An evolutionary system of mineralogy, Part VIII: The evolution of metamorphic minerals

Native elements

Two polymorphs of carbon (C), graphite and diamond, with 105 and 10 occurrences in Online Materials^[1] Table S3, respectively, are the only native element minerals that occur in any significant abundance in metamorphic rocks, though more than two dozen rare native elements and metal alloys are listed in Online Materials^[1] Table S1. Graphite most frequently occurs in carbon-rich metapelites (Landis 1971; Diessel et al. 1978; Buseck and Huang 1985). However, graphite is also reported to occur as a product of retrograde metamorphism of diamond, at times as euhedral pseudomorphs (Pearson et al. 1989; Ferry 1992; Davies et al. 1993; Leech and Ernst 1998). In addition, Ross et al. (1991) reported an occurrence of graphite formed by the shearing of coal—an example of Phanerozoic metamorphism of a biotic protolith.

Most diamond formation occurs in the mantle by precipitation from carbon-rich fluids (Jacob and Mikhail 2022; Kjarsgaard et al. 2022). However, in some instances, micrometer-scale diamond forms during subduction, ultra-deep metamorphism, and subsequent rebound of carbon-bearing crustal wedges (Dobrzhinetskaya et al. 1995, 2022).

Sulfides

More than 50 sulfide and other chalcogenide minerals have been reported from metamorphic rocks (Online Materials^[1] Table S1). However, only four iron-bearing sulfides, pyrite (FeS₂), pyrrhotite (Fe₇S₈), chalcopyrite (CuFeS₂), and pentlandite [(Ni,Fe)₉S₈], are relatively common. Pyrrhotite is the most commonly observed sulfide in our survey, occurring in 192 metamorphic rocks. It frequently occurs with graphite in higher-grade regional metamorphic rocks (Hoschek 1984; Ferry 1992), and it may form by solid-state transformation of pyrite coupled with sulfur loss (Bowles et al. 2011). It is also a common sulfide mineral associated with skarn deposits (Einaudi and Burt 1982).

Pyrite (FeS₂) is the next most frequently reported metamorphic sulfide, with 120 occurrences in Online Materials^[1] Table S3. Pyrite is found in all metamorphic grades, including high-pressure regimes, and may form through solid-state transformation of pyrrhotite (Hall et al. 1987). Pyrite also occurs via contact metamorphism associated with skarn formation (Einaudi et al. 1981) and in shear zones (e.g., Harker 1950).

Chalcopyrite, the only copper-bearing mineral among the most common metamorphic phases, is reported from 99 rocks in our survey. Most occurrences are as minor grains in regionally metamorphosed igneous and sedimentary lithologies (Ferry 1984, 1992, 1994; Ferry et al. 2001). It is likely that chalcopyrite, as an opaque and often micrometer-scale phase, is underrepresented in our study.

The 21 occurrences of pentlandite, the only common nickel-bearing metamorphic minerals, are recorded exclusively from ultramafic lithologies (Ferry 1995; Ferry et al. 2005).

In addition, it should be noted that sphalerite (ZnS) occurs in 9 of the rocks surveyed—a number that likely significantly underestimates the frequency of this opaque and typically minute zinc-bearing phase.

Oxides

Oxide minerals occur in the entire range of metamorphic rocks. More than 50 species, most of them relatively rare, have been reported (Online Materials^[1] Table S1). We list 13 mineral kinds among the most frequently encountered metamorphic minerals.

The simple oxide rutile (TiO₂) is reported in 297 of the metamorphic rocks from our study (Online Materials^[1] Table S3), most often in regional metamorphic rocks by transformation of prior Ti-bearing phases, including titanite, ilmenite, and titaniferous micas and magnetite (Bowles et al. 2011), but also in a range of pyrometamorphic (Grapes 2006), contact metamorphic (Reverdatto 1973), and shear zone (Harker 1950) rocks.

Additionally, among the more common simple oxides is corundum (Al₂O₃; 85 occurrences), observed most frequently in silica-poor lithologies subjected to high temperature, notably in contact metamorphic zones. Al-rich lithologies with prominent corundum, often in association with spinel, mullite, cordierite, and/or sanidine, are known as “emery.” Corundum also occurs in pyrometamorphosed limestone xenoliths (Joplin 1968; Grapes 2006), as well as in high-grade regional metamorphic rocks (Harker 1950; Augustithis 1985).

Periclase (MgO) is a common mineral in calcite/dolomite marbles, often in association with brucite (Carpenter 1967; Bowles et al. 2011). As early as 1940, Bowen described periclase as part of an evolutionary thermal metamorphic sequence (Bowen 1940). We record 36 occurrences of periclase, primarily in limestone xenoliths and contact metamorphic environments (Augustithis 1985; Ferry and Rumble 1997; Grapes 2006).

Hematite (Fe₂O₃), represented by 24 occurrences in our tabulation, forms from Fe-rich protoliths in oxidized environments. Contexts of metamorphic hematite include xenoliths, contact environments, iron formations, and regional metamorphism (Bowles et al. 2011).

We record 18 occurrences of the zirconium oxide baddeleyite (ZrO₂) in both contact and regional metamorphic contexts (Ferry 2007). That number that may underrepresent the frequency of baddeleyite because of its typical low modal abundance and small grain size.

The most frequently encountered metamorphic oxides are members of the spinel group, which occur in 672 (24%) of the 2785 metamorphic rocks we tabulated. With the general formula [(Mg,Fe²⁺)(Al,Fe³⁺,Cr³⁺,Ti)₂O₄], the spinel group encompasses a complex range of solid solutions (Bowles et al. 2011), of which four end-members are among the most commonly reported metamorphic minerals: magnetite (Fe2+Fe23+O4), spinel (MgAl₂O₄; although “spinel” may also refer to the mineral group rather than the species in some modes), chromite (Fe2+Cr23+O4), and hercynite (Fe²⁺Al₂O₄). In addition, several reports cite “pleonaste,” which refers to Fe²⁺-bearing intermediate compositions of the spinel-hercynite solid solution-examples that we include with spinel.

Magnetite is by far the most frequently reported metamorphic oxide, occurring in 429 (15%) of the 2785 metamorphic rocks we compiled, and in a wide range of metamorphic contexts (Joplin 1968; Reverdatto 1973; Grapes 2006). Magnetite often forms by thermal alteration of ferric iron oxide/hydroxides, as well as through the introduction of Fe-rich fluids—a situation in which the distinction between metamorphism and metasomatism may be blurred.

The Mg-Al oxide spinel, with 263 occurrences in our list, is found in numerous contact and regional metamorphic contexts, principally as a high-temperature mineral in metacarbonates and Al-rich protoliths (Harker 1950; Botha 1983; Carswell 1990; Grapes 2006).

Hercynite, which forms in high-grade metamorphic environments, often in association with corundum, mullite, and/or sillimanite, is represented by 34 examples in Online Materials^[1] Table S3. Metamorphic contexts include xenolith, contact, and regional environments (Harker 1950; Grapes 2006; Bowles et al. 2011).

We list 12 occurrences of metamorphic chromite—a number that likely underestimates this most common chromium mineral because it is opaque and easily mistaken for other Fe-bearing oxides. Chromite is most often associated with ultramafic lithologies, particularly ophiolites.

Four titanium-bearing double oxides are listed among the 94 most common metamorphic phases. We record ilmenite (FeTiO₃) from 181 xenoliths, contact metamorphic rocks, or regional metamorphic formations, notably forming via alteration of mafic and Fe-bearing lithologies (Reverdatto 1973; Carswell 1990; Grapes 2006). Geikielite (MgTiO₃), the magnesium analog of ilmenite, occurs in 17 contact and regional metasedimentary rocks in Online Materials^[1] Table S3. Geikeilite protoliths include carbonates, calc silicates, pelites, and sandstones. Perovskite (CaTiO₃) is reported from 12 rocks, primarily in xenoliths and contact metamorphic environments with impure limestone (Murdoch 1951; Fulignati et al. 2000; Grapes 2006). Pseudobrookite (Fe23+TiO5), which we record in 15 primarily xenolith and contact metamorphic rocks, forms most often by the high-temperature oxidation of ilmenite (Agrell and Langley 1958; Smith 1969; Basta and Shaalan 1974).

