Impacts of Black-Hole-Forming Supernova Explosions on the Diffuse Neutrino Background

Ken’ichiro Nakazato Faculty of Arts and Science, Kyushu University, Fukuoka 819-0395, Japan Ryuichiro Akaho Faculty of Science and Engineering, Waseda University, Tokyo 169-8555, Japan Yosuke Ashida Department of Physics and Astronomy, University of Utah, Salt Lake City, UT 84112, USA Takuji Tsujimoto National Astronomical Observatory of Japan, Mitaka, Tokyo 181-8588, Japan
Abstract

Flux spectrum, event rate, and experimental sensitivity are investigated for the diffuse supernova neutrino background (DSNB), which is originated from past stellar collapses and also known as supernova relic neutrino background. For this purpose, the contribution of collapses that lead to successful supernova (SN) explosion and black hole (BH) formation simultaneously, which are suggested to be a non-negligible population from the perspective of Galactic chemical evolution, is taken into account. If the BH-forming SNe involve the matter fallback onto the protoneutron star for the long term, their total emitted neutrino energy becomes much larger than that of ordinary SNe and failed SNe (BH formation without explosion). The expected event rate according to the current DSNB model is enhanced by up to a factor of two due to the BH-forming SNe, depending on their fraction and the neutrino mass hierarchy. In any case, the operation time required to detect the DSNB at Hyper-Kamiokande would be reduced by such contribution.

Neutrino astronomy (1100) — Supernova neutrinos (1666) — Core-collapse supernovae (304) — Massive stars (732) — Neutron stars (1108) — Black holes (162) — Galaxy chemical evolution (580)

1 Introduction

Supernova (SN) explosions supply the elements synthesized inside stars, serving as the building blocks for the next generation of stars, planets, and life. Understanding the properties of progenitors and explosion mechanisms of SNe is crucial for unraveling the history of the Universe. In particular for the study of core-collapse SNe, neutrinos are expected to be a powerful tool. Stars with masses larger than similar-to\sim8Msubscript𝑀direct-productM_{\odot}italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT experience core collapse at the end of their evolution, leading to the emission of a huge amount of neutrinos. In fact, neutrinos from SN1987A, which is a core-collapse SN appeared in the Large Magellanic Cloud, were successfully detected (Hirata et al., 1987; Bionta et al., 1987; Alekseev et al., 1987). In case that a next Galactic SN occurs, currently operating neutrino detectors, such as Super-Kamiokande (SK), are expected to detect statistically sufficient number of neutrino events (e.g., Abbasi et al., 2011; Abe et al., 2016; Kashiwagi et al., 2024) and various insights into SN explosions will be derived from the neutrino observations (Abe et al., 2021a; Suwa et al., 2022, 2024; Nagakura & Vartanyan, 2022; Harada et al., 2023a).

Another approach to detecting neutrinos from SNe is to concentrate on the cosmic background. Neutrinos from distant SNe have been accumulated to form the diffuse SN neutrino background (DSNB), or SN relic neutrino (SRN) background. The upper bounds on the ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT flux have been provided by detectors such as SK (Bays et al., 2012; Zhang et al., 2015; Abe et al., 2021b), KamLAND (Gando et al., 2012; Abe et al., 2022), and SK-Gd (Harada et al., 2023b), which is the SK experiment with gadolinium-loaded water (Beacom & Vagins, 2004). Furthermore, a larger-volume water Cherenkov detector, Hyper-Kamiokande (HK), is currently under construction (Abe et al., 2018), and other types of detectors such as liquid scintillators and argon/xenon-based detectors are planned. Predictions of the DSNB detection are exhibited for these future detectors including JUNO and DUNE (Priya & Lunardini, 2017; MØller et al., 2018; Sawatzki et al., 2021; Li et al., 2022; Suliga et al., 2022).

Theoretical estimations of the DSNB flux and predictions of the event rates have been investigated for a long time (Bisnovatyi-Kogan & Seidov, 1982; Krauss et al., 1984; Dar, 1985). Since models of DSNB involve a wide range of physical and astrophysical factors, improvements to the model have been made in many aspects. The spectrum of neutrinos emitted from a core-collapse SN depends on the progenitor, particularly its initial mass and metallicity. For estimating the DSNB flux, the average spectrum weighted by the initial mass function (IMF) is often used (Totani & Sato, 1995), and a large sample of progenitor models is currently under consideration (Horiuchi et al., 2018; Kresse et al., 2021). The variation of IMF is also being discussed (Ziegler et al., 2022; Aoyama et al., 2023; Ashida et al., 2023). The evolution of progenitors may be affected by binary interactions (Horiuchi et al., 2021). The cosmic SN rate, or star formation history, is deduced from astronomical observations (Totani et al., 1996; Hartmann & Woosley, 1997; Malaney, 1997; Kaplinghat et al., 2000; Horiuchi et al., 2009; Mathews et al., 2014) and the cosmic chemical evolution contributes to the metallicity distribution of progenitors (Nakazato et al., 2015). The emission of neutrinos from SNe itself involves uncertainty, particularly in the late phase (Nakazato, 2013; Ashida & Nakazato, 2022; Ekanger et al., 2022, 2024). Flavor mixing caused by neutrino oscillations is a factor that influences the event rates of DSNB (Ando et al., 2003; Galais et al., 2010), and the effects of exotic physics in the neutrino sector are also being investigated (Ando, 2003; Fogli et al., 2004; de Gouvêa et al., 2020, 2022; Tabrizi & Horiuchi, 2021; Iváñez-Ballesteros & Volpe, 2023). Furthermore, the DSNB flux may be related to other cosmic background radiation such as MeV gamma rays (Strigari et al., 2005; Anandagoda et al., 2023) and non-thermal high-energy neutrinos (Ashida, 2024). Incidentally, neutrinos emitted from accretion disks formed around SNe may contribute to the cosmic background radiation (Schilbach et al., 2019; Wei et al., 2024). The basics of DSNB are covered in several previous reviews. (Ando & Sato, 2004; Beacom, 2010; Lunardini, 2016; Mathews et al., 2020; Ando et al., 2023).

In the present paper, we focus on the stellar core collapses which lead to black hole (BH) formation. Neutrinos are emitted from the BH-forming collapse, as well as the ordinary core-collapse SNe, and these neutrinos are in the same energy regime as the DSNB (Iocco et al., 2005; Nakazato et al., 2006; Lunardini, 2009). In previous studies on the DSNB, two scenarios for the core collapse of massive stars are considered: those resulting in an ordinary SN explosion, leaving a neutron star (NS), and those leading to BH formation without an explosion, which are known as failed SNe. This limitation on the scenario of core collapse makes sense because, according to numerical studies, successful explosions that make BHs are predicted as a relatively rare event (see Fig. 13 of Sukhbold et al., 2016). Recently, however, the examples for this case are investigated in detail (e.g., Burrows et al., 2023). Furthermore, some astronomical phenomena such as GRB 980425/SN1998bw (Iwamoto et al., 1998) and W50/SS 433 (Poutanen et al., 2007) are likely to be associated with core-collapse SNe that result in the formation of BHs. Hereafter, we refer to this subset as BH-forming SNe.

The perspective of nucleosynthesis implied from chemical evolution at the early Galaxy has significantly raised the importance of contribution from BH-forming SNe. This suggests two types of BH-forming SNe, each of which is implied to be a non-negligible population (e.g., Nomoto et al., 2006). First, some studies propose that core-collapse SNe with large explosion energy (greater-than-or-equivalent-to\gtrsim1052 erg), which are referred to as hypernovae, should promote chemical enrichment with its contributed fraction as much as 50% of massive (>>> 20 Msubscript𝑀direct-productM_{\odot}italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT) stars to account for the observed abundance of some heavy elements including Zn among Galactic halo stars (Kobayashi et al., 2006, 2020). These hypernovae are generally considered to be BH-forming SNe that generate a jet-like explosion powered by energy from the rotating BH. In addition, hypernovae could play a key role as the dominant site of νp𝜈𝑝\nu pitalic_ν italic_p-process nucleosynthesis that can account for a large portion of abundances of Mo and Ru for low-metallicity stars (Fujibayashi et al., 2015; Sasaki et al., 2022).

Another candidate of BH-forming SNe is the so-called faint SNe whose luminosity is very low due to a negligibly small ejection of synthesized 56Ni (e.g., Heger et al., 2003). Faint SNe have been highlighted as the origin of a subset of carbon-enhanced metal-poor (CEMP) stars in Galactic halo (Nomoto et al., 2013). An event frequency of faint SNe could be approximately inferred from the fraction of this kind of stars against halo stars, which is estimated to be 20% among stars with [Fe/H] 2absent2\leq-2≤ - 2 (Placco et al., 2014). In addition to these arguments, there is a recent report that the inclusion of contribution from faint SNe with tens of percent leads to better agreement with the observed abundance of stars in the solar neighborhood (Pignatari et al., 2023). In the end, BH-forming SNe that possibly emerge as hypernovae or faint SNe, are suggested to be a non-minor population that could be, as not an unrealistic case, counted up to a half of all core-collapse SNe.

