NO STRIPPED HYDROGEN IN THE NEBULAR SPECTRA OF NEARBY TYPE Ia SUPERNOVA 2011fe^*

Despite the fact that Type Ia supernovae (SNe Ia) were used to discover the accelerating universe (Riess et al. 1998; Perlmutter et al. 1999), the physical nature of their progenitor systems remains theoretically ambiguous and observationally elusive (for a review see Wang & Han 2012). It is commonly accepted that SNe Ia result from the thermonuclear explosion of a carbon–oxygen white dwarf (WD) in a close binary system, but the nature of the binary companion and the sequence of events leading to the SN explosion are still uncertain. There are two dominant models: the double degenerate (DD) scenario, in which the companion is also a WD (Tutukov & Yungelson 1979; Iben & Tutukov 1984; Webbink 1984), and the single-degenerate (SD) scenario, in which the companion is a non-degenerate object such as a main-sequence (MS) star, a red giant (RG), a sub-giant, or an He star (Whelan & Iben 1973; Nomoto 1982). There are additional minor variations on these basic models of the progenitor system for both the SD channel (e.g., Justham 2011; Wheeler 2012) and the DD channel (e.g., Thompson 2011; Shappee & Thompson 2012). These uncertainties about the progenitor systems for SNe Ia remain a substantial problem for understanding the systematic errors in using SNe Ia to study cosmology (e.g., Wood-Vasey et al. 2007).

One of the observational signatures of the SD model is that the material from the companion should be stripped when struck by the SN ejecta (Wheeler et al. 1975), leading to both immediate signatures from the impact of the SN ejecta on the companion (e.g., Marietta et al. 2000; Meng et al. 2007; Pakmor et al. 2008; Pan et al. 2012b; Liu et al. 2012) and major changes in the companion's future evolution (Podsiadlowski 2003; Shappee et al. 2012; Pan et al. 2012a). Recently, the hydrodynamic simulations of Pan et al. (2012b) and Liu et al. (2012) have shown that ∼0.1–0.2 M_☉ of solar-metallicity material is expected to be removed from MS companions by the impact of the SN ejecta. These hydrodynamic simulations show that this material will be embedded in low-velocity supernova debris with a characteristic velocity of ≲1000 km s⁻¹. However, the line profiles from the stripped material are somewhat uncertain because this material will be asymmetrically stripped (Liu et al. 2012), and the orientation of the binary relative to our line of sight, at the time of explosion, is unknown. This material will be hidden in early-time spectra by higher velocity, optically thick, iron-rich ejecta, but will then appear in late-time, nebular phase spectra (≳ 250 days; Mattila et al. 2005) as the higher velocity ejecta become optically thin.

Only a handful of nebular phase, high signal-to-noise (S/N) SNe Ia spectra have been published in the literature, with the strongest limits on late-time hydrogen flux coming from Mattila et al. (2005) and Leonard (2007). Mattila et al. (2005) obtained late-time low-resolution spectra of SN 2001el, modeled the emission from solar-metallicity material stripped from a non-degenerate companion, and used both to place an upper limit of ≲0.03 M_☉ on the presence of this material. Leonard (2007) obtained deep, medium-resolution, nebular-phase, multi-epoch spectra of SN 2005am and SN 2005cf at distances ∼37 Mpc and ∼32 Mpc, respectively. Using the approach of Mattila et al. (2005), he placed a ≲0.01 M_☉ upper limit on the presence of any low-velocity solar-abundance material in both SNe. These observations seem to firmly rule out MS companions for these three SNe.

However, a more exotic scenario which might evade these constraints has since been proposed by Justham (2011). In this model, a WD is spun up by the mass it accretes from its non-degenerate companion, allowing the WD to remain stable above the Chandrasekhar mass and giving the companion time to evolve and contract before the SN explosion. This leaves a smaller and more tightly bound companion at the time of explosion, reducing the amount of stripped material. It therefore is of significant interest to put even stronger constraints on the amount of stripped material for other SNe Ia.

