A Brief Introduction to Core Collapse Supernovae

Core collapse supernovae (CCSNe) are the swansong of massive stars (greater than 8 solar masses at birth), whose cores collapse once they have exhausted all their fuel and can no longer compete against gravity.

Most of the energy of CCSNe (of order 10e53 erg - that is 10e46 J if you’re not an astronomer) is released in the form of neutrinos; ∼1 % of the total energy serves to accelerate the ejecta (kinetic energy),and only ∼0.01 % is converted into electro-magnetic radiation (Smartt, 2009). Still,they can be as bright as a billion suns and sometimes even outshine their host galaxy.The explosive deaths of massive stars enrich the interstellar medium with the intermediate and heavy elements that were created in the progenitors or the supernova themselves. Some of these elements (e.g oxygen) are crucial to form planets and give rise to life as we know it. The shock and energy released in the explosion causes positive and negative feedback with the nearby interstellar medium, playing a crucial role in star formation and in shaping galaxies (e.g. Hensler 2011; Scannapieco et al. 2008).


The Supernova classification…

… is a mess.

It is often the case with historical taxonomies. The classification has been refined and extended over time as objects with different spectral and light-curve characteristics were observed. Because a picture is worth a thousand words, here is my representation of the supernova classification.

Figure 1: My schematic representation of the supernova classification.

Figure 1: My schematic representation of the supernova classification.

Initially, supernovae were divided into type I and type II according to whether their spectra was hydrogen poor or hydrogen rich (mainly Balmer lines), respectively (Minkowski 1941). Type I supernovae are sub-divided into the Ia, Ib and Ic categories. The spectra of type Ia supernovae are characterised by strong Si II and Ca II H&K. Type Ib supernovae, on the other hand, do not show strong silicon features, but exhibit significant helium lines. Lastly, type Ic supernovae show no helium, and silicon absorption features are much less prominent than in type Ia supernovae (Filippenko, 1997).

Additionally, some type Ic supernovae exhibit very broad and blended spectral features, owing to very high ejecta speeds (e.g. SN 2002ap, see Mazzali et al. 2002), and are labelled broad-lined type Ic (Ic-bl). Some of these Ic-bl supernovae have also been associated with long Gamma-Ray Bursts (GRBs) and X-Ray Flashes. SN 1998bw (Patat & et al., 2001) was the first convincing candidate for the association of a supernova to a long gamma-ray-bursts. Since then, many other examples have been studied with various GRB energies – e.g. SN 2003dh / GRB030329, (Stanek & et al., 2003); SN 2006aj / XRF060218, (Sollerman et al., 2006); SN 2010bh / GRB100316D, (Chornock et al., 2010).

Hydrogen rich (type II) supernovae are further divided into the type II-P, II-L, IIb and IIn. The type II-P and II-L are defined according to whether their lightcurve shows a plateau or decay linearly (in magnitudes), respectively (Filippenko, 1997). In type IIb supernovae, hydrogen features are present at early days, but diminish over time as helium lines gain in strength. Finally, the spectra of type IIn SNe exhibit narrow (few hundred km/s ) hydrogen emission lines, caused by interaction of the supernova shock with circumstellar medium (dust and gas around the star).

Popular explosion mechanisms

When a massive star reaches the end of its life and its core has exhausted its fuel, the inward force exerted by gravity is no longer balanced. As a result, the core collapses to a neutron star or blackhole, and a supernova explosion ensues (or a failed supernova). It was initially thought that the bounce of the forming neutron star launched a shock wave through the in-falling outer layers causing the explosion; however, this shock is insufficient to unbind the stellar material. One of the most popular ways to shock revival is through tapping of the energy released in the form of neutrinos as the neutron star forms (neutrino heating; for a review see Janka 2012).

Furthermore, an alternative model involving jets have also been proposed in order to explain the existence of GRB-SNe: the collapsar model. In this scenario, the core of the massive star collapses to a black hole and accretion onto the black hole taps the rotational energy of the star via magnetic coupling, resulting in collimated jets which power the explosion and yield the GRB (e.g Woosley 1993; Woosley & MacFadyen 1999). Consequently this model requires the progenitor star to retain a high level of angular momentum for jet production (j > 3 × 10 16 ergs/cm^2 – Woosley 1993). Recent studies have shown that for central engines of short enough lifetime, the jets may fail to break out of the envelope of the progenitor, but would impart sufficient energy to drive a rapid expansion, hence yielding type Ic-bl supernovae with gamma-ray-burst counterpart (Bromberg et al. 2011, Lazzati et al. 2012).

Likely Progenitors

Figure 2: Relative numbers of core collapse SNe as calculated from table 3 of Shivvers et al. (2017) and fig. 9 of Li et al. (2011).

Amongst CCSNe, type II-P are the most prevalent, representing about 57 percent of the population by volume (see Figure 2– Shivvers et al. 2017). Their (single star) progenitors are red supergiants which would have started their lives with masses between 8 and ∼20 Msun (Smartt, 2009). Direct evidence as been found through the analysis of pre-images revealing red supergiant stars at the location of the SN (e.g. Van Dyk et al. 2003, 2012), and in some cases the late images taken after the SN event showed that the suspected progenitor had disappeared, like in the case of Type II-P SN 2003gd (Maund & Smartt, 2009). The presence of hydrogen in type IIn SNe points to progenitors of a similar mass range, although some type IIn SNe are known to have progenitors of higher mass (e.g SN 2005gl; Gal-Yam & Leonard 2009). Type II-L SNe may also have slightly higher ZAMS progenitors masses than most type II-P, around 20M, although more data is required (Branch & Wheeler, 2017).

