Excluding super-soft X-ray sources as progenitors for four Type Ia supernovae in the Large Magellanic Cloud
J. Kuuttila, M. Gilfanov, I. R. Seitenzahl, T. E. Woods, F. P. A. Vogt
MMNRAS , 1–9 (2018) Preprint 24 December 2018 Compiled using MNRAS L A TEX style file v3.0
Excluding super-soft X-ray sources as progenitors for fourType Ia supernovae in the Large Magellanic Cloud
J. Kuuttila , M. Gilfanov , , I. R. Seitenzahl , , T. E. Woods and F. P. A. Vogt , Max Planck Institute for Astrophysics, Karl-Schwarzschild-Str. 1, Garching b. M¨unchen 85741, Germany Space Research Institute, Profsoyuznaya 84/32, 117997, Moscow, Russia School of Science, University of New South Wales, Australian Defence Force Academy, Canberra, ACT 2600, Australia Research School of Astronomy and Astrophysics, Australian National University, Canberra, ACT 2611, Australia Institute for Gravitational Wave Astronomy and School of Physics and Astronomy, University of Birmingham,Birmingham B15 2TT, United Kingdom European Southern Observatory, Av. Alonso de C´ordova 3107, 763 0355 Vitacura, Santiago, Chile ESO Fellow
Accepted XXX. Received YYY; in original form ZZZ
ABSTRACT
Type Ia supernovae are vital to our understanding of the Universe due to their usein measuring cosmological distances and their significance in enriching the interstellarmedium with heavy elements. They are understood to be the thermonuclear explo-sions of white dwarfs, but the exact mechanism(s) leading to these explosions remainsunclear. The two competing models are the single degenerate scenario, wherein awhite dwarf accretes material from a companion star and explodes when it reachesthe Chandrasekhar limit, and the double degenerate scenario, wherein the explosionresults from a merger of two white dwarfs. Here we report results which rule out hot,luminous progenitors consistent with the single degenerate scenario for four youngType Ia supernova remnants in the Large Magellanic Cloud. Using the integral fieldspectrograph WiFeS, we have searched these remnants for relic nebulae ionized by theprogenitor, which would persist for up to ∼ years after the explosion. We detectedno such nebula around any of the remnants. By comparing our upper limits withphotoionization simulations performed using Cloudy, we have placed stringent upperlimits on the luminosities of the progenitors of these supernova remnants. Our resultsadd to the growing evidence disfavouring the single degenerate scenario. Key words:
ISM: supernova remnants – supernovae: general – white dwarfs – X-rays:binaries
Type Ia supernovae (SNe Ia) are runaway thermonuclear ex-plosions of carbon-oxygen white dwarfs (CO WDs) (see e.g.Hillebrandt et al. 2013; Maoz et al. 2014, for reviews). SNeIa can be used for distance measurements on cosmic scalesdue to a correlation between their peak luminosity, the rateof decline after maximum light, and the colour at maximum.This so-called Phillips relation (Phillips 1993; Phillips et al.1999) has been used to show the acceleration of the Uni-verse’s expansion (Riess et al. 1998; Perlmutter et al. 1999).SNe Ia are also important to the chemical evolution of galax-ies, since a typical SN Ia enriches the interstellar medium(ISM) with ∼ (cid:12) of iron and a similar amount of otherelements (Matteucci & Greggio 1986; Wiersma et al. 2011).Despite their significance to our understanding of theUniverse, the formation channel for SNe Ia is still uncertain.The two leading models are the single degenerate (SD) sce- nario, in which a single WD reaches the critical carbon igni-tion density through accretion from either a main-sequenceor an evolved companion star (Whelan & Iben 1973); and thedouble degenerate scenario, in which the explosion resultsfrom the merger of a binary pair of WDs (Iben & Tutukov1984).In the double degenerate scenario, the progenitor sys-tem is typically too faint to be detectable with current in-struments prior to the explosion. In the single degeneratescenario, however, progenitor systems should be detectableboth before and, most importantly, after the explosion. Inparticular, if steady nuclear burning of hydrogen occurs onthe surface of a WD, as expected from the single degeneratechannel, then in the most efficient regime for mass accu-mulation the accreting WD should reveal itself as a strongsuper-soft X-ray source (SSS) (van den Heuvel et al. 1992;Kahabka & van den Heuvel 1997).The first of these SSSs were discovered in the Large © a r X i v : . [ a s t r o - ph . H E ] D ec J. Kuuttila et al.
