First detection of the 2+ → 0+ ground-state transition in the β decay of 20 F O. S. Kirseboma,∗, M. Hukkanenb , A. Kankainenb , W. H. Trzaskab , K. Andersena , L. Caneteb , J. Cederkällc , T. Enqvistd , T. Eronenb , H. O. U. Fynboa , S. Geldhofb , R. de Grooteb , D. G. Jenkinse , A. Jokinenb , P. Joshie , R. Julinb , A. Khanamb,∗∗, J. Kostensalob , P. Kuusiniemid , I. Mooreb , M. Muncha , D. Nesterenkob , S. V. i Onsèsf , J. D. Ovejazf , H. Penttiläb , I. Pohjalainenb , M. Reponenb , S. Rinta-Antilab , K. Riisagera , A. de Roubinb , P. C. Srivastavag , J. Suhonenb , J. A. Swartza , O. Tengbladf , M. Vilenb , J. Äystöb a Department of Physics and Astronomy, Aarhus University, DK-8000 Aarhus C, Denmark of Jyvaskyla, Department of Physics, FIN-40014 University of Jyvaskyla, Finland c Department of Physics, Lund University, SE-22100 Lund, Sweden d University of Oulu, Oulu Southern Institute, FIN-90014, Finland e Department of Physics, University of York, York YO10 5DD, United Kingdom f Instituto de Estructura de la Materia, CSIC, E-28006 Madrid, Spain g Department of Physics, Indian Institute of Technology, Roorkee 247667, India b University Abstract We report the first detection of the second-forbidden, non-unique, 2+ → 0+ , ground-state transition in the β decay of F. A low-energy, mass-separated 20 F beam produced at the IGISOL facility in Jyväskylä, Finland, was implanted in a thin carbon foil and the β spectrum measured using a magnetic transporter and a plastic-scintillator detector. The +0.00 branching ratio inferred from the observed β yield is [1.10 ± 0.21(stat) ± 0.17(sys)−0.11 (theo)] × 10−5 corresponding to log f t = 10.47(11), making this the strongest known second-forbidden, non-unique transition. The experimental result is supported by shell-model calculations and has important astrophysical implications. 20 Keywords: 20 F, β decay, second-forbidden, non-unique 1. Introduction 5 10 15 20 Recent studies have highlighted the importance of the second-forbidden, non-unique, electron-capture transition 25 from the 0+ ground state of 20 Ne to the 2+ ground state of 20 F for the final evolution of Super-AGB stars [1, 2, 3]. The strength of the transition is, however, not well constrained, neither experimentally nor theoretically, making an experimental determination highly desirable. The strength may be determined from the branching ratio of the inverse 2+ → 0+ transition in the β decay of 20 F (Fig. 1), but this transition is not easily detected as it is masked by the much faster, allowed, 2+ → 2+ transition to the firstexcited state in 20 Ne. Indeed, previous attempts to detect the 2+ → 0+ transition have been unsuccessful [6, 7, 8] yielding a rough upper limit of ∼ 10−5 on the branching ratio [8]. The β-decay endpoint energies for the groundstate and first-excited state transitions are 7.025 MeV and 5.391 MeV, respectively [4], leaving a rather narrow energy window for the detection of the ground-state transition.1 In this Letter, we report the first successful measurement of the second-forbidden, non-unique, 2+ → 0+ transition ∗ Corresponding author address: Aalto University, P.O. Box 11000, FIN-00076 Aalto, Finland Email address: oliskir@phys.au.dk (O. S. Kirsebom) 1 The endpoint energies are known to sub-keV precision [4, 9]. ∗∗ Present Preprint submitted to Physics Letters B in the β decay of 20 F and present the results of shell-model calculations which provide additional support for our result. The astrophysical implications for the evolution of Super-AGB stars are dealt with elsewhere [10]. 