Marine Geophysical Researches 21: 351–385, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands. 351
The high temperature reaction zone of the Oman ophiolite: new ﬁeld data,
microthermometry of ﬂuid inclusions, PIXE analyses and oxygen isotopic
Thierry Juteau1, Gilles Manac’h1, Olivier Moreau1, Christophe L´
ecuyer2& Claire Ramboz3
1UMR 6538 ‘Domaines oc´eaniques’, IUEM / UBO, place Nicolas Copernic, 29280 Plouzan´e, France
2Ecole Normale Sup´erieure de Lyon, France
3ISTO-CNRS, Orl´eans, France
Received 10 February 2000, accepted 31 August 2000
Key words: Oman ophiolite, hydrothermalcirculation, epidosites, plagiogranites, ﬂuid inclusions, oxygen isotopes
The present study is focused on the so-called High Temprature Reaction Zone of the Oman ophiolite, a thin zone located between the roots of
the sheeted dyke complex and the high-level gabbros marking the roof of the fossil magma chambers. The distribution of diabases, chloritised
dykes, spilitized dykes and epidosites (in the order of increasing hydrothermal alteration) was studied along continuous outcrops of the lower
sheeted dyke complex in three selected areas. The Muwaylah section, in the Haylayn massif, representing a fossil axial discontinuity, is the most
massively epidotised area, with epidosite zones of 3–15 m wide and an average spacing of 12 m. In this area, there are two directions of dyking,
and a massive sulﬁde deposit (Daris prospect). The wadi Andam area (Samail massif), representing a much less tectonised site, close to a mantle
diapir and located probably midway between the tip and the center of an accretion segment, is also well epidotised, with epidosites spacing of
10–25 m. The wadi Salahi area probably represents the central part of a fossil accretion segment and is by far the least altered site, in spite of
being located quite beneath the Zuha sulﬁde prospect. This conﬁrms that the hot ascending ﬂuids in major discharge zones are strongly focused.
Fluid inclusions from about 50 samples collected in the plagiogranites, gabbros and sheeted dykes of the Muwaylah and wadi Falah areas (SE
Haylayn massif), consist of: (1) monophase (liquid) inclusions, generally stretched and deformed; (2) vapour-dominated, low-salinity, 2-phase
inclusions, with average salinity of 3.8 wt% eq. NaCl and an average Th of 370 ◦C; (3) liquid-dominated, low-salinity, 2-phase inclusions, with
an average salinity of 4.2 wt% eq. NaCl and an average Th of 325 ◦C, and (4) liquid-dominated, high-salinity inclusions, containing a solid
halite daughter phase, dissolving at higher temperature (292–441 ◦C) than homogenisation of the ﬂuid phases (230–403 ◦C). One plagiogranite
sample collected in the Muwaylah area (Haylayn massif) is particularly rich in quartz-epidote hydrothermal veins. Xenomorphic globular
quartz is rich in high-salinity, brine-rich inclusions, with halite cubes dissolving at very high temperatures (358–496 ◦C) yielding very high
salinities (43 to 59.2% NaCl eq.), and with liquid-vapour homogenisation of 275 ◦C on average. Molecular Raman Spectroscopy analyses have
conﬁrmed the aqueous nature of these inclusions and the absence of detectable CO2,NH
2and SO2. Daughter solid phases other than NaCl
were determined as hematite and anhydrite, and a third phase is hydroxyl-bearing (amphibole?). PIXE analyses on six brine-rich inclusions
allowed to detect signiﬁcant but variable contents in Cl, Fe, Mn, K, Ca, Zn and Br. Copper, remarkably, was never detected, and two measured
Cl/Br ratios are close to that of seawater. The measurement of oxygen isotopic ratios of the ﬂuids extracted from some ﬂuid inclusions and of
associated host-minerals (quartz, epidote) suggest that both seawater-derived and magma-derived ﬂuids have mixed in the High Temperature
Reaction Zone. PIXE data yield a similar conclusion, based on the contrasted Fe and Mn-contents and Cl/Br ratios of the analysed inclusions.
However, the oxidised and Cu-depleted nature of the brine-rich inclusions suggests that the magma-derived or seawater-derived brines are
residual liquids that have degassed. Present V-Xproperties of most NaCl-saturated inclusions do not keep the record of the boiling process, as
they homogenise by halite disappearance and not by vapour disappearance. Probably they have been modiﬁed by several post-trapping changes,
for instance by necking.
