Planets in Binary Star Systems

The Kepler space telescope has been staring at a patch of sky to look for the dimming of light when a planet passes in front of a star. One of the interesting results coming from Kepler is the discovery of planets orbiting pairs of stars, more commonly known as binaries. These circumbinary planets are exciting because star formation tends to produce stars in pairs or groups. These binary stars are common, so finding planets among them is encouraging as it suggests planetary systems are also quite common. In this blog post I'll talk about some of the recent Kepler results and prior research on planet formation in binary star systems.

The Kepler-35 system. Credit: Lynette Cook / extrasolar.spaceart.org

How do planets form?
Planets are believed to form in gas and dust disks around stars. When we first started thinking on how the Earth and the solar system formed we realized that all the planets lie roughly in the same plane, what we call the ecliptic. The big planets were far from the Sun, whereas the smaller ones were close in. We developed the disk hypothesis to explain this scenario. Here's how it goes: the early Sun was surrounded by a disk of gas and dust. The material in the disk was used to form planets, but far from the early Sun the disk temperatures were low allowing many ices like water ice and carbon dioxide ice to exist. This facilitated the formation of large planets like Jupiter and Saturn, whereas close to the Sun the disk lacked these ices and the planets, like Earth and Venus, ended up smaller. Since the planets all formed from the disk, they all share the same plane, similar to the rotation axis of the Sun, all as a consequence of conservation of angular momentum.

This picture is nice and simple, but it has been challenged since we started finding planets in the mid-90s. One of the first problems was that most extrasolar planets (or exoplanets) orbited very close to their stars, but were very massive. These 'hot Jupiters', as they are called, have masses comparable to that of Jupiter in our own solar system, but orbit much closer than Mercury is to the Sun, whipping around their stars in a matter of days. For comparison, Mercury takes 88 Earth-days to go around the Sun, whereas Jupiter takes about 11 Earth-years. These hot Jupiter are very easy to detect and so were the first such exoplanets detected. We've revised the disk model of planet formation to include things like planet migration, where planets form far away and drift inwards, to explain how these objects form.

The HAT-P-11 system.
Credit: Subaru Telescope,
National Astronomical Observatory of Japan (NAOJ)

More recently, we've started to measure the plane of exoplanet orbits relative to the spin axis of their stars. We expected these to be aligned, but much to our surprise some systems are very different or are orbiting in the opposite fashion as the star is spinning. This again challenges the simple disk model I outlined in the first paragraph. Planets migrating inward wouldn't change their orbital plane, so perhaps dynamical interaction between multiple planets is more important here.






As more and more planets are discovered, we find new ways to challenge our models. This is how science grows: we develop a model, test it, and refine it, continuing until we can satisfactorily explain the phenomenon we observe. Here's one other system that's difficult to explain:

The Kepler-20 system. Credit: David A. Aguilar (CfA)

Kepler-20 depicted above hosts 5 planets, but the radius and masses of these cycle between big and small rather than having all of them nearly equal size or having them segregated by mass. This is yet another scenario that challenges our ideas on planet formation.
But enough about planet formation, let's talk about binaries.

What do we know about disks in binary systems?
Many young stars possess these gas and dust disks, also known as protoplanetary disks. Furthermore, roughly half of all stars are in binary or multiple systems. These systems possess more than one main star and the secondary star(s) will influence the formation and evolution of any other objects in the system.

Theoretical and numerical studies suggest that secondary stars will disperse protoplanetary disks as the gravitational influence of the secondary creates gaps or truncates the disks. This has also been seen in observations. Studies of binary protoplanetary systems do show lower disk masses than around single stars and more widely separated binaries. The cutoff appears to be in binaries with separations of 100 astronomical units or less. An astronomical unit, or AU, is the distance between the Earth and the Sun and corresponds to about 150 million kilometers. This defines the scales for studies of planetary systems and disk studies. For comparison, in our solar system Mercury is at 0.39 AU and Neptune is at 30 AU from the Sun.

My own research has dealt with debris disks, which are older disks on planetary evolution timescales. I wrote a brief blog post about debris disks some time ago. They also show similar trends as the younger protoplanetary disks and are consistent with the theoretical picture. Most stellar binaries have separations of order 20-40 AU and would very readily disrupt any disks forming in the system. If a planet is to form in a binary system, the two stars must be very widely separated (hundreds of AU) or very closely separated.

The V4046 Sgr circumbinary disk; the two stars are 0.04 AU apart. 
Credit: David A. Aguilar (CfA)

Are any planets known in binary systems?
This is the big question and we already know the answer to it: YES. While many studies avoid searching for planets in binary systems (due to the complications these pose), about 10-20% of planet bearing stars have a more distant companion. These companions tend to be very widely separated so it's unlikely that they have influenced the evolution of the system in any significant way. For a long time, it seemed that the secondary stars were at least 20 AU away from the primary, and more usually at least 100 AU. However, there were a few cases were circumbinary planets were announced. These only increased in number with Kepler.

Circumbinary planets and Kepler results
A handful of circumbinary planet candidates were known prior to Kepler. These were discovered by looking at eclipsing binaries and accurately timing the eclipses. If you understand the system well, you can accurately predict when one star or the other passes in front of, or eclipses, the other. However, in a few cases the eclipse did not occur at the expected time. There is a discrepancy between the observed and calculated eclipse time which suggest something else is in the system. An unseen planet, for example, can tug on the stars and mess up the timing. This is an indirect probe of the system, as the planet is not actually observed.

Kepler searches for planets by staring at stars and monitoring how bright they are. When a planet passes in front of the star, or transits, the star will become slightly fainter. Kepler has detected thousands of candidate planets and many of those are being confirmed as true planetary systems. One particularly exciting planet system, in my opinion, is Kepler-16, depicted below.

The Kepler-16 system. Credit: NASA/JPL-Caltech/T. Pyle

Kepler-16 was the first case where a planet was seen to transit two stars, which also mutually eclipse one another. This is undeniable evidence for a circumbinary planet. The two stars are separated by only 0.2 AU, but the planet is more than 3 times farther away with an orbit at 0.7 AU. The system is quite compact, but the stars are smaller than the Sun so the temperature of the planet (and any moons) is likely to be too cold to support liquid water and life as we know it. Since then, Kepler has discovered two more circumbinary planets: Kepler-34b and Kepler-35b. It looks like circumbinary planets aren't so rare after all!

While more and more circumbinary planets are being found, we're finding that these follow the theoretical expectations and disk observations. That is, the stars are very closely separated compared to the planet location. Hence, planets like Star Wars' Tatooine, may form out there if a planet-forming disk surrounds a very close pair of stars. Considering the fact that hundred of billions of planets may be in our galaxy (see here), then even with the restrictions imposed by a secondary star, there must be millions of circumbinary planets out there.

Bad Astronomy has an article on the Kepler-34 and 35 systems (also Kepler-16) in case you want to do further reading.