Metamorphism of Mafic Rocks

CHAPTER 9
Metamorphism of Mafic Rocks
9.1
Mafic Rocks
Metamorphic mafic rocks (e. g., mafic schists and gneisses, amphibolites) are
derived from mafic igneous rocks, mainly basalts and andesites, and of lesser
importance, gabbros (Chap. 2). Metamorphic assemblages found in mafic
rocks are used to define the intensity of metamorphism in the metamorphic
facies concept (Chap. 4).
Extrusive mafic igneous rocks, basalts and andesites comprise by far the
largest amount of mafic rocks and greatly outweigh their plutonic equivalents
(Tables 2.1 and 2.2). Basalts and andesites occur as massive lava flows, pillow
lavas, hyaloclastic breccias, tuff layers, sills, and dykes. Basaltic rocks consti-
tute a major portion of the oceanic crust and most basalts appear to have
been subjected to ocean-floor metamorphism immediately after formation at
a spreading ridge. When transported to a continental margin via plate mo-
tion, oceanic mafic rocks are again recrystallized at or near convergent plate
junctions; the alteration in mineralogy depends on whether the oceanic crust
was subducted under a continental plate or was abducted onto continental
crust. On the other hand, andesitic rocks and other related calc-alkaline vol-
canics as well as associated graywacke are the dominant lithologies within is-
land arcs and Pacific-type continental margins. These rocks are subjected to
alteration by hydrothermal fluids as evidenced by present-day geothermal ac-
tivity in many island arc districts or are transformed during burial and oro-
genic metamorphism.
Metamorphosed mafic rocks are very susceptible to changes in tempera-
ture and pressure. This is the reason why most names of individual meta-
morphic facies are derived from mineral assemblages of this rock group,
e. g., greenschist, amphibolite, granulite, blueschist, and eclogite (cf. Chaps. 2
and 4). In addition, mafic rocks that are metamorphosed under very weak
conditions below the greenschist and blueschist facies often show systematic
variations in mineralogy that permits a further subdivision into characteris-
tic metamorphic zones. All these distinct low-grade zones may be given sep-
arate metamorphic facies names if one wishes. However, we prefer to use the
expression subgreenschist facies for very-low-grade conditions of incipient
metamorphism.
K. Bucher et al., Petrogenesis of Metamorphic Rocks

© Springer-Verlag Berlin Heidelberg 2002
280 Chapter 9 Metamorphism of Mafic Rocks
9.1.1
Hydration of Igneous Mafic Rocks
Basalts and gabbros have solidus temperatures in the order of 1200 oc. Con-
sequently, hydrates are not typical members of the solidus mineralogy of ba-
salts and other mafic rocks. At the onset of metamorphism, mafic igneous
rocks and pyroclastics are, therefore, in their least hydrated state as opposed
to "wet" sedimentary rocks that start metamorphism in their maximum hy-
drated state. Because newly formed minerals in metamafics at low tempera-
ture are predominantly hydrous phases, access of water is absolutely essential
to initiate metamorphism. Otherwise, igneous rocks will persist unchanged
in metamorphic terrains. Partial or complete hydration of mafic rocks may
occur during ocean-floor metamorphism in connection with hydrothermal
activity in island arcs or during orogenic metamorphism where deformation
facilitates the influx of water. In metamorphism of mafic igneous rocks it is
not self-evident that the condition Ptotal= PH2 o is continuously maintained
during the prograde reaction history. On the other hand, worldwide experi-
ence with mafic rocks shows that partial persistence of igneous minerals and
microstructures is widespread in subgreenschist facies rocks and it is typical-
ly absent from the greenschist facies upward. Primary igneous mesostruc-
tures such as magmatic layering and pillow structures may be preserved even
in eclogite and granulite facies terrains. Coarse-grained gabbroic rocks have
the best chance to conserve primary igneous minerals up to high-grade
metamorphic conditions. Gabbro bodies often escape pervasive internal de-
formation and this, in turn, prevents access of water, hampers recrystalliza-
tion and hinders hydration of the igneous minerals.
As pointed out in Chapter 3, dehydration reactions are strongly endother-
mic. Consequently, hydration of basalt is exothermic and releases large
amounts of heat. The heat of reaction released by replacing the basalt assem-
blage Cpx + Pl by a collection of low-temperature hydrates such as prehnite +
chlorite+ zeolites could raise the temperature by as much as 100 oc (in a
heat-conserving system). Another interesting aspect of exothermic reactions
is their self-accelerating and feedback nature. Once initiated, they will pro-
ceed as long as water is available. The increasing temperature will initially
make the reactions go faster, but it will also bring the reaction closer to its
equilibrium where it eventually will stop. Also note that the reactions that
partially or completely hydrate igneous mafic rocks are metastable reactions
and the reactions tend to run far from equilibrium. Consider, for instance,
the reaction diopside +anorthite+ H 2 0 =>chlorite+ prehnite±quartz that was
mentioned above. In the presence of water at low temperature, this reaction
will always run to completion and the product and reactant assemblage will
never reach reaction equilibrium. The unstable or metastable nature of the
hydration reactions that replace high-T anhydrous igneous assemblages by
low-T hydrate assemblages has the consequence that low-grade mafic rocks
often show disequilibrium assemblages. Unreacted high-T igneous assem-
blages may occur together with various generations of low-T assemblages of
the subgreenschist facies in intimate spatial association. The hydration pro-
9.1 Mafic Rocks 281
cess commonly leads to the development of zoned metasomatic structures.

Examples are: networks of veins and concentric shells of mineral zones that
reflect progressive hydration of basaltic pillow lavas. The highly permeable
inter-pillow zones serve as aquifers for hydration water and the hydration
process progresses towards the pillow centers. The nature of the product as-
semblages depends not only on the pressure and temperature during the hy-
dration process, but also on the chemical composition of the hydration fluid.
C0 2 in the fluid plays a particularly important role. COz-rich aqueous fluids
may result in altered basalts with modally abundant carbonate minerals (cal-
cite, ankerite). COz-rich fluids also tend to favor the formation of less hy-
drous assemblages compared with pure H2 0 fluids. The cation composition
of aqueous fluids that have already reacted with large volumes of basalt is
dramatically different from that of seawater. Fluid that moves fast through
permeable fractured rocks will have a different composition than fluid that
trickles slowly through porous basalt. Fluid-rock interaction typically also
leads to extensive redistribution of chemical elements in piles of basaltic pil-
low lavas. This may still be witnessed in high-grade terrains. In the Saas-Zer-
matt ophiolite complex of the Central Alps, for example, mesoscopic pillow
lava structures are often preserved with pillow cores of 25 kbar eclogite and
inter-pillow material of nearly monomineralic glaucophane. Finally, the al-
teration products of the fluid-basalt interaction process will, of course, reflect
these compositional differences in the hydration fluid.
Heterogeneous hydration results also in features such as incipient to exten-
sive development, sporadic distribution, and selective growth of low-grade
minerals in vesicles and fractures, and the topotaxic growth of these minerals
after igneous plagioclase, clinopyroxene, olivine, hornblende, opaques, and
volcanic glass. From the viewpoint of igneous petrologists, these minerals are
often referred to as "secondary minerals". Depending on the effective bulk
composition of local domains of the rock, different associations of secondary
minerals may develop in vesicles, veins and irregular patches replacing pri-
mary phases even within a single thin section. Furthermore, relic igneous
phases such as plagioclase and clinopyroxene are common. From the view-
point of metamorphic petrologists the low-grade hydrates and carbonates,
the "secondary minerals", constitute the stable low-grade assemblage and the
starting material of prograde metamorphism. In the conceptually simplest
case, no relic igneous minerals remain in the mafic rock and its igneous as-
semblage has been completely replaced by a stable low-T assemblage in its
maximum hydrated state.
9.1.2
Chemical and Mineralogical Composition of Mafic Rocks
The characteristic composition of alkali-olivine basalt is listed in Table 2.1

and of MOR basalt in Table 2.3. Mid-ocean ridge basalt (MORB) that is pro-
duced along ocean spreading centers is by far the most common type of
igneous mafic rocks. It may serve as a reference composition for all mafic
rocks discussed in this chapter. Mafic rocks are characterized by Si02 con-
tents of about 45-60 wto/o and are also relatively rich in MgO, FeO, CaO and
Al 2 0 3 . It is general custom in petrology to call igneous rocks with 45-
52 wto/o of Si0 2 basic, and their metamorphic derivatives are then called
metabasic rocks or, in short, metabasites. Metabasalts represent the most
commonly encountered group of metabasites. Andesitic rocks, on the other
hand, contain higher Si0 2, Ah0 3 , and alkalis, but lower MgO, FeO, and CaO
than basaltic compositions (e.g., Carmichaell989); basaltic andesites and an-
desites belong to the intermediate igneous rocks, defined by Si0 2 contents of
52-63 wto/o.
Igneous mafic rocks contain appreciable amounts of at least the following
eight oxide components: Si0 2, Ti0 2 , Ah0 3 , Fe 2 0 3 , FeO, MgO, CaO, and Na 2 0.
K2 0, H 2 0, and C0 2 may also be present in small amounts. These components
are stored in relatively few different minerals. The mineralogical inventory of
the mafic protolith comprises the major constituents plagioclase and clino-
pyroxene. In addition, quartz, orthopyroxene, olivine and nepheline can be
present in various associations with Cpx +Pl. Plagioclase and clinopyroxene
are the prime and most common minerals of most mafic rocks and many
gabbros and basalts contain more than 90-95 volo/o of these two minerals.
Other basalts and gabbros may be composed of Pl + Cpx + Opx + Qtz or
Pl + Cpx + 01 + Ne; troctolites contain plagioclase+ olivine only; anorthosites
more than 90 volo/o plagioclase, and so on. However, all rocks are always
combinations of a few different mineral species. A large variety of minor and
Table 9.1. Minerals and compositions in metabasaltic rocks
Nesosilicates Sheet silicates

Garnet (Fe,Mg,Cah(A1,Fe)zSi30t 2 Muscovite KA13Si 30 10 (0H)z
Celadonite KMgAISi 40t 0(0H)z
Sorosilicates Paragonite NaA1 3Si30 10 (0H)z
Kyanite Phlogopite KMg3 [AlSi30t 0](0Hh
Zoisite Biotite K(Mg,Fe,Al,Tih
Epidote [(Al,SihOJO](OH,F,CI)z
Pwnpellyite Clinochlore Mg5 AI 2Si 30JO(OH) 8
Vesuvianite Chlorite (Fe,Mg) 5A}zSi 30 10 (0H) 8
Lawsonite Prehnite Ca 2 A1zSi30to(OH)z
Chloritoid
Tectosilicates
Pyroxenes Quartz Si0 2
Diopsid CaMgSi 20 6 Anorthite CaAI 2 SizOs
Jadeite NaAISi 2 0 6 Albite NaAISi 30 8
Hyperstene (Mg,Fe)zSiz06 Analcite NaAISi2 0 6 • H2 0
Omphacite (Ca,Na)(Mg,Fe,AI)Si 2 0 6 Scapolite C~(A1 2 Si 2 0sh<C0 3 , S04, Cl 2 )
Amphiboles Zeolites
Tremolite Ca 2 MgsSis0dOH)z Laumontite CaAizSi4012 ·4 HzO
Actinolite Caz(Fe,Mg)sSisOzz(OH)z Heulandite CaAizSi701s·6 HzO
Glaucophane Na2 (Fe,Mg) 3 (A1)zSi 8 0zz(OH)z Stilbite CaAlzSi701s · 7 HzO
Barroisite (Ca,Na)z(Fe,Mg,AI) 5 Wairakite CaAlzSi 4 0 1z · 2 HzO
SisOzz(OH)z
Tschermakite Ca 2 (Fe,Mgh(Al)z Carbonates
(AlzSi60zz(OH)z Calcite CaC03
Hornblende (Na,K)Ca 2(Fe,Mg,Al)s Aragonite CaC0 3
(Si,A1) 8 0 22 (0H,F,CI)z Dolomite Ca(Mg,Fe)(C0 3 )z
9.1 Mafic Rocks 283
accessory minerals including ilmenite, magnetite, spinel, garnet, and even

hornblende and biotite can be found in igneous mafic rocks. The latter two
hydrates are often poor in OH-groups as a result of oxidation and/or halogen
substitution.
The complex chemical bulk rock composition of basalts will be redistrib-
uted from the few igneous mineral species into a variety of new minerals
during metamorphism. The high calcium content of basalts that is stored in
Cpx and Pl results in the formation of numerous separate calcium-bearing
metamorphic minerals. This is the basic difference to metapelitic rocks
(Chap. 7) where CaO is very low and does not form separate calcic phases
but rather is found as a minor component in ordinary Fe-Mg minerals such
as garnet. The most important minerals found in metamafic rocks are listed
in Table 9.1 as a refresher.
9.1.3
Chemographic Relationships and the ACF Projection
The composition space that must be considered in an analysis of phase rela-

tionships in metamafic rocks is rather complex and requires eight or more
system components. This follows from above and from the chemical analysis
of MOR basalt in Table 2.3. A graphical analysis of phase relationships in
metamafic rocks is often accomplished by means of the ACF projection. The
principles of the ACF projection were introduced in Chapter 2.5.3 and the
general chemography and a sample ACF diagram are shown in Fig. 2.10. The
ACF diagram is used, in general, for the display and analysis of phase rela-
tionships that involve calcic minerals. Because of its importance for the sub-
sequent treatment of metamafic rocks, some important features of the ACF
projection will be briefly reviewed.
The following discussion makes extensive use of Fig. 9.1 a that is a pseu-
do-3D display of the NACF tetrahedron, and Fig. 9.1 b, a conventional ACF
diagram. The ACF diagram represents, as usual, a mole fraction triangle and
displays the three composition variables Al 2 0 2 , CaO and FeO. Any mineral
that is composed of only these three components can be directly shown on
an ACF diagram (e. g., hercynite ). Any other mineral composition is pro-
jected onto the ACF triangle or is used as a projection point. In the latter
case, the mineral must be present in all assemblages in order to deduce
meaningful phase relationships. ACF diagrams are projections from Si0 2 and
this means that at any given P-T condition the stable polymorph of Si02
(e. g., quartz) must be present in all assemblages. This restriction imposed by
the conventional ACF diagram often causes problems when dealing with
metamorphic silica-poor igneous mafic rocks such as troctolites, nepheline
basalts, olivine basalts and olivine gabbros. However, other meaningful dia-
grams may be designed for such quartz-free metamafics following the sug-
gestions given in Chapter 2.5. ACF diagrams are also projections from H 2 0
and C0 2 in order to permit the display of hydrates and carbonates. This
means that a fluid phase of a specified composition must be present (or the
chemical potentials of H 2 0 and C0 2 must be defined otherwise) under the
ACF-projection
a
NaSi N (K)
zeolites Lws An •----+--::;c:=~':-:---t---=-• Ab, Jd,
Zo _ (Kfs)
Prh
c cal
M(F)
b +quartz
+ paragonite (albite, jadeite)
+ muscovite (K-feldspar)
+ C0;1- Hi) fluid
wol lastonite
calcite --- - - - - - - - - '11--l''----"'------'"..
to Phi and Gin (from Ms+Pg)

