1. Li(d,xn) reaction integral & differential σ exp. Neutron transmission experiment D 2 O( 3 He,xn) source reaction Integral benchmark of neutron cross sections D 2 O(p,xn) source NPI Participation on IFMIF Project Activation of accelerator hardware p&d variable energy beams Cross-section measurement 7 Li(p,n) neutron source reactions

2. NPI Neutron Source, High-power D 2 O Target Station the slit- free beam guide quadropole triplet, beam-diagnostics, part of pressure- and water flow control system target chamber holder for irradiated samples heavy-water flow tube separated vacuum hardware

4.  The analysis definitely points out the necessity of updating the specific cross sections in the IEAF-2001 library above 15 MeV. 186 W(n,np) 185 Ta Cross Section Benchmarking of 185 Ta Production (185Ta, T1/2 = 49 min) EAF-2001 Library C/E 4.0 ± 0.2 Reaction Path 186W(n,np) – 100% First benchmark data on 185Ta production IEAF-2001 overestimates 186W(n,np+d) cross sections above 15 MeV The only on measurement of XS up to 22 MeV is rather old (1959) Integral benchmark tests of activation cross section ● 7-35 MeV (IFMIF relevant) neutron energy region

5. P. Bém, M. Götz, M. Honusek, E. Šimečková and F. Veselý Nuclear Physics Institute Řež, Czech Republic Workshop on HEC Electronics Tests 2009, February 12-13, MPI Munich February 12-13. 2009, MPI Münich Workshop on HEC-E Tests °1 HEC Electronics Tests (1) Neutron Beam at NPI Cyclotron status (2) Next Steps low-energy neutron contribution to neutron beam s-to-s variance of faulty neutron flux - RADMON to DFM tests cryostat - radiation safety - support and adjustment procedure - measurement of faulty neutron flux - (new D 2 O target)

6. (1) Neutron Beam at NPI Řež Cyclotron, status February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 2 ● p(37 MeV)+D 2 O neutron source reaction ● determination of spectral flux density (SFD) at sample positions ○ MCNP(X) simulation ○ dosimetry-foil method ● spectral flux at 3 mm sample position ○ deconvolution procedure by the SAND-II code ○ uncertainty of adjusted spectrum ● variance of spectral flux with source-to-sample distance ○ simple corrections for HEC samples

7. Spectral neutron yield from p+D 2 O reaction □ spectral yield measured at 37 MeV proton beam ■ resulting data □ extends up to 32 MeV □ mean (fluence-averaged) neutron energy of 13.9 MeV. ■ simulate the dominant part of IFMIF source spectrum □ high integral yield 5x10 10 n/sr/s/μA □ evaluated flux density up to 10 11 n/cm 2 /s ■ provide suitable tool for validation of ACS at IFMIF energies ● p(37 MeV)+D2O neutron source reaction February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 3

8. 1 -water cooled protecting diaphragm with interchangeable carbon slit 2 - insulator, 3 - repeler of electrons, 4 - inlet of heavy water flow, 5 – Mo/Ta foil, 6 - target chamber samples February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 4 NPI neutron source, heavy-water-flowing target chamber beam integrator insulated sample holder Al(DF) beam monitor

9. ■ because of intensity reason samples are inserted close to the source ■ spectra of source reaction are measured at large distance (PLG Arrangement) ■ PLGA data are usualy not sufficient to determine the spectral flux at sample position due to averaging of observabeles over sample- and target dimensions ■ MCNPX fails to predict p-D2O data correctly since LA-150h with inbuild analytical model does not represent DDCS for the D(p,xn) reaction 37MeV protons D 2 O target Φ30x16 mm x r DDCS for the D(p,n) reaction February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 5 ● spectral flux density (SFD) at sample positions ○ MCNP(X) simulation PLGA

10. proton beam D 2 O target Φ30x16 mm ● spectral flux density (SFD) at sample positions ○ dosimetry-foil method, experiment x beam current integrator February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 6 Al(DF) monitor ■ a stack of metallic foils of Al, Ti, Fe, Co, Ni, Y, Nb, Lu, Au and Bi diameter of 12-14 mm thicknesses ranging from 0.05 to 0.3 mm ■ irradiated for 9 to 12 hours in several runs and different x position. ■ neutron flux monitors for time profile of the neutron source strength and normalization of dat from different runs □ Al(DF) monitors used □ the proton beam current and total charge recorded by a current-to-frequency converter and scaler on PC