2+ 7.025 20 F 19 t1/2 = 11.163(8) s 5.788 1− −6 5.621 3− 8 .2 (6 ) × 10 −5 4.967 2− 0 .999907 1.634 2+ 0.0 0+ < 7 × 10 −7 < 1 .5 × 10 β 1 .1 (3 ) × 10 −5 20 Ne 10 Figure 1: 20 F β-decay scheme [4, 5] including the newly observed ground-state transition. Energies are in MeV relative to the 20 Ne ground state. April 9, 2018 2. Experimental setup 35 40 Below, we briefly describe the experimental setup and procedures; a detailed account will be published separately [11]. The experiment was performed at the IGISOL-4 facility of the JYFL Accelerator Laboratory in Jyväskylä, 65 Finland [12]. Radioactive ions of 12 B and 20 F were produced via (d, p) reactions on targets of B and BaF2 at energies of 9 MeV and 6 MeV, respectively. The ions were accelerated to 30 keV and separated based on their massto-charge ratio, before being guided to the experimental 70 station where they were stopped in a thin (50 µg/cm2 ) carbon foil. The detection system, shown in Fig. 2, consisted of a Siegbahn-Slätis type intermediate-image magnetic electron transporter combined with an energy-dispersive detector. Such an arrangement is well suited for the measure- 1) 2) 3) 4) Plasti s intillator LaBr3 (Ce) rystal Shield Anti-positron bae Plasti Al Fe Brass 4 1 , 3 Coils 0 75 3. Data analysis and results 80 By only allowing electrons within a relatively narrow energy band to reach the detector, the magnetic transporter effectively “carves out” a slice of the β spectrum. This is clearly seen in Fig. 3 which shows energy spectra measured at three different magnetic-field strengths. The central energy selected by the magnetic transporter is apeβ ' 7.72I˜ + 3.01I˜2 MeV, where proximately given by E ˜ I is the electrical current expressed as a fraction of the maximum current provided by the power supply (700 A). 10 m 2 γ 20 F β Figure 2: Schematic diagram of the setup. 45 50 55 three detectors (veto, front and main) are plastic scintillators read out with silicon photomultipliers (SiPM). The detector dimensions represent a compromise between the requirement to fully stop a significant fraction of the most energetic electrons (the nominal range of 7-MeV electrons in plastic is 35 mm) and the requirement to minimize the cosmic-ray exposure. A small LaBr3 (Ce) crystal placed inside the shield on the centre axis is used to detect the 1.63MeV γ ray from the decay 20 F, thereby providing absolute normalisation of the β spectrum. Finally, a baffle placed at the centre of the magnet prevents positrons, which spiral in the opposite direction of electrons, from reaching the detector thereby eliminating positron emitters as a potential source of background, while only reducing the electron flux by 11%. ment of rare ground-state transitions in nuclear β decays as the effective solid angle of the detector, and hence the count rate, is greatly increased by the focussing action of the magnetic field. Furthermore, and equally important, the shield on the centre axis prevents γ rays and electrons from excited-state transitions from reaching the detector. This essentially eliminates βγ summing and ββ pile-up as sources of background and leads to an improved sensitivity towards the ground-state transition. The magnetic transporter was constructed at the JYFL Accelerator Laboratory in 1980s [13], but has been fully refurbished for the present experiment. The β detector has 85 the shape of a cylinder and consists of a 5-mm thick outer detector, used as a veto against cosmic rays, and a 45 mm × 45 mm inner detector, used to measure the full energy of the electrons. The inner detector is further subdivided into a 5-mm thick front detector and a 40-mm thick main 90 detector to provide additional discriminatory power. All 2 Counts / 100 keV / 105 decays 30 60 103 20 F ~ I = 35.3% Exp. Sim. 102 12 B ~ I = 67.7% 12 B ~ I = 79.0% 10 1 10−1 10−2 0 1 2 3 4 5 6 7 8 9 10 E β (MeV) Figure 3: Comparison of experimental and simulated energy spectra obtained at 35.3% (20 F), and 67.7% and 79.0% (12 B) of the maximum electrical current. The 20 F data have been subject to both veto and front cut, while the 12 B data have been subject only to the veto cut. The spectrum obtained at I˜ = 35.3% shows a central slice of the allowed β spectrum of 20 F, while the spectra obtained at 67.0% and 79.0% show slices of the β spectrum of 12 B which has an end-point energy of 13.37 MeV. The spectrum obtained at 35.3% has been cleaned by requiring that no coincident signal is recorded in the veto detector (veto cut) and that the energy