Keywords

8.1 The Concept and the Chemistry of High Pressure Minerals

The distinction of high-pressure and high-temperature minerals from minerals that form under less extreme conditions requires criteria that define pressures and temperatures as either high or low. It is useful to examine the effect of the two parameters, pressure and temperature, initially as separate. We find that the range of energy that is compatible with the crystalline state of matter that involves changes only in pressure exceeds by far the range of changes induced only by temperature, with regard to the crystalline state: The materials with the highest melting points melt between 2000 and 3000 K at ambient pressure (Adachi and Imanaka 1998). These temperatures correspond to energies in the range of ¼−1/3 eV/at, if we simply multiply the temperatures with the Boltzmann constant. However, the change in energy that occurs upon compressing mantle peridotite from the shallow lithosphere to the core mantle boundary over an interval of about 136 GPa of pressures is in the range of 1.6 eV/at: With an approximate bulk composition of Mg2SiO4, ¾ of the Earth’s mantle are oxygen as constituent chemical species and within this approximation the compression of the O2− anion dominates the increase of the electronic contribution to the inner energy of bulk silicate Earth over the entire range of compression by amount and size (Tschauner 2022a). Between 0 and 136 GPa (the pressure of the core mantle boundary) the contraction of the crystal radius of O2− is from about 1.26 to 1.16⋅10−10 m (Tschauner 2022a) (Fig. 8.1a), hence: 4/3π⋅Δr3 ⋅ 1.36⋅1011 N/m2 = 2.50⋅10–19 J/at = 1.56 eV/at (of O2−). Yet, throughout this range of pressure mantle rock remains in the solid state along the average geotherm. Thus, within the range of the solid state, pressure as a parameter allows for changes in energy several times larger than temperature, even within the limited range of conditions that occur inside Earth. Since the melting points of solids generally increase with pressure, a regime of high temperatures that corresponds to energy changes of ≥ 1 eV/at and that is compatible with the solid state occurs only at sufficiently high pressures.

Fig. 8.1
Two graphs, a and b. The first graph plots crystal radii versus pressure. The trend is declining with the division of the graph into vertical halves of high-pressure regimes 1, 2, and 3. The second graph plots r Cryst r B whole cube versus pressure. The trend is declining. Values are estimated.

Crystal radii of geochemically abundant elements as functions of pressure. a: Radii of K,Mg,Ca, Al, Si and O2− in different bond coordination by O2−. Radii of O are corrected for coordination by the cations. b: Ionic volumes normalized by the Bohr radius rB. Reconstructive transitions to high-pressure minerals and phases are indicated by arrows. Data are from Tschauner (2022a) with additions based on data by Dewaele et al. (2012), Levien et al. (1980), Lazarz et al. (2019), Ko et al. (2022), Richet et al. (1988). The pressure dependencies are r(K[8]) = 1.62(2)–0.003(1)P; r(Ca[6]) = 1.143(3)–0.00239(8)P; r(Ca[7,8]) = 1.188(7)–0.00149(8)P; r(Ca[9,10]) = 1.319(8)–0.0020(2)P; r(Mg[6]) = 0.856(7)–0.0015(1)P; r(Mg[12]) = 1.11(3)–0.0024(1); r(Al[6]) = 0.669(2)–0.00108(4)P; r(Si[4]) = 0.373(9) + 0.0020(3)P; r(Si[6]) = 0.567(3)–0.00095(9)P. P in GPa and radii in Å

Full size image

High-pressure and high-temperature minerals involve constituent chemical species whose valence electron configuration is energeticaly favourable at the pressures and temperatures of formation of these minerals but are unfavourable or unstable at low pressures and ambient conditions (Tschauner 2019). Radial valence electron distributions, that is: ionic radii and crystal radii, are sufficient to define these criteria. Ionic and crystal radii represent spherical spatial averages over a multitude of different bond states (Rahm et al. 2020; Tschauner 2022a, b). Although the radii neglect the actual bond states of the individual compounds and structures, they allow for assessing types of structure that are assumed by solids of very different composition or stability fields. This point is well illustrated by the successful application of tolerance factors and similar criteria that correlate composition with crystal structure types (e.g. in Li et al. 2004) and their evolution with pressure (e.g. Manjon et al. 2007).

