679

The Canadian MineralogistVol. 37, pp.679-690 (1999)

UVAROVITE IN PODIFORM CHROMITITE:THE MOA-BARACOA OPHIOLITIC MASSIF. CUBA

JOAQUIN PROENZA

Departamento d.e Geologia, Instituto Superior Minero Metalilrgico de Moa,Las Coloradas s/n, 83320 Moa, Cuba and Departament de Cristal.lografia, Mineralogia i Dipdsits Minerals,

Universitat de Barcelona, C/ Marti i Franquis s/n, E-08028 Barcelona, Catalonia, Spain

JESUS SOLE AND JOAN CARLES MELGAREJO$

Departament de Cristal.lograJia, Mineralogia i Dipdsits Minerals, Universitat de Barcelona,C/ Marti i Franquis s/n, E-08028 Barcelona, Catalonia, Spain

Assrnlcr

The chromitite pods of the Moa-Baracoa massif, in the eastern ophiolitic belt of Cuba, contain pre-existing gabbro sills. Thisassociation is affected by two processes of hydrothermal alteration. The chromitites and the hosting dunites and harzburgites areaffected first by regional serpentinization; a second alteration, represented by chloritization accompanied with formation offerrian chromite, is mainly located in the pods and their immediate vicinity. The altered chromitite pods and enclosed gabbro sillsiue cross cut by millimeter-wide veins. The vein filling consists of a sequence of clinochlore, uvarovite, chromian clinochlore,rutile, titanite and calcite. Uvarovite also occurs in the vicinity of veins Uvarovite is concentrically zoned, covering compositionsin the uvarovite-grossular solid solution series between Uva17 and Uvae:; the andradite component is very low. These composi-tions suggest a complete miscibility along the grossular-uvarovite join at relatively low temperature. On the basis of the mineralsequence and mineral chemistry (major and trace elements), the uvarovite crystals, as well as the vein assemblage, formed by alow-temperature leaching, Ca probably from the gabbro sills, and Cr and Al from the chromite dunng the formation of ferrianchromite Cr and A1 would have been mobile only at the scale of a pod during this process.

Keywords: uvarovite, zoning, hydrothermal, chromitite, ophiolite, Moa-Baracoa massif, Cuba

SorratraeJns

Les lentilles de chromitite du massif de Moa-Baracoa, dans la partie orientale de la ceinture ophiolitique de Cuba, contientdes filon-couches de gabbro. Cette association a subi les effets de deux stades d'alt6ration hydrothermale. Les lentilles dechromitite et la dunite et la harzburgite encaissantes ont d'abord subi l'influence d'une serpentinisation r6gionale. Une deuxidmealtdration a caus6 une chloritisation et la formation de chromite ferrique, surtout dans les lentilles et leurs environs. Les lentillesde chromitite alt6r6es et les f,rlon-couches de gabbro sont recoup6s par des veines millim6triques tapiss6es de clinochlore,uvarovite, clinochlore chromifbre, rutile, titanite et calcite. L'uvarovite se trouve aussi tout prbs des veines. Les cristauxd'uvarovite montrent une zonation concentrique, et les compositions s'6talent entre uvarovite Uva63 et grossulaire Uva17. Lateneur en andradite est trbs faible. Ces compositions font penser qu'il y a une miscibilitd compldte entre grossulaire et uvarovited temp6rature relativement faible. A la lumibre de la s6quence parag6ndtique et de la composition des min6raux (6l6ments majeurset traces), les cristaux d'uvarovite, ainsi que I'assemblage des veines, se sont form6s par lessivage d faible temp6rature, le calciumvenant probablement des filon-couches de gabbro, et le Cr et Al contribu6s par la chromite lors de la formation de la chromiteferrique Le chrome et I'aluminium n'auraient 6t6 mobiles qu'd l'6che11e d'une lentille pendant ce processus.

(Traduit par la R6daction)

Mots-clds'. uvarovite, zonation, altdration hydrothermale, chromitite, ophiolite, massif de Moa-Baracoa, Cuba.

s E-mail address: joanc@natura.geo ub es

680 THE CANADIAN MINERALOGIST

INrnooucrroN

The uvarovite end-member of the garnet group,Ca:Crz(SiO+):, has not yet been found in nature. Natu-ral uvarovitic garnet typically contains significantamounts of both grossular and andradite components.Such compositions have been recorded in diverse geo-logical settings: chromitites associated with basic andultrabasic rocks (Chakraborty 1968, Jan et al. 1984,Graham et al. 1996, Proenza & Melgarejo 1996,Melcher e t a l . 1997) , rod ing i tes (Mogess ie &Rammlmair 1994), skarns (von Knorring et al. 1986,Pan & Fleet 1989), and kimberlites (Meyer & Boyd1972, Sobolev et al. 1973). The range of envfuonmentsin which it is found suggests a broad field of stability(Kalamarides & Berg 1988).

Our aim in this paper is to provide new chemical andstructural data on this important garnet, and to describeits mineral association and petrological significance rnthe Mayari-Baracoa ophiolitic belt, Cuba.

