Terrestrial planet atmospheres may not form as we thought

The development of life depends on a planet's atmosphere. We show that planets are able to acquire carbon, oxygen, and nitrogen from second generation discs produced by comet and asteroid belts. Such discs are now observable thanks to the Atacama Large Millimeter Array.

Like Comment
Read the paper

I have been working on gas surrounding mature main sequence stars for quite a while now. This whole work started when I realised this could be important for creating the atmospheres of exoplanets; but had been completely neglected in previous studies. What if the planets capture most of the gas, what would happen to their atmospheres? I started some back-of-the-envelope calculations and realised that the effect may be very significant for small terrestrial-like planets and may still be visible in giant atmosphere spectra and/or composition. Now I needed to get a model of how gas is captured by planets and there was already quite a lot of literature about that for planets embedded in most massive younger protoplanetary discs (this is where planets form initially at <10Myr). Late gas discs (>10 Myr) are less massive and I had to adapt the model to account for that as well as for other subtleties. The result is striking, planets embedded in these ubiquitous long-lasting late discs will capture a lot of gas, which is going to affect their primordial atmospheres and may lead to final atmospheres that are dominated by gas coming from late discs rather than outgassing or exocomet impacting the planets. This is the short answer to the problem but things are indeed more complicated.

We have now discovered more than 4000 exoplanets, which are planets orbiting other stars than our Sun. We have started to characterise their atmospheres (for about 50 of them so far) thanks to the most powerful telescopes, which revealed that they vary widely. It is commonly believed that the origin of this diversity comes from different formation mechanisms and system histories. Exoplanet atmospheres are a mix of different gases, those present at an early stage when the planet formed, those ejected later from their interiors through volcanism, and those that came from cometary impacts.

The new work we presented in our Nature Astronomy paper shows that there is a fourth fundamental process that participates in forming atmospheres, which may dominate all others: the late accretion of gas. It is thanks to the ALMA array of telescopes observing in the sub-millimetre that gas was detected around stars well after planets had formed in their initial cocoon called protoplanetary disc. The planets will be embedded in this late gas and capture it for tens to hundreds of millions of years. This gas is probably released when planetesimals (large rocks dozens of kilometres in diameter) that orbit at the outer edge of planetary systems collide together. The released gas, carbon monoxide, and probably some water as well as more complex molecules will evolve towards the interior of the system, crossing the planets orbits, which can then capture this gas.

The process described above is very effective and allows a planet to accumulate an atmosphere up to several millions times the mass of the Earth’s, which is large enough to form Earth-like or Venus-like atmospheres as well as thicker atmospheres as those of mini-Neptunes (with a size in between Earth and Neptune and that are mainly observed close to their host star so far, i.e. much closer than our Neptune). Volatile species such as carbon and nitrogen can only be released at large distances in these late gas discs and are later delivered to closer-in planets while gas spreads inwards. Bringing these volatiles from large distances is rather unique and could potentially be favourable to the development of the first building blocks of life. These volatiles are indeed essential for a planet’s habitability (e.g. presence of liquid water and of a long-term temperate climate thanks to CO2 or some other greenhouse gases) and favours the emergence of life from prebiotic chemistry based on these volatiles.

 Moreover, we find that this late gas accretion is so efficient that it can also affect the composition of giant planets (similar to Jupiter in our Solar System) by adding heavier molecules to their hydrogen-rich atmospheres (i.e., it increases their metallicity). It could also change the carbon-to-oxygen ratio in these atmospheres, which is a fundamental ratio used to characterise atmospheres and how they formed. Giant planets will provide the crucial data to confirm our scenario and it will be done thanks to the new James Webb Space Telescope and also from the ground with the upcoming Extremely Large Telescope.  

Last but not least, we find that our work could lead to a new way of detecting planets around old systems. Indeed, late gas accretion being very efficient, there should be a deep gap or even a cavity beyond a planet (similar to the gaps observed in much younger protoplanetary discs with ALMA) as it will capture most of the incoming gas and astronomers could see holes in the gas discs in between the accreting planet and its host star. Using the high resolution of ALMA could allow to see these holes (there are already some hints from low-resolution images) and indirectly detect planets at locations (1-100 au) that other detection methods (transit, radial velocity) cannot reach. Even a low mass planet such as an Earth may potentially create these cavities, which may allow us to discover old Earth-like planets thanks to these discs (contrary to young and massive planets that can be seen in protoplanetary discs).

This work shows that the last steps to form planetary atmospheres are essential and should not be neglected anymore as it may be the key to deliver a lot of carbon to atmospheres, which may be critical to the development of life.


Terrestrial planet accreting volatiles from a late gas disc
Terrestrial planet accreting volatiles from a late gas disc (Credits: Sylvain Cnudde / Paris Observatory / LESIA)


Kral Quentin

Astronomer, Paris Observatory

No comments yet.