WEBVTT 00:00.240 --> 00:05.840 Gigantic black holes existed in a cosmic  eyeblink after the Big Bang. That should 00:05.840 --> 00:13.608 be impossible. The James Webb Space Telescope has  either made the problem much worse, or solved it. 00:08.160 --> 00:13.360 by celebrating the space time which contains them  all. Links to the merch store in the description. 00:13.360 --> 00:18.880 Black holes can grow fast. Exponentially fast.  The more mass they consume, the stronger their 00:18.880 --> 00:24.080 pull and the faster they can eat. The largest  supermassive black holes seen in the late 00:24.080 --> 00:30.480 universe can be several billion suns in mass. But  they’ve also had several billion years to grow, 00:30.480 --> 00:35.840 typically during brief periods as bright  quasars when they feed on the incandescent 00:35.840 --> 00:41.360 whirlpools of infalling gas. But then we  started to find such enormous black holes 00:41.360 --> 00:49.680 at earlier and earlier times in quasars shining  out of the first billion years or less of cosmic time. Now, 00:49.680 --> 00:55.920 even exponential growth doesn’t seem fast  enough to explain how these monsters formed. 00:55.920 --> 01:00.880 The James Webb Space Telescope only made  things worse, finding faint specks of 01:00.880 --> 01:05.360 reddened light, dubbed little red dots, that  seem to be powered by outsized black holes 01:05.360 --> 01:10.400 in the first several hundred million years of  cosmic history. What’s going on here? Is the 01:10.400 --> 01:15.680 early universe forming gigantic black holes  in ways we haven’t imagined? Do we completely 01:15.680 --> 01:20.160 not understand how black holes grow? Is  the universe older than we thought? Well, 01:20.160 --> 01:26.480 recently JWST found a very unusual  little red dot that may solve everything. 01:26.480 --> 01:31.600 We assume that supermassive black holes grow  from the much smaller black holes left behind 01:31.600 --> 01:37.760 when massive stars die (stellar black holes). The first generation of stars were probably especially enormous 01:37.760 --> 01:43.920 and so left behind especially large seed black  holes—let’s say 100 times the mass of the sun, 01:43.920 --> 01:49.440 compared to 5-15 solar masses in the modern  universe. At the theoretical maximum accretion 01:49.440 --> 01:56.320 rate, black holes double in size every 33 million  years. So that 100 solar mass black hole is going 01:56.320 --> 02:01.040 to take at least 700 million years to give  us a billion solar mass supermassive black 02:01.040 --> 02:08.560 holes. But we see these gigantic monster, and larger,  from within the first 700 million years … so, 02:08.560 --> 02:14.480 sounds like it’s just barely possible to go from  stellar to supermassive in that time? Well, not even. 02:14.480 --> 02:20.320 The first stars took at least 100 million years to  form after the Big Bang, probably more. And it’s 02:20.320 --> 02:26.400 hard to come up with scenarios in which a black  hole feeds without any interuption for so long. 02:26.400 --> 02:32.480 Something needs to give. There are two obvious possible solutions to the problem of overly massive early 02:32.480 --> 02:39.200 supermassive black holes: 1) Black hole seeds  were much larger than expected. or 2) Black holes 02:39.200 --> 02:44.560 grew much faster than we thought possible.  So, start bigger and/or grew faster. There 02:44.560 --> 02:49.440 are various ways to start big—like having  ridiculously gargantuan stars in the very 02:49.440 --> 02:53.920 early universe, or even collapsing clouds  of pristine hydrogen directly into black 02:53.920 --> 03:00.400 holes without going through a stellar stage.  Now growing faster requires creative ways to get 03:00.400 --> 03:05.360 around the “theoretical” feeding limit that  I mentioned, and which we’ll come back to. 03:05.360 --> 03:08.960 But when the first overly-massive black holes were  discovered around the turn of the millennium, 03:08.960 --> 03:14.320 some moderate rethinking of black hole formation  and growth seemed enough to explain the issue. 