Natural and experimental high-pressure, shock-produced terrestrial and extraterrestrial materials

Abstract

Hypervelocity impacts are among the fundamental phenomena occurring during the evolution of the solar system and are characterized by instantaneous ultrahigh pressure and temperature. Varied physicochemical changes have occurred in the building blocks of celestial bodies under such extreme conditions. The constituent material has transformed into a denser form, a high-pressure polymorph. The high-pressure polymorph is also thought to be the constituent of the deep Earth’s interior. Hence, experiments using a high-pressure and temperature generating apparatus have been conducted to clarify its crystal structure, pressure–temperature stability range, and transformation mechanisms. A natural high-pressure polymorph (mineral) is found from terrestrial and extraterrestrial rocks that experienced a hypervelocity impact. Mineralogists and planetary scientists have investigated high-pressure minerals in meteorites and rocks near terrestrial craters over a half-century. Here, we report brief reviews about the experiments producing high-pressure polymorphs and then summarize the research histories of high-pressure minerals occurring in shocked meteorites and rocks near terrestrial craters. Finally, some implications of high-pressure minerals found in impact-induced shocked rocks are also mentioned.

Graphic abstract

1 Introduction

High-pressure minerals have considered the major constituents of the Earth’s deep interior, and their physicochemical properties and transformation mechanisms are essential parameters to understand the structure and dynamics of the Earth’s depths. Therefore, mineral physicists who work on high-pressure minerals have been making long-standing efforts to clarify the structural changes and phase equilibria using a multianvil apparatus and a laser-heated diamond anvil cell (LHDAC).

Mineralogists and planetary scientists have tried to find high-pressure minerals in meteorites (Fig. 1) and rocks near terrestrial impact craters because we cannot get high-pressure minerals directly from the Earth’s deep interior except for inclusions in diamond crystals (Moore and Gurney 1985; Walter et al. 2011; Pearson et al. 2014) and mantle xenoliths (Collerson et al. 2000). Indeed, the first high-pressure mineral was discovered in a terrestrial impact crater; Chao et al. (1960) found a high-pressure polymorph of SiO₂, then named coesite, from the Barringer crater in the USA, and Chao et al. (1962) then found a second SiO₂ polymorph, named stishovite, from the same crater (Fig. 2). A. El Goresy (1934–2019), who was a world-leading mineral physicist/meteoriticist working on shock metamorphism, discovered a new allotropic form of carbon, named chaoite in memory of the contributions of E. C. T. Chao for his pioneer works, from the Rice crater in Germany (El Goresy and Donnay 1968).

Fig. 1

A heavily shocked meteorite (The Monze L6 ordinary chondrite). Black-colored veins (white-colored arrows) are shock-melt veins

Fig. 2

The accumulated number of a new high-pressure mineral by year. The discovery of major minerals is shown in the figure. The introduction years of instruments are also shown. Coe: coesite, Sti: stishovite, Rwd: ringwoodite, Wds: wadsleyite, Maj: majorite, Aki: akimotoite, Bdg: bridgmanite, Sei: seifertite. cXRD: conventional X-ray diffraction, TEM: transmission electron microscopy, FIB: focused ion beam, sXRD: synchrotron X-ray diffraction, EBSD: electron backscattered diffraction

The first discovery of a high-pressure mineral from a meteorite is the (Mg,Fe)₂SiO₄ ringwoodite (Binns et al. 1969), and the second and third cases are the (Mg,Fe)SiO₃ majorite (Smith and Mason 1970) and wadsleyite (Price et al. 1983) (Fig. 2; Table 1). High-pressure minerals in shocked rocks are heterogeneously distributed on a micrometer scale, and their crystal sizes are nano–micrometers across. In most cases, the occurrence of high-pressure minerals in shocked meteorites is restricted in the vicinity of shock-induced melt veins (shock-melt veins) and melt pockets up to several millimeters in width (Fig. 1). Technical difficulties on the characterization of such minute crystals shrunk further efforts to search for new high-pressure minerals and caused a long stagnation since the 1970s (Fig. 2).

Table 1 Inventory of high-pressure minerals

The breakthrough was brought by the introduction of a transmission electron microscope equipped with an X-ray energy-dispersive spectrometer (TEM-EDS) (Fig. 2). TEM-EDS analysis enables textural observations, crystallographic scrutinies, and chemical composition analyses at nano–microscales. A. El Goresy, M. Chen, T. Sharp, and their colleagues showed a beautiful ringwoodite grain having dense stacking faults in shock-melt veins of the Sixiangkou L6 ordinary chondrite by TEM (Chen et al. 1996). Subsequently, Sharp et al. (1997) and Tomioka and Fujino (1997) almost simultaneously found natural (Mg,Fe)SiO₃ with the ilmenite structure (named akimotoite later) from the Tenham L6 ordinary chondrite by using TEM-EDS analysis. In the same contribution, these authors also reported an electron diffraction pattern of (Mg,Fe)SiO₃ with the perovskite structure (named bridgmanite later by another set of authors) along with akimotoite (Tomioka and Fujino 1997). A. El Goresy and his colleagues also found a silica mineral with scrutinyite (α-PbO₂)-type structure from the shergottite Shergotty by TEM-EDS analysis (Sharp et al. 1999). We cannot directly observe a high-pressure polymorph of (Mg,Fe)₂SiO₄, poirierite, without high-resolution TEM analysis because of its complex and ultrafine crystals within the host olivine (Tomioka et al. 2021).

The second breakthrough was caused by the introduction of in situ analysis techniques: (i) micro-fabrication and -pickup technology, a focused ion beam (FIB) system, (ii) focused synchrotron X-ray diffraction (sXRD) techniques, and iii) electron backscattered diffraction (EBSD) analysis. An ultrathin film (thickness: ~ 100 nm) had been prepared for TEM-EDS analysis and was prepared conventionally by Ar-ion-milling and ultramicrotomy. However, a target containing high-pressure minerals for TEM-EDS analysis is submicron-sized. We must minimize the destruction of a valuable meteorite sample during the pickup process of such a small portion. The introduction of an FIB system overcame the difficulty (e.g., Miyahara et al. 2008; Wirth 2009). sXRD analysis, which uses a high-power X-ray beam, is a powerful tool to identify the crystal structure of submicron-sized material. SEM-EBSD analysis can identify minerals in meteorites on the surface of a polished petrographic section in a short time without sample destruction. The number of new high-pressure minerals exponentially increased since the introduction of these new analysis techniques (Fig. 2). Furthermore, the application of laser micro-Raman spectroscopy makes the characterization of high-pressure minerals in shocked rocks very rapid and easy. By the means of a 1-µm diameter laser beam, such a technique allows the identification of fine-grained minerals in polished petrographic thin sections within several 10 s, thus enabling systematic investigations of high-pressure minerals (e.g., Ohtani et al. 2004; Miyahara et al. 2020). However, meticulous interpretation of a Raman shift is necessary for identification because some high-pressure minerals have similar Raman shifts (e.g., Baziotis et al. 2018). Cross-check by another analysis technique should be carried out when necessary.

Here we review the syntheses of high-pressure minerals and their discoveries in natural samples in the last six decades. In this review, a high-pressure mineral is termed as a mineral that is stable only at the Earth’s mantle pressure conditions (above ~ 1 GPa) as listed in Table 1. Their formation mechanisms, representative phase diagrams, and typical diagnostic Raman spectra are summarized in Gasparik (2003) and Tomioka and Miyahara (2017). Concise but comprehensive introductions of shock-produced high-pressure minerals are also available in Sharp and DeCarli (2006), Morrison and Hazen (2021) and on our database website (https://sites.google.com/site/highpmineral).

2 High-pressure minerals in meteorites

2.1 High-pressure polymorphs of (Mg,Fe)₂SiO₄

Olivine [(Mg,Fe)₂SiO₄] is one of the major constituents of meteorites, and its high-pressure polymorphs are found from ordinary chondrites, lunar meteorites, Martian meteorites, and carbonaceous chondrites (Tomioka and Miyahara 2017). High-pressure polymorphs for olivine (α-phase), spinelloid-type (Mg,Fe)₂SiO₄ (β-phase), spinel-type (Mg,Fe)₂SiO₄ (γ-phase), spinelloid-type Mg₂SiO₄ (ε-phase), and orthosilicate Mg₂SiO₄ (ζ-olivine) are reported. The fayalite components of olivine for H, L, and LL group ordinary chondrites are 16–23, 23–26, and 27–32 mol%, respectively (Weisberg et al. 2006). The fayalite components of olivine in Martian meteorites and carbonaceous chondrites are widespread and higher than ordinary chondrites and lunar meteorites. Hence, Fe-rich high-pressure polymorphs of (Mg,Fe)₂SiO₄ dominate in Martian meteorites and carbonaceous chondrites compared with other meteorites.

2.1.1 β-(Mg,Fe)₂SiO₄: Wadsleyite–asimowite

Ringwood and Major (1966a) conducted high-pressure experiments in the Mg₂SiO₄–Fe₂SiO₄ system using the early-stage Bridgman anvil apparatus at 12–20 GPa and 900 °C to confirm an olivine-spinel transition. Mg₂SiO₄ with a spinelloid structure was first reported as a distorted spinel phase in their experimental samples based on X-ray diffraction (XRD) analysis (Ringwood and Major 1966a). The crystal structure of the distorted spinel phase was better studied four years later and regarded as a new orthorhombic phase (β-Mg₂SiO₄) 8% denser than forsterite (Ringwood and Major 1970).

