WorksPhysics and Astronomy
Solar-System Abundances of -Nuclides Probe Collective Neutrino Oscillations in Supernovae
Abstract
Direct evidence for collective neutrino oscillations in core-collapse supernovae remains elusive. We show that this quantum phenomenon leaves a footprint on the abundance pattern of proton-rich nuclides in the solar system. Modeling the -process using a progenitor, we map out the dependence of the total yields on the starting radius of the oscillations, self-consistently coupling hydrodynamics and nucleosynthesis. The oscillations boost key -nuclides (, ) and long-lived by up to two orders of magnitude, bringing their abundances into agreement with the observations. The best match is found when oscillations commence within of the proto-neutron star surface, indicating fast collective oscillations.
Introduction
Just outside the protoneutron star (PNS) in a core-collapse supernova (CCSN), the dense gas of streaming neutrinos creates extreme conditions unattainable in terrestrial laboratories. Coherent – scattering in this system couples the flavor states of neutrinos on different trajectories, driving collective flavor oscillations [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13]. Analytical and numerical studies of this many-body quantum phenomenon have uncovered very rich physics, including fast collective oscillations (FCO) [14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31] driven purely by – coherent scattering and slow collective oscillations (SCO) [32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45] driven by the interplay between the – coherent scattering and neutrino-mass-induced transformations. Yet, a complete theoretical description of this phenomenon in realistic CCSN conditions remains elusive, due to the extreme complexity of the problem rooted in the nonlinearity of oscillations, the presence of disparate length- and timescales [46, 47, 48], and essential multi-dimensional dynamics [13, 49]. It is then crucial to identify any observables that carry imprints of FCO or SCO, especially those that are sensitive to how far from the PNS surface the oscillations develop. In this Letter, we argue that the desired probe is provided by the pattern of certain proton-rich isotopes observed in the Solar System. Moreover, we show that the oscillations may be necessary to explain the observations.
It has long been known that if collective oscillations were to occur in the hot bubble surrounding the PNS, they could impact nucleosynthesis there [34, 50, 51]. These early papers considered the -process (rapid neutron capture) and invoked sterile neutrinos. It has since been shown, however, that collective oscillations can occur in the hot bubble even with only active neutrinos [35]. Combining representative oscillation scenarios with nucleosynthesis calculations then showed that the -process yields can depend sensitively on where flavor mixing starts [52]. This makes nucleosynthesis a promising complementary probe to neutrino signal detected on Earth, as the latter only reflects the final state of the neutrino evolution.
Our understanding of the type of nucleosynthesis that can occur in the hot bubble has also advanced. The and spectra predicted by modern CCSN simulations tend to yield proton-rich conditions in the neutrino-driven outflow from the PNS surface that forms the hot bubble, especially once the effect of weak magnetism and recoil [53, 54, 55] is accounted for [56]. This is, in fact, a welcome development. While the -process may efficiently proceed in neutron star binary mergers [57, 58, 59], one also needs to identify a site where -nuclides are made. The origin of these isotopes, especially and , which are rather abundant in the Solar System, has been a long-standing mystery [60, 61]. The hot bubble in a CCSN turns out to be a promising site, supplying not only the proton-rich conditions, but also a large flux of streaming electron antineutrinos. The latter helps solve [62, 63, 64] the famous “waiting points problem” of the classic -process [65]. This neutrino-assisted mechanism is called the -process [62].
While there are several recent notable investigations of the impact of FCO on the CCSN explosion mechanism [66, 67, 68, 69, 70, 71, 72, 73, 74], there is only a single study on how the -process is affected by FCO [75]. There is, likewise, a limited number of studies exploring the effects of SCO on the -process [76, 77, 78]. While these papers contain important findings, they do not treat the impact of the flavor transformations on the hydrodynamics self-consistently, which is required to tackle the problem.
**Analysis requirements.—**Flavor oscillations change the spectra of and , by mixing them with the original spectra of and , with . This impacts the -process in three ways. First, it changes the rates of neutron-producing reactions during the proton-capture phase of the -process. This is a far effect of the oscillations, as it occurs hundreds of km from the PNS surface [75, 76, 77, 78]. Second, oscillations modify the electron fraction in the outflow, set by the reactions and . Since the unoscillated energy hierarchy is typically larger for – than –, mixing hardens more than , enhancing relative to and increasing [75]. This is only relevant if the oscillations occur within 10–20 km from the PNS surface, where is set (see Figure 1). It is thus considered a near effect. Third, as another near effect, flavor conversion increases the heating rate [75].
Figure 1. Electron fraction as a function of radius for our benchmark model at , shown for the outflow without neutrino mixing and with mixing occurring at various distances from the PNS surface.
The last effect energizes the outflow and deserves special consideration. Two key physical factors enter here: (i) Effective -process requires a certain matching between the processes that occur at different stages of the outflow [79, 80]. The number of seeds produced during the stage of quasi-statistical equilibrium () has to be in a certain ratio with the number of neutrons produced in the reactions in the -nuclide formation stage (). Refs. [80, 81] established the hydrodynamic conditions for optimal yields: a heavy PNS mass and a subsonic outflow persisting for several seconds. (ii) As argued in [82], the outflow in a CCSN, under realistic conditions, can be either subsonic or supersonic, depending on a combination of relevant parameters. Because of this near-critical nature of the outflow, the extra heating induced by neutrino oscillations might cause supersonic transitions, depressing the yields and counteracting the effect of boosted . To assess the effect of flavor mixing on the yields then requires us to recompute the hydrodynamics of the outflow for each oscillation scenario.
**Methods.—**We use as benchmark a spherically symmetric long-term supernova simulation of a progenitor with a PNS from the Garching CCSN Archive [83, 84, 85]. This state-of-the-art simulation meets the hydrodynamic requirements for efficient , as outlined above, yet was not tailored for the -process by its authors. We take as input the PNS mass , time-dependent PNS radius, (unoscillated) neutrino luminosities and spectra, and the progenitor profile [86]. We model the luminosity evolution with the fit from Ref. [85], and we fix spectral moments at their values at post-bounce time in the SN simulation (when the -process peaks), assuming pinched Fermi-Dirac distributions [87, 88, 89].
As mentioned above, we are compelled to recompute the hydrodynamics for each oscillation scenario. We do so with the framework of Ref. [81], which solves general-relativistic steady-state hydrodynamics in spherical symmetry, with an equation of state including the baryon gas and variable radiation degrees of freedom. Neutrino interactions with both nucleons and electrons are included, and time-dependent boundary conditions at the PNS surface and the hot-bubble’s edge (with proper merging into the colder outflow) are imposed, as discussed in Supplementary Material (SM) A, B. We construct the time evolution through a sequence of snapshots, representing particles launched 0.1 s apart.
Following existing literature [75], we model flavor conversion as complete instantaneous equilibration at a distance from the PNS surface. Beyond this radius, we replace the original spectra with the equilibrated ones, combining all flavors equally across the energy modes: , thereby modifying neutrino energy moments and luminosities (see SM C). [90] This modifies both neutrino heating (thus hydrodynamics) and nucleosynthesis. We vary as a free parameter to quantify the -process’s sensitivity to conversion location.
Nucleosynthesis is computed by post-processing tracer trajectories with the reaction network SkyNet [91, 92], incorporating modifications from Ref. [81] as detailed in SM A (including medium-enhanced triple- [93, 94]). We further modify the code to include distance-dependent flavor mixing (SM C). Each trajectory is launched at after the core bounce and evolves for years in SkyNet, with conditions evaluated at , where is SkyNet’s evolution time (starting from zero at launch). Beyond the benchmark model, we perturb the PNS and progenitor’s properties to test the robustness of our qualitative findings across different regimes.
**Oscillation-driven physical effects.—**The impact of neutrino oscillations on the -process depends on how far from the PNS flavor mixing takes place. If conversions occur at (“far conversions”), only the rate of neutron production during the proton-capture phase is appreciably changed, while and are not affected. For the model we use in this paper, the rates are enhanced by .
Conversions at smaller radii (“near conversions”) induce all three effects simultaneously: enhancement of rates, up to increase in (and also the mass outflow rate ), and boosted . The impacts on and exhibit similar dependence on (see SM C).
Figure Figure 1 shows at a representative post-bounce time for several values of the flavor equilibration radius. Notice that the important weak-magnetism and recoil corrections [53, 54, 55] are included here and in what follows. With the original neutrino spectra, approaches an equilibrium value of (solid violet). Neutrino mixing increases , rendering the outflow more proton-rich, consistent with Ref. [75]. This enhancement is stronger the closer mixing occurs to the PNS. For , increases up to (solid yellow), while there is negligible impact for .
Notice that earlier studies found to be conducive to successful -process, e.g. [79, 80, 81]. That flavor mixing brings close to this desired value is, in fact, not an accident of the specific simulation we are considering and rather insensitive to the details of flavor conversion. As can be readily shown (see SM A), in the limit when the and spectra are identical, reaches . Since the original and spectra are close, any flavor mixing dynamics that transfers substantial flux into boosts to near the desired value . For instance, were the outcome of near conversion to be spectral swaps of the kind studied in Refs. [37, 41, 42, 95–98], the resulting shift would be comparable.
