Multiple growth mechanisms of jadeite in Cuban metabasite

WALTER V. MARESCH1,*, CHRISTIANE GREVEL1,2, KLAUS PETER STANEK3, HANS-PETER SCHERTL1

and MICHAEL A. CARPENTER4

1 Institut fur Geologie, Mineralogie, Geophysik, Ruhr-Universitat Bochum, Universitatsstr 150,44801 Bochum, Germany

*Corresponding author, e-mail: walter.maresch@rub.de2 TUV Rheinland, Am Grauen Stein, 51105 Cologne, Germany

3 Institute of Geology, TU Bergakademie Freiberg, Bernhard-von-Cotta-Street 2, 09596 Freiberg, Germany4 Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EQ, UK

Abstract: Samples of rocks reported in the literature to be jadeite jade from the subduction-zone complex of the Escambray Massif incentral Cuba have been studied by optical and transmission electron microscopy, electron microprobe and hot-cathode cathodolu-minescence (CL) microscopy. Although these rocks are indeed rich in jadeite, the bulk rock composition generally conforms toMORB, with Na2O enriched by . 3 wt% and CaO depleted by .2 wt%. Al2O3 contents are unchanged. These changes are attributedto early pre-subduction spilitization of the ocean-floor protolith. Relics of magmatic augite preserving an ophitic texture are common.Disequilibrium textures are the rule. Extensively recrystallized rocks show fine, felty intergrowths of predominantly Al-richglaucophane and jadeite, the latter with rims and patches of omphacite. TEM observations indicate extensive replacement ofpyroxene by amphibole. Glaucophane developed rims of magnesiokatophorite and edenite. Chlorite and epidote are also present.Late development of actinolite, chlorite, epidote and albite is observed. Quartz is present. Less recrystallized samples with numerouslarge (.1.5 mm) grains of augite show several types of sodic and sodic-calcic clinopyroxene development: (1) Topotacticreplacement of magmatic pyroxene by jadeite and omphacite along a broad front encroaching upon the augite grain from the rockmatrix. Jadeite dominates where presumably plagioclase was formerly present. Omphacite dominates where augite is internallyreplaced along cleavage and fractures. Late chlorite, taramite and ferropargasite replace these pseudomorphs. (2) Former plagioclaselaths of the ophitic fabric are replaced by jadeite together with lesser omphacite in epitactic relationship with the enclosing augite.Former plagioclase-augite grain boundaries remain preserved. Late pumpellyite is associated with the omphacite. (3) Jadeite þomphacite þ pumpellyite þ chlorite with irregular grain boundaries dominate in the rock matrix between the augite relics, withidiomorphic crystals of epidote scattered throughout and in chlorite–epidote clusters. Pumpellyite is interpreted to be a late retrogradeproduct. Quartz is present. (4) Jadeiteþ omphaciteþ chlorite assemblages, in which monomineralic sheaf-like jadeite aggregates arecommon, fill very thin (500–1500 mm) fractures criss-crossing the sample, including ophitic augite remnants. Cathodoluminescencemicroscopy shows that jadeite in the veins is distinctly different from CL in the other types of jadeite, showing features like oscillatorygrowth zoning indicative of crystallization from a fluid. Generally omphacite develops irregularly along jadeite rims, but recrys-tallization may lead to pairs with straight grain boundaries suggestive of phase equilibration. Comparison with published solvusrelationships suggests temperatures of 425–500 �C. This unusual occurrence of different types of jadeite in a metabasic rock suggeststwo contrasting sources. The first – in the rock matrix, as topotactic alteration of igneous pyroxene and as plagioclase replacementepitactically growing on augite – can be explained as due to local domain equilibration in a rapidly subducted ‘‘spilitized’’ gabbroicrock. The second, in very thin fracture fillings, conforms to an origin as a crystallization product from a pervasive fluid. Conceivably,‘‘pooling’’ of the fluids flowing through the fractures in larger cavities could lead to larger masses of jadeitite. These have not yet beenconclusively documented in the Escambray Massif.

Key-words: jadeite, jadeitite, omphacite, pyroxene topotaxy, pyroxene epitaxy, Escambray Massif, Cuba.

1. Introduction

Jadeite jade is a rare rock type world-wide. In a recentsummary, Harlow et al. (2007) cite � 14 described occur-rences. Recent findings in Iran, Cuba and Hispaniola canbe added to this still very short list (Oberhansli et al., 2007;

Schertl et al., 2007, 2008; Garcıa-Casco et al., 2009;Cardenas-Parraga et al., 2010; Baese et al., 2010). In theNew World, jadeite jade was probably already known andused for tools and ornaments in Mesoamerica (southernMexico through Guatemala to Honduras and Nicaragua)almost 3500 years ago (Harlow et al., 2011, and references

Jadeitite:new occurrences, new data,

new interpretations

0935-1221/12/0024-2179 $ 8.55DOI: 10.1127/0935-1221/2012/0024-2179 # 2011 E. Schweizerbart’sche Verlagsbuchhandlung, D-70176 Stuttgart

Eur. J. Mineral.

2012, 24, 217–235

Published online December 2011

therein). The most important source of New World jade isin the middle Motagua Valley, Guatemala, where it occursin serpentinite melanges on both sides of the MotaguaFault. Harlow et al. (2011) have provided a recent com-prehensive update of these occurrences and show thatknown sources north of the fault extend over 200 km, andover 11 km to the south.

As recounted by Harlow et al. (2011), knowledge ofGuatemalan jade sources was lost by the original inhabi-tants of Mesoamerica. Early re-discoveries in the early1950s then ushered in a new era of research on jadeititesand jadeite-bearing rocks that gained momentum afterhurricane-induced new outcrops and new archeologicalfinds (e.g., Seitz et al., 2001; Harlow et al., 2003, 2004,2007; Harlow & Sorensen, 2005). This interest is two-fold.Harlow & Sorensen (2005), Harlow et al. (2007), andSorensen et al. (2006, 2010) present cogent argumentsfor the formation of jadeite jade by precipitation from ahigh-pressure aqueous fluid in a subduction zone channel.Consequently jadeitites can be viewed as records of fluid-driven or fluid-assisted mass transfer processes in suchsettings. On the other hand, pre-Columbian jade (or possi-bly jade) artifacts have been mentioned from a number ofAntillean islands (e.g., Harlow et al., 2006; Garcıa-Cascoet al., 2009, and references therein), raising the question ofwhether Guatemala represents the only source for thisarcheological jade. As already surmised by Harlow et al.(2006) on the basis of geological and petrological simila-rities to Guatemala, recent findings in Cuban (Garcıa-Casco et al., 2009; Cardenas-Parraga et al., 2010) andHispaniolan (Schertl et al., 2007, 2008; Baese et al.,2010) serpentinite melanges now extend possible sourceareas further to the east along the Greater Antilles.

In their description of the new Cuban occurrences ineastern Cuba, Garcıa-Casco et al. (2009) also point to ashort petrographic report of rare jadeitite pebbles, cobblesand boulders in river deposits of the Escambray Massif incentral Cuba by Millan & Somin (1981), suggesting thatthis occurrence may have been largely overlooked so far.Based on optical microscopy, Millan & Somin (1981)described essentially monomineralic rocks composed ofjadeitite with minor clinozoisite, lawsonite and albite.Some samples contain relics of magmatic clinopyroxenein part replaced by fine-grained aggregates of jadeite,leading Millan & Somin (1981) to suggest that the jadei-tites formed by transformation of small bodies of basicintrusives.

Additional petrographic and analytical data of a suite ofsamples from one of the localities described by Millan &Somin (1981) have been available in an unpublished thesis(Grevel, 2000) or in abstract form for some time (Grevelet al., 1998; Maresch et al., 2007). None of these samplescollected together with G. Millan at this site can be con-sidered to be true jadeitites; they are jadeite-bearing meta-gabbroic rocks. Nevertheless, in the context of currentdiscussions on the origin of jadeitite, these samples havebecome important. Sodic and sodic-calcic pyroxene can beobserved as a topotactic replacement of igneous ophiticaugite, as a replacement of the former plagioclase in the

ophitic texture in epitactic continuity with the augite, as apervasive recrystallization product in the matrix, and as apresumed precipitate from aqueous fluid in late rock frac-tures, all in a single sample. Considering that access to thesample area is now very difficult, we will present theavailable microanalytical, cathodoluminescence and elec-tron microscope data in this paper.

2. Geological setting

The Escambray Massif is located in central Cuba at thenorthern margin of the Caribbean Sea (Fig. 1). In geody-namic terms, the geology of Cuba represents a key elementin our attempts to understand the complex plate tectonicevolution of the Caribbean area (e.g., Garcıa-Casco et al.,2008; Pindell & Kennan, 2009; Stanek et al., 2009; Pindellet al., 2012, and numerous references therein). During theLate Cretaceous and Paleogene the oceanic GreatAntillean island arc and its associated subduction-accre-tion complex interacted and collided with the Yucatanmargin and an intra-oceanic sedimentary prism extendingto the southeast of it (‘‘Caribeana’’ of Garcıa-Casco et al.,2008), before docking onto the Bahamas platform in theEocene. This suture zone is well-preserved in Cuba. In thenorth and northwest of Cuba, the rocks can be related to thecontinental margins of the Bahamas Platform and theYucatan Peninsula, whereas in the south rocks of theCretaceous oceanic island arc and associated ophiolitesare found.

To the rear of this continent–arc collisional suture, andassociated with the Cretaceous island arc, tectonic win-dows expose metamorphic rocks throughout the southernpart of the Island of Cuba (e.g., Garcıa-Casco et al., 2008;Stanek et al., 2009). With an exposed area of about 1800km2, the Escambray Massif (Fig. 1) is the largest suchmetamorphic complex of the Greater Antilles. The massifforms two morphological domes, the western Trinidad andthe eastern Sancti Spiritus dome (Somin & Millan, 1974).Considerable discussion has accompanied attempts to pro-vide a tectono-metamorphic nomenclatural framework forthe Escambray Massif, with ideas evolving from earlylithological concepts to the recognition that the Massifrepresents a complex nappe pile (Somin & Millan, 1981;Millan et al., 1985a, b; Millan-Trujillo, 1997; Stanek et al.,2000, 2006, 2009; Schneider et al., 2004). On the basis ofinterdisciplinary structural, petrological and geochronologi-cal work in the Sancti Spiritus dome, Stanek et al. (2006)recognized four major tectono-metamorphic nappes (Fig. 1),which can be discriminated on the basis of their contrastingpressure-temperature-time-paths. From bottom to top theseare the intermediate- to high-pressure Pitajones, Gavilanesand Yayabo nappes, which are in turn overlain by the low-pressure Mabujina nappe. The lower three represent anaccretionary complex composed of both oceanic and con-tinental margin sedimentary rocks. The uppermostMabujina nappe represents the basal section of theCretaceous volcanic arc (see summaries by Stanek et al.,

218 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

2006, 2009; Garcıa-Casco et al., 2008; and referencestherein). The nappe pile is characterized by multiple thrustplanes as well as out-of-sequence thrusts.