Natural β-Mg₂SiO₄ with a quench texture was discovered in the black-colored shock-melt veins of the Tenham and Peace River L6 ordinary chondrites by TEM observations (Putnis and Price 1979). The mineral was named wadsleyite after A. D. Wadsley (Price et al. 1983). Later systematic investigations revealed that, in most cases, wadsleyites in ordinary chondrites have crystallized from a shock-produced chondritic melt simultaneously transformed by a solid-state reaction from olivine entrained in the chondritic melt (Miyahara et al. 2020). The former always coexists with low-Ca pyroxene, and the latter occurs at the grain boundaries or fractures in the olivine grains (Fig. 3).

Fig. 3

Backscattered electron (BSE) images of wadsleyite. a A chondritic fragment in the shock-melt vein of Yamato (Y) 000973 L5 ordinary chondrite. b A high-magnification image of a white-colored box in a. A part of olivine in the fragment has been replaced with polycrystalline wadsleyite. The matrix of the shock-melt vein consists of wadsleyite and low-Ca pyroxene. Ol: olivine, Pyx: Low-Ca pyroxene, Wds: wadsleyite, Jd: jadeite

Although pure Fe₂SiO₄ with a spinelloid structure (β-Fe₂SiO₄) has not been synthesized, Finger et al. (1993) obtained single wadsleyite crystals with 40% of fayalite component at 15.2 GPa and 1973 K. A mineral with Fe dominance over Mg in the octahedral sites was found as inclusion in the shock-induced Fe–Ni droplets of the Suizhou L6 ordinary chondrite and the Quebrada Chimborazo 001 CB carbonaceous chondrite and was named asimowite after P. D. Asimow (Bindi et al. 2019).

2.1.2 γ-(Mg,Fe)₂SiO₄: Ringwoodite–ahrensite

The Mg₂SiO₄–Fe₂SiO₄ solid solution with a spinel structure (γ-phase) in the compositional range of Fe/(Mg + Fe) = 0.15–0.75 was reported experimentally by Ringwood and Major (1966a). As for the end-member composition, γ-Fe₂SiO₄ was first synthesized at 3 GPa, 660 °C by a squeezer apparatus (Ringwood 1958a), while the convincing evidence for γ-Mg₂SiO₄ was first reported later via synthesis at 25 GPa and 1000 °C by a Kawai-type multianvil apparatus (Ito et al. 1974). Mg₂SiO₄ with a spinel structure (γ-Mg₂SiO₄) was also obtained first experimentally by Ringwood and Major (1966a). Binns et al. (1969) found purple/blue isotropic grains in the shock-melt vein of the Tenham L6 ordinary chondrite by optical microscopic observations and obtained a powder XRD pattern from the grains. The latter could be assigned to γ-Mg₂SiO₄ and the mineral was named ringwoodite after A. E. Ringwood (Binns et al. 1969). Jeanloz (1979) claimed that ringwoodite is merely a complex mineral aggregate, misidentified as a spinel phase. Against the criticism, Putnis and Price (1979) presented conclusive evidence for the existence of ringwoodite, as they provide a TEM image of a single ringwoodite crystal in the shock-melt vein of the Tenham chondrite.

Most ringwoodite in shocked meteorites is a polycrystalline assemblage (across < 1–2 μm) (Xie and Sharp 2004; Xie et al. 2006; Chen et al. 2007; Miyahara et al. 2010; Sharp et al. 2015) (Fig. 4a). Some olivine-ringwoodite (or ahrensite) assemblage composed of a ringwoodite (Rwd) (or ahrensite (Ahr)) sheet with a specific crystallographic orientation to preserve the close-packed oxygen layers of both structures—(100)_Ol//{111}_Rwd/Ahr—is also found in an olivine (Ol) grain in contact with a shock-melt vein of the Yamato (Y)-791384 L6 ordinary chondrite (Miyahara et al. 2010) (Fig. 4b) or in the shock region of a grain of the Khatyrka CV3 carbonaceous chondrite (Hollister et al. 2014).

Fig. 4

Bright-field (BF)-TEM images of ringwoodite. a A ringwoodite crystal with a dense stacking fault in the Y-790729 L6 ordinary chondrite. b An olivine grain replaced with ringwoodite in the Y-791384 L6 ordinary chondrite. A part of the olivine grain in contact with the shock-melt vein is replaced with polycrystalline ringwoodite. Also, thin layers of ringwoodite occur in the olivine grain. The (100) of olivine is parallel with the {111} of ringwoodite. c A ringwoodite crystal with a dendritic texture in the Y-74445 L6 ordinary chondrite. Ol: olivine, Rwd: ringwoodite

γ-Ni₂GeO₄ as an analog of Fe₂SiO₄ with a spinel structure (γ-Fe₂SiO₄) was synthesized by Ringwood (1958b). Natural γ-Fe₂SiO₄ was discovered in the shock-melt vein of the Umbarger L6 ordinary chondrite by TEM observations (Xie et al. 2002b). Later, it was also found from the melt-pocket of the olivine-phyric shergottite Tissint by a combined powder XRD and EBSD analysis and named ahrensite after T. Ahrens (Ma et al. 2016). Like ringwoodite, most ahrensite grains are also polycrystalline assemblages, and some olivine-ahrensite assemblages consist of an ahrensite sheet with a specific crystallographic orientation, namely (001)_Ol//{111}_Ahr (Miyahara et al. 2016). Unlike wadsleyite, most ringwoodite–ahrensite have formed through a solid–solid-state transformation in ordinary chondrites and Martian meteorites (Walton et al. 2014; Ma et al. 2016; Miyahara et al. 2016, 2020; Tomioka and Miyahara 2017). In contrast, some ringwoodites in ordinary chondrites have formed from olivine melt or chondritic melt (Xie et al. 2002b; Miyahara et al. 2008, 2009, 2020). Ringwoodite crystallized from a chondritic melt has a dendritic texture (Fig. 4c).

An interesting point is that ringwoodite–ahrensite exhibits the normal spinel structure, with Mg(Fe) in the octahedral A site and Si in the tetrahedral B site. Through A and B site-disorder, the inverse spinel has four-coordinated A cations and the six-coordinated site hosts a mixture of A and B cations. This process affects the density and impedance contrast across the boundaries in the transition zone and seismic-wave velocities in this portion of the Earth. Bindi et al. (2018a) reported the first synthesis at high pressure (20 GPa) and temperature (1600 °C) of a Cr-bearing ringwoodite with a completely inverse-spinel structure. The chemical, structural, and computational analysis confirmed the stability of inverse ringwoodite. The degree of inversion in these minerals is crucial, as discussed by Ma et al. (2019c), because it could influence the inferred conditions of the meteorite during shock-melt cooling.

When an oceanic plate subducts along a trench into the Earth’s deep interior, hydrous minerals such as clay minerals and hydroxides are transported into the Earth’s depths, resulting in the formation of dense hydrous minerals. Many kinds of hydrous high-pressure polymorphs such as δ-AlOOH have been synthesized through high-pressure and -temperature experiments (e.g., Ohira et al. 2014; Bindi et al. 2020d). High-pressure experiments prove that nominal anhydrous ringwoodite and wadsleyite can also contain a small amount of water in their structures (< 2–3 wt.% H₂O) (e.g., Kohlstedt et al. 1996; Inoue et al. 1998).

A natural hydrous ringwoodite (> 1.4 wt.% of H₂O) was found as an inclusion in terrestrial diamond (Pearson et al. 2014). A hydrous ringwoodite (< 1132 ppm H₂O) is also found near the melt-pocket of the shergottite Tissint (Hallis et al. 2017). Mars explorations by rovers and landers indicate that water has existed extensively on the subsurface of Mars (e.g., Bandfield 2007). Some impact events might occur under hydrous conditions. The shergottites NWA 7397 and 10416 have evidence for both shock metamorphism and aqueous alteration (Herd et al. 2016; Yoshida et al. 2021). Hydrous high-pressure minerals may have been overlooked in shocked meteorites.

2.1.3 ε-Mg₂SiO₄: Poirierite

Based on the topological analyses of the crystal structures of the olivine polymorphs in shocked meteorites, shear mechanisms without long-range ionic diffusion are posited to promote their polymorphic transformations. The transformation model also predicted the possible occurrence of an intermediate spinelloid structure (ε-phase) (Madon and Poirier 1983). After the prediction over three decades, the ε-phase was found as metastable intergrowth within ringwoodite grains in the shock-melt vein of the Tenham L6 ordinary chondrite by meticulous high-resolution (HR)-TEM observations (Tomioka and Okuchi 2017). The ε-phase was also found within ringwoodite and wadsleyite grains from the Suizhou L6 and Miami H5 ordinary chondrites, respectively. Single-crystal XRD analysis of the ε-phase in the Suizhou L6 ordinary chondrite, as well as results of first-principles’ calculations, confirmed that the ε-phase has an orthorhombic unit cell with the smallest dimensions among the preexisting spinelloid structures, and this phase was named poirierite after J.-P. Poirier (Tomioka et al. 2021) (Fig. 5). The topotaxial intergrowths of poirierite with wadsleyite/ringwoodite in shocked chondrites support the idea proposed by Madon and Poirier (1983) where poirierite is a way point in the shear transformations between olivine and wadsleyite/ringwoodite (e.g., olivine becomes ringwoodite via poirierite).

