Stellar 36,38 Ar (n,γ ) 37,39 Ar reactions and their effect on light neutron-rich nuclide synthesis
M. Tessler, M. Paul, S. Halfon, B. S. Meyer, R. Pardo, R. Purtschert, K. E. Rehm, R. Scott, M. Weigand, L. Weissman, S. Almaraz-Calderon, M. L. Avila, D. Baggenstos, P. Collon, N. Hazenshprung, Y. Kashiv, D. Kijel, A. Kreisel, R. Reifarth, D. Santiago-Gonzalez, A. Shor, I. Silverman, R. Talwar, D. Veltum, R. Vondrasek
SStellar , Ar ( n, γ ) , Ar reactions and their effect on lightneutron-rich nuclide synthesis
M. Tessler, M. Paul, ∗ S. Halfon, B. S. Meyer, R. Pardo, R.Purtschert, K. E. Rehm, R. Scott, M. Weigand, L. Weissman, S.Almaraz-Calderon, M. L. Avila, D. Baggenstos, P. Collon, N. Hazenshprung, Y. Kashiv, D. Kijel, A. Kreisel, R. Reifarth, D. Santiago-Gonzalez,
4, 8
A. Shor, I. Silverman, R. Talwar, D. Veltum, and R. Vondrasek Racah Institute of Physics, Hebrew University, Jerusalem 91904, Israel Soreq NRC, Yavne 81800, Israel Department of Physics and Astronomy,Clemson University, Clemson, South Carolina 29634, USA Argonne National Laboratory, Argonne, Illinois 60439, USA Physics Institute, University of Bern, 3012 Bern, Switzerland Goethe University Frankfurt, Frankfurt 60438, Germany Department of Physics, University of Notre Dame, Notre Dame, Indiana 46556, USA Department of Physics and Astronomy,Louisiana State University, Baton Rouge, Louisiana 70803, USA (Dated: November 5, 2018) a r X i v : . [ nu c l - e x ] A ug bstract The Ar( n, γ ) Ar ( t / = 35 d) and Ar( n, γ ) Ar (269 y) reactions were studied for the firsttime with a quasi-Maxwellian ( kT ∼
47 keV) neutron flux for Maxwellian Average Cross Section(MACS) measurements at stellar energies. Gas samples were irradiated at the high-intensity Soreqapplied research accelerator facility-liquid-lithium target neutron source and the Ar/ Ar and Ar/ Ar ratios in the activated samples were determined by accelerator mass spectrometry atthe ATLAS facility (Argonne National Laboratory). The Ar activity was also measured by low-level counting at the University of Bern. Experimental MACS of Ar and Ar, corrected tothe standard 30 keV thermal energy, are 1.9(3) mb and 1.3(2) mb, respectively, differing fromthe theoretical and evaluated values published to date by up to an order of magnitude. Theneutron capture cross sections of , Ar are relevant to the stellar nucleosynthesis of light neutron-rich nuclides; the two experimental values are shown to affect the calculated mass fraction ofnuclides in the region A=36-48 during the weak s -process. The new production cross sections haveimplications also for the use of Ar and Ar as environmental tracers in the atmosphere andhydrosphere. Ar and Ar are among the rare stable nuclides for which no exper-imental neutron-capture cross sections exist above thermal energy. While the abundancesof , Ar in terrestrial atmospheric argon are very low relative to Ar (produced mainlyfrom K decay [1, 2]), Ar (84.59%) and Ar (15.38%) are the major argon isotopes inthe solar system [3] and likely so in stellar matter. They are expected, together with thebranching point Ar, to play a role in nucleosynthesis of light neutron-rich nuclei ( e.g. S, Ar, K), believed to be produced during the weak s -process phase of stellar evolution[4, 5]. The K ( t / =1.248(3) Gy [6]) nuclide, in particular, is an important cosmo- orgeochronometer and was used to estimate the age and duration of the s -process as ∼
10 Gy[7, 8]. K can be produced also in explosive oxygen burning [9] as a primary nucleosyn-thesis product in a massive star of initially pure hydrogen while the (secondary) s -processproduction of K requires initial abundances of heavy species. A better understanding ofAr cross sections will help clarify the relative primary vs. secondary production of K. Ina different realm of study, the half-life of Ar ( t / =35.011(19) d [10]) makes this isotopean ideal chronometer for studying circulation and mixing [11], and that of Ar (269(3) y[12]) for dating groundwater [13, 14] and ocean water up to about 1000 years [15]. Theatmospheric steady state concentrations of Ar and Ar are mainly determined by thespallation reactions Ar( n, n ) Ar and Ar( n, n ) Ar and at lower neutron energies bythe Ar( n, γ ) Ar and Ar( n, γ ) Ar reactions [11]. The latter are also relevant for theestimation of anthropogenic emissions from nuclear installations or for nuclear explosionmonitoring [16].We measured the Ar and Ar neutron capture cross sections by activation with quasi-Maxwellian neutrons produced by the Li( p, n ) reaction at the superconducting linear accel-erator of Soreq applied research accelerator facility (SARAF) [17, 18] and the Liquid-LithiumTarget (LiLiT) [19, 20]. The activation products Ar and Ar were counted offline by ac-celerator mass spectrometry (AMS); Ar production was also determined by Low-LevelCounting (LLC). Neutron irradiation of separate Ar and Ar samples was performed atthe pneumatic transfer tube (rabbit) of the Soreq IRR-1 nuclear reactor in order to re-measure the respective thermal neutron capture cross sections. Preliminary results of theseexperiments were reported in [21, 22].Enriched Ar, Ar and mixed Ar+ nat