The tectonically lowermost Pitajones nappe consists ofmonotonous carbonate- and quartz-mica schists with abun-dant boudins of marble, metagabbro, and greenschist in itsupper part. Maximum preserved equilibration pressuresare about 8 kbar, while correlated temperatures varybetween 410 and 520 �C (Grevel, 2000; Stanek et al.,2006). The overlying Gavilanes nappe, considered as thesource of the samples investigated in this study (seebelow), is a tectonic mega-melange containing quartzite(occasionally with the rare high-pressure mineral deerite),blueschist, eclogite, amphibolite, and serpentinized ultra-basic rocks in a general matrix of impure metacarbonates,quartz-mica schists, and marbles. Preserved maximumpressure (P)-temperature (T) conditions in the variouslithologies vary from 15 to 25 kbar and 500 to 660 �C(Grevel, 2000, Schneider et al., 2004; Garcıa-Casco et al.,2006). Lenses of ‘‘high-grade’’ serpentinite with antigoriteare common (Auzende et al., 2002), indicating that theseserpentinized ultramafic rocks shared the same meta-morphic history as the other lithologies. They are oftenassociated with metabasic lithologies, and serpentinitemelanges with blocks of eclogite in serpentinite matrixhave been described (Schneider et al., 2004; Garcıa-Casco et al., 2006; Stanek et al., 2006). Therefore thisnappe contains both rocks of a subducted passive marginand fragments of subducted oceanic lithosphere (see

summaries by Garcıa-Casco et al., 2006, 2008; Staneket al., 2006). The Yayabo unit is restricted to the north-eastern part of the Sancti Spiritus dome (Fig. 1), but inter-calated slivers of this unit have also been reported from theupper parts of the Gavilanes unit (Somin & Millan, 1981).The Yayabo unit comprises fine- to coarse-grained, epi-dote-bearing garnet-amphibolites with barroisitic amphi-boles. Maximum pressures of 13–15 kbar are indicated(Grevel, 2000). Low-pressure/high-temperature, amphibo-lite- to greenschist-facies rocks characterize the Mabujinaunit, the tectonically highest nappe of the EscambrayMassif. Peak PT-conditions of �7 kbar at 620–700 �C(Grevel, 2000) support the regional temperature estimatesof Somin & Millan (1981).

3. Sampling

Guided by G. Millan we took samples in 1994 from asmall, dried-out, 3–4 m wide tributary of the UnimazoRiver, at ca. 450 m elevation in a densely forested area(Fig. 1; UTM coordinates of the locality are 637 418,242 2851). The slightly slumped banks of the tributaryexpose friable greenschists of the Pitajones nappe, whichcan be interpreted to represent the in situ nappe unit. Theserocks contain the typical greenschist assemblage of chlor-ite, actinolite, epidote, albite and titanite, � quartz, �white mica, � rutile. Characteristic porphyroblasts ofalbite can impart a typical spotted appearance. No

5 km

C

b

a

Fig. 1. Geological setting. (a) Regional situation in the Caribbean area. [G] Guatemalan jadeitite occurrences; [E] Escambray massif; [C]Sierra del Convento jadeitite occurrences; [R] Rio San Juan Complex jadeitite occurrences. (b) Geological sketch map of the easternEscambray massif (Sancti Spiritus dome) with sample locality. (c) Simplified geological profile (line N-S in part (B)) with sample locality.

Jadeite formation in metabasic rock 219

clinopyroxene or blue amphibole is observed. In contrast,the subangular to rounded cobbles filling the dry streambed present tough, massive, fine-grained, grey- to bluish-green rocks. Thin (�1.5 mm) light-coloured veinlets irre-gularly cross-cutting the rocks, light-coloured diffuseblotches and scattered small dark crystals � 1.5mm indiameter are the most conspicuous macroscopic feature.According to Millan (Personal communication, 1994),these were the rocks described by Millan & Somin(1981) as jadeitites. They are the subject of this report.Based on mapping and structural analysis (Stanek et al.,2006), the source of these blocks is assumed to be theoverlying Gavilanes nappe, which in this area representsa klippe-like remnant (Fig. 1). No in situ exposures of theserocks were found. Although lawsonite is known from sev-eral localities in the Gavilanes unit (Grevel, 2000), anddescribed by Millan & Somin (1981) from their jadeite-bearing samples, no lawsonite was found in the presentsuite of samples.

4. Analytical methods

Bulk rock chemistry was obtained by X-ray fluorescencewith a Philips PW 1400 spectrometer. Uncertainties areestimated to be � 1 % for major and � 10 % for minorelements of importance. FeO was determined potentiome-trically (Ungethum, 1965), and Fe2O3 by difference. H2Oand CO2 were analyzed coulometrically (Johannes &Schreyer, 1981).

Microprobe analyses were obtained with a Cameca SX50 at 15 kV acceleration voltage and 10 nA beam current,20 s counting time (10 s on background). High-resolutionelement distribution maps of small sample areas wereobtained in beam-scan mode at 15 kV, 40 nA, and 140ms counting time per pixel. Data correction followed thePAP procedure (Pouchou & Pichoir, 1984). The standardsemployed were natural pyrope for Si, Al, Mg; naturalspessartine for Mn; natural jadeite for Na; natural rutilefor Ti; synthetic andradite for Fe, Ca; synthetic potassiumsilicate and barium silicate glass for K and Ba, respec-tively; synthetic oxides of Cr, Ni, Cu, and Zn for therespective elements. Calculation procedures for mineralformulae were as follows: pyroxene – 6 oxygens and 4cations; sodic and sodic-calcic amphibole – 23 oxygensand 13 cations without Ca,Na,K; actinolite – 23 oxygensand Fetot ¼ Fe2þ; epidote-group minerals – 12.5 oxygensand Fetot ¼ Fe3þ; chlorite – 28 oxygens and Fetot ¼ Fe2þ;biotite – 22 oxygens and Fetot ¼ Fe2þ; pumpellyite – 8cations without hydrogen.

The cathodoluminescence (CL) examinations were doneusing a ‘‘hot cathode’’ CL microscope (type HC1-LM)developed at the Ruhr-University Bochum (Neuser,1995). This device allows a comparative investigation ofthin sections using transmitted light and an electron beam.We employed a beam energy of 14 keV and a beam currentdensity of�9 mA/mm2 on the sample surface. The pictureswere taken with a digital camera system (DX30 C from

KAPPA optoelectronics GmbH) with a high sensitivity atlow light conditions, requiring only 5–10 s per exposure.At a resolution of 1300 � 1030 points a maximum of 10frames/s is available in progressive scan mode.

The TEM samples were cut from standard petrographicthin sections, using 2.3 mm copper discs for support, andion-beam thinned. They were then examined in a JEOLJEM 100CX microscope operating at 100 kV.

5. Petrography

Mineral abbreviations used in this paper follow theupdated compilation of Fettes & Desmons (2007): Ab ¼albite, Act ¼ actinolite, Amp ¼ amphibole, An ¼anorthite, Aug ¼ augite, Bt ¼ biotite, Chl ¼ chlorite, Czo¼ clinozoisite, Di ¼ diopside, Ep ¼ epidote, Fe2-Prg ¼ferropargasite, Gl ¼ glaucophane, Jd ¼ jadeite, Mg-Ktp ¼magnesiokatophorite, Ne ¼ nepheline, Omp ¼ omphacite,Pmp ¼ pumpellyite, Qtz ¼ quartz, Tmt ¼ taramite, Ttn ¼titanite

Thin-section observation confirms the suggestion ofMillan & Somin (1981) that these rocks represent formerbasic magmatic intrusives. The macroscopically recogniz-able dark crystals are seen to be relict magmatic augitewith recognizable ophitic texture (Fig. 2a). Sample M592is an example with abundant augite relics (Fig. 2b),whereas sample M591 appears to represent a more highlyoverprinted variety with only a few remnants of augite and

Fig. 2. (a) Augite relic with conspicuous ophitic texture. Formerplagioclase laths now consist of jadeite þ omphacite þ pumpellyite(M592; XPL; width of image ¼ 750 mm). (b) Microphoto of M592showing augite relics in a fine-grained matrix consisting mainly ofjadeite þ omphacite þ pumpellyite þ chlorite þ epidote (XPL;width of image ¼ 1250 mm). (c) Microphoto of M591 showinglarge glaucophane crystal with frayed edges and chlorite þ calcicamphibole alteration. The felty matrix consists mainly of glauco-phane þ jadeite as well as omphacite þ chlorite þ epidote (PPL;width of image ¼ 280 mm). (d) Microphoto of M592 with jadeititeveinlets consisting of jadeiteþ omphaciteþ chlorite (XPL, width ofimage ¼ 3.3 cm).

220 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

with the characteristics of a blueschist (Fig. 2c). Asdescribed below, both samples are characterized by abun-dant examples of disequilibrium textures. Grain boundariesare complex and irregular and replacement textures con-trolled by the crystal chemistry of precursor minerals arecommon. Both samples exhibit irregular cross-cutting vein-lets up to� 1.5 mm in thickness rich in jadeite (Fig. 2d). Thefollowing descriptions will focus on these two samples.Because of the very fine-grained nature of these rocks, theoptical microscope by itself is not a sufficient tool for apetrographic description, and therefore the results from qua-litative microprobe data and cathodoluminescence micro-scopy (e.g., Schertl et al., 2004) will be integrated into thissection.