Although the detailed explosion mechanisms and formation processes of BHs are not identified for these BH-forming SNe, they should accompany the emission of a huge amount of neutrinos as well as the ordinary core-collapse SNe and failed SNe. In this study, we investigate the impact of BH-forming SNe on the DSNB flux updating our previous study (Ashida et al., 2023). This paper is organized as follows. In § 2, we describe the spectral model of neutrinos emitted from the BH-forming SNe. While the dynamics of BH-forming SNe includes uncertainties, we focus on the case induced by fallback mass accretion with a long duration. The formulation of the DSNB flux is presented in § 3. Issues concerning rates of core-collapse SNe and failed SNe based on the model of galactic chemical evolution are also provided. In § 4, we investigate the DSNB event rate of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT. Furthermore, the experimental sensitivities at HK are evaluated. Finally, a summary and discussion are provided in § 5.

2 Neutrino emission from fallback induced BH formation

In this section, we consider the neutrino emission from the BH-forming supernovae. Unfortunately, their dynamics is still not well understood. In particular, the time interval from the core bounce to BH formation, which corresponds to the duration of the neutrino emission, has notable impacts; generally the total energy of emitted neutrinos gets larger for a longer duration (Kresse et al., 2021). To address this, we assume two extreme cases: one where a BH is formed dynamically on short timescales of O(1)𝑂1O(1)italic_O ( 1 ) s, and another where the moderate fallback causes the mass of the protoneutron star (PNS) to exceed the maximum mass, resulting in the formation of a BH at later stages on timescales of >O(10)absent𝑂10>\!O(10)> italic_O ( 10 ) s. Since the former case is similar to that of the failed SNe, we adopt the same spectral model of the emitted neutrinos as the failed SNe for the prompt BH-formation case. On the other hand, as for the case of fallback induced BH formation, we construct a neutrino spectrum in the present paper. In the following, we describe the model of emitted neutrinos from fallback induced BH formation.

We combine the model of stellar core collapse in Nakazato et al. (2021) and the model of fallback mass accretion in Akaho et al. (2024) for the evaluation of the neutrino spectrum from the fallback induced BH formation. In Akaho et al. (2024), the neutrino luminosity emitted by fallback mass accretion onto a PNS with the gravitational mass of 1.98M1.98subscript𝑀direct-product1.98M_{\odot}1.98 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT, whose baryon mass corresponds to 2.35M2.35subscript𝑀direct-product2.35M_{\odot}2.35 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT, is provided.

So as to estimate the neutrino spectrum emitted during the early dynamical phase, we utilize the core-collapse simulation of a 30M30subscript𝑀direct-product30M_{\odot}30 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT progenitor in Nakazato et al. (2021). We integrate the neutrino emission of this model up to the point where a baryon mass of the PNS reaches 2.35M2.35subscript𝑀direct-product2.35M_{\odot}2.35 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT. While three models with different nuclear equation of state (EOS) are presented in Nakazato et al. (2021), we adopt the model with the Togashi EOS (Togashi et al., 2017) in this study. Note that, the Furusawa-Togashi EOS (Furusawa et al., 2017) is used in the fallback model of Akaho et al. (2024) and it is different from the Togashi EOS in the low density regime. Nevertheless, the impact of their difference is minor on the neutrino emission during the early dynamical phase (Sumiyoshi et al., 2023).

In the fallback mass accretion model of Akaho et al. (2024), the inside PNS is assumed to have an isotropic temperature of T=2𝑇2T=2italic_T = 2 MeV. In contrast, the temperature of the PNS is TO(10)similar-to𝑇𝑂10T\sim O(10)italic_T ∼ italic_O ( 10 ) MeV when the baryon mass reaches 2.35M2.35subscript𝑀direct-product2.35M_{\odot}2.35 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT in Nakazato et al. (2021). Therefore, in order to bridge the temperature gap, we perform the cooling simulation of the PNS and assess the neutrino emission. For this purpose, we adopt the structure of PNS obtained by the core-collapse simulation in Nakazato et al. (2021) as the initial condition. As for the EOS, Furusawa-Togashi EOS is adopted. Our numerical methods for the PNS cooling are similar to those employed in Suzuki (1994) and Nakazato et al. (2013). As a result of the simulation, we find that the PNS cools to T2similar-to𝑇2T\sim 2italic_T ∼ 2 MeV over a period of 200 s. Since the effect of convection is not taken into account in this evaluation, the cooling time might be shorter. Nevertheless, since the total energy of neutrinos emitted from the PNS cooling stems from the binding energy of PNS, the time-integrated spectrum is insensitive to the cooling time.

Figure 1: Time integrated spectra of (a) νesubscript𝜈𝑒\nu_{e}italic_ν start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT, (b) ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT, and (c) νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT, where νx=νμ=ν¯μ=ντ=ν¯τsubscript𝜈𝑥subscript𝜈𝜇subscript¯𝜈𝜇subscript𝜈𝜏subscript¯𝜈𝜏\nu_{x}=\nu_{\mu}=\bar{\nu}_{\mu}=\nu_{\tau}=\bar{\nu}_{\tau}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT = italic_ν start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT = over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT = italic_ν start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT = over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT, for the fallback induced BH formation model. Orange, blue, and green histograms correspond to the components of the early dynamical phase, the PNS cooling, and the fallback mass accretion, respectively.
Table 1: Properties of emitted neutrinos for the models considered in this study.
Eνedelimited-⟨⟩subscript𝐸subscript𝜈𝑒\langle E_{\nu_{e}}\rangle⟨ italic_E start_POSTSUBSCRIPT italic_ν start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT end_POSTSUBSCRIPT ⟩ Eν¯edelimited-⟨⟩subscript𝐸subscript¯𝜈𝑒\langle E_{\bar{\nu}_{e}}\rangle⟨ italic_E start_POSTSUBSCRIPT over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT end_POSTSUBSCRIPT ⟩ Eνxdelimited-⟨⟩subscript𝐸subscript𝜈𝑥\langle E_{\nu_{x}}\rangle⟨ italic_E start_POSTSUBSCRIPT italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT end_POSTSUBSCRIPT ⟩ Eνe,totsubscript𝐸subscript𝜈𝑒totE_{\nu_{e},{\rm tot}}italic_E start_POSTSUBSCRIPT italic_ν start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT , roman_tot end_POSTSUBSCRIPT Eν¯e,totsubscript𝐸subscript¯𝜈𝑒totE_{\bar{\nu}_{e},{\rm tot}}italic_E start_POSTSUBSCRIPT over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT , roman_tot end_POSTSUBSCRIPT Eνx,totsubscript𝐸subscript𝜈𝑥totE_{\nu_{x},{\rm tot}}italic_E start_POSTSUBSCRIPT italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT , roman_tot end_POSTSUBSCRIPT
Model explosion remnant (MeV) (1052superscript105210^{52}10 start_POSTSUPERSCRIPT 52 end_POSTSUPERSCRIPT erg)
Ordinary core-collapse SN successful NS 9.2 10.9 11.8 4.47 4.07 4.37
BH-forming SN, (i) fallback induced successful BH 11.8 13.6 10.9 19.48 18.50 12.07
BH-forming SN, (ii) prompt successful BH 16.1 20.4 23.4 6.85 5.33 2.89
Failed SN failed BH 16.1 20.4 23.4 6.85 5.33 2.89

Note. — Eνidelimited-⟨⟩subscript𝐸subscript𝜈𝑖\langle E_{\nu_{i}}\rangle⟨ italic_E start_POSTSUBSCRIPT italic_ν start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT end_POSTSUBSCRIPT ⟩ and Eνi,totsubscript𝐸subscript𝜈𝑖totE_{\nu_{i},{\rm tot}}italic_E start_POSTSUBSCRIPT italic_ν start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT , roman_tot end_POSTSUBSCRIPT are the average and total energies of the time-integrated neutrino signal for νisubscript𝜈𝑖\nu_{i}italic_ν start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT, where νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT represents the average of νμsubscript𝜈𝜇\nu_{\mu}italic_ν start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT, ν¯μsubscript¯𝜈𝜇\bar{\nu}_{\mu}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT, ντsubscript𝜈𝜏\nu_{\tau}italic_ν start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT, and ν¯τsubscript¯𝜈𝜏\bar{\nu}_{\tau}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT.

We also incorporate the neutrinos emitted from the fallback mass accretion based on Akaho et al. (2024). Here, we consider the fallback mass accretion of 0.002Ms10.002subscript𝑀direct-productsuperscripts10.002M_{\odot}\,{\rm s}^{-1}0.002 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT roman_s start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT onto a PNS with the gravitational mass of 1.98M1.98subscript𝑀direct-product1.98M_{\odot}1.98 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT. Since, based on their adopted EOS, the maximum baryon mass of the NSs is 2.70M2.70subscript𝑀direct-product2.70M_{\odot}2.70 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT and baryon mass of our PNS is 2.35M2.35subscript𝑀direct-product2.35M_{\odot}2.35 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT, it takes 175 s to form a BH by the accretion of 0.35M0.35subscript𝑀direct-product0.35M_{\odot}0.35 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT. Thus, the total amount of neutrinos emitted from the fallback mass accretion would be evaluated by integrating the emission rate of this model over 175 s. However, it is noted that, since the neutrino luminosity shown in Akaho et al. (2024) includes the emission from the PNS in quasi-steady state, we should subtract its contribution to avoid duplication with the contribution of PNS cooling evaluated above. As shown in Fig. 7 of Akaho et al. (2024), the neutrino luminosity and the mass accretion rate exhibit a linear relationship with an offset. This offset can be evaluated by extrapolation and regarded as the emission from the PNS. Therefore, we subtract the offset spectrum to incorporate the net contribution of the fallback mass accretion. Owing to the subtraction, the total amount of neutrinos emitted from the fallback mass accretion is insensitive to the choice of mass accretion rate, provided that the accreted mass is fixed to 0.35M0.35subscript𝑀direct-product0.35M_{\odot}0.35 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT.