The high temperature reaction zone: oceanic versus
Submarine hydrothermal systems observed at mid-
ocean ridges are thought to be composed of three
main parts (Alt, 1995): (a) a widespread and dif-
fuse recharge zone, where cold seawater enters the
crust, is heated progressively and reacts at increas-
ing temperatures as it penetrates downwards into the
crust; (b) a so-called High Temperature Reaction Zone
(HTRZ), located just aboveaxial heat sources (magma
chambers or hot crystalline rocks), where seawater is
transformed to a hot hydrothermal ﬂuid after intense
chemical exchanges with the rocks making up the roof
of the accretion magma chambers (diabases, gabbros,
plagiogranites); (c) a discharge zone through which
the hot and buoyant hydrothermal ﬂuid rises adiabati-
cally and rapidly toward the seaﬂoor, the upward ﬂow
being either diffuse or strongly focused. In this latter
case, black-smoker activity and massive sulﬁde de-
posits, rich in copper and zinc, may be observed on
Subsurface HTRZ should be typically located at a
depth of 1.5–2.5 km below the seaﬂoor, where steep
temperature gradients develop at the roof of accretion
magma chambers (Fehn et al., 1983; Sleep, 1991;
Nehlig, 1993; Alt, 1995). This level corresponds pet-
rographically to the base of the sheeteddyke complex
(when it exists), and to the high-level gabbros. In
modern oceans, this level is difﬁcult to observe. Only
gabbros from slow-spreading ridges, or collected in
oceanic fracture zones, could be sampled by sub-
mersible (Kelley and Delaney, 1987), or drilled by
the DSDP/ODP (Dick et al., 1991; Mevel and Can-
nat, 1991). On faster spreadingridges, the HTRZ level
is poorly known, with the exception of the base of
Hole ODP 504B on the Costa Rica Rift, which has
reached the upper part of the HTRZ (Alt et al., 1995,
1996; Laverne et al., 1995; Kelley et al., 1995). It is
thus clear that the technical limitations of deep crustal
drilling restrict the direct observation and study of the
deepest parts of the oceanic hydrothermal systems in
On the other hand, well-preserved ophiolites, such
as the Troodosophiolite in Cyprus, the Bay of Islands
ophiolite in Newfoundland or the Semail ophiolite in
Oman expose, sometimes in exceptionnallygood con-
ditions, the root zones of fossil oceanic hydrothermal
systems. Several authors have pointed out the com-
plexity of that speciﬁc zone located between the lower
part of the sheeted dyke complex and the high-level
foliated gabbro unit, where the magmatic foliation
steepens, and where the ﬁrst diabase dykes appear
(Rosencrantz, 1983; Rothery, 1983; Lippard et al.,
1986; Juteau et al., 1988a,b; Nehlig, 1989; Nicolas
and Boudier, 1991; Nehlig et al., 1994). This thin
zone (hundred of meters thick) is characterized by an
extreme thermal gradient (of the order of 5◦C/m), and
is regarded as a thermal boundary between the con-
vective magma chamber system below, and the main
convective hydrothermal system above (Nicolas and
Boudier, 1991; Nehlig, 1993). It is characterized by
mutual intrusive relationships between diabase dykes,
high-level gabbros and diorites, plagiogranite veins,
dykes and bodies, by abundantxenoliths of altered di-
abase within plagiogranites and gabbros,and by many
pegmatitic, amphibole-rich patchy bands of gabbros
and diorites (hydrated recrystallization zones).
The Oman ophiolite
The well-preserved Oman ophiolite (Figure 1) offers
the best opportunity for a three-dimensional study of
ﬂuid circulation in fossil oceanic crust. It is a huge
thrust sheet of Cretaceous oceanic lithosphere, more
than 500-km long, 50-km wide, and up to 15-kmthick,
ideal for observation of both vertical and along-strike
continuous sections, over long distances. Its internal
stratigraphy and structure has been studied in detail
(e.g. Pallister and Hopson, 1981; Boudier and Cole-
man, 1981; Lippard et al., 1986; Nicolas et al., 1988;
Juteau et al., 1988a, 1988b; Nicolas, 1989), and con-
forms to the classical Penrose Conference deﬁnition:
the mantle sequence, composed of harzburgitic tec-
tonites, is overlain by a crustal sequence, 4 to 6-km
thick, including from bottom to top, (a) a well lay-
ered cumulate gabbro unit; (b) a high-level gabbro unit
with minor plagiogranite bodies; (c) a robust diabase
sheeted dyke complex, about 1200 to1500 m thick; (d)
an extrusive sequence, mainly made of basaltic pillow
lavas and subordinate massive lava ﬂows, overlain by
pelagic sediments, including a lower volcanic unit and
an upper volcanic unit(Alabaster et al., 1982; Pﬂumio,
The products of a fossil black-smoker-type hy-
drothermal activity may be observed all along the
ophiolite belt in about 25 major prospects and occur-
rences of massive sulﬁde deposits (Figure 1), which
were emplaced on the seaﬂoor immediately after erup-
tion of MORB-type basaltic lavas making up the lower
volcanic unit (Alabaster et al., 1980; Alabaster and
Pearce, 1985; Lescuyer et al., 1988). One sulﬁde de-
posit (Aarja mine) has been shown to have developed
slightly off-axis, but still related to the main accre-
tion stage (Haymon et al., 1989), and some prospects,
such as the Daris prospect (eastern Haylayn block),
are located in the upper volcanic unit (Lescuyer et al.,
1988), probably related to an off-axis mantle diapir
(Reuber et al., 1991). The textures and mineral com-
positions of the ores are quite similar to those of the
East Paciﬁc Rise massive sulﬁdes deposited by black
smoker activity (Ixer et al., 1984; Haymon et al.,
1984; Lescuyer et al., 1988). These analogies are en-
hanced by the discovery of fossil tube worm casts in
the Oman mineralizations, similar to those described
57° E58° E
Hayl al Safil
Main massive sulphide deposits
Wadi Falah Wadi Andam
Figure 1. Schematic map of the Oman ophiolite, showing: the names of the main ophiolite massifs, the surfaces covered by the mantle sequence
(white) and the crustal sequence (dotted) respectively; the location of the main massive sulﬁde deposits, mined or prospected (black triangles);
and the location of the ﬁve areas described or mentioned in this study (from North to South, wadis Salahi, Haymiliyah, Hawqayn, Falah and
around present-day active black smokers (Haymon
et al., 1984).
Previous studies have shown that the Oman mas-
sive sulﬁde deposits: (1) are structurally controlled by
normal faults parallel to the sheeted dyke complex, as
in the Troodos ophiolite (Haymon et al., 1989; Les-
cuyer et al., 1988), and (2) lay over stockwork veins
and alteration zones that can be traced down into the
sheeted dyke complex (Lescuyer et al., 1988; Hay-
mon et al., 1989; Nehlig, 1989; Nehlig et al., 1994).