Gror-ocrcal SBrrrNc

The northem part of the island of Cuba is character-ized by a 1000-km-long Upper Jurassic - Lower Creta-ceous ophiolitic belt (Iturralde-Vinent 1996). Theeastern part, the Mayari-Baracoa ophiolitic belt (Fig. 1),is 170 km long, l0 to 20 km wide, and comprises twoallochthonous massifs: Mayari-Cristal and Moa-Baracoa. Both massifs contain abundant podiform de-posits of chromit i te (Thayer 1942, Guild 1947,Murashko & Lavandero 1989, Proenza et al.1997a,7999, Proenza 1998). The Mayari-Cristal massif, the

westem part of the Mayari-Baracoa belt, is mainly com-posed of serpentinized ultrabasic rocks (harzburgite andinterlayered harzburgite and dunite), with microgabbrosand a diabase dike complex overlying the unit (Iturralde-Vinent 1996). Chromitite pods occur within the dunite;the chromite is relatively Cr-rich, with Crl(Cr + Al) =0.75 (metallurgical grade).

The Moa-Baracoa massif is also mainly composedof serpentinized ultrabasic rocks; harzburgite gradesupward to harzburgite and dunite. In the uppermost partof the mantle section, sills and dikes of gabbro are com-mon; this suite is covered by a crustal section compris-ing, from bottom to top, banded gabbro, isotropicgabbro, gabbro with diabase dikes and, finally, basaltand chert. The chromitite pods, commonly with duniticenvelopes, occur in the uppermost part of the mantle-crust transition zone (MTZ). Many chromitites encloseand replace the older gabbro sills parallel to the elonga-tion of the pods (Fig. 2). This feature is typical of mostchromite deposits of the Moa-Baracoa district (Proenzaet al. 1997a, 1999), as well as of other districts in cen-tral Cuba (Flint et al. 1948\ and at Coto in the Zambalesophiolite in Philippines (Leblanc & Violette 1983).These chromitites are made up of Al-rich chromite orchromian hercynite, with Cri(Cr+Al) = 0.45 (refractorygrade).

Uvarovitic garnet related to chromitites has beenfound only in the Moa-Baracoa massif (Grild 1947,Kenarev 1966, Lewis et al. 1994,Proenza &. Melgarejo1996). Consequently, the Moa-Baracoa chromite oreswill be described with greater detail. However, therrabsence in Mayari-Cristal massif has genetic implica-tions as discussed later.

MAYARI.CRISTAL "\

Frc. 1. Geological sketch of the northwestern Caribbean ophiolites (modified from Wadge et al. 7984).

UVAROVITE IN PODIFORM CHROMITITE, CUBA 681

" " nTYt" v .Effx*;l>*"

i:ti::;::t:ir;l:rdur

i : r i l : ; lW!: ; : : i i i ; i ! i i i l

v v u u v

har*urgite

F\c 2 Schematicsketchofthechromitebodies,showingthegabbrosillsandthetwosetsof veins

Tnr, CHnoN4tre Onns n Moe-Beneco,qAND THE Uvanovrre OccunnpNcBs

A first stage in the formation of the chromitite podsinvolved the magmatic processes that produced the de-posit. Primary mineralization is formed by chromite plusinterstitial olivine and, as small inclusions in chromite,minor amounts of clinopyroxene, amphibole and pla-gioclase. This association replaced the enclosed pre-chromitite gabbro sills, as well as the hosting dunite andharzburgite (Proenza e t al. 1997 a, 1999, Proenza 1 99 8).

The texture of the chromitite is massive, grading insome deposits to disseminated, and displays pull-apartfractures. This type of fracture is very common inpodiform deposits of chromite, and it is visible in handsamples as well as on the scale of the deposit (i.e.,Leblanc & Nicolas 1992).

Subsolidus hydrothermal processes produced sec-ondary minerals that widely replaced the above mineralassociation. Two main stages of alteration can be dis-tinguished. The first stage was regional, and affectedthe"chroriitite pods ani the hosirocks (dunite andharzburgite), replacing the olivine grains by an qssem-blage of serpentine plus magnetite. The second stagetook place largely in the chromitite pods and their vi-cinity, and comprises chromite alteration into ferrianchromite ("ferritchromit"), coupled with replacement ofserpentine by Cr-poor clinochlore (clinochlore-I,Proenza et al. 1997b). This type of alteration is poorlydeveloped in harzburgite or dunite occurring far fromthe pods, and is restricted to a millimeter scale in thevicinity of the small grains of accessory chromianspinel.

The chromitite bodies of the Moa-Baracoa massifare cross-cut by two sets of mineralized joints (Fig. 3).Uvarovitic garnet appears only within the first set.

Uvarovite appears also in an alteration zone, up to sev-eral centimeters in width, that replaces the hostingchromitite along these fractures. In this case, thechromite grains are partly corroded and replaced byuvarovite, and the replacement of chromite progressesalong the grain boundaries and the pull-apart fractures.

Frc. 3. Schematic cross-section of the garnet-bearing veins.

Symbols: 1) chromite, 2) fertran chromite, 3) clinochlore-I, 4) uvarovite, 5) clinochlore-Il (chromian clinochlore)'

6) titanite, T) rutile, 8) calcite, and 9) clinochlore-Ill.