Magellanic Cloud (LMC) using the Einstein Observatory(HEAO-2, Long et al. 1981). SSSs are typically luminous( L bol (cid:38) − erg s − ) and characterized by effective tem-peratures of T eff ∼ K (Greiner 2000). Due to their hightemperatures and luminosities, SSSs emit significant fluxesof UV and soft X-ray photons, which will strongly ionizeany surrounding interstellar gas. This in turn will create acharacteristic ionization nebula with a typical (“Str¨omgren”)radius of R S = (cid:32) π N ph n α (cid:33) ≈ (cid:18) N ph s − (cid:19) (cid:18) n ISM − (cid:19) − , (1)where α is the recombination coefficient, N ph is the numberof ionizing photons per second, and n ISM is the number den-sity of the ISM (Rappaport et al. 1994; Woods & Gilfanov2016). In addition to the direct emission from the centralSSS, significant ionization nebulae can be also produced bythe emission from the accretion disc when the accretion rateis too low for the hydrogen fusion to be ignited (Woods et al.2017).To date, only one such emission-line nebula has beendetected, surrounding a SSS in the LMC known as CAL 83(Remillard et al. 1995). These authors, in fact, conductedalso imaging observations of nine other known SSSs in theLMC and SMC, but did not detect nebulae around them.Later, Gruyters et al. (2012) observed a part of the nebulaaround CAL 83 using the VLT/VIMOS and detected for thefirst time also the He ii ii (54.4 eV) requires a hot ( (cid:38) K)ionizing source, making the He ii emission an easily recogniz-able signal of an accreting WD with steady nuclear burningon its surface. Models for SSS nebulae indeed predict thatthese nebulae should be bright in He ii iii ]5007˚A , making them distinct from other astrophysical neb-ulae (Rappaport et al. 1994; Woods & Gilfanov 2016). Forthis reason, the He ii ii emission of galaxies to the expectedemission from population synthesis models (Woods & Gil-fanov 2013; Johansson et al. 2014).If a single degenerate SN Ia progenitor spends a signif-icant amount of time as a SSS prior to the explosion, thenthe ionized nebula should remain detectable after the WD isdestroyed in the explosion, until the majority of the ionizedgas has recombined. If the gas in the nebula is initially fullyionized (i.e. n e ≈ n ISM ), the typical hydrogen recombinationtime can be estimated as τ rec = (cid:16) n e α B ( H , T ≈ K ) (cid:17) − ≈ × (cid:18) n ISM − (cid:19) − years , (2)where α B ( H , T ) is the Case B recombination coefficient(Woods & Gilfanov 2016; Woods et al. 2017). For heliumthe corresponding time-scale is ≈ × years (see e.g. Pe-quignot et al. 1991). Searching for these relic nebulae aroundyoung supernova remnants and determining the ionizationstate of the surrounding gas can thus be used effectively toconstrain the properties of the progenitor.Previous efforts in determining the ionization state ofthe gas around SNRs have focused on the forward shocks. Many shock fronts around SNe Ia remnants are so-calledBalmer-dominated shock fronts, where the optical emissionis dominated by both broad and narrow Balmer line emis-sion. This emission is understood to be the result of theshock interacting with the surrounding neutral hydrogen,and can thus be used in estimating the ionized fraction ofhydrogen (Ghavamian et al. 2000, 2001, 2003). This methodhas recently been used to place stringent upper limits onthe luminosity of the progenitors for SNe Ia remnants in theGalaxy and LMC (Woods et al. 2017, 2018). Any methodused to determine the ionization state of the gas is very sen-sitive to the density of the gas, but the expanding shocks canalso be used to determine the density of the surrounding gas(Badenes et al. 2007; Yamaguchi et al. 2014).Here we report a different method for constraining thenature of SNe Ia progenitors. We have searched directlyfor the relic ionization nebulae around four known SNe Iaremnants in the LMC (see Table 1 for list of sources). Us-ing integral field spectroscopy, we searched for and did notfind any He ii ∼ years before the explosions.This paper is organized as follows: In Sec. 2 we describeour observations and the data reduction procedure. In Sec. 3we describe the methods used in this paper, specifically thespectral extraction (Sec. 3.1) and the Cloudy simulations(Sec. 3.2). In Sec. 4 we describe our results and then discussthem and the possible implications in Sec. 5. We have observed the four LMC SN Ia remnants SNR 0509-67.5, SNR 0505-67.9, SNR 0509-68.7, and SNR 0519-69.0with the Wide Field Spectrograph (WiFeS) mounted on theNasmyth A focus of the Australian National University 2.3 mtelescope at the Siding Spring Observatory (Dopita et al.2007, 2010). SNR 0505-67.9, SNR 0509-68.7, and SNR 0519-69.0 were observed on the nights of 2014 December 18–20(P.I.: Seitenzahl; Proposal ID: 4140118) and SNR 0509-67.5was observed on 2015 December 13 (P.I.: Seitenzahl; Pro-posal ID: 4150145). Here we provide only a short summaryof the data reduction method, which is also described indetail by Dopita et al. (2016) and Ghavamian et al. (2017).The observations were performed in the ‘binned mode’,which provided us a field of view of 25 ×
35 spatial pixels (orspaxels), each of them (cid:48)(cid:48) × (cid:48)(cid:48) in angular size. The instrumentis a double-beam spectrograph providing simultaneous andindependent channels for both the blue and red wavelengthranges. We used the B3000 and R7000 gratings, providinga spectral resolution of R = 3000 ( ∆ v ≈
100 km s − ) in theblue wavelength range (3500–5700 ˚A) and R = 7000 ( ∆ v ≈
45 km s − ) in the red (5300–7000 ˚A).SNR 0509-67.5, SNR 0509-68.7, and SNR 0519-69.0were observed in a mosaic of two overlapping fields, and SNR0505-67.9 was observed with ten fields in order to cover thewhole remnant. Each field was observed in × s expo- MNRAS , 1–9 (2018) xcluding SSSs as SNe Ia progenitors sures, with × s blank sky exposures, which were sub-tracted from the two co-added frames for each field.The data were reduced with the PYWIFES v0.7.3pipeline (Childress et al. 2014a,b), which provided us a wave-length calibrated, sensitivity corrected, and photometricallycalibrated data cube. The final mosaics were then combinedfrom the individually reduced cubes, with the respectivealignment of each field in the mosaic derived by comparingthe reconstructed continuum frames from the red cubes withthe Digitized Sky Survey 2 red band image of the area. Thefinal mosaic for SNR 0505-67.9 has dimensions of (cid:48)(cid:48) × (cid:48)(cid:48) ,and for the three other sources the dimensions are (cid:48)(cid:48) × (cid:48)(cid:48) ,which correspond to fields of . × . pc and . × . pc,respectively, assuming a distance of 50 kpc to the LMC. In order to study the properties of the possible nebulaearound the observed supernova remnants with as high sen-sitivity as possible, we extracted spectra from large areassurrounding each remnant. Since the expected brightnessof an emission line decreases as a function of distance (seeFig. 