deposited in the front detector is between 650–1600 keV (front cut). The spectra 100 105 110 115 120 125 130 135 140 145 3 Counts / 10 keV 3 ×10 105 20 104 10 103 1.4 10 1.6 1.8 2 10 0 1 2 3 4 5 6 7 8 9 E γ (MeV) Figure 4: γ spectrum obtained at 67.7% of the maximum electrical current. The arrow indicates the 1.63-MeV line from the decay of 20 F and the inset shows a Gaussian fit to the peak. Counts / 100 keV 95 obtained at 67.0% and 79.0% have also been subject to the veto cut, but the front cut could not be applied because the 12 B measurements were performed with an earlier version of the β detector, identical to the one described in Sec. 2 except that the inner detector is not divided into a front and a main section. The spectral shape is seen to be well reproduced by the GEANT4 simulation, especially at the lower current setting. Larger deviations are found at the higher current settings, but the accuracy remains sufficient for our needs. The 20 F and 12 B data (only a subset of which is shown in Fig. 3) and data obtained with an absolutely calibrated 207 Bi source, have been used to validate the absolute accuracy of the GEANT4 simulations all the way up to eβ = 8.0 MeV. For the raw data (i.e. before application E of the veto and front cuts) we find good agreement between the experimental and simulated peak efficiency (the number of counts in the full-energy peak normalized to the number of decays) across the full energy range. However, the simulation tends to over-estimate the peak efficiency for the cleaned data, partly due to the presence of what appears to be electronic cross-talk between the inner and the outer detectors, but also due to inaccuracies in the modelling of the stopping process in the detector eβ < 3.0, volumes. The required correction is small for E but grows with increasing energy and amounts to a factor of 1.25(7) at 6.0 MeV. Additionally, we observe small temporal variations in the peak efficiency throughout the experiment. These variations occur in response to slight changes in beam optics which affect the source geometry, and amount to a 13%-spread in peak efficiency which we include as a systematic uncertainty on the final result. Long-duration measurements were performed at the current settings I˜ = 67.7% (67 h) and 70.5% (38 h) to search for a signal in the energy region 5.4–7.0 MeV, and at 79.0% (37 h) to demonstrate that any signal detected at the two lower settings did not persist above 7.0 MeV. The average 20 F implantation rate for these measurements was 11 kHz. Additionally, background data were collected at150 67.7% and 70.5% (183 h). The γ and β spectra obtained at 67.7% are shown in Fig. 4 and 5, respectively. In the γ spectrum we clearly see the well-resolved 1.63-MeV line from the decay of 20 F which is used for absolute normalisation. The cosmic-ray background dominates the raw β155 spectrum above 5.4 MeV while electrons from the allowed 2+ → 2+ transition produce the the bump centered at 5.0 MeV and the continuum below it. In the signal region the cosmic-ray background is reduced by a factor of 105 by the veto cut. The front cut removes another factor of160 3.35 resulting in an overall reduction factor of 352. On the other hand, about 2/3 of the β particles survive the cuts, a figure which would have been even larger in the absence of cross-talk between the inner and outer detectors. Fig. 6 shows a zoom-in on the cleaned spectrum in165 the region of interest. The spectra obtained at 67.7% and 70.5% reveal a clear excess of counts in the signal region when compared to the background spectrum. Equally im- No cuts Veto and front cut Fit region 104 103 102 10 1 0 1 2 3 4 5 6 7 8 9 10 E β (MeV) Figure 5: β spectrum with cuts (gray) and without (yellow) obtained at 67.7% of the maximum electrical current. The dashed box indicates the region used for the likelihood fits shown in Fig. 6. portant, no excess of counts is observed above the signal region in the data obtained at 79.0%. For example, the spectrum at 67.7% has 55 ± 7 counts between 