WorksPhysics and Astronomy
First measurement of 87Rb(α, xn) cross sections at weak r-process energies in supernova ν-driven ejecta to investigate elemental abundances in low-metallicity stars
Abstract
Observed abundances of Z ~ 40 elements in metal-poor stars vary from star to star, indicating that the rapid and slow neutron capture processes may not contribute alone to the synthesis of elements beyond iron. The weak r-process was proposed to produce Z ~ 40 elements in a subset of old stars. Thought to occur in the ν-driven ejecta of a core-collapse supernova, (α, xn) reactions would drive the nuclear flow toward heavier masses at T = 2-5 GK. However, current comparisons between modelled and observed yields do not bring satisfactory insights into the stellar environment, mainly due to the uncertainties of the nuclear physics inputs where the dispersion in a given reaction rate often exceeds one order of magnitude. Involved rates are calculated with the statistical model where the choice of an α-optical-model potential (αOMP) leads to such a poor precision. The first experiment on 87Rb(α, xn) reactions at weak r-process energies is reported here. Total inclusive cross sections were assessed at Ec.m. = 8.1 - 13 MeV (3.7 - 7.6 GK) with the active target MUlti-Sampling Ionization Chamber (MUSIC). With a N = 50 seed nucleus, the measured values agree with statistical model estimates using the αOMP Atomki-V2. A re-evaluated reaction rate was incorporated into new nucleosynthesis calculations, focusing on ν-driven ejecta conditions known to be sensitive to this specific rate. These conditions were found to fail to reproduce the lighter-heavy element abundances in metal-poor stars.
Introduction
The oldest, metal-poor, stars in the Milky Way and in close dwarf galaxies have been investigated over the past decade to bring forth the presence of elements heavier than iron. At these sites, the observed chemical abundances ([15, 51] (Côté et al. 2019)) hint that the rapid neutron capture process ([56]), expected to produce half of nuclei beyond Fe, should take place in early galactic ages. An active site of this nucleosynthesis has been recently found with the
Tabulated Rb(,xn) reaction rates
observation of -process elements in the kilonova following a binary neutron star merger (NSM) ([24, 58, 63]). However, it occurs in galactic evolution too rarely and too late to explain the observed chemical abundances in old stars ([26] (Côté et al. 2019)). Other explosive stellar environments like magnetorotationally-driven supernovae ([66, 40, 53, 52]), collapsars ([57]), neutrino()-driven winds in core-collapse supernovae (CCSNe) ([18, 20]) are being investigated.
An enhancement of the elements around the first -process peak, with –, was also observed in a subset of metal-poor stars ([59, 32, 55]). These abundance patterns, e.g. Figure 7 ([55]) and Table 2 ([44, 45]), call for an additional mechanism that must occur in early galactic ages. Two have been put forward: the weak -process ([38, 46, 22, 2, 8, 10, 44] (Bliss et al. 2020)) also known as the -process ([35]), and the neutrino-proton process (-process) ([16, 39]). Both are expected to take place in -driven ejecta of stellar explosions like CCSNe or NSMs, depending on the neutron-richness of the ejected material. We focus here on the weak -process in -driven winds of core-collapse supernovae.
Figure 7. Comparison between nucleosynthesis models and elemental abundance ratios of metal-poor stars from the compilation of Psaltis et al. [44]. The four astrophysical conditions are presented in Table 4 (Appendix). The error bars correspond to the uncertainty for the ratios based on the prior reaction rate. The calculations with the new experimentally constrained rate are presented as the red points. The new uncertainty based on the present reaction rate is smaller than the size of the points. The selected ejecta conditions fail to reproduce abundance observations in several metal-poor stars.
| (MeV) | Range | Gamow temperature (GK) | (mb) | Uncertainties (%) statistical | Uncertainties (%) systematic |
| [13.27, 12.73] | 7.6 | 420(17) | 1.01 | 3.95 | |
| [12.73, 12.19] | 7.2 | 355(19) | 1.13 | 5.37 | |
| [12.19, 11.64] | 6.7 | 291(11) | 1.28 | 3.71 | |
| [11.64, 11.09] | 6.2 | 192(8) | 1.62 | 4.32 | |
| [11.09, 10.54] | 5.8 | 109(6) | 2.19 | 5.40 | |
| [10.54, 9.98] | 5.4 | 67.4(38) | 2.81 | 4.87 | |
| [9.98, 9.41] | 5.0 | 32.9(20) | 4.08 | 4.62 | |
| [9.41, 8.84] | 4.5 | 10.7(10) | 7.24 | 6.18 | |
| [8.84, 8.27] | 4.1 | 3.77(66) | 12.3 | 12.4 | |
| [8.27, 7.70] | 3.7 | 0.92(31) | 25.0 | 23.6 |
Table 2. Reported are the cross sections () of the reactions measured for a given effective center-of-mass energy (). The uncertainty of includes the contributions of measured energy losses and incident beam energy. The energy range mentioned covers the strip. Statistical and systematic uncertainties of are given.
In the aftermath of the collapse, temperature decreases while the neutron-rich matter is expanding away from the compact neutron star and, at some point, the nuclear statistical equilibrium breaks down. Regardless of the thermodynamics conditions (expansion time scale, entropy, electron fraction) in these extreme winds, parametric 1-dimension modelling studies ([8, 10] (Bliss et al. 2020)) show that (i) radiative -capture reactions are balanced by their reciprocal photo-disintegration reactions, (ii) decays occur more slowly than the expansion time scale of tens of ms, and (iii) and reactions are faster to fall out of equilibrium than other -induced and -induced reactions at temperatures of – GK. Hence, the nucleosynthesis pathway should stay relatively close to stability, and the nuclear reaction flow toward the elements from Fe to Mo should be driven by reactions where are the typical cases. Several sensitivity studies ([10, 44, 45] (Bliss et al. 2020)) have shown that model-to-observations comparisons of abundances around Sr in metal-poor stars are currently inadequate to firmly constrain the thermodynamics conditions of the -driven ejecta.
