Origin of magnetite veins in serpentinite from the Late ... Azzer.pdf · tite deposits in highly serpentinized peridotite of the Bou-Azzer ophiolite, Anti-Atlas, Morocco. The present

www.elsevier.com/locate/jafrearsci

Journal of African Earth Sciences 46 (2006) 318–330

Origin of magnetite veins in serpentinite from the Late ProterozoicBou-Azzer ophiolite, Anti-Atlas, Morocco: An implication

for mobility of iron during serpentinization

Hisham A. Gahlan a,*, Shoji Arai a, Ahmed H. Ahmed b, Yoshito Ishida a,Yaser M. Abdel-Aziz c, Abdellatif Rahimi d

a Department of Earth Sciences, Kanazawa University, Kakuma-machi, Kanazawa 920-1192, Japanb Central Metallurgical Research and Development Institute, Helwan, Cairo, Egypt

c Department of Geology, Faculty of Science, Assiut University, Assiut, Egyptd Department of Geology, Faculty of Science, Chouaib Doukkali University, El Jadida, Morocco

Received 16 September 2005; received in revised form 12 March 2006; accepted 11 June 2006Available online 1 August 2006

Abstract

Vein-like deposits of low-TiO2 (<0.03 wt%) magnetite are present in the mantle section of Bou Azzer ophiolite, Anti-Atlas, Morocco.Two types of magnetite veins, I and II, can be recognized: magnetite is fibrous in Type I, but is stout or idiomorphic in Type II. Mode ofoccurrence and petrography indicate they had formed filling the open space of cracks. Magnetite is Ni-bearing in Type I, and is Ni-freeand sometimes includes Mn-rich chromian spinel in Type II. Magnetite is commonly accompanied by serpentine mainly and magnesite inType I, and by chlorite (clinochlore), serpentine, talc and lesser amount of garnet (andradite) in Type II. In situ trace-element analysisshows depletion with Al, Si, V, Cr and Zn in the vein magnetite relative to the disseminated one in wall serpentinite. Positive slope ofPGE distribution pattern and Au enrichment were the result of hydrothermal activity. The formation of magnetite veins was apparentlyassociated with enrichment of magnetite component in chromian spinel. The iron was supplied from olivine upon serpentinization whichaccompanied the obduction of the Bou-Azzer complex. The iron mobility may have been enhanced by high water/rock ratio of theserpentinization.� 2006 Elsevier Ltd. All rights reserved.

Keywords: Magnetite; Serpentinite; Proterozoic; Bou-Azzer; Ophiolite

1. Introduction

As is well known, magnetite is a nearly ubiquitous acces-sory mineral in a wide variety of igneous and metamorphicrocks. There are many data dealing with the iron oxidedeposits associated with felsic to mafic igneous rocks (e.g.Hitzman et al., 1992; Barton and Johnson, 1996) irrespec-tive of their origin and setting. However, only a scarcity ofstudies have been published on the magnetite deposits thatare embedded within serpentinized ultramafics (e.g. Singh

1464-343X/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jafrearsci.2006.06.003

* Corresponding author. Tel.: +81 90 3760 1976; fax: +81 76 264 6545.E-mail address: [email protected] (H.A. Gahlan).

and Srivastava, 1980; Zucchetti et al., 1988; Diella et al.,1994). A variety of origins have been proposed for them;magmatic (e.g. Singh and Srivastava, 1980; Cocker,1990), metamorphic (e.g. Paraskevopoulos and Economou,1980; Diella et al., 1994) and/or hydrothermal origins (e.g.Zucchetti et al., 1988). We found unique vein-like magne-tite deposits in highly serpentinized peridotite of the Bou-Azzer ophiolite, Anti-Atlas, Morocco.

The present paper aims to describe in detail the geolog-ical, petrographical and geochemical characteristics of twotypes of magnetite veins in serpentinite, and to discuss theirorigin. In addition, we describe in brief the host peridotite,because only limited number of studies on peridotites havebeen published for the Precambrian ophiolites (e.g. Quick,

mailto:[email protected]

H.A. Gahlan et al. / Journal of African Earth Sciences 46 (2006) 318–330 319

1990; Bonavia et al., 1993; Liipo et al., 1995; Vuollo et al.,1995; Suita and Strieder, 1996; Ahmed et al., 2001; Gahlan,2003; Kusky, 2004).

Fig. 2. Schematic lithological column of the Bou-Azzer ophiolite in thestudy area.

2. Geological outline

The Bou-Azzer district is the southernmost ophioliteinlier within the central part of the Moroccan Anti-AtlasMountains, about 300 km east of Agadir city (Fig. 1). Itis situated along the Major Anti-Atlas Fault (Choubert,1947), which makes the contact between the two majorstructural units: the West African Eburnian craton(1900–1650 Ma) to the southwest and the Pan-African oro-genic belt (680–580 Ma) to the northeast (Saquaque et al.,1989; Ennih and Liegeois, 2001). The ophiolite complex ofBou-Azzer forms an elongate belt trending WNW–ESEand covers about 300 km2. According to Leblanc (1981),the Bou-Azzer ophiolite is a piece of Upper Proterozoicoceanic lithosphere (788 ± 9 Ma) obducted onto the north-ern margin of the West African craton during the Pan-Afri-can orogeny. The Bou-Azzer ophiolite has been onlyweakly dismembered and offers a continuous sequencefrom the mantle section (2000 m thick) through the maficcrust (layered gabbro 500 m thick) upward to the sedimen-tary cover (1500 m thick) overlying basalts (500 m thick)(Leblanc, 1976; Bodinier et al., 1984).

The mantle section of Bou-Azzer ophiolite is dominatedby totally serpentinized harzburgite. Less abundant duniteoccurs as lenses and bands within harzburgite. Weobserved small-scale chromitite pods, pyroxenite dykesand late-stage wehrlite intrusions (Fig. 2). The Bou-Azzermantle section is characterized by its high degree of ser-pentinization, and chrysotile asbestos has been exploited,especially along the borders of dialogite body (Leblancand Billaud, 1982). Occasionally the serpentinized harz-burgites exhibiting a tectonic fabric from N55 �W–SE to

Fig. 1. Simplified geological map of the Bou-Azze

E–W with moderate to steep dipping. The dunite bodiesare commonly concordant to sub-concordant (strikingfrom N10 �W–SE to N55 �W–SE) with the surroundingharzburgite foliation. The chromitite pods (striking N50�W–SE) with dunite envelopes are also parallel with theharzburgite foliation. By contrast, pyroxenite is discordantto the harzburgite structure.

Two types of vein-like magnetite deposits (Types I andII) have been discovered and described for the first timewithin the mantle section of the Bou-Azzer ophiolite(Fig. 3A and B). All magnetite veins show a clear-cut con-tact with the host rocks (serpentinized harzburgite)(Fig. 3A and B). No thermal effects were recognized onthe host serpentinites in the field. Veins are thin, <2 cmfor Type I and <20 cm for Type II, and are volumetrically

r district modified from Bodinier et al. (1984).

