Let’s start with the familiar and take a journey to the center of the Earth. The rocky super-Earth cutout in Figure 5.1a is a good illustration to the way ahead.
The outermost layer is the solid crust. What it’s made of depends in part on where we stand. On our own planet, if we begin our journey in California, the crust will be a layer of rocks rich in silica, like granite, that goes about 20 kilometers deep. If we begin our journey on the Pacific Ocean floor near the Hawaiian Islands, the crust will be a layer of basalt rocks denser than silica, such as olivine, that goes only about 5 kilometers deep. On average the Earth’s crust is 30 kilometers thick, thinner than that, as we have seen, under oceans, and up to 60 kilometers thick under continental mountain ranges. Our journey to the center is 6,400 kilometers long, so the crust of 30 kilometers is a very brief introduction.
As we go deeper the temperature rises steadily, as miners know all too well, as does the pressure due to all the rocks above us. Consequently, below the crust is a region of partial melting (the lava of many volcanoes originates there) that quickly becomes the mantle, a thick layer of hot rock, often described as molten. This is actually a misnomer. Yes, on Earth’s surface rock at 2,000 degrees flows like liquid from volcanoes, but under the enormous pressure deep in the mantle that same rock is more like cold honey: malleable and extremely slow to flow.9 The mantle is also in a state that resembles boiling in slow motion (called convection)—the mantle is so viscous that bubbles in it take millions of years to float to the top.10 However, this is short compared to the life of the Earth, so on a geological (or planetary) scale of time there is a lot of churning and mixing going on. The temperature at the bottom of Earth’s mantle is about 3,700 degrees; that heat is what drives the flow toward the surface. Meanwhile, colder and denser rock near the surface sinks and flows down, dissolving completely in the process. We can see some of the effects of all this churning on the surface, as this convection pushes around the fractured pieces of the crust—known as the tectonic plates—slowly rearranging the continents.
The mantle is Earth’s largest layer, some 2,860 kilometers thick, and takes up 84 percent of the planet’s volume. The mantle is much denser than the continental rocks, and consists mostly of a mineral called perovskite: a dark dense rock rich in iron and magnesium that is more than 50 percent denser than granite.11
Below the mantle, some 2,900 kilometers below the surface, we encounter the core, consisting of pure iron and iron alloys. The temperature here is very high—5,000 to 7,000 Kelvin—comparable to the temperature on the surface of the Sun. The pressure is very high too. The core is equal in thickness to the mantle, but, being in the center, it is only about 15 percent of the planet’s volume. Earth’s core consists of two parts: an outer liquid iron core and a solid one below it. Of course, the outer core is liquid in the same sense as the mantle is.
FIGURE 5.1. The interior of a super-Earth. The left image (a) shows the interior of a rocky super-Earth that resembles the interior of the Earth; on the right, the image (b) shows an ocean or water super-Earth with most of its water in a solid form.
We have learned all this from studying earthquakes. The best view of the interior is afforded by analyzing the paths earthquake waves take as they travel through the Earth. Seismographs all over the planet record where and how fast the different waves produced by a single earthquake arrive after bouncing inside the Earth. (This is not too different from the way ultrasound images the inside of the human body from outside.) The full picture is put together by adding to this our knowledge of Earth’s magnetic field, heat flow, gravity studies from spacecraft, and laboratory experiments on rocks under high pressure.
I alluded to how some of these structures form in Chapter 2, but the full picture is a bit more complicated than I’ve explained thus far. The preplanet structure—the “seed” of a planet—consists of solids (mostly silicates) and volatiles (such as water and ammonia), with trace amounts of hydrogen and noble gases. Due to the energy of the accretion process and the constant collisions with large solid bodies, this seed is thoroughly molten. (Some of Earth’s internal heat is a relic of this process.) In this state the structure differentiates. Iron and siderophile elements (high-density transition metals that like to bond with iron) precipitate from the silicate mix and sink under their own weight to form the core in the center. The remaining silicate minerals will remain in a mantle with the less dense ones closer to the top. Volatiles that are left over after hydrating the mantle minerals will rise to the surface and atmosphere.
