I read (what I could understand of) this article in which authors Yuan and Steinle-Neumann "...use advanced quantum mechanical simulations on silicate and metallic melts showing that hydrogen becomes increasingly more incorporated in metal over silicate at high pressure and temperature, conditions under which Earth's core formed. Therefore, hydrogen can be present in the core in high abundance. [...] The Earth's early accretion events, in particular magma ocean formation, had a great impact on the chemical and thermal evolution of the Earth. Segregation of metal from silicate to form a core during the magma ocean stage has removed elements besides iron and nickel from the mantle[...]. Light elements such as H, C, O, and Si enter the core as a consequence of this process." The article goes on to present a model for how light elements near the planet's surface could end up being transported to the core, leaving a mostly silicate crust with both light elements and metals mostly transported down to the core: light elements form weak bonds with iron under high pressures and are gravitationally sorted with the iron. "We find that hydrogen is weakly siderophile at low pressure (20 GPa and 2,500 K), and becomes much more strongly so with pressure, suggesting that hydrogen is transported to the core in a significant amount during core segregation and is stable there."


How did diffuse hydrogen in the near-vacuum of the planetary accretion disk find itself bound, chemically or gravitationally, to the forming Earth in the first place?

As I understand it, the mechanism proposed for incorporation into the core necessitates extreme pressures and temperatures - which is good for explaining how it got from deep in the molten surface down to the core, but doesn't seem to fit planet formation in the first place. Why didn't the hydrogen swept out by proto-Earth simply boil off of the upper atmosphere as fast as it came in, never finding itself in a pressure regime sufficient to become chemically bound to iron?

It seems to me that intermediate chemical bonding of hydrogen into a less volatile form could explain it. I've read elsewhere that very little of Earth's water was swept out during accretion, but perhaps silanes and alkanes or other hydrogen-rich compounds might have formed in the accretion disk, been swept out, and then been broken up and re-bound with other chemicals in the extreme environment of the accreting planet. Or there might have been enough free Si, C, and other chemicals with an affinity for hydrogen in the atmosphere or surface of proto-Earth to bond with hydrogen, and then break up when convected down into higher temperature and pressure regimes. My chemistry knowledge is pretty weak, so I don't know if either explanation is plausible.


1 Answer 1


The article you've shared doesn't assume the hydrogen in the core came from the hydrogen in the protosolar nebula.
In fact it ignores all modern literature on planet formation, and just cites lab experiments and molecular dynamics simulations on the topic of hydrogen partitioning between a Si & O-rich phase and a Fe-rich phase. The hydrogen in their scenario is assumed to come from water in the mantle, as you see in their fig. 1, there is barely any to start with.
The amount of hydrogen that they assume is therefore generally consistent with a hydrogen-poor (relative to solar) initial condition for the mantle: If we look at where we have to start out in the protosolar nebula, we land at a hydrogen-rich $\rm [H/O]\approx 2\times10^{3}$ (usually taken from the solar atmosphere as proxy).
In their simulations however, they start with a 'rocky' composition in one phase, i.e. olivine-type rocks somewhere along $\rm Fe_x Mg_{1-x}Si_yO_z$ and essentially pure iron in the other phase. They have 12 H atoms diffusing around for 377 atoms total, so that gives a total H content in the simulation domain of 3%. Their quoted result in the abstract of 1% in the metal-rich phase then isn't all that exciting, I guess?

But let us contain a key result here, that the ratio of hydrogen to oxygen of course would be very different now to the protosolar nebula, i.e. $[12/100]\sim1$%, where this low number compared to the protosolar value that you might have expected, is coming from the assumption of initial water content of the rock.
Note that I am comparing atom numbers here, where their 1% is a weight-percent.

If you are interested, I can add a bit of discussion on how volatiles might get to the planet in the first place during the accretion process, but that would happen in the next few days.

  • $\begingroup$ I would definitely like that. Can you @me in a comment when you update, so I get a notification? $\endgroup$
    – g s
    Apr 30 at 0:27
  • 1
    $\begingroup$ @gs You can get notifications about edits yourself by clicking the "Follow" link at the bottom of the post. $\endgroup$ Apr 30 at 14:57
  • $\begingroup$ Feels like you're ready to submit a comment reply to GRL! $\endgroup$
    – Gimelist
    May 15 at 11:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.