Fig. 5

TEM images of poirierite in the Miami H5 ordinary chondrite. a BF-TEM image of a wadsleyite grain partly replaced by poirierite. b High-resolution (HR) TEM image of poirierite. Lattice fringes with 0.8 nm spacings correspond to d₀₀₁. Wds: wadsleyite, Poi: poirierite

2.1.4 ζ-olivine

ζ-olivine was found in the black-colored shocked-olivine in the chassignite Northwest Africa (NWA) 2737 and lherzolitic shergottite NWA 1950 (Beck et al. 2006; van de Moortèle et al. 2007). Based on HRTEM observations and first-principles’ calculations, the ζ-olivine consists of quasi ideally close-packed oxygen sublattice with disordered cation sites and is regarded as a metastable high-pressure form of α-(Mg,Fe)₂SiO₄ produced at high-pressure and relatively low-temperature conditions (Guyot and Reynard 1992; Beck et al. 2006; van de Moortèle et al. 2007). A mineral name has not been given to the ζ-olivine. Takenouchi et al. (2019) reproduced the black-colored shocked-olivine that potentially contains the ζ-phase by shock recovery experiments.

2.1.5 High-pressure dissociation of (Mg,Fe)₂SiO₄

Phase equilibrium diagram in the system Mg₂SiO₄-Fe₂SiO₄ predicts that olivine dissociates into bridgmanite + magnesiowüstite (+ stishovite) around 23–25 GPa (Ito and Takahashi 1989), although the dissociation reaction depends on temperature and chemical composition. Miyahara et al. (2011b) found evidence for the dissociation reaction in olivine grains in contact with the shock-melt vein of the olivine-phyric shergottite Dar al Gani (DaG) 735. Granular bridgmanite + magnesiowüstite (Fig. 6a) and lamellar bridgmanite + magnesiowüstite (Fig. 6b) are found in the dissociated olivine grains (with bridgmanite almost vitrified). In most cases, Martian meteorites are heavily shocked compared with ordinary chondrites (e.g., El Goresy et al. 2013; Miyahara et al. 2020). Hence, the high-pressure dissociation of olivine is found mainly in shocked Martian meteorites (Fritz and Greshake 2009; Walton et al. 2014; Miyahara et al. 2019). On the other hand, Bindi et al. (2020c) found Fe-rich silicate perovskite (hiroseite) as the dissociation products of olivine in the Suizhou chondrite L6 ordinary chondrite. High-pressure experiments suggest that olivine melts incongruently into magnesiowüstite and liquid above 8 GPa and 2100 °C (Kato et al. 1998; Ohtani et al. 1998). Natural dissociation of olivine by incongruent melting is found from the Kamargaon L6 ordinary chondrite and Apollo 15 regolith breccia (Tiwari et al. 2021; Satta 2021).

Fig. 6

BF-TEM images of dissociated olivine in the shergottite DaG 735. a Granular bridgmanite + magnesiowüstite, b lamellar bridgmanite + magnesiowüstite. Most bridgmanite have been vitrified. Bdg: bridgmanite, Mw: magnesiowüstite

2.2 High-pressure polymorphs of (Mg,Fe,Ca)SiO₃

2.2.1 (Mg,Fe)SiO₃-Mg₃Al₂Si₃O₁₂: Majorite-pyrope

The pyroxene is the second major constituent of chondritic meteorites and the Earth’s upper mantle. Because of the four main end-member components of pyroxenes (enstatite: MgSiO₃, ferrosilite: FeSiO₃, diopside: CaMgSi₂O₆, and hedenbergite: CaFeSi₂O₆), the pressure–temperature phase equilibria of pyroxene are more complicated than that of olivine. A natural high-pressure polymorph of the (Mg,Fe)SiO₃ pyroxene was first discovered in ordinary chondrites. Mason et al. (1968) discovered aggregates with an olivine composition in the Coorara L6 ordinary chondrite using powder XRD and identified their crystal structure as a garnet structure. The garnet composition was initially misidentified because of overlaps of their neighboring ringwoodite grains. A more careful examination showed that their actual chemical composition is close to (Mg,Fe)SiO₃ (Smith and Mason 1970). Two years before the discovery, the mineral was also synthesized from glass with an Al-bearing MgSiO₃ composition at 20 GPa and ~ 900 °C (Ringwood and Major 1966b). Generally, garnet has only aluminum in its oxygen 6-coordinated sites, while the newly found garnet has a unique characteristic when the structure has Si > 3 atoms per formula unit. Indeed, by the Tschermak substitution (2Al³⁺ ⇔ Mg²⁺  + Si⁴⁺) in the octahedral sites on the MgSiO₃-Mg₃Al₂Si₃O₁₂ (pyrope) join, we can find octahedral silicon in this structure, which is symptomatic of a very high-pressure regime. The new high-pressure mineral found in Coorara was named majorite (Smith and Mason 1970) after A. Major. Al-rich majorite is one of the dominant phases in the matrices of shock-melt veins (e.g., Chen et al. 1996; Miyahara et al. 2020). According to the phase diagram of the Allende CV3 carbonaceous chondrite, Al-rich majorite is thought to have crystallized as a liquidus phase from chondritic melt in the pressure range of 14–26 GPa (Agee et al. 1995; Chen et al. 1996).

Al-rich majorite has a cubic symmetry as usual for silicate garnets, while almost Al-poor/free majorite in the system MgSiO₃–FeSiO₃ has a tetragonal distorted symmetry caused by the ordering of Mg and Si in the octahedral sites (e.g., Kato and Kumazawa 1985; Angel et al. 1989; Heinemann et al. 1997; Nakatsuka et al. 1999; Tomioka et al. 2002). Natural Ca-rich tetragonal majorite as a dissociation product of subcalsic pyroxene was found in the host-rock fragments in shock-melt veins in the Tenham L6 ordinary chondrite (Xie and Sharp 2007; Tomioka et al. 2016). Polycrystalline low-Ca tetragonal majorite was also found in the same chondrite (Fig. 7) (Tomioka et al. 2016). Following a comparison between the degrees of cation ordering in synthetic and natural majorites, the cooling rate of shock-melt veins during shock metamorphism was constrained to be 10³ °C/s (Tomioka et al. 2002, 2016).

Fig. 7

BF-TEM image of a monomineralic aggregate of tetragonal majorite in the Tenham L6 ordinary chondrite. The majorite grains formed in a solid-state transformation of low-Ca pyroxene entrapped into a shock-melt vein

Majorite, which was formed by melt crystallization and solid-state transformation, has been reported in many chondrites and Martian meteorites (shergottite) (Tomioka and Miyahara 2017). As for terrestrial impact structure, the Ries crater also yielded Al-rich majorites with various amounts of Na, Fe, K, Ca, and Fe in shocked amphibolite (Stähle et al. 2010, 2017). Majorite was crystallized from amphibolite melt or chlorite-rich melt induced by an impact.

2.2.2 MgSiO₃–FeSiO₃: Akimotoite-hemleyite

Trigonal MgSiO₃ phase with the ilmenite structure phase was synthesized at 17.5–27.5 GPa and 1000–1400 °C by an LHDAC (Liu 1976) following the finding of hexagonal MgSiO₃ by Kawai et al. (1974). About 20 years after the syntheses, the natural occurrence of the phase was reported in the shock-melt veins of the Acfer 040 L5–6 and the Tenham L6 ordinary chondrites based on TEM observations (Sharp et al. 1997; Tomioka and Fujino 1997). The matrix of the shock-melt veins in Acfer 040 consists mainly of MgSiO₃-rich amorphous grains (~ 2 µm in size), ringwoodite, and MgSiO₃-ilmenite. The MgSiO₃-ilmenite occurs as plate-like grains and contains considerable amounts of the Na₂O, Al₂O₃, and FeO components (Sharp et al. 1997), while the MgSiO₃-rich amorphous material contains significant amounts of Na₂O, Al₂O₃, CaO, and FeO. The assemblages and chemical compositions of high-pressure minerals suggest that they have crystallized from shock-induced melt over 26 GPa (Sharp et al. 1997). In Tenham, clinoenstatite [(Mg,Fe)SiO₃], fragments of the host rock entrained in the shock-melt veins, is associated with aggregates of submicron-sized grains and no other minerals in their interstices. Their chemical composition is identical to that of the host clinoenstatite (Tomioka and Fujino 1997) (Fig. 8). The ilmenite phase is thought to have formed directly from the host clinoenstatite without melting. The MgSiO₃-ilmenite was named akimotoite after S. Akimoto (Tomioka and Fujino 1999). Akimotoite has been found in chondrites and shergottites (Xie and Sharp 2004; Imae and Ikeda 2010; Tomioka and Miyahara 2017), whereas they have not been reported in impact craters.

