American Mineralogist, Volume 78, pages 158-177, 1993

Disequilibrium decomposition and breakdown of muscovite in high P-? gneisses,Betic alpine belt (southern Spain)

ArlroNro Gnncia-Casco, ANroNro SANcnnz-Nlvls, R.lrlnr, Lurs Tonnns-Ror,oANDepartamento de Mineralogia y Petrologia and I.A.G.M., Consejo Superior de Investigaciones Cientificas,

Universidad de Granada, Fuentenueva s/n, 18002-Granada, Spain

Ansrn-lcr

Muscovite from anatectic leucogneisses of the Torrox gneiss complex (Betic alpine belt,southern Spain) has an unusually wide range in composition. Primary matrix crystals areSirich (up to 6.66 atoms pfu, normalized to 20 O atoms and 4 OH) and contain abundantstructurally controlled biotite + quartz intergrowths that range in size from a few nano-meters to tens of micrometers. Diffi.rsion halos, depleted in Si, Fe, and Mg, developed inmuscovite around these intergrowths. Also, a continuous compositional trend of decreas-ing Si, Fe, Mg, and Ti and increasing Al, Na, and K was formed by primary and recrys-tallized matrix muscovite grains, the latter forming the end product of this evolution, withSi contents as low as 6.14 atoms pfu. Inspection of TEM images of the finer biotiteintergrowths reveals that (l) these intergrowths represent a variety of frozen stages ofgrowth, with interfaces that changed from coherent and semicoherent to incoherent uponcoarsening, and (2) that biotite-quartz nucleation was probably influenced by deformationdefects in the host muscovite. A third textural variety of muscovite appears as large,homogeneous, pegmatitic grains with low Si contents (6.13 atoms pfu), which underwentpartial low-P breakdown to fibrolite * andalusite * potassium feldspar + biotite. Theprimary and recrystallized muscovite crystals do not bear indications of breakdown.

The textures and compositional features associated with the decomposition of the phen-gitic components of primary muscovite, and with the breakdown of the pegmatitic variety,suggest that the overstepping of the respective reaction boundaries was rapid. This isconsistent with independent estimates of the P-T-t trajectory of the Torrox gneiss complex,which was characterized by rapid and marked decompression of approximately l0 kbarat high ?"(from >10 to 2kbar, and from 650 to 600'C) and was followed by very rapidcooling (>200 "C/m.y.) at low P. Phengitic decomposition took place during decompres-sion and deformation, resulting in a wide range of intermediate compositions, the inter-grown biotite * quartz assemblage, and the diffirsion halos. The reaction textures and theresidual composition of primary muscovite are interpreted in terms of a change in therate-limiting step from surface- to diffusion-controlled growth upon coarsening of productphases. The absence ofpotassium feldspar in the product phase assemblage, which conflictswith the stable equilibrium relationships and mass balance calculations, indicates that Kdiffused to the matrix. Although decomposition proceeded probably far from equilibrium(i.e., strongly overstepped), it is suggested that it followed the theoretical behavior expectedfor muscovite solid solution, indicating that P substantially influences the Si content ofhigh-grade muscovites similarly to what is well known in lower Z environments. Themetastable persistence of muscovite outside the subsolidus stability of muscovite + quartzis explained by a combination of (l) a P-T path that evolved near the muscovite-outisograd, (2) high rates ofcooling and decompression at low P, and (3) moderate displace-ments in the P-T location of the equilibrium reaction due to compositional differences ofthe phases involved, most probably the fluid phase.

IurnooucrroN

Because of their widespread occurrence in many meta-morphic and igneous rocks, muscovite solid solutions areimportant in the description and interpretation of naturalsystems. The major deviations of muscovite from its ide-al composition, KrAloAlrSi6Or0(OH)o, are normally as-

0003-004x/93l0 1 02-0 1 58$02.00

cribed to two main solid solutions, the paragonite-mus-covite series, involving the NaK , exchange component,and the phengite series, involving the Tschermak (tk,

SiMgALr) exchange component, which relates muscoviteto the leucophyllite component KrMgrAlrSi'Oro(OH)o.These substitutions are known to be strongly influencedby intensive and extensive parameters and to display an

1 5 8

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES 159

antipathetic behavior in muscovite solid solutions (seeGuidotti, 1984, for a review). Natural phengitic musco-vite also includes Fe2* and Fe3*, involving the FeMg_,and the Fe3*Al , exchanges (Wise and Eugster, 1964), thelatter less well constrained and documented.

The deviation of muscovite from ideal composition asa result of the Tschermak substitution has been studiedin natural, experimental, and theoretical model systems.For a given system, composition, and phase assemblage,a consistent decrease in the Si content of natural mus-covite occurs with decreasing P and increasing T (e.g.,Ernst, 1963; Cipriani et al., l97l; Guidotti and Sassi,1976; Guidotti, 1984). The system composition and phaseassemblage also exert an important control on the extentof substitution: at constant P-Z conditions, muscovitefrom strongly aluminous compositions, bearing one ormore Al-saturating phases, is more Al-rich and Si-poorthan muscovite from less aluminous compositions thatcoexist with potassium feldspar (Guidotti, 1973; Miya-shiro and Shido, 1985). The change in muscovite com-position due to the Tschermak exchange within the lim-iting equilibrium assemblage Bt + Qtz + Kfs + H,O(mineral abbreviations after Kretz, 1983) has been inves-tigated by Velde (1965, 1967), Monier and Robert (1986a),and Massonne and Schreyer (1987). The model net-trans-fer reaction (J.B. Thompson, 1982b) describing the di-variant equilibrium (KMASH system) is

' ""t:ll"?;:(oH)'

+ .*,13'"Y,*l;,.'

: 4KAlSi.O8 + 6SiO,potassiumfeldspar quartz

+ KrMguAlrSi6Oro(OH)4 + 4HrO (l)

written in terms "f

th. ;*;.rmak exchange vector toemphasize the increasing Al contents of the micas on de-composition (J.B. Thompson, 1979; A.B. Thompson,1982). Si isopleths for Reaction I (funther referred to asphengitic muscovite decomposition) have low positivedPldZslopes (Velde, 1965, 1967 Massonne and Schrey-er, 1987), indicating a strong pressure efect on the sub-stitution, which is in good agreement with natural meta-morphic muscovite.

The presence of reaction textures resulting from Re-action I helps constrain P-T paths (Massonne andSchreyer, 1987). Given the geometry of Si isopleths, re-action textures and associated compositional heteroge-neities in muscovite are often found in high-P blueschistand eclogite facies metamorphic rocks that were subject-ed to decompression (e.g., Heinrich, 1982; Saliot andVelde, I 9 82; Franz et al., I 98 6 ; Evans and Patrick, I 9 87).Indications of instability of high-Z phengitic muscoviteare not so common; however, Ferrow et al. (1990) havepresented evidence for exsolution of a l0-A phyllo-silicate, perhaps celadonite, within granitic muscovite. Thelack ofreaction textures caused by Reaction I in high-Imuscovite is the result of (l) its low Si content, in most

cases ranging from 6.1 to 6.3 atoms pfu (either meta-morphic, e.g., Evans and Guidotti,1966; Guidotti, 1973,1978a;Tracy,1978; Fletcher and Greenwood, 1979; Hol-daway et al., 1988; or igneous, e.g., Guidotti, 1978b; Mil-ler et al., 198 1; Lee et a1., 198 l; Kistler et al., 198 l; Price,1983; Monier and Robert, 1986b; Sevigny et al., 1989),(2) the positive dP/dT slopes of reaction isopleths, incombination with P-lnpaths dominated by cooling, or (3)favorable kinetics under slow changes in P-T conditionsand availability of fluids. However, that heterogeneousmuscovite populations often develop within single sam-ples, and even within individual crystals, in cooling plu-tonic bodies (e.g., Miller et al., I 98 I ; Speer, I 984; Monieret al., 1984; Konings et a1., 1988) indicates that musco-vite equilibration is a sluggish process under high- to me-dium-temperature conditions.

The subsolidus upper stability of muscovite in SiOr-saturated rocks is described by the univariant equilibri-um (KASH system)

K,Al6Si6Oro(OH)" + 2SiO,muscovtte quanT

2AlrSiO5 + 2KAlSi3O8 + 2H2O (2a)aluminumsilicate potassiumfeldspar

which is commonly considered as the low-Z limit (iso-grad) for high-grade rocks. In spite ofthe discontinuousnature of Reaction2a, muscovite can be found coexistingwith Kfs + Qtz + aluminum silicate across a zone ofpersistence in high-Z rocks subjected to prograde meta-morphism at low to intermediate P. This zone might re-sult from (l) kinetic factors (Kerrick et al., l99l), (2) rheheterogeneous nature of fluids (e.g., Speer, 1982) as Re-action 2a is displaced to lower T at the condition Prro 1P,* (e.9., Kerrick, 1972), or (3) variations in the compo-sition of the solid phases. Experimental and theoreticalstudies on Reaction 2a show that, under low to moderateoverstepping of the reaction boundary, muscovite shouldrapidly react to completion (e.g., Ridley and Thompson,1986; Schramke et al., 1987; Kerrick et al., 1991) andthat the displacements of Reaction 2a in the P-Z spacedue to variations in the composition of muscovite wouldnot account for this zone ofpersistence. The reaction

K,Al5Si6O'0(OH)o + 2x(SiMgAl_,) + (2 - 4x)SiO,muscovite Tschemak's substitution quartz

: (2 - 2x)Al,SiO, + (2 - 2xl3)KAlSi,O,aluminum silicate potassium feldspar

+ (x/3)K,Mg5Al,Si6O,0(OH), + (2 - 2x/3)H,Obiotire

(2b\

which includes the phengite components in muscoviteand the production of a ferromagnesian phase (i.e., bio-tite, KMASH system), occurs at a point only a few de-grees higher than Reaction 2a (A.B. Thompson, 1982),but this displacement is counterbalanced by the effect of

N Malagu ides I Neogene Bas ins

F Alpujarride" ll:l Fly"ch

Medi te r ranean Sea , 1oo Km ,

I! Pl io-Quaternary

ffi Vinuela Formation

El::l Flysch Unit

Alpujarr ide uDit !

fulYelez-Velaga

ffi Salarea

- TorroxLf,=l Gneiss CoEplex

ffi Competa

aN S ier ra Te jeda

M e d i t e r r a n e a n S e a

(.&

r60 GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-ZGNEISSES

Fig. 1. Geologic sketch map of the Velez Malaga-Sierra Te-jeda massif in the central Betic zone of the Betic Cordilleras.Location of TGC (Torrox gneiss complex) is shown. Lithologicsequences are not differentiated within units. Based on an un-published map by R.L. Torres-Rold6n (see Torres-Rold6tt, 1979;Elorza and Garcia-Dueflas, 1981; Sanz de Galdeano, 1989, foran explanation of these regional terrns and further geologic in-formation).

Na in the micas and feldspars (e.g., Chatte{ee and Froese,1975). Indications for Reaction 2 (either 2a or 2b, furtherreferred to as muscovite breakdown) are not common inhigh-P, medium-to-high ?" rocks subjected to decom-pression, indicating that the P-Zpaths followed by high-Procks are normally dominated by cooling at intermediateto low pressure.

In this paper we document what appears to be a rarenatural case ofphengitic muscovite decomposition (prod-uct assemblage Bt + Qtz) and breakdown (product as-semblage And or fibrous sillimanite + Kfs + Bt) in highP-T anatectic leucogneisses (Torrox gneiss complex,southern Spain). Complex reaction microtextures and widecompositional heterogeneities developed within thesemuscovites as the result of a tectonic setting that causedthe original high P-Iassemblages to undergo rapid near-isothermal decompression at high temperature.