These stockwork zones are themselves rooted within
epidosite-rich zones, developed in the lower part of
the sheeted dyke complex.
The sheeted dyke complex as a whole is deeply
affected by greenschist facies hydrothermal meta-
morphic recrystallizations (albite, actinolite, chlorite,
quartz, epidote, sphene...), and is crosscut by a dense
net of epidote-quartz-sulﬁde veins, which in more
than 90% of the cases are parallel to the dyke mar-
gins (Nehlig and Juteau, 1988a, 1988b; Nehlig, 1989;
Nehlig et al., 1990; Nehlig, 1993). Veins from the
base of the sheeted dyke complex contain mainly
epidote and quartz, whereas those from the top are
characterized by an increase in quartz and dissemi-
nated sulﬁdes. Most of the veins are less than 2 mm in
width, and the vein density increases drastically from
the high-levelgabbros to the sheeted dyke complex.
This study is focused on the fossil High Tem-
perature Reaction Zone of the Oman ophiolite. We
present below new ﬁeld, mineralogical, geochemical
and ﬂuid inclusion studies, made on the rocks typ-
ical of that speciﬁc level: plagiogranites, high-level
gabbros, diabases and associated epidosite veins.
Distribution and intensity of hydrothermal
alteration in the lower sheeted dyke complex
Choice of study areas
Nehlig et al. (1994) have shown that the intensity of
alteration within the sheeted dyke complex is highly
variable in detail, and may be described through four
major lithologies, termed respectively diabase, spilite,
mineralized spilite and epidosite:
– Diabase rocks are the least altered and by far the
most common facies. They exhibit an albite-chlorite
(+ sphene + Fe-Ti oxides) hydrothermal paragene-
sis, coexisting with Ca-plagioclase and clinopyroxene
– Spilite rocks have an albite-chlorite-quartz (+ acti-
nolite + sphene + Fe-Ti oxides + epidote) hydrother-
mal paragenesis, without primary magmatic relicts.
Primary magmatic textures, however, are perfectly
– Mineralized spilites exhibit high quartz/albite and
chlorite/actinolite ratios, and are crosscut by abun-
dant quartz-sulﬁde veins. They also have recognizable
– Epidosites are ﬁne- to medium-grained rocks, char-
acterized by an epidote-quartz paragenesis, which re-
places totally the primary magmatic phases. Magmatic
textures have totally disappeared, due to complete
and repeated dissolution/precipitation processes. Un-
der the microscope, these rocks consist of a mosaic
of anhedral to prismatic epidote associated with lo-
bate quartz crystals, with a few percent of chlorite and
intergrowths of sphene and skeletal Fe-Ti oxides.
Nehlig et al. (1994) also noted that the volume
of mineralized spilites and epidosites signiﬁcantly in-
creases along vertical zones underlying massive sul-
We have tried to quantify better the relative vol-
umetric proportions of these four lithologies, along
good exposures of the lower part of the sheeted dyke
complex, in three carefully selected areas:
(1) The Samrah – wadi Andam area, in the Samail
massif (Figure 2A), exposes excellent outcrops of
sheeted dykes, high-level foliated gabbrosand associ-
ated plagiogranites on the eastern margin of the Samail
massif, a few km to the East of the Maqsad mantle
diapir (Ceuleneer et al., 1988). This is one of the areas
where Nicolas and Boudier (1991) have described ‘di-
abase protodykes’, brownish gabbro dykes of variable
thicknesses (from 10 cm to one meter or more), hav-
ing microgranular margins against a foliated gabbro
matrix. These ‘protodykes’ mark the level of rooting
of the sheeted dyke complex. Following these authors,
the Samrah – wadi Andam area would be initially lo-
cated on the north-eastern ﬂank of the Maqsad ridge
segment, somewhere between the central part and the
southern tip of this segment (Figure 2B).
(2) The wadi Hawqayn – wadi Falah, in the SE part of
the Haylayn massif (Figure 3A), is probably the best
example of a fossil axial discontinuity recognizable in
the Oman ophiolite. This area is interpreted as a typi-
cal axial segmentation zone of the fossil oceanic ridge:
the tip of a propagator comingfrom SE has generated a
second generation of dykes oriented N130◦, crosscut-
ting older dykes oriented North-South (Juteau et al.,
1988b; Reuber et al., 1991; see also Nicolas et al., this
(3) The wadi Salahi area, in the Salahi massif (Fig-
ure 4A), is located just beneath the major gossan of
Zuha (Figure 1), a former massive sulﬁde deposit due
to a vigorous black-smoker-type hydrothermal activity
on the seaﬂoor (Regba et al., 1991; Pﬂumio, 1991).
Secondary recrystallizations of the dykes produce a
variety of colors, ranging from dark grey for the least
altered diabases, to pale green and yellowish green for
the epidosites. In outcrop, the color, grain size and
cohesion of the rocks are the main criteria for appreci-
ating their relative degree of hydrothermal alteration.
After examining numerous broken fresh surfaces in
the ﬁeld and in thin sections, we adopted the follow-
ing classiﬁcation, which follows Nehlig et al. (1994)
and reﬂects the relative intensity of hydrothermal
alteration of the dykes:
1. Diabase dykes: they are grey-blue in color, their
groundmass is dotted with disseminated chlorite clus-
ters, and some discrete epidote crystals. The rock
is massive and diﬁcult to break, producing angular
2. Chloritised dykes: they are blueish in color, and
chlorite invades the groundmass.
