New portal to the O 15 (α,γ) Ne 19 resonance triggering CNO-cycle breakout

C. Wrede, B. E. Glassman, D. Pérez-Loureiro, J. M. Allen, D. W. Bardayan, M. B. Bennett, B. A. Brown, K. A. Chipps, M. Febbraro, C. Fry, M. R. Hall, O. Hall, S. N. Liddick, P. O'Malley, W. J. Ong, S. D. Pain, S. B. Schwartz, P. Shidling, H. Sims, P. ThompsonH. Zhang

Research output: Contribution to journalArticlepeer-review

11 Scopus citations

Abstract

The O15(α,γ)Ne19 reaction is expected to trigger the initial path for breakout from the CNO hydrogen-burning cycles to the rapid proton capture (rp) process in type I x-ray bursts on accreting neutron stars. The thermonuclear reaction rate has a major impact on models of type I x-ray burst observables and it depends on the small α-particle branching ratio, Γα/Γ, of the 4.03 MeV state in Ne19. Attempts to measure Γα/Γ by populating the 4.03 MeV state using nuclear reactions have only led to strong upper limits. In the present work, we report the first experimental evidence that the 4.03 MeV Ne19 state is populated in Mg20 β-delayed proton emission. This new channel has the potential to provide the necessary sensitivity to detect a finite value of Γα/Γ.

Original languageEnglish
Article number032801
JournalPhysical Review C
Volume96
Issue number3
DOIs
StatePublished - Sep 29 2017