Frpr,o AND PETRoLocrcAL SETTTNG

The Torrox gneiss complex (TGC) is a heterogeneousgneissic body that is located in the central segment of theBetic Cordilleras, about 50 km east of M6laga; the areaincludes several Alpujarride, Malaguide, and flysch tec-

tonic units (Fig. l). The TGC crops out (-3 km'z and>500 m thick) at the bottom of the upper Alpujarrideunit in the area (Torrox unit; Fig. l), beneath a thicksequence of amphibolite facies graphitic schists. Both thegraphitic schists and the TGC gneisses experienced a sim-ilar deformational and metamorphic history during theAlpine orogeny. The latest penetrative synmetamorphicdeformation event resulted in the dominant megascopicmylonitic foliation (Dr, of Cuevas et al., 1989). Assem-blages in the system A-F-M-Ti in the graphitic metape-lites adjacent to the gneissic body include pre-S, Ms * St+ Grt + Bt + Ky * Rt, syn-S, fibrous sillimanite + St+ Bt + Ilm, and post-S2 And + Bt + Ilm + Crd. Allthree aluminum silicate polymorphs coexist in most sam-ples, and a complex mineral growth and dissolution his-tory can be deduced from the textures and compositionsof the minerals. Fibrous sillimanite + Bt + Ilm and And+ Bt + Ilm pseudomorphs after staurolite, Bt * Ms +Pl after garnet, Crd + Bt after garnet and staurolite, andreplacement of kyanite by andalusite indicate that lowerP assemblages evolved from higher P assemblages bymeans of staurolite and garnet breakdown reactions upondecompression. Comparable examples of frozen disequi-librium are common in other Alpujarride sequences(Loomis, 197 6, 197 9; Torres-Rold6n, 197 4, 1 98 1). Uni-form cooling ages for numerous minerals in several iso-topic systems (cooling ages cluster around 2l + 2 Ma;Zncket a l . , 1989, 1992;Moni6 eta l . , l99 la, l99lb) sug-gest that rapid cooling followed rapid decompression,precluding complete reequilibration at low pressure.

Within the TGC the main type of rock is a medium-grained granitic leucogneiss. In addition to quartz, sodicplagioclase, and potassium feldspar, the other majorphases are muscovite and biotite (up to l0-200/o) and, insome cases, garnet (<50/o). Accessory minerals includeabundant apatite and tourmaline, whereas zircon, ilmen-ite, rutile, kyanite, sillimanite, andalusite, and dumorti-erite are less commonplace. Three main varieties of leu-cocratic gneiss can be differentiated in terms of theirtextural, phase-assemblage, and outcrop characteristics:muscovite biotite garnet banded gneiss, augen and por-phyritic muscovite biotite gneiss, and muscovite garnetbiotite aplites containing muscovite, tourmaline, andspessartine garnet \Mith pegmatitic sizes. The porphyriticmuscovite biotite gneiss contains centimeter-sized alu-minous enclaves (restites) with a primary assemblage ofBt + Ky + Rt + Grt + Ap + graphite, which is over-printed by sillimanite, andalusite, ilmenite, and musco-vite. Nongranitic rocks also appear interbedded with thegneisses, including graphitic pelitic layers with variableamounts of muscovite, quartz, biotite, garnet, aluminiumsilicates, feldspars, ilmenite, rutile, graphite, apalite, zir-con, and tourmaline. All these rock types occur as strong-ly foliated layers or lensoid bodies, with the exception ofsome porphyritic muscovite biotite gneisses and the ap-lites. The latter form concordant layers and centimeter-to meter-sized pods and dikes with relationships that arediffuse to sharply crosscutting, suggesting an origin by

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES t 6 1

short-range melt segregation. Due to strong deformationand alpine metamorphism (which includes small to mod-erate amounts of partial melting), as well as its lithologicheterogeneity, the origin of the TGC as a whole is stilluncertain. Alternative hypotheses range from a metased-imentary derivation to a prealpine anatectic gneiss.

Sartrpr,n DEScRIPTIoN

The muscovite populations from four banded musco-vite biotite garnet gneiss samples were studied. Phase as-semblages in the four samples are fairly homogeneous. Inaddition to the main granitic matrix assemblage Qtz +Pl + Kfs, these samples consist of Ms * Bt + Grt andabundant, coarse, rounded, apatite crystals (up to 0.5 mmin size). Potassium feldspar also forms megacrysts thatinclude idiomorphic plagioclase crystals with inversezoning and xenomorphic cores. These megacrysts showintense deformation and sigmoid tails. Matrix biotite isnot abundant (ca. 5olo), and occurs parallel to the mainfoliation defined by muscovite. Medium- to fine-grainedgarnet is always present (<50/o), has corroded rims, andis partly replaced by medium- to fine-grained aggregatesof Ms + Bt + Pl + Qtz and by late retrograde (green)biotite. Accessory phases include tourmaline and veryscarce fine-grained rutile and ilmenite. No aluminum sil-icate polymorphs were detected in the matrix of the an-alyzed samples.

In all samples muscovite is more abundant than bio-tite. With the exception of scarce retrograde aggregatesafter potassium feldspar, three main textural types ofmuscovite habits can be distinguished: primary deformedmatrix crystals, recrystallized matrix grains, and largepegmatitic crystals. Primary muscovite crystals have tex-tural characteristics that suggest equilibrium with the en-closing granitic gneiss. In the samples investigated, de-formation textures and microfabrics are commonplace,but muscovite still has a medium- to coarse-grainedbooklet habit (0.5-2 mm), and it has clean, sharp con-tacts with feldspars. Associated with primary crystals arebiotite-quartz (+ minor occasional rutile) intergrowthson the optical to TEM scale (10-30 pm to l0 3 ;,rrl, s€ebelow), oriented parallel to (001) planes of the host mus-covite. Recrystallized muscovite grains have resulted fromshear deformation and grain-size reduction of primarycrystals. They appear as randomly oriented fine-grainedgrains (<0.5 mm) associated with primary deformedcrystals. Some of these recrystallized grains also containthin intergrowths of biotite, similar to but much lessabundant than those within primary grains. Pegmatiticcrystals, up to 5 cm long and 1.5 cm across, appear lesscommonly within the muscovite biotite garnet bandedgneisses, and particularly in one of the samples selectedfor detailed study (T506). These larger muscovite crystalsare typically associated with coarse tourmaline and formthe core of sparse augenJike structures, with a micro-granular outer rim ofquartz, potassium feldspar, and al-bite. The augen structures, which are deformed and lie

parallel to the foliation of the rock, are interpreted aspegmatitic segregations that formed close to the HrO-saturated granitic solidus. These pegmatitic muscovitesamples bear partial pseudomorphs consisting of coarse-grained And or fibrous sillimanite + Kfs + Bt, which arenot observed in the primary and recrystallized matrixmuscovite.

ANl.LvrrcAl, PROCEDURES

An automated Cameca SX50 electron microprobe(University of Granada), operated at 20 kV, was used todetermine mineral compositions. Data were reduced us-irg a Q-p-z procedure supplied by the manufacturer (seePouchou and Pichoir, 1985, for details), using simple ox-ides (Al,O., FerO., MnTiOo, MgO) and silicates (albite,orthoclase, wollastonite) as calibration standards. De-tailed wavelength scans indicated no appreciable amountsof other elements such as Ba, Sr, and Cl in the micas. Fmight be present in low amounts, but could not be reli-ably estimated on a routine basis. In the case of musco-vite, special care was taken to locate the electron beamon areas where no intergrown phases appeared at the scaleof optical and backscattered electron (BSE) images. Formuscovite, biotite, and feldspars, operating parameterswere optimized to avoid alkali loss, after the results of anumber of experiments, as illustrated in Figure 2.It ac-cordance with the results of these experiments, both Naand K were always counted first (simultaneously) andbeam current and probe sizes were combined so as to usecunent densities of ca. 0.8 nA pm 'z, for which no appre-ciable decrease in counting statistics was expected duringtheir counting times (15-25 s). However, large currentdensities (up to 20 nA pm ') had to be used for smallbiotite intergrowths (5 nA, 0.25-pm2 probe size). Repre-sentative mica compositions are listed in Table I (seeAppendix Tables I and 2t for the complete set of 66 spotanalyses of muscovite studied and representative com-positions of coexisting phases, respectively).

The TEM images were obtained with a Zeiss EM10C(University of Granada), operated at 100 keV. An objec-tive aperture of 40 pm, corresponding to a minimum doo,value of 0.35 nm, was selected to allow adding 001 dil:fracted and 000 transmitted beams to achieve a compro-mise between amplitude contrast and phase contrast inthe images. TEM specimens, selected from specially pre-pared thin sections, were thinned using Ar ion-beamtechniques to obtain electron-transparent edges. Quali-tative analyses were performed with a Jeol 1200-EXequipped with a LINK energy-dispersive X-ray spec-trometer (University of C6diz) operated at 120 keV andwith a spot size of 100 nm to help with the identificationof TEM-scale intergrowths within primary muscovite.

' A copy of Appendix Tables 1 and2 may be ordered as Doc-ument AM-93-514 from the Business Ofrce, Mineralogical So-ciety of America, 1130 Seventeenth Street NW, Suite 330,Washington, DC 20036, U.S.A. Please remit $5.00 in advancefor the microfiche.

t 62 GARCIA.CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

Biotite

I

E - ! f f i tr

K (Ka)

Time (seconds)Fig. 2. Plots showing the efect of the duration of electron

bombardment on measured X-ray intensities of Na and K inmicas. The diagrams represent the percent variation in countingrates (counts/second) during successive l0- and 20-s intervalsrelative to the initial 10-s integration period. The data were ob-tained during single continuous analyses on the same spot ofmicas from sample T3 l3 referred to in the text, using a constantacceleration potential of 20 kV. The results ofexperiments usingprobe current densities of 0.8 nA/pm' (open symbols) and 20nA/pm2 (solid symbols) are shown for comparison. The plotsindicate that no significant alkali mobilization takes place duringthe initial 50 s when moderate current densities of about I nA/pm2 are used (e.g., 20 nA with a probe diameter of about 5-6pm). For biotite, the same applies up to much higher probe cur-rent densities, which allowed the use of a probe size of approx-imately 0.25 pm2 to analyze small biotite intergrowths withinrnuscovite.

MrNnn,ll, coMposrrroN

Muscovite

A continuous compositional spectrum is formed by theprimary and recrystallized grains (Figs. 3, 4). The groupof primary muscovite samples exhibits the widest inter-and intragrain compositional range. Recrystallized grainsshow a much narrower intergrain compositional range,and individual crystals are homogeneous within analyti-

0.04.O

r'hl+0EI

Fig. 3. (a) Multicationic diagram including relevant mica end-members to illustrate the continuous nature of the coupled cat-ion substitutions underlying the compositional variations withinthe investigated muscovite populations. Abbreviations are ex-plained in the text and Figure 12, except eas, which stands forthe eastonite end-member [Krt6rAlrMg"AloSioO'o(OH)o]. 0) En-larged section of the plot in (a), including a regtession on theprimary and recrystallized grains. Note the progressive lncreasein the trioctahedral contents as compositions deviate from themuscovite end-member. Higher apparent trioctahedral contentsof the pegmatitic type are inferred to be caused by high Fe3*/Fe2* ratios. (c) Compositional variation within a representativesingle primary crystal from sample T316.

cal uncertainty. Because of the heterogeneous nature ofindividual primary crystals, we have used the completeset of analyses reported in Appendix Table l, which rep-resents single spot compositions located within compo-sitionally homogeneous portions at the scale of BSE im-ages. Compositional changes in the matrix grains (primaryand recrystallized) include a decrease in Si (6.66-6.14atoms pfu), Fe (0.46-0.10), Mg (0.33-0.05), Ti (0.13-0.01), and an increase in I4rAl (1.34-1.86), IurAl (3.17-3.84), Na (0.10-0.20), K(l.62-1.71), and total interlayeroccupancy (1.7 4-1.94). The octahedral occupancy com-pares well with other published analyses of muscovite (cf.Mil ler et al., l98l;Guidotti, 1984) and decreases (4.10-4.00) as primary muscovite becomes depleted in Fe andMg (Fig. 4). The good correlations found among Si, Mg,Fe, Fe * Mg, and I6rAl (Fig. 4, Table 2) indicate that theprogress ofReaction I is responsible for the largest frac-tion of the observed compositional variations in matrixmuscovite. On the other hand, the pegmatitic muscovitegrains form a homogeneous compositional group (Figs.3, 4). Their composition overlaps with recrystallized grainshaving the lowest Si values (mean 6.13 atoms pfu, Tablel); however, pegmatitic muscovite has a lower alkali con-tent (Fig. 4).

The presence of Fe3* in matrix muscovite is impliedby the relatively poor correlations of Fe with Si andt6rAl,as compared with those of Mg, and the positive correla-tion between Fe and total octahedral cations (Fig. 4, Ta-ble 2). Normalizingto 44 negative charges with Fe3t un-accounted for results in an overestimation of all cationsin the structural formula, thus increasing the calculated

In+a 1.0+U

oE o.b

80ta

100

l.i 80

P. ooc9

8 4 0

n5 t mm

+) ll0

7 ltx,

o s ,

00tn

100

4.5

O Pr imarya Pegrnatiticc Recrystal l ized

Biotite

tr Ttr tr

I

ma (rc)

I

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

TABLE 1. Representative analyses of muscovite (single spots) and biotite

1 6 3

Muscovite

Intergrown'T31 6Prm.