Augite relics up to 1.5 mm in size constitute a majorcomponent of sample M592 (Fig. 2a, b). The rock matrixbetween the augite relics is composed of finely and irregu-larly intergrown jadeite, omphacite, pale pumpellyite andchlorite with complex grain boundaries. Idiomorphic crys-tals of epidote are found scattered throughout the matrixand also in distinct epidote-chlorite clusters. The latter tendto occur where augite relics are less abundant, suggestingthat these clusters may have evolved from the breakdownof magmatic pyroxene. Minor albite appears to be a latesecondary product. Titanite is an accessory phases. Raregrains of quartz are found in the matrix between the augiterelics with no obvious spatial relationships to any of theother matrix minerals. The most distinctive feature of thisrock is the replacement of magmatic augite and of theplagioclase laths of the ophitic intergrowth by jadeite andomphacite (Fig. 3c–e). Augite, jadeite and omphacite arein optical continuity (Fig. 2a). Such topotactic replacementof magmatic augite by omphacite was already documentedby electron microscopy in a blueschist from Turkey(Carpenter & Okay, 1978). Petrographic descriptionsexist from other localities (e.g., Essene & Fyfe, 1967;Black, 1974). In the present case it is noteworthy that thesodic pyroxene topotactically replacing augite is in opticalcontinuity not only with the magmatic pyroxene, but alsowith sodic pyroxene replacing plagioclase, the latter casethen representing epitactic growth on the augite/sodic-pyr-oxene pseudomorph. Furthermore, in sample M 592omphacite only predominates where augite grains arealtered along cleavage planes and fractures from within(Fig. 3b). Jadeite predominates in the former plagioclaselaths, with omphacite and associated pumpellyite generallyonly patchily developed at jadeite grain boundaries. Theoriginal contacts between augite and plagioclase are pre-served, thus mimicking and preserving the original ophitictexture (Fig. 3c, d). Irregular aggregates of jadei-teþomphacite and minor pumpellyite encroach upon theaugite crystal along an irregular reaction front from therock matrix (Fig. 3e), which was presumably plagioclase-rich in the original metabasic rock. Quartz has not beenfound associated with jadeite growth from the augite relicsor the enclosed plagioclase laths. Late blue-green sodic-calcic amphibole and chlorite replace the complex augitepseudomorphs, with amphibole growth again controlled bythe pyroxene structure. The crystallographic c-axes of both

Fig. 3. Back-scattered electron (BSE) images: (a) Matrix of sampleM591, indicating complex irregular intergrowth of constituent miner-als; (b) Augite relic with predominant omphacite alteration (M592);(c, d) Jadeite-dominated replacement of plagioclase mimickingigneous ophitic texture and also encroaching on augite fromrock matrix at the right-hand side of the image in d (M592);(e) Remnants of zoned augite (note variations in BSE brightness inaugite due to primary variations in Mg/Fe2þ ratios – see Fig. 7) withJd-rich replacement of plagioclase inclusion in augite (right-center),JdþOmp replacement surrounding both augite remnants and ompha-cite-dominated replacement from within grain to the left (M592).

Jadeite formation in metabasic rock 221

phases are parallel and amphibole growth occurs preferen-tially in this direction. Augite is preferentially affected, butamphibole may also replace topotactically intergrownsodic pyroxene. A younger jadeiteþ omphaciteþ chloriteassemblage fills the macroscopically evident fracturescross-cutting the rock (Fig. 2d). These can even be seento dissect the ‘‘ophitic’’ augite relics. No obvious interac-tion between these vein fillings and the rest of the rock isevident.

The more bluish macroscopic colour of sample M591 isdue to abundant glaucophane, both as xenomorphic discretecrystals and together with intimately intergrown jadeite andomphacite to constitute the bulk of the felty rock matrix (Fig.2c). TEM analysis (see below) indicates that glaucophane hasto a large extent replaced earlier magmatic augite and sodicpyroxene. Chlorite and hypidiomorphic to xenomorphic epi-dote-group minerals are also present. The back-scatteredelectron microprobe image of Fig. 3a illustrates the pervasivelack of an equilibrium fabric. Grain boundaries are complexand irregular intergrowth textures are wide-spread.Omphacite and jadeite are finely intergrown, but omphacitealso tends to form thin, irregular rims at jadeite crystal mar-gins, suggesting that omphacite replaced earlier jadeite.Glaucophane crystal edges are frayed and show reaction toyounger chlorite and calcic amphibole. Albite also appears tobe a late secondary product filling interstices in the matrix.Titanite, biotite, apatite, chalcopyrite and pyrite are accessoryphases. Only a few grains of quartz have been identified in therock, with no obvious preferred spatial relationship to any ofthe other phases. Only minor relics of magmatic augite , 10mm in size are observed.

Cathodoluminescence microscopy (Fig. 4a–f) revealsthat jadeite replacing ophitic augite and jadeite growingin the matrix of sample M592 show very similar CL char-acteristics. Greenish and reddish to violet hues

predominate, suggesting compositional differences, butthe distribution is very irregular (Fig. 4a, b). The mottledappearance is likely due to the finely intergrown ompha-cite, which should not show optical CL due to the presenceof iron acting as a quencher element. Jadeite in the matrixof sample M591 is much more homogeneous in colour,with greenish hues predominating (Fig. 4c). In contrast,jadeite growing in the veinlets cross-cutting the rock arecoarser, forming interlaced sheaves (Fig. 4d). The CL col-ours are homogeneous, with a violet to blue growth genera-tion clearly discernible from a green one (Fig. 4e, f).Although the two are in places intimately intergrown, thegreen generation appears younger, being concentrated inthe centre of the veins and growing at the terminations ofviolet-blue sheaves (Fig. 4f).

6. Characterization by transmission electronmicroscopy (TEM)

The very fine scale of the disequilibrium and replacementfeatures observed optically with the polarizing microscopesuggests that further more detailed characterization bytransmission electron microscopy (TEM) is called for.

The pyroxene in M591 has been partially replaced byamphibole giving complex microstructures, the abundanceof which suggests that this sample at one time containedmuch more pyroxene than at present. Some of the pyroxeneis certainly C-face-centered, i.e., magmatic augite and/ormetamorphic jadeite. Some of the pyroxene may be P-typeordered omphacite, but it is not possible to be conclusiveabout this on the basis of electron diffraction data alone,because of overlapping reflections from P omphacite andC amphibole. Abundant mixed layer chlorite can beidentified.

Fig. 4. Cathodoluminescence (CL) images. (a) Relict augite crystal (cf., Fig. 2a) with luminescing jadeite-rich laths after plagioclase in augitewith no CL (M592, width of image ¼ 1.3 mm). (b) CL overview of M592 with augite relics (black) and luminescing jadeite-rich matrix(width of image¼ 2.8 mm). (c) Green luminescence of jadeite in M591 rock matrix (width of image¼ 2.8 mm). (d, e) Same view of a jadeite-rich veinlet in M592 under crossed polarizers (d) and in CL mode (e) (width of image¼ 2.8 mm). (f) CL image of jadeite-rich veinlet in M592showing zoned crystals (width of image ¼ 1.3 mm).

222 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

In sample M592, TEM study confirms that C-face-cen-tered augite, C-face-centered jadeite and P omphacite havethe same crystallographic topotactic relationship, i.e.,these pyroxenes give identically oriented diffraction pat-terns. The textures observed between these three pyrox-enes are very similar to those described by Carpenter &Okay (1978) in a low-temperature (� 350 �C) blueschistsample from Turkey, where augite was replaced by aegir-ine-rich omphacite. In the present case, the replacementmechanism of augite by omphacite and/or jadeite alsoappears to have been the formation of veins at an advan-cing front through each crystal. Dark-field images obtainedwith hþk odd reflections from the omphacite regions showfine-scale antiphase domain textures with rather variableantiphase boundary distributions (Fig. 5). In diffractionpatterns from these areas, the hþk odd reflections areweak and slightly diffuse; there appears to be some inten-sity at positions violating the n-glide of the expected P2/nspace group. This observation would be in line with theTurkish example, where Carpenter & Okay (1978) showedthat omphacite originally formed as a metastable disor-dered phase with probable C2/c symmetry. Ordering, andthe C! P lattice transformation, then proceeded irrever-sibly via the metastable intermediate P2/c and P2 struc-tures before attaining stable ordered P2/n symmetry.

In sample M592 an exsolution texture was also observedin some areas of C-type pyroxene that resembles the earlystages of spinodal decomposition (Fig. 6). Without addi-tional analytical data it is not possible to decide whetherthis unusual feature is found in the magmatic augite relicsor in jadeite-rich pyroxene. Both possibilities would repre-sent rare examples. Exsolution is conceivable within anoriginal magmatic augite that has just cooled through thelimb of the quadrilateral pyroxene solvus (e.g., Lindsley,1980). Alternatively, it could be occurring within a jadeite-rich crystal which entered the jadeite-omphacite

miscibility gap during cooling. Exsolution in jadeite-richcrystals during slow cooling of the Dabie ultra-high pres-sure terrain has been reported by Wu et al. (2002), but thetextural relationship between the coexisting phases was notdescribed. In sample M592 the amplitude of the composi-tion modulations is low, as indicated by the low contrast inFig. 6, so that the exsolution process has been caught at anearlier stage than is represented by the spinodal textures inomphacite described by Carpenter (1980).

7. Mineral chemistry

Representative mineral analyses are given in Table 1.

7.1. Pyroxene

The compositions of the magmatic and metamorphic pyrox-enes in samples M591 and M592 are summarized in Fig. 7and 8 according to the classification scheme of Morimotoet al. (1988). Pyroxenes classified as ‘‘Quad’’ are shown inboth figures. The magmatic relics in M592 are augite witha primary magmatic zonation. Homogeneous cores andmore Mg-rich homogeneous rims are separated by anabrupt transition. Augite chemistry is typical for ocean-floor or volcanic-arc basalts (Nisbet & Pearce, 1977).There is a slight tendency towards diopside in Fig. 8.Although some mixed analyses due to the pervasive fine-scale alteration to jadeite and omphacite cannot be ruledout, this shift reflects compositional equilibration towardomphacitic compositions (e.g., Fig. 3), and in the Quad-Jadeite-Aegirine plot of Fig. 8 there is a clear spread ofanalyses away from typical magmatic compositionstoward jadeite and aegirine. This effect is even moreobvious in the few very small , 10 mm relics that couldbe measured in sample M591. In the ‘‘Quad’’ projection of

1 µm

Fig. 5. Dark-field image (hþk odd reflection) showing antiphasedomains in a region of omphacite surrounded by C-centered pyrox-ene (black). Sample M592.

1 µm

Fig. 6. Bright-field image of C-centered pyroxene (i.e., augite orjadeite) with fine-scale spinodal exsolution texture in two orienta-tions. Sample M592.

Jadeite formation in metabasic rock 223

Tab

le1

.R

epre

sen

tati

ve

mic

rop

rob

ean

aly

ses.