Results of near flavor conversion.—Oscillation-induced changes to (Figure 1) and (explicitly shown in SM C) are largest as flavor conversion occurs close to the PNS. Thus, we begin by comparing two extreme cases: no mixing and flavor equilibration occurring above the PNS surface ().
The left panels of Figure 2 show the evolution of the outflows launched at different for our benchmark model. Without mixing, the outflow remains subsonic throughout the cooling phase (upper left). For (lower left), the additional heating induces a transonic transition at , and the termination shocks, visible as discontinuities in the radial temperature profiles around , persist until . Such emergence of transonic outflows from a modest increment of heating illustrates the near-critical nature of CCSN hydrodynamics [82], and, practically, the necessity of self-consistently solving the hydrodynamics for small .
Figure 2. Left: Radial temperature profiles for outflows launched at , depicted here in steps of . The green and yellow bands correspond to the seed-formation and -nuclide formation stages of the -process, respectively [79, 81]. Right: Time evolution of the production rates of Mo (red) and Ru isotopes (blue), and (black), with vertical dotted lines delimiting the time window where the outflows contain a supersonic transition. Top (bottom) panels show the unmixed case (flavor equilibration occurring above the PNS surface).
The right panels of Figure 2 show the corresponding production rates as a function of time of several -nuclides and the long-lived radionuclide , a key diagnostic of the -process’s efficiency [61, 80, 81].[99] Without mixing (upper right), -nuclides are efficiently produced in an optimal window of – s in the early cooling phase, while is produced later. With km (lower right), the synthesis of Mo and Ru isotopes is enhanced by factors of – and –, respectively, reflecting higher and (anti)neutrino capture rates. Importantly, the supersonic transition sets in only after the production rates have peaked (vertical dotted line at s). While the onset of mildly supersonic conditions noticeably accelerates the post-peak decline of Mo and Ru nucleosynthesis, the production rates remain higher than the original (unmixed) case at each . Oscillations increase production by as much as two orders of magnitude. In contrast to -nuclides, has a significantly broadened production window and distinct temporal behavior.
Yields as a function of mixing radius.—To compare with observations, we next compute the total yields in a given explosion by integrating over time. We present our results as time-averaged yields normalized by the total mass ejected
With this, we systematically investigate the dependence of on the mixing location, the first time such a study is undertaken in the literature. Figure 3 shows versus mass number for near (left) and far (right) conversions for different values of . For the -nuclides labeled on the upper axis, corresponds to of the indicated isotope, the only relevant species at that produced by the -process.
Figure 3. Time-averaged yields for our benchmark model under flavor conversion occurring at various radii from the PNS. Curves show versus mass number ; diamonds mark the corresponding abundance. Left: “Near” conversion, i.e., mixing within 30 km from the PNS surface, which modifies and and also enhances neutron production rates. Right: “Far” conversion, i.e., mixing at larger radii, affecting only neutron production during the -nuclide formation stage. Black arrows indicate the minimum abundances needed to reproduce the solar values (see main text and SM).
From the left panel of Figure 3, we see that the largest enhancement is obtained when km. Here, for , flavor equilibration enhances the yields up to orders of magnitude relative to the unmixed case (solid violet) for both -nuclides (curves) and (diamonds), with factors of for and for . This reflects the enhancement in the production rates illustrated in Figure 2. Counterintuitively, despite larger values at smaller (Figure 1), the cases with km yield similar . This is due to the competing effect of the increase in heating: mixing closer to the PNS triggers an earlier transonic transition that shortens the optimal production window, partially offsetting the -driven gain.
For km the -yields are strongly sensitive to the value of . Here, no supersonic transition takes place, and the effect of is not counterbalanced by hydrodynamics. As expected, the yields fall monotonically with increasing , approaching a factor of 2–3 enhancement relative to the unmixed case for km (see SM D).
For km (right panel of Figure 3) the yields are nearly independent of . In this regime, and are effectively not modified, leaving the far effect as the only relevant process, which operates at the -nuclide production stage, starting roughly at 300 km (see Figure 2, top-left panel). At larger , enhancement diminishes, becoming negligible for km where -nuclide production ends.
**Comparison with solar abundances.—**Black arrows in Figure 3 mark solar-abundance thresholds from meteoritic data [100] for –. These are computed by requiring that -process mass fractions, when diluted in the total ejecta, exceed solar values by one order of magnitude [79, 81] (see SM D). The unmixed case falls short by approximately one order of magnitude. Far conversions, increasing yields by only factors of a few, remain insufficient to reach solar thresholds. In contrast, near conversions enable agreement with solar abundances [79, 81]: our model successfully meets solar requirements of the key Mo and Ru isotopes for –, while for , , as well as , become viable.
**Discussion.—**Near flavor conversion makes our model an efficient -process site. Modest perturbations of the benchmark parameters preserve our qualitative results: when conversion occurs sufficiently close to the PNS surface, yields reach the solar thresholds, provided that the PNS and progenitor remain sufficiently massive. This is demonstrated for a representative case in SM E, obtained by reducing the progenitor mass to and modestly increasing the front shock velocity to . In this model, lower confining pressure triggers an earlier transonic transition, clipping the optimal window even without neutrino mixing. Flavor conversion leads to similar enhancement factors of the yields: for , the transonic transition occurs earlier, yet an early subsonic interval persists wherein boosted and enhance production. Though reduced relative to the benchmark, yields of , , and still exceed solar thresholds for .
More generally, flavor mixing enhances yields in all cases we explored, even when (mild) transonic transitions already occur near the beginning of the cooling phase. However, the -process is inefficient in strongly supersonic conditions [80, 81], rendering mixing-induced enhancements insufficient to reach solar thresholds. An exploration of different models is left for future work [101].
Our systematic study establishes that near conversion—triggered by FCO (see e.g., Refs. [9, 10, 12] for recent reviews) within from the PNS surface [21, 22, 23, 24, 25]—increases yields by more than one order of magnitude, allowing agreement with solar observations for sufficiently massive progenitors and PNS. By contrast, far conversion, associated with SCO at – [35, 36, 37, 38], enhances yields by only factors of a few. Mixing at larger distances, i.e. , has negligible impact [102]. These results suggest that FCO may play a crucial role in establishing the observed abundances of light -nuclides in the solar system.
The relative importance of far and near conversions depends on the closeness of unoscillated and spectra in both energy and luminosity during cooling. Similar spectra are commonly observed in state-of-the-art simulations [84, 103, 104, 105, 106, 107], while large differences typically occur in simulations neglecting inelastic neutrino-nucleon scattering corrections [77] (see Fig. 14 in Ref. [108]). The GARCHING group finds an average energy difference [84, 103], slightly smaller than the one observed by other groups (–) [104, 105, 106, 107]. Our far-conversion results agree qualitatively with Ref. [76], which assumes based on simulations from Ref. [109]. By contrast, adopting greater – spectral differences (), Refs. [77, 78] found substantially larger enhancements in -capture rates and -yields. Thus, the impact of far conversion is sensitive to the details of neutrino transport, which motivates careful future investigations, given its importance to the estimation of -yields in the presence of oscillations.
**Conclusions.—**We saw that flavor oscillations close to the PNS can transform the nature of the outflow and with it the efficiency of the -process. This requires the recomputation of the hydrodynamics for each oscillation scenario, while also including the oscillation effects on the electron fraction and the neutron production rate. Such a self-consistent treatment of the hydrodynamics and nucleosynthesis, as well as the systematic investigation of the dependence of the yields on the starting radius of the oscillations, are presented here for the first time.
Our calculations confirm that the yields of the -process are sensitive to the starting radius of collective oscillations. We find that the impact of the oscillations is greatest when they occur within of the PNS surface, in which case the yields of the -nuclides in the crucial – mass range are boosted by up to – times. In comparison, when the neutrino mixing radius is , the enhancement is an order of magnitude smaller and, moreover, the sensitivity to the exact location of the oscillations is reduced.
Crucially, when the oscillations occur close to the PNS surface, and only then, the predicted abundances line up with the solar-system observations. This applies to all -isotopes up to , as well as to , which also serves as a probe of the efficiency of the -process. Thus, collective oscillations may be responsible for the observed pattern of these isotopes. Notably, the agreement is reached entirely within the physical conditions of a modern CCSN simulation, without tuning any aspects of explosion, such as the electron fraction, neutrino fluxes, entropy per baryon, or the outflow profile, the first such example in the literature.
Given today’s best available information on the dynamics of collective oscillation and on the spectra of different neutrino flavors, our results might be taken as an indication that the relevant mechanism is fast collective. Further investigations incorporating collective oscillation dynamics self-consistently into CCSN simulations are needed to investigate the robustness of this finding.