Pressure shifts compounds into structure types which are generally assumed by compounds of chemical species with higher nuclear charge number at ambient pressure (Shannon and Prewitt 1969). For instance, bridgmanite, the high-pressure polymorph of MgSiO3 is isotypic with perovskite, CaTiO3, davemaoite, the high-pressure polymorph of CaSiO3, is isotypic with tausonite, SrTiO3, stishovite SiO2 is isotypic with rutile, TiO2, the high-pressure minerals lingunite, liebermannite, stöfflerite assume the structure of hollandite KMn3+Mn4+3O8 and so on (see Table 8.1). This general trend has been interpreted as result of the stronger compression of the anions relative to the cations (Downs and Prewitt 1998), but it also indicates relative changes of cation ionic radii with pressure (see Fig. 8.1). A quantitative concept of these pressure effects allows for correlating mantle geochemistry with high-pressure mineralogy and petrology. The effect of pressure on the crystal radii is shown in Fig. 8.1a for K+, Mg2+, Ca2+, Al3+, Si4+, and O2− in different bond coordination (henceforth, formal valences are not specified and bond coordination is given in angular brackets). The following observations are made: (a) The O-anion exhibits initially a marked non-linear compression converging towards weaker linear compression. (b) All cations exhibit linear contraction over the examined pressure intervals (Fig. 8.1a) within uncertainties. Only Si[4] expands with pressure. (c) Heavier cations like K and Ca are more compressible than lighter ones like Mg, Al, and Si. d) The higher the charge the lesser the pressure effect, see Fig. 8.1, caption). (e) A general trend for the pressure dependence of crystal radii with bond coordination is not seen for the available data. In part these basic pressure-induced trends have been noticed previously (Shannon and Prewitt 1969; Downs and Prewitt 1998; Gibbs et al. 2012) but actual compressibilities were only recently reported for some bonded radii (Gibbs et al. 2012), some crystal radii (Tschauner 2022a), for non-bonding radii (Rahm et al. 2020) and, by means of corresponding states, for Wigner–Seitz radii of elemental metals (Tschauner 2022b). In the caption of Fig. 8.1 we give the compressibilities for Mg,Al,Si,K,Ca which are based on an augmented set of data and where the change of the O-anion radius with anion coordination (Shannon 1976) is taken into account in calculating the cation crystal radii. It is noteworthy that the fitted crystal radii at 1 bar match Shannon’s radii very well although 1 bar radii were not used as fix points (see caption Fig. 8.1). The only fitted 1 bar radius that deviates from Shannon’s radii is that of Si[6].

Table 8.1 List of all approvedminerals from the high-pressure regimes hP-II and -III. Some petrologically related incipient high-pressure minerals (Tschauner 2019) from regime hP-I are listed also (printed in italics). Endmember composition, first reference of the approved mineral or announcement by the CNMNC, density of endmembers and the density of the stable polymorph at reference conditions are given. The reported densities of the type specimens of these minerals may be different if they contain noticeable amounts of other components. In Occurrences, references are given only if different from the type material reference. (natural high-pressure phases that are not approved minerals)
Full size table

Figure 8.1a shows that with increasing pressure a regime controlled by strong nonlinear O2− contraction is followed by a regime of reduced, nearly linear contraction. In this regime contraction of larger cations K and Ca is more prominent than that of O (Fig. 8.1a). The border between the two regimes coincides with the Si[4] → [6] transition and, thus, delineates the boundary between low- and intermediate-pressure silicates (Tschauner 2019) on one side, and high-pressure silicates on the other side (Table 8.1). Here we define these two regimes as ‘high-pressure I’ and ‘high pressure II’ (Fig. 8.1). The radii of Ca[8] and [10] in CaO-B1 and in davemaoite interpolate to the 1 bar crystal radii of Ca[7] and [9], respectively (Fig. 8.1 caption). If one accepts the notion that radii represent spherical spatial averages of valence electron configurations (Rahm et al. 2020; Tschauner 2022a,b), this coordination change suggests a gradual change of the valence electron configuration for Ca over about ½ Mbar of linear compression.