This is mainly due to the variations resulting from the uncertainties of reaction rates which have been poorly measured so far. Without experimental information, reaction rates are estimated within the Hauser-Feshbach (HF) framework ([19, 50]). This model is justified for the mass region of interest and the involved stellar temperatures ( GK) which correspond to the energy region of high nuclear level density in the compound nucleus. Since the decay of the latter is independent of its formation mechanism, the probability that the reaction occurs can be expressed as , i.e. the product of the transmission coefficient of the particle into the seed nucleus () and the transmission coefficient of the -exit channel () that is normalized to all exit channels (). The open exit channels are rays and multiple particles (, , , , ...). However, at weak -process energies located far above the emission threshold, -exit channels dominate and, so, the term reduces to 1. Hence, only the coefficient is relevant for statistical model estimates of cross sections.
The coefficient is derived from an -Optical-Model Potential (OMP), see e.g. [36]. Several OMPs are available, to quote the standard ones ([34, 41, 13, 27]), with the folding approach Watanabe (1958), ([5, 36]). Deviations between reaction rates of one to two orders of magnitude have been observed while testing different OMPs. This is illustrated in Figure 1 where several reaction rates, normalized to the referenced values of ReaclibV2.2 Cyburty et al. (2010), are presented along temperatures relevant for the weak -process. The reaction rates were calculated with the code Talys ([28, 29]) where only the chosen OMP was varied. Note that the ratio resulting from the OMP ([34]) (solid cyan curve Figure 1) deviates from unity even though the same OMP was used for the reaction rate calculations in ReaclibV2.2 (Cyburty et al. (2010)), but with the statistical model code NON-SMOKER ([48]) and technical differences exist between the two codes.
Figure 1. Ratio between the reaction rate calculated with the Talys code Koning & Rochman [28]; ); Koning et al. [29] and the rate from ReaclibV2.2 Cyburt et al. (2010) along temperatures where reactions impact the nucleosynthesis occurring in CCSNe -driven winds. Standard OMPs were considered to calculate the reaction rate, i.e. those from McFadden & Satchler [34] shown in solid cyan, from Nolte et al. [41] shown in solid blue, from Demetriou et al. [13] with the dispersive mode shown in solid green, from nucleon potential Koning & Delaroche [27] with the folding approach Watanabe [62] shown in dotted green, from Avrigeanu et al. [5] shown in dotted blue, and from Mohr et al. [36] shown in solid black. The maximal deviations range from a factor 10 at GK to a factor 100 at GK.
Measured data allow us to test the statistical model predictions based on available OMPs for cross sections with the goal to improve the precision of nucleosynthesis calculations for the weak -process. In this respect, statistical model estimates based on the latest OMP Atomki-v2 ([36]) appeared to be consistent with some recent measurements performed on stable nuclei located in the weak -process mass region. These are ([47, 43, 60, 25, 42]) for which the deviations between experimental and calculated values of cross sections are within a factor . However, a recent study on Sr at (Fougères et al. [14]) measured a cross section that is systematically lower ( %) than statistical model estimates with the OMP Atomki-V2. This, together with the general lack of data available on the neutron-rich side to test statistical model predictions, push for more nuclear physics measurements.
The comprehensive sensitivity studies of Bliss et al. (2020); ); Psaltis et al. [44] have identified the key reactions that strongly contribute to the uncertainties of the yields derived from the modelling of CCSNe -driven winds. These works have determined which elemental abundances (and by how much) are affected by a given reaction rate under certain thermodynamics conditions in the -driven ejecta, and thus suggest a selection of experiments that would best help constrain model-to-observations comparisons for abundances in old stars. Such experimental efforts on measuring weak -process cross sections have already started with stable beams Szegedi et al. [60]; ); Kiss et al. [25]; ); Ong et al. [42]; ); Fougères et al. [14]. Many more, but with unstable neutron-rich beams or targets, have yet to be assessed. Heavy-ion accelerators like ATLAS (US) or FRIB (US) are now enabling such experimental programs at weak -process energies ( MeV/u) thanks to the high intensities available for beams of the desired neutron-rich isotopes.
Among the stable cases of astrophysical relevance that remain to be measured at astrophysical energies, the reaction was found to impact , , , abundances by factors of (and Sr/Zr, Y/Zr, Nb/Zr) elemental ratios in a handful of CCSNe -driven winds conditions ([44] (Bliss et al. 2020)). Located at the closed shell , this should further test statistical model predictions with the up-to-date OMP Atomki-V2 Mohr et al. [36]. To our knowledge, cross sections were only measured above the weak -process energy region ( MeV, GK) with uncertainties of % Riley & Linder [54]. The reaction open at GK has not been investigated at all. An inclusive measurement of the two reactions would directly constrain the OMP since the summed contribution of the and channels depends solely on the OMP.
Responding to the need of data at weak -process energies and of constraints on the OMP, this work presents the first measurement of the total cross sections of both reactions at MeV ( GK). The inclusive excitation function was measured with the active gaseous technique. The measured cross sections were then compared to statistical model estimates of the reaction rate determined in the temperature range relevant for the weak -process in CCSNe -driven winds. The impact of this newly constrained reaction rate was assessed on elemental abundances by nucleosynthesis calculations in CCSNe -driven ejecta presenting thermodynamics conditions known to be sensitive to this specific rate. The resulting predictions were finally compared to the abundances observed in a set of old stars.
