Gamma-ray emission in alpha-particle reactions with C, Mg, Si, Fe
J. Kiener, J. Bundesmann, I. Deloncle, A. Denker, V. Tatischeff, A. Gostojic, C. Hamadache, J. Röhrich, H. Benhabiles, I. Bougaoub, A. Coc, F. Hammache, J. Peyré
GGamma-ray emission in alpha-particle reactions with C, Mg, Si, Fe
J. Kiener , J. Bundesmann , I. Deloncle , A. Denker , V. Tatischeff , A. Gostojic , C. Hamadache ,J. R¨ohrich , H. Benhabiles , I. Bourgaoub , A. Coc , F. Hammache , R. Mezhoud , J.Peyr´e CSNSM Orsay, CNRS/IN2P3 and Univ. Paris-Sud, France Protonen in de Therapie, Helmholtz-Zentrum Berlin, Germany Univ. M’HAMMED BOUGARA Boumerd`es, Algeria and IPN Orsay, CNRS/IN2P3 and Univ. Paris-Sud, France ∗ (Dated: January 13, 2020)Cross sections for the strongest γ -ray emission lines produced in α -particle reactions with C,Mg, Si, Fe have been measured in the range E α = 50 - 90 MeV at the center for proton therapyat the Helmholtz-Zentrum Berlin. Data for more than 60 different γ -ray lines were determined,with particular efforts for lines that are in cross section compilations/evaluations with astrophysicalpurpose, and where data exist at lower projectile energies. The data are compared with predictionsof a modern nuclear reaction code and cross-section curves of the latest evaluation for gamma-rayline emission in accelerated-particle interactions in solar flares. I. INTRODUCTION
Cross sections for the emission of nuclear γ -ray lines are a basic ingrediant for the analysis and interpretation ofobservations in the low-energy γ -ray band of astrophysical sites with an important population of accelerated ions.The latest compilation of Murphy et al. [1], aimed mainly at de-excitation lines produced in accelerated-particleinteractions in solar flares, features 248 cross-section curves for γ -ray lines emitted in proton, He and α -particlereactions with abundant isotopes in the solar atmosphere. The curves are typically given from reaction thresholdto several hundred MeV per nucleon, which is the important energy range for the accelerated particle populationsin solar flares. For most of the curves, however, data exist only at proton energies below about 25 MeV and for α -particles below about 10 MeV per nucleon and the extrapolation to higher energies relied on calculations with thenuclear reaction code TALYS [2].Of particular importance are the most prominent lines that often stand out in observed spectra from strong solarflares. They can be used to deduce ambient abundances and the energetic-particle composition and energy spectrum[3, 4]. These lines are from the deexcitation of the first or second excited state in the most abundant species: C, N, O, Ne, Mg, Si, Fe. The cross section curves for the emission of these lines in proton and α -particleinteractions with these nuclei typically show a broad maximum around 15-20 MeV, due to inelastic scattering reactions,reaching the several hundred mb range. This is followed by an nearly exponential fall-off to higher energies for protonreactions while a second, very broad maximum around 60 MeV is predicted for α -particle reactions. This maximummay be explained by fusion-evaporation reactions like ( α ,dnp) or ( α ,2n2p), where the residual nucleus is identical tothe target nucleus.This second cross section maximum is important in solar flares with relatively high accelerated α -particle to protonratios α /p and hard energy spectra. However, there are practically no experimental data available for the emission of γ -ray lines in α -particle reactions above E α = 40 MeV. We therefore decided to measure cross sections for the emissionof strong lines in α -particle reactions around the predicted second cross section maximum. In the first experiment, weconcentrated on the elements C, O, Mg and Fe. The chosen energy range, E α = 50 - 90 MeV also continues to higherenergies the data ( E α ≤