At a mere 6.4 Mpc (Shappee & Stanek 2011), the "plain vanilla" Type Ia SN (SN Ia; as described by Wheeler 2012) is an ideal target for obtaining improved constraints. SN 2011fe was discovered less than one day after explosion by the Palomar Transient Factory (Law et al. 2009) in the Pinwheel Galaxy (M101), a well-studied, nearby face-on spiral. This SN is the nearest SN Ia in the last 25 years and the early detection and announcement allowed extensive multi-wavelength follow-up observations (Nugent et al. 2011). Additionally, SN 2011fe is only slightly reddened and is surrounded by a relatively "clean" environment (Patat et al. 2011). Brown et al. (2012) and Bloom et al. (2012) used the very early time UV and optical observations to constrain the shock heating that would occur as the SN ejecta collide with the companion, ruling out giants and MS secondaries more massive than the Sun. These constraints appear to rule out SD models; however, these studies both rely on Kasen (2010), which assumes a Roche-lobe overflowing secondary with a typical stellar structure and depends on the orientation of the binary relative to our line of sight. However, Meng et al. (2007) pointed out that the secondary will be left smaller and more compact than a typical MS star after binary evolution, which would produce a more subtle photometric signal at even earlier times which could have been missed by early-time observations.

In this study, we place the strongest limits yet on the presence of Hα emission in the nebular spectrum of an SN Ia. We obtained a very high S/N spectrum of SN Ia 2011fe 274 days after maximum B-band light using the Multi-Object Double Spectrograph (MODS) on the 8.4 m Large Binocular Telescope (LBT). We generally followed the analysis procedures used by Leonard (2007), so our constraints should be directly comparable to earlier results. In Section 2 we describe our observations, in Section 3 we derive our Hα limits, and we summarize our findings in Section 4.

We obtained five high-S/N spectra of SN 2011fe from 73 to 274 days after maximum B-band light using the first of the MODS (Pogge et al. 2010) on the LBT. The two channels of MODS allowed us to obtain wide spectral coverage (3200–10000 Å) in a single exposure, which will make these observations useful for many future studies. Details of the observations and the flux standards used are listed in Table 1. All observations were taken through a 1'' wide slit. To perform basic CCD reductions on our spectra, we followed the "MODS Basic CCD Reduction with modsCCDRed" manual.⁵ We then performed cosmic-ray rejection using L.A.Cosmic (van Dokkum 2001) and combined each channel's two-dimensional spectra for each epoch. Next, we extracted the one-dimensional sky-subtracted spectra using the apall task in IRAF.⁶ Each spectrum was then wavelength and flux calibrated. We did not attempt to correct small-scale telluric absorption bands as performed by Leonard (2007). This will not adversely affect our results because the telluric absorptions occur on wavelength scales smaller than the Hα emission expected from the SN. This also avoids adding any noise from telluric spectra to our SN spectra.