Ar gas samples were filled into Ti spheres (10mm outer diameter, 0.2 mm thick Ti shell) [23]. Due to the thermodynamical properties of3
ABLE I. Samples used and the results of the A +1 Ar/ A Ar ratios. Ar and Ar gas samples [27]were enriched to 99.935% and 99.957% for the respective isotopes. The final Ar/ Ar ratios wereobtained by taking a weighted average of the AMS and LLC results (Fig. 2). Sphere Ar/ nat
Arratio for sphere 54 (52b) is 11.7 (10.2). For more details see the Supplemental Material [28].Sphere A Ar (mg) A +1 Ar/ A Ar ratio39 (LiLiT) Ar (24.5) 8 . × −
52a (reactor, Cd) Ar (19.9) 1 . × −
60 (reactor) Ar (22.6) 3 . × −
59 (LiLiT) Ar (19.5) 4 . × −
54 (reactor) ,nat Ar (12.8) 8 . × −
52b (reactor) ,nat Ar (15.8) 1 . × − Ar, the filling was made by successive compression with a custom-made piston and cryogenicpumping in order to achieve the required pressure ( ∼
30 bar). The samples used are listed inTable I. For the samples irradiated at SARAF-LiLiT (Table I), each gas sphere was placedwith a 25 mm-diameter Au foil (12.5 µ m thick), used as a neutron fluence monitor in anevacuated chamber downstream of LiLiT (Fig. 1). LiLiT consists of a windowless film ofliquid lithium (1.5 mm thick, 18 mm wide) flowing at 2-3 m/s, serving as both the neutron-producing target and the kW-power beam dump for the incident ∼ ∼
30% of the outgoing neutrons. The proton beam energy, measured byRutherford back scattering off a Au target after the acceleration module, was found to be1932 ± ±
3) keV for the Ar ( Ar) irradiation. A proton beam energy spread of ∼ Ar ( Ar) irradiation was found; this offset was accountedfor in our simulations. The neutron yield was continuously monitored with a fission-product4 time (sec) c p s c u rr en t ( m A ) neutron energy (keV) × × × d N / d E ( n / k e V / m C ) FIG. 1. (Color online) (right) Diagram of the Liquid-Lithium Target (LiLiT) and activationtarget assembly. The ( ∼ ∼ Ar sample (black) anda fit in the range E n ∼ −
110 keV with a Maxwell-Boltzmann flux (red) at kT ∼
47 keV. (bottomleft) Count rate (left y-axis) of fission chamber (see text) and calibration to proton current (righty-axis) during the Ar run. ionization chamber [26], located ∼
80 cm downstream the target at 0 ◦ . The fission chambercount rate was calibrated to beam current (at low intensity) using a Faraday cup located ∼ ∼ Ar ( Ar) irradiation (Fig. 1).The Ar nuclide decays by pure electron capture with no γ -ray emission; Ar is notablefor its role in Davis’ solar neutrino experiment [29] where its production via Cl( ν e , e − ) Arwas detected by Auger electron counting. We detected and counted for the first time Arby Accelerator Mass Spectrometry (AMS) at the ATLAS facility of Argonne National Lab-5ratory to measure the Ar/ Ar ratio of the irradiated samples. Ar gas was directly fedfrom the sphere container into an Electron Cyclotron Resonance (ECR) ion source througha remote-controlled sapphire leak valve. , Ar ions were extracted from the ion sourceand accelerated alternately through ATLAS at an energy of 6 MeV/ u by appropriate scalingof all accelerator elements. It was found necessary to strip the Ar ions and count Ar (fully stripped) in order to suppress the Cl (Z=17) background. Stripping was done with a200 µ g/cm C foil at an intermediate stage of the ATLAS linear accelerator. The strippingprocess (normally not used in AMS measurements at ATLAS) however produced an isotopefractionation and the effective beam transmission efficiency (1 . × − ) was determinedby interpolation between the measured Ar and Ar transmissions. The Ar ions werecounted using a ∆E-E telescope of Si detectors, 50 and 300 µ m thick, respectively, showingbackground-free spectra; the detection sensitivity in the present experiment was Ar/Ar ∼ − (see Supplemental Material [28]).The Ar activity of the same samples was also determined by ultra-low-level counting(LLC) in a second stage. Stainless steel vials containing ∼ aliquots of the sameactivated samples were shipped to the University of Bern. Each gas was quantitativelytransferred into a 100 cm copper proportional counter which was then filled with P6 gas(6% methane + 94% commercial Ar-free argon) to a pressure of ∼ Ar activitywas measured by detecting Auger electrons in an underground LLC laboratory during 1-2days [16, 30]. Energy calibration was performed with copper K-shell X-rays (E=8.133 keV)induced by an external Am γ source. The Ar peak was identified at the K-capture decayenergy of 2.82 keV [31] and integrated by means of a Gaussian fit [28]. The amount of Ar inthe sample was determined, after Ar counting, from the filling pressure of the detector andthe Ar/ Ar ratio measured by mass spectrometry [32] using established procedures. Theoverall uncertainty of 8% of the final Ar/ Ar ratio is dominated by counting statisticsand the uncertainties of counting yield (5%) [28]. A comparison of the Ar/ Ar ratiosmeasured by AMS and LLC is illustrated in Fig. 2.Accelerator Mass Spectrometry of Ar has been previously performed at ATLAS [33, 34].A high ion energy is essential for the separation and discrimination of Ar from the extremelyintense source background of the stable K isobar. In our experiment, the ECR was operatedat low power to reduce as much as possible impinging of the plasma onto the chamber walls,believed to be a source of K contamination. , , Ar ions were accelerated to 6 MeV/ u ,6 A r / A r r a t i o -13 ×10 -13 ×10 -10 AMSLLC