Maqsad mantle diapir
Ridge trend defined
by sheeted dikes
of the Samrah area
Flat lying layered
Figure 2a–b. Wadi Andam area. A. Geological map of the Samail massif, after Nicolas and Boudier, this issue (simpliﬁed). 1: extrusive series.
2: sheeted dike complex. 3: upper gabbros, amphibole gabbros. 4: lower, layered gabbros. 5: wehrlites. 6: high temperature harzburgites. 7:
low temperature harzburgites. 8: thrust faults. 9: strike of diabase dikes. M: location of the Maqsad mantle diapir, and inferred ridge axis.
The rectangle indicates the location of the area studies (wadi Andam area). B. Skematic structural map of the Samrah-wadi Andam area, in the
south-eastern part of the Samail massif, exposing the root zone of the sheeted dike complex. The geometrical relationship between the magmatic
foliation in high-level gabbros and the sheeted dikes suggests that this area is derived from the SE ﬂank of the Maqsad fossil accretion segment,
as shown in the interpretation sketch. After Nicolas and Boudier (1991). The rectangle indicates the location of the studies area.
3. Spilitic dykes: they have a green-blue color, with
penetrative recrystallizations of chlorite, albite and
4. Epidosites: they have a yellowish-green color,
the primary texture is totally destroyed by dissolu-
tion/precipitation processes, and replaced by epidote
(dominant) and quartz. The rock has a weak cohesion,
is friable and breaks easily.
Field observations and measurements
Wadi Andam (Samail block, Maqsad diapir)
Two continuous sections were studied along the lower
sheeted dyke complex, 100-m and 160-m-long, re-
spectively (Figure 2C). They include 80% dykes and
20% gabbro screens.
– Section 1 (northern section, 100 m long). Figure
2D shows the distribution of alteration lithologies. It
appears clearly that the alteration is zoned around the
most altered dykes (epidosites). They are bounded
sucessively on both sides by spilitized dykes, then
by chloritised dykes, and ﬁnally by diabases. These
epidosites are represented by meter-sized dykes that
are highly altered. The average distance between two
epidosites is about 8–10 m in the northern half of
the section (narrow epidosites), and 20–25 m in the
southern half (thicker epidotised zones, 5–7 m wide,
typically several dykes affected).
– Section 2 (southern section, 160 m long). Figure2E
shows that epidosites are scarcer in this section, with
an average spacing of about 25 m. Spilitized dykes
do not occur. A few dykes show partial epidotization,
either in the core, or along the margins.
0 20 40 60 80 100 120
Figure 2c–d. Wadi Andam area. C. Detailed geological map, showing the hydrothermal reaction zone in the eastern part of the Samail massif.
1: sheeted dike complex. 2: high-level gabbros with more than 30% dikes. 3: high-level gabbros. 4: plagiogranites. 5: laminated gabbronites.
6: layered olivine gabbros. 7: quaternary deposits. 8: stream beds. 9: faults. 10: four-wheeled vehicle tracks. 11: location of sections 1 and 2,
detailed in D and E. D and E. Distribution of the alteration facies in the lower sheeted dike complex along section 1 and section 2 respectively
(wadi Andam area). The four main alteration facies are indicated by different labels and column heights.
Figure 3a–b. Muwaylah oasis area. A. Simpliﬁed structural and geological map of the Haylayn massif, showing the main features linked to an
axial discontinuity of the fossil oceanic ridge. H: Haymiliyah magma chamber, supposed to represent the dying tip of a western axial segment.
P: Laminated gabbronites, supposed to represent the propagating tip of an eastern axial segment, intruding an older crust (Reuber et al., 1991).
This axial discontinuity is marked by a fan-shaped orientation of the sheeted-dike complex. The propagator area is marked by two directions of
diking. The rectangle indicates the studied area, around Muwaylah oasis. 1: lower volcanic unit. 2: sheeted dike complex. 3: high-level gabbros
and plagiogranite bodies. 4: magmatic breccias. 5: laminated gabbronorites. 6: layered olivine gabbros. 7: wehrlites and dunites. 8: mantle
harzburgites. B. Detailed geological map of the hydrothermal zone in the SE part of the Haylayn massif, aroundMuwailah oasis, between wadi
Hawqayn (to the West), and wadi Falah (to the East), with location of samples MU 94-45 and MU 94-46. 1: sheeted dike complex. 2: sheeted
dike complex (>30%) and screens of high-level gabbros. 3: high-level gabbros. 4: plagiogranite bodies. 5: laminated gabbronorites. 6: observed
faults. 7: average strike and dip of diabase dikes. 8: quaternary deposits. 9: location of the geological section detailed in C.
90 100 110
Dikes N0°Dikes N130°
Figure 3c. Muwaylah oasis area. C. Distribution of the alteration facies in the sheeted dike complex, along a dry valley close to Muwaylah
oasis (Haylayn massif). The four main alteration facies are indicated by different labels and column heights.
Muwaylah section (Haylayn block)
A splendid section along the lower sheeted dyke com-
plex, in a small dry valley located between Wadi
Hawqayn and Wadi Falah provides one km of con-
tinuous outcrop, along a 10-m high cliff (Figures 3A,
At a large scale (kilometer scale), our measure-
ments indicate that the N130◦dykes crop out over
693 m (from a total outcrop 900 m long), and repre-
sent 77% of the total volumeof the dyke complex, the
North-South dykes representing 23% and appearing as
screens of 10–30 m, with an average spacing of about
70 m. Along this section, 28.5% of the total volume
is massively epidotised. Epidosites develop mainly in
the N130◦dykes (89%), and their density is enhanced
at the southern part of the section. Table 1 gives the
relative percentages of epidosites in this section.