Fig. 8

BF-TEM image of akimotoite (Aki) in the Tenham L6 ordinary chondrite. The akimotoite grains formed in a solid-state transformation from low-Ca pyroxene (clinoenstatite: Cen) with (100) twin lamellae in the host rock entrapped into a shock-melt vein

Recently, the Fe-rich ilmenite with Fe/(Mg + Fe) = 0.56 was also discovered in an unmelted portion of the heavily shocked L6 ordinary chondrite Suizhou associated with olivine, clinoenstatite, and Fe-bearing pyroxene (Bindi et al. 2017). The occurrence suggests that the new phase formed by a Fe-rich clinoenstatite—formed by the thermal metamorphism in its parent body—was subsequently shock-transformed to Fe-rich akimotoite. The mineral was named hemleyite after R. J. Hemley (Bindi et al. 2017). So far, the FeSiO₃ solubility into akimotoite increases up to ~ 10 mol% with increasing pressure from 20 to 25 GPa at 1100 °C based on a phase equilibrium study (Ito and Yamada 1982), whereas at the FeSiO₃ content above 10 mol%, bridgmanite and majorite are stable rather than akimotoite at a higher pressure and lower temperature, respectively (Ohtani et al. 1991). Therefore, hemleyite would be metastably formed during a short pressure and temperature excursion in the shock event.

2.2.3 MgSiO₃–FeSiO₃: bridgmanite-hiroseite

Silicate perovskite has been first synthesized from pyrope at 30 GPa and above 800 °C in the Al-bearing system by an LHDAC (Liu 1974). Subsequently, pure MgSiO₃ perovskite was synthesized at 30 GPa and ~ 1000 °C by an LHDAC (Liu 1975) and 28 GPa and ~ 1000 °C by a Kawai-type multianvil apparatus (Ito and Matsui 1978). These phases have an orthorhombic symmetry with the same structure as perovskite (CaTiO₃). The mineral is of particular interest to geophysicists because the phase occupies ~ 77 vol% of the Earth’s lower mantle (Irifune 1993).

In nature, (Mg,Fe)SiO₃ grains (~ 0.3 µm in size) adjacent to a clinoenstatite grain were identified in the Tenham L6 ordinary chondrite to be crystalline as a perovskite-structure phase using TEM (Tomioka and Fujino 1997). The perovskite phase has the same chemical composition as clinoenstatite in the host rock, which is considered to have formed in a solid phase transition as well as akimotoite in the Tenham chondrite. Fine-scale intergrowth of (Mg,Fe)SiO₃-perovskite and magnesiowüstite, as post-ringwoodite dissociation products of olivine, were also found in the shergottite DaG 735 (Miyahara et al. 2011b). The presented TEM works did not give full crystallographic data on the perovskite phase because of its large susceptibility to electron beam radiation. However, synchrotron X-ray micro-diffraction finally gave data to determine the unit cell of (Mg,Fe)SiO₃-perovskite in the Tenham chondrite. The mineral was named bridgmanite after P. W. Bridgman (Tschauner et al. 2014).

Recently, a Fe-rich silicate perovskite phase with Fe/(Mg + Fe) = 0.59 was discovered in a shock-melt vein in the Suizhou chondrite L6 ordinary chondrite based on XRD, TEM, and micro-Raman spectroscopy (Bindi et al. 2020c). Interestingly, Fe nanoparticles less than 100 nm in size were found embedded in the host perovskite phase that has both ferric and ferrous iron. That is the first evidence for a subsolidus charge disproportionation reaction in natural high-pressure minerals: 3Fe²⁺  → Fe⁰ + 2Fe³⁺. The mineral was named hiroseite after K. Hirose (Bindi et al. 2020c). So far, bridgmanite has not been found in terrestrial impact craters.

It is noteworthy that a bridgmanite-like mineral was discovered in the Acfer 094 carbonaceous chondrite (ungrouped C2). The mineral is ~ 300 nm in size and has an oxygen isotope anomaly (Vollmer et al. 2007). Electron diffraction patterns of the mineral indicate a superstructure of the orthorhombic perovskite structure. A possible explanation of the formation process of the novel mineral is that presolar MgSiO₃-pyroxene particles floating in interstellar space have transformed into the perovskite-like phase by interstellar shock waves before incorporation into its parent body.

CaSiO₃ with the perovskite structure was also confirmed at 16 GPa and 1500 °C by an LHDAC (Liu and Ringwood 1975). It is not an atmospheric pressure quenchable phase but vitrified during decompression. As for shocked meteorites, CaSiO₃-rich amorphous phase coexisting with Ca-rich majorite was reported in a shock-melt vein of the Y-75100 H6 ordinary chondrite (Tomioka and Kimura 2003). The diopside in the host rock would have initially dissociated into CaSiO₃ with the perovskite structure plus Ca-rich majorite during prograde shock metamorphism, and then, the perovskite phase became amorphous during the subsequent retrograde stage. Crystalline CaSiO₃ perovskite with an orthorhombic cell was found in a terrestrial environment as inclusion in a super-deep diamond from South African Cullinan kimberlite (Nestola et al. 2018). More recently, cubic CaSiO₃ perovskite was also reported in a diamond from Botswanan Orapa kimberlite, and then named davemaoite after H.-K. Mao (Tschauner et al. 2020a).

2.2.4 Other high-pressure minerals with the pyroxene compositions

The MgSiO₃ component has various pyroxene-structure polymorphs with different crystal symmetries caused by configurational changes of the SiO₄ tetrahedral chains at high pressures and temperatures: low-temperature clinoenstatite (space group P2₁/c), orthoenstatite (Pbca), protoenstatite (Pbcn), high-temperature clinoenstatite (C2/c), and high-pressure clinoenstatite (C2/c) (Gasparik 1990). High-pressure clinoenstatite was found by in situ XRD study at pressures above ~ 7 GPa (Angel et al. 1992; Shinmei et al. 1999); however, it cannot be recovered to the ambient condition as it inverts to low-temperature clinoenstatite during decompression. Recently, evidence for high-pressure clinoenstatite was found in the shock-melt vein of the Grove Mountains 022,115 L6 ordinary chondrite by HRTEM (Guo et al. 2020). The occurrence of high-pressure clinoenstatite embedded in amorphous silicate and associated with majorite and ringwoodite suggests that the assemblage was initially crystallized as majorite at 20–23 GPa and over 1800 °C and was subsequently back-transformed to high-pressure clinoenstatite during retrograde shock metamorphism. High-pressure clinoenstatite would have been metastably frozen because of extremely rapid cooling (Guo et al. 2020).

A novel Fe, Al-bearing MgSiO₃-rich phase was reported in the silicate glass matrix of shock-melt veins in Tenham L6 ordinary chondrite by TEM (Xie et al. 2011). The mineral has an acicular morphology (≤ 0.5 µm in length) and its electron diffraction pattern indicates an olivine structure. A possible explanation to reconcile the discrepancy between the olivine structure and the pyroxene stoichiometry of the phase is the presence of cation vacancies at the M site in the olivine structure. The estimated density of 3.32 g/cm³ is slightly larger than those of olivine and pyroxenes; therefore, the mineral would have metastably crystallized from chondritic melt under high pressure during rapid cooling (Xie et al. 2011).

2.3 High-pressure polymorphs of feldspar

2.3.1 Jadeite + silica

Albitic plagioclase is one of the major constituents of equilibrated ordinary chondrites. Albite (NaAlSi₃O₈) dissociates into jadeite (NaAlSi₂O₆) + quartz (SiO₂), jadeite + coesite (SiO₂), and jadeite + stishovite with increasing pressure and temperature (Liu 1978). However, these dissociation products do not occur in the shock-melt veins of equilibrated ordinary chondrites. In most cases, jadeite solely occurs in originally plagioclase grains (e.g., Ohtani et al. 2004; Ozawa et al. 2009; Miyahara et al. 2013b). Jadeite shows several kinds of crystal habits, namely, dendritic, granular, stringlike, and polycrystalline (Ozawa et al. 2009; Miyahara et al. 2020), which depend on the formation processes. Even in the large-scale shock metamorphism, pressure duration is not long enough (< several seconds) to form coesite or stishovite along with jadeite because of its sluggish nucleation (Kubo et al. 2010). However, the jadeite + coesite assemblage has replaced an albitic plagioclase grain in some ordinary chondrites (Miyahara et al. 2017, 2020). On the other hand, Ma et al. (2020) propose that highly defective jadeite (Na,Ca,v)(Al,Si)Si₂O₆ (v = vacancy) with excess silicon has formed in an albitic plagioclase grain of some ordinary chondrites.

2.3.2 Tissintite

Plagioclase in Martian and lunar meteorites is richer in anorthite components compared with equilibrated ordinary chondrites (Papike 2018). Tissintite is clinopyroxene-type (Ca,Na,v)AlSi₂O₆ and found in an anorthite grain in contact with the melt-pocket of the shergottite Tissint (Ma et al. 2015). Tissintite was synthesized from amorphous plagioclase under 6–8 GPa and 1273–1623 K (Rucks et al. 2018). Most tissintite grains in shocked meteorites show a dendritic or spherulitic texture, suggesting rapid crystallization from anorthitic melt (Fig. 9).

