Here's a thought experiment that has unsettled astrophysicists: Start the clock at the beginning of time. Form a black hole in the usual way, through the collapse of a massive star. To make it grow, force-feed it with gas, which will resist being devoured by heating up and dispersing as it nears the black hole's maw. Try to grow a black hole fast enough to explain the ones that existed in the real universe when it was just a billion years old: monsters a billion times the mass of the sun that drive the powerful beacons called quasars.
"It's very difficult," says Amy Reines, an observational astronomer at the National Optical Astronomy Observatory in Tucson, Arizona. Astronomer Nico Cappelluti of Yale University is more definitive. "There is no way to grow such a massive black hole from an ordinary stellar black hole," he says. But he and others see hints of a faster route involving primordial gas clouds, as they described last week at the winter meeting of the American Astronomical Society in Grapevine, Texas.
Some theorists had already suggested that, instead of coming from collapsed stars, the behemoth black holes in the early universe could have gotten a head start. Huge gas clouds left by the big bang might have quickly shrunk under their own gravity and, instead of splintering into many stars, condensed into black hole seeds 10 thousand to 100 thousand times heavier than the sun. Those seeds would have grown further, to billions of solar masses, by sucking in stars and gas. But although a few candidates for such objects have been timidly proposed, these "direct collapse" black holes would be hard to spot and harder to confirm with current technology.
A new clue to their possible existence, which Cappelluti's team presented at the meeting, is an eerie concordance found in views of the distant universe that rely on different wavelengths. Along with the familiar cosmic microwave background—the afterglow of the big bang—the distant universe is suffused with an infrared background, thought to come from galaxies and stars too faint and far away to see. With all known galaxies and stars scrubbed away, the infrared background is about 20 times splotchier than expected. Similarly splotchy is the cosmic x-ray background, emitted by matter falling into black holes in the distant universe. And in work Cappelluti and his colleagues are now preparing for publication, they show that some of the infrared and x-ray patchiness matches, across a swath of sky about three times larger than previously tested.
"The more we look at it, the more confirmation we find," Cappelluti says. As to the cause, "you need to have sources that are very powerful in the x-ray, very powerful in the infrared, and do not emit anything in any other band." Direct collapse black holes, gobbling up dense gas clouds in the early universe, could fit the bill.
Another lead comes from the Chandra Deep Field South, an image created by a space-based x-ray telescope that observed the same patch of sky for a cumulative 81 days. Released at the meeting, the image shows more than 2000 black holes glowing brightly as they swallow up matter. These probably aren't newborn direct collapse black holes—they are too close and recent. But just as important is what can't be seen: the fainter glows from smaller black holes, slowly putting on weight, as expected if supermassive black holes were born star-sized and grew gradually. "I think it's a hint" that monster black holes might have sprung into existence quickly, through direct collapse, says Niel Brandt, an astronomer at Pennsylvania State University in State College who led the team interpreting the data.
Neither clue is definitive, and theorists' models don't give clear guidance about what observers should look for next to confirm that direct collapse black holes exist. "The model predictions," says Brandt, "are a bit squishy." And although astronomer Asantha Cooray of the University of California, Irvine, thinks the correlation between the infrared and x-ray sky is real, there might be an easier way to explain it: loose halos of stars around galaxies. "We confirm the measurement, but we cannot confirm the interpretation," Cooray says.
Astronomers plan other assaults on the question—for example, surveys of dwarf galaxies, among the universe's most pristine. They may offer "some insight into how the first seeds were formed, and have a memory of the seeding scenario," Reines says. If black hole seeds come from stars, the process should have given every dwarf galaxy its own supermassive black hole. But if seeds are born big, under special conditions, they should be rarer, and many dwarf galaxies would lack a black hole.
Directly detecting a direct collapse black hole, though, may require future instruments. The upcoming James Webb Space Telescope, for example, might see one condensing in infrared light, whereas the proposed Lynx x-ray telescope might spot a newborn black hole, fresh from a collapsing cloud, snacking on its first meals.