How do Rocky Planets Form?

In my previous post, I argued that no simple physical process leads to the rocky planets having an atmosphere. Something was missing, and to explain what, I need to discuss how rocky planets formed in the first place. The standard theory starts with a distribution of planetesimals through space, and these gravitationally accrete. For me there are several things wrong with this theory. The first is there is no known mechanism to get the planetesimals, which can be considered as medium-sized asteroids, say something like 50 km radius, although they would not usually be round. Various theories as to how dust could form planetesimals invariably fail, the problem being if you can persuade dust to accumulate, abrasion turns it back to dust because it has no strength. If you can get to rocks, then these are argued to attract and form rubble piles, but if two such piles collide, what happens next will be like a snooker break. A pile of stones does not join together, let alone a mass of sand, without some other agent.

The next problem is the time taken to form the planets. Early computations on accretion, omitting any degradation, had this as about 100 million years (100 My). When it became obvious that this was wrong (e.g. the moon probably formed after 30 – 60 My) the time got reduced, although nobody said how. We are now reasonably confident that Mars formed within three million years, and no collisional theory as yet can get it formed that quickly. One problem is the small bodies that form initially are entrained in gas and hence in circular orbits, when they never meet. If we assume a physical process, all accretion is at the same rate, leaving aside concentration as a function of distance. As the bodies grow, to collide they need some degree of eccentricity, and the greater the eccentricity, the more violent the collisions. There are examples of such collisions that have occurred in the asteroid belt; almost inevitably the net result is fragmentation, not growth, and the so-called asteroid families remain.

There are also some questions to answer, such as why is Mars so small? Why is there so little between Mars and Jupiter? If you argue that Jupiter upset the mass in this zone, why are the asteroids in almost circular orbits, apart from the “families”?

So, if there is no obvious physical mechanism to start rocky planet accretion, as a chemist I had to think in terms of chemistry. To start, there are three relevant time periods. The first is during stellar accretion, when mass pours into the forming star. The disk temperature now falls off from the star according to about r^-0.75, with temperatures around Earth reaching about 1600 degrees C. Study of such disks elsewhere suggest this stage takes about 1 My. Then the rate of inflow from the disk more or less decreases by about three to seven orders of magnitude, and the disk cools to temperatures similar to those today, with the main source of energy being radiant heat from the star. This second stage can last somewhere between 1 – 30 My. Eventually the star ignites its fusion reactions and there is a significant outflowing of mass, which clears away the accretion disk and any small dust. At this time, there is also a radial distribution of certain isotopes, thus the amount of 17 and 18 oxygen will vary by a few ppm radially, which gives us clues as to where various rocks, etc came from.

The conclusion is, most of the material on any given body in the solar system accreted from material from its own zone, i.e. there was none of the mixing expected from standard collisional theory. The Moon accreted from material in Earth’s zone, or from Earth material. Of course there are problems with such a statement, as we only have samples from Earth, Moon, Mars and some asteroids. So, what are the options remaining for accretion?

The first is if a body can get big enough very quickly. The early solar system had a certain level of 26Al, an isotope that decays with a half-life of 75,000 years. If enough could be trapped inside a large enough body, the heat generated would be enough to melt silicates, and you would end up with a large rock. The problem is to get the body sufficiently big in the first place. However, if you have any mechanism to get a rocky body of sufficient size in the first quarter My or so, that heating could make the object stronger.

My concept is a little different. As the dust in the early accretion disk approaches the forming star it gets hotter. Silicates a little above bright red heat start to get sticky, so dust will agglomerate into small stones. Hotter still, and some liquids start to form, and an important point is that different types of liquids are often mutually immiscible, so you will get phase separation. There are some key temperatures. Specifically, about 1550, iron will liquefy, and globs of iron will merge with other such globs. Above this temperature, iron filled bodies will form, and these may get covered in silicates. About 1250 degrees C, calcium silicate starts to phase separate; about 1500 degrees a series of calcium aluminosilicates start to phase separate, while about 1800 degrees C silicates themselves start to dramatically lose viscosity. The closer to the star, the larger the boulders, and plausibly most of Mercury could have accreted by this mechanism.

The problem now is to accrete the stones into larger bodies in the second stage, i.e. with lower temperatures and some disk gas remaining. Significant collisions will lead to fragmentation and dust formation, and we observe such dust, even in old disks. A larger body orbits with Keplerian motion, but the smaller ones get entrained in the gas and gradually fall towards the star. Accordingly, larger bodies will get continually struck gently by smaller objects, and any such grouping will collect dust. Now, the important point is in certain zones that experienced appropriate temperatures there will be phases of certain materials such as calcium silicate and calcium aluminosilicates, and these are hydraulic cements. Accordingly, water vapour will set the cements. Mars would have to rely only on essentially the calcium silicate type cement, which sets with one mole equivalent of water. Earth would be in an optimal position. There would be iron containing boulders, and the cements would include the calcium aluminosilicates, which can set up to fifteen mole equivalents of water. Venus is in a less desirable zone from the point of view of cement, because the higher temperatures from the star would inhibit the setting, nevertheless, once underway, the concentration of boulders is higher than around Earth.

What evidence is there for this? Perhaps the most impressive is that Earth has continents. These are granitic/feldsic, and are based on aluminosilicates. The reason is that granite and feldspar have relative densities about 2.5 – 3, whereas the basalt that makes up the bulk of the planet has a density > 3. So the continents float on the mantle, like icebergs. Venus is hard to get evidence, however there are two modest continents that are likely to be granitic. Mars has generally much lower levels of aluminium, and has no granite as far as we can tell, although on Syrtis Major there are kilometer-thick sheets of plagioclase, which can be thought of a s a sort of mix of granite and basalt. This also accounts for the atmospheres, but that is for the next Monday post.

Is this mechanism realistic? The Japanese Space Agency sent a mission to the small asteroid Itokawa, and it turned out that it had seemingly formed and had never been struck by anything of significance since. I am attaching an image, which I interpret as two larger boulders having met and been joined together by a concrete in the middle, and the result also grew somewhat by small material landing on it and being cemented on. But then I am biased, having formed this theory. Your thoughts?



4 thoughts on “How do Rocky Planets Form?

  1. I must have misunderstood something — the Moon formed only 30-60 million years ago? Or are you saying it took 30-60 million years for the Moon to form, but this happened 4.5 billion years ago?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s