WT[W]1FI=.rr.'.,::Tl

l:i':jill:l.r:l:iiil

:t*--,JF;";--i

FTlr+rTi

Ett-::-:1-1

i;,;":'l

W

684 THE CANADIAN MINERALOGIST

parameters were detemined using the Rietveld methodimplemented in the FULLPROF program (Rodriguez-Carvajal 1990). The pattern was refined in space groupIa3d (Deer et al. 1992), but it is imporlant to note thatother cubic space groups also are possible for garnetcompositions of this type (l.e., Pinet & Schmidt 1993).

MrNBner- Cuelrrsrny

Major elements

A representative set of microprobe results is givenin Table 1. Their compositions cover a large interual ofuvarovite-grossular solid solution, with a very low an-dradite component (Fig. 9). Garnet present as intersti-tial grains from the alteration zone of chromitite has thesame composition as gamet from the veins. The Fe con-tent is low, and charge-balance considerations suggestthat all the iron is trivalent; nevertheless, it is difficultto assign the correct charge to such small amounts ofFe. The Ti was assigned to occupy octahedral positions.The amounts of V, Mn and Mg are very low or belowthe detection limit, and these garnet compositions cantherefore be represented essentially as solid solutionsof the binary series grossular-uvarovite, with 17 to 63mo1.70 uvarovite.

TABLE I REPRESENTATI\IERESULTS OFELECTRON-MICROPROBE ANALYSES OF UVAROVITIC GARNET,

MOA_BARACOA MASSIF. CIIBA

S a m p l e l 2 3 4 5 6 7 8SiO, (wt,%) 38 67 38 60 38 85 37 81T i o 2 0 4 4 0 8 3 0 4 7 0 9 5Al2o3 15_90 1534 1346 1008

vzo: 0 13 0 26 0 18 023Cr2O3 7 28 8 87 10 39 15 0lFe2O3 069 020 042 03ECaO 3621 3605 3591 3566Total 99.32 100.14 9969 100 11

Slructml foroula based on 8 cations

Si (aptu) 3 0l 299 3 04 3 00T i 0 0 3 0 0 5 0 0 3 0 0 6Ar 1.46 | 40 124 0.94v 0 0 1 0 0 2 0 0 1 0 0 1Cr 045 054 064 0 .94F e 0 0 4 0 0 1 0 0 2 0 0 2C a 3 O 2 2 9 9 3 0 1 3 0 3

3'7 49 3'7 73 37 64 38.033 1 4 1 5 7 1 9 3 2 0 8

6 79 12 07 9,26 8 62

0.72 0r9 038 03 ' l

16 87 I 1 78 15.8s t6 22

0 '15 037 034 0 .3034.60 3s s7 35 06 3491

10036 99.28 10047 10052

3.02 2 99 2.99 3.030 19 0 .09 0 .12 0 130 6 4 1 1 3 0 8 7 0 8 10 0 5 0 0 1 0 0 2 0 0 21 0 7 0 7 4 1 0 0 1 0 20 0 5 0 0 2 0 , 0 2 0 0 2298 302 298 2 .98

UvaGrsAdr

2292 2 '7ss 3346 49.01 s933 388s 5220 54577459 ',7105 6464 490'7 3559 5935 454.1 43212 0 7 0 5 9 1 3 0 t . t 1 2 5 2 t t 5 1 0 8 0 , 9 7

Go 0.42 0.81 0.60 0.76 2.55 0.65 1.26 ]rzsl,2,3, 4 md,5 ftactxe-filliug uvarovite6, 7 md 8 disseminated uv4ovite gnins close to the veins in cbromititeEnd-memben molecules calculated afur Rickwood (1968)

Uva

Grs AdrFIG. 9. Plot of compositions of chromian garnet from the Moa-Baracoa Massif, Cuba, in

terms of mol Ea, compued to literature data. 1) Reaume Township (Duke & BonardiI982),2)Labrador (Kalamarides & Berg 1988), 3) White River (Pan & Fleet 1999),4)various localities in Canada (Dunn 1978), 5) Luikonlahti (von Knorring et al. 1986),6)Outukumpu (von Knorring et al. 1986),7) Jijal Complex, Pakistan (Jan et al 7984).

UVAROVITE IN PODIFORM CHROMITITE, CUBA 685

Infrared-absorption and Raman analysis of the gar-net was performed on two samples to check for the pres-ence of structurally bound OH. The results were notconclusive with IR, but the Raman spectra shows twobands that may correspond to a very small amount ofH2O. On the other hand, the chemical compositionsobtained by electron microprobe agree with those ex-pected of anhydrous garnet, because the tetrahedrallycoordinated position is nearly always completely filledwith Si.

The zoning is oscillatory, as is easily detected byvariations in hues of green. These variations in colorreflect variations in the Al:Cr ratio, i.e., the proportionof uvarovite and grossular components (Fig. 10). Thtskind of zoning is common in chromian gamet (Ian et aI.1984). The Cr-rich zones are also the richest in Ti andV. A comparable Cr-Ti correlation has also been ob-served in the chromian andradite from Reaume Town-ship, Ontario (Duke & Bonardi 1982).