Nuclear experiment
Earlier experimental investigations on reactions have tested the standard OMPs in the mass region relevant for the weak -process ([60, 42]), including at closed-shell nuclei ([1, 14]), and reported constraints on modelled elemental abundances in CCSNe -driven winds ([25]). The stellar reaction rates are determined for charged-particle induced reactions at energies MeV above the neutron emission threshold, and, so, most attempts aim at directly assessing high cross sections (0.1–100 mb). To quote recent works, this can be achieved through (i) integrated measurements with the active target technique in inverse kinematics [42, 14], or (ii) -delayed--ray measurements with the activation technique in normal kinematics [43, 60, 25], or (iii) recoil--ray prompt coincidence measurements at a mass separator using direct reactions in inverse kinematics [1]. The present work on Rb() reactions was carried out using the experimental technique (i).
Method
The active target technique relies on the detection medium being also the target material. Cross sections are straightforwardly measured while detecting the nuclei of both the entrance and exit channels of the reaction. This technique presents many advantages: an increased target thickness, a detection efficiency of %, a self-normalization of the cross section. In the case of -induced reactions, the technique can be employed in inverse kinematics with active helium-gaseous targets being ionization chambers [23, 7] (Avila et al. 2017), time-projection chambers [33, 31, 6], or arrays that combine gaseous and semi-conductor detectors [30]. The excitation function is probed at different center-of-mass energies () while the incident mono-energetic beam is slowing down in the gaseous volume. The reachable range and resolution in are governed by incident beam energy, gas pressure, and detector segmentation. Most high-profile seed nuclei in explosive stellar nucleosynthesis are radioactive, short-lived, elements. The inverse kinematics method used here allows us to investigate them thanks to the delivered radioactive beams.
Depending on the detector sensitivity, the experiment focuses on selecting the heavy recoils or light ejectiles as well as following energy losses or reconstructing reaction vertices and kinematics in order to identify the reaction channel of interest. Since energy loss of ions in matter varies as the square of the atomic number, () reactions on nuclei can be observed via the first method: the measurement of heavy-ion energy losses as they pass through the active gaseous volume. Due to the change between the beam and the heavy recoil, a sharp local increase in energy loss is a signature of a reaction occurring. With electrically-segmented (stripped) detectors in a single direction [11, 23, 7], the search for () events may follow:
the PID method, a global analysis that looks at energy losses summed over several strips , similar to the Particle IDentification (PID) with silicon detectors and their variants,
the Traces method, a local analysis that tracks energy loss per strip.
It is sufficient to analyze energy losses relative to the averaged value found for the beam in order to identify reaction events. However, the absolute energy losses of the beam particles must be known to determine the center-of-mass energies associated with the reaction of interest. The precision in the determination of the beam energy losses directly impacts the energy resolution of the measured excitation function. Energy losses of heavy ions in matter can be estimated with Monte Carlo calculations that integrate tabulated stopping powers. However, large deviations ( %) have been observed at low energies between tables from standard libraries [21, 67] (Wieck et al. 2018). Hence, performing also a measurement of the beam energy losses in the gaseous detector is essential in active-target experiments that aim at assessing cross sections.
The technique described here was chosen to determine the Rb() reaction cross sections. Guided by past successful experiments [3, 42, 14] (Avila et al. 2017), the measurement was carried out in inverse kinematics at ATLAS accelerator with the electrically-segmented MUlti Sampling Ionization Chamber (MUSIC) [11].
Set-up
The Rb stable beam delivered by the Argonne Tandem Linac Accelerator System ATLAS had a charge state of , an energy of 4.61(1) MeV/u, and an average intensity of pps. The beam energy was measured from the time-of-flight value for three resonator pairs located upstream of the MUSIC detector. The beam purity was assessed from the energy losses in the entrance of the detector. A single contaminant was observed with an intensity of about 1/7 that of the requested beam. It was considering the magnetic rigidity matching and the measured energy losses. The MUSIC detector was filled with pure He gas at 555 Torr held by Ti foils of 1.30(5) mg/cm thickness at the entrance and exit sides for the beam. The beam energy loss after passing through the entrance foil was measured to be 46.9(9) MeV. The anode of MUSIC is segmented into 18 strips of equal width (1.578 cm) along the beam axis, see Figure 1 of Carnelli et al. (2015). In the data-acquisition system of digital nature (DAQ), each channel was self-triggering, and the beam rate was monitored to ensure the DAQ stability. Less than 10 % pile-up events were observed as expected for such a gaseous detector operated at rates of tens of kHz. In the data analysis, a cut was applied on the energy losses in the first two strips to fully separate the beam events from contaminant events.
Energy losses of in the Ti foils and gas were measured with a depleted silicon surface-barrier detector that was mounted downstream of MUSIC. Note that to correct for the pulse-height defects related to heavy ions in Si detectors ([65]), the energy calibration of the Si detector was performed in-beam by measuring the beam at several energies (1.0 − 4.6 MeV/u) in the Si detector. Regarding the energy loss measurement, the beam energy was measured after the MUSIC detector: first, only with the Ti foils but no gas, second the detector was gradually filled with gas where the beam energy at a given pressure was recorded. The measured energy losses were reproduced, i.e within 2.0 %, by Monte-Carlo simulations using the mean value of the stopping powers of the ATIMA table in LISE++ ([61]) and of the tables in SRIM ([67]).
Overall, the beam energy losses per strip were of 11−13 MeV with a resolution of 8 % (FWHM) due to the strip spatial extension.
