The composition of the Earth

Geological sampling and laboratory measurements of the properties of rocks and minerals show that P wave velocities maybe associated with composition. The continental crust consists mainly of granitic rocks with gabbro near the bottom. There is no granite on the floors of the deep oceans, the crust there being entirely basalt and gabbro. The mantle below the Mohorovicic discontinuity consists primarily of the dense ultrabasic rock peridotite. The thickness of the crust varies from about 33 km under continents (even more under mountains) to about 5 km under ocean floorsю

The structure of the outermost 700 km of the Earth as revealed by the velocity of S waves is shown in Figure 3.13. The uppermost region is called the lithosphere. a slab about 70 km thick in which the continents are embedded. Its lower boundary is marked by an abrupt decrease in shear wave velocity. The lithosphere is characterized by high velocity and efficient propagation of seismic waves, implying solidity and strength. Below the lithosphere is the asthenosphere (or zone of weakness).

3.13: The structure of the outermost 700 km of the Earth illustrated by a plot of S-wave velocity against depth

It is also a low velocity zone where seismic waves are attenuated more strongly than anywhere else in the Earth. These features of the asthenosphere may indicate partial melting to the extent of 1 to 10 per cent. The velocity and density in both the lithosphere and asthenosphere would be satisfied by a peridotitic composition (ie predominantly of the mineral olivine), the boundary at 70 km being the phase change separating the solid lithosphere from the weaker asthenosphere.

3.12: Diagrammatic section through the entire structure of the Earth, showing the names and certain physical properties the various layers into which the planet is divided. The data for the subsurface is based almost entirely on seismic evidence outlined in this chapter

The asthenosphere extends to a depth of about 250 km, where the rocks again become solid. There is also a slight increase of velocity with depth because of increasing pressure. There is then a very rapid increase in velocity and density through a narrow zone at a depth of about 400 km. This increase is too rapid to be accounted for by a composition change; a change of phase (ie a closer packing on the atomic level) is required.

This theoretical explanation was verified in 1969 independently by two researchers who subjected olivine to high pressures and found that at a critical pressure and temperature its atoms take up a more compact arrangement, changing into the spinel structure. These critical conditions of temperature and pressure occur at a depth of about 400 km in the Earth. There is another region of rapid change at around 650 km following a region of small gradual changes with depth.

This zone was first indicated by shock wave experiments, and in 1974 workers in Japan and the USA reached these pressures in static compression tests and found that olivine breaks down into dense, simple oxides of iron, silicon and magnesium. Recent diamond-anvil experiments at the Australian National University, Canberra, and at the Carnegie Institution of Washington have helped establish the perovskite type of crystal structure as a major form for mantle rock above 0.3 megabar, equivalent to a depth of about 650 km.

Geochemical and geophysical evidence indicates that the mantle is composed largely of oxygen, silicon, magnesium and iron, but there is less agreement about how these elements are grouped to form minerals and how the mineral structures adjust to the different pressures and temperatures at different depths. A likely combination of mantle minerals is the three silicates, olivine, pyroxene and garnet. It has been suggested that the sudden large change in density at 650 km in part results from the atoms in the olivine crystal adjusting to the pressure by changing into the denser perovskite form, in which the oxygen atoms are closer to each other. The lower mantle extending from about 650 km to the core at a depth of about 2885 km shows little change in either composition or phase, and velocity and density increase fairly smoothly with increasing pressure.

Shear waves have never been observed in the outer part of the core, which is thus considered to be liquid. However, evidence for the existence of an inner core, which is solid, has steadily increased. In particular it has been shown that the free oscillation data demand a solid inner core in which the average shear velocity is about 3.6 km per second. The observation of shear waves in the inner core would establish its rigidity. In 1972 shear waves were claimed to have been seen on a seismograph, and a value for their velocity of about 2.95 km per second was deduced. Unfortunately this cannot be reconciled with the value of 3.6 km per second from the free oscillation data. It is not clear exactly what phase had been identified on the seismograph.

All evidence indicates that the inner core consists primarily of iron and nickel and the outer core of molten iron alloyed with 8-20 per cent of a light element, giving an average atomic number of about 23. Shock wave data for iron indicate that both densities and bulk moduli in the outer core are less than those of iron under equivalent conditions (Figure 3.14), although their gradients through the outer core are consistent with gross chemical homogeneity (ie uniform intermixing of iron with a lighter, more compressible element or compound). Both densities and bulk moduli for the inner core are compatible with those of iron, suggesting that the inner/outer core boundary is likely to be a compositional as well as a phase boundary.

There is no firm evidence as to the identity of the light element in the outer core. It must be reasonably abundant, miscible with iron and possess chemical properties that would allow it to enter the core. The prime candidates are silicon and sulphur; more recently oxygen has been suggested.

3.14: Shock wave data for certain materials under varying degrees of compression for testing hypotheses of the composition of the Earth's mantle and core. Seismic evidence limits the properties of mantle and core to the shaded areas. The dunite. consisting almost entirely of magnesium-rich olivine, most closely approximates the properties of the mantle, while an iron-silicon alloy comes closest to matching those of the core

Seismology only gives us a 'snapshot' of the interior of the Earth as it is today and gives no information about its structure in the past or its evolution. The existence of a liquid outer core and solid inner core raises the question of the Earth's origin. Has the Earth always had such a structure, or has it evolved over geological time? This question cannot be divorced from the much broader question of the origin of the Earth and solar system.

It is usually assumed that the embryo Earth was homogeneous and later differentiated into crust, mantle and core. Inhomogeneous models have also been suggested in which the (mainly iron) core formed first, the silicate mantle being deposited upon it later. There is one constraint, however, that can be imposed upon the possible evolution of the Earth's core. It is generally believed that the Earth's magnetic field is due to some form of electromagnetic induction in the fluid, electrically conducting outer core.

Rocks as old as about three thousand million years have been found to possess remanent magnetization, and so the Earth must have had a molten outer core comparable to its present size at least that long ago. A major problem in understanding the evolution of the Earth is the means by which it could have heated up sufficiently to lead to a (predominantly iron) molten outer core that long ago.

Fairly short periods (about twenty million years after accretion) are now favoured as being the time required to form the core. An upper limit is about five hundred million years so that core formation should have been essentially complete during the first tenth of the Earth's history. Deducing the subsequent evolution from an approximately homogeneous all-fluid core into the present state of a solid (mainly iron) inner core and fluid outer core containing some light alloying element is another problem. This transformation may have proceeded much more slowly and still may not be complete. It should be noted that gravitational energy would be released by the growth of the inner core and this may provide the main source of energy for driving the geodynamo.