Acknowledgments
The work of all authors at SLAC was supported by the U.S. Department of Energy under contract number DE-AC02-76SF00515. AVP acknowledges partial support from the U.S. Department of Energy under contract number DE-FG02-87ER40328 at the University of Minnesota, and would also like to thank SLAC for their hospitality and support during the period of completion of this project. IPG is supported by NSF Physics Frontier Center Award number 2020275. PM thanks the Neutrino Theory Network Program Grant under award number DE-AC02-07CHI11359, which supported this work while she was at UC Berkeley. Additionally, PM is grateful for the support of the Cambridge-Infosys AI Center for funding her current research.
References
Solar-System Abundances of p-Nuclides Probe Collective Neutrino Oscillations in Supernovae
In this Supplemental Material (SM), we describe the methods employed in modeling the hydrodynamics, nucleosynthesis, and flavor equilibration prescription. We present in detail the model parameters extracted from the GARCHING supernova simulation, define the production factors used to assess the agreement with solar abundances, and extend our analysis in the main text to mildly supersonic models.
Methods: hydrodynamics and nucleosynthesis
We compute nucleosynthetic yields following a two-step approach, akin to Refs. [80, 81]. First, we construct the trajectories of tracer particles for a number of time snapshots. Then, we obtain the yields by post-processing each tracer trajectory with the open-source nuclear reaction network SkyNet [91].
Tracer trajectories are divided into two phases. In the first, described as a steady-state neutrino-driven outflow (NDO), material accelerates from the protoneutron star (PNS) surface and subsequently decelerates through interaction with the surrounding ejecta. In the second phase, the material joins the homologous expansion behind the forward shock and co-moves with it.
Whether the outflow becomes supersonic and develops a termination shock, or instead remains subsonic throughout, is not known a priori. The system often lies close to criticality [82], as our benchmark model demonstrates, with the character of the outflow determined by the interplay of neutrino heating, PNS properties, and the confining pressure imposed by the progenitor profile. Rather than postulating a supersonic or subsonic solution, we solve the hydrodynamic equations as a boundary value problem consistent with the confining pressure, at the edge of the hot bubble, inferred from the progenitor profile and front-shock velocity at each temporal snapshot. This method is described in detail in Ref. [81]. To incorporate the effect of neutrino oscillations, we posit a radius after which the neutrino flavors equilibrate (see SM C), which maximizes the effect of neutrino mixing, and vary this radius as a parameter. Concretely, this means that the neutrino spectra which enter the neutrino capture terms in the outflow calculation as well as in SkyNet, are discretely modified before and after this mixing radius. This rather minimal prescription of positing a radius of equilibration does offer us a straightforward way of studying the radial dependence of oscillations’ impact on nucleosynthesis agnostic of the microphysics.
The main ingredients of our model for the outflow have been delineated in the Letter and detailed in [81]. Here we accentuate a few aspects. Although our hydrodynamic equations assume the steady-state condition, we consider the time evolution of the boundary conditions—including PNS radius, luminosities, confining pressure—and compute the outflow every 0.1 s. This yields a series of snapshots. Our Equation of State (EoS) incorporates variable radiation degrees of freedom and the monatomic baryon gas. A key ingredient in determining the NDO is the neutrino net heating rate , computed as in Ref. [81]. We include (i) charged-current absorption and emission on free nucleons, and , (ii) elastic neutrino-electron scattering , and (iii) cooling from electron-positron annihilation , following the prescription in Refs. [116, 117]. Charged current absorptions, with a heating rate scaling with the third moment of the neutrino energy distribution—two powers from the cross section and one from the deposited energy—are the main heating processes and also determine the neutron-to-proton ratio in the outflow. Since SN simulations typically provide only the first two energy moments, we parametrize neutrino spectra (see SM B) to reconstruct higher-order moments required to compute .
Nucleosynthesis is computed by post-processing each tracer with SkyNet, using nuclear reaction rates from the ReacLib library, version 2.2 [118], with the modifications described in Ref. [81] (see Sec. 8 and App. G therein), including for instance time-evolving neutrino luminosities, the blueshift of neutrino temperature and luminosities, medium-enhanced triple- reaction rates from Refs. [93, 94], and the correct decay rate for [80]. Additionally, for this analysis we further modify the code to account for neutrino flavor conversion at a mixing radius . For each tracer, we start nucleosynthesis runs at MeV, using as SkyNet inputs the time evolution of its temperature , density , and its location , along with a constant PNS mass and time-dependent PNS radius , assumed to be equal to the neutrinosphere radius. The time evolution of the electron fraction is computed by SkyNet and it does incorporate weak-magnetism (WM) and recoil corrections with the method described in Ref. [81].
In the absence of mixing, approaches the asymptotic value given by [53, 55, 119]
where is the nucleon mass, , , and , with the -th moment of the energy distribution. Note that this expression is derived using an approximation to the exact cross section, only including terms up to first order in and [119]. For a more exact treatment, and how it reduces to this expression at leading order, see Eqs. (22)–(25) from Ref. [55]. In presence of flavor conversion, the asymptotic value of increases and depends on the radius at which conversion occurs, as discussed in the main Letter.
To incorporate flavor mixing, we modify SkyNet to include and in addition to and , which were the only species considered in our previous analyses. As input, we provide time-dependent unmixed neutrino luminosities and spectra. Each (anti)neutrino spectrum is modeled using a fit that enables the reconstruction of higher-order energy moments from those provided by supernova simulations (see SM B). Starting from the original spectra of , , , and , we compute distance-dependent spectra and determine the corresponding mixed quantities at the radius , as described in SM C.
Model parameters from simulations
We model the NDO using input quantities extracted from simulations performed with the PROMETHEUS-VERTEX neutrino-hydrodynamics code [120]. This code solves the fully energy- and velocity-dependent neutrino transport equations for all six species of neutrinos and antineutrinos, employing a state-of-the-art treatment of neutrino interactions [109, 121, 122]. As discussed in Ref. [84], these simulations include a one-dimensional treatment of PNS convection through a mixing-length description of the convective fluxes [7]. They also account for the presence of muons in the hot PNS and the associated muonic-neutrino interactions [122]. However, the resulting differences between the luminosities and spectra of and neutrinos remain relatively small throughout most of the PNS cooling phase.
For our benchmark model, we adopt parameters from a 1D SN simulation from a progenitor [86] performed by the GARCHING collaboration [83]. This model possesses the qualitative features required for efficient nucleosynthesis [80], specifically a massive PNS (with baryonic mass of ) and subsonic outflows. We notice that such heavy PNS masses naturally arise in multi-dimensional simulations with massive progenitors [103, 123].
In this model, the evolution of the PNS radius (solid line in Figure S1) is well fitted by
Figure S1. Evolution of the PNS radius as a function of post-bounce time for our benchmark model, parametrized according to Eq. (3).
where denotes the post-bounce time in seconds and the coefficients are , and . We adopt this as the near boundary condition for our outflow calculation. The far boundary conditions are fixed in our model by taking a front-shock (FS) velocity of , consistent with the simulation results.
To assess the impact of an early transition from subsonic to supersonic outflows, we also perturb our benchmark model by considering a slightly lighter progenitor of [86], with , while keeping the PNS properties fixed. A more detailed discussion about this model is reported in SM E.
Our reference SN simulation employs the SFHo EoS [124, 125] and corresponds to the 1.93-SFHo model in Refs. [84, 85]. We describe below its main features.
The presence of convection accelerates PNS cooling and modifies the time evolution of the neutrino luminosity. As shown in Ref. [126], a kink in the neutrino light curve when plotted in a doubly logarithmic scale marks the end of convection in the mantle region of the PNS, while convection remains active in the deeper high-density core. This feature typically occurs at [84, 85, 127]. Consequently, the neutrino luminosity in the cooling phase cannot be accurately described by a simple exponential or power-law decay when convection is present.
Following Ref. [85], we parametrize the neutrino luminosity as
for each (anti)neutrino flavor , with representing non-electron species. The parameters used in this work are taken from Tabs. V and VI of Ref. [85] for electron and non-electron (anti)neutrinos, respectively. We mention here that Tab. VI in Ref. [85] reports the parameter values for . However, as discussed in Ref. [85], the differences between and are small, and we therefore use and as representative of and .
To simplify the implementation of the time dependence of in the nuclear-reaction network SkyNet [91], we assume a common time evolution for all (anti)neutrino species. Specifically, the parameters , , and are taken as the averages over all species. With this prescription, the optimal production window for the process occurs at in our benchmark model. We therefore fix the normalization constants in order to reproduce the luminosities of all species at . The fit parameters for in the model are listed in Table S1.
| [ erg/s] | 10.61 | 11.14 | 12.90 | 13.56 |
| 0.584 | 0.584 | 0.584 | 0.584 | |
| [s] | 9.31 | 9.31 | 9.31 | 9.31 |
| 3.95 | 3.95 | 3.95 | 3.95 | |
| [MeV] | 10.80 | 13.58 | 13.10 | 13.70 |
| [MeV] | 148.03 | 235.25 | 223.34 | 244.98 |
| [MeV] | 3.21 | 4.09 | 4.14 | 4.36 |
| 1.17 | 0.97 | 0.06 |
Table S1. Input parameters for neutrino luminosities (upper blocks) and energy spectra (lower blocks) of all (anti)neutrino species used for our benchmark model.