Brucite [Mg(OH)₂], the only relatively common hydroxide mineral in metamorphic rocks, occurred in 31 rocks of our survey. It is typically the consequence of hydration of periclase in altered Ca-Mg carbonates (Nakajima et al. 1992; Ferry and Rumble 1997; Ferry et al. 2002; Bowles et al. 2011).

Carbonates

At least 70 carbonate minerals have been reported from metamorphic rocks (Online Materials^[1] Table S1), though only seven species occur with any significant frequency. By far the most abundant carbonates are calcite (CaCO₃), dolomite [CaMg(CO₃)₂], and the intermediate composition Fe-dolomite [dolomite with a significant ankerite CaFe(CO₃)₂ component], with 645, 152, and 116 occurrences in our compilation, respectively. Calcite and dolomite most often occur in metamorphosed limestones and other carbonate-bearing protoliths (Harker 1950; Reverdatto 1973; Chang et al. 1996; Grapes 2006). In many instances, these phases form through recrystallization of prior carbonates to form a marble (Chang et al. 1996; Philpotts and Ague 2009). Fe-dolomite (often reported as “ankerite” in the petrologic literature) is found primarily in regionally metamorphosed sediments (Ferry 1992, 1994, 2007; Ferry et al. 2015) where it formed through carbonation reactions during metamorphism/ metasomatism (Spooner and Fyfe 1973).

Metamorphic magnesite (MgCO₃) with 16 occurrences is characteristic of altered ophiolites, where it occurs in association with talc and serpentine. It typically forms via the carbonation of Mg-bearing oxides and silicates (Chang et al. 1996).

Aragonite, a high-pressure form of CaCO₃, was recorded in 15 rocks of eclogite, blueschist, and ultrahigh-pressure facies carbonate rocks (Carswell 1990; Carswell and Compagnoni 2003; Philpotts and Ague 2009).

Two silicate carbonates, spurrite [Ca₅(SiO₄)₂(CO₃)] with 51 examples and tilleyite [Ca₅Si₂O₇(CO₃)₂] with 16 examples, arise when calcite and wollastonite react at high temperature (Tuttle and Harker 1957; Zharikov and Shmulovich 1969).

Phosphates

Of the more than 50 metamorphic phosphates recorded in Online Materials^[1] Table S1, only the calcium phosphate apatite [Ca₅(PO₄)₃(F,OH)] is widely reported, with 155 occurrences in Online Materials^[1] Table S3. We lump two common species, fluorapatite and hydroxylapatite, which are rarely differentiated in reports of metamorphic mineral modes. The majority of these apatite occurrences are in high-grade metamorphosed mafic igneous rocks, including granulites and eclogites (Harker 1950; Joplin 1968; Carswell 1990). In addition, apatite has been reported from contact metamorphic and shear environments (Harker 1950; Joplin 1968).

Silicates

Silicates constitute the majority of metamorphic minerals, both volumetrically and in terms of diversity. In Online Materials^[1] Tables S1 and S2 we record 746 silicate mineral species, corresponding to 418 root natural kinds, of which 66 are frequently encountered metamorphic phases. In the following sections we review these more common silicates.

Nesosilicates or orthosilicates

Orthosilicates, with silicon exclusively in insular SiO44− structural groups, are characteristic minerals in environments with relatively low Si, notably those associated with carbonate, calc-silicate, or aluminous protoliths (Deer et al. 1982). We detail 21 orthosilicates in addition to the orthosilicate-carbonate mineral spurrite, described above. Of these 22 phases, 16 contain essential Ca and/or Al.

Olivine Group: We recognize four members of the olivine group [(Mg,Fe,Ca)₂SiO₄] as important metamorphic minerals (Deer et al. 1982). The Mg olivine forsterite (ideally Mg₂SiO₄) is reported in 173 of the rocks we surveyed, notably via contact, regional, or high-pressure metamorphism of silica-poor ultramafic (Springer 1974; Pinsent and Hirst 1977) or carbonate-bearing (Weeks 1956; Schreyer et al. 1972; Suzuki 1977) protoliths. We also distinguish olivine as intermediate Mg-Fe compositions with 0.3 < Fe/(Fe+Mg) < 0.7, which are common in metamorphosed mafic and ultramafic rocks (53 occurrences; Reverdatto 1973; Ferry et al. 1987).

Fayalite (ideally Fe₂SiO₄) is much less common, occurring in 10 rocks with Fe-rich protoliths, including mafic rocks and iron formations (Joplin 1968; Simmons et al. 1974; Floran and Papike 1978; Carswell 1990). Note that metamorphic Fe-dominant olivines with intermediate compositions are less common than examples close to either end-member (Deer et al. 1982).

Monticellite (CaMgSiO₄), with 77 occurrences, is commonly found in contact metamorphic environments with siliceous carbonate lithologies, often forming with increasing temperature at the expense of diopside, forsterite, and/or wollastonite (Bowen 1940; Turner 1967; Deer et al. 1982). Monticellite frequently co-occurs with forsterite, as the solid solution between these two olivine group minerals is limited (Warner and Luth 1973).

Garnet Group: The garnet group is represented by four relatively common metamorphic phases, occurring in 745 (27%) of 2785 rocks in our survey. Garnets collectively display a significant compositional range, typically with solid solutions among two or three end-members (Deer et al. 1982; Chiama et al. 2020, 2022). Ideal end-members of these minerals are almandine [Fe32+Al2(SiO4)3], andradite [Ca3Fe32+(SiO4)3], grossular [Ca₃Al₂(SiO₄)₃], and pyrope [Mg₃Al₂(SiO₄)₃], often with a significant spessartine [Mn32+Al2(SiO4)3] component, as well, though true Mn-dominant spessartine is recorded in only 9 occurrences in our compilation (Woodland 1939; Roy 1965; Jan and Symes 1977).

In some instances, such as pyrope-almandine-spessartine (“pyralspite”) from ecologites and other high-grade metamorphic rocks, grossular-andradite (“grandite”) from the contact metamorphism of carbonate-bearing sediments, and contact metamorphic garnets in the grossular-spessartine-almandine field (Shimazaki 1977), the compositional ranges among end-members may be continuous, thus warranting lumping of species into a single metamorphic mineral kind. However, until cluster analysis (Gregory et al. 2019; Boujibar et al. 2021; Hystad et al. 2021) can be performed on a wide range of garnet compositions from known paragenetic environments, we will treat these five types of metamorphic garnet separately.

Almandine, with 367 occurrences in Online Materials^[1] Table S3, is the commonest garnet in metamorphic rocks (Harker 1950; Joplin 1968; Reverdatto 1973; Botha 1983; Augustithis 1985). Most almandine forms in a regional metamorphic context, derived from mafic or pelitic protoliths (Atherton 1964; Deer et al. 1982), including high-pressure examples from blueschist (Coleman and Lee 1963; Banno and Matsui 1965), eclogite (Coleman et al. 1965), and granulite (Buddington 1952; Eskola 1952) facies. In addition, almandine from contact metamorphism of pelites is not uncommon (Tilley 1926; Stewart 1942), while it also occurs in some metamorphosed iron formations (Klein 1966).

Pyrope’s 192 entries are overwhelmingly from high-pressure metamorphic environments, in many instances from ecologite-grade rocks with mafic precursors, most commonly in association with omphacite (Carswell 1990). Metamorphic pyrope typically has a significant almandine component (Deer et al. 1982).

