Compositional effects on the vibrational properties of (Mg,Fe)O

One of the main constituents of Earth's lower mantle is (Mg,Fe)O periclase. Iron-poor (Mg1-xFex), with x<40, is of interest because it has been found to be in equilibrium with phase assemblages consisting of CaSiO3-perovskite and (Mg,Fe)(Si,Al)O3-perovskite under shallow lower mantle conditions (e. g., Irifune 1994). However, the partitioning of iron in deeper mantle assemblages is not certain. Nevertheless, the preferential partitioning of iron into periclase has been observed in the presence of perovskite and post-perovskite (e.g., Sakai et al. 2010). Further enrichment of iron may occur in the deepest regions due to interactions with the liquid iron-dominant outer core or fractional crystallization from a magma ocean (Labrosse et al. 2007). Although several measurements exist for iron-poor periclase and a few exist for iron-rich periclase, very few vibrational studies have been done on the intermediate composition. We present nuclear resonant inelastic x-ray scattering (NRIXS) measurements on (Mg0.5Fe0.5)O at ambient conditions. The 57Fe-weighted phonon density of states (PDOS) was determined from the raw data. Using the low energy region of the PDOS and the density of (Mg0.5Fe0.5)O, the Debye sound velocity (VD) is determined. With known bulk modulus, density, and VD, the seismically relevant compressional (VP) and shear (VS) are calculated. Good agreement is found when comparing to past ambient pressure studies on iron-bearing (Mg,Fe)O using various methods (e.g., Jacobsen et al. 2004; Lin et al. 2006; Wicks et al. 2010). Other parameters derived from the PDOS are also useful in obtaining a full understanding of the (Mg,Fe)O solid solution. These parameters include the average Lamb-Mössbauer factor (fLM), vibrational specific heat per atom at constant volume (cV), the vibrational entropy per atom (Svib), the mean force constant (D), and the vibrational kinetic energy (EK) of the 57Fe nucleus. We will discuss these parameters in the context of previous NRIXS results for (Mg,Fe)O. We find that our NRIXS results from (Mg0.5Fe0.5)O complement the parameters determined for iron-rich (Struzhkin et al. 2001; Wicks et al. 2010) and iron-poor (Lin et al. 2006; Chen et al. 2010) periclase samples. References: Chen, B. et al. (2010), Fall AGU. Irifune, T (1994), Nature, 370, 131-133. Jacobsen, S., et al. (2004), PNAS, 101, 5867-5871. Labrosse, S., et al. (2007), Nature, 450, 866-869. Lin, J.-F. et al. (2006), GRL 33, L22304. Sakai, T., et al. (2010), PCM, 37, 487-496. Struzhkin, V. V., et al. (2001), PRL, 87, 255501. Wicks, J. K. et al. (2010), GRL, 37, L15304.