The resulting time evolution of in our model is shown in the left panel of Figure S2, exhibiting the hierarchy at any given time. The small difference between and is mainly related to WM effects [55], which increase neutrino interaction rates and lower the rates for antineutrinos (see, e.g., the discussion in Ref. [89]). These effects also lead to small differences between the and spectra, as discussed below.
Figure S2. Time evolution of the neutrino luminosities (left panel) and production spectra at (right panel) for (solid black), (dashed black), (solid red), (dashed red), in our model.
Following the default option in SkyNet, in this analysis we model neutrino spectra with a normalized “pinched” Fermi-Dirac distribution [87, 88]
where the spectral parameters and are obtained directly from the same SN simulation, and is fixed by the luminosity. This parametrization provides an accurate fit to simulated spectra [89] and enables the reconstruction of higher-order energy moments from the moments supplied by the simulations.
Since SN simulations provide only the first and second moments of the energy distributions, we relate the effective neutrino temperature and the degeneracy parameter to these quantities as
where are the Fermi integrals defined as
where is the polylogarithm of order .
Since and vary only mildly during the first s after core bounce (see, e.g., Figure 1 of Ref. [84]), we use time-independent spectral parameters and . These are fixed by matching and to the values extracted from the simulations at the reference time s. The adopted values of , , , and for our benchmark model are reported in Table S1.
The right panel of Figure S2 shows the production spectra (in MeV s) for electron (black) and non-electron (red) neutrinos (solid lines) and antineutrinos (dashed lines) at s, normalized to reproduce the corresponding luminosities at that time. For each flavor, neutrino spectra peak at lower energies than the corresponding antineutrino spectra, with about 25% larger than and about 4% larger than . The few-percent difference between the and spectra arises from WM effects; in their absence the two spectra would be essentially identical.
With the assumed values of neutrino luminosity and spectra, the equilibrium value of in the NDO (see Eq. (2)) is approximately 0.57.
Mixing of neutrino flavors
As a benchmark for the effect of collective neutrino oscillations, we consider complete flavor equilibration, which transforms distributions as
where, under the assumption of flavor equipartition, the mixed (anti)neutrino distribution function receives equal contributions from () and the individual heavy-lepton (anti)neutrinos. As discussed in the main text, our results remain valid provided that oscillations convert a substantial fraction of the heavy-lepton-flavor (anti)neutrino flux into electron (anti)neutrinos. Assuming the pre-mixing spectra are of Fermi-Dirac shape as reported in Eq. (5), the mixed luminosities are
and the energy moments are
with as the Fermi-Dirac integral of power defined in Eq. (7). An analogous expression for follows by replacing () everywhere in Eq. (10).
The left panel of Figure S3 shows the unmixed (black) and mixed (orange) production spectra of electron neutrinos (solid) and antineutrinos (dashed) at . The contribution of non-electron (anti)neutrinos—whose spectra are more similar (see the right panel of Figure S2)—reduces the difference between neutrino and antineutrino spectra after mixing. The resulting mixed luminosities and spectra yield via Eq. (2) and are used in the outflow calculation described in SM A.
Figure S3. Left: Production spectra at for unmixed (black) and mixed (orange) electron neutrinos (solid curves) and antineutrinos (dashed curves), in our model. Right: Ratio of neutrino to antineutrino spectra for the unmixed electron flavor (black) and non-electron flavors (red), as well as for the electron flavor in the mixed case (orange) in the model. The vertical orange lines correspond to the average energy (solid) and its inverse ratio with the second (dashed) and third (dotted) moments of the mixed electron antineutrino’s energy distribution.
A more quantitative comparison is obtained by considering the ratio of neutrino to antineutrino spectra, shown in the right panel of Figure S3 for the unmixed electron flavor (black), non-electron flavors (red), and the mixed electron flavor (orange). In the mixed case, the spectrum receives a 1/3 contribution from electron flavors and 2/3 from non-electron species, so the ratio lies between the unmixed cases. Vertical lines indicate the average energy (solid), as well as the parameters (dashed) and (dotted) for the mixed antineutrino spectra. These quantities enter the determination of and in Eq. (2). Due to the similarity between and spectra, the ratio of mixed to spectra is close to unity near and , explaining why is obtained.
Larger values of enhance the neutrino net heating rate and consequently . The left panel of Figure S4 shows the radial profile of at near the PNS for our benchmark model, for flavor mixing at different distances from the PNS surface. For each mixing scenario, deviations from the unmixed case are negligible at , while at larger radii increases by due to higher neutrino luminosities and harder spectra. Enhanced closer to the PNS accelerates the outflow and increases .
Figure S4. Left: Neutrino net heating rate as a function of radius at for flavor conversion at different distances from the PNS surface. Right: Relative enhancement of the mass outflow rate as a function of launch time for different mixing radii . Vertical lines denote the beginning of the transonic case for (solid), (dashed) and (dash-dotted). Color coding follows Figure 3.
The right panel of Figure S4 shows the time evolution of enhancement relative to the unmixed case for different mixing radii . The relative enhancement decreases with time: it is for , for , and for larger mixing radii. The increased also raises the outflow velocity, inducing a transonic transition at (vertical solid line), (dashed line), and (dash-dotted line). For , the transonic transition occurs at progressively later times as increases. For , the transonic transition disappears and the velocity enhancement remains below relative to the unmixed case.
Integrated yields and production factors
To compare yields with solar-system abundances [100], we compute time-integrated yields. For a nuclide , the time-averaged abundance is defined as [80]
where and denote the instantaneous abundance and mass outflow rate, respectively. The resulting time-averaged yields are shown for the benchmark model in Figure 3 and for the model in the left panel of Figure S8.
Figure S8. Left: Time-averaged yields for the 18.6 model under flavor conversion occurring at different radii from the PNS. Curves show versus mass number ; diamonds mark the corresponding yield. Black arrows indicate the minimum abundances needed to reproduce the solar values. Color coding follows Figure 3. Right: Production factors of -nuclides in the 18.6 model, for the unmixed case (upper right panel) and flavor mixing occurring 1 km above the PNS ( km—lower right panel). In each panel, the cyan band indicates the co-production region, and the dashed orange line marks the threshold corresponding to an overproduction factor of 10, required to account for solar abundances.
Time-averaged yields increase as mixing occurs closer to the PNS, as discussed in the main text. Figure S5 shows enhancement factors for -nuclides (curves) and (diamonds) under near (left) and far (right) conversions. The enhancement grows with mass number in all scenarios.
Figure S5. Enhancement factors for time-averaged yields under flavor equilibration at various radii, for near (left) and far (right) conversions. Curves show enhancement versus mass number ; diamonds mark enhancement. Vertical lines denote (solid red), (dashed red), (solid blue), and (dashed blue). Color coding follows Figure 3.
For near conversions, the cases nearly coincide, yielding factors of for and for . At , enhancements remain within a factor of 2 of these values: for and for . The cases produce smaller boosts: and , respectively.
Far conversions yield more modest increases. For –, enhancement factors remain . At larger radii, where conversion occurs during the -nuclide production phase, enhancement diminishes further (dotted purple line, ) and becomes negligible for (dashed violet line).
The increase differs qualitatively. Near conversion enhances its abundance by up to 2 orders of magnitude, with factors of for . Far conversion produces a factor of increase at , and enhancement persists even beyond 1000 km, reaching 1.4 at (violet diamond).
From these abundances we define the isotopic production factor [79, 80, 81]
where is the solar-system mass fraction of isotope , and is the time-averaged mass fraction obtained from the SkyNet output, with . Solar abundances are taken from meteoritic measurements of carbonaceous chondrites [100], following Ref. [81].
Production factors quantify both relative and absolute abundances. Accounting for astrophysical uncertainties, nuclides are considered co-produced if , where is the largest production factor in a given model [79, 128]. Absolute yields are characterized by the overproduction factor [79, 80, 129]
where the ratio is the astrophysical dilution factor, with the total mass driven out in the NDO and the total mass ejected in the explosion. Values are typically required to reproduce solar abundances [79, 80, 129]. The ejected mass is fixed by the underlying SN simulation and is therefore independent of the mixing prescription, with for the progenitor and for the progenitor. Flavor mixing increases neutrino heating in the hot bubble and thus raises the mass-outflow rate . As a result, flavor equilibration occurring closer to the PNS yields a larger total outflow mass, , and a greater dilution factor. For the model, we find (dilution factor ) without mixing, rising to (dilution factor ) for the case. For the model, we obtain (dilution factor ) in the unmixed case, and (dilution factor ) for .
