017 γ/cm2 /s and right part is natural decay.

1 93

Zr Y 93 Nbm 85 Kr 81 Kr

Log (N) (arb. unit)

90

1e-10

1e-20

1e-30 16

17

18 2 Log(φ) (γ/cm /s)

19

Figure 18: Flux dependence for transmutation of

25

20

93 Zr

with CINDER’90

5.1.3

Transmutation of

137 Cs

No transmutation estimations have been done for 137 Cs in Saclay but we can still compare CINDER’90 results with Matsumoto’s work. The amount of 136 Cs (figure 19) production is almost the same but the amount of 135 Cs is by two orders of magnitude higher than in Matsumoto’s work due to (γ,2n) reaction taken into account in our case. For the same reason as previously more 137 Cs was transmuted and more daughter nuclei are produced. Results for other values of the flux can be found in appendix C. 5.2

Work left

Although HMS-ALICE distinguishes cross sections for residual in ground, first excited and second excited state at the moment we have only inserted the total cross section leading to the ground state. HMS-ALICE does not give cross section for light emitted particles. So for most of the nuclides these information are not available in the present library. Evaluated data from GNASH have not been added so far but they will be implemented in the near future. Cross sections from the IAEA should be weighted by a typical Bremsstrahlung flux to increase the accuracy of the calculations for dedicated applications. In the case of the absorption cross section a parametrization for actinides is studied and will be soon added to HMS-ALICE.

26

Variation of radioactive nuclei for 17

137

Cs

2

Flux : 10 γ/cm /s 1 137

Cs Cs Cs 134 Cs 133 Ba 132 Cs 131 Cs 131 Ba 136 135

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (days)

Figure 19: Transmutation of 137 Cs. Left part is the irradiation at 1017 γ/cm2 /s and right part is natural decay.

1 137

Cs Cs 134 Cs 133 Ba 135

Log (N) (arb. unit)

1e-10

1e-20

1e-30

1e-40 16

17

18 2 Log(φ) (γ/cm /s)

19

Figure 20: Flux dependence for transmutation of

27

20

137 Cs

with CINDER’90

6 6.1

Photofission neutron emission Simulation of prompt neutron emission

Before producing delayed neutron emission, we should be able to reproduce prompt neutron emission for which we have more data. As we begin we assure that the value of νp (Eγ ) for photofission was equal to the value of νp (En ) for neutron induced fission minus the binding energy of one neutron. This energy was taken equal to the threshold of the (n,2n) reaction for the A-1 nucleus (e.g. 234 U in the case of 235 U). We use the following approach to calculate prompt neutron emission : νp (Eγ , A, Z) = νp (En − Eth (n, 2n), A − 1, Z), where Eγ is the energy of the incident photon, En the energy of the incident neutron in the case of neutron induced fission and EB is the value of the threshold of the (n,2n) reaction. Another method consists of using the value of νp (En ) for photofission of the same nucleus. The binding energy is here taken equal to the threshold of (γ,1n) reaction. The formula becomes then νp (Eγ , A, Z) = νp (En − Eth (γ, 1n), A, Z) The comparison for different actinides is presented figure 21. In this case the IAEA data which come from the Russian BOFOD evaluation are sometimes the same as Caldwell data [Cal80]. Caldwell data come from [Ber86] and [Cal80] and Veyssiere evaluations from [Vey73]. Veyssiere evaluation is also based on value for prompt neutrons emission after neutron induced fission. For energies around the giant dipole resonance (GDR) we have a good approximation of the value of νp (E) for 235 U if we change the value of the binding energy. But this works not so well for other nuclides as we can see on the same figure. We do not reproduce the energy dependence. More accurate calculations should be done, for example with GNASH. 6.2

Delayed neutrons sources

Simulation done with CINDER’90 and its new library shows that one minute-irradiation of non-fissile stable nuclides by a photon flux with a 28

235

238

U

U

5 Caldwell data 234 U shifted by Eth(n,2n) 235 U shifted by Eth(γ,n) IAEA evaluation

Caldwell data 237 U shifted by Eth(n,2n) 238 U shifted by Eth(γ,n) IAEA evaluation Veyssiere evaluation

4

4

4 3

2

5

10 15 Energy (MeV)

237

20

3

3

10 15 Energy (MeV)

