91av

WHAT did the universe look like shortly after the big bang? Cosmologists have already uncovered important clues about its history from the tepid afterglow of the big bang fireball. But not all the universe’s secrets are locked up in the cosmic microwave background. Some, it seems, are literally written in the stars. And one in particular may reveal the nature of the very first stars to form out of the big bang’s cooling debris. “The first stars blew up and vanished around 13 billion years ago, so the prospect of learning anything about them is staggering,” says Timothy Beers, an astronomer at Michigan State University in East Lansing. Ultimately, this single star could shed light on the shadiest moment in the universe’s history – when it transformed from a gas-filled void to a starry cosmos.

The star in question is called HE 0107-5240, but its lacklustre name conceals a wealth of astonishing features. Aged 13.5 billion years, it is the oldest star ever found, so ancient it formed before the galaxies and probably most of the elements around us today. Located 36,000 light years away in the halo of the Milky Way, HE 0107-5240 is in our own cosmic backyard. And its surface has inherited the simple elements churned out by the very first of its ilk. To astronomers, it is like finding a 5000-year-old Egyptian pharaoh living in New York. Better still, a pharaoh who remembers the Neanderthals 30,000 years earlier.

Just after the big bang 13.7 billion years ago, the universe consisted mostly of hydrogen and helium, along with a smidgen of lithium. The very first stars to congeal from this primordial gas would have looked nothing like the stars we see today (91av, 9 September 2000, p 25). Weighing hundreds of times more than our sun, they burnt their fuel at breakneck speed, which made them 10 million times as bright. And unlike modern stars, these monsters were born with no heavy elements at all.

Despite their bland make-up, nuclear reactions in the cores of these stars would have forged heavier elements, such as carbon, nitrogen and oxygen. When the giants exploded, these elements would have spewed into space, providing the raw materials for the next generation of stars, including HE 0107-5240. “The elements [that the monster stars] produced during their short lifetimes would be preserved in stars like HE 0107-5240,” says Beers. If he is right, we may finally be able to learn about the lost generation of stars that blew up and disappeared a mere 200 million years after the big bang.

Astronomers believe that somewhere in the Milky Way lurk ancient stars over 10 billion years old. The key to spotting these relics is the light they emit. Each element in a star absorbs or emits light at a characteristic wavelength, resulting in a unique spectrum. Ancient stars contain hardly any “metals” – a term used by astronomers to mean elements heavier than helium. So all you need to do to find a metal-poor star is look for its metal-poor starlight.

It sounds straightforward, but hunting for the real oldies among the 200 billion or so stars in the Milky Way is like looking for a needle in a haystack. To date, the most successful strategy has been to observe and sift through the spectra of millions of individual stars. And from the late 1970s to the early 1990s, Beers and his colleagues used this approach to build up the largest ever sample of highly metal-deficient stars. In the early 1990s, astronomers in Germany at the University of Hamburg and the European Southern Observatory near Munich extended the technique to look at fainter stars and larger areas of the sky.

Together, the two teams have measured the spectra of 4 million stars. To pick out promising metal-poor candidates, they selected those that had little or no trace of calcium, a marker for low iron abundance. These candidate stars were then studied in greater detail using larger telescopes. “It was then that one star stood out,” says Beers. This was HE 0107-5240, a yellow giant 36,000 light years away. With a surface temperature of 5100 kelvin, it is a little cooler than the sun but a few hundred times as bright.

In November 2001, Beers and his colleagues, Norbert Christlieb of the University of Hamburg and Mike Bessell of the Australian National University in Canberra used the Anglo-Australian Telescope at Siding Springs, New South Wales, to study HE 0107-5240. Its spectrum showed only the slightest hint of calcium, indicating an extremely low abundance of metals. But the resolution was not good enough to identify exactly which metals were present, so the team requested time on the Very Large Telescope on Cerro Paranal in Chile to make follow-up measurements. During a mammoth six-hour observation, they built up a high-resolution spectrum of the star. What they found stunned them.

HE 0107-5240 contains 200,000 times less iron than the sun and 200 times less than the previous record holder for the most iron-deficient star. What’s more, it has hardly any detectable metals – only nine compared with between 25 and 30 in other ancient stars that contain very little iron. “It is very encouraging to find such a low-metallicity star,” says John Cowan of the University of Oklahoma. “There is a strong possibility there are other such objects, which will allow us to study element building in the early universe.”

