From the March 2019 issue

How high-speed stars escape the galaxy

Astronomers have found dozens of hypervelocity stars on a one-way trip out of the Milky Way and into intergalactic space.
By | Published: March 14, 2019 | Last updated on May 18, 2023
If a binary star system ventures too close to the supermassive black hole at the center of the Milky Way, the black hole’s gravity could capture one of the stars and toss the other into the galaxy’s halo at speeds of 1 million mph or more. Astronomers think this scenario explains most of the Milky Way’s two dozen or so hypervelocity stars.
Astronomy: Roen Kelly
Stars bob and weave in and out of the Milky Way Galaxy’s spiral arms like cars speeding through rush-hour traffic. But a snapshot of the nighttime sky makes it appear that these luminaries are as fixed as the great pyramids of Egypt. Of the estimated 200 billion to 400 billion stars that call our galaxy home, however, a tiny fraction of hot, massive ones stand out. Gravitational interactions have revved them up to speeds double or even triple that of the Sun. These so-called hyper­velocity stars race through the Milky Way so quickly that they are destined to break free of the galaxy’s gravitational embrace.

Our local solar neighborhood is in constant motion, participating in a mostly orderly flow shared by the vast majority of the stars revolving around the Milky Way’s center. But a small number of fast-moving suns break this overall pattern. Astronomers often find these “runaway stars” fleeing youthful clusters.

Hypervelocity stars take these speeds to a new level. Over the past decade, scientists have discovered a couple of dozen of these speed demons. Nearly all are B-type luminaries with masses between two and five times that of the Sun and surface temperatures above 18,000° F (10,000 kelvins).

Most lie in the Milky Way’s outer halo at least 150,000 light-years from the galactic center and move at velocities of more than 700,000 mph (1.1 million km/h). Such a star could zip from Earth to the Moon in barely 20 minutes and could traverse 1,000 light-years in a million years.

And all of them, by definition, have velocities high enough to eventually escape the gravitational clutches of our galaxy. The advent of newground- and space-based technology should soon uncover more hypervelocity stars and give us a closer look at those already known.

The first glimpse

Theorist Jack Hills of the Los Alamos National Laboratory in New Mexico first predicted hypervelocity stars in 1988, but astronomers didn’t find one until 2005. Warren Brown of the Harvard-Smithsonian Center for Astrophysics (CfA) and colleagues stumbled upon SDSS J090745.0+024507 while targeting faint blue stars in the galaxy’s halo. This particular star lies some 350,000 light-years from the center of the Milky Way and has a radial velocity (its speed directly away from the galaxy’s center) of 1.51 million mph (2.42 million km/h). It made its way from the Milky Way’s core to the outer galactic halo in only 140 million years.

Theorists think that Sagittarius A* (pronounced “A-star”) — the supermassive black hole at the Milky Way’s center — accelerates most of these hypervelocity stars, but astronomers also are interested in whether any high-velocity stellar interlopers exist. Such stars could be making their way into the Milky Way in the same way that others are leaving. And other researchers wonder if some hypervelocity stars could be ejected from dense stellar clusters or by supernova explosions.

Gravity accelerates these stars to their phenomenal speeds. The basic explanation is a “three-body exchange” between a binary pair of stars and a black hole, says Brown. The black hole captures one of the stars into a tight orbit and slings the other out of the galaxy.

“This gravitational slingshot could potentially eject stars at a speed that approaches the speed of light,” says Avi Loeb, a theoretical astrophysicist at Harvard University. “These stars traverse cosmological distances and are the ultimate hypervelocity stars. They [are] not yet observed, but we made predictions about their abundance as a function of speed.”

CfA astronomer Scott Kenyon says it is unclear how many hypervelocity stars exist in our galaxy. He estimates there are probably several hundred of them with three to five times the Sun’s mass. “We use spectroscopic techniques to estimate a distance and thus a position in the Milky Way,” he says. “We then compare the radial velocity to the velocity needed to escape the galaxy.” And all of these stars are on track to exit the Milky Way within a few hundred million to a billion years.