Tio2Ta2O5

Nb205

Al2Or

Cr2O1

MgoCaOMrOFeOFe2O3

NioNazoKzoE oTotal

Some differences in Cr content exist among the three si (aptu)generations of chlorite. The early-formed chlorite does NAt'not contain Cr (Table 2). I t can be classif ied as 'Alclinochlore, after Hey (1954). The second generation of richlorite, which overgrows uvarovite, has Cr contents rabetween 2.5 and 7.5 wt%o Cr2O3, and is classified as Tchromian chlinoclore (Table 2). The third generation is il"again Cr-free. In all three generations, Fe content is very c"low, and nickel values are lower than 0.2 wtTo NiO, with Mnan average as low as 0.057o. Rutile crystals are of end- Fe2-member composition (Table 2). Titanite contains 39Vo Fer*

TiOz and 107o AlzOz, with variable amounts of Cr, up Ni

ro l.4vo Cr2o j. )"

TABLE 2 REPRESENTATI\,IE RBSIILTS OFFT FCTRON-MICROPROBE ANALYSES OF CLINOCHLORE'

CHROMIAN CLINOCHLORE, TITANITE AND RUTIIJ,MOA_BARACOA MASSIF, CIJBA

S m p l e l 2 3 4 5 6 7 8SiOr(vt,%) 31.'16 32.14 27.92 26.81 31.40 30.95

001 0.02 0,00 000 3913 3852 9977 9942- 000 002

0 0 7 0 0 0

0 8 2 0 6 3 0 1 1 0 . 1 6

0 3 4 1 3 8 0 0 8 0 0 3

0.00 0.01 0.00 0.03

2E 55 28 ,16 0 03 0 00

0.00 0.00 0 00 0 07

0 0 0 0 0 00 0 0 0 0 6

0 0 1 0 2 2 0 1 9

0 0 0 0 0 1 0 0 30 00 0.03 0 02

128a 1255 1242

s98r 99,64 9902 10025 99.6s 9962 99't9

1832 l'1 96 2020 2050

0 0 6 0 0 7 5 7 8 7 0 5

3594 35 88 31 13 3020

0 - 0 1 0 0 3 0 0 0 0 0 0

0 0 0 0 0 2 0 0 0 0 . 0 0

0 5 9 0 8 0 1 7 8 1 8 1

0 0 90020.01

12 8'799 68

s 9 2 5 9 9 5 3 4 5 1 8

2 0 8 2 0 1 2 6 6 2 8 2

1 9 5 1 9 3 1 8 9 1 8 5

0 0 0 0 0 0 0 0 0 0 0 0

102 1 .01 _ _

0.03 0.02 0 00 0.00095 0 .95 100 100

0 0 0 0 0 00.00 0.00

001 004 000 000000 000 000 0000 9 9 0 9 9 0 0 0 0 0 0000 000 0 .00 0000,00 0 00

o o , o o t o t t9 . 9 9 9 9 6 8 8 70 0 0 0 0 1 0 0 00 0 0 0 0 0 0 0 0

0 0 9 0 1 2 0 2 8

1 0 88 6 90 0 00 0 0

o-29

0.00 0 00

001 0 .00 003 0030 0 r 0 0 0 0 . 0 1 0 0 1000 000 0.01 0.00 - - - -

I *d ,=ltt""Ll-" (*tit*"l",iltted on the basis of 28 O)

3 md 4: chromim clinochlore (cations calculated on the basis of 28 O)

5 md 6: titmite (cations calrulated on the bdis of 3 cat')

7 md 8; rutile (cations caleulat€d on the basis of2 O)

| 2 3 4 5 6 1 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9AnalYsed Points

Rim Core -

Frc. 10. a) Back-scattered-electron (BSE) image of oscillatory zoning in uvarovite. The line indicates the profile of the crystal

analyzed. b) Profile of electron-microprobe data along the line 1-19.

sEo

X

t6

t4

12

10

8

2

0

686 THE CANADIAN MINERALOGIST

Trace elements

One sample of garnet was analyzed for 31 trace ele-ments by ICP-MS. As far as we know, quantitative dataofthis type have not been published before. As shownin Table 3, the Moa-Baracoa uvarovite has no detect-able amounts of Pb and Ce, a few ppm of Li, Co, Cu,Ga, Rb, Sr, M, Sn, Cs, Ba, Hf, Ta, Th and U, and,20Oppm of Ni. The V contents (1000 ppm) is similar to theconcentration obtained by electron microprobe. Lan-thanide contents are relalively low. not farfrom beingchondr i t i c {F ig . l t ) .

Compared with the semiquantitative data of Wan &Yeh (1984), the levels of Ni, Cu, Nb, Sn, Ba and pb areof a similar order of magnitude, but levels of Co, Sr andLa are lower in the uvarovite from Moa. The lanthanumcontent reported by Wan & Yeh (1984) is probably toohigh because it implies a concentration of lanthanidesten thousand times the chondritic values, which is ques-tionable in such garnet.

DerpnvnnrroN oF THE CBlr- Penaueren

We obtained a cell parameter a of I1.922 A usingall reflections. This value is a weighed average of thedifferent values of the cell parameter in this slightlyzoned gamet. It can be compared to values given byDeer et al. (1992) for the end-member samets. If we

assume a linear relationship between cell parameter andcomposition in the solid solution uvarovite-grossular,and neglect the small andradite component, the cell pa-rameter obtained is equivalent to an average composi-tion of Uvaa7Grs53.