03:14.320 --> 03:20.080 But as our telescopes got better we saw too-big  black holes at greater and greater distances, 03:20.080 --> 03:24.080 our re-theorizing had to do a  lot more work to explain them. 03:24.080 --> 03:30.400 And then came JWST and the little  red dots and things got much worse. 03:30.400 --> 03:36.960 Little red dots are tiny, compact objects  that appear extremely red in JWST images of 03:36.960 --> 03:43.840 the distant universe. And many of them have broad  emission lines, especially from hydrogen. Emission 03:43.840 --> 03:49.680 lines are usually sharp spikes that occur at very  specific wavelengths due to electron transitions 03:49.680 --> 03:56.480 in specific atoms. These lines can be broadened  if the gas producing the line is moving very fast, 03:56.480 --> 04:02.160 due to the Doppler effect. These broad emission  lines are classic features of quasars, where the 04:02.160 --> 04:07.600 gas is moving at enormous speeds, either orbiting  or flowing towards or away from the black hole. 04:07.600 --> 04:12.720 These features are a standard way to measure  the mass of a distant supermassive black 04:12.720 --> 04:17.680 hole. If we assume that the gas is moving fast  because it’s in orbit around that black hole, 04:17.680 --> 04:22.560 then the width of the line gives the orbital  speed and hence black hole mass. Some 04:22.560 --> 04:28.240 assumptions included. And I’ll come back to those. Based on the line widths in the little red dots, 04:28.240 --> 04:34.880 black hole masses from millions to hundreds  of millions of Suns have been measured. 04:34.880 --> 04:40.480 Now, those aren’t quite billions of Suns like  the biggest seen in the early universe, 04:40.480 --> 04:45.360 but they’re definitely in the supermassive  category. And these black holes are also way, 04:45.360 --> 04:50.240 way too big for the galaxies that they live in.  Supermassive black holes are always found in the 04:50.240 --> 04:54.720 centers of galaxies, and in the modern universe  they appear to have grown in relative lock-step 04:54.720 --> 05:00.320 with their galaxies over the age of the universe.  Bigger galaxies have bigger central black holes. 05:00.320 --> 05:06.160 But the host galaxies of the little red dots  are tiny compared to their black holes. That 05:06.160 --> 05:12.240 suggests that the central black hole massively  outpaced its galaxy as it grew, or started out 05:12.240 --> 05:18.880 ridiculously large. So far JWST and its little  red dots are just making the problem worse. 05:18.880 --> 05:26.640 Then along comes MoM-BH-1*. It was first  detected my JWST in its PRIMER survey in 2023, 05:26.640 --> 05:32.000 and identified as weird and followed up by  Rohan Naidu and collaborators. They presented 05:32.000 --> 05:38.720 the results of their JWST spectroscopy and  their interpretation in Nature very recently. 05:38.720 --> 05:44.160 This particular little red dot existed a  mere 660 million years after the Big Bang. 05:44.160 --> 05:49.440 It has a spectrum that on one hand looks like a  classic little red dot, with broad emission lines 05:49.440 --> 05:55.920 that suggest hydrogen gas moving at more than  3000 kilometers per second. That screams accreting 05:55.920 --> 06:02.800 black hole. On the other hand, it also kind of  looks like a star. It has what we call a Balmer 06:02.800 --> 06:07.840 break—a huge chunk missing from its spectrum due  to absorption by hydrogen, just as we see in … 06:07.840 --> 06:14.720 In emission lines—like the broad lines seen in our  quasars—an energized cloud of gas emits specific 06:14.720 --> 06:20.560 wavelengths of light as electrons in its atoms  drop down in energy levels. But if your gas cloud 06:20.560 --> 06:27.040 happens to obscure another source of light, those  same specific wavelengths can also be sucked out 06:27.040 --> 06:33.680 of that background light as electrons in the gas  absorb photons to jump up in atomic energy levels. 