11. 27Al (n,α+)24Na natTi (n,x)48Sc, (n,x)46Sc, (n,x)47Sc natFe (n,x)56Mn, (n,x)51Cr, (n,x)54Mn 59Co (n,α+)56Mn, (n,p)59Fe, (n,2n)58Co, (n,3n)57Co natNi (n,x)57Co, (n,x)58Co, (n,x)60Co, (n,x)57Ni 89Y (n,2n)88Y 93Nb (n,2n)92mNb 197Au (n,2n)196Au, 197Au(n,3n)195Au, (n,4n)194Au February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 7 ● spectral flux density (SFD) at 3 mm sample positions ○ dosimetry-foil method, gamma-spectrometry ■ The induced γ-ray activities were measured by high purity germanium detector (HPGe) at several cooling times from 1 h up to 100 days afterwards. ■ Finally, 26 activation reactions were employed for the neutron spectrum adjustment.

12. ● spectral flux density (SFD) at 3 mm sample positions ○ deconvolution procedure by the SAND-II code ■ modified SAND-II unfolding code data up to 55 MeV added from EAF-2005 ■ initial guess neutron spectrum at 37 MeV measured by scintillation unfolding technique supplied by MCNPX simulation below 7 MeV ■ the adjustment procedure changes the shape and absolute value of the initial spectra bringing the Acal/Aexp ratios close to unity ■ the iterative procedure converges after 5-10 cycles when the standard relative deviation between calculated C and experimental E activities (summed over all 25 detectors) reach its minimum of 6-8%. □ C/E results are displayed along energy weighted with product of neutron flux and corresponding CS. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 8

13. □ 3-6% in 7 to 25 MeV energy range correspond to mean square deviation of C/E ratios from unity □ 20-30% above 25 MeV reflects mean deviations of EAF-2005 cross sections from measurement □ no correct evaluation below 7 MeV could be carried out further relevant DF data needed ● spectral flux density (SFD) at 3 mm sample position ○ uncertainty of adjusted spectrum result from the ones of measured activities and dosimetry cross sections The SAND code does not deal with this problem uncertainty were estimated as follows 5th-iteration 0201030 MeV n/cm 2 /11.6 μA February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 9

14. Reaction rates ratio from irradiation at different s-to-s distances. □ selected set of foils were irradiated at different s-t-s distances □ RR ratio were formed and normalized to the longest s-t-s distances □ the exponential fit was done to resulting data February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 10 ● variance of flux with source-to-sample distance

15. Reaction rates ratio from irradiation at different s-to-s distances. ■ Using data at 3 mm distance and the quantitative form of observed averaging effects, the spectral flux at arbitrary distance from target is determined with an accuracy of 10-15%. □ the exponential fit was done to resulting data from reactions with lowest threshold 27Al(n,α) (upper curve) and highest one - 59Co(n,3n), 197Au(n,4n) (lower curve) ■ due to the averaging effect the deviation of flux density from the 1/R 2 law is different for reactions with different energies February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 11 ● variance of spectral flux with source-to-sample distance

16. Reaction rates ratio for reaction products from irradiation at 3 and 156 mm s-to-s distance □ identical set of foils was activated simultaneously at two s-to-s distance □ the ratio of reaction rates calculated for individual reactions are displayed along characteristic energies ■ the value of rates corresponds to the drop of flux density with distance ■ change of drop for data with different energy indicates the space-energy averaging effect on the spectral flux February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 12 ● variance of spectral flux with source-to-sample distance ■ Using data at 3 mm distance and the quantitative form of observed averaging effects, the spectral flux at arbitrary distance from target is determined with an accuracy of 10-15%.

18. Recommended displacement damage functions for NIEL scaling in Silicon ► Neutron Irradiation of Electronic Set-up Modules ● low-energy neutron contribution of SF needed energy range of p+D 2 O source IFMIF ACS benchmark tests ATLAS-HEC

19. ► Neutron Irradiation of Electronic Set-up Modules ● low-energy neutron contribution of SF Two guess spectra were obtained from MCNPX calculation and combination of this calculation with measurements. Inspite the deference in energy shape the final adjusted spectra agree each others.