Spot 1

T336Rcrt'

Spot 79

T506Pegm

Spot 12Mean

( n : 1 1 )Mean

( n : 2 )

sio,Tio,Alro3Cr,O3FeOMnOMgoCaONa.OK.o

TotalH,O"

ratAl16lAl

CrTiFeMnMgt6t>

NaKn 2t>

48 961.27

28.110.003 6 70 0 61.650.01o.429.68

93.824.407

6.6611.3393.1690.0000.1 300.4170.0070.3344.O570 0010 1121.6811 794

+ c . / o0.21

36.030.031.010 0 20 2 60.010.77

1 0 1 794.274.472

o . t J o

1.8643.8300.0030.0210.1 130.0030.0524.0230.0010.2001.7401.942

46.s30 . 1 0

36.870.001.490.020.260 0 00.68

1 0 . 1 396.06

+ . c c o

6.1241 8763.8430.0000.0100.1640.0020.0514.0690.0000.1741.7001.874

34 4322.867

19.3390 0 1 4

24 2670.1 484.4070.0160 1958.875

94.5613.823

5.4012.599o.9770.0020.3383.1840.0201.0305.5510.0030.0591.7761.838

0.6490.208o.4620.0160.7750.0270.2030.0170.0460.0710.3350.019

0.0830.0830.0390.0020.0240 . 1 1 30 0040.0430 0570.0030.0140 0190.013

34.6953.338

19.452no

22.6140 1654.8000.0100 1388.691

93.9003.835

5.4252.5751 .010nd

0.3922.9570.0221 . 1 1 95.5010.0020.0421.7341.777

0.0360.0040.0930.0100.0330.0360.1320.007

0.0130.0130.021

0.0070.0010.0010 0200.0050.0020.0100.0100 019

0.1460.0670.148

Note.' Total Fe expressed as FeO Cations normalized to 20 O atoms and 4 OH; n : number of analyses; nd : not determined. Petrographic groupsof muscovite: Prm : primary, Rcrt : recrystallized, and Pegm : pegmatitic grains (see Appendix Table 1 for the complete set of analyses). Matrixbiotite represents an average of high-Ti (>0.3 atoms pfu) compositions (see text), and the intergrown biotite corresponds to the biggest lamellaeoccurring within the primary muscovite crystal of Figure 5.

. Analyses used in the mass balance calculations (Eq 7 in the texo.-- Calculated HrO wt% on the basis of 4 OH pfu

octahedral occupancies. The recrystallized and pegmatit-ic grains have a higher octahedral occupancy (mean of4.03 and 4.07, respectively) than expected as a continu-ation of the trend formed by the primary grains and rel-ative to their Fe + Mg contents (Figs. 3, 4). This suggeststhat the Fe3*/Fe2* ratio in these muscovite grains is high-er. The presence of Fe3* might also account for the ratioFe/Mg > I in our samples, which increase from the pri-mary (mean 1.178, o :0.236) to the recrystallized (mean1.808, o : 0.282) and pegmatit ic (mean 3.214, o: 0.505)grains, as compared with the experimentally determinedgreater relative solubility of Mg in muscovite (Velde, 1965,1967; Monier and Robert, 1986a). Any rigorous evalua-tion of the Fe3* contents in the muscovite is precludedby the probable occurrence of octahedral occupancy inexcess of 4 atoms pfu (Monier and Robert, 1986a; Mas-sonne and Schreyer, 1987), as the amount of the calcu-lated Fe3* (normalizing to eight tetrahedral plus four oc-tahedral cations and 44 negative charges) is proportionalto the octahedral occupancy (normalizingto 22 O). How-ever, the formulae that include calculated Fe3* can beconsidered as end-member estimates. For matrix mus-covite, these calculations resulted in a slight modificationof the compositional ranges of the major elements andlittle or no change of the minor elements (compare theranges given above with Si : 6.63-6.12, .6tS1 : 3.82-

' " b 0 6 2 6 4 6 5 6 8

- - 6 0 6 2 6 4 6 6 6 8 6 0 6 2 6 4 6 6 6 a

si si si

Fig. 4. Selected bivariate diagrams showing the composi-tional spectrum of the entire muscovite data set. Circles : pri-mary grains. Squares : recrystallized grains. Triangles : peg-matitic grains. The regression lines do not include the pegmatiticsrarns.

0 5

o 4

0 3

D 2

o l

0.!

0 1 5

0 1 0

o 0 5

*

a

0 9

0 6

0 3

6 2 6 4

Si

3 4 3 6 3 a 4 0Lnlot

1 8 0

1 7 5

170

0 2 0

0 1 8

0 1 6

0 1 4

o t z

164 GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

Tret-e 2. Pearson correlation matrix of selected variables for primary and recrystallized muscovite

si rorAl Ti Fe Mo r6t> Na K n2r> Na/K Fe + Mg

Sir6lAlTiFeMg16t>NaKF2t>Na/KF e + M g

1.000-0.932 1.000

0.813 0.888 1 .0000.765 -0.911 0.6740.975 -0.978 0.8510.225 -0.408 0.098

-0.838 0.853 -0.849-0 538 0.321 -0.275-0.814 0.649 -0.610-0.781 0.826 -0.827

0.893 -0.978 0.784

1.0000.8s4 1.0000.715 0.363

-0.671 -0.863-0.196 -0.453-0.465 0.758-0.657 0.818

0.970 0.955

1.000- 0.149 1 .000-0 .156 0214 1 .000-0.192 0 630 0.893

0.130 0.991 0.0800.576 -0 786 -0.322

1.0000.519 1 .000

-0 .619 -0 .758 1000

Note: Number of observations: 61.

3. l2,Mg: 0.33-0.05, T i : 0 .13-0.01, Na : 0.20-0.10,and K : 1.76-1.61), indicating that only minor errorsare introduced in the other components by the chosennormalization, with Fe,o, as Fe2*. This is not the case withregard to the pegmatitic muscovite, whose octahedral oc-cupancy is largely overestimated as a probable conse-quence ofvery high Fe3*/Fe,t ratios (see below).

Other minerals

Biotite also exhibits strong intergrain heterogeneities(Appendix Table 2). Matrix crystals are Ti-rich (>0.3atoms pfu), intermediate Ti contents are found in biotitepseudomorphs (+ Ms + Pl) aftergarnet, and low-Ti (<0.1atoms pfu) crystals are retrograde pseudomorphs aftergarnet. This indicates that garnet-consuming reactions areresponsible for the change in composition of the matrixbiotite crystals. The decrease in Ti involves a parallelincrease in tulAl, Fe, and octahedral occupancy and a de-crease in Mg. On the other hand, the composition of in-tergrown biotite within primary and pegmatitic musco-vite is relatively Ti-rich (0.39 and 0.25 atoms pfu,respectively, Table I and Appendix Table 2). Garnet isnormally unzoned, being rich in Fe and Ca (xo," > 0.2)and poor in Mg and Mn (xro" < 0.03, Appendix Table 2).Lower Ca and higher Mn contents are recorded in somefine-grained crystals and in thin rims (up to 50 pm) of thecoarser grains. This zonation is crosscut by the Ms + Btf Pl pseudomorphs. Plagioclase is rich in albite (Appen-dix Table 2). The compositional variations within indi-vidual samples are related to several stages of gro\\th,involving a progressive increase in the anorthite content.In sample T313, crystals included in Kfs megacrysts ap-pear with a slight reverse zoning with xenomorphic cores(;". : 0.09) and idiomorphic overgrowths (xo, : 0.13),whereas matrix grains are homogeneous (xo, : 0.14).Within the pseudomorphs after garnet, plagioclase bearshigher anorthite contents (x"" : 0. I 9). Potassium feldsparalso has minor compositional intergrain heterogeneitieswithin individual samples (Appendix Table 2), beingslightly less albitic in the fine-grained matrix grains (x^o: 0. I 9-0. I 8), as opposed to the larger grains (x^o : 0.23-0.21), except when the latter appear shghtly exsolved (e.g.,T 3 1 3 , x o o : 0 . 1 8 ) .

MrcnorBxrunps

Primary and recrystallized muscovite

The compositional heterogeneities within individualprimary muscovite grains are generally associated withthe biotite-quartz intergrowths. Halos depleted in Si, Fe,and Mg and enriched in Al develop close to these inter-growths (Fig. 5), indicating that they were produced byphengitic decomposition of primary muscovite, in agree-ment with the observed bulk compositional changes ofmatrix muscovite (Fig. a). The discontinuous and irreg-ular outer rims appear to be related to the formation oflarger biotite grains and overgrowths and to grain-sizereduction and recrystallization at the edges of the crystals(Fig. 5A), although internally recrystallized grains are alsoapparent (Fig. 5,{, 5B). Since recrystallized grains are theend product of the main compositional trend of primarycrystals bearing Bt + Qtz intergrowths (Figs. 3, 4), theircompositions can be directly related to the same contin-uous reaction affecting the primary muscovite. In the fol-lowing paragraphs microtextural observations related tothe mechanism of generation of the biotite-quartz inter-growths are summarized, shedding light on the kineticsof primary muscovite decomposition.

Heterogeneous nucleation. A variety ofdefects and de-fective areas appear over broad areas of primary mus-covite grains. Most of these are deformational structuresresulting from the shear deformation suffered by theserocks, and they include abundant stacking faults, elasticbending ofthe layers, edge dislocations (Fig. 6), and localpull-apart microstructures (Fig. 7). Because the storedelastic and surface energy reduces the activation energyrequired for the formation of a nucleus of critical size,defects are favorable nucleation sites (e.9., Putnis andMcConnell, 1980; Lasaga, l98l; Ridley and Thompson,1986). In our images, this is reflected by the abundanceof small intergrowths (i.e., abundant nucleation sites). Aparticular case was found in relation with pull-apart mi-crostructures. These microstructures formed as voids andmicrocracks along (001) planes (Fig. 7), probably alongthe interlayer region, which contains large and less stronglybonded cations (Veblen, 1983). In the example illustratedin Figure 7, the plastic strain is stored as two opposite

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES 1 6 5

Fig. 5. BSE images and qualitative element traverses from aprimary rnuscovite crystal from sample T336 bearing abundantBt + Qtz intergrowths. (A) Low-magnification view of the crys-tal. The abundant biotite lamellae (bright white zones) and quartzinclusions (dark rods) parallel to the (001) planes of muscovitestand out clearly. Muscovite appears as various shades of gray.Note the irregular outer rim, and the patchy diffirsion zoningwithin muscovite. Recrystallized muscowite grains appear as high-angle inclusions and at the edges of the primary muscovite. (B),

(C), and (D) Enlarged area marked by arrow in A showing Feand Mg depletion halos (dark gray) in muscovite (gray) close tointergrown biotite (white), and a tabular dark recrystallized mus-covite grain. (E) and (F) Enlarged area marked by arrow in Ashowing elongated rods ofquartz (black) and biotite laths (white)and the depletion in Si and enrichment in Al of muscowite (gray)

close to the intergrowths. The qualitative element traverses areexpressed in counts per second.

t66 GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

Fig. 6. TEM images of a composite intergrowth, within aprimary muscovite host, consisting of two 200-A biotite lamel-lae separated by an inner low-contrast (quartz?) band. (A) Low-magnification electron micrograph of this threefold microstruc-ture. (B) An enlarged image of the region outlined in A showingabundant defective areas (thin arrows) consisting ofelastic bend-ing ofthe layers and edge dislocations in muscovite. Biotite la-mellae show only minor deformation as compared with mus-covite. The semicoherent muscovite-biotite interface is indicatedby the thicker arrows: no change in orientation along a and bdirections of the micas across the interface is inferred from theabsence of marked contrast.

edge dislocations, so that the initial continuity betweenthe two muscovite layers afected by the pulling apart isactually disrupted. In this figure, two poorly defined sin-gle fringes (the arrowed darker fringe inside the lighterarea oflow electron density) appear attached to the twoedge dislocations disrupting muscovite. These extra fringescould be single atomic rows of a quenched Angstrdm-sized phase (biotite?) growing in the void as the micro-structure migrates through the climb of the edge dislo-cations. An alternalive interpretation is that these apparentextra fringes do not actually result from an extra layer,but rather from the Fresnel fringes at the edges of themica along these gaps (see also Eggleton and Buseck,I 980).

Semicoherent growth. TEM images further confirm theevidence for structurally controlled (interface-controlled)growth of biotite, and to a lesser extent quartz, parallelto (001) planes of the micas. However, variable degreesof apparent coherency relative to the c* direction (Figs.6, 8, 9) were found to be related to the grain size ofthebiotite lamellae, and thus to different stages of growth.