Fig. 9.1 . ACF projections
9.1 Mafic Rocks 285
condition of the diagram. This again may cause pain, particularly at low me-
tamorphic grades, because the universal presence of a fluid phase is not un-
avoidably obvious in anhydrous igneous protolith rocks (see discussion
above). Note that pumpellyite (Pmp) is on the ACF surface in Fig. 9.1 a and
its composition projects between zoisite and prehnite seen from the clino-
chlore composition.
ACF diagrams also represent projections parallel to the MgFe_ 1 exchange
vector. This is a rigid and lamentable feature of ACF diagrams. All complex
relationships, continuous and discontinuous reactions arising from variations
in the Fe/Mg ratio of ferro-magnesian minerals cannot be properly analyzed
and understood with the aid of such diagrams. The AF binaries on ACF and
AFM diagrams are identical. Consequently, many of the relationships dis-
cussed in Chapter 7 on metapelites are also valid here. The AF binary is ex-
panded to the AFM triangle in Chapter 7 and, in order to include Ca-miner-
als, to the ACF triangle in this chapter. An ACFM tetrahedron would permit
a rather thorough analysis of phase relationships in metamafic rocks. How-
ever, it is very inconvenient to work with three-dimensional composition
phase diagrams.
The MgFe_ 1 projection on ACF diagrams has the consequence that cross-
ing tie-line relationships among ferro-magnesian silicates (with contrasting
XFe) and Ca-minerals do not necessarily represent a discontinuous reaction
relationship but rather span a composition volume in the ACFM space. An
example (Fig. 9.1b): the four minerals garnet, hornblende, kyanite, and zo-
isite can be related to the reaction 2 tschermakite + 1 kyanite = 2 garnet+
2 zoisite + 1 quartz+ 1 H2 0. This important reaction in high-pressure amphi-
bolites is not discontinuous, as may erroneously be concluded from Fig. 9.1 b.
The four minerals actually occupy corners of a composition volume in ACFM
space and may occur as a stable assemblage over a certain P-T interval. The
reaction formulated above is thus in reality a continuous reaction. Another
effect of the MgFe_ 1 projection is that Fe and Mg end members of a given
mineral species project to identical positions (and of course any intermediate
XFe composition of that mineral too, e. g., pyrope and almandine). The conse-
quences of the MgFe_ 1 projection on ACF diagrams must always be kept in
mind.
About 3 wt% of Na2 0 is present in typical basalts and sodium-bearing
minerals are important in metamafics (plagioclase, amphibole, mica, pyrox-
ene; Table 2.3). If one wants to represent sodic phases on ACF diagrams, one
may expand the ACF triangle of Fig. 9.1 a (that actually is an ACM figure,
but this is equivalent, see above) to a tetrahedron with NaA10 2 as an addi-
tional corner. In this corner we will find the projected compositions of albite,
jadeite and analcite (Qtz, H2 0 projection). In a similar fashion, K-feldspar
will project to a KA10 2 corner of an analogous K tetrahedron. End-member
paragonite and muscovite project onto the AN and AK binary, respectively.
All plagioclase compositions project onto the ACN ternary. The composi-
tional variation in plagioclase is mainly related to the NaSiAL 1Ca_ 1 exchange
that connects albite and anorthite in Fig. 9.1a. The same exchange compo-
nent can also be found in, for example, amphibole, pyroxene and margarite.
End-member celadonite, glaucophane and phlogopite (biotite) project onto

the MN and MK binary, respectively. The micas and amphibole show strong
P-T- and assemblage-dependent compositional variations along the
MgSiAL 1AL 1-exchange direction. This tschermak exchange is parallel to the
AM-(AF) binary. Phengites project onto the AMN (AFN) ternary and the sol-
id solution series connects Ms and Cel. Sodic phases are often depicted on
ACF diagrams by projecting from the NaA10 2 apex of the ACFN tetrahedron
onto the ACF triangle. This means then that Ab, Jd, or Anl (whatever is
stable at the pressure and the temperature of the diagram) is present as an
extra mineral or that the potentials of these phase components are fixed. It
can be seen in Fig. 9.1 a that pure albite cannot be shown on ACF diagrams,
but that any Ca-bearing plagioclase will project to the An point. This means
that the consequences of the plagioclase composition on phase relationships
in metamafic rocks cannot be analyzed on ACF diagrams. Ms and Pa will
project to the A-apex of ACF diagrams, whereas phengites will occupy the
entire AF binary when projected from Kfs (Ab). However, Kfs is very rare in
metamafics and consequently Kfs projections are not much used. In analogy
to AFM diagrams, one may project from the white micas rather than from
the feldspar components, particularly for blueschist facies rocks. In this case,
the feldspars do not project onto the ACF plane with the exception of an-
orthite; phengite does not project onto the ACF plane either, whereas glauco-
phane projects onto the AF binary with a negative A-coordinate, like biotite
projected from muscovite (Fig. 9.1 a, b). In metamafic rocks, just one of the
minerals, K-mica, phengite or biotite, is generally present. This makes it pos-
sible to ignore the small amount of potassium that is stored in just one extra
minor phase. One may also choose to project from albite (jadeite) and mus-
covite. The consequences of the countless alternatives can be understood by
carefully studying Fig. 9.1 a. Also note that not all possible compositional var-
iations of minerals are shown in Fig. 9.la (e. g., Jd-Di solution). Whatever
projection you prefer for your own work, it is important to specify the pro-
jections on any of your phase diagrams carefully.
Another complication when dealing with metamafic rocks is the fact that
they tend to be much more oxidized than, for example, metapelites. Redox
reactions that involve phase components with ferric iron are common and
important. Notably, in garnet (andradite), amphibole (e.g., ribeckite, crossite)
and pyroxene (e.g., acmite), a considerable amount of the total iron is usual-
ly present in the trivalent state. However, a considerable substitution of
Fe 2 0 3 for Al2 0 3 can also be found in most of the low-grade Ca-Al hydrosili-
cates. The presence of epidote in rocks always signifies the presence of Fe3 +
in the total iron. Magnetite is a widespread oxide phase in metamafics. A
separate treatment of redox reactions will not be given here, however. One of
the effects of variable Fe3 + /Fe 2 + ratios in minerals is a further increase of
the variance of the considered assemblages. Consequently, the assemblages
occur over a wider P-T range compared to the situation where all iron is pre-
sent as Fe2 +. One may also construct ACF diagrams for a fixed oxygen activ-
ity. For instance, the coexistence of hematite and magnetite in a rock at P
and T will fix a02 (as shown in Chap. 3) and the conditions will be rather
9.1 Mafic Rocks 287
oxidizing. Thus, a significant amount of the total iron of the rock will be pre-
sent as Fe3 +. The effects on the minerals and mineral compositions can be
related, to a large extent, to the Fe3 +ALI exchange. This exchange makes epi-
dote from clinozoisite, magnetite from hercynite and ribeckite from glauco-
phane (for example). By considering ACF projections as projections parallel
to the Fe3 +ALI exchange vector, one may represent important minerals in
metamafics such as epidote, grandite garnet, magnetite and crossite. Note
that magnetite does not project to the F-apex but to the same projection
point as hercynite and Mg-Al-spinel.
Carbonates are widespread and abundant in low-grade metamafics. Fe-
bearing calcite and ankerite (dolomite) are predominant. Carbonates are also
often present in high-grade rocks. Carbonates can be displayed on ACF dia-
grams as explained above. However, any reactions that involve carbonates
are in general mixed volatile reactions and must be analyzed accordingly
(see Chaps. 3, 6 and 8). The presence of much C0 2 in an aqueous fluid phase
also has effects on pure dehydration reactions that do not involve carbonate
minerals (Chap. 3). Compared with pure H 2 0 fluids, in COrrich fluids hy-
drates (e. g., chlorite, amphibole) may also break down at much lower tem-
peratures in rocks that do not even contain carbonates.
Basalts also contain Ti0 2 in the % range. Titanium is mainly shelved in
one major Ti-phase, i.e., titanite, ilmenite, or rutile. If two of these minerals
are present in the rock, Ti-balanced reactions may be useful for P-T esti-
mates. An example: Ky+ 3 Il + 2 Qtz = 3 Rt + Grt. Ti is also found in consider-
able amounts in rock-forming silicates (biotite, amphiboles, garnet, etc.) with
unavoidable consequences for solution properties and equilibrium conditions
of reactions.
In conclusion, the composition of metamafic rocks is rather complex and
the minerals that typically occur in the assemblages show extensive chemical
variation along several exchange directions. This makes comprehensive
graphical analysis of phase relationships in metamorphic mafic rocks a diffi-
cult task. The complex chemical variation of solid-solution minerals can be
simplified for graphical analysis by projecting parallel to some of the ex-
change components. The choice of projection depends entirely on the prob-
lem one wants to solve and the kind of rocks one is working with. For many
metamafic rocks it has turned out that a very advantageous and powerful
projection can be made from Si0 2 , NaAl0 2 , H 2 0 and C0 2 , parallel to MgFe_I
and Fe3 +ALI onto the Al 2 0rCaO-FeO mole fraction triangle. This ACF pro-
jection will be used below.
The chemical complexity of mafic rocks makes it difficult to discuss phase
relationships by means of chemical subsystems and comprehensive petroge-
netic grids and phase diagrams in P-T space (as, for example, in previous
chapters). Note, however, the complexity of mafic rocks can be quantitatively
analyzed but this must be done for each given suite of rocks individually.
The presentation below is thus mainly a discussion of the ACF system and
MORB composition with some important reactions discussed separately
where necessary.
9.2
Overview of the Metamorphism of Mafic Rocks
The best overview of the metamorphism of mafic rocks can be gained from
the metamorphic facies scheme shown in Fig. 4.2. The characteristic assem-
blages in metabasalts are used for the definition of metamorphic facies and
serve as a reference frame for all other rock compositions. The assemblages
of metamafic rocks that are diagnostic for the different facies are given in
Chapter 4. From Fig. 4.2 it is evident that basalt undergoing prograde meta-
morphism along a Ky- or Sil-type path will first show diagnostic assemblages
of the subgreenschist facies; later it will become a greenschist then an am-
phibolite and finally end up as a mafic granulite. High-pressure low-tempera-
ture (HPLT) metamorphism turns basalt first into blueschist and then into
eclogite. Any geologic process that brings basalts to great depth (>50 km)
will result in the formation of eclogite. In contact metamorphism, basalts are
metamorphosed to mafic hornfelses. Partial melting of H2 0-saturated meta-
mafics starts at a temperature that is significantly higher than in metagrani-
toids and metapelites.
Prograde metamorphism of mafic rocks produces sequences of mineral
zones that can be compared with mineral zones defined by minerals in meta-
pelites. Figure 9.2 shows sequences of minerals formed in mafic and pelitic
rocks by prograde metamorphism from northern Michigan as an example.
Metapelites contain muscovite and quartz throughout and the sequence is of
the low-pressure type with andalusite and sillimanite as aluminosilicates. The
staurolite zone along the Sil geotherm is narrow, as expected from Fig. 7.3.
The corresponding prograde mineral zonation in metabasites shows a series
of features that are very characteristic.
• There are very few different mineral species present in metamafic rocks.
In the greenschist facies metamafic rocks contain: albite+ chlorite+ actino-
lite+ epidote~ greenschist. In the amphibolite facies the minerals are: pla-
gioclase (>Ann) +hornblende± biotite± epidote~ amphibolite.
• Most minerals occur over many of the mineral zones defined by meta-
pelites.
• The characteristic prograde changes in metabasites pertain to the compo-
sition of plagioclase and amphibole.
• Plagioclase systematically changes its composition from albite at low grade
to more calcic plagioclase (andesine in the example of Fig. 9.2). The tran-
sition from albite to oligoclase is abrupt and marks a sharp mappable
boundary in the field. The discontinuous nature is caused by a miscibility
gap in the plagioclase system. This discontinuity can be used to define the
greenschist - amphibolite facies boundary. Along a Ky-type path the oligo-
clase boundary coincides with the staurolite zone boundary that marks
the beginning of the amphibolite facies in metapelites. In low-pressure
metamorphism (e. g., Fig. 9.2) staurolite occurs for the first time inside the
amphibolite facies {Fig. 7.3).
• Amphibole systematically changes its composition from actinolite at low
grade to alkali- and aluminum-bearing hornblende at high grade (the
9.2 Overview of the Metamorphism of Mafic Rocks 289
Metamorphic facies Greenschist

I Amphibolite
Mineral zoning Chlorite Biotite Gamet Staurolite Sillimanite
Metamafites
Albite
Albite-oligoclase
Oligoclase-andesine
•••••••
....
Andesine
Epidote I
Actinolite •••••• green and
..
blue-green green brown
Hornblende
Chlorite
Calcite ••••••••• I I I
green-brown brown
Biotite II II
Muscovite •••••••• ••••••••• ••••••••• •••••••••

Quartz I I I I I I I I
••••••••• ......... .••.•..•. •..•..•..
Metapelites
Chlorite I I I I I I I
Muscovite
Biotite
Gamet
Staurolite
.....
I I I I I I
andaluslte sillimanite
Alumosilicate
Chloritoid
clastic oligoclase
Plagioclase
Quartz
I
I
Fig. 9.2. Progressive mineral changes in northern Michigan. (James 1955)
hornblende color information in Fig. 9.2 relates to the changing mineral

composition).
• Quartz is only occasionally present and, hence, ACF diagrams must be
used with care.
• Biotite is present as an extra K-bearing mineral from the greenschist to
the upper amphibolite facies. The mineral systematically changes its com-
position during prograde metamorphism.
• Calcite is present in low-grade rocks but is used up by mixed volatile reac-
tions in prograde metamorphism.
A comprehensive representation of the effects of metamorphism on mafic

rocks is shown in Fig. 9.3. The figure depicts the characteristic assemblages
in metabasalts in the form of ACF diagrams representative for the respective
position in P- T space. The ACF diagrams are arranged along three typical
paths of metamorphism and the 9 oC/km geotherm limits the geologically ac-
cessible P- T space. The aluminosilicate phase diagram is given for reference.
N
'D
12~---------------------,--~------~ 0
~
eclogite
A
I
10 ' ' c ' - -.. - amphibolite-

Cpx Opx granulite
.... I
basalt com osition
+ excess quartz + water

y . . . ,.,
I
+Phe
+Pg
Lws
transition
garnet
ca KM / granulite
Ep
-
.c 8
~
J Cal Cpx Hbl
-
!
:::1
Cl)
greenschist- pyroxene
Cl) amphibolite granulite
rt
transition eif§./
! ~al Dol ':lbl
s/\
ll. 6 ~ r: c-1 AKY ..,I If
"' ....
'D
:::
"'S'
4 1 I t"'hl / '\. Pl tf ' \~ 'A""l.
, ~- ..... ~"''"' ~--·~~-~- I lg...
'0
e:
"'5!
0
~{;nl ~~-.
......
Ca~l ~ ~ ~
. Dol Cal Dol Cal Dol Hbl Wo
• Cpx Opx :::
2 :tl
"',.,
100 200 300 400 500 600 700 800 900 ,.,S'
Temperature (°C) ~
"'
Fig. 9.3. Metamorphism of mafic rocks (metabasalts) represented by ACF diagrams
9.3 Subgreenschist Facies Metamorphism 291
The inset in the upper left corner shows the composition of typical mid-
ocean ridge basalt (Pl+Cpx±Opx±Qtz). Many metabasalts are expected to
lie compositionally within the horizontally ruled field. On the ACF diagrams
of Fig. 9.3, the assemblages that represent the approximate composition of
MORB have been shaded. Metabasalts that have been strongly metasomatized
during ocean-floor metamorphism may project to other, often more Ca-rich,
compositions in the ACF diagram. In the following presentation, the details
of the phase relations in each of the represented ACF chemographies will be
discussed separately.
9.3
Subgreenschist Facies Metamorphism
9.3.1
General Aspects and a Field Example
ACF chemographies #1 and #2 (Fig. 9.3) represent mineral assemblages typi-

cal of the subgreenschist facies. At very low grade, the characteristic assem-
blage includes albite+ chlorite+ carbonate+ a variety of zeolites. On the ACF
chemography #1 zeolite is typified by laumontite but any of the four zeolites
listed in Table 9.1 would project to the same point. In addition, clay (e.g.,
smectite, vermiculite) and "white mica'' (illite, sericite) may be present as
well as kaolinite. At slightly higher grade (ACF chemography #2), the zeolites
are replaced by the minerals prehnite, pumpellyite and epidote in a number
of different combinations. Still albite, chlorite and carbonates are major
minerals in prehnite-pumpellyite rocks. Pyrophyllite may be present instead
of kaolinite. Zeolite-bearing metamafics are representative for temperatures
below about 150-200°C; metabasalts with prehnite and pumpellyite are most
characteristic for the temperature range 150-300 °C.
The zoneography of minerals in metamafic rocks from the Tanzawa moun-
tains in central Japan is shown in Fig. 9.4. The metamorphic terrain has been
subdivided into five zones ranging from lower subgreenschist facies to am-
phibolite facies. The terrain is of the low-pressure sillimanite type. The first
two zones are characterized by the presence of zeolites in the assemblages:
low-T zeolites in zone I and high-T zeolites in zone II. The zone I zeolites
are typically stilbite and heulandite; zone II zeolites are laumontite and wair-
akite. This is understandable by reference to Table 9.1 where one can see that
zone I zeolites clearly contain more crystal water (zeolite water) than zeolites
of zone II. The transition of zone I to zone II is therefore related to ordinary
dehydration reactions. For the same reason, it is also clear that wairakite is
typically the last zeolite in prograde metamorphism. Analcite occurs in both
zones sporadically. Mixed-layer clays are typically present in the first two
zones. Plagioclase is present in all zones; its composition is that of albite in
the first four zones. The protolith is SiOrrich, and quartz is present in ex-
cess throughout (ACF diagrams). No zeolites survive in zone III, which is
characterized by three new metamorphic minerals: pumpellyite, prehnite and
epidote. Also chlorite has transformed into a well-defined ferro-magnesian
Mineral zoning zone I zone II zone Ill zone IV zoneV
Clinoptililite ~~·· X X
Stilbite
1-
X X
Heulandite
Mordenite I I I I I I I I I
Chabazite X X
Laumontite X X X
--
Thompsonite I I I I I I I I I I
Wairakite
Yugawaralite
Analcime I I I I I I I I I I I I I I I I I I I I I I I
Montm. -verm. ~Ill

Verm. -chlorite II
Chlorite I I I I I I
Sericite
Biotite
Pumpellyite
I I I I I
I
I I I I I I I I I
I I I I I I ..
I I I I I I I I
Prehnite I II
Epidote II II
Piemontite I I I I I I I I I I I I I
Actinolite
Hornblende
Cummingtonite I I I I I I I I I
Diopside I l l -
Ca-garnet ••••
Plagioclase
An10 An20 An30
Opalline silica I I I I
Quartz
Magnetite •••• I I I I I I I I I I
......... ......... .........