5.8–6.8 MeV while the background spectrum, downscaled to account for the longer measurement time, only has 30 ± 3 counts in the same region. Furthermore, an analysis of the temporal distribution of the counts between 5.8–6.8 MeV reveals a clear correlation with the 20 F implantation rate, which varied by more than a factor of two during the experiment, while the temporal distribution of the counts above 7.0 MeV shows no such correlation. Thus, the observed signal is consistent with being due to the hitherto unobserved, second-forbidden, ground-state transition in the β decay of 20 F. To quantify this statement and obtain an accurate determination of the branching ratio, we perform a likelihood fit to the data between 5.0–8.0 MeV, in which we allow the normalisations of the simulated energy spectra of the al- Exp. 2++ → 0++ 2 →2 Backgr. Fit 2 10 (a) ~ I = 67.7% 20 F (2.4 × 109) 67 h 185 10 1 103 (b) ~ I = 70.5% 20 F (1.5 × 109) 38 h 102 goodness of fit worsens to χ2 /N = 171.8/113 = 1.52 corresponding to a significance level of only Pχ2 >171.8 = 0.0003, providing clear evidence for a positive observation. In Fig. 7 we show the dependence of the goodness of fit on the assumed end-point energy of the forbidden transition. The best fit is clearly obtained by adopting the actual end-point energy of 7.025 MeV. If we require a significance χ 2/ N Counts / 100 keV 103 10 1.45 N = 112 1.4 1.35 1% 1 1.3 103 (c) ~ I = 79.0% 20 F (1.8 × 109) 37 h 102 1.25 5% 1.2 10 1.15 6 7 8 103 (d) ~ I = 67.7%, 70.5% Backgr. 183 h 102 5 5.5 6 6.5 7 7.5 8 E β (MeV) 195 Figure 6: Likelihood fits to the experimental data in the energy region 5.0–8.0 MeV obtained with beam (a–c) and without (d), at current settings focused on the region of interest (67.7% and 70.5%) and immediately above it (79.0%). The contribution of the secondforbidden, ground-state transition in the β decay of 20 F is shown by200 the thin, solid (blue) curve. 180 11 12 190 1 175 10 Figure 7: Dependence of the goodness of fit on the assumed endpoint energy of the forbidden transition. The 5% and 1% significance levels are shown by the dashed horizontal lines. The actual end-point energy of 7.025 MeV is indicated by a star. 10 170 9 End-point energy (MeV) 1 lowed and forbidden transitions to vary freely while the background is modelled by a simple exponential function 205 with two free parameters. While the normalisation of the allowed transition is, in principle, fixed, in practice it is necessary to allow the normalisation to vary because the GEANT4 simulation becomes inaccurate in the low-energy tail of the trasmission window. We also allow for a small (< 50 keV) constant energy shift to account for inaccuracies in the energy calibration and allow for small varia-210 tions in the intrinsic detector resolution which is around 5% (FWHM). Assuming an allowed shape for the forbidden transition, a simultaneous fit to the four spectra shown in Fig. 6 yields a branching ratio of 1.10(21) × 10−5 and a goodness of fit of χ2 /N = 133.9/112 = 1.20 corresponding to an acceptable significance level of Pχ2 >133.9 = 0.076. If,215 on the other hand, we fix the branching ratio to zero, the 4 level of 1% or above, end-point energies above 10 MeV and below 6.2 MeV are excluded by the fit; if the cut is made at 5%, the range of acceptable end-point energies shrinks to 6.7–8.0 MeV. Thus, it seems unlikely that an unknown β − -unstable beam contaminant should be the cause of the observed signal. This is well in line with expectations as 20 F is the only β emitter with mass 20 produced by the 19 F(d, p) reaction at 6 MeV. Measurements performed on neighboring masses and on mass 40 were used to rule out the possibility that the signal was due to a β emitter with a mass different from 20, transmitted to the setup through the tails of the acceptance window of the dipole magnet, as a molecule or as a doubly-charged ion. Taking into account the uncertainties related to the normalisation of the β spectrum discussed above, our result for the branching ratio of the forbidden transition is [1.10 ± 0.21(stat) ± 0.17(sys)]×10−5 , which translates into log f t = 10.47(11) when statistical and systematical uncertainties are added in quadrature. 