Fe-valence in magmatic clinopyroxene and the redox state of oceanic basalts – perspectives from natural samples and experiments

The redox state of magmatic systems controls important physico-chemical properties and processes on the Earth, such as the composition of volcanic gases, the rheology of magma, and the transport and deposition of critical metals. In natural silicate magma, Fe is the most abundant multivalent element, and the redox state of the system can be determined if the relationship between Fe valence and fO₂ is known. Popular oxybarometers (e.g., Fe³⁺/Fe^T in glass or Fe-Ti oxide pairs) can accurately determine the redox state of magmatic systems but suffer limitations such as beam damage during analysis or the requirement of specific phases to be present.

Clinopyroxene is a common igneous mineral that plays a key role in chemical cycling on Earth and is found in igneous rocks ranging from near-primary basalts to rhyolites. Clinopyroxene can incorporate both Fe²⁺ and Fe³⁺ and may capture a record of magmatic redox state upon crystallisation. However, attempts to quantify how Fe valence varies in clinopyroxene as a function of redox state are limited, partly due to the inability to routinely measure Fe valence using electron probe microanalysis (EPMA).

To remedy this, we are exploring the utility of the “flank method” [1] and conventional stoichiometric estimates for determining Fe valence using EPMA. Using a suite of Mössbauer calibrated clinopyroxene standards, preliminary “flank method” analyses demonstrate that the FeO content of clinopyroxenes ranging from diopside (FeO^T = 3 wt%) to aegirine (FeO^T = 28 wt%) can be determined with a RMSE of 0.03 wt%. Additional standards with FeOT from 2 – 10 wt% are being collated to improve the performance of this method when applied to augitic clinopyroxene. Furthermore, it is possible to obtain 3σ uncertainties on Fe³⁺/Fe^T of 3-7% using conventional stoichiometric estimates if EPMA analyses are sufficiently precise [2].

Using high-precision EPMA and stoichiometric estimates of Fe³⁺, we demonstrate that clinopyroxene crystals in oceanic basalts from the Reykjanes Ridge, Iceland and the Canary Islands have Fe³⁺/Fe^T of 0.1 – 0.7. Olivine-glass and olivine-spinel pairs constrain the redox state of these oceanic basalts to be equivalent to FMQ, FMQ+1.5 and FMQ+2, respectively, in line with independent estimates from the literature. The partitioning of Fe³⁺ between clinopyroxene and melt (K_D Fe³⁺cpx-melt) ranges from 0.50 – 1.4 in tholeiitic to alkali basalts, and we show that clinopyroxene Fe³⁺/Fe^T increases concomitantly with estimates of redox state.

However, there is currently limited experimental data in which Fe³⁺ has been measured with sufficient accuracy or precision to fully understand the controls on Fe³⁺ partitioning in basaltic systems, precluding the use of clinopyroxene as a probe for magmatic redox at present. An experimental campaign is currently underway to help refine models of Fe³⁺ partitioning, ultimately contributing to the development of a clinopyroxene based Fe-oxybarometer, and to shed light on the poorly defined role of Fe³⁺ in the chemical evolution of basaltic magmatic systems.

 

[1] Hofer & Brey, 2007. Am Min, 92, pp.873-885.

[2] Neave et al., 2024. CMP, 179, 5. doi: 10.1007/s00410-023-02080-2