Hematite I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
Pyrite I I I I I I I I I ......... I I I I I I I I I
Calcite I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
Fig. 9.4. Occurrence of some metamorphic minerals in Tanzawa Mountains, central Japan
(Seki et al. 1969). Mineral zonation in the Tanzawa terrain, Japan. Zone I Stilbite (clinoptilo-
lite)-vermiculite; zone II laumontite-"mixed layer" -chlorite; zone III pumpellyite-prehnite-
chlorite; zone IV actinolite-greenschist; zone V amphibolite. X only in veins
sheet silicate. Sericite is present sporadically. Several zeolite- and chlorite-

consuming reactions are responsible for the generation of the new meta-
morphic minerals Prh, Pmp, and Ep. In zone IV, typical greenschist facies as-
semblages replace the zone III mineralogy. In particular, pumpellyite and
prehnite disappear and the characteristic greenschist facies mineral associa-
tion chlorite+ epidote+ actinolite+ albite± quartz becomes dominant. In zone
V, the minerals are diagnostic for the amphibolite facies. Calcic plagioclase
and hornblende are the dominant minerals but biotite, garnet and Cpx may
appear as well. In low-pressure terrains such as the Tanzawa Mountains,

cummingtonite or other Fe-Mg-amphiboles typically occur together with
calcic amphiboles (hornblende).
9.3.2
Metamorphism in the CASH and NCMASH Systems
Quantitative phase relationships in the simple CASH system in low-grade

metamorphism are shown in Fig. 9.5. Some of the mineral stability fields are
highlighted on the figure. The reaction stoichiometries can be found in
Table 9.2. The predicted sequence of Ca-zeolites in progressively metamorp-
hosed metamafic rocks is: stilbite, heulandite, laumontite and wairakite. This
sequence is clearly consistent with the sequence of observed mineral zones in
the Tanzawa Mountains shown in Fig. 9.4. It also follows from Fig. 9.5 that
lawsonite favorably forms in terrains that are characterized by high-pressure
low-temperature geotherms. In fact, lawsonite is an important mineral in
blueschist terrains.
Figure 9.6 depicts phase relations in the more complex NCMASH system
involving the following 14 minerals: actinolite, plagioclase, chlorite, epidote,
garnet, heulandite, laumontite, lawsonite, paragonite, prehnite, pumpellyite,
quartz, stilbite, and wairakite. The relationships in Fig. 9.6 are based on ther-
modynamic data for Mg end-member phase components in the NCMASH
system. The deviations of average natural mineral compositions from the
model end-member compositions have been accounted for by incorporating
appropriate activity terms. However, amphiboles, chlorite, epidote, and pum-
pellyite from low-grade metabasites have large variations in their chemical
composition, depending on mineral assemblage, rock composition, and meta-
morphic grade. It is therefore important to realize that the petrogenetic grid
in Fig. 9.6 is valid only for an average metabasite (meta-MORB). The effect
of variable mineral composition on the P-T stability fields of various index
minerals may be considerable and will be evaluated separately.
Solid solution in zeolites may also be important, especially the "plagio-
clase" substitution NaSiCa_ 1AL 1 . However, rather poorly constrained thermo-
dynamic data for zeolites and limited chemical data do not justify including
activity terms for zeolites.
Figure 9.6 is a projection from albite+ chlorite (of fixed average composi-
tion)+ quartz+ water, because these minerals are ubiquitous in low-grade me-
tabasites. At very low temperatures (T < 200 °C), however, this represents a
simplification because smectite may be present, either as a discrete phase or
as interstratified layers in chlorite. Thus it is fortunate that chlorite is not in-
volved in reactions limiting the stability fields of zeolites.
Below follows a brief discussion of the typical P-T conditions under which
some diagnostic index minerals may form and subsequently decompose in
prograde metamorphism of metamafic rocks. The discussion is based on
Figs. 9.5 and 9.6. Reaction stoichiometries are given in Table 9.2.
Analcite is commonly found in quartz-bearing diagenetic and low-grade
metamorphic rocks. The analcite-quartz assemblage, along with stilbite +
-...
-
~ 4
~
100 200 300 400 500

Temperature (°C)
Fig. 9.5. Low-grade metamorphism in the CASH system. Prehnite field shaded
projected from
Qtz, Ab , Chi
(a=0 .06) and H 2 0
5
-~
greenschist
facies
-
C'CI
.c 4
~
~
:::::1
"'"'~ 3
pumpellyite
c..
pumpellyite -
2 + prehnite
150 200 250 300 350 400
Temperature (°C)
Fig. 9.6. Low-grade metamorphism in the NCMASH system. Prehnite, pumpellyite and
prehnite +pumpellyite fields shaded
Table 9.2. Reaction stoichiometries of reactions shown in the figures in this chapter and
reactions discussed in the text
Subgreenschist facies
CASH system (Fig. 9.5)
(1) Stb=Lws+5 Qtz+5 H 20
(2) Lmt=Lws+2 Qtz+2 H 20
(3) Stb=Lmt+3 Qtz+3 H 20
(4) Stb=Hul+H20
(5) Hul=Lmt+3 Qtz+2 H20
(6) Hul=Wa+3 Qtz+4 H20
(7) Lmt=Wa+2 H20
(8} Wa=Lws+2 Qtz
(9) Lws =An+ 2 H20
(10) Wa=An+2 Qtz+2 H20
(ll) Stb+Grs=2 Prh+4 Qtz+5 H20
(12) 2 Prh = Lws+Grs+Qtz
(13) 5 Prh=2 Zo + 2 Grs+3 Qtz+4 H2 0
(14) 5 Lws+Grs=4 Zo+Qtz+8 H20
(15) 2 Lws+Prh=2 Zo+Qtz+4 H2 0
(16) Prh+2 Lmt = 2 Zo+5 Qtz+8 H2 0
(17) Prh+2 Wa=2 Zo+5 Qtz+4 H20
NCMASH system (Fig. 9.6)

(18) 4 Hul+Tr = 3 Prh+Chl+24 Qtz + 18 H20
(19) 4 Lmt+Tr = 3 Prh+Chl+12 Qtz + 10 H2 0
(20} 20 Pmp+3 Tr+6 Qtz=43 Prh+7 Chl+2 H2 0
(21} 86 Lmt+17 Tr=30 Pmp+ll Chl+267 Qtz+212 H20
(22} 86 Lws+17 Tr = 30 Pmp+ll Chl+95 Qtz+40 H2 0
(23} 5 Pmp+3 Qtz=7 Prh+3 Zo+Chl+5 H20
(24) 6 Lmt+17 Prh+2 Chl = 10 Pmp+21 Qtz + 14 H20
(25) 14 Lmt + 5 Pmp = 17 Zo+Chl+32 Qtz+61 H2 0
(26} 86 Stb + 17 Tr = 30 Pmp + ll Cbl + 525 Qtz + 470 H20
Subgreenschist to greenschist facies transition (Fig. 9.6)

Essential reactions
(27} 5 Prh+Chl+2 Qtz = 4 Zo+Tr+6 H2 0
(28) 25 Pmp+2 Chl+29 Qtz =7 Tr+43 Zo+67 H20
Additional reactions
(29} 4 Wa+Ab = Pg+2 Zo+10 Qtz+6 H 2 0
(30} 4 Lws+Ab = Pg+2 Zo+2 Qtz+6 H 2 0
(31) 14 Lws+5 Pmp = 17 Zo+Chl+4 Qtz+33 H 20
(32) 4 Lmt+Ab = Pg+2 Zo+lO Qtz+14 H20
Reactions involving carbonates (examples)

(33} 3 Chl+10 Cal+21 Qtz =2 Zo+3 Tr+10 C0 2 +8 H20
(34) 15 Pmp+9 Qtz+4 C0 2 = 4 Cal+25 Zo+3 Tr+37 H20
Greenschist facies
CMASH reactions (Fig. 9.7)
(35) 2 Zo+5 Prl = 4 Mrg+l8 Qtz+2 H20
(36} Mrg+2 Qtz+2 Zo=5 An+2 H20
(37) Mrg+Qtz=An+And+H 2 0
(38) 4 Mrg + 3 Qtz = 2 Zo+5 Ky+3 H2 0
(39} 2 Chl+2 Zo+2 Qtz = 2 Tr+5 Ky+7 H2 0
(40) 6 Zo+7 Qtz+Chl = 10 An+Tr+6 H20
(41) 2 An+Chl+4 Qtz=Tr+3 Sil+3 H2 0
Example of reaction producing tschermak component

(42) 7 Chl+14 Qtz+l2 Zo=12 Tr+25 TS (A1 1Al 1Si... 1Mg_ 1 )+22 H20
Table 9.2 (continued)
Mica-involving reactions
(43) 16 Tr+ 25 Ms = 25 Phi+ 16 Zo+ 1 Chl + 77 Qtz+4 H2 0
(44) 4 Tr+6 Chi+25 Cel=25 Phl+4 Zo+63 Qtz+26 HzO
(45) 1 Chi+ 4 Cel = 3 Phi+ 1 Ms+ 7 Qtz+4 HzO
(46) 2 Chl+2 Zo+5 Ab+4 Qtz=2 Tr+S Pg+2 H2 0
Carbonate-involving reactions (examples)

(47) 1 Ms+3 Qtz+8 Dol+4 H20= I Phl+l Chi+8 Cal+8 C0 2
(48) 9 Tr+2 Cal+15 Ms=lS Phi+lO Zo+42 Qtz+2 C0 2 +4 H20
(49) 19 Cal+3 Chl+ll C0 2 = IS Dol+2 Zo+3 Qtz+ll H 20
Greenschist-amphibolite facies transition

(SO) 4 Chi+l8 Zo+21 Qtz=5 Ts-amphibole+26 An+20 H20
(51) 7 Chi+ 13 Tr+ 12 Zo + 14 Qtz= 25 Ts-amphibole + 22 H20
(52) Ab+ Tr= Ed +4 Qtz
(53) 12 Zo+15 Chl+18 Qtz=8 Grs+25 Prp+66 HzO
Amphibolite facies
(54) 4 Tr+3 An=3 Prp+ll Di+7 Qtz+4 H20
(55) 1 Ts-amphibole+6 Zo+3 Qtz= 10 An+4 Di+4 H2 0
(56) 3 Ts-amphibole+7 Ky=6 An+4 Prp+4 Qtz+3 H2 0
(57) 7 Ts-amphibolie + 7 Sil + 4 Qtz = 14 An+ 4 Ath + 3 H20
(58) 7 Tr=3 Ath+l4 Di+4 Qtz+4 H20
Amphibolite-granulite facies transition

(59) Tr=2 Cpx+3 Opx+Qtz+H 20
(60) Ts=Cpx+3 Opx+An+HzO
(61) Tr+7 Grs=27 Cpx+Prp+ 6 An+ H20
(62) Tr+Grs=4 Cpx+Opx+ An+H 20
Granulite facies
(63) 4 En+! An= I Di+1 Qtz+l Prp
Blueschist facies (Fig. 9.9)

(64) Tr+ 10 Ab+2 Chl=2 Lws+S Gln
(65) 6 Tr+50 Ab+9 Chl=25 Gln+6 Zo+7 Qtz+ 14 H20
(66) 13 Ab+3 Chi + I Qtz=5 Gln+3 Pg+4 H20
(67) 12 Lws+1 Gln=2 Pg+1 Prp+6 Zo+5 Qtz+20 HzO
(68) 4 Pg+9 Chl+16 Qtz=2 Gln+l3 Prp+38 HzO
(69) 10 Pg+3 Chi+ 14 Qtz=5 Gin+ 13 Ky+ 17 H2 0
Blueschist-eclogite facies transition (Figs. 9.9 and 9.10)

(70) Jd+Qtz=Ab
(71) I Gln+1 Pg=1 Prp+3 Jd+2 Qtz+2 H20
Granulite- and amphibolite-eclogite facies transition
(72) CaTS+Qtz=An
(73) 2 Zo+Ky+Qtz=4 An+H 20
(74) I Grs+2 Ky+ I Qtz=3 An
(75) 4 Tr+3 An=3 Prp+ll Di+7 Qtz+4 H2 0
(76) 4 En+ I An=1 Prp+l Di+l Qtz
(77) 3 Di+3 An= I Prp+2 Grs+3 Qtz
(78) I An+2 Di=2 En+ 1 Grs+ 1 Qtz
Eclogite facies
(79) Pg=Jd+Ky+HzO
(80) 6 Zo+4 Prp+7 Qtz=13 Ky+12 Di+3 H20
(81) Tlc+Ky=2 En+Mg-TS+2 Qtz+H 20
(82) Grs+Prp+2 Qtz=3 Di+2 Ky
(83) Grs + 6 Opx= 3 Di + Prp
quartz, can be used as an indicator of the initial stage of zeolite facies meta-
morphism. In the NASH model system, analcite decomposes in the presence
of quartz to albite according to the reaction: Anl + Qtz = Ab + H2 0 (the reac-
tion is not listed in Table 9.2). The equilibrium conditions of the analcite
breakdown reaction very nearly coincide with the three reactions that limit
the stilbite field towards higher temperature in Fig. 9.5 (reactions 1, 2, and
3). In order to avoid overcrowding of the figure, the analcite reaction has not
been shown in Fig. 9.5 explicitly. Natural analcite, however, often shows ex-
tensive solid solution with wairakite, thus expanding the stability field of
analcite to higher temperature. This has the consequence that the P- T fields
of analcite and heulandite overlap.
Stilbite, a widespread zeolite mineral, dehydrates to heulandite, laumontite
or lawsonite with increasing temperature and successively higher pressures
in the CASH model system (Fig. 9.5). In nature, however, stilbite is sometimes
associated with heulandite and/or laumontite, as, e. g., in geothermal areas in
Iceland. Although little is known about the composition of stilbite in low-grade
metamorphic rocks, this feature may result from solid-solution effects.
Heulandite is one of the most common zeolites besides analcite and lau-
montite. In Fig. 9.5, its stability field is located at temperatures around 200 oc
and at pressures below 2 kbar, and limited by the stability fields of three
other zeolites, i.e., stilbite towards low-T (reaction 4), wairakite towards
high-T (reaction 6; not shown in Fig. 9.5, reaction is stable below 1 kbar
pressure), and laumontite towards high P- T (reaction 5). Heulandite and lau-
montite seem to be equally abundant in ocean-floor thermal, hydrothermal,
and burial metamorphism. Laumontite, but not heulandite, has been de-
scribed from subduction zone metamorphism, e. g., from the Franciscan
Complex of California and the Sanbagawa Belt of Japan (for a summary, see
Liou et al. 1987). This indicates that heulandite is a low-pressure zeolite con-
sistent with the topology shown in Fig. 9.5.
Natural heulandite commonly contains Na or K or both substituted for Ca,
and its chemical composition may extend to those of clinoptilolite and alkali-
clinoptilolite. This effect, together with other variables (e. g., pore-fluid
chemistry), may explain the considerable overlap between the stability re-
gions of heulandite and other zeolites.
Laumontite is widespread in ocean-floor, burial, and subduction zone
metamorphism. In Fig. 9.5, its field of stable occurrence is limited by five
univariant reactions. With increasing temperature, laumontite is formed from
heulandite at lower pressure and from stilbite at higher pressure. The upper
pressure limit of laumontite with respect to lawsonite is about 3 kbar [reac-
tion (2)], and the upper temperature limit with respect to wairakite [reaction
{7)] and epidote (zoisite; reaction 16) is about 230-260 oc. Compositionally,
laumontite is near the ideal composition, CaA12 Si4 0 12 • 4 H 2 0, but solid solu-
tion in heulandite and wairakite will allow the coexistence of Lmt + Hul and
Lmt+Wa.
The presence of laumontite is often taken to indicate low-grade meta-
morphic conditions, and thus its lower thermal stability limit is of interest.
In the CASH model system, this limit is about 180 °C. From field evidence,
however, it has been inferred that laumontite may have formed at tempera-
tures as low as 50-100 °C (e. g., Boles and Coombs 1975).
Wairakite is restricted to areas of relatively high geothermal gradients, in-
cluding geothermal systems and areas with intrusive igneous bodies. In the
CASH model system, wairakite is produced from heulandite or laumontite at
temperatures between 220 and 260 a c. Wairakite has the highest thermal sta-
bility of any zeolite and, in the presence of excess albite+ quartz, its upper
thermal stability is between 260 and 380 a c. As mentioned earlier, wairakite
shows considerable solid solution with analcite and this may explain the nat-
ural occurrence of Wa + Anl and Wa + Lmt ( + Ab + Qtz).
Lawsonite is one of the most definitive of the blueschist suite of minerals.
Pumpellyite is a very common mineral in metamafics and is found in var-
ious associations in low-grade metamafic rocks. The shaded pumpellyite field
in Fig. 9.6 shows the P-T range of the assemblage Pmp + Qtz + Chl + H 2 0 for
average pumpellyite and chlorite composition. The reactions forming and
consuming pumpellyite are listed in Table 9.2. In rocks that depart from the
average meta-MORB, pumpellyite may occur over a much larger or much
smaller P-T interval. However, for all rock compositions, the center of the
pumpellyite fields clusters around 220 °C±20o and 4±2 kbar. Most pumpel-
lyite-bearing rocks probably formed close to 200-250 oc and at pressures of
2-3 kbar.
Prehnite occurs widespread in low-grade metamafic rocks as a part of the
matrix assemblage but also very often in late veins and open fracture spaces.
The shaded prehnite fields in Figs. 9.5 and 9.6 clearly suggest that the miner-
al is diagnostic for low-pressure conditions. The upper pressure limit of pre-
hnite is about 5 kbar but a more typical value is probably around 3 kbar.
The prehnite +pumpellyite assemblage is shown as a dark shaded field in
Fig. 9.6. The assemblage may be found in meta-MORE's at 200-250 oc at
pressures of 2-3 kbar in association with quartz and chlorite (of the average
composition). Depending on rock composition (and the resulting average
mineral compositions at low grade), the two minerals (Pmp + Prh) may not
have a common field of occurrence at all. However, most metamafics may be
able to generate the Prh + Pmp assemblage. The assemblage has, however, un-
der any circumstances an extremely narrow P-T field where it may be stable.
The cumulative conditions for the co-occurrence of Pmp + Prh + Chl + Chl is
about 200-280 oc and 1-4 kbar. In many metamorphic terrains Pmp + Prh
can be found in rocks of the subgreenschist facies. This is not surprising be-
cause of the natural variation of bulk composition of mafic rocks and the
geotherms followed by prograde metamorphism will almost certainly pass
through the mentioned cumulative P-T field of Pmp + Prh. The higher-grade
portion of the subgreenschist facies is therefore characterized by assemblages
that involve: Pmp, Prh, Wa, Ep, Chl, Ab, Pg and Qtz.
9.4 Greenschist Facies Metamorphism 299
9.3.3
Transition to the Greenschist Facies
The transition to the greenschist facies is marked by the first appearance of

