Planet Formation Around Close Binaries

by Paul Gilster on October 25, 2010

Planets around binary stars fascinate me, doubtless because of Alpha Centauri’s proximity and the question of whether there are planets there. About ten percent of the planets we’ve found around main sequence stars are found in binary systems, and most of these binaries have wide separations, in the range of 100 to 300 AU. But, like Alpha Centauri, close binaries remain promising targets. I’m looking at a new paper by Andras Zsom, Zsolt Sándor and Kees Dullemond (Max-Planck-Institute für Astronomie) dealing with early stage planet formation in binaries, and they’re quick to note that planets in close binary systems put constraints on planet formation theories.

After all, if we find planets in these systems, our planet forming theories have to produce satisfactory explanations for their existence. Does core accretion, then, work in these environments? We can look to close binaries with planets, systems like Gamma Cephei (separation 18.5 AU), GL 86 (18.4 AU), HD 41004 (23 AU) and HD 196885 (17 AU). What Zsom et al. are interested in is the question of the growth and fragmentation of dust in binary systems, the gas and dust disk around the primary star being perturbed by the secondary, and their paper compares their model — with and without eccentricity — to the dust population in disks around single stars.

Core Accretion and Its Problems

Think about how planets form under the core accretion model. Tiny dust grains eventually form planetesimals, which then form protoplanetary cores that over time gain a gaseous envelope (or not, depending on the presence of gas), and undergo a chaotic impact phase until their orbits are finally stable. It’s a complicated picture that binary stars make even more treacherous, and in several ways. For one thing, contemplate what a stellar companion does to the dust disk:

The tidal torques of the companion generate strong spiral arms in the disk around the primary. Angular momentum is transferred to the binary orbit which will truncate and restructure the disk… This dynamical effect of the secondary has several consequences which might influence planet formation: it decreases the lifetime of the disk, increases the temperature of the disk and modifies the stable orbits around the primary.

But it’s not just the early disk that’s affected. Here the authors go into what happens to the evolving planetesimals, and how their fate may shape planet formation:

The evolution of planetesimals is also influenced in a binary system. The perturbation of the secondary increases the relative velocity of the planetesimals and/or creates unstable regions where the planetary building blocks cannot maintain a stable orbit as shown by e.g., Heppenheimer (1978), Whitmire et al. (1998), Thébault et al. (2004). The increased relative velocities can then lead to the disruption of the planetesimals. However, Marzari & Scholl (2000) showed that the combined effects of the gravitational perturbation and the gas drag may increase the efficiency of their accretion by reducing their relative velocity and produce, later on, terrestrial planets. Marzari et al. (2009) calculated the relative velocity of planetesimals in highly inclined systems and concluded that planet formation appears possible for inclinations as high as 10◦ , if the separation between the stars is larger than 70 AU. The region where planetesimals can accumulate into protoplanets shrinks consistently for lower binary separations.

Of course, it’s the closer binaries that we’re interested in today. This paper homes in on the earliest stages of core accretion, the key issue being: can planetesimals form in the first place, or is the disk too thoroughly disrupted to allow the process? Some theories suggest that a ‘pressure bump’ forming around the snow-line could concentrate particles so that planetesimals could form in the inner system, but the authors’ simulations show that this effect quickly disappears in infant systems, so quickly that the needed planetesimals never emerge.

Various other models are out there that could concentrate particles as needed for the early stages of planet formation — the authors run through the possibilities — but they’re still under investigation and may be rendered unworkable by the presence of the secondary star. As the authors note, “Particle concentration mechanisms in general need to cope with the continuous gravitational stirring and perturbations of the secondary.” And we’re by no means sure that core accretion can function within these models when a secondary star is present.

A Different Formation Model?

If that’s the case, what about the planets we’ve already found around close binaries? Are they the result not of core accretion but gravitational instability, the alternate model of planet formation? In the latter, unstable regions in the protoplanetary disk form clumps of gas and dust which eventually coagulate into a single core. In this model, knots of matter collapse rapidly, to form planets in a much shorter time frame.

The gas giants we’ve found around close binaries thus far could be the result of gravitational instability. But here we have to fall back on the limits of our observations, for as some readers have already been discussing in the comments to a previous post, close binaries are tricky targets. We have yet to find low-mass planets around close binaries, but this may simply be because of the limits of our current techniques, and because of an observational bias that has kept such systems out of many early surveys. But if further work rules out core accretion in these environments, then the alternate model — gravitational instability — becomes the focus of more targeted study.

How dust turns into planetesimals, and thence to planets, is the subject of intense ongoing work. And if the Zsom paper has you discouraged about the prospects around Alpha Centauri, we’ll look tomorrow at another new paper that tackles early disk issues and comes to a different conclusion about what may be possible there. For today, though, the paper is Zsom et al., “The First Stages of Planet Formation in Binary Systems: How Far Can Dust Coagulation Proceed?” accepted at Astronomy & Astrophysics and available as a preprint. You may want to check the comments to my recent article on the SIM mission, where this paper first came up (and thanks to Centauri Dreams regulars andy and spaceman for the discussion).

tzf_img_post

Dave Moore October 25, 2010 at 20:27

The Palomar High-precision Astrometric Search for Exoplanet Systems (Astrometry of binaries with 10 to 40 au separation) has discovered, with a high degree of probability that the B component of HD 176051 (aka GJ 738) has a 1.5+ Jupiter mass planet in a 2.78 year orbit.

