Discovery of coesite and stishovite in eucrite

Significance

Quartz and/or cristobalite in eucrite were transformed into denser minerals, coesite and stishovite, under transient high-pressure and high-temperature conditions. Coesite and stishovite probably formed simultaneously under pressures of similar magnitudes but under different temperature conditions. The expected age of the dynamic event that formed coesite and stishovite is ca. 4.1 Ga ago, which is inconsistent with the predicted formation age (ca. 1.0 Ga) of the impact basins on 4 Vesta.

Abstract

Howardite–eucrite–diogenite meteorites (HEDs) probably originated from the asteroid 4 Vesta. We investigated one eucrite, Béréba, to clarify a dynamic event that occurred on 4 Vesta using a shock-induced high-pressure polymorph. We discovered high-pressure polymorphs of silica, coesite, and stishovite originating from quartz and/or cristobalite in and around the shock-melt veins of Béréba. Lamellar stishovite formed in silica grains through a solid-state phase transition. A network-like rupture was formed and melting took place along the rupture in the silica grains. Nanosized granular coesite grains crystallized from the silica melt. Based on shock-induced high-pressure polymorphs, the estimated shock-pressure condition ranged from ∼8 to ∼13 GPa. Considering radiometric ages and shock features, the dynamic event that led to the formation of coesite and stishovite occurred ca. 4.1 Ga ago, which corresponds to the late heavy bombardment period (ca. 3.8–4.1 Ga), deduced from the lunar cataclysm. There are two giant impact basins around the south pole of 4 Vesta. Although the origin of HEDs is thought to be related to dynamic events that formed the basins ca. 1.0 Ga ago, our findings are at variance with that idea.

Howardite–eucrite–diogenite meteorites (HEDs) are the largest group among the achondrites. Although the origin of HEDs is still under debate (1, 2), the similarities between the reflectance spectra of HEDs and the spectra of one of the largest asteroids in the asteroid belt 4 Vesta and dynamic considerations indicate that HEDs originated from 4 Vesta (3–5). The Dawn mission supports this prediction. It has been revealed that many craters exist on 4 Vesta (6, 7), which suggests heavy meteoroid bombardment. The existence of a high-pressure polymorph in a shocked meteorite provides clear evidence for a dynamic event on its parent body (8). Some recent studies propose that 4 Vesta, similar to the Moon, might have suffered from late heavy bombardment (9–11). However, to date, no high-pressure polymorph has been found in HEDs. We now report clear evidence of high-pressure polymorphs of silica, coesite, and stishovite from eucrite.

We envisaged that some eucrites might contain high-pressure polymorphs because it is expected that the surface of 4 Vesta consists mainly of eucrite. We obtained one of the eucrites, Béréba, to clarify a dynamic event occurring on 4 Vesta, using the high-pressure mineral inventory. The Béréba sample used in this study has many shock-induced melt (hereafter referred to as shock melt) veins (Fig. 1A), implying that it was heavily shocked. Major constituent minerals in the host rock of Béréba are low-Ca pyroxene (Fs_59–63En_34–37Wo_2–3), augite (Fs_25–32En_29–31Wo_38–44), plagioclase (An_86–92Ab_7–14Or_0–1), silica, minor kamacite, ilmenite, chromite, and Ca-phosphate. Most of the low-Ca pyroxene has exsolution lamellae of augite. Plagioclase now transformed into maskelynite partly and/or completely. Flow-like textures appear in some maskelynite. Mixing between plagioclase and pyroxene occurs in the flow-like textures, suggesting that the feldspar was once melt quenched to maskelynite (12).

Fig. 1.

Back-scattered electron images of silica grains in Béréba. (A) Low-magnification image of a shock-melt vein. Quartz (and/or cristobalite) grains exist in and around the shock-melt veins. Feldspar partially transforms to maskelynite. (B) Silica grains with a network-like texture. Coesite, stishovite, and silica glass (+ minor quartz) coexist in the silica grain. (C) Silica grains with a lamellae-like texture. Stishovite and silica glass coexist in silica grains. (D) High-magnification image of the outlined section in C. Coe, coesite; Fd, feldspar; Pyx, pyroxene; Qtz, quartz; Si-gla, silica glass; Sti, stishovite.