the diagnostic assemblage actinolite+ epidote in the presence of chlorite (of
average composition), albite and quartz. The assemblage is produced from
prehnite decomposition at low pressures (< 3 kbar) and pumpellyite decom-
position at higher pressure (>3 kbar). Both reactions [reactions (27) and
(28), Table 9.2] produce the typical greenschist facies mineralogy at a tem-
perature of about 280±30 oc at pressures below 6 kbar. Along the characteris-
tic Ky- and Sil-type path, the first occurrence of actinolite+ epidote+ chlo-
rite+ albite+ quartz defines the beginning of the greenschist facies. The as-
semblage is shown on chemography #3 in Fig. 9.3. Chemography #3 also sug-
gests that the greenschist facies assemblage may also form from dolomite- or
calcite-involving reactions (an example is reaction 33). As carbonates are pre-
sent in many low-grade mafic rocks, these mixed volatile reactions are im-
portant in removing carbonate minerals from metamafic rocks at an early
stage of prograde metamorphism and connecting to the greenschist facies.
Typically, these reactions consume chlorite and carbonate and produce epi-
dote+ actinolite. The reactions are of the maximum-type mixed volatile reac-
tions (see Chap. 3). Note, however, that pumpellyite- and prehnite-consuming
reactions produce rather than consume carbonates. Reaction (34) is an exam-
ple reaction that limits pumpellyite in COrbearing fluids and produces the
greenschist facies assemblage. This means that, in the presence of COrrich
fluids, greenschist facies assemblages may appear at lower temperature com-
pared with pure H2 0 fluids.
At high pressures (e.g., along a low-T subduction geotherm), the sub-
greenschist facies assemblages are replaced by assemblages that are diagnos-
tic for the blueschist facies. These rocks contain glaucophane, lawsonite,
paragonite, epidote, and other minerals in various assemblages. The blue-
schist facies assemblages and blueschist metamorphism will be discussed be-
low (see Sect. 9.7).
9.4
Greenschist Facies Metamorphism
9.4.1
Introduction
Greenschist facies mafic rocks are, as the name suggests, green schists. The
green color of greenschists results from the modal dominance of green
minerals in the rocks notably chlorite, actinolite and epidote. The most char-
acteristic assemblage found in greenschists is: chlorite+ actinolite+ epidote+
albite± quartz. The term greenschist is exclusively reserved for a schistose
chlorite-rich rock derived from mafic igneous rocks that were metamorphos-
ed under greenschist facies conditions (hence contains the diagnostic assem-
blage: Chl +Act+ Ep + Ab ± Qtz). Note, however, pumpellyite+ epidote+ chlo-
rite schists of the subgreenschist facies are green schists as well, although
not in greenschist facies. Serpentinites may be green schists of the greens-
chist facies but they are not greenschists.
Chemographies #3 and #4 (Fig. 9.3) represent the typical range of assem-
blages in the greenschist facies. In addition to the most important minerals
named above, some further minerals may be found in greenschists in minor
amounts. Their occurrence mostly depends on the bulk composition of the
protolith and the details of the hydration history prior to greenschist facies
metamorphism. Greenschists commonly lost all relic structures from pre-
vious metamorphic and magmatic stages. It is typical and ordinary to find
mineralogically and structurally perfectly equilibrated greenschists.
9.4.2
Mineralogical Changes Within the Greenschist Facies
9.4.2.1
Reactions in the CMASH System
The difference between chemography #3 and #4 (Fig. 9.3) is the presence of

pyrophyllite and phengite in the lower greenschist facies and of kyanite and
biotite in the upper greenschist facies. The phase relationships in the ASH
system that covers the changes in the A-apex of ACF diagrams has been dis-
cussed in Chapter 7 and they are displayed in Fig. 7.2. The reactions in the
ASH system also represent the backbone of Fig. 9.7 that depicts some phase
relationships in the CMASH system (using the same average mineral compo-
sitions as in Fig. 9.6, unless indicated otherwise). The reaction stoichiometry
of reactions shown in Fig. 9.7 is listed in Table 9.2. Note, however, that not
all equilibria shown are listed in Table 9.2; in particular, the details of the
phase relations in the sillimanite field of Fig. 9.7 have been omitted because
we are concerned with the greenschist facies here.
It is evident from Fig. 9.7 that the first reaction relevant for mafic rocks in
greenschist facies is reaction (35). It replaces epidote+ pyrophyllite by margar-
ite in quartz-bearing rocks. The shaded margarite field in Fig. 9.7 is created by
the stable reaction (35) and it is terminated by other reactions that will be dis-
cussed below. Margarite is not shown in Fig. 9.3 (ACF figure). Figure 9.7 sug-
gests that large central portions of the greenschist facies are within the stability
field of margarite. However, margarite in greenschists is not very common. The
reason for this apparent mismatch is that most metamafic rocks fall composi-
tionally into the shaded Chl-Ep-Act triangle in Fig. 9.3. However, if bulk com-
position is more aluminous than the epidote-chlorite tie line, then the stable
assemblages may be read from Fig. 9.7. The sequence is (for Qtz-saturated
rocks following a Ky-type path): Prl+Ep~Mrg+Ep~Ky+Ep.
Figure 9.7 also suggests that the assemblage epidote+ kyanite is diagnostic
for relatively high pressures (>about 5 kbar). The Ep + Ky assemblage is pro-
duced by reaction (38). Other reactions that terminate the margarite +quartz
assemblage in greenschists are reactions (36) and (37). The assemblage acti-
nolite (amphibole)+ kyanite commences with the higher greenschist facies. It
excess quartz Zo +
and H20 Prl
-
6
"-
co
-
.c
~
~ 4
::s
(/J
(/J
~
ll.2
200 300 400 500 600

Temperature (°C)
Fig. 9.7. Phase relationships in the greenschist facies. Margarite field shaded
is diagnostic for pressures typically above 5-6 kbar. The composition of am-
phibole that occurs stable with kyanite is, however, far removed from com-
mon actinolite and reaction (39), that produces the Act+ Ky assemblage, runs
close to the amphibolite facies boundary.
Inside the Chl-Ep-Act triangle of Fig. 9.3, there are also reactions that
modify the typical assemblage of the greenschist facies. The processes are
best explained by using reaction (42) as an example. The reaction describes
the production of tschermak component that is taken up by amphibole and
chlorite. Both minerals systematically change their Al content across the
greenschist facies. The effect can be shown on ACF diagrams by a displace-
ment of the Chl + Ep +Act triangle along the TS vector. Note, however, con-
current rotations of FM relationships cannot be displayed on ACF diagrams.
In general, chlorite and actinolite become more aluminous across the green-
schist facies. AF relationships are complex and can be evaluated along the
lines described in Chapter 7 (metapelites). Redox reactions may change the
composition of epidote and other Fe-bearing minerals.
9.4.2.2
Reodions Including Micas
Muscovite (in general K-white mica, sericite, phengite) is the most common
K-bearing mineral in low-grade metamafic rocks. This is also true for micas
in the lower greenschist facies. However, in the middle of the greenschist fa-
cies ( -400 °C), biotite appears for the first time in metamafics replacing K-
white mica. This is shown in Fig. 9.3 (chemographies #3 and #4). Biotite for-
mation can be modeled by reaction (43). It involves all greenschist minerals
and it transfers the Ms component to the Phl component. The equilibrium of
reaction (43) is close to 600 oc in the pure KCMASH system and it is rather
insensitive to pressure. For real mineral compositions of greenschist facies
rocks, biotite formation from reaction (43) takes place around 400 oc
(aPhi= 0.05; aMs = 0.2; all others see Fig. 9.6). The first biotite that appears in
prograde metamorphism of mafic rocks is often green under the microscope.
This is an indication of high Fe3 + in low-grade biotite and demonstrates the
importance of redox reactions in low-grade metamafics (so be aware of that
when calculating phase relationships involving biotite).
Biotite production is a continuous process. The biotite-producing reac-
tion (44) is similar to reaction (43). However, it consumes the celadonite
component in K-white mica in order to produce biotite. The reaction leaves
behind a white mica that is closer to muscovite end-member composition.
This is in accordance with field evidence that K-white mica becomes progres-
sively depleted in Cel component and enriched in Ms component during pro-
grade metamorphism along a Ky-type path. The effect can also be under-
stood by investigating reaction (45), where celadonite component is con-
sumed in white mica and biotite+ muscovite components are formed in pro-
grade metamorphism.
Again, biotite forms in metamafics at about 400-450°C. K-white mica is
typically celadonitic (phengite, sericite) in low-grade rocks; it becomes more
muscovitic (muscovite) at the upper end of the greenschist facies. However,
the small amount of potassium in mafic rocks is usually bound in K-white
mica below 400 oc and in biotite at temperatures above 400 oc, and there is
not much overlap of the two micas.
Paragonite occurs in greenschists at high pressures. The equilibrium con-
ditions of reaction (46) are extremely sensitive to small compositional varia-
tions in the protolith and the resulting metamorphic minerals. The typical
actinolite+ paragonite assemblage forms, however, at pressures above about
6 kbar, and this pressure is insensitive to temperature.
Garnet may form in mafic schist from reactions very similar to the reac-
tions that form garnet in pelitic schist. The first garnet is manganiferous and
contains little grossular component. Garnet appears in the uppermost part of
the greenschist facies for the first time, but garnet is not really typical in
mafic rocks until the amphibolite facies.
Stilpnomelane is a characteristic mineral in many low-grade mafic schists
and, if present, is diagnostic for lower greenschist facies conditions (or blue-
schist facies). The brown color of this pleochroic mineral often resembles
biotite under the microscope. Stilpnomelane is replaced by green biotite at

around 400 oc. Mafic rocks containing both stilpnomelane and biotite occur
over a narrow temperature interval near 400 °C.
Carbonate minerals may still be present in greenschists. Both dolomite
(ankerite) and calcite can be present and the carbonates participate in many
mixed volatile reactions that involve the characteristic greenschist facies sili-
cates, Chl, Ep, and Act. The reactions may also involve micas such as reac-
tion (47) that replaces K-white mica by biotite in Dol- and Cal-bearing rocks.
Another example is reaction (48), where the assemblage actinolite+ musco-
vite+ calcite is replaced by epidote+ biotite. Reaction (49) involves two carbo-
nates in rocks containing epidote+ actinolite. Some of the carbonate-involv-
ing reactions are also relevant for marbles (CMS-HC system) or for marly
rocks. These have been discussed in Chapters 6 and 8, respectively, and we
will not present an extensive discussion of phase equilibria in carbonate-
bearing greenschists here. There is, however, a great potential for useful in-
formation contained in carbonate-bearing greenschists.
9.4.3
Greenschist-Amphibolite Facies Transition
The greenschist facies assemblage experiences two basic changes as meta-

morphic temperatures approach about 500 °C:
• Albite disappears and it is replaced by plagioclase. The composition of the
first Ca-bearing plagioclase is typically An 17 (oligoclase).
• Amphibole becomes capable of taking up increasing amounts of aluminum
and alkalis. Actinolite disappears and it is replaced by alkali-bearing alu-
minous hornblende.
The combined transformation results in the replacement of the Ab +Act pair

by the Pl + Hbl pair, i.e., in other words, replacement of greenschist by am-
phibolite. In orogenic metamorphism along a Ky-type path, the greenschist-
amphibolite facies transition occurs at temperatures of about 500°C (5 kbar).
Anorthite component is produced by a series of continuous reactions.
However, albite present in greenschist does not continuously change its com-
position along the albite-anorthite binary. At the conditions of the upper
greenschist facies the plagioclase solid-solution series is not continuous but
rather shows several miscibility gaps. The first Ca-bearing plagioclase that
forms in this manner is an oligoclase and its typical composition is An 17 • Be-
cause of the abrupt appearance of oligoclase, due to the miscibility gap in
the plagioclase series, the first appearance of plagioclase with An 17 can be
mapped in the field as an oligoclase-in isograd in mafic schists. This isograd
defines the beginning of the amphibolite facies; it separates greenschist facies
from amphibolite facies terrains.
The systematic compositional changes in the amphibole solid-solution se-
ries are generally more continuous in nature (although miscibility gaps do
exist there as well). The amphibole composition changes basically by: (1)
taking up tschermak component produced by a series of continuous reac-
tions, (2) incorporation of edenite component produced by albite-consuming

reactions, and (3) the unavoidable FM exchange. Other effects are related to
Ti incorporation and redox reactions.
The resulting amphibole is a green tschermakitic to pargasitic hornblende
that is now present together with plagioclase. The two minerals constitute,
by definition, an amphibolite and define the amphibolite facies. Chemogra-
phy #5 (Fig. 9.3) shows this new situation; hornblende has changed its com-
position along the TS exchange direction and plagioclase shows up on ACF
projections. In the beginning, chlorite and epidote may still by present in
amphibolites.
The most important mineralogical changes at the greenschist-amphibolite
transition can be related to reaction (50). The reaction consumes epidote and
chlorite from the greenschist assemblage and produces anorthite component
of plagioclase and tschermak component of amphibole. The reaction will
eventually consume all epidote or all chlorite in the rock, leaving behind a
chlorite- or epidote-bearing amphibolite. All three FM minerals of the
greenschist facies assemblage are consumed by reaction (51). It produces
tschermak component and consumes actinolite component of the amphibole.
Reaction (52) describes the formation of the edenite component of the am-
phibolite facies hornblende. The combined effect of the three reactions (51),
(52) and (53) is that the chlorite+epidote assemblage gradually disappears,
plagioclase becomes increasingly calcic and amphibole systematically changes
its composition from actinolite to alkali- and aluminum-bearing green horn-
blende.
Garnet may also appear at the transition to the amphibolite facies. Its for-
mation is accomplished by similar reactions as in metapelitic rocks
(Chap. 7). Garnet grows initially mostly at the expense of chlorite. Chlorite
decomposition contributes the bulk of the almandine and pyrope compo-
nents of garnet. Epidote-consuming reactions produce much of the grossular
and andradite components found in garnets in metamafic rocks. As in meta-
pelitic rocks, low-grade garnets are generally rich in manganese and the
strong fractionation of Mn into garnet makes the mineral appear in mafic
schists at temperatures as low as 450 °C. Reaction (54) consumes chlorite and
epidote and produces both grossular and Fe-Mg-garnet components, which
results in the typical ternary Ca-Fe-Mg-garnets found in mafic rocks.
9.5
Amphibolite Facies Metamorphism
9.5.1
Introduction
The amphibolite facies is characterized by chemographies #6, #7 and #8

