≡ Menu

Unusual Planets in a Close Binary System

The three Alpha Centauri stars get more and more interesting as we begin to discover planets around them, and the hope of finding planets in the habitable zone around Centauri A or B continues to drive research. Alpha Centauri could be thought of as a close binary with a distant companion, since we’re still not absolutely sure whether Proxima Centauri is gravitationally bound to the system. Learning more about binary systems, in any case, is interesting in itself but also may open windows into our nearest stellar neighbors.

Thus the discovery of planets in the binary system HD 87646 draws my attention. Here we have a primary star, HD 87646A, about 12 percent more massive than the Sun that is some 22 AU away from another star, HD 87646B, the latter about 10 percent less massive than the Sun. Translated into local terms, that would be something like having another star at about the distance Uranus is in our Solar system.


Image: The HD 87646 system, seen here in adaptive optic imaging from Palomar Observatory. Credit: Jian Ge/Bo Ma.

Compared to the Alpha Centauri system, the numbers are close. Centauri A and B orbit a common center of mass, separated by an average distance of 23.7 AU. But this is a highly elliptical orbit, with the average encompassing a swing between 36 AU and a close 11.4 AU, the latter not much farther than Saturn in our system. Proxima Centauri’s separation is wide, some 15000 AU plus or minus 700. If it is indeed bound to Centauri A and B, its orbital period is on the order of half a million years.

We learn in new work from Jian Ge (University of Florida) and postdoc Bo Ma that the HD 87646 binary contains planets, and unusual ones at that. MARVELS-7b (the name stands for Multi-object APO Radial Velocity Exoplanet Large-area Survey, a part of the Sloan Digital Sky Survey-III program) is 12 times the mass of Jupiter. A second world, MARVELS-7c, is 57 times Jupiter’s mass. We’re right on the edge here between giant planet and brown dwarf. Both these massive objects orbit the primary star at 0.1 and 1.5 AU respectively.

These are worlds that would have had to accumulate far more dust and gas than the kind of circumstellar disks we’re familiar with can offer, an indication that they must have formed through a different mechanism. The systems’ stability also raises questions about the formation process. One possibility considered in the paper is that the two planets formed as stars with their hosts, the result of the fragmentation of a large molecular cloud into four pieces. But the authors doubt that fragmentation on such a small scale can occur.

But if the paper casts doubt on the fragmentation model above, it also doubts the core accretion model, in which the gradual buildup of a planetary core pulls in gas from the circumstellar disk. In the HD 87646 system, forming two massive objects through core accretion would require a disk more massive than is usually found in close binaries.

That leaves disk instability as the formation model. Here gas in a high mass disk collapses quickly to form gas giant planets or, in this case, a possible brown dwarf. While core accretion models demand million year timeframes, disk instability can produce a planet within a few thousand years, assuming a circumstellar disk could cool quickly enough to create a bound object that continues to condense. The method can help us understand how massive planets could form at large distances from their star. But can it help us in a close binary situation?

The process is problematic for HD 87646:

Considering the primordial disk mass should be a lot more massive than the planet and brown dwarf mass, the disk is likely to be gravitationally unstable throughout the disk. This is consistent with gravitational instability leading to planet formation. Although several advantages exist for the disk instability model, we consider that such an explanation for the formation of the HD 87646 system should be taken with caution because whether disk instability can be triggered in the present of a close stellar companion remains an issue under debate…

Exactly so, and we also need to learn how the giant planet MARVELS-7b moved into its current low eccentricity orbit. The authors make the case that the planet formed in the disk and migrated inward to its current position, while the brown dwarf (MARVELS-7c) moved because of gravitational interactions in the system to a higher eccentricity orbit. The complexity of the interaction between other objects in the system and the smaller of the two binary stars leads the authors to call for future Gaia observations to constrain these orbits.

We’re left with disk instability as the best formation candidate, while conceding that HD 87646 still has secrets we need to unravel. On a broader front, it’s interesting to note that this work marks the eleventh detection of substellar companions in a binary system with separations of about 20 AU. We have no such detection for Centauri A or B and have to remove the putative world Centauri Bb from this list, but the remaining systems are all going to help us make sense of what we see in our stellar neighbors. HD 87646 is the first multiple planet system detected in binaries as close as this, but we can assume we’ll soon be adding to the catalog.

The paper is Bo Ma et al., “Very Low-Mass Stellar and Substellar Companions to Solar-like Stars From MARVELS VI: A Giant Planet and a Brown Dwarf Candidate in a Close Binary System HD 87646,” Astronomical Journal Vol. 152, No. 5 (7 October 2016). Abstract / preprint.


Comments on this entry are closed.

  • DJ Kaplan November 3, 2016, 12:38

    What’s the apparent angular separation for HD 87646’s elements?