In Fig. 1, we show the neutrino spectra of individual components: the early dynamical phase, the PNS cooling, and the fallback mass accretion. A substantial amount of νesubscript𝜈𝑒\nu_{e}italic_ν start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT and ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT are emitted from the fallback mass accretion. These neutrinos have high average energies and they are mainly produced by electron capture and positron capture. Fallback mass accretion produces the high-temperature environment where thermal electrons and positrons are created and supplies enormous free protons and neutrons which capture electrons and positrons, respectively. In contrast, the amount of νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT (=νμ=ν¯μ=ντ=ν¯τabsentsubscript𝜈𝜇subscript¯𝜈𝜇subscript𝜈𝜏subscript¯𝜈𝜏=\nu_{\mu}=\bar{\nu}_{\mu}=\nu_{\tau}=\bar{\nu}_{\tau}= italic_ν start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT = over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT = italic_ν start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT = over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_τ end_POSTSUBSCRIPT) emitted from the fallback mass accretion is much smaller than those of νesubscript𝜈𝑒\nu_{e}italic_ν start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT and ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT. According to Akaho et al. (2024), the primary process for emitting νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT is nucleon-nucleon bremsstrahlung, which is efficient in the high density regions inside the PNS, and the luminosity of νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT is hardly dependent on the accretion rate. Therefore, due to the subtraction considered in this study, the contribution of fallback mass accretion is minor for νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT.

In Table 1, the average and total energies of neutrinos emitted from the fallback induced BH-forming SN are compared with other scenarios taken into account in this study. The total emission energy of neutrinos, summing all flavors, amounts to 8.63×10538.63superscript10538.63\times 10^{53}8.63 × 10 start_POSTSUPERSCRIPT 53 end_POSTSUPERSCRIPT erg. Since the binding energy of the maximum-mass NS is 8.78×10538.78superscript10538.78\times 10^{53}8.78 × 10 start_POSTSUPERSCRIPT 53 end_POSTSUPERSCRIPT erg for the EOS adopted here, our fallback induced BH formation can be interpreted as maximizing the emission energy of neutrinos from a single stellar collapse111Except for the collapse of supermassive stars with greater-than-or-equivalent-to\gtrsim10M3superscriptsubscript𝑀direct-product3{}^{3}M_{\odot}start_FLOATSUPERSCRIPT 3 end_FLOATSUPERSCRIPT italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT (e.g., Nakazato et al., 2006).. On the other hand, as already mentioned, we assume that the neutrino signal of the prompt BH-forming SN is same as the failed SN (BH formation without explosion). The total emission energy of neutrinos, summing all flavors, is 2.37×10532.37superscript10532.37\times 10^{53}2.37 × 10 start_POSTSUPERSCRIPT 53 end_POSTSUPERSCRIPT erg, which is 28% of the fallback induced BH formation model. Then, we consider that the uncertainty in the neutrino emission from BH-forming SNe is accounted for both extremes.

3 DSNB flux Model

Following Ashida et al. (2023), we describe the DSNB flux as

dΦ(Eν)dEν=c0zmaxdzH0Ωm(1+z)3+ΩΛ×\displaystyle\frac{d\Phi(E_{\nu})}{dE_{\nu}}=c\int^{z_{\rm max}}_{0}\frac{dz}{% H_{0}\sqrt{\Omega_{\rm m}(1+z)^{3}+\Omega_{\Lambda}}}\timesdivide start_ARG italic_d roman_Φ ( italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) end_ARG start_ARG italic_d italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT end_ARG = italic_c ∫ start_POSTSUPERSCRIPT italic_z start_POSTSUBSCRIPT roman_max end_POSTSUBSCRIPT end_POSTSUPERSCRIPT start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT divide start_ARG italic_d italic_z end_ARG start_ARG italic_H start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT square-root start_ARG roman_Ω start_POSTSUBSCRIPT roman_m end_POSTSUBSCRIPT ( 1 + italic_z ) start_POSTSUPERSCRIPT 3 end_POSTSUPERSCRIPT + roman_Ω start_POSTSUBSCRIPT roman_Λ end_POSTSUBSCRIPT end_ARG end_ARG ×
[RSN(z){(1fBHSN)dNCCSN(Eν)dEν+\displaystyle\left[R_{\rm SN}(z)\left\{(1-f_{\rm BHSN})\frac{dN_{\rm CCSN}(E^{% \prime}_{\nu})}{dE^{\prime}_{\nu}}+\right.\right.[ italic_R start_POSTSUBSCRIPT roman_SN end_POSTSUBSCRIPT ( italic_z ) { ( 1 - italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT ) divide start_ARG italic_d italic_N start_POSTSUBSCRIPT roman_CCSN end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) end_ARG start_ARG italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT end_ARG +
fBHSNdNBHSN(Eν)dEν}+RBH(z)dNBH(Eν)dEν],\displaystyle\left.\left.f_{\rm BHSN}\frac{dN_{\rm BHSN}(E^{\prime}_{\nu})}{dE% ^{\prime}_{\nu}}\right\}+R_{\rm BH}(z)\frac{dN_{\rm BH}(E^{\prime}_{\nu})}{dE^% {\prime}_{\nu}}\right],italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT divide start_ARG italic_d italic_N start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) end_ARG start_ARG italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT end_ARG } + italic_R start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_z ) divide start_ARG italic_d italic_N start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) end_ARG start_ARG italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT end_ARG ] , (1)

where c𝑐citalic_c is the speed of light and the cosmological constants are Ωm=0.3089subscriptΩm0.3089\Omega_{\rm m}=0.3089roman_Ω start_POSTSUBSCRIPT roman_m end_POSTSUBSCRIPT = 0.3089, ΩΛ=0.6911subscriptΩΛ0.6911\Omega_{\Lambda}=0.6911roman_Ω start_POSTSUBSCRIPT roman_Λ end_POSTSUBSCRIPT = 0.6911, and H0=67.74subscript𝐻067.74H_{0}=67.74italic_H start_POSTSUBSCRIPT 0 end_POSTSUBSCRIPT = 67.74 km sec-1 Mpc-1. The neutrino energy at a detector, Eνsubscript𝐸𝜈E_{\nu}italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT, and a source, Eνsubscriptsuperscript𝐸𝜈E^{\prime}_{\nu}italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT, are related to the redshift of the source, z𝑧zitalic_z, as Eν=(1+z)Eνsubscriptsuperscript𝐸𝜈1𝑧subscript𝐸𝜈E^{\prime}_{\nu}=(1+z)E_{\nu}italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT = ( 1 + italic_z ) italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT, where the range of redshift is set to 0zzmax=50𝑧subscript𝑧max50\leq z\leq z_{\rm max}=50 ≤ italic_z ≤ italic_z start_POSTSUBSCRIPT roman_max end_POSTSUBSCRIPT = 5. As for the neutrino sources, we consider ordinary core-collapse SNe, BH-forming SNe, and failed SNe, whose spectra are denoted as, dNCCSN(Eν)/dEν𝑑subscript𝑁CCSNsubscriptsuperscript𝐸𝜈𝑑subscriptsuperscript𝐸𝜈dN_{\rm CCSN}(E^{\prime}_{\nu})/dE^{\prime}_{\nu}italic_d italic_N start_POSTSUBSCRIPT roman_CCSN end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT, dNBHSN(Eν)/dEν𝑑subscript𝑁BHSNsubscriptsuperscript𝐸𝜈𝑑subscriptsuperscript𝐸𝜈dN_{\rm BHSN}(E^{\prime}_{\nu})/dE^{\prime}_{\nu}italic_d italic_N start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT, and dNBH(Eν)/dEν𝑑subscript𝑁BHsubscriptsuperscript𝐸𝜈𝑑subscriptsuperscript𝐸𝜈dN_{\rm BH}(E^{\prime}_{\nu})/dE^{\prime}_{\nu}italic_d italic_N start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT respectively. For dNCCSN(Eν)/dEν𝑑subscript𝑁CCSNsubscriptsuperscript𝐸𝜈𝑑subscriptsuperscript𝐸𝜈dN_{\rm CCSN}(E^{\prime}_{\nu})/dE^{\prime}_{\nu}italic_d italic_N start_POSTSUBSCRIPT roman_CCSN end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT and dNBH(Eν)/dEν𝑑subscript𝑁BHsubscriptsuperscript𝐸𝜈𝑑subscriptsuperscript𝐸𝜈dN_{\rm BH}(E^{\prime}_{\nu})/dE^{\prime}_{\nu}italic_d italic_N start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT, we adopt the same models as Ashida et al. (2023), but we only investigate the case of Togashi EOS. The model of an ordinary core-collapse SN corresponds to the case where a 1.32M1.32subscript𝑀direct-product1.32M_{\odot}1.32 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT NS is formed from the collapse of a 15M15subscript𝑀direct-product15M_{\odot}15 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT progenitor, while the model of a failed SN corresponds to the case where a BH is formed without an explosion from the collapse of a 30M30subscript𝑀direct-product30M_{\odot}30 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT progenitor (Ashida & Nakazato, 2022). As stated in § 2, we investigate two cases: (i) fallback induced and (ii) prompt BH-forming SNe (Table 1) for dNBHSN(Eν)/dEν𝑑subscript𝑁BHSNsubscriptsuperscript𝐸𝜈𝑑subscriptsuperscript𝐸𝜈dN_{\rm BHSN}(E^{\prime}_{\nu})/dE^{\prime}_{\nu}italic_d italic_N start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT ( italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUPERSCRIPT ′ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT.