Fig. 3. Field photographs of magnetite veins of Types I and II in the area of Wadi Ait-Ahmane, Late Proterozoic Bou-Azzer ophiolite complex, Anti-Atlas, Morocco. (A) Type I magnetite vein within serpentinized harzburgite. (B) Type II magnetite vein within serpentinized harzburgite, with markedstriation on the vein wall. (C) Close-up view of the wall of Type I magnetite vein. Note the clear fibrous texture. (D) Close-up view of a slab of Type Imagnetite vein perpendicular to the wall, showing the S-c fabric as a result of dextral heterogeneous simple shear. Note the light green co-precipitatedserpentine aggregates. (E) Close-up view of a sawed surface of Type II magnetite vein showing a massive character. Note the light green serpentine,possible, wall-rock aggregates and the millimeter-size pores. (F) Close-up view of a pin-hole void covered with idiomorphic magnetite crystals in Type IIvein.

320 H.A. Gahlan et al. / Journal of African Earth Sciences 46 (2006) 318–330

insignificant relative to the host mass. Both Type I and IIveins commonly show high relief on the host serpentinite(Fig. 3). Cross-cutting or network veining is common inType I veins, while brecciation is characteristic of Type IIveins, containing serpentinite wall-rock fragments. Thereare no systematic relations between attitudes of the veinsand the foliation of the host serpentinites. Magnetite grainsare megascopically fibrous in Type I veins (Fig. 3C) andsometimes showing S-C fabric with co-precipitated serpen-tine aggregates (Fig. 3D). They are idiomorphic, with con-spicuous light green serpentine aggregates (possibleserpentinite wall-rock fragments) in Type II veins(Fig. 3E). An unusual pin-hole texture occurs at Type IIveins, where the idiomorphic magnetite crystals (<5 mmacross) with well-developed facets (Fig. 3F) are lining thepin-hole voids (<5 cm across). Additionally, the massiveType II magnetite vein is characterized by the presence ofirregular-shaped void spaces (pore spaces) (Fig. 3E).

3. Petrography

3.1. Peridotites

Serpentinite of harzburgite parentage is mainly composedof serpentine (lizardite/chrysotile and rarely antigorite),

magnetite, chlorite, stichtite, chromian spinel and sulfide(chalcopyrite). The relic texture indicates that the initialharzburgite had protogranular texture (Fig. 4A and B).The bastite shows low modal abundances (�10%) of theinitial orthopyroxene. Chromian spinel (61% in volume)is vermicular to anhedral (Fig. 4C) and replaced partly orcompletely by stichtite (Fig. 4D) (e.g. Ashwal and Cairn-cross, 1997).

Serpentinized dunite is homogeneous in appearance andcomposed of serpentine (lizardite/chrysotile with subordi-nate amounts of antigorite), chromian spinel and chlorite.It shows mesh to nonpseudomorphic textures (Fig. 4E)(e.g. O’Hanley, 1996). Kinking in chlorite reflects a post-serpentinization deformation (Fig. 4F). Chromian spinel(62% in volume) is euhedral to subhedral (Fig. 5A) andless altered than that in harzburgite. Magnetite is especiallydeveloped in core rather than rim of individual cells of themesh-textured serpentine.

3.2. Magnetite veins

Microscopically, Type I veins are simple and consist ofmagnetite, serpentine and hematite in the order of abun-dance. Magnetite is commonly fibrous, sometimes radiat-ing toward the vein center (Fig. 5B), and contains

Fig. 4. Photomicrographs of serpentinized peridotites wall rocks of magnetite veins. PL, plane-polarized light. CL, cross-polarized light. (A) Bastite (Opx)after orthopyroxene with marked interstitial habit possibly indicating the protogranular texture of harzburgite. PL. (B) Bastite (Opx) after deeply embayedorthopyroxene with smooth curved grain boundaries indicating the protogranular texture of harzburgite. PL. (C) Vermicular chromian spinel inserpentinized harzburgite indicating non-significant strain after formation. CL. (D) Stichtite partially replacing chromian spinel grain (Sp) in serpentinizedharzburgite. PL. (E) Antigorite blades showing flame texture (Francis, 1956) or thorn texture (Green, 1961) replacing olivine (now serpentinized) in dunite.CL. (F) Kinked chlorite flakes in dunite. CL.


serpentine and magnesite inclusions. Step-like micro-faultsare characteristic of Type I veins (Fig. 5C). Serpentine(antigorite) is anhedral and interstitial to magnetite, andusually encloses fine magnetite crystals.

Type II veins are more complicated in petrography thanType I ones. The former consist of magnetite, serpentine,chlorite, talc, fine-grained talc-carbonate aggregate, garnet(andradite), Mn-rich chromian spinel and hematite. Mag-netite occurs as anhedral to euhedral crystals (Fig. 5D).Serpentine is mainly antigorite (crosshatching needle-likecrystals) with subordinate amounts of lizardite/chrysotile.The subhedral to euhedral garnet (andradite) (Fig. 5E)sometimes poikilitically includes magnetite and vice versa.The euhedral crystals of Mn-rich chromian spinel are char-acteristic of the Type II veins (Fig. 5F).

4. Geochemistry

4.1. Methods

Major- and trace-element concentrations in serpenti-nites were determined by XRF (X-ray fluorescence spec-

trometry; Rigaku 3270 X) at Department of EarthSciences, Kanazawa University. We adopted the methodsof Nakada et al. (1985) and Tamura et al. (1989) formajor- and trace-element analyses, respectively (Table1). Detection limits are 1 ppm for trace-elements. Losson ignition (LOI) was determined by heating the pow-dered sample for 5 h at 800 �C. Mineral analysis formagnetite, silicates and chromian spinel in magnetiteveins and serpentinites was carried out with a JEOL elec-tron-probe micro-analyzer JXA-8800 (WDS) at the Cen-ter for Cooperative Research of Kanazawa University.Analytical conditions were 20 kV accelerating voltage,20-nA probe current and 3 lm probe beam diameter.The raw data were corrected with an online ZAF pro-gram. Fe3+ and Fe2+ in spinel phases were calculatedassuming spinel stoichiometry. All iron is assumed tobe ferrous in silicates except andradite. Mg# is Mg/(Mg + Fe2+) atomic ratio for minerals and Mg/(Mg + to-tal Fe) atomic ratio for bulk rocks. Cr# is Cr/(Cr + Al)atomic ratio. Selected analyses of magnetite, chromianspinel and the associated carbonates and silicates arelisted in Tables 2–4.

Fig. 5. Photomicrographs of Bou-Azzer serpentinized peridotites and magnetite veins. Reflected plane-polarized light except for (E), which was bytransmitted light combined with reflected one. (A) Euhedral chromian spinel from completely serpentinized dunite. (B) Folded serpentine vein along thecentral part of a Type I magnetite vein. Note the radiation of magnetite crystals toward the vein center indicating precipitation from the vein wall inward.(C) Step-like micro-fractures, one of the petrographical characteristics of Type I magnetite vein. (D) Octahedral magnetite crystals in Type II vein. (E)Subhedral andradite (And) marginally replaced by antigorite needles in Type II vein. (F) Euhedral Mn-rich chromian spinel inclusion within Type IImagnetite vein.


Trace-element concentrations were determined onmagnetites from veins and serpentinites as well as onandradite by a laser ablation (193 nm ArF excimer: Micr-olas GeoLas Q-plus)-inductively coupled plasma massspectrometer (Agilent 7500 S) (LA–ICP–MS) at the Incu-bation Business Laboratory Center of Kanazawa Univer-sity (Tables 5 and 6, respectively). Analytical details areshown in Morishita et al. (2005). Each analysis was per-formed by ablating 50 lm diameter spots for magnetiteand andradite at 5 Hz with energy density of 8 J/cm2 perpulse. The primary calibration standard was NIST SRM612 glass, with selected element concentration from thepreferred values of Pearce et al. (1997). The raw dataacquired from the andradite were quantified by using theCaO contents (determined by EPMA) as an internal stan-dard, following a protocol essentially identical to that out-lined by Longerich et al. (1996). While, a semi-quantitativeanalysis was applied for magnetites, because there was noappropriate internal standard element.

