The deep interior of our planet holds many secrets, but among the most profound is the story of how Earth’s oxidation state—essentially, its chemical balance between reduced and oxidized components—was established during its formative years. Recent groundbreaking research sheds new light on this complex narrative by focusing on bridgmanite, the mineral believed to comprise the majority of Earth’s lower mantle. This research reveals that the ferric iron (Fe³⁺) content within bridgmanite played a pivotal role in determining the oxidation state of the entire mantle, and by extension, the Earth itself.
When the Earth first formed, the upper mantle was likely equilibrated with iron-rich metal descending toward the core. This process would have stripped it of much of its oxidized iron, leaving it predominantly in the ferrous (Fe²⁺) state and correspondingly low in Fe³⁺ content. Today’s measurements, however, reveal that the upper mantle exhibits a Fe³⁺ to total iron ratio (Fe³⁺/ΣFe) ranging from approximately 0.02 to 0.06. This discrepancy has long perplexed geoscientists, raising questions about how the upper mantle became more oxidized than initial core formation would suggest.
A popular hypothesis posits that this rise in upper mantle oxidation occurred through mixing with oxygen-rich material from the lower mantle. Within the lower mantle, bridgmanite formation is accompanied by a process called ferrous iron charge disproportionation. In this reaction, ferrous iron (Fe²⁺) spontaneously converts into ferric iron (Fe³⁺) and metallic iron (Fe⁰), the latter of which, due to its metallic nature, can separate and sink into the core. Such a mechanism implies that the lower mantle itself became enriched in ferric iron and that through large-scale mixing, this oxygen-rich signature was imparted to the upper mantle.
However, this explanation encounters a key problem: partial separation of the resulting iron metal in the lower mantle is difficult to reconcile with geological evidence. If all the metallic iron separated efficiently and was lost to the core, then the Fe³⁺/ΣFe ratio in bridgmanite would be nearly 0.5 to 0.6 according to previous estimates, which would lead the upper mantle to be far more oxidized—three to five times its observed present-day value. Conversely, if no metal separation occurred, the lower mantle would not become oxygen-rich enough to oxidize the upper mantle at all. This paradox highlights an incomplete understanding of bridgmanite’s ferric iron content during its crystallization from the early magma ocean.
Recent investigations have provided refined measurements indicating that bridgmanite crystallized at the top of the lower mantle from silicate melt in equilibrium with metallic iron has a Fe³⁺/ΣFe ratio of approximately 0.14 to 0.25. This range is notably lower than the 0.5 to 0.6 observed under modern geotherm conditions, suggesting that the ferric iron content of bridgmanite during Earth’s formative magma ocean phase was relatively constrained. Furthermore, these studies show that pressure has negligible influence on this ratio in bridgmanite within pyrolitic systems, implying that the ferric iron content established during initial crystallization persisted along the solidus of the lower mantle as that region cooled and crystallized.
Building on these experimental results, researchers developed a detailed thermodynamic model, incorporating data for ferropericlase’s iron component, to compute bridgmanite’s Fe³⁺/ΣFe ratio along a mantle solidus estimated from an average of existing literature. This model assumes an oxygen fugacity—an effective measure of the availability of oxygen—at about two log units below the iron-wüstite buffer (IW −2.0), consistent with conditions hypothesized for the early Earth’s magma ocean. The model yields an average bridgmanite Fe³⁺/ΣFe ratio of approximately 0.17, indicating that the lower mantle as a whole would have an iron oxidation state around 0.09. When combined with the upper mantle, largely lacking in ferric iron, this results in a whole mantle oxidation state in the vicinity of 0.07, which aligns well with upper mantle observational estimates.