3), we used the area between the outer edge of theforward shock ( ∼ pc from the centre) and a distance ofabout 5 pc from the approximate geometrical centre of eachsource for SNR 0509-67.5, SNR 0509-68.7, and SNR 0519-69.0; for SNR 0505-67.9 the corresponding values are 7–9pc and 10 pc. We avoided any areas with residuals fromforeground star subtraction. As an illustrative example, seeFig. 1, where we show the SNR 0519-69.0 remnant with thespectral extraction area marked.For each source, the spectra were averaged over thespecified area, corrected for the average redshift of 277.5km s − , which was measured from the H β and [O iii ] 4959,5007 ˚A emission lines (LMC peculiar velocity is 262.2 kms − ; McConnachie 2012), and dereddened using the averageLMC extinction curves of Weingartner & Draine (2001) witha carbon abundance b c = × − and using the H columndensities for each source listed in table 1. An example spec-trum is shown in Fig. 2.No He ii ii β , were inside thewindow.Then, because the nebular He ii emission is expectedto be narrower than the instrumental resolution and wouldthus be spectrally unresolved (Gruyters et al. 2012), we as-sumed a gaussian line with a fixed width corresponding tothe instrumental resolution of B3000 ( ∼ km s − ), andusing a chi-squared test we calculated the minimum ampli-tude of a line, which would be statistically separable fromthe estimated noise level. The line was taken to be distinctfrom the noise, when adding the line on the spectrum in-creased the χ value by 9, corresponding to σ , or 99.7 % confidence. The flux of such a gaussian line was then takento be the upper limit of the possible He ii We computed a grid of numerical photoionization modelswith Cloudy (v17.01; Ferland et al. 2013). We assumed aspherically symmetric and static configuration with the cen-tral ionizing source emitting a blackbody spectrum, whichprovides a reasonable approximation of the ionizing emis-sion of nuclear-burning WDs, except far into the Wien tail(Chen et al. 2015; Woods & Gilfanov 2016). The effectivetemperature of the ionizing radiation was varied from to K and the bolometric luminosity from to ergs − with logarithmically evenly spaced steps. In light of thepre-shock densities of the remnants shown in Table 1, thedensity of the ambient gas was kept fixed at either 0.5, 1, or2.4 cm − , while dust was neglected. The metallicity of thegas was set to Z = . Z (cid:12) , where Z (cid:12) is the solar metallicity,based on the average results of several studies on the metal-licity of the ISM around many LMC SNRs, including forexample SNR 0505-67.9 and SNR 0519-69.0 (Hughes et al.1998; Maggi et al. 2016; Schenck et al. 2016).The calculations were performed in three different wayswith regard to the gas temperature: in the first case, the am-bient gas temperature was calculated self-consistently andthe calculations were stopped when the gas temperaturedropped below 3000 K. While this is an idealized assump-tion, this case offers a possibility to study as an examplean isolated situation, where the only source of energy is thecentral ionizing source. In reality, there is a diffuse emissionfield originating from stars and other sources in addition tothe central ionizing source. Depending on the strength ofthe diffuse emission field, and properties of the gas, suchas density, the ISM has been historically classified roughlyinto three different phases: a hot and very low density phase( n ∼ − . cm − , T ∼ K), a warm low density phase( n ∼ − . cm − , T ∼ K), and a cold dense phase( n ∼ . cm − , T ∼ K) (McKee & Ostriker 1977). In thehot phase the gas is already ionized and any possible SSSwould not then change the ionization state of the ISM. Inthe cold phase the central SSS would be the main source ofenergy and an ionization nebula would be clearly detectable.This phase corresponds mostly to the self-consistent temper-ature calculations. However, the estimated gas density limitsof the SNRs studied here point mostly to the warm phase.Thus, to include the contribution from the diffuse emissionin our simulations, we ran the calculations with a fixed gastemperature in addition to calculating it self-consistently.Although the relatively low temperatures of the warm lowdensity phase are not expected to contribute significantlyto the ionization of He + due to its high ionization potential(54.4 eV), we ran the calculations with the temperature setto either 5000 K or 10000 K in order to test the effect of thegas temperature on the He ii (cid:15) i ( r ) of a line i as a function of the distance r from , 1–9 (2018) J. Kuuttila et al.
Table 1.
List of observed sources with relevant properties.Source Size (pc) Age (yrs) n (cm − ) N H (10 cm − ) a ReferencesSNR 0509-67.5 4 400 ±
120 0.4–0.6 1.64 ± ∼ ± ±
20 1–2.5 3.09 + . − . van der Heyden et al. (2002); Williams et al. (2014)SNR 0519-69.0 4 680 ±
200 2.4 ± ± DEM L71 N103B a Maggi et al. (2016)
5h 09m 33s 31s 29s67 31 D e c . ( J )
5h 05m 48s 43s 38s 33s67 52
08m 57s 08m 59s 09m 01s5h 09m 03s
R.A. (J2000)
68 43 D e c . ( J )
5h 19m 38s 36s 34s 32s
R.A. (J2000)
69 01 Figure 1.
All supernova remnants in H α with WiFeS. Top row from left to right: SNR 0509-67.5 and SNR 0505-67.9; bottom row fromleft to right: SNR 0509-68.7 and SNR 0519-69.0. The spectral extraction areas are outlined with the black dashed lines. the ionizing source. This can be used to find the surfacebrightness of a line i: SB i ( r ) = ∫ l (cid:15) i ( r ) π d l , (3)where we have integrated along the line of sight l through the emission nebula. Examples of the He ii − K and luminosities 10 − erg s − areshown in Fig. 3.From these surface brightness profiles, we calculated theaverage surface brightness of the He ii MNRAS , 1–9 (2018) xcluding SSSs as SNe Ia progenitors S u r f a c e b r i g h t n e ss ( e r g s c m a r c s e c ) [OIII]HHeII HeII
Figure 2.
An example spectrum of the interstellar gas ahead ofthe forward shock around SNR 0519-69.0. On the y-axis is themean surface brightness in units 10 − erg s − cm − arcsec − andon the x-axis is the wavelength in units of ˚A. The inset figureshows the spectrum in more detail around the 4686 ˚A wavelength.In blue are marked the brightest emission lines H β and [O iii ] ,and the red dashed lines indicate the 4686˚A wavelength. Radius (pc) S u r f a c e b r i g h t n e ss ( e r g s c m a r c s e c ) K, 10 erg s K, 10 erg s K, 10 erg s K, 10 erg s Figure 3.
