How Do Stars Get Kicked Out of Globular Clusters?

Globular clusters are densely-packed collections of stars bound together gravitationally in roughly-shaped spheres. They contain hundreds of thousands of stars. Some might contain millions of stars.

Sometimes globular clusters (GCs) kick stars out of their gravitational group. How does that work?

There are a few things that can cause GCs to eject stars. Gravitational scattering, supernovae, tidal disruption events, and physical collisions could all be responsible. Whatever’s behind it, the gradual ejection of stars from GCs is an established phenomenon.

The evidence for stellar ejection from GCs is in the tidal tails that stream out from them.

Palomar 5 is a globular cluster being torn apart by the Milky Way. Palomar 5 is the white blob at the center, and the orange is streams of stars. The yellow line with arrows represents the cluster's orbit around the Milky Way. Image Credit: Odenkirchen, Grebel, et al. 2002/Sloan Digital Sky Survey
Palomar 5 is a globular cluster being torn apart by the Milky Way. Palomar 5 is the white blob at the center, and the orange is streams of stars. The yellow line with arrows represents the cluster’s orbit around the Milky Way. Image Credit: Odenkirchen, Grebel, et al. 2002/Sloan Digital Sky Survey

A new study based on data from the ESA’s Gaia mission aims to understand how GCs eject stars. Its title is “Stellar Escape from Globular Clusters I: Escape Mechanisms and Properties at Ejection.” It’s been submitted to the Astrophysical Journal, and the lead author is Newlin Weatherford, an astronomy Ph.D. student at Northwestern University in Illinois.

“Recent exquisite kinematic data from the Gaia space telescope has revealed numerous stellar streams in the Milky Way (MW) and traced the origin of many to specific MWGCs, highlighting the need for further examination of stellar escape from these clusters,” the authors write. This study is the first of a series, and the authors examine all the escape mechanisms and how each one contributes to GC star loss.

GCs are some of the oldest stellar associations in the Milky Way. Individual GC stars are also older and have lower metallicity than the Milky Way’s general population. Nearly all galaxies host GCs, and in spiral galaxies like ours, the GCs are mostly found in the halo. The Milky way hosts more than 150 of them. Astronomers used to think that stars in a GC form from the same molecular cloud, but now they know that that’s not true. GCs contain stars of different ages and metallicities.

GCs are different from their cousins, the open clusters (OCs). OCs are most often found in the disks of spiral galaxies, have more heavy elements, and are less dense and also smaller than GCs. OCs have only a few thousand stars, and there are more than 1100 of them in the Milky Way.

NGC 6441 is one of the most luminous and massive globular clusters in the Milky Way. Image Credit: ESA/Hubble & NASA, G. Piotto
NGC 6441 is one of the most luminous and massive globular clusters in the Milky Way. Image Credit: ESA/Hubble & NASA, G. Piotto

GCs are unique, and astronomers consider them tracers of galactic evolution. Thanks largely to the ESA’s Gaia spacecraft, we know more about GCs. Gaia helped reveal the presence of numerous stellar streams coming from the Milky Way’s globular clusters. As the authors explain in their paper, “These drawn-out associations of stars on similar orbits are likely debris from disrupted dwarf galaxies and their GCs, shorn off by Galactic tides during accretion by the MW (Milky Way.)”

Gaia did more than spot these streams. It was able to connect some streams to specific GCs. “Gaia’s exquisite kinematic data has firmly tied the origins of ~10 especially thin streams to specific MWGCs,” the authors write. The Palomar 5 GC and its streams are well-known examples. The streams are excellent tracers of the Milky Way’s evolution. (Palomar 5 gained even more notoriety in astronomy recently when a 2021 paper found more than 100 black holes in its center.)

Observations of these types of tails, both from stars ejected from GCs, and from interacting and merging galaxies, are an extremely active area of research. There are many astounding images of these interactions. But as the authors point out, “… the theoretical study of stellar escape from GCs has a longer history.” Astronomers have come up with different mechanisms for these escapes, and this paper starts with a review of each one.