Figure 21: νp (Eγ) for

20

2

5

235 U, 238 U, 237 Np

29

20

Pu

Caldwell data 238 Pu shifted by Eth(n,2n) IAEA

5

4

5

10 15 Energy (MeV)

239

4

2

5

Np

Caldwell data 236 Np shifted by Eth(n,2n) 237 Np shifted by Eth(γ,n) Veyssiere evaluation

5

2

10 15 Energy (MeV)

and

239 Pu.

20

maximum energy of 8 MeV does not create irradiation products which are delayed neutron emitters. Some very light nuclides (He, Li) are missing in the library but they will be soon added. If we use a photon flux with a maximum energy of 20 MeV the only neutron emitter we create is 9 Li. It comes from (γ,p) reactions on 10 Be (threshold 19.6 MeV). The quantity of 17 N via 18 O created during one minute irradiation is also very small. The estimation of the delayed neutron emission via the photofission process depends on the knowledge of photofission yields. Calculations of these are in progress and preliminary comparison with experimental data is promising. These yields will be tabulated and inserted into CINDER’90 data library in the future.

30

7

Conclusion

The construction of a photonuclear library was tried with the different tools we have access to. The IAEA library gives a number of good quality cross sections and we could quickly complete it for missing elements thanks to HMS-ALICE. In the case we need more precision for particular nuclei we used GNASH as we did for 237 Np. Many work still has to be done to complete and update this new library before it can be used for confident activation calculations in the case of photonuclear reactions. Photofission yields, prompt and delayed neutron data from photofission are still to be prepared. Our preliminary results by the use of the above data library with CINDER’90 are in a reasonable agreement with previous work done on transmutation of long lived fission fragments.

31

8

Acknowledgement

I would like to thank Mark B. Chadwick, my supervisor for all the time he spent to explain me how to use GNASH, William B. Wilson, who helped me to build CINDER’90 library during my stay at LANL, Marshall Blann, who passed me his HMS-ALICE code even if he did not consider it as finished version. I would like to add a special thank to Kay Grady who helped me a lot with the administrative work.

32

References [Ben02] La transmutation de d´echets par r´eactions photonucl´eaires S. Benomard BhSc report : unpublished, 2002, CEA Saclay [Ber86] Photofission and photoneutron cross sections and photofission neutron multiplicities for 233 U, 234 U, 237 Np and 239 Np B. L. Berman and al. Physical Review C 34 (6) 2201 (1986) [Bla96] New compound decay model M. Blann Phys. Rev C 54 1341 (1996) [Cal80] Experimental determination of photofission neutron multiplicities for 235 U, 236 U, 238 U and 232 Th using monoenergetic photons J. T. Caldwell, E. J. Dowdy Nucl. Sci. Eng. 73 153 (1980) [Die88] Atlas of photoneutron cross sections obtained with monoenergetic photons S. S. Dietrich, B. L. Berman Atomic data and nuclear data tables 38, 199-338 (1988) [Gia03] Transmutation of 90 Sr, 93 Zr and 237 Np in a photon flux Giacri M.-L., Ridikas D., Benomard S. Work in progress [IAE00] Handbook on photonuclear data for applications Cross sections and spectra IAEA-TECDOC-Draf No 3 [Lit03] Photofission delayed neutron simulation capability Verification and validation Robert C. Little, Mark B. Chadwick, Charles A. Gouldind private communication [Mat88] Calculations of gamma ray incineration of 90 Sr and 137 Cs T. Matsumoto Nuclear Instruments and Methods in Physics Research A268, 234243 (1988) 33

[Mol95] Nuclear Ground-State Masses and Deformations P. M¨oller Atomic Data and Nuclear Data Tables, 59 (1995) 185-383 also Los Alamos Preprint LA-UR-93-3083 [Rid03] Renewed interest in photonuclear reactions Ridikas, M.-L. Giacri, P. Bokov Proceedings of Int. Conf. AccApp’03, San Diego, USA, 1-5 June 2003 [RSI91] RSIC Computer Code Collection 1991. ORIGEN2.1–Isotope Generation and Depletion Code Matrix Exponential Method. RSIC repport CCC-371 [Saf03] Photofissions experiments at CEA H. Safa Seminar at LANL, April 28 2003 [Vey73] A study of photofission and photoneutron processes in giant dipole resonance of 232 Th, 238 U and 237 Np A. Veyssi`ere and al. Nuclear Physics A199 45 (1973) [You96] Comprehensive nuclear model calculation Theory and use of GNASH code P. G. Young, E. D. Arthur, M. B. Chadwick ”Nuclear reaction data and nuclear reactor - Physics, design and Safety - Vol 1”, International centre for theoretical physics, Trieste, Italy, 15 April-17 May 1996.