20. Reaction rates ratio for reaction products from irradiation at 3 and 156 mm s-to-s distance □ identical set of foils was activated simultaneously at two s-to-s distance □ the ratio of reaction rates calculated for individual reactions are displayed along characteristic energies ■ the value of rates corresponds to the drop of flux density with distance ■ change of drop for data with different energy indicates the space-energy averaging effect on the spectral flux February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 2 ► Neutron Irradiation of Electronic Set-up Modules ● variance of SFD with source-to-sample distance

21. Neutron flux determination, DF experiments

22. As seen in Fig. 5, the adjustment procedure changes the shape and absolute value of the initial spectra bringing the Acal/Aexp ratios close to unity. On this way, the iterative procedure converges after 5-10 cycles when the standard relative deviation between calculated and experimental activities summed over all 25 detectors reach its minimum of 6-8%. Two final spectra obtained from two guess spectra are in a good agreement. The mean deviations amount 12% above 4 MeV, where the initial spectra have essentially different energy dependences and absolute values, since one of them was measured and another was calculated. Below 1-3 MeV the guess spectra were calculated by MCNPX and undergo the essential increasing during the adjustment procedure. In this energy domain only 197Au(n,γ) and 60Co(n,γ) cross sections affect on final spectrum. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 2

23. Neutron flux determination, DF experiments Nezapomenout na kapitolu o průchodu hmotou a taky použít obrázky dole

24. The uncertainties of adjusted spectrum stems from those of the measured activities (≈ 4%) and dosimetry cross sections. Since the SAND-II code does not deal with this problem we estimated them as follows. In the energy range 5 to 25 MeV, where we have 21 dosimetry detectors, the uncertainties of the spectrum could be assessed as a mean square deviation of Acal/Aexp ratios from unity, which amounts to 6%. Above 25 MeV, where only two reactions 59Co(n,3n)57Co and 197Au(n,4n)196Au were used in the adjustment procedure, a dominant contributor is the cross section uncertainty of these reactions. We estimate them at a level of 20 - 30% as mean deviations of EAF-2007 cross sections from measured ones. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 2

25. Conclusions An activation foil method has been used for the neutron spectrum determination from the p-D2O neutron source at Rez cyclotron facility. 11 elements and 25 dosimetry reactions having thresholds and effective cross sections covering the whole energy range up to 37 MeV were selected. The available evaluated nuclear data libraries and experimental data were carefully screened. The EAF-2007 library with 211 energy group structure was selected representing all desirable dosimetry cross sections in the energies from thermal to 55 MeV. For the determination of the neutron spectra from the induced γ- activities the adjustment code SAND-II was applied after having it modified for the use of the selected dosimetry cross sections above 20 MeV. Two guess spectra were obtained from MCNPX calculation and combination of this calculation with measurements. Inspte the deference in energy shape the final adjusted spectra agree each others. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 2

26. 27 Al(n,α) 24 Na reaction rates at different distance from D2O target Guess spectrum for the DF unfolding iterative procedure target-to sample distance (mm) February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 2

27. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 8 ■ reconstruction of the neutron spectrum from γ-activities of the unstable nuclides produced by the neutron activation reactions in a set of set of dosimetry reactions sensitive to the different energy ranges of the spectra are needed. the materials (dosimetry foils or detectors). ■ The international dosimetry reaction file IRDF-2002 [6] was taken as a guide since it contains the evaluated cross sections and its uncertainties based on numerous measurements and specially selected for the dosimetry purposes up to 20 MeV. ■ Above this energy, the cross section data are available in the European Activation File EAF- 2007 [7] up to 55 MeV, ENDF/B-VII [8] up to 30-150 MeV and the Intermediate Energy Activation File IEAF-2001 [9] up to 150 MeV. ■ Table 1 lists the elements and neutron reactions which we selected for this purpose ● determination of spectral flux density at sample position ○ Methodological approach of the dosimetry foil method

