Singular Equilibria in Paragonite Blueschists, Amphibolites and Eclogites
J. F. MOLINA1*AND S.POLI2
1DEPARTAMENTO DE MINERALOGA Y PETROLOGA, CAMPUS DE FUENTENUEVA S/N, UNIV. GRANADA, 18002 GRANADA, SPAIN2DIPARTIMENTO SCIENZE DELLA TERRA, UNIVERSITÀ DI MILANO, VIA BOTTICELLI 23, 20133 MILANO, ITALY
RECEIVED JULY 20, 1997; REVISED TYPESCRIPT ACCEPTED FEBRUARY 2, 1998
Chemographic constraints from high-pressure mafic rocks of the Nevado-Filábride Complex (Betic-Rif mountain belts, SE Spain) and from worldwide high-pressure terrains have been used to unravel phase relationships between amphibole, garnet, clinopyroxene, and albite in the system Na2O-CaO-FeO-MgO-Al2O3-SiO2-H2O (NCFMASH) at quartz-, paragonite-, clinozoisite- and water-saturated conditions. Extensive analysis of element partitioning between amphibole and clinopyroxene suggests that the Fe-Mg partition coefficient is low at low-temperature and high-pressure conditions (`high-temperature' blueschists-low-temperature eclogites) and a progressive increase is achieved by an increase in temperature, or by a decrease in pressure, or by a decrease in the sodium content of the clinopyroxene. Mass-balance analysis of paragonite eclogites and amphibolites suggests that at least two singular reactions are possible in NCFMASH: clinopyroxene + paragonite + quartz = amphibole + albite + clinozoisite + water (garnet-absent), and amphibole + clinozoisite = garnet + clinopyroxene + paragonite + quartz + water (albite-absent). Albite-absent singular equilibrium is relevant for the blueschist-eclogite transition, whereas garnet-absent equilibrium may control the eclogite to amphibolite transition. A complex arrangement of divariant reactions, singular reactions and conjugate non-degenerate reactions arises in topologies at the blueschist-eclogite-amphibolite transitions. Schreinemakers' analysis in P-T diagrams, and mixed-variable T-XM[prime]-XC[prime] and P-XM[prime]-XC[prime] diagrams suggest that complex pressure-temperature paths are not mandatory if a recurrent appearance and disappearance of similar phase assemblages is observed in paragonite-bearing mafic rocks. The P-T location of the singular curve net is discussed in the context of Nevado-Filábride Complex mafic rocks.
The relevance of singular equilibria has been largely recognized in chemical systems of interest for the Earth Sciences. After the early experiments of Konovalov on azeotropic or congruent transformations, Ricci, (1951) exhaustively discussed the fundamentals of singularities. Two types of singularities were recognized: the first, called type m and most relevant, is associated with azeotropic points, and the second, called type k, is related to the intersection of solvi with compositional loops, as occurs in the alkali feldspar system at low pressures (Morse, 1970). Examples and discussion of both type m and type k singular points in igneous systems have been given by
Morse, (1980, e.g. see chapters 16 and 18). The widespread occurrence of singularities in igneous systems is obviously related to the extremely large compositional range of silicate liquids. The well-known concept of the `thermal barrier' or `thermal divide' implies the occurrence of singular equilibria, and their importance in the petrogenesis of granites [e.g. modelled in the binary system kalsilite-silica (Lindsley, 1966)] or basalts [e.g. in the binary system forsterite-silica or in the ternary system forsterite-silica-diopside or forsterite-silica-anorthite (Presnall et al., 1979)] has been widely discussed.
In contrast, the significance and implications of singular point nets at subsolidus conditions have been largely undervalued. Although
Korzhinskii, (1963) analysed azeotropic transformations in subsolidus open systems, only recently have applications to metamorphic assemblages and theoretical developments been discussed (Connolly & Trommsdorff, 1991; Abart et al., 1992).
A number of lines of evidence, however, suggest the occurrence of singularities in a wide range of metamorphic rocks. In metapelites, Ballèvre et al., (1989) demonstrated that the reversal of the Fe-Mg distribution coefficient between garnet and staurolite in kyanite-bearing assemblages implies degeneracies in an AFM compositional diagram and that, consequently, singular curves are generated. Labotka, (1987) found that Fe-Mg reversals between amphibole and biotite coexisting with plagioclase are responsible for the appearance of garnet-hornblende assemblages in K-bearing mafic schists. Degeneracy along such a so-called `garnet + hornblende isograd' is again necessarily related to a singular point net topology in a P-T diagram.
Actually, in mafic systems the large compositional variability of both whole rocks [Mg metagabbros, metabasalts, Fe-Ti metagabbros; see
Mottana et al., (1990)] and mineral phases (amphibole, clinopyroxene, garnet) render singular equilibria very likely. Unfortunately, the application of singular point net analysis to mafic rocks involving such complex solid solutions as amphibole has been hindered by the overwhelming number of possible topologies (see also
Zingg, 1995).
Despite the fact that amphibolite-eclogite and blueschist-eclogite facies transformations in mafic rocks are commonly regarded as the result of relatively simple divariant reactions (Ridley, 1984), the possible occurrence of singular equilibria greatly complicates the phase relationships. In this case, phase diagrams which neglect singular equilibria are incomplete and may lead to the incorrect interpretation of natural phase assemblages. Using natural constraints recorded both in the mafic schists of the Nevado-Filábride Complex (SE Spain) and in similar case studies published in the recent literature, this paper intends to show that singular point nets may explain the development of complex sequences of mineral assemblages at blueschist-amphibolite-eclogite transition zones.
To demonstrate how the occurrence of singular point nets can be retrieved from chemographic constraints we illustrate in Fig. 1 the basic principles concerning the generation of type m singularities (Ricci, 1951). Phase relationships between four solutions in a binary system are shown as a function of a compositional variable X and two potentials Y(1) and Y(2), e.g. pressure and temperature or chemical potentials for open systems. X(A) and X(B) represent the amount of X in the coexisting phases A and B for a bulk composition (dashed plane) selected in the region of the compositional space where an azeotrope develops. The first Y(2)-X section shows a peritectic reaction where the assemblage B + C forms from A with increasing Y(2). In the second section the congruent transformation A -> B appears as a result of the relative displacement of A, B and C compositions as a function of Y(1). As the singularity is formed (open circle) a point of extremum (a maximum or a minimum) in both potentials is generated, following the generalized Konovalov rule as defined by
Hillert, (1985). The locus of extrema in a Y(1)-Y(2) diagram is a singular curve. Figure 1 also shows schematically differences between singular points, invariant points, and indifferent crossings.
Figure1. Schematic diagram of phase relationships in a binary system involving four solutions, A, B, C, and D. X is a compositional variable, whereas Y(1) and Y(2) are independent potential variables, e.g. pressure, temperature or chemical potentials for open systems. The distribution of X between phase A and B is represented in an X(A) vs X(B) diagram for a bulk composition represented by the dashed plane in Y(2) vs X sections. A and B compositions considered at each Y(1)-Y(2) condition are connected by a dotted line in Y(2) vs X sections. Kd, i.e. distribution coefficient of X between A and B, equals one when the azeotrope intersects the dashed plane representing the bulk composition selected. Y(1) vs Y(2) section bears a type m singularity, a singular univariant reaction, an invariant point, and a so-called `indifferent crossing'.
It is expected to be a common behaviour that the location of the azeotropic point in the composition space moves along the Y(1)-Y(2) curve. As a result, the coexisting phases A and B in a fixed bulk composition show a variable composition as a function of progressive variation in azeotropic composition. In the case of a suitable bulk composition (dashed plane in Fig. 1), a reversal in the distribution of X between A and B can be observed, and the condition Kd = 1 represents a `picture' of the extremum in the X(A) vs X(B) plane at fixed Y(1) and Y(2). However, in nature it is very unlikely that we will be able to recover such records of the azeotrope. As an example, Ballèvre et al., (1989) were able to demonstrate a reversal in KdFe/Mg between garnet and staurolite in kyanite-bearing metapelites, after the analysis of a number of assemblages from seven different terrains, but none is a direct observation of the azeotropic mixture. Nevertheless, information from the Kd reversal is sufficient to infer the existence of singular equilibria. In a simple binary system, reversal of element partitioning clearly means that the condition Kd = 1 occurs. However, in more complex systems, `reversals' are not easily represented and attainment of azeotropes should be demonstrated using a matrix form [equations (3)-(5) of
Abart et al., (1992), and
Labotka, (1987)].
To test the existence of singularities in mafic systems, we investigated the Fe2+-Mg partition coefficients between the ferromagnesian mineral phases commonly found in mafic systems (i.e. garnet, amphibole and clinopyroxene).