In eq. (1), RSN(z)subscript𝑅SN𝑧R_{\rm SN}(z)italic_R start_POSTSUBSCRIPT roman_SN end_POSTSUBSCRIPT ( italic_z ) and RBH(z)subscript𝑅BH𝑧R_{\rm BH}(z)italic_R start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_z ) are rates of successful SNe and failed SNe, respectively, as functions of redshift. We adopt RSN(z)subscript𝑅SN𝑧R_{\rm SN}(z)italic_R start_POSTSUBSCRIPT roman_SN end_POSTSUBSCRIPT ( italic_z ) and RBH(z)subscript𝑅BH𝑧R_{\rm BH}(z)italic_R start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_z ) deduced from the model of galactic chemical evolution in Tsujimoto (2023), which are also investigated in Ashida et al. (2023) as a reference model (see Table 1 of that paper). This model exhibits two distinct features. Firstly, the stellar IMF depends on the type of galaxies (e.g., Hopkins, 2018); the early-type galaxies, which are formed in bursty star formation, have a flat IMF (a slope index of a power law x=0.9𝑥0.9x=-0.9italic_x = - 0.9) and efficiently eject heavy elements while the late-type galaxies have the Salpeter IMF (x=1.35𝑥1.35x=-1.35italic_x = - 1.35). Secondly, the upper bound on the mass of core-collapse SN progenitors is 18M18subscript𝑀direct-product18M_{\odot}18 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT (e.g., Smartt, 2009, 2015; Sukhbold et al., 2016; Kresse et al., 2021); a mass range of progenitors is 8–18Msubscript𝑀direct-productM_{\odot}italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT for core-collapse SNe and 18–100Msubscript𝑀direct-productM_{\odot}italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT for failed SNe. The predicted redshift evolution of RSNsubscript𝑅SNR_{\rm SN}italic_R start_POSTSUBSCRIPT roman_SN end_POSTSUBSCRIPT is in better agreement with the measured rates. On the other hand, RBHsubscript𝑅BHR_{\rm BH}italic_R start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT corresponds to the rate of BH formations without SN explosions. In the present study, core-collapse SNe are classified into two categories: ordinary core-collapse SNe, which leave NSs, and BH-forming SNe. For simplicity, we assume that the fraction of BH-forming SNe, denoted as fBHNSsubscript𝑓BHNSf_{\rm BHNS}italic_f start_POSTSUBSCRIPT roman_BHNS end_POSTSUBSCRIPT, does not depend on the redshift.

In Tsujimoto (2023), RSN(z)subscript𝑅SN𝑧R_{\rm SN}(z)italic_R start_POSTSUBSCRIPT roman_SN end_POSTSUBSCRIPT ( italic_z ) and RBH(z)subscript𝑅BH𝑧R_{\rm BH}(z)italic_R start_POSTSUBSCRIPT roman_BH end_POSTSUBSCRIPT ( italic_z ) are calculated by converting from observationally estimated cosmic star formation rate (SFR). For this purpose, SFRs of Madau & Dickinson (2014) and Hopkins & Beacom (2006), which are referred to as MD14 and HB06, respectively, are used. We also investigate the both cases in this paper.

Since neutrinos undergo flavor oscillations before the detection, we take into account the so-called MSW effect (Wolfenstein, 1978; Mikheyev & Smirnov, 1985) following Nakazato et al. (2015). The survival probability of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT, P¯¯𝑃\bar{P}over¯ start_ARG italic_P end_ARG, depends on the neutrino mass hierarchy (Dighe & Smirnov, 2000) as P¯=cos2θ12cos2θ13¯𝑃superscript2subscript𝜃12superscript2subscript𝜃13\bar{P}=\cos^{2}\theta_{12}\cos^{2}\theta_{13}over¯ start_ARG italic_P end_ARG = roman_cos start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_θ start_POSTSUBSCRIPT 12 end_POSTSUBSCRIPT roman_cos start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_θ start_POSTSUBSCRIPT 13 end_POSTSUBSCRIPT for the normal mass hierarchy (NH) and P¯=sin2θ13¯𝑃superscript2subscript𝜃13\bar{P}=\sin^{2}\theta_{13}over¯ start_ARG italic_P end_ARG = roman_sin start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_θ start_POSTSUBSCRIPT 13 end_POSTSUBSCRIPT for the inverted mass hierarchy (IH), where θ12subscript𝜃12\theta_{12}italic_θ start_POSTSUBSCRIPT 12 end_POSTSUBSCRIPT and θ13subscript𝜃13\theta_{13}italic_θ start_POSTSUBSCRIPT 13 end_POSTSUBSCRIPT are mixing angles. Since recent measurements for them are sin2θ120.31superscript2subscript𝜃120.31\sin^{2}\theta_{12}\approx 0.31roman_sin start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_θ start_POSTSUBSCRIPT 12 end_POSTSUBSCRIPT ≈ 0.31 and sin2θ130.02superscript2subscript𝜃130.02\sin^{2}\theta_{13}\approx 0.02roman_sin start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT italic_θ start_POSTSUBSCRIPT 13 end_POSTSUBSCRIPT ≈ 0.02 (Workman et al., 2022), we set P¯=0.68¯𝑃0.68\bar{P}=0.68over¯ start_ARG italic_P end_ARG = 0.68 for NH and P¯=0.02¯𝑃0.02\bar{P}=0.02over¯ start_ARG italic_P end_ARG = 0.02 for IH in this study.

In Fig. 2, the ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT flux estimated by the DSNB model described above is compared with the latest experimental upper bounds. The largest flux is provided by the model with NH, HB06 SFR, and fallback induced BH formation for the following reason. In the case of fallback induced BH formation, the binding energy of the maximum-mass NS is converted to the total emission energy of neutrinos. Since, as already stated, the fallback mass accretion emits a much larger amount of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT than νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT, the terrestrial DSNB flux is larger for NH, which has higher survival probability of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT than IH. If the fraction of BH-forming SNe is fBHSN=0.5subscript𝑓BHSN0.5f_{\rm BHSN}=0.5italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.5 for NH, the DSNB flux exceeds the upper bounds in several energy bins. As shown in Fig. 3, where the integrated fluxes with Eν>17.3subscript𝐸𝜈17.3E_{\nu}>17.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT > 17.3 MeV are compared with the 90% C.L. upper limits and best-fit results in Abe et al. (2021b), the models of fallback induced BH formation has a constraint of fBHSN<0.45subscript𝑓BHSN0.45f_{\rm BHSN}<0.45italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT < 0.45 (0.26) for the case with NH and MD14 (HB06) SFR.

Figure 2: Spectra of cosmic background ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT flux from this study compared with the 90% confidence level upper bounds from SK with pure (Abe et al., 2021b) and gadolinium-loaded (Harada et al., 2023b) water, and KamLAND (Abe et al., 2022). Models with different mass hierarchy and SFR are shown in each panel: (a) NH and MD14, (b) NH and HB06, (C) IH and MD14, and (d) IH and HB06. Solid, dashed, and dot-dashed lines show the spectra estimated with fBHSN=0.5subscript𝑓BHSN0.5f_{\rm BHSN}=0.5italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.5, 0.1, and 0, respectively, where fBHSNsubscript𝑓BHSNf_{\rm BHSN}italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT is the fraction of successful SN explosions that form a BH. Orange and blue lines correspond to (i) fallback induced and (ii) prompt BH-formation cases, respectively, in Table 1.
Figure 3: The integrated ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT flux with Eν>17.3subscript𝐸𝜈17.3E_{\nu}>17.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT > 17.3 MeV estimated by our DSNB models in comparison with the best-fit values and their ±1σplus-or-minus1𝜎\pm 1\sigma± 1 italic_σ uncertainties shown in grey solid lines and shaded regions, respectively, and the 90% observed upper limits shown in red dot-dashed lines from SK with pure water (Abe et al., 2021b). Note that these experimental constraints are associated with the spectral models in Nakazato et al. (2015) while they are insensitive to the adopted spectrum models. Models with different mass hierarchy are shown in each panel: (a) NH and (b) IH. Solid and dashed lines represent the spectra for the models with MD14 SFR and HB06 SFR, respectively. Orange and blue lines correspond to (i) fallback induced and (ii) prompt BH-formation cases, respectively, in Table 1.