FIG. 2. (Color online) Comparison of the Ar/ Ar ratio (at the end of irradiation) measuredby AMS (black) and LLC (red). similarly as described before and Ar ions were analyzed in the Enge gas-filled magneticspectrograph [35], which physically separates Ar from beam contaminants, e.g. K and S , which have close-by m/q values (Fig. 3). The accelerator transmission efficiency for Ar (0.40(3)) was interpolated between those of Ar and Ar [28].The ratios r = A +1 Ar/ A Ar at the end of irradiation are determined by r = N A +1 (cid:15) t qe − i A e λt cool where N A +1 is the number of A +1 Ar detected, (cid:15) is the detector efficiency (measured to be0.91(3) for Ar due to grid shadowing in the spectrograph focal-plane detector), t thecounting time, q is the ion charge state (18 for Ar and 8 for Ar), e is the electroniccharge in coulomb, and i A the A Ar q + beam intensity (nanoampere); λ = ln (2) t / is the A +1 Ardecay constant and t cool is the time between the end of irradiation and counting. The finalresults of the A +1 Ar/ A Ar ratios for all gas samples are presented in Table I.In the reactor irradiations, two small Au samples were attached to the Ar and Arspheres for neutron monitoring, using 98.65(9) b [37] for the
Au thermal neutron capturecross section. A minor correction for the epithermal activation of Au was applied, using the
Au activity measured for a gas sphere entirely shielded with 1 mm thick Cd. In contrast tothe Ar sample, two Ar samples irradiated at the reactor (Table I) were mixed with nat
Ar7 K Ar+nLiLiT S Ar Ca K Ar S Ca Focal plane position (channel number) D E ( c h a nn e l nu m b e r ) FIG. 3. Identification spectrum of Ar ions in the detector measured for the LiLiT irradiated Ar gas (top) and for non-irradiated Ar gas (bottom). The horizontal axis represents dispersionalong the focal plane and the vertical axis a differential energy loss signal measured in the fourthanode of the focal-plane ionization chamber [36]. to use Ar ( σ th ( Ar)=0.66(1) b [37]) as an internal neutron monitor in addition to the Aumonitors; excellent agreement was obtained between the two neutron fluence calibrations[28]. The , Ar measured thermal capture cross sections are listed in Table II. Uncertainties(1 σ ) for Ar ( Ar) are 3% (2%) and 7% (11%) from the neutron fluence and atom ratiodeterminations, respectively. 8or the LiLiT irradiated samples, the average experimental cross section, σ exp , is obtainedby σ exp = r Φ n , where Φ n is the effective neutron fluence (n/cm ). In view of the complexgeometry of the gas sphere irradiation, Φ n is calculated as Φ n = (cid:80) l n V where V (0.46 cm ) isthe gas sphere’s volume, l n is the length a neutron travels inside the Ar gas and (cid:80) l n is thesum of the lengths traveled by all the neutrons inside the Ar gas sphere during the irradiation. (cid:80) l n is calculated by a detailed simulation (see below), taking a statistically representativesample of neutrons and scaling by the Au activity. The validity of the expression (cid:80) l n V forthe neutron fluence, Φ n , was confirmed by comparing the value calculated in this way forthe Au (planar) monitor with its measured activity; experimental and calculated valuesagree within 0.5%. The values of Φ n (n/cm ) and σ exp for Ar ( Ar) are 6.2(1) × (4.22(9) × ) and 1.4(1) mb (0.95(10) mb), respectively. Uncertainties (1 σ ) for the Ar( Ar) experimental cross section σ exp are 2% (2%) and 7% (11%) from the neutron fluenceand atom ratio determinations, respectively.The experimental cross section measured in our experiments is an energy-averaged valueover the neutron spectrum and interpretation in terms of a Maxwellian Averaged CrossSection (MACS) requires knowledge of the shape of the spectrum. The integral neutronspectrum seen by the targets under the irradiation conditions of the experiment is howevernot measurable. Instead we rely on detailed simulations using the codes SimLiT [50] for thethick-target Li( p, n ) neutron yield, and GEANT4 [51] for neutron transport (Fig. 1) [52].The SimLiT-GEANT4 simulations have been carefully benchmarked in separate experimentsand excellent agreement with experimental time-of-flight and (differential and integral) en-ergy spectra was obtained [50, 52, 53]. The simulated neutron spectrum, dn sim dE n is well fittedin the range E n ∼ −
110 keV ( ∼
90% of the incident neutrons) by a Maxwell-Boltzmann(MB) flux v dn MB dE n ∝ E n exp ( − E n /kT ) with kT ∼
47 keV (Fig. 1). The quantitative normal-ization of the neutron spectrum, dn sim dE n , was obtained by comparing the experimental numberof Au nuclei (measured by gamma activity with a high-purity germanium detector) in theAu foil monitor with the number of