At smaller scale (120-m long, hectometric scale),
Figure 3C shows that, as in wadi Andam, there is
a clear and systematic zonation of alteration on both
sides of the most epidotised dykes. These epidosites
are 3–5-m wide (typically 2–3 dykes together), with
an average spacing of 12 m. There are incipient sulﬁde
stockworks at dyke margins and in cracks crosscut-
ting them. Two panels about 15 m-wide are intensely
epidotised, with a spacing of about 60 m, compa-
rable to the spacing observed by Nehlig (1989) in
Table 1. Statistical measurements on the orientations and alteration
facies of the diabases dikes in the lower part of the sheeted dike
complex, along a 900 m long continuous outcrop (small dry valley,
North of Muwaylah oasis, SE Haylayn massif).
Total N 130°N 0°
Cumulative distances 900 692.5 207.5
% of dikes 76.94 23.06
% of epidotized zones 28.5 89.28 10.72
relative % of epidotization 33.65 11.33
wadi Haymiliyah, in the same area (but on the other
Figure 4D shows the distribution of rock types in
wadi Falah area, some km to the East of Muwaylah
Wadi Salahi section
Our detailed study was done in the lowest part of the
sheeted dyke complex, just east of (and stratigrapically
above) the central graben occupied by plagiogranites
(Figure 4A). We expected to ﬁnd epidosite-rich zones
Figure 3d. Muwaylah oasis area. D. Geological map of the hydrothermal reaction zone along wadi Falah (east of Muwaylah oasis), with
location of sample DA 94-08. 1: sheeted dike complex. 2: sheeted dike complex (60%) and screens of high-level gabbros (40%). 3: high-level
gabbros (with <10% isolated dikes). 4: plagiogranite bodies. 5: laminated gabbronorites. 6: observed faults. 7: average strike and dip of diabase
dikes. 8: quaternary deposits. ç: stream beds. 10: four-wheeled vehicle tracks.
Figure 4a. Wadi Salahi area. A. Geological m ap of the hydrothermal reaction zone in the wadi Salahi area (Salahi massif). 1: lower extrusives. 2:
pillow breccias. 3: sheeted dike complex. 4: high-level gabbros (with <30% isolated dikes). 5: pagiogranite bodies. 6: laminated gabbronorites.
7: average strike and dip of diabase dikes. 8: observed and inferred faults. 9: quaternary deposits. 10: location of the geological section detailed
in B. A1. Inset showing the location of the studied section with respect to the Zuha peospect (massive sulﬁde deposit) and associated stockxerks
visible in the landscape. 1: extrusives. 2: sheeted dike complex and gabbros. 3: plagiogranites. 4: ancient quarternary terraces. 5: faults. 6: strike
and dip of lava ﬂows or diabase dikes. 7: mineralized stockworks.
Figure 4. B. Distribution of the alteration facies in the lower sheeted dike complex, along the wadi Salahi section. The four main alteration
facies are indicated by different labels and column heights.
Wadi Andam 1
Wadi Andam 2
0 10 20 30 4050 607080 90100
% of different types of dikes
% Diabases % Chloritized dikes % Spilitized dikes % Epidotized dikes
Figure 5. Relative volumetric proportions of the four main alteration facies described in the text, in the lower part of the sheeted dike complex
of the three studied areas. The average length of the compared sections is about 100 m.
Figure 6. Detailed geometry of a curved normal fault observed at the root of the sheeted dike complex, along wadi Haymiliyah (north ofHuwayl
village, Haylayn massif). The fault, striking parallel to the dikes, is injected with late diabase dikes, themselves crosscut by high-temperature
hydrothermal veins made of quartz, epidote and chlorite. This kind of tectonic feature, trending parallel to the dike complex, is interpreted as
being contemporaneous of the main oceanic accretion episode. 1: sheeted dikes. 2: syntectonic injections of diabase dikes. 3: tectonized and
hydrothermalized diabases (spilitic facies). 4: remobilized hydrothermal epidote. 5: quartz-epidote veins along dikes margins. 6: fault planes.
in this area, since it is located stratigraphically below
the Zuha prospect. However the section studied along
wadi Salahi is practically devoid of epidosites, and is
certainly the least altered of the three studied areas.
It appears that the studied section is not vertically in
line with the Zuha stockworks, which are well visible
on outcrops and in the landscape, just north of wadi
Salahi (see Figure 4A1). A slightly oblique shift of
wadi Salahi, about 1 km southwestward, is enough to
get the sheeted dyke complex outside the main feed-
ing discharge zones. This conﬁrms the idea that major
discharge zones feeding black-smoker activity on the
seaﬂoor are strongly focused (Alt, 1995).
Along this section, about 80-m long, spilitized and
chloritised dykes alternate with diabases, and epido-
tised dykes are practically absent (Figure 4B). The
average spacing between the most altered zones (spili-
tized dykes) is about 15 m in average. As in the
previous sections, a more or less symmetrical zoning
of the alteration intensity is visible around the most
Table 2. Relative proportions of the four alteration facies described
in the text, in the three studied areas.
Salahi Muwaylah And-1 And-2
% Diabases 40.09 23.92 22.92 19.49
% Metadiabases 37.92 18.71 35.25 20.67
% Spilites 20.67 18.97 11.78 27.05
% Epidosites 1.32 38.4 30.05 32.79
altered zones. Dyke margins are again the privileged
pathways for hydrothermal ﬂuids, although numerous
millimetric transverse veins may be observed.