In this study, the focus of our interest was silica. The silica grain is up to ∼300 μm across. The chemical compositions of the silica grains (especially the impurities present, such as Al) were measured by electron microprobe analysis (Table S1). Raman spectroscopy analyses indicate that the original silica grains are quartz and cristobalite. Representative Raman spectra are shown in Fig. S1. Some silica grains adjacent to or near the shock-melt veins have network-like textures (Fig. 1B). Raman spectroscopy analyses further indicate that the network-like textures include the high-pressure polymorphs of silica, coesite, and stishovite, along with quartz. Transmission electron microscopy (TEM) images indicate that the network-like texture consists of a fine-grained granular coesite assemblage (Fig. 2) and minor lamellar stishovite. Amorphous (or poorly crystallized) silica exists between coesite and stishovite grains. Lamellae-like textures are observed in some silica grains adjacent to or near the shock-melt veins (Fig. 1 C and D), similar to a transition texture from quartz (or cristobalite) to stishovite (13, 14). TEM images indicate that the silica grains with lamellae-like texture include lamellar stishovite (Fig. 3). We also investigated pyroxene and plagioclase in and around the shock-melt veins, but their high-pressure polymorphs (e.g., majorite, akimotoite, jadeite, and lingunite) were not detected.

Fig. 2.

TEM images of silica grain having a network-like texture (Fig. 1B). (A) High-angle annular dark-field image of the network-like texture. The network-like texture is a fine-grained granular coesite grain assemblage. A small amount of lamellar stishovite accompanies some network-like texture. (B) TEM image of the outlined section in A. Quartz crystals accompany most of the coesite grain assemblage. Coe, coesite; Fe, metallic iron; Qtz, quartz; Si-gla, silica glass.

Fig. 3.

TEM images of silica grain having a lamellae-like texture (Fig. 1D). (A) High-angle annular dark-field image of the lamellae-like texture, consisting of stishovite and silica glass. (B) TEM image of the outlined section in A. (C) Selected area electron diffraction pattern corresponding to stishovite. Si-gla, silica glass; Sti, stishovite.

High-pressure polymorphs of silica have been found in lunar meteorites, Martian meteorites, and carbonaceous chondrite (14–17), and in terrestrial impacted rocks (13, 18). Coesite is thermodynamically stable above ∼2.5 GPa (19). Stishovite can be easily synthesized in shock experiments (20, 21). On the other hand, the formation of coesite is not easily achieved in a dynamic process. This is because phase transformation from quartz to coesite is sluggish because of a high kinetic barrier (22–24). The formation mechanism of shock-induced coesite has been explained as follows (i): crystallization in the solid state from vitrified silica, or (ii) crystallization from silica melt (13, 25). Nanosized coesite grain assemblages accompanying minor quartz and/or stishovite grains occur in the network-like textures of silica grains in Béréba (Figs. 1B and 2). Coesite that occurs in lunar meteorites is also a nanosized crystal assemblage embedded in silica glass (17), which is suggestive of a quench crystal from silica melt. When quartz is deformed in a piston cylinder at 2.7–3.0 GPa under ambient temperature conditions, coesite forms along melted ruptures in the deformed quartz (26).

The network-like texture of Béréba indicates a flow-like texture; it would be a mixture of silica and plagioclase because small amounts of calcium and aluminum, as well as silicon, are contained in the texture. When a silica grain was shocked, it was ruptured and, simultaneously, melting took place because of friction along the ruptures. The crystallization of coesite from silica melt would have a lower kinetic barrier than a solid–solid phase transformation. Nanosized coesite grain assemblages would form from the melted silica. On the other hand, the silica grain around the melted silica was not heated, by friction, beyond its melting temperature. The unmelted silica portions indicate a smooth surface or lamellae-like texture, and consist of amorphous silica or stishovite. Static high-pressure synthetic experiments indicate that quartz and/or cristobalite transform to stishovite (or seifertite), or become amorphous, through a pressure-induced phase transformation at high pressure but under low-temperature conditions (27, 28). Stishovite in Béréba is formed through a solid–solid phase transition from original quartz/cristobalite. TEM observations with electron-beam irradiation indicate that stishovite vitrified immediately, implying that stishovite is very sensitive to heating. Some stishovite might become amorphous during adiabatic decompression.

The pressure conditions in a silica grain of Béréba would be homogeneous; whereas, the temperature conditions are heterogeneous because some portions are heated beyond the melting temperature and melting takes place along the fractures. The Clapeyron slope between coesite and stishovite, deduced from laboratory high-pressure static experiments, is positive (29). Accordingly, both coesite and stishovite probably formed under similar pressure conditions, but coesite crystallized under a higher temperature conditions from the silica melt, whereas stishovite formed under lower temperature conditions in a solid state. The coexistence of coesite and stishovite would be due to the significant temperature gradient in the silica grain.