28. DosimetryHalf LifeEnergy, MeVUpper Energy Limit, MeV ReactionT 1/2 (σ > σ max /2)IRDF 2002ENDF/B-VIIExp.Data Reactions with cross sections maxima below 10 MeV 60 Co(n,γ) 60 Co5.27 y< 10< 20 197 Au(n,γ) 198 Au2.70 d< 10< 20< 30< 20 115 In(n,n’) 115m In4.49 h2 – 11< 20 Reactions with cross sections maxima between10 and 20 MeV 27 Al(n,α+) 24 Na14.9 h8 – 18< 20< 150< 50 59 Co(n,2n) 58 Co70.9 d13 - 27< 20 < 75 197 Au(n,2n) 196 Au6.18 d10 – 20< 20< 30< 38 Reactions with cross sections maxima above 20 MeV 197 Au(n,3n) 195 Au186.1 d20 – 30-< 30< 28 197 Au(n,4n) 194 Au38.0 h30 – 40-< 30< 38 209 Bi(n,3n) 207 Bi31.2 y20 – 30-< 150< 40 Some of 25 dosimetry reactions used for the white neutron spectrum determination up to 40 MeV at NPI p-D2O source indication of ■ decay half life of residual nuclei ■ energy range where cross section exceeds its half maximum ■ upper energy limits of IRDF-2002, ENDF/B-VII libraries and available experimental data.

29. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 8 The reactions listed in Table 1 are splitted in three groups ■ having the cross sections maximum below 10 MeV, only one low threshold reaction 115In(n,n’)115mIn and several non threshold neutron capture ones ■ between 10 and 20, and above 20 MeV. The reactions with emissions of charged particles or two neutrons have the maximum of the cross sections at energies 10 to 20 MeV, the excitation functions being represented in the IRDF-2002 file only up to 20 MeV. ■ Those sensitive to the neutrons above 20 MeV are usually multiple nucleons emission reactions, whose cross sections were neither precisely measured nor evaluated yet. ● determination of spectral flux density at sample position ○ Methodological approach of the dosimetry foil method the 197Au(n,2n)196Au reaction as an example: ■ only few independent measurements do exist above 20 MeV, ■ experimental uncertainties up 50% make impossible to prefer EAF-2007, ENDF-VII or IEAF-2001 libraries ■ the EAF-2007 library was finally chosen, agreeing with IRDF-2002 data below 20 MeV reasonably representing experimental points above this energy

30. February 12-13. 2009, MPI Münich Workshop on HEC-E Tests 8 To unfold the neutron spectrum we used the SAND-II code in a modified version. Originally this code and the supplemented dosimetry cross sections library (640 groups format) were developed to adjust fission type neutron spectra to the measured reactions rates [10]. Later the main subroutine of this package was modified for input the cross sections in 80 group structure covering the energy range 0.1 to 1000 MeV [11]. Since both mentioned approaches do not optimally cover the energy range of interest, we modified this code further to enable the use of nuclear data from the EAF-2007 library having 211 groups and covering the energies from 0.1 eV to 55 MeV. Further modifications concern the calculations of the sensitivity energy domain of dosimetry detectors.

31. 3p + D2O neutron source spectrum determination A 37 MeV proton beam current of ≈ 12 μA bombarded the heavy water target, Fig. 2, and produced a white neutron spectrum which is similar to that of the IFMIF neutron source except it has a cut-off at 35 MeV [1]. Since the samples were located at the distance 2-3 mm from the target bottom, the neutron spectrum has to be known there. In present experiment a stack of pure metallic foils of Al, Ti, Fe, Co, Ni, Y, Nb, In, Lu, Au and Bi having a diameter of 12-14 mm and thicknesses ranging from 0.05 to 0.3 mm were irradiated for 9 to 12 hours in several runs at this position. The mean proton beam current and total charge were recorded and used for the subsequent normalization of results obtained in different runs. The induced γ-ray activities were measured by high purity germanium detector (HPGe) at several cooling times from 1 h up to 100 days afterwards. These data were used to calculate specific disintegration rate of reaction residual nuclei at the end of the irradiation and then saturated ones Aexp, which could be reached during constant and infinitely long irradiation. Finally, 26 activation reactions highlighted in grey in the Table 1 and additionally were employed for the neutron spectrum adjustment.

