A database of spectral energy distributions of progenitors of core-collapse supernovae
aa r X i v : . [ a s t r o - ph . S R ] F e b Research in Astronomy and Astrophysics manuscript no.(L A TEX: db˙preCCSNe˙rev3.tex; printed on February 8, 2021; 1:37)
A database of spectral energy distributions of progenitors ofcore-collapse supernovae ⋆ Zhong-Mu Li and Cai-Yan Mao Institute for Astronomy and History of Science and Technology, Dali University, Dali 671003, China; [email protected]
Received 20xx month day; accepted 20xx month day
Abstract
This paper presents a database of the spectroscopic- and photometric- spectral en-ergy distributions (spec-SEDs and phot-SEDs) of the progenitors of core-collapse supernovae(CCSNe). Both binary- and single-star progenitors are included in the database. The databasecovers the initial metallicity ( Z ) range of 0.0001–0.03, mass range of 8–25 M ⊙ , binary massratio range of 0–1, and orbital period range of 0.1–10000 days. The low-resolution spec-SEDsand phot-SEDs of single- and binary-star CCSN progenitors are included in the database.These data can be used for studying the basic parameters, e.g., metallicity, age, initial andfinal masses of CCSN progenitors. It can also be used for studying the effects of differentfactors on the determination of parameters of CCSN progenitors. When the database is usedfor fitting the SEDs of binary-star CCSN progenitors, it is strongly suggested to determine themetallicity and orbital period in advance, while it is not necessary for single-star progenitors. Key words: (stars:) supernovae: general < Stars — Astronomical Databases — (stars:)binaries: general < Stars
As the main kind of supernovae (SNe), core-collapse supernova (CCSN) is the explosion that attend thedeath of massive stars. CCSN is a singularly important phenomenon in the universe for two main reasons.First, CCSNe are principal drivers of cosmic chemical evolution. Most heavy elements heavier than hydro-gen (H) and helium (He), except those around the iron peak, were synthesized by CCSNe. Second, theyare possibly related to the rapid neutron capture process (r-process), which produced many of the extremelyheavy elements above atomic mass of approximately 70. CCSNe have observed kinetic energies of typically ∼ ergs, and their integrated luminosities are usually 1-–10% of this value (Smartt 2009). The explosionof CCSNe has been a perennial challenge in theoretical astrophysics for decades. So does the progenitor ofCCSNe. ⋆ The database and SED fitting code are available at GitHub.
Li & Mao
The progenitor of CCSNe is fundamental for understanding CCSNe, in particular, for their explosions.However, it is still far from well understanding the formation and properties of CCSN progenitors. It iswell-known that the minimum initial mass that can produce a CCSN is about 8 M ⊙ , according to the directdetections of red supergiant progenitors and the most massive white dwarf progenitors. The maximum initialmass is less than about 25 M ⊙ , because the majority of massive stars above 20 M ⊙ may collapse quietlyto black holes and that the explosions remain undetected. The progenitors of CCSNe have been widelystudied, but the results are actually model dependent (Smartt 2009). The common image to form a CCSNprogenitor is as follows. The H in stellar cores converts to He in stellar evolution. If a star is sufficientlymassive, heavier elements such as carbon, oxygen, nickel, nitrogen, magnesium, silicon and iron (C, O,Ne, N, Mg, Si and Fe) are subsequently produced in nuclear synthesis reactions. For stars more massivethan 8 M ⊙ , either an O-Ne-Mg core (Poelarends et al. 2008) or a Fe core (Woosley et al. 2002) will formeventually and would cause an SN explosion (Lisakov 2018).Most studies of CCSN progenitors so far take single-star models. For example, the nearest progenitor,SN1987A, which is in the Large Magellanic Cloud (LMC), is shown to be evolved from a single star withinitial mass in the region of 14—20 M ⊙ . Using detailed stellar evolutionary codes such as MESA (Paxtonet al. 2011), many works evolved massive star models from the main sequence until core collapse (e.g.,Lisakov et al. 2018). These works put forward a lot of the studies of CCSN