Surface brightness (in units of erg s − cm − arcsec − )profiles for He ii and 10 K, andthe solid and dashed lines have luminosities of 10 and 10 ergs − , respectively. The density of the gas was set to 1 cm − andthe gas temperature was calculated self-consistently. in the data extraction range of 4–5 pc (or 7–10 pc for SNR0505-67.9) for each point in the temperature–luminositygrids. Then, comparing the upper limits acquired from theWiFeS observations to the grids of simulated brightnesses,we can constrain the luminosity as a function of the assumedemission temperature of the central ionizing sources. The σ upper limits for the surface brightness of the He ii ii emission was detected ahead of the forwardshock in any source and the derived upper limits are within Table 2. σ upper limits on the He ii × − erg s − cm − arcsec − )SNR 0509-67.5 4.2SNR 0505-67.9 4.7SNR 0509-68.7 5.7SNR 0519-69.0 5.3 a factor of two from each other. How well the progenitorproperties can be constrained, depends, however, on the sizeof the remnant and the density of the surrounding gas. SNR0505-67.9 is much older and thus much larger than the otherthree SNRs. As shown in Fig. 3, the expected surface bright-ness decreases with the distance from the ionizing source,making it harder to constrain the progenitor luminosities ofolder and larger SNRs.To transform the surface brightness upper limits to lim-its on the progenitor luminosities, we compared the resultsto the Cloudy simulations, as explained in Sec. 3.2. The up-per limits on the bolometric luminosity as a function of theassumed emission colour temperature for each source areshown in Fig. 4. In this figure, the parameter space aboveeach line is ruled out, and the area below is unconstrained.Here the temperature is calculated self-consistently and thegas density is set to 1 cm − . The limits of the three youngand small remnants are all almost the same. SNR 0505-67.9deviates from the others mostly because of its greater size;the surface brightness around SNR 0505-67.9 is studied at adistance of ≈ (cid:12) to 1.4 M (cid:12) . All of these mod-els lie well above the derived upper limits for all the SNRsstudied in this paper. In Fig. 4 are also shown the parameterranges for four well-known super-soft X-ray sources locatedin the Magellanic clouds: 1. CAL 87 (LMC); 2. 1E 0035.4-7230 (SMC); 3. RX J0513.9-6951 (LMC); and 4. CAL 83(LMC) (Greiner 2000). All of these four SSSs lie in the ruled-out regions of the three young sources, with the latter threeSSSs having similar temperatures and luminosities as thenuclear-burning WD models. The upper limit of the largestremnant, SNR 0505-67.9, overlaps with the parameter rangeof CAL 87, but one should note that CAL 87 (number 1 inFig. 4), which has the lowest claimed luminosity of the four,is viewed almost edge-on, meaning that its unobscured lu-minosity is likely much higher (Ness et al. 2013).In Fig. 4, the results for each source are shown withthe gas density of the simulations set to 1 cm − , but asmentioned before, the results are affected by the assumedgas density of the simulations. To test this effect, we ran thesimulations also with the gas density set either to 0.5 or 2.4cm − , which correspond to the upper limits of the densityaround SNR 0509-67.5 and SNR 0519-69.0 (see Table. 1),respectively. The effect of the density on the results is shownin Fig. 5. As is evident from this figure, the highest densityprovides the least constraining limits, while the low and mid MNRAS000
Surface brightness (in units of erg s − cm − arcsec − )profiles for He ii and 10 K, andthe solid and dashed lines have luminosities of 10 and 10 ergs − , respectively. The density of the gas was set to 1 cm − andthe gas temperature was calculated self-consistently. in the data extraction range of 4–5 pc (or 7–10 pc for SNR0505-67.9) for each point in the temperature–luminositygrids. Then, comparing the upper limits acquired from theWiFeS observations to the grids of simulated brightnesses,we can constrain the luminosity as a function of the assumedemission temperature of the central ionizing sources. The σ upper limits for the surface brightness of the He ii ii emission was detected ahead of the forwardshock in any source and the derived upper limits are within Table 2. σ upper limits on the He ii × − erg s − cm − arcsec − )SNR 0509-67.5 4.2SNR 0505-67.9 4.7SNR 0509-68.7 5.7SNR 0519-69.0 5.3 a factor of two from each other. How well the progenitorproperties can be constrained, depends, however, on the sizeof the remnant and the density of the surrounding gas. SNR0505-67.9 is much older and thus much larger than the otherthree SNRs. As shown in Fig. 3, the expected surface bright-ness decreases with the distance from the ionizing source,making it harder to constrain the progenitor luminosities ofolder and larger SNRs.To transform the surface brightness upper limits to lim-its on the progenitor luminosities, we compared the resultsto the Cloudy simulations, as explained in Sec. 3.2. The up-per limits on the bolometric luminosity as a function of theassumed emission colour temperature for each source areshown in Fig. 4. In this figure, the parameter space aboveeach line is ruled out, and the area below is unconstrained.Here the temperature is calculated self-consistently and thegas density is set to 1 cm − . The limits of the three youngand small remnants are all almost the same. SNR 0505-67.9deviates from the others mostly because of its greater size;the surface brightness around SNR 0505-67.9 is studied at adistance of ≈ (cid:12) to 1.4 M (cid:12) . All of these mod-els lie well above the derived upper limits for all the SNRsstudied in this paper. In Fig. 4 are also shown the parameterranges for four well-known super-soft X-ray sources locatedin the Magellanic clouds: 1. CAL 87 (LMC); 2. 1E 0035.4-7230 (SMC); 3. RX J0513.9-6951 (LMC); and 4. CAL 83(LMC) (Greiner 2000). All of these four SSSs lie in the ruled-out regions of the three young sources, with the latter threeSSSs having similar temperatures and luminosities as thenuclear-burning WD models. The upper limit of the largestremnant, SNR 0505-67.9, overlaps with the parameter rangeof CAL 87, but one should note that CAL 87 (number 1 inFig. 4), which has the lowest claimed luminosity of the four,is viewed almost edge-on, meaning that its unobscured lu-minosity is likely much higher (Ness et al. 2013).In Fig. 4, the results for each source are shown withthe gas density of the simulations set to 1 cm − , but asmentioned before, the results are affected by the assumedgas density of the simulations. To test this effect, we ran thesimulations also with the gas density set either to 0.5 or 2.4cm − , which correspond to the upper limits of the densityaround SNR 0509-67.5 and SNR 0519-69.0 (see Table. 1),respectively. The effect of the density on the results is shownin Fig. 5. As is evident from this figure, the highest densityprovides the least constraining limits, while the low and mid MNRAS000 , 1–9 (2018)