Heavy metal

According to Beers, the total lack of any metals heavier than nickel or iron tells us something important about the nuclear processes at work billions of years ago. Researchers believe two types of nuclear reaction forge heavy elements in stars. The first process occurs in the cores of mature stars where nuclei slowly absorb neutrons released in nuclear reactions. Eventually these nuclei become unstable and decay radioactively, converting neutrons into protons. In this way, ever-heavier nuclei are created, all the way up to lead and bismuth.

Even heavier elements such as uranium are created when a massive star comes to the end of its life and dies in a supernova explosion. Vast numbers of neutrons are released in the burning fireball and are rapidly absorbed by nuclei to produce a wealth of metals. Because HE 0107-5240 contains none of these metals, says Beers, astronomers think it comes from a time before these nuclear reactions started in earnest. The question is when.

Many other ancient stars do contain elements heavier than iron. Today, theorists can track their evolution using computer models of the nuclear reactions in their cores. From these simulations, they have worked out that the stars most deficient in iron are about 13 billion years old. But when they applied the same technique to HE 0170-5240, they discovered it is another 0.5 billion years older still. So the nuclear reactions that form heavy elements did not even start until the universe was between 0.2 and 0.7 billion years old.

The discovery of such a milestone in the history of the universe is a beautiful result. But not everything about the spectrum of HE 0107-5240 made sense. Beers and his colleagues found it contained another bombshell. “The star has 10,000 times as much carbon relative to iron as the sun,” says Beers. “In fact, it has the highest relative carbon abundance of any known star.” How could the surface of such an ancient star contain huge amounts of carbon, yet few heavy metals? It’s like finding a 100-year-old with skin as soft as a baby’s.

Different research groups have proposed a number of possibilities. Yuzuru Yoshii’s team at the University of Tokyo in Japan believes that during its lifetime HE 0107-5240 has passed through dense interstellar clouds, enriched with heavy metals flung out by younger supernovae. These clouds may have enveloped the ancient star, increasing its surface carbon content. But Beers is sceptical. “The difficulty with this scenario is that, with the right stellar history, it is possible to mimic any surface abundances at all,” he says. “It may be the right scenario but it is difficult to test.”

Meanwhile Masayuki Fujimoto of Japan’s Hokkaido University and his colleagues suggest that the carbon originated in the core of HE 0107-5240. According to theoretical models of ultra-low-metal stars, the convection currents that churn the matter in a star’s interior become stronger at certain points in its history. In the sun, for instance, convection currents only move material in the outer 25 per cent. However, during the helium-burning phase in an ultra-low-metal star, convection currents could reach down into its heart. This would dredge up carbon and other elements such as nitrogen and oxygen from the star’s interior to the surface.

Fujimoto’s theory looks promising: the surface of HE 0107-5240 has 200 times as much nitrogen as the sun. But this scenario also predicts high levels of oxygen, which is frustratingly difficult to spot. That’s because the light emitted by oxygen atoms in stars is absorbed by oxygen in the Earth’s atmosphere. Instead, researchers are looking for evidence of oxygen by finding the hydroxyl radical, which is made of a hydrogen and oxygen atom linked together. Because its fingerprints occur in a different part of the spectrum, they should be visible in modern telescopes. Astronomers have already bagged time on the Very Large Telescope to hunt for hydroxyl and they are now analysing the results.

But Beers is not convinced by Yoshii’s or Fujimoto’s interpretations. He prefers a third, more exciting scenario for the unusual carbon abundance. He believes the enhanced carbon reflects nuclear processes going on in the very first stars whose explosions provided the raw material for HE 0107-5240. “My bet is that the very first carbon in the universe was made by some nuclear process different from the one that makes it in today’s universe and it is this carbon we are seeing in these primitive stars,” he says. Beers doesn’t know what this process might be, though, since the lost generation of stars were totally different from modern stars.

It’s a bold claim and others are not so sure. Cowan agrees that we are seeing material ejected from the very first stars. But he suspects these lightweight elements are mixed with carbon forged in the star’s core and dredged up to the surface. “Nature isn’t simple,” he says. “In these stars with unusual carbon but low heavy elements, I think we are seeing the results of two different processes.”