DrscussroN AND CoNcLUSroNs

The garnet-bearing veins are confined to the interiorof the chromitite bodies. These veins can be interpretedto be the response to deformation of a rigid body (thechromitites) enclosed by more ductile rocks (theserpentinites). Their formation took place well after theserpentinization process, as is suggested by Bamba(1984) for some alpine podiform deposits of chromitein Turkey.

The textures of the idiomorphic crystals, as well astheir position in the veins, indicate that they formed froma hydrothermal fluid in an open space, along fractures.The mineral association points to a formation by a low-temperature hydrothermal process. The existence of apure Fe-poor clinochlore that cocrystallized with uvaro-vite in the veins suggests the possibility to estimate thetemperatures of formation of this assemblage usingchlorite geothermometry. The first generation ofclinochlore yields a low temperature, about l8G-190'C,using the geothermometer of Hillier & Velde (1991),and230-240C using the geothermometer of Kranidiotis

Uo

(t)

10.00

1.00

0.10

2t

""

a Uvarovite

€- Sillofgabbro

LaCe PrNd SmEuGdTbDyHoErTmYbLu

FIG 1 1. Plot of the REE data normalized to chondrite for both uvarovite and sabbro sill

UVAROVITE IN PODIFORM CHROMITITE, CUBA 687

L i 51 Sn 03V 1060 CB 0 l lCo 2.6 Ba 06Ni 238 lrf 127Cu 15 Ta 002Ga 40 Pb ndRb 03 Th 009S r 1 0 U 0 4N b 0 1

TABLE3 TRACE.ELEMENTCONCENTRATIONSIN I'VAROVITEFROM TI{E MOA.BARACOA MASSIF. CUBA

LrCePrNdSmEuGdIbDyHoErTEYbLu

most feasible explanation for the oscillatory zoning ofthe Moa-Baracoa garnet.

A portion of the compositional field of uvaroviticgarnet published to date is plotted in Figure 9. It is sig-nificant to note that the suite from the Moa-Baracoamassif contains compositions not yet reported in the liferature on natural examples of the solid solution uvaro-vite-grossular, despite the fact that these compositionshave already been synthesized (Huckenholz & Knittelr9'ts).

Kalamarides & Berg (1988) discussed in detail thegaps in this series. On the basis of their own data, ex-perimental work by Huckenholz ( I 975) and Huckenhoiz& Knittel (1975), and data supplied by Duke & Bonardi(1982) and Jan et al. (1984), they concluded thatugrandite garnet shows complete solid-solution at hightemperatures (800-1550'C) and variable pressures(1 bar to 10 kbar). They concluded, also, that a compo-sitional gap exists between uvarovite and grossular atlow temperatures. The paragenesis and hydrothermalmode of occulrence at Moa-Baracoa indicate that thegarnet crystals formed at relatively low temperature andlow pressure. Our data suggest a complete miscibilityalong the grossular-uvarovite join (i.e., very low levelof the andradite component) at low temperatures. Thepotential of a miscibility gap at greater andradite com-Donent cannot be discarded from the data presented.^Ganguly

(1976) presented a model of the solid solutionin the system grossular - uvarovite - andradite. Its cal-culated spinodal at 400'C does not intersect the uvaro-vite-grossular join. Because the Moa-Baracoa suite wasvery likely formed below this temperature, our data donot conflict with this diagram.

These veins formed from a Ca- and Al-bearing hy-drothermal fluid. The Al may have been extracted dur-ing the replacement of chromite by ferrian chromite(Proenza et al.199'7b, Proenza 1998), according to thereaction:

aluminian chromite + ferroan antigorite - ferrianchromite + chromian clinochlore

The activity of Ca in this fluid led to the develop-ment of uvarovite, and the development of titaniteinstead of rutile. The source of Ca-rich fluids inophiolites is controversial. Mittwede & Schandl (1992)

suggested liberation of Ca linked to the phenomenon ofserpentinization. In our case, the Ca-rich fluids enteredinto the system much later than the serpentinization re-action, during low-temperature hydrothermal alterationthat replaced serpentine by chlorite. Since the ultrabasicrocks hosting the chromitite pods display a very limitedextent of chloritization, the composition of chlorite- anduvarovite-producing fluids seems to be controlled by thecomposition of the assemblage of chromitites plus theenclosed gabbro sills. The enclosed gabbros could havebeen an alternative source for Ca. These gabbros werehydrothermally altered, and plagioclase crystals were

0 451n d

0 171'l oll0 360d a r u0 46E0 093o 677o 172o 6't30 t230 9990 200

Analysis by ICP-MS tr d : oot detwled, All values re in ppm

& Maclean (1987). The estimated temperatures for thesecond generation of clinochlore (which is associatedwith uvarovite) are 280-300'C using the first geother-mometer. The second yield higher temperatures, about340-3'10"C. Therefore, the significant increase in the Crcontent of the clinochlore of the second generation re-sults in unrealistically higher temperatures than thoseof the first generation. The temperatures obtained usingboth geothermometers in the Cr-free clinochlore of thefirst generation, however, are in the range of those ob-tained by Christidis et al. (1998) for chlorite formationin the Vourinos ophiolitic complex in Greece.