06:33.680 --> 06:38.800 That happens with stars—the stellar atmosphere  absorbs specific wavelengths of light coming from 06:38.800 --> 06:44.800 deeper in the star. Because it’s only the light that’s  able to excite an electron that gets absorbed, 06:44.800 --> 06:52.640 this leads to a spectrum with many narrow gaps. Sufficiently energetic photons can completely free 06:52.640 --> 06:59.600 an electron from its atom. So in the right  conditions, all photons above this so-called 06:59.600 --> 07:07.200 ionization energy get absorbed by the gas. That translates  to the absorption of all photons below a specific 07:07.200 --> 07:12.800 wavelength corresponding to that energy. The Balmer limit is the wavelength below which 07:12.800 --> 07:19.920 all photons get absorbed by hydrogen gas—364.6  nanometers. More precisely, it’s the absorption 07:19.920 --> 07:24.880 limit for hydrogen with electrons sitting in  the second energy level. But for many stars, 07:24.880 --> 07:29.840 their outer atmosphere is full of hydrogen  in this state and so all light from deeper 07:29.840 --> 07:34.240 in the star that’s beyond the Balmer limit  gets sucked up in the atmosphere and doesn’t 07:34.240 --> 07:43.600 escape. The result is a spectrum like this… We  call this 364.6nm transition the Balmer break. 07:43.600 --> 07:49.440 Right, good. So, this is the spectrum  for MoM-BH-1*. It has a very clear Balmer 07:49.440 --> 07:55.040 break. It looks like a star, but actually with a  stronger Balmer break than almost all stars. It 07:55.040 --> 08:00.000 also looks like a quasar with its broad  emission lines. So, what’s going on? 08:00.000 --> 08:05.520 Well it could be that this Balmer break is really from  the galaxy surrounding the quasar. After all, 08:05.520 --> 08:10.160 galaxies are made of stars and stars can have  Balmer breaks. But there are other hydrogen 08:10.160 --> 08:15.920 absorption lines in the spectrum that suggest  that we’re looking through an enormous depth of 08:15.920 --> 08:22.080 hydrogen at colossal densities—something the  size of a solar system with density between 08:22.080 --> 08:27.520 a billion and a hundred billion atoms per cubic  centimeter. That’s not what stars look like. It’s 08:27.520 --> 08:33.520 not even what many stars added together look like.  This thing looks like a single vast atmosphere 08:33.520 --> 08:40.160 surrounding … something. Presumably the black  hole. This type of object—a star-like envelope 08:40.160 --> 08:45.600 powered by an accreting black hole has been  dubbed the black hole star. Not sure if the name 08:45.600 --> 08:52.080 will stick, but it's catchier than little red dot. We’re used to seeing accreting black holes that look more 08:52.080 --> 08:58.240 like this. A black hole surrounded by a relatively  thin whirlpool—accretion disk—of stuff pouring in 08:58.240 --> 09:03.120 to feed it. Most of the light we see comes from  the inner part of the disk where its infall 09:03.120 --> 09:08.800 results in crazy speeds, and friction between  adjacent streams heats the disk to an incandescent 09:08.800 --> 09:14.960 glow. Most of that light has a relatively free  path away from the disk, and so we only see strong 09:14.960 --> 09:21.440 absorption in rare cases, and never this sort  of complete cocooning of the inner light source. 09:21.440 --> 09:27.920 So apparently quasars can look weird in the early universe.  But that weirdness may actually answer a lot 09:27.920 --> 09:37.360 of our original questions. This cocoon changes what light can  get out, but it also changes what can get in. 09:37.360 --> 09:42.400 Now earlier I mentioned a “theoretical limit” to  how fast black holes can grow. As stuff falls 09:42.400 --> 09:48.160 into a black hole it heats up and radiates. But  that same radiation pushes back on the infalling 09:48.160 --> 09:54.400 material and can slow and even stop accretion. In  the most obvious scenario, the infalling stuff is 09:54.400 --> 09:59.120 ionized hydrogen and the outgoing radiation is  scattering off its electrons. Then there’s a 09:59.120 --> 10:04.160 very clean limit for how brightly black holes  can radiate and how fast they can grow. It’s 10:04.160 --> 10:11.440 called the Eddington limit. The vast majority of  accreting black holes we’ve measured are below 10:11.440 --> 10:17.200 this limit. It takes some creative tweaking of the  accretion physics to get much above this limit. 