Magnetite veins of Types I and II were also analyzed forPGEs (Os, Ir, Ru, Rh, Pt, Pd), Ni, Cu and Au (Table 7).

The analysis was conducted by using ICP–MS, after theNi-sulfide fire assay collection at the Genalysis LaboratoryServices, Australia. Detection limits are 2 ppb for Os, Ir,Ru, Pt and Pd, 1 ppb for Rh, 5 ppb for Au, and 1 ppmfor Ni and Cu.

4.2. Bulk rock chemistry of serpentinite

The serpentinite samples were carefully selected foranalysis to avoid mesoscopic cross-cutting veins or frac-tures filled with carbonates. They are characterized by highand variable Mg #, from 0.89 to 0.94, and a highlyrestricted range of SiO2 (39–41 wt.%). It is noteworthy thatthe serpentinite shows depletion in iron (FeO* down to4.7 wt.%) relative to magnesium, approaching the magne-tite veins (Fig. 6; Table 1). They exhibit a marked depletionin TiO2 (almost nil), K2O (60.01 wt.%), Na2O (0.02–0.06 wt.%), Al2O3 (<0.6 wt.%) and CaO (<0.1 wt.%). Theyare very poor in Rb (61 ppm), Y (=1 ppm), Nb (=1 ppm)and Zr (4–7 ppm), and to lesser extent in Sr (7–18 ppm).Cu, Pb, Sr and Ba in the wall rock are more depleted

Table 1Major (wt.%) and trace-elements (ppm) analyses of Bou-Azzer serpenti-nized harzburgite

Sample no. Wall of Type I Wall of Type II

716A 722 750C 750D

SiO2 40.54 38.94 41.18 38.82TiO2 n.d. n.d. n.d. n.d.Al2O3 0.45 0.55 0.55 0.36FeO* 4.73 8.12 5.89 8.58MnO 0.03 0.11 0.18 0.19MgO 40.45 38.21 39.77 37.42CaO 0.03 0.03 0.11 0.08Na2O 0.05 0.05 0.02 0.06K2O n.d. n.d. 0.01 n.d.P2O5 n.d. n.d. n.d. n.d.LOI 13.03 13.79 11.89 14.16Total 99.32 99.80 99.59 99.66Mg# 0.94 0.89 0.92 0.89

Trace-elements (in ppm)

Ni 2500 2800 3500 3000Cu 6 6 29 44Zn 75 111 72 53Pb n.d. 6 31 11Th 1 2 1 3Rb n.d. n.d. n.d. 1Sr 7 9 18 13Y 1 1 1 1Zr 4 4 4 7Nb 1 1 1 1Co 44 60 53 61Cr 3400 3400 4300 2700V 37 27 36 20Ba n.d. n.d. 4 37

n.d.: Not detected. FeO* as total iron. LOI: loss on ignition.Mg# = Mg/(Mg + Fe*) atomic ratio.

able 2epresentative microprobe analyses (wt.%) of magnetite in Type I and II

eins and the disseminated magnetite in their wall rock serpentinites

Type I Type II

Vein Wall Serp Vein Wall Serp

iO2 0.01 0.11 0.11 0.19iO2 0.00 0.01 0.00 0.01l2O3 0.01 0.01 0.00 0.00r2O3 0.00 0.02 0.02 0.02eO 30.45 30.29 30.08 29.03e2O3 68.59 68.45 67.01 66.65nO 0.06 0.10 0.13 0.26gO 0.23 0.29 0.04 0.54

aO 0.02 0.01 0.01 0.02a2O 0.04 0.01 0.03 0.02

2O 0.02 0.02 0.02 0.02iO 1.14 0.12 0.08 0.10

otal 100.57 99.45 97.54 96.85

umber of ions on the basis of 4 (O)

i 0.000 0.004 0.004 0.007i 0.000 0.000 0.000 0.000l 0.000 0.000 0.000 0.000r 0.000 0.001 0.001 0.001e2+ 0.977 0.979 0.993 0.961e3+ 1.980 1.990 1.990 1.985n 0.002 0.003 0.005 0.009g 0.013 0.017 0.003 0.032

a 0.001 0.000 0.000 0.001a 0.000 0.001 0.002 0.001

0.000 0.000 0.000 0.000i 0.035 0.004 0.003 0.003

otal 3.009 3.000 3.001 3.000


around Type I magnetite than around Type II one (Table1).

The Bou-Azzer serpentinites contain appreciableamounts of compatible trace elements; 2700–4300 ppmCr, 44–61 ppm Co, 2500–3500 ppm Ni, 20–37 ppm V and53–111 ppm Zn, which are relatively immobile during alter-ation (Hebert et al., 1990).

4.3. Mineral chemistry

4.3.1. Magnetite

All the magnetites are very poor in other cations espe-cially Ti (Fig. 7; Table 2). NiO content of magnetite ismarkedly different between the two types; nickeliferouswith 0.4–1.14 wt.% of NiO in Type I, whereas almost Ni-free in Type II. Microprobe analysis indicates slightlyhigher MgO content of magnetite in Type I veins (0.1–0.4 wt.%) than in Type II (60.1 wt.%). The disseminatedmagnetite grains within the wall serpentinite of Type IIveins contain higher amounts of MgO (0.2–1.15 wt.%).LA–ICP–MS analysis clearly demonstrates the differencein magnetite chemistry between the two types; magnetiteis richer in Mg, Co and Ni, and poorer in V and Zn in TypeI than in Type II (Table 5). The analysis detected B, K, Ti,

TRv

STACFFMMCNKN

T

N

STACFFMMCNKN

T

Cr, Cu, Ge, As, Rb, Sr and Ba only on magnetite of TypeII veins (Table 5). By contrast, the disseminated magnetiteswithin the wall serpentinite are richer in Al, Si, V, Cr andZn than those of the veins (Table 5).

4.3.2. Chromian spinel

Chromian spinel in the wall serpentinized harzburgite ishigh in Cr2O3 content, and the Cr# ranges from 0.6 to 0.8(Figs. 8 and 9). Mg# shows a weak negative correlationwith Cr# (Fig. 9). TiO2 and NiO contents and [Fe3+/(Fe3+ + Cr + Al) atomic ratio] are generally very low,0.03 and 0.1 wt.% and 0.1, on average, respectively (Table3). Chromian spinel is characteristically higher in ferrousand ferric irons in serpentinites around Type II veins thanthose around Type I (Table 3; Fig. 10).

The high MnO and ZnO contents (up to 16.4 and2.1 wt.%, respectively) (Table 3, No. 2) are characteristicof chromian spinel from the serpentinite around Type IIveins. However, much higher MnO content (up to18 wt.%) is recorded in the Mn-rich chromian spinel inclu-sions within Type II magnetite (Table 3, No. 6). From thecore to ferritchromite rim (Table 3, No. 3 and 4, respec-tively) Al2O3, Cr2O3 and MgO decrease while FeO* (totaliron), MnO, NiO, V2O3 and ZnO increase (Table 3). Wefound unusual Ni, Mn, Zn-rich ferritchromite in the wallserpentinite of Type II veins (Table 3, No. 5).