The great majority of 162 grossular occurrences in Online Materials^[1] Table S3 arise from contact metamorphism of calcareous rocks, often in association with diopside and/or wollastonite (Watters 1958; Reverdatto 1973), or in regional metamorphic formations, also with carbonate-bearing protoliths (Tilley 1927; Sylvester and Anderson 1976). In addition, 28 occurrences of the Ca-Fe³⁺ garnet andradite arise predominantly from contact and regional metamorphism of calc-silicate rocks (Harker 1950; White 1959; Shedlock and Essene 1979). In several instances, contact metamorphic garnets have so-called “grandite” compositions intermediate between grossular and andradite (Coombs et al. 1977; Tulloch 1979).

Three additional Ca-Mg orthosilicates, bredigite [Ca₇Mg(SiO₄)₄] with 6 occurrences (Tilley and Vincent 1948; Grapes 2006), larnite (Ca₂SiO₄) with 24 occurrences (Deer et al. 1986), and merwinite [Ca₃Mg(SiO₄)₂] with 38 occurrences (Larsen and Foshag 1921; Reverdatto 1973), are frequently found in contact metamorphosed calc-silicate protoliths, often in association with the calc-silicates melilite, rankinite, and spurrite (Joplin 1968; Deer et al. 1986; Grapes 2006). Merwinite is also reported as an ultrahigh-pressure mantle phase (Zedgenizov et al. 2014).

We lump three compositionally similar IMA species—humite, clinohumite, and hydroxylclinohumite—into humite [Mg_7–9(SiO₄)₄(F,OH)₂]. Members of the humite group differ in the ratios of two structural modules, one of forsterite composition [Mg₂(SiO₄)] and the other of brucite composition [Mg(OH,F)₂]. Reports of metamorphic mineral modes seldom distinguish between humite [Mg₉(SiO₄)₄(OH,F)₂] and clinohumite [Mg₇(SiO₄)₄(OH,F)₂], nor between the OH- and F-dominant species. We record 21 occurrences of humite, all of which are characteristic of the contact metamorphism of dolomite-bearing sediments (Tilley 1951; Joplin 1968; Deer et al. 1982). Note that two other members of the humite group (Van Valkenburg 1961), norbergite [Mg₃(SiO₄)₄(OH,F)₂] and chondrodite [Mg₅(SiO₄)₄(OH,F)₂], are also contact metamorphic minerals, but did not appear as common phases in our tabulations of metamorphic rock modes.

Three aluminosilicate (Al₂SiO₅) polymorphs, andalusite, kyanite, and sillimanite, are abundant constituents of many metapelites, with 146, 102, and 235 occurrences in Online Materials^[1] Table S3, respectively. These phases, which can coexist at their invariant triple point (~500 °C and 0.4 GPa; Hodges and Spear 1982; Bohlen et al. 1991; Pattison 2001), have received special attention for their ability to document the pressure-temperature regimes of their host rocks (Barrow 1893; Zen 1969; Deer et al. 1982; Whitney 2002; Philpotts and Ague 2009). Sillimanite, the high-temperature, low-pressure polymorph, is a common phase in various metapelites subjected to hornblende hornfels, granulite, and pyrometamorphic (sanidinite) conditions (Reverdatto 1973; Botha 1983; Grapes 2006). Mullite [Al_4+2xSi_2–2xO_10–x (x ≈ 0.4)] is also a high-temperature, low-pressure orthosilicate that we record from 62 pyrometamorphic rocks, often in association with sillimanite (Grapes 2006).

Andalusite forms in pelitic protoliths at low pressure and moderate temperature (<770 °C), notably from albite-epidote hornfels and hornblende hornfels facies (Read 1923; Guitard 1965; Reverdatto 1973). Kyanite, the highest-pressure crustal polymorph of Al₂SiO₅, is frequently encountered in regional and high-pressure metamorphic rocks with aluminous precursors (Carswell 1990; Carswell and Compagnoni 2003). The aluminosilicates may record either prograde or retrograde metamorphism. For example, Lal (1969) described andalusite and kyanite formed via retrograde metamorphism from cordierite-bearing rocks, and Gates and Speer (1991) record retrograde kyanite after sillimanite in metapelite shear zones.

Chloritoid [(Fe²⁺,Mg,Mn²⁺)Al₂O(SiO₄)(OH)₂], with 46 occurrences in our tabulation, is most commonly formed by regional or high-pressure metamorphism of pelitic rocks (Joplin 1968; Carswell 1990). We lump three IMA-CNMNC-approved species, chloritoid [Fe²⁺Al₂O(SiO₄)(OH)₂], magnesiochloritoid [MgAl₂O(SiO₄)(OH)₂], and ottrelite [Mn²⁺Al₂O(SiO₄)(OH)₂], because they form a continuous solid solution and they are rarely differentiated in reports of metamorphic rock modes. The broad pressure-temperature stability field of chloritoid leads to a wide range of assemblages, from low-grade, clay-mineral- and phengite-bearing facies to high-grade rocks with kyanite, pyrope-almandine, and/or staurolite (Halferdahl 1961).

Staurolite [(Fe²⁺,Mg)₂Al₉Si₄O₂₃(OH)] is another common phase derived by regional or high-pressure metamorphism of pelitic sediments. Its 52 occurrences in Online Materials^[1] Table S3 reflect a range of P-T conditions of formation, from low-grade assemblages with chloritoid and quartz, medium-grade assemblages with almandine and kyanite, and high-grade assemblages with sillimanite and plagioclase (Deer et al. 1982; Augustithis 1985; Carswell 1990). Staurolite is also observed in the contact metamorphism of pelites (Reverdatto 1973; Grapes 2006).

Titanite (CaTiSiO₅; commonly reported as “sphene”) occurs as a minor phase in 186 metamorphic rocks in our survey in various contexts (Harker 1950; Joplin 1968; Reverdatto 1973; Carswell 1990). Zircon (ZrSiO₄), another volumetrically minor phase, is listed in 66 of 2785 rocks in our tabulation, including a wide range of contact and regional metamorphic lithologies (Joplin 1968; Carswell 1990; Grapes 2006). Titanite and zircon are particularly durable accessory minerals that are widespread in igneous and sedimentary formations; therefore, their occurrence in metamorphic rocks sometimes derives from protolith minerals that have been little altered.

Though not among the more common metamorphic orthosilicates, willemite (Zn₂SiO₄) is an important mineral in some metamorphosed Pb-Zn deposits, such as the sillimanite-grade deposits at Franklin, New Jersey (Pinger 1950; Frondel 1990). In this case, which may be relevant to minerals in many metamorphic environments, willemite occurs both as a major phase in the ore and as a secondary phase in thin hydrothermal veins. These two generations of willemite, furthermore, have distinct properties: both forms are fluorescent, but only the secondary willemite has persistent luminescence as a consequence of its greater arsenic content (Rakovan and Waychunas 1996). With their distinct modes of formation and attributes, these coexisting forms of willemite represent two different mineral kinds in our evolutionary system.

Sorosilicates or disilicates

Sorosilicates incorporate the double-tetrahedron pyrosilicate group (Si2O76−). We find 8 sorosilicate root natural kinds, corresponding to at least 35 IMA-CNMNC-approved species, among the 94 most frequently encountered metamorphic mineral kinds. All of these phases, in addition to the disilicate-carbonate mineral tilleyite described earlier, are calcium-bearing minerals that occur most frequently in the contact metamorphic zones of limestone and dolomite.

