New Scientist July 22, 2006 SECTION: FEATURES; Cover Story; Pg. 42-45 HEADLINE: Last exit to Andromeda Some of the Milky Way's stars are leaving. But before they do, they're going to tell us the galaxy's deepest secrets, says Maggie McKee A HUNDRED and sixty million years ago, a star was coursing casually through space when it came across a black hole. It could have been a disaster for the star, but it was lucky: it wasn't quite close enough to be swallowed. Instead it was flung away across space at an unimaginable speed. In December 2004, Warren Brown, an astronomer at the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts, saw this very star. It was travelling away from Earth at 850 kilometres a second - fast enough for it to escape the galaxy altogether. When Brown told his colleagues what he had in his telescope's sights, they thought he'd lost his mind. "You have a what?!" yelled Margaret Geller of the Smithsonian Astrophysical Observatory, also in Cambridge. Brown understood her reaction. Most stars in the region move no faster than 240 kilometres per second. They certainly don't have the kind of speed that can counteract the gravitational pull of all the matter in the galaxy. "It was just absurd," he says. That was the first time anyone had seen anything like this; less than two years later we know of nine of these "hypervelocity" stars racing out of the Milky Way - and the galaxy's loss may be the astronomer's gain. For a start, it's likely that these stars come from the centre of the galaxy, so studying them could provide important clues to what exactly is down there. It is difficult to observe the centre directly because the whole region is shrouded in dust, but the paths of hypervelocity stars already hint that our galaxy may have not one black hole at its centre, but two. And that's not the only dark secret waiting to be uncovered. Measuring where and how fast the stars are going could provide a means of tracing the shape of the dark matter halo that envelops our galaxy. We can use that to test our best theory of how galaxies form, as well as finally pinning down the nature of some of the universe's missing mass. That encounter with the black hole may have been bad news for the star, but it could be very good news for our understanding of the cosmos. The idea that a black hole might slingshot a star is not a new one. Physicist Jack Hills of Los Alamos National Laboratory in New Mexico was the first to realise this could happen; in 1988 he published a paper discussing how the colossal black hole that seemed to lurk at the centre of the Milky Way could separate pairs of stars that came within about 0.1 per cent of a light year of it. One star would be captured into a tight orbit and the other would be flung outwards at speeds of up to 4000 kilometres per second . Even though Hills's research was published in Nature (vol 331, p 687), it generated surprisingly little interest. Science fiction writer Paul McAuley apparently learned of the research, as a hypervelocity star was featured as a weapon of mass destruction in his 1993 novel Eternal Light - the star being launched at our solar system by a race of unfriendly aliens. There was no reaction to the idea of hypervelocity stars from Hills's fellow astronomers, however. Perhaps no one got excited because the idea of finding a hypervelocity star seemed no more plausible than the plot of McAuley's book. For example, theorists Qingjuan Yu of the Canadian Institute for Theoretical Astrophysics in Toronto and Scott Tremaine of Princeton University calculated in 2003 that the Milky Way's black hole should kick a star to extreme speeds just once every 100,000 years. That would mean there are only 1000 of these stars in the galaxy. The odds of actually locating one among the Milky Way's 100 billion stars seemed like the classic needle in a haystack problem, says Hills. "I approached several astronomers who could have done it, but nothing seemed to come of it." Sometimes, it seems, you just have to rely on luck. Brown wasn't looking for a hypervelocity star, and he had no knowledge of any attempts to find one; he and his colleagues were searching for small galaxies that had merged with the Milky Way. Because the galaxy's outer reaches are dominated by dark matter and old, red stars, they decided to focus their search on blue stars, which would stand out against the background. They picked the stars using data from the Sloan Digital Sky Survey, which is mapping a quarter of the sky. Then they took spectra of the stars with the 6.5-metre Multiple Mirror telescope in Arizona to measure how quickly the stars were moving towards or away from Earth. The one in Brown's sights was moving pretty darn quickly, it seemed. Brown knew that neutron stars, the collapsed remnants of a massive star, could be pushed to similar speeds by the supernova explosions that spawned them. But his star was just not old enough to be a corpse - it appeared to be a massive young star no more than a few hundred million years old. The researchers soon arrived at a very reasonable conclusion. "A massive black hole is the only way you can produce it," says Geller. It wasn't long before they came across Hills's theoretical prediction. And very soon the excitement of the discovery spurred other researchers to re-examine their archived data. Within months, two European groups announced that they had identified two more hypervelocity stars set to escape the galaxy. So far that figure has climbed to nine. The researchers expect another handful of discoveries this year, as Brown has so far only measured the velocities of half of the 500 blue stars identified in the Sloan data. Hills is thrilled by the roster of discoveries. And he's not the only one. Plenty of astronomers are over the moon at what these stars might reveal. By calculating the stars' distances from the galactic centre and their apparent speeds, for example, astronomers can estimate when each star was shot away from the centre. Each of the hypervelocity stars seems to have been ejected at a different time, ranging from 30 million to 160 million years ago. If these stars formed near the black