Skip to main content

The 1st stars in the universe formed earlier than thought

The first stars in the universe formed even earlier than astronomers had thought, a new study suggests.

Researchers probing the early universe found no sign of first-generation stars in galaxies that existed just 500 million to 1 billion years after the Big Bang.

"These results have profound astrophysical consequences, as they show that galaxies must have formed much earlier than we thought," study lead author Rachana Bhatawdekar, a research fellow at the European Space Agency (ESA), said in a statement.

Related: Peering back to the Big Bang & early universe (images)

Artist's illustration of the early universe. (Image credit: ESA/Hubble, M. Kornmesser, and NASA)

Bhatawdekar and her colleagues used the NASA/ESA Hubble Space Telescope, NASA's Spitzer Space Telescope and the European Southern Observatory's Very Large Telescope in Chile to hunt for "Population III" stars in a variety of distant galaxies.

Population III stars were the first suns to form in our 13.8-billion-year-old universe, and they're identifiable by their unique composition: just hydrogen, helium and lithium, the only elements around immediately after the Big Bang. Heavier elements were forged in the cores of these stars and their successors.

(The somewhat confusing moniker results from the fact that astronomers had already classified the stars of our own Milky Way galaxy into two groups before considering their super-old cousins. "Population I" stars, such as Earth's sun, are rich in heavy elements, and "Population II" stars are considerably less so.)

The research team took advantage of a phenomenon called gravitational lensing to bring their hard targets into view. In each case, they used a giant galaxy cluster in the foreground as a sort of magnifying glass, allowing them to study small, distant and incredibly faint galaxies. 

It has taken the light from these background galaxies 12.8 billion to 13.3 billion years to reach Earth — meaning that these objects are time capsules harboring lots of information about the early universe, including what types of stars were shining back then.

"We found no evidence of these first-generation Population III stars in this cosmic time interval," Bhatawdekar said. 

This image from the NASA/ESA Hubble Space Telescope shows the galaxy cluster MACS J0416. This is one of six galaxy clusters being studied by the Hubble Frontier Fields program, which together have produced the deepest images of gravitational lensing ever made. Scientists used intracluster light (visible in blue) to study the distribution of dark matter within the cluster.

This image from the NASA/ESA Hubble Space Telescope shows the galaxy cluster MACS J0416. This is one of six galaxy clusters being studied by the Hubble Frontier Fields program, which together have produced the deepest images of gravitational lensing ever made.  Scientists used intracluster light (visible in blue) to study the distribution of dark matter within the cluster. (Image credit: NASA, ESA, and M. Montes (University of New South Wales))

Population III stars and the first galaxies must therefore be older still — so old that they're beyond Hubble's reach. But NASA's $9.8 billion James Webb Space Telescope, which is scheduled to launch next year, may be able to spot them, study team members said.

The new results, which were presented this week at the 236th meeting of the American Astronomical Society and will be published in an upcoming issue of the journal Monthly Notices of the Royal Astronomical Society, shed other light on the early universe as well. 

For example, low-mass, faint galaxies like the ones observed in the new study were probably responsible for "cosmic reionization," Bhatawdekar and her colleagues said. In this process, which began perhaps 400 million years after the Big Bang, radiation split the hydrogen atoms pervading the universe into their constituent protons and electrons. Reionization was a big cosmic transition, and getting a better handle on how it happened could help astronomers better understand our universe's structure and evolution, scientists have said

Mike Wall is the author of "Out There" (Grand Central Publishing, 2018; illustrated by Karl Tate), a book about the search for alien life. Follow him on Twitter @michaeldwall. Follow us on Twitter @Spacedotcom or Facebook

OFFER: Save 45% on 'All About Space' 'How it Works' and 'All About History'!

For a limited time, you can take out a digital subscription to any of our best-selling science magazines for just $2.38 per month, or 45% off the standard price for the first three months.View Deal

Join our Space Forums to keep talking space on the latest missions, night sky and more! And if you have a news tip, correction or comment, let us know at: community@space.com.

  • rod
    FYI, some discussion about the first stars and metal recycling in the universe, https://forums.space.com/threads/how-many-times-has-our-matter-cycled-through-the-stellar-life-cycle.31756/
    Reply
  • dfjchem721
    Seems that we keep reading about everything forming earlier and earlier than previously proposed. Galaxies, stars, black holes, all apparently forming much earlier then BB physics (had) predicted, or so it seems. Well, as most "models" predicted. One must assume (hope?) that the folks who have been telling us all this stuff have some wiggle room up their theories.