The most common metamorphic sorosilicates are from the diverse epidote group (Deer et al. 1986; Armbruster et al. 2006). We lump 4 monoclinic species (including Fe³⁺-bearing clinozoisite) into epidote [Ca₂(Al₂Fe³⁺)[Si₂O₇][SiO₄]O(OH)] with 149 occurrences; 5 rare-earth element-bearing epidote group minerals into allanite [(CaCe)(AlAlFe²⁺)O[Si₂O₇][SiO₄](OH)] with 16 occurrences (though certainly under-reported); and orthorhombic zoisite [Ca₂Al₃[Si₂O₇][SiO₄]O(OH)] with 156 occurrences. In addition, Mn-bearing piemontite [Ca₂Al₂Mn³⁺(Si₂O₇)(SiO₄)O(OH)] is an important phase in metamorphosed manganese deposits, though we record only 7 occurrences. Metamorphic epidote is found most commonly via contact metamorphism of carbonate-bearing sediments and mafic igneous rocks (Joplin 1968; Reverdatto 1973), but also in regional (Harker 1950), high-pressure (Carswell 1990), and xenolith (Grapes 2006) contexts. Note, however, that it may be difficult to distinguish metamorphic epidote and zoisite (see below) from examples formed by metasomatism (Joplin 1968).

Zoisite generally forms at lower metamorphic grades than epidote, though it can coexist with epidote in medium-grade regional metamorphic rocks derived from calcareous sediments or mafic igneous rocks (Myer 1966; Ackermand and Raase 1973; Raith 1976). Zoisite is also common in kyanite-bearing ecologite (Carswell and Compagnoni 2003), where it may form by a prograde reaction from lawsonite (Deer et al. 1986).

We lump two species, lawsonite and itoigawaite, into the root kind lawsonite [(Ca,Sr)Al₂(Si₂O₇)(OH)₂·H₂O], with 11 occurrences in Online Materials^[1] Table S3. Lawsonite forms exclusively in high-pressure blueschist or eclogite facies rocks (Philpotts and Ague 2009), commonly in association with glaucophane and an epidote group mineral, either zoisite or epidote (Coleman et al. 1965; Carswell 1990).

Rankinite (Ca₃Si₂O₇), with 20 occurrences, is found exclusively in contact metamorphic rocks, commonly in association with larnite, melilite, spurrite, and/or wollastonite (Reverdatto 1973; Grapes 2006).

The melilite group includes the solid solution between åkermanite [Ca₂(Al₂SiO₇)] and gehlenite [Ca₂(MgSi₂O₇)], as well as alumoåkermanite [(Ca,Na)₂(Al,Mg,Fe²⁺) (Si₂O₇)]—minerals that we lump into the root mineral kind melilite. With 108 occurrences in Online Materials^[1] Table S3, melilite is a common mineral in pyrometamorphosed siliceous limestone and dolomite, particularly at pyroxene hornfels and sanidinite facies (Reverdatto 1973; Grapes 2006), often forming at the expense of diopside or anorthite (Bowen 1940; Reverdatto 1970).

Pumpellyite [Ca₂(Al,Fe²⁺,Fe³⁺)₃(Si₂O₇)(SiO₄)(OH,O)₂·H₂O] lumps 9 closely related species of Ca-Al-Fe sorosilicates that are found most frequently in the low-grade zeolite and pumpellyite-prehnite facies of regional metamorphism. We list 10 occurrences, all of which occur in low-grade metapelites (Joplin 1968; Botha 1983; Augustithis 1985). However, Deer et al. (1986) note that Al-rich pumpellyite also occurs in blueschist facies, and Fe- and Mn-rich pumpellyite may occur in mineralized skarn zones.

We lump 10 IMA-CNMNC-approved species, most of which are rare compositional variants, into vesuvianite [(Ca,Na)₁₉(Al,Mg,Fe)₁₃(SiO₄)₁₀(Si₂O₇)₄(OH,F,O)₁₀]. We record 15 occurrences in contact metamorphosed limestone, in which it is a characteristic skarn mineral, commonly in association with diopside, grossular, and/or wollastonite (Harker 1950). Vesuvianite occurs less commonly in regional metamorphosed limestones (Tilley 1927; Deer et al. 1982). Note that, as with the example of epidote, it may be difficult to distinguish vesuvianite formed by metamorphism vs. metasomatism (Joplin 1968).

Cyclosilicates

Members of the cordierite, tourmaline, and osulmilite groups are relatively common metamorphic cyclosilicates. We lump the species cordierite (with Mg) and sekaninaite (with Fe²⁺) into the root mineral kind cordierite [(Mg,Fe²⁺)₂Al₄Si₅O₁₈], which, with 395 occurrences in Online Materials^[1] Table S3, is among the most common minerals in contact and regionally metamorphosed pelites (Joplin 1968; Reverdatto 1973; Botha 1983; Grapes 2006). Deer et al. (1986) detail a wide range of cordierite parageneses, including pyrometamorphosed xenoliths, contact metamorphosed argillaceous sediments, and a range of regional metamorphic facies, including low-pressure, high-temperature assemblages with andalusite; moderate-pressure assemblages with sillimanite and garnet; and high-pressure assemblages with kyanite.

Tourmaline [(Na,Ca,□)(Mg,Al,Fe²⁺,Fe³⁺,Ti⁴⁺)₃(Al,Fe³⁺,Mg)₆(Si₆O₁₈)(BO₃)₃ (OH)₃(O,F)] is the only common boron-bearing mineral in metamorphic rocks (Deer et al. 1986; Henry et al. 2011; Henry and Dutrow 2012). We lump 18 IMA-CNMNC-approved species of the tourmaline group (Online Materials^[1] Table S1), all of which have been reported from regional metamorphic environments. The 47 tourmaline occurrences listed in Online Materials^[1] Table S3, including several examples of tourmaline-dominant tourmalinites, are from metapelites (Harker 1950; Joplin 1968). Joplin (1968) suggests that metamorphic tourmaline occurs in 3 distinct ways: as a remnant mineral of the protolith, through metamorphism of a borate-containing lithology, or as the result of boron metasomatism.

Osumilite [(K,Na)(Fe²⁺,Mg)₂(Al,Fe³⁺)₃(Si,Al)₁₂O₃₀] was reported in 13 of our mineral modes, typically from ultrahigh-temperature metamorphosed pelites, in which it commonly occurs with cordierite, orthopyroxene, sanidine, and/or sillimanite (Harley 2021).

Inosilicates

Among the 94 relatively common metamorphic minerals listed in Table 2, 16 are chain silicates, including several members of the pyroxene (7 kinds) and amphibole (7 kinds) groups (Deer et al. 1997a, 1997b). Inosilicates are one of the most common classes of metamorphic minerals, occurring in 1387 (50%) of the 2785 metamorphic rocks in Online Materials^[1] Table S3.

Pyroxene Group: We consider 7 root kinds of pyroxene group single-chain silicates, most of which lie in or near the [(Ca,Mg,Fe)₂Si₂O₆] quadrilateral (Deer et al. 1997a).

Orthopyroxene lumps 281 occurrences of orthorhombic pyroxenes, most often described as enstatite (the Mg end-member) or “hypersthene” (with Mg~Fe²⁺) but sometimes “bronzite” (with Mg > Fe²⁺) or ferrosilite (the Fe²⁺ end-member), always lying close to the MgSiO₃–Fe²⁺SiO₃ binary. Orthopyroxene most often occurs in granulite, ecologite, and UHT facies of metamorphosed ultramafic and mafic igneous rocks, in which it is often the most abundant mafic phase (Joplin 1968; Augustithis 1985; Carswell and Compagnoni 2003). It also occurs in lower-grade regional metamorphic rocks (Joplin 1968; Botha 1983), with iron-rich examples in metamorphosed iron formations (Kranck 1961; Simmons et al. 1974). Orthopyroxene is not uncommon in contact metamorphic environments, including pyrometamorphosed xenoliths (Reverdatto 1973; Grapes 2006). In some instances, Mg-rich orthopyroxene is associated with carbonate minerals (Schreyer et al. 1972; Ohnmacht 1974).