All Fe2+-Mg partition coefficient data from natural and synthetic mafic assemblages suggest that garnet has the highest affinity for Fe2+ for a large range of P-T conditions (Ellis & Green, 1979; Graham & Powell, 1984). According to the data of
Ellis & Green, (1979) and
Graham & Powell, (1984), both KdFe/MgGar/Cpx and KdFe/MgGar/AmphKdGar/AmphFe/Mg, where KdA/Bi/jKdi/jA/B= [i/j]A/[i/j]B (see Appendix A for abbreviations), decrease with temperature. Whereas the compositional dependence of KdFe/MgGar/Cpx has been largely discussed [i.e. Kd decreases with decreasing almandine and grossular components in garnet (Pattison & Newton, 1989) and increasing jadeite content in clinopyroxene (Koons, 1984)], there is still a poor knowledge of element partitioning between amphibole and the other ferromagnesian silicates of blueschist, amphibolite and eclogite facies rocks. The relatively large uncertainty in the systematics of both KdFe/MgGar/AmphKdGar/AmphFe/Mg and KdFe/MgAmph/CpxKdAmph/CpxFe/Mg variation with mineral composition, pressure and temperature is related to the variable Fe3+/Fetot ratio, which is seldom determined directly. The Fe/Mg distribution coefficients of 100 amphibole-clinopyroxene pairs from 28 studies on world-wide metamorphic terrains and three experimental studies were calculated (Fig. 2a-c) following the normalization methods for amphibole suggested by
Robinson et al., (1982), Trzcienski et al., (1984) (see their fig. 4), and
Schumacher, (1997). Because, generally, more than one method verifies the stoichiometric constraints (e.g. Robinson et al., 1982), for each analysis we give the range of acceptable KdFe/MgAmph/CpxKdAmph/CpxFe/Mg values [see also J. C. Schumacher, (1991, , 1997), for further discussion]. As a general rule, minimum ferrous iron contents are given by the recalculation schemes, which assume that all cations except Ca, Na and K occupy 13 sites or that all cations except Ca, Na, K and Mn occupy 13 sites. Maximum ferrous iron contents are obtained with all Fe as Fe2+. Some glaucophanic amphiboles can be normalized only assuming that all cations except K occupy 15 sites. However, it should be noted that the variations displayed in Fig. 2 are extreme inasmuch as all Fe as Fe2+ represents only a chemical limit but most amphiboles are expected to contain some Fe3+ [see fig. 7 of
Hawthorne, (1983), and R. Schumacher, (1991)].
Figure2. Dependence of the natural logarithm of the ratio [Fe2+/Mg]Amph/[Fe2+/Mg]Cpx (ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg]) on pressure, temperature and clinopyroxene composition. (a) Temperature dependence for amphibole-clinopyroxene pairs from low-pressure eclogites and amphibolites in the pressure range from 5 to 12 kbar; (b) amphibole-clinopyroxene pairs from eclogites equilibrated at pressures from 14 to 25 kbar; (c) inverse relationship between ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] and sodium content in clinopyroxene. Data sources: HT-eclogites and granulites-
Krogh, (1980, 1982), Heinrich, (1986), Hirajima et al., (1988), Krogh et al., (1990), Poli, (1991), O'Brien, (1993), Okay, (1995), Janák et al., (1996), Clarke et al., (1997), Sisson et al., (1997), and
Kohn & Spear, (1990) and references therein; LT-eclogites-
Okrusch et al., (1978), Krogh, (1980), Maresch & Abraham, (1981), Gil-Ibarguchi & Ortega-Gironés, (1985), Koons, (1986), Schliestedt, (1986), Oh et al., (1991), Zhou et al., (1993), Krogh et al., (1994), Okay & Kelly, (1994), Bröcker & Day, (1995); M&H72, Makanjuola & Howie, (1972); M&others92, Messiga et al., (1992); B67, Binns, (1967); Sch93, Schmidt, (1993); P93, Poli, (1993); S&J96, Skjerlie & Johnston, (1996). Error bars stand for the maximum range of variation caused by different recalculation schemes adopted for amphibole. Symbols are placed at minimum Fe2+, mostly obtained assuming that all cations except Ca, Na and K occupy 13 sites or that all cations except Ca, Na, K and Mn occupy 13 sites. Maximum Fe2+ are obtained with all Fe as Fe2+.
Because commonly there is a large uncertainty in the pressure estimates for mafic assemblages, the variations of ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] with temperature are represented for two relatively large pressure ranges, i.e. 5-12 kbar (Fig. 2a) and 14-25 kbar (Fig. 2b). Values of ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] for low-temperature and intermediate-pressure conditions approach unity. Intermediate-pressure terrains (Fig. 2a) show an increase of ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] with temperature. For high-pressure metamorphic regimes, the correlationln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg]-T is less evident. Actually, Fig. 2c demonstrates that there is an overall dependence ofln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] on the Na content in clinopyroxene. Iron preferably partitions into amphibole when clinopyroxene is poor in sodium. Experimental data of
Schmidt, (1993) on two-clinopyroxene assemblages (filled triangles) confirm that amphibole + Na-rich clinopyroxene pairs show values of ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] lower than those of amphibole + omphacite pairs. This behaviour can be attributed to the preference of the M2 site for Fe2+ in jadeitic clinopyroxenes, such that Fe2+/(Fe2+ + Mg) increases as Ca/(Ca + Na) decreases (Koons, 1984, fig. 3; Koons, 1986). At moderate Na contents (0·4-0·6 a.p.f.u.) ln[KdFe/MgAmph/CpxKdAmph/CpxFe/Mg] seems to vary inversely with temperature. Amphibole-clinopyroxene pairs with the lowest KdFe/MgAmph/CpxKdAmph/CpxFe/Mg are from jadeite-glaucophane-bearing assemblages of Sifnos (Okrusch et al., 1978), Sesia Zone (Koons, 1986) and NW Turkey (Okay & Kelly, 1994), and from eclogites of the Cordillera de la Costa Belt (Sisson et al., 1997), Dabie Shan (Okay, 1995) and Adula Nappe (Heinrich, 1986). In summary, from Fig. 2a-c it can be concluded that values of KdFe/MgAmph/CpxKdAmph/CpxFe/Mg are low at low-temperature and high-pressure conditions (`high-temperature' blueschists-low-temperature eclogites) and a progressive increase occurs by an increase in temperature, or by a decrease in pressure, or by a decrease in jadeite content of clinopyroxene.
Such systematic variations in Fe-Mg partitioning between amphibole, clinopyroxene and garnet are not only of interest for geothermobarometric applications, but also suggest the possible occurrence of degeneracies. However, as most phases belong to the seven-component model system Na2O-CaO-FeO-MgO-Al2O3-SiO2-H2O (NCFMASH), the attainment of KdFe/MgAmph/CpxKdAmph/CpxFe/Mg= 1 is not a unique condition for degeneracies. As mafic schists metamorphosed at blueschist, amphibolite and eclogite facies conditions are commonly saturated in quartz, epidote and paragonite (Guidotti, 1987), we will show in the next section that chemographic relationships can be analysed in a condensed ternary model system.
Assuming water-saturated conditions, the chemographic relations among assemblages coexisting with paragonite + epidote + quartz can be examined on a CaO'-FeO'-MgO' (C'-F'-M') compositional diagram (paragonite-CFM chemography, Fig. 3) after a component transformation following the matrix given in Appendix B. Once the compositions of the saturated phases epidote, mica and plagioclase are fixed to pure clinozoisite, paragonite and albite, respectively, grossular (Gro), pyrope (Py) and almandine (Alm) plot at the three corners of the triangle, and grossular, jadeite and albite coincide. Compositional fields of amphibole, clinopyroxene and garnet solid solutions from paragonite-epidote mafic assemblages are shown in Fig. 3. Mafic bulk compositions ranging from Mg gabbros to Fe-Ti gabbros mostly plot within the field of sodic-calcic amphiboles, in agreement with the large amounts of amphibole which commonly occur in mafic rocks. Amphiboles from assemblages formed at the blueschist-eclogite transition plot close to the Mg glaucophane-Fe glaucophane (Gl-FGl) join, whereas those from assemblages approaching the eclogite-amphibolite transition show a larger compositional variability and higher AlIV and NaA occupancies. As a consequence, most amphiboles from intermediate- to high-pressure terrains plot within a (Gl-FGl)-Mg pargasite-Fe pargasite (Parg-FParg) quadrilateral. Clinopyroxenes from relatively low-temperature eclogites show minor Tschermak substitution and plot near the omphacite-Fe-omphacite (Omph-FOmph) join. In the pressure-temperature region considered, garnets are almandine rich and the grossular proportion commonly does not exceed ~30% (Fig. 3). Consequently, reactions which occur in Ca-rich compositions (rodingites and calc-schists) will not be considered here.
Figure3. Paragonite-CFM projection constructed following the procedure described in Appendix B. Pale grey shaded fields are the maximum compositions expected for garnets, amphiboles and clinopyroxenes in blueschists, amphibolites and eclogites. Dashed and dotted fields represent the compositional variation of garnets, amphiboles and clinopyroxenes considered in the database used in Fig. 2. (See Appendix A for symbols.)