Table 3Representative microprobe analyses (wt.%) of spinels and stichtite in the wall serpentinites around Type I and II veins

Rock type Type I Type II Types I and II

Serpentinite Serpentinite Mt veins SerpentiniteAnalysis no. 1 2 3 4 5 6 Stichtite

SiO2 0.02 0.33 0.02 0.04 0.00 0.01 0.28TiO2 0.01 0.02 0.00 0.02 0.03 0.04 0.00Al2O3 14.84 14.10 13.06 1.03 3.50 12.11 0.61Cr2O3 51.32 40.50 45.55 29.05 38.50 48.61 19.82FeO 19.98 8.63 23.13 27.02 8.93 6.08 0.00Fe2O3 3.56 12.29 11.18 29.02 26.41 7.42 4.02MnO 0.55 16.40 0.53 11.68 9.28 18.02 0.03MgO 9.17 3.50 7.09 0.28 3.06 4.23 40.68CaO 0.01 0.01 0.01 0.02 0.18 0.02 0.00Na2O 0.01 0.10 0.03 0.01 0.16 0.09 0.00K2O 0.02 0.03 0.01 0.02 0.01 0.00 0.02NiO 0.03 0.31 0.16 0.35 1.34 0.05 0.16ZnO 0.57 2.12 0.36 0.24 7.93 2.07 n.d.V2O3 0.14 0.19 0.11 0.42 0.00 0.11 n.d.Total 100.22 98.52 101.24 99.19 99.31 98.85 65.61Mg# 0.47 0.42 0.37 0.03 0.44 0.54 0.95Cr# 0.70 0.66 0.70 0.95 0.88 0.73YFe 0.04 0.13 0.12 0.48 0.32 0.07

Mg# = Mg/(Mg + Fe2+) atomic ratio. n.d.: Not detected.Cr# = Cr/(Cr + Al) atomic ratio.YFe = Fe3+/(Fe3+ + Al + Cr) atomic ratio.

Table 4Representative microprobe analyses (wt.%) of the associated minerals within magnetite veins

Mineral Type I Serp Type II Serp Andaradite Clinochlore Cr-clinochlore Talc

SiO2 41.60 42.42 35.98 30.77 30.38 63.87TiO2 0.00 0.01 0.02 0.01 0.02 0.01Al2O3 0.13 0.73 0.26 18.07 16.76 0.12Cr2O3 0.04 0.08 0.00 0.64 2.10 0.00FeO* 1.42 3.48 27.32 3.12 0.75 1.24MnO 0.02 0.18 0.34 0.02 0.01 0.03MgO 36.34 35.56 0.21 31.61 31.90 30.17CaO 0.04 0.15 33.37 0.01 0.00 0.01Na2O 0.00 0.03 0.00 0.00 0.00 0.06K2O 0.02 0.03 0.02 0.02 0.01 0.02NiO 0.15 0.31 0.00 0.14 0.15 0.21Total 79.76 82.97 97.52 84.44 82.13 95.73Mg# 0.98 0.94 0.95 0.99 0.98

Mg# = Mg/(Mg + Fe2+) atomic ratio.FeO* = Fe2O3 for andradite.


4.3.3. Silicates and carbonates

Serpentine, chlorite and talc are all high inMg# > 0.94 (Table 4). There is no systematic change inchemistry of serpentine between the veins and wall rockserpentinite. Serpentine is higher in MnO, NiO andAl2O3 contents in Type II veins than in Type I veins(Table 4). Chlorite is classified as clinochlore (Hey,1954) and is sometimes chromiferous (Table 4). Andra-dite is chemically homogeneous and is very low inMnO and Al2O3 (Table 4). Trace-element analysis byLA–ICP–MS on andradite shows a moderate heterogene-ity between core and rim (Table 6). In the chondrite-nor-malized distribution pattern, LRRE are generally higherthan HREE except Y, which shows a marked positive

anomaly higher than the chondrite value. Eu shows aslightly positive anomaly and a gentle negative slopefrom Eu to Lu (Fig. 11).

4.4. Bulk-rock PGE geochemistry

Type I and II magnetite veins show very low PGE con-tents (

PPGE = 24 and 13 ppb, respectively). They exhibit

nearly flat chondrite-normalized PGE patterns (Fig. 12),with a slightly positive slope from Ir to Pd [(Pd/Ir)ratio P 1]. PGE data of serpentinite and a separate of dis-seminated magnetite crystals (Leblanc and Fischer, 1990)are shown for comparison (Fig. 12). All show similarly flatdistribution patterns; however, magnetites of Types I and

Table 5Representative LA–ICP–MS trace-elements (ppm) analyses of magnetitein Type I and II veins, and the disseminated magnetite in their wall rockserpentinites

Type I Type II

Vein Wall Serp Vein Wall Serp

Idiomorphic Sp-bearing

Li 0.4 Nil 1 1 2B Nil 8 Nil 16 17Na Nil Nil Nil Nil 18Mg 1000 32,000 97 130 47,000Al Nil 860 Nil 7 3500Si Nil 4300 Nil Nil 12,000K Nil Nil Nil 350 720Sc Nil 4 Nil Nil 2Ti Nil Nil 3 9 24V 1 180 15 16 100Cr Nil 6700 Nil 110 18,000Mn 400 500 470 320 1600Co 720 200 4 2 92Ni 6000 1700 1100 350 2600Cu Nil Nil Nil Nil 1Zn 3 85 20 18 530Ga Nil 1 Nil Nil 1Ge Nil Nil 1 8 NilAs Nil Nil Nil 3 18Rb Nil Nil 1 1 3Sr Nil 0.3 Nil 4 13Y Nil Nil Nil Nil 0.5Zr Nil Nil Nil Nil 2Nb Nil Nil Nil Nil 0.1Ag Nil Nil Nil Nil 0.2In Nil Nil Nil Nil 0.1Sn Nil Nil Nil Nil 180Ba Nil 0.5 Nil 14 31La Nil Nil Nil Nil 1Ce Nil 0.3 Nil Nil 1Pr Nil Nil Nil Nil 0.1Nd Nil Nil Nil Nil 0.2W Nil Nil Nil Nil 0.5Pb Nil 2 Nil Nil 4

Idiomorphic: octahedral magnetite crystals.Sp-bearing: Mn-rich chromian spinel bearing magnetite.

Table 6Representative LA–ICP–MS trace-elements (ppm) analyses of andraditecore and rim in Type II magnetite veins

Rim Core Detection limit

B 28 51 3.30Al (wt.%) 0.2 0.2 0.01Si (wt.%) 36 36 0.01Ca (wt.%) 34 34 0.01Sc 2 1 0.30Ti Nil 3 2.20V 1 4 0.20Mn 2600 2300 0.80Fe (wt.%) 27 27 0.02Co Nil 0.3 0.10Ni 2 28 0.35Cu 1 2 0.80Zn 6 65 1.30Ga 0.4 0.5 0.15Ge 16 18 1.50Rb Nil Nil 0.60Sr 0.1 0.4 0.05Y 6 9 0.03Zr 0.1 0.2 0.02Ba 0.3 1 0.02La 4 3 0.01Ce 34 27 0.02Pr 8 8 0.01Nd 40 46 0.07Sm 6 10 0.10Eu 4 6 0.03Gd 4 7 0.06Tb 0.3 0.5 0.01Dy 1 2 0.07Ho 0.2 0.3 0.01Er 0.4 0.6 0.05Tm 0.1 0.1 0.02Yb 0.4 0.5 0.07Lu 0.1 0.1 0.02W 36 22 0.10Pb 0.1 Nil 0.03Th Nil 0.1 0.01

Values in wt.% are obtained by EPMA.