Hypervelocity star HE 0437–5439 could have a complex origin story. Some astronomers think it started as a triple star made up of a tight binary system and an outlying singleton (top left). When the trio approached the black hole in the Milky Way’s core, the hole captured the outer member and threw out the binary at hypervelocity (top right). The ejected pair evolved until one component became a red giant and the two spiraled closer (bottom left). Eventually, the duo merged into a single “blue straggler” (bottom right).
Astronomy: Roen Kelly

To B or not to B

So far, almost all known hypervelocity stars are B-type suns on the main sequence, the period in their lives when they produce energy by fusing hydrogen into helium in their core. Because such stars live no more than several hundred million years, you wouldn’t expect to find any on the galaxy’s fringes.

“These B stars should not exist there,” says Brown. “There’s no star formation in the outer Milky Way halo. It’s a dead region — the galaxy’s halo contains the galaxy’s globular clusters and old, metal-poor, low-mass stars. Unless it was ejected there, you would never expect to see a B star traveling at those speeds in the outer halo.”

But there they are. The best explanation for their existence involves a binary star that ventures too close to a massive black hole, says Hagai Perets, an astrophysicist at the Technion – Israel Institute of Technology in Haifa. The black hole captures one star into a highly eccentric orbit and ejects the other as a hypervelocity star.

A decade after Brown and his colleagues discovered the first such star, however, there’s still a dearth of data on bona fide hypervelocity stars across the whole sky. Astronomers can measure any given star’s radial velocity by examining its spectrum — light spread out into its constituent wavelengths. If an object is moving toward us, its spectral lines shift to shorter wavelengths; if it’s moving away, the lines swing to longer wavelengths. The higher the velocity, the greater the shift.

Although this sort of spectral analysis is straightforward for nearby stars, it becomes far more difficult for distant suns in the Milky Way’s outer halo. Even large telescopes can’t gather enough of their light. That’s why astronomer Ulrich Heber of the University of Erlangen-Nuremberg in Germany thinks there are probably several low-mass hypervelocity stars just waiting to be discovered. Although these diminutive objects live longer than their B-type cousins, they radiate much less light and so can’t be seen out to as great a distance. Still, they would be easier to detect than the even fainter white dwarf remnants of any dead B-type star.

Los Alamos theorist Jack Hills predicted hypervelocity stars in 1988, but astronomers didn’t find one until 2005.

On the move

Once astronomers know a star’s radial velocity as viewed from Earth, they can calculate how fast it’s moving relative to the galaxy’s center. But even this tells only half the story. To directly link a wayward star in the galaxy’s outer fringes to its theoretical point of origin at the supermassive black hole in the Milky Way’s core, observers also must determine the star’s motion across our line of sight. This so-called proper motion is even harder to measure precisely than radial velocity.

Astronomers determine proper motion by observing the shift in an object’s position relative to more distant objects. For a hypervelocity star, this means measuring its movement in relation to background galaxies or quasars, a process that takes years.

Despite their breakneck speeds, hypervelocity stars have proper motions of less than 1 milliarcsecond per year. (One milliarcsecond equals 0.000000005°, or the angular size of a dime seen from about 2,300 miles [3,700 kilometers] away.)

Ground-based surveys are accurate to only about 5 milliarcseconds per year, so proper-motion studies for hypervelocity stars must be done from space. That’s where the European Space Agency’s (ESA) Gaia mission comes in. This astrometric observatory — designed to measure precise positions and radial velocities of some 1 billion stars — is yielding proper motions accurate to within 0.1 milliarcsecond per year. In the next year or two, Gaia should provide superb proper motions for known hypervelocity stars and new candidates.

These observations will, in theory, help astronomers determine much more about these stars’ points of origin. Although researchers think most originated in interactions with Sagittarius A*, they still debate whether some might be interlopers from outside the galaxy. Perhaps they made their way into the Milky Way’s outer halo in a stream of stars from a tidally disrupted dwarf galaxy. Or maybe the Milky Way’s satellite galaxy, the Large Magellanic Cloud (LMC), ejected some into our galaxy’s halo.