Fig. 7. Shear pull-apart microstructure parallel to the mus-covite layers. Two edge dislocations within muscovite disruptthe structure that probably grew through a climb process. Thetwo poorly defined extra fringes that appear attached to thesecoupled edge dislocations might represent single atomic rows ofa growing phase (biotite?) in the void or Fresnel fringes at theedges of the mica along these gaps.

The coincidence of the c* axes of muscovite and verythin biotite lamellae is indicated by lattice fringes withl0-A spacing in both phases (Fig. 8). Since Figure 8 isnot a structure image, no direct information can be gainedrelative to the structural continuity along other direc-tions. However, the lack of boundaries with sharp con-trast, such as those related to stacking faults of micas orother phyllosilicates (e.g., Veblen, 1983, his Fig. 2), whichwould indicate a change in orientation in a and b direc-tions of the lamellae (D. R. Veblen, personal communi-cation, 1991), suggests the development of semicoherent(001) interfaces for this very thin lamella (i.e., during theinitial stages of growth).

The above structural relations suggest that (00 l) planesmight have acted as optimal phase boundaries (Robinsonetal., l97l) for both micas, in spite of the fact that theirdifferent a and b dimensions would make an exact fit ofboth structures along (001) planes impossible (see Iijimaand Zhu. 1982. for coherent biotite-muscovite inter-growths perpendicular to mica layers). The a and b di-mensions are related to the coplanarity of the basal O ofthe tetrahedral layers and the geometry ofthe octahedralcoordination in micas and are a function of the octahe-dral cations and occupancy (Zussman, 1979; Weiss et al.,1985). Thus, coherence was favored during initial stagesof growth by the higher phengite contents of the hostmuscovite (i.e., by its greater a atd b dimensions; Zuss-man, 1979; Massonne and Schreyer, 1986), although itwas not maintained with further decomposition as the cand D dimensions of muscovite shrink. With continueddecomposition and biotite gro"rth, the elastic strain fieldsat the semicoherent interfaces would be expected to be-

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES 167

200

Fig. 8. Semicoherent biotite lamella (dark) within muscovlre.Continuity ofa and b directions in both micas across the inter-faces is inferred from the absence of sharo boundaries betweenthe two phases.

come too large and would lead to the development ofdefects and increasingly incoherent interfaces. This al-lows a reduction of the strain energy but increases thesurface energy (Putnis and Mcconnell, 1980) and is il-lustrated by the bending ofthe layers close to the inter-faces of relatively small lamellae (Fig. 6). Figure 6A showsa threefold structure consisting of an inner low contrastband (quartz?) surrounded by two biotite lamellae (eachabout 200 A thick) included in muscovite. Abundant edgedislocations involving (001) planes in muscovite (Fig. 68)indicate the lack of perfectly coherent boundaries, al-though the dislocations might in fact be an artifact of thebending of the layers causing apparent shifting of thefringes (Guthrie and Veblen, 1989, 1990). As noted pre-viously, the orientations ofthe a and b directions ofad-jacent muscovite-biotite pairs in this figure also appearto be the same across the two interfaces. The symmetryof this structure suggests that nucleation took place in aplanar structure (stacking fault?) away from which thegrowth of biotite had taken place and where quarlz is nowlocated. Since the biotite layers bear only minor defor-mation compared with the host muscovite, their nucle-ation on defective areas of this type and the accommo-dation ofboth phases appoar to have benefited from thealready deformed structure of muscovite. The develop-ment of semicoherent to incoherent interfaces is henceinferred to have resulted from both nucleation and growthon defective areas and the early stages ofthe coarseningof previously semicoherent biotite intergrowths.

Coarsening. The coincidence ofthe c* axes is lost, andlow angle grain boundaries develop, when biotite lamel-lae get thicker (more than about 500 A). The advance ofreplacement across the (001) planes of the muscovitewould imply that the diffusion of decomposing productsproceeded approximately perpendicular to mica layers,

Fig. 9. Low magnification TEM image of intergrown biotitelamellae of various sizes. The thicker one (bottom) bears irreg-ular boundaries and is surrounded by quartz rims'

as is indicated by diffusion halos around the intergrownphases (Fig. 5). Regardless of being an unfavorable dif-fusion pathway (e.g., Fortier and Giletti, 199 1), this pro-cess could have been favored because of the minimiza-tion of the interfacial energy, at least while the migratingphase boundaries maintained crystallographic semico-herency. The development ofincoherent boundaries andwide depletion halos in the host muscovite must have ledto a change from the initial surface-controlled semicoher-ent growth to a growth controlled by volume diffusion.The presence of thick biotite lamellae rimmed by quartz(Fig. 9) suggests mass transfer by diffusion of dissolvedspecies in solution through the grain interfaces and growthby precipitation from a fluid.

The above observations indicate that the progress ofthe decomposition reaction at a particular site was con-trolled by the rate of volume diffusion, although a com-plex picture of the reaction pathway is envisaged fromthe large number of intergrowths (i.e., nucleation sites),their wide range of sizes and degree of coherency (i.e.,stages ofgrowth), and the role ofthe continuous hetero-geneous deformation and recrystallization accompanyingmuscovite decomposition. In any case, the close spatialrelationship between the intergrowths and the diffusionhalos in muscovite, the fine-grained nature of the inter-growths, and their heterogeneous distribution within thecore of the primary muscovite crystals (Fig. 5) suggestthat garnet decomposition was not directly related to thegrowth of the intergrown phases.

Pegmatitic muscovite breakdown

The reaction textures associated with the pegmatiticgrains are also distinctive (Fig. 10). Fibrolitic sillimaniteis associated with potassium feldspar in thin reaction rimsaround the muscovite crystals (Fig. 10B), whereas coarse

1 6 8 GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

Fig,. 10. Optical and BSE images of muscovite breakdowntextures within a pegmatitic crystal from sample T506. (A) Op-tical image ofthe chessboard intergrowths ofpotassium feldspar(black and dark gray) and andalusite (gray and light gray) locatedin the core of the pegrnatitic muscovite crystal (white). Partialoptical continuity is indicated by simultaneous extinction ofdif-ferent potassium feldspar grains (crossed polarizers). (B) Opticalimage of fibrous sillimanite (fbr) and potassium feldspar aroundthe periphery of a pegmatitic muscovite crystal in contact withq\aftz; a wider spot of intergrown andalusite and potassiumfeldspar (lower right corner) is unrelated to the fibrous silliman-

andalusite and chessboard intergrowths of And + Kfsdevelop towards the cores of the crystals (Fig. l0B, l0C).The chessboard intergrowths are rimmed by potassiumfeldspar and scarce biotite growing parallel to the (001)planes of the host muscovite (Fig. 10C, lOD). The prod-uct assemblages and textural analysis indicate that thismuscovite reacted through the breakdown Reaction 2b,proceeding initially at the grain contacts between mus-covite and quartz to produce small amounts of Kfs +fibrous sillimanite, which partly isolated the cores ofmuscovite crystals from the matrix. The fact that the mostextensive breakdown took place inside the muscovitecrystals, away from the surrounding quartz, implies a flu-

ite + potassium feldspar envelope (plane-polarized light). (C)and (D) BSE images of the potassium feldspar + andalusitechessboard spot shown in A. Skeletal andalusite (black) appearsas blocky areas intergrown with potassium feldspar (light gray)and biotite (white) in the central area of the spot (C). Towardthe rims (D), biotite and potassium feldspar lamellae are orient-ed parallel to the (001) planes ofthe host muscovite, causing theragged appearance ofthe contact between the reactant muscoviteand the product assemblage. Note the presence within the prod-uct assemblage ofirregular patches ofunreacted muscovite (darkergray) associated with andalusite (thicker white arrows).

id-rich environment to facilitate the transport of Si with-in the pegmatitic crystals; otherwise, these cores of mus-covite (considered as a SiOr-subsaturated subsystem)should not have reacted to produce andalusite.

Pegmatitic muscovite lacks indications of decomposi-tion processes analogous to those already described forprimary muscovite (i.e., through Reaction l), probablydue to its limited compositional deviations from puremuscovite. BSE inspection and qualitative traversesshowed no compositional heterogeneities in the musco-vite that borders the product phases, indicating that sig-nificant continuous net exchanges were not involved andhence supporting the discontinuous nature of Reaction

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES t69

2b (cf. A.B. Thompson, 1982). What is notable is that (1)these muscovite samples did not react to completion, and(2) the matrix muscovite crystals lack any textural evi-dence for breakdown through Reaction 2b. These factsconstrain the interpretation of these reaction textures interms of the amount of overstepping required for the for-mation of critically sized nuclei (see below).

P-I-t rvor,urroN

Garnet-biotite (GARB) thermometry (Thompson, I 976;Ferry and Spear, I 978; Hodges and Spear, I 982; Perchukand Larent'eva, 1983; Ganguly and Saxena, 1984; In-dares and Martignole, 1985; Berman, 1990) and garnet-plagioclase-biotite-muscovite (GPBM) barometry (Ghentand Stout, l98l; Hodges and Crowley, 1985; Hoisch,1990) were selected for P-T estimations in the analyzedbanded gneiss (Fig. I l). Difficulties in the application ofboth equilibria to these rocks arise from the composi-tional heterogeneities displayed by the intervening phases,such that conditions of equilibrium cannot be demon-strated. The compositions used, considered to approachearly equilibration conditions, are the homogeneous high-Ca garnet core, high-Ti matrix biotite, high-Si primarymuscovite, and matrix plagioclase (Table l).

A large range of temperatures (600-750 "C) are ob-tained when different calibrations of the GARB Fe-Mgexchange equilibrium and activity expressions are ap-plied to the analyzed samples. Figure I I presents the end-member Testimates, together with error bands (2o) basedon analytical error propagation. As a general result, thehighest temperatures are obtained when garnet activitymodels (Hodges and Spear, 1982; Ganguly and Saxena,1984; Berman, 1990) are included to correct the experi-mental calibration of Ferry and Spear (1978). Composi-tional corrections to account for deviations ofbiotite fromthe pure Mg-Fe system (Indares and Martignole, 1985)result in lower temperature estimates (in the range of 600-650 'C), whose error bands overlap with solutions of boththe experimental calibrations (Ferry and Spear, 1978;Perchuk and Larent'eva, 1983), and the empirical cali-brations (Thompson, 1976). These latter estimates areconsidered to approach more closely the early equilibra-tion conditions. The empirical calibrations of the GPBMequilibrium of Hodges and Crowley (1985) and Hoisch(1990, his Reaction R6) yield pressure estimates in therange of 12-14 and 13.5-17 kbar, respectively (calculat-ed at 650 'C; the molar volume of grossular calculatedfollowing Newton and Haselton, l98l). These estimatesare considered to have low accuracy because the com-positions used by these authors in their regression cal-culations differ significantly from those present in the an-alyzed banded gneisses, particularly in the case of themicas (note that Hodges and Crowley, 1985; and Hoisch,1990, used pelitic assemblages bearing any of the AlrSiO,polymorphs, so that the micas are saturated in Al). Inaddition, a large uncertainty arises from the possible in-accuracies ofother equilibria used for the regression cal-ibrations (Hodges and Crowley, 1985, suggested a maxi-

r (c) r (c)

Fig. 11. (a) P-Tdiagram showing selected mineral and melt-ing reactions within the KNaFMASH (dashed curves) andKFMASH (solid curves) systems. All reactions after A.B.Thompson (1982), except the Si isopleths for Reaction I in thetext (numbers without boxes) that are after Massonne andSchreyer ( I 987) and the aluminum silicate transitions (after Hol-daway, I 97 I ). The remaining reactions are labeled following thenumbers of A.B. Thompson (1982, his Fig. 7, written with thelow-Iassemblage to the left) as follows: 33': Fe-Ms + Ab + Fe-Bt + Ksp + Qtz + V: L, 5': Fe-Ms * Ab + Qtz: Kfs + Als+ Fe-Bt + Y,34': Kfs + Ab + Qtz + Fe-Bt + V: L, 37': Fe-Ms + Ab + Qtz : Fe-Bt + Kfs + Als + L, whereAls : AluSiOs,V: HzO, and L: liquid; Reactions 33 and 5 (KFMASH sys-tem) are similar to 33' and 5' (KNaFMASH system), respec-tively, but excluding Ab. Reaction 5 in this figure is Reaction2b in the text. The dotted arrow indicates the path followed bythe studied rocks, inferred from thermobarometric estimates(below), the compositional range of muscovite, and the break-down product ofpegmatitic grains. (b) (c), and (d) P-f plots ofequilibria for three of the samples studied. The three curvesshown for each biotite + garnet equilibrium (dashed curves la-beled I&M : Indares and Martignole (1985), model A; solidcurves labeled H&S : Hodges and Spear (1982) correspond tothe calculated curve for the selected compositions (central line)and to 2o error bands, based on analytical error propagation onK.[:(Mg/Fe)G"/(Mg/Fe)B']. The curves shown for garnet + pla-gioclase + biotite + muscovite equilibria (labeled GPBM, Hodgesand Crowley, 1985) represent the selected compositions withline types, according to plagioclase type as follows: solid curve: matrix; dashed and dotted curves (sample T313): rim andcore, respectively, of plagioclase inclusion within potassiumfeldspar phenocryst. In all plots, the Si isopleth of Reaction Icorresponding to Si : 6.6 atoms pfu is shown for reference.

mum precision of +2 kbar and tl00 "C for a given

calculation). Because of these problems of precision andaccuracy, no proper error propagation calculations wereattempted, and Figure I I presents the results using thecalibration of Hodges and Crowley (1985). These esti-

15

! r o

170 GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

mates do nonetheless indicate pressures greater than l0kbar, in agreement with the Si content of matrix mus-covite. The graphical celadonite barometer of Massonneand Schreyer ( 1 987, as represented in our Fig. I I a) yieldsca. I I kbar (at 650 'C) with the maximum Si content ofthe analyzed primary muscovite (6.66 atoms pfu).