Because our spectra were taken under non-photometric conditions with a relatively narrow (1'') slit, it is necessary to scale the fluxes of our spectra to place them on an absolute flux scale. To do this, we performed aperture photometry on our Sloan r' (Fukugita et al. 1996) acquisition images using the IRAF package apphot. Because the SN was saturated in the acquisition image of the first epoch, we must treat that spectrum separately. For the remaining epochs, we used the four brightest stars in the image to put our photometry on the Sloan Digital Sky Survey (SDSS; York et al. 2000) Data Release 7 (DR7; Abazajian et al. 2009) magnitude system. Our r'-band magnitudes are reported in Table 1. We then scale the spectrum so that its synthetic r'-band photometry matches the computed r'-band aperture photometry. For the first epoch we scaled the spectrum such that a synthetic R-band magnitude (Bessell & Murphy 2012) matches the R-band magnitude for its epoch found by interpolating the combined R-band light curves of Richmond & Smith (2012) and Munari et al. (2012). To convert from the AB magnitude system (Oke 1974) of our synthetic photometry to the Vega magnitude system reported in Richmond & Smith (2012) and Munari et al. (2012), we use the conversion presented in Blanton & Roweis (2007). The factor by which each spectrum was multiplied is reported in Table 1. From the last four epochs' spectra, which are calibrated from the acquisition image but are also covered by the combined R-band light curves of Richmond & Smith (2012) and Munari et al. (2012), we estimate that our absolute flux calibration in the R band (where Hα is located) is accurate to 10% or better. In Figure 1 we show the BVRI light curves of Richmond & Smith (2012) and Munari et al. (2012) as compared to our synthetic BVRI photometry from the spectra for illustrative purposes. Figure 2 presents our calibrated spectra with obvious telluric features marked. Finally, we compare our nebular phase spectrum to the spectra presented in Leonard (2007), which provided the previous best limits on Hα. As can be seen in Figure 3, our spectrum has a substantially higher S/N than the previous studies. These observations allow us to place even stronger limit on Hα emission as we discuss in Section 3.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

In our search for Hα emission in the spectrum of SN 2011fe, we closely follow the methods presented in Leonard (2007). We define a continuum by smoothing our spectra on scales large compared to the expected Hα feature, then subtract off this continuum and examine the residuals. We searched for Hα emission within ±1000 km s⁻¹ (±22 Å) about Hα at the redshift of M101, 0.000804 ± 0.000007 (de Vaucouleurs et al. 1991). We smoothed the spectrum with a second-order Savitsky–Golay smoothing polynomial (Press et al. 1992) with a width of 60 Å. Our spectrum requires a smaller smoothing scale than that employed by Leonard (2007) for a good continuum fit. It is, however, still significantly larger than the expected velocity width of any Hα feature. Additionally, we smoothed over the telluric feature at 6510–6525 Å before smoothing the rest of the spectrum because the feature was affecting the continuum determination near Hα.

Our nebular phase spectrum and continuum fit are shown in the top panel of Figure 4 in the vicinity of Hα and are binned to their approximate spectral resolution.

Zoom In Zoom Out Reset image size

We then subtracted the continuum from the binned spectrum and examined the residuals, shown in the bottom panel of Figure 4, for narrow Hα emission. There is no evidence for any Hα emission in the spectrum. The closest possible emission is the hump in the smoothed continuum at ∼6575–6595 Å, which is discussed in Section 3.2. More broadly, we found no narrow emission lines at any wavelength, including regions around Hβ, [O i] λλ6300, 6364, [O ii] λλ7319, 7330, [O iii] λλ4959, 5007, [Ca ii] λλ7291, 7324, and [N ii] λλ6548, 6583.

3.1. Statistical Limit

Following Leonard & Filippenko (2001) and Leonard (2007), we compute a 3σ upper bound on the equivalent width as

$$\begin{eqnarray} W_\lambda (3\sigma) &=& 3 \Delta \lambda \ \Delta I \sqrt{W_{\rm{line}} / \Delta \lambda \ } \sqrt{1 / B} \nonumber \\ &=& 3 \Delta I \sqrt{W_{\rm{line}} \ \Delta X}, \end{eqnarray} \tag{ 1 }$$

where Δλ is the width of a resolution element (in Å), ΔI is the 1σ rms fluctuation of the flux around a normalized continuum level, W_line is the FWHM of the expected spectral feature (in Å), B is the number of bins per resolution element in the spectrum, and ΔX is the bin size of the spectrum (in Å). For the spectrum shown in Figure 4, ΔI = 0.0024 Å and ΔX = 4 Å leading to W_λ(3σ) = 0.067 Å for W_line = 22 Å.