4. Theoretical calculations In order to validate our experimental result, we have performed shell-model calculations using two ab initio approaches known as the in-medium similarity renormalization group (IM-SRG) [14] and the coupled-cluster effective interaction (CCEI) [15]. For comparison, we have also used 225 Branching ratio × 105 220 the phenomenological USDB interaction [16]. The calculations were done with the code NUSHELLX [17]. The formalism used for the calculation of the forbidden, nonunique β-decay transitions is that of Ref. [18]. Details on the implementation of the formalism, including the use of the next-to-leading-order nuclear-matrix elements, can be found in Ref. [19]. To check the accuracy of the wave functions, we first calculate the electromagnetic properties 20 F and 20 Ne and compare to experimental data. As seen in Tables 1 and 2, all three interactions reproduce the magnetic dipole moments reasonably well, and two of the interactions (CCEI and USDB) also do a fairly good job at reproducing the electric quadrupole moments. The accuracy of the wave Level µexp a 20 F 2+ 1 2+ 1 4+ 1 +2.09335(9) +1.08(8) +1.5(3) 20 Ne a Refs. b Using IM-SRG µtheo b CCEI USDB +2.171 +1.036 +2.086 +2.183 +1.037 +2.095 +2.092 +1.020 +2.052 Qexp a 20 F 2+ 1 2+ 1 0.0547(18) −0.23(3) 20 Ne a Refs. b Using 230 235 240 245 −0.0515 +0.0359 Qtheo CCEI USDB +0.0729 −0.1578 +0.0679 −0.1576 1.2 1 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 gA Figure 8: Dependence of the calculated branching ratio on the value adopted for the effective axial-vector coupling constant, gA . The dots mark the values required to reproduce the experimental half-life of 20 F. The experimental branching ratio is shown by the horizontal band (statistical and systematic uncertainties added in quadrature). The dashed lines show the effect of using the shape predicted by the shell-model calculations instead of the allowed shape. b IM-SRG 1.4 0.2 0.6 255 Level 1.6 0.4 Table 2: Comparison of experimental and theoretical quadropole moments in units of eb. Nucleus 1.8 0.6 250 [20, 21]. the free g factors for sd shell nuclei. USDB CCEI IM-SRG Exp. (allowed shape) Exp. (theo. shape) 2 0.8 Table 1: Comparison of experimental and theoretical magnetic dipole moments in units of µN . Nucleus 2.2 260 [22, 23]. ep = 1.5e and en = 0.5e. functions can also be tested by comparing the predicted half life of 20 F to the experimental value of 11.163(8) s [4]. For IM-SRG and USDB, the experimental half life is repro-265 duced using an effective axial-vector coupling constant of gA = 1.06, while gA = 1.12 must be used for CCEI. This is well in line with previous studies in the sd shell, which have found that an effective gA value close to unity is needed to reproduce the experimental half-lives [24, 25, 26]. Since our calculations predict several key observables270 in 20 F and 20 Ne, we expect to obtain reasonably accurate estimates also for the second-forbidden, non-unique 2+ → 0+ transition. The results for the branching ratio are shown in Fig. 8 where it is seen that the branching ratio depends both on the interaction and the assumed value for275 gA . Overall, there is good agreement between theory and experiment with theory underestimating the experimental value somewhat. Using the values for gA that give the correct half life for the allowed transition, the branching ratio comes out as 5.3 × 10−6 (USDB), 5.4 × 10−6 (CCEI) and280 7.7 × 10−6 (IM-SRG). 