We show the production factors of different -nuclides for our benchmark model in Figure S6 for mixing occurring at different distances from the PNS surface, comparing “far” conversion (left panel) and “near” conversion (right panel). Each panel also displays the production factor of the extinct radionuclide (green diamond), normalized to of the solar abundance. The associated error bars reflect the inferred production ratio , constrained to lie in the range – to account for the solar-system abundance of , into which decays [61, 130, 131] [112].
Figure S6. Production factors of -nuclides (purple dots) and (green diamond with error bars) in the 20 model, under flavor conversion occurring at different radii from the PNS. In each panel, the cyan band indicates the co-production region, and the dashed orange line marks the threshold required to satisfy solar observations. The uncertainty on arises because is an extinct radionuclide: its solar abundance cannot be measured directly, but is inferred from meteoritic data to correspond to a production fraction within –.
Production factors increase as mixing occurs closer to the PNS. For far conversion, Mo and Ru isotope production factors increase by more than a factor of 2 from the unmixed case to km, while results at km and 40 km are the same. For near conversion, production factors increase monotonically from km to 1 km due to enhanced and .
In all cases, has the highest production factor. The co-production band (cyan) includes all isotopes with . Without mixing, all species with fall below this threshold. The increase induced by far conversion is not enough to reach the co-production region. For km, production factors approach co-production, and for km, isotopes up to are co-produced. For km, flavor mixing increases Mo and Ru isotopes by a factor , by , and enhances and by and more than 3 orders of magnitude, respectively.
The horizontal dashed orange line marks the threshold required to reproduce solar abundances [Eq. (13)]. Mixing increases through enhanced heating, yielding larger and thus slightly lowering this threshold for smaller . In the unmixed scenario, nuclides with are neither co-produced nor sufficiently abundant to meet solar constraints. Far conversion cannot reach the solar threshold (all nuclides remain below the orange line), while Mo and Ru isotopes satisfy solar observations for km. For km, Mo and Ru isotopes—as well as within uncertainties—are co-produced and meet solar-abundance requirements.
From the overproduction factor, we derive the solar-abundance thresholds shown in Figure 3 and Figure S8 by inverting the condition [79, 80, 129]:
For , the minimum threshold corresponds to normalizing its production to of the abundance, while the black dot denotes the benchmark normalization of .
To remain conservative, we always show only for the unmixed case, which yields the largest thresholds relative to the corresponding mixed scenarios.
Mildly-supersonic models
A natural question one has, after seeing the results of our benchmark model, is whether the conditions of this simulation, albeit taken off the shelf, are “fine-tuned” or coincidental so that when mixing is introduced, the supersonic transition enters only after the nucleosynthesis has mostly concluded. Had the transition occurred earlier and nucleosynthesis operated in a supersonic environment, there could be a real competition between the favorable influence of increased and the adverse effect of the supersonic regime.
To this end, we perturb the far boundary condition of our benchmark model by adopting a progenitor of with km/s, while retaining the PNS properties and neutrino spectra. This lowers the confining pressure of the hot bubble and increases the outflow’s velocity, inducing a supersonic transition within the first seconds of the cooling phase without the introduction of neutrino oscillations. Considering such a mildly supersonic model affords us insights into the other side of near-critical supernova outflows.
Figure S7 compares the time evolution of the benchmark model (left panels) and the perturbed model (right panels) during the cooling phase ( s). The upper panels show the evolution of the outflows in the unmixed (panels and ) and in the km cases (panels and ), while the lower panels display the corresponding production rates as a function of the launch time. In contrast to the benchmark model, this (panel ) model exhibits a transonic transition at s, followed by a return to subsonic outflows at s with the decline of neutrino luminosities. Flavor equilibration accelerates the onset of this transonic transition, which takes place at s when mixing occurs 1 km above the PNS surface (panel ). This presents an interesting study as the induced supersonic regime now intrudes on the early cooling phase, when the process is most efficient.
Figure S7. Upper panels: Radial temperature profiles for outflows launched at s, depicted here in steps of 0.5 s. The green and yellow bands correspond to the seed-formation and -nuclide formation stages of the -process, respectively [79, 81]. Upper panels: Time evolution of the production rates of Mo (red) and Ru isotopes (blue), and (black), with vertical dotted lines delimiting the time window where the outflows contain a supersonic transition. In each set of plots, the left (right) panels refer to the 20 (18.6 ) progenitor, while the upper (lower) panels are dedicated to the unmixed ( km) case.
What is the impact on nucleosynthesis? As shown in the lower panels of Figure S7, without mixing, the production rates in the optimal window (prior to the transonic transition at s) are smaller by a factor of compared to the benchmark model (panel ). When mixing occurs 1 km above the PNS, the supersonic transition occurs near the end of the optimal production window in the model (panel ), whereas in this model (panel ) it clips the synthesis of -nuclides, preventing the process from attaining its peak rates.
Despite this hydrodynamic effect, in both models mixing at km enhances the production rates by a factor of for Mo and for Ru isotopes during the optimal window. The s reduction in the duration of this window is subdominant relative to this enhancement from the increased and . Thus, in both the and models, effective neutrino mixing near the PNS renders higher nucleosynthetic yields. The production of (black lines) is similarly improved in both models, with flavor equilibration at km extending the production window up to s.
The time-integrated results are reported in Figure S8. Its left panel shows the time-averaged abundances for the model, comparing the unmixed case (solid black line) with scenarios of varying radii of flavor equilibration, given by the displacement from the PNS radius . The qualitative behavior is similar to that of the benchmark model (Figure 3): mixing magnifies yields, generally as the equilibration radius approaches the PNS. For – km, yields increase by less than one order of magnitude, while when flavor conversion occurs sufficiently close to the PNS ( km), yields increase by up to a factor of relative to their original values. Notably, the km case produces slightly larger yields than the km scenario. This occurs because the earlier transonic transition in the km case partially counteracts the larger boost, resulting in a net reduction compared to km. Overall, the yields from the 18.6 model are systematically lower by about one order of magnitude compared to the corresponding cases in the benchmark model. Consequently, the time-averaged yields of and exceed the solar-abundance thresholds (blue arrows) only if mixing occurs at km, i.e., at smaller radii than for the 20 case, where km is sufficient. A similar conclusion applies to production. In the absence of neutrino mixing, the predicted abundance lies well below the solar threshold (see the black diamond in the left panel of Figure S8), whereas mixing at small radii (see, e.g., the red diamond corresponding to km) brings agreement with observations.
Finally, the right panel of Figure S8 shows the production factors of the -nuclides (purple dots) and (green diamond with error bars), as defined in SM D, for the unmixed case (upper panel) and for mixing at km (lower panel). As in the benchmark model, exhibits the largest production factor. Without neutrino oscillations, yields beyond are strongly suppressed, while with flavor equilibration at km, and lie within a factor of below the co-production band (cyan region), indicating non-negligible production within uncertainties. The horizontal dashed orange line denotes the threshold required to reproduce solar abundances, corresponding to (see Eq. (13) and the discussion below). Mixing enhances the mass outflow rate through increased heating, leading to a larger and thus a slightly lower threshold in the km case relative to the unmixed scenario (see SM D). Overall, production factors in the 18.6 model are smaller than in the benchmark 20 case (see Figure S6), implying that -nuclides with are severely underproduced without mixing, while with mixing only species up to exceed the solar threshold.
In summary, the 18.6 model has reduced yields relative to the benchmark, yet effective neutrino mixing still supports efficient process. Mo and Ru isotopes are produced at non-negligible levels—just below the co-production band yet above the solar threshold. In both models, mixing salvages the otherwise underproduced , putting it close to the co-production band (within ) and satisfying observational constraints.
References
- [1]J. T. Pantaleone, Neutrino oscillations at high densities, Phys. Lett. B287 (1992) 128.DOI
- [2]G. Sigl and G. Raffelt, General kinetic description of relativistic mixed neutrinos, Nucl. Phys. B406 (1993) 423.
- [3]B. H. J. McKellar and M. J. Thomson, Oscillating doublet neutrinos in the early universe, Phys. Rev. D49 (1994) 2710.pubmed.ncbi.nlm.nih.gov/10017264
- [4]S. Samuel, Neutrino oscillations in dense neutrino gases, Phys. Rev. D 48 (1993) 1462.pubmed.ncbi.nlm.nih.gov/10016384
- [5]Y. Z. Qian and G. M. Fuller, Neutrino-neutrino scattering and matter enhanced neutrino flavor transformation in Supernovae, Phys. Rev. D51 (1995) 1479 [astro-ph/9406073].
- [6]H. Duan, G. M. Fuller and Y.-Z. Qian, Collective neutrino oscillations, Annual Review of Nuclear and Particle Science 60 (2010) 569.
