The contribution of white dwarf kicks to the population of free-floating planets
We explored an alternative scenario for the formation of free-floating planets (FFPs) in old planetary systems, linked to the recoil impulse when a white dwarf (WD) forms from an asymptotic giant branch (AGB) star. Observations confirm that WDs get a gentle kick during shell ejection. We show that while only ~1% of known planets have their gravitational bonds directly broken, in more than 40% of long-period multi-planet systems this kick triggers dynamical instability, leading to planet ejection in about half the cases. Additionally, planets leaving the system experience strong and prolonged heating from the increased luminosity of the AGB star. Due to low ejection velocities (a result of the kick's weakness), these heated FFPs linger near their parent stars for several million years, which enables their association. Thus, the contribution of WD kicks forms a distinct, observationally identifiable subpopulation of FFPs, making up a few percent of all Galactic FFPs, which is relevant for the upcoming Roman Galactic Exoplanet Survey.
Context
Free-floating planets — exoplanets without parent stars — are a numerous but still not fully understood class of objects. Most proposed formation mechanisms, such as direct collapse or ejection in young systems, operate at early evolutionary stages. However, little attention has been paid to the role of late stellar evolution. Observations of gravitational microlensing have already identified many FFP candidates, and the upcoming Galactic survey by Roman will significantly increase their numbers. Modern detection methods for exoplanets, like the transit method (e.g., the Kepler mission led by William Borucki and TESS), are strongly biased toward short-period orbits, whereas FFPs are mostly discovered through microlensing.
Methods
To assess the impact of white dwarf kicks, the empirical kick velocity distribution derived for exoplanetary systems from Gaia observations was used (peak at about 0.75 km/s), and the kick was assumed to occur instantaneously at the end of the asymptotic giant branch (AGB) phase. Based on an analytical derivation of the fraction of directly disrupted orbits and the Petrovich (2015) stability criterion, the probability of instability in known multi-planet systems was estimated. Calculations were supplemented by n-body numerical simulations with the REBOUND code for 53 systems from the NASA exoplanet archive, whose host stars are massive enough to become white dwarfs within the age of the universe. The modeling was limited to 5 million years after the kick.
Results
Direct release of planets by kicks affects only about 1.3% of the known exoplanet population, which is expected due to observational bias toward close-in orbits. However, in 41% of multi-planet systems, according to the stability criterion, the kick leads to orbit crossing or violation of stability conditions; numerical simulations show ejections in 47% of cases within 5 million years. As an example, in the outer Solar System, the Uranus–Neptune pair exhibits a 50% probability of instability after a similar kick. Such events produce FFPs with characteristic velocities of about 0.75 km/s, which remain within a projected distance of no more than 5 pc from the former star for ~8 million years in half of the cases. Moreover, during the AGB phase, planets were subjected to intense heating: for a Jupiter-like planet, the temperature could reach 700 K. After ejection, such a giant cools according to a power law, staying above 400 K for many millions of years. The total contribution of the white dwarf kick channel to the overall FFP population of the Galaxy is estimated at a few percent, which is significant for the statistics of future surveys.
Implications
The existence of warm, slowly receding FFPs in the vicinity of young white dwarfs opens a new observational window. These objects can be distinguished from other FFPs by their infrared excess and spatial correlation with white dwarfs. This changes approaches to searching for and classifying exoplanets, and underscores the importance of accounting for late stellar evolution in models of planetary system dynamics. Data from future telescopes like James Webb and Hubble (via archival observations) could be used to test these predictions.
Future development
The upcoming survey of the Galaxy by Roman using microlensing will greatly expand the sample of FFPs, and identifying the subpopulation from white dwarf kicks will become possible when combined with infrared observations, for instance, with spectroscopic instruments on future telescopes. Development of planetary cooling models and accounting for the diversity of their internal structures will allow more accurate estimates of the visibility timescales for such warm FFPs. As the Galaxy ages, the fraction of this channel will increase, affecting the mass function of FFPs.
Impact
The results will impact observational exoplanetology, stellar astrophysics (especially the dynamics of open clusters and stellar streams), and the planning of search strategies for microlensing-oriented missions.
Next steps
Full-scale n-body simulations with longer evolutions and realistic orbital distributions are needed, along with a targeted infrared survey of a few dozen young white dwarfs to search for excess thermal signatures at distances up to 10 pc.
Key open problems
This work connects the unresolved questions of free-floating planet origins with the physics of mass loss and kicks during white dwarf formation. The exact kick mechanism, its dependence on initial stellar mass, and the influence of mutual inclinations on the long-term stability of multi-planet systems remain open.
🎯 If the Sun turned into a white dwarf right now, Jupiter would heat up to 700 K — hotter than the surface of Venus — and would stay warm longer than human civilization has existed.
Key numbers
- Kick velocity: 0.75 km/s
- Fraction of unstable multi-planet systems: 41–47%
- Time within projected 5 pc of host (half the sample): 8 million years
- Temperature of Jupiter during the Sun's AGB phase: 700 K
- Estimated contribution to the overall FFP population: a few percent