Martinet et al, bracketed the decomposition reaction of dolomite to aragoni
te + magnesite between 5.0 and 5.5 GPa at 600 OC using in-situ XRD in a mul
tianvil apparatus. They then extrapolated their results using the available
thermodynamic data such that the breakdown reaction occurs at similar to1,
000 degreesC at 6 GPa. If correct, this extrapolation has implications for
the type(s) of stable carbonate(s) in the mantle. For example, the reaction
magnesite + diopside = dolomite + enstatite would be metastable at P > 7 G
Pa relative to the aragonite-containing analog, and thus aragonite, rather
than dolomite, might be stable near the solidus of Iherzolite. This extrapo
lation would also imply that aragonite and magnesite would be stable in gar
net + clinopyroxene assemblages at pressures in the stability field of diam
ond. To test this extrapolation, experiments were conducted at higher tempe
ratures in a multiple-anvil apparatus. Results of initial experiments were
consistent with the results of Martinet et al. At higher temperatures, howe
ver, the reaction occurs at higher pressures than predicted by Martinet et
al., requiring significant curvature to the reaction boundary. This curvatu
re (increasing dP/dT with increasing T) is consistent with increasing disor
der in the dolomite with increasing temperature, increasing the stability o
f dolomite relative to aragonite plus magnesite. The experimental results r
estrict aragonite + magnesite-bearing assemblages to low temperature. Arago
nite will not be present in near-solidus Iherzolite, and can be present in
orthopyroxene-free eclogites and pyroxenites only at low temperature.