GeminiFocus 2013 Year in Review | 页面 23

large and independent sets of optical spectra have reached the same conclusion that 20-30 percent of SNe Ia harbor unprocessed carbon. Meanwhile, there had been no detection of carbon in the near-infrared spectra of a normal SN Ia. The issue appeared settled. It was up to the theorists to find the right combination white dwarf binary systems and explosion mechanism to reproduce the observed rate of carbon detection. The situation changed, however, in 2011. Using the high-quality GNIRS spectra and a more sophisticated spectrum modeling technique, we were able to detect carbon in SN 2011fe, a first in the near-infrared wavelengths for a normal SN Ia. In Figure 2, we show the comparison between observed and model spectra. The near-infrared carbon line we studied is relatively isolated and ideally located between two magnesium lines. Our model spectra show that the presence of carbon is required to produce the observed “flattened” profile near 1.03 microns. Furthermore, the time-series GNIRS observations indicate that the influence of carbon increases with time (Figure 2). The carbon line in the optical, on the other hand, usually disappears very early, requiring that the supernova be discovered at a very young age. We propose that the delay in the onset of the near-infrared carbon feature can be explained simply by the change in the ionization condition. The “flattened” profile caused by the presence of carbon in SN 2011fe appears to be common in normal SNe Ia. This suggests that many SNe Ia harbor unprocessed carbon. Again, the low rate of detection in the optical may be caused by the difficulty of obtaining spectra within a few days of the explosion. Since the conclusion of ubiquitous unprocessed carbon would have profound implications for our understanding of SNe Ia explosions, we are currently conducting a careful survey of the near-infrared carbon feature in our growing sample of nearinfrared spectra. The Main Driver of the Luminosity-decline Rate Relation A landmark paper by Wheeler et al. (1998) identified the strong and relatively isolated absorption feature near 1.05 microns as magnesium. They predicted that the velocity of this feature would decrease rapidly and then settle to a constant velocity. The prediction is finally confirmed 15 years later, as our SpeX and GNIRS spectra caught the rapid decline Figure 3. Time evolution of SN Ia near-infrared magnesium velocity. The magnesium velocity of the GNIRS SN 2011fe spectra underwent a rapid decline and an extended period of constant velocity. Note that SN 1999by is a spectroscopically peculiar SN Ia, much like SN 1991bg. The magnesium velocity of normal SNe Ia all show similar constant behavior as that of SN 2011fe. As the supernova ejecta expands, the temperature decreases. The optical carbon line in its first ionized state then gradually recombines into neutral carbon which forms the ever stronger neutral carbon feature in the near-infrared. Due to this fortuitous delay in its appearance, the near-infrared neutral carbon feature is potentially a superior probe of unprocessed material to the more commonly used optical feature. January2014 2013 Year in Review GeminiFocus 21