- [7]A. Mirizzi, I. Tamborra, H.-T. Janka, N. Saviano, K. Scholberg, R. Bollig, L. Hüdepohl and S. Chakraborty, Supernova Neutrinos: Production, Oscillations and Detection, Riv. Nuovo Cim. 39 (2016) 1 [1508.00785].DOI
- [8]S. Chakraborty, R. Hansen, I. Izaguirre and G. Raffelt, Collective neutrino flavor conversion: Recent developments, Nucl. Phys. B 908 (2016) 366 [1602.02766].DOI
- [9]I. Tamborra and S. Shalgar, New Developments in Flavor Evolution of a Dense Neutrino Gas, Ann. Rev. Nucl. Part. Sci. 71 (2021) 165 [2011.01948].DOI
- [10]F. Capozzi and N. Saviano, Neutrino Flavor Conversions in High-Density Astrophysical and Cosmological Environments, Universe 8 (2022) 94 [2202.02494].DOI
- [11]S. Richers and M. Sen, Fast Flavor Transformations, in Handbook of Nuclear Physics (I. Tanihata, H. Toki and T. Kajino, eds.), pp. 1–17. Springer, 2023. 2207.03561. DOI.
- [12]M. C. Volpe, Neutrinos from dense environments: Flavor mechanisms, theoretical approaches, observations, and new directions, Rev. Mod. Phys. 96 (2024) 025004 [2301.11814].arxiv.org/abs/2301.11814
- [13]L. Johns, S. Richers and M.-R. Wu, Neutrino Oscillations in Core-Collapse Supernovae and Neutron Star Mergers, Ann. Rev. Nucl. Part. Sci. 75 (2025) 399 [2503.05959].DOI
- [14]R. F. Sawyer, Speed-up of neutrino transformations in a supernova environment, Phys. Rev. D 72 (2005) 045003 [hep-ph/0503013].DOI
- [15]R. F. Sawyer, The multi-angle instability in dense neutrino systems, Phys. Rev. D 79 (2009) 105003 [0803.4319].DOI
- [16]R. F. Sawyer, Neutrino cloud instabilities just above the neutrino sphere of a supernova, Phys. Rev. Lett. 116 (2016) 081101 [1509.03323].DOI
- [17]B. Dasgupta, A. Mirizzi and M. Sen, Fast neutrino flavor conversions near the supernova core with realistic flavor-dependent angular distributions, JCAP 02 (2017) 019 [1609.00528].DOI
- [18]I. Izaguirre, G. Raffelt and I. Tamborra, Fast Pairwise Conversion of Supernova Neutrinos: A Dispersion-Relation Approach, Phys. Rev. Lett. 118 (2017) 021101 [1610.01612].DOI
- [19]S. Chakraborty, R. S. Hansen, I. Izaguirre and G. Raffelt, Self-induced flavor conversion of supernova neutrinos on small scales, JCAP 01 (2016) 028 [1507.07569].DOI
- [20]S. Chakraborty, R. S. Hansen, I. Izaguirre and G. Raffelt, Self-induced neutrino flavor conversion without flavor mixing, JCAP 03 (2016) 042 [1602.00698].DOI
- [21]H. Nagakura, T. Morinaga, C. Kato and S. Yamada, Fast-pairwise collective neutrino oscillations associated with asymmetric neutrino emissions in core-collapse supernova, 1910.04288.
- [22]T. Morinaga, H. Nagakura, C. Kato and S. Yamada, Fast neutrino-flavor conversion in the preshock region of core-collapse supernovae, Phys. Rev. Res. 2 (2020) 012046 [1909.13131].DOI
- [23]S. Abbar, H. Duan, K. Sumiyoshi, T. Takiwaki and M. C. Volpe, Fast Neutrino Flavor Conversion Modes in Multidimensional Core-collapse Supernova Models: the Role of the Asymmetric Neutrino Distributions, Phys. Rev. D 101 (2020) 043016 [1911.01983].DOI
- [24]R. Glas, H. T. Janka, F. Capozzi, M. Sen, B. Dasgupta, A. Mirizzi and G. Sigl, Fast Neutrino Flavor Instability in the Neutron-star Convection Layer of Three-dimensional Supernova Models, Phys. Rev. D 101 (2020) 063001 [1912.00274].DOI
- [25]H. Nagakura, L. Johns, A. Burrows and G. M. Fuller, Where, when, and why: Occurrence of fast-pairwise collective neutrino oscillation in three-dimensional core-collapse supernova models, Phys. Rev. D 104 (2021) 083025 [2108.07281].DOI
- [26]I. Padilla-Gay, I. Tamborra and G. G. Raffelt, Neutrino Flavor Pendulum Reloaded: The Case of Fast Pairwise Conversion, Phys. Rev. Lett. 128 (2022) 121102 [2109.14627].
- [27]M. Zaizen and H. Nagakura, Simple method for determining asymptotic states of fast neutrino-flavor conversion, Phys. Rev. D 107 (2023) 103022 [2211.09343].DOI
- [28]H. Nagakura and M. Zaizen, Connecting small-scale to large-scale structures of fast neutrino-flavor conversion, Phys. Rev. D 107 (2023) 063033 [2211.01398].
- [29]Z. Xiong, M.-R. Wu, M. George, C.-Y. Lin, N. K. Largani, T. Fischer and G. Martínez-Pinedo, Fast neutrino flavor conversions in a supernova: Emergence, evolution, and effects, Phys. Rev. D 109 (2024) 123008 [2402.19525].DOI
- [30]D. F. G. Fiorillo and G. G. Raffelt, Theory of neutrino slow flavor evolution. Part I. Homogeneous medium, JHEP 04 (2025) 146 [2412.02747].DOI
- [31]D. F. G. Fiorillo and G. G. Raffelt, Theory of neutrino slow flavor evolution. Part II. Space-time evolution of linear instabilities, JHEP 06 (2025) 146 [2501.16423].arxiv.org/abs/2501.16423
- [32]V. A. Kostelecky and S. Samuel, Self-maintained coherent oscillations in dense neutrino gases, Phys. Rev. D 52 (1995) 621 [hep-ph/9506262].DOI
- [33]S. Samuel, Bimodal coherence in dense selfinteracting neutrino gases, Phys. Rev. D 53 (1996) 5382 [hep-ph/9604341].
- [34]S. Pastor and G. Raffelt, Flavor oscillations in the supernova hot bubble region: Nonlinear effects of the neutrino background, Phys. Rev. Lett. 89 (2002) 191101 [astro-ph/0207281].DOI
- [35]H. Duan, G. M. Fuller and Y.-Z. Qian, Collective neutrino flavor transformation in supernovae, Phys. Rev. D 74 (2006) 123004 [astro-ph/0511275].DOI
- [36]H. Duan, G. M. Fuller, J. Carlson and Y.-Z. Qian, Simulation of Coherent Non-Linear Neutrino Flavor Transformation in the Supernova Environment. 1. Correlated Neutrino Trajectories, Phys. Rev. D 74 (2006) 105014 [astro-ph/0606616].
- [37]H. Duan, G. M. Fuller, J. Carlson and Y.-Z. Qian, Coherent Development of Flavor in the Supernova Environment, Phys. Rev. Lett. 97 (2006) 241101 [astro-ph/0608050].DOI
- [38]S. Hannestad, G. G. Raffelt, G. Sigl and Y. Y. Y. Wong, Self-induced conversion in dense neutrino gases: Pendulum in flavour space, Phys. Rev. D 74 (2006) 105010 [astro-ph/0608695]. [Erratum: Phys.Rev.D 76, 029901 (2007)].arxiv.org/abs/astro-ph/0608695
- [39]B. Dasgupta and A. Dighe, Collective three-flavor oscillations of supernova neutrinos, Phys. Rev. D 77 (2008) 103002 [0712.3798].
- [40]H. Duan, G. M. Fuller and Y.-Z. Qian, Stepwise spectral swapping with three neutrino flavors, Phys. Rev. D 77 (2008) 085016 [0801.1363].DOI
- [41]A. Friedland, Self-refraction of supernova neutrinos: mixed spectra and three-flavor instabilities, Phys. Rev. Lett. 104 (2010) 191102 [1001.0996].DOI
- [42]H. Duan and A. Friedland, Self-induced suppression of collective neutrino oscillations in a supernova, Phys. Rev. Lett. 106 (2011) 091101 [1006.2359].pubmed.ncbi.nlm.nih.gov/21405613
- [43]I. Padilla-Gay, H.-H. Chen, S. Abbar, M.-R. Wu and Z. Xiong, Flavor equilibration of supernova neutrinos: Exploring the dynamics of slow modes, Phys. Rev. D 112 (2025) 043039 [2505.11588].
- [44]D. F. G. Fiorillo and G. G. Raffelt, Theory of neutrino fast flavor evolution. Part I. Linear response theory and stability conditions, JHEP 08 (2024) 225 [2406.06708].arxiv.org/abs/2406.06708
- [45]D. F. G. Fiorillo and G. G. Raffelt, Theory of neutrino fast flavor evolution. Part II. Solutions at the edge of instability, JHEP 12 (2024) 205 [2409.17232].
