By “second-wave planets” I mean a planet forming long after the protoplanetary phase of the star system (it’s not the technical term, and I’m not even entirely sure there IS one). Can that happen? It’s entirely possible, and astronomers may have now seen evidence of that.
I normally deal with young stars, whose planet formation capabilities are well-known (if not exactly well-understood in detail). The protoplanetary phase goes something like this:
- Stars start surrounded by a cloud of gas
- that gas breaks apart into multiple collapsing cloudlets
- a self-gravitating spherical mass (or two, or five) starts forming in the center gathering mass as material spirals in
- angular momentum starts producing a disk of material around the rapidly heating protostar; its magnetic field catches some material and forms jets coming off the poles.
- the protostar heats up enough that the winds coming off of the star blow away the gas and some of the dust OR the gas and dust just run out.
- the resulting solar system objects collide with each other until everything that would collide, does.
Somewhere between steps 3 and 5, planets are formed in orbit of the star (directly from the nebula? gas eddies in the disk? dust collisions collecting material? All of the above? Nobody knows). In any case, it’s the presence of gas and dust – raw materials in the disk – that allow them to form.
Stars end up surrounded by dust again toward the end of their lives. This is the point where the cores of stars are running out of hydrogen, so they start collapsing until helium fusion starts (with a hydrogen shell also fusing – think layers, like one of those novelty jawbreakers). As the core contracts and heats up massively, the outer layers start floating off into space; and they form a dusty disk of OUTBOUND material, rather than inbound.
Now astronomers have seen one of these disks. This was a difficult thing to do because old dying stars of this type are kind of rare, and not gathered into obvious clumps like the nearby young stars are. The one studied in this particular paper, IRAS08544-4431, is 4,000 light years away in the constellation Vela. To get the images, the astronomers have used interferometry on the European Southern Observatory’s Very Large Telescopes, which allowed them to combine the mirrors of multiple large telescopes to resolve fine details as if they had a 150+ meter wide mirror (the process of interferometry requires an enormous amount of precision mechanical and optical engineering, but nothing like the difficulty of actually building a 150-meter-wide mirror). The particular wavelength of light they were looking at was specifically good at finding the emissions from the carbon monoxide gas mixed in with the dust, so what stood out in the image was the ring of dust around the star that had been puffed out in some earlier episode, and the star itself (because stars emit at all wavelengths in the optical and infrared), AND the dying star’s smaller K-type companion, which appears to have gained a new accretion disk courtesy of the material sloughed off the other star.
The actual paper ends there, but in the various press releases, the authors speculate that all this dust may represent the right conditions for planets to form… again. Second-wave planets, if you will. As I mentioned above, the exact mechanism(s) by which planets form around young stars are still up for debate, but the authors make the argument that the ring they’ve seen is qualitatively similar to the rings seen around young stars. Dust and gas surrounding the smaller K type star may be enough to form planets again.
Technically speaking, we’ve already seen second-wave planets. In 1992, Alexander Wolszczan and Dave Frail conducted incredibly accurate studies of the pulsar PSR B1257+12, and found two (eventually three) planet-mass objects orbiting the star: one that’s half the mass of Mercury, two that are four times the mass of the Earth. It’s considered doubtful that these objects survived the catastrophic destruction of the supernova explosion at such close range, so they probably formed afterwards. In this case, it’s thought that the planets are actually the shredded remains of a star that once orbited the now-pulsar. That’s an entirely different method of planet formation than what is being proposed here, though. The process by which stars become white dwarfs is downright gentle in comparison. On the whole, this extra round of star formation would be more like regular young-star star formation, except that the material is from the dying star, and not a nebula.
There may be a better example of what we’re seeing here. The bright dying star in the IRAS08544-4431 system will eventually become a white dwarf. We already know of planets in systems that include a white dwarf – the GJ 86 system includes a K-type star (A) orbited by a planet (Ab), and a white dwarf (B) that USED to be the brighter star of the pair. If this second-wave planet formation theory is true, it might be that GJ 86 A got two chances to form planets, and the planet might be a more recent addition (I don’t know how you’d test that, though). I’ll admit I’m not a theorist, so there may (and probably are) other complications to this that would rule out planets formed from stellar ejecta, but it’s an intriguing idea.