Comparison between the three studied areas
Figure 5 and Table 2 present a direct comparison be-
tween the relative proportions of the alteration facies
in the three studied areas, showing the heterogeneous
distribution of the intensity of alteration along the
strike of the fossil oceanic ridge:
The most altered section is the Muwaylah sec-
tion, interpreted to be a fossil axial discontinuity. It is
likely that in such areas, various superposed tectonic
events have favoured and enhanced the circulation of
ﬂuids, as observed today along the East Paciﬁc Rise
(Bougault et al., 1990).
The two sections studied along wadi Andam are al-
most as altered as the wadi Muwaylah area (especially
Section 2). They are interpreted to represent discharge
zones linked to the vicinity of a mantle diapir, prob-
ably located somewhere between the central part and
the tip of an accretion segment.
Finally the wadi Salahi section is the least al-
tered. This section is located on the ﬂanks of the
main discharge zones and stockworks which fed the
Zuha massive sulﬁde deposit, and which are obviously
More detailed studies would be necessary to have
a precise idea of the distribution of the intensity of the
hydrothermal alterations along the strike of the Oman
ophiolite. The ﬁeld data presented here show however
that axial discontinuities along this fossil oceanic ridge
could be as hydrothermally active as the central part of
the accretion segments.
Evidence of ridge axis extensional tectonics
Normal faults, striking parallel to the sheeted dykes
and crosscutting them at low angle are common in
all the studied areas. They may be marked by quartz-
cemented breccias (Nehlig, 1989; Nehlig et al., 1990).
In the wadi Andam area, in the wadi Falah area, in
Muwaylah oasis area, a number of fault planes in-
jected with hydrothermal recrystallisations are well
visible, and were obviously used as pathways by the
In wadi Haymiliyah, north of Huwayl terraces,
listric normal faults are particularly evident (Figure 6).
In this area where the dykes are very regularly ori-
ented, sheeted dyke blocks are tilted 15–20◦with
respect to their original attitude, along sheared normal
fault planes, forming dark green, epidote- and quartz-
rich tectonic lenses, crosscut by abundant later chlorite
veins. Locally thin diabase dykes with chilled margins
were injected into these fault planes, as shown in Fig-
ure 6, suggesting that these tectonic structures were
developed during the main oceanic accretion stage.
Study of ﬂuid inclusions from the hydrothermal
Fluid inclusions were studied in 8 samples selected
from a collection of about 50 samples from the ‘re-
action zone’ outcropping in the eastern area of the
Haylayn block, specially along wadi Falah and in the
Muwaylah area. These samples include plagiogran-
ites, upper gabbros and diabases from the lower part
of the sheeted dyke complex (Table 3). The ﬂuid
inclusions were studied mainly in quartz crystals de-
veloped in hydrothermal veins or vugs, and also in
some epidote crystals. Secondary inclusions in pla-
gioclase crystals could not be analyzed successfully
because of their small size.
One plagiogranite sample of the Muwaylah area,
particularly rich in 3-phase, brine-rich ﬂuid inclu-
sions, has been more intensively studied (sample MU
94–45, see location on Figure 3B; Moreau, 1998).
Plagiogranite MU 94-45 is mainly made of plagio-
clase and quartz (more than 80%), and of amphibole,
chlorite, epidote, treillis magnetite (i.e., with ilmenite
exsolutions), sphene and apatite. Quartz appears un-
der several aspects: primary quartz is xenomorphic,
whereas secondary quartz is idiomorphic, often asso-
ciated with epidote. Late ﬁbrous quartz ﬁlls microﬁs-
sures. Plagioclase is purely albitic (An94 to An99),
and is probably completely secvondary. Amphibole is
mainly an actinolitic hornblende, also secondary (the
magmatic trend in such rocks is: pargasite-pargasitic
hornblende-edenitic hornblende, Spulber and Ruther-
ford, 1982). These amphiboles are Cl-poor (between
0 to 0.12% Cl, average: 0.06%) and they do not show
the Cl-enrichment of some oceanic amphiboles (more
than 3%, Mevel, 1984; Vanko, 1986). Chlorite is a
secondary phase, often developed at amphibole mar-
gins. This phase is typically a pychnochlorite, with
homogeneous Fe/Fe+Mg ratios, around 0.34 in aver-
age. Epidote is idiomorphic, with a pistachite-content
(Fe/Fe+Al) ranging from 21 to 32% (average: 25%).
Fe-Ti oxides are made of magnetite containing il-
menite lamellae, altered to rutile and sphene.Apatite
is abundant, as small translucent needles. This miner-
alogy is very typical of most plagiogranites from the
Table 3. Nature and modal mineralogical composition of the main studied samples (selection).
AND-94-07 Wadi Andam Quartz-epidote
vein in an epidotized
AND-94-16 Wadi Andam Epidotized and mineralized
diabase dike margin
MU-94-13 Muwaylah Quartz-epidote
vein in fault plane
MU-94-42 Muwaylah Epidotized plagiogranite-
diabase dike contact
MU-94-45 Muwaylah Stockwork in plagiogranite
MU-94-46 Muwaylah Quartz vein in isotropic
DA-94-06 Wadi Falah Plagiogranite
rich in epidote-quartz blebs
DA-94-08 Wadi-Falah Epidote-quartz bleb
DA-94-09 Wadi Falah Plagiogranite devoid
DA-94-10 Wadi Falah Plagiogranite
(rich in diabase inclusions)
DA-94-11 Wadi Falah "Pustule"-
DA-94-13 Wadi Falah Leucogabbro
DA-94-21 Wadi Falah Fine-grained
DA-94-22 Wadi Falah Hydrothermal vein
in isotropic gabbro
SA-94-23 Wadi Falah Plagiogranite
SA-94-37 Wadi Sahali Quartz-epidote
vein in plagiogranite
Sample Location Lithology Mode Micro Fluid
Some ﬂuid inclusions were studied in one sam-
ple from wadi Andam (Table 3). Samples from wadi
Salahi area were not studied (lack of time).