After these high-pressure initial conditions, decom-pression and lower Iconditions are indicated by the low-er Ca and higher Mn contents of garnet rims, either ifthese are considered as a growth characteristic (cf. Green,1977) or a difusion-induced feature at relatively high ?",and by the Pl-Bt-Ms pseudomorphs after garnet, whichcrosscut this zonation. The compositional heterogeneitiesand reaction textures of muscovite also indicate that rel-atively large changes in P affected these rocks since theearly crystallization and equilibration of primary mus-covite. The decrease in Si content (6.66-6.14 atoms pfu)accompanying the decomposition and recrystallization ofmatrix muscovite would indicate a change in pressurefrom ca. ll to 2-3 kbar (Massonne and Schreyer, 1987),whereas the breakdown of pegmatitic muscovite to And+ Kfs + Bt pseudomorphs through Reaction 2b indicatesthat the rocks reached low pressures at relatively hightemperatures, i.e., ca. 600'C and 2 kbar (Chatterjee andJohannes, 1974 A.B. Thompson, 1982) (Fig. I l).

Although large errors may be associated with the cal-culated pressures and temperatures, we infer roughly l0kbar of decompression with minor cooling (50-100 "C)(Fig. I la). The presence of both deformed and crosscut-ting aplite bodies indicates that continuous ductile defor-mation was afecting the Torrox gneiss complex whilepartially molten, i.e., that the main deformational eventstarted at high P-T and proceeded to subsolidus condi-tions. The processes of decomposition and recrystalliza-tion affecting the primary muscovite also suggest that de-formation was taking place during decompression, asrecorded by the nucleation of biotite in deformation de-fects. This close relationship between deformation andnear isothermal decompression implies that the main de-formation recorded in the TGC is associated with crustalextension rather than thickening (e.g., Thompson andRidley, 1987), opposite to the interpretation ofCuevas etal. (1989), who associate it with nappe emplacement. In-deed, the occulrence of rapid synmetamorphic extension-al tectonics associated with the main deformation eventis consistent with the clustering of cooling ages of severalisotopic chronometers (Rb/Sr and aoArl3eAr in micas andfeldspars) around 2l + 2 Ma in rocks from the TGC andadjacent metapelites Qnck et al., 1989, 1992; Moni6 etal., 199 1a). As an example, specimen T337 , a porphyriticmuscovite biotite gneiss not presented in this paper thatalso contains Si-rich pnmary and low-Si recrystallizedmuscovite, yields cooling ages of 22.4 + 0.7 Ma (Rb/Srmuscovite-whole rock), 19.0 + 0.7 Ma (4oArl3eAr mus-covite), 20.3 ! 0.3 (aoAr/3eAr biotite), and 20.0 + 0.6(low-Z 40Ar/3eAr potassium feldspar), and the matrixmuscovite from sample T316 yields a 40Ar/3eAr age of19.4 + 0.4 Znck et al.. 1989. 1992: Moni6 et al.. l99la,

and unpublished data). Continued rapid decompressionis suggested by the depositional age of 19 + I Ma for thenappe-sealing sediments of the La Vifiuela formation(Gonz6lez-Donoso et al., 1982), which includes bouldersof medium-grade graphite schists belonging to the Torroxunit (see Fig. l). These data indicate very high rates ofcooling (100-300 'Clm.y.) and uplift (2.5 to > 5 km/m.y.)for the late low-P stage of the decompression path (Zecket al., 1989, 1992; Moni6 et al., l99la).

Drscussroill: MoDEL DECoMPOSITIoN ANDBREAKDOWN REACTIONS

Phengitic decomposition

Qualitative reaction pathway. One significant feature ofthe biotite-quartz intergrowths within primary muscoviteis the lack of a distinct minimum size (i.e., l0-6-10 'zmm

along c* for the case ofbiotite), indicating that nucleationdid not end as the reaction progressed (see Ridley, 1985).The fact that new nucleii were continuously generatedduring decomposition of primary muscovite is explainedby (l) the continuous nature of the decomposition reac-tion, (2) a high decompression rate at intermediate tohigh pressure, and (3) a change from surface- to diffusion-controlled growth of the thicker intergrowth with theprogress ofreaction. Textural analysis ofthe reactant pri-mary muscovite and product biotite and quartz indicatesthat the development of depletion halos within musco-vite close to the biotite-quartz intergrowths hamperedfurther growth and even caused the cessation ofdecom-position at a particular growing site when the halos hadreached a critical width (in the range of 5-10 pm). Thelarge free energy changes ofreactions involving the relicthigh-Si compositions are proportional to the degree ofoverstepped conditions (Ridley and Thompson, 1986) andshould have preserved and even increased the nucleationrate. This favored the progress of reaction in these not-yet-decomposed areas by further nucleation (either oncontinuously generated deformation defects or at ran-dom) and semicoherent growth of biotite. At the hightemperature of decomposition, this reaction pathway ne-cessitates a high decompression rate; otherwise furthergrowth of previously formed interglowths would havebeen favored rather than an extremely high number ofintergrowths. This resulted in the heterogeneous spatialdistribution and sizes of the product phase and of thediffusion-induced zoning in the reactant muscovite.

Under these circumstances, decomposition must haveproceeded irreversibly as the relict composition of mus-covite in these not-yet-decomposed areas was increasing-ly displaced from equilibrium during decompression (cf.Ridley and Thompson, 1986, their Fig. 4). The lack ofquantitative compositional data on the finer biotite in-tergrowths precludes the evaluation of a hypotheticalcrystallization of biotite with a disequilibrium composi-tion; however, the absence of conspicuous potassiumfeldspar within the intergrown assemblage could be con-sidered an indication of a metastable decomposition pro-

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES t 7 l

Tleue 3. Components and net-transfer reactions (KTMT|ASHsystem) describing the decomposition of primarymuscovite

Phasecomponents Additive components Exchange components

Muscovite (Ms) lGAl6Si6O,o(OH)4 (Ms) Tk, Prt, di-tri, Ti-SplBiotite (Bt) K,Mg6Al,Si6O,o(OH)4 (pht) Tk, prt, di-tri, Ti !Potassium

feldspar (Kfs) KAlSi3Os (Or)Quartz (Qtz) SiO, (Otz)Rutile (Rt) TiO, (R0Fluid H.O

Net-transfer reactions (mole units)3 Ms + 6Tk : 4 Or + 6Otz + Phl + 4H,O6 Or + 6 Pr l :24 Qtz3 M s + 6 d i - t r i : 3 P h l3 M s + 6 T i - S p l : 3 P h l + 6 T i f ]3 Ms + 6Ti-Spl : Phl + 6 Rt + 4 Or + 4H2O

@

oFeMg_, € sry- offiJ1

0.50

0.25

A/ofe; Exchange component abbreviations (reterence is made to selectedexperimental investigations): Tk : {4rsit61Mgl41Al 1t6rAl 1 (Tschermak's sub-stitution; Velde, 1965, 1967; Monier and Robert, 1986a; Massonne anoSchreyer, 1987), Ti-Spl : r6rMgt6rTir6rAl , (named titanium spinel as in Dy-mek, 1983), Prl : I41Sil1'r!F2rK_1r41Al I (pyrophyllite; vetde, 1969; Rosen-berg, 1987); di-tri : t6rMg 3r6tAl 2t618-, (di-trioctahedral; Monier and Robert,1986a; Massonne and Schreyer, 1987), Ti ! : Tir6rUMg_2 (Ti !; Forbesand Flower, 1974; Abrecht and Hewitt, 1988).

cess, since Reaction I predicts detectable amounts of thisphase. Nonetheless, as will be shown below, we interpretthis major divergence among the expected and observedphase assemblage produced by the decomposition of pri-mary muscovite in terms of (1) the effect of other com-ponents of muscovite in the final product assemblage,and (2) diffirsion of K to the matrix.

Mass balance calculations and net-transfer reactions.The bulk decomposition reaction can be evaluated fol-lowing the algebraic approach of J.B. Thompson (1982a,1982b), in terms of a set of independent net-transfer re-actions among independently variable components de-fined as linearly independent exchange vectors and ad-ditive components (Table 3). For the solid solutions ofmuscovite and biotite, with muscovite and phlogopite,respectively, as the additive components, the number ofindependent exchange vectors to be defined is six, sincethe number of stoichiometric and charge-balance con-straints that must be satisfied is four (i.e., total positivecharge : total negative charge : 44, Si + t4rAl : 8, I6lAl+ Ti + Fe2* * Mg + t61fl : 6, andK + Na + u2ra: 2),and the number of compositional variables in the phasesis ten (cf. Labotka, 1983; Hewitt and Abrecht, 1986). Formuscovite, in addition to the exchange vectors fm(FeMg ,) and nk (NaIC,), the exchange componentsshown in Table 3 were selected to account for the ob-served correlations among the components (Fig. 4, Table2), and following the suggestions by Hewitt and Abrecht(1986). For biotite, the same exchange components wereselected, except for the titanium spinel exchange, whichhas been substituted for the Ti-vacancy exchange (Table3) because the amount of Ti clearly controls the amountofoctahedral vacancies (see Appendix Table 2).

Since exchange reactions do not substantially modifythe modal abundances of the phases involved in the re-

qdlcp D

A

0.04

0.000.15

0.10 o

o o Aprl

0.03

pa tr

6 O

0.55 0.60 0.65 0.70 0.75 0.80 0.85rllS

Fig. 12. Bivariate plots illustrating the compositional varia-tions within the investigated muscovite populations (symbols asin Figs. 3, 4). The regression lines do not include the pegmatiticcrystals. The seven-component cation basis (KNaFMTiAS) hasbeen transformed into the molecular basis resulting from theoperation ofthe exchange components mentioned in the text andTable 3 on the muscovite additive component (see Greenwood,1975; J.B. Thompson, 1982a, for details on the procedure). Inaddition to muscovite, the ensuing molecular species are leuco-phyllite [cp, K.Mg.AlrSi8Or0(OH)4], titanium muscovite [Ti-ms, KrMgrTirAlrsi6oro(OH)ol, pyrophyllite [prl, AloSirOro(OH)o],phlogopite (phl, K,MgAl,Si6O,0(OH)ol, and paragonite (pa,NarAluSiuOro(OH)ol. Negative values of the leucophyllite com-ponent are obtained in the case ofthe pegmatitic grains, wherethe estimated trioctahedral and Ti components account for moreMg + Fe than its actual value. In this case, higher Fe3*/Fe'?*ratios may have led to an overestimation of the octahedral oc-cupancy when total Fe is expressed as Fe2*.

actions (J.B. Thompson, 1982b), those involving the ex-change compononts fm and nk can be excluded from con-sideration by condensation of the KNaFMTiASH systeminto the KMTiASH system (i.e., projecting from FeMg_,and NaIl,). This procedure is warranted by the fact thatthe doublets Fe-Mg and Na-K behaved similarly upondecomposition of primary muscovite (Figs. 4, l2). The

0.000.25

0.15

0.05

-0.050.08

( 1 )(3)(4)(5)(6)

0.05

0.000.06

0.00o.r2

0.04 L-0.50

0.08

^ o o au ^ o

t 72

five independent net-transfer reactions (Reactions 1, 3,4, 5, 6, Table 3) relating the I I selected components with-in the KMTiASH system can be used to describe thechanges in muscovite composition and modal propor-tions of the product phases, but only to the extent thatthe selected exchange components describe composition-al changes not biased by the normalization procedure. Itmust be noted that, in practice, the selection of a definiteset of independent components is not a simple task forcomplex solid solutions because their number and naturedepend on (1) the degree ofcompleteness ofthe analyses(e.g., the availability of independent determinations ofFe3*, H, and O), (2) the type of normalization used in thestructural formulae (e.9., 22 O atoms vs. 8 tetrahedralplus 4 octahedral cations in muscovite), and (3) particularcrystal-chemical choices (e.g., toli vs. rarTi). Additionally'once a given composition has been normalized, consid-erable freedom exists in the selection given the largenumber of potential exchange vectors (Hewitt andAbrecht, 1986). These limitations make the selected setof components nonunique, and some of the exchangesmight be nonexistent (see below), but it is considered use-ful to model the compositional changes in muscovite dur-ing phengitic decomposition.