4 Event Rate and Experimental Sensitivity

In this section, we investigate the event rate spectra and the experimental sensitivity to our DSNB models following Ashida & Nakazato (2022). Water Cherenkov detectors, such as SK and HK, detect the DSNB via inverse beta decay (IBD) of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT:

ν¯e+pe++n.subscript¯𝜈𝑒𝑝superscript𝑒𝑛\bar{\nu}_{e}+p\to e^{+}+n.over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT + italic_p → italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_n . (2)

Thus, the DSNB event rate is calculated as

dNevent(Ee+)dEe+=NpσIBD(Eν)dΦν¯edet(Eν)dEν,𝑑subscript𝑁eventsubscript𝐸superscript𝑒𝑑subscript𝐸superscript𝑒subscript𝑁𝑝subscript𝜎IBDsubscript𝐸𝜈𝑑subscriptsuperscriptΦdetsubscript¯𝜈𝑒subscript𝐸𝜈𝑑subscript𝐸𝜈\frac{dN_{\rm event}(E_{e^{+}})}{dE_{e^{+}}}=N_{p}\cdot\sigma_{\rm IBD}(E_{\nu% })\cdot\frac{d\Phi^{\rm det}_{\bar{\nu}_{e}}(E_{\nu})}{dE_{\nu}},divide start_ARG italic_d italic_N start_POSTSUBSCRIPT roman_event end_POSTSUBSCRIPT ( italic_E start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ) end_ARG start_ARG italic_d italic_E start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT end_ARG = italic_N start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT ⋅ italic_σ start_POSTSUBSCRIPT roman_IBD end_POSTSUBSCRIPT ( italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) ⋅ divide start_ARG italic_d roman_Φ start_POSTSUPERSCRIPT roman_det end_POSTSUPERSCRIPT start_POSTSUBSCRIPT over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT end_POSTSUBSCRIPT ( italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) end_ARG start_ARG italic_d italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT end_ARG , (3)

where σIBD(Eν)subscript𝜎IBDsubscript𝐸𝜈\sigma_{\rm IBD}(E_{\nu})italic_σ start_POSTSUBSCRIPT roman_IBD end_POSTSUBSCRIPT ( italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) is the IBD cross section taken from Strumia & Vissani (2003) and dΦν¯edet(Eν)/dEν𝑑subscriptsuperscriptΦdetsubscript¯𝜈𝑒subscript𝐸𝜈𝑑subscript𝐸𝜈d\Phi^{\rm det}_{\bar{\nu}_{e}}(E_{\nu})/dE_{\nu}italic_d roman_Φ start_POSTSUPERSCRIPT roman_det end_POSTSUPERSCRIPT start_POSTSUBSCRIPT over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT end_POSTSUBSCRIPT ( italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT ) / italic_d italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT is terrestrial flux of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT. The positron energy Ee+subscript𝐸superscript𝑒E_{e^{+}}italic_E start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT is related to the neutrino energy Eνsubscript𝐸𝜈E_{\nu}italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT as Ee+=EνΔc2subscript𝐸superscript𝑒subscript𝐸𝜈Δsuperscript𝑐2E_{e^{+}}=E_{\nu}-\Delta c^{2}italic_E start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT = italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT - roman_Δ italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT, where ΔΔ\Deltaroman_Δ is a neutron-proton mass difference, and Npsubscript𝑁𝑝N_{p}italic_N start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT represents the number of free protons contained in the fiducial volume of the detector, which is Np=1.5×1033subscript𝑁𝑝1.5superscript1033N_{p}=1.5\times 10^{33}italic_N start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT = 1.5 × 10 start_POSTSUPERSCRIPT 33 end_POSTSUPERSCRIPT for SK and Np=12.6×1033subscript𝑁𝑝12.6superscript1033N_{p}=12.6\times 10^{33}italic_N start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT = 12.6 × 10 start_POSTSUPERSCRIPT 33 end_POSTSUPERSCRIPT for HK.

The event rate spectra at SK from our DSNB models with MD14 SFR are shown in Fig. 4. Incidentally, HK has a similar-to\sim8.4 times higher event rate than SK and the model with HB06 SFR has a similar-to\sim1.24 times higher event rate than that with MD14 SFR. If there are no BH-forming SNe (fBHSN=0subscript𝑓BHSN0f_{\rm BHSN}=0italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0), the expected number of IBD signal events with 17.3<Eν<31.317.3subscript𝐸𝜈31.317.3<E_{\nu}<31.317.3 < italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT < 31.3 MeV is 180 (170) for the model with NH (IH) and MD14 SFR at HK over 10 yr.222The event number is reduced to 50–60 when the detection efficiency including neutron tag is taken into account (Ashida et al., 2023). If the contributions of fallback induced BH-forming SNe are included and fBHSN=0.5subscript𝑓BHSN0.5f_{\rm BHSN}=0.5italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.5 is assumed, the event number increases to 340 for NH and 230 for IH. On the other hand, in the case of prompt BH-forming SNe and fBHSN=0.5subscript𝑓BHSN0.5f_{\rm BHSN}=0.5italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.5, the event number is 260 for NH and 210 for IH. The impact on the event number is largest for the case with NH and fallback induced BH-forming SNe. However, for the other cases also, the event numbers increase due to the BH-forming SNe while the impacts are not so large. This is because high-energy neutrino emission in the early dynamical phase is more efficient compared to ordinary core-collapse SNe. In any case, the inclusion of BH-forming SNe favors the detection of DSNB. Incidentally, it reduces the number of IBD signal events with Eν<13.3subscript𝐸𝜈13.3E_{\nu}<13.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT < 13.3 MeV, where many background events exist at HK, for the case with IH and prompt BH-forming SNe. This is because low-energy neutrinos are mainly emitted from the cooling of the PNS, which is not included in the prompt BH formation case.

Refer to caption
Figure 4: Predicted DSNB event rate spectra at SK (a water volume of 22.5 kton) per year with different choices of the BH-forming SNe model and fBHSNsubscript𝑓BHSNf_{\rm BHSN}italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT for NH (top) and IH (bottom) and MD14 SFR. The notation of lines is the same as in Fig. 2.
Figure 5: Operation time required to detect the DSNB at HK as a function of the fraction of BH-forming SNe, fBHSNsubscript𝑓BHSNf_{\rm BHSN}italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT, based on our models with different mass hierarchy at different C.L.: (a) NH and 2σ𝜎\sigmaitalic_σ, (b) IH and 2σ𝜎\sigmaitalic_σ, (c) NH and 3σ𝜎\sigmaitalic_σ, and (d) IH and 3σ𝜎\sigmaitalic_σ. The notation of lines is the same as in Fig. 3.

Now we move on to the experimental sensitivity. The expected upper bound on the integrated flux of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT with 17.3<Eν<31.317.3subscript𝐸𝜈31.317.3<E_{\nu}<31.317.3 < italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT < 31.3 MeV is calculated as (Abe et al., 2021b; Ashida & Nakazato, 2022)

Φlim=NlimTNpσ¯IBDϵsig,subscriptΦlimsubscript𝑁lim𝑇subscript𝑁𝑝subscript¯𝜎IBDsubscriptitalic-ϵsig\Phi_{\rm lim}=\frac{N_{\rm lim}}{T\cdot N_{p}\cdot\bar{\sigma}_{\rm IBD}\cdot% \epsilon_{\rm sig}},roman_Φ start_POSTSUBSCRIPT roman_lim end_POSTSUBSCRIPT = divide start_ARG italic_N start_POSTSUBSCRIPT roman_lim end_POSTSUBSCRIPT end_ARG start_ARG italic_T ⋅ italic_N start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT ⋅ over¯ start_ARG italic_σ end_ARG start_POSTSUBSCRIPT roman_IBD end_POSTSUBSCRIPT ⋅ italic_ϵ start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT end_ARG , (4)

where Nlimsubscript𝑁limN_{\rm lim}italic_N start_POSTSUBSCRIPT roman_lim end_POSTSUBSCRIPT is the upper bound on the number of events for the given operation time T𝑇Titalic_T. The IBD cross section at Eν=24.3subscript𝐸𝜈24.3E_{\nu}=24.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT = 24.3 MeV is used for the averaged cross section σ¯IBDsubscript¯𝜎IBD\bar{\sigma}_{\rm IBD}over¯ start_ARG italic_σ end_ARG start_POSTSUBSCRIPT roman_IBD end_POSTSUBSCRIPT. In the following, we consider the sensitivity at HK. The signal efficiency assumed in this study (ϵsigsubscriptitalic-ϵsig\epsilon_{\rm sig}italic_ϵ start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT) is taken from the former SK analysis (Abe et al., 2021b), as was done in our previous study (Ashida & Nakazato, 2022), which is around 20% to 30% depending on energy. As is done at SK, we assume that the IBD reaction of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT is identified by the coincidence of the Cherenkov light emitted from the positron and the 2.2 MeV γ𝛾\gammaitalic_γ ray emitted from neutron capture on hydrogen. This method is called neutron tagging and the signal efficiency becomes not very high due to the low energy of this γ𝛾\gammaitalic_γ ray.