progenitors. However, single-starmodels ignore the observed fact that a lot of stars are in binaries. In fact, many of the supernova progenitorsare possibly binaries. For example, CCSN SN1993J needs a binary of 15+14 M ⊙ with an initial orbitalperiod of 5.8 years to explain the observed UBVRCIc SED (Humphreys & Davidson 1994). A potentialsurviving companion of the type Ia supernova Tycho Brahe’s 1572, was found around the position of theexplosion (Ruiz-Lapuente et al. 2004).According to stellar evolution theory, binary stars evolve in a substantially different way from singlestars if their components are not too far from each other. Binary interactions are therefore very importantfor stellar evolution (Podsiadlowski et al. 1992). Binary evolution affects both the population synthesis of alarge amount of stars (Podsiadlowski et al. 2002; Belczynski et al. 2008; Han et al. 2007; Li & Han 2008;Zhang & Li 2006; Eldridge et al. 2017; Farrell et al. 2020) and the detailed models of a small numberof stars. If binary evolution is taken instead of single-star evolution, CCSN progenitors could differ a lotin mass, age, radius, and composition (Farrell et al. 2020; Zapartas et al. 2020), which would impact theresulting SN and its remnant obviously.There has been a long history for the study of binary-star progenitors. For example, Eldridge et al.(2008) investigated the effect of massive binaries on stellar populations and SN progenitors using a detailedstellar evolution code (Eldridge & Tout 2004) and the single-star evolution equation of Hurley et al. (2000).Their theoretical predictions from binary-star models agree with the observed ratios of the Type Ib/c SNrate to the Type II SN rate at different metallicities, but the single-star models predict a lower relative ratefor Type Ib/c SNe than the observation. This implies that many CCSNe stem from binary-star progenitors.However, their results cannot be used for studying the SEDs of CCSNe progenitors because SEDs were notcalculated and the binary parameter-space resolution of that work is rather low. Waldman (2008) studied thethe heaviest models which do not encounter CCSN, and obtain some similar results with previous results EDs of CCSN Progenitors 3 (e.g., Woosley et al. 2007; Umeda & Nomoto 2008). The properties of Type Ib and IIb SN progenitors thatare produced by stable mass transfer in binary systems were explored using the MESA stellar evolutioncode (Paxton et al. 2011; Yoon et al. 2017) from an initial primary mass in the range of 10—18 M ⊙ at solarand LMC metallicities. However, only two metallicities were considered and it is not enough for manystudies of CCSN progenitors, in particular the SED studies.Besides binarity, stellar rotation and magnetic fields were also not considered in most CCSN progenitorstudies, although they have some effects on the formation and properties of progenitors (e.g., Heger et al.2000; Meynet & Maeder 2007; Woosley & Janka 2005; Langer 2012). In fact, there is still a long way togo, because the effect of stellar rotation and magnetic field remains very uncertain (e.g., Powell & M¨uller2020).In the studies of CCSNe, there have been a few good algorithms, e.g., Supernova Identification (SNID).Such codes can be used to identify the type of an SN spectrum and to determine its redshift and age(Blondin & Tonry 2007). However, there is no comprehensive SED database to determine the propertiesof different kinds of CCSN progenitors yet. This hampers many studies, e.g., the identification of the CCSNprogenitor on pre-explosion images. This work therefore aims to build a database of the photometric- andspectroscopic- spectral energy distributions (phot- and spec-SEDs) of CCSN progenitors. Both photo-SEDsand spec-SEDs are concerned here because they are the main approaches to estimate the properties of CCSNprogenitors. This is the first attempt to give the predicted SEDs of CCSN progenitors.The structure of paper is as follows: in section 2, we introduce the parameter ranges of stars and thecalculation of stellar evolution. Then in section 3, we present the phot- and spec-SEDs of CCSN progenitors.Next, in section 4, we apply the database to some mock progenitors with phot-SEDs. Finally, we concludeand discuss this