Copy the page URI to the clipboard
Cernok, Ana
(2015).
URL: https://epub.uni-bayreuth.de/2719/
Abstract
Silica, SiO2, exists in a wide range of pressure and temperature conditions, accommodating its simple chemical composition by numerous polymorphs. More than 30 stable or metastable phases of silica are known, most of which occur at ambient to moderate pressures (<9 GPa). Understanding polymorphism and densification mechanisms in silica is not only important from a geomaterial point of view, but it also has relevant implications in material sciences, technology and industry (glass, ceramics, etc.). The pathways of phase transitions are often influenced by thermodynamically metastable polymorphs that intermediate or hinder formation of thermodynamically stable phases.
This work is focused on understanding high pressure behaviour of two important silica minerals, coesite and cristobalite. Both of them are framework silicates comprised of fully polymerized corner sharing SiO4 tetrahedra. Coesite, thermodynamically stable above ca. 2.5 GPa and at temperatures in excess of 500 °C, hereafter referred to as coesite-I, is the densest known polymorph with silicon atoms tetrahedrally coordinated to oxygen. It can be found as a high-pressure mineral in rocks related to meteorite impact sites, in ultra-high pressure metamorphic rocks or in kimberlites (mantle derived rocks). Cristobalite is a high-temperature (> 1470 °C), low-pressure polymorph of silica that has a subordinate and rather exotic terrestrial occurrence among silica phases (volcanic rocks, chert, etc.), but it is the predominant SiO2 polymorph in various planetary materials (meteorites, lunar rocks, interplanetary dust particles, etc.). Exact pathways of pressure-induced transitions in coesite and cristobalite, as well as the structures of their high-pressure polymorphs have been poorly understood until now, despite being investigated by a number of studies over the past 30 years.
In this study, the response of coesite and cristobalite to compression was investigated at pressures exceeding 50 GPa using diamond-anvil cells by means of in situ Raman spectroscopy, synchrotron single-crystal X-ray diffraction, and analyses of the recovered samples by various techniques. Structural and Raman spectroscopic studies reveal that coesite-I (monoclinic C2/c, Z=16) undergoes two phase transitions (I->II->III) and does not become amorphous at least up to ~51 GPa. A reversible, displacive phase transition to coesite-II (P21/n) near 23 GPa is likely driven by the extreme shortening (0.05 Å or 3.2%) of the shortest and the most compressible Si-O bond, related to the stiff 180º Si-O-Si angle. The unit cell of the novel polymorph is doubled along the b-axis with respect to that of the initial coesite-I and contains all Si atoms in tetrahedral coordination. Further Si-O compression down to an extremely short distance of ~1.52 Å prompts subsequent structural changes, with the formation of a triclinic phase at ~31 GPa, coesite-III. The second transition (coesite-II tocoesite-III) is also reversible but with a large hysteresis. According to the abrupt change in Raman spectra this is likely a first order phase transition which leads to a very distinct structure with the lowest-symmetry. Despite all the efforts, the structure of the polymorph coesite-III remains unresolved. Samples recovered from the quasi-hydrostatic experiments carried out up to ~51 GPa, show the structure of the initial coesite, but those compressed between the diamond anvils (uniaxial stress) appear amorphous. The very short Si-O bond found in coesite-I and its high compressibility is a good example of how such bonds are restricted to the (almost) linear Si-O-Si geometry, and appear highly unfavorable in other Si-O-Si arrangements. Uncommon for other silica polymorphs, coesite-I demonstrates high-pressure behaviour governed by two simultaneous compressional mechanisms: polyhedral tilting along with Si-O-Si bond-angle reduction on the one hand and Si-O bond-length compression with polyhedral distortion on the other hand.
Regarding α-cristobalite, the study demonstrates that it responds differently to high pressures depending on the degree of the hydrostaticity. Under highly hydrostatic conditions, the initial structure of cristobalite is preserved. When the crystal experiences even slight stresses during an experiment, transformation sequence leads to cristobalite X-I at ~11 GPa – a monoclinic P21/n polymorph with silicon atoms in octahedral coordination. The structure and formation of this novel polymorph was a long-standing enigma up to now. The likely reconstructive transition that involves increase in coordination number of silicon from four in cristobalite to its six-fold coordinated polymorph does not require any thermal activation; however, the high-pressure polymorph cannot be preserved at ambient conditions. No other silica polymorph was found to transform to an octahedra-based structure on cold compression at such low pressures (~11 GPa) and this structure could be accommodated in a (quasi)-hydrostatic environment where temperature is not sufficient to form the thermodynamically stable stishovite. In non-hydrostatic conditions in the presence of uniaxial stress, cristobalite eventually transforms to seifertite-like SiO2, which is quenchable. Thus, according to our results, presence of seifertite may not always require the minimum shock pressures equal to that of thermodynamic equilibrium (~80 GPa) as it can be clearly formed at much lower pressures.
Both coesite and cristobalite follow the same densification path initially by undergoing a displacive phase transition to a slightly distorted structure of reduced symmetry. The most striking difference in response to compression of coesite and cristobalite is their reaction to (uniaxial) stress: coesite becomes amorphous when compressed in non-hydrostatic conditions; cristobalite, on the other hand, transforms directly to quenchable seifertite, a post-stishovite polymorph of silica. This may have important implications on occurrence and preservation of coesite and cristobalite in natural samples which have been exposed to non-hydrostatic compression, such as impact-related terrestrial rocks or shocked meteorites. Crystalline or amorphous metastable phases derived from coesite or cristobalite under high-pressure conditions are of particular interest because they are often used as potential tracers of peak transient pressures (stress) reached in processes such as impacts or faulting. The novel metastable polymorphs reveal compressional mechanism of silica minerals and set important constraints on the very complex phase diagram of SiO2.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)