Au nuclei calculated in the detailed simulation of theentire setup (see [52] for details).We calculate the MACS at a given thermal energy kT with the procedure developed in[52, 54], using the expression M ACS ( kT ) = √ π C H - F ( kT ) σ exp where the correction factor9 H - F ( kT ) is given by C H - F ( kT ) = ´ ∞ σ ( E n ) E n e − EnkT dE n ´ ∞ E n e − EnkT dE n / ´ ∞ σ ( E n ) dn sim dE n dE n ´ ∞ dn sim dE n dE n . (1) σ ( E n ) may have coherent contributions from compound-resonances and (weakly energy de-pendent) direct captures (DC). We note here that σ exp includes all contributions in theexperimental energy range; we use in Eq. (1) the Hauser-Feshbach model for the energydependence of σ ( E n ) in the wider MB range and estimate the additional uncertainties as-sociated with direct capture. In order to account for the sensitivity to the low density ofavailable compound states in , Ar, we apply different codes [28]: TENDL-2014 [55], -2015 [48], -2017 [56] and TALYS-1.8 [57] with a microscopic level density and average the C H - F ( kT ) values obtained; the greater of 20% of the correction or their standard deviationis attributed to the MACS corrections. It should however be noted that the extrapolationof the MACS to different thermal energies and determination of their uncertainties weremade using a limited number of theoretical models, due to the total absence of experimentalknowledge of resonances in the , Ar compound nuclei. We also add an estimated inde-pendent 15% uncertainty from s-wave and p-wave DC contributions. Detailed calculationsof the correction factor and its uncertainties will be included in an expanded version of thisLetter. Our MACS values and uncertainties are listed in Table II and compared to existingtheoretical values.The experimental MACS values (Table II) obtained in this work are notably differentfrom previous calculations. Fig. 4 shows the , Ar( n, γ ) reaction rates ( N A (cid:104) σv (cid:105) ) basedon our measurements and extrapolation to different temperatures, compared to the ratesadopted so far [49]. In order to show the potential effect of these experimental rates on stellarnucleosynthesis, we performed a single-zone network calculation using physical conditionsappropriate for the He core burning phase of a massive star in which the new , Ar( n, γ )rates are used, leaving all others unchanged [49]. The calculations are done using the single-zone NucNet Tools reaction network code [58] starting at the H-burning phase with solarabundances [3] and continuing into a single-zone He core burning (T= 300 MK, density of1 kg/cm ). Substantial (10-50%) changes in the calculated mass fractions for neutron-richlight nuclides between S and Fe are observed (Fig. 5), reminiscent of the sensitivityobserved in the weak s -process region (A ∼ Ar itself is observed to increase by a factor of ∼
10 due to its10
ABLE II. Comparison of the experimental thermal cross sections and MACS(30 keV) obtainedin this work to theoretical and evaluated data.Year [Ref.] Ar Arthermal cross section (b)1950 [38] 6.5(10)1968 [39] 5.0(8)1952 [40] 0.8(2)2006 [37] 5.2(5) 0.8(2)This work 3.9(3) 0.68(8)MACS(30 keV) (mb)1978 [41] 6.7 2.61983 [42] 82000 [43] 14 3.92005 [44] 24.6 8.072011 [45–47] 8.86 0.1372015 [48] 8.48 2.82Kadonis [49] 9.0(15) 3.0(3)This work (30 keV) 1.9(3) 1.3(2)This work (47 keV) 1.4(2) 0.92(16) lower measured capture-cross section. Especially interesting is the ∼
45% decrease in thecalculated mass fraction of the important cosmo/geo-chronometer K implying a weakercontribution of the secondary s -process relative to primary production. As shown in Franket al. [60], the mass fraction of K differs considerably over time whether it is primaryonly or secondary only. For example, with a larger primary production of K which is the11 -1 r ea c t i on r a t e , N A < σ v > ( c m / m o l e / s )
10 20 30 40 50 60 70 80 90thermal energy (keV)10 -1 Kadonisexperimental0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1temperature (GK) Ar Ar FIG. 4. (Color online) Comparison of the Ar (top) and Ar (bottom) ( n, γ ) reaction rates( N A (cid:104) σv (cid:105) ) extracted from this work (red) to the Kadonis [49] recommended values (black). Thedashed curves encompass the estimated 1 σ uncertainty. dominant initial heat generator in Earth-like exoplanets, considerable heating would occurin these worlds even early in the Galaxy history.The measurements of the Ar( n, γ ) cross sections affect also the calculation of the natural Ar background activity in the atmosphere, the interpretation of Ar emission rates inunderground nuclear explosion monitoring [16] and the investigation of atmospheric aircirculation [11]. The detection of Ar by AMS demonstrated here opens the way to analternative method for the monitoring of environmental samples [22]. Similarly to Ar,the Ar( n, γ ) Ar reaction contributes to the Ar production rate in the atmosphere [11]12 .20.40.60.81 r a t i o ( e x p ./t heo . ) S Cl Ar Ar Ar K K K Ca Ca Ca Sc Ti Ti -7 -6 -5 -4 m a ss f r a c t i on theoreticalexperimental FIG. 5. (top) Comparison of the mass fractions calculated for stable nuclei between S and Fechanging by >