Several reaction channels were energetically allowed in the active region: the Rutherford and inelastic scatterings referred to as , the channel with MeV, the with MeV, and the channel with MeV. Additionally, the channel with MeV was opened at the highest energies ( MeV). The dominant mechanism is with calculated cross sections in excess of 1 barn (estimates from LISE++). Monte-Carlo simulations of the set-up indicated that and events would overlap with events. But statistical model cross sections from the Talys code revealed that, at present energies, reactions should be favoured, by several orders of magnitude , over and reactions. Possible contamination from the latter was thereby negligible compared to the statistical uncertainty (1 − 25 %) of the measured cross sections. Simulated energy losses of recoils differ by less than a hundred keV: this is far below the detector energy resolution. Hence, the exit channel could not be isolated from the exit channel. The measured inclusive cross sections were compared with statistical model cross sections. This allowed us to extract their respective contributions at MeV.
The PID method was first employed to identify and scattering events which occurred in each strip of the MUSIC detector. This is illustrated for the reactions taking place in the strip 3 on the left of Figure 2. As detailed previously, two conditions were required: energy losses measured in the first two strips corresponded to , and energy losses measured in the strip 3 were higher than the ones in the strip 2. Two isolated regions are noticed but the one at higher energies is unambiguously associated to events considering of the respective recoils. Then, the Traces method was used to count events. Such traces of reactions occurring in the strip 3 are shown with the solid red curves on the right of Figure 2. A third condition was used here with a cut applied on the region (Figure 2 left). Several traces of unreacted beam events (black curves) and of beam-like scattering events at low angles (dotted red curves) are also presented on the right of Figure 2. Similarly to Ong et al. [42], scattering reactions at large angles were observed with a higher variation of than reactions.
Figure 2. Application of the PID method (left) and Traces method (right) to identify and quantify reactions occurring in the strip 3 of the MUSIC detector. Left panel: sums of energy losses are shown for events selected on an incoming beam and a sharp increase in energy loss measured in the strip 3. The energy losses per strip were normalized to the averaged value of 12 MeV found for the beam. The isolated region at corresponds to events, and scattering events are noticed below. Right panel: individual traces of calibrated energy losses along the strip number are shown for the unreacted beam (black curves), the (solid red curves) and few at low angles (dotted red curves) reactions. The events were obtained with a cut applied on the associated region in the PID plot (left).
Results
The total cross sections of the reactions are presented along the measured effective center-of-mass energies () of 8.09 – 13.01 MeV in Figure 3 and detailed in Table 2 (Appendix). To determine , the center-of-mass energies deduced from the measurement of the beam energy losses in the MUSIC detector were corrected for the thick-target yield (Szegedi et al. [60]). Shown with the red points, horizontal (energy) uncertainties of the measured cross sections correspond to the spatial extension of the strip and vertical uncertainties include both statistical and systematic contributions. Note that at low energies, the systematic contribution strongly increased because the and channels became more entangled in both the PID method and Traces method. Due to the too small difference in between Rb and Y relative to the total atomic number, the above effect puts a low-energy limit on the present set-up to assess weak r-process cross sections.
Figure 3. Comparison of cross sections as obtained from the present measurement and from the statistical model estimates. The common abscissa is the effective center-of-mass energy (). Upper panel: experimental data are shown as red points. The calculated cross sections including the OMP Atomki-V2 Mohr et al. [36] (black curve) are dominated by the channel (dotted black curve) at low energy MeV. The channel (dotted blue curve) starts to contribute () at MeV, and dominates at high energy MeV. The calculations scaled to experimental data are shown (dotted red curve). This corresponds to the best fit () of the weighted sum of the and calculations to measurements: the deviations are 0.89(9) and 0.87(14), respectively. Statistical model estimates, only 10 % higher than measurements, well reproduce the cross sections at – 13 MeV. The scaled calculations allow to access cross sections at lower astrophysical energies (T3.7 GK). Uncertainties contributions of the measured cross sections are detailed in the text and Table 2 (Appendix). The coloured band corresponds to the uncertainty of the fit for the scaled calculations. Past work on by Riley & Linder [54] is presented (green points). Lower panel: deviations between scaled calculations and measured data scatter around 0 10 %, and are not energy dependent.
The measured cross sections are compared to statistical model cross sections in Figure 3 (upper panel). Calculated and cross sections were obtained with the Talys code (Koning et al. [29]; Koning & Rochman [28]) using the OMP Atomki-V2 (Mohr et al. [36]). The results, shown as the dotted black and blue curves, respectively, were summed up to obtain the black curve representing the cross sections (black curve). A fit was then performed between the measurement and the weighted sum of the and calculations, and the best result (dotted red curve) was obtained for a relative deviation of 0.89(9) and 0.87(14), respectively. The observed deviations were found consistent between the two exit channels, i.e. they are equal within uncertainties. This was expected since both cross sections are governed by the same OMP. It should be noted that using a scaling factor per exit channel or a common factor of 0.88 resulted in negligible changes in the associated reaction rate. This new experimental result supports the use of OMP Atomki-V2 to reliably predict cross sections at astrophysical energies in weak r-process. The results of the reaction measured at high energy by Riley & Linder [54] are also shown in Figure 3 (green points). Agreeing with the present work at MeV where the exit channel is suppressed, it confirms the robustness of statistical model estimates based on OMP Atomki-V2.
Once scaled, the calculated cross sections were found to differ only by % in comparison to experimental data, see Figure 3 (lower panel). No systematic trend was noticed along . The energy dependence of the experimental data was found to be properly reproduced by statistical model estimates. Therefore, information at energies lower than measured values could be assessed by using the derived constant scaling factor of 0.89(9) to the lowest energies for the reaction of interest.