Min

eral

Au

ga

Jda

Om

pa

Am

ph

ibo

le

An

aly

sis

11

30

2/1

81

11

53

/82

11

30

6/9

11

33

5/5

51

13

03

/41

13

07

/16

19

71

22

9c/

10

:Gln

b1

13

05

/8:A

ctc

11

30

3/6

:M

g-K

tpb

11

33

5/4

7T

mtb

11

33

5/4

6F

e 2-P

rgb

Sam

ple

M5

91

M5

92

M5

91

M5

92

M5

91

M5

92

M5

91

M5

91

M5

91

M5

92

M5

92

SiO

25

2.0

85

3.1

15

7.5

45

8.3

55

5.5

25

5.2

35

6.0

45

3.1

14

5.8

03

9.1

54

0.1

9T

iO2

0.0

40

.31

0.5

20

.14

0.1

30

.22

0.4

40

.25

0.2

90

.30

0.1

0A

l 2O

31

.99

1.5

11

8.7

02

4.7

89

.69

10

.81

9.4

33

.66

10

.06

17

.08

13

.71

Cr 2

O3

0.0

00

.14

0.0

50

.07

0.0

00

.02

0.0

90

.00

0.0

60

.06

0.0

7F

e 2O

34

.21

0.9

32

.98

0.0

04

.19

4.6

33

.02

0.0

03

.31

4.1

23

.64

FeO

6.4

54

.92

1.0

70

.77

2.8

83

.94

10

.21

11

.99

13

.01

16

.35

18

.36

Mg

O1

0.6

61

8.5

52

.39

0.1

67

.35

5.3

99

.26

14

.69

10

.84

5.6

16

.27

Mn

O0

.42

0.1

50

.00

0.0

40

.21

0.2

00

.30

0.3

40

.32

0.1

90

.30

CaO

22

.18

19

.47

4.0

20

.54

12

.43

10

.87

1.0

81

1.4

19

.36

8.1

49

.97

Na 2

O1

.69

0.1

51

2.6

81

4.5

97

.42

8.3

27

.02

1.8

94

.23

5.5

24

.22

K2O

0.0

00

.00

0.0

00

.00

0.0

00

.00

0.0

40

.22

0.3

30

.28

0.3

7T

ota

l9

9.7

29

9.2

49

9.9

49

9.4

49

9.8

29

9.6

29

6.9

39

7.5

69

7.6

19

6.8

09

7.2

0S

i1

.95

51

.94

91

.99

11

.98

31

.99

31

.99

2S

i7

.88

27

.68

76

.77

95

.99

86

.20

7A

lIV0

.04

50

.05

10

.00

90

.01

70

.00

70

.00

8A

lIV0

.11

80

.31

31

.22

12

.00

21

.79

32

.00

02

.00

02

.00

02

.00

02

.00

02

.00

08

.00

08

.00

08

.00

08

.00

08

.00

0A

lVI

0.0

43

0.0

15

0.7

54

0.9

76

0.4

03

0.4

52

AlV

I1

.44

60

.31

10

.53

41

.08

20

.70

2T

i0

.00

10

.00

90

.01

40

.00

40

.00

30

.00

6T

i0

.04

60

.02

70

.03

20

.03

40

.01

2C

r0

.00

00

.00

40

.00

10

.00

20

.00

00

.00

1C

r0

.01

00

.00

00

.00

70

.00

70

.00

9F

e3þ

0.1

22

0.0

26

0.0

78

0.0

00

0.1

13

0.1

26

Fe3þ

0.3

20

0.0

00

0.3

72

0.4

75

0.4

23

Mg

0.5

96

1.0

15

0.1

23

0.0

08

0.3

93

0.2

90

Mg

1.9

41

3.1

69

2.3

92

1.2

81

1.4

44

Fe2þ

0.2

08

0.1

51

0.0

31

0.0

22

0.0

87

0.1

19

Fe2þ

1.2

01

1.4

51

1.6

24

2.0

95

2.3

71

Mn

0.0

13

0.0

05

0.0

00

0.0

01

0.0

06

0.0

06

Mn

0.0

36

0.0

42

0.0

40

0.0

25

0.0

39

Ca

0.8

92

0.7

66

0.1

49

0.0

20

0.4

78

0.4

20

5.0

00

5.0

00

5.0

00

5.0

00

5.0

00

Na

0.1

23

0.0

11

0.8

51

0.9

68

0.5

16

0.5

82

Ca

0.1

63

1.7

70

1.4

84

1.3

37

1.6

50

2.0

00

2.0

00

2.0

00

2.0

00

2.0

00

2.0

00

NaB

1.8

37

0.2

30

0.5

16

0.6

63

0.3

50

2.0

00

2.0

00

2.0

00

2.0

00

2.0

00

NaA

0.0

78

0.3

00

0.6

98

0.9

77

0.9

13

K0

.00

80

.04

00

.06

20

.05

40

.07

20

.08

60

.34

00

.76

01

.03

10

.98

5

224 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

Tab

le1

.

Min

eral

Czo

dE

pid

ote

dC

hlo

rite

eB

ioti

tef

Pm

pg

An

aly

sis

11

30

2/1

61

13

02

/17

11

33

5/7

61

13

06

/3A

l-p

oo

r1

13

91

/2A

l-ri

ch1

11

54

/1A

l-p

oo

r1

11

51

/9A

l-ri

ch1

13

06

/10

11

49

9/3

2S

amp

leM

59

1M

59

1M

59

2M

59

1M

59

1M

59

2M

59

2M

59

1M

59

2S

iO2

39

.00

38

.06

38

.13

28

.37

25

.40

28

.00

25

.29

37

.61

37

.35

TiO

20

.08

0.0

60

.25

0.0

30

.03

0.0

10

.08

0.3

00

.03

Al 2

O3

32

.85

27

.98

26

.47

17

.22

20

.73

17

.08

20

.47

14

.09

26

.79

Cr 2

O3

0.0

00

.00

0.0

00

.08

0.0

00

.00

0.0

20

.00

Fe 2

O3

1.0

97

.59

9.3

0F

eO2

2.4

52

3.3

52

3.6

82

5.4

11

8.9

64

.47

Mg

O0

.00

0.0

00

.02

18

.79

16

.13

17

.18

15

.26

13

.43

1.4

6M

nO

0.2

50

.17

0.0

30

.38

0.3

40

.26

0.3

40

.12

0.0

9C

aO2

4.5

92

4.4

22

4.3

40

.04

0.1

70

.14

0.1

30

.29

22

.07

Na 2

O0

.03

0.0

40

.04

0.0

00

.25

0.8

6K

2O

0.0

20

.00

0.0

50

.00

9.4

70

.02

To

tal

97

.86

98

.28

98

.54

87

.40

86

.19

86

.45

86

.99

94

.52

93

.14

Si

2.9

71

2.9

63

2.9

79

Si

5.9

05

5.4

05

5.9

36

5.3

92

Si

2.8

90

Si

2.9

80

AlIV

0.0

29

0.0

37

0.0

22

AlIV

2.0

95

2.5

95

2.0

64

2.6

08

AlIV

1.1

10

Ti

0.0

02

3.0

00

3.0

00

3.0

00

8.0

00

8.0

00

8.0

00

8.0

00

4.0

00

Al

2.5

19

AlV

I2

.92

12

.53

12

.41

6A

lVI

2.1

30

2.6

05

2.2

04

2.5

35

AlV

I0

.16

6M

g0

.17

4C

r0

.00

00

.00

00

.00

0T

i0

.00

50

.00

50

.00

10

.01

3T

i0

.01

7F

e2þ

0.2

98

Fe3þ

0.0

63

0.4

45

0.5

46

Cr

0.0

13

0.0

00

0.0

00

0.0

03

Cr

0.0

00

Mn

0.0

06

2.9

84

2.9

76

2.9

62

Mg

5.8

31

5.1

18

5.4

31

4.8

50

Mg

1.5

38

5.9

79

Ti

0.0

04

0.0

04

0.0

14

Fe2þ

3.9

08

4.1

55

4.1

99

4.5

30

Fe2þ

1.2

18

Ca

1.8

86

Mg

0.0

00

0.0

00

0.0

03

Mn

0.0

67

0.0

61

0.0

47

0.0

62

Mn

0.0

08

Na

0.1

34

Mn

0.0

16

0.0

11

0.0

02

Ca

0.0

09

0.0

38

0.0

33

0.0

29

2.9

48

K0

.00

2C

a2

.00

82

.03

72

.03

7N

a0

.01

20

.01

60

.01

60

.00

0C

a0

.02

42

.02

22

.02

82

.05

22

.05

6K

0.0

05

0.0

00

0.0

14

0.0

00

Na

0.0

37

11

.98

01

1.9

98

11

.94

41

2.0

22

K0

.92

80

.98

9

Act

,ac

tin

oli

te;

Au

g,

aug

ite;

Czo

,cl

ino

zois

ite;

Fe 2

-Prg

,fe

rro

par

gas

ite;

Gln

,g

lau

cop

han

e;Jd

,ja

dei

te;

Mg

-Ktp

,m

agn

esio

kat

op

ho

rite

;O

mp

,o

mp

hac

ite;

Pm

p,

pu

mp

elly

ite;

Tm

t,ta

ram

ite.

No

rmal

izat

ion

pro

ced

ure

sfo

rfo

rmu

laca

lcu

lati

on

:a6

ox

yg

ens

and

4ca

tio

ns;

b2

3o

xy

gen

san

d1

3ca

tio

ns

wit

ho

utC

a,N

a,K

;c

23

ox

yg

ens

and

Feto

t¼

Fe2þ

;d

12

.5o

xy

gen

san

dF

etot¼

Fe3þ

;e

28

ox

yg

ens

and

Feto

t¼

Fe2þ

;f 2

2o

xy

gen

san

dF

etot¼

Fe2þ

;g8

cati

on

sw

ith

ou

th

yd

rog

en.

(Continued)

Jadeite formation in metabasic rock 225

Fig. 7 these plot as diopside compositions, but in Fig. 8they are seen to be quite Na- and Al-rich.

Possible problems of mixed analyses are also evidentfrom Fig. 8, and are a logical consequence of the intimateintergrowths observed between jadeite and omphacite(Fig. 3, 9, 10). Although there are compositional clustersconforming to jadeite and omphacite, intermediate ana-lyses are found. Pyroxenes reach Jd99 in sample M592,but only Jd79 in M591. No systematic compositional trendscould be identified for jadeite and omphacite occurring as

topotactic alteration products of augite, as matrix compo-nents or as vein fillings. Figures 9 and 10 indicate that withtime irregular jadeite/omphacite intergrowths may recrys-tallize to take on a quasi-equilibrium fabric with regulargrain boundaries. Analyses of coexisting sodic pyroxenepairs from settings such as in Fig. 10 probably represent theclosest approach to local equilibrium available. Two such‘‘quasi-equilibrium’’ pairs are indicated in Fig. 8. ForM591 these are two omphacite compositions correspond-ing to Jd72Di15Hd5Aeg8 and Jd43Di35Hd11Aeg11. ForM592 a jadeite s.s. with composition Jd81Di9Hd2Aeg8

coexists with an omphacite corresponding toJd45Di34Hd14Aeg7 (Fig. 8). Figure 9 also demonstratesthat although the contacts between augite relics and thejadeite-rich lamellae are generally sharp, small embay-ments of omphacite are seen to grow locally on bothsides of the contact.

If equilibrium can be assumed, coexisting sodic pyrox-enes offer the possibility of using the solvus to obtaininformation on the temperatures of equilibration. In theirdetailed study of coexisting sodic pyroxenes in jadeite jadesof the Sierra del Convento in Cuba, Garcıa-Casco et al.(2009) showed that the sodic pyroxene populations therefollow linear trends from almost pure jadeite to fictive end-member compositions between Di90Hd5Aeg5 andDi80Hd10Aeg10. These trends are shown in projection inthe Quad-Jd-Aeg diagram of Fig. 8. Although there is con-siderable scatter of sodic pyroxene compositions in Fig. 8,especially in jadeite, a similar tendency appears indicated inFig. 8, so that in Fig. 11 the two ‘‘quasi-equilibrium’’ sodicpyroxene pairs of Fig. 8 are compared to the pseudobinaryT-X sections calculated by Garcıa-Casco et al. (2009).Notwithstanding the considerable approximations involved,the M592 pair suggests equilibration temperatures of475–500 �C. The M591 omphacite suggests 425–450 �C,whereas the coexisting jadeite-rich pyroxene offers no plau-sible fit to any of the pseudosections. This composition does,however, also deviate considerably from the assumed linearJd to Di-Hd-Aeg trend (Fig. 8).