hole, as many astronomers believe, this also reveals a time line of the types of stars born in the turbulent galactic centre. "You can learn about the history of star formation in a way that's not accessible with other techniques," says Scott Kenyon of the Smithsonian Astrophysical Observatory. The stars might also bolster the evidence that a second black hole up to 10,000 times as massive as the sun may lurk at the core of the galaxy (New Scientist , 21 June 2003, p 24). That's because hypervelocity stars require a three-body interaction to rocket up to their phenomenal speeds. This interaction could be between a pair of stars and a black hole, or a pair of black holes and a passing star. One of the stars discovered by a team led by Heiko Hirsch of the University of Erlangen-Nuremberg in Germany hints at the latter, more exciting possibility. The star, called US 708, is extremely rich in helium. Such stars are rarely found in pairs, suggesting the lone star might have been flung outward by a binary black hole. There are other surprises: one of the stars, called HE 0437-5439, may not even have originated in the Milky Way, and raises questions about the existence of a black hole in a galaxy far, far away . Meanwhile, the other seven are promising to shine light into the darkest areas of our own galaxy. One dark spot is the crowded centre of the Milky Way. There is so much dust there that astronomers have trouble seeing anything, but what they do see puzzles them. For example, the central black hole appears to have young, blue stars orbiting it just a fraction of a light year away. That's so puzzling because stars are thought to condense from cold clouds of molecular gas, and the intense gravity of the black hole - which has a mass 3.5 million times greater than the sun - should prevent condensation here. "The turbulence and the strong gradient of gravitational acceleration should tear the clouds apart," says Pierre Demarque of Yale University. In other words, he says, there simply shouldn't be any young stars anywhere near. Nevertheless, using spectral observations and measurements of how fast the stars are rotating, researchers have confirmed that at least two of the hypervelocity stars are relatively young. Only by finding more hypervelocity stars to study will astronomers be able to get closer to solving this mystery. Meanwhile, in an even murkier part of the galaxy, there's a more pressing conundrum. Every galaxy is thought to lie inside a halo of dark matter, the presence of which can only be detected through its gravitational pull on normal matter. Dark matter is thought to make up about 20 per cent of the universe, according to the latest observations of the cosmic microwave background radiation, yet we know almost nothing about it. Hypervelocity stars could help change that. According to our best ideas, dark matter is made of cold, slow-moving particles that began clumping together just after the big bang. These tiny clumps eventually grew into dark matter "haloes" that pulled stars together to form galaxies; the one around the Milky Way is about 1.5 million light years across. The cold dark matter (CDM) model says that these haloes should be elongated, like American footballs. That is because the haloes should have grown more along the direction in which they first began clumping, billions of years ago. This ovoid shape must exist if we are right about how structures such as galaxies form, how gravity acts on matter, and how matter flows. "If this is not so, our current understanding of how structure forms is just not correct," says Oleg Gnedin of The Ohio State University in Columbus. It's not just a problem for our understanding of structure formation. If the haloes turn out to be spherical, it could spell doom for all our CDM models. Although "warm" dark matter theories do exist, and do predict a more spherical shape to the haloes, there is too much evidence against warm dark matter for this to be the most likely explanation for any roundness in the halo, Gnedin says. In other words, find a spherical halo and it is back to the drawing board for dark matter research. So how can we tell what shape it is? "This is where hypervelocity stars are really indispensable," says Gnedin. The furthest hypervelocity star lies 360,000 light years away - in the outer reaches of the halo where dark matter outweighs normal matter by a factor of six. The others are scattered through the halo at distances closer to Earth. Because hypervelocity stars are thought to come from the galactic centre, astronomers already know their approximate paths and, since their motion is affected by the gravity of matter around them, pinning down their exact paths could reveal the distribution of matter - including dark matter - in the galaxy's halo. To do this, astronomers will need to have measurements of the stars' barely perceptible side-to-side movements. Gnedin believes these can be obtained with the Hubble Space Telescope; starting this month, he and his colleagues will use Hubble to image five of the stars. Assuming Hubble survives for a few more years, they will use it to image the stars again in 2008 to check how much the galactic speedsters have moved relative to background stars. By comparing the stars' observed motions with predictions, the researchers will be able to calculate their distances to an accuracy of 1 per cent. If their 3D motion suggests the stars really do come from the galactic centre, it would give a huge boost to the idea that the stars really were accelerated by a black hole, and the team would finally be able to measure the shape of the galactic halo. Of course, if the origin of hypervelocity stars turns out to be somewhere other than the centre of the Milky Way, then we will have no answers. "If they don't point back to the galactic centre, then we cannot put any constraints on the halo because we don't know their starting point," says Gnedin. Then we will have yet another mystery to solve. "It means there's some other phenomenon in the galaxy that can shoot stars out with such a huge velocity," he says. Let's hope it's not a race of xenophobic aliens with a grudge.