Looking at the reduced cross section () and energy () (Gomes et al. [17]; Mohr et al. [37]) allows to directly compare total cross sections of charged-particle reactions whatever the beam, target and energy are. These two parameters are given by and . A common trend was observed for -induced reactions around in Mohr et al. [37]. This is illustrated in Figure 4 where the result of the reaction fits well to this trend, similarly to several nuclei with . If confirmed to be the case for even more nuclei, this trend may also be of use to infer weak -process cross sections not yet measured.
Figure 4. Evolution of the reduced cross section () as a function of the reduced energy () for –144 nuclei produced in -induced reactions at low energy from (Mohr et al. [37]; Oprea et al. [43]; Kiss et al. [25]; Rapp et al. [47]; Ong et al. [42]; Fougères et al. [14]). Present results on (blue circles) follow the apparent trend.
Astrophysical implications
Thermonuclear reaction rates
The cross sections were measured at the energies corresponding to the Gamow temperatures of 3.7–7.6 GK. Calculations of the associated thermonuclear reaction rate were performed with the code EXP2RATE Rauscher (2025). They included the cross sections from statistical model estimates based on the Atomki-V2 OMP Mohr et al. [36] scaled by the constant factor of 0.89(9) deduced from the experiment. Recommended values of the reaction rate at temperatures where the weak -process impacts the nucleosynthesis in -driven ejecta of CCSNe are reported in Table 1. The lower and higher lim-
| (GK) | Low | Recommended | High |
| 1.6 | |||
| 2.0 | |||
| 2.5 | |||
| 3.0 | |||
| 3.5 | |||
| 4.0 | |||
| 4.5 | |||
| 5.0 | |||
| 5.5 | 1.95 | 2.17 | 2.39 |
| 6.0 | 7.45 | 8.39 | 9.44 |
Table 1. Low, recommended, and high thermonuclear rates of the reaction, in units of , for temperatures relevant for the weak -process.
EXP2RATE Rauscher (2025). They included the cross sections from statistical model estimates based on the Atomki-V2 OMP Mohr et al. [36] scaled by the constant factor of 0.89(9) deduced from the experiment. Recommended values of the reaction rate at temperatures where the weak -process impacts the nucleosynthesis in -driven ejecta of CCSNe are reported in Table 1. The lower and higher lim-
its of the reaction rate were derived from uncertainties of the scaling factor that range from at measured energies ( MeV) to a factor 2 at the threshold energy of the reaction ( MeV). The reaction rate was found to be exceeded by the reaction rate at GK. In similar calculations to assess the rate of the latter reaction, the statistical model cross sections were weighted by the constant scaling factor found for the exit channel (0.87(14)). Note that the tabulated rates of the two reactions for each value of the default temperature grid of the Talys code are given in Table 3 (Appendix).
| (GK) | Recommended | Low | High | Recommended | Low | High |
| 0.10 | ||||||
| 0.15 | ||||||
| 0.20 | ||||||
| 0.25 | ||||||
| 0.3 | ||||||
| 0.35 | ||||||
| 0.40 | ||||||
| 0.45 | ||||||
| 0.50 | ||||||
| 0.60 | ||||||
| 0.70 | ||||||
| 0.80 | ||||||
| 0.90 | ||||||
| 1.0 | ||||||
| 1.2 | ||||||
| 1.4 | ||||||
| 1.6 | ||||||
| 1.8 | ||||||
| 2.0 | ||||||
| 2.25 | ||||||
| 2.5 | ||||||
| 2.75 | ||||||
| 3.0 | ||||||
| 3.25 | ||||||
| 3.5 | ||||||
| 3.75 | ||||||
| 4.0 | ||||||
| 4.25 | ||||||
| 4.5 | ||||||
| 4.75 | ||||||
| 5.0 | ||||||
| 5.5 | ||||||
| 6.0 | ||||||
| 6.5 | ||||||
| 7.0 | ||||||
| 7.5 | ||||||
| 8.0 | ||||||
| 8.5 | ||||||
| 9.0 | ||||||
| 9.5 | ||||||
| 10 |
Table 3. Recommended, low, and high thermonuclear rates of the and reactions, in units of , along the default temperature grid of the Talys code.
The evolution of the recommended rate of the reaction rate as a function of temperature is shown in Figure 5 (violet curve). The status prior to the measurement, shown as the cyan band, was obtained from calculations which included standard OMPs (see Figure 1), low (high) limits corresponding to Demetriou et al. [13] (Nolte et al. [41]). The precision of the reaction rate is drastically increased at –5 GK, i.e. current uncertainties based on experimental data are –15% whereas prior uncertainties based on statistical model calculations are a factor of 100–10. The referenced rate from ReaclibV2.2 (Cyburt et al. 2010) is also given in Figure 5 (black curve). At weak -process temperatures, the two rates agree within a factor of 2. This is in line with the initial development where the ReaclibV2.2 rate was observed to be close to statistical model estimates based on the Atomki-V2 OMP (see Figure 1).
Figure 5. Evolution of the thermonuclear reaction rate of as a function of temperature relevant for the weak -process. The recommended reaction rate (violet curve) was determined using the code EXP2RATE based on cross sections from Talys+Atomki-V2 scaled by the constant factor of 0.89. Resulting uncertainties are % at GK. The prior rate, shown as the cyan band, was evaluated from Talys calculations based on standard OMPs (Figure 1, uncertainties of 100 - 10). The reaction rate from ReaclibV2.2 (black curve) agrees with the rate here reevaluated (within a factor of 2).
