Planets are born around young stars in formation. They arise from the rotating discs of remains of material, which are left over after the birth of the star itself. Then, the surroundings of the star, already with their planets emerging, are “cleaned” of diffuse material and stay relatively clear. Until a few years ago we thought that discs were exclusive to those early stellar stages, but then it was discovered that no, that stars in advanced phases again had discs of material around them. And that’s where the question arises: could there be second-generation planets?
When sun-like stars run out the fuel of their nuclei, they start a decline consisting of various stages. One of them is the red giant phase, in which it swells considerably and begins to expel its material in the form of layers, as in a slow wave of gas molecules and dust grains. This is where stellar winds push that material out. The star continues to “get rid” of its layers, reaching the highest mass loss phase, the Asymptotic Giant Branch, or AGB.
Many stars that have already crossed this stage (called post-AGB) in binary systems (pairs of stars orbiting each other) have a disk made up of gas and dust that revolves around both stars. We know that they exist, but we ignore the details of their formation, structure and evolution, although surprising similarities have been found with discs that revolve around young stars. 
There is research that considers the possibility that these gravitationally linked dust discs exist in many binary stars in advanced stages. In fact, the data is even used the other way around: the presence of a disk indicates that it can be a binary system with a post-AGB star.
The Red Rectangle
In 2003, a team led by Valentín Bujarrabal, a researcher at the National Astronomical Observatory (OAN-IGN), discovered that the planetary nebula known as the Red Rectangle had a rotating disc . This nebula, studied with the IRAM interferometer, also launches jets of material at low speed and has a complex structure in which there is a binary star system whose main star is a post-AGB.
Until early 2015, only this rotating disc had been clearly studied and identified. The second disc of this detected type was the one orbiting around other evolved star: AC Herculis. From this moment on, scientists suspected that these discs play a key role in late stellar evolution and were abundant around evolved stars.
Bujarrabal, principal investigator of both works, states that these, and other published results, are part of a long collaboration maintained by the OAN team with the Institute of Astronomy of Leuven: “We were the first to demonstrate the existence of discs rotating around old stars, as normally the material around them, which has been ejected by them, is expanding. Some of our latest observations, particularly using ALMA and VLTI, are really spectacular and contain a huge amount of information about these amazing objects.“
The data that ALMA is providing will be decisive in the future to compare this type of disc with those around young stars. At the moment, thanks to the VLTI interferometer, the most accurate image of a disc around an evolved star, IRAS 08544-4431, formed by a red giant and a less evolved one, has been obtained.
The image is impressive: the dust ring surrounding the stars is clearly visible. From all these observations it has been inferred that the discs surrounding old stars are very similar to the discs around the young stars. And if they’re so similar, could planets form?
Wait a minute, let’s rewind.
Born and die
We are at the moment just before the stars are born (our protagonists are always mid-sized stars). The molecular cloud, laden with gas and dust, is compacted at some points where matter ends up condensing and ignition begins in the stellar nuclei. Around them, these young stars have discs on which planets can eventually form. Once formed, there may even be a “dance” of planets, called planetary migration, that causes some to change orbit around their star until the system stabilizes.
The stars normally live their hydrogen consumption stage in the nucleus, until it is finished and drift begins. The red giant phase is so overwhelming (because of its huge size and impressive increase in luminosity) that the planets will most likely end up slashed, pushed or broken. It will all depend on the distance that separates the planet from its star and, again, the planetary migrations between orbits.
Of course, planets could overcome that phase and still exist (very battered, yes). And this has led to quite a few confusions when it came to determining whether a star (as in BP Piscium’s case) was young or old – the same thing happened with Gomez’s Hamburguer, which was believed to be an old star until it was studied in depth and was seen to be a young star that probably has a protoplanet orbiting around it.
But would a second generation of planets be possible? What specific conditions should be given? Could the remains of first-generation planets, along with the materials of the second disc, form new planets? How long would it take to do so? Could the energy of the dying white dwarf left in the center feed that system?
As for the possible existence of second generation planets, there are already studies related to the first exoplanets detected, discovered in 1992, which also revolved around a pulsar: PSR B1257+12. But in this case, we are not talking about the death of sun-like stars (between one and eight solar masses) but a hypothesis about something that could have happened after the death as a supernova of a much more massive star .
Moreover, in the environments of evolved mid-mass stars, in the final stages of their lives, so far, no planets have been found in formation. Scientists theorize about which environment would be most appropriate, which parameters they should meet, but nothing has yet been detected that can confirm these hypotheses.
Meanwhile, researchers like Bujarrabal continue to study these second-generation discs to determine how they form and what their final destination is.
 Valentin Bujarrabal himself published a beautiful report entitled “The Colloquium of the Nebulae”, inspired by Cervantes’ “The Colloquium of Dogs”, in which the Rectangle Nebula and the Pumpkin Nebula have a dialectical encounter. You can find it in the yearbook of the National Astronomical Observatory (OAN-IGN) of 2014 and in this link (in Spanish).
 Some argue that a millisecond pulsar could explode in the form of a Quark-type nova. This still hypothetical object, the result of a previous explosion of supernova, would generate a disk capable of forming planets. This theory could explain the existence of the planets around the PSR B1257+12 pulsar.
Originally published in Spanish on the Naukas website: “¿Planetas de segunda generación?” (2016/04/26).