Making planets isn't easy*.
Stars form in the centers of disks of material that can stretch for many tens of billions of kilometers. The stuff from the disk falling into the center forms the star, but farther out where it's cooler is where planets tend to form. Usually, for stars this young planetesimals — chunks of rock and metal and ice from 1–100 km in size — likely dominate, the precursors to actual planets.
But a lot can go wrong. Too close to a star and it gets cooked. Too far and there may not be enough material to gather to create an actual planet.
The neighborhood matters, too. Most stars form in open clusters, some with tens of thousands or more stars. Astronomers have thought for some time that just where the star is located in the cluster makes a big difference in whether it can form planets or not, and that conjecture now has new support from a series of observations of a huge, relatively nearby cluster.
YeGADS. Pretty, ain't it?
Westerlund 2 is about 13,500 light years away, and huge; the total mass of stars in it is about 36,000 times the Sun's! Most stars that form are lower mass than the Sun, so this means there could be over 100,000 stars being born there. They're divided into two distinct clumps, which you can see in the Hubble image above; one is smaller and above the larger one. Careful analysis has shown the two clumps are about the same age, between 0.5 and 2 million years old. So these are very young stars indeed.
To the lower left you can see a mix of glowing gas and darker dust, the basic building material of stars. The dust clouds are elongated, pointed toward the middle of the cluster. That's real, and due to the most massive stars in the cluster. Some of those stars, called O and B stars, are beasts, huge and luminous. There's a binary in there called WR20a with two stars each about 80 times the mass of the Sun orbiting each other; together they blast out over a million times the energy the Sun does! Yikes.
These stars emit a lot of high-energy ultraviolet light, which eats away at the dust in a process called photoevaporation. Basically, those stars are eroding the material around them (note the central part near the stars is mostly clear of gas and dust).
And that, it turns out, is central to our story.
Using the Hubble Space Telescope, astronomers observed the stars in the cluster for three years, looking for changes in their brightness. Very young stars like these are known to be highly variable, and measuring that brightness change can tell you a lot about the stars, including how fast they rotate, whether they have starspots (like sunspots), and how they grow, and also give physical characteristics of the material around them.
The astronomers looked at a total of 9,268 stars in Westerlund 2, ranging in mass from dinky red dwarfs with 1/10th the mass of the Sun up to very luminous stars almost 6 times the Sun's mass. They found that 30% of those stars are variable for various reasons.
Of those, 5% are what they call dippers — the light from the star suddenly dips, then goes back up again. These are very likely from stars with disks we see very nearly edge-on, and disk material is circling around and blocking some of the starlight from our point of view.
In other words, these are the stars that could be forming planets even as we watch.
And this is where things get very interesting indeed: When they plotted where these stars were, they tend to be in the outskirts of the cluster, avoiding the central regions. In fact, they don't seem to exist in the center of either clump making up the overall cluster!
Why? The obvious explanation is, again, photoevaporation. The massive stars tend to be in the centers of the clumps (close encounters between stars are common, and when that happens the more massive stars tend to sink to the cluster center, and the lower mass ones move outward; this process is called mass segregation). So the center of the cluster is where most of that lethal UV comes from, and, apparently, lower mass stars that form there get their planet-forming disks destroyed by that light.
Well, you know the real estate cliché: location, location, location not near massive OB stars in a dense cluster or else your protoplanetary disk will get vaporized by a flood of ultraviolet light.
Most real estate agents shorten the phrase though. Makes sense.
Our Sun has planets (I leave the proof as a logic exercise for the reader), so if it too formed in a cluster maybe it was in the suburbs, or it was in a smaller cluster with fewer massive stars. We don't know, but we're learning more about how this works all the time.
Westerlund 2 is a fascinating place, and a spectacular one, too. But given what lurks at its heart(s), I'm glad it's so far away.
*Just ask Carol Marcus.