Three Ca-bearing clinopyroxenes, diopside [Ca(Mg,Fe²⁺)Si₂O₆; 409 occurrences in Online Materials^[1] Table S3), hedenbergite (CaFe²⁺Si₂O₆; 16 occurrences), and the ternary solid solution augite [(Ca,Mg,Fe²⁺)₂Si₂O₆, typically with 0.5 < Ca/(Mg+Fe) < 0.9; 150 occurrences], are common in a wide variety of metamorphic rocks (Deer et al. 1997a). Pyroxenes close to the continuous CaMg–CaFe²⁺ solid solution between diopside and hedenbergite are most typical of thermally metamorphosed carbonate and calc-silicate rocks, occurring in xenoliths and contact metamorphic contexts (Grapes 2006). We define diopside broadly to include most intermediate Mg-Fe compositions (e.g., “salite” and “ferrosalite”), as well as Al-bearing “fassaite.” More than 80% of occurrences of diopside, the most abundant pyroxene in our survey, arise from contact metamorphism. Diopside also occurs in regional and high-pressure metamorphic rocks, with several examples from amphibolite facies (Harker 1950; Augustithis 1985) and eclogite facies (Carswell 1990; Carswell and Compagnoni 2003).

Hedenbergite displays much the same parageneses as diopside, but with Fe-rich protoliths (Joplin 1968; Augustithis 1985). We debated whether to lump these two end-members, but hedenbergite appears to form a discrete cluster of metamorphic clinopyroxenes with low Mg. Cluster analysis of igneous and metamorphic clinopyroxenes represents an important future research goal.

The IMA-CNMNC-approved species augite, including clinopyroxenes in the [(Ca,Mg,Fe²⁺)₂Si₂O₆] system with Ca occupying between ~25 and ~45 atom percent of the Ca-Mg-Fe sites, is equivalent to our root kind augite. The 150 occurrences in Online Materials^[1] Table S3, while mostly from contact metamorphism or pyrometamorphism of ultramafic/mafic lithologies (Reverdatto 1973; Grapes 2006), also include representatives derived by regional metamorphism of mafic and intermediate igneous protoliths (Harker 1950; Joplin 1968; Ferry et al. 1987).

Three Na-bearing kinds of clinopyroxene, aegirine [(Ca,Na)(Fe³⁺,Mg,Fe²⁺) Si₂O₆; 16 occurrences], jadeite (NaAlSi₂O₆; 34 occurrences), and omphacite [(Ca,Na) (Mg,Fe,Al)Si₂O₆; 151 occurrences] are especially characteristic of high-pressure metamorphic environments. Aegirine, in which we lump two IMA-CNMNC-approved species aegirine (NaFe³⁺Si₂O₆) and aegirine-augite [(Ca,Na)(Fe³⁺,Mg,Fe²⁺)Si₂O₆], is the least common of these phases in metamorphic rocks, being found primarily in the context of mafic and intermediate igneous rocks subjected to ultrahigh pressure and ecologite facies (Carswell 1990), though aegirine also occurs via contact metamorphism of alkaline rocks (Reverdatto 1973). A complication is the formation of aegirine through sodium metasomatism of prior pyroxenes (Moore 1973; Deer et al. 1997a).

Jadeite is a diagnostic phase found exclusively in high-pressure metamorphic environments, including blueschist facies, eclogite facies, and ultrahigh-pressure metamorphic rocks (Carswell 1990). It often forms through the iconic reaction albite → jadeite + quartz (Deer et al. 1997a, and references therein). Jadeite commonly incorporates up to 15 mol% of an aegirine/omphacite component; however, a significant compositional gap separates jadeite from these phases, which often coexist in high-pressure assemblages (Coleman and Clark 1968).

Omphacite, which represents a solid solution among aegirine, diopside, and jadeite, is a relatively common phase in high-pressure metamorphic rocks. All 151 occurrences in our tabulation were reported from blueschist, eclogite, or ultrahigh-pressure environments, most often with ultramafic or mafic protoliths and often in association with glaucophane, pyrope/almandine, and quartz/coesite (Carswell 1990; Carswell and Compagnoni 2003).

Pyroxenoid Group: Two members of the inosilicate pyroxenoid group, wollastonite (CaSiO₃; 158 occurrences) and the Mn-bearing rhodonite [CaMn₃Mn(Si₅O₁₅); 8 occurrences, not listed among the top 94 phases], are most commonly associated with skarn zones. Almost all wollastonite reports are from carbonate or calc-silicate protoliths subjected to pyroxene hornfels or sanidinite facies metamorphism (Reverdatto 1973; Grapes 2006).

Rhodonite, which lumps 4 closely related species of Mn pyroxenoids, is most frequently encountered in the high-pressure metamorphic environments of Mn-rich protoliths (Carswell 1990), notably by reaction of rhodochrosite (MnCO₃; Hori 1962), though it can also form through Mn metasomatism (Bilgrami 1956).

Amphibole Group: The amphibole group of double-chain silicates, which boasts more than 110 IMA-CNMNC-approved species (https://rruff.info/ima; accessed 13 January 2023), is likely the most diverse of all mineral structure types (Deer et al. 1997b; Hawthorne et al. 2012). Here, we provisionally lump 55 amphibole species known to occur in metamorphic rocks into 7 root mineral kinds. It should be noted, however, that the variety of amphibole parageneses, coupled with the extensive and complex solid solutions and miscibility gaps among many species, render any suggestion of amphibole mineral kinds tentative, at best. In the context of metamorphism, we have yet to determine if different facies, different protoliths, effects of metasomatism, prograde vs. retrograde formation, and other factors may yield numerous distinct combinations of paragenesis and attributes. We require data resources with analyses of tens of thousands of well-characterized amphibole specimens, coupled with advanced methods of cluster analysis (Boujibar et al. 2021; Hystad et al. 2021). Such an epic endeavor could represent a lifetime of fruitful study for an ambitious young mineralogist.

Anthophyllite [□(Mg,Fe²⁺)₂(Mg,Fe²⁺,Fe³⁺,Al)₅(Si,Al)₈O₂₂(OH)₂; with 52 occurrences in Online Materials^[1] Table S3], lumps 6 species of the complex anthophyllite/ferro-anthophyllite/gedrite/ferro-gedrite solid solution of orthorhombic amphiboles (Ferré 1989). An unresolved question regards the possible presence of a miscibility gap in this system between Al-rich (at times with Na) and Al-poor orthoamphiboles (Hawthorne et al. 1980; Spear 1982). If so, then at least two root mineral kinds would be warranted. Most of the examples in our compilation arise from hornblende-hornfels or pyroxene-hornfels facies contact metamorphism of ultramafic/mafic igneous rocks or pelitic sediments, often in association with biotite, cordierite, and quartz (Reverdatto 1973). We also record several instances of anthophyllite in regional metamorphic settings, including amphibolite and granulite facies metamorphism of pelites and ultramafic rocks (Joplin 1968).

The closely related clinopyroxenes cummingtonite [□Mg₂Mg₅Si₈O₂₂(OH)₂; with 12 occurrences] and grunerite [ ◻Fe22+Fe52+Si8O22(OH)2; with 8 occurrences, hence not listed among the top 94] are the monoclinic polymorphs of anthophyllite and ferro-anthophyllite. All but one of the cummingtonite examples in our tabulation, with Mg/ (Mg+Fe) generally >0.4 (i.e., in some instances with Fe > Mg), are from metapelites subjected to hornblende-hornfels facies contact metamorphism (Reverdatto 1973). The more iron-rich grunerite examples, by contrast, are primarily from amphibolite or granulite facies regionally metamorphosed iron formations (Joplin 1968; Kimball and Spear 1984). Therefore, we provisionally distinguish these two closely related mineral kinds based on paragenetic mode, even though they may display continuous solid solution between the Mg and Fe²⁺ end-members. It should be noted that as a result of miscibility gaps, cummingtonite and grunerite often occur in assemblages with multiple amphiboles, including calcic hornblende, Al-bearing anthophyllite, and/or a sodic amphibole (Deer et al. 1997b).