Mass balance calculations were done by transformingthe high-Si primary muscovite cation composition (Prm,Spot I in Table 1) into the coordinate system defined byrecrystallized muscovite (Rcrt, Spot 79 in Table l), in-tergrown biotite (mean in Table l), and stoichiometricquartz, potassium feldspar, rutile, and HrO. The com-positions were recast into the condensed KMTiASH sys-tem and expressed in terms of oxy-equivalent units (24O units) to approximate modal abundances (cf. Bradyand Stout, 1980; Thompson, 1982a). The resultingequatron

Ms (Prm) :0.647 Ms (Rcrt) + 0.158 Bt-r 0.089 Qtz + 0.086 Kfs+ 0.005 Rt + 0.016 H,o (7)

predicts detectable amounts of product potassium feld-spar (i.e., the major solid product assemblage would beformed by 47o/o Bt, 27o/o Qtz, and 260lo Kfs). As the pro-portions of phases expected from Reaction l, in oxy-equivalent units, are 350/o Bt, l8o/o Qtz, and 47o/o Kfs, thenet effect of other components of muscovite upon decom-position has been a decrease ofKfs and an increase in Btand Qtz, whether or not subsequent diffusion modifiedthe predicted values. Recasting the compositions in Re-action 7 in terms of the components shown in Table 3yields (in mole units)

3 Ms + 4.510 Tk + l l30 Pr l+ 0.500 di-tri + 0.990 Ti Spl

: 1 .342 Phl + 0.527 Ti t r + 2.185 Or+ 9.031 Qtz + 0.463 Rt + 3.315 H,O. (8)

Reaction 8 can be considered as a weighted bulk decom-position reaction made up of the simpler net-transfer Re-actions l, 3, 4,5, and 6 corrected by % of the correspond-

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P.TGNEISSES

ing stoichiometric coefficients of the exchange vectors onthe left side of Equation 8 (note that the Ti-Spl exchangehas not been split into its corresponding values involvedin Equations 5 and 6, but the correction factors for theseequations equal'/u of the stoichiometric coefficients of theTi-! and Rt components, respectively). The normalizedvalues of these coefficients in Equation 8 are 63.30/o Tk,15.80/o Prl, 7 .0o/o di-li, and 13.9o/o Ti-Spl, indicating thatthe phengitic muscovite decomposition in Reaction I ac-counts for most of the product phases and changes incomposition of the matrix muscovite. However, thechanges in Ti and the trioctahedral components in mus-covite through Reactions 5 and 4, respectively, contrib-uted mostly in the production of biotite, and the changein the alkali content of muscovite, monitored by the py-rophyllite exchange through Reaction 3, contributed tothe production of more quartz and less potassium feld-spar. This is illustrated in the condensed AKM projectionof Figure 13, where the observed compositional trend islocated in the middle of the trends anticipated for eachof the four exchange vectors operating in muscovite (orfive net-transfer reactions). Since the composition of pri-

mary muscovite undergoing decomposition (assemblageA in Fig. 13) appears displaced towards the Al-Phl-Msjoin of the Al-Phl-Ms-Kfs triangle, lower amounts of Kfswould have been produce d (ca. 25o/o of the product solidphases vs.47o/o predicted by Reaction l, in oxy-equiva-lent units). This is largely the result ofthe increasing al-kali contents in matrix muscovite as decomposition pro-ceeds (Figs. 4, l2).

The above results are probably inaccurate because theydepend on the value ofthe selected set ofexchange vec-tors, which, in turn, critically depends on the value of thenormalization procedure and the lack of estimates of in-dependent Fe3* and H. Particularly important in this con-text is the lack of H estimates and the existence of thepyrophyllite component in muscovite. As long as the in-terlayer cations correlated with other compositional vari-ables in the studied muscovite samples (Table 2), carewas taken to avoid alkali loss during the microprobe anal-yses (Fig. 2); other elements, such as Ba, Sr, or Cs, werefound to be below the normal quantitation limits (i.e., inamounts less than about 0.01 wto/o), and the calculatedA-site deficiency (ranging from 0.256 to 0.058 atoms pfu)must be explained as vacancies (Velde, 1969; Rosenberg,1987) or the possible occutrence of HrO* or HrO in theinterlayer sites (i.e., the hydronium substitution HrO*I(-';see Dyar et al., 1991, and references therein). The nega-tive correlation ofinterlayer cations and Si (Fig. 4, Table2) would favor the occurrence of actual vacancies (i.e.,

the pyrophyllite substitution) but only to the extent thatthe hydronium contents had not been sensitive to thechanges in P-T that triggered muscovite decomposition.This cannot be ruled out, given the evidence for H andO variations in biotite with metamorphic grade and as-semblage as reported by Dyar et al. (1991). It is worth-while to note that the pegmatitic muscovite, believed tohave formed close to the HrO-saturated granite solidus,

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES t73

bears low alkali contents as compared with those of therecrystallized matrix muscovite, suggesting higher H con-tents in the interlayer sites. In fact, if HrOr or HrO fillthe interlayer vacancies, reactions ofthe type

9 Or + 6H3O*K , [Ms] : 1.5 Ms + 18 etz+ 6H,O (9)

can also explain both the progressive increase of alkalications in matrix muscovite with the progress of decom-position and a decrease in the evolved potassium feldsparcomponents relativo to that expected through Reactionl. Thus, although we have followed other workers (e.g.,Wang and Banno, 1987) in modeling the A-site changesthrough the pyrophyllite substitution due to the lack ofindependent estimations of H, the calculated pyrophyllitecomponent (Fig. 12) and the relative contribution ofEquation 3 to the bulk decomposition Reaction 8 mustbe seen at best as a composite estimate of the effect ofboth the pyrophyllite and the hydronium substitutions.Nonetheless, the compositional data presented confirmthat variations in the interlayer sites attended the phen-gitic decomposition of matrix muscovite.

The lack of potassium feldspar occurring as inter-growths within primary muscovite grains contradicts thepredictions of phase equilibrium considerations and themass balance calculations (Eqs. 7, 8). The preferred ex-planation is that K had diffused to the groundmass, themore so when an aqueous fluid phase evolved as decom-position proceeded. Diffusion of K through primary mus-covite crystals can be rationalized as a hydrated com-pound dissolved in the fluid, implying that hydrolysisreactions like

6 Or + 4I{2O : Ms + 12 Qtz + 4KOH (fluid) (10)

actually represented steps of the molecular reactions. Inthis regard, the increase in the alkali content of primarymuscovite as decomposition proceeds can be envisagedto have taken place through exchange with the K-bearingfluid through the simple reaction

KOH (fluid) + H3O*K , (Ms) : 2H,O (l l)

favoring the operation ofthe hydronium substitution inthese micas. More precise compositional data are neededto evaluate the relative importance of the pyrophylliteand hydronium substitutions.

Muscovite breakdown

Rubie and Brearley (1987) and Brearley and Rubie(1990, e.g., their Fig. 5) observed topotactic and skeletaltextures in muscovite that sufered disequilibrium break-down by metastable melting in laboratory experiments(see also Brearley, 1986, for a natural case of metastablebreakdown). These authors obtained complete break-down of muscovite within 2-20 weeks for HrO-saturatedand HrO-undersaturated systems, under strongly over-stepped experimental conditions (50-200 .C) at low p (lkbar), and observed that the stable Kfs + Sil + Bt assem-blage developed only under HrO-undersaturated condi-

(K,Na)AIo.lcp,Ti-m"

1F",Mg;oFig. 13. Condensed AKM plot of analyses of primary (cir-

cles) and recrystallized (squares) muscovite, high-Ti (Ti > 0.3atoms pfu; circle) matrix, and intergrown (star) biotite (Table 1),projected through SiOr, TiOr, and HrO, and showing the sche-matic phase relations (KMASH system) relevant to primarymuscovite decomposition [abbreviations as in Fig. 12, exceptAl-phl: aluminum phlogopite, Krt5rAl,MgrAlrSirOro(OH)ol. Thethree-phase field A, defined by potassium feldspar, muscovitebearing the maximum Si content (6.66 atoms pfu), and hightitanium matrix biotite represents the early equilibration assem-blage. As decomposition of primary muscovite proceeds, thethree-phase field expands (arrow), as represented by the three-phase field B defined by potassium feldspar, low-Si primarymuscovite and intergrown biotite. The compositional deviationsof muscovite undergoing decomposition from the msJcp jointoward pyrophyllite and phlogopite end-members would pro-duce less potassium feldspar upon decomposition relative to theamounts predicted by Reaction I (note also that the dewiationsof muscovite due to the Tk and Ti-Spl exchanges are collineardue to the projection through TiOr). When projected into thisdiagram, the compositions of the heterogenous matrix biotiteoverlap the aluminous biotite end-member.

tions. These results would indicate rapid growth for thetopotactic and chessboard textures of the And + Kfs* Bt pseudomorphs developed in the pegmatitic mus-covite grains (Fig. l0), although the strongly oversteppedconditions for the Equilibrium 2 investigated by Rubieand Brearley (1987) and Brearley and Rubie (1990) makeboth cases not strictly comparable. However, Schramkeet al. (1987) obtained the stable assemblage And + Kfsunder moderate overstepping (0-100 'C, at 0.5-5 kbar)and HrO-saturated conditions and concluded that Reac-tion 2 should quickly proceed to completion during pro-grade metamorphism, even at conditions very close tothe equilibrium boundary, if the rate-controlling step (asurface-controlled rate involving andalusite surfaces; see

AI.O,

t74 GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-ZGNEISSES

also Kerrick et al., l99l) remains unchanged through thereaction progress. Also, Ridley and Thompson (1986) es-timated a wide range of temperature overstepping re-quired for nucleation in the breakdown of muscovite (Ms+ Qtz : Sil + Kfs + HrO) as a function of interfacialenergy and based on selected seeded and unseeded ex-perimental studies at 2 kbar, but they suggested the gen-eralized values of ca. l0 'C and I kbar for dehydrationreactions. These experimental and theoretical evidencesfor low overstepping required for nucleation and high re-action rates conflict with the fact that Reaction 2 did notprogress to completion in most of the observed pegma-titic muscovite crystals. But perhaps even more surpris-ing is the fact that primary and recrystallized matrix mus-covite crystals do not bear any textural evidence forbreakdown through Reaction 2, in spite of the presenceof incompatible fine-grained quartz intergrown withinthem.

The interpretation of the sequential growth of fibroussillimanite and andalusite in terms of a change from thesillimanite to the andalusite stability fields with the re-action progress is further complicated by the metastablenature of fibrolitic sillimanite, which might have grownoutside the sillimanite stability field (cf. Kerrick, 1990).However, the decompression path followed by these rocks(evolving from the kyanite, to the sillimanite, to the an-dalusite stability fields, see Fig. 11) is consistent with ini-tial growth of fibrous sillimanite close to Equilibrium 2.The progress of the reaction, involving andalusite, mustalso have taken place close to the equilibrium boundary,otherwise muscovite would have reacted completelywithin the andalusite stability field, according to the ex-perimental evidence cited above. This and the fact thatthe amount of overstepping of Equilibrium 2 was notadequate for the activation of nucleation in the primaryand recrystallized matrix crystals indicate that cooling ac-companied decompression at this stage of the P-Z evo-lution and that the P-f path evolved close to Equilibrium2. Additionally, kinetic factors afecting the reactionpathway (e.g., a change from surface-controlled to diffu-sion-controlled growth?) could help explain the lack ofcomplete breakdown. However, to explain the contrast-ing behavior of matrix and pegmatitic muscovite crystals,modest displacements in the P-?" space of Equilibrium 2due to compositional differences of the phases, includingthe fluid, seem to be required. The virtually identicalcompositions of recrystallized and pegmatitic grains inthe analyzed samples (Fig. a) and the fact that muscovitebreakdown textures have been observed only in apliticand pegmatitic bodies within the TGC suggest that vari-ations in the local fluid composition could be the majorfactor affecting the onset of breakdown. Important com-positional variables of the local fluids in equilibrium withthese rocks include the activity of ionic species (i.e., K*,Na*, H*; e.g., Wintsch,1975) and of B and F (Manningand Pichavant, 1983; Pichavant, 1987). The onset ofthebreakdown of the pegmatitic muscovite should be pro-moted at higher P by a lowered 4"ro caused by the pres-

ence of B or F (note the coexistence of tourmaline) or theinfiltration from the matrix of a fluid with a lowered a*-/a,, ratio (Wintsch, 1975, his Fig. 2) in the pegmatitic

stnrcture.