- [46]N. F. Bell, A. A. Rawlinson and R. F. Sawyer, Speedup through entanglement: Many body effects in neutrino processes, Phys. Lett. B573 (2003) 86 [hep-ph/0304082].DOI
- [47]A. Friedland and C. Lunardini, Do many particle neutrino interactions cause a novel coherent effect?, JHEP 10 (2003) 043 [hep-ph/0307140].DOI
- [48]A. Friedland, B. H. J. McKellar and I. Okuniewicz, Construction and analysis of a simplified many-body neutrino model, Phys. Rev. D73 (2006) 093002 [hep-ph/0602016].DOI
- [49]A. Mezzacappa, E. Endeve, O. E. Bronson Messer and S. W. Bruenn, Physical, numerical, and computational challenges of modeling neutrino transport in core-collapse supernovae, Liv. Rev. Comput. Astrophys. 6 (2020) 4 [2010.09013].arxiv.org/abs/2010.09013
- [50]Y.-Z. Qian, G. M. Fuller, G. J. Mathews, R. Mayle, J. R. Wilson and S. E. Woosley, A connection between flavor mixing of cosmologically significant neutrinos and heavy element nucleosynthesis in supernovae, Phys. Rev. Lett. 71 (1993) 1965.pubmed.ncbi.nlm.nih.gov/10054549
- [51]J. T. Pantaleone, Neutrino flavor evolution near a supernova’s core, Phys. Lett. B 342 (1995) 250 [astro-ph/9405008].
- [52]H. Duan, A. Friedland, G. McLaughlin and R. Surman, The influence of collective neutrino oscillations on a supernova r-process, J. Phys. G 38 (2011) 035201 [1012.0532].DOI
- [53]C. J. Horowitz and G. Li, Nucleosynthesis in Supernovae, Phys. Rev. Lett. 82 (1999) 5198 [astro-ph/9904171].DOI
- [54]C. J. Horowitz and G. Li, Charge conjugation violating neutrino interactions in supernovae, Phys. Rev. D 61 (2000) 063002 [astro-ph/9908219].DOI
- [55]C. J. Horowitz, Weak magnetism for anti-neutrinos in supernovae, Phys. Rev. D 65 (2002) 043001 [astro-ph/0109209].
- [56]This subdominant but important effect boosts the νₑ cross sections on nucleons compared to those of ν̄ₑ.
- [57]M. Arnould, S. Goriely and K. Takahashi, The r-process of stellar nucleosynthesis: Astrophysics and nuclear physics achievements and mysteries, Phys. Rept. 450 (2007) 97 [0705.4512].DOI
- [58]D. Kasen, B. Metzger, J. Barnes, E. Quataert and E. Ramirez-Ruiz, Origin of the heavy elements in binary neutron-star mergers from a gravitational wave event, Nature 551 (2017) 80 [1710.05463].DOI
- [59]F.-K. Thielemann and J. J. Cowan, The r-process: history, required conditions, astrophysical sites, and observations, European Physical Journal A 62 (2026) 105 [2601.17246].
- [60]M. Arnould and S. Goriely, The p-process of stellar nucleosynthesis: astrophysics and nuclear physics status, Phys. Rept. 384 (2003) 1.DOI
- [61]T. Rauscher, N. Dauphas, I. Dillmann, C. Fröhlich, Z. Fülöp and G. Gyürky, Constraining the astrophysical origin of the p-nuclei through nuclear physics and meteoritic data, Reports on Progress in Physics 76 (2013) 066201 [1303.2666].
- [62]C. Frohlich, G. Martinez-Pinedo, M. Liebendorfer, F. K. Thielemann, E. Bravo, W. R. Hix, K. Langanke and N. T. Zinner, Neutrino-induced nucleosynthesis of a>64 nuclei: the νp-process, Phys. Rev. Lett. 96 (2006) 142502 [astro-ph/0511376].
- [63]J. Pruet, R. D. Hoffman, S. E. Woosley, H. T. Janka and R. Buras, Nucleosynthesis in early supernova winds. 2. the role of neutrinos, Astrophys. J. 644 (2006) 1028 [astro-ph/0511194].
- [64]S. Wanajo, The ν-process in neutrino-driven winds, Astrophys. J. 647 (2006) 1323 [astro-ph/0602488].
- [65]H. Schatz, L. Bildsten, A. Cumming and M. Wiescher, The Rapid Proton Process Ashes from Stable Nuclear Burning on an Accreting Neutron Star, Astrophys. J. 524 (1999) 1014 [astro-ph/9905274].DOI
- [66]J. Ehring, S. Abbar, H.-T. Janka, G. Raffelt and I. Tamborra, Fast Neutrino Flavor Conversions Can Help and Hinder Neutrino-Driven Explosions, Phys. Rev. Lett. 131 (2023) 061401 [2305.11207].DOI
- [67]J. Ehring, S. Abbar, H.-T. Janka, G. Raffelt and I. Tamborra, Fast neutrino flavor conversion in core-collapse supernovae: A parametric study in 1D models, Phys. Rev. D 107 (2023) 103034 [2301.11938].DOI
- [68]H. Nagakura, Roles of Fast Neutrino-Flavor Conversion on the Neutrino-Heating Mechanism of Core-Collapse Supernovae, Phys. Rev. Lett. 130 (2023) 211401 [2301.10785].DOI
- [69]T. Wang and A. Burrows, The Effect of the Fast-Flavor Instability on Core-Collapse Supernova Models, 2503.04896.
- [70]T. Wang and A. Burrows, The Effect of the Fast-flavor Instability on Core-Collapse Supernova Models. II. Quasi-equipartition and the Impact of Various Angular Reconstruction Methods, Astrophys. J. 997 (2026) 325 [2511.20767].arxiv.org/abs/2511.20767
- [71]K. Mori, T. Takiwaki, K. Kotake and S. Horiuchi, Three-dimensional core-collapse supernova models with phenomenological treatment of neutrino flavor conversions, Publ. Astron. Soc. Jpn. 77 (2025) L9 [2501.15256].arxiv.org/abs/2501.15256
- [72]M. Gogilashvili and I. Tamborra, Flavor Conversion Enhances or Suppresses Supernova Explodability Independent of the Progenitor Mass, 2605.18972.DOI
- [73]M. Gogilashvili and I. Tamborra, Neutrino Flavor Conversion Shapes the Rate of Failed Core-collapse Supernovae, 2605.16504.DOI
- [74]R. Akaho, H. Nagakura, W. Iwakami, S. Furusawa, A. Harada, H. Okawa, H. Matsufuru, K. Sumiyoshi and S. Yamada, Bifurcated Impact of Neutrino Fast Flavor Conversion on Core-Collapse Supernovae Informed by Multiangle Neutrino Radiation Hydrodynamics, Phys. Rev. Lett. 136 (2026) 191002 [2601.08269].DOI
- [75]Z. Xiong, A. Sieverding, M. Sen and Y.-Z. Qian, Potential Impact of Fast Flavor Oscillations on Neutrino-driven Winds and Their Nucleosynthesis, Astrophys. J. 900 (2020) 144 [2006.11414].DOI
- [76]G. Martinez-Pinedo, B. Ziebarth, T. Fischer and K. Langanke, Effect of collective neutrino flavor oscillations on νp-process nucleosynthesis, Eur. Phys. J. A 47 (2011) 98 [1105.5304].
- [77]H. Sasaki, T. Kajino, T. Takiwaki, T. Hayakawa, A. B. Balantekin and Y. Pehlivan, Possible effects of collective neutrino oscillations in three-flavor multangle simulations of supernova νp processes, Phys. Rev. D 96 (2017) 043013 [1707.09111].
- [78]A. B. Balantekin, M. J. Cervia, A. V. Patwardhan, R. Surman and X. Wang, Collective Neutrino Oscillations and Heavy-element Nucleosynthesis in Supernovae: Exploring Potential Effects of Many-body Neutrino Correlations, Astrophys. J. 967 (2024) 146DOI
- [79]S. Wanajo, H.-T. Janka and S. Kubono, Uncertainties in the ν p-process: supernova dynamics versus nuclear physics, Astrophys. J. 729 (2011) 46 [1004.4487].DOI
- [80]A. Friedland, P. Mukhopadhyay and A. V. Patwardhan, Successful νp-process in neutrino-driven outflows in core-collapse supernovae, JCAP 02 (2025) 005 [2312.03208].DOI
- [81]A. Friedland, D. J. Li, G. Lucente, I. Padilla-Gay and A. V. Patwardhan, νp-process in core-collapse supernovae: imprint of general relativistic effects, JCAP 02 (2026) 067 [2508.02055].arxiv.org/abs/2508.02055
- [82]A. Friedland and P. Mukhopadhyay, Near-critical supernova outflows and their neutrino signatures, Phys. Lett. B 834 (2022) 137403 [2009.10059].DOI
- [83]Garching core-collapse supernova research archive, https://wwwmpa.mpa-garching.mpg.de/ccsnarchive/.