40 MeV), obtained in previous experiments at the Orsay tandem-van-de-Graaff accelerator[5, 6].
II. EXPERIMENT
The experiments have been done at the center for proton therapy of the Helmholtz-Zentrum Berlin in two campaignsin 2015 and 2016. Pulsed beams of α particles with E α = 50 MeV in 2015 and E α = 60, 75 and 90 MeV in 2016 havebeen produced using the Van-de-Graaff injector and the K-132 separated sector cyclotron of the Helmholtz center.They were directed onto self-supporting target foils of C, Mg, Si and Fe inside a spherical chamber of 15 cm diameter ∗ Electronic address: [email protected] a r X i v : . [ nu c l - e x ] J a n made of stainless steel and stopped in a thick copper block about 4 m downstream of the target chamber. The targetchamber was equipped with two glass windows perpendicular to the beam direction, where one of them served foroptical monitoring of the beam spot through the use of an alumina target. In both campaigns, 4 detectors, oneLaBr3:Ce scintillation detector and three HP-Ge detectors in a wide angular range and at large distances of typically50-70 cm from the target, have been used for the detection of the emitted γ rays.The pulsed beams were produced with relatively low repetition rates of ∼
80 kHz to ∼ Ba,
Cs, Co and
Eu for energies below about 1.4 MeV and extensive Geant simulations to extrapolate the detectorefficiencies to higher energies and also to obtain precise detector response functions. htEEntries 9.364181e+07Mean x 1493Mean y 4131RMS x 1615RMS y 64.06 htEEntries 9.364181e+07Mean x 1493Mean y 4131RMS x 1615RMS y 64.06 T2:E2
FIG. 1: Timing channel (T2) versus energy channel (E2) of events in the LaBr :Ce detector recorded during the irradiation ofthe C target with a 90-MeV α -particle beam. The detector signals were treated with standard NIM modules, providing a signal for the deposited energy andanother signal for the timing of the event. The timing signal was defined as the time difference between the detectorsignal and the beam pulse signal, both determined with constant fraction discriminators fed into a time-to-amplitudeconverter NIM module. The energy and timing data of the 4 detectors were then digitized and recorded in event-by-event mode and a time stamp of 100 ns resolution using an acquisition system of FAST ComTec [7], providing alsothe dead times of the 4 energy and 4 timing channels.Figure 1 shows a zoom in the timing vs. energy plane of data recorded by the LaBr :Ce detector in the irradiationof the C target with a 90-MeV α -particle beam. The slightly inclined, horizontal branch at T2 ∼ γ rays induced by beam interactions in the target. At E2 ∼ ∼ ∼ ∼ γ rays, have been identified to come from interactions of secondary particles, mostly neutronsprobably, inside the stainless-steel walls of the reaction chamber. Events arriving still later are from interactions infurther away beam tubes and supporting structures and from secondary neutrons interacting in the LaBr :Ce crystal.Beam-induced γ rays from interactions in the beam stop are in this histogram around T2 ∼ c o un t s energy (keV) FIG. 2: Compton-subtracted spectrum of the HP-Ge detector at 90 o to the beam direction from irradiation of the Mg targetwith 75-MeV α particles. III. DATA ANALYSIS AND RESULTS
Spectra of prompt γ rays from the target were obtained by a selection of events within a narrow band around thevisible horizontal branchs in the energy-timing plane from prompt target γ rays as shown in Figure 1. With theLaBr :Ce detectors, events from interactions of secondary particles could be completely excluded by the selection,while for the HP-Ge detectors, with a time resolution not better than 3-4 ns, not all events from secondary-particleinteractions in the target chamber walls could be separated. Their contributions, essentially lines from the first fewexcited states of Fe and Cr, could however be estimated from their time profile in the LaBr :Ce detector. Thiswas important for cross-section determinations with the Fe target.Integrals for narrow γ -ray lines could be directly taken from these spectra, while for broader lines and line complexes,Compton-subtracted spectra were used. These spectra were obtained by an unfolding procedure made possible bydedicated measurements with radioactive sources and extensive Geant simulations for the detector response functionsin the two experimental setups. The accuracy of the Compton subtraction from about 8 MeV down to the 511-keVline was estimated to be better than 20%. An example of Compton-subtracted spectrum is shown in Figure 2.Some of the most important lines for each target, broad lines (like E γ = 4439 keV in C) or narrow ones (e.g. E γ = 1368 and 1779 keV in Mg and Si, respectively) were merged with other lines in a complex structure and neededline-shape calculations to determine their intensity. Examples for the 1779-keV line and the 4439-keV line are shownin Figures 3, 4. The uncertainties related to these decompositions is added to the statistical and other systematicuncertainties like the detector efficiencies. In cases of lines very close in energy, like the 4439-keV line of C and the4445-keV line of B, the sum has been used for the line cross section determination.Differential cross section data from the 3 HP-Ge detectors have been obtained for about 60 different γ -ray lines,and for most of them at all α -particle energies. With the LaBr :Ce detectors, however, owing to their