(Fig. 9.3). The minerals plagioclase and hornblende make up the bulk volume
of amphibolites. All other minerals that can be present, such as quartz, epi-
dote, muscovite, biotite, garnet, and Cpx, are modally subordinate. Calcite
can be found in some amphibolites. Mineralogical changes within the amphi-
9.5 Amphibolite Facies Metamorphism 305
bolite facies mostly result from reactions that run continuously over a wide
P-T range. The main effect of these continuous reactions is seen in systemat-
ic variations in the composition of the two master minerals, plagioclase and
hornblende. The continuous reactions also cause epidote-clinozoisite to de-
crease in modal amount and eventually they disappear. Muscovite, if it sur-
vived from the greenschist facies, also gradually disappears. Garnet persists
and becomes modally more important up-grade (garnet amphibolites); clino-
pyroxene appears at higher temperatures in the amphibolite facies.
In principle, given a MORB composition and the composition of plagio-
clase and hornblende, the P-T conditions of equilibration are uniquely de-
fined. However, at present, experimental data do not permit a rigorous treat-
ment of plagioclase-hornblende relationships. In particular, solution proper-
ties of amphiboles are still poorly known and few end-member phase com-
ponents are well constrained. Surprisingly, the low-temperature behavior of
the plagioclase system is also not quantitatively known (e. g., quantitative
thermodynamic description of the various miscibility gaps and structural
transitions along the Ab-An binary at low temperature).
9.5.2
Mineralogical Changes Within the Amphibolite Facies
In prograde orogenic metamorphism of metabasalts, the rocks contain horn-

blende and plagioclase at the beginning of the amphibolite facies (-500 oC),
as explained in Section 9.4.3. In addition, amphibolites may still contain
some epidote and/or chlorite that have not been completely consumed by re-
actions that produced anorthite and tschermak components. Biotite may be
present as well. The same continuous reactions that produced the amphibo-
lite facies mineralogy initially continue to consume chlorite and epidote
within the lower part of the amphibolite facies. Eventually, chlorite complete-
ly disappears at about 550 oc and epidote is not typically found in amphibo-
lites that were metamorphosed to 600 oc. Some of the epidote- and chlorite-
consuming reactions produce garnet that, in general, becomes modally more
important with increasing grade. Chemography #6 is characteristic for the
central portion of the amphibolite facies around 600 oc. Here, amphibolites
contain plagioclase (typically andesine) and green hornblende± garnet± bio-
tite. At still higher temperatures, clinopyroxene appears in amphibolites (che-
mography #7). Cpx of the diopside-hedenbergite series usually appears
around 650 oc (along the Ky path) and a reaction that produces it is listed as
reaction (54). The reaction simultaneously produces Cpx and garnet; it is
typical of relatively high pressures. However, Cpx is also often found in am-
phibolites lacking garnet. A typical assemblage is Hbl + Pl + Cpx + Bt. Reac-
tion (55) produces diopside and anorthite components from amphibole and
epidote or zoisite.
This extremely important reaction has several significant effects: (1) it
continuously consumes epidote or clinozoisite that may still be present in
mid-amphibolite facies rocks and eventually eliminates them; (2) the reaction
consumes amphibole component, a process typical for the higher amphibolite
facies; (3) the reaction produces Cpx that appears in higher-grade amphibo-
lites; and (4), last but not least, the reaction produces even more anorthite
component that is incorporated in plagioclase in high-grade amphibolites.
Plagioclase in high-grade amphibolites becomes progressively more calcic
and in the upper amphibolite facies andesine-labradorite compositions are
typical of plagioclase in amphibolites (the range is An 30 -An70 ). Note, how-
ever, that phase relations in the plagioclase system are complex and little un-
derstood, so if you find a bytownite or anorthite (An 95 plagioclase) in an
upper amphibolite facies terrain, this is no reason to be confused.
The first appearance of clinopyroxene in amphibolites can be used to de-
fine the boundary to the upper amphibolite facies. Reactions in the upper
amphibolite facies, such as reactions (54) and (55), begin to break down am-
phibole components and replace them with pyroxene components. This is a
continuous process, however, and the first appearance of Cpx in mafic rocks
is usually not a sharp isograd in the field. Nevertheless, the "Cpx-in" transi-
tion zone marks the beginning of the upper amphibolite facies and a repre-
sentative temperature is about 650 °C. In water-saturated environments, meta-
mafic rocks show the first structural evidence in the field of local partial
melt formation, migmatization and appearance of quartzo-feldspathic seams,
patches, veins and similar mobilisate structures in the upper amphibolite fa-
cies. The processes remove quartz, plagioclase and biotite from the rocks
and transfer them to a melt phase.
At higher pressures than that of the ordinary Ky-geotherm (Fig. 9.3), am-
phibolites may contain the diagnostic kyanite +hornblende assemblage. The
Ky+ Hbl pair may have formed by reaction (39) earlier in the course of pro-
grade metamorphism (see Fig. 9.7). The link to the more common plagio-
clase+ garnet assemblage found in amphibolites is given by reaction (56).
The reaction replaces the garnet+ plagioclase tie line of chemographies #6
and #7 by the kyanite +hornblende tie line of chemography #8. The continu-
ous dehydration reaction runs typically between chemographies #6 and #8 in
Fig. 9.3 for typical metabasalt compositions. Therefore, kyanite-bearing am-
phibolites are diagnostic of high-pressure amphibolite facies (pressures typi-
cally greater than 7 kbar). Remember, however, that the Ky-Hbl-Grt-Pl as-
semblage is not normally coplanar (as one may erroneously conclude from
ACF figures such as Fig. 9.3) but rather defines a phase volume that occurs
over a relatively wide P-T interval.
9.5.3
Low-Pressure Series Amphibolites
At low metamorphic grades, mafic rocks that were metamorphosed along a

Ky-type and a Sil-type path, respectively, are very similar in mineralogy. The
most significant difference may be found in the composition of chlorite and,
especially, K-white mica. The celadonite component of K-white mica is rela-
tively sensitive to pressure variations and it is controlled by processes such
as reaction (45). The tschermak component in mica (and in principle also
9.5 Amphibolite Facies Metamorphism 307
chlorite or biotite) can be used to monitor pressure conditions in low-grade

metamorphism.
The transition to the amphibolite facies at low pressures is similar to the
one described in Section 9.4.3. In extremely Al-rich rocks, andalusite may ap-
pear instead of kyanite (chemography #10). However, andalusite +amphibole
is not very common. The phase relationships shown in Fig. 9.7 suggest that
the transition to the amphibolite facies occurs at slightly lower temperatures
in low-pressure metamorphism. This is reasonable because the reactions that
produce amphibolite from greenschist are ordinary continuous dehydration
reactions. The lower boundary of the amphibolite facies along a Sil-path is
typically at about 450 oc (at 3 kbar). Reactions (36), (37), (40) and the anda-
lusite equivalent of reaction (41) pass through the andalusite field in Fig. 9.3.
The reactions generally produce anorthite and amphibole components from
greenschist facies mineralogy. At 550 oc (Figs. 9.3 and 9.7), epidote and
chlorite typically disappeared from low-pressure amphibolites and the char-
acteristic assemblage is hornblende+ plagioclase± andalusite ±biotite.
Compared with orogenic metamorphism along a Ky-path, clinopyroxene
forms at significantly lower temperatures in low-pressure amphibolites
(600 oc or lower). Chemography #11 is characteristic of amphibolites that
formed in the range 3-4 kbar. It shows the usual hornblende+ plagioclase
assemblage together with clinopyroxene and biotite. Sillimanite may be present
in Al-rich amphibolites and the Sil + Hbl pair forms a stable, though rare,
assemblage. However, garnet is scarce or even absent in low-pressure amphibo-
lites.
Chemography #12 is representative for amphibolites in the upper amphi-
bolite facies at low pressures (Sil path). The most significant feature is the
presence of ferro-magnesian amphiboles in addition to the calcic amphiboles
(hornblende). The types of Fe-Mg amphiboles that occur together with horn-
blende include anthophyllite, gedrite and cummingtonite. In some amphibo-
lites, as many as three different amphibole species may be present such as
hornblende, gedrite and anthophyllite. The abbreviation "Oam" for orthoam-
phibole in Fig. 9.3 includes all Fe-Mg amphiboles and also gedrite. Phase re-
lationships in such multi-amphibole rocks can be very complex. Miscibility
gaps in various amphibole series as well as structural changes in amphiboles
complicate the picture. However, such low-pressure amphibolites possess a
great potential for detailed analysis of relationships among minerals and as-
semblages and for the reconstruction of the reaction history of low-pressure
amphibolites. Fe-Mg-amphiboles may form by a number of different conceiv-
able mechanisms. The first and obvious one is reaction (57) that links che-
mography #11 to chemography #12. The reaction breaks down the hornblen-
de+ sillimanite assemblage and produces anorthite component that increases
the An content of plagioclase and also produces anthophyllite component
that is a major component in any Fe-Mg-amphibole. As metamorphic grade
increases, reaction (58) becomes more important in producing anthophyllite
component. The reaction consumes calcic amphibole component and replaces
it with clinopyroxene and Fe-Mg-amphibole. The reaction equilibrium of re-
action (58) in metabasaltic rocks involves phase components of the typical
assemblage: Hbl + Oam + Cpx + Pl ± Qtz. As a general rule of thumb, the as-
semblage is characteristic for the temperature range between 650 and 750 °C
along the Sil geotherm.
9.5.4
Amphibolite-Granulite Facies Transition
Much of the quartz that has been produced in the subgreenschist facies to
greenschist facies metamorphism has been consumed by continuous reac-
tions in the amphibolite facies. Many amphibolites are quartz-free (use ACF
diagrams with care!). Most of the water bound in hydrous minerals has been
released by reactions during prograde metamorphism. Metabasaltic rocks
that have been progressively metamorphosed to 700 ac contain Pl + Hbl ±
Cpx± Grt± Bt (Ky-path) and Pl + Hbl± Cpx± Oam± Bt (Sil path). Amphibole
is the last remaining hydrous phase and further addition of heat to the rock
by tectono-thermal processes will ultimately destroy the amphibole. The
transition from a hydrous amphibolite facies assemblage to a completely an-
hydrous granulite facies assemblage is gradual and takes place over a tem-
perature interval of at least 200 ac (from about 650-850 °C). The first clear
and unequivocal indication that granulite facies conditions are reached is the
appearance of orthopyroxene in Cpx-bearing quartz-free rocks. Granulite fa-
cies is obvious if Opx turns up in quartz-bearing amphibolites. Orthopyrox-
ene is most common in low-pressure mafic granulites. At higher pressures,
the typical anhydrous granulite facies assemblage is Pl + Cpx + Grt ± Qtz. The
reactions that eventually produce this assemblage are continuous in nature
and the temperature interval is very wide. Hbl gradually decreases in modal
amount, leaving behind the anhydrous granulite facies assemblage. Note,
however, that Pl + Cpx + Grt may also occur in the amphibolite facies and that
the assemblage as such is by no means diagnostic for granulite facies condi-
tions. If pressure conditions are too high for Opx to form, then the ultimate
granulite facies condition is reached when the last crystal of amphibole dis-
appeared from the rock. Reactions (59)-(62) consume the major amphibole
components tremolite and tschermakite and produce pyroxene, garnet and
anorthite components and, hence, represent the transition of the amphibolite
to the granulite facies. Reaction (60), in particular, is of great importance. It
decomposes hornblende and makes the two-pyroxene assemblage that is di-
agnostic for the granulite facies in quartz-absent mafic rocks. It produces, in
addition, anorthite component that ultimately leads to the calcic plagioclase
found in mafic granulites.
What happens to the biotite? Biotite is a subordinate hydrous phase in
high-grade amphibolites. The small total amount of potassium stored in bio-
tite will be taken up by pyroxene and ternary feldspar or it is used to form a
Kfs-Ab-Qtz melt in the migmatite-producing process in high-grade meta-
morphism of mafic rocks. The total amount of water that can be stored in
high-grade amphibolites is very small (in the order of 0.4 wt% H2 0 or less).
Consequently, mafic rocks are less susceptible to partial melting compared
with their metapelitic or metagranitoid counterparts.
9.6 Granulite Facies and Mafic Granulites 309
9.6
Granulite Facies and Mafic Granulites
Granulite facies metamorphism of mafic rocks is represented by chemogra-

phies #9 and #13 in Fig. 9.3. At high pressures the characteristic assemblage
is: plagioclase+ clinopyroxene (augite)+ garnet. At lower pressures the typical
granulite facies assemblage is: plagioclase+ clinopyroxene (augite)+ orthopy-
roxene (hyperstene). The two assemblages are linked by the important reac-
tion (63). The reaction separates a field with pyroxene granulites at pressures
below about 5-7 kbar from a field with garnet granulites at pressures above
5-7 kbar and below the eclogite field at very high pressures (at 800 °C). The
continuous nature of reaction (63) generates a wide overlap zone with both
garnet and Opx present in Pl + Cpx rocks. If strictly concerned with meta-
MORE's, however, the overlap is smaller and the boundary occurs close to
6-7 kbar at 800 oc.
Because high-grade brown hornblende is an extremely stable mineral, very
high temperatures are necessary to destroy the last hydrous phase in mafic
rocks. Completely anhydrous mafic granulites form normally above 850 °C or
at even higher temperatures. Dehydration of amphibole, on the other hand,
is often aided by interaction of the rocks with a foreign fluid phase that is
poor in H2 0. Alternatively, dehydration of amphibole may be triggered by
partial melting and removal of H 2 0 in a melt phase; see also Chapter 7
(metapelitic granulites), where some aspects of granulite facies metamorph-
ism are discussed. Figure 9.8 shows some of the relationships that are impor-
tant in high-grade and granulite facies metamorphism of mafic rocks.
Figure 9.8 is a schematic temperature versus fluid composition diagram at an
approximate total pressure of 6 kbar. Along the 750 °C isotherm, three
different assemblage fields intersect. If H 2 0-rich fluids are present (section
A), the mafic rocks undergo partial melting and mafic migmatites are char-
acteristic for such conditions. Intermediate fluid compositions are consistent
with amphibolite facies rocks with the typical mineralogy: Pl + Hbl + Cpx
(section B). Migmatite structures are absent. Fluids low in H 2 0 coexist with
granulite facies rocks at the same temperature and the characteristic assem-
blage is: Pl + Cpx + Opx ± Grt (section C). Note that, in real rocks, the
transition zone between the granulite and amphibolite facies assemblages
(Fig. 9.8) covers a wide range of fluid composition. The important message
of Fig. 9.8 is, however, that at the same temperature (e. g., 750 °C) mafic mig-
matites, amphibolites and mafic granulites may be found intimately asso-
ciated in a metamorphic complex or terrain depending on the composition
of the fluid phase present in the volume of rocks under consideration. The
situation may be complicated by variations of fluid composition over time at
a particular site. Also note that the first melt in mafic rocks forms if pure
H 2 0 fluids are present (wet solidus in Fig. 9.8). With increasing temperature,
migmatites form with fluids that contain decreasing amounts of water. The
amphibolite facies assemblage has a temperature maximum in Fig. 9.8. The
actual value of the maximum temperature for amphibolite can be correlated
with the amphibole-out isograd that is at about 900 oc (at 6 kbar) in pure
mafic
migmatites granulite-facies rocks
' / maximum T lor amphlbol ltes
PI + Cpx + Opx ± Grt

c
--
00
...f...
;:,
«< amphibolite-facies rocks
G)
c.
E
G)
1- p = 6 kbar
fluid composition
Fig. 9.8. Schematic isobaric temperature versus fluid composition diagram showing amphi-
bolite-granulite facies relationships
H2 0 fluids and may be closer to 850 °C in mixed volatile fluids at the same
pressure.
Although completely anhydrous mafic granulites are widespread (e. g., in
the Jotun nappe and the Bergen arcs of the Scandinavian Caledonides), many
mafic granulites still contain prograde hornblende (as opposed to retrograde
post-granulite Hbl). The transitional nature of the amphibolite to granulite
facies transition that occurs over a wide P-T range requires a definition of
granulite facies in mafic rocks. In low-pressure terrains this is a simple task:
the first appearance of orthopyroxene in mafic rocks (quartz-free or quartz-
bearing does not matter) marks the onset of granulite facies conditions. Ma-
fic rocks will, at this stage, always contain clinopyroxene as a major mineral.
The assemblage orthopyroxene+ clinopyroxene will then be diagnostic for
two-pyroxene granulites. The case is more difficult at high pressures where
Opx does not form in mafic rocks and the typical assemblage Pl + Cpx + Grt
may also be present in the amphibolite facies or hornblende persists in mafic
rocks to very high temperatures {1000 °C at 10 kbar). Obviously, in this case,
the definition of an amphibolite-granulite facies boundary is impossible on
the basis of mafic rocks alone (see comments in Chap. 4 on metamorphic
grade).
In addition to the minerals discussed so far, mafic granulites may also
contain more exotic minerals such as sapphirine or scapolite in various as-
9.7 Blueschist Facies Metamorphism 311
semblages. Also, hercynitic spinel is commonly present in mafic granulites

and its presence is diagnostic for low-pressure conditions ( < 4 kbar).
At very low pressures, such as represented by chemography #14, pyroxene
hornfelses form by high-temperature contact metamorphism of mafic rocks.
Ultimately, the mineralogy of the protolith basalt is expected to form in ul-
tra-high-temperature contact metamorphism of mafic rocks. The assemblage
Pl+Cpx+Opx±Spl (chemography #14) is then identical to the basalt miner-
alogy shown in the upper left corner of Fig. 9.3. The upper boundary of
granulite facies metamorphism of mafic rocks is given by the dry liquidus
for basalt ( -1200 °C),
Geological causes of granulite facies metamorphism at low pressures are
commonly attributed to intrusions of mafic (gabbroic/basaltic) or charnocki-
tic magma into the continental crust (see references and further reading of
this chapter and Chap. 7). Large volumes of dry or COrrich hot magma
from the mantle or lower crust serve as heat source for extensive high-tem-
perature metamorphism at shallower crustal levels. The dry or COrrich na-
ture of the magma facilitates dehydration of the crustal rocks. There is little
tectonic activity associated with this type of granulite facies metamorphism.
After heating to high temperature the rocks cool essentially isobarically at
the crustal level at which they reside. Many granulite facies terrains are char-
acterized by isobaric cooling paths (and perhaps a counter-clockwise P-T
loop; see also Chap. 3). In general, the exhumation of granulite facies terrains
requires a later contractional orogenic event that involves stacking and tilting
of crustal slices and, e. g., the formation of fold nappe structures that ulti-
mately permit the erosion surface to intersect with granulites of the middle
or lower crust. If water becomes available during slow cooling, the high-
grade assemblages may be completely erased and granulites may lose their
memory of the granulite facies event. In particular, if Opx disappears from
mafic granulites during amphibolitization, it is very hard to tell that the rock
ever went through a granulite facies episode.
9.7
Blueschist Facies Metamorphism
9.7.1
Introduction
Blueschists are rocks that contain a significant amount of blue alkali-amphi-