Funding

et al. Wrede C. 1,2 * Glassman B. E. 1,2 † Pérez-Loureiro D. 2 ‡ Allen J. M. 3 Bardayan D. W. 3 Bennett M. B. 1,2 Brown B. A. 1,2 Chipps K. A. 4,5 Febbraro M. 4,5 Fry C. 1,2 Hall M. R. 3 Hall O. 3 Liddick S. N. 2,6 O'Malley P. 3 Ong W.-J. 1,2 Pain S. D. 4 Schwartz S. B. 1,2 Shidling P. 7 Sims H. 8 Thompson P. 4,5 Zhang H. 1,2 Department of Physics and Astronomy, 1 Michigan State University , East Lansing, Michigan 48824, USA National Superconducting Cyclotron Laboratory, 2 Michigan State University , East Lansing, Michigan 48824, USA Department of Physics, 3 University of Notre Dame , Notre Dame, Indiana 46556, USA 4 Oak Ridge National Laboratory , Oak Ridge, Tennessee 37831, USA Department of Physics and Astronomy, 5 University of Tennessee , Knoxville, Tennessee 37996, USA Department of Chemistry, 6 Michigan State University , East Lansing, Michigan 48824, USA Cyclotron Institute, 7 Texas A & M University College Station , Texas 77843, USA 8 University of Surrey , GU2 7XH, Guildford, UK * [email protected][email protected][email protected] September 2017 29 September 2017 96 3 032801 14 February 2017 ©2017 American Physical Society 2017 American Physical Society The O 15 ( α , γ ) Ne 19 reaction is expected to trigger the initial path for breakout from the CNO hydrogen-burning cycles to the rapid proton capture ( r p ) process in type I x-ray bursts on accreting neutron stars. The thermonuclear reaction rate has a major impact on models of type I x-ray burst observables and it depends on the small α -particle branching ratio, Γ α / Γ , of the 4.03 MeV state in Ne 19 . Attempts to measure Γ α / Γ by populating the 4.03 MeV state using nuclear reactions have only led to strong upper limits. In the present work, we report the first experimental evidence that the 4.03 MeV Ne 19 state is populated in Mg 20   β -delayed proton emission. This new channel has the potential to provide the necessary sensitivity to detect a finite value of Γ α / Γ . National Science Foundation http://dx.doi.org/10.13039/100000001 NSF http://sws.geonames.org/6252001/ http://sws.geonames.org/6254928/ PHY-1102511 PHY-1419765 PHY-1404442 U.S. Department of Energy http://dx.doi.org/10.13039/100000015 DOE http://sws.geonames.org/6252001/ http://sws.geonames.org/4138106/ DE-SC0016052 DE-NA0003221 DE-NA0000979 Introduction. Thermonuclear runaways can occur periodically on the surface of a neutron star that is accreting matter from a hydrogen-rich companion star in a close binary system. These events are frequently observed using space-based x-ray telescopes and classified as type I x-ray bursts [1] . Models show that hydrogen burning through the hot carbon-nitrogen-oxygen nucleosynthesis cycles occurs during the initial stages of the burst while temperatures are sufficiently low that the elemental composition of the material is contained below mass number A = 20 [2] . Once sufficiently high temperatures of ≈ 0.4 GK are reached, the rate of the O 15 ( α , γ ) Ne 19 reaction ( Q = 3528.5 ± 0.5 keV [3] ) is expected to become high enough to trigger a nucleosynthesis breakout path from the hot CNO cycles to higher masses, initiating a chain of rapid proton captures and β decays known as the r p process [4] . The O 15 ( α , γ ) Ne 19 reaction rate is expected to determine the temperature and density at which the breakout occurs [2] and, therefore, varying the rate in models of these events can lead to dramatic differences in the predicted x-ray burst light curves and nucleosynthesis ashes [5–8] . A reliable O 15 ( α , γ ) Ne 19 reaction rate for use in the simulations is needed to extract meaningful physics and astrophysics from observations of these extreme cosmic laboratories. Unfortunately, the thermonuclear O 15 ( α , γ ) Ne 19 reaction rate has a large experimental uncertainty [7] . While it has been determined that a single resonance at E c . m . = 505.8 ± 1.0 keV [3,7,9,10] corresponding to a Ne 19 excited state at 4.03 MeV ( J π = 3 / 2 + [7] ) dominates the reaction rate, the strength of the resonance, ω γ , is unknown. It is not currently possible to measure that resonance strength directly because a O 15 ( T 1 / 2 = 122 s) rare-isotope beam of sufficient intensity is not available to bombard a helium target and measure the yield. Fortunately, the resonance strength can be constructed by combining measurements of the level lifetime τ and the small α -particle branching ratio Γ α / Γ using the following expression: (1) ω γ = 2 ℏ τ Γ α Γ 1 − Γ α Γ ≈ 2 ℏ τ Γ α Γ . Three successful measurements of the level lifetime using the Doppler shift attenuation method have yielded consistent finite values that are sufficiently precise for this astrophysical application [10–12] . However, attempts to measure the branching ratio by populating the 4.03 MeV state using nuclear reactions have proved to be more challenging, leading only to strong upper limits [7] of Γ α / Γ < 6 × 10 − 4 [13] , Γ α / Γ < 4.3 × 10 − 4 [14] , and Γ α / Γ = ( 2.9 ± 2.1 ) × 10 − 4 [15] . In the present work, we introduce and substantiate a novel approach to measure the branching ratio of the 4.03 MeV Ne 19 state via nuclear β decay. This new β decay portal has the potential to provide more sensitive measurements than reaction-based methods. Considering that Na 19 is unbound to proton emission, causing