Table 7Bulk contents of PGEs (ppb), Au (ppb), and Cu and Ni (ppm) of Type Iand II magnetite veins

Type I Type II Serp DMt

Os <2 <2 4 5Ir <2 <2 3.5 1.9Ru 5 6 5.4 12Rh <2 2 1 1Pt 11 5 4 6Pd 8 <2 2.5 4.6

Total PGE 24 13 20 30

Au 8 13 9 10Pd/Ir >1 �1 0.71 2.42Ni 9200 1100 n.d. n.d.Cu 2 272 n.d. n.d.

Averaged compositions of PGEs and Au for serpentinites (Serp) and thedisseminated magnetite crystals (DMt) are listed after Leblanc and Fischer(1990) for comparison. n.d.: Not determined.


II have slightly positive slopes from Os to Rh. Type II mag-netite vein is richer in Cu than Type I one (Table 7).

5. Discussion

5.1. Characteristics of the serpentinized peridotites around

magnetite veins

The possible mineral assemblage (olivine–tremolite–antigorite–ferritchromite/magnetite) before the latest ser-pentinization is consistent with the greenschist faciesmetamorphism (Evans, 1977) that had been partly imposedon the Bou-Azzer ophiolite (e.g. Clauer, 1976; Leblanc,1981). The metamorphism and subsequent alteration havesubstantially modified the primary petrological and chem-ical characteristics of the Bou-Azzer peridotite. We were,however, able to identify the primary lithologies by the

Fig. 6. Variation of FeO* (total iron) versus Mg# of the bulk serpenti-nized harzburgite wall of magnetite veins. The arrow refers to thedepletion in iron toward the magnetite veins.

Fig. 7. Comparison of major-element compositions of magnetite; Type Iand II veins and their wall serpentinite (disseminated).

Fig. 8. Relations between Cr2O3 and Al2O3 (wt.%) of chromian spinel inserpentinized harzburgite around magnetite veins.

Fig. 9. Relations between Cr# and Mg# of chromian spinel in serpen-tinized harzburgite around magnetite veins.


aid of relic textures and chromian spinel chemistry andmorphology (e.g. Arai, 1980; Liipo et al., 1995; Matsumotoand Arai, 2001). The relic protogranular texture and ver-micular shape of chromian spinel indicate that harzburgitehas not experienced any significant deformation (Fig. 4A–C).

Serpentinization may be an essentially isochemical pro-cess (Coleman and Keith, 1971; Donaldson, 1981; Shiga,1983, 1987). Bulk serpentinite shows high Mg# accompa-nied by enrichment of Ni, Cr and Co, and a marked deple-tion in the incompatible elements reflecting the depletedand highly refractory nature of the initial peridotite (e.g.Roberts, 1992). Ca and mobile incompatible trace ele-

ments, however, have been also leached from the perido-tites, if any, during serpentinization. Ni, Cr, Co, V andperhaps Zn are compatible and relatively ‘‘immobile’’ dur-ing alteration (e.g. Hebert et al., 1990; Pereira et al., 2003).The high Ni, Cr and Co of the bulk serpentinites combinedwith the high Cr# (up to 0.8) of the intact spinels suggesthigh degrees of partial melting (e.g. Gulacar and Delaloye,1976; Dick and Bullen, 1984; Arai, 1994; Ishiwatari et al.,2003; Ahmed et al., 2005). The highly depleted harzburgitewith high Cr# (P0.7) of spinel is very common to the Pre-cambrian ophiolites (e.g. Quick, 1990; Liipo et al., 1995;

0.05

0.10

0.15

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.1 0.2 0.3 0.4 0.5 0.6

Chromian Spinel

Type II wall

Type I wall

Mg/(Mg + Fe2+) atomic ratio

Fe3+

/(F

e3++

Cr+

Al)

ato

mic

rat

io

0.000.0

Alteration trend

Fig. 10. Relations between Mg# and [Fe3+/(Cr + Al + Fe3+) atomicratio] of chromian spinel in serpentinized harzburgite around magnetiteveins.


Vuollo et al., 1995; Ahmed et al., 2001; Gahlan et al.,unpublished), and not uncommon in the Phanerozoic ones(e.g. England and Davies, 1973; Arai, 1997; Tamura et al.,1999; Ishiwatari et al., 2003 and references therein). Thehighly depleted Bou-Azzer mantle harzburgite with high-Cr# and low-TiO2 spinel suggests mantle wedge or sub-arc mantle origin (e.g. Arai, 1992).

5.2. Genesis of the magnetite veins from Bou-Azzer ophiolite

5.2.1. Source of Fe for the magnetite veins

The tectonic emplacement of Bou-Azzer ophiolite wasaccompanied by low-grade regional metamorphism (e.g.Leblanc, 1981). Fluids composed largely of meteoric waterswere activated by that regional metamorphism. The hydro-thermal fluids caused serpentinization and mobilized Feand the other elements from the ferromagnesian mineralsin the initial peridotite of Bou-Azzer serpentinite and pre-cipitated magnetite on decrease of temperature (e.g. Park

Fig. 11. Chondrite-normalized trace-element patterns of andradite. Note theNormalization values are of McDonough and Sun (1995).

and MacDiarmid, 1964). The internal source of Fe maybe shown by its distribution in serpentinite; the highMg# and low FeO* content of the serpentinized harzburg-ite close to magnetite veins probably reflect extraction ofiron from the wall rocks (Fig. 6; Table 1). The hydrother-mal fluids have utilized shears, cracks and tension joints intransportation and concentration processes. Therefore, thecracks that are filled with magnetite as veins are borderedby Fe-deficient harzburgite (Fig. 6) (e.g. Ingerson, 1954).

Due to the absence of Cl- and S-bearing phases in Bou-Azzer magnetite veins, we suggest that iron was transportedas an aqueous colloidal solution of ferrous hydroxide.Upon dehydration, the ferrous hydroxide is oxidized tomagnetite (e.g. Shand, 1947), according to the equation:3Fe (OH)2 = Fe3O4 + 2H2O + H2. Presence of carbonatesand stichtite in and around the magnetite veins, respec-tively, indicates an involvement of CO2.

The concentration of Fe as magnetite veins is associatedby adding magnetite component in chromian spinel uponalteration (Fig. 10). Iron released from olivine on serpent-inization could enrich chromian spinel with magnetite com-ponent (e.g. Barnes, 2000). In ordinary serpentinizedperidotites, the Fe released from olivine was precipitatedas in situ fine discrete magnetite particles. In the Bou-Azzerharzburgite, however, Fe as well as Mn and Zn may havebeen more mobile possibly due to higher water/rock ratioduring serpentinization (Gahlan and Arai, submitted forpublication) and was precipitated as magnetite veins. Fer-rous and ferric irons (Fig. 10), MnO and ZnO contentsof chromian spinel in serpentinite are much lower aroundType I veins than Type II (Table 3). This may be due toa higher water/rock ratio for the latter. The manganeseenrichment in chromian spinel is usually attributed to alter-ation or metasomatism (e.g. Abzalov, 1998; Shcheka et al.,2001).