7.2. Other minerals

As a general feature, amphibole dominates over pyroxenein M591, but is only a minor constituent in M592. Followingthe classification scheme of Leake et al. (1997), the amphi-boles of sample M591 are Al-rich glaucophane, i.e., Al-richer than the ‘‘crossite’’ of classical usage. Occasionalrims of magnesiokatophorite or edenite (i.e., straddling theboundary of the calcic to sodic-calcic amphibole groups) areobserved. Secondary actinolite replaces glaucophane atfrayed crystal edges. Late amphibole replacing augite insample M592 is taramite of the sodic-calcic group, but Ca-contents lie between 1.3 and –1.5 apfu, and gradation to aferroparagasite composition is observed.

Defining clinozoisite as Ca2Al3Si3O12(OH) and epidoteas Ca2Al2Fe3þSi3O12(OH) (e.g., Armbruster et al., 2006),then the zoned epidote-group crystals of sample M591 areclinozoisite with epidote contents ranging from 6 mol% in

Diopside Hedenbergite

Augite

Pigeonite

Clinoenstatite Clinoferrosilite

M 591M 592

Fe2Si2O6Mg2Si2O6

Ca2Si2O6

rimscores

Fig. 7. Classification of relict magmatic pyroxenes according toMorimoto et al. (1988). Note the distinct core-rim zonation in mag-matic pyroxenes of sample M592.

100

90

80

70

60

50

40

30

20

10

100

90

80

70

60

50

40

30

20

10

100 90 80 70 60 50 40 30 20 10

Jadeite Aegirine

Aegirine-AugiteOmphacite

AegirineJadeite

Quad

M 591M 592

AB C

Fig. 8. Classification of all analyzed pyroxenes according toMorimoto et al. (1988). Connected large open circles: ‘‘quasi-equi-librium’’ sodic-pyroxene pair in M591. Connected large filledsquares: ‘‘quasi-equilibrium’’ jadeite-omphacite pair in M592.Lines A, B, C represent, as discussed in the text, the projections oflinear compositional trends from end-member jadeite toDi90Hd5Aeg5, Di80Hd10Aeg10 and Di60Hd20Aeg20, respectively.

226 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

the cores to 45 % in the rims. The idiomorphic epidote-group minerals in sample M592 are very homogeneous,with epidote contents ranging from 54 to 68 mol%.

Considering that amphibole and pumpellyite (seebelow) compositions are generally quite homogeneous ineach sample, it is noteworthy that the compositions of theferromagnesian mineral chlorite are not. Different genera-tions of growth or domain equilibria could be a reason. Asoutlined in the following section, there is clear evidencethat the protolith was affected by spilitization prior tosubduction, and a first generation of chlorite could alreadyhave formed at this time. TEM observations also suggestthat mixed-layer chlorite is abundant, so that EMPA resultsof microscopically identified chlorite should be interpretedwith some caution. Nevertheless, in almost all cases AlIV�(AlVI þ 2Ti þ Cr), so that octahedral vacancies or Fe3þ

contents appear to be negligible (e.g., Zane et al., 1998).

The Tschermak’s substitution accounts for most of thecompositional variability observed (Fig. 12). No systema-tic trends with respect to mode of occurrence (associatedminerals, matrix vs. late alteration of glaucophane, etc.)could be found in either sample. Chlorite is Mg-dominant,as expected for metabasic rocks (Zane et al., 1998), withchlorites from sample M591 clustering at higher Mg/(MgþFe) values than those from M592. This differenceis actually the opposite of what might be expected from thebulk Fe and Mg contents of the two samples.

Pumpellyite is found only in sample M592, both in therock matrix and in the pseudomorphs after ophitic augite,where it is closely associated with omphacite. The composi-tion is homogeneous and Al-rich (Table 1). Following Hatertet al. (2007,2008), a typical mineral formula can be written as(Ca1.89Na0.13)2.02(Al0.52Fe2þ

0.30Mg0.17Mn0.01)1.00Al2.0(SiO4)(Si2O7)(OH)2.45�0.55H2O.

Fig. 9. Element distribution maps obtained in beam-scan mode (sample M592, image dimensions 82� 82 mm; 256� 256 pixels; colour scalebar indicates total counts per pixel). Detail of augite grain in contact with matrix and with a plagioclase lamella pseudomorph (lower left).Note irregular jadeite/omphacite intergrowth (cf. Fig. 10).

Jadeite formation in metabasic rock 227

Al always slightly exceeds 50 % of the M(1) site, thusdefining the composition unequivocally as pumpellyite-(Al)(see discussion in Hatert et al. 2008). Within analyticalerror, no Fe3þ is required to maintain pumpellyite stoichio-metry. The composition and homogeneity of pumpellyiteare significant features for a metabasic rock. Although theformation of Al-rich pumpellyite seems to be favouredtoward higher pressures into the blueschist facies and tohigher temperatures transitional to the greenschist facies(Cortesogono et al., 1984, and references therein), thesetrends are often obscured by the local chemical environ-ment. Relict plagioclase as a precursor phase offers such afavourable local environment for the growth of Al-richpumpellyite (Cortesogono et al., 1984, and referencestherein). In sample M592 Al-rich pumpellyite is found insuch plagioclase pseudomorphs, but its occurrence there and

elsewhere in the rock appears to be even more intimatelytied to omphacite, suggesting that pumpellyite may be analteration product of this phase throughout the rock. Such apossibility places critical constraints on the metamorphicevolution of the rock as discussed further below,

Plagioclase in both samples is almost pure albite(Ab99An01). Rare small flakes of biotite (,10 mm insize) in sample M591 are slightly Mg-dominant phlogo-pite-annite solid solutions displaced approximately 20–25mol% toward siderophyllite-eastonite (Rieder et al., 1998).

8. Bulk rock composition

Available chemical data on samples M591 and M592 arelisted in Table 2. Schneider et al. (2004) have presented

Fig. 10. Element distribution maps obtained in beam-scan mode (sample M592, image dimensions 82 � 82 mm; 256 � 256 pixels; colourscale bar indicates total counts per pixel). Detail of jadeite-omphacite intergrowth with beginning development of regular grain boundariessuggestive of approach toward an equilibrium fabric (cf. Fig. 9).

228 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

much more extensive major, trace element, REE as well asNd and Sr isotope data for several metabasites (eclogites)from the Gavilanes unit, from which samples M591 andM592 are considered to have been derived. Schneider et al.(2004) conclude that metabasites in the Gavilanes unit areof two distinct types. Those found in a serpentinite-mel-ange context derive from a MORB-type protolith. Othersamples from intercalations within metasedimentary rocksappear to possess a distinctly different calc-alkaline affi-nity. The Al2O3 contents of samples M591 and M592compare well with the 14.84 wt% of the MORB-typesample studied by Schneider et al. (2004) and are distinctlylower than the Al2O3 contents (20.12 and 20.16 wt%) ofthe calc-alkalic types, suggesting a MORB-type protolithfor the present samples as well.

In Fig. 13 the bulk rock compositions of M591 andM592 are compared to average N-MORB, using an isoconplot adapted from Grant (1986). For both samples themajor oxides Al2O3, SiO2, MgO and FeO* all lie on orclose to a 1:1 isocon, i.e. suggesting relatively constantmass during any modification of a presumed N-MORBprotolith. In both samples Na2O is highly enriched withrespect to N-MORB and CaO is clearly lower, as are TiO2,P2O5, and K2O. There is considerable scatter in some trace

elements, but Zr, Cr, and V conform to the isocon; Ba isenriched with respect to N-MORB, but contents are stilllow. All these features could be the result of hydrothermalocean-floor alteration processes (e.g., Staudigel, 2003),fluid-driven metasomatic exchange during subductionand high-pressure metamorphism (e.g., Sorensen et al.,1997; Bebout, 2007), or a combination of both.Analyzing these trends with the data at hand is not straight-forward. Following Sorensen et al. (1997), we note thatalthough the Ba contents and Ba/TiO2 ratios (Table 2, Fig.13) are somewhat higher than in pristine N-MORB, thisenrichment is still in the general range of subduction-zoneMORB protolith and can be explained by an at most 1 %contamination of a sediment-equilibrated subduction-zonefluid (Sorensen et al., 1997).

Although K2O should be positively correlated with Baduring low-temperature hydrothermal alteration of oceaniccrust (Staudigel, 2003) or syn-subduction metasomatic pro-cesses (Sorensen et al., 1997; Bebout, 2007), this is notobserved in samples M591 and M592. However, potassiumis leached from basalts by hydrothermal high-temperaturealteration reactions (Staudigel, 2003), and we suggest thatthe large increase in Na2O and decrease in CaO is a dis-tinctive chemical characteristic produced by classical pro-cesses of spilitization of a MORB protolith (e.g., Graham,1976; Fettes & Desmons, 2007), where magmatic plagio-clase is albitized and the ferromagnesian minerals are lar-gely replaced by chlorite. The magmatic fabric of the rock ispreserved. It is also important to note that the increase inNa2O is not accompanied by a similarly significant increasein Al2O3, so that no net gain of a ‘‘jadeite component’’ withrespect to N-MORB can be seen in M591 and M592. Theeffect of the thin jadeitite veinlets on the bulk compositionof the rock appears to be negligible.

We conclude from the data on bulk chemistry that theprotolith of M591 and M592 was spilitized chlorite-richocean-floor basalt, with abundant albitized plagioclaseavailable as a precursor mineral for jadeite formation duringsubduction metamorphism via local domain reequilibration.Although the assumption of a spilite protolith subjected toessentially isochemical recrystallization provides a bulk-compositional template for exploring isochemical pres-sure-temperature (pseudosection) phase diagrams, this

T (°C)

M592

M591

C c2/ C c2/P n2/Diopside Omp Jd

500

700

600

800

400

300

0.00 1.000.800.600.400.20

XJd

DiopsideJadeite(A)

(B)(C)

?

Fig. 11. Comparison of the two ‘‘quasi-equilibrium’’ pyroxene pairsof Fig. 8 with phase relations calculated for the binary join Di-Jd(Green et al., 2007) and various T-X locations of the omphacite-jadeite solvus limbs as calculated by Garcıa-Casco et al. (2009) forpseudobinary trends Di0.9Hd0.05Aeg0.05 (curves A), Di0.8Hd0.1Aeg0.1

(curves B) and Di0.6Hd0.2Aeg0.2 (curves C). A, B and C correspondto the projected trends shown in Fig. 8. A and B taken from Fig. 7 ofGarcıa-Casco et al. (2009), C interpolated from the results of Garcıa-Casco et al. (2009) and the phase relations Jd-Di-Aeg of Green et al.(2007) at 500 �C.