- [84]D. F. G. Fiorillo, M. Heinlein, H.-T. Janka, G. Raffelt, E. Vitagliano and R. Bollig, Supernova simulations confront SN 1987A neutrinos, Phys. Rev. D 108 (2023) 083040 [2308.01403].DOI
- [85]G. Lucente, M. Heinlein, H.-T. Janka and A. Mirizzi, Simple fits for the neutrino luminosities from protoneutron star cooling, Phys. Rev. D 110 (2024) 063023 [2405.00769].DOI
- [86]T. Sukhbold, T. Ertl, S. E. Woosley, J. M. Brown and H. T. Janka, Core-Collapse Supernovae from 9 to 120 Solar Masses Based on Neutrino-powered Explosions, Astrophys. J. 821 (2016) 38 [1510.04643].DOI
- [87]H.-T. Janka and W. Hillebrandt, Neutrino emission from type II supernovae : an analysis of the spectra., Astron. Astrophys. 224 (1989) 49.
- [88]H.-T. Janka and W. Hillebrandt, Monte Carlo simulations of neutrino transport in type II supernovae., Astron. Astrophys. Suppl. Ser. 78 (1989) 375.
- [89]M. T. Keil, G. G. Raffelt and H.-T. Janka, Monte Carlo study of supernova neutrino spectra formation, Astrophys. J. 590 (2003) 971 [astro-ph/0208035].DOI
- [90]This prescription should be viewed as a representative, approximate limiting case, since flavor equilibration is constrained by lepton-flavor number conservation [110, 111]. However, the resulting asymptotic value of Yₑ is likely to vary within a narrow range around ∼ 0.6. A further quantitative investigation will follow in future work [101].
- [91]J. Lippuner and L. F. Roberts, SkyNet: A modular nuclear reaction network library, Astrophys. J. Suppl. 233 (2017) 18 [1706.06198].DOI
- [92]‘SkyNet’ source code available at: https://bitbucket.org/jlippuner/skynet/src/master/.
- [93]M. Beard, S. M. Austin and R. Cyburt, Enhancement of the Triple Alpha Rate in a Hot Dense Medium, Phys. Rev. Lett. 119 (2017) 112701 [1708.07204].DOI
- [94]S. Jin, L. F. Roberts, S. M. Austin and H. Schatz, Enhanced triple-α reaction reduces proton-rich nucleosynthesis in supernovae, Nature 588 (2020) 57.DOI
- [99]Studies prior to Refs. [80, 81] used insufficient 92Nb production to argue against νp-process viability [61, 112, 113].
- [100]K. Lodders, Solar System Abundances and Condensation Temperatures of the Elements, ApJ 591 (2003) 1220.DOI
- [101]A. Friedland, D. Li, G. Lucente, P. Mukhopadhyay, I. Padilla-Gay and A. Patwardhan, in preparation.
- [102]Oscillations at these distances might arise from the Mikheyev-Smirnov-Wolfenstein (MSW) effect, a manifestation of the coherent forward scattering of neutrinos with the electrons in the medium [114, 115]. This becomes only relevant when the outflow reaches densities of about 10^3 g/cm3 at distances of thousands of kilometers from the PNS. These densities are reached, however, at late times t ≳ 6 s [82] i.e. after the optimal window of the νp-process.
- [103]H.-T. Janka, Long-Term Multidimensional Models of Core-Collapse Supernovae: Progress and Challenges, Ann. Rev. Nucl. Part. Sci. 75 (2025) 425 [2502.14836].arxiv.org/abs/2502.14836
- [104]T. Fischer, G. Martinez-Pinedo, M. Hempel and M. Liebendorfer, Neutrino spectra evolution during proto-neutron star deleptonization, Phys. Rev. D 85 (2012) 083003 [1112.3842].DOI
- [105]K. Nakazato, K. Sumiyoshi, H. Suzuki, T. Totani, H. Umeda and S. Yamada, Supernova Neutrino Light Curves and Spectra for Progenitor Stars: From Core Collapse to Proto-neutron Star Cooling, Astrophys. J. Suppl. 205 (2013) 2 [1210.6841].DOI
- [106]T. Fischer, G. Guo, K. Langanke, G. Martinez-Pinedo, Y.-Z. Qian and M.-R. Wu, Neutrinos and nucleosynthesis of elements, Prog. Part. Nucl. Phys. 137 (2024) 104107 [2308.03962].DOI
- [107]L. Choi, A. Burrows and D. Vartanyan, Predicted neutrino signal features of core-collapse supernovae, Phys. Rev. D 111 (2025) 123038 [2503.07531].
- [108]B. Müller, H.-T. Janka and A. Marek, A New Multi-dimensional General Relativistic Neutrino Hydrodynamics Code for Core-collapse Supernovae. II. Relativistic Explosion Models of Core-collapse Supernovae, Astrophys. J. 756 (2012) 84 [1202.0815].DOI
- [109]R. Buras, M. Rampp, H. T. Janka and K. Kifonidis, Two-dimensional hydrodynamic core-collapse supernova simulations with spectral neutrino transport. 1. Numerical method and results for a 15 solar mass star, Astron. Astrophys. 447 (2006) 1049 [astro-ph/0507135].
- [116]Y. Z. Qian and S. E. Woosley, Nucleosynthesis in neutrino driven winds: 1. The Physical conditions, Astrophys. J. 471 (1996) 331 [astro-ph/9611094].
- [117]K. Otsuki, H. Tagoshi, K. Kajino and S.-y. Wanajo, General relativistic effects on neutrino driven wind from young, hot neutron star and the r process nucleosynthesis, Astrophys. J. 533 (2000) 424 [astro-ph/9911164].DOI
- [118]R. H. Cyburt, A. M. Amthor, R. Ferguson, Z. Meisel, K. Smith, S. Warren, A. Heger, R. D. Hoffman, T. Rauscher, A. Sakharuk, H. Schatz, F. K. Thielemann and M. Wiescher, The JINA REACLIB Database: Its Recent Updates and Impact on Type-I X-ray Bursts, Astrophys. J. Supplement Series 189 (2010) 240.DOI
- [119]A. Burrows and T. A. Thompson, Neutrino - matter interaction rates in supernovae: The Essential microphysics of core collapse, astro-ph/0211404.DOI
- [120]M. Rampp and H.-T. Janka, Radiation hydrodynamics with neutrinos: Variable Eddington factor method for core collapse supernova simulations, Astron. Astrophys. 396 (2002) 361 [astro-ph/0203101].DOI
- [121]H.-T. Janka, Explosion Mechanisms of Core-Collapse Supernovae, Ann. Rev. Nucl. Part. Sci. 62 (2012) 407 [1206.2503].DOI
- [122]R. Bollig, H. T. Janka, A. Lohs, G. Martinez-Pinedo, C. J. W. and T. Melson, Muon Creation in Supernova Matter Facilitates Neutrino-driven Explosions, Phys. Rev. Lett. 119 (2017) 242702 [1706.04630].DOI
- [123]R. Bollig, N. Yadav, D. Kresse, H. T. Janka, B. Müller and A. Heger, Self-consistent 3D Supernova Models From −7 Minutes to +7 s: A 1-bethe Explosion of a ∼19 M⊙ Progenitor, Astrophys. J. 915 (2021) 28 [2010.10506].DOI
- [124]A. W. Steiner, M. Hempel and T. Fischer, Core-collapse Supernova Equations of State Based on Neutron Star Observations, Astrophys. J. 774 (2013) 17 [1207.2184].DOI
- [125]M. Hempel and J. Schaffner-Bielich, A statistical model for a complete supernova equation of state, Nuclear Physics A 837 (2010) 210 [0911.4073].DOI
- [126]L. F. Roberts, G. Shen, V. Cirigliano, J. A. Pons, S. Reddy and S. E. Woosley, Proto-Neutron Star Cooling with Convection: The Effect of the Symmetry Energy, Phys. Rev. Lett. 108 (2012) 061103 [1112.0335].DOI
- [127]A. Pascal, J. Novak and M. Oertel, Proto-neutron star evolution with improved charged-current neutrino–nucleon interactions, Mon. Not. Roy. Astron. Soc. 511 (2022) 356 [2201.01955].DOI
- [128]J. Bliss, A. Arcones and Y.-Z. Qian, Production of Mo and Ru isotopes in neutrino-driven winds: implications for solar abundances and presolar grains, Astrophys. J. 866 (2018) 105 [1804.03947].
- [129]S. E. Woosley, J. R. Wilson, G. J. Mathews, R. D. Hoffman and B. S. Meyer, The r process and neutrino heated supernova ejecta, Astrophys. J. 433 (1994) 229.DOI
- [130]T. Iizuka, Y.-J. Lai, W. Akram, Y. Amelin and M. Schönbächler, The initial abundance and distribution of 92Nb in the Solar System, Earth and Planetary Science Letters 439 (2016) 172 [1602.00986].
- [131]Y. Hibiya, T. Iizuka, H. Enomoto and T. Hayakawa, Evidence for Enrichment of Niobium-92 in the Outer Protosolar Disk, Astrophys. J. Lett. 942 (2023) L15.DOI









