Mantle behaviour and the influence of minerals

To most geologists minerals are a means to an end. Identifying them and working out their relative proportions in a rock provides a quick means of assessing its rough chemical composition. Textural relations between minerals help work out the sequence of processes that were involved in its evolution, and in the case of metamorphic minerals what pressure and temperatures were involved. In the case of the Earth’s mantle, however, mineralogy comprises only one or two abundant minerals – olivine and pyroxene at shallow depths, and the mineral perovskite (MgSiO3) at depths greater than about 670 km – and dominates the mantle’s physical properties and bulk behaviour. There are distinct, narrow zones or discontinuities that separate different seismic properties and these have long been considered to represent changes in mineralogy of the more or less uniform bulk composition of the mantle. The most likely phase transition is from olivine + pyroxene to perovskite, in response to increasing pressure, thought to occur at about 670 km down. That transition was confirmed by high-pressure experiments, but whether that simple mineralogy persists down to the outer core has remained a mystery. Using tiny diamond anvils in a laser-heated furnace to create the enormous pressures at depths up to 2700 km is fraught with technical difficulties, but Kei Hirose and Shigeaki Ono of the Japan Marine Science and Technology Centre have finally achieved them (see Cyranoski, D. 2006. Magical mantle tour. Nature, v. 440, p. 1108-1110).

Hirose and Ono discovered that perovskite itself collapses to produce another, more tightly-packed molecular structure – post-perovskite with a sheet-like structure. This phase transition occurred experimentally under conditions that characterise the thin D” layer just above the core-mantle boundary. Seismic tomography has suggested that a number of weird things happen there. For instance, seismic S waves near the CMB have different speeds according to their direction of travel, and even accelerate in some parts. The platy structure of post-perovskite, unlike the more regular perovskite, is likely to create such physical anisotropy, especially if grains are aligned. The mineral, when iron enters its structure, may also help to explain thin (5-40 mm) zones in the D” layer in which seismic wave speeds fall by 5 to 30% compared with expected values (Mao, W.L. et al. 2006. Iron-rich post-perovskite and the origin of ultralow-velocity zones. Science, v. 312, p. 564-565). When first detected by seismic tomography, these zones had been assumed to involve regions in which partial melting occurred. It also seems that the phase transition is temperature- as well as pressure-dependent, so that post-perovskite could form at shallower depths in cooler regions. Being denser than its parent, that could result in sinking: like slab-pull at shallow depths, such a gravitational force would contribute to whole mantle convection by displacing hotter D” material. That in turn would ‘flip’ through the phase transition in the reverse fashion to become less dense, perhaps encouraging the initiation of rising plumes.

Sure enough, what might seem to be a boring bit of exotic mineralogy promises to exert some control over speculation on what happens at the bottom of the mantle. But it is too early to say how seminal the discovery might be – the errors in the experiments correspond to a depth range of about 350 km. On top of that, other experiments need to be conducted under these extremely difficult conditions, such as finding out if post-perovskite can chemically interact with the iron-rich outer core, and if its electrical properties are in some way different from those of better-understood perovskite.

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 )

Connecting to %s