Fig. 9

BF-TEM image of tissintite in the shergottite Tissint. Spherulitic tissintites have occurred in an original plagioclase grain in contact with the melt-pocket of Tissint. The interstices are filled with amorphized plagioclase (probably because of melting). Tis: Tissintite, Mas: maskelynite

2.3.3 Lingunite–Stöfflerite–Liebermannite

Hollandite-type KAlSi₃O₈ was synthesized first at 12 GPa and 900 °C by Ringwood et al. (1967). Also, Liu (1978) synthesized hollandite-type NaAlSi₃O₈ above 20 GPa at ~ 1000 °C using an LHDAC. In contrast, subsequent experimental works in the system NaAlSi₃O₈-KAlSi₃O₈ reported that the hollandite phase is stable only in K-rich composition (Na/(Na + K): <  ~ 0.4), while calcium ferrite (CaFe₂O₄)-type NaAlSiO₄ + stishovite appears at Na-rich composition (Na/(Na + K): >  ~ 0.4) above ~ 20 GPa and 1000 °C (Yagi et al. 1994; Liu 2006). This discrepancy was reconciled by a recent transformation kinetics study that clarified that hollandite-type NaAlSi₃O₈ is a metastable phase crystallized from amorphous plagioclase glass (Kubo et al. 2017).

Hollandite-type NaAlSi₃O₈ replacing albitic feldspar was found in the shock-melt veins of the Sixiangkou and Tenham L6 ordinary chondrites (Gillet et al. 2000; Tomioka et al. 2000; Xie and Sharp 2004). A small amount of K-feldspar has also been found in some Martian and lunar meteorites. Hollandite-type KAlSi₃O₈ was found in the melt-pocket of the basaltic shergottite Zagami (Langenhorst and Poirier 2000). Hollandite-type NaAlSi₃O₈ and KAlSi₃O₈ were named lingunite after L. G. Liu (Liu and El Gorsey 2007) and libermannite after R. C. Liebermann (Ma et al. 2018), respectively. Lingunite was also found from the doleritic rocks of Lockne impact structure (Agarwal et al. 2016). Hollandite-type CaAl₂Si₂O₈ has yet to be synthesized but was recently found in a melt-pocket of the basaltic shergottite NWA 856 and named stöfflerite after D. Stöffler (Tschauner et al. 2021).

2.3.4 Donwilhelmsite–zagamiite

A calcium aluminum silicate phase (CaAl₄Si₂O₁₁) was synthesized by Irifune et al. (1994) and tentatively named the Ca–Al–Si-rich (CAS) phase. Its crystal structure with the CaAl₄Si₂O₁₁ composition was determined to be of the hexagonal BaFe₂O₄-type based on single-crystal XRD (Gautron et al. 1999). A mineral with (Ca_x,Na_1–x)Al_3+xSi_3-xO₁₁ composition was discovered in some shergottites such as Zagami and NWA 856 by micro-Raman spectroscopy (Beck et al. 2004; El Goresy et al. 2013). The mineral is intergrown with acicular stishovite grains in melt pockets. A natural CAS phase was also found from the feldspathic lunar meteorite Oued Awlitis 001 and was well characterized by three-dimensional electron diffraction pattern analysis (Fritz et al. 2020). This mineral was named donwilhelmsite after D. E. Wilhelms (Fritz et al. 2020). Ma et al. (2019a) also found a hexagonal BaFe₂O₄-type CaAl₂Si_3.5O₁₁ in the melt-pocket of Zagami and NWA 856 and named it zagamiite. Zagamiite is expected to be a crystallization product from a plagioclase-rich melt with some clinopyroxene components because it has small amounts of Fe and Mg and a substantial deficit of Na (Ma et al. 2019a).

2.3.5 Dense plagioclase glass

Experimentally shocked plagioclase becomes an amorphous phase with a high refractive index without loss of short-range ordering in its structure (Milton and de Carli 1963; Kitamura et al. 1977). The first identification of such a dense plagioclase glass is from the shergottite Shergotty and has been called maskelynite after N. S. Maskelyne (Tschermak 1872). Similar dense plagioclase glass has been then commonly found in strongly shocked chondrites and achondrites (e.g., Stöffler et al. 1986, 1991). The glass almost retains the morphology of the original plagioclase but loses its crystallinity. Therefore, it is easy to be identified by its optical isotropy under a polarizing microscope.

One of the big debates about maskelynite is its formation process. Some maskelynite shows a flowing texture or mixing texture with other melts (schlieren), suggesting that the maskelynite has formed by the quenching of plagioclase shock-induced melt (El Goresy et al. 2013). By contrast, some maskelynite has retained original compositional zoning, suggesting that the maskelynite has formed through a solid–solid-state phase transformation (Jaret et al. 2015).

Tschermak (1883) originally described that maskelynite has a non-vesicular flow texture resulting from melting and liquid migration followed by solidification. Based on the historical background, only material with evidence for melting like schlieren should be called maskelynite, whereas material without evidence of melting should be called diaplectic plagioclase glass. Normal glass is quenched from a melt which is still liquid after pressure release (Stöffler 1984). In case that the formation process (solid-state amorphization or melting) is not clear, but its high density can be assumed by its higher refractive index than normal plagioclase glass under an optical microscope (Stöffler et al. 1986, 1991), we suggest the term “dense plagioclase glass” instead of “maskelynite” and “diaplectic plagioclase glass” to avoid confusion.

2.4 High-pressure polymorphs of silica

The synthesis and naming histories of high-pressure polymorphs of silica, except for seifertite, are mentioned in a later section because their first discoveries are from a terrestrial impact crater. Coesite, stishovite, and seifertite are high-pressure polymorphs of silica. Quartz is rare but cristobalite and tridymite are abundant in meteorites (Papike 2018). Hence, most high-pressure polymorphs of silica have occurred as replacements of cristobalite or tridymite.

As a discovery from a meteorite, Weisberg and Kimura (2010) found coesite in the Gujba CB carbonaceous chondrite for the first time by Raman spectroscopy. Coesite is mainly found in basaltic lunar meteorites and eucrites (Miyahara et al. 2014; Tomioka and Miyahara 2017; Kayama et al. 2018) but very rarely from enstatite chondrite and Martian meteorite (Kimura et al. 2017; Hu et al. 2020). TEM observations revealed that these coesite crystals are granular and always coexist with silica glass or quartz (Miyahara et al. 2013a, 2014) (Fig. 10). Coesite is also found as one of the decomposition products of albitic plagioclase in shocked ordinary chondrites (Miyahara et al. 2017, 2020).

Fig. 10

BF-TEM image of coesite in the eucrite Béréba. Spherical coesite crystals have formed in cristobalite or tridymite grains in contact with the shock-melt veins of Béréba. A part of the original silica has become amorphous. Coe: coesite, Si-Gla: silica-glass

As for stishovite replacing silica minerals, it was discovered from the lunar meteorite Asuka-881757 for the first time (Ohtani et al. 2011). Stishovite crystals are acicular and always coexist with silica glass, the same as coesite (Ohtani et al. 2011; Miyahara et al. 2013a, 2014; Kayama et al. 2018). Stishovite is mainly found in basaltic lunar meteorites, eucrites, and shergottites (Tomioka and Miyahara 2017) and very rarely from carbonaceous chondrite (Bindi et al. 2012) and iron meteorite (Holtstam et al. 2003). Some stishovites have formed through the high-pressure decomposition of plagioclase in some shocked shergottites (Langenhorst and Poirier 2000; Beck et al. 2004; El Goresy et al. 2013). Kaneko et al. (2015) discovered stishovite from regolith breccia Apollo 15,299. This is the first case of a high-pressure mineral being discovered in extraterrestrial rocks by sample return missions.

Scrutinyite (α-PbO₂)-type silica was synthesized in shock recovery experiments of quartz up to 90 GPa (German et al. 1973). Its natural analog was found from the basaltic shergottite Shergotty (Sharp et al. 1999; Dera et al. 2002) and named seifertite after F. Seifert by El Goresy et al. (2008). Seifertite is also found from other shergottites and basaltic lunar meteorites (El Goresy et al. 2013; Miyahara et al. 2013a). In situ XRD studies show that seifertite is stable in the pressure range of 120–260 GPa (Murakami et al. 2003; Kuwayama et al. 2005), while Kubo et al. (2015) clarified that seifertite has metastably formed from cristobalite even above ~ 11 GPa. Therefore, seifertite is unlikely an indicator of megabar shock pressures in shergottites and lunar meteorites. Seifertite in backscattered electron images always shows a tweed-like texture, which is one of the diagnostic features for its identification (El Goresy et al. 2008). Like stishovite, seifertite crystals exhibit an acicular morphology and always coexist with silica glass (Miyahara et al. 2013a) (Fig. 11). El Goresy et al. (2000) also found an unnamed post-stishovite phase with the baddeleyite (ZrO₂)-like structure from one shergottite.