The difference in geochemical signature between Type Iand II magnetite veins suggests that the involved brine waschemically different between the two types of veins. Thechemical signature of andradite suggests that the brinefor Type II vein was initially rich in B, Si, Ca, Mn, Feand Y, and poor in Al (Table 6 and Fig. 11). It is believedthat the small-scale difference between the andradite coreand rim (Table 6) is the result of: either self-organization

slightly but clear difference in REE contents between the core and rim.

Fig. 12. Chondrite-normalized platinum-group element patterns of serp-entinites (S), disseminated magnetite crystals (DMt), and Type I (I) andType II (II) magnetite veins. Data of (S) and (DMt) are after Leblanc andFischer (1990) for comparison. PGE and Au values for normalization arefrom Naldrett and Duke (1980).


during growth or influx of fresh fluid. The sign and magni-tude of Eu anomaly is related to the redox condition. Themarkedly positive Y anomaly can be attributed to the pref-erential retention of Y in the hydrothermal fluid.

PGE can be slightly redistributed and concentrated dur-ing the late stage hydrothermal activity (e.g. Dillon-Leitchet al., 1986; Rowell and Edgar, 1986). Au, PPGE and, tolesser extent, Ru are easily leached by hydrothermal solu-tions during serpentinization, while Os and Ir remain inchromian spinel (e.g. Barnes et al., 1985; Oberger et al.,1988; Bai et al., 2000). The Au and PPGE enrichment rel-ative to Os and Ir in Bou-Azzer magnetite veins (Fig. 12) isconsistent with the hydrothermal origin. The PGE signa-ture of Bou-Azzer magnetite veins (especially of Type II)is consistent with Keays et al. (1982) hypothesis that hydro-thermal deposits are depleted in Ir and to lesser extent Pdrelative to magmatic ones.

The serpentine mineral species (lizardite/chrysotilerarely antigorite) co-precipitated with magnetite indicates100–300 �C (Wenner and Taylor, 1971) for the formationof Bou-Azzer magnetite veins. This is consistent with thelow temperature of formation for the hydrothermal (Cu–U–Au–REE) magnetite deposits (e.g. Hitzman et al.,1992 and references therein).

5.2.2. Mechanism of the magnetite veins formation

The Bou-Azzer magnetite vein deposits are unique andunequivocally of hydrothermal origin. Absence of thermaleffects on the wall serpentinite may deny the possibility ofinvolvement of high-temperature melts. Their featuresinclude hydrothermal textures as veins (e.g. Frondel andBaum, 1974); open-space filling and brecciation suggestthat fluid pressures were locally higher than lithostaticpressure (e.g. Jebrak, 1997; Kent et al., 2000). Growth ofidiomorphic magnetite (Type II) along the void’s wallsindicates precipitation from aqueous fluid, rather thanfrom gas alone (Sillitoe and Burrows, 2002). The open-space features in Type II magnetite veins are believed to

represent the passage of brine, not solely gas, at the timeof magnetite deposition (Sillitoe and Burrows, 2002).

Radiation of the fibrous Type I magnetite (Fig. 5B)toward the vein center may indicate its growth from thevein walls toward the center. Distortion of magnetitefibers or folding of serpentine veins along the centralplane of Type I magnetite veins (Fig. 5B) indicates thatthe formation of magnetite was accompanied by deforma-tion. This is also supported by S-c fabric (Fig. 3D) as fieldevidence of the heterogeneous simple shear. The relationof serpentine (antigorite) (and andradite for Type II veins)enclosing fine magnetite crystals, and vice versa, indicatesthat all of them were precipitated from the same fluidsimultaneously.

The fibrous and idiomorphic habits of magnetite inType I and Type II veins, respectively, were strongly con-trolled by interrelated variables such as diffusion rate ofinvolved components, growth rate of involved minerals,temperature, degree of supersaturation and viscosity ofinvolved fluids (e.g. Jackson, 1958; Kirkpatrick, 1975).Hirano and Somiya (1976) found a very important roleplayed by hydrogen in producing euhedral magnetite crys-tals grown from an aqueous fluid. Furthermore, Tiller(1977) also emphasized the importance of hydrogen con-tent of the gas phase in producing euhedral crystals ingeneral.

6. Conclusions

(1) The magnetite veins are found in serpentinized harz-burgite of the Proterozoic Bou-Azzer ophiolite andare considered to be epigenetic deposits formed alongwith the serpentinization process during and after theobduction of the ophiolite complex.

(2) Magnetite is fibrous, Ni-rich and Cu-poor in Type Iveins, and is stout or idiomorphic, Ni-poor, Cu-rich,and contains garnet (andradite) and Mn-rich chro-mian spinel in Type II veins.

(3) The hydrothermal solution for Type I magnetite veinsis slightly different in chemistry from that for Type II.

(4) The source of iron was internal, supplied from olivineupon serpentinization. The mineralizing hydrother-mal fluids utilized shears, cracks and tension jointsfor transportation and precipitation of iron, whichtransported as ferrous hydroxide. The intensity oftransportation and precipitation of iron was mainlygoverned by the fluid/ rock ratio, which apparentlylower in Type I area than Type II area.

Acknowledgements

This paper is taken from a part of the Ph.D. thesis of thefirst author. One of the authors (H.A.G.) wishes to expresshis sincere appreciation to Dr. M. Shirasaka, Mr. K. Koiz-umi and Mr. H. Inoue for their help in the laboratory. Spe-cial thanks for the staff members of Bou-Azzer district


mine for accommodation during our field work. We aregrateful to Prof. M. Economou-Eliopoulos, anonymous re-viewer and Prof. S. Muhongo for their constructive com-ments and the editorial handling, respectively.

References

Abzalov, A.Z., 1998. Chrome-spinel in gabbro–wehrlite intrusions ofPechenga area, Kola peninsula, Russia: emphasis on alterationfeatures. Lithos 43, 109–134.

Ahmed, A.H., Arai, S., Attia, A.K., 2001. Petrological characteristics ofpodiform chromitites and associated peridotites of the Pan AfricanProterozoic ophiolite complexes of Egypt. Mineralium Deposita 36,72–84.

Ahmed, A.H., Arai, S., Abdel-Aziz, Y.M., Rahimi, A., 2005. Spinelcomposition as a petrogenetic indicator of the mantle section in theNeoproterozoic Bou Azzer ophiolite, Anti-Atlas, Morocco. Precam-brian Research 138, 225–234.

Arai, S., 1980. Dunite–harzburgite–chromitite complexes as refractoryresidue in the Sangun-Yamaguchi zone, western Japan. Journal ofPetrology 21, 141–165.

Arai, S., 1992. Chemistry of chromian spinel in volcanic rocks as apotential guide to magma chemistry. Mineralogical Magazine 56, 173–184.

Arai, S., 1994. Characterization of spinel peridotites by olivine-spinelcompositional relationships: review and interpretation. ChemicalGeology 113, 191–204.

Arai, S., 1997. Control of wall-rock composition on the formation ofpodiform chromitites as a result of magma/peridotite interaction.Resource Geology 47, 177–187.

Ashwal, L.D., Cairncross, B., 1997. Mineralogy and origin of stichtite inchromite-bearing serpentinites. Contributions to Mineralogy andPetrology 127, 75–86.

Bai, W., Robinson, P.T., Fang, Q., Yang, J., Yan, B., Zhang, Z., Hu, X-F.,Malpas, J., 2000. The PGE and base-metal alloys in the podiformchromitites of the Luobusa ophiolite, southern Tibet. CanadianMineralogist 38, 585–598.

