High-pressure phase transitions and equations of state in NiSi. I.Ab initiosimulations

Abstract

First-principles calculations have been used to determine the equation of state and structural properties of NiSi up to pressures equivalent to that in the Earth's inner core. At atmospheric pressure, the thermodynamically stable phase is that with the MnP structure (as found experimentally). At high pressures, NiSi shows phase transformations to a number of high-pressure polymorphs. For pressures greater than ∼250 GPa, the thermodynamically stable phase of NiSi is that with the CsCl structure, which persists to the highest pressures simulated (∼500 GPa). At the pressures of the Earth's inner core, therefore, NiSi and FeSi will be isostructural and thus are likely to form a solid solution. The density contrast between NiSi and FeSi at inner-core pressures is ∼6%, with NiSi being the denser phase. Therefore, if a CsCl-structured (Fe,Ni)Si alloy were present in the inner core, its density (for the commonly assumed nickel content) might be expected to be ∼1% greater than that of pure FeSi.

High-temperature ab initio calculations on FeSi and NiSi at conditions relevant to small planetary cores

“…More recently, a detailed experimental investigation of the NiSi phase diagram to ~65 GPa using both MAP and LHDAC techniques (Dobson et al 2016) determined the boundaries between the MnP, Pmmn, ε-FeSi and CsCl phases. In particular, the ε-FeSi to CsCl boundary was found to have a Clapeyron slope of -67 MPa/K, with the ε-FeSi + CsCl + liquid invariant point occurring at ~33 GPa and ~2125 K. (this study also found that, when the MnP-structured material is compressed at 300 K, a transition to a further metastable phase of NiSi occurs, somewhere between 35 and 60 GPa, with the high-pressure structure corresponding to that labelled Pnma-II by Vočadlo et al (2012) for which the transition was predicted to occur at ~42 GPa).…”

“…Stability of NiSi-structured phases in FeSi at 0 K Some of the stable NiSi structures seen in Vočadlo et al (2012) were a result of spontaneous transformations after relaxation of the starting structure. In order to replicate these calculations as closely as possible, the same starting structures have been used here, with the exception of the NiAs, 'anti'-NiAs and NaCl structures, which were all found to be unstable by a large margin in NiSi.…”

“…Therefore, the structures that have been considered are the MnP, 'anti-MnP', Pbma-I and WC structures, as well as the new Pmmn phase found by Wood et al (2013). Further details of these structures can be found in Vočadlo et al (2012) and Wood et al (2013). The calculated energy-volume values were fitted to third-order BirchMuraghan equations of state to obtain the enthalpy-pressure values.…”

“…More recent static computer simulations by Gavryushkin et al (2015) confirmed the stability of the Pmmn phase but also suggested that a tetragonally distorted (a/c ~ 0.8) CsCltype structure formed above 213 GPa, becoming fully cubic above 522 GPa. In the static simulations of Vočadlo et al (2012), it was, however, noticed that the enthalpy difference between the ε-FeSi structured form of NiSi and the thermodynamically stable phases was as small as 8-12 meV/atom for part of the pressure range considered, suggesting that the ε-FeSi structure might become stable at finite temperatures. A subsequent study using LH-DAC and synchrotron X-ray diffraction by Lord et al (2012) confirmed that this was indeed the case, with a transformation to the ε-FeSi structure being observed at 12.5 GPa and 1550 K, prior to a transformation to the cubic CsCl structure at 46 GPa and 1900 K; no evidence for the large tetragonal distortion suggested by Gavryushkin et al (2015) was found in the X-ray diffraction patterns of the quenched samples.…”

“…Following this work, Wood et al (2013) found a new stable phase of NiSi in experiments carried out in a MAP. This new phase of NiSi, with Pmmn symmetry (an orthorhombic distortion of the tetragonal CuTi structure), had not originally been considered by Vočadlo et al (2012), but further ab initio calculations revealed that this Pmmn phase was indeed more stable than either of the P4/nmm, Pbma-I or Pnma-III (FeB) phases, resulting in a much simpler phase diagram at 0 K. The new phase stability sequence in NiSi, therefore, became MnP → Pmmn → CsCl, with transitions occurring at 21 and 264 GPa (Wood et al 2013). More recent static computer simulations by Gavryushkin et al (2015) confirmed the stability of the Pmmn phase but also suggested that a tetragonally distorted (a/c ~ 0.8) CsCltype structure formed above 213 GPa, becoming fully cubic above 522 GPa.…”