An intermediate-mass black hole might be lurking within a dense stellar cluster — a discovery that could point toward how these oddities form.

Computer generated image of a black hole causing distortions in the surrounding light. Alain Riazuelo
Computer-generated image of a black hole.
Alain Riazuelo

Every Friday Bülent Kızıltan heads to a local coffee shop. There, Kızıltan (Harvard-Smithsonian Center for Astrophysics), who is a self-described “overly curious person,” meets with scientists, artists and philosophers from different disciplines across the Cambridge and Boston area. His goal is to enrich himself intellectually.

That’s how he found himself deep in conversation with a cognitive scientist one afternoon who introduced him to a mathematical tool called Kullback–Leibler divergence. Originally developed by two mathematicians and cryptanalysts at the National Security Agency in the 1950s, the tool attempts to extract information from incomplete data.

Perhaps it was the stimulating effects of the freshly brewed cup of joe, but Kızıltan quickly realized that this tool could also be used when observing a dense stellar cluster — a messy environment chockfull of hundreds of thousands of stars where the center is hidden from view. It was the first step toward a paper published today in the journal Nature that provides the best evidence yet that a globular cluster might harbor an intermediate-mass black hole.

"Weighing" a Hidden Black Hole

Few intermediate-mass black holes have been observed before. In fact, skeptics might argue that none of these black holes — which theoretically lie in the range between those that are just a few times the mass of our Sun and those that weigh millions to billions of times the Sun — even exist.

The issue is two-fold. First, there’s no obvious, direct way for these objects to form. Second, the evidence that they exist has been hazy at best.

But with this new mathematical tool in hand, Kızıltan knew he could do better. He set his eyes on the globular cluster 47 Tucanae, located roughly 15,000 light-years away in the southern constellation Tucana. The cluster, which appears roughly the size of the full Moon, is one of the most massive globular clusters in the galaxy, with millions of stars.

47 Tucanae
The globular cluster 47 Tucanae is the second brightest in the sky and contains tens of thousands of stars — and possibly an intermediate-mass black hole.
NASA / ESA /Hubble Heritage (STScI/AURA) / Hubble Collaboration. Acknowledgment: J. Mack (STScI) and G. Piotto (Univ. of Padova, Italy)

In order to probe those stars, Kızıltan and his colleagues first built a realistic model of 47 Tucanae. Then, they measured the locations of 25 pulsars throughout the globular cluster. These rapidly rotating neutron stars sweep electromagnetic radiation across Earth with every spin at a rate so reliable they rival atomic clocks. This (plus the fact that they emit radio waves, as opposed to visible light) enabled the team to probe deeper into the typically invisible depths of the globular cluster.

The pulsars, they found, are on the move. They’re spinning around the center of the cluster rapidly — too rapidly given the number of central stars alone. Using their new mathematical tool, Kızıltan and his colleagues compared those pulsars’ oddly quick movements with their simulation in order to determine that an intermediate-mass black hole, some 2,200 times the mass of the Sun, must be lurking in the cluster’s center, governing the pulsars’ motions.

Runaway Collisions or Dangerous Tangos?

But Cole Miller (University of Maryland), who was not involved in the study, doesn’t think that the new evidence will convince every skeptic.

“Most scientists would like to see strong evidence,” he says. “Because if these objects exist, they will have implications for the formation of the earliest massive black holes. They will have implications about stellar dynamics. And they will have implications about the evolution of very massive stars. All of this is frontier research and you don't want to be jumping on the band wagon too early.”

The main issue is that astronomers are at a loss to explain how these intermediate-mass black holes can form. Miller’s favorite hypothesis enlists runaway stellar collisions. Essentially, the heaviest stars sink to the center of the globular cluster, where they are most likely to collide and merge with other heavy stars, forming even bigger stars. And because a star’s gravitational pull increases with its size and mass, so does its chance of further stellar collisions. This effect snowballs until a massive star rests in the center of the globular cluster.

But how this massive star evolves is not fully understood, says Kayhan Gültekin (University of Michigan) who was not involved in the study. “Standard stellar theory doesn’t apply to these objects. A lot of turning this merged star-like object into a black hole is based on analogies with stellar evolution rather than detailed calculations of how it would actually evolve.”

An artist's concept shows two merging black holes. NASA
An artist's concept shows two merging black holes.

Luckily, the competing theory doesn’t need to invoke strange astrophysics. Here, stars evolve into black holes, which sink to the center of the cluster where one enters into a binary orbit with another. Through interactions with other black holes, that duo shrinks until it merges, creating a more massive black hole. This single black hole picks up another companion and the process begins anew, until there’s only one massive black hole in the center of the globular cluster.