To translate this equivalent width constraint into a constraint on the amount of material stripped from a non-degenerate companion, we follow the analysis of Mattila et al. (2005). Mattila et al. (2005) estimate that 0.5 M_☉ of solar-abundance material 380 days after explosion produces an Hα feature with peak luminosity of ∼3.36 × 10³⁵ ergs s⁻¹ Å⁻¹. Accounting for the distance to M101 and Galactic extinction toward M101 of E(B − V) = 0.009 mag (Schlegel et al. 1998) and assuming R_V = 3.1, the expected Hα peak flux from 0.5 M_☉ of stripped material in SN 2011fe is 6.71 × 10⁻¹⁷ erg s⁻¹ cm⁻² Å⁻¹. To allow for comparison with Leonard (2007), we initially approximate the Hα feature as a Gaussian with FWHM 22 Å. This feature would then have an equivalent width of W_λ(0.05 M_☉) = 5.8 Å. Scaling linearly from the equivalent width of the Hα emission line to the amount of stripped material, we place an upper limit on the amount of solar-abundance material in SN 2011fe of 5.8 × 10⁻⁴ M_☉.

3.2. Conservative Limit

We obtained five deep, medium-resolution spectra of the nearby "plain vanilla" SN Ia 2011fe with LBT/MODS from 73 to 274 days after maximum B-band light. With the last nebular phase spectrum we place the deepest flux limits on narrow Hα emission yet for an SN Ia. Determining the late-time Hα emission in SNe Ia spectra requires difficult radiative transfer calculations, but linearly scaling from the models of Mattila et al. (2005), our limit translates into an upper limit on the mass of solar-abundance material of ≲0.001 M_☉, an order of a magnitude smaller than previous limits (Leonard 2007). However, two important theoretical questions remain. First, the opacity of the high-velocity iron-rich ejecta should be recomputed to confirm when the SN ejecta becomes optically thin. Second, the excitation of Hα emission by gamma-ray deposition should be modeled in detail assuming various velocity profiles with differing amounts of stripped companion material. These issues notwithstanding, our mass limit poses a significant challenge to more exotic SD models proposed to evade previous constraints. Additionally, our limit strongly rules out all MS and RG companions for which hydrodynamic simulations of the SN ejecta's impact are presented in the literature.

We thank Douglas C. Leonard for discussions and for providing the nebular phase spectra of SN 2005am and SN 2005cf. We also thank Chris Kochanek, Todd Thompson, Jennifer van Saders, Gisella DeRosa, Dale Mudd, and Joe Antognini for discussions and encouragement. We also thank the anonymous referee for his/her helpful suggestions. B.J.S. was supported by a Graduate Research Fellowship from the National Science Foundation. B.J.S. and K.Z.S. are supported in part by NSF grant AST-0908816. This Letter used data obtained with the MODS spectrographs built with funding from NSF grant AST-9987045 and the NSF Telescope System Instrumentation Program (TSIP), with additional funds from the Ohio Board of Regents and the Ohio State University Office of Research. The LBT is an international collaboration among institutions in the United States, Italy, and Germany. LBT Corporation partners are The Ohio State University, and The Research Corporation, on behalf of the University of Notre Dame, University of Minnesota and University of Virginia; the University of Arizona on behalf of the Arizona university system; Istituto Nazionale di Astrofisica, Italy; LBT Beteiligungsgesellschaft, Germany, representing the Max-Planck Society, the Astrophysical Institute Potsdam, and Heidelberg University. This research used the facilities of the Italian Center for Astronomical Archive (IA2) operated by INAF at the Astronomical Observatory of Trieste. The SDSS is managed by the Astrophysical Research Consortium for the Participating Institutions. Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan Foundation, the Participating Institutions, the National Science Foundation, the US Department of Energy, the National Aeronautics and Space Administration, the Japanese Monbukagakusho, the Max Planck Society, and the Higher Education Funding Council for England. The SDSS Web Site is http://www.sdss.org/. This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. This research has made use of NASA's Astrophysics Data System Bibliographic Services.