34

A

Transmutation of

90

Sr 90

Variation of radioactive nuclei for Sr 17

2

Flux : 10 γ/cm /s 1 90

Sr Sr Y 88 Y 89 Zr 87 Y 88 Zr 85 Sr 89 90

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (days)

Figure 22:

90 Sr

: Flux 1017 γ/cm2 /s 90

Variation of radioactive nuclei for Sr 18

2

Flux : 10 γ/cm /s 1 90

Sr Sr 90 Y 88 Y 89 Zr 87 Y 88 Zr 85 Sr 84 Rb 83 Rb 83 Sr 82 Sr 81 Kr 89

Log (N) (arb unit)

Log (N) (arb unit)

0.001

0

20

40

60

500

1000

1500

Time (days)

Figure 23:

90 Sr

: Flux 1018 γ/cm2 /s

35

2000

2500

3000

90

Variation of radioactive nuclei for Sr 19

2

Flux : 10 γ/cm /s 1 90

Sr Sr Y 88 Y 89 Zr 87 Y 88 Zr 85 Sr 84 Rb 83 Rb 83 Sr 82 Sr 81 Kr 79 Kr 89 90

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (years)

Figure 24:

B

90 Sr

Transmutation of

137

: Flux 1019 γ/cm2 /s

Cs

Variation of radioactive nuclei for 17

137

Cs

2

Flux : 10 γ/cm /s 1 137

Cs Cs Cs 134 Cs 133 Ba 132 Cs 131 Cs 131 Ba 136 135

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

Time (days)

Figure 25:

137 Cs

: Flux 1017 γ/cm2 /s 36

2000

2500

3000

Variation of radioactive nuclei for 18

137

Cs

2

Flux : 10 γ/cm /s 1 137

Cs Cs Cs 134 Cs 133 Ba 132 Cs 131 Cs 131 Ba 129 Cs 127 Xe 126 I 125 I 136 135

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (days)

Figure 26:

137 Cs

: Flux 1018 γ/cm2 /s

Variation of radioactive nuclei for 19

137

Cs

2

Flux : 10 γ/cm /s 1 137

Cs Cs Cs 134 Cs 133 Ba 132 Cs 131 Cs 131 Ba 129 Cs 127 Xe 126 I 125 I 124 I 136 135

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

Time (days)

Figure 27:

137 Cs

: Flux 1019 γ/cm2 /s

37

2000

2500

3000

C

Transmutation of

93

Zr 93

Variation of radioactive nuclei for Zr 17

2

Flux : 10 γ/cm /s 1 93

Zr Sr Y 89 Zr 89 Sr 88 Y 87 Y 85 Sr 90

0.01

90

Log (N) (arb unit)

0.0001

1e-06

1e-08

1e-10

1e-12

0

20

40

60

500

1000

1500

2000

2500

3000

Time (days)

Figure 28:

93 Zr

: Flux 1017 γ/cm2 /s 93

Variation of radioactive nuclei for Zr 18

2

Flux : 10 γ/cm /s 1 93

Zr Sr 90 Y 89 Zr 89 Sr 88 Y 87 Y 85 Sr 93 Nbm 85 Kr 88 Zr 84 Rb 83 Rb 83 Sr 82 Sr 81 Kr 90

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

Time (days)

Figure 29:

93 Zr

: Flux 1018 γ/cm2 /s

38

2000

2500

3000

93

Variation of radioactive nuclei for Zr 19

2

Flux : 10 γ/cm /s 1 93

Zr Sr Y 89 Zr 89 Sr 88 Y 87 Y 85 Sr 93 Nbm 85 Kr 88 Zr 84 Rb 83 Rb 83 Sr 82 Sr 81 Kr 90 90

Log (N) (arb unit)

0.001

1e-06

1e-09

1e-12

0

20

40

60

500

1000

1500

Time (days)

Figure 30:

93 Zr

: Flux 1019 γ/cm2 /s

39

2000

2500

3000