For a certain confidence level (C.L.), Nlimsubscript𝑁limN_{\rm lim}italic_N start_POSTSUBSCRIPT roman_lim end_POSTSUBSCRIPT is obtained as the excess of the observation over the background expectation, where the statistical and systematic uncertainties of the background events are taken into account. The background at HK in higher energies (Eν>17.3subscript𝐸𝜈17.3E_{\nu}>17.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT > 17.3 MeV) mainly stems from atmospheric neutrinos and is classified into two categories: neutral-current quasielastic (NCQE) interactions and others (non-NCQE). In the NCQE interactions, a neutrino often knocks a neutron out of an oxygen nucleus, where the γ𝛾\gammaitalic_γ ray emitted from the deexcitation of the residual nucleus and the knocked out neutron mimic the positron signal from IBD. The size of the NCQE background assumed in the present study is scaled from that at the past SK experiment (Abe et al., 2021b). The systematic uncertainty of the NCQE background is assumed to decrease year by year (see Table 1 of Ashida & Nakazato, 2022) due to expected efforts in accelerator neutrino and nuclear experiments (Abe et al., 2019; Ashida et al., 2024; Tano et al., 2024). As for the non-NCQE background, we consider the following interactions: charged-current interaction of atmospheric ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT, which produces a positron and a neutron as in IBD, charged-current muon neutrino interaction, and some of neutral-current interactions involving a low-energy pion. In the latter two interactions, a visible positron from the decay of an invisible muon becomes background in the DSNB search because a neutron produced by the parent reaction may be tagged or, even without a neutron, an accidental coincidence with noise might occur. The size and systematic uncertainty of the non-NCQE backgrounds are again extrapolated from the SK experiment in this study. In addition, we also take into account the background due to spallation of oxygen nuclei induced by energetic atmospheric muons, which produces radioactive isotopes and leads to misidentification as IBD events. HK is expected to suffer from the four times higher spallation background event rate per volume compared to SK due to the shallower depth of its construction site (Abe et al., 2018). Not only the isotopes with β𝛽\betaitalic_β+n𝑛nitalic_n decay, such as 9Li, increase, but also the likelihood of accidental coincidences becomes higher. We then increase both 9Li and accidental backgrounds by a factor of four compared to the ones in Ashida & Nakazato (2022)333This is rather a conservative estimation because the accidental background in the energy range of the present analysis is not necessarily made by spallation but partially by atmospheric events as well..

Using ΦlimsubscriptΦlim\Phi_{\rm lim}roman_Φ start_POSTSUBSCRIPT roman_lim end_POSTSUBSCRIPT obtained by eq. (4), we evaluate the operation time required to detect the DSNB at HK for our models. This is shown in Fig. 5 as a function of fBHSNsubscript𝑓BHSNf_{\rm BHSN}italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT for different cases and C.L. We found that the DSNB detection would be achieved within at most similar-to\sim6 yr at 2σ𝜎\sigmaitalic_σ and similar-to\sim20 yr at 3σ𝜎\sigmaitalic_σ, and including the BH-forming SNe reduces the required operation time in any case. Furthermore, the impact of BH-forming SNe is significant in some cases; the operation time required for 3σ𝜎\sigmaitalic_σ is shortened by half with fBHSN=0.2subscript𝑓BHSN0.2f_{\rm BHSN}=0.2italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.2 for the model of fallback induced BH formation, NH, and MD14 SFR. Otherwise, the upper bound on fBHSNsubscript𝑓BHSNf_{\rm BHSN}italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT would be provided within the expected operational period of HK.

5 Summary and Discussion

In this study, we have constructed a new DSNB model that includes the contribution of BH-forming SNe, which lead to both a successful SN explosion and BH formation simultaneously. According to studies on Galactic chemical evolution and nucleosynthesis, the population of BH-forming SNe is implied to be non-negligible in accounting for the observed abundance of some heavy elements. Since the detailed dynamics and neutrino emission of BH-forming SNe are uncertain, we have considered two extreme cases: fallback-induced BH formation and prompt BH formation. In the first scenario, a longer duration until BH formation ensures significant neutrino emission, and the total energy of emitted neutrinos in our model is consistent with the binding energy of a maximum-mass NS. On the other hand, in the second scenario, the shorter duration results in reduced neutrino emission. The rates of successful SNe (the sum of ordinary core-collapse SNe and BH-forming SNe) and failed SNe (BH formation without SN explosions) have been based on the model of galactic chemical evolution in Tsujimoto (2023). As a result, we have found that the contribution of BH-forming SNe enhances the flux and event rate of the DSNB at high energies (Eν>17.3subscript𝐸𝜈17.3E_{\nu}>17.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT > 17.3 MeV). In particular, the impacts are largest in the case of fallback-induced BH formation with neutrino oscillation in NH since the fallback mass accretion onto a PNS emits a much larger amount of ν¯esubscript¯𝜈𝑒\bar{\nu}_{e}over¯ start_ARG italic_ν end_ARG start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT than νxsubscript𝜈𝑥\nu_{x}italic_ν start_POSTSUBSCRIPT italic_x end_POSTSUBSCRIPT. In this case, the expected event rate at Eν>17.3subscript𝐸𝜈17.3E_{\nu}>17.3italic_E start_POSTSUBSCRIPT italic_ν end_POSTSUBSCRIPT > 17.3 MeV doubles if the fraction of BH-forming SNe is fBHSN=0.5subscript𝑓BHSN0.5f_{\rm BHSN}=0.5italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.5. Furthermore, with fBHSN=0.2subscript𝑓BHSN0.2f_{\rm BHSN}=0.2italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT = 0.2, the operation time required for 3σ3𝜎3\sigma3 italic_σ detection at HK is shortened by half, assuming MD14 SFR. Similarly, in other cases, the required operation time is also reduced due to the contribution of BH-forming SNe.

Concerning the treatment of BH-forming SNe, there is room for further improvement. In this paper, we have avoided specifying the dynamics of BH-forming SNe and instead focused on discussing two extreme cases of neutrino emission. In actual conditions, the amount of neutrino emission may lie between these two extremes or vary widely. While numerical examples are still limited, the model of Burrows et al. (2023) has a short time to BH formation of less than 2 s, which can be considered to be similar to our prompt BH formation model. Incidentally, Kresse et al. (2021) shows that even failed SNe may take around 10 s to form a BH. The time to BH formation is heavily dependent on the efficiency of fallback mass accretion, which is significantly determined by the structure of the progenitor. Therefore, it is worthwhile to investigate the neutrino emissions from BH-forming SNe using the same progenitor models employed in studies of nucleosynthesis and chemical evolution.

In the present study, we assume that BH-forming SNe reside within the mass range of 8–18M18subscript𝑀direct-product18M_{\odot}18 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT according to the observed implication (Smartt, 2009, 2015). In contrast, the arguments based on nucleosynthesis/chemical evolution have been done under the hypothesis that the progenitor masses of BH-forming SNe would be larger than 20M20subscript𝑀direct-product20M_{\odot}20 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT (e.g., Kobayashi et al., 2006; Pignatari et al., 2023). Thus, there is a clear inconsistency between the two. While our results depend mainly on the fraction of BH-forming SNe, i.e., fBHSNsubscript𝑓BHSNf_{\rm BHSN}italic_f start_POSTSUBSCRIPT roman_BHSN end_POSTSUBSCRIPT, for the given rate of successful SNe, i.e., RSNsubscript𝑅SNR_{\rm SN}italic_R start_POSTSUBSCRIPT roman_SN end_POSTSUBSCRIPT, regardless of their progenitor masses, it is worthwhile to discuss how the observed upper mass bound (18M18subscript𝑀direct-product18M_{\odot}18 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT) can be reconciled with their masses (>20Mabsent20subscript𝑀direct-product>20M_{\odot}> 20 italic_M start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT) implied from the theoretical argument based on nucleosynthesis.

One possible explanation is the observational bias. Many BH-forming SNe may be unnoticed if they exhibit systematically lower peak luminosities than ordinary core-collapse SNe. This possibility is quite plausible for the case of faint SNe (Nomoto et al., 2006). Even for hypernovae, such a non-detection is possible owing to their jet-like explosions; in most cases, the jets do not direct to us and the corresponding SNe seem to exhibit low brightness. Another possible solution is the metallicity-dependent frequency of BH-forming SNe; they exclusively emerge in a low-metallicity environment, which is provided by the limited regions in the local Universe. This possibility is in particular expected for hypernovae, because their progenitors are considered to be fast-rotating massive stars (e.g., Iwamoto et al., 1998) and a low metallicity helps to retain enough angular momentum (e.g., Woosley & Heger, 2006). In addition, hypernovae could be closely connected to long gamma ray bursts (Galama et al., 1998), whose emergence is indeed biased toward low-metalliicty (Z0.30.5Zless-than-or-similar-to𝑍0.30.5subscript𝑍direct-productZ\lesssim 0.3-0.5Z_{\odot}italic_Z ≲ 0.3 - 0.5 italic_Z start_POSTSUBSCRIPT ⊙ end_POSTSUBSCRIPT) galaxies (Fruchter et al., 2006; Vergani et al., 2015). These arguments propose that the redshift evolution of metallicity for individual galaxies could be one of the key factors including the IMF for counting the DSNB flux, as done by Nakazato et al. (2015).

This work is supported by Grants-in-Aid for Scientific Research (JP18H01258, JP20K03973, JP23H00132, JP24K07021), Grant-in-Aid for Scientific Research on Innovative Areas (JP19H05811), and Grant-in-Aid for Transformative Research Areas (JP24H02245) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan, and by NSF Grant No. PHY-2309967.