CoNcr,uorNc REMARKS

The instability of the leucophyllite component in ma-trix muscovite upon decompression is in accordance withother investigations on natural and experimental systemsunder subsolidus conditions. However, the high Si valuesfound in the primary muscovite of the TGC stand asexceptional when compared with typical compositions ofhigh-grade muscovite (e.g., Si : 6.2-6.3; Miller et al.,1981). Further. both the estimated P-Tconditions of ear-ly equilibration (600-650 .C, > 10 kbar) and the presence

of concordant and crosscutting aplites strongly suggestthat primary muscovite within the banded gneisses had

been in equilibrium with a melt during this stage. Ex-amples of anatectic granites either emplaced or generated

at high pressure and bearing high-Si muscovite are notcommon. However, the intrusive anatectic leucogranitesemplaced in kyanite-bearing peraluminous graniticgneisses in the Higher Himalayas (Searle and Fryer, 1986)also bear muscovite with high Si contents (6.48 atomspfu), which is consistent with our observations and in-terpretation of the TGC as an instance of high-pressuremelting within a markedly thickened crust.

The data presented in this paper indicate that the com-positional behavior of high-?" muscovite is P-sensitive'resembling that of lower grade environments, althoughvery high P conditions and high decompression rateswould possibly be required to detect this effect in naturalinstances because of the increased probability that com-plete reequilibration takes place at low pressure duringhigh-I emplacement or exhumation of granites and an-atectites. In the P-T-t path, section radiometric evidencewas summarized, indicating that high decompression ratescharacteized the low-P portion (<2-3 kbar) of the P-7path depicted in Figure I l Although more difficult toevaluate quantitatively, the reaction textures, and specif-ically the evidence for continuous nucleation and thepreservation of relict Si-rich compositions in primarymuscovite from the TGC, may be taken as a strong in-dication that high decompression rates were also preva-

lent during the earlier stage. This makes conceivable thatthe age of the pressure peak in the Alpujarride series wouldnot have been far in time from the Iow-P cooling agesclustering at 2I + 2 Ma. In this regard it is worth men-tioning that a aoAr/3eAr age of 25.4 + 0.4 Ma was re-ported by Moni6 et al. (1991b), after a low-Z phengite

coexisting with carpholite, aragonite, and chloritoid in asample from the Alpujarride Trevenque unit (located 50km to the northeast of the TGC), which was meant tomark the "end of the underthrusting processes in the Al-pujarride nappes and the beginning of exhumation." Ifthat were the case, an average uplift rate of about l0 km/m.y. is easily derived for the whole decompression tra-jectory of Figure I l, which illustrates what kind of ex-

treme geological conditions accompanied the productionof the muscovite textures reported in this study.

AcxNowr,nlcMENTS

We wish to thank C V. Guidotti, D.R. Veblen, and H J. Massonne fortheir suggestions and comments on an earlier version ofthe manuscript,and J.T. Cheney, G. Guthrie, and Associate Editor J.B. Brady for theirconstructive reviews, which contributed substantially to the article's im-provement. Thanks are also given to D.S. Silverberg for his commentsand help with the English version and M.A. Hidalgo Laguna for his carefulattention to the microprobe analyses. The results presented in this paperform part ofa Ph.D thesis by A.G.C. This study has received financialsupport from the Spanish CAICYT (project PB89-0017, A.G.C. andR.L.T.R., and project PB87-022801, A.S.N.) and from the Junra de An-dalucia (research groups 4065, A.S.N, and 4072, A,G.C. and R.L.T.R.).

RnrnnnNcrs crrEDAbrecht, J., and Hewitt, DA. (1988) Experimental evidence on the sub-

stitution of Ti in biotite. American Mineralogist, 73, 1275-1284.Berman, R G. (1990) Mixing propenies of Ca-Mg-Fe-Mn gamets. Amer-

ican Mineralogist, 7 5, 328-344Brady, J.B., and Stout, J.H. (1980) Normalization of thermodynamic

properties and some implications for graphical and analytical problemsin petrology American Journal ofScience, 280, 173-189.

Brearley, A.J. (1986) An electron optical srudy ofmuscovite breakdownin pelitic xenoliths during pyrometamorphism. Mineralogical Maga-zine. 50. 385-397.

Brearley, A.J., and Rubie, D C (1990) Effects of HrO on the disequilib-rium breakdown of muscovite + quartz Joumal of Petrol ogy, 31, 925-956 .

Chatterjee, N D., and Froese, E. (1975) A thermodynamic study of thepseudobinary join muscovite-paragonite in the system KAlSi.Os-NaAlSi,Or-Al,O,-SiOr-H,O. American Mineralogist, 60, 985-993.

Chatterjee, N.D., and Johannes, W. (1974) Thermal stability and standardthermodynamic propert ies of synthet ic 2M, -muscovi te,KAlr[AlSi]O,o(OH)rl. Contributions to Mineralogy and petrology, 48,89- l 14.

Cipriani, C., Sassi, F P, and Scolari, A (l 97 l) Metamorphic white micas:Definition of parageneric fields. Schweizerische mineralogische und pe-trographische Mitteilungen, 5 l, 259-302.

Cuevas, J., Navarro-Vil6, F., and Tubia, J M. (1989) Interpr6tation descisaillements ductiles vers le NE dans les gneiss de Torrox (ComplexeAlpujarride, Cordilldres B6tiques). Geodinamica Acta, 3, 107- 1 I 6.

Dyar, M.D , Colucci, M T., and Guidotti, C.V. (1991) Forgotron majorelements: Hydrogen and oxygen variation in biotite from metapelites.Geology, 19, 1029-1032.

Dymek, R.F. (1983) Titanium, aluminum and interlayer cation substi-tutions in biotite from high-grade gneisses, West Greenland. AmericanMineralogist, 68, 880-899.

Eggleton, R.A , and Buseck, P.R. (1980) High resolution electron mrcros-copy offeldspar weathering. Clays and Clay Minerals, 28,1l3-178.

Elorza, J.J., and Garcia-Dueflas, V. (1981) Hoja y memoria explicativade V6lez-M6laga. Mapa Geol6gico de Espafra, escala l:50000. InstitutoGeol6gico y Minero de Espafla, Madrid.

Ernst, W.G. (1963) Significance of phengitic micas from low grade schistsAmerican Mineralogist, 48, 1357 -137 3

Evans, B.W., and Guidotti, C.V. (1966) The sillimanite-potash feldsparisograd in western Maine, USA. Contributions to Mineralogy and Pe-trology, 12,25-62.

Evans, B.W., and Patrick, B E (1987) Phengite-.if in high-pressure meta-morphosed granitic orthogneisses, Seward Peninsula, Alaska. CanadianMineralogist, 25, 141-1 58.

Ferrow, E.A., Iondon, D., Goodman, K.S, and Veblen, D.R. (1990) Sheetsilicates of the Lawler Peak granite, Arizona: Chemistry, structuralvariations, and exsolution. Contributions to Mineralogy and Petrology,1 0 5 , 4 9 1 - 5 0 1 .

Ferry, J M., and Spear, F.S. (1978) Experimental calibration of the par-titioning of Fe and Mg between biotite and garnet Contributions toMineralogy and Petrology, 66, l13-117.

t75

Fletcher, C.J.N., and Greenwood, H J. (1979) Metamorphism and struc-ture of the Penfold Creek Area, near Quesnel l-ake, British Columbia.Journal of Petrology, 20, 7 43-7 9 4.

Forbes, W.C., and Flower, M F.J (1974) Phase relations of titan-phlog-opite, KrMgTiAl,Si,O,o(OH)": A refractory phase in the upper mantle?Earth and Planetary Science Irtters, 22,60-66.

Fortier, S.M, and Giletti, B.J. (1991) Volume self-diffusion ofoxygen inbiotite, muscovite, and phlogopite micas. Geochimica et Cosmochimi-ca Acta, 55, 1319-1330.

Franz, G, Thomas, S., and Smith, D.C. (1986) High-pressure phengitedecomposition in the Weissenstein eclogite, Miinchberger Gneiss Mas-sif, Germany. Contributions to Mineralogy and Petrology, 92, 71-85

Ganguly, J., and Saxena, S K (1984) Mixing properties ofaluminosilicategarnets: Constraints from natural and experimental data, and applica-tions to geothermo-barometry. American Mineralogist, 69, 88-97 .

Ghent, E.D., and Stout, M.Z. (1981) Geobarometry and geothermometryof plagroclase-biotite-garnet-muscovite assemblages. Contributions toMineralogy and Petrology, 7 6, 92-97.

Gonz6lez-Donoso, J M, Linares, D., Molina, E., Serrano, F., and Vera,J.A. (1982) Sobre la edad de la Formaci6n de la Vifruela (CordillerasB6ticas, Provincia de M6laga). Boletin de la Real Sociedad Espafrolade Historia Natural, 80, 255-275.

Green, T.H (1977) Garnet in silicic liquids and its possible use as a P-Tindicator. Contributions to Mineralogy and Petrology, 65, 59-67.

Greenwood, H.J (1975) Thermodynamically valid projections of exten-sive phase relationships. American Mineralogist, 60, l-8.

Guidotti, C V. (1973) Compositional vanation in muscovite as a functionof metamorphic grade and assemblage in metapelites from N.W. Maine.Contributions to Mineralogy and Petrology, 42,33-42.

- (1978a) Compositional variation of muscovite in medium- to high-grade metapelites of northwestern Maine. American Mineralogist, 63,878-884

-(1978b) Muscovite and K-feldspar from two-mica adamellite innorthwestern Maine: Composition and petrogenetic implications.American Mineralogist, 63, 7 50-'7 53.

-(1984) Micas in metamorphic rocks. In Mineralogical Society ofAmerica Reviews in Mineralogy, 13,357-468.

Guidotti, C.V., and Sassi, F.P (1976) Muscovite as a petrogenetic indi-cator mineral in pelitic schists. Neues Jahrbuch liir Mineralogie Ab-handlungen, 127, 97 -142.

Guthrie, G.D., Jr., and Veblen, D.R. (1989) High-resolution transmissionelectron microscopy of mixedJayer illite/smectite: Computer simula-tions. Clays and Clay Minerals, 37, l-11.

- (l 990) Interpreting one-dimensional high-resolution transmissionelectron micrographs of sheet silicates by computer simulations Amer-ican Mineralogisr, 7 5, 276-288

Heinrich, C A ( I 982) Kyanite-eclogite to amphibolite facies evolution ofhydrous mafic and pelitic rocks. Adula nappe, Central Alps Contri-butions to Mineralogy and Petrology, 81, 30-38.

Hewitt, D.4., and Abrecht, J. (1986) Limitations on the interpretation ofbiotite substitutions frorn chemical analyses of natural samples. Amer-ican Mineralogist, 7 1,, 1 126-1 128

Hodges, K V., and Crowley, P D. (1985) Enor estimation and empiricalgeobarometry for pelitic systems. American Mineralogist, 70, 702-7 09.

Hodges, K.V, and Spear, F.S. (1982) Geothermometry, geobarometry,and the AISiO, triple point at Mt. Moosilauke, New Hampshire.American Mineralogist, 67, I l l8-1134.

Hoisch, T.D. (1990) Empirical calibration of six geobarometers for themineral assemblage quartz + muscovite + biotite + plagioclase +garnet. Contributions to Mineralogy and Petrology, 104, 225-234.

Holdaway, M.J. (1971) Stability of andalusite and the aluminum-silicatephase diagram. American Journal of Science, 27 l, 97 -131.

Holdaway, M.J., Dutrow, B.L., and Hinton, R.W. (1988) Devonian andCarboniferous metamorphism in west-central Maine: The muscovite-almandine geobarometer and the staurolite problem revisited. Ameri-can Mineralogist, 7 3, 20-47 .

Iijima, S., and Zhu, J. (1982) Electron microscopy ofa muscovite-bioiteinterface. American Mineralogist, 67, | | 9 5-1205.