bole with a very high proportion of glaucophane end-member component.
Such rocks are, in an outcrop or hand specimen, blue in color. Note, how-
ever, that pure glaucophane is colorless under the microscope. If a rock con-
tains a blue pleochroic amphibole in thin section it means that it is not a
pure glaucophane and appreciable amounts of ribeckite and other sodic fer-
ro-ferric amphibole components are important and the overall composition
of the amphibole is crossite (amphibole name recently disapproved by IMA,
however, I like it!). Note also that Cl-rich hornblende may have a blue color
under the microscope. However, if mafic rocks of basaltic bulk composition
...........
20
not
18 geolllnnl
16
14
-... 12
-
cu
..Q 10
.:tt.
f!! 8
:I
en
en
f!! 6
ll.
4
0 100 200 300 400 500 600 700 800 900 1000
Temperature CC)
Fig. 9.9. High-pressure metamorphism overview
contain sodic amphibole, the rocks are likely to have been metamorphosed
under the conditions of the blueschist facies.
The general P- T field of the blueschist facies is delineated in Fig. 4.2. The
blueschist facies is represented in Fig. 9.3 by chemographies #16 and #17.
The lower-pressure version of the blueschist facies #16 is characterized by
glaucophane (sodic amphibole)+ lawsonite+ chlorite assemblages. In the
higher-grade blueschist facies #17, typical metamafics contain various assem-
blages among glaucophane (sodic amphibole), zoisite (clinozoisite-epidote),
garnet, paragonite, phengite, chlorite, talc, kyanite, rutile, ankerite, and other
minerals.
A detailed rendering of the blueschist field is shown in Fig. 9.9. It is
bounded towards the subgreenschist and greenschist facies by reactions (64),
(65) and (66). Its high-pressure boundary is given by reaction (71). Blue-
schist facies terrains are worldwide associated with subduction zone meta-
morphism that typically occurs along destructive plate margins where ocean-
ic crust (basalt) is recycled to the mantle under continental lithosphere. Two
typical geotherms related to subduction tectonics are depicted in Fig. 9.9.
The slow subduction geotherm only marginally passes through the blueschist
field and eclogite assemblages are produced at pressures as low as 13-
14 kbar. If subduction is fast, prominent blueschist facies terrains may devel-

op and eclogite formation does not occur until pressures of 18-20 kbar are
reached (corresponding to subduction depths of 50-60 km). Because there
are natural limits to attainable subduction velocities in global tectonic pro-
cesses, the P-T conditions in the upper left wedge in Fig. 9.9 are not accessi-
ble conditions for rock metamorphism.
9.7.2
Reactions and Assemblages
As discussed above (Fig. 9.9), reactions (27) and (28) mark the boundary be-
tween the subgreenschist facies and the greenschist facies. The boundary be-
tween the subgreenschist facies and the blueschist facies is given by reac-
tion (64). This reaction replaces the low-grade assemblage actinolite+ chlo-
rite+ albite by the high-pressure assemblage glaucophane+ lawsonite. The re-
action is H2 0-conserving, and as such independent of water pressure or the
composition of the fluid phase. Equilibrium of reaction (64) is shown in
Fig. 9.9 for average mineral compositions found in low-grade equivalents of
basaltic rocks.
Point A in Fig. 9.9 marks an important P-T region where blueschist,
greenschist and subgreenschist facies assemblages meet. It is at about 8 kbar
and 300 a c. The P-T region is defined by the approximate intersection of re-
actions (22), (28), {30), (31), (64), (66) and (70). Note, however, that point A
is not an invariant point in Schreinemakers' sense but rather a relatively nar-
row P-T range of locations of various invariant points generated by the inter-
section of continuous reactions in the NCMASH system.
The blueschist facies is separated from the greenschist facies by reac-
tions (65) and (66). The stable equilibrium of these reactions is shown in
Fig. 9.9. Reaction (65) consumes the low-grade assemblage chlorite+actino-
lite +albite and replaces it by the most typical blueschist facies assemblage
glaucophane+ epidote (zoisite). Reaction (65) connects the points A and Bin
Fig. 9.9, respectively. It nearly coincides with another very important reac-
tion, reaction (66). The reaction replaces the greenschist facies assemblage
chlorite+ albite by the characteristic blueschist assemblage glaucophane+
paragonite. The combined effect of reactions (65) and (66) is that the chlo-
rite+ albite+ actinolite assemblage, that is diagnostic for low-grade metama-
fic rocks, disappears and the blueschist assemblage glaucophane+ epidote+
paragonite is formed. If for any reason some albite should still be present in
the rocks after all the chlorite has been used up, that albite will be destroyed
by reaction (70). The reaction boundary of reaction (66) roughly coincides
with the diagonally ruled boundary that separates a plagioclase present from
a plagioclase absent region (Fig. 9.10). Like in the eclogite facies, blueschists
do not contain stable albite (plagioclase). In most low-grade mafic rocks,
however, chlorite is much more abundant than albite. Consequently, reac-
tions (65) and (66) will normally destroy all albite.
The glaucophane+ paragonite assemblage is replaced by omphacite +gar-
net as a result of reaction (71). The reaction terminates the blueschist field
not attainable
20 geotherms
-c-
eo
.c
-
15
..lll::
200 300 400 500 600 700 800
Temperature (°C)
Fig. 9.10. Selected mineral equilibria in high-pressure metabasites. =plagioclase present-
absent transition zone and coexistence of Na-pyroxene, plagioclase and quartz. The heavy
dashed line is the approximate high-pressure boundary for plagioclase in mafic rocks. On
the high-T side of reaction (71) and at the high-P side of the dashed line=field of eclogites.
aH2 o= 1; ap~ = 0.9 in reaction (78), other activities in reaction (78) correspond to those of
reaction (70), arrp = 0.3 and a1d=0.4 in reaction (71)
towards higher pressures where it is grading into the eclogite field with the di-
agnostic omphacite +garnet assemblage. The sodium present in mafic igneous
rocks is stored at low grade in albite feldspar, in the blueschist facies in alkali
amphibole (glaucophane, crossite) and paragonite, and in the eclogite facies in
sodic pyroxene (omphacite, jadeite). Reaction (71) has a negative slope on a P-
T diagram (Figs. 9.9 and 9.10) and terminates at point B towards lower pressure
where it becomes metastable relative to feldspar-involving reactions. The neg-
ative slope of reaction (71) means that the most favorable conditions for the de-
velopment of extensive blueschist facies terrains are met in tectonic settings of
very rapid subduction of cold (old) oceanic crust.
Inside the blueschist facies field, some additional reactions may modify
the assemblages. Most important is reaction (67). It terminates the lawson-
ite+ glaucophane assemblage that is represented by chemography #16 and
garnet shows up in blueschists towards higher pressure (chemography #17).
Another important garnet producer in the blueschist facies is reaction (68).
Reaction (69) generates the relatively rare glaucophane+kyanite assemblage
that is diagnostic for very low-T high-P glaucophane schists.
The continuously running reactions in the blueschist facies field result in
a systematic, also gradual, change in the mineralogical and modal composi-
tion of metabasaltic rocks during prograde metamorphism. Let us follow a

metabasaltic rock with a subgreenschist facies assemblage that undergoes
subduction at a rate intermediate to the slow and fast subduction geotherms
in Fig. 9.9. Such a geotherm will cover a temperature interval in the blue-
schist facies field from 250-500°C and a pressure range from 7-16 kbar. At
pressures between 6-8 kbar, lawsonite+ alkali amphibole will begin to replace
the low-grade chlorite+ actinolite+ albite assemblage. The alkali amphibole is
initially fairly Fe-rich, containing high proportions of ribeckite component.
As pressure further increases, albite gradually disappears and amphibole be-
comes increasingly sodic and depleted in actinolite component. Sodium
gradually replaces calcium on the M4-position of the amphibole structure by
an exchange mechanism that can be formulated as the glaucophane exchange
component: NaM 4 AlCaM 4 _ 1Mg_ 1 • The exchange leaves the A-site of the amphi-
bole structure untouched. The continuous reactions produced the lawsonite+
crossite +chlorite+ paragonite± phengite assemblage that is most characteris-
tic for pressures around 10 kbar. At still higher pressures, lawsonite gradually
disappears, alkali amphibole becomes increasingly enriched in glaucophane
component, zoisite and epidote become important members of the assem-
blages, and chlorite continuously decreases in modal amount. The character-
istic assemblage for this intermediate blueschist facies stage is: glaucophane
+epidote+ chlorite+ paragonite± phengite. Typical pressures for that assem-
blage are in the range 12-14 kbar (along the geotherm we are presently look-
ing at). With increasing pressure, garnet becomes a member of the assem-
blage. Chlorite further decreases in modal amount giving rise to increasing
amounts of garnet. High-pressure blueschists most typically contain: glauco-
phane+ epidote (clinozoisite, zoisite) +garnet+ paragonite± phengite. Other
minerals that might be present in high-P low-T blueschists include Mg-chlor-
itoid, talc and kyanite. For a discussion of systematic compositional changes
in the only important potassic mineral in blueschists, i.e., phengite, with in-
creasing pressure you should consult Chapter 7.8. Also note that blueschists
often contain carbonate minerals that give rise to mixed volatile reactions
that are not discussed here. Carbonates found in blueschists include calcite
and its high-pressure polymorph aragonite, ankerite, dolomite, breunerite
and magnesite. The origin of carbonates in blueschists is related to either
ocean-floor metamorphism of basalt and metasomatism associated with it or
they are derived from carbonate sediment deposited concurrently with the
extrusion of pillow basalt lavas on the ocean floor.
The blueschist facies assemblages are gradually replaced with the typical
eclogite facies assemblage omphacite +garnet at pressures greater than about
14-16 kbar. Because high-pressure blueschists typically contain garnet at
these conditions, it is the appearance of omphacite (sodic pyroxene) that
marks the transition to the eclogite facies. The boundary is rather gradual
and there is a wide pressure range where glaucophane+ paragonite+ epi-
dote+ garnet+ omphacite may occur in a stable association. Strictly speaking,
this assemblage is diagnostic for the eclogite facies. Along an intermediate
high-P low-T geotherm, typical eclogite assemblages are formed at about
16-18 kbar and a corresponding temperature of about 500°C.
9.8
Eclogite Facies Metamorphism
9.8.1
Eclogites
Eclogites are metamorphic mafic rocks that contain the stable mineral pair
garnet and omphacite in significant modal amounts. Eclogites are free of
plagioclase. Eclogites are very dense rocks with a density even greater than
some ultramafic mantle rocks (I? eclogite > 3300 kg m - 3 ). Because of the high
density of eclogites, their origin can be readily related to very high-pressure
conditions during formation. Omphacite +garnet is the diagnostic assem-
blage in metabasaltic rocks that recrystallized above about 12-14 kbar pres-
sure outside the stability field of plagioclase. Typical eclogites form at pres-
sures of 18-22 kbar and higher. Many of the eclogites found in collisional
mountain belts such as the Alps or the Scandinavian Caledonides still display
clear compositional and structural evidence for being derived from basaltic
lavas. In the Zermatt region of the Central Alps, for example, preserved
MORB compositions and pillow lava structures may be found in eclogites
that formed at pressures greater than 20 kbar corresponding to a subduction
depth of 60 km. This, in turn, reveals compelling evidence that surface rocks
(basaltic lava) can be transported by active tectonic processes to depths of
60 km below the surface (or in the case of coesite-bearing eclogites to
> 100 km).
The P-T field of the eclogite facies shown in Fig. 9.9 can be accessed in
prograde metamorphism from all three neighboring facies fields, blueschist,
amphibolite and granulite facies, depending on the tectonic setting. In sub-
duction tectonics, eclogites originate from blueschists, as outlined above.
These types of subduction-related eclogites often register very high pressures
of formation.
In a setting of continent-continent collision, the continental crust often at-
tains twice its normal thickness and the deeper parts of the double-crust are
exposed to pressures in the range 12-24 kbar. Any mafic rocks of suitable
composition in the deeper part of the double-crust could be transformed to
eclogite (Fig. 9.9). An example of such a setting is the collision of the Eur-
asian and the Indian plate and the resulting double-crust formation of the
Tibetan plateau. A Ky-type geotherm, typical of collision tectonics, is shown
in Fig. 9.9. It can be seen that eclogites in such a setting will be created from
amphibolites. Eclogites that form in such a setting of crustal stacking and
nappe formation do not record extremely high pressures of formation. Pres-
sures in the range 14-18 kbar are most characteristic.
If collision is accompanied by substantial magmatic heat transfer to the
crust, eclogites may form from granulites and the "hot" geotherm shown in
Fig. 9.9 may be more appropriate in this situation. Mantle-generated basaltic
magmas often sample pieces of rocks along the conduits during ascent to the
surface. The rock fragments typically found as xenolith in lavas of basalt vol-
canoes include ultramafic rocks from the mantle, granulites and eclogites.
9.8 Eclogite Facies Metamorphism 317
The association is characteristic of abnormally hot geotherms in conjunction

with collision and magmatism, initial stages of extension or magmatism
alone. In such settings, eclogites may have been created from mafic granu-
lites or directly crystallized from basaltic magma at great depth. The pressure
range of high-temperature eclogites is very large; however, typical pressures
are in the same range as for the eclogites formed from amphibolites. Basaltic
magmas may, on the other hand, also collect some eclogite xenoliths from
great depth in the mantle and transport them to the surface. The ultimate
origin of these deep eclogite samples is ambiguous, however. They also may
represent pieces of recycled former oceanic crust.
From the comments above it follows that eclogites may form in a wide
range of tectonic settings and the eclogite facies comprises the widest region
of temperature and pressure conditions of any of the metamorphic facies
fields. The temperature ranges from about 400 to 1000 oc. It may be reason-
able to distinguish three general types of eclogites depending on geological
setting and temperature of formation:
1. Low-temperature eclogites are related to subduction tectonics and form
from blueschists (high-pressure low-temperature eclogites= LT eclogites).
2. Intermediate-temperature eclogites form in continent collision settings
from amphibolites (medium-temperature eclogites= MT eclogites).
3. High-temperature eclogites form in settings of collision or extension where
the geotherm is abnormally "hot" due to magmatic heat transfer from the
mantle (e. g., basalt magma underplating), from mafic granulites or crys-
tallize directly as eclogites from mafic magma (high-temperature eclog-
ites= HT eclogites).
The three different genetic types of eclogite are also characterized by typical
mineral associations because of dramatically different temperatures of forma-
tion and because of widely varying H 2 0 pressures associated with their for-
mation. While the LT eclogites that formed in subduction zones often contain
modally large amounts of hydrate minerals, HT eclogites often hold "dry" as-
semblages.
In dealing with high-pressure rocks, important issues also include the
kind of mechanism by which they return to the surface, what kind of de-
tailed P-T path they follow on their way to the surface, and, most impor-
tantly, what kind of modifications the rocks experience as they are returned
to the surface. For example, if an eclogite formed by a subduction mecha-
nism and equilibrated at 650 °C and 20 kbar, it is crucial for the fate of the
eclogite assemblages whether the rocks reach the surface via the blueschist
field (cooling and decompression occur simultaneously), via the amphibolite
and greenschist fields (first decompression then cooling), or via the granulite
field (first heating and decompression then cooling). Most eclogites display
some mineralogical impact of reactions along the return path to the surface.
The access of H 2 0 along that path is, of course, essential whether or not the
high-P assemblages are conserved or extensive retrogression occurs. On the
other hand, as will be shown below, LT eclogites often contain abundant hy-
drate minerals at peak pressure. Such rocks will undergo extensive dehydra-
tion reactions that will eventually eliminate the eclogite assemblage if decom-
pression occurs without concurrent cooling.
9.8.2
Reactions and Assemblages
Reactions that connect the blueschist facies to the low-temperature eclogite

facies were discussed in Section 9.7.2. The most typical assemblages in LT
eclogites involve the following minerals (bold= compulsory minerals; ± = op-
tional minerals): garnet+ omphacite ± zoisite ± chloritoid ± phengite ±para-
gonite± glaucophane± quartz± kyanite ±talc± rutile± dolomite.
Omphacite is a sodic high-pressure clinopyroxene that is composed of the
major phase components jadeite, acmite, diopside and hedenbergite (a solid
solution of: NaA1Si 2 0 6 -NaFe3 + Si2 0 6 -CaMgSi2 0 6 -CaFeSi2 0 6 ). The M2 site of
the pyroxene structure is filled with about 50% Na and 50% Ca in typical
omphacite. LT eclogites often contain pure jadeite together with omphacite.
Omphacite contains commonly minor amounts of chromium giving the
mineral a grass-green color. This is particularly the case in metagabbros.
Note that Cr-diopside in garnet-peridotite also shows a characteristic grass-
green color. However, because ultramafic rocks contain very little sodium,
the jadeite component in Cpx of high-pressure ultramafic rocks is generally
low.
Garnet in eclogite is a solid solution of the major phase components al-
mandine, pyrope and grossular (a solid solution of: Fe3AhSi 3 0 12-Mg3 Ah
Si30 12 -Ca3Al2Sh0 12). However, due to its extreme refractory nature, garnet
in eclogite can be inherited from earlier stages of metamorphism or be relics
from the protolith (e.g., garnet-granulite).
Chloritoid in LT eclogites is very magnesium-rich and XMg may be as high
as 0.5. Chloritoid and paragonite are diagnostic minerals in LT eclogites. The
two minerals are removed by various reactions from eclogites at higher tem-
peratures and are typically absent in MT- and HT eclogites.
Phengite and other sheet silicates in LT eclogites derived from gabbros also
often contain small amounts of chromium giving them a green color (Cr-
phengite = fuchsite).
9.8.2.1
Amphibolite and Granulite to Eclogite Facies Transition
Amphibolites and granulites contain plagioclase feldspar that is rich in an-