In the Nevado-Filábride Complex, the high-pressure metamorphosed mafic rocks appear in two different metamorphic units. Detailed descriptions and thermobarometric estimates have been given by
Gómez-Pugnaire, (1979), Gómez-Pugnaire & Fernández-Soler, (1987), Puga et al., (1989), Jabaloy et al., (1993), Gómez-Pugnaire et al., (1994), Cámara, (1995), Molina, (1995), López Sánchez-Vizcaíno et al., (1997), and Molina et al. (in preparation). In the lower unit, coronitic and foliated eclogites and omphacitites (garnet-absent mafic rocks) display the paragonite-bearing assemblages Gar + Bar + Omph + Pg + Ep + Q + Ru (Gar-Bar-Omph assemblage) and Bar + Omph + Pg + Ep + Q + Ru (Bar-Omph assemblage), respectively. These assemblages are overprinted by the amphibolite facies assemblage pargasite + albite ± epidote. Barroisite is overgrown by pargasitic amphibole, which coexists with albite in symplectites on clinopyroxene sites, and coronas of epidote + pargasite develop on garnet. In the upper unit, mafic schists show the paragonite assemblages Gar + Bar + Pg + Ep + Q + Ru (Gar-Bar assemblage) and Gar + Bar + Ab + Pg + Ep + Q + Ru + Sph (Gar-Bar-Ab assemblage). Barroisite is overgrown by magnesio-hornblende. Albite appears either as small inclusion-free crystals parallel to the foliation or as large porphyroblasts with abundant inclusions of amphibole, epidote, rutile and sphene oriented parallel to the matrix foliation, suggesting an important post-kinematic albite recrystallization. The bulk compositions of rocks showing Gar-Bar-Ab and Gar-Bar-Omph assemblages are very similar (Table 1), although those which display eclogite assemblages sensu stricto have slightly higher Mg/(Mg + Fe2+) ratios.
Table1. Chemical analyses and CIPW norm of bulk compositions of mafic schists from the Nevada-Filábride Complex.
Microprobe analyses of mineral phases characteristic of Gar-Bar-Omph and Gar-Bar-Ab assemblages are given in Table 2a-d. Mineral analyses were performed using a CAMECA SX-50 by wavelength dispersive (WDS) electron microprobe operating at 15 kV accelerating voltage and 20 nA sample current (University of Granada) using ZAP-PAP correction methods, and an ARL-SEMQ (CNR, Milan) using a ZAF (MAGIC IV) matrix correction. Both natural and synthetic standards were used. Amphibole compositions from Gar-Bar-Omph assemblages (assuming that all cations except Ca, Na, K, and Mn occupy 13 sites) are barroisites and magnesio-katophorites (Leake, 1978) and those from Gar-Bar-Ab assemblages are barroisites. Calcic amphiboles in overgrowths are mainly pargasites and hastingsites in the former and magnesio-hornblende in the latter. Once tremolite, tschermakite, pargasite and glaucophane are chosen as linearly independent components of the amphibole solid solution, the compositional variations of amphibole can be described in AlVI + Fe3+ + 2Ti vs NaM4 and AlVI + Fe3+ + 2Ti - NaM4 vs NaA plots (Fig. 4a and b). In paragonite-albite-bearing assemblages the compositional change toward magnesio-hornblende compositions is produced by a decrease both in NaM4 and in NaA coupled to the increase in AlVI + Fe3+ + 2Ti - NaM4. Pargasite-rich compositions appear in eclogitic assemblages as a result of a decrease in NaM4 and an increase in NaA following the compositional trend of pargasite substitution (Fig. 4b) (see Table 2a). Garnets are invariably rich in almandine and show slightly higher pyrope contents in eclogite-garnet (Table 2b, Fig. 5). In both assemblages the pyrope content increases and grossular content decreases toward the rim. Clinopyroxenes are omphacites with Fe3+ contents ranging from 0·14 a.p.f.u. in the cores to ~0·09 a.p.f.u. in the rims (Table 2c). AlIV is negligible, suggesting low equilibrium temperatures (Carpenter, 1979; Newton, 1986). Mg/(Mg + Fe2+) ratios decrease slightly towards the rim, ranging from 0·77 to 0·72. Epidotes in both assemblages show pistacite contents [Xps = Fe3+/(Fe3+ + Al- 2)] ranging from 0·68 in cores and inclusions in garnets to 0·50 in rims (Table 2d). Albite, sphene and rutile have nearly ideal end-member compositions. Paragonite is also very close to its ideal end-member composition and it contains ~0·1 a.p.f.u. of K in eclogite schists.
Table2a. Representative microprobe analyses of amphiboles (cations based on 23 oxygens).
Table2b. Representative microprobe analyses of garnets (cations based on 12 oxygens).
Table2c. Representative microprobe analyses of clinopyroxenes (cations based on six oxygens).
Table2d. Representative microprobe analyses of epidote (cations based on 12.5 oxygens).
Figure4. Amphibole compositional trends in the Nevado-Filábride Complex. [wdbullf], amphibole analyses from garnet-barroisite-albite assemblages; [circle], from eclogitic assemblages, garnet-barroisite-omphacite.
Figure5. Ternary diagram for garnet. All analyses plot in the group C eclogite field of
Coleman et al., (1965). Symbols as in Fig. 4.
Metamorphic conditions for Gar-Bar-Omph and Gar-Bar-Ab assemblages from the Nevado-Filábride Complex were estimated by multiequilibrium calculations (Molina et al., in preparation) using thermodynamic data from the database of
Holland & Powell, (1990, datafile created 1994), the equation of state for water of
Holland & Powell, (1991), and solution models from
Berman, (1990) for garnet, from
Holland, (1990) for clinopyroxene, and an ideal model for epidote. Assuming, as a first approximation, water-saturated conditions, garnet + clinopyroxene + amphibole + paragonite + epidote equilibrated at 527 ± 57°C and 20 ± 5 kbar, whereas the amphibolite facies assemblage, garnet + albite + paragonite + amphibole + epidote, equilibrated at 565°C ± 94°C and 11 kbar ± 4 kbar.
Bulk compositions from paragonite-bearing foliated eclogites and albite-paragonite amphibolites (A and B stars in Fig. 6b) plot in the overlapping Gar-Amph-Cpx and Gar-Amph-Ab phase fields. These chemographic relations suggest three possible reactions:
Gar1 + Cpx = Amph2 + Ab
(1)
Gar1 + Cpx = Gar2 + Amph2
(2)
Amph1 + Gar1 + Cpx = Amph2
(3)
(where 1 and 2 stand for minerals in paragonite-bearing eclogites and albite-paragonite amphibolites, respectively), but only the crossing tie line reaction (1) unequivocally relates the two assemblages considered. Reactions (2) and (3) simply result from the compositional displacement of garnet, clinopyroxene and amphibole because of the change in the metamorphic conditions, but, as shown in fig. 2 of
Gordon et al., (1991), a correct mass-balance should only include reaction (1) as the actual `equilibrium' responsible for the topological difference between the two assemblages.
Figure6. Phase relations of paragonite-bearing assemblages at variable KdFe/MgAmph/CpxKdAmph/CpxFe/Mg. (a) Kd < 1, experiments of
Poli, (1993): [utrif], experiments at 550°C and 14 kbar; [circle], experiment at 650°C and 22 kbar. (b) Kd ~1, Nevado-Filábride Complex: [circle], eclogitic assemblages (1); [wdbullf], albite-bearing mafic assemblages (2). A, bulk composition of paragonite-bearing eclogites in the Nevado-Filábride Complex; B, bulk composition of albite-paragonite-bearing amphibolites. (c) Kd >1, Margarita Island (Maresch & Abraham, 1981): [squf], albite-bearing mafic assemblages; [squ], eclogitic assemblages. (d) Kd 1, Hubei Province (
Zhou et al., 1993): T, eclogites (sample HB-6.5); M, eclogites from Spitsbergen (Hirajima et al., 1988) (sample 24-23).
The compositional and environmental conditions attained by the Nevado-Filábride mafic rocks are clearly not unique. Similarly, in Margarita Island (Maresch, 1971; Maresch & Abraham, 1981) a continuous transition from Gar-Bar-Ab to Gar-Bar-Omph assemblage was recognized in the metamorphic zones of La Rinconada Group. Such zones were suggested to represent the prograde transition from a non-eclogite to an eclogite-bearing terrain according to
Maresch, (1977). Metamorphic conditions for mafic assemblages in Margarita Island range from 450 to 525°C, at pressures from 11·5 to 13·5 kbar (Maresch & Abraham, 1981).