The complex zoning surfaces of the Moa-Baracoagarnet do not seem to be linked to dissolution processes,but rather to growth. Moreover, their formation at a verylow temperature prevents solid-state diffusion as an ef-fective mechanism to explain this zoning (Shore &Fowler 1996). As stated by Jamtveit et al. (1995), inugrandite-type garnet in skarns, the compositions andmorphological properties of a hydrothermal mineral arecontrolled by: a) bulk composition of fluid, b) localtransport processes near the fluid-solid interface, andc) surface kinetics. Changes in the chemistry of the min-eralizing solution are unlikely in this case, because theveins are strictly related to the inner parts of the chromitebodies. This fact suggests that the fluid phase precipr-tated in an essentially closed system. The compositionof the fluid, therefore, must have been strongly con-trolled by the composition of both chromitites and thegabbro sills included in the chromitite pods. Conse-quently, it is difficult to explain oscillatory zoning bychanges in the bulk composition of the fluid.

In the opinion of Jamtveit et al. (1995) and Ivanovaet al. (1998), covariation among major elements andsimilarities among crystals suggest that zoning in gar-net is confrolled mainly by extemal forces (flow ratesor chemical changes), with only minor participation ofsurface kinetics. Although the gamet in our suite dis-plays the same type of correlations, we believe that aself-organization process (e.g., Putnis et al. 1992) is the

688

converted to prehnite and epidote; clinopyroxene crys-tals are replaced by fibrous tremolite.

As pointed out above, the uvarovitic garnet as wellas the chromitite pods occur in direct association withgabbro sills in the Moa-Baracoa massif (Proenza 1998,Proenza et al. 1999). Likewise, the lanthanide pattems(Fig. 11) of gabbros and gamet reflect a similarity intheir light rare-earth element contents, with a similarpositive Eu anomaly. The positive Eu anomaly in thefresh gabbros is due to Eu2+ enrichment in plagioclaseowing to the high partition coefficient (Dplacioclaseniquid)of this element (Hamois et al. 1990i). As during the earlystages of vein infilling, it is uvarovite that is by far thedominant calcium-bearing phase, Eu can also be ac-commodated in the garnet structure, replacing calcium.On the other hand, the heavy rare-earth enrichment inthe garnet is an expression ofthe general preference ofthese elements for the garnet structure (i.e., Hanson1980). The garnet thus seems to have inherited its lan-thanide pattern from the gabbro sills during the leach-ing process that affected both the chromitite bodies andthe cross-cutting sills of gabbro.

We conclude, therefore, that the Moa-Baracoauvarovite was formed by the leaching of Ca from thegabbro sills, and Cr and Al from chromite. The processis associated with low-temperature formation of chlo-rite and ferrian chromite in the chromitite pods. Exten-sive crystallization of chlorite and uvarovite is reStrictedto the chromitite pods. This fact can be attributed to alow mobility of Cr and Al during low-temperature hy-drothermal processes operating on ophiolites. These re-sults agree with the low mobility of Cr described inmany environments (1.e., S6nchez-Vizcaino et al. 1995,and references therein), and explain the conlmon occur-rence of uvarovite in Moa-Baracoa massif. In contrast,uvarovite is absent in Mayari-Cristal massif. Althoughthis massif underwent the same processes ofserpentinization and chloritization as at Moa-Baracoa,gabbro sills or other Ca-rich rocks do not occur. Hence,in the Mayari-Cristal massif, the supply of Ca to thehydrothermal fluids was lower and, consequently,uvarovite did not form.

Acrc{owr-socBN4sNrs

We thank the staff at the Mercedita mine, and spe-cially D. Rev6 and G. Rodriguez, for their aid in thefield work. All the analyses were performed at the"serveis Cientifico-Tdcnics" from ihe University ofBarcelona (Spain). We thank Drs. X. Llovet and J.Garcia Veigas for their help with the electron-micro-probe analyses, Dra. E. Pelfort, wirh the ICP-MS mea-surements, Dr. R. Fontarnau, with the the BSE imaging,Dr. T. Jawhari-Colin, with the Raman analyses, and Dra.N. Ferrer, with the IR analyses. We also thank Drs. J.H.Berg, F. Gervilla, S. Gali, G. Franz and R.F. Martin fortheir valuable discussions and criticism of the manu-script. This work was partially supported by the Span-

ish Instituto de Cooperaci6n Iberoamericana with an ICIgrant for J.P. (Ph.D. thesis).

RBrennNcBs

BeMsa, T. (1984): A model illustrating the formative processof the podiform chromite deposits in some alpine orogenicterains. Syngenesis and epigenesis in the formation ofmineral deposits (A. Wauschkuhyn, C. Kluth & R.A. Zirnmer-mann, eds.). Springer-Verlag, Berlin, Germany (507-518).

Cne.rneronrv, K.L (1968): Mineralogical note on thechrome-chlorite (KZimmererite) and chrome-garnet(uvarovite) from the chromite deposits of Kalrangi, Orissa,lndia. Am Mineral. 53, 962-965.