orthite component. Reaching the eclogite facies from the amphibolite facies
or the granulite facies, the reactions that destroy feldspar at high pressures
must also decompose the anorthite component of plagioclase.
Reaction (72) produces Ca-tschermak component that is taken up by om-
phacite. The equilibrium of this reaction is nearly independent of tempera-
ture and it can be used for reliable pressure estimates. Reaction (73) replaces
anorthite by zoisite and kyanite. This reaction is the most important an-
orthite-consuming reaction in high-pressure metamorphism of both MT and
HT eclogites. The reaction is a hydration reaction that consumes H 2 0 as well

as anorthite component. In the absence of water, anorthite will not decom-
pose to zoisite. However, if water is not available as solvent, then reaction ki-
netics are so slow and transport distances are so small that albite and other
low-pressure phases or phase components will also survive the high pres-
sures. For example, in an olivine gabbro from the Alps (Allalin gabbro, see
below), the primary igneous minerals survive pressures in excess of 20 kbar
at 650-700 oc without forming eclogite except where water had access to the
rocks. Another example is from the Bergen Arcs (Norwegian Caledonides), a
basement nappe complex that experienced P -20 kbar and T -700 °C during
Caledonian metamorphism. It was found that the mineralogy of Precambrian
mafic granulite is perfectly preserved over large areas except along an anas-
tomosing network of shears where the access of water permitted the forma-
tion of the stable rock under the Caledonian metamorphic conditions, i.e.,
MT eclogite.
The equilibrium conditions of reaction (73) are shown in Fig. 9.10. It runs
roughly parallel to the albite breakdown reaction (70). Reaction (74) decom-
poses anorthite component and produces grossular component in eclogite
garnet together with kyanite and quartz (Fig. 9.10). The slope of the reaction
is similar to those of reactions (70) and (73).
Reaction (70) was used in Chapter 3 to illustrate some general principles
of chemical reactions in rocks. In high-pressure mafic rocks it is the most
important of all reactions. It limits a plagioclase-present from a plagioclase-
absent region of the P-T space (Fig. 9.10). If all three minerals are present
(quartz, plagioclase and sodic pyroxene), the equilibrium conditions of reac-
tion (70) can be used for pressure estimates (if the temperature can be esti-
mated from a thermometer, e.g., Fe-Mg distribution between garnet and om-
phacite). If quartz is not present, a maximum pressure can be estimated from
the pyroxene-feldspar pair. If feldspar is not present (the normal case in
eclogites), the omphacite+quartz assemblage can be used to estimate a mini-
mum pressure for eclogite formation by assuming unit activity for the phase
component albite in the equilibrium constant expression for reaction (70).
Other important reactions of the amphibolite (or granulite) to eclogite tran-
sition zone must consume calcic amphibole or pyroxene. An example is reac-
tion (75) that destroys tremolite and anorthite component (hornblende+ pla-
gioclase= amphibolite) and produces garnet+ pyroxene. Reactions (76)-(78)
are examples of anorthite-consuming reactions that are relevant for the tran-
sition from the granulite to the HT eclogite facies.
As a result of the reactions that produce medium-temperature eclogites
(from amphibolites) the typical mineralogy found in meta-MORB is: garnet
+ omphacite ± zoisite (clinozoisite) ± phengite ± kyanite ±amphibole (Na-Ca
amphibole)± quartz (at very high pressure coesite) ±rutile.
High-temperature eclogites (from granulites) contain typically: garnet+
omphacite ± kyanite ±orthopyroxene± amphibole± quartz (at very high pres-
sure coesite) ±rutile.
9.8.2.2
Reactions in Eclogites
Prograde metamorphic mineral reactions that affect eclogites are basically or-
dinary dehydration reactions. The reactions transform assemblages involving
hydrous minerals of LT eclogites and substitute them with less hydrous as-
semblages of MT eclogites and ultimately HT eclogites. Examples are the
paragonite breakdown reaction (79), the zoisite breakdown reaction (80), and
the replacement of talc+ kyanite by orthopyroxene [reaction (81)]. Other re-
actions gradually remove chlorite from quartz-free MT eclogites, some of
them producing Opx. Reaction (80) replaces zoisite and garnet and produces
kyanite + Cpx. This continuous reaction relates MT to HT eclogites. The reac-
tion ultimately leads to the change from chemography #18 to #19 shown in
Fig. 9.3.
Other important reactions in HT eclogites are reactions (82) and (83). The
last reaction is important in the relatively rare Opx-bearing eclogites that
occur, e. g., in the Western Gneiss region of the Scandinavian Caledonides.
Phengite is the only K-bearing phase in LT eclogites. At higher temperatures,
the mica continuously decomposes under production of K-amphibole compo-
nent. The K-bearing Na-Ca-amphiboles found in high-grade eclogites are
very stable and remain the prime K-carrier in HT eclogites.
9.8.3
Eclogite Facies in Nonbasaltic Mafic Rocks
Mafic or basic rocks cover a very wide range in bulk composition. Typical
eclogites with garnet and omphacite as the dominant minerals usually devel-
op from basaltic protoliths. The above discussion of eclogites refers to
MORB-type mafic rocks. However, many gabbros are compositionally dissim-
ilar to MORB. In particular, gabbros are commonly (but not exclusively; e. g.,
ferro-gabbros) much more Mg-rich compared with MORB. It can be expected
that high-pressure equivalents of troctolites and olivine-gabbros may develop
mineral assemblages that are quite different from normal eclogite facies gar-
net+omphacite rocks. In fact, Mg-rich olivine gabbro metamorphosed at
25 kbar and 650 oc may not contain garnet+ omphacite at all as a peak as-
semblage. The low XFe of the protolith prevents the formation of garnet. An
example of LT eclogite metamorphism of an olivine gabbro from the Alps
will be presented in Section 9.8.3.1. Such mafic rocks were metamorphosed
under eclogite facies conditions; however, they are not eclogites. We also re-
commend avoiding confusing rock names like pelitic eclogites or pelitic
blueschists. The proper expressions for such kind of rocks are metapelites in
blueschist facies or pelitic gneisses in eclogite facies (see Chap. 7.7).
Anorthosites, also basic rocks, cannot be converted into eclogites by high-
pressure metamorphism because the rocks contain more than 90-95% pla-
gioclase. Plagioclase of the composition An 70 will be converted into a jadeite
+ zoisite + kyanite +quartz assemblage in the LT- and MT eclogite facies.
9.8.3.1
Blueschist and Eclogite Facies Metamorphism of an Olivine Gabbro:
a Case History from the Central Alps
The 2-km gabbro lens forms the Allalinhorn between Zermatt and Saas and
is referred to as the Allalin gabbro in the literature (Bearth 1967; Chinner
and Dixon 1973; Meyer 1983a,b). The gabbro belongs to a typical ophiolite
complex that once formed the oceanic lithosphere of the Mesozoic Tethys
ocean. The ophiolites make up a part of the South Pennine nappes of the
Alps. During early Alpine orogeny, the ophiolites were subducted to great
depth and metamorphosed under LT eclogite facies conditions. Pillow basalts
were turned first into ordinary LT eclogites, later retrogressed to blueschists,
and finally were overprinted by the Tertiary metamorphism that has reached
the greenschist facies in the Zermatt area. Abundant eclogite garnet repre-
sents the only relic mineral from the eclogite stage in locally strongly retro-
gressed eclogites. The garnet-bearing greenschists of the Zermatt-Saas ophio-
lites demonstrate the very refractory nature of garnet.
The Allalin gabbro is an Mg-rich olivine gabbro with irregular layers of
troctolite. The succession of minerals that formed in the gabbro through its
entire P-T history is given in Fig. 9.11. The data shown in Fig. 9.11 are based
on Meyer (1983a). The P-T diagram in the upper left corner of Fig. 9.11
shows the path followed by the gabbro from the magmatic crystallization, via
the LT eclogite facies stage to the present-day erosion surface. The meta-
morphic history of the gabbro can be subdivided into various stages that will
be briefly explained. (A) Crystallization of the igneous assemblage. In uncle-
formed domains within the gabbro lens, where water never had access to the
rock, igneous minerals are often well preserved. (B) Onset of subduction of
the partly cooled gabbro (or of the completely cooled gabbro that has been
partly altered by oceanic metamorphism: Barnicoat 1995; Barnicoat and Cart-
wright 1997). Associated with stage B is the formation of typical coronites.
They form because olivine+ plagioclase becomes an incompatible assemblage
[Chap. 5, reaction (15)]. The up to 4-cm-large olivine megacrysts react with
matrix plagioclase. The reaction products (orthopyroxene and garnet) form
concentric monomineralic shells around the primary olivine with garnet
forming the outer shell toward the matrix. This remains the only garnet that
ever formed in the LT eclogite facies gabbro. The Opx-Grt coronas were re-
peatedly modified afterwards by the progress of metamorphism. The most
characteristic mineral of this stage is jadeite. It forms from albite component.
It is found in matrix plagioclase only as extremely small grains along frac-
tures and cracks that provided access for water to the igneous plagioclase
grains. Note that there is no quartz in the rock at any stage. Quartz that is
produced from albite breakdown [reaction (70)] dissolved in the aqueous
metamorphic fluid. It was transported to the reaction sites in the nearby
corona structures where it was used up by olivine-consuming reactions. (C)
During the first stage of the eclogite facies, the characteristic assemblage of
LT eclogites form; Cr-omphacite pseudomorphs augite, the assemblage jade-
ite+ zoisite + kyanite replaces plagioclase; kyanite +talc+ chloritoid pseudo-
high-pressure cycle greenschist facies cycle

(eoalp•ne, cretaceous) (mesoalpine, tertiary)
Fig. 9.11. Minerals in the eclogite facies Allalin gabbro, Swiss Alps (Meyer 1983a)
morph olivine and the coronas. The coarse-grained igneous gabbro structure
is often perfectly preserved. Note that Mg-rich chlorite is present in the ec-
logite-stage assemblage. The 650 oc and 25 kbar LT eclogite facies metagabbro
contains a number of hydrate minerals (Chi, Cld, Tic, Zo) and the high-pres-
sure mineral assemblage owes its very existence to the access of water to the
gabbro during the subduction process. (D) Access of water is made possible
by deformation Dl (brittle fractures healed with Mg-chloritoid and euhedral
omphacite!). Deformation Dl induces, locally, a complete recrystallization of

the rock but usually leaves the magmatic gabbro structure intact (microfrac-
tures and cracks). (E) The high-pressure stage 2 is characterized by the for-
mation of pure glaucophane, Cr-paragonite and phengite. The assemblage is
characteristic of the blueschist facies. Note, however, that the blueschist facies
here represents a retrograde overprint on an LT eclogite facies assemblage.
Its extent depends, again, on the availability of water and it is irregularly de-
veloped. (F) A transitional stage is associated with the return to shallower
levels of the lithosphere that is characterized by fine-grained reaction prod-
ucts of local hydration such as margarite, preiswerkite and the Na-Ca-am-
phibole barroisite; however, spinel, corundum and diaspore are also formed
at this stage. The barroisitic amphibole links the blueschist facies glauco-
phane to the greenschist facies actinolite. Barroisite is typical and widespread
in the retro-eclogites of the Zermatt area. The F-stage assemblages are often
found in symplectites that pseudomorph and replace eclogite stage minerals.
Note also that plagioclase (oligoclase) returned to the rock. (G, Hand I) The
Tertiary greenschist facies overprint is accompanied by two distinct phases
of deformation. The impact of these stages on the Allalin gabbro varies from
place to place and ranges from virtually nothing to complete retrogression.
The greenschist facies deformation and hydration is often localized along
shear zones and deformation is ductile in contrast to the eclogite-stage defor-
mation that has been brittle. The typical assemblage of the greenschist stage
is albite+ zoisite (epidote)+ actinolite+ chlorite+ biotite. The assemblage is
characteristic for conditions around 4-5 kbar and 400-450 a c. (K, L) The
minerals of stages of the subgreenschist facies are mainly found in veins (al-
bite veins, calcite, prehnite in veins).
The Allalin gabbro also contains sulfides, oxides and other accessory
minerals that formed and were stable at various stages of the P-T history of
the rock. This example of a complex reaction history and successive develop-
ment of minerals and mineral assemblages in the Allalin gabbro shows that:
• Deformation is essential in metamorphic processes. It allows water to en-
ter dry igneous rocks. In the absence of water, igneous and metamorphic
assemblages will survive under even extreme P-T conditions (e. g., igneous
Pl +Aug+ Ol at 25 kbar and 650 °C).
• Mafic rocks that undergo eclogite facies metamorphism do not necessarily
become eclogites. The bulk composition of many gabbros may not be fa-
vorable for the formation of garnet.
• LT eclogite facies mafic rocks that form by subduction of oceanic litho-
sphere are characterized by the presence of a wide variety of hydrous
minerals at peak-pressure conditions of the eclogite facies (Zo, Chl, Tlc,
Cld, Phe).
• LT eclogite facies rocks may pass through the blueschist facies on their re-
turn path to the surface. The blueschist facies overprint is then retrograde
in nature and adds another selection of hydrous minerals to the high-pres-
sure low-temperature rock (Pg, Gln).
• All stages (A through L) of the complex P-T path followed by the Allalin
gabbro and the reaction history involving a large number of minerals can
be preserved in a single rock body of not more than about 2 km in extent.