This finding sets an important upper limit for the oxygen content of the lower mantle, because under sub-solidus conditions—below the melting point defined by the solidus—metallic iron generated by disproportionation would not have been able to efficiently segregate into the core. Several plausible physical mechanisms existed to facilitate the loss of this metal to the core during Earth’s infancy. Firstly, the crystallization adiabat of the cooling magma ocean, closely mirroring the solidus, means that droplets or diapirs of metal liquid could percolate through already crystallized solid mantle material without becoming trapped. Secondly, experimental studies demonstrate that liquid iron wets grain boundaries under the pressure and temperature conditions prevalent in the mantle, allowing percolation that might be enhanced by mantle convection stresses or the aggregation of droplets into larger bodies sinking toward the core.
Moreover, metallurgical experiments suggest the solidus of iron metal itself lies very close to the silicate mantle solidus, providing a plausible lower bound to the depth and temperature range over which metallic iron could migrate downward before solidifying and halting further downward flow. Additionally, as the magma ocean crystallized, it is thought to have reached saturation in iron sulfide, a dense, metal-rich liquid that eventually separated to the core in what is known as the "Hadean matte." This sulfide melt may have acted as a scavenger, alloying with and removing disproportionated metallic iron from the silicate mantle, thereby influencing the Earth’s early oxidation state.
The model developed in this new study is limited in its ability to extend into melting regimes of the silicate mantle. However, based on the strong temperature dependence of ferric iron content in bridgmanite, the Fe³⁺/ΣFe ratio at crystallization is close to but probably slightly lower than that at the solidus. The close match between the whole-mantle oxidation state produced by model homogenization at the solidus and the observed ratios in the present upper mantle reinforces the view that the lower mantle played a key and active role in the oxidation state evolution following core formation.
Some theories have suggested that changes in the redox state of the lower mantle could lead to variations in density sufficient to drive buoyancy and mantle convection. Yet, the narrow range of Fe³⁺/ΣFe ratios predicted across the entire lower mantle today and during the early Earth suggests any such redox-driven density differences were likely minimal, and unlikely to exert a major influence on mantle dynamics.
Before bridgmanite began to crystallize, some ferrous iron disproportionation likely occurred within the molten silicate itself. Loss of resulting metal during this stage would have elevated ferric iron content in the magma. Yet, new high-pressure experiments indicate that the ferric content generated in the liquid phase was less than that produced as bridgmanite crystallized. Consequently, ongoing disproportionation during crystallization was essential to fix the ferric iron balance in the solid lower mantle. Limited metal retention in the mantle at near-solidus conditions remains a possibility, and further research is required to unravel the efficiency of metal segregation in Earth’s early deep interior.
Importantly, the current understanding does not conflict with observations that the modern lower mantle is saturated with iron metal. As the mantle cooled below the solidus, progressively more iron metal would have been generated to balance rising ferric iron contents in bridgmanite due to decreased temperatures. Quantitative estimates suggest that approximately 0.2 weight percent of metallic iron has formed near the top of the lower mantle through this charge disproportionation as the Earth’s interior cooled from the early magma ocean state to present conditions.
These contemporary conditions exhibit a slightly higher oxygen fugacity than during initial bridgmanite crystallization, reflecting temperature-dependent chemical equilibria, including increased nickel content in iron metal alloys and greater iron concentration in ferropericlase. Notably, the Fe³⁺/ΣFe ratio in bridgmanite is observed to be highest near the top of the lower mantle, decreasing at greater depths due to changes in iron spin states within ferropericlase, which influence iron partitioning between minerals.
This comprehensive synthesis brought forth by recent research not only resolves longstanding inconsistencies between mantle oxidation states but also highlights the critical role of bridgmanite’s ferric iron content as a controlling factor in Earth’s chemical evolution. It suggests that the oxidation state of our planet was far from static and instead dynamically shaped by intricate mineralogical processes during the cooling of its primordial magma ocean. Ultimately, Earth’s oxidation narrative is a direct reflection of the complex interplay between mineral physics, thermodynamics, and geochemical partitioning deep within its hidden mantle.
Subject of Research:
Article Title:
Article References:
Wang, F., Wang, L., Fei, H. et al. Bridgmanite’s ferric iron content determined Earth’s oxidation state. Nat. Geosci. (2025). https://doi.org/10.1038/s41561-025-01725-0
Image Credits: AI Generated