it to decay on strong-interaction time scales, its β decay cannot be used to populate the 4.03 MeV state of Ne 19 . This may be the reason that β decay, in general, has apparently been overlooked as an experimental method to investigate the O 15 ( α , γ ) Ne 19 reaction rate. However, the 4.03 MeV state of Ne 19 is energetically accessible in the β -delayed proton decay of Mg 20 ( T 1 / 2 = 91.4 ms [16] , Q EC = 10.627 MeV [3,17,18] ) through Na 20 and, therefore, it may be populated with significant intensity (Fig.  1 ) [19] , but it has never been detected. The β -delayed proton decay of Mg 20 is already known [16,20] to populate low-lying states of Ne 19 including the ground state and the first five excited states up to an excitation energy of 1.62 MeV. In order to be energetically allowed, Mg 20 decay to the seventh excited state of Ne 19 at 4.03 MeV would have to proceed through Na 20 states above 6223 keV excitation energy. While these Na 20 states include the strongly populated isospin T = 2 isobaric analog state (IAS) at 6498 keV [16,20,21] , it is unlikely that the IAS would have a significant proton branch to feed the 4.03 MeV state: proton emission from the IAS is isospin forbidden and the c.m. energy for the transition to the 4.03 MeV Ne 19 state is only 275 keV, so it should also be suppressed by the Coulomb barrier. Let us, therefore, consider the other Na 20 states that are sufficiently high in energy to emit protons to populate the 4.03 MeV state of Ne 19 and sufficiently low in energy to be populated in Mg 20   β decay. The Mg 20   β decay feeding of T = 1 Na 20 states above 6223 keV was recently measured to be 0.67 ± 0.09 % using Mg 20   β -delayed proton decay [16] . If even a small fraction of this Na 20 feeding would undergo proton emission to populate the 4.03 MeV Ne 19 level then a variety of experimental techniques could be used to provide sensitive measurements of Γ α / Γ . We have carried out an experiment to search for the population of the 4.03 MeV state of Ne 19 via the β -delayed proton- γ decay of Mg 20 . 10.1103/PhysRevC.96.032801.f1 1 FIG. 1. Simplified Mg 20 ( T 1 / 2 = 91.4 ms [16] , Q EC = 10627 keV [3,17,18] ) β decay scheme focusing on the transitions relevant to the present work (blue online). Energies are shown in units of keV. The proton separation energy of Na 20 is 2190 keV and the α -particle separation energy of Ne 19 is 3529 keV [3] . Branches are given as intensities in percent of Mg 20   β decays. The values for the Mg 20 ( β + ν ) Na 20 branches are adopted from Ref.  [16] and the value for the Na * 20 ( p ) Ne 4033 * 19 branch shown is from the present work. Experiment. The experiment [21] was carried out at Michigan State University's National Superconducting Cyclotron Laboratory (NSCL) and employed a procedure similar to that of our previous β decay experiments [22–26] . A fast radioactive Mg 20 beam was produced using projectile fragmentation of a 170 MeV/u, 60 pnA Mg 24 primary beam from the Coupled Cyclotron Facility. The beam impinged upon a 961 mg / cm 2 Be 9 target, which transmitted the Mg 20 reaction products to the A1900 fragment separator. The A1900 separated Mg 20 ions from other fragmentation products by magnetic rigidity [27] . Rates of up to 4000 Mg 20 ions s − 1 were delivered to the experimental setup. Beam ions were cleanly identified by combining the time of flight with energy loss. The energy loss was measured using a 300 - μ m -thick silicon detector located ≈ 70 cm upstream of the counting station. The time of flight was measured over a 25 m path between a plastic scintillator at the focal plane of the A1900 and the Si detector. In order to mitigate radiation damage to the Si detector, it was extracted while running with the full Mg 20 beam intensity. These production runs were interleaved with attenuated beam-intensity runs during which the Si detector was inserted for particle identification. The average composition of the beam delivered to the experiment was found to be 34% Mg 20 with the contaminant isotones Ne 18 ( T 1 / 2 = 1.7 s, 24%), F 17 ( T 1 / 2 = 64  s, 12%), O 16 (stable, 22%), and N 15 (stable, 8%) (these values have been refined since [21] ). The Mg 20 ions were implanted to a depth of ≈ 10 mm in a 25-mm thick plastic scintillator. The scintillator recorded the ion implantations and their subsequent β decays with sufficient energy resolution to discriminate between the two. The Segmented Germanium Array (SeGA) of high-purity Ge detectors [28] surrounded the scintillator in two coaxial 13-cm radius rings consisting of eight detectors apiece and it was used to detect γ rays. Signals were processed using the NSCL digital data acquisition system [29] . The SeGA spectra were gain-matched to produce cumulative spectra using the strong γ -ray lines from room-background activity with transition energies of 1460.851 ± 0.006 keV (from K 40 decay) [30] and 2614.511 ± 0.010  keV (from Tl 208 decay) [31] as reference points, providing an in situ first-order energy calibration. In order to reduce the room-background contribution to the γ -ray spectra, a β -coincident γ -ray spectrum was produced by requiring coincidences with β -particle signals from the implantation scintillator (Figs.  2 and 3 ). Lines with well-known transition energies of 1633.602 ± 0.015 ,   3332.84 ± 0.20 ,   6129.89 ± 0.04 ,   8239 ± 4 , and 8640 ± 3 keV [32,33] from the β -delayed γ (and α - γ ) decays of Na 20 (the daughter of Mg 20   β decay) were observed with high statistics and used together with the two room-background lines for a more extensive energy calibration. Small corrections for the energy carried by daughter nuclei recoiling from γ -ray emission were applied throughout the calibration procedure. 