10:17.200 --> 10:21.760 A lot of the work in understanding how early  black holes could have gotten so big is about 10:21.760 --> 10:27.760 finding new ways to break the Eddington limit—of  enabling what we call super-Eddington accretion. 10:27.760 --> 10:31.360 Well, our new little red dot  gives us solid evidence that 10:31.360 --> 10:34.960 the Eddington limit has been thoroughly shattered. 10:34.960 --> 10:39.280 One of the assumptions in determining  the Eddington limit is that the radiation 10:39.280 --> 10:45.440 eventually escapes the object. It can take a long  time for any one photon to fight its way through the 10:45.440 --> 10:51.440 surrounding gas, but as long as it does ultimately  escape, then it’s done the maximum work in pushing 10:51.440 --> 10:58.320 that gas outwards. But what if some of the light  never escapes? In the case of an extremely dense, 10:58.320 --> 11:03.680 infalling cocoon, the amount of time taken for  a photon to escape can exceed the amount of 11:03.680 --> 11:10.480 time it takes the gas to reach the black hole. That means photons get swept down and swallowed by the black 11:10.480 --> 11:17.520 hole before they can exert maximum outward push.  This is a recipe for super-Eddington accretion. 11:17.520 --> 11:21.680 Astrophysicists have speculated on this  sort of process for the extreme growth 11:21.680 --> 11:28.800 rates of early black holes. But MoM-BH*-1 may  be the most direct evidence we’ve discovered. 11:28.800 --> 11:34.560 But this doesn’t solve our problem because our  estimate of the black hole mass in this object 11:34.560 --> 11:41.040 is still huge, certainly for such a small galaxy. If this  and all the other little red dot black holes 11:41.040 --> 11:46.000 are so big, when did they get this big? There  should also be a population of smaller black 11:46.000 --> 11:52.880 holes with crazy accretion rates. But in fact,  this one object may solve that problem too. 11:52.880 --> 11:58.880 As I said earlier, black hole masses for distant  quasars are typically measured from the breadth 11:58.880 --> 12:04.640 of their emission lines because these indicate  orbital velocities around black holes. But what 12:04.640 --> 12:10.960 if the emission line breadth is NOT due to orbital  motion? In the case of such a densely cocooned 12:10.960 --> 12:18.640 source, photons trying to escape bounce many, many  times on their way out. That scattering blurs and 12:18.640 --> 12:24.000 broadens the usually sharp spike of an emission  line. The authors of this study explored how much 12:24.000 --> 12:29.120 broader a broad emission line might become  in this sort of scenario. They figure that 12:29.120 --> 12:37.200 the velocity could be over-estimated by a factor  of 5, so a 600km/s orbital velocity gets blurred 12:37.200 --> 12:45.040 out so it looks like 3000km/s. And because the  inferred black hole mass goes as velocity squared, 12:45.040 --> 12:51.280 a million solar mass black hole can look  like a 25 million solar mass monster. 12:51.280 --> 12:56.240 Now, a million suns worth of black hole is still  technically supermassive, but it’s borderline. 12:56.240 --> 13:02.640 It’s much more reasonable for such an early  cosmic age and such a small host galaxy. They are 13:02.640 --> 13:07.600 black holes that, at the rate of super-Eddington  accretion we also might be seeing, can sensibly 13:07.600 --> 13:13.840 give us the colossal early supermassive black  holes that have been vexing us for decades now. 13:13.840 --> 13:18.720 I should also add that this interpretation  of the data is just that: an interpretation. 13:18.720 --> 13:23.680 We’ll need more observations and to find  more objects like this to corroborate this 13:23.680 --> 13:28.720 story of the black hole star as part of the  answer to the problem of gigantic early black 13:28.720 --> 13:37.276 holes. But it is a step further in understanding the  strange state at the beginning of spacetime. 13:28.720 --> 13:33.760 story of the black hole star as part of the  answer to the problem of gigantic early black 13:28.720 --> 13:37.276 holes. But it is a step further in understanding the  strange state at the beginning of spacetime.