work in section 5. This work aims to supply an SED database with a large parameter coverage and reasonable resolution, sothe parameter ranges are wider than most previous works. In detail, stellar metallicity ( Z ) covers a rangeof 0.0001–0.03. Stars from metal-poor to metal-rich kinds are included. The zero-age main-sequence massrange of single stars is set to 8–25 M ⊙ , because most CCSN progenitors have main-sequence masses inthis range. This range is similar to some theoretical studies (e.g., Straniero 2018), and larger than someobservational results (8.5—16.5 M ⊙ , e.g., Smartt 2009 and references therein). The same range is set tothe total mass ( M + M ) of two binary components. The range of the mass ratio of secondary to primaryof binaries, q , is set to 0–1. The orbital period ( P ) of a binary changes from 0.1 to 10 days. In fact, withinthe current age of the universe, the evolution of binaries with periods longer than 10 days is similar tothe counterparts with a period of 10 days. The intervals of M , q and log P of main-sequence stars are setto 0.1, 0.1, 0.5, respectively. Two values (0.3 and 0.7) are chosen for the eccentricity ( e ) of binary stars,as previous studies (e.g., Hurley et al. 2002) have shown that e affects the final results somewhat slightly.Although the evolution of massive stars is sensitive to metallicity, binarity, rotation, and possibly magnetic Li & Mao
Table 1: Parameter ranges and steps of zero-age main sequence stars of CCSN progenitors. q , P and e arefor binary stars only. Stellar mass M mean the total main-sequence mass of a single or binary star. Forbinaries, the masses of primary and secondary components are calculated by M = M / (1 + q ) and M = q × M . Parameter Range Step Unit Note Z M M ⊙ single star and binary star q P ) -1.0–4.0 0.5 days binary star e fields, rotation and magnetic field are not taken into account in this work because of their huge uncertainties.Table 1 lists the parameter ranges and steps of zero-age main sequence stars, which are taken by this work.Note that the mass range of this work is similar to most previous studies and findings. For example,Smartt (2009) investigates a mass range of 8–25 M ⊙ . Lisakov et al. (2018) takes 12, 25 and 27 M ⊙ intheir work. Dessart et al. (2010) performed some radiation-hydrodynamic simulations and indicate that theprogenitor main-sequence masses inferior to ∼ M ⊙ , and the range of 25-–30 M ⊙ is not supported bythe narrow width of OI 6303–6363 ˚ A in Type II-P SNe with nebular spectra. Langer (2012) gave the likelyminimum initial mass range of massive stars at solar metallicity as 8–12 M ⊙ , according to Poelarends et al.(2008). In close binaries, this limit depends on other initial system parameters such as metallicity and orbitalperiod. The mass limit at solar metallicity can be as high as 15 M ⊙ (Wellstein et al. 2001). This work models the parameters of CCSN progenitors with a reasonable resolution, via a rapid population-synthesis code, BSE (Hurley et al. 2000, 2002). It takes some fitting formulae based on the reliable stellarevolutionary tracks of Pols et al. (1998). In addition to all aspects of single-star evolution, binary interac-tions including mass transfer, mass accretion, common-envelope evolution, collisions, supernova kicks andangular momentum loss mechanisms have been taken into account by this code, and the calculation result issimilar to some detailed stellar evolution codes (Langer 2012). This code is fast for modeling the populationof a large number of single or binary stars. It is widely used in many stellar population synthesis studies,e.g., Zhang & Li (2006); Han et al. (2007); Li et al. (2012, 2013, 2016); Luo & Li (2018); Li & Mao (2018).It has also been used by some previous works, e.g., Eldridge & Tout (2004) and Eldridge et al. (2008),to reproduce the observed trends such as the distribution of well-studied SN progenitors in the metallicityversus initial mass plane, and the ratio of the Type Ib/c SN rate to the Type II SN rate. The code makes itpossible to cover large ranges of parameters in the studies of CCSN progenitors. Although there are smalluncertainties ( ≤ EDs of CCSN Progenitors 5
Table 2: Input parameters used for stellar evolution.