Unfortunately, there are issues with this theory too. In the interactions between the black hole binary and the third black hole, energy is transferred to the third black hole. And if enough energy is transferred, the third black hole could get kicked out of the system entirely — a huge problem when you’re trying to build a massive black hole.

Although there are other theories that could explain how intermediate-mass black holes arise, only these two rely on the densely packed environment of a globular cluster.

“The fact that this black hole exists in 47 Tucanae, might mean that one or both of those theories is correct,” says Gültekin, who is hopeful that both theories will receive rejuvenated attention in light of the latest discovery.

B. Kızıltan et al. “An Intermediate-Mass Black Hole in the Centre of the Globular Cluster 47 Tucanae.”Nature, February 9, 2017.


Image of Randy Pfeiffer

Randy Pfeiffer

February 13, 2017 at 1:01 pm

All supermassive black holes must have evolved from conditions with no black hole. Along the way, their masses would be in the intermediate range. Hopefully, this potential discovery leads to development of mathematical tools capable of representing all stages of BH evolution in any setting.

If supermassive BHs accrete through a different process than proposed for intermediate-mass BHs, it would be helpful to see an article comparing the two processes. Without knowing the math, it seems natural that supermassive BHs should evolve through an intermediate-mass stage similar to what's described in this finding.

You must be logged in to post a comment.

Image of StanR


February 14, 2017 at 11:44 pm

The above link to Kullback-Leibler divergence did not work.
This seems to:

You must be logged in to post a comment.

Image of kiparsky


July 3, 2017 at 1:17 pm

Where is this coffee shop? They are clearly making a very effective cup of coffee!

You must be logged in to post a comment.

Image of DC0


November 10, 2017 at 2:33 pm

Why a black star is the alternative to a black hole.
How does the Chandrasekhar limit prevent the creation of intermediate black stars?

Very little is known about the formation of black holes. An article, recently published in the Journal of High Energy Physics, Gravitation and Cosmology, in which I discuss a computer model that poses an alternative process to the accepted theory of the formation. It can be found at .
Due to a unique requirement of creating black holes by freezing time and space from the inside out, the conventional method of deriving results from general relativity could not be used. Instead I used the Newtonian model, while factoring in relativistic corrections derived from general relativity which includes relative contraction of both space and time. This mathematical model, written in Excel and Visual Basic, takes about a day to run. It calculates time dilation profiles, density cross sections and red shift factors as it forms a “black star”. I use the term black star because the calculations show that during the collapse no singularity is formed and there is no event horizon as in the accepted theory of black holes.
When a white dwarf or the remnant core of a supernova gets to a size that is just above the Chandrasekhar limit (1.44 solar masses), the pressure will overcome the electron degeneracy pressure. This happens at a density of 1x10^9 kg/m^3. It will then collapse down to a neutron star resulting in densities starting at 3.5 x 10^15 kg/m^3. The associated pressure at this density is just high enough to support neutron matter. During this contraction, the decreasing gravitational potential caused time to relatively slow.
For a remnant greater than 2.2 solar masses, while it is still contracting, this gravitational potential causes time at the center to relatively freeze and stop the contraction before the pressure gets high enough to stop it, as it would in a neutron star. Then, above this frozen matter, as the contraction continues, the time freeze moves toward the surface, stopping the contraction of the rest of the remnant, creating a black star.
The creation of black stars, using this model, lead to the discovery of why we do not have intermediate black stars. During the collapse of a white dwarf, it transitions from degenerate white dwarf matter, at a density of 1x10^9 kg/m^3, to neutron matter which starts at a density of 3.5x 10^15 kg/m^3. This matter has just crossed a span of densities that cannot be supported. The gap between these two densities relate to the answers of questions like:
1. Why is there such a large size gap between stellar black stars and super massive black stars?
2. Why are there no stellar black stars below 2 solar masses?
3. Why are there no supernova created stellar black stars above 15 solar masses?
4. Why does the smallest super massive black star start at 50,000 solar masses?
5. Are super massive black stars made before or after the existence of first-generation stars in a galaxy?
These questions cannot be answered using a model of black holes that have a singularity. They are answered by this model of the formation of black stars. The information produced by this model agrees with observation when available.

You must be logged in to post a comment.

Image of DC0


November 13, 2017 at 4:02 pm

Calculations in my above-mentioned article, show that black stars (black holes) cannot exist between about 50 and 50,000 solar masses. I would define the nonexistent intermediate black stars as being between these two limits. This range of black star sizes requires densities of matter that cannot exist. These densities are the same as those found when matter transitions from degenerate white dwarf matter to neutron matter. I think that if black stars are suspected in this intermediate range, another explanation that does not require them, needs to be found.

You must be logged in to post a comment.

You must be logged in to post a comment.