This process, called planetary differentiation, is quick in geological terms and works for planetary bodies as small as big asteroids just a few miles across. Iron meteorites—pieces of pure iron alloy that orbit around the Sun until one day they fall to Earth for us to find them—originate in the differentiated iron cores of asteroids that were later smashed up by collisions with other asteroids. So, although we have no samples of our Earth’s iron core, and no good prospects to get them anytime soon, iron meteorites are excellent proxies. Differentiation is an orderly and predictable process thanks to our knowledge of chemistry and mineral properties under pressure.
Some super-Earths, the rocky ones, develop quite similarly, although the pressure in the mantle is almost tenfold higher and different varieties of minerals form. Other super-Earths, the oceanic ones, are totally exotic beasts, with oceans that are 100 kilometers deep overlying a dense hot solid water, called ice VII.
It might seem ridiculous to refer to this water as ice, given that it is at a searing temperature of 1,000 K, but under such high pressures, it forms. Water—H2O—has a familiar structure and formula, but our familiarity with it can make us overlook the fact that it is actually very complex. One key feature is that the oxygen atom in its molecule does not share electrons equally with its two hydrogen atoms; the result is that the molecule ends up with an asymmetric distribution of electrical charge. Imagine the tiny water molecules like small magnets, except with three poles (a negative O and two positive H’s). Water molecules interact with each other because the positive charge near a hydrogen atom of one molecule bonds weakly with the negative charge near the oxygen atom of another molecule. Many such weak bonds together can form a strong structure if the temperature becomes low enough to allow it. Thus common water ice is formed, dubbed ice Ih or hexagonal ice. In common ice the weak bonds between the molecules cause the molecules to form rings (mostly hexagons) that leave lots of empty space in between. The empty space gives it a lower density than liquid water, and so—as you know from a glass of ice water—it floats.
Under high pressure the density of water increases as the molecules are forced closer together; the bonds are bent to form tighter rings, which also interpenetrate. That makes the water solid, almost irrespective of the temperature, and much denser than the liquid phase.12 The high-pressure ices that exist at high temperatures are known as ice VII, X, and XI; these are the ice phases we expect to find inside oceanic super- Earths. 13 These ices are still less dense than rocks, however, so an oceanic super-Earth will be less dense than a rocky one of the same mass.
Ocean planets might be very common in the Universe because water is very common in the low-temperature environments where planets form and evolve.14 This might be especially true for super-Earths, which can retain volatiles more easily thanks to their larger mass and surface gravity.15 In order for a planet to become an ocean planet, it should form with or obtain at least 10 percent of its mass in water. Ammonia could be mixed in, but water is by far the dominant volatile chemical we see among the materials in protoplanetary disks. For comparison, Earth’s oceans are just about 0.02 percent of its mass. However, a much greater amount of water could be incorporated into Earth’s mantle.16 For that reason I assume a much larger fraction of water (greater than 10 percent) to produce a separate uninterrupted layer of water surrounding a planet (Figure 5.1b).
An ocean planet, regardless of its surface temperature, should have the same layers inside: an iron cor
e surrounded by a silicate-rich mantle that transitions into the hot water ice. The latter will become liquid water near the surface (the last 100 kilometers or so). The surface of the liquid water ocean will be covered with ice Ih, if the planet is far from its star and cold, like Jupiter’s moon Europa. If the planet is close to its star and hot at the surface, the liquid ocean will transition into a thick hot steam atmosphere. If the planet has moderate temperatures such as we have on Earth, the water ocean will resemble Earth’s, but there will be no continents or basalt tectonic plates under it. The interior of the ocean planet will remain under the control of the planet’s internal reservoir of heat. The transition between silicate mantle and hot water ice happens with a small change in density but no change in temperature, and the two materials have similarly high viscosity. Like the silicate mantle, the hot water ice “mantle” convects slowly.