0.40

0.45

0.50

0.55

0.60

0.65

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70Al(IV)- 2 (apfu)

Mg

/(M

g+

Fe

)

M591

M592

Fig. 12. Compositional variation of chlorites in terms of Mg/(MgþFe) vs. the Tschermak’s substitution (referred to ideal(Mg,Fe2þ)6Al2(Si6Al2)O20(OH)16).

Jadeite formation in metabasic rock 229

approach has not been successful. Abundant partially alteredrelict phases at all stages of metamorphic development,especially of magmatic augite in M592, make it difficult toquantify effective bulk compositions during pressure-tem-perature evolution. If the bulk rock compositions in Table 2are used without modification, Na2O-enrichment appears tolead to a wide P-T stability field for two pyroxenes (ompha-cite and an aegirine-rich sodic pyroxene) for a hypotheti-cally completely equilibrated rock.

9. Discussion

9.1. Summary of inferred phase relationships

Considering the many stranded reaction relationships withtheir relict reactants and the evidence for domain equili-bration as a driving force for recrystallization, definingpossible equilibrium assemblages and reaction relation-ships between them is difficult and in part speculative. Atbest, a sequence of phase growth can be outlined for eachsample. In M592 the augite relics appear to representdomains where local equilibrium predominated. Jadeite isa dominant phase that replaces both augite and especially

former plagioclase. Although omphacite growth may havebeen in part coeval with jadeite, it is generally concentratedin the interstices between jadeite grains and forms rims onthese. The internal omphacite-dominated replacement ofaugite grains is also inferred to post-date the early jadeite-dominated replacement front encroaching on augite fromthe rock matrix. In the latter, finely and irregularly inter-grown jadeite, omphacite, pumpellyite, chlorite and epi-dote are found together, but these phases can not representan equilibrium assemblage. As in the augite relics, much ofthe omphacite appears to be replacing jadeite. The cross-cutting idiomorphic epidote-group minerals and the chlor-ite-epidote clusters grew at a later stage. Thus an earlyjadeite þ chlorite � omphacite growth stage appears indi-cated, although chlorite should already have formed in theoriginal spilite. The occurrence of pumpellyite is enig-matic. Although it is found intimately intergrown withother minerals of the rock matrix, the thermal history ofM592 outlined in the next section indicates that the P-Tstability field of this phase must have been overstepped byat least 100 �C, so that we are led to consider pumpellyiteto be a late phase formed mainly from omphacite on theretrograde path. Late albite could also belong to this retro-grade growth stage. The Al-rich amphiboles taramite andferropargasite and associated chlorite replacing augiterelics probably represent recrystallization at the highesttemperatures reached, suggestive of the epidote-amphibo-lite facies. No formation of any amphibole until this verylate stage in metamorphic development is observed inmetabasite M592.

Table 2. Major (wt%) and selected minor (ppm) element concentra-tions for samples M591 and M592.

M591 M592

SiO2 49.65 49.24TiO2 1.04 1.23Al2O3 15.01 14.84Fe2O3 3.16 2.92FeO 6.12 6.46MnO 0.21 0.15MgO 6.62 7.60CaO 9.07 9.07Na2O 6.11 6.38K2O 0.05 0.01P2O5 0.09 0.00H2Oþ 2.04 2.39CO2 0.26 0.14Sum 99.43 100.43Ba 47 64Co 69 69Cr 126 277Cu 70 80Ga 15 12Nb 5a 3a

Ni 52 106Pb 9a 4a

Rb 9a 12Sn 6a 9a

Sr 194 64V 282 258Y 34 37Zn 33 76Zr 69 72

Note: aValue within uncertainty limits.

100

10050

50

00

5MgO

5CaO

5 Al O2 3

SiO2

10Na O2

3FeO*

200K O2

100P O2 5

20TiO2

N-M

OR

B

M 591

M 592

Ba

Cr/5

Cu

Ni/2

lossNi/2

Cr/5

Sr/2

Sr/2

V/10

2Y

Co

Zn ZnZr

jadeite-bearing metabasalts

gain

2 x N-MORB

cons

tant m

ass

0.5

x N

-MO

RB

Fig. 13. Isocon diagram (Grant, 1986) comparing element concen-trations of samples M591 and M592 (Table 2, anhydrous basis) to N-MORB. Major-element oxides are given in wt% oxide and traceelements in ppm. Scaling factors are indicated. N-MORB data aretaken from Basaltic Volcanism Study Project (1981) and Sun &McDonough (1989).

230 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

In sample M591, jadeite þ chlorite � omphacite alsoappear to have dominated primary metamorphic recrystal-lization from the presumed spilitized magmatic protolith.Omphacite may in part be coeval with but definitelyreplaces jadeite. Recrystallization is much more pervasivethan in sample M592, and TEM observations indicate thatearly sodic pyroxene and augite were later extensivelyaltered to glaucophane. Clinozoisite overgrows the feltyglaucophane þ jadeite þ chlorite matrix. Minor albite,actinolite and a second generation of chlorite are clearlyyounger, replacing glaucophane and sodic pyroxene.

The jadeite þ omphacite þ chlorite assemblages in thelate veinlets are the youngest features observed.

In their early petrographic account, Millan & Somin(1981) describe a rock composed essentially of single-phase jadeite with isolated grains of clinozoisite, lawsoniteand albite. Such assemblages were not encountered in thisstudy.

9.2. Metamorphic evolution of the jadeite-bearingmetabasites

Although the Gavilanes unit is a heterogeneous mega-melange, a number of P-T paths are available for variousblocks in the melange that help to constrain possible P-Ttrajectories for samples M591 and M592. These data aresummarized in Fig. 14. Maximum reported pressures forthe various rocks vary between 15 and 25 kbar. Maximumreported temperatures vary between � 500 and 660 �C.Prograde trajectories are poorly defined, but the P-T pathsare of ‘‘hair-pin’’, clockwise type, so that progradegeotherms between 5 and 10�/km are indicated for manyof them. These lie in the stability field of jadeite in thepresence of quartz (Fig. 14). The various members of themelange were collected during exhumation, and the exhu-mation paths can be seen to converge to a common trend(Fig. 14) that shows cooling during depressurization alonga geotherm of � 12 �/km. The ‘‘coolest’’ path is indicatedby rare deerite-bearing quartzites of the metasedimentarymelange matrix (Grevel et al., 2006). The ‘‘warmest’’ pathhas been deduced for an eclogite from a serpentinite mel-ange by Garcıa-Casco et al. (2006). Thus exhumationduring active subduction is indicated (Schneider et al.,2004; Grevel et al., 2006; Garcıa-Casco et al., 2006, 2008).

The above regional constraints allow some reasonableconclusions to be reached for the P-T development of M591and M592. Jadeite-omphacite equilibration (Fig. 11) pro-vides information on minimum temperatures reached. Atemperature of 475–500 �C is suggested for sample M592,although it is not clear whether such equilibration occurredduring the prograde or the retrograde path of the rock. Forsample M591 a somewhat lower temperature of 425–450�C is obtained, but only the omphacite provides a reason-ably fit to the sodic pyroxene solvus. Figure 14 indicatesthat the intergrown pumpellyite in the matrix of sampleM592 could be stable up to 400 �C on the basis of the end-member experiments of Schiffman & Liou (1980), butavailable pseudosection calculations actually indicate that

for metabasite compositions the stability field of pumpel-lyite is even more restricted. The calculations of Willneret al. (2009) and Baziotis et al. (2009), for instance, showfor various different metabasite compositions that duringprograde development pumpellyite should normally reactout between 300 to 350 �C and 5 to 10 kbar. Consideringthat Al-rich amphiboles grew in M592, such conditionsseem unrealistically low, and we consider pumpellyite tobe a phase that crystallized during the retrograde path (seeabove).

As noted in the previous section, attempts to calculateisochemical P-T diagrams, i.e. pseudosections, were notsuccessful, mainly because the effective bulk compositioncould not be properly constrained. Nevertheless, classicalmetamorphic-facies field distributions (e.g., Evans, 1990)show that the P-T trajectories of Fig. 14 pass through acritical region between 8–12 kbar and 400–500 �C whereepidote-blueschist, epidote-amphibolite and greenschist-facies fields are in close proximity. Al-rich amphibolesfound in both samples suggest that conditions of the epi-dote-amphibolite facies were reached. Depending on bulkand mineral compositions, the actual phase field bound-aries can be quite variable in this P-T region (e.g., Evans,1990; Baziotis et al., 2009). The trend of the P-T trajec-tories in Fig. 14 allows the reasonable conclusion that theprograde development of M591 and M592 occurred aboveor at last very close to the Ab ¼ Jd þ Qtz reaction curve,although this conclusion cannot be unequivocally corrobo-rated from petrographic observation. Minor quartz is pre-sent in the matrix of both samples, but its relationship tojadeite is unclear. No quartz is seen as a reaction product inthe plagioclase pseudomorphs, and SiO2 must have been

Fig. 14. Summary of available P-T paths for samples from theGavilanes unit from Grevel (2000), Grevel et al. (2006) andSchneider et al. (2004). Pyroxene reaction curves from Waterwieseet al. (1995) and Gasparik & Lindsley (1980). Pumpellyite-out curvefor Ca4Al5MgSi6O21(OH)7 composition from Schiffman & Liou(1980). Lawsonite-out curve from Harley & Carswell (1990),based on Chatterjee et al. (1984).

Jadeite formation in metabasic rock 231

transported to and consumed in the rock matrix by othermineral reactions, perhaps those forming chlorite and epi-dote. Early jadeite þ chlorite � omphacite growth isassumed for both samples during this prograde path, withthe formation of jadeite presumably facilitated by the pre-sence of albitized plagioclase. Sample M591 equilibratedto a large extent in the epidote-blueschist facies beforepassing through the epidote-amphibolite and then thegreenschist facies. Sample M592 shows no evidence ofprograde amphibole development, be it sodic or calcic,until epidote-amphibolite conditions are reached.However, the formation of sodic-calcic taramite therecalls for conditions close to the blueschist facies.

Although lawsonite was not observed in our samples(cf., Millan & Somin, 1981), Fig. 14 indicates that thestability field for this phase is traversed by most progradeP-T trajectories and encountered by many retrograde P-Tpaths between 500 and 400 �C. Given appropriate bulkcompositions, lawsonite-bearing assemblages are possible.