Parameter or process Symbol Value NoteReimers mass-loss coefficent η α λ M ⊙ dispersion in the Maxwellian for SN kick speed σ β ∝ v the wind accretion efficiency factor xi 1.0Bondi-Hoyle wind accretion factor acc2 1.5fraction of accreted matter retained in nova eruption ǫ nov γ -1.0 When evolving stars, some default values of BSE code are taken for the input parameters, because theyhave been checked by the developer and widely used in different works. They are listed in Table 2 to helpthe readers to understand the physical processes in the evolution of CCSN progenitors. Note that there is animportant difference between single stars and close binary components. Close binary components undergomass transfer following Roche lobe over-follow but single stars do not. Mass transfer can occur betweentwo binary components including different types. White dwarfs are the only degenerate objects able to filltheir Roche lobes for a significant amount of time without breaking up. Thus dynamical mass transfer froma white dwarf can occur in binary evolution. Mass accretion on to degenerate objects is important bothduring Roche lobe overflow and when material is accreted from the wind of the companion. Accretion isassumed to be restricted by the Eddington limit. Two binary components can merge to a single remnant insome cases. Besides nondegenerate stars and white dwarfs, neutron stars and black holes can also merge.This will increase the mass and possibly change the type of the remnants. The angular momentum loss thatis caused by both gravitational radiation and magnetic braking is considered by the BSE code. One can readTout et al. (1997); Pols et al. (1998); Hurley et al. (2002) for more details about the treatment of stellarevolutionary processes.The progenitors of CCSNe are massive stars that evolve very fast. An C-O core or Fe core is finallyformed and its mass grows with stellar evolution, up to the effective Chandrasekhar mass ( ≥ about 1.26 M ⊙ ). Once the core attains this critical mass, unstable gravitational collapse and explosion ensues. Thiswork gives the properties of CCSN progenitors at the moment of the explosion. The CCSN progenitors arefound by comparing the change of star type and the mass of stellar core, which is similar to the methodof Eldridge et al. (2008). The phot- and spec-SEDs, age, mass, effective temperature, luminosity, gravita- Li & Mao tional acceleration, and radius are given for each progenitor. If the progenitor is a binary, orbital period andeccentricity are also given.A standard and widely used stellar spectral library, BaSeL 3.1 (Lejeune et al. 1997, 1998), is usedfor transforming the stellar evolutionary parameters to spec-SEDs when calculating the SEDs of CCSNprogenitors. The SEDs cover a large wavelength from ultraviolet (UV) to medium infrared (MIR), whichis suitable for most multi-band studies. This library is a comprehensive hybrid library of synthetic stellarspectra based on three original grids of model atmosphere spectra. It covers the largest possible ranges instellar parameters (T eff , log g , and [M/H]) and provides flux spectra with useful resolution on an uniformgrid of wavelengths. The standard library has been calibrated and its consistency has been tested carefully. Inparticular, the library spectra was conformed to the empirical color temperature relations, successfully. Afterthe calculation of spec-SEDs, the phot-SEDs are calculated from spec-SEDs, by taking the AB photometrysystem. All magnitudes of progenitors are calibrated using the data of Vega. The properties are given with the same format for single- and binary-star progenitors. A single-star progen-itor is regarded as a binary-star progenitor with a zero-mass component. The case of single-star progenitorsis relatively simple, but it becomes much more complicated when including binary-star evolution. Thelikely important role of binary-star evolution to the formation of SN progenitors remains to be thoroughlyexplored (see e.g., Cantiello et al. 2007). Overall, these complications make the final mass, radius and ageof the CCSN progenitors uncertain (Dessart et al. 2010). For a general use purpose, the age, mass, effectivetemperature, luminosity, gravitational acceleration, radius, and star type of progenitors are included in thefinal database. Here we show the results that are calculated via BSE code (Hurley et al. 2002), as mostworks used this code. However, the similar results are also calculated using an updated version of the code(Spera et al. 2019).Figs. 1–3 show the distribution of CCSN progenitors in various spaces. Fig. 1 shows the progenitordistribution in the initial mass versus final mass plane. We observe some difference between single- andbinary-star progenitors clearly. The final mass of single-star progenitor is lower than about 12 M ⊙ , but thatof binary-star progenitor can be as large as twice as some secondaries accreted masses from their primaries.Note that different stellar evolution codes usually