10.1103/PhysRevC.96.032801.f2 2 FIG. 2. Energy spectra of SeGA events. The upper spectrum shows all SeGA events in coincidence with events of any energy in the scintillator, including both ion implantations and β decays. This spectrum includes prompt γ rays from nuclear reactions and β -delayed γ rays. The lower spectrum selects events in coincidence with events depositing less than 10 MeV in the scintillator. This spectrum includes β -delayed γ rays and excludes prompt γ rays. γ -ray photopeaks are labeled by the nuclide in which the γ -ray transition occurs. First and second 511-keV γ -ray escape peaks are labeled by one and two asterisks, respectively. 10.1103/PhysRevC.96.032801.f3 3 FIG. 3. (b): the points show the energy spectrum of SeGA events in coincidence with events depositing less than 10 MeV in the scintillator. This spectrum includes β -delayed γ rays and excludes prompt γ rays. The spectrum is identical to the lower spectrum in Fig.  2 , but the binning is different. The error bars associated with the data points are statistical. The smooth line is a functional fit to the data comprised of a Gaussian function added to a linear background. (a): the points show the difference between the data and the linear background component of the fit shown in (b). The smooth line is the Gaussian function derived from the fit shown in (b). The efficiency of the scintillator to detect β decays in coincidence with γ rays was investigated using the SeGA spectra. Comparing the integrals of known β -delayed γ decay lines in the cumulative singles spectrum to the integrals of the corresponding lines in coincidence with scintillator events yielded the efficiency. By considering several such data points, a uniform efficiency of 90 ± 1 % was found for the β decays of Mg 20 to Na 20 ,   Na 20 to Ne 20 , and for the β -delayed proton decay of Mg 20 to Ne 19 . The photopeak efficiency of the SeGA array was determined using measurements with a standard Eu 154 calibration source and geant 4 Monte Carlo simulations. The source was placed on the front face of the scintillator at the center of the SeGA array. It provided absolute efficiency calibration points up to an energy of 1.6 MeV. The geant 4 simulation included the gross features of the experimental geometry and was found to overestimate the absolute photopeak efficiencies by a constant scale factor of 1.03. The relative efficiencies from geant 4 were found to be very accurate and were, therefore, used to interpolate and extrapolate the measured absolute efficiencies to other energies. Discussion. Previously known Ne 19   γ rays from the β -delayed proton decay of Mg 20 [9,20] were observed at 238, 275, 1232, and 1298 keV. In addition, the known Ne 19   γ rays at 1261, 1269, and 1340 keV [9] were observed for the first time in Mg 20   β decay. All of these γ rays are from deexcitations of the five lowest energy excited states of Ne 19 at 238, 275, 1508, 1536, and 1616 keV. Several of the γ -ray peaks were conspicuously Doppler broadened [23,34] due to the recoil of Ne 19 following proton emission from Na 20 and, due to their complex shapes, we reserve a quantitative discussion of those peaks for a more detailed report. We did not observe the population of the sixth Ne 19 excited state at 2795 keV ( J π = 9 / 2 + ) , likely because the allowed β decays of Mg 20 ( J π = 0 + ) populate 0 + and 1 + states of Na 20 , which would need to emit ℓ ≥ 4 protons to feed the 2795 keV Ne 19 state; these proton emissions should be strongly suppressed by the centrifugal barrier. Lastly, and most importantly in the context of the present work, we observed evidence for the population of the seventh Ne 19 excited state at 4.03 MeV in the form of a 4.03 MeV γ -ray peak corresponding to its deexcitation by a transition to the Ne 19 ground state (Figs.  2 and 3 ). The 4.03-MeV γ -ray transition is already known to have a 80 ± 15 % branch deexciting the 4.03 MeV state [9] . To avoid assumptions about the magnitude of Doppler broadening of the peak, we used a simple Gaussian function to fit the peak with the width, centroid, and amplitude as free parameters. In the fit, the peak was summed with a linear background described by two free parameters over the range shown in Fig.  3 . The χ -squared value per degree of freedom for the fit was χ 2 / ν = 67.4 / 47 . Including an additional parameter to describe the curvature of the background did not improve the fit, suggesting that there may be small fluctuations in the background beyond statistical ones. To account for the fluctuations, we inflated the statistical uncertainties of all quantities extracted from the fit by a factor of χ 2 / ν . We also performed a separate fit of the data that was unweighted by the statistical error bars of each bin. An unweighted fit is justified to a good approximation in this case because every bin carries roughly the same statistical