J. Kuuttila et al.
Log Effective temperature (K) L o g B o l o m e t r i c l u m i n o s i t y ( e r g s ) Figure 4. σ upper limits on the bolometric luminosity as afunction of the assumed emission colour temperature for the pro-genitors of the four SNRs studied here. The blue, green, black,and red lines show the upper limits for SNR 0509-67.5, SNR 0505-67.9, SNR 0509-68.7, and SNR 0519-69.0, respectively. In all casesin this figure, the ambient gas density is set at 1 cm − , the tem-perature is calculated self-consistently, and the calculations ter-minated when the temperature dropped below 3000 K. For com-parison, the black dotted lines show the accreting nuclear-burningWD models of Wolf et al. (2013) with the mass increasing from0.51 M (cid:12) on the left to 1.4 M (cid:12) on the right. For ease of read-ing, only every second model is labelled. The black dash-dottedboxes represent the parameter ranges of four well-known SSSs: 1.CAL87; 2. 1E 0035.4-7230; 3. RX J0513.9-6951; and 4. CAL 83(Greiner 2000). Log Effective temperature (K) L o g B o l o m e t r i c l u m i n o s i t y ( e r g s ) Figure 5. σ upper limits on the bolometric luminosity of theprogenitor of SNR 0519-69.0 with different densities. The densi-ties 0.5, 1, and 2.4 cm − are shown in red dashed, blue solid, andgreen dot–dashed lines, respectively. Also shown are the nuclearburning WD models and SSSs, as in Fig. 4. density limits differ only slightly from each other, with themid density limits being the most constraining.In addition to the density, we tested how the assumedtemperature of the gas affects the results. This effect isdemonstrated in Fig. 6, where we show the upper limits withthe temperature either calculated self-consistently, fixed at5000 K, or fixed at 10000 K. From this figure one can see Log Effective temperature (K) L o g B o l o m e t r i c l u m i n o s i t y ( e r g s ) KT = 10 K Figure 6. σ upper limits on the bolometric luminosity for SNR0519-69.0 with different electron temperatures. The temperaturesare either calculated self-consistently (red dashed line) or fixed at5000 (blue solid) or 10000 K (green dot–dashed). The density isset to 1 cm − in all cases. Also shown are the nuclear burning WDmodels and SSSs, as in Fig. 4. that the chosen simulation temperature affects the resultsonly very little. The reason for this is the high ionizationpotential of He II (54.4 eV), which requires much higher en-ergies than available in a typical warm interstellar medium.The upper limits on the bolometric luminosity as a func-tion of the assumed emission colour temperature for the pro-genitor of SNR 0519-69.0 are also shown in Fig. 7. Based onthe analysis presented in this paper, the parameter spaceabove the blue line is ruled out. For comparison, the upperlimits for the same source derived by Woods et al. (2018) areshown in the same figure with a black dashed line. For effec-tive temperatures higher than ∼ K, our analysis providessignificantly tighter constraints on the bolometric luminos-ity than that of Woods et al. (2018), who derived the limitsusing the Balmer-dominated forward shocks of the super-nova remnant (see also e.g. Ghavamian et al. 2003; Woodset al. 2017). On the other hand, for temperatures lower than ∼ K, the work of Woods et al. (2018) provides lower up-per limits on the luminosity than ours, because the incidentradiation field does not possess significant amount of pho-tons with sufficient energies to ionize He + ions, causing thisregime to be poorly constrained by our work, while Woodset al. (2018) rely on the ionization of hydrogen, which re-quires considerably lower photon energies. In Fig. 7 is alsoshown for comparison the upper limits derived from pre-explosion archival Chandra
X-ray data for SN2011fe, whichhas the lowest upper limits of the ten SNe Ia studied byNielsen et al. (2012). The upper limits for SN2011fe areslightly lower than our results for SNR 0519-69.0 in hightemperatures ( (cid:38) K), for the part that there exists datafor SN2011fe. In the high temperature regime our resultsbecome less constraining, because increasing the photon en-ergies leads to less efficient ionizing of the ambient gas, whichis due to the ionizing cross section of a hydrogen-like ion de-creasing as a function of energy, approximately as σ ∝ E − (Hummer & Seaton 1963). MNRAS , 1–9 (2018) xcluding SSSs as SNe Ia progenitors Log Effective temperature (K) L o g B o l o m e t r i c l u m i n o s i t y ( e r g s ) Figure 7. σ upper limits on the bolometric luminosity as afunction of the emission colour temperature for the progenitor ofSNR 0519-69.0. In this figure, the ambient gas density is set at 1cm − , the temperature is calculated self-consistently and the cal-culations terminated when the temperature dropped below 3000K. The solid blue line shows the upper limits derived in this pa-per, and for comparison the black dashed line shows the upperlimit for SNR 0519-69.0 derived by Woods et al. (2018) using theBalmer-dominated shocks. The green dashed line shows the up-per limit from pre-explosion archival X-ray data for SN 2011fe(Nielsen et al. 2012). Also shown are the nuclear burning WDmodels and SSSs, as in Fig. 4. The super-soft X-ray sources have long been suggested aspossible progenitors for Type Ia supernovae. However, re-cent studies have constrained their viability as a progenitorchannel both for large populations (Di Stefano 2010; Gil-fanov & Bogd´an 2010; Woods & Gilfanov 2013; Johanssonet al. 2014; Woods & Gilfanov 2016) and individual super-nova remnants (Nielsen et al. 2012; Woods et al. 2017, 2018;Graur & Woods 2018). In this paper, we have presented anovel method for constraining supernova progenitor prop-erties, and using this method, we have strongly disfavouredthe super-soft progenitor channel for four Type Ia supernovaremnants in the Large Magellanic Cloud.With this method we have focused only on the He ii β and [O iii ] 5007˚A, present in the spectra of the ISMaround the SNRs, as is evident from Fig. 2. These emissionlines, while expected to be bright in a SSS nebula (Rappa-port et al. 1994), are present also in a typical warm ISM inthe LMC (e.g. Pellegrini et al. 2012) and thus with these linesone encounters the problem of disentangling the ionizationcaused by the possible progenitor from the contributions ofother sources, such as the diffuse background and the shockemission (Smith et al. 1994; Ghavamian et al. 2000). For ex-ample, in the case of the most luminous allowed (by the He ii analysis) source with a temperature of 10 K, the predictedH β emission line brightness is a factor of 5 lower and [O iii ]5007˚A brightness is 10 times lower than observed aroundSNR 0519-69.0.Our results add to the growing body of evidence sup- porting the double degenerate scenario as a progenitor chan-nel for these remnants. For SNR 0519-69.0 Edwards et al.