Fe2+/Mg partitioning in the Nevado-Filábride Complex and in Margarita Island are similar (Fig. 6b and c), and a KdFe/MgAmph/CpxKdAmph/CpxFe/Mg >1 confirms the relevance of reaction (1). None the less, in the Nevado-Filábride Complex the Fe/Mg ratios of amphibole and clinopyroxene are very similar, KdFe/MgAmph/CpxKdAmph/CpxFe/Mg approaches unity, and therefore the garnet coefficient in reaction (1) is expected to be very low. To perform a more rigorous analysis avoiding the problem of dimensional reduction associated with graphical methods, the algebraic technique of singular value decomposition (SVD, Fisher, 1989) was employed to analyse the phase relations among the mafic assemblages. Si, Ti and H were not included as components in the calculations inasmuch as they can be balanced by quartz, rutile and water, respectively. Eclogite and albite-paragonite-bearing assemblages were found to intersect as we observed in the projected three-component space (Fig. 6), therefore suggesting that such a change in mineral assemblage should not be ascribed to differences in bulk composition but most probably to a reaction mechanism. Reactions which involve albite and clinopyroxene (see also Table 3) are in agreement with reaction (1) and as suggested in paragonite-CFM chemography, coefficients for garnet are either zero or very low. The conclusion is that the eclogite-amphibolite transition in the Nevado-Filábride mafic schists is the result of clinopyroxene- and paragonite-consuming reactions that produce amphibole and albite with only very minor involvement of garnet. This agrees with the degenerate, singular reaction
Cpx = Amph + Ab.
[Gar]S
In contrast, Fig. 6a [e.g. experiments reported by
Poli, (1993)] demonstrates that the disappearance of clinopyroxene when KdFe/MgAmph/CpxKdAmph/CpxFe/Mg<1 implies a terminal reaction of the form:
Cpx = Amph + Gar + Ab
(4)
where, opposite to reaction (1), garnet is expected to be a product of the eclogite to amphibolite facies transformation.
Table3. Stoichiometric coefficients for reactions discussed in the text as deduced from mineral compositions of the Nevado-Filábride Complex (Table 2).
The paragonite-bearing assemblages Gar + Gl + Pg + Cz-Zo + Q (garnet-glaucophane assemblage) and Gar + Gl + Omph + Pg + Cz-Zo + Q (garnet-glaucophane-omphacite assemblage) are common in mafic rocks from high-grade blueschist facies or low-temperature eclogite facies (New Caledonia:
Yokoyama et al., 1986; Clarke et al., 1997; Sifnos Island, Greece:
Okrusch et al., 1978; Schliestedt, 1986; Schliestedt & Okrusch, 1988; Sesia-Lanzo Zone, Italy:
Compagnoni, 1977; Lombardo et al., 1977; Reinsch, 1979; Syros Island, Greece:
Ridley, 1984; Hubei Province, Central China:
Zhou et al., 1993; Alanya Nappe, Turkey:
Okay, 1989; Franciscan Complex, California:
Oh & Liou, 1990; Oh et al., 1991) and have been also reported in intermediate-temperature eclogite facies (Spitsbergen:
Hirajima et al., 1988). These high-grade blueschist assemblages show a continuous transition towards `eclogite' (garnet + omphacite) assemblages. Reinsch, (1979) and
Ridley, (1984) suggested that the continuous divariant reaction
Gl + Ep = Gar + Omph + Pg + Q + H2O
(5)
is responsible for the blue-amphibole consumption and the generation of garnet + clinopyroxene eclogitic assemblage. However, it should be noted that reaction (5) is divariant in NCFMASH only if Fe and Mg are independent components of the system. Chemographic analysis of some metamorphic terrains (Fig. 6d) shows that collinearity between sodic or sodic-calcic amphibole, garnet and clinopyroxene is approached (Spitsbergen:
Hirajima et al., 1988), or even attained (Hubei Province:
Zhou et al., 1993, sample HB-6.5) and consequently reaction (5) degenerates to the singular equilibrium
Amph = Gar + Cpx.
[Ab]S
In conclusion, chemographic analysis indicates that the stoichiometries of reactions involving amphibole and clinopyroxene change continuously as a function of pressure, temperature and bulk composition, and may reverse both along the eclogite to amphibolite facies and blueschist to eclogite facies transitions.
Phase relationships in paragonite-bearing rocks were discussed by
Maresch & Abraham, (1977, , 1981), Reinsch, (1979), Ridley, (1984), Yokoyama et al., (1986), Hirajima et al., (1988), Okay, (1989) and
Evans, (1990). Yokoyama et al., (1986) established a multi-bundle topology in the NCFMASH system with three invariant points. This topology shows a chlorite-absent invariant point in the high-pressure and relatively high-temperature portion of the grid from which two relevant reactions emanate. A chlorite- and glaucophane-absent reaction accounts for the eclogite-amphibolite transition, in agreement with the mass-balance calculations previously discussed, and a chlorite- and albite-absent reaction accounts for the blueschist-eclogite transition in agreement with the results of
Reinsch, (1979), Ridley, (1984) and
Okay, (1989).
Therefore, blueschist-eclogite-amphibolite transitions can be assumed to be produced essentially under chlorite-conservative or chlorite-absent conditions for a wide range of mafic compositions and metamorphic conditions. Below, a topological analysis of this chlorite-absent region is performed, on the basis of the reaction mechanisms previously described.
In this section, only the non-degenerate reactions in the NCFMASH system are considered. The topological consequences of degeneracies will be analysed in a subsequent section. Following the chemographic constraints imposed by the Fe/Mg partitioning among the phases, Schreinemakers' analysis of the relationships between amphibole, garnet, clinopyroxene, paragonite, albite, epidote, quartz and water is shown in Fig. 7a. Each of the subsystems NCFASH and NCMASH bears one invariant point, four univariant reactions (the right side is the high-temperature side):
Amph + Cz = Gar + Cpx + Pg + Q + W [Ab]Fe, [Ab]Mg
Amph + Cz + Pg + Q = Gar + Ab + W [Cpx]Fe, [Cpx]Mg
Cpx + Pg + Q = Ab + Amph + Cz + W [Gar]Fe, [Gar]Mg
and six divariant assemblages. In the NCFMASH system, univariant reactions become divariant and the non-terminal univariant reaction
Cpx + Gar + Pg + Q = Ab + Amph + Cz + W (1[prime])
accounts for the eclogite-amphibolite transition.
Figure7. Fig. 7.(a) Topology of reactions in the NCFMASH system saturated in quartz, epidote, paragonite and water for KdFe/MgAmph/Cpx >1. Dashed fine lines locate the T-X and P-X pseudosections. (b) T-XMg pseudosection at pressure P'. (c) P-XMg pseudosection at temperature T'. (See text for discussion.)
The Clapeyron slope of these reactions is qualitatively evaluated (Table 3) using volumes and entropies of mineral phases from the database of
Holland & Powell, (1990, revised 1994) and the equation of state for water of
Holland & Powell, (1991). Ideal mixing models were used for solid solutions. However, stoichiometry is not unique and depends on the amphibole and clinopyroxene composition assumed. To provide a reasonable evaluation of reaction coefficients (Table 3), we use the compositions of mineral assemblages from the Nevado-Filábride Complex (Table 2). In agreement with previous considerations by
Maresch & Abraham, (1981), Yokoyama et al., (1986) and
Poli, (1993), reaction (1€) has a moderate dP/dT slope (12·2 bar/K) and indicates the dominant role of pressure in the transition from eclogite to amphibolite.
The variation of Fe2+/Mg ratios in mineral phases from the divariant assemblages consistent with the KdFe/MgAmph/CpxKdAmph/CpxFe/Mg>1 are shown in the T-XMg and P-XMg pseudosections of Fig. 7b and c (Thompson, 1976a, 1976b). According to these pseudosections, along the reaction (1€), the Fe2+/Mg ratios of coexisting mineral phases decrease with increasing temperature, as the invariant point in the Fe subsystem (I) lies at lower temperatures than the invariant point for the Mg subsystem (II). The stable extension of reaction (1€) must lie between the stable extensions of the reactions [Gar]Mg and [Amph]Fe, on the basis of Schreinemakers' principles, Fe2+/Mg partition coefficients and the assumed absence of miscibility gaps in amphibole. Reaction (1€) will terminate at invariant points I and II as imposed by reactions defined in the NCFASH and NCMASH subsystems.
This topology explains, as a first approximation, the reaction mechanisms which control the stability of high-pressure mafic assemblages. All divariant reactions are responsible for an increase of paragonite abundance with increasing pressure. In contrast, the temperature effect on paragonite stability can be the opposite. If the reaction [Gar], [Cpx] or [Amph] occurs, then an increase in temperature will cause a decrease in paragonite abundance, whereas reaction [Ab] will cause an increase in paragonite. This behaviour is simply related to the negative slope of reaction [Ab]. A further consequence of the arrangement shown in Fig. 7 is a generalized increase of XMg in the coexisting phases with temperature increase (Fig. 7b), but a differential response to pressure. XMg of garnet, amphibole and clinopyroxene will decrease for a pressure increase if the reaction [Gar], [Cpx] or [Amph] operates (Fig. 7c), but will increase in the case of reaction [Ab]. The lowermost loop in Fig. 7b relating amphibole, garnet and clinopyroxene basically corresponds to fig. 5 of
Ridley, (1984).
As a consequence, paragonite disappearance in Gar-Amph-Cpx eclogites at high pressure and high temperature (albite-absent) has to be determined by a discontinuous reaction, e.g. a kyanite-bearing reaction, inasmuch as the [Ab] divariant reaction cannot account for paragonite breakdown with increasing pressure or temperature. In contrast, within the albite stability field, paragonite may disappear through a set of continuous reactions.
The paragonite-CFM chemography from Fig. 8 shows that two singular reactions are possible in the NCFMASH system when Fe-Mg partitioning data from high-pressure terrains and phase relations previously discussed are considered (Fig. 6):
Cpx + Pg + Q = Amph + Ab + Cz + W
[Gar]s
and
Amph + Cz = Gar + Cpx + Pg + Q + W.
[Ab]s
Purely on a topological basis, there are a very large number of possible arrangements. Consequently, we will discuss here only two examples which are feasible from a thermodynamic point of view and which may explain phase relations at the blueschist-eclogite-amphibolite transitions.
Figure8. Possible collinearities consistent with Fe-Mg partitioning data and with phase relalationships. Degenerate reaction [Gar]s is possible at the limit KdFe/MgAmph/CpxKdAmph/CpxFe/Mg = 1.
The singular univariant equilibrium [Gar]s at the eclogite-amphibolite transition
As previously shown, KdFe/MgAmph/CpxKdAmph/CpxFe/Mgdecreases with increasing pressure, which suggests the possible attainment of KdFe/MgAmph/CpxKdAmph/CpxFe/Mg = 1, in a range of pressure and temperature conditions comparable with the metamorphism of mafic rocks from the Nevado-Filábride Complex. The resulting collinearity of clinopyroxene, amphibole and albite in the paragonite-CFM projection leads to the appearance of the univariant degenerate reaction [Gar]s in the grid of the system NCFMASH. This necessarily implies a variation of Kd along the reaction (1€). When inversion of Kd occurs, the singular point appears, the univariant singular curve [Gar]s departs from it, and garnet moves on the high-temperature, low-pressure side of reaction (1€). As a consequence, the conjugate equilibrium (Abart et al., 1992)
Cpx + Pg + Q = Gar + Amph + Ab + Cz + W (1")
connects the singular point with the invariant point I in NCFASH (Fig. 9a), at conditions where KdFe/MgAmph/CpxKdAmph/CpxFe/Mg <1. Because of the observed positive dependence of KdFe/MgAmph/CpxKdAmph/CpxFe/Mg with temperature (Fig. 6), we will consider only the topology which has reaction (1) on the low-temperature side of the singular point. Following the rules discussed by
Connolly & Trommsdorff, (1991) and Abart et al., (1992), the reaction [Gar]s has to lie on the high-pressure side of reaction (1), and, according to the arrangement of reactions around invariant point I, it will generate two indifferent crossings with the reactions [Amph]Fe and [Ab]Fe (Fig. 9a).
Figure9. Schematic Schreinemakers' projection of one of the possible arrangements of the singular reaction [Gar]s and of the reactions (1€) and (1") in the system NCFMASH saturated in quartz, epidote, paragonite and water. Dashed fine lines locate the three-dimensional phase diagram P-XC€-XM€.(b) shows schematic three-dimensional phase diagram P-XC'-XM€ of the Schreinemakers' projection at temperature T'. (See text for discussion.)
Along the singular reaction, an extremum point must arise in T-X-X and P-X-X sections [see fig. 15 in
Hillert, (1985), and
Connolly & Trommsdorff, 1991]. This is represented for the topology of Fig. 9a in the hypothetical three-dimensional phase diagram P-XC'-XM' of Fig. 9b, where Xi = i/(C€ + F€ + M€). The two three-phase triangles Cpx-Amph-Ab degenerate into a straight line at the point of maximum P3. In nature, a bulk-rock chemistry that shows a singular reaction such as [Gar]s is unlikely. Bulk compositions along the join Ab-Cpx-Amph (composition 1 in Fig. 9b) would be required. Nevertheless, a relatively large range of bulk compositions may help us to infer the occurrence of singular curves. Let us consider, as an example, bulk composition 2 in Fig. 9b. At P = P1, amphibole and clinopyroxene coexist, but with increasing pressure the three-phase loop Amph + Cpx + Ab is crossed and albite appears. A further increase in pressure at P = P2 will cause the disappearance of clinopyroxene and occurrence of the assemblage albite + amphibole, which suggest the transformation to amphibolite facies. However, clinopyroxene will appear and albite disappear again as P reaches P3, where the singular reaction occurs and albite will no long coexist with amphibole. Actually, the main feature to be expected in common bulk-rock chemistries is possibly a sequential corona of similar assemblages which should not be ascribed to unusual variations of pressure-temperature conditions, but to loops which are conjugate to singular reaction curves.
The singular univariant equilibrium [Ab]s at the blueschist-eclogite transition
Reinsch, (1979), Ridley, (1984), Yokoyama et al., (1986) and
Okay, (1989) indicated that the divariant reaction [Ab] is responsible for the blueschist-eclogite transition. However, it should be noted from Fig. 8 that the divariant reaction [Ab] in the NCFMASH system may degenerate to the singular univariant equilibrium [Ab]s. As previously shown, glaucophane, clinopyroxene and garnet were found to be almost collinear in eclogites from the Hubei Province (Zhou et al., 1993), and Spitsbergen (Hirajima et al., 1988), suggesting that the degenerate reaction [Ab]s is also possible in nature and may account for the blueschist-eclogite transition.
One of the possible topologies which account for the occurrence of a singular curve [Ab]s at the blueschist-eclogite transition is shown in Fig. 10a. In this topology, the reactions (1€), [Ab]Fe and [Ab]Mg border the P-T region where the singular reaction [Ab]s lies. We chose to let the singular point coincide with invariant point II, to simplify the chemographic arrangement, although in principle it can occur anywhere along [Ab]Mg univariant reaction in NCMASH.
Figure10. Schematic Schreinemakers' projections of one the possible arrangements of the reaction [Ab]s in the system NCFMASH saturated in quartz, epidote, paragonite and water. Dashed fine lines locate the three-dimensional phase diagram T-XC€-XM€.(b) shows schematic three-dimensional phase diagram T-XC€-XM€ at pressure P'. (See text for discussion.)
The three-dimensional phase diagram T-XA€-XM€ of Fig. 10b shows an extremum at T = T2, where three distinct amphibole-garnet-clinopyroxene assemblages form. All of these three-phase assemblages vanish on the [Ab]Mg univariant reaction and amphibole finally breaks down. The displacement of phase compositions toward Mg-rich compositions again may cause the sequential appearance and `breakdown' of one of the phases. If we consider the bulk composition 3 (Fig. 10b), the assemblage Gar + Amph occurs at T = T1. At T = T2 the singular reaction develops two new three-phase assemblages (Amph + Gar + Cpx) but no change will be observed in the bulk chemistry 3. With temperature increase the following sequence occurs: Amph + Gar + Cpx (at T2 < T < T3) -> Gar + Cpx (at T = T3) -> Amph + Gar + Cpx (at T3 < T < T4) -> Gar + Amph (at T = T4) -> Amph + Gar + Cpx (at T4 < T < T5) -> Gar + Cpx. At this last step the blueschist to eclogite transformation is completed. This in no way means that complete transformation to eclogite facies only takes place close to the NCMASH system because the diagram just described is largely schematic and all transformations may well occur in a very different XMg region. However, we want to emphasize that complex sequences of mineral replacements, with several stages of growth of amphibole + garnet ± clinopyroxene ± plagioclase symplectites are common in mafic rocks close to the blueschist-eclogite-amphibolite transitions (e.g. Messiga & Scambelluri, 1991; Messiga et al., 1992). Such complex textural relationships may well be interpreted as evidence of degenerate topologies, and complex P-T trajectories are not mandatory.
The validity of Fig. 10b is not restricted to the stability field of albite, because singularities occur in albite-absent compositions, and at pressures higher than the reaction albite = jadeite + quartz, three-phase fields such as garnet-clinopyroxene-albite (upper left corner in Fig. 10b) simply degenerate to a garnet-clinopyroxene high-variance assemblage. Therefore, although we do not discuss here the arrangement of the topologies presented with albite breakdown, such a framework is substantially valid both for the transition zone and for eclogites sensu stricto.
The interpretation of the transformations from blueschist to eclogite facies and from eclogite to amphibolite facies still relies upon a poor knowledge of the thermodynamic properties of amphibole solid solutions. Although recent experimental studies (Poli, 1993; Jenkins, 1994; Sharma,1996) may help to decipher the mineral chemistry of amphibole, the phenomena previously discussed cannot be predicted from calculations using available thermodynamic data, and therefore cannot be anchored to a P-T diagram.
The main reactions which control the stability of paragonite and epidote at water-saturated conditions are shown in Fig. 11. The P-T field for the assemblage paragonite + zoisite + quartz + water is delimited at low pressure by the reaction forming margarite + plagioclase (Franz & Althaus, 1977), at high temperature by the solidus (Lambert & Wyllie, 1972), and at high pressure by the paragonite breakdown reaction to kyanite, jadeite and quartz (Holland, 1979).
Figure11. Approximate pressure and temperature conditions for the singular equilibria [Gar]s and [Ab]s. Continuous lines are experimentally determined reactions of relevance to define the field of paragonite eclogites and amphibolites: (1) and (2), Holland, (1979); (3)
Franz & Althaus, (1977); (4)
Poli, (1993); (5)
Holland, (1980); basalt wet solidus from
Lambert & Wyllie, (1972). Boxes indicates P-T ranges for paragonite-bearing eclogites and albite-paragonite assemblages from the Nevado-Filábride Complex (Molina et al., in preparation). Facies fields after
Spear, (1993).
Reactions (1), (1[prime]), (1"), and [Gar]s, i.e. reactions active at the eclogite to amphibolite facies transformation, occur within the divariant field of sodic clinopyroxene in equilibrium with albite (Holland, 1980). Reaction [Ab]s is expected at conditions where the appearance of the garnet + clinopyroxene pair commonly suggests the attainment of temperatures not lower than ~400°C. However, a more precise determination of the pressure and temperature location of the singular reactions [Gar]s and [Ab]s is inhibited by the poor knowledge of the non-ideality of amphibole solid solution.
A first-order estimate of the [Delta]S/[Delta]V for singular reactions suggests a low dP/dT slope for the reaction [Gar]s (~11 bar/K, Table 3), as expected for the eclogite-amphibolite transformation, whereas the reaction [Ab]s shows a rather steep negative slope (-57 bar/K), again consistent with the schematic blueschist to eclogite facies transition as represented by
Spear, (1993).
Systematic studies of KdFe/MgAmph/CpxKdAmph/CpxFe/Mg in amphibole-clinopyroxene symplectites may help us to understand whether singular point nets are realistic in mafic systems. As it was previously shown that the eclogite to amphibolite facies transformation in the Nevado-Filábride Complex approaches the reaction [Gar]s (Table 3), it is suggested that singular reactions are not restricted to unusual conditions, but might be the rule rather than the exception.
The quantitative positioning of P-T-XC€-XM€ loops may solve the compositional dependence of the blueschist to eclogite facies transformation. In principle, in Fe-rich systems the facies transition could be produced by the simple divariant reaction suggested by
Ridley, (1984), whereas in Mg-rich systems the transition might well be governed by the loop which is conjugate to the [Ab]s singular reaction (Fig. 10b). Therefore, a different mechanism can be responsible for the blueschist-eclogite transition in Fe-Ti and Mg gabbros.
Singular point nets are expected to involve amphibole not only at high pressure, as previously discussed, but also at high temperature, namely, on the solidus. It has been shown [see the review by
Gilbert et al., (1982)] that a complex arrangement of the incongruent melting reactions of pargasite appears in a system where the component H2O is diluted, e.g. by CO2. Melting experiments in basaltic and andesitic systems showed that the amphibole stability field exhibits a temperature maximum with decreasing H2O activity, at constant total pressure (e.g. Holloway & Burnham, 1969, , 1972; Nesbitt & Hamilton, 1970; Eggler, 1972). The experimentalstudy by
Holloway, (1973) on pargasite melting in thesystem Na2O-CaO-MgO-Al2O3-SiO2-H2O-CO2 demonstrated that at low X(H2O), i.e. at low X(H2O)fluid and, hence, low X(H2O)melt(Burnham & Davis, 1971), the reaction
is necessarily related to X(H2O)melt > X(H2O)amph. As a result, a temperature extremum has been shown on T-X(H2O)fluid sections [see p. 255 of
Gilbert et al., (1982)], which indicates that collinearity between melt, pargasite and the anhydrous assemblage has to occur. Such a temperature extremum results in a reversal in theX(H2O) distribution between the phases and consequently a singular equilibrium is generated. Although topological consequences of such a reversal have still not been explored, it should be noted that the analysis of singular point nets cannot be undervalued in interpreting the petrogenesis of mafic rocks over a wide range of P-T conditions.
The likelihood of singular equilibria in rocks of mafic composition is maximized by the large compositional variation both of whole-rock chemistry and of mineral chemistry of amphibole, garnet, and clinopyroxene. Extreme states should also be favoured by the large non-ideality of all of these solid solutions. Although direct findings of compositional degeneracies are expected to be rare in nature, reversals of element partitioning and the recurrent appearance and disappearance of similar phase assemblages may provide a strong indirect indication that singular point nets are relevant in the petrogenesis of high-pressure rocks.
Phase relationships in amphibolites and eclogites from the Nevado-Filábride Complex and other metamorphic terrains suggest that the singular reaction
J.F.M. is very grateful to F. Bea and M. T. Gómez-Pugnaire, who encouraged him to carry out the post-doctoral project. We are indebted to R. Abart, J. C. Schumacher and F. S. Spear for thoughtful and constructive reviews. I. Guerra, P. G. Montero and J. M. Fernández-Soler are also thanked for their technical assistance. This work has been financially supported by the Spanish Interministry Commission for Science and Technology (CICYT) projects AMB93-0535 and PB95-1266 and by FPI-(PN90) and FPI en el Extranjero grants of the Spanish Ministry of Education and Science. S.P.'s work was supported by MURST-40% grants and CNR-96.04174.CT11/96.00412.CT05 funding.
Abart, R., Connolly, J. A. D. & Trommsdorff, V. (1992). Singular point analysis, construction of Schreinemakers projection for systems with a binary solution. American Journal of Science 292, 778-805.
Ballèvre, M., Pinardon, J. L., Kienast, J. R. & Vuichard, J. P. (1989). Reversal of Fe-Mg partitioning between garnet and staurolite in eclogite-facies metapelites from Champtoceaux Nappe (Brittany, France). Journal of Petrology30, 1321-1349.Berman, R. G. (1990). Mixing properties of Ca-Mg-Fe-Mn garnets. American Mineralogist75, 328-344.Binns, R. A. (1967). Barroisite-bearing eclogite from Naustdal, Sogn og Fjordane, Norway. Journal of Petrology8, 349-371.Bocchio, R., De Capitani, L., Liborio, G., Maresch, W. V. & Mottana, A. (1996). Equilibration conditions of eclogite lenses from Isla Margarita, Venezuela: implications for the tectonic evolution of the metasedimentary Juan Griego Group. Lithos37, 39-59.Bröcker, M. & Day, W. (1995). Low-grade blueschist facies metamorphism of metagreywackes, Franciscan Complex, northern Califonia. Journal of Metamorphic Geology13, 61-78.Burnham, C. W. & Davis, N. F. (1971). The role of H2O in silicate melts. I. P-V-T relations in the system NaAlSi3O8 to 10 kilobars and 1000°C. American Journal of Science270, 54-79.Cámara, F. (1995). Estudio cristalquímico de minerales metamórficos en rocas máficas del Complejo Nevado-Filábride (Cordilleras Béticas). Ph.D. Thesis, Granada University.Carpenter, M. A. (1979). Omphacites from Greece, Turkey, and Guatemala, composition limits of cation ordering. American Mineralogist64, 102-108.Clarke, G. L., Aitchison, J. C. & Cluzel, D. (1997). Eclogites and blueschists of Pam Peninsula, NE New Caledonia: a reappraisal. Journal of Petrology38, 843-876.Coleman, R. G., Lee, D. E., Beatty, L. B. & Brannock, W. W. (1965). Eclogites and eclogites, their differences and similarities. Geological Society of America Bulletin76, 483-508.Compagnoni, R. (1977). The Sesia-Lanzo Zone: high pressure-low temperature metamorphism in the Austroalpine continental margin. Rendiconti della Società Italiana di Mineralogia e Petrografia33, 355-374.Connolly, J. A. D. & Trommsdorff, V. (1991). Petrogenetic grids for metacarbonate rocks, pressure-temperature phase diagrams projection for mixed volatile systems. Contributions to Mineralogy and Petrology108, 93-105.Droop, G. T. R., Lombardo, B. & Pognante, U. (1990). Formation and distribution of eclogites facies rocks in the Alps. In: Carswell, D. A. (ed.) Eclogite Facies Rocks. New York: Blackie, pp. 225-259.Eggler, D. H. (1972). Amphibole stability in H2O-undersaturated calc-alkaline melts. Earth and Planetary Science Letters15, 28-34.Ellis, D. J. & Green, D. H. (1979). An experimental study of the effect of Ca upon garnet-clinopyroxene Fe-Mg exchange equilibria. Contributions to Mineralogy and Petrology71, 13-22.Evans, B. W. (1990). Phase relations of epidote-blueschists. Lithos25, 2-23.Fisher, G. W. (1989). Matrix analysis of metamorphic mineral assemblages and reactions. Contributions to Mineralogy and Petrology102, 69-77.Franz, G. & Althaus, E. (1977). The stability relations of the paragenesis paragonite-zoisite-quartz. Neues Jahrbuch für Mineralogie, Abhandlungen130, 159-167.Gilbert, M. C., Helz, R. T., Popp, R. K. & Spear, F. S. (1982). Experimental studies of amphibole stability. In: Veblen, D. R. & Ribbe, P. H. (eds) Amphiboles: Petrology and Experimental Phase Relations. Mineralogical Society of America, Reviews in Mineralogy9B, 229-353.Gil-Ibarguchi, J. I. & Ortega-Gironés, E. (1985). Petrology, structure and geotectonic implications of glaucophane-bearing eclogites and related rocks from Malpica-Tuy (MT) Unit, Galicia, Northwest Spain. In: Smith, D. C., Franz, G. & Gebauer, D. (eds) Chemistry and Petrology of Eclogites. Chemical Geology50, 145-162.Gómez-Pugnaire, M. T. (1979). La evolución del metamorfismo alpino en el Complejo Nevado-Filábride de la Sierra de Baza (Cordilleras Béticas, España). Ph.D. Thesis, Granada University.Gómez-Pugnaire, M. T. & Fernández-Soler, J. M. (1987). High-pressure metamorphism in metabasites from the Betic Cordilleras (SE Spain) and its evolution during the Alpine orogeny. Contributions to Mineralogy and Petrology95, 231-244.Gómez-Pugnaire, M. T., Franz, G. & López Sánchez-Vizcaino, V. (1994). Retrograde formation NaCl-scapolite in high pressure metaevaporites from the Cordilleras Béticas. Contributions to Mineralogy and Petrology116, 448-461.Gordon, T. M., Ghent, E. D. & Stout, M. Z. (1991). Algebraic analysis of the biotite-sillimanite isograd in the File Lake area, Manitoba. Canadian Mineralogist29, 673-686.Graham, C. M. & Powell, R. (1984). A garnet-hornblende geothermometer, calibration, testing and application to the Pelona Schist, Southern California. Journal of Metamorphic Geology2, 13-31.Guidotti, C. V. (1987). Micas in metamorphic rocks. In: Bailey, S.W. (ed.) Micas. Mineralogical Society of America, Reviews in Mineralogy13, 357-467.Hawthorne, F. C. (1983). The crystal chemistry of the amphiboles. Canadian Mineralogist21, 173-480.Heinrich, C. A. (1986). Eclogite facies regional metamorphism of hydrous mafic rocks in central Alpine Adula Nappe. Journal of Petrology27, 123-154.Heinrich, C. A. & Althaus, E. (1988). Experimental determination of the reactions 4 lawsonite + 1 albite = 1 paragonite + 2 zoisite + 2 quartz + 6 H2O and 4 lawsonite + 1 jadeite = 1 paragonite + 2 zoisite + 1 quartz + 6 H2O + 1 jadeite. Neues Jahrbuch für Mineralogie, Monatshefte11, 516-528.Hillert, M. (1985). Principles of phase diagrams. International Metals Reviews30, 45-67.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. Lithos22, 75-97.Holland, T. J. B. (1979). Experimental determination of the reaction paragonite = jadeite + kyanite + H2O and internally consistent thermodynamic data for part of the system Na2O-Al2O3-SiO2-H2O with application to eclogites and blueschists. Contributions to Mineralogy and Petrology68, 293-301.Holland, T. J. B. (1980). The reaction albite = jadeite + quartz determined experimentally in the range 600-1200°C. American Mineralogist65, 129-134.Holland, T. J. B. (1990). Activities of components in omphacitic solid solutions. An application of Landau theory to mixtures. Contributions to Mineralogy and Petrology105, 446-453.Holland, T. J. B. & Powell, R. (1990). An enlarged and updated internally consistent dataset with uncertainties and correlations, the system K2O-Na2O-CaO-MgO-MnO-FeO-Fe2O3-Al2O3-TiO2-SiO2-C-H2-O2. Journal of Metamorphic Geology8, 89-124.Holland, T. J. B. & Powell, R. (1991). A Compensated-Redlich-Kwong (CORK) equation for volumes and fugacities of CO2 and H2O in the range 1 bar to 50 kbar and 100-1600°C. Contributions to Mineralogy and Petrology109, 265-273.Holloway, J. R. (1973). A model for melting of a hydrous mineral with a mixed-volatile fluid-I. Experimental results to 8 kbar. Geochimica et Cosmochimica Acta37, 651-666.Holloway, J. R. & Burnham, C. W. (1969). Phase relations and compositions in basalt-H2O-CO2 under Ni-NiO buffer at high temperatures and pressures. Geological Society of America, Abstracts and Programs7, 104-105.Holloway, J. R. & Burnham, C. W. (1972). Melting relations of basalt with equilibrium water pressure less than total pressure. Journal of Petrology13, 1-29.Jabaloy, A., Galindo-Zaldívar, J. & González-Lodeiro, F. (1993). The Alpujárride-Nevado-Filábride extensional shear zone, Betic Cordillera, SE Spain. Journal of Structural Geology15, 555-569.Janák, M., O'Brien, P. J., Hurai, V. & Reutel, C. (1996). Metamorphic evolution and fluid composition of garnet-clinopyroxene amphibolites from the Tatra Mountains, Western Carpathians. Lithos39, 57-79.Jenkins, D. M. (1994). Experimental reversal of the aluminum content in tremolitic amphiboles in the system H2O-CaO-MgO-Al2O3-SiO2. American Journal of Science294, 593-620.Kohn, J. M. & Spear, F. S. (1990). Two new geobarometers for garnet amphibolites, with applications to southeastern Vermont. American Mineralogist75, 89-96.Korzhinskii, D. S. (1963). Extreme conditions and their significance in mineral systems. In: Vinogradov, S. P. (ed.) Chemistry of the Earth's Crust (Volume 1). Proceeding of the Geochemical Conference Commemorating the Centenary of Academician V. I. Vernadskii's Birth. Academy of Sciences of the USSR. Institute of Geochemistry and Analytical Chemistry [English translation: Amoils, R. (ed.) (1966). Jerusalem: Israel Program for Scientific Translations, pp. 62-86.]Koons, P. O. (1984). Implications to garnet-clinopyroxene geothermometry of non-ideal solid solution in jadeitic pyroxenes. Contributions to Mineralogy and Petrology88, 340-347.Koons, P. O. (1986). Relative geobarometry from high-pressure rocks of quartzofeldspathic composition from Sesia Zone, Western Alps, Italy. Contributions to Mineralogy and Petrology93, 322-334.Krogh, E. J. (1980). Geochemistry and petrology of glaucophane-bearing eclogites and associated rocks from Sunnfijord, Western Norway. Lithos 13, 355-380.Krogh, E. J. (1982). Metamorphic evolution of Norwegian country-rock eclogites, as deduced from mineral inclusions and compositional zoning in garnets. Lithos15, 305-321.Krogh, E. J., Andresen, A., Bryhni, I., Broks, T. M. & Kristensen, S. E. (1990). Eclogites and polyphase P-T cycling in the Caledonian Uppermost Allochthon in Troms, northern Norway. Journal of Metamorphic Geology8, 289-309.Krogh, E. J., Oh, C. W. & Liou, J. G. (1994). Polyphase and anticlockwise P-T evolution for Franciscan eclogites and blueschists from Jenner, California, USA. Journal of Metamorphic Geology12, 121-134.Labotka, T. C. (1987). The garnet + hornblende isograd in calcic schists from an andalusite-type regional metamorphic terrain, Panamint Mountains, California. Journal of Petrology28, 323-354.Lambert, I. B. & Wyllie, P. J. (1972). Melting of gabbro (quartz eclogite) with excess water to 35 kilobars, with geological applications. Journal of Geology80, 693-708.Leake, B. E. (1978). Nomenclature of amphiboles. Canadian Mineralogist16, 501-520.Lindsley, D. H. (1966). P-T projection for part of the system kalsilite-silica. Carnegie Institution of Washington, Yearbook67, 153-158.Lombardo, B., Compagnoni, R., Fiora, L. & Facchinelli, A. (1977). Composition of some sodic pyroxenes from the eclogitic micaschist of lower Val d'Aosta (Sesia-Lanzo Zone, Western Alps). Rendiconti della Società Italiana di Mineralogia e Petrografia33, 375-387.López Sánchez-Vizcaíno, V., Connolly, J. A. D. & Gómez-Pugnaire, M. T. (1997). Metamorphism and phase relations in carbonate rocks from the Nevado-Filábride Complex (Cordilleras Béticas, Spain), Application of the Spn + Ru + Cal + Qtz + Gr buffer. Contributions to Mineralogy and Petrology126, 292-302.Makanjuola, A. A. & Howie, R. A. (1972). The mineralogy of glaucophane schists and associated rocks from le de Groix, Brittany, France. Contributions to Mineralogy and Petrology35, 83-118.Maresch, W. V. (1971). The metamorphism and structure of northeastern Margarita Island, Venezuela. Ph.D. Thesis, Princeton University.Maresch, W. V. (1977). Similarity of metamorphic gradients in time and space during metamorphism of the La Riconada group, Margarita Island, Venezuela. GUA Papers of Geology9, 110-111.Maresch, W. V. & Abraham, J. (1977). Chemographic analysis of reaction textures in an eclogite from the Island of Margarita (Venezuela). Neues Jahrbuch für Mineralogie, Abhandlungen130, 103-113.Maresch, W. V. & Abraham, J. (1981). Petrography, mineralogy, and metamorphic evolution of an eclogite from the Island of Margarita, Venezuela. Journal of Petrology22, 337-362.Messiga, B. & Scambelluri, M. (1991). Retrograde P-T-t path for the Voltri Massif eclogites (Ligurian Alps, Italy), some tectonic implications. Journal of Metamorphic Geology9, 93-109.Messiga, B., Tribuzio, R. & Caucia, F. (1992). Amphibole evolution in Variscan eclogite-amphibolites from the Savona crystalline massif (Western Ligurian Alps, Italy), controls on the decompressional P-T-t path. Lithos27, 215-230.Molina, J. F. (1995). Evolución petrológica y geoquímica de los esquistos máficos del Complejo Nevado-Filábride, Cordilleras Béticas, España. Ph.D. Thesis, Granada University.Morgan, B. A. (1970). Petrology and mineralogy of eclogite and garnet amphibolite from Puerto Cabello, Venezuela. Journal of Petrology11, 101-145.Morse, S. A. (1970). Alkali feldspars with water at 5 kbar pressure. Journal of Petrology11, 221-251.Morse, S. A. (1980). Basalts and Phase Diagrams. An Introduction to the Quantitative Use of Phase Diagrams in Igneous Petrology. Berlin: Springer-Verlag.Mottana, A., Carswell, D. A., Chopin, C. & Oberhänsli, R. (1990). Eclogite facies mineral parageneses. In: Carswell, D. A. (ed.) Eclogite Facies Rocks. New York: Blackie, pp. 14-52.Nesbitt, R. W. & Hamilton, D. L. (1970). Crystallization of an alkali-olivine basalt under controlled PO2, PH2O conditions. Physics of the Earth and Planetary Interiors3, 309-315.Newton, R. C. (1986). Metamorphic temperatures and pressures of Group B and C eclogites. Geological Society of America, Memoir164, 17-30.O'Brien, P. J. (1993). Partially retrograded eclogites of the Münchberg Massif, Germany: records of a multi-stage Variscan uplift history in the Bohemian Massif. Journal of Metamorphic Geology11, 241-260.Oh, C. W. & Liou, J. G. (1990). Metamorphic evolution of two different eclogites in the Franciscan Complex, California, USA. Lithos25, 41-53.Oh, C. W., Liou, J. G. & Maruyama, S. (1991). Low-temperature eclogites and eclogitic schists in Mn-rich metabasites in Ward Creek, California; Mn and Fe effects on the transition between blueschist and eclogite. Journal of Petrology32, 275-301.Okay, A. I. (1989). An exotic eclogite/blueschist slice in a Barrovian-style metamorphic terrain, Alanya Nappes, southern Turkey. Journal of Petrology30, 107-132.Okay, A. I. (1995). Paragonite eclogites from Dabie Shan, China, re-equilibration during exhumation? Journal of Metamorphic Geology13, 449-460.Okay, A. I. & Kelly, S. P. (1994). Tectonic setting, petrology and geochronology of jadeite + glaucophane and chloritoid + glaucophane schists from north-west Turkey. Journal of Metamorphic Geology12, 455-466.Okrusch, M., Seidel, E. & Davis, E. N. (1978). The assemblage jadeite-quartz in the glaucophane rocks of Sifnos (Cyclades Archipelago, Grece). Neues Jahrbuch für Mineralogie, Abhandlungen132, 284-308.Pattison, D. R. M. & Newton, R. N. (1989). Reversed experimental calibration of the garnet-clinopyroxene Fe-Mg exchange thermometer. Contributions to Mineralogy and Petrology101, 87-103.Pognante, U. (1988). Early alpine eclogitisation in talc/chloritoid bearing Mg-metagabbros and in jadeite-Fe-omphacite bearing metatrondhjemites from the ophiolites of the western Alps. Rendiconti della Società Italiana di Mineralogia e Petrologia43, 687-704.Poli, S. (1991). Reaction spaces and P-T paths, from amphibole eclogite to greenschist facies in the Austroalpine domain (Oetzal Complex). Contributions to Mineralogy and Petrology106, 399-416.Poli, S. (1993). The amphibole-eclogite transformation, an experimental study on basalt. American Journal of Science 293, 1061-1107.Presnall, D. C, Dixon, J. R., O'Donnell, T. H. & Dixon, S. A. (1979). Generation of mid-ocean ridge tholeiites. Journal of Petrology20, 3-36.Puga, E., Diaz de Federico, A., Fediukova, E., Bondi, M. & Morten, L. (1989). Petrology, geochemistry and metamorphic evolution of the ophiolitic eclogites and related rocks from the Sierra Nevada (Betic Cordilleras, Southeastern Spain). Schweizerische Mineralogische und Petrographische Mitteilungen69, 435-455.Reinsch, D. (1979). Glaucophanites and eclogites from the Val Chiusella (Sesia-Lanzo-Zone, Italian Alps). Contributions to Mineralogy and Petrology70, 257-266.Ricci, J. E. (1951). The Phase Rule and Heterogeneous Equilibrium. New York: Dover.Ridley, J. (1984). Evidence of a temperature-dependent `blueschist' to `eclogite' transformation in high-pressure metamorphism of metabasic rock. Journal of Petrology25, 852-870.Robinson, P., Spear, F. S., Schumacher, J. C., Laird, J., Klein, C., Evans, B. W. & Doolan, B. L. (1982). Phase relations of metamorphic amphiboles: natural occurrence and theory. In: Veblen, D. R. & Ribbe, P. H. (eds) Amphiboles: Petrology and Experimental Phase Relations. Mineralogical Society of America, Reviews in Mineralogy9B, 1-228.Schliestedt, M. (1986). Eclogite-blueschist relationships as evidenced by mineral equilibria in the high-pressure metabasic rocks of Sifnos (Cycladic Islands) Greece. Journal of Petrology27, 1437-1459.Schliestedt, M. (1990). Occurrences and stability conditions of low-temperature eclogites. In: Carswell, D. A. (ed.) Eclogite Facies Rocks. New York: Blackie, pp. 160-179.Schliestedt, M. & Okrusch, M. (1988). Meta-acidites and silicic meta-sediments related to eclogites and glaucophanites in Northern Sifnos, Cycladic Archipelago, Greece. In: Smith, D. C. (ed.) Developments in Petrology: Eclogites and Eclogite-facies Rocks. Amsterdam: Elsevier, pp. 291-334.Schmidt, M. W. (1993). Phase relations and compositions in tonalite as a function of pressure: an experimental study at 650°C. American Journal of Science293, 1011-1060.Schumacher, J. C. (1991). Empirical ferric iron corrections: necessity, assumptions, and effects on selected geothermobarometers. Mineralogical Magazine 55, 3-18.Schumacher, J. C. (1997). Appendix 2. The estimation of the proportion of ferric iron in the electron-microprobe analysis of amphiboles. Canadian Mineralogist 35, 238-246.Schumacher, R. (1991). Compositions and phase relations of calcic amphiboles in epidote- and clinopyroxene-bearing rocks of the amphibolite and lower granulite facies, central Massachusetts, USA. Contributions to Mineralogy and Petrology108, 196-211.Sharma, A. (1996). Experimentally derived thermochemical data for pargasite and reinvestigation of its stability with quartz in the system Na2O-CaO-MgO-Al2O3-SiO2-H2O. Contributions to Mineralogy and Petrology125, 263-275.Sisson, V. B., Eartan, I. E. & Avé Lallemant, H. G. (1997). High-pressure (~2000 MPa) kyanite- and glaucophane-bearing pelitic schist and eclogite from Cordillera de la Costa Belt, Venezuela. Journal of Petrology38, 65-83.Skjerlie, K. & Johnston, D. (1996). Vapor-absent melting from 10 to 20 kbar of crustal rocks that contain multiple hydrous phases: implications for anatexis in the deep to very deep continental crust and active continental margins. Journal of Petrology37, 661-691.Spear, F. S. (1993). Metamorphic Phase Equilibria and Pressure-Temperature-Time Paths. Washington, DC: Mineralogical Society of America.Thompson, A. B. (1976a). Mineral reactions in pelitic rocks, I. Prediction of P-T-X(Fe-Mg) phase relations. American Journal of Science 276, 401-424.Thompson, A. B. (1976b). Mineral reactions in pelitic rocks, II. Calculation of some P-T-X(Fe-Mg) phase relations. American Journal of Science 276, 425-454.Trzcienski, W. E., Jr, Carmichael, D. M. & Helmstaedt, H. (1984). Zoned sodic amphibole: petrologic indicator of changing pressure and temperature during tectonism in the Bathurst area, New Brunswick, Canada. Contributions to Mineralogy and Petrology85, 311-320.Yokoyama, K., Brothers, R. N. & Black, P. M. (1986). Regional eclogite facies in the high-pressure metamorphic belt of New Caledonia. Geological Society of America, Memoir164, 407-423.Zhou, G., Liu, J., Eide, E. A., Liou, J. G. & Ernst, W. G. (1993). High-pressure/low-temperature metamorphism in northern Hubei Province, central China. Journal of Metamorphic Geology11, 561-574.Zingg, A. J. (1995). Continuous reactions with amphibole, garnet and plagioclase. Journal of Metamorphic Geology13, 431-443.
