Annu. Rev. Astron. Astrophys. 1997. 35: 309-355 Copyright © 1997 by Annual Reviews. All rights reserved

4.4. Is There Helium in SNe Ic?

What evidence do we have that helium is truly absent from the spectra of genuine SNe Ic? Do some (or even most) SNe Ic actually have weak He I lines? [Filippenko (1991b) demonstrated how difficult it can be to distinguish between SNe Ib and SNe Ic, especially if the explosion date is unknown.] If so, what does the helium tell us about the progenitors and their evolutionary histories?

The bright, nearby SN Ic 1994I in NGC 5194 (M51) has begun to shed light on these issues. As shown by Filippenko et al (1995b), strong He I 10,830 absorption was visible during the first month past maximum brightness (Figure 11). Moreover, the Na I D 5892 absorption line may have been somewhat contaminated by He I 5876 (see also Clocchiatti et al 1996c), as evidenced by the weak notch in the blue wing of the Na I D absorption on April 18, 1994. Based on the optical region alone, SN 1994I is clearly a SN Ic; nevertheless, its atmosphere cannot be completely devoid of helium, as is most convincingly demonstrated by the He I 10,830 line. Clocchiatti et al (1996c), Clocchiatti & Wheeler (1997) showed that weak optical He I lines appear to be present in several other classical SNe Ic, including SN 1987M (see also Jeffery et al 1991). Indeed, the line at ~ 5520 Å in the first four spectra of Figure 10 (most easily visible in the third spectrum) is identified as He I 5876. Thus, the presence of at least some helium is a common property of the progenitors of SNe Ic.

Figure 11. Montage of spectra of SN 1994I in NGC 5194 (cz = 500 km s-1), from Filippenko et al (1995b). Epochs (days) are given relative to maximum B brightness (April 8, 1994). The late-time spectra are significantly contaminated by gas and early-type stars in the host galaxy; note the blue continuum, as well as the Balmer absorption and emission lines. Blueshifted He I 10,830 is prominent at early times, and the transition to the nebular phase is rapid.

Nomoto et al (1994; see also Iwamoto et al 1994) suggested that the progenitor of SN 1994I was a 2.2-M C-O core formed as a consequence of two stages of mass transfer in a binary system; during the second stage, helium was lost to the close companion (most likely an O-Ne-Mg white dwarf). Woosley et al (1995), on the other hand, invoked only the first stage of mass transfer, during which the hydrogen envelope is lost to the companion. Depending on the initial mass of the star, the resulting helium star has a mass in the range 4-20 M, but subsequent mass loss through winds is very efficient and makes the final mass of the C-O star always converge to the narrow range 2.26-3.5 M. In both cases, the explosion mechanism was iron core collapse, as in SNe II, but the mass of ejected helium is rather different: ~ 0.01 M (Nomoto et al 1994) or 0.1-0.3 M (Woosley et al 1995).

Baron et al (1996) found no direct evidence for helium (upper limit of ~ 0.1 M) in a preliminary analysis of the optical spectra of SN 1994I, but they did not include the He I 10,830 line and non-LTE effects. Naively, if 0.1 M of helium is moving at 16,500 km s^-1 (as observed for the He I 10,830 absorption minimum), the corresponding kinetic energy (~ 0.3 × 10⁵¹ erg) already seems excessive; this may favor the Nomoto et al (1984) model, which requires a factor of 10 less helium. On the other hand, Nomoto et al made a specific prediction about the late-time spectrum of SN 1994I: Emission lines of calcium should exhibit velocities of up to ~ 10,000 km s^-1, and the oxygen lines would be even broader because oxygen is concentrated in the outermost gas layers (Iwamoto et al 1994). This is not the case (Filippenko et al 1995b); despite consisting of the more closely spaced doublet, the [Ca II] line is broader than [O I], and the most rapidly moving oxygen has v 7000 km s^-1.