Microthermometric analyses of ﬂuid inclusions were
carried out on a Chaixmeca heating and freezing
stage, according to the procedures deﬁned by Roed-
der (1984). The stage was calibrated using stan-
dard synthetic products during heating steps. Freez-
ing stage temperatures were calibrated using ﬂuid
inclusions in natural (Camperio Quartz: −56.6 ◦C)
and synthetic (−0.6◦C) quartz. Heating stage tem-
peratures were calibrated using about ten synthetic
compounds of known melting temperatures, ranging
from 45◦C to 398 ◦C. Freezing and heating measure-
ments were made on individual inclusions in order
to obtain corresponding homogenisation temperatures
and ice-melting temperatures (ﬂuid salinities). All
freezing data were obtained before heating to avoid
decrepitation. Homogenisation and dissolution tem-
peratures were measured during progressive heating of
the sample in order to avoid decrepitation and/or phase
The measurement accuracy is estimated to be ±
5◦C around 400◦C, and ±0.2 ◦Cbelow0
dersaturated and saturated solution salinities were ob-
tained using the equations published by Bodnar (1993)
and Gunter et al. (1983), including a standard er-
ror estimated to ±0.1 wt% eq. NaCl. Calculation
of trapping temperatures is based on data by Bodnar
Molecular Raman microspectroscopy (Dhamelincourt
et al., 1979) was used to try and identify solid phases
in ﬂuid inclusions of plagiogranite MU 94-45, and also
check the presence of volatile species other than wa-
ter in the inclusions. The analyses were done by J.M.
Beny (ISTO-CNRS Laboratory at Orléans), using the
Dilor XY multichannel Raman microprobe.
The trace-element composition of 6 individual brine-
rich ﬂuid inclusions from sample MU 94-45 was
analysed by the nondestructive PIXE (Particle In-
duced X-Ray Emission) method. This method is based
on the spectrometry of X-Rays produced by atoms
bombarded with beams of charged particles (protons)
issued from particle accelerators. It allows all elements
with 11<Z<60 to be analysed in a matrix (Na is not
detected). A signiﬁcant advantage of this method is
the penetration depth of protons in silicates (around
70 µm with a 3 MeV proton beam), which allows the
ion-content of silicate-hosted ﬂuid inclusions placed
along the beam path to be analysed, provided that they
are located 5 to 30 µm-deep below the surface (An-
derson et al., 1989; Ryan et al., 1991, 1993; Volﬁnger
et al., 1997). Concentrations of the species dissolved
in inclusion ﬂuids (Table 8) are calculated from the in-
tensity of each characteristic X-ray line in the spectra.
The attenuation of the proton energy along the beam
path and the absorption of the emitted X photons in the
host matrix and in the inclusion ﬂuid are calculated re-
ferring to a database, as a function of the composition
and density of the host and of the inclusion ﬂuid.
The ﬂuid inclusion analyses were performedusing
the nuclear microprobe at the CEA-LPS Laboratory
at Saclay (Engelmann and Revel, 1991). A 3.5 MeV-
proton beam, with a size of about 5 ×5µm2,was
used. Beam currents varied between 300 and 500 pA.
In order to decrease the transmission of X-rays emit-
ted from the matrix, a 135-µm thick Be-ﬁlter and a
100-µm thick Al-ﬁlter with a hole drilled at its center
(diameter of the hole: 500 µm) were placed in front of
the Si(Li) detector. The inclusions were analysed in a
200-µm thick section. All were located at a depth of
about 10 to 20 µm in quartz, in order to prevent their
decrepitation under the beam. The quartzmatrix close
to the chosen inclusions was systematically analysed
in order to verify that it does not contain any de-
tectable metallic or non-metallic element. The spectra
were processed using the multi-layer procedure of the
GUPIX programme (Maxwell et al., 1995), with the
help of Marcel Volﬁnger (ISTO-CNRS, Orléans). The
depth to the top of the analysed cavities in the host-
mineral was measured with a precision of ±1µm
using a DRM Leitz microscope equiped with a motor-
ized stage. The ﬂuid inclusion thickness could only be
estimated optically. For treating the spectra, the com-
postion of the inclusion ﬂuid was arbitrarily ﬁxed at 60
wt% eq. NaCl (this ﬁxes its density), for lack of hav-
ing previously characterized the analysed inclusions
by microthermometry. This parameter however does
not signiﬁcantly affect the calculated concentrations,
as the energy of protons is only weakly attenuated in
brines and X-photons are little absorbed in liquids.
Figure 7. Main types of quartz-hosted ﬂuid inclusions developed in the plagiogranites, high-level gabbros and hydrothermal veins of the
Oman ophiolite reaction zone. Microphotographs taken in plagiogranite MU 94-45: A. Vapor-dominated, salt-poor, two-phase inclusions. B.
Liquid-dominated, salt-poor, two-phase ﬂuid inclusion, salinity close to that of seawater. C. Three-phase, brine-rich ﬂuid inclusions (halite
cube, vapor bubble and aqueous liquid). D, E. Four-phase, birne-rich inclusions (halite cube +a second smaller solid phase, vapor bubble and
aqueous liquid). F. Teared and stretched two-phase, liquid-dominated ﬂuid inclusion.