give different final masses. The difference among theresults can be as large as 6 M ⊙ (see the comparison of results of, e.g. Heger et al. 2000; Lisakov 2018;Lisakov et al. 2018).Fig. 2 shows the progenitor distribution in the age versus initial mass plane. We observe that the age ofsingle-star progenitors decreases with increasing initial mass. Meanwhile, the case of binary-star progenitoris much more complicated. In particular, some binary-star progenitors have significantly older ages thanthose single-star progenitors. The reason is that a long time is needed for the mass exchange of thesebinary-star progenitors.Fig. 3 shows the progenitor distribution in the gravity versus effective temperature plane. We see thatmany primaries of binary-star CCSN progenitors locate in the high-temperature ( T eff > K ) area whilethere are much less secondary components in this region. This implies that primaries contribute more to EDs of CCSN Progenitors 7
Fig. 1: Final mass as a function of initial mass for solar-metallicity CCSN progenitors. Green points arefor single-star progenitors, while red points and blue circles for the primary and secondary components ofbinary-star progenitors. Gray triangles are for the total mass of binary-star progenitors.the combined SEDs at short-wavelengths, which is verified by the example of Fig. 4. That figure givesthe contributions of two components of binary-star progenitors to the combined SED. From the figure wealso find that both two components of binary-star progenitors contribute to the combined SEDs at long-wavelengths.When comparing the final masses of CCSN progenitors to the results of Lisakov (2018) and Dessartet al. (2010), the results of this work (calculated via BSE) are found to consistent with those calculated viaMESA code (Paxton et al. 2011). Table 3 shows the results of different works.
This section presents the spec-SEDs of both single- and binary-star CCSN progenitors. Because thedatabase is as large as 1.4 GB, only some examples are shown here. One can see Figs. 5–7 for the spec-SEDsof a few example progenitors with metallicities of 0.001, 0.02 (solar metallicity) and 0.03, and total main-sequence masses of 9.1 M ⊙ (solid lines) and 20 M ⊙ (dashed lines). We read that there is obvious differencebetween the SEDs of single- and binary-star progenitors, even though they have the same metallicity andtotal mass. This suggests that different results will be possibly obtained when fitting to the observed SEDsusing single- and binary-star progenitor models. Li & Mao
Fig. 2: Age as a function of initial mass for solar-metallicity CCSN progenitors. Colors have the samemeanings as in Fig. 1.Table 3: Comparison of the final mass and age of CCSN progenitors of this work to two previous works(Lisakov 2018; Dessart et al. 2010). Subscripts ‘BSE’ corresponds to the results of this work, while ‘MESA’and ‘Woosely’ denote the results that are calculated via MESA (Paxton et al. 2011) and Woosley et al. (2002)codes, respectively. All models have the metallicity of Z = 0.02. M init is initial main sequence mass. M init M BSE M MESA M Woosley
Age
BSE
Age
MESA M ⊙ M ⊙ M ⊙ M ⊙ Myr Myr13.0 11.5 11.1 17.7 15.315.0 10.0 11.9 12.64 14.3 12.517.0 9.9 14.2 12.0 10.719.0 10.1 13.6 10.5 9.421.0 7.3 8.6 9.3 8.523.0 8.1 8.1 8.5 7.725.0 8.3 8.6 12.53 7.8 7.2
This section shows the phot-SEDs of CCSN progenitors. Such SEDs are usually more useful for the studiesof CCSN progenitors. All phot-SEDs are calculated from the spec-SEDs. The AB system are adopted,for a purpose of wide applications. As a result, the AB magnitudes in
F U V , N U V , u , g , r , i , z , J , H , EDs of CCSN Progenitors 9
Fig. 3: Distribution of CCSN progenitors in the plane of gravity (log g ) versus effective temperature ( T eff ). T eff is in K . Red filled and black open circles are for the primaries and secondaries of binary-star CCSNprogenitors, respectively. Fig. 4: Contributions of two components to the combined SED of a binary-star CCSN progenitor. Towcomponents have the solar metallicity, and initial masses of 11.54 M ⊙ and 3.46 M ⊙ . Green and black linesare for the SEDs of primary and secondary components respectively, while red line is for the combinedSED. Ks , W , W , and W bands are calculated. This makes it possible to study the phot-SEDs of CCSNprogenitors in a wide wavelength range. Figs. 8–10 show some examples of the phot-SEDs of single-and binary-star progenitors. Fig. 8 shows the phot-SEDs of single-star progenitors, for eight metallicitiesfrom 0.0001 to 0.03. It is shown that single-star progenitors with various masses and metallicities usuallyhave different phot-SED shapes. Some massive progenitors with metallicity poorer than 0.001 have UV-upturn phot-SEDs. However, there is obvious overlap for the phot-SEDs of single-star progenitors. Thisimplies that the metllicity and main-sequence mass of such progenitors can be determined via fitting to theobserved SEDs, but the uncertainties of the results of some progenitors will be possibly large. This agreeswith previous studies on SNe such as SN1987A. Similarly, Figs. 9 and 10 show the phot-SEDs of someexample binary-star CCSN progenitors. This section applies the database to some mock CCSN progenitors. The phot-SEDs of mock progenitorsare fitted using the database. Each phot-SED consists of the magnitudes in
