A Tantalizing Signal From the Early Universe

Astronomers have found evidence of the first stars igniting in cold, dark gas, a detection with huge implications for our understanding of the cosmos.

Artist's rendering of the first stars in the universe, orbs of blue light emanating outward in reddish space
Artist's rendering of the first stars in the universe (N.R. Fuller / National Science Foundation)

Near the beginning, not long after the Big Bang, the universe was a cold and dark place swirling with invisible gas, mostly hydrogen and helium. Over millions of years, gravity pulled some of this primordial gas into pockets. The pockets eventually became so dense they collapsed under their own weight and ignited, flooding the darkness with ultraviolet radiation. These were the very first stars in the universe, flashing into existence like popcorn kernels unfurling in the hot oil of an empty pan.

Everything flowed from this cosmic dawn. The first stars illuminated the universe, collapsed into the black holes that keep galaxies together, and produced the heavy elements that would make planets and moons and the human beings that evolved to gaze upon it all.

This epoch in our cosmic history has long fascinated scientists. They hoped that someday, using technology that was calibrated just right, they could detect faint signals from that moment. Now, they think they’ve done it.

Astronomers said Wednesday they have found, for the first time, evidence of the earliest stars. Using a table-sized radio instrument in the desert in Western Australia, the researchers detected radio emissions from the cold hydrogen that interacted with brand-new stars in that stage of the early universe.

Astronomers from Arizona State University, MIT, and the University of Colorado at Boulder, funded by the National Science Foundation, spent more than a decade trying to find this signal, calibrating and recalibrating the technology. Their results were published in Nature.

“This is a huge milestone,” says Judd Bowman, the Arizona State University astronomer who led the effort. “If you really look at our cosmic origins as humans, what are all the events in the universe that had to happen for us to be there? The first rung in that ladder is these first stars.”

The nature of this signal suggests a new estimate for when the first stars emerged: about 180 million years after the Big Bang, slightly earlier than many scientists expected, but still within the expectations of theoretical models. “This number will become the gold standard for when the first light by the first stars came about in the universe,” says Anna Frebel, an astrophysicist at MIT who studies the early stars of the universe, and who was not involved in this research. “Knowing this number is important for many branches of astrophysics, so it’s exciting to finally have a measurement.”

The astronomers can’t see the actual light from the first stars, but they know they’re there because they can detect the stars’ effects on the surrounding medium, the cold hydrogen gas. In the early universe, that hydrogen gas was pretty much just floating along. When the first stars appeared, things changed quickly. The new light, in the form of ultraviolet radiation, shot through the gas, altering the state of the single electron in the hydrogen atoms. This prompted the hydrogen to start absorbing energy from the background radiation that filled the universe, left behind after the Big Bang. When the atoms absorbed this energy, their electrons became excited. Then, to return to their original state, the electrons sent the energy back out.

Theoretical models predicted that this tiny jump could be detected at a certain radio frequency—and this is what the astronomers detected.

Bowman and his colleagues spent years fine-tuning their radio antenna and receiver, which together resemble a glass coffee table. Their instrument is designed to absorb all the radio waves coming from the sky of the southern hemisphere and weed out waves coming from other sources, including the entire Milky Way galaxy and any Earth technologies operating at similar frequencies. Bowman says they detected the signal in 2015, but spent the next two years making sure it was what they thought it was.

The ground-based radio spectrometer at the Murchison Radio-Astronomy Observatory in Western Australia (CSIRO Australia)

It certainly helped that the instrument was located at the Murchison Radio-Astronomy Observatory (MRO). The observatory is managed by CSIRO, the national science agency of Australia, which limits the use of radio transmitters within about 160 miles of the site. Theoretical models had predicted the signal of the primordial hydrogen gas would overlap with the frequencies used by FM radio stations. Someone broadcasting Top 40 too close to the site could very well have drowned out the message coming from the beginning of time.

The nature of the radio waves Bowman and his colleagues detected mostly matches theoretical predictions, but not everything lines up. When they tuned their instrument to listen to the frequency for hydrogen gas that models predicted, they didn’t hear anything. When they decided to search in a lower range, they got it. But the signal they found was stronger than expected. That meant that the hydrogen gas in the early universe was much much colder—perhaps nearly twice as cold—than previously estimated.

This is where things get even more interesting. Only two things could explain why the temperature of the gas was lower than predicted. There may have been more background radiation present after the Big Bang than astronomers previously thought, but they say that’s unlikely. The more plausible explanation involves an interaction between hydrogen gas and another cosmic mystery: dark matter, the invisible stuff that scientists say makes up most of the universe. “The idea is that the dark matter is colder than the gas,” Frebel says. “By interacting, the gas transfers some heat to the dark matter to warm it up a little. Hence the gas we observe has been cooled.”

This possibility is explored in a second paper, also published in Nature on Wednesday, by Rennan Barkana, an astrophysicist at Tel Aviv University who first proposed this potential interaction. If Barkana’s theory bears out, scientists may be able to figure out some of the properties of these dark-matter particles.

This would be huge for the field. The existence of dark matter has so far only been inferred indirectly from observational data. Astronomers can see the effects of dark matter by studying certain gravitational effects on the light of massive clusters of galaxies. But these effects only allow astronomers to map big chunks of dark matter, not individual particles, says Priyamvada Natarajan, a theoretical astrophysicist at Yale, who was not involved in the research.

“It’s like looking at a huge sand dune,” Natarajan says. “You know how the sand dune is assembled, you can show how it dissipates in the wind, how it’s washed out by water. But you don’t know what a grain of sand is made of.”

Natarajan says the study of the early signal detected by Bowman and his team could help theoretical astrophysicists uncover new, fundamental physics beyond our current understanding. In order for the observed cooling to make sense, the dark-matter particle would need to be much lighter than previously thought.

“Theorists are going to be busy,” Natarajan says.

Bowman says other teams around the world have been working to build and design instruments to detect this signal from the early universe, and he expects they should be able to confirm the results in the coming months.

“Only confirmation by other groups and experiments will truly usher in the new era of detecting hydrogen from the very first observable phase in the universe,” Frebel says. “As exciting and informative it is, we need to be cautious until it has been confirmed by others before entering it into the textbooks.”

Further study will let astronomers learn more about the first stars of the universe. “As we move forward, we can pull out the actual properties of the stars using models,” Bowman says. “Stars of a certain mass will affect our absorption signal.”

Bowman says it’s unlikely astronomers will be able to see any earlier into the history of stars than this, at least in our lifetimes, with the current state of radio technology. After all, the earlier they go, the less there is to see. This might seem like a rather sad prospect. But a quick glance at the night sky, far from the glow of city lights, should extinguish any feelings of disappointment. There may be nothing to see before the first stars flickered on, but there is plenty to witness after they did.