This is an interesting problem but this paper has a few fundamental flaws. 1) The background on the current thinking of the Fe-isotope contents of the bulk silicate earth is out of date. 2) If the mantle were hotter, which it surely was in the past, post-perovskite would not be stable at the core-mantle boundary (does perovskite display a similar behavior?). Lastly (3) the mantle is not in equilibrium with the core at the CMB (Ni and Co partitioning).
The idea behind the paper is neat (using the partial density of states to predict isotope fractionation) unfortunately it left the reading group wanting more.

The idea behind the paper (pyroxenite vs. peridotite sources) is intriguing but the paper left myself as well as others in the reading group unconvinced. Figure 2c was particularly problematic; the authors asserted that there data was dissimilar from that of Wang and Eiler 2008, a conclusion I would not draw from looking at the same plot. On a more general note, no matter what the Os or O isotope composition of the sample they were able to explain the result successfully using the Sobolev model. Could they have included other elements (and isotopes) in their model and if so could it still explain the data? It seems to me that there are a lot of knobs to turn to fit this data (a delta 18O of 0 permil for layer 3 crust seems low…) an ... read more

Suggested paper