10% at the end of a single-zone calculation modeling He burning in a massive star,using literature rates [49] (solid circles) or replacing the , Ar( n, γ ) rates with the experimentalvalues from this work (solid squares). We observe a smoother distribution of mass fractions in thevicinity of Ar when using the experimental cross sections. (bottom) Ratio of the mass fractionsusing experimental and literature reaction rates as above. and determines the initial value for the use of Ar as a groundwater dating chronometer[13, 14, 61].In summary, first measurements of the neutron capture cross sections of Ar and Arat stellar energies were performed. The experimental value for Ar, in particular, is smallerthan the one adopted so far from theoretical calculations and evaluations by a factor of ∼
10. Nucleosynthesis calculations for the weak s -process regime using the measured crosssections are shown to increase the mass fraction of Ar by a factor of ∼
10 and lower theresidual mass fraction of neutron-rich nuclides in the region A=36-48 by 10 to 50%. The , Ar( n, γ ) cross sections affect the interpretation of environmental monitoring using Aror Ar as geophysical tracers.We would like to thank the SARAF and LiLiT (Soreq NRC) and the ATLAS operation13taffs for their dedicated help during the experiments. This work was supported in partby the Israel Science Foundation (Grant No. 1387/15), by the Pazy Foundation (Israel),the Israel Ministry of Science (Eshkol Grant No. 18145), the US Department of Energy,Office of Nuclear Physics, under Award No. DE-AC02-06CH11357. D.S.G. acknowledgesthe support by the U.S. Department of Energy, Office of Nuclear Physics, under AwardNo. DE-FG02-96ER40978. This research has received funding from the European ResearchCouncil under the European Unions’s Seventh Framework Program (FP/2007-2013)/ERCGrant Agreement No. 615126. ∗ Corresponding author: [email protected][1] C. F. von Weizs¨acker, The possibility of a dual β -decomposition in potassium, Phys. Z. ,623 (1937).[2] E. Anders and T. Owen, Mars and Earth: origin and abundance of volatiles, Science , 453(1977).[3] K. Lodders, Solar system abundances and condensation temperatures of the elements, Astro-phys. J. , 1220 (2003).[4] R. D. Hoffman, S. E. Woosley, T. A. Weaver, T. Rauscher, and F.-K. Thielemann, The reactionrate sensitivity of nucleosynthesis in type II supernovae, Astrophys. J. , 735 (1999).[5] R. Reifarth, K. Schwarz, and F. K¨appeler, The stellar neutron-capture rate of S: the originof S challenged, Astrophys. J. , 573 (2000).[6] J. Chen, Nuclear data sheets for A = 40, Nucl. Data Sheets , 1 (2017).[7] E. M. Burbidge, G. R. Burbidge, W. A. Fowler, and F. Hoyle, Synthesis of the elements instars, Rev. Mod. Phys. , 547 (1957).[8] H. Beer and R. D. Penzhorn, Measurement of the neutron capture cross section of Ar-40 andan s-process analysis from S-34 to Ca-42, Astron. Astrophys. , 323 (1987).[9] D. D. Clayton, Handbook of the Isotopes in the Cosmos (Cambridge University Press, Cam-bridge, England, 2003).[10] J. Cameron, J. Chen, B. Singh, and N. Nica, Nuclear data sheets for A = 37, Nucl. DataSheets , 365 (2012).[11] H. H. Loosli, H. Oeschger, and W. Wiest, Argon 37, argon 39, and krypton 81 in the atmo- phere and tracer studies based on these isotopes, J. Geophys. Res. , 2895 (1970).[12] B. Singh and J. A. Cameron, Nuclear data sheets for A = 39, Nucl. Data Sheets , 225(2006).[13] J. A. Corcho Alvarado, R. Purtschert, F. Barbecot, C. Chabault, J. Rueedi, V. Schneider, W.Aeschbach-Hertig, R. Kipfer, and H. H. Loosli, Constraining the age distribution of highlymixed groundwater using Ar: A multiple environmental tracer ( H/ He, Kr, Ar, and C) study in the semiconfined Fontainebleau Sands Aquifer (France) Water Resour. Res. ,W03427 (2007).[14] H. H. Loosli, A dating method with Ar, Earth Planet. Sci. Lett. , 51 (1983).[15] P. Schlosser, B. Kromer, G. ¨Ostlund, B. Ekwurzel, G. B¨onisch, H. H. Loosli, and R. Putschert,On the C and Ar Distribution in the Central Arctic Ocean: Implications for Deep WaterFormation, Radiocarbon , 327 (1994).[16] R. Riedmann and R. Purtschert, Natural Ar concentrations in soil air: implications formonitoring underground nuclear explosions, Environ. Sci. Technol. , 8656 (2011).[17] A. Kreisel et al. , Phase-I proton/deuteron linac beam operation status, in Proceedings ofLinac 2014, Geneva (Switzerland) , (WEIOB02, 770, 2014), and references therein, http://accelconf.web.cern.ch/AccelConf/LINAC2014/papers/weiob02.pdf .[18] I. Mardor et al. , The Soreq Applied Research Accelerator Facility (SARAF): overview, researchprograms and future plans, Eur. Phys. J. A , 91 (2018).[19] S. Halfon et al. , High-power liquid-lithium jet target for neutron production, Rev. Sci. Instr. , 123507 (2013).[20] S. Halfon et al. , Note: Proton irradiation at kilowatt-power and neutron production from afree-surface liquid-lithium target, Rev. Sci. Instr. , 056105 (2014).[21] M. Paul et al. , Nucleosynthesis reactions with the high-intensity SARAF-LiLiT neutron source,Proc. Sci., INPC2016 ( ) 139; https://pos.sissa.it/281/139/pdf .[22] M. Paul et al. , Positive-ion accelerator mass spectrometry at ATLAS: peaks and pits, in Pro-ceedings of the Fourteenth International AMS Conference (AMS14), 2017, Ottawa (Canada) ;(to be published.[23] G. Rupp, D. Petrich, F. K¨appeler, J. Kaltenbaek, B. Leugers, and R. Reifarth, High pressuregas spheres for neutron and photon experiments, Nucl. Instrum. Methods Phys. Res., Sect. A , 152 (2009).