The two families of super-Earths have planets that are diverse in size and amount of water. These characteristics depend on the mixture of elements present as the planet forms. From studying the spectra of many stars, we know that the amount of iron and other heavy elements will be different in different planetary systems. We already know that where in the proto-planetary disk a planet forms also matters. So, among the rocky planets we could find super-Mercurys—planets that have as much as 70 percent of iron core inside, like our planet Mercury.17 Or we could find super-Moons—planets that have no iron core, just an iron-rich mantle and perhaps a water layer.18
There is a third possible family of super-Earths and terrestrial planets—carbon planets. These would be extremely rare, as they require more carbon than oxygen to be present in the planet-forming mixture.19 Normally carbon is half as abundant as oxygen, as we saw in Figure 2.1. But astronomers have observed rare stars in which carbon is more plentiful than oxygen. A planet that forms from such a mixture will be different—it will have a mantle rich in silicon carbide and graphite in its interior.20 It will still have a precipitated iron core, but its overall size and the chemistry on its surface and crust will be very different. Silicon carbide is a very hardy substance—we use it to make durable ceramics, the disk brakes of sports cars, and tools for other high-stress environments. So volcanism, tectonics, and weathering are going to be minimal on carbon planets. Also, carbon planets are likely deficient in water.
Figure 5.2 shows some imagined family portraits of super-Earths and compares them to Earth and Neptune. The relative sizes of the super-Earths are accurate to the best of our theoretical models. Carbon planets are not shown in Figure 5.2. They will have intermediate sizes between rocky and ocean planets. The Neptune-like giant planet orbiting the star Gliese 436 is shown for comparison too.21
Finally, in the approximate mass range for super-Earths we discover small planets with relatively large amounts of hydrogen and helium—perhaps up to 10 percent by mass—just as we see on Neptune. Over all, of course, the planets would be smaller than Neptune: call them mini-Neptunes. It is still unknown where the transition occurs from planets rich in solids to planets with increasingly more massive hydrogen-helium gas envelopes. Theory gives us multiple possible solutions and no clear choices, so we’ll need the observations to show us nature’s preferences, and NASA’s Kepler is well on its way to provide them. Fortunately, it appears that studying the colors of such planets (a.k.a. spectroscopy) will allow separating the mini-Neptunes from the super-Earths.22
Even this doesn’t exhaust all the possibilities of terrestrial planets in the Universe. Another possible planet would simply be a bare iron core! Under most conditions, we would not expect to see such a thing because a planet body has to be assembled first (at which point it would mostly be silicate, as we’ve seen) and then differentiate, and only then have the iron precipitate in a core. The silicate mantle, without which there would be no iron core, would have to be stripped away to leave behind a bare iron core. That’s easy to do to a small body—a twelve-mile-wide asteroid, for example. Super-Earths are a different matter; they are so massive that even two super-Earths colliding head-on would not shatter to bits and expose blobs of pure iron.23 Even under such stress, their gravitational reach stretches far beyond their surface and brings rocks back in. In fact, our own planet Mercury, as small as it is compared to Earth (just 1/18 of Earth’s mass), is reported to have survived such a head-on collision early in its history and still retains most of its mantle. Nevertheless, iron super-Earths probably do exist, and we may have already discovered them—the pulsar planets.
FIGURE 5.2. Comparison of planet sizes among different super-Earths, as well as giants.
Pulsar planets were discovered by Alex Wolszczan and Dale Frail in 1992, marking the dawn of the hunt for exoplanets. As the name suggests, these planets orbit a pulsar—an ultradense, very massive (more than our Sun), fast spinning neutron star; a remnant from the explosion of a spent massive star, known as a supernova. We do not know their sizes (and there is no obvious way to measure them in the near future), nor do we have a good idea how they could have formed. Pulsar planets appear to have accumulated from the iron-rich debris left after the supernova explosion. Heavy-metal planets!