Artist's impression of the thin stream of stars torn from the Phoenix globular cluster, wrapping around our Milky Way (left). Red giant stars make up a significant portion of the stream and helped astronomers map it. Credit: James Josephides (Swinburne Astronomy Productions) and the S5 Collaboration.
Artist’s impression of the thin stream of stars torn from the Phoenix globular cluster, wrapping around our Milky Way (left). Red giant stars make up a significant portion of the stream and helped astronomers map it. Credit: James Josephides (Swinburne Astronomy Productions) and the S5 Collaboration.

The authors divide escape mechanisms into two categories: Evaporation and Ejection. Evaporation is gradual, while ejection is more abrupt. The following are brief descriptions of each of the ejection methods, beginning with the Evaporation category.

Two-Body Relaxation: the motions of each body induce granular perturbations that create exchanges in energy and momentum in the bodies. Over time, stars can be ejected from GCs.

Cluster mass loss: stars lose mass over time, and that can affect the gravitational binding that holds stars in the cluster.

Sharply time-dependent tides: MWGCs orbit the Milky Way in eccentric and inclined orbits. The galactic tide will be stronger at some points in the orbit. The changing gravity can allow stars to exit the GCs.

The second broad category is Ejection. These are events typically involving single stars that are ejected rapidly and dramatically.

Strong Encounters: a close passage between two or more bodies that provides a strong enough kick to eject a star.

(Near)-Contact Recoil: encounters so close that tides, internal stellar processes, and/or relativistic effects are relevant. This includes collisions and gravitational waves.

Stellar Evolution Recoil: This includes the powerful forces unleashed when a star goes supernova, for example, or when a black hole or neutron star is formed.

Since there was no way to go and observe a statistically significant number of GC ejections, the team of researchers took what data was available and performed simulations. They used what’s called the CMC Cluster Catalog.

The study is concerned with the two types of GCs: non-core collapsed and core-collapsed. They’re different from each other and are a fundamental property of GCs, so the team simulated both types.

Core collapse in GCs occurs when the more massive stars in a GC encounter less massive stars. This creates a dynamic process that, over time, drives some stars out of the center of the GC towards the outside. This creates a net loss of kinetic energy in the core, so the remaining stars in the GCs core take up less space, creating a collapsed core.

This figure from the study shows the number of escaped single stars and stellar objects for the archetypal core-collapse GCs and non-core-collapse GCs. The x-axis is unlabelled but measures time in Gyrs. Each black marker is two Gyrs. Dashed lines are results from core-collapsed GCs, while solid lines are non-core-collapsed GCs. The plotted lines are colour coded according to the legend at the top. As the figure shows, most ejected stars are main-sequence stars, mirroring the population of the GCs themselves. Image Credit: Weatherford et al. 2022.
This figure from the study shows the number of escaped single stars and stellar objects for the archetypal core-collapse GCs and non-core-collapse GCs. The x-axis is unlabelled but measures time in Gyrs. Each black marker is two Gyrs. Dashed lines are results from core-collapsed GCs, while solid lines are non-core-collapsed GCs. The plotted lines are colour coded according to the legend at the top. As the figure shows, most ejected stars are main-sequence stars, mirroring the population of the GCs themselves. Image Credit: Weatherford et al. 2022.

An important astronomical principle plays a role in the team’s results. Two-body relaxation is a fundamental aspect of stellar associations that has far-reaching effects. It’s a complicated topic, but it basically describes the ways that stars in stellar associations, such as GCs, interact gravitationally and share kinetic energy with each other. It shows that star-to-star interactions drive GCs to evolve during the lifetime of the galaxy they’re attached to.

Not surprisingly, the researchers found that two-body relaxation plays a powerful role. That conclusion lines up with the established theory. “Consistent with longstanding theory and numerical modelling, we find that two-body relaxation in the cluster core dominates the overall escape rate,” they write.