Fig. 11

BF-TEM image of seifertite in the basaltic lunar meteorite Northwest Africa (NWA) 4734. Lath-like seifertite crystals have formed in cristobalite grains in contact with the shock-melt veins of NWA 4734. Most parts of the original cristobalite grains have become amorphous. Sei: seifertite, Si-Gla: silica-glass

2.5 Others

2.5.1 High-pressure polymorphs of phosphate minerals

A Ca₃(PO₄)₂ mineral with a Ba₃(PO₄)₂-structure [γ-Ca₃(PO₄)₂] was synthesized by Murayama et al. (1986). Xie et al. (2002a) reported that merrillite [Ca₉NaMg(PO₄)₇] has transformed into the γ-Ca₃(PO₄)₂ phase in the shock-melt vein of the Suizhou L6 ordinary chondrite. The γ-Ca₃(PO₄)₂ phase was named tuite after G. Tu (Xie et al. 2004).

2.5.2 High-pressure polymorphs of chromite

CaTi₂O₄-type FeCr₂O₄ phase was synthesized by Chen et al. (2003a), and a natural one was found in the Suizhou L6 ordinary chondrite (Chen et al. 2003b). The CaTi₂O₄-type FeCr₂O₄ phase was named xieite after X. Xie (Chen et al. 2008). Harmunite (CaFe₂O₄)-type FeCr₂O₄ phase was synthesized by Chen et al. (2003a), and a natural one was found in the Suizhou L6 ordinary chondrite (Chen et al. 2003b). The mineral was named chenmingite after M. Chen (Ma et al. 2019b).

2.5.3 High-pressure polymorphs of other oxides

LiNbO₃-type FeTiO₃ phase was first reported in the Ries crater (Dubrovinsky et al. 2009). The phase was also found in the Suizhou L6 ordinary chondrite and named wangdaodeite after D. Wang (Xie et al. 2020). Schreyerite-type (Ti⁴⁺Fe²⁺)Ti₃⁴⁺O₉ phase was found in the melt-pocket of the eucrite NWA 8003 and named vestaite after the name of its parental asteroid Vesta (Pang et al. 2018). Sr₂Tl₂O₅-type Fe²⁺₂(Fe²⁺Ti⁴⁺)O₅ phase and GdFeO₃-type FeTiO₃ have been recently reported from the Martian meteorite Shergotty and named feiite after Y. Fei and liuite after L.-G. Liu, respectively (Ma et al. 2021b). In the same meteorite, CaTi₂O₄-type (Fe²⁺)(Fe²⁺Ti⁴⁺)O₄ (simply Fe₂TiO₄) with a post-spinel structure was also discovered (Ma et al. 2021a) and was named tschaunerite after O. Tschauner. CaGa₂O₄-type CaAl₂O₄ was discovered in a Ca–Al-rich inclusion (CAI) of the NWA 470 CH3 carbonaceous chondrite and named domitryivanovite after D. A. Ivanov (Mikouchi et al. 2009). The mineral was also reported in an unusual Ca–Al–Fe rich inclusion in the Vigarano CV3 carbonaceous chondrite (Maruyama and Tomioka 2011). Previous high-pressure experiments suggest that domitryivanovite is stable above ~ 2 GPa (Ito et al. 1980).

2.5.4 High-pressure polymorphs of other high-pressure (Mg,Fe)-silicates

Interestingly, a new phase found as a unique grain in a shock-melt vein of the Suizhou L6 ordinary chondrite has the composition of (Mg,Fe)₅Si₂O₉ and was named elgoresyite after A. El Goresy (Bindi et al. 2021). The crystal structure of this new iron–magnesium silicate, never obtained in laboratory experiments so far, is the same as the iron oxide Fe₇O₉, strongly suggesting that silicates also form the ((Mg,Fe)O)_m+n(SiO₂)_n series that are isostructural to iron oxides via (Mg²⁺,Fe²⁺) + Si⁴⁺  = 2Fe³⁺ substitution. The newly found iron–magnesium silicate is a potential constituent mineral in rocky planets with relatively high MgO + FeO content.

2.5.5 High-pressure polymorph of phosphide

Co₂Si-type (Fe,Ni)₂P was discovered in the Onello iron meteorite and named allabogdanite after A. N. Bogdanova (Britvin et al. 2002). The mineral was regarded to be a high-pressure polymorph of barringerite that is stable above 6 GPa and 900 °C based on previous high-pressure experiments (Dera et al. 2008) and natural occurrences in several other iron meteorites (Britvin et al. 2019). However, further detailed experimental and theoretical studies clarified that the mineral is also stable at ambient pressure below 500 °C (Litasov et al. 2020). Hence, the results suggest a possibility that the formation of allabogranite in iron meteorites does not require shock metamorphism.

2.5.6 High-pressure polymorphs of carbon

The first natural occurrence of diamond in shocked materials was reported, as in the Novo-Urei ureilite, in the late nineteenth century (Kunz 1888) and have currently been found in more than 10 ureilites (e.g., Goodrich et al. 2004; Miyahara et al. 2015; Nestola et al. 2020; Barbaro et al. 2021). Experimental works on the synthesis of diamond for industrial purposes succeeded by static compression of graphite over 3 GPa by Bundy et al. (1955) and shock-synthesized at 30 GPa by DeCarli and Jamieson (1961). Diamonds in ureilites are also expected to be formed under transient high-pressure condition shock metamorphism (e.g., Berkley et al. 1976; Nakamuta and Aoki 2000; Nakamuta and Toh 2013; Nestola et al. 2020). On the other hand, some diamonds might form under static high-pressure conditions in the deep interior of an ureilite parent body (Urey 1956; Miyahara et al. 2015; Nabiei et al. 2018) or at low pressure through chemical vapor deposition (Fukunaga et al. 1987).

2.6 Natural quasicrystals

One of the most remarkable findings of exotic extraterrestrial materials has been the discovery of natural quasicrystals (QCs) (Bindi et al. 2009, 2012). They have been detected as rare materials in the Khatyrka meteorite, a peculiar CV3 carbonaceous chondrite (MacPherson et al. 2013) containing Cu, Al-metallic compounds. QCs, short for quasiperiodic crystals, are solids that are able to violate the conventional rules of crystallography because their structure is quasiperiodic rather than periodic (Bindi et al. 2020b). The first quasicrystalline phase found in nature, icosahedrite Al₆₃Cu₂₄Fe₁₃ (Fig. 12) (Bindi et al. 2009, 2011), displayed a five-fold symmetry in two dimensions and icosahedral symmetry in three dimensions. Then, a second QC, decagonite Al₇₁Ni₂₄Fe₅ (Bindi et al. 2015a, 2015b), was found in the same meteorite and it was the first mineral to exhibit crystallographically forbidden decagonal symmetry. Both icosahedrite and decagonite, however, showed compositions matching those of synthetic quasicrystalline phases found earlier (Tsai et al. 1987, 1989) in the laboratory at standard pressure.

Fig. 12

BSE images of icosahedrite in the Khatyrka CV3 carbonaceous chondrite. a Grain 126A of Khatyrka CV3 carbonaceous chondrite. A red dashed box indicates the region to be enlarged in b. b The area where there is the metal assemblage containing the icosahedral phase (icosahedrite). c The enlarged image of the region indicated by a red dashed box in the panel b. d Electron backscatter diffraction pattern of the icosahedral phase obtained from the region marked with a cross in the inset. The pattern matches that predicted for a face-centered icosahedral quasicrystal

The origin of these unusual metallic alloys is enigmatic as they contain metallic aluminum that forms under highly reducing conditions not normally found in nature. Furthermore, the puzzling combination of metallic aluminum, a refractory lithophile element, and copper, a moderately volatile siderophile or chalcophile element, makes these alloys even more mysterious. When first reported, a plausible explanation was that the samples were byproducts of some laboratory or industrial process. Previous and successive studies, however, provided compelling evidence that the quasicrystals are natural and from a common meteoritic source: (i) ¹⁴C-dating of materials from undisturbed clay layers where some of the samples were collected yielded 6.7–8.0 ka BP, (ii) the Cu–Al metallic alloys were found to be intimately intermixed with oxides/silicates with nonterrestrial oxygen isotope composition (Hollister et al. 2014; Lin et al. 2017), (iii) clear evidence was found of high-pressure-induced phase transitions requiring transient conditions consistent with an asteroidal collision event, at least 5 GPa and 1200 °C, sufficient to melt and rapidly quench the Al–Cu bearing alloys (Asimow et al. 2016), (iv) noble gas measurements confirmed that the shock event reached pressure above 5 GPa and occurred at least hundreds of million years ago (Meier et al. 2018), and (v) robust petrographic and chemical evidence established that some metallic alloy grains (including QCs) found in the samples predated the shocks. Moreover, the recent discovery of a CO-type chondritic spherule from the Nubian desert, Sudan (Suttle et al. 2019), containing the same assemblage of aluminum, iron, and copper and with a morphology remarkably similar to Khatyrka provided further support and independent evidence that these samples were formed in outer space.

By means of laser ablation inductively coupled plasma mass spectrometry measurements of the trace element contents of the two naturally occurring quasicrystalline minerals, Al₆₃Cu₂₄Fe₁₃ icosahedrite and Al₇₁Ni₂₄Fe₅ decagonite, Tommasini et al. (2021) hypothesized a three-stage model for the formation of the quasicrystals and their inclusions: a high-temperature condensation stage and a low-temperature vapor-fractionation stage to make nanoparticles, followed by a third stage that leads to the formation of quasicrystals incorporating the two different types of nanoparticles and their incorporation into the CV chondrite parent body of the Khatyrka meteorite.