References

  • Abbasi et al. (2011) Abbasi, R., Abdou, Y., Abu-Zayyad, T., et al. 2011, A&A, 535, A109, doi: 10.1051/0004-6361/201117810
  • Abe et al. (2016) Abe, K., Haga, Y., Hayato, Y., et al. 2016, Astroparticle Physics, 81, 39, doi: 10.1016/j.astropartphys.2016.04.003
  • Abe et al. (2018) Abe, K., Abe, K., Aihara, H., et al. 2018, arXiv e-prints, arXiv:1805.04163, doi: 10.48550/arXiv.1805.04163
  • Abe et al. (2019) Abe, K., Akutsu, R., Ali, A., et al. 2019, Phys. Rev. D, 100, 112009, doi: 10.1103/PhysRevD.100.112009
  • Abe et al. (2021a) Abe, K., Adrich, P., Aihara, H., et al. 2021a, ApJ, 916, 15, doi: 10.3847/1538-4357/abf7c4
  • Abe et al. (2021b) Abe, K., Bronner, C., Hayato, Y., et al. 2021b, Phys. Rev. D, 104, 122002, doi: 10.1103/PhysRevD.104.122002
  • Abe et al. (2022) Abe, S., Asami, S., Gando, A., et al. 2022, ApJ, 925, 14, doi: 10.3847/1538-4357/ac32c1
  • Akaho et al. (2024) Akaho, R., Nagakura, H., & Foglizzo, T. 2024, ApJ, 960, 116, doi: 10.3847/1538-4357/ad118c
  • Alekseev et al. (1987) Alekseev, E. N., Alekseeva, L. N., Volchenko, V. I., & Krivosheina, I. V. 1987, Soviet Journal of Experimental and Theoretical Physics Letters, 45, 589
  • Anandagoda et al. (2023) Anandagoda, S., Hartmann, D. H., Fryer, C. L., et al. 2023, ApJ, 950, 29, doi: 10.3847/1538-4357/acc84f
  • Ando (2003) Ando, S. 2003, Physics Letters B, 570, 11, doi: 10.1016/j.physletb.2003.07.009
  • Ando et al. (2023) Ando, S., Ekanger, N., Horiuchi, S., & Koshio, Y. 2023, Proceedings of the Japan Academy, Series B, 99, 460, doi: 10.2183/pjab.99.026
  • Ando & Sato (2004) Ando, S., & Sato, K. 2004, New Journal of Physics, 6, 170, doi: 10.1088/1367-2630/6/1/170
  • Ando et al. (2003) Ando, S., Sato, K., & Totani, T. 2003, Astroparticle Physics, 18, 307, doi: 10.1016/S0927-6505(02)00152-4
  • Aoyama et al. (2023) Aoyama, S., Ouchi, M., & Harikane, Y. 2023, ApJ, 946, 69, doi: 10.3847/1538-4357/acba87
  • Ashida (2024) Ashida, Y. 2024, arXiv e-prints, arXiv:2401.12403, doi: 10.48550/arXiv.2401.12403
  • Ashida & Nakazato (2022) Ashida, Y., & Nakazato, K. 2022, ApJ, 937, 30, doi: 10.3847/1538-4357/ac8a46
  • Ashida et al. (2023) Ashida, Y., Nakazato, K., & Tsujimoto, T. 2023, ApJ, 953, 151, doi: 10.3847/1538-4357/ace3ba
  • Ashida et al. (2024) Ashida, Y., Nagata, H., Mori, M., et al. 2024, Phys. Rev. C, 109, 014620, doi: 10.1103/PhysRevC.109.014620
  • Bays et al. (2012) Bays, K., Iida, T., Abe, K., et al. 2012, Phys. Rev. D, 85, 052007, doi: 10.1103/PhysRevD.85.052007
  • Beacom (2010) Beacom, J. F. 2010, Annual Review of Nuclear and Particle Science, 60, 439, doi: 10.1146/annurev.nucl.010909.083331
  • Beacom & Vagins (2004) Beacom, J. F., & Vagins, M. R. 2004, Phys. Rev. Lett., 93, 171101, doi: 10.1103/PhysRevLett.93.171101
  • Bionta et al. (1987) Bionta, R. M., Blewitt, G., Bratton, C. B., et al. 1987, Phys. Rev. Lett., 58, 1494, doi: 10.1103/PhysRevLett.58.1494
  • Bisnovatyi-Kogan & Seidov (1982) Bisnovatyi-Kogan, G. S., & Seidov, Z. F. 1982, Soviet Ast., 26, 132
  • Burrows et al. (2023) Burrows, A., Vartanyan, D., & Wang, T. 2023, ApJ, 957, 68, doi: 10.3847/1538-4357/acfc1c
  • Dar (1985) Dar, A. 1985, Phys. Rev. Lett., 55, 1422, doi: 10.1103/PhysRevLett.55.1422
  • de Gouvêa et al. (2020) de Gouvêa, A., Martinez-Soler, I., Perez-Gonzalez, Y. F., & Sen, M. 2020, Phys. Rev. D, 102, 123012, doi: 10.1103/PhysRevD.102.123012
  • de Gouvêa et al. (2022) —. 2022, Phys. Rev. D, 106, 103026, doi: 10.1103/PhysRevD.106.103026
  • Dighe & Smirnov (2000) Dighe, A. S., & Smirnov, A. Y. 2000, Phys. Rev. D, 62, 033007, doi: 10.1103/PhysRevD.62.033007
  • Ekanger et al. (2022) Ekanger, N., Horiuchi, S., Kotake, K., & Sumiyoshi, K. 2022, Phys. Rev. D, 106, 043026, doi: 10.1103/PhysRevD.106.043026
  • Ekanger et al. (2024) Ekanger, N., Horiuchi, S., Nagakura, H., & Reitz, S. 2024, Phys. Rev. D, 109, 023024, doi: 10.1103/PhysRevD.109.023024
  • Fogli et al. (2004) Fogli, G. L., Lisi, E., Mirizzi, A., & Montanino, D. 2004, Phys. Rev. D, 70, 013001, doi: 10.1103/PhysRevD.70.013001
  • Fruchter et al. (2006) Fruchter, A. S., Levan, A. J., Strolger, L., et al. 2006, Nature, 441, 463, doi: 10.1038/nature04787
  • Fujibayashi et al. (2015) Fujibayashi, S., Yoshida, T., & Sekiguchi, Y. 2015, ApJ, 810, 115, doi: 10.1088/0004-637X/810/2/115
  • Furusawa et al. (2017) Furusawa, S., Togashi, H., Nagakura, H., et al. 2017, Journal of Physics G Nuclear Physics, 44, 094001, doi: 10.1088/1361-6471/aa7f35
  • Galais et al. (2010) Galais, S., Kneller, J., Volpe, C., & Gava, J. 2010, Phys. Rev. D, 81, 053002, doi: 10.1103/PhysRevD.81.053002
  • Galama et al. (1998) Galama, T. J., Vreeswijk, P. M., van Paradijs, J., et al. 1998, Nature, 395, 670, doi: 10.1038/27150
  • Gando et al. (2012) Gando, A., Gando, Y., Ichimura, K., et al. 2012, ApJ, 745, 193, doi: 10.1088/0004-637X/745/2/193
  • Harada et al. (2023a) Harada, A., Suwa, Y., Harada, M., et al. 2023a, ApJ, 954, 52, doi: 10.3847/1538-4357/ace52e
  • Harada et al. (2023b) Harada, M., Abe, K., Bronner, C., et al. 2023b, ApJ, 951, L27, doi: 10.3847/2041-8213/acdc9e
  • Hartmann & Woosley (1997) Hartmann, D. H., & Woosley, S. E. 1997, Astroparticle Physics, 7, 137, doi: 10.1016/S0927-6505(97)00018-2
  • Heger et al. (2003) Heger, A., Fryer, C. L., Woosley, S. E., Langer, N., & Hartmann, D. H. 2003, ApJ, 591, 288, doi: 10.1086/375341
  • Hirata et al. (1987) Hirata, K., Kajita, T., Koshiba, M., et al. 1987, Phys. Rev. Lett., 58, 1490, doi: 10.1103/PhysRevLett.58.1490
  • Hopkins (2018) Hopkins, A. M. 2018, PASA, 35, e039, doi: 10.1017/pasa.2018.29
  • Hopkins & Beacom (2006) Hopkins, A. M., & Beacom, J. F. 2006, ApJ, 651, 142, doi: 10.1086/506610
  • Horiuchi et al. (2009) Horiuchi, S., Beacom, J. F., & Dwek, E. 2009, Phys. Rev. D, 79, 083013, doi: 10.1103/PhysRevD.79.083013
  • Horiuchi et al. (2021) Horiuchi, S., Kinugawa, T., Takiwaki, T., Takahashi, K., & Kotake, K. 2021, Phys. Rev. D, 103, 043003, doi: 10.1103/PhysRevD.103.043003
  • Horiuchi et al. (2018) Horiuchi, S., Sumiyoshi, K., Nakamura, K., et al. 2018, MNRAS, 475, 1363, doi: 10.1093/mnras/stx3271
  • Iocco et al. (2005) Iocco, F., Mangano, G., Miele, G., Raffelt, G. G., & Serpico, P. D. 2005, Astroparticle Physics, 23, 303, doi: 10.1016/j.astropartphys.2005.01.004
  • Iváñez-Ballesteros & Volpe (2023) Iváñez-Ballesteros, P., & Volpe, M. C. 2023, Phys. Rev. D, 107, 023017, doi: 10.1103/PhysRevD.107.023017
  • Iwamoto et al. (1998) Iwamoto, K., Mazzali, P. A., Nomoto, K., et al. 1998, Nature, 395, 672, doi: 10.1038/27155
  • Kaplinghat et al. (2000) Kaplinghat, M., Steigman, G., & Walker, T. P. 2000, Phys. Rev. D, 62, 043001, doi: 10.1103/PhysRevD.62.043001