The dynamic event that formed both coesite and stishovite is the most intense impact incident recorded in Béréba. We could constrain a pressure condition recorded in Béréba using high-pressure polymorphs. It is expected that the stable pressure field of stishovite depends on the alumina content, based on high-pressure synthetic experiments using alumina-bearing silica (30). However, the silica present in Béréba contains hardly any aluminum (<0.2 wt% as Al₂O₃; Table S1). We adopted a phase diagram deduced from static high-pressure synthetic experiments using pure silica to estimate a shock-pressure condition. Considering the phase diagram, the coexistence of coesite and stishovite suggests that the pressure condition recorded in Béréba would range from ∼8 to ∼13 GPa (29, 31).

Bulk-rock Pb–Pb isotope age indicates not only an older age (4.521 ± 0.0004 Ga) but also a younger age (ca. 4.1 Ga) (32). When shock-induced glass veinlets in the bulk rock are removed from a crushed bulk-rock sample, the relatively younger age is removed, implying that the relatively younger isotope ages are closely related to an event that formed the shock-melt vein (32). U–Pb systems are hosted in feldspar (32). Plagioclase in Béréba melted and quenched to maskelynite glass. Induced by a dynamic event, the U–Pb systems would be disturbed through the melting and quenching of plagioclase to maskelynite glass. Accordingly, the relatively younger isotope age (ca. 4.1 Ga) corresponds to a dynamic event that formed the shock-melt veins, including high-pressure polymorphs, and induced melting and quenching of plagioclase to maskelynite glass. It is expected that heavy meteorite bombardment occurred on the Moon around 3.8–4.1 Ga ago (33). The existence of high-pressure polymorphs in Béréba may support the idea that a parent body of HEDs also suffered from such a cataclysm, as deduced from ⁴⁰Ar–³⁹Ar isotope age distributions (9–11).

Two giant impact basins (Rheasilvia and Veneneia) on 4 Vesta are depicted by the Dawn mission. Model calculation and crater chronology obtained by the Dawn mission reveal that the giant impact basins would have formed around 1.0 Ga ago (4, 34). Fragments were launched from 4 Vesta by the giant impact-basin formations, and became the Vesta family in the asteroid belt. In addition, some of them fell into the Earth as HEDs. The dynamic events that formed two giant impact basins are the most catastrophic events to occur on 4 Vesta since it formed. On the other hand, the most intense dynamic event recorded in Béréba is ca. 4.1 Ga ago, which contradicts the launch time of HEDs from 4 Vesta. Most resetting timings of ⁴⁰Ar–³⁹Ar isotopic ages (9, 10) do not coincide with the HEDs formation model deduced from model calculation and crater chorology. Assuming that HEDs originate from 4 Vesta, HEDs would not originate from the impact event that formed the Rheasilvia or Veneneia basins.

Methods

The Béréba sample studied here was obtained through Bruno Fectay and Carine Bidaut of www.meteorite.fr. A polished Béréba chip sample was prepared for the study, and petrological observation was carried out with an optical microscope. The mineralogy was determined using a laser micro-Raman spectrometer, JASCO NRS-5100. A microscope was used to focus the excitation laser beam (green laser; 532 nm). We used a field-emission gun scanning electron microscope, JEOL JSM-71010, for fine textural observations. An accelerating voltage of 15 kV was used. Chemical compositions of minerals were determined using the wavelength-dispersive procedure with a JEOL JXA-8800M electron microprobe analyzer. Analyses were carried out using an accelerating voltage of 15 kV, a beam current of 10 nA and a defocused beam of 1–10 μm.

Slices for TEM observations were prepared using a focused ion beam (FIB) system, JEOL JEM-9320FIB. JEOL JEM-2010 and JEM-2100F transmission electron microscopes operating at 200 kV were used for conventional TEM observations and selected area electron diffraction. The JEOL JEM-2100F is equipped with a scanning TEM (STEM) mode and a JEOL energy dispersive X-ray spectroscopy (EDS) detector system. The chemical compositions of individual minerals were obtained by EDS under STEM mode. The compositions were corrected using theoretically determined k factors. The unused Béréba samples are stored at Department of Earth and Planetary Systems Science, Graduate School of Science, Hiroshima University, Japan.