The compression of radii by 10–30% (Fig. 8.1a) is well within the range of differences between radii of different chemical species or different valences of the same species. In Fig. 8.1b ionic volumes rcryst3 are normalized by the cube of the Bohr radius rB. The volumes of Si, Al, Mg, and Ca in six-fold coordination by O2− are approximately one-,two, four- and ten-times rB3 (lines in Fig. 8.1b). Sixfold Mg, Al, and Ca approach Si[6] between 170–180 GPa by extrapolation of their linear pressure-depencencies, Ca[8-10] between 290–300 GPa. At those pressures, the contraction of the O anion is small, thus, volume reducing transitions either have to involve a change in cation coordination and valence electron structure or a change in valence of O (Zhu et al. 2013). Significant volume reduction may involve hybridization of inner shell electrons with the valence electron states. This tentative ‘ultra-high pressure regime’ is labeled as ‘high-pressure III’ in Fig. 8.1a. The bridgmanite-ppv transition (Murakami et al. 2004; Ono and Oganov 2004) may indicate the onset of this regime (although without hybridization of inner and valence shell electrons). Consequently, we classify high-pressure minerals based these three regimes as hPI, hPII, hpIII. However, the process is generally not as straightforward: reconstructive pressure-induced phase transitions appear to reset the electron density. In CaO the transition from the NaCl- to the CsCl-type around 40–60 GPa (Richet et al. 1988) resets the normalized volume of Ca[8] to that of Ca[6] at ambient pressure and so does the coordination change of Ca upon formation of davemaoite (Fig. 8.1b, red arrows). The transitions from Mg[6] to [12] and from Si[4] to [6] also increase the ionic volume (Fig. 8.1b, black and blue arrows). Hence, bulk volume contraction upon those transitions is result of the increased bond coordination of both, cation and anion, which generally allows for denser structural arrangements of the atoms (Downs and Prewitt 1998). The positive pressure-dependence of Si[4] and its volume smaller than r3B indicate indirectly the extensive overlap of Si–O binding orbitals. The reset of high-pressure crystal radii to larger radii upon high-pressure phase transitions is indicative of the changes in valence electron configuration, if we allow the radii to represent spherically symmetric spatial averages of these configurations (Rahm et al. 2020; Tschauner 2022a,b). This case becomes interesting, when high-pressure transitions induce radii that match those of other elements at low pressure: For instance, the crystal radius of Mg in CaIrO3-type MgSiO3 matches the crystal radius of Ca[6] extrapolated to the transition pressure of ~ 120 GPa (Fig. 8.1b). K[6] intersects Ca[9,10] between 32 and 40 GPa, Mg[12] intersects Ca[6] around 20 GPa and Ca[8] between 60 and 80 GPa. There is no known mineral where Ca[6] would substitute for Mg[6] around 20 GPa but the substitution Ca + Fe for Mg + Al in bridgmanite has been proposed to occur above 60 GPa in experimental work (Ko et al. 2022). Type davemaoite, CaSiO3, contains a noticeable amount of K and Fe (Tschauner et al. 2021b, 2022a), consistent with a formation in the range of 20–30 GPa (Fig. 8.1a,b). Coordination changes reset the crystal ionic volumes (see above, Fig. 8.1b) but this effect is only indirectly expressed in solid solutions through changes in crystal chemical compatibility. In consequence, some but not all intersections of relative ionic volumes (rcryst/rB)3 match the formation of high-pressure minerals or pressure-induced chemical substitution. The underlying chemical selection rules are beyond the topic of this chapter. Even at high pressure entropic components remain important and the phase diagrams do not simply reflect a sequence of pressure-induced transformations but include minerals and mineral assemblies that occur at combined elevated pressure and temperature (e.g. in Fig. 8.2). This is the case at least within the hPI and hPII regimes.

Fig. 8.2
A phase plane of temperature versus pressure plots 6 increasing solid lines for melt pocket adiabat, bulk rock adiabat, and more. It also plots a few broken dashed and dotted lines in an increasing trend. The line for bulk rock adiabat is at the bottom.

Shock release path of melt pockets in the Tissint Shergottite. The pocket shown in Fig. 8.3a contains dense glass in the center and its main cooling occurred within the stability field of bridgmanite (red dotted lines). Another pocket contains intergrowth of pigeonite and fayalite in its center indicating cooling at much lower pressure (red and yellow lines). Thermodynamic phase boundaries are indicated for Fe2SiO4 (green, dashed) and the simplified CMS system (black), adiabats of the shock-generated melt (yellow and red lines) and the bulk rock (green) bracket the cooling paths. Data are taken from Ma et al. (2016)

Full size image

8.1.1 High Pressure Minerals–Their Occurrences

Minerals from the high-pressure regime I (‘hpI’) are found in high-grade metamorphic rocks such as eclogites and in xenoliths of garnet peridotites from below 60 km depth in the upper mantle. Several excellent reviews about these occurrences are available and it is not necessary to recapitulate this work here. Some of these intermediate pressure minerals are presented here along with the discussion of high-pressure minerals hPII and -III (Table 8.1). The occurrence of high-pressure minerals in Earth in the deep Earth is beyond direct access to us. However, four sources of these minerals have been found: meteorites, whose parent bodies have experienced strong shock-metamorphism by asteroid collisions, (b) terrestrial rocks that have experienced shock metamorphism through asteroid impacts, (c) inclusions in terrestrial diamonds. In addition (d) regimes of high pressures and temperatures occur in the ejecta of novae and supernovae part of whose debris is conserved as presolar grains in primitive meteorites, interplanetary and interstellar dust.