Several calcic clinoamphiboles are common constituents of metamorphic rocks. Tremolite [ ◻Ca2(Mg5.0−4.5Fe0.0−0.52+)Si8O22(OH,F)2; with 122 occurrences] is typically an almost pure Ca-Mg phase (i.e., low Fe²⁺) formed from ultramafic/mafic igneous or calc-silicate sediments in contact, regional, and high-pressure metamorphic environments. Most of the examples in Online Materials^[1] Table S3 are from muscovite-hornfels, hornblende-hornfels, or pyroxene-hornfels contact metamorphic zones, in which tremolite is associated with diopside, dolomite, grossular, talc, and/or other Ca-Mg phases (Reverdatto 1973). We lump the OH^–- and F^–-bearing species, which display a complete solid solution and share the same paragenesis.

Actinolite [□Ca₂(Mg,Fe²⁺)₅Si₈O₂₂(OH,F)₂; with 74 occurrences] lumps the species actinolite and ferro-actinolite, spanning a range 0.9 > Mg/(Mg+Fe²⁺) > 0 (Deer et al. 1997b). Though chemically and structurally similar to tremolite, actinolite is distinguished both by its greater Fe²⁺ content and by its common association in metapelites or metabasites with biotite, epidote or zoisite, and/or quartz in high-pressure, regional, or contact metamorphic settings (Harker 1950; Joplin 1968; Botha 1983; Carswell 1990).

We lump 26 IMA-CNMNC-approved metamorphic species of Ca-(±Na,K)-clinoamphiboles into hornblende [(Na,K)Ca₂(Mg,Fe²⁺,Al,Fe³⁺)₅(Si,Al)₈O₂₂ (OH,F,Cl)₂; with 387 occurrences]. This complex group displays significant compositional plasticity, with solid solutions among Na, K, and vacancies in alkali sites; Mg-Fe²⁺-Fe³⁺-Al in octahedral sites; and Al-Fe³⁺-Si in tetrahedral sites, as well as among OH, F, and Cl (Deer et al. 1997b; Hawthorne et al. 2012; see Online Materials^[1] Tables S1 and S2). Hornblende, a defining phase in hornblende-hornfels and amphibolite facies rocks, appears in numerous metamorphic environments, including contact metamorphism (albite-epidote to sanidinite facies; Reverdatto 1973; Grapes 2006), regional metamorphism (greenschist to granulite facies; Harker 1950; Joplin 1968), and high-pressure metamorphism (blueschist to eclogite facies; Reverdatto 1973; Carswell 1990). Hornblende protoliths, similarly, span a wide range of igneous, sedimentary, and metamorphic rocks. Hornblende in metamorphic rocks commonly coexists with other amphiboles, including anthophyllite, cummingtonite, and grunerite (Deer et al. 1997b). Given the complexity of this mineral group, cluster analyses of numerous hornblende samples based on composition, paragenesis, and mineral associations, would doubtless reveal many distinct kinds of hornblende.

We lump a wide range of Na-Ca clinoamphiboles into richterite, defined here as [(□,Na)(NaCa)(Mg,Fe²⁺,Al,Fe³⁺)₅(Si,Al,Fe³⁺)₈O₂₂(OH)₂]. Although we record only 12 occurrences in Online Materials^[1] Table S3, most of which were originally described as barroisite [nominally (□NaCa)(Mg₃Al₂)(Si₇Al)O₂₂(OH)₂] or winchite [(□NaCa) (Mg₄Al)Si₈O₂₂(OH)₂], we lump 16 IMA-CNMNC-approved species into the root natural kind richterite, while acknowledging that much more work is needed to fully characterize these minerals and their associated parageneses. Most of the richterite occurrences that we record are metamorphosed mafic rocks from eclogite facies, almost always in association with omphacite, pyrope, and rutile (Binns 1967; Carswell 1990), though it is reported from regionally metamorphosed basalt, as well (Iwasaki 1960).

The sodium amphibole glaucophane [□Na₂(Mg,Fe²⁺)₃Al₂Si₈O₂₂(OH)₂; with 79 occurrences] lumps 2 species, glaucophane and ferro-glaucophane, which form a Mg-Fe²⁺ solid solution. All of the examples in Online Materials^[1] Table S3 are from high-pressure metamorphic rocks (most commonly blueschist or eclogite facies, but also ultrahigh-pressure facies) of mafic/intermediate igneous rocks or Mg-bearing sediments (Augustithis 1985; Carswell 1990).

One additional inosilicate group, sapphirine [(Mg,Fe²⁺,Al,Fe³⁺)₈O₂(Al,Si)₆O₁₈, 28 occurrences], is important as a key indicator of the temperature (>900 °C) of ultra high-temperature metamorphic rocks formed from ultramafic protoliths (Monchoux 1972; Deer et al. 1997a; Carswell and Compagnoni 2003; Harley 2021). It commonly occurs in association with orthopyroxene and sillimanite.

Phyllosilicates

Mica Group: With 1244 occurrences in Online Materials^[1] Table S3 (45% of the 2795 rocks surveyed), the mica minerals are prominent constituents of many metamorphic lithologies (Guidotti 1984; Fleet 2003). We lump 15 IMA-CNMNC-approved species into five root mineral kinds of micas: biotite, phlogopite, phengite, muscovite, and paragonite. However, there undoubtedly exist many more kinds of metamorphic micas, the identification of which will require the construction of extensive mica databases and application of cluster analysis (Hazen 2019). In particular, widely occurring metamorphic biotite and muscovite will likely be split into multiple natural kinds.

We define the familiar group of dark-colored, Fe²⁺-bearing trioctahedral mica species as biotite [ KFe22+(Fe2+,Mg,Mn2+)(Si,Al,Fe3+)2Si2O10(OH,F,Cl)2; 822 occurrences]. We lump 6 IMA-CNMNC-approved species, which are themselves rarely identified in the petrographic literature. Few minerals occur in as a diverse array of metamorphic environments as biotite, which we record from low-pressure pyrometamorphic, contact, regional, and high-pressure metamorphism, typically of pelites, but also of ultramafic, mafic, intermediate, acidic, and (rarely) agpaitic igneous rocks, as well as Fe-bearing impure carbonate and calc-silicate protoliths (Harker 1950; Joplin 1968; Reverdatto 1973; Botha 1983; Carswell 1990; Fleet 2003; Grapes 2006).

We also lump 6 Mg-dominant species of trioctahedral micas, which are generally lighter in color than biotite, as phlogopite [KMg₂(Mg,Fe²⁺,Mn²⁺,Fe³⁺,Ti⁴⁺) (Si,Al,Fe³⁺)₂Si₂O₁₀(OH,F)₂; with 78 occurrences]. Though often combined with biotite in some descriptions of mica (e.g., Fleet 2003), and consequently sometimes reported as biotite in the petrologic literature, phlogopite displays distinct mineral associations in metamorphosed high-Mg, low-Fe environments, including ultramafic and dolomitic carbonate protoliths (Joplin 1968; Reverdatto 1973). Owing to their diverse parageneses and compositional range, trioctahedral micas represent yet another mineral group that is ripe for investigation by cluster analysis.

The dioctahedral aluminous mica muscovite [K(Al,Fe³⁺,Cr)₂(Si₃Al)O₁₀(OH)₂; 515 occurrences], commonly reported with the varietal names illite, phengite, or sericite, is most characteristic of regionally and contact metamorphosed pelites (Reverdatto 1973; Philpotts and Ague 2009), which represent most of the occurrences recorded in Online Materials^[1] Table S3. Muscovite, among the first phases to form during diagenesis of clay minerals (Fleet 2003), also occurs in a wide range of other contexts, including contact and regional metamorphosed arkosic, calc-silicate, and impure carbonate sediments (Philpotts and Ague 2009), as well as various igneous protoliths (Harker 1950; Carswell 1990).