(2012) ruled out all post-main-sequence stars as possiblesurviving ex-companions and thus claim that among thepublished single-degenerate models, only the super-soft X-ray source model is a possibility for this remnant. In addi-tion, SNR 0519-69.0 has a tilted axisymmetric morphologyand high oxygen abundance, which points to an oxygen-richmerger (Kosenko et al. 2010, 2015). Taking these togetherwith our results, which rule out a SSS as a plausible progen-itor, it seems clear that the only viable origin of SNR 0519-69.0 was the merger of two white dwarfs. Similarly for SNR0509-67.5 Schaefer & Pagnotta (2012) ruled out all possiblesurviving companion stars in the centre of the remnant, andthus ruled out all single degenerate scenarios as a progenitorchannel for this remnant, which is in good agreement withour results.Our results disfavour SSSs as possible progenitors, butwe made some simplifications along the way, which shouldbe considered in detail. Firstly, we assumed that the lumi-nosity remained constant throughout stable accretion andnuclear burning, although in reality these sources exhibitcomplex variability. However, for variable sources, the pa-rameter of interest is the time-averaged luminosity, whichdetermines the average ionization state of the gas, and thusgiven a sufficiently long time-scale, the system can be wellapproximated with a constant luminosity case (Chiang &Rappaport 1996; Woods et al. 2017). The detailed structureof ionization nebulae may change based on the behaviourof the central source, for example in the case of nova out-bursts. Such cases, and the time variability of the source andnebulae, will be addressed in future studies.Secondly, the calculations were carried out in steady-state, i.e. assuming an equilibrium state between ionizationand recombination, where the central source continues tosupply the nebula with ionizing photons. This is obviouslynot the case for SNRs, where the possible central ionizingsource has exploded and the emission has ceased. Neverthe-less, this is a reasonable assumption in the case of youngSNRs, where the age ( < ∼ yrs).This argument raises the question, however, of whether therecould be a long delay between the explosion and the ionizingphase. This can be achieved with spin-up/spin-down models(Justham 2011), where the accreting WD is spun up becauseof the accreted angular momentum. Because of the high spinrates, the mass of the WD can increase beyond the criticalmass, and only after accretion has ceased and the spin rateof the WD has decreased can the WD explode as a super-nova. If the spin-down time is longer than the recombinationtime, this model can produce super-Chandrasekhar single-degenerate explosions surrounded by neutral gas. In addi-tion, by the time of the explosion, the donor star may haveexhausted its stellar envelope and become a WD, renderingit difficult to detect in post-explosion companion searches(Di Stefano et al. 2011). In fact, such super-Chandrasekharexplosions would be preferentially overluminous, “1991T-like” events (Fisher et al. 1999). This is thought to be thecase for SNR 0509-67.5, which Rest et al. (2008) showedto be a 1991T-like event using its light echoes, a result atwhich Badenes et al. (2008) also arrived independently, us-ing the remnant dynamics and X-ray spectroscopy. Super- MNRAS000
X-ray data for SN2011fe, whichhas the lowest upper limits of the ten SNe Ia studied byNielsen et al. (2012). The upper limits for SN2011fe areslightly lower than our results for SNR 0519-69.0 in hightemperatures ( (cid:38) K), for the part that there exists datafor SN2011fe. In the high temperature regime our resultsbecome less constraining, because increasing the photon en-ergies leads to less efficient ionizing of the ambient gas, whichis due to the ionizing cross section of a hydrogen-like ion de-creasing as a function of energy, approximately as σ ∝ E − (Hummer & Seaton 1963). MNRAS , 1–9 (2018) xcluding SSSs as SNe Ia progenitors Log Effective temperature (K) L o g B o l o m e t r i c l u m i n o s i t y ( e r g s ) Figure 7. σ upper limits on the bolometric luminosity as afunction of the emission colour temperature for the progenitor ofSNR 0519-69.0. In this figure, the ambient gas density is set at 1cm − , the temperature is calculated self-consistently and the cal-culations terminated when the temperature dropped below 3000K. The solid blue line shows the upper limits derived in this pa-per, and for comparison the black dashed line shows the upperlimit for SNR 0519-69.0 derived by Woods et al. (2018) using theBalmer-dominated shocks. The green dashed line shows the up-per limit from pre-explosion archival X-ray data for SN 2011fe(Nielsen et al. 2012). Also shown are the nuclear burning WDmodels and SSSs, as in Fig. 4. The super-soft X-ray sources have long been suggested aspossible progenitors for Type Ia supernovae. However, re-cent studies have constrained their viability as a progenitorchannel both for large populations (Di Stefano 2010; Gil-fanov & Bogd´an 2010; Woods & Gilfanov 2013; Johanssonet al. 2014; Woods & Gilfanov 2016) and individual super-nova remnants (Nielsen et al. 2012; Woods et al. 2017, 2018;Graur & Woods 2018). In this paper, we have presented anovel method for constraining supernova progenitor prop-erties, and using this method, we have strongly disfavouredthe super-soft progenitor channel for four Type Ia supernovaremnants in the Large Magellanic Cloud.With this method we have focused only on the He ii β and [O iii ] 5007˚A, present in the spectra of the ISMaround the SNRs, as is evident from Fig. 2. These emissionlines, while expected to be bright in a SSS nebula (Rappa-port et al. 1994), are present also in a typical warm ISM inthe LMC (e.g. Pellegrini et al. 2012) and thus with these linesone encounters the problem of disentangling the ionizationcaused by the possible progenitor from the contributions ofother sources, such as the diffuse background and the shockemission (Smith et al. 1994; Ghavamian et al. 2000). For ex-ample, in the case of the most luminous allowed (by the He ii analysis) source with a temperature of 10 K, the predictedH β emission line brightness is a factor of 5 lower and [O iii ]5007˚A brightness is 10 times lower than observed aroundSNR 0519-69.0.Our results add to the growing body of evidence sup- porting the double degenerate scenario as a progenitor chan-nel for these remnants. For SNR 0519-69.0 Edwards et al.