    As more and more of these "early birds" are found, one can only wonder what the earliest really is. It would appear that we still don't have the faintest glimmer, figuratively and literally!

    Could there be an inherent flaw in the prevailing BB dogma? Stay tuned.........
    Reply
  • Helio
    dfjchem721 said:
    Seems that we keep reading about everything forming earlier and earlier than previously proposed. Galaxies, stars, black holes, all apparently forming much earlier then BB physics (had) predicted, or so it seems. Well, as most "models" predicted.
    Did the models claim 400 million years was too early for formation?

    One must assume (hope?) that the folks who have been telling us all this stuff have some wiggle room up their theories.
    Yes, that's very likely, but future observations (e.g. JWST) will improve the unwanted degrees of wiggle. Part of the problem is trying to understand how having only hydrogen and a little helium would form a star, how big they could be, etc. These stars don't exist even in the old globular clusters.

    Could there be an inherent flaw in the prevailing BB dogma? Stay tuned.........
    I suspect all is in order, but we still need observations to help build better models.
    Reply
  • dfjchem721
    Helio said:
    I suspect all is in order, but we still need observations to help build better models.

    Helio, I have seen a number of proposals for the first galaxies, stars, etc over the years. For galaxies, it seems that many started around 1 billion years after BB, or they used to. Then they got younger and younger. It seems like every year or so they are re-modeled based on new observations rather than theory. Clearly empirical evidence is (or should be) defining.

    So , have been watching the time-frame on these "original" objects appear to go down over time, not in "strict" keeping with most BB models (at those times). Not even sure what the latest story is, nor does anyone else I suspect.

    As you suggest, the primary aspects of BB physics is probably correct. Details are the devil as usual. Hoping for the JWST to tell us more, assuming it ever gets up there.
    Reply
  • Helio
    dfjchem721 said:
    Helio, I have seen a number of proposals for the first galaxies, stars, etc over the years. For galaxies, it seems that many started around 1 billion years after BB, or they used to. Then they got younger and younger. It seems like every year or so they are re-modeled based on new observations rather than theory. Clearly empirical evidence is (or should be) defining.
    That's sounds logical. The objective evidence available to tweak that early period is very scant, though the overall age does come with strong evidence.

    So , have been watching the time-frame on these "original" objects appear to go down over time, not in "strict" keeping with most BB models (at those times). Not even sure what the latest story is, nor does anyone else I suspect.
    The article seems to strongly discount the 500M to 1B yr period for early formation. These stars and galaxies, regardless of their formation period, must exist to explain today's stars and galaxies, reionization, etc., so it's just a matter of time before equipment and techniques get us to where we need to be.

    I am a little surprised that more isn't known. I recall a U Texas paper that, among other claims, revealed a galaxy estimate for the universe of roughly 2 trillion. It wasn't that long ago the estimate was 250 billion. They too used lensing and incredible subtraction techniques to let us see those more distant galaxies.

    It is interesting how the 13.8 B yr. age is so accurate, though it was 13.7 B years a year ago or so. But that is due to very detailed evidence found in things like the CMBR, so the BBT requires each piece of the puzzle to be better resolved even though we have all the side pieces in place, so to speak. This isn;t surprising given how broad BBT is.
    Reply
  • dfjchem721
    Yes, the existing data is very poor, after reading up on Pop-III stars. They are in fact hypothetical, and deemed (by many) to have been formed exclusively as super giants, with 100s of SMs. Other concepts put them in a range of sizes, with some of them still existing today. It seems that this is all based on metallicity.

    I recall reading, not sure where, that some smaller stars could have formed very early, during the formation of Pop-III stars, and that a few of them still exist. Seem to recall that one super-low metallicity star was observed in a binary in the Milky Way, but don't recall the details.

    There is no proof that the earliest stars (Pop-III) were all super-giants and now long gone. Was surprised to hear that there are none in globular clusters, but then that observation may simply be defined as size, rather then metallicity? How can you measure the metallicity of all (e.g.) 100,000 stars in a cluster?

    I do recall the formation of blue stars in globular clusters (GCs), by the fusion of two old yellow stars. The combined mass of hydrogen was enough to light them up as blue giants. Is it possible that some of the stars in GCs are actually fused Pop-III stars that started smaller than models suggest they should?