10. Conclusions

On the basis of the available data, we suggest the followingevolutionary scenario for samples like M591 and M592:

(1) The protolith corresponds to MORB, but spilitizationis indicated by a clear increase in Na2O, coupled with adepletion of CaO in the rock (e.g., Graham, 1976).Spilitization is a complex process, but albitization ofigneous plagioclase is a characteristic feature. Augite relicsof 1.5 mm are abundant in M592, so that original grainsizes corresponding to a fine-grained gabbro appear likely.Ophitic texture is common.(2) The rocks were not deformed during subduction, andjadeite grew readily at the expense of albitized plagioclase.The enrichment of Na2O and depletion in CaO during spili-tization may also have provided an effective local bulkcomposition between abundant augite relics that was con-ducive to jadeite nucleation and growth. Domain equilibra-tion appears to have dominated, and ophitic textures werepreserved. Jadeite and minor omphacite crystallized in for-mer plagioclase laths enclosed in augite in epitactic relation-ship to the host augite. Jadeite and more abundant omphacitegrowing from plagioclase in the rock matrix encroachedupon augite grains from the outside. Omphacite dominatedwhen augite was replaced along cleavage planes or fracturesfrom within, as described by Carpenter & Okay (1978). Allthree pyroxenes are generally in crystallographical (optical)continuity, either as a result of the direct topotactic replace-ment of augite, or because of the epitactic overgrowth onaugite of pyroxene replacing plagioclase.(3) At increasing temperatures, more omphacite replacedearlier jadeite along grain boundaries. Where sufficientwater became available, reequilibration to amphibole-richassemblages as in M591 was possible; in other cases thepyroxene-rich assemblages persisted.

(4) The samples were exhumed along steep P-T trajec-tories of approximately 12 �/km during active subduction.This may be the reason for the excellent preservation ofjadeite and omphacite.(5) During exhumation, brittle fractures cross-cutting therocks were filled by jadeite-rich assemblages. These displaythe typical features of open-system precipitation from anaqueous fluid known from well-studied jadeite jade occur-rences (e.g., Harlow & Sorensen, 2005; Sorensen et al.,2006, 2010; Harlow et al., 2007). Cathodoluminescenceimaging indicates the growth of more than one generationof jadeite in these veins. These jadeite jade veinlets suggestthat a pervasive fluid rich in jadeite-component must havebeen available, but the timing of their formation and thesource of the fluids is unclear. The jadeitite-filled brittlefractures cut through the rock matrix as well as the augitepseudomorphs with no obvious interaction between the veinmaterial and the enclosing rock. There are no obvious com-positional gradients in the rock toward the veins. Withoutadditional information from REE spectra and trace elementdistributions it is not possible to decide whether the fluidswere derived from a completely external source, wereexpelled from the rock itself at near-peak conditions, orwere ‘‘re-injected’’ at lower temperatures from an aqueousfluid equilibrated at higher temperatures with the metaba-sites themselves.

We conclude that aqueous fluids capable of precipitat-ing jadeite jade were present during exhumation of theGavilanes unit of the Escambray Massif, but we couldnot find any jadeitite veinlets exceeding 1–2 mm in thick-ness. The interpretation of Millan & Somin (1981) thatmetabasic rocks were in the process of being transformedto jadeite jade reflects domain equilibration and enhancedearly jadeite growth in a spilitized, Na2O-enriched meta-basic rock. Even if the enrichment in Na2O could lead tostabilization of omphacite and aegirine-rich jadeitite insuch a metabasic rock after reaching full equilibrium, apossibility that could not be unequivocally tested in thisstudy, there was clearly no net enrichment of jadeite com-ponent in the bulk composition, and no trend from a meta-basic toward a jadeite jade composition.

Acknowledgements: We thank H.-J. Bernhardt andR. Neuser, Bochum, for invaluable help with the electronmicroprobe and with hot-cathode cathodoluminescencemicroscopy, respectively. Arne Willner gave decisive inputon the ‘‘pumpellyite enigma’’. We are also grateful to TadaoNishiyama, Antonio Garcıa-Casco and George Harlow fortheir thoughtful reviews that helped us clarify some impor-tant points and improve the presentation as a whole.

References

Armbruster, T., Bonazzi, P., Akasaka, M., Bermanec, V., Chopin, C.,

Giere, R., Heuss-Assbichler, S., Liebscher, A., Menchetti, S.,

Pan, Y., Pasero, M. (2006): Recommended nomenclature of

epidote-group minerals. Eur. J. Mineral., 18, 551–567.

232 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

Auzende, A.-L., Devouard, B., Guillot, S., Daniel, I., Baronnet, A.,

Lardeaux, J.-M. (2002): Serpentinites from central Cuba: petrol-

ogy and HRTEM study. Eur. J. Mineral., 14, 905–914.

Baese, R., Maresch, W.V., Schenk, V., Schertl, H.-P. (2010):

Geochemistry of jadeitites and jadeite-lawsonite rocks from

the Rio San Juan Complex (northern Dominican Republic):

Interaction between jadeitite-forming fluids and meta-

morphic rocks within a subduction channel. 88th Annual

Meeting German Mineralogical Society, CD-ROM, Abstract

09-T06, 119.

Basaltic Volcanism Study Project (1981): Basaltic volcanism on the

terrestrial planets. Pergamon Press Inc., New York, 1286 p.

Baziotis, I., Proyer, A., Mposkos, E. (2009): High-pressure/low-tem-

perature metamorphism of basalts in Lavrion (Greece): implica-

tions for the preservation of peak metamorphic assemblages in

blueschists and greenschists. Eur. J. Mineral., 21, 133–148.

Bebout, G.E. (2007): Metamorphic chemical geodynamics of sub-

duction zones. Earth Planet. Sci. Lett., 260, 373–393.

Black, P.M. (1974): Mineralogy of New Caledonian metamorphic

rocks III. Pyroxenes. Contrib. Mineral. Petrol., 45, 281–288.

Cardenas-Parraga, J., Garcıa-Casco, A., Nunez-Cambra, K.,

Rodrıguez-Vega, A., Blanco-Quintero, I.F., Harlow, G.E.,

Lazaro, C. (2010): Jadeitite jade occurrence from the Sierra del

Convento melange (eastern Cuba). Bol. Sociedad Geol. Mex.,

62, 199–205.

Carpenter, M.A. (1980): Mechanisms of exsolution in sodic pyrox-

enes. Contrib. Mineral. Petrol., 71, 289–300.

Carpenter, M.A. & Okay, A. (1978): Topotactic replacement of

augite by omphacite in a blueschist rock from north-west

Turkey. Mineral. Mag., 42, 435–438.

Chatterjee, N.D., Johannes, W., Leistner, H. (1984): The system

CaO-Al2O3-SiO2-H2O: new phase equilibria data, some calcu-

lated phase relations, and their petrological applications.

Contrib. Mineral. Petrol., 88, 1–13.

Cortesogono, L., Lucchetti, G., Spadea, P. (1984): Pumpellyite in

low-grade metamorphic rocks from Ligurian and Lucanian

Apennines, Maritime Alps and Calabria (Italy). Contrib.

Mineral. Petrol., 85, 14–24.

Essene, E.J. & Fyfe, W.S. (1967): Omphacite in Californian meta-

morphic rocks. Contrib. Mineral. Petrol., 15, 1–23.

Evans, B.W. (1990): Phase relations of epidote-blueschists. Lithos,

25, 3–23.

Fettes, D. & Desmons, J. eds. (2007): Metamorphic rocks: a classi-

fication and glossary of terms: recommendations of the

Internationall Union of Geological Sciences Subcommission

on the systematic of metamorphic rocks. Cambridge

University Press, Cambridge, 242 p.

Garcıa-Casco, A., Torres-Roldan, R.L., Iturralde-Vinent, M.,

Millan, G., Nunez Cambra, K., Lazaro Calisalvo, C.,

Rodrıguez Vega, A. (2006): High-pressure metamorphism of

ophiolites in Cuba. Geol. Acta, 4, 63–88.

Garcıa-Casco, A., Iturralde-Vinent, M.A., Pindell, J. (2008): Latest

Cretaceous collision ? accretion between the Caribbean Plate

and Caribeana: origin of metamorphic terranes in the Greater

Antilles. Int. Geo. Rev., 50, 781–809.

Garcıa-Casco, A., Vega, A.R., Parraga, J.C., Iturralde-Vinent, M.A.,

Lazaro, C., Quintero, I.B., Agramonte, Y.R., Kroner, A., Cambra,

K.N., Nunez Cambra, K., Millan, G., Torres-Roldan, R.S.,

Carrasquilla, S. (2009): A new jadeitite jade locality (Sierra del

Convento, Cuba): first report and some petrological and arche-

ological implications. Contrib. Mineral. Petrol., 158, 1–16.

Gasparik, T. & Lindsley, D.H. (1980): Phase equilibria at high

pressure of pyroxenes containing monovalent and trivalent

ions. Rev. Mineral., 7, 309–339.

Graham, C.M. (1976): Petrochemistry and tectonic setting of

Dalradian metabasaltic rocks of the S.W. Scottish Highlands.

J. Geol. Soc., 132, 61–84.

Grant, J.A. (1986): The isocon diagram – a simple solution to

Gresen’s equation for metasomatic alteration. Econ. Geol., 81,

1976–1982.

Green, E., Holland, T.J.B., Powell, R. (2007): An order-disorder

model for omphacitic pyroxenes in the system jadeite-diop-

side-hedenbergite-acmite, with applications to eclogitic rocks.

Am. Mineral., 92, 1181–1189.

Grevel, C. (2000): Druck- und Temperaturentwicklung der meta-

morphen Deckeneinheiten des Escambray Massives, Kuba.

Dr. rer. nat. Thesis, Ruhr-Universitat Bochum, Germany,

169 p.

Grevel, C., Maresch, W.V., Stanek, K.P., Grafe, F., Millan, G.

(1998): Petrology and PT-data of high-pressure metamorphic

rocks from the Escambray Massif, Cuba. Ber. Dtsch. Min.

Ges., Beih. z. Eur. J. Mineral., 10, 110.

Grevel, C., Maresch, W.V., Stanek, K.P., Grafe, F., Hoernes, S.

(2006): Petrology and geodynamic significance of deerite-bear-

ing metaquartzites from the Escambray Massif, Cuba. Mineral.

Mag., 70, 527–546.

Harley, S.L., Carswell, D.A. (1990): Experimental studies on the

stability of eclogite facies mineral parageneses. in ‘‘Eclogite

Facies Rocks’’, D.A. Carswell, ed., Blackie and Son Ltd.,

London, 53–82.

Harlow, G.E. & Sorensen, S.S. (2005): Jade (nephrite and jadeitite) and

serpentinite: metasomatic connections. Int. Geol. Rev., 47,

113–146.

Harlow, G.E., Sisson, V.B., Ave Lallemant, H.G., Sorensen, S.S.,

Seitz, R. (2003): High-pressure, metasomatic rocks along the

Motagua Fault Zone, Guatemala. Ofioliti, 28, 115–120.

Harlow, G.E., Hemming, S.R., Ave Lallemant, H.G., Sisson, V.B.,

Sorensen, S.S. (2004): Two high-pressure – low-temperature

serpentine–matrix melange belts, Motagua fault zone,

Guatemala: a record of Aptian and Maastrichtian collisions.

Geology, 32, 17–20.

