The Paradox of Mantle Redox

See allHide authors and affiliations

Science  06 May 2005:
Vol. 308, Issue 5723, pp. 807-808
DOI: 10.1126/science.1110532

This article has a correction. Please see:

Redox reactions (those involving reduction or oxidation) occur in many everyday processes, from photosynthesis and metabolism to fuel combustion and household cleaning. They also play a critical role in many geological systems. Processes on Earth's surface are intimately linked to the oxidation state of the mantle through the geochemical cycles of elements such as carbon, sulfur, oxygen, and hydrogen. Recent studies have advanced our understanding of the oxidation state of the mantle, elucidating the redox relations within Earth and their consequences for global processes.

The term “oxidation state” has caused some confusion in the geological literature, because it has two different meanings in the context of mantle properties. First, it is used to indicate the valence state of elements, for example, divalent iron (Fe2+) and trivalent iron (Fe3+). Second, it is used to indicate the chemical potential of oxygen, more commonly referred to as oxygen fugacity. High oxygen fugacity means oxidizing conditions, whereas low oxygen fugacity implies reducing conditions.

In everyday experience, these two definitions of oxidation state are almost always coupled: Oxidizing conditions favor the formation of Fe3+ (for example, rust on a car), whereas reducing conditions favor the formation of Fe2+ or even metallic iron (Fe0). However, paradoxical behaviors can arise when solids are present, because crystal structures impose additional constraints: Some minerals incorporate almost no Fe3+ even under oxidizing conditions, whereas others incorporate Fe3+ even under reducing conditions. A classic example is iron oxide, FexO, which always contains a measurable amount of Fe3+ in its crystal structure, even under reducing conditions where metallic iron is stable.

Studies of mantle rocks show that the oxygen fugacity of the upper mantle is relatively high (1), even though the abundance of oxidized iron (Fe3+) is low (2) (see the figure). How can we reconcile these apparently contradictory observations? The answer lies in the unfavorable energetics of defect incorporation in olivine, the most abundant mineral in the upper mantle. This property leads to an almost negligible Fe3+ concentration in olivine even under relatively oxidizing conditions (3). Fe3+ is readily incorporated into the minerals spinel and garnet, but because they are at least 1/10th as abundant as olivine, their presence causes only a small increase in Fe3+ abundance. On the other hand, the Fe3+ concentrations of spinel and garnet determine oxygen fugacity through mineral equilibria (1); the resulting oxygen fugacity of the upper mantle is therefore relatively high.

The relation between Fe3+ abundance and oxygen fugacity is also paradoxical in the lower mantle, but in an opposite sense. For decades it was assumed, based on the behavior of the upper mantle, that the dominant phase of the lower mantle, magnesium silicate perovskite, contains iron primarily as Fe2+. But high-pressure experiments on perovskites containing aluminum (a small amount of which is believed to be present in the lower mantle) have shown this assumption to be wrong (4). Subsequent studies, including those on natural samples (5), have confirmed that the combined substitution of Al3+ and iron into magnesium silicate perovskite stabilizes Fe3+. Fe3+ concentrations are high even under extremely reducing conditions (6).

The oxidation state of Earth's mantle.

The paradoxical relation between oxygen fugacity (left) and Fe3+ abundance (right) arises from the aversion of upper-mantle olivine and the strong affinity of lower-mantle perovskite for Fe3+. Oxygen fugacity is expressed relative to the conditions where fayalite, magnetite, and quartz are in equilibrium (the FMQ buffer); positive values are more oxidizing, and negative values are more reducing. Fe3+ concentrations are expressed as weight % Fe2O3. The high Fe3+ concentration of the lower mantle is balanced by ∼1 weight % metallic iron (6). Diamond formation in the lower mantle may arise from redox gradients in subducted material (18); other heterogeneities in mantle oxidation state are likely but not shown. The composition of subducted slabs is simplified to that of mid-ocean ridge basalt (MORB). Numerical estimates of oxygen fugacity and Fe3+ abundance are from (7, 15).

The high Fe3+ concentration in lower-mantle magnesium silicate perovskite has at least two important consequences. First, the Fe3+ concentration in the dominant lower mantle phase cannot be ignored; about 50% of the iron in the perovskite is expected to be Fe3+. The properties of the lower mantle may therefore differ from those predicted by aluminum-free experiments.

The difference between Fe2+ and Fe3+ in mantle minerals goes beyond the removal of an electron. In mantle minerals, iron occurs most frequently as Fe2+, and can thus substitute for Mg2+ without affecting the charge balance of an individual crystal. In contrast, substitution of Fe3+ causes an initial charge imbalance, because there are no other abundant elements with the same charge. Balancing the charge requires effects such as coupled substitution (for example, 2Fe3+ for Mg2+ + Si4+), the creation of defects, and/or the addition or loss of volatile elements such as hydrogen. Many mantle properties are sensitive to such effects; electrical conductivity, elasticity, and trace-element partitioning have been found to vary substantially with trivalent cation (Fe3+, Al3+) concentration (7).

The second consequence of high Fe3+ in lower-mantle magnesium silicate perovskite is that oxidation of iron must be balanced by a corresponding reduction reaction. The most likely reaction is self-reduction of iron according to 3Fe2+ → Fe0 (metal) + 2Fe3+, which results in the production of ∼1 weight % metallic iron in the lower mantle (6).

Although such a metal phase would probably evade geophysical detection, the potential geochemical consequences are immense, including the resolution of two long-standing conundrums: how the upper mantle became oxidized, and why the abundance of siderophile (metal-loving) elements in the mantle is so high (6).

The evolution of the oxidation state of the mantle through time has been a focus of hot debate, particularly because it relates to the rise of atmospheric oxygen and the origin of life (8). The abundance of redox-sensitive trace elements provides a window to the ancient mantle. Results have shown that the oxidation state of the upper mantle has remained essentially constant for ∼3.5 billion years (9, 10), implying that the upper mantle became oxidized relatively early in its history.

One plausible scenario for this oxidation is the removal during core formation of ∼10% of the metallic iron that formed through self-reduction when magnesium silicate perovskite first became stable within the accreting Earth (6). The removal of 10% of this metallic iron from the lower mantle would cause a net increase in Fe3+ in the lower mantle that, following mantle convection, would raise the oxygen fugacity of the upper mantle to current values. If the remaining metallic iron took up a portion of the siderophile elements, the redistribution of these elements throughout the mantle could account for their unusually high abundance (11) and remove the need for a late-stage addition of meteoritic material to the accreting Earth (12).

The mantle is connected to the atmosphere and the hydrosphere through geochemical cycles of the volatile elements. Subduction of oxidized sediments and exhalation of volcanic gases are a driving force in these cycles (13), but the nature of their coupling to the oxidation state of the mantle remains unclear.

Techniques to measure the oxidation state have improved rapidly in the recent past as a result of increased spatial resolution and the development of new proxies for oxygen fugacity. Coupled with focused research initiatives (14), the use of these techniques should improve our knowledge of the temporal and spatial evolution of the oxidation state of the mantle. Such information is crucial to understanding the rise in atmospheric oxygen that occurred ∼2.3 billion years ago (15), and may reveal whether atmospheric oxygen and perhaps life itself can be attributed to the peculiar properties of the perovskite structure.

References and Notes

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
View Abstract

Stay Connected to Science

Navigate This Article