Elemental abundances
The impact of the newly-constrained reaction rate on the weak -process abundances around was studied via extensive nucleosynthesis calculations. We used the four thermodynamical conditions MC12, MC13, MC15 and MC16 from Bliss et al. (2020) (see Table 4), which have shown sensitivity to the reaction rate in the works of Bliss et al. (2020); Psaltis et al. [44], with the same reaction network setup as in Psaltis et al. [44]. With the exception of MC13, the electron fraction is high and two trajectories (MC15, MC16) have a high entropy ().
| CCSNe -wind trajectory | ||||||
| CCSNe -wind trajectory | ( nucleon) | (ms) | ||||
| 12 | 0.48 | 85 | 9.7 | |||
| 13 | 0.43 | 64 | 35.9 | |||
| 15 | 0.48 | 103 | 20.4 | |||
| 16 | 0.49 | 126 | 15.4 |
Table 4. Astrophysical conditions of CCSNe -driven winds and abundances at impacted by the reaction rate.
Figure 6 shows the nucleosynthesis results for each of the aforementioned weak -process thermodynamical conditions. Each panel displays the final isotopic pattern after 1 Gy using the ReaclibV2.2 rate for the reaction as a baseline case (black line). Additionally, we calculated the range of abundances using the prior reaction rate based on standard OMPs (cyan shaded region). Furthermore, we repeated the calculations using the new experimental reaction rate (Table 3) together with its constrained uncertainty (magenta shaded region). The use of the experimental reaction rate for reduces the uncertainty in the yields of the species to a few percent. By comparison, the use of the prior rate results in uncertainties of for some conditions.
We also compared the new nucleosynthesis results with the elemental abundance ratios – Sr/Zr and Y/Zr – of metal-poor stars that exhibit an excess of Sr - Ag relative to Solar values, as reported by Psaltis et al. [44]. Despite showing some sensitivity to the uncertainty in the reaction rate, the four astrophysical conditions we investigated were unable to reproduce the observed abundance ratios, consistent with the findings of Psaltis et al. [44]. However, as illustrated in Figure 7, our current experiment has significantly reduced the uncertainty in these abundance ratios. This improvement in one of the major nuclear physics uncertainties allows us to confidently rule out these conditions of the -driven winds as potential candidates for reproducing the observed abundance ratios in these metal-poor stars.
Conclusion
The reaction has been reported to impact abundances which should be produced by the weak -process ongoing in -driven winds after core-collapse supernovae. The present work provides the first experimental insight into this reaction at astrophysical temperatures. The total cross sections of reactions were measured at GK with the active gaseous target MUSIC. Measured values were found to be highly consistent with statistical model estimates based on the OMP Atomki-V2. This led to a robust re-evaluation of the thermonuclear rates of the and reactions. The experimental inputs indeed constrained, by two-to-three orders of magnitude, the reaction rate at weak -process temperatures, i.e. the present uncertainties of 15 - 65 % are to be compared with the prior uncertainties of 10 - 100. This test of the input to the statistical model calculations for cases at further supports the potential Atomki-V2 in its predictive power of -induced reaction cross sections. In the near future, more and more constraints on weak -process reactions are expected from the ongoing experimental program with the MUSIC detector harvesting the neutron-rich beams available at the ATLAS and FRIB accelerators.
The impact of the re-evaluated rate of the reaction was assessed in detailed nucleosynthesis calculations for different -driven trajecto-

Figure 6. Isotopic abundance patterns for four different sets of -driven ejecta corresponding to the conditions MC12, MC13, MC15, and MC16 detailed in Table 4 (Appendix). The calculations use the setup of Psaltis et al. [44]. Each panel shows the abundance pattern as a function of mass number for three cases: calculations using the ReaclibV2.2 rate (black), the rate of the present work (magenta), and the previously recommended rate (cyan). The bottom panels illustrate the relative differences between each calculation and the one using the ReaclibV2.2 rate. The new constrained reaction rate reduces the uncertainty in isotopic production for . See the text for further details.
ries of various thermodynamics conditions, introduced in previous impact studies ([44] (Bliss et al. 2020)). Under the investigated conditions, the experimental reaction rate reduces the production uncertainty of species to . As expected, from previous studies, our comparison with observed abundance ratios did not yield any matches. However, it has helped eliminate these -driven conditions as the candidates, narrowing down the possibilities for reproducing the peculiar abundance patterns in the Sr-Ag region observed in metal-poor stars in the Galaxy.
Acknowledgments
The authors thank the support of the ATLAS beam and detection physicists. This material is based upon work supported by the U.S. Department of Energy, Office of Science, Office of Nuclear Physics, under Contract No. DE-AC02-06CH11357. This research used resources of Argonne National Laboratory’s ATLAS facility, which is a DOE Office of Science User Facility. A.P. acknowledges support from the U.S. Department of Energy, Office of Science, Office of Nuclear Physics, under Award Number DE-SC0017799 and Contract Nos. DE-FG02-97ER41033 and DE-FG02-97ER41042. P.M. would like to acknowledge the support of NKFIH (grant K134197). S.B. acknowledges support from the Institute for Basic Science, South Korea (IBS), under the Grants No. IBS-R031-D1.
Cross-section data
Tabulated reaction rates
Astrophysical conditions used in nuclear network calculations
Table 4. Astrophysical conditions of CCSNe -driven winds and abundances at impacted by the reaction rate, as taken from Bliss et al. (2020): the electron mass fraction , the entropy per baryon , and the expansion timescale . The uncertainty in the elemental ratios is based only on the variation of the rate within its new experimental uncertainties.