We also include the fine-grained, Si-rich white mica phengite (110 occurrences) as a separate kind, even though it falls under IMA’s definition of muscovite. Phengite is most commonly associated with high-pressure metamorphism (Carswell 1990).

The sodium trioctahedral mica paragonite [NaAl₂(Si₃Al)O₁₀(OH)₂; 67 occurrences] is characteristic of high-pressure eclogite facies metamorphism of mafic igneous rocks, often in association with glaucophane, kyanite, omphacite, and/or pyrope (Carswell 1990). Paragonite, often intimately intermixed with phengite (with which it has limited solid solution; e.g., Thompson and Thompson 1976; Guidotti et al. 1994), also occurs in low- and medium-grade metapelites, in which it can form by both prograde and retrograde reactions (Chatterjee 1970; Guidotti 1984; Guidotti and Sassi 1998; Fleet 2003). Paragonite also frequently co-occurs with the so-called “brittle mica” margarite [CaAl₂(Si₂Al₂)O₁₀(OH)₂] in a wide range of metamorphic grades of metasediments (Guidotti 1984; Fleet 2003); however, margarite is relatively rare in comparison to the micas described above, being reported from only one of the metamorphic rocks in our survey (Carswell and Compagnoni 2003, Table 2 therein).

Other phyllosilicates: More than 30 other IMA-CNMNC-approved layer silicates occur in metamorphic rocks (Online Materials^[1] Table S1). Most of these minerals (e.g., apophyllite group, gillespite, pyrophyllite, stilpnomelane, zussmanite) occur only rarely in metamorphic rocks. Note that we do not list diagenetically formed clay minerals as metamorphic phases (Wilson 2013); they will be considered further in Part IX of this series. However, chlorite, prehnite, serpentine, and talc are included in our list of 94 relatively common metamorphic phases (Deer et al. 2009).

Chlorite [(Mg,Fe²⁺)₅(Al,Fe³⁺)(Si₃AlO₁₀)(OH)₈; 339 occurrences] encompasses 3 IMA-CNMNC-approved species of Mg-Fe²⁺-Al-(Fe³⁺) layer silicates: chamosite, clinochlore, and sudoite. Chlorite is common in pelites subjected to greenschist and amphibolite facies metamorphism (Harker 1950; Joplin 1968; Botha 1983), as well as from muscovite-, hornblende-, and pyroxene-hornfels facies contact metamorphism of pelites and basic igneous rocks (Joplin 1968; Reverdatto 1973), often in association with albite, biotite, muscovite, and/or quartz. In addition, Coleman et al. (1965) and Carswell (1990) record more than a dozen examples of chlorite in eclogite facies high-pressure metamorphism. Chlorite forms through various pathways, including prograde and retrograde metamorphism, both with and without external aqueous fluids. As such, chlorite represents a mineral whose often uncertain parageneses grade continuously from regional or “burial” metamorphism to metasomatism to hydrothermal alteration (Deer et al. 2009).

We lump 5 IMA-CNMNC-approved species, including three structural variants of Mg₃Si₂O₅(OH)₄ (antigorite, chrysotile, and lizardite), aluminous amesite, and Fe- bearing greenalite, into serpentine [(Mg,Fe²⁺,Al,Fe³⁺)₃(Al,Si)Si(OH)₄; 68 occurrences]. Most examples in Online Materials^[1] Table S3 are from low- to moderate-grade regional metamorphism of ultramafic/mafic igneous or pelitic protoliths (Joplin 1968; Philpotts and Ague 2009), in which they form primarily by retrograde/hydrothermal reactions from olivine and Mg-rich pyroxene or by prograde metamorphism of serpentinite (Deer et al. 2009, and references therein). Given its varied modes of formation, coupled with multiple polymorphs, cluster analysis of metamorphic serpentine is warranted.

Talc [Mg₃Si₄O₁₀(OH)₂; 73 occurrences] is characteristic of Mg-rich protoliths over a wide range of pressure-temperature condition. Examples include thermal metamorphism of dolomite-bearing sediments (Tilley 1948; Reverdatto 1973; Augustithis 1985), high-pressure (blueschist and eclogite facies) metamorphism of basic igneous rocks (Chopin 1981; Carswell 1990), and greenschist to amphibolite grade regional metamorphism of ultramafic rocks (Harker 1950; Joplin 1968). Of special note are high-pressure to ultrahigh-pressure (>0.6 GPa) talc-kyanite-(quartz/coesite) assemblages known as “whiteschists” (Schreyer 1977), which form in the Mg-Al-Si-H (“MASH”) system, at times with P_H2O approximately equal to the total pressure.

Prehnite [Ca₂Al(Si₃Al)O₁₀(OH)₂; 16 occurrences], though most familiar as a hydrothermal phase associated with zeolites in amygdaloidal basalt, is also common in the eponymous prehnite-pumpellyite facies of low-grade regional metamorphism (Coombs 1960; Philpotts and Ague 2009). Prehnite, often in association with chlorite and quartz, is also present occasionally in metamorphosed basic igneous and pelitic sedimentary rocks from zeolite to lower amphibolite grades (Harker 1950; Joplin 1968; Botha 1983), at times the result of retrograde reactions (Coombs 1993).

Tectosilicates

A wide range of framework silicates, including the silica group, feldspars, feldspathoids, and zeolites (Deer et al. 2001, 2004), occur in metamorphic rocks, with one or more examples reported in 1595 (67%) of the 2785 rocks surveyed in Online Materials^[1] Table S3. We focus attention on 10 mineral kinds that occur most frequently.

Silica Group: Four silica group minerals—quartz, high-pressure coesite, and high-temperature cristobalite and tridymite—span the entire range of metamorphic environments, occurring in all but the most Si-deficient rocks (Deer et al. 2004).

Quartz (SiO₂), with 1353 occurrences in our study (49% of rocks in Online Materials^[1] Table S3), is the most common metamorphic mineral. It occurs in pyrometamorphosed xenoliths, contact metamorphic rocks, metamorphosed iron-manganese formations, high-pressure and regional metamorphic settings, metasomatized rocks, and shear zones (Online Materials^[1] Table S3 and references therein), often by recrystallization of protolith quartz (Deer et al. 2004). Quartz is stable in all crust and upper mantle pressure-temperature regimes except above ~2.7 GPa, where it transforms to coesite, or at low pressure above ~850 °C, where cristobalite and tridymite are the stable silica phases.

Coesite (SiO₂; 23 occurrences) is restricted to ultrahigh pressure (>2.7 GPa) metamorphic environments, where it is typically associated with kyanite, omphacite, and/or pyrope (Carswell and Compagnoni 2003). Several reports describe coesite as inclusions in upper mantle phases, including pyrope (Chopin 1984; Schertl et al. 1991) and diamond (Stachel et al. 2022, and references therein).

Cristobalite (10 occurrences) and tridymite (44 occurrences) are high-temperature, low-pressure polymorphs of SiO₂ that occur almost exclusively in sedimentary rocks that have been thermally metamorphosed (pyroxene hornfels or sanidinite facies) by basic igneous rocks (Agrell and Langley 1958; Reverdatto 1973; Black 1989; Grapes 2006). These two minerals co-occur in 7 of the 10 reported rocks with cristobalite. Tridymite and cristobalite also are associated in some burning coal deposits with temperatures that may exceed 1100 °C (Bustin and Mathews 1982; Grapes 2006), and therefore are the consequence of Phanerozoic biological precursors (to be considered in Part XII).

Feldspar Group: Metamorphic feldspar group minerals display compositions close to two binary solid solutions (Deer et al. 2001): the Na-Ca plagioclase feldspars and the Na-K alkali feldspars. In both instances we suggest modifications of the nomenclature approved by the IMA-CNMNC.

In the case of the plagioclase series [(CaAl,NaSi)AlSi₂O₈], we identify albite as compositions close to NaAlSi₃O₈ [Na/(Na+Ca) > 0.85 and often with >10 mol% KAlSi₃O₈], anorthite as close to CaAl₂Si₂O₈ [Ca/(Ca+Na) > 0.85], and plagioclase as having intermediate compositions between ~An₂₀ and ~An₇₀ as valid root mineral kinds. We justify this division based on the existence of the so-called peristerite and Huttenlocker miscibility gaps between ~An_2–17 and An_65–88 respectively. As a consequence, several authors have recorded coexisting albite-plagioclase and plagioclase-anorthite pairs (Evans 1964; Botha 1983).

Albite, with 177 occurrences in Online Materials^[1] Table S3, is observed in a wide range of thermal, regional, and high-pressure metamorphic environments, with both igneous and sedimentary protoliths (Harker 1950; Joplin 1968; Reverdatto 1973; Philpotts and Ague 2009). Anorthite (78 occurrences) is more restricted in its occurrences, being found primarily as a contact metamorphic mineral derived from calc-silicate and carbonate-bearing sediments (Grapes 2006), though it is also found in regionally metamorphosed mafic and calc-silicate rocks from amphibolite to granulite facies (Joplin 1968). Plagioclase (809 occurrences), like quartz, albite, and kspar (see below), is not a particularly diagnostic phase in metamorphic rocks because it occurs across the full spectrum of thermal, regional, and high-pressure metamorphic environments, with an equally broad range of igneous, sedimentary, and metamorphic protoliths. For a given protolith, the anorthite content of plagioclase tends to increase with metamorphic grade (Deer et al. 2001). Note that most reports of “plagioclase” in the older metamorphic literature lack compositional information; thus, some of these occurrences may fit our definitions of albite or anorthite.

The alkali feldspars are complicated by the existence of three K-rich (KAlSi₃O₈) variants—the higher-temperature (>500 °C) monoclinic sanidine and two lower-temperature triclinic phases, microcline and orthoclase, which are often reported as “kspar” in the literature of metamorphic petrology. An additional consideration is that alkali feldspars of intermediate compositions often exsolve Na- and K-rich phases, typically reported as “perthite.” In our study, we adopt the name kspar for microcline and orthoclase, and record both albite and kspar for perthite.

Sanidine (102 occurrences) is most commonly found in thermally metamorphosed rocks of pyroxene hornfels or sanidinite grade (Grapes 2006), though it also has been reported from ultrahigh-pressure metamorphism of pelites (Carswell 1990; Carswell and Compagnoni 2003) and ultrahigh-temperature regimes (Harley 2021).

Kspar (411 occurrences) has been reported from thermal (zeolite to pyroxene hornfels facies), regional (amphibolite to granulite facies), and high-pressure (eclogite to ultrahigh-pressure facies) metamorphic environments. Protoliths for kspar include mafic, acidic, and agpaitic igneous rocks and arkosic, pelitic, and carbonate-bearing sedimentary rocks (Harker 1950; Joplin 1968; Reverdatto 1973; Carswell 1990; Deer et al. 2001).

At the temperatures of UHT metamorphism (>900 °C), an additional complication is the occurrence of Ca-Na-K ternary feldspars, which typically exsolve to a perthite with coexisting plagioclase and alkali feldspar lamellae (Harley 2008; Harley 2021, Fig. 20 therein).

Scapolite Group: Three species of the scapolite group are lumped as scapolite [(Na,Ca)₄Al₃(Al,Si)₃Si₆O₂₄(CO₃,SO₄,Cl), with 33 occurrences]. Scapolite is not infrequently observed in medium- to high-grade contact (hornblende-hornfels and pyroxene hornfels facies) and regional (amphibolite and granulite facies) metamorphosed pelites and calc-silicate sediments (Joplin 1968; Reverdatto 1973), and has also been reported from the albite-epidote hornfels facies contact metamorphism of amygdaloidal basalt (Joplin 1968).

Zeolite Group: The zeolite facies is the lowest pressure-temperature regime of metamorphism, with temperatures <200 °C at pressures <0.3 GPa (Philpotts and Ague 2009). Most zeolite minerals form via low-temperature aqueous processes, including fluid interactions with cooling basalt and authigenesis (Deer et al. 2004). Nevertheless, some zeolite occurrences are attributed to metamorphism, sensu stricto. Though not sufficiently abundant to include among the most common metamorphic phases, analcime (2 occurrences), laumontite (1), and wairakite (2), as well as 6 undifferentiated reports of “zeolite,” were listed in modes of low-grade metamorphosed mafic igneous rocks and pelites (Joplin 1968).

Feldspathoid Group: None of the members of the sub-silicic feldspathoid framework silicate group is common in metamorphic rocks. Nevertheless, kalsilite, leucite, and nepheline (with 5, 2, and 4 occurrences, respectively) are representative of undersaturated pyrometamorphosed calc-silicates (Grapes 2006).

To these framework silicates, we add silicate glass [(Si,Al,Ca,Mg,Fe)O; SiO₂ >70 wt%; with 68 occurrences] as an important yet often poorly characterized metamorphic phase in thermal metamorphic environments, particularly of arkosic sandstones and pelites (Reverdatto 1973; Grapes 2006). Two varietal names of metamorphic glass are “buchite,” which forms when a silica-rich pelitic rock is altered by igneous contact, and “porcellanite,” a glass derived from pyrometamorphosed clay, marl, shale, or bauxite (Grapes 2006). Melting and glass formation may occur as low as ~650 °C at 05. GPa in a granite protolith, or >1000 °C at low pressure and dry conditions. Grapes (2006) details how “Si-rich glass” in many pyrometamorphic zones typically contains significant Al₂O₃ and alkalis, a consequence of quartz-feldspar melting. Note that pure SiO₂ melts at 1700 °C—a temperature only attainable by lightning strikes or bolide impacts.

Rarer Metamorphic Minerals: In addition to the 94 mineral kinds outlined above, 661 other mineral kinds occur rarely as trace phases in metamorphic rocks, as documented by reports in numerous primary sources and compilations, notably Anthony et al. (1990–2003) and references cited in https://mindat.org and https://rruff.info/ima (both accessed 20 January 2023). Most of these scarce minerals, which are listed in Online Materials^[1] Table S1, were not recorded from any of the 2785 metamorphic rock modes in Online Materials^[1] Table S3.

However, a few of the less common minerals in Online Materials^[1] Table S1, were also noted in one or more metamorphic rock modes. Among these minor minerals, listed alphabetically, are alleghanyite (1 occurrence), analcime (2), anhydrite (2), ankerite (6), ardennite (1), arsenopyrite (1), axinite (4), bornite (1), braunite (7), bredigite (6), brownmillerite (5), bustamite (4), calzirtite (3), celestine (1), chondrodite (4), cuspidine (2), deerite (2), diaspore (1), fluorite (1), fluormayenite (5), friedelite (1), galaxite (1), giuseppeite (1), grunerite (8), hausmannite (1), hillbrandite (1), hügbomite (2), ilvaite (2), jacobsite (1), kalsilite (5), kornerupine (1), kutnohorite (1), laumontite (1), leucite (2), margarite (1), monazite (3), nepheline (4), norbergite (1), piemontite (7), pigeonite (2), pyrophanite (1), pyroxmangite (2), qandilite (3), rhodocrosite (2), rhodonite (8), riebeckite (4), scawtite (1), siderite (9), sonolite (1), sphalerite (9), spessartine (9), stilpnomelane (3), suenoite (5), tephroite (1), thompsonite (3), uvarovite (2), wairakite (2), and “zeolite” (6).