    • DJ Kaplan November 3, 2016, 14:28

      Never mind, got it: 401 ± 12 mas. Compared to 2 arcseconds for the Alpha Cen system.

      Very interesting system at any rate.

  • Ashley Baldwin November 3, 2016, 13:52

    Thanks Paul. Very topical and though provoking .

    “Stability of multi planet systems in binaries ” , Marzari and Gallini , Arxiv , 16 Sept 2016, looked at just this issue recently and is a good accompanying read . Their work showed that it was almost impossible for multi planet systems to exist for binaries closer than 20 AU. For 20-3o AU the chances of stable orbits for more two or more planets was less than 10 % even for a system with a low eccentricity such as 0.1, the lowest they plotted and an important factor in orbital stability for all separations though diminishing with distance . For higher eccentricities the likekihood of multiple orbits dropped off precipitously unti semi major axes of 60 AU or so when numbers increased dramatically .

    The problem is that close binary systems and especially those with high eccentricity leads to powerful destabilising resonances between planets in different orbits around individual binary constituents AND also between individual planets of one constituents system and the companion star . To have two planets in orbit around one star of this system eccentricity must be around or below 0.1 , which is well below the average ( 0.3- 0.4) found to date. Although possessing a semi major axis of 23.7 AU, Alpha Centauri has an eccentricity of 0.5 unfortunately .

    I’m still hoping that the current Alpha Centauri planet Bb is a spurious finding given its unpleasant proximity to the star and the fact I fear that otherwise Sun like Alpha Centauri A and B may be limited to a maximum of just one planet each . Hopefully in their habitable zones , which do exist within the maximum outer individual stable orbits of each star .

    • Andrew Palfreyman November 4, 2016, 0:11

      Thanks for that final sentence – a key datapoint and one which encourages re. the Centauri system.

      • Ashley Baldwin November 7, 2016, 13:13

        I’m on your side ! I was disappointed at the findings of this study, no less for its robust methodology . I fear that essentially it’s findings are accurate . The “current” Alpha Centauri Bb is being increasingly discredited though, so hopefully there is still room for a habitable zone planet around each star, which lets face it , we would all be happy with ( single planet stable orbits out to about 2.3 AU , nicely covering the complete hab zones of a G2 and K1 star ) . This especially as given they would likely be terrestrial in nature as anything bigger would likely have been found by now. If there were indeed two , it’s a strange place to imagine living.

    • andy November 7, 2016, 15:09

      That study mainly considers inner planets that are already located fairly far out (3 AU for the majority of the discussion and examples presented). Two-planet systems with inner planets closer to the star should still be stable.

      Their equation 3 gives an estimate of the limiting binary semimajor axis over a range of parameters. Inverting the equation to solve for the location of the inner planet and using the parameters of the Alpha Centauri system, I get a location of 1.38 AU for the outermost location of the inner planet that would allow a two-planet system to be stable. As a caveat, this is outside the parameter range for which the equation was derived (inner planet locations 2–4 AU).

      • Ashley Baldwin November 8, 2016, 12:30

        Fine by me , and all within the habitable zones in which case more good news. Let’s hope any planetary systems arent too compact and near the star as a result . Just 11AU at closest is not a lot. Turning it round again , it’s interesting how even compact planetary systems ,as seem to be the vogue around many stars, don’t have more issues with inter orbital resonances regardless of being in a close binary or not . But their existence suggests not.

      • Ashley Baldwin November 8, 2016, 12:35

        It’s surprising they didn’t do that too given the thrust of paper. I would point it out to them and see what they say. I didn’t get the impression this was about Alpha Centauri in any way but it’s got to be relevant to their work though given that’s it’s both the nearest and highest profile binary of all with even more riding on it .

      • Ashley Baldwin November 8, 2016, 12:47

        So based around luminosities of 1.52 Sun and 0.5 Sun ,bthe Hab zones for A and B should be about 1.169- 1.722 and 0.665- 0.98 AU . All clear for B but cutting it close for A.( as part of a multi planet system anyway. )

        • andy November 8, 2016, 15:21

          Remember, it is only the inner planet that is considered by that equation. If the habitable zone planet is the outer planet, then you’re probably fine for a multiplanet system.

          One further point about that study: the planets being considered are Jupiter-mass (mentioned at the start of section 2). Anything of that mass in an S-type orbit around Alpha Centauri A or B would have been discovered by now (at least I think so, not sure what the limits are for orbits that are closer to being face-on). Smaller planets are likely to be less disruptive, therefore better for the prospects of multiplanet systems. Unfortunately they don’t explore the situation with different planetary masses, presumably because of the difficulties of fully exploring the huge parameter space of the problem.