References
- [1]Angus, C., Williams, M., Andreyev, A., et al. 2023, EPJ Web Conf., 279, 11003, doi: 10.1051/epjconf/202327911003DOI
- [2]Arcones, A., & Montes, F. 2011, ApJ, 731, 5, doi: 10.1088/0004-637X/731/1/5DOI
- [3]Avila, M. L., Rehm, K. E., Almaraz-Calderon, S., et al. 2016, Phys. Rev. C, 94, 065804, doi: 10.1103/PhysRevC.94.065804DOI
- [5]Avrigeanu, V., Avrigeanu, M., & Mănăilescu, C. 2014, Phys. Rev. C, 90, 044612, doi: https://doi.org/10.1103/PhysRevC.90.044612
- [6]Ayyad, Y., Abgrall, N., Ahn, T., et al. 2020, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 954, 161341, doi: https://doi.org/10.1016/j.nima.2018.10.019
- [7]Blankstein, D., Bardayan, D., Allen, J., et al. 2023, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 1047, 167777, doi: https://doi.org/10.1016/j.nima.2022.167777
- [8]Bliss, J., Arcones, A., Montes, F., & Pereira, J. 2017, J. Phys. G: Nucl. Part. Phys., 44, 054003, doi: https://doi.org/10.1088/1361-6471/aa63bd
- [10]Bliss, J., Witt, M., Arcones, A., Montes, F., & Pereira, J. 2018, ApJ, 855, 135, doi: https://link.aps.org/doi/10.3847/1538-4357/aaadbeDOI
- [11]Carnelli, P., et al. 2015, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 799, 197
- [13]Demetriou, P., Grama, C., & Goriely, S. 2002, Nuclear Physics A, 707, 253, doi: https://doi.org/10.1016/S0375-9474(02)00756-X
- [14]Fougères, C., Avila, M. L., Jayatissa, H., et al. 2024, Phys. Rev. C, 109, 065805, doi: 10.1103/PhysRevC.109.065805
- [15]Frebel, A. 2018, Annual Review of Nuclear and Particle Science, 68, 237, doi: https://doi.org/10.1146/annurev-nucl-101917-021141
- [16]Fröhlich, C., Martínez-Pinedo, G., Liebendörfer, M., et al. 2006, Phys. Rev. Lett., 96 142502, doi: https://doi.org/10.1103/PhysRevLett.96.142502
- [17]Gomes, P. R. S., Lubian, J., Padron, I., & Anjos, R. M. 2005, Phys. Rev. C, 71, 017601, doi: https://doi.org/10.1103/PhysRevC.71.017601
- [18]Hansen, C. J., Montes, F., & Arcones, A. 2014, ApJ, 797, 123, doi: 10.1088/0004-637X/797/2/123
- [19]Hauser, W., & Feshbach, H. 1952, Phys. Rev., 87, 366, doi: 10.1103/PhysRev.87.366
- [20]Horowitz, C. J., Arcones, A., Côté, B., et al. 2019, J. Phys. G: Nucl. Part. Phys., 46, 083001, doi: https://doi.org/10.1088/1361-6471/ab0849
- [21]Hubert, F., Bimbot, R., & Gauvin, H. 1990, Atomic Data and Nuclear Data Tables, 46, 1, doi: https://doi.org/10.1016/0092-640X(90)90001-Z
- [22]Izutani, N., Umeda, H., & Tominaga, N. 2009, ApJ, 692, 1517, doi: 10.1088/0004-637X/692/2/1517
- [23]Johnstone, J., Kumar, R., Hudan, S., et al. 2021, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 1014, 165697, doi: https://doi.org/10.1016/j.nima.2021.165697
- [24]Kasen, D., Metzger, B., Barnes, J., Quataert, E., & Ramirez-Ruiz, E. 2017, Nature, 551, 80, doi: 10.1038/nature24453
- [25]Kiss, G. G., Szegedi, T. N., Mohr, P., et al. 2021, ApJ, 908, 202, doi: 10.3847/1538-4357/abd2bc
- [26]Kobayashi, C., Mandel, I., Belczynski, K., et al. 2023, The Astrophysical Journal Letters, 943, L12, doi: 10.3847/2041-8213/acad82
- [27]Koning, A., & Delaroche, J. 2003, Nuclear Physics A, 713, 231, doi: https://doi.org/10.1016/S0375-9474(02)01321-0
- [28]Koning, A., & Rochman, D. 2012, Nuclear Data Sheets, 113, 2841, doi: https://doi.org/10.1016/j.nds.2012.11.002
- [29]Koning, A., Rochman, D., Sublet, J.-C., et al. 2019, Nuclear Data Sheets, 155, 1, doi: https://doi.org/10.1016/j.nds.2019.01.002
- [30]Koshchiy, E., Blackmon, J., Rogachev, G., et al. 2017, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 870, 1, doi: https://doi.org/10.1016/j.nima.2017.07.030
- [31]Koshchiy, E., Rogachev, G., Pollacco, E., et al. 2020, Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 957, 163398, doi: https://doi.org/10.1016/j.nima.2020.163398
- [32]Mashonkina, L., Christlieb, N., Barklem, P. S., et al. 2010, Astronomy and Astrophysics, 516, A46, doi: 10.1051/0004-6361/200913825
- [33]Mauss, B., et al. 2019, Nucl. Instrum. Methods Phys. Res. Sect. A: Accel. Spectrom. Detect. Associated Equip., 940, 498