The predominant assemblage that is found in a single piece of rock or
within a small area of a given outcrop depends on how much water en-
tered the rock at a given P- T stage.
• The availability of water is, in turn, related to the local extent of deforma-
tion. Given a certain stage, if much water is available, the formation of the
maximum hydrated assemblage for that stage is possible. Water is con-
sumed by hydration reactions that tend to dry out the rock. Water must
be continuously supplied in order to keep the reactions running that pro-
duce the maximum hydrated assemblage at that stage. The access for the
water to the originally dry rock is provided by micro-fractures and other
brittle deformation during high-pressure metamorphism and by pervasive
ductile deformation during greenschist facies metamorphism.
References
Abbott RN Jr (1982) A petrogenetic grid for medium and high grade metabasites. Am
Mineral 67:865-876
Ahn JH, Peacor DR, Essene EJ (1985) Coexisting paragonite-phengite in blueschist eclogite:
a TEM study. Am Mineral 70:1193-1204
Austrheim H, Griffin WL (1985) Shear deformation and eclogite formation within granulite
- facies anorthosites of the Bergen Arcs, western Norway. Chern Geol 50:267-281
Banno S (1986) The high pressure metamorphic belts in Japan: a review. Geol Soc Am Mem
164:365-374
Barnicoat AC (1995) High pressure metamorphism and sea-floor alteration: a control on
H2 0-release patterns during subduction. Boch Geol Geotechn Arb 44:17-18
Barnicoat AC, Cartwright I (1997) The gabbro-eclogite transformation: an oxygen isotope
and petrographic study of west Alpine ophiolites. J Metamorph Geol15:93-104
Battey MH, McRitchie WD (1975) The petrology of the pyroxene-granulite facies rocks of
Jotunheimen, Norway. Norsk Geol Tidsskrift 55:1-49
Bearth P (1963) Chloritoid und Paragonit aus der Ophiolith-Zone von Zermatt-Saas Fee.
Schweiz Mineral Petrogr Mitt 43:269-286
Bearth P (1967) Die Ophiolithe der Zone von Zermatt-Saas Fee. Beitr Geol Karte Schweiz
NF 132:130 pp
Begin NJ (1992) Contrasting mineral isograd sequences in metabasites of the Cape Smith
Belt, northern Quebec, Canada: three new bathograds for mafic rocks. J Metamorph Geol
10:685-704
Bevins RE, Robinson D (1993) Parageneses of Ordovician sub-greenschist to greenschist fa-
cies metabasites from Wales, UK. Eur J Mineral 5:925-935
Bohlen SB, Liotta JJ (1986) A barometer for garnet amphibolites and garnet granulites. J
Petrol 27:1025-1034
Boles JR, Coombs DS (1975) Mineral reactions in zeolitic Triassic tuff, Hokonui Hills, New
Zealand. Geol Soc Am Bull 86:163-173
Brothers RN, Yokoyama K (1982) Comparison of high pressure schists belts of New Caldo-
nia and Sanbagawa, Japan. Contrib Mineral Petrol 79:219-229
Brown EH (1974) Comparison of the mineralogy and phase relations of blueschists from
the North Cascades, Washington, and greenschists from Otago, New Zealand. Geol Soc
Am Bull 85:333-344
Brown EH (1977) The crossite content of Ca-amphibole as a guide to pressure of meta-
morphism. J Petrol18:53-72
Carmichael RS (1989) Practical handbook of physical properties of rocks and minerals.
CRC Press, Boca Raton, 834 pp
Chinner GA, Dixon JE (1973) Some high-pressure parageneses of the Allalin gabbro, Valais,
Switzerland. J Petrol 14:185-202
Coleman RG, Lee DE, Beatty LB, Brannock WW (1965) Eclogites and eclogites: their differ-
ences and similarities. Geol Soc Am Bull 76:483-508
References 325
Coombs DS, Ellis AJ, Fyfe WS, Taylor AM (1959) The zeolite facies, with comments on the
interpretation of hydrothermal syntheses. Geochim Cosmochim Acta 17:53-107
Cooper AF (1972) Progressive metamorphism of metabasic rocks from the Haast Schist
group of southern New Zealand. J Petrol 13:457-492
Cox RA, Indares A (1999) Transformation of Fe-Ti gabbro to coronite, eclogite and amphi-
bolite in the Baie du Nord segment, Maicouagan Imbricate Zone, eastern Grenville Pro-
vince. J Metamorph Geol 17:537-555
Crawford ML (1966) Composition of plagioclase and associated minerals in some schists
from Vermont, USA, and South Westland, New Zealand, with inferences about the peri-
sterite solvus. Contrib Mineral Petrol 13:269-294
Donato MM (1989) Metamorphism of an ophiolitic tectonic melange, northern California
Klamath Mountains, USA. J Metamorph Geol 7:515-528
El-Shazly AK (1994) Petrology of lawsonite-, pumpellyite- and sodic amphibole-bearing
metabasites from north-east Oman. J Metamorph Geol 12:23-48
Ernst WG (1972) Occurence and mineralogical evolution of blueschist belts with time. Am J
Sci 272:657-668
Eskola P (1921) On the eclogites of Norway. Skr Vidensk Selsk Christiania, Mat-Nat Kl I
8:1-118
Evans BW (1990) Phase relations of epidote-blueschists. Lithos 25:3-23
Evans BW, Brown EH (1986) Blueschists and eclogites. Geological Society of America Mem-
oir, The Geological Society of America, Boulder, Colorado, 423 pp
Ferry JM (1984) Phase composition as a measure of reaction progress and an experimental
model for the high-temperature metamorphism of mafic igneous rocks. Am Mineral
69:677-691
Fett A, Brey GP (1995) Partitioning of AI, Cr and Fe between garnet, clinopyroxene and ru-
tile in natural eclogites and high pressure experiments. Eur J Mineral 7:64
Forbes RB, Evans BW, Thurston SP (1984) Regional progressive high-pressure metamor-
phism, Seward peninsula, Alaska. J Metamorph Geol 2:43-54
Frey M, De Capitani C, Liou JG (1991) A new petrogenetic grid for low-grade metabasites. J
Metamorph Geol 9:497-509
Frimmel HE, Hartnady CJH (1992) Blue amphiboles and their significance for the history of
the Pan-African Gariep belt, Namibia. J Metamorph Geol 10:651-670
Gibson GM (1979) Margarite in kyanite- and corundum-bearing anorthosite, amphibolite,
and hornblendite from central Fiordland, New Zealand. Contrib Mineral Petrol 68:171-
179
Griffin WL, Austrheim H, Brastad K, Bryhni I, Krill A, Mork MBE, Qvale H, Torudbakken
B (1983) High-pressure metamorphism in the Scandinavian Caledonides. In: Gee DG,
Sturt BA (eds) The caledonide orogen - Scandinavia and related areas. Wiley, Chichester
Harte B, Graham CM (1975) The graphical analysis of greenschist to amphibolite facies
mineral assemblages in metabasites. J Petrol 16:347-370
Helms TS, McSween HY Jr, Labotka TC, Jarosewich E (1987) Petrology of a Georgia Blue
ridge amphibolite unit with hornblende+ gedrite+ kyanite +staurolite. Am Mineral
72:1086-1096
Hirajima T, Hiroi Y, Ohta Y (1984) Lawsonite and pumpellyite from the Vestgotabreen For-
mation in Spitsbergen. Norsk Geol Tidsskrift 4:267-273
Hirajima T, Banno S, Hiroi Y, Ohta Y (1988) Phase petrology of eclogites and related rocks
from the Motalafjella high-pressure metamorphic complex in Spitsbergen (Arctic Ocean)
and its significance. Lithos 22:75-97
Holland TJB (1979 a) Experimental determination of the reaction paragonite= jadeite+
kyanite + H 2 0, and internally consistent thermodynamic data for part of the system
Na2 0-Al 2 0 3 -Si0rH 2 0, with applications to eclogites and blueschists. Contrib Mineral
Petrol 68:293-301
Holland TJB (1979b) High water activities in the generation of high pressure kyanite eclo-
gites of the Tauern Window, Austria. J Geol 87:1-27
Holland TJB, Richardson SW (1979) Amphibole zonation in metabasites as a guide to the
evolution of metamorphic conditions. Contrib Mineral Petrol 70:143-148
Hosotani H, Banno S (1986) Amphibole composition as an indicator of subtle grade varia-
tion in epidote-glaucophane schists. J Metamorph Geol 4:23-36
Humphris SE, Thompson G (1978) Hydrothermal alteration of oceanic basalts by seawater.
Geochim Cosmochim Acta 42:107-125
James HL (1955) Zones of regional metamorphism in the Precambrian of northern Michi-
gan. Geol Soc Am Bull 66:1455-1488
Johnson CD, Carlson WD (1990) The origin of olivine-plagioclase coronas in metagabbros

from the Adirondack Mountains, New York. J Metamorph Geol 8:697-717
Kawachi DE (1975) Pumpellyite-actinolite and contiguous facies metamorphism in the
Upper Wakatipu district, southern New Zealand. N Z J Geol Geophys 17:169-208
Klein U, Schumacher JC, Czank M (1996) Mutual exsolution in hornblende and cumming-
tonit: compositions, lamellar orientations, and exsolution temperatures. Am Mineral
81:928-939
Konzett J, Hoinkes G (1996) Paragonite- hornblende assemblages and their petrological sig-
nificance: an example from the Austroalpine Schneeberg Complex, Southern Tyrol, Italy.
J Metamorph Geol14:85-101
Leake BE (1964) The chemical distinction between ortho- and para-amphibolite. J Petrol
5:238-254
Liogys VA, Jenkins DM (2000) Hornblende geothermometry of amphibolite layers of the
Popple Hill gneiss, north-west Adirondack Lowlands, New York, USA. J Metamorph Geol
18:513-530
Liou JG, Maruyama S, Cho M (1987) Very low-grade metamorphism of volcanic and volca-
niclastic rocks - mineral assemblages and mineral facies. In: Frey M (ed) Low tempera-
ture metamorphism. Blackie, Glasgow, pp 59-113
Lopez S, Castro A (2001) Determination of the fluid-absent solidus and supersolidus phase
relationships of MORB-derived amphibolites in the range of 4-14 kbar. Am Mineral
86:1396-1403
Lucchetti G, Cabella R, Cortesogno L (1990) Pumpellyites and coexisting minerals in differ-
ent low-grade metamorphic facies of Liguria, Italy. J Metamorph Geol 8:539-550
Markl G, Bucher K (1997) Eclogites from the early Proterozoic Lofoten Island, N Norway.
Lithos 42:15-35
Messiga B, Scambelluri M, Piccardo GB (1995) Chloritoid-bearing assemblages in mafic sys-
tems and eclogite-facies hydration of alpine Mg-Al metagabbros (Erro-Tobbio Unit, Li-
gurian Western Alps). Eur J Mineral 7:1149-1167
Messiga B, Kienast JR, Rebay M, Piccardi P, Tribuzio R (1999) Cr-rich magnesiochloritoid
eclogites from the Monviso ophiolites (Western Alps, Italy). J Metamorph Geol 17:287-
299
Meyer J (1983a) Mineralogie und Petrologie des Allalingabbros. Doctoral dissertation. Doc-
toral dissertation. Univ Basel, 329 pp
Meyer J (1983 b) The development of the high-pressure metamorphism in the Allalin meta-
gabbro (Switzerland). Terra Cognita 3:187
Meyre C, Puschnig AR (1993) High-pressure metamorphism and deformation at Trescol-
men, Adula nappe, Central Alps. Schweiz Mineral Petrograph Mitt 73:277-283
Meyre C, De Capitani C, Partsch JH (1997) A ternary solid solution model for omphacite
and its application to geothermobarometry of eclogites from the Middle Adula nappe
(Central Alps, Switzerland). J Metamorph Geol 15:687-700
Miyashiro A (1979) Metamorphism and metamorphic belts. Allen and Unwin, London,
492 pp
Mongkoltip P, Ashworth JR (1986) Amphibolitization of metagabbros in the Scottish High-
lands. J Metamorph Geol 4:261-283
Neuhoff PS, Fridriksson T, Arn6rsson S, Bird DK (1999) Porosity evolution and mineral
paragenesis during low-grade metamorphism of basaltic lavas at Teigarhorn, eastern Ice-
land. Am J Sci 299:467-501
Newton RC (1986) Metamorphic temperatures and pressures of group B and C eclogites.
Geol Soc Am Memoir 164:17-30
O'Brien PJ (1993) Partially retrograded eclogites of the Miinchberg Massif, Germany: re-
cords of a multistage Variscan uplift history in the Bohemian Massif. J Metamorph Geol
11:241-260
Okay AI (1994) Sapphirine and Ti-clinohumite in ultra-high-pressure garnet-pyroxenite and
eclogite from Dabie Shan, China. Contrib Mineral Petrol 116:145-155
Patrick BE, Day HW (1989) Controls on the first appearance of jadeitic pyroxene, northern
Diablo Range, California. J Metamorph Geol 7:629-639
Pattison DR (1991) Infiltration-driven dehydration and anatexis in granulite facies metagab-
bro, Grenville Province, Ontario, Canada. J Metamorph Geol 9:315-332
Platt JP (1993) Exhumation of high-pressure rocks: a review of concepts and processes. Ter-
ra Nova 5:119-133
Pognante U, Kienast J-R (1987) Blueschist and eclogite transformations in Fe-Ti gabbros: a
case from the western Alps ophiolites. J Petrol28:271-292
References 327
Powell WG, Carmichael DM, Hodgson CJ (1993) Thermobarometry in a sub-greenschist to

greenschist transition in metabasites of the Abitibi greenstone belt, Superior Province,
Canada. J Metamorph Geol 11:165-178
Russ-Nabelek C (1989) Isochemical contact metamorphism of mafic schist, Laramie An-
orthosite Complex, Wyoming: amphibole compositions and reactions. Am Mineral
74:530-548
Sanders LS (1988) Plagioclase breakdown and regeneration reactions in Grenville kyanite
eclogite at Glenelg, NW Scotland. Contrib Mineral Petrol 98:33-39
Seki Y, Oki Y, Matsuda T, Mikami K, Okumura K (1969) Metamorphism in the Tanzawa
Mountains, central Japan. J Jpn Assoc Mineral Petrol Econ Geol 61:1-29, 50-75
Selverstone J, Spear FS (1985) Metamorphic P-T paths from pelitic schists and greenstones
from the south-west Tauern Window, eastern Alps. J Metamorph Geol 3:439-465
Smelik EA, Veblen DR (1989) A five-amphibole assemblage from blueschists in northern
Vermont. Am Mineral 74:960-964
Smith DC, Lappin MA (1989) Coesite in the Straumen kyanite-eclogite pod, Norway. Terra
Nova 1:47-56
Soto Jl (1993) PTMAFIC: software for thermobarometry and activity calculations with mafic
and ultramafic assemblages. Am Mineral 78:840-844
Spear FS (1980) NaSi«CaAl exchange equilibrium between plagioclase and amphibole. An
empirical model. Contrib Mineral Petrol 72:33-41
Springer RK, Day HW, Beiersdorfer RE (1992) Prehnite-pumpellyite to greenschist facies
transition, Smartville Complex, near Auburn, California. J Metamorph Geol10:147-170
Starkey RJ, Frost BR (1990) Low-grade metamorphism of the Karmutsen Volcanics, Van-
cover Island, British Columbia. J Petrol 31:167-195
Terabayshi M (1993) Compositional evolution in Ca-amphibole in the Karmutsen metaba-
sites, Vancouver Island, British Columbia, Canada. J Metamorph Geol11:677-690
Thieblemont D, Triboulet C, Godard G (1988) Mineralogy, petrology and P-T-t path of Ca-
Na amphibole assemblages, Saint-Martin des Noyers formation, Vendee, France. J Meta-
morph Geol 6:697-716
Thompson AB (1971) Pc02 in low grade metamorphism: zeolite, carbonate, clay mineral,
prehnite relations in the system CaO-Al2 0rC0 2 -H 2 0. Contrib Mineral Petrol 33:145-161
Thompson JB (1991) Modal space: applications to ultramafic and mafic rocks. Can Mineral
29:615-632
Triboulet C, Thieblemont D, Audren C (1992) The (Na-Ca) amphibole-albite-chlorite-epi-
dote-quartz geothermobarometer in the system S-A-F-M-C-N-H2 0. 2. Applications to
metabasic rocks in different metamorphic settings. J Metamorph Geol 10:557-566
Wenk E, Keller F (1969) Isograde in Amphibolitserien der Zentralalpen. Schweiz Mineral
Petrograph Mitt 49:157-198
Wolke C, Truckenbrodt J, Johannes W (1995) Beginning of dehydration-melting in amphi-
bolites at P =10 kbar. Eur J Mineral 7:273
Zen E-A (1961) The zeolite facies: an interpretation. Am J Sci 259:401-409

Metamorphism of Mafic Rocks

Uploaded by

Copyright:

Available Formats

Metamorphism of Mafic Rocks

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Metamorphism of Mafic Rocks

Uploaded by

Copyright:

Available Formats

CHAPTER 9

Metamorphism of Mafic Rocks

K. Bucher et al., Petrogenesis of Metamorphic Rocks

cess commonly leads to the development of zoned metasomatic structures.

The characteristic composition of alkali-olivine basalt is listed in Table 2.1

Table 9.1. Minerals and compositions in metabasaltic rocks

Nesosilicates Sheet silicates

accessory minerals including ilmenite, magnetite, spinel, garnet, and even

The composition space that must be considered in an analysis of phase rela-

to Phi and Gin (from Ms+Pg)

End-member celadonite, glaucophane and phlogopite (biotite) project onto

Metamorphic facies Greenschist

Mineral zoning Chlorite Biotite Gamet Staurolite Sillimanite

Actinolite •••••• green and

Muscovite •••••••• ••••••••• ••••••••• •••••••••

Fig. 9.2. Progressive mineral changes in northern Michigan. (James 1955)

hornblende color information in Fig. 9.2 relates to the changing mineral

A comprehensive representation of the effects of metamorphism on mafic

10 ' ' c ' - -.. - amphibolite-

+ excess quartz + water

ACF chemographies #1 and #2 (Fig. 9.3) represent mineral assemblages typi-

Mineral zoning zone I zone II zone Ill zone IV zoneV

Montm. -verm. ~Ill

......... ......... .........

sheet silicate. Sericite is present sporadically. Several zeolite- and chlorite-

appear as well. In low-pressure terrains such as the Tanzawa Mountains,

Quantitative phase relationships in the simple CASH system in low-grade

100 200 300 400 500

150 200 250 300 350 400

NCMASH system (Fig. 9.6)

Subgreenschist to greenschist facies transition (Fig. 9.6)

Reactions involving carbonates (examples)

Example of reaction producing tschermak component

Table 9.2 (continued)

Carbonate-involving reactions (examples)

Greenschist-amphibolite facies transition

Amphibolite-granulite facies transition

Blueschist facies (Fig. 9.9)

Blueschist-eclogite facies transition (Figs. 9.9 and 9.10)

The transition to the greenschist facies is marked by the first appearance of

The difference between chemography #3 and #4 (Fig. 9.3) is the presence of

200 300 400 500 600

biotite under the microscope. Stilpnomelane is replaced by green biotite at

The greenschist facies assemblage experiences two basic changes as meta-

The combined transformation results in the replacement of the Ab +Act pair

tions, (2) incorporation of edenite component produced by albite-consuming

The amphibolite facies is characterized by chemographies #6, #7 and #8

In prograde orogenic metamorphism of metabasalts, the rocks contain horn-

At low metamorphic grades, mafic rocks that were metamorphosed along a

chlorite or biotite) can be used to monitor pressure conditions in low-grade

Granulite facies metamorphism of mafic rocks is represented by chemogra-

PI + Cpx + Opx ± Grt

semblages. Also, hercynitic spinel is commonly present in mafic granulites

Blueschists are rocks that contain a significant amount of blue alkali-amphi-

14 kbar. If subduction is fast, prominent blueschist facies terrains may devel-

200 300 400 500 600 700 800

tion of metabasaltic rocks during prograde metamorphism. Let us follow a

The association is characteristic of abnormally hot geotherms in conjunction

Reactions that connect the blueschist facies to the low-temperature eclogite