In a recent speculative essay, Bindi et al. (2020a) reported that although the only known examples of natural quasicrystals are from the Khatyrka meteorite, this does not necessarily mean that quasicrystals must be extremely rare in the universe. The authors present several reasons why quasicrystals may prove to be among the most ubiquitous minerals found in the universe.

2.7 Significance of high-pressure minerals from meteorites

Ringwoodite can be identified easily by an optical microscope because it shows purple/blue color in transmitted light. Hence, in the classic shock classification criterion (S1–S6), the identification of ringwoodite in ordinary chondrites became a criterion for very strongly shocked (S6) (Stöffler et al. 1991). Many kinds of high-pressure minerals were found from ordinary chondrites after the shock classification criterion was proposed. The criterion was revised, Stöffler et al. (2018) defining that not only ringwoodite but also wadsleyite, ahrensite, majorite, akimotoite, bridgmanite, and lingunite may be present in ordinary chondrite that satisfies shock stages S5–S6.

High-pressure melting experiments of the Allende CV3 carbonaceous chondrite and peridotite were conducted, and their phase equilibrium diagrams were obtained using their run products (Zhang and Herzberg 1994; Agee et al. 1995). There are similarities in the species, assemblages, chemical compositions, and grain sizes of high-pressure minerals between the recovered samples and constituents in shock-melt veins. Based on the similarities, Chen et al. (1996) estimated the shock pressure and temperature recorded in a shock-melt vein using the phase equilibria diagrams. This scheme was applied to several ordinary chondrites, eucrites, Martian, and lunar meteorites to estimate the shock pressure and temperature (e.g., Xie et al. 2003, 2006; Ohtani et al. 2004; El Goresy et al. 2013; Pang et al. 2016; Zhang et al. 2021).

The major disruption that occurred on the early parent body of ordinary chondrites was deduced by the systematic investigations of high-pressure minerals (Miyahara et al. 2020). Most ordinary chondrites have experienced thermal metamorphism after accumulation into the early parent body. Based on the maximum metamorphic temperature and cooling rate recorded in each ordinary chondrite, the early parent body of ordinary chondrites had an onion shell structure before its major disruption (e.g., Trieloff et al. 2003). Assuming the onion shell model, shock pressure should be the highest at the surface (least thermally metamorphosed one: petrographic type 3) of the parent body. However, unexpectedly, the shock pressure estimated by high-pressure minerals increases with increasing petrographic type from 3 to 6 (petrographic type 6 is the highest thermally metamorphosed one corresponding to the deepest interior of the parent body) (Miyahara et al. 2020). The results suggest that the apparent lower shock pressure on the surface would be explained by spallation at the opposite side of the parent body to the impact point.

Recent studies point out that shock metamorphism recorded in shocked meteorites is not simple. Some ordinary chondrites seem to record a dual impact because (i) a shock-melt vein is penetrated by another shock-melt vein (Hu and Sharp 2016) and (ii) a high-pressure polymorph includes another high-pressure polymorph indicating different shock pressures (Miyahara et al. 2011a). The dual impact (or dual heating event) record is also deduced from radioisotope ages in some shocked meteorites (e.g., Treiman et al. 2007).

The formation of a high-pressure mineral and resetting of a radioisotope’s age depend on the thermal history. However, the thermal history recorded in a shock-melt vein or melt-pocket is also complex. Pressure decreases drastically (several microseconds’ order) but temperature decreases gradually (several seconds’ order) during an adiabatic decompression stage, resulting in low-pressure and high-temperature conditions. If the low-pressure and high-temperature conditions are prolonged, the back transformation of a high-pressure mineral occurs, which underestimates the shock pressure (Kimura et al. 2004; Hu and Sharp 2017; Miyahara et al. 2019). However, a back-transformation mechanism of a high-pressure mineral and its kinetics have not been deeply investigated except for a few works (Ming et al. 1991; Fukimoto et al. 2020; Miyahara et al. 2021). We must work on these issues to trace a pressure–temperature-time path recorded in a shock-melt vein or melt-pocket as future work.

Phase transformation is kinetically controlled. Ohtani et al. (2004) suggested estimating the magnitude of an impact event recorded in an ordinary chondrite using the grain-growth kinetics of ringwoodite. The grain sizes of ringwoodite crystals in shocked ordinary chondrites are similar to each other (across < 1–2 μm) (Ohtani et al. 2004; Miyahara et al. 2010; Kato et al. 2017). Hence, the estimated duration of shock pressure is almost the same (< several seconds) (Ohtani et al. 2004; Xie et al. 2006; Sharp et al. 2015). Also, the phase transition rates of plagioclase and silica are applied to estimate the duration of shock pressure (< ~ 5 s) in some ordinary chondrites and lunar meteorites (Kubo et al. 2010; Miyahara et al. 2013a, 2017; Ozawa et al. 2014). On the other hand, recent time-resolved femtosecond diffractometry using an X-ray free-electron laser revealed that ringwoodite can form from laser shock compressed olivine through ultrafast lattice shearing in nanoseconds (Okuchi et al. 2021).

3 High-pressure minerals in impact craters

3.1 High-pressure polymorphs of carbon

In terrestrial impact structures, diamond was found in the Barringer (Ksanda and Henderson 1939), Ries (Hough et al. 1995), Zapadnaya (Gurov et al. 1995), and Popigai (Koeberl et al. 1997) craters. A previous high-pressure experiment succeeded in synthesizing nano-polycrystalline diamond (NPD) with grain sizes less than 100 nm by direct conversion from pure polycrystalline graphite at 12–25 GPa and 2300–2500 °C (Irifune et al. 2003). The NPD was brought to material scientists’ attention because it has a larger stiffness than a single-crystal diamond. Its natural analog was also discovered from the Popigai impact crater (Ohfuji et al. 2015).

A modified diamond structure with a hexagonal cell was synthesized by a high-pressure experiment (Bundy and Kasper 1967). The natural hexagonal diamond was first found in the Barringer crater and was named lonsdaleite after K. Lonsdale (Frondel and Marvin 1967). It is a very enigmatic mineral. It was also discussed that it should not exist as a discrete mineral because it could correspond to a faulted and twinned cubic diamond (Németh et al. 2014). In the Popigai crater, both diamond and lonsdaleite were found in impact melt rock and breccia. These high-pressure phases of carbon have been formed from graphite in the target gneissic rocks by lattice-shear (martensitic) transformation mechanisms of the graphite structure by shock metamorphism (Koeberl et al. 1997; Nakamuta and Toh 2013). The formation of diamond and lonsdaleite was estimated at 35–60 GPa based on shock effects on the host-rock minerals.

A novel carbon crystal was reported in the Ries crater. The crystal shows a diffraction pattern different from those of preexisting crystalline carbon polymorphs (graphite, diamond, or lonsdaleite) and named chaoite, as mentioned previously (El Goresy and Donnay 1968). A similar allotropic form of carbon was synthesized during the sublimation of pyrolytic graphite (Whittaker and Kintner 1969). The mineral is thought to have a crystal structure comprising chains with alternating single and triple bonds (–C≡C–), which is called the carbyne structure. Five variations of carbynes were also found from the Allende and Murchison carbonaceous chondrites (Whittaker et al. 1980). The finding was once questioned and claimed as misidentified phases by powder XRD as a mixture of quartz and Fe-smectite (Smith and Buseck 1982). However, subsequent experimental studies ensure the existence of carbynes (e.g., Chalifoux and Tykwinski 2010; Pan et al. 2015). First-principles’ calculations clarified that carbyne is about twice as stiff as the stiffest known materials under tension (Liu et al. 2013); therefore, chaoite is also a potentially important mineral for material science.

Another novel allotropic carbon was discovered in a shocked gneiss from the Popigai impact crater (El Goresy et al. 2003). Carbon platelets, embedded in quartz and cristobalite, consist of graphite and lonsdaleite and an unknown carbon phase. The unknown phase has considerably high polishing hardness overcoming diamond powder. Synchrotron XRD clarified that the phase has a cubic cell with the space group Pm-3m, although its density and the detailed crystal structure have not yet been obtained. Its origin and formation conditions are also unknown because of a lack of experimental and theoretical studies.

Finally, Németh and Garvie (2020) have recently reported the discovery of cage-like nanostructured carbonaceous materials, including carbon nano-onions and bucky-diamonds, formed through extraterrestrial impacts in the Gujba CB_a carbonaceous chondrite.

3.2 High-pressure polymorphs of silica

Dense SiO₂ crystals are the first synthesized minerals among silicates and oxides related to the Earth and planetary materials. Coesite and stishovite were first synthesized by static compression by Coes (1953) and Stishov and Popova (1961), respectively, and stishovite was also shock-synthesized by DeCarli and Milton (1965). Phase equilibria studies based on static high-pressure experiments clarified that coesite and stishovite are stable above ~ 2 GPa and ~ 8 GPa, respectively. Stishovite was shock-synthesized at pressure over 12 GPa, while there is no convincing report that coesite can be formed by shock recovery experiments. Smaller shock pressure duration (10^–6 s) in laboratory shock experiments than that in terrestrial impact cratering (> 10^–3 s) (Kieffer 1971) and sluggish kinetics in the quartz to coesite transformation would cause difficulty in the artificial production of the mineral (e.g., Nagai et al. 1997; Perrillat et al. 2003).