“There’s an unbound B star, HE 0437–5439, that’s near the Large Magellanic Cloud that could either be from the LMC or the Milky Way,” says Brown. “HE 0437–5439 is moving away from us, and we don’t know if it’s angled our way or to the LMC.” If this star originated in the LMC, it might be the smoking gun for a previously undetected intermediate-mass black hole that ejected the star at hypervelocity.

Hypervelocity stars don’t tend to be photogenic because they glow dimly due to their distances. The Hubble Space Telescope captured HE 0437–5439 (center), which lies in the constellation Doradus near the Large Magellanic Cloud.
NASA/ESA/O. Gnedin (University of Michigan)/W. Brown (Harvard-Smithsonian CfA)
Although the source remains elusive, Brown argues that the star is a remnant of a binary system ejected from the Milky Way. His scenario begins with a trio of stellar companions: a tightly bound binary pair in orbit with a more-distant sun. The three got caught up in the destructive gravity of Sagittarius A* and paid the price. The supermassive black hole captured the system’s outlying member and ejected the other two, which remained an intact pair.

As the pair hurtled away from the galactic center, the more massive star eventually evolved into a red giant. As it swelled, the two stars spiraled together and merged into an even bigger “blue straggler.” This delayed formation process for HE 0437–5439 is the best way to get a B-type main sequence star to its current position some 200,000 light-years from the Milky Way’s center. Otherwise, this particular star would have evolved off the main sequence long ago.

“Only Sagittarius A* can explain the fastest B-type hypervelocity stars,” says Brown. “Other processes eject different types of stars at different speeds.”

The oddball stars

Heber, for one, studies some of these special cases. And he has concluded that nature has found a way to generate hypervelocity stars that doesn’t rely solely on interactions with the galaxy’s supermassive black hole. He suggests alternatives that include satellite galaxies disrupted by the Milky Way’s tidal forces, binary supernovae, and ejections from star clusters.

Heber preferentially targets relatively low-mass stars that have evolved into bloated red giants. Such stars burn helium in their cores rather than hydrogen. “We find that most of our candidates are unlikely to have been launched from the galactic center,” says Heber. “We are eagerly awaiting the Gaia astrometric measurements, which will allow us to trace their trajectories back to their place of origin much more accurately, maybe to a stellar stream, a cluster, or a spiral arm.”

Right now, says Kenyon, astronomers have two good models for generating hypervelocity stars: disrupting a binary star system that passes too close to a black hole, and disrupting a binary during a supernova explosion. In the second scenario, two stars revolve around each other in a tightly bound orbit. When the higher-mass one reaches the end of its life and its core collapses, it triggers a supernova that can liberate its lower-mass companion. The exploded star’s collapsed remnant — either a neutron star or a black hole — and the previously bound main sequence star then go their separate ways. This mechanism will work wherever young stars hang out, including inside youthful star clusters.

The neutron star RX J0822–4300 is a prime example. In 2012, astronomers clocked it moving at 1.5 million mph (2.4 million km/h). The explosion that created Puppis A — a supernova remnant some 7,000 light-years from Earth in the southern constellation Puppis — launched the stellar remnant onto this trajectory. Astronomers think this supernova was a lopsided explosion, and the neutron star headed one way while much of the supernova debris went in the opposite direction.

Unfortunately, astronomers estimate that they would have to observe some 10,000 normal core-collapse supernovae to find one hypervelocity supernova. Scientists don’t think such explosions create all hypervelocity supernovae, however.

A case in point is US 708, the fastest-known hypervelocity star with a speed of 2.7 million mph (4.3 million km/h). This helium-rich star is spectral type O, and one of the hottest stars known in the Milky Way’s halo. Judging by its trajectory, it almost certainly did not originate in the galaxy’s center.

Astronomers think it was once part of an ultracompact binary system. Its companion was a massive white dwarf near the limit of what these stars can weigh. When US 708 evolved into a red giant, it transferred much of its hydrogen envelope onto the white dwarf, eventually triggering a statistically rarer type Ia supernova that sent US 708 onto a hypervelocity trajectory.