5 We have also calculated the half life of the 2+ → 0+ transition in the β decay of 36 Cl which was the only secondforbidden, non-unique transition measured in the sd shell prior to this work. For this case, the choice of interaction has a large impact on the calculated strength, resulting in partial half-lives ranging from (6–7) × 104 yr (USDB) to ∼ 107 yr (CCEI), which may be compared to the experimental half life of 3.07(2) × 105 yr. The very poor performance of the CCEI interaction in this case is not surprising as the interaction has not been designed for the mass region near A = 36. On the other hand, the poor performance of the USDB interaction (a factor of five deviation for 36 Cl compared to a factor of two for 20 F) is unexpected and poses the question whether the good agreement for 20 F is in part coincidental. 5. Discussion The experimental branching ratio given in Sec. 3 was obtained assuming that the 2+ → 0+ transition has an allowed shape even though significant deviations from the allowed shape can occur. An assumption about the shape had to be made in order to convert the observed β yield between 5.4–7.0 MeV to a total β yield, as the experimental data have very limited sensitivity towards the spectral shape. To quantify the model dependency introduced by this assumption, we have repeated our analysis using the shapes predicted by our shell-model calculations and find that the branching ratio is reduced by 10%, improving the agreement with the theoretical predictions somewhat. We include this as a theoretical uncertainty on our final result. With a log f t value of 10.5, the 2+ → 0+ transition in the β decay of 20 F is, to our knowledge, the strongest second-forbidden, non-unique transition ever measured, as 285 290 295 300 305 310 the 27 previously known cases have log f t values ranging from 10.6 to 14.2 [27]. Shell-model calculations are known to reproduce the strengths of second-forbidden, unique transitions in the sd shell within a factor of two or better [28, 29]. However, little appears to be known about the accuracy for330 second-forbidden, non-unique transitions in the sd shell, such as the 2+ → 0+ transition in the β decay of 20 F, though we should expect the accuracy to be worse or, at the best, comparable. The astrophysical importance of the 2+ → 0+ transition was first pointed out by Ref. [1]335 and a shell-model calculation based on the USDB interaction was published soon after by the same team [30]. The branching ratio was erroneously given as 1.3 × 10−6 while the result was in fact 1.3 × 10−5 using Wood-Saxon radial wave functions. Moreover, when Harmonic-Oscillator340 wave functions were used, the value obtained was in agreement with the value of 5.3 × 10−6 obtained in this Letter. It would be of great interest to clarify the origin of the factor of five discrepancy between theory and experiment in the case of 36 Cl (see Sec. 4), but this requires a345 careful theoretical analysis which is beyond the scope of this Letter and is left for future work. Notwithstanding these issues, it is encouraging that the calculations give correct order-of-magnitude estimates for two transitions with markedly different strengths that span almost three orders of magnitude (log f t = 10.5 in the case of 20 F com-350 pared to log f t = 13.3 in the case 36 Cl). 6. Conclusion 355 + + The second-forbidden, non-unique, 2 → 0 groundstate transition in the β decay of 20 F has been observed for the first time. The detection was made possible by the development of a dedicated experimental setup consisting360 of Siegbahn-Slätis type intermediate-image magnetic electron transporter combined with a plastic-scintillator telescope. The branching ratio was determined to be 365 h i +0.00 1.10 ± 0.21(stat) ± 0.17(sys)−0.11 (theo) × 10−5 , 315 320 325 in agreement with upper limits established in previous ex370 periments and implying log f t = 10.47(11), which makes this the strongest second-forbidden, non-unique transition ever measured. This remarkable result is supported by shell-model calculations which yield correct order of mag-375 nitude estimates not only for this transition, but also for the much weaker 2+ → 0+ transition in the β decay of 36 Cl (log f t = 13.3). The result has important astrophysical implications which are discussed elsewhere [10]. While380 the inferred branching ratio was found to be fairly insensitive to the assumed spectra shape, the impact on the astrophysical electron-capture rate could be substantial (a factor of 4–10 according to Ref. [1]), and hence future ex-385 periments should aim to provide improved constraints on the shape. 6 Acknowledgements We thank G. Martínez-Pinedo and A. Idini for enlightening discussions on the topic of shell-model calculations of forbidden β-decay transitions and for informing us about the typo in Ref. [30]. We are greatly indebted to the technical staff at the JYFL laboratory and Aarhus University who contributed with their time and expertise to the refurbishment of the spectrometer. We would also like to thank P. Schotanus and E. Bodewits at SCIONIX Holland B.V. for a fruitful collaboration on the design of the β detector. We thank F. Lyckegaard for making the BaF2 targets, H. Kettunen for preparing the catcher foil and E. Nacher for providing technical support with the GEANT4 simulations. Finally, we thank T. Kibedi for valuable advice and encouragement in the early phases of the project. This work has been supported by the Academy of Finland under the Finnish Centre of Excellence Programme 2012– 2017 (Nuclear and Accelerator Based Physics Research at JYFL) and the Academy of Finland grants No. 275389, 284516, 312544 and XXXXX. OSK acknowledges support from the Villum Foundation. References [1] G. Martínez-Pinedo, Y. H. Lam, K. Langanke, R. G. T. Zegers, C. Sullivan, Astrophysical weak-interaction rates for selected A = 20 and A = 24 nuclei, Phys. Rev. C 89 (2014) 045806. doi:10.1103/PhysRevC.89.045806. URL http://link.aps.org/doi/10.1103/PhysRevC.89.045806 [2] J. Schwab, E. Quataert, L. Bildsten, Thermal runaway during the evolution of ONeMg cores towards accretioninduced collapse, Monthly Notices of the Royal Astronomical Society 453 (2) (2015) 1910–1927. arXiv:http://mnras. oxfordjournals.org/content/453/2/1910.full.pdf+html, doi:10.1093/mnras/stv1804. URL http://mnras.oxfordjournals.org/content/453/2/ 1910.abstract [3] J. Schwab, L. Bildsten, E. Quataert, The importance of urcaprocess cooling in accreting one white dwarfs, Monthly Notices of the Royal Astronomical Society 472 (3) (2017) 3390– 3406. arXiv:/oup/backfile/content_public/journal/mnras/ 472/3/10.1093_mnras_stx2169/1/stx2169.pdf, doi:10.1093/ mnras/stx2169. URL +http://dx.doi.org/10.1093/mnras/stx2169 [4] D. R. Tilley, C. M. Cheves, J. H. Kelley, S. Raman, H. R. Weller, Energy levels of light nuclei, A =20, Nuclear Physics A 636 (3) (1998) 249–364. doi:10.1016/S0375-9474(98)00129-8. [5] E. L. Wilds, Jr., R. H. France III, J. E. McDonald, Z. Zhao, M. Gai, Upper limits on the first-forbidden rank-one β decay of 20 F, Phys. Rev. C 76 (2007) 018501. doi:10.1103/PhysRevC.76.018501. URL https://link.aps.org/doi/10.1103/PhysRevC.76. 018501 [6] C. Wong, Beta decay of f20 , Phys. Rev. 95 (1954) 761–764. doi:10.1103/PhysRev.95.761. URL http://link.aps.org/doi/10.1103/PhysRev.95.761 [7] S. S. Glickstein, R. G. Winter, Phys. Rev. 1281–1283doi:10. 1103/PhysRev.129.1281. [8] F. P. Calaprice, D. E. Alburger, Beta spectrum of f20 , Phys. Rev. C 17 (1978) 730–738. doi:10.1103/PhysRevC.17.730. URL http://link.aps.org/doi/10.1103/PhysRevC.17.730 [9] M. Wang, G. Audi, F. Kondev, W. Huang, S. Naimi, X. Xu, The ame2016 atomic mass evaluation (ii). tables, graphs and references, Chinese Physics C 41 (3) (2017) 030003. URL http://stacks.iop.org/1674-1137/41/i=3/a=030003 390 395 400 405 410 415 420 425 430 435 440 445 450 455 [10] O. S. Kirsebom, in preparation. 