The type IIb, Ib, and Ic(-bl) SNe are collectively called stripped envelope SNe, owing to the nature of their progenitor stars which have been stripped of their outer H-rich envelopes to varying degrees. Together they represent ∼30 percent of CCSNe. Type IIb SNe progenitors have lost most of their hydrogen envelope, retaining less than 0.5 M of hydrogen (Smith et al., 2011), whereas type Ib SN progenitors must have lost their hydrogen envelope entirely. Furthermore, the stars yielding type Ic SNe must have shed both their hydrogen and their helium layers.

In single stars model, the best candidates for the progenitors of stripped-envelope SNe are Wolf-Rayet stars (WR). The WR class is named after Wolf and Rayet who identified stars characterised by broad-emission lines (Wolf & Rayet, 1867), which are the result of their fast (a few hundred to a few thousand km/s ), dense (∼10 −5 M /yr) winds. To reach the WR phase, an O star must have been heavily stripped during luminous blue variable (LBV) or read super-giant (RSG) phase and in most cases have lost its hydrogen envelope. This evolution can only be attained by the most massive stars (M ZAMS >25M – Crowther 2007).

There are three main sub-types of WR stars. Nitrogen sequence WR stars (WN) are characterised by very strong N iii-v emission lines; these stars are hydrogen poor but helium rich. They are anticipated to end their lives as type Ib or potentially type IIb SNe. Carbon sequence WR stars (WC) show very prominent N III and C IV emission lines, and oxygen sequence WR stars (WO) are defined by their strong O vi λλ3811 − 34 emission lines. Both WC and WO are hydrogen and helium poor, and WO stars in particular are thought to be very close to core helium exhaustion (e.g Langer 2012, although see Tramper et al. 2013); they are much rarer than WC stars. The degree of stripping of WC/WO stars make them candidates for the progenitors of type Ic SNe.

In the case of Ic-bl SNe driven by jets in a scenario akin to that of a collapsar, the progenitor must retain sufficient angular momentum. This may be made easier if the progenitor has low metallicity (the winds are line driven, the lower metallicity implies less mass loss and therefore less angular momentum loss). This is supported observationally as Ic-bl SNe are found preferentially in low-metallicity environments, with GRB-SNe being found at lower metallicities than Ic-bl SNe without GRB counterpart (e.g Modjaz et al. 2008; Levesque et al. 2010; Graham & Fruchter 2015).

The single star progenitor scenario, however, can be difficult to reconcile with observations and stellar evolution. At solar and sub-solar metallicity the mass loss rate of single WR stars could result in final stellar masses producing SN Ib/c with light-curves too broad and ejecta masses too high to fit observations (Yoon, 2015). Additionally the WC/WN ratio is between 0.1 and 1.2 (at SMC and Milky Way metallicity, respectively – Crowther 2007), whereas the Ic/Ib ratio as ∼2 (Smartt, 2009). Furthermore, the presence of a binary companion to the progenitor of the type IIb SN 1993J was confirmed by Maund et al. (2004).

Figure 3: Cartoon of a possible binary route to type Ib and type Ic SNe. Inspired by fig. 6 of (Yoon, 2015).

The binary route has received a lot of attention in recent years, since binary interactions can strip donor stars effectively, and we know that most massive stars are in binaries (Sana et al., 2012). Mass transfer with a secondary star through Roche Lobe overflow could strip the primary sufficiently to produce a type Ib (M He > 0.14M ; Hachinger et al. 2012) or a type IIb. Additionally, the progenitors that follow this evolutionary path could have ZAMS as low as 12.5M (Yoon, 2015). It can be difficult to strip the primary star sufficiently to yield a type Ic SNe, but if the progenitor is the secondary star undergoing mass transfer to a compact companion, the mass loss rate could be high enough to strip the helium layers efficiently before the SN explosion (Yoon, 2015). An example of a binary route yielding type Ib/c SNe is given in Figure 3). Eldridge et al., 2008 pointed out that a combination of both single and binary systems shows better agreement with the rate of explosion of CCSNe. In the past decade, increasing evidence has shown that the binary route is not only important, but dominates the evolution towards stripped-envelope SNe: Smith et al. (2011) showed that stripped-envelope SN rates could not be solely explained by single star evolution, and more recently Prentice et al. (2019) found that the average ejecta mass of stripped-envelope SN was 2.8±1.5Msun , which suggests that a majority of these SN arose from lower mass progenitors with M_ZAMS < 25Msun .

The shape of supernovae

For a long time, the ejecta of SNe were assumed to be spherical, however, indirect evidence called this belief into question, such as the high velocity kick of pulsars (Lyne and Lorimer 1994), the asphericity of young SN remnants (Manchester, 1987), or light echoes (Grefenstette et al. 2014). In 1982, Shapiro and Sutherland were the first to propose polarimetry as a way to probe the geometry of the unresolved SNe envelopes, and, nearly 25 years later, spectropolarimetric data seemed to demonstrate that virtually all CCSNe were aspherical (Wang and Wheeler, 2008). Analytical models and simulations have shown that effects necessary for successful explosions such as magnetic fields, rotation and various instabilities, yield non-spherical explosions (e.g. Blondin et al. 2003; Burrows et al. 2006; Takiwaki et al. 2016). Hence, characterising the geometry of supernovae is crucial to testing explosion models, and to that effect, spectropolarimetry is a decisive tool. You can find a brief introduction to specpol here.

Was this useful? Something not clear?

Let me know!

Hit me up on Twitter @Sydonahi! You can also checkout my contact detail on this page —> click me <—

All illustrations available in HD: Feel free to ask me!