F U V , N U V , u , g , r , i , z , J , H , Ks , W , W , and W bands. It is found that the main-sequence mass, age and metallicity of mostsingle-star progenitors can be reproduced as a whole, although a few progenitors are not reproduced well EDs of CCSN Progenitors 11
Fig. 5: Spec-SEDs of example CCSN progenitors. The metallicity Z is 0.001. Solid and dashed lines arefor 9.1 M ⊙ and 20 M ⊙ models respectively. “single” denotes single-star progenitors, while orbital periodnumbers denote binary-star progenitors. Blue, green, red, and purple lines are for orbital periods of 3162,100, 10, and 10000 d, while black line is for single-star progenitors.Table 4: CCSN progenitor models for Figs. 9 and 10. “No.” means the line number in two figures. m and m are in M ⊙ , and P is in days. No. m m P e
No. m m P e
Fig. 9 Fig. 101 8.18 0.82 3162 0.3 1 8.18 0.82 10000 0.32 8.18 0.82 10000 0.3 2 16.36 1.64 10000 0.33 7.50 1.50 1000 0.3 3 19.09 1.91 3162 0.34 10.91 1.09 3162 0.3 4 17.50 3.50 316 0.35 13.64 1.36 3162 0.3 5 20.00 2.00 3162 0.36 16.36 1.64 3 0.7 6 8.57 3.43 3162 0.37 16.36 1.64 3162 0.3 7 9.23 2.77 3162 0.38 10.91 1.09 10000 0.39 10.00 2.00 3162 0.310 9.23 2.77 3162 0.311 13.64 1.36 10000 0.312 12.50 2.50 10000 0.32 Li & Mao
Fig. 6: Similar to Fig. 5, but for a solar metallicity of Z = 0.02.because of the metallicity and mass degeneracy. As an example, Fig. 11 shows the comparison of input andreproduced masses of single-star progenitors.However, the main-sequence masses of most binary-star progenitors are not reproduced correctly, ifall parameters are free in the SED fitting (Fig. 12). The fitted main-sequence masses of most binary-starprogenitors are much lower than the real values. This is caused by the degeneracy among mass, metallicityand orbital period. In order to find a reliable method for determining the main-sequence masses of binary-star progenitors, the cases of fixed metallicity or fixed period are tested, but the uncertainties in resultsare still large. Finally, the case of fixed metallicity and period gives satisfactory results (see Fig. 13). Thismeans that if one wants to determine the masses of bianry-star progenitors reliably, the metallicity andorbital period (initial or final values) are suggested to be determined in advance. This paper presents a new database of SEDs of the single- and binary-star CCSN progenitors. Both the phot-and spec-SEDs of progenitors are included in the database. The database covers wide ranges of metallicity(0.0001–0.03), main-sequence mass (8–25 M ⊙ ), component mass ratio (0–1), binary period (0.1–10 days),and two eccentricities (0.3 and 0.7). It is then applied to the phot-SEDs of some mock CCSN progenitors.Our investigation leads to the following conclusions: – The database of spec- and phot-SEDs of CCSN progenitors can be used for the studies of progenitorproperties, and the difference between binary- and single-star progenitors. The results are consistent
EDs of CCSN Progenitors 13
Fig. 7: Similar to Fig. 5, but for a metallicity of Z = 0.03.with those calculated via MESA code (Paxton et al. 2011), but the database is model dependent. Thusit is better for statistical studies such as population synthesis. It can be potentially used for the identi-fication of CCSN progenitors in large surveys. Stellar evolutionary code can affect the results, but therelative results are usually similar. For example, when we use the code of Spera et al. (2019) to calculatethe SEDs instead of Hurley et al. (2002), similar results are shown (see Fig. 14). – Binary-star CCSN progenitors have much more complicated parameter spaces than single-star pro-genitors. It leads to much larger uncertainties in the determination of progenitor properties includingcomponent masses, total mass, metallicity and period. – Binaries with component masses less massive than 8 M ⊙ can form CCSN progenitors, although singlestars less massive than this value cannot lead to CCSN. – When the SED database is used for determining the properties of CCSN progenitors, whether the pro-genitor is single or binary star affects the result accuracy significantly. If progenitors are single stars,the initial and final mass can be determined well for most progenitors via phot-SEDs from FUV to W3bands. However, the results will be not reliable for binary-star progenitors, if metallicity, mass, and pe-riod are set as free parameters of fit. In order to get reliable results, the metallicity and binary period(initial or final periods) are needed to be measured using other methods. If these two parameters areknown, the masses of binary components can be determined well via SED fitting.
