The Ab initio study proposes two systems as low-pressure analogues for the post-post-perovskite transitions in the Mg-Si-O system, which are fundamentally important for understanding the deep interiors of super-Earths but are challenging to study experimentally. Mg-Ge-O and Na-Mg-F systems are very useful for experimental investigation of the post-post-perovskite phases, Mg2SiO4 and MgSi2O5, respectively.

Background

Bridgmanite, MgSiO₃ perovskite, is the most abundant mineral in the lower mantle of the Earth. It transforms into the post-perovskite (PPV) structure close to core-mantle boundary conditions (~125 GPa). Although the PPV structure is the final form of MgSiO₃ in the Earth, a natural question arises: what is next? What is the post-PPV transition? This question has gained importance, since many potential ‘super-Earths’ – terrestrial exoplanets whose masses are up to ~13 Earth masses (or less than one Uranus/Neptune mass) – have been discovered recently. Pressures in their mantles can reach up to ~2.5 TPa; such ultra-high pressures are expected to induce post-PPV transitions in the Mg-Si-O system. A series of predictions of post-PPV phase transitions in MgSiO₃ and other stoichiometries in the Mg-Si-O system have been made and confirmed by several calculations. MgSiO₃ should undergo the following three-stage dissociation: MgSiO₃ PPV → Mg₂SiO₄ + MgSi₂O₅ → Mg₂SiO₄ + SiO₂ → MgO + SiO₂. When MgSiO₃ coexists with MgO or SiO₂, they should recombine into Mg₂SiO₄ or MgSi₂O₅, respectively. Predictions of these recombinations are unexpected. Mg₂SiO₄ polymorphs dominate the Earth’s upper mantle. The spinel form of Mg₂SiO₄ (ringwoodite) breaks down into MgSiO₃ perovskite and MgO, defining the upper boundary of the Earth’s lower mantle. These predictions have been utilised in numerical modelling of the interiors of super-Earths. However, despite the importance of the post-PPV transitions in super-Earths, these predictions have not been confirmed experimentally. This is because the predicted transition pressures are extremely high, over ~500 GPa. Such high pressures are still difficult to achieve with diamond-anvil-cell experiments. Without experimental confirmation, the computational predictions might be seen as just a pie in the sky. For this reason, studies of ‘low-pressure-analogues’ (LPAs) of key high-pressure phases are important. Such analogue materials are expected to display similar phase relations and structure-property relations to the experimentally inaccessible high-pressure forms. Today, in this age of exoplanetary discoveries, the search for LPAs is more active than ever.

Research outcome

This research presents an ab initio investigation of the pressure-induced behavior of MgGeO₃- and NaMgF₃-perovskite, traditional LPAs of MgSiO₃ bridgmanite. It is found that, although neither MgGeO₃ nor NaMgF₃ are perfect LPAs of MgSiO₃, they are useful for several reasons: (1) both display the PPV transition observed in MgSiO₃; (2) starting at ~20 GPa the Na-Mg-F system produces a novel NaMg₂F₅ phase by dissociation of NaMgF₃ or by its compression with MgF2. Instead of following this pattern, the Mg-Ge-O system produces a Mg₂GeO₄ by dissociation of MgGeO₃ or by its compression with MgO starting at ~200 GPa. Such pressures are routinely accessible in laser-heated diamond-anvil cells today; (3) like MgSiO₃, both systems ultimately dissociate into the binary AX and BX₂ compounds, confirming this trend in ternary systems first predicted in MgSiO₃. It should be noted here that very recently the predicted post-PPV transitions in NaMgF₃ were reported experimentally (Dutta et al., Proc. Nat. Acad. Sci. 116, 19324 (2019)). Based on the LPAs this research proposed, experimental studies on the post-PPV transitions will be facilitated, leading to further understanding of the deep interiors of super-Earths.