weight. The unweighted fit intrinsically captures both statistical and systematic fluctuations of the background in the uncertainties of the fit parameters. The values and uncertainties from the unweighted fit were found to be almost identical to those from the weighted fit with inflated uncertainty. By adopting the results from the weighted fit with inflated uncertainties, the integral of the peak was found to be 2684 ± 503 counts: 5.3 standard deviations above the expected background level. The measured γ -ray energy of 4033.4 ± 1.7 keV in the laboratory reference frame corresponds to an excitation energy of 4033.8 ± 1.7 keV, which is in good agreement with the evaluated literature value of 4034.3 ± 0.9 keV [7,9,10] for the Ne 19 transition. Using the integral of the peak and applying the scintillator and SeGA efficiency calibrations anchored by the known intensity of the strong 984 keV Na 20 line [16,20] yields an intensity of 0.0125 ± 0.0020 % for the 4.03 MeV γ ray in Mg 20   β decay. This value corresponds to a β -delayed proton feeding of 0.0156 ± 0.0038 % for the 4.03 MeV level of Ne 19 after the 20% γ -decay branch of this level to excited states [9] is taken into account. Our experiment was not sensitive to the weaker branches via β - γ or β - γ - γ coincidences. The measured intensities are compatible with the 0.67% Mg 20   β decay feeding of isospin T = 1 Na 20 levels that are energetically allowed to feed the 4.03 MeV Ne 19 state [16] . In particular, the measurements suggest that approximately 2% of the proton emissions from these levels feed the 4.03 MeV Ne 19 state rather than lower lying Ne 19 states, which is consistent with expectations based on a simple barrier penetration model. This is the first detection of the population of the 4.03 MeV Ne 19 state via β decay and it opens a potentially sensitive new channel that can now be exploited to measure Γ α / Γ . Each event will involve a β - p - α decay sequence in which the proton carries ≈ 0.5 –1.0 MeV of kinetic energy and the α particle shares ≈ 0.5 MeV with the O 15 recoil. By taking advantage of coincidences between the proton and the α particle (and potentially, but not necessarily, the O 15 recoil), background events can be strongly suppressed. This measurement could be realized by thermalizing Mg 20 in a time-projection chamber (TPC), for example, and identifying the individual decay products inside using their characteristic Bragg curves. Alternatively, Mg 20 could be trapped in vacuum using electromagnetic fields and the decay products could be observed with surrounding detectors. Either of these methods could yield an efficiency approaching 100% for the detection of the events of interest. Considering a Mg 20 production rate of 4000 per second (already realized at NSCL, for example, in the present experiment) and the 0.0156% feeding of the 4.03 MeV Ne 19 level in Mg 20   β decay, this state will be populated 37 times per minute on average. Assuming Γ α / Γ = 3 × 10 − 4 [15] , approximately 16 α -particle emissions from this level would occur every day. A week-long experiment would yield on the order of 100 events, corresponding to 10% statistical precision on the value of Γ α / Γ assuming an efficient detection system with negligible background. If the model-dependent value of Γ α from Ref.  [35] is adopted instead, then the count-rate estimate is reduced by a factor of ≈ 3 . Potential backgrounds will have to be assessed carefully for specific experimental configurations, but the unique signatures of the events of interest including the particle identities, their energies, and the coincidence condition should enable a strong suppression of background events. In the case of a TPC measurement, a special signature is available to identify the events of interest: a relatively dense energy deposition from the α -particle emission located at the base of the proton's Bragg curve. The present value for the feeding of the 4.03 MeV Ne 19 level will be necessary to normalize the value of Γ α / Γ in future measurements if a sensitive γ -ray spectrometer is not employed. Next-generation rare-isotope-beam facilities currently under construction will yield orders of magnitude more Mg 20 enabling precision studies. Conclusions. We have reported the first experimental evidence for the population of the 4.03 MeV state of Ne 19 via Mg 20   β -delayed proton emission. We find that the 4.03 MeV state is populated in 0.0156% of Mg 20   β decays, providing a new portal for sensitive measurements of the α -decay branching ratio, which determines the conditions for breakout from the hot CNO cycles during type I x-ray bursts on accreting neutron stars. Acknowledgments. We gratefully acknowledge the NSCL staff for working on the data acquisition system and providing the Mg 20 beam. This work was supported by the U.S. National Science Foundation under Grants No. PHY-1102511, No. PHY-1419765, and No. PHY-1404442, the U.S. Department of Energy, Office of Science, under Award No. DE-SC0016052, and the U.S. Department of Energy, National Nuclear Security Administration under Awards No. DE-NA0003221 and No. DE-NA0000979.

Fingerprint

Dive into the research topics of 'New portal to the O 15 (α,γ) Ne 19 resonance triggering CNO-cycle breakout'. Together they form a unique fingerprint.

Cite this