This figure from the study shows binary objects ejected in the simulation. The number of objects is on the y-axis, and time in two-Gyr increments is on the x-axis. Dashed lines are results from core-collapsed GCs, while solid lines are non-core-collapsed GCs. Image Credit: Weatherford et al. 2022.
This figure from the study shows binary compact objects ejected in the simulation. The number of objects is on the y-axis, and time in two-Gyr increments is on the x-axis. Dashed lines are results from core-collapsed GCs, while solid lines are non-core-collapsed GCs. Image Credit: Weatherford et al. 2022.

They also found that “… central strong encounters involving binaries contribute especially high-speed
ejections, as do supernovae and gravitational wave-driven mergers.” This also lines up with other research.

This figure from the study shows binary objects containing a compact object and a main-sequence or giant star ejected from GCs. The number of objects is on the y-axis, and time in two-Gyr increments is on the x-axis. Dashed lines are results from core-collapsed GCs, while solid lines are non-core-collapsed GCs. Image Credit: Weatherford et al. 2022.
This figure from the study shows binary objects containing a compact object and a main-sequence or giant star ejected from GCs. The number of objects is on the y-axis, and time in two-Gyr increments is on the x-axis. Dashed lines are results from core-collapsed GCs, while solid lines are non-core-collapsed GCs. Image Credit: Weatherford et al. 2022.

But one of their results is new. It concerns three-body binary formation (3BBF.) 3BBF is when three bodies collide to form a new binary object. “We have also shown for the first time that three-body binary formation plays a significant role in the escape dynamics of non-core-collapsed GCs typical of those in the MW. BHs are an essential catalyst for this process,” they write. “3BBF dominates the rate of present-day high-speed ejections over any other mechanism,” they explain, as long as significant numbers of BHs remain in the GCs core. 3BBFs also produce a significant number of hypervelocity stars.

In their conclusion, the authors explain that “… this study provides a broad sense of the escape mechanisms and demographics of escapers from GCs,” while also noting that the results are “not immediately comparable to Gaia observations.” That’s why this work is the first in a series of papers. In their follow-up paper, they intend to integrate the trajectories of escaped stars and construct their velocity distributions to reproduce tidal tails. After that work, they hope that they’ll have a clearer understanding of how stars escaping from GC contribute to galactic evolution.

In a third paper, they intend to “… identify likely past members (‘extratidal candidates’) of specific MWGCs and directly compare the mock ejecta from our cluster models to the Gaia data.” This will get closer to some of the core questions surrounding GCs and the Milky Way’s evolution: how do stellar streams form? How many BHs are there in GCs? What role do supernovae play?

“Ultimately, we hope to better understand stellar stream formation and, in an ideal case, leverage the new
observables from Gaia to better constrain uncertain properties about MWGCs, such as BH content, SNe kicks, and the initial mass function, which affect ejection velocities and the cluster evaporation rate.”

The ESA's Gaia spacecraft doesn't make a lot of headlines in regular media because it doesn't take gorgeous images. But as this study shows, its contribution to important topics like galactic evolution can't be overstated. Artist's impression of the ESA's Gaia Observatory. Credit: ESA
The ESA’s Gaia spacecraft doesn’t make a lot of headlines in regular media because it doesn’t take gorgeous images. But as this study shows, its contribution to important topics like galactic evolution can’t be overstated. (Those who know, know.) Artist’s impression of the ESA’s Gaia Observatory. Credit: ESA

This study is an interesting look at how a number of natural phenomena all contribute to galactic evolution. The evolution of individual stars, how individual stars interact gravitationally and how they form binary objects, the tidal interactions between globular clusters and their host galaxies, two-body relaxation, and even three-body binary formation. Throw in supernovae and hypervelocity stars.

Each one of these topics can form the basis of an entire career in astrophysics. It’s easy to see why follow-up studies are needed. Once they’re completed, we’ll have a much better picture of how galaxies, specifically our own Milky Way, evolve.

More: