Journal of Petrology Advance Access originally published online on June 3, 2005
Journal of Petrology 2005 46(10):2129-2144; doi:10.1093/petrology/egi050
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
A Quaternary Solution Model for White Micas Based on Natural Coexisting Phengite–Paragonite Pairs
DEPARTMENT OF EARTH SCIENCES, BASEL UNIVERSITY, BERNOULLISTRASSE 32, CH-4056 BASEL, SWITZERLAND
RECEIVED NOVEMBER 25, 2003; ACCEPTED APRIL 12, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
A thermodynamic model for the quaternary white mica solid solution with end-members muscovite–Mg-celadonite–paragonite–Fe-celadonite (Ms–MgCel–Pg–FeCel) is presented. The interaction energies for the MgCel–Pg join, the FeCel–Pg join and the ternary interactions were obtained from natural coexisting phengite–paragonite pairs. Phengite–paragonite pairs were selected based on the criteria that their chemical compositions may be represented as a linear combination of the model end-member compositions and that the respective formation conditions (350–650°C, 4–21 kbar) are accurately known. Previously published excess free energy expressions were used for the Ms–Pg, Ms–MgCel and Ms–FeCel binaries. The suggested mixing model was tested by calculating multicomponent equilibrium phase diagrams. This proved to be particularly well suited to reproduce compositional variations of white micas from amphibolite-facies metapelites.
KEY WORDS: white mica; solution model; equilibrium phase diagrams
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
White mica is a ubiquitous phase in low- to medium-grade metamorphic pelites. In the early work of Thompson (1957)
white mica was treated as an excess phase of fixed (muscovite) composition giving rise to the well-known AFM projection. It was soon recognized (Thompson, 1957; Guidotti, 1973; Thompson & Thompson, 1976) that this treatment was oversimplified in many cases, because it did not account for the compositional variations of white micas and the coexistence of potassium- and sodium-rich white micas in metapelites.
Taking muscovite as an end-member, several chemical substitutions may occur in white mica (see Fig. 1). The compositional join between the muscovite [KAl2(AlSi3O10)(OH)2] and paragonite [NaAl2(AlSi3O10)(OH)2] end-members is represented by the Na = K substitution. Two Tschermak-type substitutions, SiMg = AlVI + AlIV and SiFe2+ = AlVI + AlIV, lead to the theoretical Mg-celadonite [KAlMg(Si4O10)(OH)2] and Fe-celadonite [KAlFe(Si4O10)(OH)2] end-members. White mica with a chemical composition along the muscovite–celadonite join is common in metapelites and is referred to as phengite. If phengite is high in sodium, the celadonite substitution is usually less pronounced (Guidottti & Sassi, 1976).
|
At high temperatures the muscovite–paragonite solvus tends to close, whereas at low temperatures the solvus opens and a potassium- and a sodium-rich white mica may coexist (see also Guidotti, 1984
). The temperature dependence of the K–Na partitioning works well as a qualitative geothermometer, whereas its use as a quantitative geothermometer is not very successful (e.g. Guidotti & Sassi, 1976; Guidotti, 1984). Experiments showed that the solvus on the ideal Ms–Pg join widens with increasing pressure (e.g. Chatterjee & Flux, 1986). Guidotti et al. (1994a) showed that pressure alone cannot account for the opening of the solvus between potassium- and sodium-rich white micas, and that the celadonite component in phengite also has an influence on the Na–K partitioning between coexisting paragonite and phengite.
Since the pioneering work of Guidotti & Sassi (1976) it has been known that the solvus between sodium- and potassium-rich white micas opens with increasing celadonite content. This is documented by the fact that the phengite limb of the solvus approaches successively more potassium-rich compositions with increasing celadonite component (Katagas & Baltazis, 1980; Enami, 1983; Grambling, 1984; Guidotti, 1984). The celadonite content in phengite, in turn, increases with increasing pressure but is also sensitive to mineral assemblage (Guidotti & Sassi, 1998).
Mixing models exist for the binary muscovite–paragonite (Eugster et al., 1972; Chatterjee & Froese, 1975; Chatterjee & Flux, 1986; Roux & Hovis, 1996) and the muscovite–celadonite joins (Massonne & Szpurka, 1997; Coggon & Holland, 2002). A quaternary model, which covers the entire composition space that spans the muscovite, paragonite, Mg-celadonite and Fe-celadonite end-members, has, however, not been calibrated so far.
We suggest a quaternary mixing model that covers a major part of the composition space of phengite and paragonite. Our model is based on existing binary interaction parameters from the literature and additional binary and ternary interaction terms, which were derived from the analysis of coexisting phengite and paragonite from natural rocks. Our model accounts for the combined pressure and bulk-rock composition effect on the K–Na partitioning between coexisting Phe and Pg. Application of our model in phase equilibrium calculations yields reasonable descriptions of phase relations in metapelites and will foster the petrological analysis of white mica-bearing assemblages by means of geothermobarometry and phase diagram calculations. This particularly concerns pelitic schists at high-pressure metamorphic conditions where sodium and potassium are mainly stored within Phe and Pg.
|
INPUT DATA |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
Published solid solution models based on experimental data
The shape of the solvus in the quaternary white mica system strongly depends on the three binary solvi along the Ms–Pg, Pg–MgCel and Pg–FeCel joins. At low-temperature conditions, each of the three binary joins exhibits miscibility gaps. It is found that the Ms–Pg solid solution model of Roux & Hovis (1996)
best fits the compositional data from our samples. This is particularly true for pairs of coexisting Phe–Pg from Al-rich high-grade (
600°C, 7 kbar) pelitic rocks from the Lepontine Alps. The model of Roux & Hovis (1996) predicts a relatively narrow miscibility gap. The miscibility gap of solution models from earlier workers (e.g. Chatterjee & Froese, 1975) is too wide to reproduce natural data (Blencoe et al., 1994, fig. 3; Roux & Hovis, 1996). For the two binaries Ms–MgCel and Ms–FeCel we used the solution model of Massonne & Szpurka (1997).
End-member calculation of natural coexisting paragonite–phengite pairs
We investigated 63 pairs of coexisting Pg–Phe from 56 samples (Table 1). Data were taken from the literature and from our own work. With the exception of references 1, 7, 10, 17, 18, 23 and 25 the data sources listed in Table 1 were also used by Guidotti et al. (1994a, table 1) for analysing the effect of the ferromagnesian components on the paragonite–muscovite solvus. Sample selection was based on the following criteria. (1) The P–T conditions of Pg–Phe equilibration must be known accurately. For the data taken from Guidotti et al. (1994a, table 1) we used their P–T estimates as conditions of Pg–Phe equilibration. (2) The chemical composition of the white micas must be a linear combination of the end-members Ms, Pg, MgCel and FeCel.
|
The mole fractions of the end-members Ms, Pg MgCel and FeCel were calculated by applying the method of least squares to the following set of equations:
 |
standard deviation for each element. In our analyses we only considered those samples that fall within a 3 band around the 1:1 correlation for each element [3 (atoms per formula unit; a.p.f.u.) is 0·066 for Si, 0·025 for Al, 0·056 for Fe, 0·054 for Mg, 0·083 for K, 0·086 for Na]. Based on this criterion 14 Phe–Pg pairs were excluded. Of the remaining 49 Phe–Pg pairs only 40 were used for the calibration of a mixing model. The following nine pairs were only used for comparative purposes and model testing: (1) three pairs from Nagel (2002), for which P–T conditions of equilibration are uncertain; (2) the five Phe–Pg pairs (XFeCel in phengite 
0·01, XMgCel in phengite 0·05) with high Na in phengite (XPg in phengite 0·29–0·37) of Irouschek (1983), which have almost ideal binary composition and thus are used to test the solvus close to the Ms–Pg join only; (3) one pair for which phengite composition gives negative end-member values.
Phengite composition
For phengite, the measured chemical compositions are compared with the compositions recalculated from the end-members in Fig. 1. Figure 2a shows that in our dataset XCel (XMgCel + XFeCel) in Phe is correlated with the Si content. This indicates that the Si content is predominantly controlled by the Tschermak substitution (SiMg/Fe2+ = AlVI + AlIV). The pyrophyllitic substitution [(Na/K)Al = 
Si] (e.g. Bousquet et al., 2002) has only a minor effect on the Si content in Phe. This is supported by the fact that the generally high K + Na content in Phe does not depend on XCel (XMgCel + XFeCel) (Fig. 2b). The recalculated Mg content in Phe shows fairly good correlation with the measured Mg content (Fig. 1c). In contrast, the recalculated Fe content of Phe is systematically lower than the measured Fe content (Fig. 1d). This discrepancy suggests that most of the selected phengites contain ferric iron. Assuming that the deviation between the analysed and recalculated Fe contents reflects the ferric iron content its maximum is 0·06 Fe3+ a.p.f.u., and most analyses of phengite contain less than 0·04 Fe3+ a.p.f.u. (Fig. 2c). The ferric to ferrous iron ratio of most phengites does not exceed 0·6. Nevertheless, fairly high Fe3+ contents are indicated for some samples (Fig. 2d). For two out of the five almost binary samples of Irouschek (1983) the Fe3+ content is up to the total Fe (0·02 and 0·04 a.p.f.u.) (Fig. 2d). In situ measurements (XANES) of the ferric iron content in ultrahigh-pressure eclogites reveal ferric–ferrous ratios in phengite in the range of 0·2–0·6 (Schmid et al., 2003). Hence, there is agreement between the magnitude of measured ferric–ferrous ratios and those suggested from stoichiometric considerations. A good correlation between the measured and the recalculated Al content is observed. This suggests that the Fe3+ content of the selected phengites does not result from the substitution of (Fe3+)VI for AlVI. Possibly the Fe3+ content in the phengites can be explained by the substitution Fe2+ + H = Fe3+ + (see Guidotti, 1984). In addition, Fig. 1c and d indicates that phengite is generally higher in Mg than in Fe. Apparently the substitution SiMg = AlVI + AlIV is preferred to the substitution SiFe2+ = AlVI + AlIV, particularly in phengite with high Si content.
|
Paragonite composition
It has repeatedly been reported (Guidotti, 1984
) that, in contrast to phengite, paragonite generally contains very little or no Mg, Fe2+ or Fe3+. Guidotti et al. (2000) reported that the incorporation of minor amounts of these cations into paragonite does not depend on pressure. Substitution of substantial amounts of magnesium and ferric or ferrous iron would destabilize paragonite as a result of crystallochemical constraints (Guidotti, 1984). In addition, it was shown experimentally that magnesium substitution is very limited in paragonite (Franz & Althaus, 1976). These findings are in line with our observations. In our dataset the average Mg content is 0·015 ( = 0·011) a.p.f.u. and the average Fetot. content is 0·024 ( = 0·016) a.p.f.u. |
CALCULATION PROCEDURE |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
In the quaternary system Ms–MgCel–Pg–FeCel we estimated the binary interaction parameters on the MgCel–Pg and on the FeCel–Pg joins, and all ternary interaction parameters from the available information on compositions and formation conditions of coexisting Pg–Phe pairs (Table 1). The remaining interaction parameters were taken from published solution models (see above).
Chemical potential expression
If the quaternary solutions are described in terms of their end-members, the following expression gives the Gibbs free energy as a function of the respective end-member concentrations, where 
Gex accounts for molecular non-ideal mixing:
 | (1) |
For Gex we used the quaternary expansion of Jackson (1989), which is based on the ternary equation of Wohl (1946, 1953). The following relation gives the quaternary excess function (Gex):
 | (2) |
 | (3) |
) thus W224 and W244 are set to zero. These simplifications yield an excess function:
 | (4) |
; Massonne & Szpurka, 1997). As we assumed symmetric mixing between MgCel–Pg and between FeCel–Pg there remain two unknown parameters in equation (4): W223 (=W233) and W334 (=W344).
The chemical potentials can be calculated from the relation
 | (5) |
 | (6) |
 | (7) |
 | (8) |
 | (9) |
 | (10) |
) indicating that the non-ideality of the Phe–Pg solvus depends primarily on pressure and not on temperature.
When minerals with close to end-member compositions are used, the calculated interaction energies are very sensitive to uncertainties in the determination of end-member contents. In this respect, the celadonite content in paragonite poses a problem, because the mole fractions of XMgCel and XFeCel in paragonite are generally small. To avoid this problem, a constant value for XMgCel and XFeCel in paragonite was assumed for all the phengite–paragonite pairs that were used for the calculation of interaction energies. The mean celadonite content of the 49 paragonites is 1·2 mol %, whereas the average proportion of Fe- and Mg-celadonite is 0·63. On average this yields 0·7 mol % Mg-celadonite and 0·5 mol % for Fe-celadonite contents in paragonite. The major problem of fitting the interaction parameters on the MgCel–Pg join and the FeCel–Pg join is the lack of data on Phe with compositions near both binary joins. However, the calculated interaction energies for the MgCel–Pg join (W223, W233) and the FeCel–Pg join (W334, W344), in particular for coexisting Phe–Pg pairs, for which the celadonite content in phengite is high, cause a miscibility gap at corresponding P–T conditions. Consequently we used binary interaction energies, which are high enough to produce miscibility gaps in order to fit for both W223 (=W233) and W334 (=W344) (open circles in Fig. 3a and b). It should be noted that the binary interaction energies used correspond to Phe–Pg pairs equilibrated at 
7 kbar where the interaction energy is a linear function of pressure. Below 7 kbar the calculated interaction energies scatter over a wide range and most interaction energies are too small to produce miscibility gaps. The determined and used values for the Margules parameters are listed in Table 2.
|
, Interaction energies used to fit the interaction parameters;
|
|
COMPARISON BETWEEN THE MODEL PREDICTIONS AND OBSERVATIONAL DATA |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
To test our model, we recalculated the phengite composition of the 40 pairs of coexisting Phe–Pg that we used to derive interaction energies. In addition, we compared model predictions and measured compositions for the samples of Irouschek (1983)
. The thermodynamic calculations were done with the program THERIAK (De Capitani & Brown, 1987; De Capitani, 1994), which calculates the stable mineral assemblage and the composition of the solid solutions for fixed values of P, T and given bulk composition. In this case the stable mineral assemblage is always composed of Phe and Pg. For the bulk composition we used the composition defined by a 1:1 mixture of the measured compositions of the coexisting phengite and paragonite. The results and 2 of the difference between the reference composition and the calculated composition are given in Fig. 4. The good correlation between measured and recalculated compositions in Fig. 4 shows that the model correctly reflects white mica phase relations. THERIAK is based on a G-minimization algorithm, which computes chemical equilibria in complex systems containing non-ideal solutions. For more details the reader is referred to the homepage of the THERIAK-DOMINO-THERTER software at http://titan.minpet.unibas.ch/minpet/theriak/theruser.html.
|
To further test our model we calculated isothermal and isobaric sections of the ternary Phe–Pg (Ms–MgCel–Pg) miscibility gap and compared these with natural data (Fig. 5). For the sake of simplicity, data points that lie inside the quaternary composition space are projected onto the Ms–MgCel–Pg ternary plane from the FeCel apex. Because this projection deviates from a thermodynamic projection, and to obtain an idea of the geometry of the quaternary solvus, we also projected sectional parts of the quaternary Phe–Pg (Ms–MgCel–Pg–FeCel) miscibility gap onto the Ms–MgCel–Pg ternary plane; the sections correspond to a constant XFeCel in phengite.
|
To illustrate the effect of P and T on the ternary solvus, calculations were carried out for five P–T regimes. The pairs of coexisting phengite–paragonite were divided into five P–T categories (Table 1, Fig. 5). The ternary sections were calculated with the program THERTER (De Capitani, 1994
) using the thermodynamic end-member properties for Ms, MgCel and Pg of Berman (1988, update 1992). Because the solvus varies only slightly within the P–T range of a distinct P–T category the sections were calculated for the lower and upper P–T boundary limiting the categories (Fig. 5). Figure 5 shows that our model is compatible with most of the selected 49 Phe–Pg pairs. The effect of the Cel component on the Phe–Pg solvus is most pronounced at high-T and low-P conditions, and only if XMgCel in Phe is relatively low (XMgCel < 0·2) (Fig. 5b). Towards lower temperatures the above effect decreases (Fig. 5a and b). As pressure increases the solvus widens and consequently the effect of the MgCel component becomes less pronounced (Fig. 5c and d).
|
COMPARING OUR SOLUTION MODEL WITH OTHER MODELS, APPLICATION AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
Comparing our solution model with other models
Because the effect of the MgCel component on the Phe–Pg solvus is most pronounced for low MgCel contents in Phe we compare our model with the Pg–Ms solvus established by Guidotti et al. (1994b)
. Those workers used pairs of natural coexisting Pg–Ms with compositions (Fig. 6a, Guidotti et al., 1994b, table A1) near the ideal Pg–Ms join to establish parametric equations that describe the T–X form of the solvus. The Ms compositions used by Guidotti et al. (1994b) deviate slightly from the end-member composition of muscovite [Fig. 6a, Si 3·0–3·14 a.p.f.u; Fm (Fe2+ + Mg + Fe3+) = 0·02–0·26 a.p.f.u.]. To compare predictions from our model with the model of Guidotti et al. (1994b) we calculated three binary T–X sections through the ternary solvus. These sections were calculated with the program DOMINO (De Capitani & Brown, 1987; De Capitani, 1994) for the reference pressure of 7 kbar given by Guidotti et al. (1994b) for their solvus. DOMINO is based on the THERIAK algorithm (De Capitani & Brown, 1987). The locations of the T–X sections are shown in Fig. 6b. A constant Si content in Phe is achieved by calculating the T–X sections on sections of constant XMgCel (Fig. 6b). The Pg limb is calculated on sections with the end-members of ideal Pg and Phe (Fig. 6b). For a celadonite (MgCel) content in the range of 0·05–0·15 (Si 3·05–3·15 p.f.u.) there is good agreement between the data of Guidotti et al. (1994b) and the model presented here in the location of the Pg limb (Fig. 6a). In addition, there is good agreement between the Ms limb of the solvus of Guidotti et al. (1994b) and the Ms limb of our model corresponding to a celadonite (MgCel) content of 0·15 (Fig. 6a).
|
Up to now, the only thermodynamic considerations that account for the effect of the celadonite component in Phe are those of Coggon & Holland (2002)
for the system Na2O–K2O–MgO–Al2O3–SiO2–H2O. Those workers proposed 52 kJ to be the interaction energy on the binary Pg–MgCel join. This high interaction energy causes a miscibility gap on the binary Pg–MgCel join, which is very wide over a large P–T range and thus may restrict the position and shape of the ternary solvus to within a relatively small compositional range. In contrast, our investigations point to a much lower interaction energy (20 kJ at 10 kbar), and thus to a smaller miscibility gap on the binary Pg–MgCel join. Despite this relative small miscibility gap the solution model presented here predicts very low celadonite contents in paragonite for the compositions of natural coexisting phengite–paragonite pairs (Figs 5 and 6b). To compare the results obtained by Coggon & Holland (2002, fig. 5b) with our model we calculated T–X sections through the ternary Ms–MgCel–Pg solvus. These were calculated for different Si contents in Phe. In general, there is agreement between the results of Coggon & Holland (2002, fig. 5b) and our model (Fig. 7). However, by comparing the Ms limbs of our model with the Ms limbs of the model of Coggon & Holland (2002) it is obvious that the latter are restricted to a relatively small compositional range, particularly at higher Si (>3·2) content in Phe (Fig. 7). As stated above, this is caused by the high interaction energy on the Pg–MgCel join.
|
Limits of the presented solution model
Because the model is calibrated for pressures up to 21 kbar at about 650°C we do not recommend an extrapolation towards much higher pressures. At low pressures and high temperatures the solution model allows complete mixing between Pg–Ms and between Pg–Cel (MgCel, FeCel). As this occurs outside the high-temperature low-pressure stability limit of paragonite, which may be given by the assemblage phengite + paragonite + sillimanite + quartz (Grambling, 1984
; see Guidotti, 1984 for further discussion), complete mixing on the above binary joins is regarded as a mathematical artefact of our model.
Application
It is well known that the compositions of phengite in pelitic rocks are affected by the metamorphic conditions (Guidotti, 1973, 1984; Guidotti & Sassi, 1976, 1998). To test whether our model predicts the compositional trends observed in Phe of pelitic rocks, we calculated the equilibrium phase diagram section (Fig. 8) and the end-member isopleths of phengite (Fig. 9) for the bulk composition corresponding to the average of 18 pelitic rock samples described by Shaw (1956) (in wt %: SiO2 61·37, TiO2 0·98, Al2O3 19·25, FeO 6·91, MnO 0·7, MgO 2·03, CaO 0·46, Na2O 1·3, K2O 3·58). Nagel et al. (2002) also used this composition to discuss aspects of phase relations concerning metapelites, which experienced decompression from eclogite-facies conditions to amphibolite-facies conditions. The calculation was performed in the system K2O–Na2O–CaO–FeO–MgO–Al2O3–SiO2–H2O with DOMINO (De Capitani & Brown, 1987; De Capitani, 1994) using the database of Berman (1988, update 1992). For staurolite and chloritoid we used the thermodynamic data of Nagel et al. (2002). For garnet we used the solution model of Berman (1990), for biotite binary (Phl, Ann) ideal mixing on site (3) was assumed, for chlorite the solution model of Hunziker (2003) was used, for feldspar we used the solution model of Fuhrman & Lindsley (1988), for staurolite the solution model of Nagel et al. (2002), and for omphacite we used the model of Meyre et al. (1997). The activity of H2O was assumed to be unity.
|
) and quartz are stable with all mineral assemblages. The stippled area outlines assemblages where phengite coexists with paragonite; the area outlined with a grey border corresponds to assemblages where feldspar is stable; the horizontal ruled area outlines assemblages where staurolite, kyanite or sillimanite is stable. Mineral abbreviations are after Kretz (1983)
|
Figure 8 presents the equilibrium phase diagram section and Fig. 9a–d shows the corresponding compositional variation of Phe given by its end-member isopleths. The stippled area in the above figures represents the P–T field where Pg and Phe coexist.
Effects of pressure and temperature on the composition of phengite
From Fig. 9a it can be seen that as temperature and pressure increases Phe becomes Na enriched up to 0·3XPg in phengite. Further temperature and pressure increase has the opposite effect as the Na content decreases and the K content increases in Phe (Fig. 9a and b). The conditions of maximum Na content in Phe are shifted towards higher temperatures as pressure increases. At the same time the maximum amount of Na decreases (see also Guidotti & Sassi, 1976, fig. 13). Because the Phe–Pg solvus widens in response to both a pressure increase and an increase of the celadonite content in Phe, the maximum Na content in Phe decreases as pressure increases. The P–T conditions of maximum Na content in Phe are defined by the low-pressure stability limit of Pg at temperatures higher than about 600°C (Figs 8 and 9a). Similar compositional trends were described by Guidotti & Sassi (1976, figs 7, 13 and 14). At high-T and low-P conditions within the stability range of phases such as St, Ky and sillimanite (Sil) (Fig. 8, area with horizontal ruling) the Na content in Phe decreases as temperature increases and increases as pressure increases (Fig. 9a). The thermally induced decrease of the Na content in Phe is probably caused by the breakdown of Na-rich Phe, which releases Al, forms Pl and produces K-enriched Phe [Guidotti & Sassi, 1976, reaction (7)]. This is in accordance with our calculation as the abundance of Phe decreases within the stability fields of St, Ky and Sill (Keller et al., 2005).
The equilibrium diagram (Fig. 8) predicts that the thermally driven reaction progress, forming high Al phases so typical for the progressive Barrow-type metamorphism, may occur in very small T intervals where continuous and discontinuous reactions take place (e.g. breakdown of Chl and St). However, in collision mountain belts, the formation of high-Al phases may occur during approximately isothermal decompression from high-pressure conditions, which has recently been discussed for the central Alps (Lepontine Dome) by Nagel et al. (2002) [see Keller et al. (2005) for further discussion].
Figure 9c and d shows the P–T-induced variation of the celadonite content. The isopleths of the MgCel content in Phe are predominantly functions of pressure within most assemblages, whereas the FeCel content in Phe decreases as temperature increases, particularly at higher pressures (Fig. 9c and d). By combining the compositional variation of the two celadonite end-members the variation of the celadonite content coincides with that discussed by Guidotti & Sassi (1976, fig. 12). Those workers stated that: (1) particularly at low temperatures the celadonite content increases as pressure increases; (2) particularly at higher pressures the celadonite content decreases as temperature increases. By comparing the isopleths of XFeCel in Phe with calculated abundance of Grt it is indicated that the FeCel content in Phe probably decreases during thermally driven growth of Grt (see Keller et al., 2005).
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
We believe that our model represents an improvement over existing solution models for white micas as it accounts for the effect of the celadonite component on the Phe–Pg solvus. When the model is used to calculate the P–T-dependent compositional variation of Phe in multicomponent systems, it gives reasonable and consistent results, in both quantitative and qualitative respects, at least for metapelites. The model should be particularly useful in modelling the phase relations and corresponding white mica composition for distinct pelitic bulk compositions, particularly for high-grade metamorphic conditions where the white micas can usually be described in terms of the end-members Ms, Pg, MgCel and FeCel.
|
ACKNOWLEDGEMENTS |
|---|
This study is supported by the Swiss National Foundation Grant 20-61814.00. The reviewers D. Nakamura, P. O'Brien, Chun-Ming Wu and an anonymous reviewer are gratefully acknowledged for their suggestions and comments. A review of J. Connolly, who does not agree with the derivation of the present solution model, is also acknowledged. We thank S. Bucher for providing us with his unpublished data. In addition, we had support from K. Waite and R. Bousquet.
* Corresponding author. Telephone: 0041/(0) 61 267 36 31. Fax: 0041/(0) 61 267 36 13. E-mail: Lukas.Keller{at}unibas.ch
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONINPUT DATACALCULATION PROCEDURECOMPARISON BETWEEN THE MODEL...COMPARING OUR SOLUTION MODEL...CONCLUSIONREFERENCES |
|---|
Ahn, J., Peacor, D. R. & Essene, E. J. (1985). Coexisting paragonite–phengite in blueschist eclogite: a TEM study. American Mineralogist 70, 1193–1204.[Abstract]
Ashworth, J. R. & Evirgen, M. M. (1984). Garnet and associated minerals in the southern margin of the Menderes Massif, southwest Turkey. Geological Magazine 121, 323–337.[Abstract]
Berman, R. G. (1988). Internally-consistent thermodynamic data for minerals in the system Na2O–K2O–CaO–FeO–Fe2O–Al2O3–SiO2–H2O–CO2. Journal of Petrology 29, 445–552.
Berman, R. G. (1990). Mixing properties of Ca–Mg–Fe–Mn garnets. American Mineralogist 75, 328–344.[Abstract]
Blencoe, J. G., Guidotti, C. V. & Sassi, F. P. (1994). The paragonite–muscovite solvus: II. Numerical geothermometers for natural, quasibinary paragonite–muscovite pairs. Geochimica et Cosmochimica Acta 58, 2277–2288.[CrossRef][Web of Science]
Bousquet, R., Goffé, B., Oberhänsli, R. & Patriat, M. (2002). The tectono-metamorphic history of the Valaisan domain from the Western to the Central Alps: new constraints on the evolution of the Alps. Geological Society of America Bulletin 114, 207–225.
Brown, E. H. & Forbes, R. B. (1986). Phase petrology of the eclogitic rocks in the Fairbanks district, Alaska. In: Evans, B. W. & Brown, E. H. (eds) Blueschists and Eclogites. Geological Society of America, Memoirs 164, 155–167.
Chatterjee, N. D. & Flux, S. (1986). Thermodynamic mixing properties of muscovite–paragonite crystalline solutions at high temperatures and pressures, and their geological applications. Journal of Petrology 27, 677–693.
Chatterjee, N. D. & Froese, E. (1975). A thermodynamic study of the pseudobinary join muscovite–paragonite in the system KAlSi3O8–NaAlSi3O8–Al2O3–SiO2–H2O. American Mineralogist 60, 985–993.[Web of Science]
Chopin, C. (1979). De la Vanois au massif du Grand Paradis, une approche pétrographique et radiochronologique de la signification géodynamique du métamorphisme de haute pression. Ph.D. thesis, Université de Paris VI, 145 pp.
Chopin, C., Henry, C. & Michard, A. (1991). Geology and petrology of the coesite-bearing terrain, Dora Maira massif, Western Alps. European Journal of Mineralogy 3, 263–291.[Web of Science]
Coggon, R. & Holland, T. J. B. (2002). Mixing properties of phengitic micas and revised garnet–phengite thermobarometers. Journal of Metamorphic Geology 20, 683–696.[CrossRef][Web of Science]
De Capitani, C. (1994). Gleichgewichts-Phasendiagramme: Theorie und Software. Berichte der Deutschen Mineralogischen Gesellschaft. Beihefte zum European Journal of Mineralogy 6, 48.
De Capitani, C. & Brown, T. H. (1987). The computation of chemical equilibrium in complex systems containing non-ideal solutions. Geochimica et Cosmochimica Acta 51, 2639–2652.[CrossRef][Web of Science]
Enami, M. (1983). Petrology of pelitic schists in the oligoclase–biotite zone of the Sanbagawa metamorphic terrain, Japan: phase equilibria in the highest grade zone of a high-pressure intermediate type of metamorphic belt. Journal of Metamorphic Geology 1, 141–161.[Web of Science]
Eugster, H. P., Albee, A. L., Bence, A. E., Thompson, J. B. & Waldbaum, D. R. (1972). The two-phase region and excess mixing properties of paragonite–muscovite crystalline solutions. Journal of Petrology 13, 147–179.
Feininger, T. (1980). Eclogite and related high-pressure regional metamorphic rocks from the Andes of Ecuador. Journal of Petrology 21, 107–140.
Ferry, J. M. (1992). Regional metamorphism of the Waits River Formation, Eastern Vermont: delineation of a new type of giant hydrothermal system. Journal of Petrology 33, 45–94.
Franceschelli, M., Mellini, M., Memmi, I. & Ricci, C. A. (1989). Sudoite, a rock-forming mineral in Verrucano of the northern Apennines (Italy) and the sudoite–chloritoid–pyrophyllite assemblage in prograde metamorphism. Contributions to Mineralogy and Petrology 101, 274–279.[CrossRef][Web of Science]
Franz, C. & Althaus, E. (1976). Experimental investigation on the formation of solid solutions in sodium–aluminum–magnesian micas. Neues Jahrbuch für Mineralogie, Abhandlungen 126, 233–253.
Fuhrman, M. L. & Lindsley, D. H. (1988). Ternary-feldspar modelling and thermometry. American Mineralogist 73, 201–215.[Abstract]
Gil Ibarguchi, J. I. & Dallmeyer, R. D. (1991). Hercynian blueschist metamorphism in north Portugal: tectonothermal implications. Journal of Metamorphic Geology 9, 539–549.[Web of Science]
Grambling, J. A. (1984). Coexisting paragonite and quartz in sillimanite rocks from New Mexico. American Mineralogist 69, 79–87.[Abstract]
Guidotti, C. V. (1973). Compositional variation of muscovite as a function of metamorphic grade and assemblages in metapelites from N. W. Maine. Contributions to Mineralogy and Petrology 42, 33–42.[CrossRef][Web of Science]
Guidotti, C. V. (1984). Micas in metamorphic rocks. In: Bailey, S. W. (ed.) Micas. Mineralogical Society of America. Reviews in Mineralogy 13, 357–467.
Guidotti, C. V. & Sassi, F. P. (1976). Muscovite as a petrogenetic indicator mineral in pelitic schists. Neues Jahrbuch für Mineralogie, Abhandlungen 127, 97–142.
Guidotti, C. V. & Sassi, F. P. (1998). Petrogenetic significance of Na–K white mica mineralogy: recent advances for metamorphic rocks. European Journal of Mineralogy 10, 815–854.
Guidotti, C. V., Sassi, F. P., Sassi, R. & Blencoe, J. G. (1994a). The effects of ferromagnesian components on the paragonite–muscovite solvus: a semiquantitative analysis based on chemical data for the natural paragonite–muscovite pairs. Journal of Metamorphic Geology 12, 779–788.[Web of Science]
Guidotti, C. V., Sassi, F. P., Blencoe, J. G. & Selverstone, J. (1994b). The paragonite–muscovite solvus: I. P–T–X limits derived from the Na–K composition of natural, quasibinary paragonite–muscovite pairs. Geochimica et Cosmochimica Acta 58, 2269–2275.[CrossRef]
Guidotti, C. V., Sassi, F. P., Comodi, P., Zanazzi, P. F. & Blencoe, J. G. (2000). The contrasting response of muscovite and paragonite to increasing pressure: petrological implications. Canadian Mineralogist 38, 707–712.[CrossRef][Web of Science]
Heinrich, C. A. (1982). Kyanite-eclogite to amphibolite facies evolution of hydrous mafic and pelitic rocks, Adula nappe, Central Alps. Contributions to Mineralogy and Petrology 81, 30–38.[CrossRef][Web of Science]
Hirajima, T., Shohei, B., Yoshikuni, H. & Yoshihide, O. (1988). Phase petrology of eclogites and related rocks from the Motalafiella high-pressure metamorphic complex in Spitsbergen (Arctic Ocean) and its significance. Lithos 22, 75–97.[CrossRef][Web of Science]
Höck, V. (1974). Coexisting phengite, paragonite and margarite in metasediments of the Mittlere Hohe Tauern, Austria. Contributions to Mineralogy and Petrology 43, 261–273.[CrossRef][Web of Science]
Hoffer, E. (1978). On the ‘late’ formation of paragonite and its breakdown in pelitic rocks of the southern Damara orogen (Namibia). Contributions to Mineralogy and Petrology 67, 209–219.[CrossRef][Web of Science]
Hunziker, P. (2003). The stability of tri-octahedral Fe2+–Mg–Al chlorite. A combined experimental and theoretical study. Ph.D. thesis, University of Basel, 162 pp.
Irouschek, A. (1983). Mineralogie und Petrographie von Metapeliten der Simano-Decke unter besondere Berücksichtigung Cordieritführender Gesteine. Ph.D. thesis, University of Basel, 205 pp.
Jackson, S. L. (1989). Extension of Wohl's ternary asymmetric solution model to four and n components. American Mineralogist 74, 14–17.[Abstract]
Katagas, C. & Baltatzis, E. (1980). Coexisting celadonite muscovite and paragonite in chlorite zone metapelites. Neues Jahrbuch für Mineralogie, Monatshefte H 5, 206–214.
Keller, L. M., Abart, R., Schmid, S. M. & De Capitani, C. (2005). Phase relations and chemical composition of phengite and paragonite in pelitic schists during decompression: a case study from the Monte Rosa nappe and Camughera–Moncucco unit, Western Alps. Journal of Petrology 46, doi:10.1093/petrology/egi051.
Koch, E. (1982). Mineralogie und plurifazielle Metamorphose der Pelite in der Adula-Decke (Zentralalpen). Ph.D. thesis, University of Basel, 201 pp.
Kretz, R. (1983). Symbols for rock-forming minerals. American Mineralogist 68, 277–279.[Abstract]
Massonne, H. J. & Szpurka, Z. (1997). Thermodynamic properties of white micas on the basis of high-pressure experiments in the system K2O–MgO–Al2O3–SiO2–H2O and K2O–FeO–Al2O3–SiO2–H2O. Lithos 41, 229–250.[CrossRef][Web of Science]
Meyre, C., De Capitani, C. & Partsch, J. H. (1997). A ternary solid solution model for omphacite and its application to geothermobarometry of eclogites from the middle Adula nappe (Central Alps, Switzerland). Journal of Metamorphic Geology 15, 687–700.[CrossRef][Web of Science]
Meyre, C., De Capitani, C., Zack, T. & Frey, M. (1999). Petrology of high-pressure metapelites from the Adula nappe (Central Alps, Switzerland). Journal of Petrology 40, 199–213.[CrossRef][Web of Science]
Nagel, T. (2002). Metamorphic and structural history of the southern Adula nappe (Graubünden, Switzerland). Ph.D. thesis, University of Basel, 103 pp.
Nagel, T., De Capitani, C. & Frey, M. (2002). Isograds and P–T evolution in the Southeastern Lepontine Dome (Graubünden, Switzerland). Journal of Metamorphic Geology 20, 309–324.[CrossRef][Web of Science]
Okay, A. I. (1989). An exotic eclogite/blueschist slice in a barrovian-style metamorphic terrain, Alanya Nappes, Southern Turkey. Journal of Petrology 30, 107–132.
Roux, J. & Hovis, G. L. (1996). Thermodynamic mixing models for muscovite–paragonite solutions based on solution calorimetric and phase equilibrium data. Journal of Petrology 37, 1241–1254.
Schmid, R., Wilke, M., Oberhänsli, R., Janssens, K., Falkenberg, G., Franz, L. & Gaab, A. (2003). Micro-XANES determination of ferric iron and its application in thermobarometry. Lithos 70, 381–392.[CrossRef][Web of Science]
Shaw, D. M. (1956). Geochemistry of pelitic rocks. Part 3: Major elements and general geochemistry. Geological Society of America Bulletin 67, 919–934.
Theye, T. & Seidel, E. (1991). Petrology of low-grade high-pressure metapelites from the external Hellenides (Crete, Peloponnese) a case study with attention to sodic minerals. European Journal of Mineralogy 3, 343–366.
Thompson, J. B. (1957). The graphical analysis of the mineral assemblages in pelitic schists. American Mineralogist 42, 842–858.[Web of Science]
Thompson, J. B. & Thompson, A. B. (1976). A model system for mineral facies in pelitic schists. Contributions to Mineralogy and Petrology 58, 243–277.[CrossRef][Web of Science]
Vidal, O., Para, T. & Trotet, F. (2001). A thermodynamic model for Fe–Mg aluminous chlorite using data from phase equilibrium experiments and natural pelitic assemblages in the 100°C to 600°C, 1 to 25 kb range. American Journal of Science 301, 557–592.
Wohl, K. (1946). Thermodynamic evaluation of binary and ternary liquid systems. Transactions of the American Institute of Chemical Engineers 42, 215–249.[Web of Science]
Wohl, K. (1953). Thermodynamic evaluation of binary and ternary liquid systems. Chemical Engineering Progress 49, 218–219.[Web of Science]
Zhang, R. & Liou, J. G. (1994). Coesite-bearing eclogite in Henan Province, central China: detailed petrography, glaucophane stability and P–T path. European Journal of Mineralogy 6, 217–233.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C.-M. WU and G. ZHAO Recalibration of the Garnet-Muscovite (GM) Geothermometer and the Garnet-Muscovite-Plagioclase-Quartz (GMPQ) Geobarometer for Metapelitic Assemblages J. Petrology, December 1, 2006; 47(12): 2357 - 2368. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||