Cunrsrmrs, G 8., EcoNolrou-Erropoulos, M., MARcopoulos,T & LAsKou, M. (1998): An unusual assemblage of high-Ti oxides and ferroan clinochlore along zones of brittledeformation in the Vourinos (Rodiani) ophiolite complex,Greece. Can. Mineral 36, 1327-1338.

DEER, W.A., Howm, R.A. & ZussrvraN, J. (1992): Rock-Form-ing Minerals. lA. Orthosilicates (second ed.). Longrnan,London. U.K.

DUKE, J.M & BoNm.or, M. (1982): Cbromian andradite fromReaume Township, Ontario. Can. Mineral. 20,49-53.

DUNN, P.J. (1978): On the composition of some Canadian green

Earnels. Can. Mineral 16,205-206.

FI-tNr, D.E., DE ALBEAR, J.F. & Gun-r, P.W. (1948): Geologyand chomite deposits of the Camagiiey district, CamagiieyProvince, Cuba. U.S. Geol. Suw., Bull. 954-8,39-63.

GrNcuI-y, J. (1976): The energetics of natural gamet solid so-lution. II- Mixing of the calcium silicate end-members.Contrib Mineral. Petrol. 55, 81-90.

Gnlrterra, I.T., FuNrlnv, B.J. & MensHlr-L, B (1996): Chem-istry and mineralogy of podiform chromitite deposits,southern NSW, Australia: a guide to their origin and evolu-ion. Mineral. Petrol. 57, 129-150.

GuIlr, P.W. (1947): Petrology and structure of the Moachromite district, Oriente province, Cuba. Trans. Am.Geophy s. U nion 28, 218-246.

H.lNsoN, G.N. (1980): Rare earth elements in petrogeneticstudies of igneous systems. Annu. Rev. Earth Planet. Sct.8,371-406

HnnNors, L., TRorrreR, J. & MonBNcy, M- (1990): Rare earthelement geochemistry of Thetford Mines ophiolitic com-plex, northern Appalachians, Canada. Contrib. Mineral.Petrol 105,433-445.

HBv, M.H. (1954): A new revision of the chlorite. Mineral.Mas.30,277-292.

HtI-r-rrn, S. & Vrros, B. (1991): Octahedral occupancy and thechemical composition of diagenetic (low-temperature)chlorites. Clay Mineral. 26, 149-168.

UVAROVITE IN PODIFORM CHROMITITE. CUBA 689

HucKENHoLz, H.G. (1975): Uvarovire stability in the CaSiO3-Cr2O3 j oin up to 70 kbzr. N e ue s J ahrb M ine ral., M o nat s h.,337 -360.

__ & KNrrrrL, D. (1975): Uvarovite: stability ofuvarovite-grossularite solid solution at low pressure.Contrib. Mineral Petrol. 49, 2ll-232.

IrunnlI-os-VnrBNT, M.A. (1996): Ofiolitas y arcos volcdnicosde Cuba. IGCP Project 364, Spec. Contib. l.

IvANovA, T.I, SHrurrNnsnc, A.G., PuNrN, Yu O, FRANK-KaurNrrsreye, O.V. & SoreLov, P.B. (1998): On thecomplex zonality in grandite gamets and implications. Min-eral. Mag 62,857- 868

J.lurvrrr, B , RecNensoorrrn, K.V. & Wooo, B.J (1995): Onthe origin of zoned grossular-andradite garnets in hydro-thermal systems. Eur. J. Mineral. 7 , 1399-1410

JeN, M.Q , WrNor-By, B F & WrsoN R N. (1984): Chromianandradite and olivine-chromite relations in a chromititelayer from the Jijal complex, northwestern Pakistan. Cdn.Mineral 22,341- 345

KaleNrlnrnes, R.I. & BERG, J.H. (1988): Coexisting Cr-richand Cr-poor gamet from a calc-silicate gneiss, Labrador.Can. Mineral. 26, 335-342

KnNenpv, V.V. (1966): Caracteristicas mineral6gicas delyacimiento Potosi, Cuba. Revista Tecnol6gica 4,3-6

KneNrororrs, P. & MAcLEAN, W.H. (1987): Systematics ofchlorite alteration at the Phelps Dodge massive sulphidedeposit, Matagami, Quebec. Econ. Geol 82, 1898-1911

LEBLANC,M &Nrcouls, A (1992): Ophioliticchromites Inr.Geol. Rev.34,653- 686

& VIoI-Brrs, J.F (1983): Distribution of aluminium-rich and chromium-rich chromite pods in ophiolitependotites. Econ. Geol. 78, 293-301.

Lrws, J F , MuNoz, J.N , PeNc, G., Cerrapos, M., Qurxrm, F(1994): Mineralogy of silicate and chrome-spinel phase inthe ophiolite rock and chromite deposits of the Moa-Baracoa Nea. Restimenes del Segundo Congreso Cubanode Geologia (Santiago de Cuba),108 (abstr.)

L6pnz SANcuEz-VrzcniNo, V., FneNz, G & G6urz-PuGlqAInB,M.T. (1995): The behavior of chromium during metamor-phism ofcarbonate rocks from the Nevado-Fildbride Com-plex, Betic Cordilleras, Spain Can. Mineral 33, 85-104.

