Comments on the Progenitor of NGC 6946-BH1

Adams et al. (2017) have proposed that the rapid decline of a transient in NGC 6946 was a failed supernova due to the core-collapse of a red supergiant to a black hole. Here I suggest that the progenitor was instead an intermediate temperature star, a yellow hypergiant, similar to objects like IRC +10420 (Oudmaijer et al. 1996; Humphreys et al. 1997, 2002) and Var A in M33 (Hubble & Sandage 1953; Humphreys et al. 1987, 2006). It was on an evolutionary track back to warmer temperatures when the collapse occurred. The properties of yellow hypergiants, including high mass loss episodes, strong stellar winds, and dusty circumstellar ejecta, are relevant to understanding some of the features in the pre-collapse light curve of the progenitor.

Figure 1 shows the observed colors and magnitudes corrected for foreground extinction. Adams et al. (2017) reject any additional local or environmental extinction in NGC 6946. Due to its low Galactic latitude, however, the foreground extinction, (A_V of 0.94 mag) is rather high. They use the same in their models. The broad-band colors of the progenitor from 1999 up to 2008 are those of an intermediate temperature star. The most complete set of UBVR photometry is from 2005.5. For example, (B − V)₀ of 0.8 mag, suggests an early-G-type supergiant with a surface temperature of about 5000 K. Similarly the other colors are consistent with a late F to early G-type supergiant with temperatures of 5000–6000 K. Two of their progenitor models (Table 2 in Adams et al. 2017) for this period indicate temperatures for the star of 6000–7000 K.

Zoom In Zoom Out Reset image size

The SED for 2005.5 fit with a simple Planck curve yields a temperature of 5000 K¹ Spitzer IRAC 3.6 and 4.5 μm observations from the same period show a significant infrared excess due to circumstellar dust and/or free–free emission. The same models (Adams et al. 2017) yield dust temperatures of ∼1700 which is unusually high. The typical condensation temperature for dust around evolved stars is ≈1000 K (Suh et al. 1990). The near-infrared flux in the SED is consistent with a strong free–free component, another characteristic of the yellow hypergiants with high mass loss and strong winds. This does not rule a contribution from dust at these and longer wavelengths. Integrating over the SED and the infrared component out to 4.5 μm yields a total luminosity of 2 × 10⁵L_⊙. Thus, I suggest a 5000–6000 K yellow hypergiant for the progenitor. This is different from a red supergiant, an M-type star on the far red side of the HR Diagram. Temperatures for RSGs are <4000 K, typically about 3600 K (Levesque et al. 2006).

A curious feature of the pre-collapse light curve is the decrease in brightness beginning in mid-2008 observed at all of the optical wavelengths during which the star also got significantly redder; indeed, with the colors of a true red supergiant before the brief outburst. Many of the yellow hypergiants are observed to have high mass loss episodes as recorded in the complex circumstellar ejecta of IRC +10420 (Humphreys et al. 1997), the brief dense wind episodes in rho Cas (Lobel et al. 2003) during which it gets redder with the appearance of TiO bands, and most interestingly in the light curve for Var A in M33 (Hubble & Sandage 1953). Var A had a high mass loss episode, but instead of getting brighter it declined several magnitudes due to a shift in its bolometric correction forming a cool dense wind with TiO bands and the formation of dust (Humphreys et al. 1987, 2006). Its cool dense wind, high mass loss state lasted about 50 yr.

The progenitor's surface may have been experiencing an event similar to Var A when the core collapse occurred. During this same time the infrared shows a slow increase in flux very likely due to the enhanced mass loss and dust formation. Adams et al. (2017) attribute the subsequent brief brightening to ejection of the supergiants envelope. Post-collapse photometry shows very blue colors (Figure 1). But it is uncertain how realistic the published photometry and colors are. Some of the reported very faint magnitudes may be equivalent to non-detections. HST and Spitzer visual and near-infrared magnitudes from 2015.7 present a different picture. The SED from this time reveals a source with most of the radiation arising from the 3 to 5 micron region. It is unclear if this is remnant dust from the progenitor's circumstellar ejecta or new dust formed from the outburst.

In summary, the pre-outburst and pre-collapse observations of NGC 6946-BH1 show that the progenitor was a yellow hypergiant probably experiencing high mass loss before the collapse. Some might question what difference does it make, yellow supergiant or red supergiant. This star was obviously highly evolved and was therefore in a post-RSG state immediately before the core collapse. Not only was its surface hotter, its radius was much smaller, and its atmosphere or envelope was denser; all factors that will make a difference in the models for the outburst and collapse.