A first-principles study of the phase transition from Holl-I to Holl-II in the composition KAlSi3O8

Abstract

The phase relation and structural evolution of Holl-I and Holl-II in the composition KAlSi 3 O 8 at 0 K have been investigated by the first-principles method up to 130 GPa. Holl-I and Holl-II are polymorphs of KAlSi 3 O 8 stable at low pressures and high pressures, respectively. The transition pressure is determined at ∼23(5) GPa, in agreement with recent experimental observations. All experimentally observed major changes associated with this phase transition such as the deviation of the γ-angle from 90°, splitting of the a-and b-axes, as well as its P-V evolution, are successfully simulated. By evaluating the effect of different Al/Si substitution mechanisms on the computing cell of Holl-I, we have found: (1) different Al/Si substitution mechanisms do not result in apparent difference in the minimized cohesive energies, suggesting a possible random distribution of Al and Si; (2) different Al/ Si substitution mechanisms lead to different powder X-ray diffraction features, which, compared to the experimentally observed powder X-ray diffraction data, implies that local non-random distribution of Al and Si exists to some extent in the Holl-I structure; and (3) the phase transition from Holl-I to Holl-II might be associated with a change in the distribution pattern of Al and Si in the structure. From the simulated compression data, we have derived K 0 = 174 GPa and V 0 = 244.82 Å 3 for Holl-I, and K 0 = 168 GPa and V 0 = 244.8 Å 3 for Holl-II (K 0 ′ fixed at 4). The larger K 0 of Holl-I is probably related to the more stable squared open tunnel delimited by the rigid tetragonal octahedral framework, which is gradually deformed by compression in Holl-II after the phase transition from Holl-I to Holl-II.

Compressional behavior of MgCr2O4 spinel from first-principles simulation

“…Compared to the experimentally-determined curves, the curve obtained with the GGA method has a slightly steeper slope whereas that obtained with the LDA method has a more gradual slope. In general, the GGA method should yield a lower K T and the LDA method should yield a higher K T , with the experimentally-constrained K T falling in between (Deng et al, 2011). The V/V 0 -P trend from Nestola et al (2014) is much closer to our GGA result, while the V/V 0 -P trend from Yong et al (2012) is much closer to our LDA result.…”

First‐principles prediction of fast migration channels of potassium ions in KAlSi3O8 hollandite: Implications for high conductivity anomalies in subduction zones

“…In view of the disordered site occupation between Al and Si, two models with and without tetragonal symmetry were considered in our simulations. Model A (Figure a) applying the tetragonal symmetry was selected from the previous work of Kawai and Tsuchiya [] and Model B (Figure S1a) without tetragonal symmetry from the work of Deng et al [].…”

“…Previous calculations on K‐hollandite investigated its equation of states, elasticity, and phase transition from a tetragonal (hollandite I) to a monoclinic phase (hollandite II) [ Mookherjee and Steinle‐Neumann , ; Caracas and Ballaran , ; Mussi et al , ; Deng et al , ; Kawai and Tsuchiya , ]. The latter revealed that K‐hollandite exhibits strong shear wave anisotropy at high pressure, and [001] dislocations are favored under the shear stress of the subducted slabs [ Mussi et al , ].…”

High‐Pbehavior of anorthite composition and some phase relations of the CaO‐Al2O3‐SiO2 system to the lower mantle of the Earth, and their geophysical implications

“…Many high‐ P studies with different methods have been conducted to disclose the high‐ P behavior of feldspar [e.g., Ringwood et al , 1967; Liu , 1978; Yagi et al , 1994; Gautron et al , 1996; Akaogi et al , 2004a; Sueda et al , 2004; Nishiyama et al , 2005; Liu , 2006; Hirao et al , 2008; Mookherjee and Steinle‐Neumann , 2009; Deng et al , 2010, 2011], and significant progress has been achieved. As summarized in Liu et al [2010a], Ab (NaAlSi 3 O 8 ) breaks down to jadeite (Jd; NaAlSi 2 O 6 ) + Qtz at ∼2.5 GPa [ Birch and LeComte , 1960; Liu , 1978; Holland , 1980], which in turn transforms into Jd + Coe at ∼3 GPa, then into Jd + St at ∼9 GPa and finally into calcium ferrite‐structured NaAlSiO 4 (CF) + St at ∼22 GPa [ Liu , 1978; Yagi et al , 1994; Tutti et al , 2000; Deng et al , 2010].…”

“…At approximately 10 GPa, these three phases, Wd, Ky and St, recombine to form a hollandite‐type phase with the composition of KAlSi 3 O 8 (Holl‐I). At ∼23 GPa, the structure of Holl‐I is deformed and an isochemical new phase forms (Holl‐II) [ Sueda et al , 2004; Nishiyama et al , 2005; Ferroir et al , 2006; Deng et al , 2011]. Compared to the density of the mantle material at similar pressures, the density of the isochemical phase assemblages of San and Wd + Ky + Coe is generally low, that of the phase assemblages of Wd + Ky + St and Holl‐I relatively high, and that of Holl‐II broadly similar.…”