Stars are not fixed objects, they have instead a complex evolution. The final stage of some types of stars are given by supernova explosions, releasing huge amounts of energy into the interstellar space. The actual problem is identifying the star that gives rise to these explosions, as telescopes are not so powerful and it is not possible to track the evolution of stars (telescopes were invented 400 years ago and the typical lifetime of a massive star is 107 years, that is, a factor of 30000 times more). In this paper, we confirmed observationally that certain types of the most violent stellar explosions (Type Ic supernovae) do not come from a single star, but from two stars, a binary system. This improves our knowledge of how stars and supernovae affect their surroundings, and hence, how galaxies evolve.
The motivation of this research was based on understanding the nature of these kinds of supernovae using an approach never used before: high-resolution data + a large sample of supernovae. To do so, we studied the molecular clouds at the supernova locations, thanks to the Atacama Large Millimeter/submillimeter Array (ALMA) telescope, the greatest telescope of its kind, located in the north of Chile at an altitude of 5000 metres. The importance of ALMA was unique in this research as it is the only telescope in the world capable of reaching resolutions of the size of molecular gas in nearby galaxies. In Fig. 1 it is possible to appreciate some examples of the environment of molecular clouds for the supernovae.
Figure 1. Example of molecular cloud environments at the supernova explosion.Caption
As our original idea was to compare the locations of different types of supernovae, we did it, however, when the sample was not completed, we worked with a reduced number of the supernovae. This gave us a totally different conclusion from the current stated in the paper (“Binary progenitor systems for Type Ic supernovae”, as in the title). Instead, we got that Type Ic supernova progenitors were very massive stars (because the molecular hydrogen environment of this population was much higher than Type II supernovae). The data at that moment consisted of only 5 Type Ic supernovae, hence, obviously the sample was biassed. Taking in consideration this last was of very importance because we ask to ourselves therefore, “when the supernova sample is not biassed anymore? In order to solve this problem we had to figure out at what sample size we were able to make a claim, independently of the results.
It was mentioned before that the novel of this research was the use of high resolution data (that we fulfilled), plus the use of a large sample of supernovae. To be concise with the second statement, it should be modified to the use of a statistical large sample of supernovae. In order to show quantitatively when the sample is biassed and it is not we had to include statistical tests.
Three statistical tests were computed: first starting from the measured gas densities of Type II supernovae, second from the measured gas density distribution in PHANGS galaxies, and the last from lifetimes of binary systems from a numerical model. In summary, the three methods gave us that for our current sample (30 Type II and Type 21 Ic supernovae), for 95% of the cases we are able to conclude that both populations of supernovae have similar masses. Running these statistical tests was essential in order to satisfy the requirements of Nature Communication science impact policy.
This effort was possible thanks to a tremendous collaboration between 21 people, based in 9 different countries (Poland, Spain, Denmark, Greece, Sweden, Italy, Germany, UK, and France).