  • Kashiwagi et al. (2024) Kashiwagi, Y., Abe, K., Bronner, C., et al. 2024, arXiv e-prints, arXiv:2403.06760, doi: 10.48550/arXiv.2403.06760
  • Kobayashi et al. (2020) Kobayashi, C., Karakas, A. I., & Lugaro, M. 2020, ApJ, 900, 179, doi: 10.3847/1538-4357/abae65
  • Kobayashi et al. (2006) Kobayashi, C., Umeda, H., Nomoto, K., Tominaga, N., & Ohkubo, T. 2006, ApJ, 653, 1145, doi: 10.1086/508914
  • Krauss et al. (1984) Krauss, L. M., Glashow, S. L., & Schramm, D. N. 1984, Nature, 310, 191, doi: 10.1038/310191a0
  • Kresse et al. (2021) Kresse, D., Ertl, T., & Janka, H.-T. 2021, ApJ, 909, 169, doi: 10.3847/1538-4357/abd54e
  • Li et al. (2022) Li, Y.-F., Vagins, M., & Wurm, M. 2022, Universe, 8, 181, doi: 10.3390/universe8030181
  • Lunardini (2009) Lunardini, C. 2009, Phys. Rev. Lett., 102, 231101, doi: 10.1103/PhysRevLett.102.231101
  • Lunardini (2016) —. 2016, Astroparticle Physics, 79, 49, doi: 10.1016/j.astropartphys.2016.02.005
  • Madau & Dickinson (2014) Madau, P., & Dickinson, M. 2014, ARA&A, 52, 415, doi: 10.1146/annurev-astro-081811-125615
  • Malaney (1997) Malaney, R. A. 1997, Astroparticle Physics, 7, 125, doi: 10.1016/S0927-6505(97)00012-1
  • Mathews et al. (2020) Mathews, G. J., Boccioli, L., Hidaka, J., & Kajino, T. 2020, Modern Physics Letters A, 35, 2030011, doi: 10.1142/S0217732320300116
  • Mathews et al. (2014) Mathews, G. J., Hidaka, J., Kajino, T., & Suzuki, J. 2014, ApJ, 790, 115, doi: 10.1088/0004-637X/790/2/115
  • Mikheyev & Smirnov (1985) Mikheyev, S. P., & Smirnov, A. Y. 1985, Yadernaya Fizika, 42, 1441
  • MØller et al. (2018) MØller, K., Suliga, A. M., Tamborra, I., & Denton, P. B. 2018, J. Cosmology Astropart. Phys, 2018, 066, doi: 10.1088/1475-7516/2018/05/066
  • Nagakura & Vartanyan (2022) Nagakura, H., & Vartanyan, D. 2022, MNRAS, 512, 2806, doi: 10.1093/mnras/stac383
  • Nakazato (2013) Nakazato, K. 2013, Phys. Rev. D, 88, 083012, doi: 10.1103/PhysRevD.88.083012
  • Nakazato et al. (2015) Nakazato, K., Mochida, E., Niino, Y., & Suzuki, H. 2015, ApJ, 804, 75, doi: 10.1088/0004-637X/804/1/75
  • Nakazato et al. (2013) Nakazato, K., Sumiyoshi, K., Suzuki, H., et al. 2013, ApJS, 205, 2, doi: 10.1088/0067-0049/205/1/2
  • Nakazato et al. (2021) Nakazato, K., Sumiyoshi, K., & Togashi, H. 2021, PASJ, 73, 639, doi: 10.1093/pasj/psab026
  • Nakazato et al. (2006) Nakazato, K., Sumiyoshi, K., & Yamada, S. 2006, ApJ, 645, 519, doi: 10.1086/504282
  • Nomoto et al. (2013) Nomoto, K., Kobayashi, C., & Tominaga, N. 2013, ARA&A, 51, 457, doi: 10.1146/annurev-astro-082812-140956
  • Nomoto et al. (2006) Nomoto, K., Tominaga, N., Umeda, H., Kobayashi, C., & Maeda, K. 2006, Nucl. Phys. A, 777, 424, doi: 10.1016/j.nuclphysa.2006.05.008
  • Pignatari et al. (2023) Pignatari, M., Trueman, T. C. L., Womack, K. A., et al. 2023, MNRAS, 524, 6295, doi: 10.1093/mnras/stad2167
  • Placco et al. (2014) Placco, V. M., Frebel, A., Beers, T. C., & Stancliffe, R. J. 2014, ApJ, 797, 21, doi: 10.1088/0004-637X/797/1/21
  • Poutanen et al. (2007) Poutanen, J., Lipunova, G., Fabrika, S., Butkevich, A. G., & Abolmasov, P. 2007, MNRAS, 377, 1187, doi: 10.1111/j.1365-2966.2007.11668.x
  • Priya & Lunardini (2017) Priya, A., & Lunardini, C. 2017, J. Cosmology Astropart. Phys, 2017, 031, doi: 10.1088/1475-7516/2017/11/031
  • Sasaki et al. (2022) Sasaki, H., Yamazaki, Y., Kajino, T., et al. 2022, ApJ, 924, 29, doi: 10.3847/1538-4357/ac34f8
  • Sawatzki et al. (2021) Sawatzki, J., Wurm, M., & Kresse, D. 2021, Phys. Rev. D, 103, 023021, doi: 10.1103/PhysRevD.103.023021
  • Schilbach et al. (2019) Schilbach, T. S. H., Caballero, O. L., & McLaughlin, G. C. 2019, Phys. Rev. D, 100, 043008, doi: 10.1103/PhysRevD.100.043008
  • Smartt (2009) Smartt, S. J. 2009, ARA&A, 47, 63, doi: 10.1146/annurev-astro-082708-101737
  • Smartt (2015) —. 2015, PASA, 32, e016, doi: 10.1017/pasa.2015.17
  • Strigari et al. (2005) Strigari, L. E., Beacom, J. F., Walker, T. P., & Zhang, P. 2005, J. Cosmology Astropart. Phys, 2005, 017, doi: 10.1088/1475-7516/2005/04/017
  • Strumia & Vissani (2003) Strumia, A., & Vissani, F. 2003, Physics Letters B, 564, 42, doi: 10.1016/S0370-2693(03)00616-6
  • Sukhbold et al. (2016) Sukhbold, T., Ertl, T., Woosley, S. E., Brown, J. M., & Janka, H. T. 2016, ApJ, 821, 38, doi: 10.3847/0004-637X/821/1/38
  • Suliga et al. (2022) Suliga, A. M., Beacom, J. F., & Tamborra, I. 2022, Phys. Rev. D, 105, 043008, doi: 10.1103/PhysRevD.105.043008
  • Sumiyoshi et al. (2023) Sumiyoshi, K., Furusawa, S., Nagakura, H., et al. 2023, Progress of Theoretical and Experimental Physics, 2023, 013E02, doi: 10.1093/ptep/ptac167
  • Suwa et al. (2022) Suwa, Y., Harada, A., Harada, M., et al. 2022, ApJ, 934, 15, doi: 10.3847/1538-4357/ac795e
  • Suwa et al. (2024) Suwa, Y., Harada, A., Mori, M., et al. 2024, arXiv e-prints, arXiv:2404.18248, doi: 10.48550/arXiv.2404.18248
  • Suzuki (1994) Suzuki, H. 1994, in Physics and Astrophysics of Neutrinos, XIII, ed. M. Fukugita & A. Suzuki, 420
  • Tabrizi & Horiuchi (2021) Tabrizi, Z., & Horiuchi, S. 2021, J. Cosmology Astropart. Phys, 2021, 011, doi: 10.1088/1475-7516/2021/05/011
  • Tano et al. (2024) Tano, T., Horai, T., Ashida, Y., et al. 2024, arXiv e-prints, arXiv:2405.15366, doi: 10.48550/arXiv.2405.15366
  • Togashi et al. (2017) Togashi, H., Nakazato, K., Takehara, Y., et al. 2017, Nucl. Phys. A, 961, 78, doi: 10.1016/j.nuclphysa.2017.02.010
  • Totani & Sato (1995) Totani, T., & Sato, K. 1995, Astroparticle Physics, 3, 367, doi: 10.1016/0927-6505(95)00015-9
  • Totani et al. (1996) Totani, T., Sato, K., & Yoshii, Y. 1996, ApJ, 460, 303, doi: 10.1086/176970
  • Tsujimoto (2023) Tsujimoto, T. 2023, MNRAS, 518, 3475, doi: 10.1093/mnras/stac3351
  • Vergani et al. (2015) Vergani, S. D., Salvaterra, R., Japelj, J., et al. 2015, A&A, 581, A102, doi: 10.1051/0004-6361/201425013
  • Wei et al. (2024) Wei, Y.-F., Liu, T., & Song, C.-Y. 2024, ApJ, 966, 101, doi: 10.3847/1538-4357/ad3824
  • Wolfenstein (1978) Wolfenstein, L. 1978, Phys. Rev. D, 17, 2369, doi: 10.1103/PhysRevD.17.2369
  • Woosley & Heger (2006) Woosley, S. E., & Heger, A. 2006, ApJ, 637, 914, doi: 10.1086/498500
  • Workman et al. (2022) Workman, R. L., Burkert, V. D., Crede, V., et al. 2022, Progress of Theoretical and Experimental Physics, 2022, 083C01, doi: 10.1093/ptep/ptac097
  • Zhang et al. (2015) Zhang, H., Abe, K., Hayato, Y., et al. 2015, Astroparticle Physics, 60, 41, doi: 10.1016/j.astropartphys.2014.05.004
  • Ziegler et al. (2022) Ziegler, J. J., Edwards, T. D. P., Suliga, A. M., et al. 2022, MNRAS, 517, 2471, doi: 10.1093/mnras/stac2748