The Large Magellanic Cloud, the largest satellite galaxy of the Milky Way, could be the source of hypervelocity star HE 0437–5439. If so, a previously undetected intermediate-mass black hole likely ejected it from a young star cluster.

Pride of the Lion

You might think that hypervelocity stars would be spread across the sky randomly, but that’s not the case. One of the biggest puzzles surrounding current observations is that half of the B-type hypervelocity stars are clumped around the constellation Leo, Brown says. Heber thinks this could mean that the galactic center ejected them preferentially in a certain direction. He says this could happen if the ejected stars came from a stellar disk surrounding the supermassive black hole.

But the observations of clustering may just reflect a lack of data. “We do not have a complete survey of the entire sky,” says Kenyon, “so maybe we are seeing a statistical fluke.” Surveys of the southern sky should clarify this issue. In particular, they will allow astronomers to study north-south asymmetries and see whether similar numbers of hypervelocity stars reside in Aquarius, the constellation opposite Leo.

New surveys with the 1.35-meter SkyMapper robotic telescope in Australia, the European Southern Observatory’s 2.6-meter VLT Survey Telescope in Chile, and the 8.4-meter Large Synoptic Survey Telescope (LSST) currently under construction in Chile will complete the search for hypervelocity stars in the southern sky.

Once these surveys are finished, astronomers can start using hypervelocity stars to study other features of the Milky Way. “Because [B-type] hypervelocity stars originate from the galactic center, their trajectories should be a straight line outward,” says Brown. “However, theorists believe that the Milky Way is surrounded by a triaxial [football-shaped] distribution of dark matter.”

This means the present trajectories of the hypervelocity stars should deviate from a straight line as they feel the gravitational pull of this unseen matter, he says. How much the trajectories deviate, and in what direction they do so, depend on the shape and orientation of the dark matter halo.

If astronomers can find 200 or so hypervelocity stars distributed all across the sky, says Kenyon, then measuring their precise trajectories could tell us how they decelerate as they travel from the galactic center into the halo. Scientists then could use these deceleration measurements as a function of position on the sky to learn whether the shape of the dark matter halo is more spherical or if it is flatter at the galaxy’s poles.

In principle, Kenyon says, if hyper­velocity stars spread out nonuniformly on the sky, the degree of their asymmetry can tell us about asymmetries in the distribution of matter in the galactic center and inner bulge.

Trajectories of hypervelocity stars should deviate from a straight line depending on the shape and orientation of the dark matter halo.

Into the future

Still, current technology limits observations to only the brightest and thus most massive hypervelocity stars. Astronomers identify candidates based on their brightness and color and then they take spectra to see if they are moving at high velocities. To extend their reach to smaller and dimmer solar-type stars, scientists will need the LSST to select hypervelocity candidates based on color and then target these suns with the next generation of extremely large telescopes to get spectra. Finally, the Gaia mission will provide precise astrometric data to verify their place of origin and determine whether they hail from the galaxy’s center, stellar streams, or elsewhere.

These solar-type stars become visible once they evolve off the hydrogen-burning main sequence and become red giants. Such stars shine brightly enough to be visible throughout the Milky Way’s halo — and perhaps beyond. Future all-sky infrared surveys conducted from space, such as NASA’s Wide Field Infrared Survey Telescope and ESA’s Euclid spacecraft, should be able to detect these aged hypervelocity stars.

Hundreds of billions of years in the future, hypervelocity stars may be the only objects beyond the galaxy we’ll be able to observe. At that stage of cosmic evolution, all the galaxies in our Local Group will have merged into a single megagalaxy. And, assuming that the Hubble expansion of the universe continues to accelerate under the relentless push of dark energy, all the galaxies outside the Local Group will disappear beyond our cosmic horizon.

As Brown has written, “The only extragalactic sources of light in the observable cosmic volume will be hypervelocity stars ejected from our galaxy. Thus, hypervelocity stars may become the primary tool for measuring the Hubble expansion.” And our window on the cosmos at large will be reduced to those few hypervelocity stars that ply space-time wholly unimpeded by the gravitational bonds of their original host galaxy.