Fig. 8: Example phot-SEDs of single-star CCSN progenitors. Black, red, green, blue, cyan, purple, yellow,and orange colors are for Z = 0.0001, 0.0003, 0.001, 0.004, 0.008, 0.01, 0.02, 0.03, respectively. Lines withthe same color but different shapes are for various masses. Acknowledgements
The authors thank Prof. Xiaofeng Wang for suggestions and Dr. Jicheng Zhang fordiscussions. This work has been supported by the Chinese National Science Foundation (No. 11863002),Sino-German Cooperation Project (No. GZ 1284), and Yunnan Academician Workstation of Wang Jingxiu(No. 201905F150106).
References
Belczynski, K., Kalogera, V., Rasio, F. A., et al. 2008, ApJS, 174, 223 2Blondin, S., & Tonry, J. L. 2007, ApJ, 666, 1024 3Cantiello, M., Yoon, S. C., Langer, N., & Livio, M. 2007, A&A, 465, L29 6Dessart, L., Livne, E., & Waldman, R. 2010, MNRAS, 408, 827 4, 6, 7, 8Eldridge, J. J., Izzard, R. G., & Tout, C. A. 2008, MNRAS, 384, 1109 2, 4, 5Eldridge, J. J., Stanway, E. R., Xiao, L., et al. 2017, PASA, 34, e058 2Eldridge, J. J., & Tout, C. A. 2004, MNRAS, 353, 87 2, 4Farrell, E. J., Groh, J. H., Meynet, G., & Eldridge, J. J. 2020, MNRAS, 494, L53 2Han, Z., Podsiadlowski, P., & Lynas-Gray, A. E. 2007, MNRAS, 380, 1098 2, 4Heger, A., Langer, N., & Woosley, S. E. 2000, ApJ, 528, 368 3, 6Humphreys, R. M., & Davidson, K. 1994, PASP, 106, 1025 2
EDs of CCSN Progenitors 15
Fig. 9: Example phot-SEDs of binary-star CCSN progenitors with metallicities of 0.0001 and 0.03. Grayarea indicates the range of all phot-SEDs of all binary-star progenitors in the database. The detailed modelparameters of these CCSN progenitors are listed in Table 4.Hurley, J. R., Pols, O. R., & Tout, C. A. 2000, MNRAS, 315, 543 2, 4Hurley, J. R., Tout, C. A., & Pols, O. R. 2002, MNRAS, 329, 897 3, 4, 5, 6, 13Langer, N. 2012, ARA&A, 50, 107 3, 4Lejeune, T., Cuisinier, F., & Buser, R. 1997, A&AS, 125, 229 6—. 1998, A&AS, 130, 65 6Li, Z., & Han, Z. 2008, ApJ, 685, 225 2Li, Z., & Mao, C. 2018, ApJ, 859, 36 4Li, Z., Mao, C., Chen, L., & Zhang, Q. 2012, ApJ, 761, L22 4Li, Z., Mao, C., Chen, L., Zhang, Q., & Li, M. 2013, ApJ, 776, 37 4Li, Z., Mao, C., Zhang, L., Zhang, X., & Chen, L. 2016, ApJS, 225 4Lisakov, S. 2018, Theses, Universit´e Cˆote d’Azur. https://tel.archives-ouvertes.fr/tel-02018238
2, 7, 8Lisakov, S. 2018, PhD thesis, Laboratoire Lagrange, UMR7293, Universit´e Nice Sophia-Antipolis,Observatoire de la Cˆote d’Azur, F-06304 Nice, France 6Lisakov, S. M., Dessart, L., Hillier, D. J., Waldman, R., & Livne, E. 2018, MNRAS, 473, 3863 2, 4, 6Luo, Q., & Li, Z. 2018, New Astron., 64, 61 4Meynet, G., & Maeder, A. 2007, A&A, 464, L11 3