8.1.1.1 High-Pressure Minerals that Form Under Dynamic Compression

This section highlights some general aspects of high-pressure minerals that form under dynamic compression rather than the physics of shock and the processes that occur during shock-metamorphism.

Presolar dust grains are subject to extensive research mostly focusing on isotopic anomalies that witness nucleonic processes inside large stars and during supernovae. These processes are beyond the stability of atomic matter and, therefore, beyond our topic. Nonetheless, the process of capturing matter in solid phases through sublimation in the cooling ejecta involves high temperatures (see Sect. 8.4) and may in part involve elevated pressures also. Because of the low density of the ejected gas the regime of high pressure at temperatures below the condensation point of solid phases is rather limited and may, for that reason, be restricted to diamond as the solid phase with the highest sublimation and melting point. Diamond is a common presolar mineral (Table 8.1). The occurrence of presolar diamonds with high density of stacking faults along [111] (Daulton et al. 1996) is consistent with formation at high dynamic stresses and stress-rates, (Armstrong et al. 2022). Periodic stacking faults along [111] lead to the formation of lonsdaleite, the 2H-polytype of diamond (Table 8.1). Metastable formation at low pressures provides an alternative explanation of presolar diamond, for instance, nano-diamond forms during combustion of acetylene. So far, no presolar high-pressure mineral other than diamond has been found.

Collision of small planetary bodies, so called ‘planetesimals’ were a process intrinsic to the early history of the solar system and have nurtured the formation the larger planets. Chondrules, that is: spherical aggregates of one or several minerals that are frequently found in many common meteorites (‘s.c chondrites’, Rubin and Ma 2017), have been suggested to be the quench products of shock-induced melting and spallation of these melt particles, but there are alternative explanations of chondrule formation (see Rubin and Ma 2017 for detailed discussion). Within the asteroid belt collisions continue to occur. For instance, one of the most common type of meteorites, L-chondrites, is debris from the disruption of a planetesimal during a collision that occurred in the asteroid belt in the Ordovicium (Greenwood et al. 2007). Principally, all meteorites that we find on Earth have experienced modifications through dynamic compression during the events that destroyed their parent body in large events or ejected them from their surface in smaller events. The range of petrographically documented shock-metamorphic processes ranges from a few GPa to > 70 GPa (Stöffler et al. 2018). These changes have been categorized based on shock-induced deformation features that have been observed both in experiments and in nature on a scale that ranges from S1 (0–5 GPa) to S6 (>70 GPa) (Stöffler et al. 2018). High-pressure minerals are observed in meteorites of the shock metamorphic categories S4 and above. States of dynamic compression during asteroid collisions are generally assessed to less than 1 s. In fact, most estimates suggest durations of 10–100 ms (Tschauner et al. 2009; Hu and Sharp 2017; Ma et al. 2016; Tomioka and Miyahara 2017), corresponding to small cratering events or collisions of small bodies (Melosh and Ivanov 2002). Within this time period pressures are beyond the stability range of most of the rock-forming minerals in those meteorites: forsterite, enstatite, feldspars. However, the kinetic barriers are high for transforming these minerals into the polymorphs or decomposition products that represent thermodynamic stability at those pressures. Thus, along the principal Hugoniot of these rocks most of these minerals only develop characteristic deformation features and high densities of defects (Stöffler et al. 2018, for the specific terminology of shock compression: See for instance Ahrens 1987). Feldspars transform into a dense glass, ‘maskelynite’, whose structure and density deviate from feldspathic glass synthesized at ambient pressure even after full relaxation of the dynamic stress state. This shock-induced amorphization of feldspars occurs above 30 GPa depending on composition and shock duration (Stöffler et al. 2018). Maskelynite is therefore a ‘diaplectic glass’ because it has not formed through quenching of a shock-induced melt but through compression of a crystalline material beyond its mechanical stability. It had been suggested that maskelynite in highly shocked meteorites has formed from melt (Chen and El Goresy 2000). However, in many such meteorites the volume fraction of maskelynite is incompatible with conservation of the bulk rock upon release from the shock-compression state if maskelynite had been molten. At very high degrees of dynamic compression the Hugoniot line of the bulk rock intersects the melt line under dynamic compression with subsequent bulk rock melting and disruption of the shocked rock upon release (Ahrens 1986; Stöffler et al. 2018). S7 level meteorites exhibit pervasive melt veins and may reflect sources close to the regions of complete melting (Fritz et al. 2017; Stöffler et al. 2018). Variations in shock levels within given meteorite classes may also reflect different distances to the impact location (Fritz et al. 2017).