Harlow, G.E., Murphy, A.R., Hozjan, D.J., de Mille, C.N., Levinson,

A.A. (2006): Pre-Columbian jadeite axes from Antigua, West

Indies: description and possible sources. Can. Mineral., 44,

305–321.

Harlow, G.E., Sorensen, S.S., Sisson, V.B. (2007): Jade deposits. in

‘‘Geology of Gem Deposits’’, L.A. Groat, ed., Mineralogical

Society of Canada Short Course Series, 37, 207–254.

Harlow, G.E., Sisson, V.B., Sorensen, S.S. (2011): Jadeitite from

Guatemala: new observations and distinctions among multiple

occurrences. Geol. Acta, 9, 363–387.

Hatert, F., Pasero, M., Perchiazzi, N., Theye, T. (2007): Pumpellyite-

(Al), a new mineral from Bertrix, Belgian Ardennes. Eur. J.

Mineral., 19, 247–253.

Hatert, F., Ernst, A.J., Burke, E. (2008): The IMA-CNMNC domi-

nant-constituent rule revisited and extended. Can. Mineral., 46,

717–728.

Johannes, W., & Schreyer, W. (1981): Experimental introduction of

CO2 and H2O into Mg-cordierite. Am. J. Sci., 281, 298–317.

Leake, B.E., Woolley, A.R., Arps, C.E.S., Birch, W.D., Gilbert, M.C.,

Grice, J.D., Hawthorne, F.C., Kato, A., Kisch, H.J., Krivovichev,

V.G., Linthout, K., Laird, J., Mandarino, J., Maresch, W.V., Nickel,

Jadeite formation in metabasic rock 233

E.H., Rock, N.M.S. (1997): Nomenclature of amphiboles: report of

the subcommittee on amphibole commission on new minerals and

mineral names. Eur. J. Mineral., 9, 623–651.

Lindsley, D.H. (1980): Phase equilibria of pyroxenes at pressures .

1 atm. Rev. Mineral., 7, 289–307.

Maresch, W.V., Grevel, C., Stanek, K.P., Carpenter, M.A. (2007):

Coexisting jadeite and omphacite in metabasites from the

Escambray Massif, Cuba. Geochim. Cosmochim. Acta, 71/15S,

A621.

Millan, G. & Somin, M.L. (1981): Litologıa, estratigrafıa, tectonica

y metamorfismo del Macizo del Escambray. Editorial academia,

La Habana, 1–104.

Millan, G. & Somin, M.L. (1985a): Condiciones geologicas de la

constitucion de la capa granito-metamorfica de la corteza ter-

restre de Cuba. Instituto de geologıa y paleontologıa, La Habana,

1–83.

Millan, G. & Somin, M.L. (1985b): Contribucion al conocimiento

geologico de las metamorfitas del Escambray y del Purial.

Academia de ciencias de cuba, La Habana, 1–74.

Millan Trujillo, G. (1997): Geologıa del macizo metamorfico

Escambray. in ‘‘Estudios sobre Geologıa de Cuba’’, G.F.

Furrazola Bermudez, & K.E. Nunez Cambra, eds., Centro

Nacional de Informacion Geologica, La Habana, 271–288.

Morimoto, N., Fabries, J., Ferguson, A.K., Ginzburg, I.V., Ross, M.,

Seifert, F.A., Zussman, J., Aoki, K., Gottardi, G. (1988):

Nomenclature of pyroxenes. Am. Mineral., 73, 1123–1133.

Neuser, R.D. (1995): A new high-intensity cathodoluminescence

microscope and its application to weakly luminescing minerals.

Bochumer Geol. u. Geotech. Arb., 44, 116–118.

Nisbet, E.G. & Pearce, J.A. (1977): Clinopyroxene compositions in

mafic lavas from different tectonic settings. Contrib. Mineral.

Petrol., 63, 149–160.

Oberhansli, R., Bousquet, R., Moinzadeh, H., Moazzen, M., Arvin,

M. (2007): The field of stability of blue jadeite: a new occur-

rence of jadeitite at Sorkhan, Iran, as a case study. Can. Mineral.,

45, 1501– 1509.

Pindell, J., & Kennan, L. (2009): Tectonic evolution of the Gulf of

Mexico, Caribbean and northern South America in the

mantle reference frame: an update. in ‘‘The Origin and

Evolution of the Caribbean Plate’’, K.H. James, M.A.

Lorente, J.L. Pindell, eds., Geological Society,London,

Special Publications, 328, 1–55.

Pindell, J., Maresch, W.V., Martens, U., Stanek, K.P. (2012): The

Greater Antillean Arc: early Cretaceous origin and proposed

relationship to Central American subduction melanges: implica-

tions for models of Caribbean evolution. Int. Geol. Rev., 54,

131–143.

Pouchou, J.L. & Pichoir, F. (1984): Un nouveau modele de calcul

pour la microanalyse quantitative par spectrometrie de rayons X.

Rech. Aerospatiale, 3, 167–192.

Rieder, M., Cavazzini, G., D’Yakonov, Y., Frank-Kamenetskii,

V.A., Guggenheim, S., Koval, P.V., Muller, G., Neiva,

A.M.R., Radoslovich, E.W., Robert, J.-L., Sassi, F.P., Takeda,

H., Weiss, Z., Wones, D.R. (1998): Nomenclature of the micas.

Can. Mineral., 36, 905–912.

Schertl, H.-P., Neuser, R.D., Sobolev, N.V., Shatsky, V.S. (2004):

UHP-metamorphic rocks from Dora Maira/Western Alps and

Kokchetav/Kazakhstan: new insights using cathodolumines-

cence petrography. Eur. J. Mineral., 16, 49–57.

Schertl, H.-P., Krebs, M., Maresch, W.V., Draper, G. (2007):

Jadeitite from Hispaniola: a link between Guatemala and

Antigua?. 20th Colloquium on Latin American Earth Sciences,

Kiel, Germany, Abstract Vol, 167–168.

Schertl, H.-P., Maresch, W.V., Krebs, M., Stanek, K.P. (2008):

Jadeitite as a monitor of subduction-zone fluids: the Rio San

Juan Serpentinite Complex (Dominican Republic). Ber. Dtsch.

Min. Ges., Beih. z. Eur. J. Mineral., 15, 197.

Schiffman, P. & Liou, J.G. (1980): Synthesis and stability relations

of Mg-Al Pumpellyite, Ca4Al5MgSi6O21(OH)7. J. Petrol., 21,

441–474.

Schneider, J., Bosch, D., Monie, P., Guillot, A., Garcıa-Casco, A.,

Lardeaux, J.M., Torres-Roldan, R., Millan Trujillo, G. (2004):

Origin and evolution of the Escambray Massif, Central Cuba: an

example of HP/LT rocks exhumed during intra-oceanic subduc-

tion. J. Metamorphic Geol., 22, 227–247.

Seitz, R., Harlow, G.E., Sisson, V.B., Taube, K.A. (2001): Formative

jades and expanded jade sources in Guatemala. Antiquity, 87,

687–688.

Somin, M.L. & Millan, G. (1974): Nekotorye tschorty struktury

mezozoiskich metamorfitscheskich tol’sch Kuby.

Geotektonika, 5, 19–30.

Somin, M.L. & Millan, G. (1981): Geologija metamorficeskich

kompleksov Kuby. Isdatelstvo Nauka, Moskva, 219 p.

Sorensen, S.S., Grossman, J.N., Perfit, M.R. (1997): Phengite-hosted

LILE enrichment in eclogite and related rocks: implications for

fluid-mediated mass tranfer in subduction zone and arc magma

genesis. J. Petrol., 38, 3–34.

Sorensen, S., Harlow, G.E., Rumble, D. (2006): The origin of

jadeitite-forming subduction zone fluids: CL-guided SIMS

oxygen isotope and trace element evidence. Am. Mineral.,

91, 979–996.

Sorensen, S.S., Sisson, V.B., Harlow, G.E., Ave Lallemant, H.G.

(2010): Element residence and transport during subduction-zone

metasomatism: evidence from a jadeitite serpentinite contact,

Guatemala. Int. Geol. Rev., 52, 899–940.

Stanek, K.P., Cobiella-Reguera, J.L., Maresch, W.V., Millan

Trujillo, G., Grafe, F., Grevel, C. (2000): Geological develop-

ment of Cuba. Z. Angew. Geol., SH1, 259–265.

Stanek, K.P., Maresch, W.V., Grafe, F., Grevel, C., Baumann, A.

(2006): Structure, tectonics and metamorphic development of

the Sancti Spiritus Dome, eastern Escambray, Central Cuba.

Geol. Acta, 4, 151–170.

Stanek, K.P., Maresch, W.V., Pindell, J. (2009): The geotectonic

story of the northwestern branch of the Caribbean arc: implica-

tions from structural and geochronological data of Cuba. in

‘‘The Origin and Evolution of the Caribbean Plate’’, K.H.

James, M.A. Lorente, J.L. Pindell, eds., Geological Society of

London, Special Publications 328, 361–398.

Staudigel, H. (2003): Hydrothermal alteration processes in the ocea-

nic crust. Treat. Geochem., 3, 511–535.

Sun, S.-S. & McDonough, W.F. (1989): Chemical and isotopic

systematics of oceanic basalts: Implications for mantle compo-

sition and processes. in ‘‘Magmatism in the Ocean Basins’’,

K.H. Saunders, M.J. Norry, eds., Geological Society Special

Publication 42, 313–345.

Ungethum, H. (1965): Eine neue Methode zur Bestimmung von

Eisen (II) in Gesteinen und Mineralen. Zeitschr. Angew. Geol.,

11, 500–505.

Waterwiese, T., Chatterjeee, N.D., Dierdorf, I., Gottlicher, J., Kroll,

H. (1995): experimental and thermodynamic study of heteroge-

neous and homogeneous equilibria in the system NaAlSiO4-

SiO2. Contrib. Mineral. Petrol., 121, 61–73.

234 W.V. Maresch, C. Grevel, K.P. Stanek, H.-P. Schertl, M.A. Carpenter

Willner, A.P., Sepulveda, F.A., Herve, F., Massonne, H.-J., Sudo, M.

(2009): Conditions and timing of pumpellyite-actinolite-facies

metamorphism in the Early Mesozoic frontal accretionary prism

of the Madre de Dios Archipelago (Latitude 50�20’S; Southern

Chile). J. Petrol., 50, 2127–2155.

Wu, X., Han, Y., Meng, D., Li, D. (2002): Discovery and implication

of the P21/n crystal structure on a nano-scale in single jadeitite

crystals. Earth Plan. Sci. Lett., 197, 165–169.

Zane, A., Sassi, R., Guidotti, C.V. (1998): New data on metamorphic

chlorite as a petrogenetic indicator mineral, with special regard

to greenschist-facies rocks. Can. Mineral., 36, 713–726.

Received 5 March 2011

Modified version received 18 August 2011

Accepted 2 November 2011

Jadeite formation in metabasic rock 235

View publication statsView publication stats