Barnes, S.J., 2000. Chromite in Komatiites, II. Modification duringgreenschist to mid-amphibolite facies metamorphism. Journal ofPetrology 41, 387–409.

Barnes, S-J., Naldrett, A.J., Gorton, M.P., 1985. The origin of thefractionation of platinum-group elements in terrestrial magmas.Chemical Geology 53, 303–323.

Barton, M.D., Johnson, D.A., 1996. Evaporitic-source model for igneous-related Fe-oxide–(REE–Cu–Au–U) mineralization. Geology 24, 259–262.

Bodinier, J.L., Dupuy, C., Dostal, J., 1984. Geochemistry of Precambrianophiolites from Bou-Azzer, Morocco. Contributions to Mineralogyand Petrology 87, 43–50.

Bonavia, F.F., Diella, V., Ferrario, A., 1993. Precambrian podiformchromitites from Kenticha Hill, southern Ethiopia. Economic Geology88, 198–202.

Choubert, G., 1947. L’accident majeur de l’Anti-Atlas. Comptes Rendusde l’Academie des Sciences (Paris) 224, 1172–1173.

Clauer, N., 1976. Geochimie isotopique du strontium des milieuxsedimentaires. Application a la geochronologie du craton oust-africain. These de Doctorat d’Etat Es-Science, Sciences Geologiques,Strasbourg, France, p. 256.

Cocker, M.D., 1990. Mineralogy, geochemistry and genesis of REE–Ti–Venriched iron oxides in the Burks Mountain complex, ColumbiaCountry, Georgia. Geological Society of America, Abstracts withProgram 22, 8.

Coleman, R.G., Keith, T.E., 1971. A chemical study of serpentinization –Burro Mountain, California. Journal of Petrology 12, 311–328.

Dick, H.J.B., Bullen, T., 1984. Chromian spinel as a petrogenetic indicatorin abyssal and Alpine-type peridotites and spatially associated lavas.Contributions to Mineralogy and Petrology 86, 54–76.

Diella, V., Ferrario, A., Rossetti, P., 1994. The magnetite ore deposits ofthe southern Aosta Valley: chromitite transformed during an alpinemetamorphic event. Ofioliti 19, 247–256.

Dillon-Leitch, H.C.H., Watkinson, D.H., Coats, C.J.A., 1986. Distribu-tion of platinum-group elements in the Donaldson West Deposit, CapeSmith Belt, Quebec. Economic Geology 81, 1147–1158.

Donaldson, M.J., 1981. Redistribution of ore elements during serpenti-nization and talc – carbonate alteration of some Archean dunites,Western Australia. Economic Geology 76, 1698–1713.

England, R.N., Davies, H.L., 1973. Mineralogy of ultramafic cumulatesand tectonites from eastern Papua. Earth and Planetary ScienceLetters 17, 416–425.

Ennih, N., Liegeois, J-P., 2001. The Moroccan Anti-Atlas: the WestAfrican craton passive margin with limited Pan-African activity.Implications for the northern limit of the craton. PrecambrianResearch 112 (3–4), 289–302.

Evans, B.W., 1977. Metamorphism of Alpine peridotites and serpentinites.Annual Reviews of Earth and Planetary Sciences 5, 397–447.

Francis, G.H., 1956. The serpentinite mass in Glen Urquhart, Inverness-shire, Scotland. American Journal of Science 254, 201–226.

Frondel, C., Baum, J.L., 1974. Structure and mineralogy of the Franklinzinc–iron–manganese deposit, New Jersey. Economic Geology 69,157–180.

Gahlan, H.A., 2003. Geology of the area of Wadi Fiqo, South EasternDesert, Egypt. M.Sc. Thesis, Assiut University, Assiut, Egypt, p. 141.

Gahlan, H.A., Arai, S., submitted for publication. Genesis of peculiarlyzoned Co, Zn and Mn-rich chromian spinel in serpentinite of Bou-Azzer ophiolite, Anti-Atlas, Morocco. Journal of Mineralogical andPetrological Sciences.

Green, D.H., 1961. Ultramafic breccias from the Musa Valley, easternPapua. Geological Magazine 98, 1–26.

Gulacar, O.F., Delaloye, M., 1976. Geochemistry of nickel, cobalt andcopper in Alpine-type ultramafic rocks. Chemical Geology 17, 269–280.

Hebert, R., Adamson, A.C., Komor, S.C., 1990. Metamorphic petrologyof ODP Leg 109, Hole 670A serpentinized peridotites: serpentinizationprocesses at a slow spreading ridge environment. In: Detrick, R.,Honnorez, J., Bryan, W.B., Juteau, T., et al. (Eds.), Proceedings of theOcean Drilling Program, Scientific Results, 106/109, College station,Texas (Ocean Drilling Program), pp. 103–115.

Hey, M.H., 1954. A new review of the chlorites. Mineralogical Magazine30, 277–292.

Hirano, S., Somiya, S., 1976. Hydrothermal crystal growth of magnetite inthe presence of hydrogen. Journal of Crystal Growth 35, 273–278.

Hitzman, M.W., Oreskes, N., Einaudi, M.T., 1992. Geological character-istics and tectonic setting of Proterozoic iron oxide (Cu–U–Au–REE)deposits. Precambrian Research 58, 241–287.

Ingerson, E., 1954. Nature of the ore-forming fluids at various stages – asuggested approach. Economic Geology 49, 727–733.

Ishiwatari, A., Sokolov, S.D., Vysotskiy, S.V., 2003. Petrological diversityand origin of ophiolites in Japan and Far East Russia with emphasison depleted harzburgite. In: Dilek, Y., Robinson, P.T. (Eds.),Ophiolites in Earth History. Geological Society, 218. Special Publica-tions, London, pp. 597–617.

Jackson, K.A., 1958. Mechanism of growth. In: Moddin, M. (Ed.), LiquidMetals and Solidification. ASM, Cleveland, Ohio, pp. 174–186.

Jebrak, M., 1997. Hydrothermal breccias in vein-type ore deposits: areview of mechanisms, morphology and size distribution. Ore GeologyReviews 12, 111–134.

Keays, R.R., Nickel, E.H., Groves, D.I., McGoldrik, P.J., 1982. Iridiumand palladium as discriminates of volcanic-exhalative, hydrothermaland magmatic nickel sulfide mineralization. Economic Geology 77,1535–1547.

Kent, A.J.R., Ashley, P.M., Fanning, C.M., 2000. Metasomatic alterationassociated with regional metamorphism: an example from the Willy-ama Supergroup, South Australia. Lithos 54, 33–62.

Kirkpatrick, R.J., 1975. Crystal growth from the melt: a review. AmericanMineralogist 60, 798–814.


Kusky, T.M., 2004. Precambrian Ophiolites and Related Rocks. Elsevier,Amsterdam, 748 pp.

Leblanc, M., 1976. Proterozoic oceanic crust of Bou-Azzer. Nature 261,34–35.

Leblanc, M., 1981. The late Proterozoic ophiolites of Bou-Azzer(Morocco): evidence for Pan African plate tectonics. In: Kroner, A.(Ed.), Precambrian Plate Tectonics. Elsevier, Amsterdam, pp. 436–451.

Leblanc, M., Billaud, P., 1982. Cobalt arsenide orebodies related to anupper Proterozoic ophiolite: Bou-Azzer (Morocco). Economic Geol-ogy 77, 162–175.