Indares, A., and Martignole, J (1985) Biotite-garnet geothermometry ingranulite facies: The influence of Ti and Al in biotite. American Min-eralogist, 70, 272-278.

GARCIA-CASCO ET AL.: MUSCOVITE IN HIGH P-ZGNEISSES

I / O GARCIA.CASCO ET AL.: MUSCOVITE IN HIGH P-TGNEISSES

Kerrick, D.M. (1972) Experimental determination of muscovite + qluartzstability with PF,o < P,"o,. American Journal of Science, Z'72,946-958.

-(1990) The AISiO' polymorphs. Mineralogical Society of AmericaReviews in Mineralogy, 22,406 p.

Kerrick, D.M., Lasaga, A.C., and Raebum, S.P. (1991) Kinetics of het-erogeneous reactions. In Mineralogical Society of America Reviews inMineralogy, 26, 583-67 1.

Kistler, R.W., Ghent, E.D., and O'Neil, J.R (1981) Petrogenesis of garnettwo-mica granites in the Ruby Mountains, Nevada. Journal of Geo-physical Research, E6, 1059 l- 10606.

Konings, R.J.M., Boland, J.N., Vriend, S.P., and Jansen, J.B.H. (1988)Chemistry of biotites and muscovites in the Abas granite, northernPortugal. American Mineralogist, 7 3, 7 54-7 65.

Kretz, R. (1983) Symbols for rock-forming rninerals. American Miner-alogist, 68, 277-279.

Labotka, T.C. (1983) Analysis ofcompositional variations ofbiotite inpelitic hornfelses from northeastern Minnesota. American Mineralo-

s ist , 68, 900-914.Lasaga, A.C. (1981) The atomistic basis of kinetics: Defects in minerals.

In Mineralogical Society of America Reviews in Mineralogy, 8, 261-319 .

l€e, D.E., Kistler, R.W., Friedman, I., and Yan Loenen, R.E. (1981) Two-mica granites of northeastern Nevada. Journal of Geophysical Re-search, 86, 10607-10616.

Loomis, T.P. (l 976) Irreversible reactions in high-grade metapelitic rocks.Journal ofPetrology, 17, 559-588.

-(1979) A natural example of metastable reactions involvinggametand sillimanite. Joumal of Petroloey,2O, 27l-292

Manning, D.A.C., and Pichavant, M. (1983) The role of fluorine andboron in the generation of granitic melts In MP. Atherton and C.D.Gribble, Eds., Mignatites, meltingand metamorphism, p. 9zt-109. ShivaPublishing, Nantwich, England.

Massonne, H.J., and Schreyer, W (1986) High-pressure syntheses andX-ray properties of white micas in the system KrO-MgO-Al'O,-SiOt-H,O. Neues Jahrbuch fiir Mineralogie Abhandlungen, 153, 177-215.

-(1987) Phengite geobarometry based on the limiting assemblagewith K-feldspar, phlogopite, and quartz. Contributions to Mineralogyand Petrology, 96, 212-224.

Miller, C.F., Stoddard, E.F, Bradfish, L.J., and Dollase, W.A. (1981)Composition of plutonic muscovite: Genetic implications. CanadianMineralogist, 19, 25-34.

Miyashiro, A., and Shido, F. (1985) Tschermak substitution in low- andmiddle-grade pelitic schists. Journal of Petology, 26, 449-487 .

Moni6, P., Torres-Rolddn, R.L., Garcia-Casco, A, and Goff6, B. (l99la)High rates of cooling in the westem Alpujarrides, Betic Cordilleras,southern Spain: A aoAr/r'Ar study. Terra Nova Abstract, 3 (suppl. 6), 8.

Moni6, P., Galindo-Zaldivar, J., Gonziiez-Lodeiro, F., Gofft, B., andJavaloy, A. (1 99 lb) ooArlsAr geochronology ofAlpine tectonism in theBetic Cordilleras (southern Spain). Journal ofthe Geological Society ofLondon, 148, 288-297 .

Monier, G., and Robert, J.L (1986a) Muscovite solid solutions in thesystem KrO-MgO-FeO-Al,O,-SiOr-HrO: An experimental study at 2kbar PH.O and comparison with natural Li-free white micas. Miner-alogical Magazine, 5O, 257 -266.

-(1986b) Titanium in muscovites from two mica granites: Substi-tutional mechanisms and partition with coexisting biotites. Neues Jahr-buch fiir Mineralogie Abhandlungen, 153, 147 -161.

Monier, G., Mergoil-Daniel, J., and Labernardidre, H. (1984) G6n6rationssuccesives de muscovites et feldspaths potassiques dans les leucogra-nites du massif de Millevaches (Massif Central frangais). Bulletin deMin6ralogie, 107, 55-68

Newton, R.C., and Haselton, H.T. (1981) Thermodynamics of the garnet-plagioclase-AlrSiOyquartz geobarometer. Advances in Physical Geo-chemistry, l,129-145.

Perchuk, L.L., and Larent'eva, I.V. (1983) Experimental investigation ofexchange equilibria in the system cordierite-gamet-biotite. Advancesin Physical Geochemistry, 3, 199-239.

Pichavant, M. (1987) Effects ofB and H,O on liquidus phase relations inthe haplogranite system at I kbar. American Mineralogist, 72,1056-1070.

Pouchou, J.L., and Pichoir, F. (1985) "PAP" dfuz) procedure for im-proved quantitative microanalysis. In J.T. Armstrong, Ed., Microbeam

Analysis, p 104. San Francisco Press, San Francisco.Price, R.C. (1983) Geochemistry of a peraluminous granitoid suite from

north-eastern Victoria, south-eastern Australia. Geochimica et Cos-mochimica Acta. 47. 31-42.

Putnis. A.. and McConnell, J.D.C. (1980) Principles of mineral behavior,257 p. Blackwell, Oxford.

Ridley, J. (19E5) The effect of reaction enthalpy on the progress of a

metamorphic reaction. Advances in Physical Geochemistry, 4, E0-97.Ridley, J., and Thompson, A.B. (1986) The role of mineral kinetics in

the development of metamorphic microtextures. Advances in Physical

Geochemistry, 5, 154-193Robinson, P., Jaffe, H.W., Ross, M., and Klein, C., Jr. (1971) Orientation

of exsolution lamellae in clinopl'roxenes and clinoamphiboles: Consid-

eration ofoptimal phase boundaries. American Mineralogist, 56, 909-939.

Rosenberg, P.H. (1987) Synthetic muscovite solid solutions in the system

KrO-AlrOr-SiOr-H'O. American Mineralogist, 7 2, 7 l6-7 23

Rubie, D.C., and Brearley, A.J. (1987) Metastable melting during the

breakdown of muscovite + quartz at I kbar. Bulletin de Min6ralogie,

I 10, 533-549.Saliot, P, and Yelde, B. (1982) Phengite compositions and post-nappe

high-pressure metamorphism in the Pennine zone ofthe French Alps.

Earth and Planetary Science Irtters, 57, 133-138.

Sanz de Galdeano, C (1989) Estructura de las Sierras de Tejeda y de

C6mpeta (Conjunto Alpujarride, Cordilleras Beticas). Revista de la So-

ciedad Geol6gica de Espafla, 2,77-E4.Schramke, J.A., Kerrick, DM., and Lasaga, A.C. (1987) The reaction

muscovite + quartz: andalusite + K-feldspar + water. Part 1. Growth

kinetics and mechanism. American Journal of Science,287 ' 517 -559.

Searle, M.P., and Fryer, B.J. (1986) Garnet, tourmaline and muscovrte-

bearing leucogranites, gneisses and migmatites of the Higher Himalayas

from Zanskar, Kulu, Lahoul and Kashmir. The Geological Society Spe-

cial Publication. 19, 185-201.Sevigny, J.H, Parrish, R.R., and Ghent, E.D' (1989) Petrogenesis ofper-

aluminous granites, Monashee Mountains, southern Canadian Cordil-

lera. Journal ofPetrology, 30, 557-581.Speer, J.A (1982) Metamorphism ofpelitic rocks ofthe Snyder Group in

the contact aureole ofthe Kigalpait layered intrusion, Labrador: Effects

of buffering partial pressures of water. Canadian Journal of Earth Sci-

ences. 19. 1888-1909.-(1984) Micas in igneous rocks In Mineralogical Society of Amer-

ica Reviews in Mineralogy, 13,229-356.Thompson, A.B. (1976) Mineral reactions in pelitic rocks II. Calculation

of some P-I'-X (Fe-Mg) phase relations. American Joumal of Science,

276,425-454.- (l 982) Dehydration melting of pelitic rocks and the generation of

HrO-undersaturated granitic liquids. American Journal of Science, 282,

t567-1595.Thompson, A.8., and Ridley, J.R. (1987) Pressure-temperature-time (P-

I-f) histories oforogenic belts. Philosophical Transactions ofthe Royal

Society of London, A 321, 27-45.Thompson, J.B., Jr (1979) The Tschermak substitution and reaction in

pelitic schists. In V.A. Zharikov, V.I Fonorev, and J.P. Korikovskii,

Eds., Problems in physicochemical petrology, p. 146-1 59. Academy of

Sciences, Moscow (in Russian)-(1982a) Compositional space: An algebraic and geometric ap-

proach. In Mineralogical Society of America Reviews in Mineralogy,

1 0 , t - 3 1 .- (1982b) Reaction space: An algebraic and geometric approach. In

Mineralogical Society of America Reviews in Mineralogy' 10, 33-52

Torres-Rold6n, R L. (1974) El metamorfismo progresivo y la evoluci6n

de la serie de facies en las metapelitas alpujrirrides al SE de Sierra

Almijara (sector central de las Cordilleras Beticas. S de Espafra)' Cua-

dernos de Geologia, 5,2l-77.-(1979D The tectonic subdivision of the Betic Zone (Betic Cordil-

lera, S. Spain): Its significance and one possible geotectonic scenario

for the westernmost Alpine Belt. American Journal of Science, 279,

l 9 -51 .

-(1981) Plurifacial metamorphic evolution of the Sierra Bermejaperidotite aureole (southem Spain). Estudios Geol6gicos, 37, 115-133.

Tracy, R. (1978) High grade metamorphic reactions and partial meltingin pelitic schists, west-central Massachusetts. American Journal of Sci-ence- z/d- l )u- l /6.

Veblen, D.R. (1983) Microstructures and mixed layering in intergrownwonesite, chlorite, talc, biotite, and kaolinite. American Mineralogist,68, 566-580.

Velde, B. (1965) Phengite micas: Synthesis, stability, and natural occur-rence. American Journal ofScience, 263, 886-913.

-(1967) Si*4 content of natural phengites. Contributions to Miner-alogy and Petrology, 14,250-258.

-(1969) The compositional join muscovite-pyrophyllite at moder-ate pressures and temperatures. Bull6tin de la Societ6 Franqaise de Mi-n6ralogie et de Cristallo$aphie, 92, 360-368.

Wang, G.F., and Banno, S. (1987) Non-stoichiometry of interlayer ca-tyions in micas from low- to middle-grade metamorphic rocks in theRyoke and the Sanbagawa belts, Japan. Contributions to Mineralogyand Petrology, 97, 313-319

t 7 7

Weiss,2., Rieder, M., Chmielovi, M., and Krajicek, J. (1985) Geometryofthe octahedral coordination in micas: A review ofrefined structures.American Mineralogist, 70, 7 47 -7 57 .

Wintsch, RP. (1975) Solid-fluid equilibria in the system KAlSi,O8-NaAlSirO8-AlrSiO5-SiO,-H'O-HCI. Journal of Petrology, 16, 57 -7 9.

Wise, W.S., and Eugster, H.P. (1964) Celadonite: Synthesis, thermal sta-bility and occurrence. American Mineralogist, 49, 1031-10E3.

7nck,H.P., Albat, F., Hansen, B.T., Torres-Rold6n, R.L., Garcia-Casco,A., and Martin-Algarra, A. (1989) A 2l + 2 Ma age for the terminationof the ductile Alpine deformation in the internal zone of the Betic

Cordilleras, south Spain. Tectonophysics, 169, 215-220.Z.eck,H.P., Moni6, P., Villa, I., and Hansen, B.T. (1992)Yery high rates

ofcooling and uplift in the Alpine belt ofthe Betic Cordilleras, southernSpain Geology, 20, 79-82.

Zussman, J. (1979) The crystal chemistry of the micas. Bulletin de Mi-n6ralogie, 102,5-13.

Mem-rscnrrr RECETvED SsprpMnrn 9, 1991Mmruscnrrr AccEprED Aucusr 26, 1992

GARCIA-CASCO ET AL.:MUSCOVITE IN HIGH P-TGNEISSES