Whereas the bulk rock of shocked meteorites only exhibits shock-induced defects and deformation features, locally temperatures are high enough to overcome the kinetic barriers of formation of stable and metastable hP-I and -II minerals. These s.c. hot spots form from collapse of pore spaces and cracks, or represent shock-induced melts that penetrate into fracture zones of the shocked bed rock with velocities that scale with the particle velocity of the shock compression state, or they form through frictional heating along shear zones within the dynamically deforming rock, similar to pseudotachylites along fault surfaces during earthquakes.

In laboratory-scale shock experiments high-pressure mineral formation has only been obtained through collapse of void space (Tschauner et al. 2009) whereas shock-induced friction experiments have not generated any high-pressure minerals (Kenkmann et al. 2000). However, the failure of the latter type of experiments may be owed to the comparatively short duration of ≤ 1 ms of the dynamic compression state in laboratory scale experiments.

In nature we find high-P I and high-P II minerals at the rims or within transformed clasts of shock melt-veins and-pockets in meteorites (see Fig. 8.2 and Table 8.1). Generally, phase occurrence follows the temperature gradient. For instance, in the martian meteorite Tissint a sequence deformed forsterite (Fo80Fay20) → nano-rwd in deformed Fo → ahrensite (out of faylitic rims of the Fo grains) → bridgmanite + wuestite → quenched melt is observed (Ma et al. 2016). (Fig. 8.2 and 8.3a; Table 8.1). In highly shocked chondrites, the highest pressure minerals observed, bridgmanite and akimotoite (Table 8.1), are found in small (≤ 50 μmø) clasts replacing enstatite, whereas larger clasts of enstatite are transformed into majorite (Table 8.1) or contain untransformed enstatite in their kernel. Similarly olivine at the border of the melt vein and in clasts within the vein is transformed to ringwoodite and wadsleyite (two references for many: Tomioka and Miyahara 2017, Hu and Sharp 2017). The melt vein matrix is composed of a jadeitic (Tomioka and Miyahara 2017; Hu and Sharp 2017; Ghosh et al. 2021) or albitic clinopyroxene (Ma et al. 2022d. (Table 8.1), periclase (Per80-90Wst 10–20), iron, and troillite, and reflects crystallization upon cooling during rarefaction (Tschauner et al. 2014, see Fig. 8.2). In Acfer 040 the shock melt vein matrix contains the high-pressure mineral akimotoite (Sharp et al. 1997, Table 8.1). The release of the dynamic compression state in the shock melt veins is controlled (a) by the release of the shock state in the meteorite parent body (spall or disruption occurs late in the release process, when the stress state drops below the Hugoniot plastic limit of the bedrock) and (b) by temperature release that is controlled by the temperature gradient between the melt and the much cooler bedrock: During dynamic compression the pressure, temperature, and latent heat of shocked melts are correlated, a marked T-gradient implies spatial differences in shock impedance which cause turbulent mixing on the time scale of the particle velocity of the shock compression state (order of few to several km/s) and this turbulent mixing controls the cooling process at high particle velocity (Fig. 8.2). The observation of bridgmanite as mineral in shock-transformed clasts in such veins defines a fiducial point of pressure and temperature and it also constrains the cooling path (Tschauner et al. 2014; Ma et al. 2016) because bridgmanite vitrifies at low pressure at very modest temperatures on fast time scales (Nishi et al. 2022). In sum, the shock release path is divided in three regimes (Fig. 8.2): (a) An initial isentropic release path, (b) a regime of rapid cooling at high pressure controlled by turbulent mixing and T-homogenization of the melt, (c) a modest to low pressure regime at temperatures below 1000 K to nearly ambient. In chondrites the bulk rock Hugoniot pressure appears generally higher than the pressures indicated by the shock melt vein minerals and it has been proposed that the latter form during rarefaction (Fritz et al. 2017; Hu and Sharp 2017). However, it should be noted that the dynamic pressure in a solid and in coexisting melt is generally not equal because part of the shock-induced change in energy is dissipated through the motion and mixing of the melt. Stress equilibration depends on shock-duration and may not be achieved on the time scale of the chondrite-shock metamorpism. In terrestrial impactites this appears to be different (see next section).