(2012) ruled out all post-main-sequence stars as possiblesurviving ex-companions and thus claim that among thepublished single-degenerate models, only the super-soft X-ray source model is a possibility for this remnant. In addi-tion, SNR 0519-69.0 has a tilted axisymmetric morphologyand high oxygen abundance, which points to an oxygen-richmerger (Kosenko et al. 2010, 2015). Taking these togetherwith our results, which rule out a SSS as a plausible progen-itor, it seems clear that the only viable origin of SNR 0519-69.0 was the merger of two white dwarfs. Similarly for SNR0509-67.5 Schaefer & Pagnotta (2012) ruled out all possiblesurviving companion stars in the centre of the remnant, andthus ruled out all single degenerate scenarios as a progenitorchannel for this remnant, which is in good agreement withour results.Our results disfavour SSSs as possible progenitors, butwe made some simplifications along the way, which shouldbe considered in detail. Firstly, we assumed that the lumi-nosity remained constant throughout stable accretion andnuclear burning, although in reality these sources exhibitcomplex variability. However, for variable sources, the pa-rameter of interest is the time-averaged luminosity, whichdetermines the average ionization state of the gas, and thusgiven a sufficiently long time-scale, the system can be wellapproximated with a constant luminosity case (Chiang &Rappaport 1996; Woods et al. 2017). The detailed structureof ionization nebulae may change based on the behaviourof the central source, for example in the case of nova out-bursts. Such cases, and the time variability of the source andnebulae, will be addressed in future studies.Secondly, the calculations were carried out in steady-state, i.e. assuming an equilibrium state between ionizationand recombination, where the central source continues tosupply the nebula with ionizing photons. This is obviouslynot the case for SNRs, where the possible central ionizingsource has exploded and the emission has ceased. Neverthe-less, this is a reasonable assumption in the case of youngSNRs, where the age ( < ∼ yrs).This argument raises the question, however, of whether therecould be a long delay between the explosion and the ionizingphase. This can be achieved with spin-up/spin-down models(Justham 2011), where the accreting WD is spun up becauseof the accreted angular momentum. Because of the high spinrates, the mass of the WD can increase beyond the criticalmass, and only after accretion has ceased and the spin rateof the WD has decreased can the WD explode as a super-nova. If the spin-down time is longer than the recombinationtime, this model can produce super-Chandrasekhar single-degenerate explosions surrounded by neutral gas. In addi-tion, by the time of the explosion, the donor star may haveexhausted its stellar envelope and become a WD, renderingit difficult to detect in post-explosion companion searches(Di Stefano et al. 2011). In fact, such super-Chandrasekharexplosions would be preferentially overluminous, “1991T-like” events (Fisher et al. 1999). This is thought to be thecase for SNR 0509-67.5, which Rest et al. (2008) showedto be a 1991T-like event using its light echoes, a result atwhich Badenes et al. (2008) also arrived independently, us-ing the remnant dynamics and X-ray spectroscopy. Super- MNRAS000 , 1–9 (2018)
J. Kuuttila et al.
Chandrasekhar-mass explosions, however, can also resultfrom double-degenerate mergers, which lack the issues facingspin-up/spin-down models, such as the scarcity of observedrapidly-spinning WDs (Di Stefano et al. 2011; Maoz et al.2014).Thirdly, in the analysis presented here, we have consid-ered only unobscured sources, where all the emitted radi-ation contributes to the ionization of the surrounding gas.The emission could, however, be obscured by a fast-movingand optically-thick stellar wind, if the WD were accret-ing at higher rates than the steady nuclear-burning regime(Hachisu et al. 1996; Wolf et al. 2013). If the wind mass-loss rate were high enough to obscure the central source,however, the wind should excavate a large ( (cid:38)
10 pc) low-density cavity around the progenitor, which should be eas-ily distinguished from the undisturbed ISM. Such large cav-ities are incompatible with the remnants’ dynamics for thethree young supernova remnants studied here (Badenes et al.2007), and the densities (see Table 1) and evolution of theremnants are consistent with expansion into a uniform andundisturbed ISM (Maggi et al. 2016). In addition to a windfrom the accreting WD, a slow and dense wind from a gi-ant companion star may obscure the ionizing radiation, ifthe mass-loss rate is (cid:38) − M (cid:12) yr − (Nielsen & Gilfanov2015). However, such a scenario is disfavoured for SNe Iaprogenitors, given the strong constrains on circumstellar in-teractions both from radio (Chomiuk et al. 2012, 2016) andX-ray observations (Margutti et al. 2012, 2014), and the lackof detected giant companions (Edwards et al. 2012; Schaefer& Pagnotta 2012; Olling et al. 2015).Therefore, we may conclude that none of the progeni-tors of the Magellanic supernova remnants considered herewere super-soft X-ray sources for a significant fraction of thelast 100,000 years preceding their detonation. Future spec-troscopic observations can extend these limits to all nearby,recent SNe Ia and supernova remnants, or in the event of adetection, provide the first measurement of the luminosityand temperature of a SN Ia progenitor. ACKNOWLEDGEMENTS
I.R.S. acknowledges support from the Australian ResearchCouncil Future Fellowship Grant FT160100028.