- [34]McFadden, L., & Satchler, G. 1966, Nuclear Physics, 84, 177, doi: https://doi.org/10.1016/0029-5582(66)90441-X
- [35]Meyer, B. S., Mathews, G. J., Howard, W. M., Woosley, S. E., & Hoffman, R. D. 1992, ApJ, 399, 656, doi: 10.1086/171957
- [36]Mohr, P., Fülöp, Z., Gyürky, G., et al. 2021, Atomic Data and Nuclear Data Tables, 142, 101453, doi: https://doi.org/10.1016/j.adt.2021.101453
- [37]Mohr, P., Kiss, G., Fülöp, Z., et al. 2013, Atomic Data and Nuclear Data Tables, 99, 651, doi: https://doi.org/10.1016/j.adt.2012.10.003
- [38]Montes, F., Beers, T. C., Cowan, J., et al. 2007, ApJ, 671, 1685, doi: 10.1086/523084
- [39]Nishimura, N., Rauscher, T., Hirschi, R., et al. 2019, Monthly Notices of the Royal Astronomical Society, 489, 1379, doi: 10.1093/mnras/stz2104
- [40]Nishimura, N., Sawai, H., Takiwaki, T., Yamada, S., & Thielemann, F.-K. 2017, Astrophys. J. Lett., 836, L21, doi: https://doi.org/10.3847/2041-8213/aa5dee
- [41]Nolte, M., Mächner, H., & Bojowald, J. 1987, Phys. Rev. C, 36, 1312, doi: 10.1103/PhysRevC.36.1312
- [42]Ong, W.-J., Avila, M. L., Mohr, P., et al. 2022, Phys. Rev. C, 105, 055803, doi: 10.1103/PhysRevC.105.055803
- [43]Oprea, A., Glodariu, T., Filipescu, D., et al. 2017, EPJ Web Conf., 146, 01016, doi: https://doi.org/10.1051/epjconf/201714601016
- [44]Psaltis, A., Arcones, A., Montes, F., et al. 2022, ApJ, 935, 27, doi: https://doi.org/10.3847/1538-4357/ac7da7
- [45]Psaltis, A., Jacobi, M., Montes, F., et al. 2024, The Astrophysical Journal, 966, 11, doi: 10.3847/1538-4357/ad2dfb
- [46]Qian, Y.-Z., & Wasserburg, G. J. 2007, Physics Reports, 442, 237, doi: https://doi.org/10.1016/j.physrep.2007.02.006
- [47]Rapp, W., Dillmann, I., Käppeler, F., et al. 2008, Phys. Rev. C, 78, 025804, doi: 10.1103/PhysRevC.78.025804DOI
- [48]Rauscher, T. 2001, Computer code NONSMOKER https://nucastro.org/websmoker.html
- [50]Rauscher, T., & Thielemann, F.-K. 2000, Atomic Data and Nuclear Data Tables, 75, 1, doi: https://doi.org/10.1006/adnd.2000.0834
- [51]Reichert, M., Hansen, C. J., Hanke, M., et al. 2020, Astronomy and Astrophysics, 641, A127, doi: 10.1051/0004-6361/201936930
- [52]Reichert, M., Obergaulinger, M., Aloy, M. A., et al. 2022, Monthly Notices of the Royal Astronomical Society, 518, 1557, doi: 10.1093/mnras/stac3185
- [53]Reichert, M., Obergaulinger, M., Eichler, M., Aloy, M. A., & Arcones, A. 2021, Monthly Notices of the Royal Astronomical Society, 501, 5733, doi: 10.1093/mnras/stab029
- [54]Riley, C., & Linder, B. 1964, Phys. Rev., 134, B559, doi: 10.1103/PhysRev.134.B559DOI
- [55]Schwerdtfeger, P., Smits, P., & Pyykkö, O. 2020, Nat Rev Chem, 4, 359–380, doi: 10.1038/s41570-020-0195-yDOI
- [56]Siegel, D. M. 2022, Nature Reviews Physics, 4, 306–318, doi: https://doi.org/10.1038/s42254-022-00439-1
- [57]Siegel, D. M., Barnes, J., & Metzger, B. D. 2019, Nature, 569, 241, doi: 10.1038/s41586-019-1136-0DOI
- [58]Smartt, S., Chen, T.-W., Jerkstrand, A., et al. 2017, Nature, 551, 75, doi: https://doi.org/10.1038/nature24303
- [59]Sneden, C., Cowan, J. J., & Gallino, R. 2008, Annual Review of Astronomy and Astrophysics, 46, 241, doi: 10.1146/annurev.astro.46.060407.145207
- [60]Szegedi, T. N., Kiss, G. G., Mohr, P., et al. 2021, Phys. Rev. C, 104, 035804, doi: 10.1103/PhysRevC.104.035804DOI
- [61]Tarasov, O., & Bazin, D. 2008, Nucl. Instr. and Meth. in Phys. Res. B, 266, 4657, doi: https://doi.org/10.1016/j.nimb.2008.05.110
- [62]Watanabe, S. 1958, Nuclear Physics, 8, 484, doi: https://doi.org/10.1016/0029-5582(58)90180-9
- [63]Watson, D., Hansen, C., Selsing, J., et al. 2019, Nature, 574, 497, doi: https://doi.org/10.1038/s41586-019-1676-3
- [65]Wilkins, B., Fluss, M., Kaufman, S., Gross, C., & Steinberg, E. 1971, Nucl. Inst. and Meth., 92, 381, doi: https://doi.org/10.1016/0029-554X(71)90414-9
- [66]Winteler, C., Käppeli, R., Perego, A., et al. 2012, Astrophys. J. Lett., 750, L22, doi: https://doi.org/10.1088/2041-8205/750/1/L22
- [67]Ziegler, J. F., Ziegler, M. D., & Biersack, J. P. 2010, Nuclear Instruments and Methods in Physics Research B, 268, 1818, doi: 10.1016/j.nimb.2010.02.091