460 [11] O. S. Kirsebom, in preparation. [12] I. Moore, T. Eronen, D. Gorelov, J. Hakala, A. Jokinen, A. Kankainen, V. Kolhinen, J. Koponen, H. Penttilä, I. Pohjalainen, M. Reponen, J. Rissanen, A. Saastamoinen, S. RintaAntila, V. Sonnenschein, J. Äystö, Towards commissioning the465 new IGISOL-4 facility, Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 317, Part B (2013) 208–213, {XVIth} International Conference on ElectroMagnetic Isotope Separators and Techniques Related to their Applications, December 2–7, 2012 at470 Matsue, Japan. doi:10.1016/j.nimb.2013.06.036. URL http://www.sciencedirect.com/science/article/pii/ S0168583X13007143 [13] R. Julin, J. Kantele, J. Kumpulainen, M. Luontama, V. Nieminen, A. Passoja, W. Trzaska, E. Verho, A setup for spectrom-475 etry of high-energy conversion electrons, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 270 (1) (1988) 74–77. doi:10.1016/0168-9002(88)90011-3. URL http://www.sciencedirect.com/science/article/pii/480 0168900288900113 [14] S. R. Stroberg, H. Hergert, J. D. Holt, S. K. Bogner, A. Schwenk, Ground and excited states of doubly open-shell nuclei from ab initio valence-space hamiltonians, Phys. Rev. C 93 (2016) 051301. doi:10.1103/PhysRevC.93.051301. 485 URL https://link.aps.org/doi/10.1103/PhysRevC.93. 051301 [15] G. R. Jansen, M. D. Schuster, A. Signoracci, G. Hagen, P. Navrátil, Open sd-shell nuclei from first principles, Phys. Rev. C 94 (2016) 011301. doi:10.1103/PhysRevC.94.011301. URL https://link.aps.org/doi/10.1103/PhysRevC.94. 011301 [16] B. A. Brown, W. A. Richter, New “usd” hamiltonians for the sd shell, Phys. Rev. C 74 (2006) 034315. doi:10.1103/PhysRevC.74.034315. URL https://link.aps.org/doi/10.1103/PhysRevC.74. 034315 [17] B. A. Brown, W. D. M. Rae, E. McDonald, M. Horoi, Computer code NUSHELLX, Michigan State University, East Lansing. [18] H. Behrens, W. Buhring, Electron Radial Wave Functions and Nuclear Beta-decay, Clarendon Press, Oxford, 1982. [19] M. Haaranen, J. Kotila, J. Suhonen, Spectrum-shape method and the next-to-leading-order terms of the βdecay shape factor, Phys. Rev. C 95 (2017) 024327. doi:10.1103/PhysRevC.95.024327. URL https://link.aps.org/doi/10.1103/PhysRevC.95. 024327 [20] N. Stone, Table of nuclear magnetic dipole and electric quadrupole moments, Atomic Data and Nuclear Data Tables 90 (1) (2005) 75 – 176. doi:https://doi.org/10.1016/j.adt. 2005.04.001. URL http://www.sciencedirect.com/science/article/pii/ S0092640X05000239 [21] Data extracted from the ENSDF database using the NNDC World Wide Web site. [22] M. D. Rydt, M. Depuydt, G. Neyens, Evaluation of the groundstate quadrupole moments of the (sd) nuclei, Atomic Data and Nuclear Data Tables 99 (4) (2013) 391 – 415. doi:https: //doi.org/10.1016/j.adt.2011.12.005. URL http://www.sciencedirect.com/science/article/pii/ S0092640X12000204 [23] N. Stone, Table of nuclear electric quadrupole moments, Atomic Data and Nuclear Data Tables 111-112 (2016) 1 – 28. doi: https://doi.org/10.1016/j.adt.2015.12.002. URL http://www.sciencedirect.com/science/article/pii/ S0092640X16000024 [24] B. H. Wildenthal, M. S. Curtin, B. A. Brown, Predicted features of the beta decay of neutron-rich sd-shell nuclei, Phys. Rev. C 28 (1983) 1343–1366. doi:10.1103/PhysRevC.28.1343. URL https://link.aps.org/doi/10.1103/PhysRevC.28.1343 7 [25] T. Siiskonen, M. Hjorth-Jensen, J. Suhonen, Renormalization of the weak hadronic current in the nuclear medium, Phys. Rev. C 63 (2001) 055501. doi:10.1103/PhysRevC.63.055501. URL https://link.aps.org/doi/10.1103/PhysRevC.63. 055501 [26] J. T. Suhonen, Value of the axial-vector coupling strength in and decays: A review, Frontiers in Physics 5 (2017) 55. doi:10.3389/fphy.2017.00055. URL https://www.frontiersin.org/article/10.3389/fphy. 2017.00055 [27] B. Singh, J. Rodriguez, S. Wong, J. Tuli, Review of logf t values in β decay, Nuclear Data Sheets 84 (3) (1998) 487 – 563. doi:https://doi.org/10.1006/ndsh.1998.0015. URL http://www.sciencedirect.com/science/article/pii/ S0090375298900151 [28] E. K. Warburton, Second-forbidden unique β decays of 10 Be, 22 Na, and 26 Al, Phys. Rev. C 45 (1992) 463–466. doi:10.1103/ PhysRevC.45.463. URL https://link.aps.org/doi/10.1103/PhysRevC.45.463 [29] G. Martínez-Pinedo, P. Vogel, Shell model calculation of the β − and β + partial half-lives of 54 mn and other unique second forbidden β decays, Phys. Rev. Lett. 81 (1998) 281–284. doi:10.1103/PhysRevLett.81.281. URL https://link.aps.org/doi/10.1103/PhysRevLett.81. 281 [30] A. Idini, A. Brown, K. Langanke, G. Martinez-Pinedo, Electron Capture Processes in Intermediate mass Stars, Proceedings of Science NIC XIII (2014) 002.