REFERENCES
Badenes C., Hughes J. P., Bravo E., Langer N., 2007, ApJ, 662,472Badenes C., Hughes J. P., Cassam-Chena¨ı G., Bravo E., 2008,ApJ, 680, 1149Chen H.-L., Woods T. E., Yungelson L. R., Gilfanov M., Han Z.,2015, MNRAS, 453, 3024Chiang E., Rappaport S., 1996, ApJ, 469, 255Childress M., Vogt F., Nielsen J., Sharp R., 2014a, PyWiFeS:Wide Field Spectrograph data reduction pipeline, Astro-physics Source Code Library (ascl:1402.034)Childress M. J., Vogt F. P. A., Nielsen J., Sharp R. G., 2014b,Ap&SS, 349, 617Chomiuk L., et al., 2012, ApJ, 750, 164Chomiuk L., et al., 2016, ApJ, 821, 119Di Stefano R., 2010, ApJ, 712, 728Di Stefano R., Voss R., Claeys J. S. W., 2011, ApJ, 738, L1 Dopita M., Hart J., McGregor P., Oates P., Bloxham G., JonesD., 2007, Ap&SS, 310, 255Dopita M., et al., 2010, Ap&SS, 327, 245Dopita M. A., Seitenzahl I. R., Sutherland R. S., Vogt F. P. A.,Winkler P. F., Blair W. P., 2016, ApJ, 826, 150Edwards Z. I., Pagnotta A., Schaefer B. E., 2012, ApJ, 747, L19Ferland G. J., et al., 2013, Rev. Mex. Astron. Astrofis., 49, 137Fisher A., Branch D., Hatano K., Baron E., 1999, MNRAS, 304,67Ghavamian P., Raymond J., Hartigan P., Blair W. P., 2000, ApJ,535, 266Ghavamian P., Raymond J., Smith R. C., Hartigan P., 2001, ApJ,547, 995Ghavamian P., Rakowski C. E., Hughes J. P., Williams T. B.,2003, ApJ, 590, 833Ghavamian P., Seitenzahl I. R., Vogt F. P. A., Dopita M. A.,Terry J. P., Williams B. J., Winkler P. F., 2017, ApJ, 847,122Gilfanov M., Bogd´an ´A., 2010, Nature, 463, 924Graur O., Woods T. E., 2018, preprint, ( arXiv:1811.04944 )Graur O., Maoz D., Shara M. M., 2014, MNRAS, 442, L28Greiner J., 2000, New Astron., 5, 137Gruyters P., Exter K., Roberts T. P., Rappaport S., 2012, A&A,544, A86Hachisu I., Kato M., Nomoto K., 1996, ApJ, 470, L97Hillebrandt W., Kromer M., R¨opke F. K., Ruiter A. J., 2013,Frontiers of Physics, 8, 116Hughes J. P., Hayashi I., Koyama K., 1998, ApJ, 505, 732Hummer D. G., Seaton M. J., 1963, MNRAS, 125, 437Iben Jr. I., Tutukov A. V., 1984, ApJS, 54, 335Johansson J., Woods T. E., Gilfanov M., Sarzi M., Chen Y.-M.,Oh K., 2014, MNRAS, 442, 1079Justham S., 2011, ApJ, 730, L34Kahabka P., van den Heuvel E. P. J., 1997, ARA&A, 35, 69Kosenko D., Vink J., Blinnikov S., Rasmussen A., 2008, A&A,490, 223Kosenko D., Helder E. A., Vink J., 2010, A&A, 519, A11Kosenko D., Hillebrandt W., Kromer M., Blinnikov S. I., PakmorR., Kaastra J. S., 2015, MNRAS, 449, 1441Long K. S., Helfand D. J., Grabelsky D. A., 1981, ApJ, 248, 925Maggi P., et al., 2016, A&A, 585, A162Maoz D., Mannucci F., Nelemans G., 2014, ARA&A, 52, 107Margutti R., et al., 2012, ApJ, 751, 134Margutti R., Parrent J., Kamble A., Soderberg A. M., Foley R. J.,Milisavljevic D., Drout M. R., Kirshner R., 2014, ApJ, 790,52Matteucci F., Greggio L., 1986, A&A, 154, 279McConnachie A. W., 2012, AJ, 144, 4McKee C. F., Ostriker J. P., 1977, ApJ, 218, 148Ness J.-U., et al., 2013, A&A, 559, A50Nielsen M. T. B., Gilfanov M., 2015, MNRAS, 453, 2927Nielsen M. T. B., Voss R., Nelemans G., 2012, MNRAS, 426, 2668Olling R. P., et al., 2015, Nature, 521, 332Pellegrini E. W., Oey M. S., Winkler P. F., Points S. D., SmithR. C., Jaskot A. E., Zastrow J., 2012, ApJ, 755, 40Pequignot D., Petitjean P., Boisson C., 1991, A&A, 251, 680Perlmutter S., et al., 1999, ApJ, 517, 565Phillips M. M., 1993, ApJ, 413, L105Phillips M. M., Lira P., Suntzeff N. B., Schommer R. A., HamuyM., Maza J., 1999, AJ, 118, 1766Rappaport S., Chiang E., Kallman T., Malina R., 1994, ApJ, 431,237Remillard R. A., Rappaport S., Macri L. M., 1995, ApJ, 439, 646Rest A., et al., 2005, Nature, 438, 1132Rest A., et al., 2008, ApJ, 680, 1137Riess A. G., et al., 1998, AJ, 116, 1009Schaefer B. E., Pagnotta A., 2012, Nature, 481, 164Schenck A., Park S., Post S., 2016, AJ, 151, 161MNRAS , 1–9 (2018) xcluding SSSs as SNe Ia progenitors Smith R. C., Raymond J. C., Laming J. M., 1994, ApJ, 420, 286Weingartner J. C., Draine B. T., 2001, ApJ, 548, 296Whelan J., Iben Jr. I., 1973, ApJ, 186, 1007Wiersma R. P. C., Schaye J., Theuns T., 2011, MNRAS, 415, 353Williams B. J., et al., 2014, ApJ, 790, 139Wolf W. M., Bildsten L., Brooks J., Paxton B., 2013, ApJ, 777,136Woods T. E., Gilfanov M., 2013, MNRAS, 432, 1640Woods T. E., Gilfanov M., 2016, MNRAS, 455, 1770Woods T. E., Ghavamian P., Badenes C., Gilfanov M., 2017, Na-ture Astronomy, 1, 800Woods T. E., Ghavamian P., Badenes C., Gilfanov M., 2018, ApJ,863, 120Yamaguchi H., et al., 2014, ApJ, 785, L27van den Heuvel E. P. J., Bhattacharya D., Nomoto K., RappaportS. A., 1992, A&A, 262, 97van der Heyden K. J., Behar E., Vink J., Rasmussen A. P., Kaas-tra J. S., Bleeker J. A. M., Kahn S. M., Mewe R., 2002, A&A,392, 955This paper has been typeset from a TEX/L A TEX file prepared bythe author.MNRAS000