Abstract(Mg,Fe)O ferropericlase‐magnesiowüstite has been proposed to host the majority of Earth's sodium, but the mechanism and capacity for incorporating the alkali cation remain unclear. In this work, experiments in the laser‐heated diamond anvil cell and first‐principles calculations determine the solubility of sodium and favorability of sodium incorporation in iron‐rich magnesiowüstite relative to (Mg,Fe)SiO3 bridgmanite. Reaction of Mg/(Mg + Fe) (Mg#) 55 and 28 olivine with NaCl at 33–128 GPa and 1600–3000 K produces iron‐rich magnesiowüstite containing several percent sodium, while iron‐rich bridgmanite contains little to no detectable sodium. In sodium‐saturated magnesiowüstite, sodium number [Na/(Na + Mg + Fe)] is 2–5 atomic percent at pressures below 60 GPa and drastically increases to 10–20 atomic percent at deep lower mantle pressures. For these two compositions, there is no significant dependence of the results on Mg#. Our calculations not only show consistent results with experiments but further indicate that such an increase in solubility and partitioning of Na into magnesiowüstite is driven by the spin transition in iron. These results provide fundamental constraints on the crystal chemistry of sodium at lower‐mantle conditions. If the sodium capacity of (Mg,Fe)O is not strongly dependent on Mg#, (Mg,Fe)O in the lower mantle may have the capacity to store the entire sodium budget of the Earth.
Read full abstract