The Great Escape: planet evolution post-main sequence

Last month Dimitri Veras and collaborators wrote a nice, and very thorough paper on what happens to planets when their parent stars die...

"The Great Escape: How Exoplanets and Smaller Bodies Desert Dying Stars",
Dimitri Veras, Mark C. Wyatt, Alexander J. Mustill, Amy Bonsor, John J. Eldridge
(MNRAS, arXiv:1107.1239)

National Geographic News Story: How Planets Can Survive a Supernova

The paper revisits an old topic I am particularly fond of: what happens to the planets orbiting a star as the star leaves the main sequence, become a giant, and sheds its mass enroute to becoming a white dwarf.

There are several aspects to the problem: close planets get engulfed by the envelope of the giant; planets further out are in a death race between the expansion of the envelope and the onset of tidal coupling between the planet and the envelope on one hand, and the expansion of the planetary orbit on the other hand (cf Rybicki and Denis); and, for outer planets, well beyond the radii at which interactions with the envelope are important, there is the orbital response of the planet to the mass loss. Radiation effects from the increased luminosity of the star may also become important (cf Villaver and Livio).

Mass loss, obviously, changes the gravitational force the planet feels from the star.

The effect of the mass loss is generally considered in two limits:

if the mass loss is sudden, "impulsive", the planet generally goes onto an eccentric orbit, possibly hyperbolic for mass loss of ~50% or more, depending on the original eccentricity of the orbit and where the planet is along the orbit. This scenario was considered by Blaauw in 1961 and is often referred to as a "Blaauw kick".

if the mass loss is slow, "adiabatic", then the solution is analytic, as the integrals of motion are conserved; this scenario has been considered extensively cf Parriott and Alcock and Debes and Sigurdsson in different contexts.

Schematic of outer planet orbit evolution as a star leaves the main sequence and undergoes mass loss (from Debes 2003).

The intermediate case, where the time scale for the mass loss is neither rapid, compared to the planetary orbital time scale, nor slow, compared to the orbital time scale, is a pain in the ass and is much neglected.
Hadjidemetriou considered the general solution in a series of papers, and now Dimitri and his collaborators have revisited the problem to consider explicit scenarios for stars of different mass and planets of different masses in a variety of orbits.

The paper considers a number of cases of current interest, including: planets around horizontal branch stars; effect of planets on planetary nebulae morphology; planets around white dwarfs; and, the prospect for dynamical ejection or stripping of planets in very wide orbits around stars.


The paper recovers the limiting solutions, and explores in detail the variety of different outcomes for the intermediate case, some of which are surprising.
In particular, they find a series of solutions for moderately rapid mass loss, where the eccentricity of the planet grows secularly reaching e > 1.

ie the planets are ejected because the specific angular momentum grows, and not so much because of the decrease in binding energy.
That is interesting.

The principal focus of the paper is exploring the role of this effect in unbinding planet from their host star, especially in view of the recent observations which support the long held suspicion that there may be very large numbers of "free floaters" out there, cf Sumi et al.

A direct interpretation of the observation implies that there may well be more free floating planets than there are stars. If you think planets mostly form around stars, this would imply more than one planet ejected per star, on average, and the question then is how?
One method, which is effective if planets form "close packed" around typical stars, is for planet-planet dynamical interactions to eject some planets from their host stars; alternatively a substantial fraction of free floating planets may be ejected by the death throes of stars.
Both mechanisms must operate, which is dominant depends on details, such as the spacing and relative masses of planets when they are formed; how many planets are formed at large distances form their host stars; and, the number of planets formed around the more massive stars, compared to the less massive stars. The mechanism in Dimitri's paper is most effective for the more massive stars, and planets formed in wide orbits.
Scattering, of course, can also place planets in wide orbits, after formation further in, so the two mechanism are also complementary.

For really wild speculation, we can then consider whether there are circumstances in which free floating planets can sustain life, cf Stevenson 1999 and Debes and Sigurdsson 2007.


More like this

Final day of Extreme Solar Systems II here at Jackson Lake, and what a beautiful day it is.. And we are live for session 8 of day 5. Thursday night there was, again, an after dinner informal discussion, with the topic being "planets in binary systems" - the panel highlighted the theoretical…
The discovery of a jovian mass planet around a metal poor horizontal branch star in a kinematic stream of stars, possibly originating from a merger with a dwarf galaxy, is an interesting conundrum. With bonus updates and links. HIP 13044 - 10' field generated from SIMBAD using Aladin HIP 13044 is…
Two Earth Mass Planet found in close orbit around a red dwarf star Gliese 581 is an M3 red dwarf, it has a mass of about a third that of the Sun, slightly lower metal content and is about the same age probably a bit older than the Sun, we think. It is about 20 light years away. Gliese 581 was…
The Pale Red Dot project has found a planet. It is a terrestrial planet, orbiting in the formal habitable zone of Proxima Centauri, the nearest star to the Solar System. This wide-field image shows the Milky Way stretching across the southern sky. The beautiful Carina Nebula (NGC 3372) is seen at…

Planets can survive supernova... but it still doesn't look likely to me that you could make the PSR B1257+12 system that way...

And are PSR B1257+12-type systems extremely rare, or is no-one looking for them or what? Surely there can't be just one system of terrestrial-mass pulsar planets out there?

Indeed you cannot make PSR B1257+12 type systems this way, I think Currie and Hansen nailed down the essence of the formation mechanism for that system.

PSR B1257+12 systems are moderately rare - about 1% of millisecond pulsars seem to have low mass planets.
For regular pulsars the fraction is more like 0.1%.

People are indeed looking for them, quite hard...

Watch this space!

Planets ejected from dying stars will be fried, partially molten and without any water or other volatiles (unless they are former ice giants like Uranus).
Also, as the planets will be as old as their ex-suns, the decay of long-lived isotopes in their cores will have petered out, as most of those isotopes will have deacyed. Therefore any tectonic forces will have slowed down or stopped altogether.

This subset of "orphan" planets are therefore unlikely to be of much interest to exobiologists, or anyone else.

The orphan planets ejected by a system in its youth are more likely to be interesting.

By Birger Johansson (not verified) on 06 Aug 2011 #permalink

Birger, if the hot star has say twice the mass of the sun initially, the lifetime of the star should be shorter than the time for the planet to lose thermal driven geology. So it is possible that quite a number of planets that aren't geologically stripped might be free floating. Im also not so sure if the high luminosity phase of the parent start lasts long enough to melt more than the planets outer few kilometers. If you did that to our planet, plate tectonics would still bring some volatiles back to the surface later on.

By Omega Centauri (not verified) on 06 Aug 2011 #permalink

Very intriguing post and comments. Actually for me its the best post I've ever read from you. Never considered there could be more 'free floaters' than actual suns. Keep up the good work and I hope it eventually gets moved into the Old Favorites list.

Watch this space!

Soon? This soon? ;-)

I seem to recall some rumours of a terrestrial-mass planet orbiting a pulsar in a globular cluster a few years back, wonder what became of that...