    It seems likely that there are a few Pop-III stars out there that are low mass and long-lived. If so, they would offer unique opportunities to study the earliest stars.
    Reply
  • Helio
    dfjchem721 said:
    Seem to recall that one super-low metallicity star was observed in a binary in the Milky Way, but don't recall the details.
    Yes, but even these are likely not considered in the Pop III class. There were no metals for the first stars since Li and Be were so rare.

    Was surprised to hear that there are none in globular clusters, but then that observation may simply be defined as size, rather then metallicity?
    Those clusters consists of very old stars, but only the less massive stars survive for very long periods, and smaller stars are more convective so as they produce heavier elements. These, I assume, would rise to the surface, and their spectrum will reveal their greater metallicity.

    How can you measure the metallicity of all (e.g.) 100,000 stars in a cluster?
    IIRC, the Sloan Survey used fiber optics that ran through metal templates to align perhaps a few hundred stars for each imaging section to get the spectrum of each star. The Hobby-Eberly a year or so ago was taking time to do something similar but on a larger scale with its 9 meter mirror, and all for spectroscopy.

    Is it possible that some of the stars in GCs are actually fused Pop-III stars that started smaller than models suggest they should?
    Good question. It's my understanding that these clusters are still enigmatic as to how they are what they are.

    It seems likely that there are a few Pop-III stars out there that are low mass and long-lived. If so, they would offer unique opportunities to study the earliest stars.
    Outside of the globular cluster regions that are normally outside the galactic plane, it might take a rogue intergalactic star to remain unaffected by the spewing of normal stars, novae, and supernovae. But it would suffer with growing metalicity over time and, being small, the convective issue as mentioned above.
    Reply
  • dfjchem721
    Noting supernova, your comments have led me to two mechanisms I was unaware of for the end-stage of "super giant" stars :

    "Photodisintegration" and "Pair-instability" supernova. The former leads to black holes and a lot of ejecta, the latter only to ejecta, no remnant. Both are apparently theoretical as I find no immediate evidence that they even occur. They would certainly be rare in our neighborhood. I can see that Pair-instability supernova are quite unique, but Photodisintegration supernova appears closely related to the standard core-collapse of smaller giant stars. I don't seem to find an exact size relating to either mechanism. Perhaps they are competing theories on the end fate of super giants.

    This is pretty wild stuff. Are there any other proposed fates for super giant stars? Any that collapse into a BH without radiation or ejecta? Seem to remember that somewhere, but it might be mish-mashed from some of my readings on this rapid course in Pop-III stars, etc.

    My first and only prior exposure to super giant stars is Eta Carinae, estimated around 150 SMs. Apparently the largest star known to exist in the Milky Way. No doubt you are aware of this star, and it is believed to be short-lived, and could go off at any moment. As this star is so large, is it expected to undergo one of the above supernova mechanisms? A remnant BH would quickly distinguish the two mechanisms, one would think.

    Very interesting story on the Sloan Survey. Will have to do some looking into that.
    Reply
  • Helio
    dfjchem721 said:
    Very interesting story on the Sloan Survey. Will have to do some looking into that.

    Wiki has a nice article:

    Wiki said:
    The spectrograph operates by feeding an individual optical fibre for each target through a hole drilled in an aluminum plate. Each hole is positioned specifically for a selected target, so every field in which spectra are to be acquired requires a unique plate. The original spectrograph attached to the telescope was capable of recording 640 spectra simultaneously, while the updated spectrograph for SDSS III can record 1000 spectra at once. Over the course of each night, between six and nine plates are typically used for recording spectra. In spectroscopic mode, the telescope tracks the sky in the standard way, keeping the objects focused on their corresponding fibre tips.

    Every night the telescope produces about 200 GB of data.
    Reply
  • rod
    In the 1970s, searches for Population III stars looked at red dwarfs (very logical because these stars burn H very slowly and last much longer than 13.8 billion years). None found. Population III star EOS modified to make them larger, burn faster, and disappear earlier in BB cosmology while seeding the early universe with r-process and s-process elements. Astronomy is correct to search for this type of star, predicted by BBN and a requirement it seems. So far Population III stars remain undiscovered, i.e. not verified like other stars. Perhaps the issue is BBN. The universe never existed in the primordial condition for BBN, thus the creation of Population III stars did not take place in the early universe. If this is the situation (Pop III simply never there), much is at stake for BB cosmology now and the search for these 1st generation stars. I prefer confirmed stellar spectra to show Pop III stars existed or are still in the universe.
    Reply