Fig. 8.3
2 microscopic images of solid particles. A, shock melt pockets of ahrensite, bridgmanite, wustite, and olivine. B, shock melt pocket of Tissintite which is in the form of a cluster of irregularly shaped particles.

High-pressure minerals ahrensite, bridgmanite, wüstite and tissintite in shock melt pockets from the Tissint Martian meteorite (Ma et al. 2015, 2016)

Full size image

In the Martian meteorite class of the shergottites shock-induced melt pockets are much more common than melt veins, indicating either a much shorter duration of the shock-state or formation within the isobaric core of impacts of much smaller scale than the L-chondrite parent body disruption. Models of the probability of escape of ejecta from the gravitation of Mars indicate that the shergottites formed at the outer region of the impact (Head et al. 2002) and indirectly support the former hypothesis. The high-pressure minerals tissintite (hpI), donwilhelmsite (hpII) and stishovite (hpII) (Table 8.1) have been reported from lunar meteorites which are all highly shocked.

8.1.1.2 Terrestrial Impactites

The thick atmosphere of Earth decelerates asteroids that are captured by Earth’s gravitation. Only objects of more than 60–100 tons, but depending on impact angle, initial velocity relative to Earth, and density, retain sufficient velocity to generate shock compression in the ground and subsequent crater formation. Many asteroids burst in the higher atmosphere. Hence, the number of terrestrial impacts is comparatively much less than that observed on the Moon or on Mars, even when taking into account that on Earth many craters have been eliminated through later tectonic processes.

Shock states in terrestrial impact craters are assessed through a shock-metamorphic scale that is primarily based on planar deformation features in quartz and feldspars, the transformation of quartz to diaplectic glass (see Stöffler et al. 2018), formation of maskelynite (see above) and high-pressure minerals (Table 8.1). In addition a scale between crater and impactor size allows for estimating dynamic compression states through hydrodynamic modeling. Shock duration in impacts on the scale of the Nördlinger Ries (ø24km), Manicouagan (ø85km), and the Chixculub impact (ø170km) is on the scale of minutes.

High pressure minerals have been found in shocked bedrock (Agarwal et al. 2016) or in xenoliths of bedrock that was trapped in impact breccias (Stähle et al. 2011, 2022) and exhibit a similar fabric as shock meteorites: heavily deformed bed rock, eventually with diaplectic silica and feldspar, and shock melt veins which contain high-pressure minerals and intermediate pressure minerals at their rims. Thus, the overall appearance of shock-metamorphic features in terrestrial impactites is similar to that of highly shocked meteorites. Differences are the result of (a) the different composition of terrestrial continental crust, compared to Martian and lunar crust and to primitive meteorites, and (b) the much longer duration of the dynamic compression state in many terrestrial impactites. In consequence of the longer shock duration the melt vein matrix can contain high-pressure minerals like majoritic garnet (Stähle et al 2011; Ma et al. 2022b) or stöffleritez and albitic clinopyroxene (Ma et al. 2022c). Because of the composition and mineralogy of terrestrial continental crust, partially different, alkaline- and alkline-earth rich high-pressure minerals like zagamiite and accessory high-pressure minerals like high-pressure polymorphs of ilmenite, rutile and zircon are observed in terrestrial impactites but have not been found in meteorites (El Goresy et al. 2010; Stähle et al. 2011; Tschauner et al. 2020a,b, see Table 8.1). Recently, water-bearing intermediate pressure minerals were reported from shock metamorphized berdrock xenoliths from the Ries (Stähle et al. 2022). Tektites are quenched melted impact ejecta (Stöffler et al. 2018). Their composition is quite similar and more controlled by ion vapor pressure than the bedrock composition (Magna et al. 2011), thus, they are carriers of high-temperature rather than high-pressure minerals. Similarly, and despite their extremely high peak shock pressures, impact melt rocks from the former isobaric core of the impact site and pyroclastic impact melt breccia (‘suevite’) show generally the imprint of their formation at high temperatures which upon release of the shock state remains high for longer time than the stress state. Thus, these impact-related rocks contain mostly high-temperature minerals although diamond has been found in suevite (El Goresy et al. 2001a, b) and xenoliths of shocked bedrock that are entrapped in suevite contain high-pressure minerals (see above, Table 8.1). Neither in terrestrial nor meteoritic shock-metamorphic mineralogy many minerals without stability field are observed: Lingunite, stöfflerite, and poirierite are the three undisputed cases (Table 8.1). This observation contrasts with the large number of more or less metastable structures that have been computed. The discrepancy is not entirely result of kinetics because of both, terrestrial and meteoritic shock-events lack these occurrences, whereas sub-ms shock experiments have yielded transitory metastable phases of silica (Luo et al. 2001). Rather, the absence of a larger number of transitory silicate phases indicates sterical hindrance of the Si[4] → [6] transformation.