There are limits, however. We wouldn’t expect, for example, to see a pure-water planet. Even if we allow for the water to be mixed with ammonia and a small percentage of other impurities, a pure super-Snowball is as unlikely as a snowball on Venus. The problem, as with the pure iron super-Earths, lies again in planet formation. There is no conceivable way to “purify” the snowflakes from the dust in the debris disk or to shield the water-rich planet from accreting rocky bodies. Collisions don’t seem to help either, and for the same reason—super-Earths are just too massive. Even in our Solar System, the most water-rich bodies are comets and distant cousins of Pluto with water-rock ratios of about 1:1. If Kepler discovers something with a mean density of water, I would bet on it being an extraterrestrial civilization’s container for water storage ... or perhaps just a mixture of hydrogen gas, water, and rocks—a mini-Neptune planet.
Now that we’ve considered this range of planets, you may be wondering if it’s fair to use the name of our beautiful planet Earth for such a plethora of exotic worlds, even if the models of these other planets are fairly based on knowledge of our own. To me, the answer is clearly yes. Our Earth is a super-Earth, part of the big family, and something general and deep unites our planet with those others. It’s so deep, in fact, that you’ll find it at the bottom of Earth’s mantle. Let’s see what it is.
CHAPTER SIX
SUPER-EARTHS
The Hardest Rocks in the Universe
Rock hunting is fun: get yourself a hammer, a chisel, and a guidebook and head to the hills. If you’re lucky (and persistent), you’ll discover a pretty specimen to join your collection on the mantle or the bookshelf at home. There are so many—a seemingly endless variety. And the average rock hunter is just scratching the surface of the planet. If only we could look inside it, what wonders might we find? Actually, if you could go deeper inside the Earth, the variety of minerals would decrease dramatically to just a handful, with one of them—perovskite—dominating the bulk of the body of our planet (40 percent of its mass).
Perovskite, despite its planetary abundance, was not noticed and classified by geologists until 1837. The discovery was set in motion in the 1820s, when Czar Nicolas I of Russia, eager to map the uncharted riches of Siberia, invited the accomplished German naturalist Alexander von Humboldt to lead an expedition into the region. Humboldt set out east in the spring of 1829, taking two associates with him. They explored the Ural Mountains and the Caspian Sea, and reached the Altai Mountains of central Asia, where China, Russia, and Mongolia converge.
Reportedly Humboldt did not enjoy the trip much, though it appears to have been very successful.1 One of his associates, Gustav Rose, was a geologist. Among the minerals he brought back was a heavy dark piece found in the Ural region near the town of Slatoust. It had not been catalogued, so Rose had naming rights as its discoverer. He named it pero
vskite (originally, peroffskite), in honor of the Russian geologist Count Lev Alexeevich von Perovski, who was also the expedition’s host in Moscow.2 Gustav Rose had no idea that he had discovered the most abundant mineral on Earth!
Rocks are aggregates of one or more minerals. Minerals are inorganic solids with a definite chemical composition. They range from a single element (e.g., gold) to very complex silicates. On planets the predominant minerals are oxides—one or more oxygen atoms bound to silicon, iron, magnesium, aluminum, and so on. For example, silica, also known as quartz, consists of one silicon atom and two oxygen atoms. In silicate minerals that are most common in rocks, one silicon atom is usually surrounded by several oxygen atoms. Depending on how the atoms arrange themselves in regular repeating patterns, different crystal structures are possible.
The perovskite mineral Gustav Rose discovered in Russia was an oxide of titanium and calcium. There is a rich group (known as the perovskite group) of minerals that all share the same unique crystal structure of connected octahedrons of silicon or titanium bound to six oxygen atoms.3 Despite the fact that Rose discovered a titanium perovskite first, Earth’s lower mantle is, by volume, almost entirely taken up by silicon perovskite minerals.
Perovskite minerals contain rare earth metals (such as lanthanum, neodymium, and niobium) as trace elements. Perovskite is famous to physicists as the mineral in which high-temperature superconductivity was first discovered in 1986 by Alex Muller and Georg Bednorz.4 Nevertheless, most people have never heard of it, and it rarely appears in the popular field guides to rocks and minerals.5 (More common is a mineral called enstatite, similar to perovskite, which can be found in a variety of rocks.) Importance need not bring fame.
The Life of Super-Earths Page 6
