Can merging white dwarfs form variable stars?

In August of 2017, the world was introduced to the idea of a kilonova, an explosion resulting from the merger of two neutron stars. LIGO and VIRGO detected gravitational waves coming from the lenticular galaxy NGC 4993, and an electromagnetic transient was observed soon after. While the idea of a kilonova had been explored for years, no direct detection had been confirmed. Suddenly, astronomers had hard data on an fascinating new type of explosion with implications for gravitational wave astronomy, our understanding of gamma ray bursts, and much more.

NGC 4993, showing Hubble Space Telescope observations of the afterglow from the kilonova in August 2017.
Even a week or two after gravitational waves from two neutron stars in NGC 4993 were detected, the Hubble Space Telescope and other observatories were able to see the afterglow. Image credit: Hubble Space Telescope/NASA/ESA.

Kilonovae aren’t the only interesting events involving double-degenerate progenitors, of course – far from it. For instance, we expect mixed-remnant collisions to also produce observable transients. Some of the most exciting possibilities, though, involve the results of these mergers. The 2017 kilonova, for instance, likely produced a massive neutron star that subsequently collapsed into a black hole. Other double-degenerate events, however, can lead to stable products.

Take, for instance, the merger of two white dwarfs in a binary system. If the combined mass is high enough, it’s possible to ignite a Type Ia supernova – a process that has thrown a wrench into the idea of using these supernovae as standard candles. However, in the case of a CO white dwarf and an He white dwarf, the resulting product might become something called an R Coronae Borealis variable (R CrB), which can survive for something on the order of 10,000 years.

A light curve of R Coronae Borealis from 1990 to 2017, exhibiting periodic brightness fluctuations by about 8 magnitudes and a large dropoff around 2007.
The light curve of R Coronae Borealis from 1990 to 2017. Note the irregular drops in brightness of about 8 magnitudes, as well as the extreme dip around 2007. Image credit: Wikipedia user Lithopsian, under the Creative Commons ShareAlike Unported 3.0 license.

R CrB stars have been know to exist for over 200 years, since their namesake yellow supergiant was discovered in 1795. In the intervening centuries, several of their properties have stood out:

  • They exhibit short variations in brightness on the order of 0.1 magnitudes, changing over about one month.
  • They are extremely hydrogen-poor (with some hydrogen-free), but may exhibit higher-than-usual levels of carbon and nitrogen.
  • From time to time, R CrB variables may dim by up to eight magnitudes, in addition to their ~0.1 magnitude fluctuations. This may be due to a dust envelope.

In short, they constitute a class of carbon- and oxygen- rich helium giants with large differences in hydrogen abundances, as well as both low- and high- amplitude variations in luminosity. The difference in hydrogen has been taken by some to indicate that a number of different formation mechanisms are at work. These mechanisms can be divided into two main classes: single-star and binary-star mechanisms.

Single-star models often involve a late thermal pulse occurring in the central star of a protoplanetary nebula, in cases where the outer helium layer is massive enough to trigger a shell flash. This should produce carbon through the triple-alpha process, as well as oxygen through secondary pathways.

The most common binary-star model is the one I alluded to at the start. In this model, a 0.6 solar mass carbon-oxygen white dwarf accretes matter from a 0.3 solar mass helium white dwarf, eventually triggering a helium shell flash and sending the product up the Hertzsprung-Russell diagram to become a yellow giant on a timescale of a few hundred years, with possible earlier hydrogen shell burning if the He dwarf is not hydrogen-free. After about 10,000 years, the giant will evolve bluewards, eventually ending its life for good.

Figure 1, Saio & Jeffrey 2002. Evolutionary tracks for an R CrB variable entering the yellow giant phase of its life.
Figure 1, Saio & Jeffrey 2002. Evolutionary tracks for an R CrB variable after accretion begins depend on its hydrogen content, but they all end up with luminosities of ~10,000 times that of the Sun.

The explanation for the pulsations observed in R Coronae Borealis variables is slightly more mundane. Radial pulsations are believed to be responsible for the ~0.1 magnitude variations, while ejected circumstellar dust likely leads to the much more dramatic dimming. These ideas are not directly tied to the progenitor models, which are motivated mostly by the peculiar chemical abundances of these low-mass yellow giants.

R CrB stars are only one class of variable star, but like all chemically peculiar stars, their abundances are clear signs of an unusual past. I’ve walked through the major binary-progenitor mechanism, but there are other ideas out there, some more exotic. Assuming that the white dwarf-white dwarf model is indeed responsible for certain subsets of these hydrogen-poor stars, however, we might have a really interesting opportunity to study the future of double-degenerate systems.


The Journey Begins

Welcome to An Eye on the Universe! This is the unofficial community blog of Astronomy Stack Exchange, a question and answer site on the Stack Exchange network devoted to helping astronomers, astronomy enthusiasts, and anyone else interested in the subject. This site is still under construction, so we apologize for the appearance, but we’ll have a pretty stellar design up shortly.

And yes, future writers, we will have LaTeX math:

\textrm{\$latex\ \textbackslash alpha\$ \ becomes\ } \alpha