MBr-csrn, F., Gnua, W., Sntou, G., THer-uaulrBn, TV. &SruMpFL, E F ( 1 997): Petrogenesis of the Ophiolitic giantchromite deposrts of Kempirsai, Kazakhstan: a study ofsolid and fluid inclusions in chromite. J. Petrol. 38. l4l9-1458

MByrn, H O A. & Bovo, F R (1972): Composition and originof crystalline inclusions in natural diamonds. GeochimCosmochim. Acta 36, 1255-1273

MIrrwEDE, S.K. & ScHANDL, E.S. (1992): Rodingites from thesouthem Appalachian Piedmont, South Carolina, USAEur. J. Mineral. 4, 7 -16.

MocEssrE, A. & RAMMLMAIR, D. (1994): Occurrence of zoneduvarovite-grossular gamet in a rodingite from the Vumbaschist belt, Botswana, Africa: implications for the origin ofrodingites. Mineral. Mag. 58, 37 5-386.

MuRAsHKo, V.I. & LeveNopRo, R.M. (1989): Chromite in thehyperbasite belt of Cuba Int. Geol Rev. 31,90-99

PnN, YulNlrwc & FI-esr, M E (1989): Cr-rich calc-silicatesfrom the Hemlo area, Ontario Can. Mineral 27, 565-57'7

PINrr, M & ScHvnr, D.C (1993): La microsp6ctrometrieRaman des gretats X3Y223O12. I La s6rie calciquenaturelle ouvarovite * grossulaire - andradite. SchweizMineral. Petrogr Mitt 73,21-40.

PnonNza, J (1998): Mineralizaciones de cromita en la fajaofiolitica Mayari-Baracoa (Cubn) Ejemplo del yacimientoMercedita. Ph.D. thesis, Univ. of Barcelona, Barcelona,Soain

Grnvrr-r-e, F. & MrlcenBro, J.C. (1997a):Compositional variations of podiform chromitites amongdifferent mining districts in the Mayari-Baracoa OphioliticBelt (eastem Ctba) In Mineral Deposits: Research andExploration Where Do they Meet? (H Papunen, ed.)Balkema, Dordrecht, The Netherlands G8'7 -490).

& BorrNnn, J.L. (1999): Al-rich and Cr-rich chromitites from the Mayari-Baracoaophiolrtic belt (eastern Cuba): consequence of interactionbetween volatile-rich melt and peridotite in suprasub-duction mantle. Econ. Geol. 94, 547 -566

& Mnlclnpro. J.C (1996): Granates de la seriegrosular ia-uvarovi ta en cromit i tas podi formes delyacimiento Mercedita (Cuba). Geogaceta 20, l5l7 -1519.

& Gpnvrr-rq, F (1997b): Proceso dealteraci6n a ferricromita en Cr-espinelas del yacimientoMercedita (Cintur6n ofiolitico Mayari-Baracoa, Cuba).BoI Soc. Espaftola Mineral. 20A,29-30.

PurNrs, A., FERNANDEZ-DIAZ, L & PRrEro, M. (1992): Experrmentally produced oscillatory zoning in the (Ba,Sr)SO+solid solution. Nature 358,

'7 43-7 45.

RrcKwooD, P.C. (1968): On recasting analyses of gamet intoend-member molecrles. C o nt rib M ine r al. P e tro l. 18. l7 5 -

198.

Roonrcunz-ClnvArAl, J (1990): FULLPROF: a program forRietveld refinement and pattern matching analysis. Sarel-lite Meeting on Powder Diffraction of the XV Congress ofthe IUCr (Toulouse), I27

SHoRE, M & Foupn, A D. (1996): Oscillatory zoning in min-erals: a common phenomenon Can Mineral.34, l11l-t 126 .

690 THE CANADIAN MINERALOGIST

Sosor-Bv, N.V., LevnBNr'ev, Y.G., PorHil-sNro, N.P. &Usova, L.V. (1973): Chrome-rich garnets from thekimberlites of Yakutia and their parageneses. Contrib. Min-e ral. P etrol. 40, 39 -52.

TsevBn, P.T. (1942): Chrome resources of Cuba. U.S. GeoI.Surv.. Bull. 93- A. l-7 4.

voN KNoRRrNc, O., CoNor-rrs, E. & ToNG, Y.L. (1986): Somemineralogical and geochemical aspects of chromium-bear-ing skarn minerals from northern Karelia, Finland. Bull.Geol. Soc. Finland 58, 27'7 -292.

WAN, H.M. & YsH, C.L. (1984): Uvarovite and grossular fromthe Fengtien nephrite deposits, eastern Taiwan. Mineral.Mag. 48,3I-37.

Weocn, G., Dtepen, G. & LEwrs, J.F. (1984): Ophiolites ofthe northern Caribbean: a reappraisal of their roles in theevolution of the Caribbean plate boundary. 1n Ophiolitesand Oceanic Lithosphere (I.G. Gass, S.J. Lippard & A.W.Shelton, eds.). Geol. Soc., Spec. PubI. 13,367-380.

Received August 20, 1997, revised manuscript accepted Feb-ruarv 15.1999.