8.1.1.3 High-Pressure Minerals from the Earth's deep mantle

Terrestrial high-pressure minerals from below 410 km depth are essential constituents of Earth but beyond our access. Only diamond and a few inclusions in diamond have been identified as pristine minerals from the deep Earth. Besides diamond the following high-pressure and intermediate-pressure minerals have been identified, that is: both their structure and composition have been described (see Table 8.1): breyite, davemaoite, deltanitrogen, ice-VII, ringwoodite, the 10 Å-phase, further garnets with high majorite component have been reported. In addition, minerals with stability fields that range from ambient to elevated or high pressure such as iron, periclase, jeffbenite, and larnite have been found (e.g. in Stachel et al. 2000). Deltanitrogen is a product of exsolution of N from diamond (Navon et al. 2017). It is remarkable that the remaining four minerals are hydrous (ice-VII) (Tschauner et al. 2018a), ringwoodite., (Gu et al. 2022), and the 10 Å-phase (Huang et al. 2020) or have been found in diamond which contain ice-VII (davemaoite, Tschauner et al. (2021b)). Garnet coexisting with the 10 Å-phase indicates a formation pressure of 14–15 GPa (Huang et al. 2020) based on the independent barometric scales by Collerson et al. (2010) and Tao and Fei (2021). Trace elements of this inclusion gave similar patterns as expected for HiMU-source region (and it is noted that Pb isotopes could not be measured along with trace elements). Because of the high yield strength of diamond, inclusions may retain elevated pressures and high-pressure crystal structures. The remnant pressure of inclusions at 300 K is the end point of a P–T path whose initial point represents the conditions of entrapment of the inclusion in the growing diamond. Reconstruction of these paths based on isochores (Schrauder and Navon 1993), isomekes (path of stress equilibrium between host and guest phase, e.g. in Anzolini et al. 2016), and paths that account for viscoelastic deformation of the hosting diamond (Wang et al. 2021) have been proposed. Chap. 7 of this book describes diamonds and their inclusions in more detail. The present discussion is constrained to intermediate-and high-pressure minerals (hPI and hPII minerals) that actually have been reported as inclusions in diamonds. Hypothetical retrograde transformation products are not discussed here. The few observations of high-pressure minerals suggest that the Earth’s water- and carbon cycle extends into the lower mantle. This point follows from the observation of hydrous minerals, ice-VII (Tschauner et al. 2018a), and ringwoodite (Gu et al. 2022), the fact that these minerals were entrapped in growing diamond, and the tentative assessment of the depth of entrapment. Furthermore, three global horizons of extensive metasomatism may exist in the Earth’s mantle are potential hosts of a rich intermediate and high-pressure mineralogy that witnesses mobilization of less common elements and are probed by diamonds. These metasomatic horizons may provide incompatible elements to the upper mantle through active and passive upwellings and are replenished through subduction. The mineralogy of the deep Earth has been thought as void of the rich variety of mineral species that occur at the Earth’s surface. Variety of species represents enrichment of less common elements. The three zones of potentially rich mineralogy in the mantle are marked by presence of fluids and melts that allow for mobility of these elements, which then may be enriched in accessory phases: (a) The lithosphere-asthenospheric boundary, (b) possibly the UM-TZ boundary, (c) the TZ-LM boundary and the shallow lower mantle. The mineralogy of the metasomatized lithosphere and the lithosphere-asthenospheric boundary is not discussed in this chapter that is dedicated to high-pressure minerals. It shall only be mentioned that minerals like the silicates Ti-and hydroxyl-clinohumite,the titanates carmichealite, priderite, and minerals of the mathiasite-haggeryite series mark a regime of high fluid mobility and enrichment of incompatible elements in the upper mantle (Haggerty 1991; Wang et al. 1999) and are related to the formation of K-rich volcanism that, in part, carries diamonds to the surface. Diamonds which form in the lithospheric mantle contain ocassionaly minerals whose constituent species are minor or trace elements in the average mante such as goldschmidtite (Meyer et al. 2019) and perovskite. A second global layer of fluid or melt or o horizon that contains regions of fluid and melt enriched in elements that are incompatible in the upper mantle has been proposed to exist at the boundary between the transition zone and the upper mantle (Bercovici and Karato 2003). This hypothesis is consistent the observation of diamond inclusions from that depth that give trace element patterns consistent with at least some types of OIB volcanites (Huang et al. 2020). The partially very alkaline-rich inclusions reported by Stachel et al. (2000) from localities in Southamerica have been hypothesized to originate in the lower mantle (Stachel et al. 2000) but experiments (Litasov et al. 2014; Bulatov et al. 2019; Fedoraeva et al. 2019), geobarometry (Anzolini et al. 2016), and the mineralogy of these inclusions (Brenker et al. 2021) indicate formation in the deep upper mantle or shallow transition zone, and rather support the hypothesis of an enriched, mobile boundary layer between transition zone and upper mantle than processes in the lower mantle. The observation of ice-VII inclusions (Tschauner et al. 2018a), hydrous ringwoodite (Gu et al. 2022), and K-rich davemaoite (Tschauner et al. 2021b, 2022d) from the deep transition zone or lower mantle suggest a third region of extensive regional mantle metasomatism between 600-860 km depth–given that the assessment of entrapment conditions is correct (Wang et al. 2021; Tschauner et al. 2021b; Gu et al. 2022). However, it is not known if these occurrences represent local, regional or global phenomena in the deep mantle.