Fig. 10: Similar to Fig. 9, but for a metallicity of Z = 0.02.Paxton, B., Bildsten, L., Dotter, A., et al. 2011, ApJS, 192, 3 2, 3, 7, 8, 13Podsiadlowski, P., Joss, P. C., & Hsu, J. J. L. 1992, ApJ, 391, 246 2Podsiadlowski, P., Rappaport, S., & Pfahl, E. D. 2002, ApJ, 565, 1107 2Poelarends, A. J. T., Herwig, F., Langer, N., & Heger, A. 2008, ApJ, 675, 614 2, 4Pols, Onno R., Schr¨oder, Klaus-Peter, Hurley, Jarrod R., Tout, Christopher A., Eggleton, Peter P. 1998,MNRAS, 298, 525 4, 5Powell, J., & M¨uller, B. 2020, MNRAS, 494, 4665 3Ruiz-Lapuente, P., Comeron, F., M´endez, J., et al. 2004, Nature, 431, 1069 2Smartt, S. J. 2009, ARA&A, 47, 63 1, 2, 3, 4Spera, M., Mapelli, M., Giacobbo, N., et al. 2019, MNRAS, 485, 889 6, 13, 20Straniero, O. 2018, European Physical Journal Plus, 133, 388 3Tout, Christopher A., Aarseth, Sverre J., Pols, Onno R., Eggleton, Peter P. 1997, MNRAS, 291, 732 5Umeda, H., & Nomoto, K. 2008, ApJ, 673, 1014 3Waldman, R. 2008, ApJ, 685, 1103 2Wellstein, S., Langer, N., & Braun, H. 2001, A&A, 369, 939 4Woosley, S. E., Heger, A., & Weaver, T. A. 2002, Reviews of Modern Physics, 74, 1015 2, 8Woosley, S., & Janka, T. 2005, Nature Physics, 1, 147 3Woosley, S. E., Blinnikov, S., & Heger, A. 2007, Nature, 450, 390 3Woosley, S. E., Heger, A., & Weaver, T. A. 2002, Reviews of Modern Physics, 74, 1015 2, 8 EDs of CCSN Progenitors 17
Fig. 11: Comparison of input ( m mock ) and reproduced ( m fit ) masses of single-star progenitors in phot-SEDfitting. Different symbols denote different metallicities.Yoon, S.-C., Dessart, L., & Clocchiatti, A. 2017, ApJ, 840, 10 3Zapartas, E., de Mink, S. E., Justham, S., et al. 2020, arXiv e-prints, arXiv:2002.07230 2Zhang, F., & Li, L. 2006, MNRAS, 370, 1181 2, 4 Fig. 12: Comparison of input ( m mock ) and reproduced ( m fit ) masses of binary-star progenitors in phot-SED fitting. The result is for the case of free metallicity and orbital period. Filled circle, open circle andpentagram are for primary mass, secondary mass and total mass respectively. EDs of CCSN Progenitors 19
Fig. 13: Similar to Fig. 11, but for binary-star progenitors with known metallicity and orbital period.
Fig. 14: Similar to Fig. 5, but for another stellar evolution calculation (Spera et al. 2019). Solid and dashedlines are for total stellar masses of 9.1 and 20 M ⊙⊙