The first finding of natural dense silica was performed in the early 1960s. E. C. T. Chao and his colleagues applied hydrofluoric acid leaching to a large amount of the Coconino sandstone of the Barringer crater, which was produced by the impact of the Canyon Diablo iron meteorite ~ 49,000 years ago. Coesite and stishovite, the high-pressure polymorphs of SiO₂, were identified in the residues by powder XRD (Chao et al. 1960, 1962). In particular, coesite is one of robust evidence for impact origin and found in over 20 impact craters. Stishovite has been found in three more impact craters, namely Popigai (Vishnevsky et al. 1975), Ries (Stöffler 1971; Stähle et al. 2007), and Vredefort (Martini 1978, 1991; Spray and Boonsue 2018), in addition to the Barringer crater.

3.3 High-pressure polymorphs of TiO₂

Rutile transforms into scrutinyite (α-PbO₂, TiO₂-II), baddeleyite (ZrO₂), orthorhombic-I, and cotunnite structures with increasing pressure (Nishio-Hamane et al. 2010). Scrutinyite-type TiO₂ with an orthorhombic cell was first synthesized by a static compression experiment (Bendeliany et al. 1966). In nature, almost pure scrutinyite-type TiO₂ was first discovered in shocked garnet gneisses from the Ries crater (El Goresy et al. 2001). The phase was also found from suevite in a drilling core of granitic basement rock from the Chicxulub crater (Kring et al. 2020), which was estimated to have been shocked at ~ 20 GPa in the Cretaceous–Paleogene boundary (Timms et al. 2019). The TiO₂-rich scrutinyite-phase had been known as mineral srilankite [(Ti,Zr)O₂], where Ti and Zr ions are disordered at the cation sites. The Ries crater also yielded almost pure TiO₂ with the scrutinyite-like structure that was named riesite after the name of the location (Tschauner et al. 2020b). Riesite occurs in shock-melt veins within xenoliths of bedrock in suevite. Srilankite has one cation site for Ti and Zr ions and one oxygen site, while there are two cation and oxygen sites each in the riesite structure, causing the monoclinic distortion. Riesite has not been synthesized, but its higher density (4.37 g/cm³) than rutile (4.25 g/cm³) suggests riesite is a high-pressure mineral. Thus, riesite has formed through the back transformation from the baddeleyite phase upon pressure release.

Baddeleyite-type TiO₂, which has 7-coordinated Ti ions, was first observed at 20 GPa, 770 °C by in situ XRD using a Kawai-type high-pressure apparatus (Sato et al. 1991) but could not be recovered at ambient condition. In contrast, in nature, baddeleyite-type TiO₂ was first found in the suevite breccia from the Ries crater and named akaogiite after M. Akaogi (El Goresy et al. 2001, 2010).

3.4 High-pressure polymorphs of MgFe₂O₄

A high-pressure phase of magnesioferrite (MgFe₂O₄) was discovered in a drill core of shocked diamond-bearing gneiss from the Xiuyan crater (Chen et al. 2019). Powder XRD and TEM clarified that the mineral has the harmunite (CaFe₂O₄) structure as same as chenmingite (FeCr₂O₄) discovered in the Suizhou L6 ordinary chondrite (Chen et al. 2003b) and the shergottite Tissint (Chen et al. 2019). The mineral was named maohokite after H.-K. Mao (Chen et al. 2019). Maohokite coexists with magnesioferrite embedded in calcite and is thought to be a decomposed product of ankerite. The high-pressure phase equilibria of MgFe₂O₄ have not been established by experimental studies. However, based on pressure and temperature conditions for the post-spinel transformation (e.g., Andrault et al. 2001), maohokite is thought to have formed at 25–45 GPa and 800–900 °C (Chen et al. 2019).

3.5 High-pressure polymorphs of ZrSiO₄

The high-pressure phase of zircon is an important key mineral to understanding the shock metamorphism in impact craters because of its robustness against weathering. The scheelite (CaWO₄)–type ZrSiO₄ was discovered in the upper Eocene impact ejecta layer in marine sediments by powder XRD and named reidite after A. F. Reid who first synthesized this phase (Glass and Liu 2001; Glass et al. 2002). Subsequently, reidite was found in five impact craters: Chesapeake Bay (Malone et al. 2010), Xiuyan (Chen et al. 2013), Rock Elm (Cavosie et al. 2015), Ries (Erickson et al. 2017), and Woodleigh (Cox et al. 2018) craters. Shock recovery experiments of zircon indicate that the transition to reidite is above ~ 30 GPa (Kusaba et al. 1985), while static high-pressure experiments demonstrated that the stability field of the phase is in the pressure range of ~ 8–25 GPa (Ono et al. 2004; Tange and Takahashi 2004). The discrepancy between two shock pressure estimations would be caused by different high-pressure timescale and/or differential stress between the static and dynamic high-pressure experiments.

3.6 Significance of high-pressure minerals from impact crater rocks

High-pressure minerals do not occur as commonly in terrestrial impact craters when compared with those in shocked meteorites. One of the possible reasons is the different cooling histories after shock pressure release between the surface materials of asteroids and the Earth’s crust. Shocked surface rocks in the parent bodies of meteorites are excavated and released to space, resulting in rapid cooling. In terrestrial impact craters, part of the surface and near-surface materials are ejected and also rapidly cooled to preserve high-pressure minerals, which are mostly inclusions in lithic impact breccia (suevite). Nevertheless, the materials in larger burial depths of the craters such as those in thick sediments and basement rocks should experience longer residual shock heating because the temperature decreased mainly by thermal conduction within the rocks. In such a case, high-pressure minerals would have been back-transformed to low-pressure minerals even if they were once formed during shock metamorphism. Another reason for the weak preservation of high-pressure minerals is aqueous alterations after the crater formation. Abundant shock-induced fractures play a role in fluid pathways, and fine-grain size and glassy materials containing high-pressure minerals are susceptible to alterations by weathering and impact-induced hydrothermal activities (Grieve et al. 1996; Kring et al. 2020).

Nevertheless, technical developments on microanalyses and constant efforts by impact petrologists enabled the findings of many high-pressure minerals from impact craters. Since the discovery of coesite in the Barringer crater, over 10 new high-pressure minerals have been found from terrestrial impact craters (Table 1). These dense minerals are direct and robust evidence for the impact origin (but not a volcanic origin) of the craters. Thus far, shock microstructures of quartz and, in particular, crystallographic orientations of planar deformation features are commonly used to evaluate shock pressures in the pressure range of ~ 10–35 GPa of sedimentary and granitic rocks in impact craters (Langenhorst 2002; Stöffler et al. 2018). Also, phase transformations of quartz such as formations of diaplectic (dense) glass, lechatelierite (normal glass), coesite, and stishovite further provide pressure constraints in a wider pressure range. A recently found superstructure of quartz from the basement rock of the Chicxulub crater (Tani et al. 2018) is another possible shock pressure indicator. The previously described high-pressure phases of carbon, oxides, and silicates would also contribute to elucidate spatial shock pressure distribution in impact structures considering the results of static and dynamic high-pressure experiments.

The attenuation of shock pressure in impact cratering is expressed by the following equation, P = P₀(R/Lp)^–b, where pressure P at the distance R from the isobaric core is a function of initial peak pressure P₀, R, the radius of impactor Lp, and attenuation coefficient of shock wave b (Mizutani et al. 1990). When we can obtain a shock pressure gradient based on pressures of different locations within an impact crater, the size of the impactor can be constrained. For this purpose, the occurrences of high-pressure minerals can be useful anchor points of shock pressures.

Shock pressure distribution during impact cratering is one of the most fundamental parameters to understand shock temperature distribution, rock deformation, and impactor’s size and velocities. The information is closely related to impact-induced phenomena such as the size and morphologies of the impact craters (Morgan et al. 2016; Riller et al. 2018), post-impact hydrothermal activities (Kring et al. 2020), atmospheric conditional changes (Artemieva et al. 2017), and biological mass extinctions (Lowery et al. 2018). Therefore, future research of high-pressure minerals will play important roles to elucidate how the impacts of extraterrestrial materials affect Earth’s environments.

4 Summary

Pioneer works in the 1960s by R. A. Binns, E. C. T. Chao, A. El Goresy, and their colleagues clarified the presence of high-pressure minerals in shocked meteorites and rocks around terrestrial craters. After their discoveries, mineralogists and planetary scientists have continued to find high-pressure minerals from natural shocked materials. The identification of new high-pressure minerals has increased drastically in the twenty-first century because of the introduction of nanoanalysis technologies. The search for high-pressure minerals in shocked materials was initiated by finding the analog materials of the deep Earth’s interior. This initial purpose is almost achieved at present. However, the crystal structures and formation mechanisms of high-pressure minerals in naturally shocked materials are more complicated than those expected based on experiments and theoretical simulations. As the next step, we aim for better quantification of P–T histories of the shocked metamorphism based on the high-pressure and high-temperature transformation kinetics and of impact histories in the solar system by combining shock petrography and radioisotope ages.