32. For the unfolding iterative procedure an initial guess neutron spectrum is needed. To get it, the spectrum at a distance of 700 cm was measured by a scintillation detector having calibrated response function [12]. This spectrum was also calculated by the MCNPX code [13] simulating the neutron source, the target and the experimental hall and employing both the LA-150 proton library [8] and in-built analytical model. The results shown in Fig. 3 indicate the large discrepancies between measurements and calculations. Further study has shown that the LA-150 proton library does not reproduce experimental double differential cross sections for the D(p,xn) reaction and hence the spectral yield from the thick heavy water target, Fig. 4. To get guess spectrum at foil stack position, i.e. 3 mm from the source bottom, the neutron spectrum measured at 700 cm distance was renormalized using 27Al(n,α)24Na reaction rates measured at different distances from the target. Below the detector threshold of 4 MeV, the spectrum was supplemented by MCNPX calculations described above. The second guess spectrum was assessed inside the dosimetry foil stack volume only by MCNPX simulation to check the sensitivity of adjustment procedure to the trial spectrum. Both initial guess spectra are shown in Fig. 5 (upper part). The final spectra obtained after the adjustment procedure are also displayed there. The bottom part of Fig. 5 shows the ratios of calculated (with initial and final spectra) to measured saturated activities Acal/Aexp for all dosimetry reactions used in the present analyses. The abscissa data show the energy weighted with the product of the neutron flux and the corresponding cross section, the horizontal bars being show the mean square deviations. In such a way, the sensitivity energy range for every reaction is clearly indicated. In particular, one can see that most of the used reactions are concentrated between 10 and 25 MeV, with a few above and below this domain.

33. As seen in Fig. 5, the adjustment procedure changes the shape and absolute value of the initial spectra bringing the Acal/Aexp ratios close to unity. On this way, the iterative procedure converges after 5-10 cycles when the standard relative deviation between calculated and experimental activities summed over all 25 detectors reach its minimum of 6-8%. Two final spectra obtained from two guess spectra are in a good agreement. The mean deviations amount 12% above 4 MeV, where the initial spectra have essentially different energy dependences and absolute values, since one of them was measured and another was calculated. It is worthwhile to note that former cuts at 32 MeV whereas the latter extends up to 37 MeV. Below 1-3 MeV the guess spectra were calculated by MCNPX and undergo the essential increasing during the adjustment procedure. In this energy domain only 197Au(n,γ) and 60Co(n,γ) cross sections affect on final spectrum. The uncertainties of adjusted spectrum stems from those of the measured activities (≈ 4%) and dosimetry cross sections. Since the SAND-II code does not deal with this problem we estimated them as follows. In the energy range 5 to 25 MeV, where we have 21 dosimetry detectors, the uncertainties of the spectrum could be assessed as a mean square deviation of Acal/Aexp ratios from unity, which amounts to 6%. Above 25 MeV, where only two reactions 59Co(n,3n)57Co and 197Au(n,4n)196Au were used in the adjustment procedure, a dominant contributor is the cross section uncertainty of these reactions. We estimate them at a level of 20 - 30% as mean deviations of EAF-2007 cross sections from measured ones. Further independent experimental validation of the unfolding procedure and its uncertainties by proton-recoil telescope technique is presently under way. In this technique the spectrum of protons kicked–off from the thin polyethylene radiator will be recorded and then converted to the neutron one using the well known n-p scattering cross section.

34. 4. Conclusions An activation foil method has been used for the neutron spectrum determination from the p- D2O neutron source at Rez cyclotron facility. 11 elements and 25 dosimetry reactions having thresholds and effective cross sections covering the whole energy range up to 37 MeV were selected. The available evaluated nuclear data libraries and experimental data were carefully screened. The EAF-2007 library with 211 energy group structure was selected representing all desirable dosimetry cross sections in the energies from thermal to 55 MeV. For the determination of the neutron spectra from the induced γ-activities the adjustment code SAND-II was applied after having it modified for the use of the selected dosimetry cross sections above 20 MeV. Two guess spectra were obtained from MCNPX calculation and combination of this calculation with measurements. Inspite the deference in energy shape the final adjusted spectra agree each others.