24] G. Feinberg, Study of the Li( p, n ) Reaction Towards Measurements of Neutron-Capture CrossSections in the Astrophysical s -process With the SARAF Accelerator and a Liquid-LithiumTarget, Ph.D. thesis, Hebrew University, 2014 (unpublished); http://arad.mscc.huji.ac.il/dissertations/W/JSL/001975970.pdf .[25] M. Paul et al. , First nuclear-astrophysics experiments with high-intensity neutrons form theliquid-lithium target LiLiT, Proc. Sci. NIC, XIII2014 ( ) 059, http://pos.sissa.it/archive/conferences/204/059/NIC%20XIII_059.pdf .[26] Model PFC16A, Centronics Ltd., UK, .[27] Trace Sciences International Corporation, .[28] See Supplemental Material (following these references) for additional information.[29] R. Davis, A review of the homestake solar neutrino experiment, Prog. Part. Nucl. Phys. ,13 (1994).[30] H. H. Loosli, M. M¨oll, H. Oeschger, and U. Schotterer, Ten years low-level counting in theunderground laboratory in Bern, Switzerland, Nucl. Instrum. Methods Phys. Res., Sect. B , 402 (1986).[31] J. P. Renier, H. Genz, K. W. D. Ledingham, and R. W. Fink, ML orbital-electron-capture ratioin Ar decay and the fraction of K α X rays in the K series of chlorine, Phys. Rev. , 935(1968).[32] Thermo Finnigan MAT 253, Thermo Electron Corporation, .[33] P. Collon et al. , Development of an AMS method to study oceanic circulation characteristicsusing cosmogenic Ar, Nucl. Instrum. Methods Phys. Res., Sect. B , 428 (2004).[34] P. Collon et al. , Reducing potassium contamination for AMS detection of Ar with anelectron-cyclotron-resonance ion source, Nucl. Instrum. Methods Phys. Res., Sect. B ,77 (2012).[35] M. Paul, B. G. Glagola, W. Henning, J. G. Keller, W. Kutschera, Z. Liu, K. E. Rehm, B.Schneck, and R. H. Siemssen, Heavy ion separation with a gas-filled magnetic spectrograph,Nucl. Instrum. Methods Phys. Res., Sect. A , 418 (1989).[36] M. Paul, B. Harss, D. Henderson, C. L. Jiang, and K. E. Rehm, A new large area focalplane detector for the ATLAS split pole spectrograph, Argonne National Laboratory, Ar-gonne, Illinois, USA, Physics Division Annual Report No. ANL-97/14, 1997 p. 79, https://inldigitallibrary.inl.gov/Reports/ANL-97-14.pdf .