lower energy Si 1779 keV Si 1779 keV fev Mg 1793 keV Mg 1809 keV Si 1794 keV energy (keV) c o un t s FIG. 3: Symbols show the line complex with the1779-keV line of Si in the Compton-subtractedspectrum of the HP-Ge detector situated at 147 o to the beam direction from irradiation of the Sitarget with 60-MeV α particles. The line ” Si1779 keV fev” on the figure means productionof the 1779-keV line by fusion-evaporation reac-tions. C 4439 keV C 4319 keV B 4445 keV energy (keV) c o un t s FIG. 4: Symbols: Compton-subtracted spec-trum around the 4.4-MeV line complex of theHP-Ge detector at 116 o to the beam directionduring irradiation of the C target with 50-MeV α particles. The shape of the 4439-keV line of C has been obtained using coupled-channelscalculations for inelastic scattering as in [8]. resolution, no cross section data could be obtained for some lines in complex structures. Finally, the γ -ray emissioncross sections have been obtained by Legendre-polynomial fits of the 3 or 4 detector data. To the resulting cross-sectionuncertainty from the fit were added the target thickness and beam charge uncertainties.Cross section curves for the strongest line of each target, here from the deexcitation of the first excited level inthe major isotopes C, Mg, Si and Fe are shown on Figure 5. The data show effectively clearly a secondcross section bump for the 1.78-MeV Si and 0.85-MeV Fe lines, approximately at the energies predicted by thecompilation [1]. The latter, adjusted to the data at lower energies, that were obtained at the tandem-Van-de-Graaffaccelerators of Washington [9, 10] and Orsay [5], however, overestimates significantly the measured cross sections.For the 1.37-MeV line of Mg, there is a hint for a weak second cross section bump, in any case significantly lesspronounced than predicted by the compilation. Its prediction for the cross section curve of the 4.44-MeV line of Cagrees reasonably with the present data.Similar disagreements can also be seen for cross sections of the other strong lines, albeit with a slightly betteragreement for the lines from the Fe target compared to the other targets. We also did calculations with the nuclearreaction code Talys [2] with default parameters for the nuclear structure and potentials. For the lines from thedeexcitation of the first few levels of the major target isotopes, they generally agree at energies below E α ∼ γ -rayemission in α -particle reactions and an extrapolation to higher projectile energies. IV. CONCLUSIONS
Cross section data for the emission of about 60 γ -ray lines could be determined for α -particle reactions with targetsof C, Mg, Si and Fe at E α = 50, 60, 75 and 90 MeV. For the strongest lines from each target, there is now with thepresent data a complete coverage of experimental data from threshold to 90 MeV. Furthermore, cross sections formany new lines could be added, where no data were available from previous experiments. Comparison of cross sectioncurves of the latest cross section compilation and nuclear reaction code calculations show significant disagreementswith the new data. This underlines the importance and necessity of experimental work at particle accelerators for theestablishment of an accurate cross section data base in a wide projectile energy range that is needed for applicationsin astrophysics, but also for other applications with energetic particles like hadrontherapy. New accelerator centersopen possibilities for future measurements in astrophysics and nuclear physices [11]. E γ = 4.44 MeV σ ( m b ) E γ = 1.37 MeV
50 100 E γ = 1.78 MeV energy (MeV) σ ( m b )
50 100 E γ = 0.85 MeV energy (MeV) FIG. 5: Gamma-ray line emission cross sections for α -particle reactions with, from left to right and from upper to lower: C,Mg, Si and Fe. Red filled circles are the present data, blue filled squares are from Ref. [5], green triangles are from Ref. [9]for C and from Ref. [10] for Mg, Si and Fe. The cyan curves are the cross section excitation functions from the cross sectioncompilation [1]. V. ACKNOWLEDGMENTS
We like to thank the staff at the section for proton therapy of the Helmholtz-Zentrum Berlin for their stronginvolvement to provide the high-quality beams in the time periods between the cancer therapy sessions and theirgenerous and efficient help in the planning and setup of the experiments.
References [1] R. J. Murphy, B. Kozlovsky, J. Kiener, and G. H. Share, Astrophysical Journal Supplement Series , 142 (2009).[2] A. J. Koning, S. Hilaire, and M. C. Duijvestijn, in
TALYS-1.0, Proceedings of the Int. Conf. on Nuclear Data for Scienceand Technology, April 22-27, 2007, Nice, France (EDP Sciences, 2008), p. 211.[3] R. J. Murphy, G. H. Share, J. E. Grove, W. N. Johnson, R. L. Kinzer, J. D. Kurfess, M. S. Strickmann, and G. V. Jung,Astrophysical Journal , 883 (1997).[4] R. J. Murphy, R. Ramaty, B. Kozlovsky, and D. V. Reames, Astrophysical Journal , 793 (1991).[5] A. Belhout, J. Kiener, A. Coc, J. Duprat, C. Engrand, et al., Physical Review C , 034607 (2007).[6] H. Benhabiles-Mezhoud, J. Kiener, J.-P. Thibaud, V. Tatischeff, I. Deloncle, et al., Physical Review C , 014605 (2019).[9] P. Dyer, D. Bodansky, D. D. Leach, E. B. Norman, and A. G. Seamster, Physical Review C , 1873 (1985).[10] A. G. Seamster, E. B. Norman, D. D. Leach, P. Dyer, and D. Bodansky, Physical Review C , 394 (1984).[11] D. Tonev, N. Goutev, and L. S. Georgiev, Journal of Physics: Conference Series724