W Aquilae — or W Aql for short — is a star that has shined brightly across my research career to date. Partly, it shines brighter because it is not just one star, but a binary system containing a dying star and a second star very similar to our Sun. The dying star is an AGB star (short for asymptotic giant branch star), and it is dying in the same way that we expect our Sun will, in about five billion years. The biggest difference in their deaths will be that the Sun will die alone, whereas the dying star in W Aql has a companion star leaving its mark on the dusty death-throes of its final moments (speaking on an astronomical time scale).

We have known since the 1960s that W Aql was made up of two stars, but it wasn’t until modern telescopes — starting with the Hubble Space Telescope — that we were able to visually separate them out. These and other recent observations showed us that, over the span of around 15 years, the two stars in W Aql were not moving very much relative to each other. That means the orbit of the two stars around each other is quite slow and must take a long time. Since we can measure that the two stars are at least 200 au apart, they shouldn’t have much of an impact on each other. Or so we thought.

New high-resolution observations from ALMA (the Atacama Large Millimetre/sub-millimetre Array, located in Chile) allowed us to examine structures in the stellar wind — gas and dust that the dying star is spewing out. This dense wind is in the process of carrying away a significant portion of the AGB star’s mass and once the star’s outer layers have been expelled, all that will be left behind is the current core of the star. Eventually, W Aql will become a planetary nebula with a white dwarf orbiting a Sun-like star at its centre.

Right now, what we see is a lot of structure in the stellar wind, some of which looks quite chaotic. We thought this might somehow be caused by the Sun-like companion ploughing through the wind, but until now we didn’t understand exactly how this might be happening. 

Generally, when we detect molecules around AGB stars, they are distributed in roundish blobs around the star, or maybe in a roundish ring if they do not form close to the star. There might be some structures that we can see inside those blobs, but they usually have some degree of symmetry to them. That’s why, when we found emission from the SiN molecule located in more of a triangular blob to one side of W Aql, we suspected something strange might be going on. After a bit more analysis, we realised that the SiN was actually located in an arc on one side of the star, but rotated so that we were seeing it edge on from Earth. You can see this in what we call the position-velocity diagram below.

SiN — a simple molecule made up of silicon and nitrogen — has not been studied much before in astrophysical contexts. To understand how it forms, and how it could have formed on only one side of the star, we turned to astrochemistry. My colleague Marie Van de Sande had determined that more SiN can form if there is a nearby source of UV light. In the context of W Aql, that could be the sunlike companion to the dying star. However, right now the sunlike star is on the opposite side of the dying star to where the SiN is located. So how could that work?

In the end, we realised that everything would make sense if the two stars had a highly eccentric orbit. In contrast to the solar system, where the planets’ orbits around the sun are close to circular, the two stars in W Aql orbit each other along a very elliptical path. With the help of hydrodynamic simulations performed by PhD student Jolien Malfait, we were able to determine the orbital properties of the stars. At their closest point, they pass within only a few au of each other, which last happened around 200 years ago. The stars take about 1000 years to complete a full orbit. The anomalous SiN emission was formed when the sunlike star passed through the dense inner region of the dying star’s stellar wind, irradiating it with UV and triggering various chemical reactions. The arc of SiN that we now observe originally formed in the wake of sunlike star, but has since expanded outwards with the rest of the stellar wind. A diagram of this process is shown below.

Now that we have a working example of patterns of molecular emission that can point to past binary interactions, we can search for similar patterns around other stars. This includes AGB stars that show evidence of binary interactions but for which we haven’t already detected an optical companion. If the companion is presently too hidden by the AGB’s dust to be seen, we could still determine some of its properties based on the molecules it helped create.