Knock, Knock. Who’s There? A Free-Floating Planet!

Editor’s Note: Astrobites is a graduate-student-run organization that digests astrophysical literature for undergraduate students. As part of the partnership between the AAS and astrobites, we occasionally repost astrobites content here at AAS Nova. We hope you enjoy this post from astrobites; the original can be viewed at

Title: Formation History of HD106906 and the Vertical Warping of Debris Disks by an External Inclined Companion
Authors: Nathaniel Moore et al.
First Author’s Institution: Georgia Institute of Technology
Status: Published in ApJ

When early astronomers theorized how planets formed, they often used the solar system as a model, mainly because that was all we had observationally available at the time. The thing is, the solar system is pretty “well behaved” — the planets are more or less in the same orbital plane, and their orbits are not too eccentric (i.e., they are closer to being circles than ellipses). However, as more exoplanets are found, astronomers begin to question their ideas for how planets formed. The binary system HD 106906, for example, has an asymmetrical debris disk and a planet that is not in the same plane as the disk and is separated from the stars by 730 astronomical units (au; for reference, Earth is 1 au away from the Sun). This system has an unusual architecture, and the authors of today’s article try to theorize how this system formed. Understanding the formation of an usual system like this one allows us to expand our knowledge of planet formation beyond the simplicity of such well-behaved systems such as our own!

illustration of the two possible scenarios for planet formation

Figure 1. The two possible scenarios for planet formation: accretion model (“bottom-up”) and gravitational instability (“top-down”). Click to enlarge.
 [NASA and A. Feild (STScI); CC BY 4.0]

There are two main theories for the formation of planets: core accretion and gravitational instability (also called the gas-collapse model). Figure 1 shows the two different scenarios. In both cases, the planets form in a protoplanetary disk, which means that the planets initially start in the same orbital plane. A system like HD 106906 challenges this notion, since it has a very massive planet far away from the disk and in a different orbital plane. The authors of this article explore the idea that the planet HD 106906 b actually formed from the disk, but a recent (about 1–5 million years ago) close encounter with a free-floating planet knocked the planet away from the disk into an eccentric orbit, and the interactions from this close encounter actually caused the disk to become more eccentric as well.

The authors explore this idea using N-body simulations (a simulation of how bodies interact over a period of time) of the system combined with simulations of how the observational data would look for this scenario. They then compare the simulations to real observations.

Companion and Disk Interactions

The authors first try to determine whether the HD 106906 system has been like this for a long time or if its current configuration is the result of a recent event. To do this, they simulate different variations of the planet’s eccentricity, inclination, and semi-major axis. For the simulations, they include the effects of radiation pressure. They also use two different central body configurations: one with a binary star system and another with a single central body and an extra J2 potential term, which emulates the binary system but is more computationally efficient. The main results from these simulations are shown in Figure 2.

simulation results after 1, 5, and 10 million years, compared to the original observations

Figure 2: The simulations at 1 million years (top left), 5 million years, (bottom left) and 10 million years (bottom right) compared to the Crotts et al. 2021 original observations. After a million years, the simulated disk is very similar to the real image. By 5 and 10 million years, the appearance (size and brightness) of the disk exceeds the observational constraints. [Moore et al. 2023]

The simulations lead the authors to conclude that the disk and planet have likely been in this configuration for only 1–5 Myr, which for the system’s age of 13 Myr is quite recent. If it had been there for longer than that, the simulations for 5 and 10 Myr would have been within the observational constraints from the real data.

Knock, Knock. Who’s There?

A cartoon of the three possible outcomes of the free-floating planet's visit.

Figure 3: The free-floating planet, represented in red, can simply “fly by” the system, leaving the original configuration mostly unchanged; be exchanged with the original planet (blue), which then gets ejected; or be captured into the system. [Moore et al. 2023]

Next, the authors simulate a close encounter between a 11±1 Jupiter-mass free-floating planet and the native planet of the HD 106906 system to see if this can cause the system’s current arrangement. Figure 3 shows the possible outcomes of simulations of this (un)expected visit. The authors simulate 100,000 initial conditions and see their outcomes. These 100,000 conditions are obtained such that the closest approach distance of the planets is less than 50 au (if it winds up being more than that, the initial condition is rejected).

The team’s final results are shown in Figure 4. From the figure, we can see that a few outcomes in which either the free-floating planet stays in the system or the native planet stays in the system agree with observations. The authors conclude that an encounter with a free-floating planet is a possible explanation for the current architecture of this system. The close encounter only reproduces observational results 0.2% of the time, but this system is quite unusual — so a low probability of a system forming like this is expected!

final simulation results of the encounter between the free-floating planet and the binary star system

Figure 4: The final results of the close encounter simulations of the free-floating planet and the HD 106906 system. The dots within the dashed square fall within the expected observational constraints of the companion (i.e., that agree with current estimates for the orbital eccentricity, semi-major axis, and inclination of the companion). The blue dots represent the outcomes that agree with observations where the native planet remains bound to the system. The red dots represent the outcomes where the free-floating planet remains bound to the system. The gray dots represent the other parameters of the 100,000 simulations. [Moore et al. 2022]

Original astrobite edited by H Perry Hatchfield.

About the author, Clarissa Do O:

I am a third-year physics graduate student at UC San Diego. I study exoplanet orbital dynamics and also work on exoplanet instrumentation. My current work is on the adaptive optics upgrade of the Gemini Planet Imager 2.0, an instrument that aims to directly image and characterize exoplanets.