Chris Sneden of the University of Texas in Austin wonders whether all this matters at all. “If one says that HE 0107-5240 was born with this peculiar composition, the story can’t be stopped at that point. One must ask how the preceding generation of stars produced so much carbon,” he says. “Whatever happens, it is marvellous that the early galaxy is turning out to have such a freak show of stars so wonderfully dissimilar to the sun.”

To have lived so long, Beers reckons HE 0107-5240 must be small because lightweight stars burn their fuel more slowly. He estimates its mass to be 80 per cent of the sun’s. Better still, he believes the presence or absence of certain heavy elements in HE 0107-5240’s spectrum may even tell us about the masses of the lost generation of first stars. Elements such as europium, thorium and uranium are made by nuclear reactions in the inferno of a supernova explosion. If the first stars were as massive as some theorists suspect, they would have left behind black holes when they exploded. The intense gravitational field of such a black hole would suck in any heavy metals before they had a chance to reach the interstellar medium. HE 0107-5240 should therefore show no signs of them. If, on the other hand, the first stars were much less massive, the gravitational force would have been much weaker and the heavy elements could have escaped. Before astronomers draw any conclusions though, they have to be sure HE 0107-5240 contains no heavy elements and not just low levels that remain undetected so far.

Remarkably, Beers thinks that the observed element abundances in HE 0107-5240 could tell us how fast the first stars were spinning. Nuclear reactions in a star are very sensitive to its temperature and density, both of which depend on the balance between gravity pulling all the gas inwards and thermal pressure pushing it out. In a spinning star the centrifugal force flings gas outwards, changing this balance. Astronomers expect different nuclear reactions in a spinning star, which would alter the elements visible on its surface.

Measuring the spectral fingerprints of ancient stars is not the only way to find out about the lost generation. They will also have left their mark on the afterglow of the big bang – the cosmic background radiation. In February results from NASA’s Wilkinson Microwave Anisotropy Probe, which has been measuring this radiation for over a year, showed that neutral hydrogen gas was re-ionised to form hydrogen ions as early as 200 million years after the big bang. Cosmologists believe this was done by intense ultraviolet radiation beamed out from the first stars.

Search for the ancients

And even though these stars are thought to be long dead, evidence of their existence can be seen today – as black holes in today’s universe. Astronomers have recently observed black holes with masses as great as a few hundred suns. These are heavier than the black holes created when modern stars die, yet lighter than the supermassive black holes thought to lurk at the centres of galaxies. The most likely explanation is that these black holes formed when the first stars exploded into supernovae at the end of their lives. “Their properties could tell us about the first stars,” says Beers.

Astronomers will need to observe many more stars like HE 0107-5240 before they can draw any definite conclusions about the first stars in the universe. So far, researchers have measured the spectra of a few tens of thousands of stars and found just one containing tiny amounts of iron. And these stars represent a mere fraction of all the stars in the galaxy. “The statistics are small,” admits Beers. “But even if only 1 in 50,000 stars in the Milky Way has an ultra-low metal abundance like HE 0107-5240, as many as a million stars of a similar nature are yet to be found.”

To find them, astronomers will need to observe tens of millions of stars. Beers believes the spectroscopy techniques used in the previous surveys are simply not good enough because they cannot be used on very faint stars. Instead, he advocates looking for stars with certain distinctive colours. Metals absorb predominantly blue light, says Beers, so low-metal stars will appear much bluer than modern stars.

A great place to start looking is the Sloan Digital Sky Survey, presently under way in New Mexico. It has already recorded the colours of many millions of stars. “Based on spectroscopic follow-up of the most interesting objects, we’re already finding stars with 100 times to 1000 times less iron than the sun,” says Beers. “And we’re confident of finding stars with 10,000 to 100,000 times less, or even lower amounts.”

Although most astronomers studying the Sloan data aren’t looking for metal-poor stars, Beers believes they will unwittingly find them. The Sloan researchers are looking for quasars, bright objects that herald the birth of newborn galaxies. But because quasars look like stars when viewed through a telescope, they can be mixed up with the ultra-low-metal stars. “About 10 per cent of the quasar candidates are not quasars but the kind of stars we are looking for,” says Beers. “The good news is that the quasar hunters are doing a job for us.”