Leblanc, M., Fischer, W., 1990. Gold and platinum group elements incobalt–arsenide ores: hydrothermal concentration from serpentinitesource-rock (Bou-Azzer, Morocco). Mineralogy and Petrology 42,197–209.

Liipo, J., Vuollo, J., Nykanen, V., Piirainen, T., Pekkarinen, L., Tuokko,I., 1995. Chromites from the early Proterozoic Outokumpu-Jormuaophiolite belt: a comparison with chromites from Mesozoic ophiolites.Lithos 36, 15–27.

Longerich, H.P., Jackson, S.E., Gunther, D., 1996. Laser ablationinductively coupled plasma mass spectrometry transient signal dataacquisition and analyte concentration calculation. Journal of Analyt-ical Atomic Spectrometry 11, 899–904.

Matsumoto, I., Arai, S., 2001. Morphological and chemical variations ofchromian spinel in dunite–harzburgite complexes from the Sangunzone (SW Japan): implications for mantle/melt reaction and chromititeformation processes. Mineralogy and Petrology 73, 305–323.

McDonough, W.F., Sun, S.-s., 1995. The composition of the earth.Chemical Geology 120, 223–253.

Morishita, T., Ishida, Y., Arai, S., Shirasaka, M., 2005. Determination ofmultiple trace element compositions in thin (<30 lm) laser of NISTSRM 614 and 616 using Laser Ablation–Inductively Coupled Plasma–Mass spectrometry (LA–ICP–MS). Geostandards and GeoanalyticalResearch 29, 107–122.

Nakada, S., Yanagi, T., Maeda, S., Fang, D., Yamaguchi, M., 1985. X-ray fluorescence analysis of major elements in silicate rocks. ScienceReport of Kyushu University 14 (3), 103–105, in Japanese with Englishabstract.

Naldrett, A.J., Duke, J.M., 1980. Platinum metals in magmatic sulfideores. Science 208, 1417–1424.

Oberger, B., Friedrich, G., Woermann, E., 1988. Platinum-group elementsmineralization in the ultramafic sequence of the Acoje ophiolite block,Zambales, Philippines. In: Prichard, H.M., Potts, P.J., Bowles, J.F.W.,Cribb, S.J. (Eds.), Geoplatinum, vol. 87. Elsevier Applied Science,London, pp. 361–380.

O’Hanley, D.S., 1996. Serpentinites. Oxford University press, New York,Oxford, 277 pp.

Paraskevopoulos, G.M., Economou, M.I., 1980. Genesis of magnetite oreoccurrences by metasomatism of chromite ores in Greece. NeuesJahrbuch fur Mineralogie Abhandlungen 140, 29–53.

Park, C.F., MacDiarmid, R.A., 1964. Ore Deposits. W.H. Freeman andCompany, San Francisco and London, 475 pp.

Pearce, N.J.G., Perkins, W.T., Westgate, J.A., Gorton, M.P., Jackson,S.E., Neal, C.R., Chenery, S.P., 1997. A compilation of new andpublished major and trace element data for NIST SRM 610 and NISTSRM 612 glass reference materials. Geostandards Newsletter 21, 115–144.

Pereira, M.D., Shaw, D.M., Acosta, A., 2003. Mobile trace elements andfluid-dominated processes in the Ronda peridotites, South Spain.Canadian Mineralogist 41, 617–625.

Quick, J.E., 1990. Geology and origin of late Proterozoic Darb Zubaydahophiolite, Kingdom of Saudi Arabia. Geological Society of AmericaBulletin 102, 1007–1020.

Roberts, S., 1992. Influence of partial melting regime on the formation ofophiolitic chromite. In: Parson, L.M., Murton, B.J., Browning, P.(Eds.), Ophiolites and their Modern Oceanic Analogues, vol. 60.Geological Society Special Publication, pp. 203–217.

Rowell, W.F., Edgar, A.D., 1986. Platinum-group elements mineralizationin a hydrothermal Cu–Ni sulfide occurrence, Rathbun Lake, North-eastern Ontario. Economic Geology 81, 1272–1277.

Saquaque, A., Admou, H., Karson, J.A., Hefferan, K.P., Reuber, I., 1989.Precambrian accretionary tectonics in the Bou-Azzer El Graara region,Anti-Atlas. Geology 17, 1107–1110.

Shand, S.J., 1947. The genesis of intrusive magnetite and related ores.Economic Geology 42, 634–636.

Shcheka, S.A., Ishiwatari, A., Vrzhosek, A.A., 2001. Geology andpetrology of Cambrian Khanka ophiolite in Primorye (Far EastRussia) with notes on its manganese-rich chromian spinel. EarthScience (Chikyu Kagaku) 55, 265–274.

Shiga, Y., 1983. Origin of nickel and cobalt in ore deposits of theKamaishi mining district, northeastern Japan. Mining Geology 33,385–398, in Japanese with English abstract.

Shiga, Y., 1987. Behavior of iron, nickel, cobalt and sulfur duringserpentinization, with reference to the Hayachine ultramafic rocks ofthe Kamaishi mining district, northeastern Japan. Canadian Mineral-ogist 25, 611–624.

Sillitoe, R.H., Burrows, D.R., 2002. New field evidence bearing on theorigin of the El Laco magnetite deposits, northern Chile. EconomicGeology 97, 1101–1109.

Singh, R.N., Srivastava, S.N.P., 1980. Ophiolites and associated miner-alization in the Naga Hills, north-eastern India. In: Panayiotou, A.(Ed.), Ophiolites-Proceedings, International Ophiolite Symposium,Cyprus 1979. Cyprus Geological Survey, Nicosia, pp. 758–764.

Suita, M.T.F., Strieder, A.J., 1996. Cr-spinel from Brazilian mafic-ultramafic complexes: metamorphic modifications. International Geol-ogy Review 38, 245–267.

Tamura, S., Kobayashi, Y., Shuto, K., 1989. Quantitative analysis of thetrace elements in silicate rocks by X-ray fluorescence method. EarthScience (Chikyu Kagaku) 43 (3), 180–185, in Japanese.

Tamura, A., Makita, M., Arai, S., 1999. Petrogenesis of ultramafic rocksin the Kamuikotan belt, Hokkaido, northern Japan. The Memoirs ofthe Geological Society of Japan 52, 53–68, in Japanese with Englishabstract.

Tiller, W.A., 1977. On the cross-pollination of crystallization ideasbetween metallurgy and geology. Physics and Chemistry of Minerals 2,125–151.

Vuollo, J., Liipo, J., Nykanen, V., Piirainen, T., Pekkarinen, L., Tuokko,I., Ekdahl, E., 1995. An early Proterozoic podiform chromitite in theOutokumpu ophiolite complex, Finland. Economic Geology 90, 445–452.

Wenner, D.B., Taylor Jr., H.P., 1971. Temperature of serpentinization ofultramafic rocks based on O18/O16 fractionation between coexistingserpentine and magnetite. Contributions to Mineralogy and Petrology32, 165–185.

Zucchetti, S., Mastrangelo, F., Rossetti, P., Sandrone, R., 1988. Serpent-inization and metamorphism: their relationships with metallogeny insome ophiolitic ultramafics from the Alps. In: Zuffar’ Days Sympo-sium in Honor of Piero Zuffardi, University of Cagliari, Cagliari,October 10–15, 1988, pp. 137–159.

Documents

Origin of magnetite veins in serpentinite from the Late ... Azzer.pdf · tite deposits in highly serpentinized peridotite of the Bou-Azzer ophiolite, Anti-Atlas, Morocco. The present