The binary is 49 light years from Earth and has a 1.07 solar mass GOV orbited by a 0.71 solar mass K1V, so is similar to the Alpha Centauri system.
The separation of the 2 stars is 19 au and the eccentricity is 0.27; although, this is not well characterized. The tidal disruption distance of the accretional disk is 0.32 au (Pichardo et al. 2005).

The preprint of the paper: THE PHASES DIFFERENTIAL ASTROMETRY DATA ARCHIVE. V. CANDIDATE SUBSTELLAR COMPANIONS TO BINARY SYSTEMS can be found at
http://arxiv.org/pdf/1010.4048v1

Michael Simmons October 26, 2010 at 0:06

Re: HD 176051
a link from the slashdot article.
http://exoplanet.hanno-rein.de/system.php?hash=ec845b906e3c7acdf4f922ece469e4ec

The image at the bottom suggests its in the habitable zone.

It seems to me, that if spin axis stability is vital to a stable environment over a very extremely long time span and if this is required for the development of intelligent life. Then large tidally locked Pandora like moon/planet seem a much more likely place to find intelligent life than a rocky inner planet.

The chances of another just right Theia impact leading to an earth moon like system with its high angular momentum, and tidal spin axis stabilisation must be extremely rare.

Ronald October 26, 2010 at 5:40

Looks promising. Even more amazing in this respect is the binary HD 188752, where separation is only 12.3 AU with an eccentricity of 0.50 (this candidate is controversial though).
If the HD 176051 binary has a separation of 18 AU and an eccentricity of 0.27, what then is their *minimum* separation? 13.9 AU, or 16.4 AU ?

Minimum separation of Alph Cen A and B is only 11.4 AU, which is really on the low side, even in comparison with any known binary planet hosts.
But even if this is too close for gas giants, there may still be smaller, terrestrial planets in the inner systems. The maximum distance at which terrestrial planets could exist in stable orbits around A and B varies, according to various models, for B from about 0.5 (Thébault) to about 2 AU (partic. Lissauer, Quintana, Guedes studies).

Ronald October 26, 2010 at 5:42

Small correction to my previous post: for the HD 176051 binary the separation is 19 AU, not 18 AU.

andy October 26, 2010 at 16:11

One of the things about the HD 176051 planet is that we don’t know whether it orbits the primary or the secondary star. The parameters quoted at the Extrasolar Planets Encyclopaedia are for the case where it orbits the secondary.

HD 188753 is in fact a triple star (the secondary is a close binary consisting of K and M-type dwarfs), unfortunately the claimed hot Jupiter planet hasn’t held up to scrutiny.

Gliese 86 (not 83!) is a weird one because it has a white dwarf secondary. Probably this is the most extreme environment yet found that has managed to produce a gas giant planet. It isn’t clear how it managed to do it.

Paul Gilster October 26, 2010 at 16:17

Thanks for the note re Gl 86 vs. 83. I’ve made the change in the original.

spaceman October 27, 2010 at 3:43

As for gravitational instability, I was under the impression that only gas giant planets could form via this mechanism. Has anyone heard if it is likely, or even possible, to form terrestrial planets by way of gravitational instability? Is there indeed a lower limit mass below which planets cannot form in this manner?

Here is a paper that looks at pre-planetary disks around binaries of a variety of separations. PLANET FORMATION IN BINARY SYSTEMS: A SEPARATION-DEPENDENT MECHANISM? Duchenne 2009

In this paper, the author determines that the tendency to find only high mass planets around tight binaries is not due to a selection effect or observational bias.

There is observational evidence that seems to support the Dullemond hypothesis: “Debris Disks around Solar-Type Stars: Observations of the Pleiades with Spitzer Space Telescope” Plavchan et al (2010)

They conclude:

“We confirm the results of Stauffer et al. (2009) that close, high-mass binary systems tend not to harbor debris disks. This behavior is probably associated with binary companions that orbit close to the zone where debris disks tend to lie (Trilling et al. 2007).”

Ronald October 27, 2010 at 5:49

What is really weird about Gl 86, is the fact that the mean separation (semi-major axis) between A and B is only 18.4 AU, with an eccentricity of almost 0.4. On top of that metallicity of A is rather low (Fe/H = -0.24).
And yet is has a gas giant of some 4 Jupiter masses in close orbit (0.11 AU). Makes one wonder about its history, maybe the two stellar components started at wider separation. Or is it also possible that a close binary somehow concentrates the primordial dust disc (i.e. not just sweeping it up)?

spaceman October 27, 2010 at 8:33

It was hypothesized in the case of HD196885 that perhaps the planets we are observing in close binaries may not be native to those systems; in other words, they may be captured from a close encounter with another planetary system. I find it hard to believe that with a growing list of these close binary planets that this unlikely event can explain them all, especially in the case of multiple planet systems. Or, perhaps the binary separation was not as great during the period of planet formation. In any case, this topic is one of the more fertile ones in all of astrophysics.

andy October 27, 2010 at 16:54

Ronald: with Gliese 86, the current orbital parameters are not really the problem. If you take into account the orbital expansion due to mass loss from the white dwarf progenitor, the original semimajor axis could easily have been around 12 AU. Plus the white dwarf progenitor would have been the more massive star, further reducing the stable region.

Nebula Gray September 7, 2011 at 10:31

Would planetary formation be feasible around extremely close binaries less than an AU apart? If they are close enough that they behave gravitationally like one star, they should only clear a small area around them with relatively unperturbed disc material within 2 AU. The dust cleared away may aid in the accretion of the nearest planet to form in the unperturbed section.

Comments on this entry are closed.

{ 1 trackback }