We often speak about planets migrating from the outer to the inner system of a star, something that helps us put ‘hot Jupiters’ in context. But what about migration within the galactic disk? It’s an idea under continuing investigation. In the absence of direct observational evidence, we infer migration and assume that older stars often come from regions with significantly different metallicity than stars in their current environment. The presumed origin would be the inner disk, which Misha Haywood defines as that part of the galaxy inside the radius from galactic center to our Sun.
Dave Moore sent me Haywood’s latest paper a few months back and I’ve been slow in getting to it because I wanted to give its conclusions further thought. It’s intriguing stuff. Haywood (Observatoire de Paris) takes note of the fact that we tend to find gas giants around stars that are rich in metals (here a pause to remind newcomers that by ‘metals,’ we mean elements higher than helium). And he wants to answer a key question: How do we know that this higher percentage of Jovian worlds detected around metal-rich stars is the result of metallicity, and not some other factor linked with their origin in the inner disk? The question is relevant, Haywood writes:
…because any measurable property of inner disk stars other than metallicity would be correlated with the presence of planet. The obvious a priori response is that metallicity is a measurable parameter, and intrinsic to the star. But there could be others however, which, although not measurable on the stars, could be no less important, such as, for example, the surface density of molecular hydrogen in the inner galactic disk regions.
The obvious next step is to find exceptions to the giant planet/metallicity correlation, and Haywood notes that we don’t see the same metallicity connection among giant stars hosting planets that we do around smaller stars. Moreover, at intermediate metallicities, giant planets seem to favor thick disk stars rather than thin disk objects.
Here we’re talking about different and distinct star populations. The ‘thin disk’ we see edge-on in images of spiral galaxies is complemented by the more diffuse ‘thick disk,’ containing older stars. The thick disk population, thinks Haywood, comes from migrating stars from the inner disk, while the metal-poor group derives from stars from the outer disk.
This takes us to an interesting place:
We are now facing the following picture: stars that come from the inner disk are noticeably rich in giant planets, while stars that come from the outer disk seems to be less favored in this respect. This new information changes considerably how we envisage the correlation between metallicity and the presence of giant planets. For the surprising point here is not the fact that most host-planet stars are metal-rich, since they come from a region where most stars are metal-rich, but the very fact that most would come from the inner disk. We are led to conclude that the distance to the galactic center must somehow play a role in setting the percentage of giant planets…
The italics above are mine, because the statement is the core of the argument. Looked at from this perspective, the correlation between metals and the presence of giant planets turns out to reflect the galactic origin of the stars. It does not imply that metallicity is the necessary cause for the formation of these planets.
But if not metallicity, what other factors can we link to the galactocentric distance of a star? One possibility is dust density, which would favor the development of planetesimals. But Haywood prefers molecular hydrogen as the answer. It is the basic ingredient for the formation of giant planets, the principal consituent of stellar disks. Moreover, we have to think in terms of where it is most abundant:
Its main structure in the Galaxy, the molecular ring, is thought to contain 70% of H2 gas inside the solar circle…, thereby providing a huge reservoir for star (H2 is known to be directly linked to star formation…) and planet formation. The most interesting aspect however, is the fact the molecular ring reaches a maximum density at 3-5 kpc from the sun, corresponding to the distance where stars with metallicity in the range (+0.3,+0.5) dex are expected to be formed preferentially.
Haywood argues that stars hosting ‘super Earths’ or Neptune-class worlds with no accompanying gas giants are less likely to have had an origin in the inner disk, and thus form in an environment less dense in molecular hydrogen. We would, then, expect no predominance of metal-rich stars among this population. Surveying twelve systems that house super-Earths or Neptune-class planets, Haywood finds that the seven with no Jovian planets have low metallicity, fitting his theory, while the five that do contain gas giants indeed show a higher proportion of metals, “…amply confirming the possibility that the first group of stars could be genuine solar radius objects, and the second wanderers from inside the Galaxy.”
The paper is Haywood, “On the Correlation Between Metallicity and the Presence of Giant Planets,” accepted at Astrophysical Journal Letters and available as a preprint.