8.2 High Temperature Minerals–Definition

The concept of induced changes in valence electron configuration works well for defining high-pressure minerals. Hence, it may be applied to high-temperature minerals as well. The regime of temperatures that induce changes in valence electron configuration is achieved for the solid state at pressures where the melting curves are sufficiently high. However, this regime is barely explored by observation in nature or by experiment. Ringwoodite-Q and ahrensite-Q are silicate spinels with partial inversion and involve a spinel endmember component Si[]SiO4 that makes up to 30 mol% in these minerals. They form as solidus phases in shock-melt pockets of picritic to komatiitic bulk composition (Table 8.1) and may be labeled a intermediate-pressure/temperature phases. In nearly all environments minerals form in paragenesis with other minerals or phases of different composition. Under conditions of very high temperature, redox reactions with gases or coexisting minerals and melts can stabilize redox states that do not occur at temperatures in the common range of igneous or metamorphic processes in the Earth’s crust. The temperature-induced intersections of redox reactions at the given O2-fugacity (Essene and Fisher 1986) provide a criterion for high-temperature minerals that is conceptionally related to the criterion for high-pressure minerals (Sect. 8.1) and describes well the occurrences of minerals in early solar condensates, tektites, fulgurites, and impact melts. It is noted that many of these minerals, carbides, silicides, alloys like cohenite and khamrabaevite (Table 8.2), are not bound to high temperatures-they occur under sufficiently reducing conditions at much lower temperatures or at high pressures as well. Some genuine high-temperature minerals like cristobalite owe their formation to large entropic components. However, the decrease of the vibrational relative to ground state energy with decreasing temperature commonly induces distortive phase transitions or order–disorder transitions such as for cristobalite, tridymite, isocubanite which convert to lower symmetric, partially ordered phases, which are observed as minerals. Many minerals that occur in former high-T environments are likely products of such transitions such as panguite and kangite (Fig. 8.4; Table 8.2). As in the case of high pressures, there are also minerals that have natural stability fields at both, low and high temperatures such as corundum, zircon, baddeleyite, thorite, thortveitite. In advance of a more rigorous classification we focus here on minerals that form at very high temperature where the relevant redox buffer reactions have stabilized valences that are not stable under typical conditions of igneous and metamorpic processes on Earth. This regime of mineral formation includes presolar minerals, minerals that formed by sublimation in the solar nebula as first or early condensates, minerals in fulgurites, tektites, and former impact melts. The use of modern micro-analysis techniques has greatly extended our knowledge about these minerals which are recognized as carriers of information about processes in the early solar nebula through their isotopic record, trace elements and formation conditions (Rubin and Ma 2017). Presolar minerals can be carriers of isotope anomalies that are result of nucleonic processes during novae or supernovae. Other high temperature minerals occur in volcanic, i.p. phreatomagmatic, environments and in pyrometamorphic rocks such as the Hatrurim formation in the Near East.

Table 8.2 Recently-identified primary high-temperature minerals in refractory inclusions from the solar nebula that have formed by sublimation (‘condensates’)
Full size table
Fig. 8.4
2 microscopic images. A includes olivine, panguite, and davisite. B, a cluster of thin, curvy structures in dark and light shades. It labels kangite, davisite, and warkite all around the spine in the center.

Ultrarefractory minerals panguite and davisite from the Allende CV3 meteorite (Ma et al. 2012), kangite, warkite and davisite from the DOM 08,004 CO3 meteorite (Ma et al. 2020)

Full size image