Authors: Andrew J. Winter, Cathie J. Clarke, Giovanni P. Rosotti
First Author’s Institution: University of Cambridge, UK
Status: Published in MNRAS (open access on arXiv)
Planets are born from the dense, dusty regions surrounding young stars called protoplanetary disks. Recently, the ALMA telescope has provided some of the most detailed images and information about protoplanetary disks to date. Understanding the evolution of these objects is crucial to constrain how and where planets form. It is thought that planets can grow from tiny dust grains that settle toward the midplane of the disk, or even quickly materialize through gravitational instabilities.
These disks don’t stay around forever. Their viscous, fluid nature means that disk material continually migrates inward, eventually falling onto the central star. At the same time, the radiation from the central star heats and erodes the disk from the inside out until eventually nothing is left except for a few planets. Determining how long the disks stay around tells us how long planets take (at most) to form.
Stars, which host the planet forming disks, tend to form in groups and so this evaporation process can also be driven by other nearby companions. Clusters and associations usually host at least a few extremely massive stars which output colossal amounts of ultraviolet (UV) radiation. The radiation can carve huge bubbles and cavities into the natal material, in addition to deteriorating nearby protoplanetary disks. This process is called photoevaporation and is caused by the acceleration of material due to the intense heat from energetic photons.
Astronomers have recently begun asking how these high radiation environments affect the evolution of protoplanetary disks. The authors of today’s paper flip this question around and instead ask what the distribution of protoplanetary disks on the sky can tell us about the environment from which they formed. Specifically, they focus on the evolution of the stellar association Cygnus OB2, home to some of the most massive stars ever found, and try to understand how the highly UV irradiated environment affects the population of disks.
A recipe for evaporation
The authors begin by constructing a model of the evolutionary history of Cygnus OB2. At the highest level, this involves an N-body simulation to track the motions of individual stars in the association. Over time, the UV radiation from the hottest and brightest stars expels any leftover gas, dramatically reducing the depth of the gravitational well and allowing some of the faster moving stars to escape. At the same time the model tracks the evolution of the individual protoplanetary disks around each star. A number of different parameters are tuned to reproduce observations of Cygnus OB2. These include the initial stellar mass, the initial velocity and positions of those stars and a number of properties which govern the expulsion of the interstellar gas.
Figure 1: Results from the model that best reproduces the observed properties of Cygnus OB2. The left plot shows the evolution of the surviving fraction of disks as a function of projected UV flux (set by the proximity of the disk to the massive stars on the sky) near the end of the simulation. The crosses indicate observed constraints. The right plot shows the x and y positions of the simulated disks (dots) relative to the most massive stars (star symbols). The disk markers are color coded by the FUV flux bins shown in the left hand plot. Open circles denote disks which have been completely evaporated.
These parameters are subject to a handful of observational constraints. The first of these is the surviving disk fraction as a function of projected UV flux. The latter quantity is set by both the proximity of a disk to a star on the sky, along with the luminosity of the star. Figure 1 shows a comparison between the modeled and observed disk fraction (left) in addition to a sky map showing the proximity of the disks to the most massive stars (right). The simulated disk fraction matches well to the observations for high UV fluxes, but doesn’t do so well for less irradiated disks. This is not terribly surprising because the model used in this study does not account for the effects of photoevaporation driven by the central star. This effect is most important in regions where the external UV driven evaporation is weak. This explains why the model overpredicts the disk fraction at low external UV fluxes.
Finally, the authors use the present day stellar mass of Cygnus OB2 to constrain the initial stellar mass that formed. To match the observed stellar mass, they find that the association must have lost about 75% of its initial stellar mass, or about 60,000 Suns worth of stellar material! In order to match all of these constraints (along with the correct disk fraction), Cygnus OB2 must have formed about 80,000 Suns worth of stellar material (distributed in clumps) and the interstellar gas must have been gradually expelled between 2 million and 500,000 years ago.
Diagnosing external photoevaporation
The authors then discuss the possibility that protoplanetary disks in a high UV radiation environment may look fundamentally different than those found elsewhere. To determine whether a population of disks have been subjected to this type of environment, the authors look at the relationship between the disk mass and the host mass of the star. There is a clear trend between these two quantities, shown in figure 2. At first glance, this trend is not entirely surprising. Massive stars tend to hold on to more disk material because they have a stronger gravitational pull. This pattern has already been observed and well-documented in other environments. It turns out, however, that if one were to reproduce figure 2 using a sample of disks from a less irradiated environment, the relation would be much shallower. The authors claim that the steeper relation found here comes from the fact that the strength of external photoevaporation depends on the mass of the host star. Around a low mass star, the gravitational potential well is much shallower and material is more easily stripped away by energetic photons.
Figure 2: The remaining disk mass from the final simulation as a function of stellar host mass. Points are color coded by the original ‘clump’ from which the star came in the initial conditions.
As this study has demonstrated, it is entirely possible to constrain the history in a stellar association by combining a catalog of protoplanetary disk detections with observed positions and velocities of the stars. Most importantly, the correlation between disk mass and host star mass looks to be a promising diagnostic indicating how much UV radiation there was in the past. A major caveat that the authors highlight is that there is a degeneracy between the gas expulsion history in the cluster and the viscosity of the protoplanetary disks. This is because a very viscous disk will quickly dump material onto the host star, depleting the amount of available material just like photoevaporation would. Regardless, the results of this study show that the high radiation environment of a stellar association appears to leave a lasting imprint on the population of protoplanetary disks. In terms of planet formation, the much steeper relation between host star mass and disk mass implies that the material available for planet formation will be significantly depleted.