37] S.F. Mughabghab,
Atlas of Neutron Resonances (Elsevier Science, New York, 2006); dataavailable online: .[38] G. E. McMurtrie and D. P. Crawford, The thermal neutron-capture cross section of A , Phys.Rev. , 840 (1950).[39] P. Wille, The decay of Ar following neutron capture in Ar, Atomkernenergie , 383(1968).[40] S. Katcoff, Thermal neutron capture cross section of A and observation of A , Phys. Rev. , 886 (1952).[41] S. E. Woosley, W. A.Fowler, J. A. Holmes, and B. A. Zimmerman, Semiempirical thermonu-clear reaction-rate data for intermediate-mass nuclei, At. Data Nucl. Data Tables , 371(1978).[42] H. Gruppelaar and H. A. J. van der Kamp, Evaluation of activation cross sections of corrosionproducts, cover-glass nuclides and other nuclides in the primary cooling circuit of a fast powerreactor, Nuclear Data for Science and Technology , edited by K. B¨ockhoff (D. Reidel PublishingCompany, Dordrecht, Holland, 1983), p. 643.[43] T. Rauscher and F.-K. Thielemann, Astrophysical reaction rates from statistical model cal-culations, At. Data Nucl. Data Tables ,1 (2000).[44] S. Goriely, Hauser-Feshbach rates for neutron-capture reactions, Ver. 8/29/2005, .[45] M. B. Chadwick, et al. , ENDF/B-VII.1 nuclear data for science and technology: cross sections,covariances, fission product yields and decay data, Nucl. Data Sheets , 2887 (2011).[46] JEFF-3.2 Library, Joint Evaluated Fission and Fusion (2014), .[47] ROSFOND-2010 Library, Institute of Physics and Power Engineering (2010), .[48] A. J. Koning et al. , TENDL-2015, https://tendl.web.psi.ch/tendl_2015/tendl2015.html .[49] I. Dillmann et al. , KADoNiS The Karlsruhe Astrophysical Database of Nucleosynthesis inStars, AIP Conf. Proc. , 123 (2006); available online: .[50] M. Friedman et al. , Simulation of the neutron spectrum from the Li( p, n ) reaction with aliquid-lithium target at Soreq Applied Research Accelerator Facility, Nucl. Instrum. Methods hys. Res., Sect. A , 117 (2013).[51] S. Agostinelli et al. , Geant4 - a simulation toolkit, Nucl. Instrum. Methods Phys. Res., Sect.A , 250 (2003).[52] M. Tessler et al. , Stellar 30-keV neutron capture in , Zr and the Zr( γ, n Zr photonuclearreaction with a high-power liquid-lithium target, Phys. Lett. B , 418 (2015).[53] G. Feinberg et al. , Quasi-stellar neutrons from the Li( p, n ) Be reaction with an energy-broadened proton beam, Phys. Rev. C , 055810 (2012).[54] M. Tessler et al. , Stellar 30-keV neutron capture in , Zr and the Zr( γ, n Zr photonuclearreaction with a high-power liquid-lithium target, Supp. Mat., Phys. Lett B , 418 (2015); https://ars.els-cdn.com/content/image/1-s2.0-S0370269315008175-mmc1.pdf .[55] A. J. Koning and D. Rochman, Modern nuclear data evaluation with the TALYS code system,Nucl. Data Sheets , 2841 (2012); available online: .[56] A. J. Koning et al. , TENDL-2017, https://tendl.web.psi.ch/tendl_2017/tendl2017.html .[57] A. Koning, S. Hilaire, and S. Goriely, TALYS-1.8, NRG-1755 ZG Petten, The Netherlands(2015), .[58] https://sourceforge.net/projects/nucnet-tools/ .[59] H. Nassar et al. , Stellar ( n, γ ) cross section of Ni, Phys. Rev. Lett. , 092504 (2005).[60] E. A. Frank, B. S. Meyer, and S.J. Mojzsis, A radiogenic heating evolution model for cosmo-chemically Earth-like exoplanets, Icarus , 274 (2014).[61] P. Collon, W. Kutschera, and Z. T. Lu, Tracing noble gas radionuclides in the environment,Annu. Rev. Nucl. Part. Sci. , 39 (2004). upplementary material We present here supplementary material referred to in the paper.
FIG. 6. Ti sphere used as container for the pressurized A Ar gas for irradiation (left), and thesphere target holder for irradiation at SARAF-LiLIT (right).TABLE III. Samples used and the results of the A +1 Ar/ A Ar ratio after the irradiation for all gassamples. Ar and Ar gas samples [27] were enriched to 99.935% and 99.957% for the respectiveisotopes. The final Ar/ Ar ratios were obtained by taking a weighted average of the AMS andLLC results (Fig. 2). Sphere A Ar gas A Ar mass (mg) Irradiation A +1 Ar/ A Ar ratio39 Ar 24.5 LiLiT 8 . × − Ar 19.9 reactor, 20 min., 80 kW, w/ Cd 1 . × − Ar 22.6 reactor, 20 min., 80 kW, w/o Cd 3 . × − Ar 19.5 LiLiT 4 . × − ,nat Ar 12.8 ( Ar/ nat
Ar = 11.7) reactor, 40 sec, 5 MW 8 . × − ,nat Ar 15.8 ( Ar/ nat
Ar = 10.2) reactor, 20 sec, 3 MW 1 . × − IG. 7. Identification spectra of the Ar counts with the ∆E-E telescope of Si detectors for theLiLiT irradiated sphere IG. 10. Same as Fig. 7 for a non irradiated Ar (blank). One Ar count detected for this sampleover 6.5 hours, likely due to a memory effect in the ion source, corresponds to a concentration Ar/ Ar=9 × − . -2 Focal plane position (channel number) D E ( c h a nn e l nu m b e r ) -1 Focal plane position (channel number) D E ( c h a nn e l nu m b e r ) -2 -1 Ar+n
LiLiT Ar Ar1366 cts. background
16 cts.
FIG. 11. Two-dimensional spectra of ∆E vs. focal plane position gated on the Ar region-of-interest in the time-of-flight, ∆E and ∆E detector parameters. ABLE IV. AMS results obtained for the LiLiT irradiated sphere ( Ar/ Ar ratio. Statisticaluncertainties are given in (). An additional systematic uncertainty of 10% is due to the Artransmission.run N (37) time (sec) i (nA) Ar/ Ar ratio40 94(10) 5207.8 1.66(1) 9 . × − . × −
43 37(6) 3866.7 1.1(1) 7 . × −
44 50(7) 3918.6 1.00(5) 1 . × −
45 55(7) 3489.1 1.50(6) 8 . × − weighted 8 . × − average run number0.05.0 × -13 × -12 × -12 A r / A r r a t i o
40 41+42 4443 45weighted average: 8.8±0.5×10 -13
LiLiT irr.sphere need to add a systematic uncertainty of 10%
FIG. 12. Repeated measurements of the Ar/ Ar ratio for the LiLiT irradiated sphere ( ABLE V. Same as Table IV for the Cd shielded sphere irradiated at the reactor ( N (37) time (sec) i (nA) Ar/ Ar ratio52 48(7) 2526.3 1.4(4) 1 . × −
53 122(11) 3135.3 1.9(1) 1 . × − weighted 1 . × − average TABLE VI. Same as Table IV for the sphere irradiated at the reactor ( N (37) time (sec) i (nA) Ar/ Ar ratio54 10990(105) 579.2 2.7(1) 4 . × −
55 9228(96) 563.2 2.85(3) 3 . × −
56 8574(93) 566.6 2.79(3) 3 . × −
57 8595(93) 567.5 2.86(7) 3 . × −
58 7293(85) 564.8 2.7(2) 3 . × −
59 8942(95) 644.9 2.81(2) 3 . × −
60 9402(97) 645.5 2.89(7) 3 . × − weighted 3 . × − average × -10 × -10 × -10 A r / A r r a t i o weighted average: 3.61±0.02×10 -10 reactor irr.sphere need to add a systematic uncertainty of 10% FIG. 13. Same as Fig. 12 for the sphere irradiated at the reactor ( Ar measurements by low-levelcounting (LLC). The peak energy correspond to the 2.82 keV K-capture decay energy of Ar. ABLE VII. LLC results of the Ar/ Ar ratios for all three samples.sample Ar/ Ar ratioLiLiT 8 . × − reactor with Cd 1 . × − reactor (w/o Cd) 3 . × − t r an s m i ss i on ( t o F C P ) Ar Ar FIG. 15. Transmission efficiency of Ar and Ar . × -13 × -13 × -13 × -13 A r / A r r a t i o
53 54 55 56 57 63 69 70 73weighted average = 4.22±0.15×10 -13
LiLiT irr.sphere
FIG. 16. Repeated measurements of the Ar/ Ar ratio for the LiLiT irradiated sphere ( ABLE VIII. Analysis of the LiLiT irradiated sphere ( Ar/ Ar ratio.run N (39) time (sec) i ( µ A) Ar/ Ar ratio53 680(19) 3584 0.69(7) 3 . × −
54 589(20) 3606 0.48(5) 4 . × −
55 756(19) 3855 0.71(7) 3 . × −
56 610(17) 3584 0.56(6) 4 . × −
57 656(18) 3538 0.71(7) 3 . × −
63 659(18) 3653 0.60(6) 4 . × −
69 783(23) 3697 0.64(6) 4 . × −
70 791(23) 3698 0.65(7) 4 . × −
73 374(14) 1826 0.63(6) 4 . × − weighted 4 . × − averageblank 2 . × − final ratio 4 . × −
13 aa including the 3.3% uncertainty of the detector efficiency. ABLE IX. AMS results obtained for the reactor irradiated sphere (54) Ar/ Ar ratio.run N (39) time (sec) i ( µ A) Ar/ Ar ratio76 91629(1392) 3628 0.42(2) 8 . × −
77 65234(1081) 3606 0.28(3) 9 . × −
78 76924(1236) 3102 0.39(4) 8 . × −
79 115416(1847) 3601 0.53(5) 8 . × −
80 103310(1616) 3542 0.48(5) 8 . × −
81 119096(1939) 3381 0.58(6) 8 . × −
82 124323(1979) 3545 0.58(6) 8 . × −
83 119768(1980) 3580 0.55(6) 8 . × −
84 115483(1880) 3587 0.53(5) 8 . × − weighted 8 . × − averageblank 2 . × − final ratio 8 . × −
11 aa including the 3.3% uncertainty of the detector efficiency. × -11 × -10 × -10 × -10 A r / A r r a t i o weighted average = 8.63±0.30×10 -11 reactor irr.sphere FIG. 17. Same as Fig. 16 for the reactor irradiated sphere ( N (39) time (sec) i ( µ A) Ar/ Ar ratio157 3597(47) 900 0.30(2) 1 . × −
158 3349(56) 908 0.25(2) 2 . × −
159 3192(46) 901 0.28(2) 1 . × −
161 3180(53) 900 0.27(2) 1 . × −
163 6674(110) 1231 0.44(3) 1 . × −
164 11207(259) 2037 0.42(3) 1 . × − weighted 1 . × − averageblank 1 . × − final ratio 1 . × −
11 aa including the 3.3% uncertainty of the detector efficiency. × -11 × -11 × -11 × -11 A r / A r r a t i o
157 158 159 161 163 164weighted average = 1.84±0.06×10 -11 reactor irr.sphere
FIG. 18. Same as Fig. 16 for the reactor irradiated sphere ( σ ( ba r n ) S. Katcoff1952 Ar norm. Oct. ’15
Au norm. Oct. ’15 Ar(n, γ ) cross sectionthermal neutron energy Ar norm. Dec. ’15
Au norm. Dec. ’15
FIG. 19. Comparison of the measured Ar thermal cross section using external
Au and internal Ar neutron monitoring.