Measurements of oxygen isotopic ratios
Extraction of the aqueous ﬂuids from the ﬂuid inclu-
sions Fluid inclusions trapped in quartz were ex-
tracted under vacuum by thermal decrepitation. Min-
eral fractions of a few millimeters in size, weighting
from1gto4g,weredegassedat100◦C for a mini-
mum of two hours under vacuum. The samples were
then thermally decrepitated up to 600◦C at a heating
rate of 60–70◦Cmin
−1. Water was then collected in
a trap held at liquid nitrogen temperature and was the
only major gaseous phase detected duringthermal de-
crepitation of quartz samples. During decrepitation, 80
to 200 µmoles of water was liberated, then transferred
to a microequilibration vessel, to which 30 µmoles
of CO2were added. The H2OandCO
amount and isotopic composition were then allowed
to exchange oxygen isotopes at 30 ◦Cforthreedays.
After this time, equilibration was complete, and the
equilibrated samples of H2OandCO
cryogenically once again. The δ18O-values of the wa-
ter samples were calculated using the mass balance
equation (1) of Kishima and Sakai (1980). Precisions
obtained during the experiments for δ18O-values of
H2O were in the range 0.2h–0.5h, depending on
2O ratio during the equilibration procedure
(Lécuyer et al., 1999).
with: αCO2-H2O=1.0402 at T =30 ◦C (O’Neil
and Adami, 1969); δ18OofCO
CO2equilibrated after three days with H2O; δ18O
of CO2(i) =δ18OofCO
2before equilibrium with
H2O=16.15±0.05 (SMOW); [CO2]and[H
the amounts of the two gases in µmoles.
Extraction of oxygen and measurement of oxygen iso-
topic ratios of the host-silicates (Table 9) Oxygen
was extracted from powders of pure fractions of quartz
and epidote separates using the BrF5method (Clayton
and Mayeda, 1963), and analyzed as CO2gas with a
VG SIRA 10 mass spectrometer at the University of
Rennes I (Table 9). Reaction times of 24 h at a tem-
perature of 650◦C allowed to obtain 100% recovery
of molecular oxygen from both quartz and epidote.
Isotopic compositions are quoted in the standard δ
notation relative to SMOW. Results from the NBS28
standard gave a mean δ18O-value of +9.4 ±0.2h.
Oxygen isotope compositions of analysed silicate min-
erals have been therefore corrected on the basis of
the accepted δ18O-value of +9.6hfor the NBS28
Microthermometry of ﬂuid inclusions
Description of ﬂuid inclusions in quartz crystals
The ﬂuid inclusions in quartz crystals appear either as
densely spaced clusters along crystallographic planes
(primary ﬂuid inclusions), or as anastomosing arrays
along microcracks (secondary ﬂuid inclusions). Their
morphologies and shapes are variable: most of them
have negative crystal shapes. Others are irregular,
stretched or teared. They may be classiﬁed into 4 main
(1) Monophase (liquid) inclusions. Observed along
microcracks, they have very irregular shapes. They
result obviously from necking or stretching of the
inclusions, most probably.
(2) Aqueous, vapour-rich, 2-phase ﬂuid inclusions
(Figure 7A). These rather large ﬂuid inclusions (20–
30 µm in diameter) have irregular shapes, and exhibit
a large gas bubble (gas ﬁlling >50% in volume). They
are randomly distributed in the host-minerals, and
appear often associated with the 3-phase, brine-rich
ﬂuid inclusions. Other inclusions, with stretched mor-
phologies, often appear associated with monophase
inclusions. Altogether, they represent about 40% of
the 2-phase inclusions examined.
(3) Aqueous, liquid-dominated, 2-phase ﬂuid inclu-
sions (Figure 7B). These are either primary inclusions
clustering at quartz crystal margins, or secondaryﬂuid
inclusions clustered along microcracks, with a small
gas bubble (gas ﬁlling <30% in volume). They are
usually tiny (2–3 µm in diameter) and difﬁcult to
(4) Aqueous, brine-rich 3-phase ﬂuid inclusions
(Figures 7C, D, E; Figure 8). These inclusions have
during later tectonic events two ﬂuid phases, and at
least one solid daughter phase. They were observed
in ﬁve samples (4 plagiogranites and one high-level
gabbro), conﬁrming previous observations by Nehlig
(1989) in the plagiogranites of wadi Haymiliyah (Hay-
layn block also). All these inclusions typically contain
a cubic solid phase, always dissolving between 350◦
and 500◦C in a homogeneous liquid, and interpreted
as halite. Some inclusions also contain a brown-red
hexagonal solid identiﬁed as hematite by Raman spec-
troscopy. Finally, a clear rounded isotropic solid,
stable at high temperature, is observed in some inclu-
sions but remains unidentiﬁed so far (oxy-hydroxide?,
amphibole?, see below).
Sylvite (KCl), generally yellowish and having a
dissolution temperature well below that of halite, was
not observed. However, K was detected and measured
in several ﬂuid inclusions from plagiogranite MU 94–
45 (Table 8). Sterner and Bodnar (1984) have shown
that sylvite is metastable and does not always nucleate
in synthetic NaCl-KCl-saturated inclusions. The bulk
salinity (“equivalent” NaCl) of the KCl-supersaturated
inclusions interpreted only from NaCl-melting tem-
peratures may then be underestimated by about 10–15
wt% compare to the salinity interpreted from both
Figure 8. Typical set of quartz-hosted, three-phase, brine-rich ﬂuid
inclusions in plagiogranite MU 94-95. The halite cube and gas
bubble are well visible in all ﬂuid inclusions.
KCl- and NaCl-melting temperatures (Kelley and De-
laney, 1987), and their NaCl-content is overestimated
by about 10% (Kelley, 1990).
The three latter kinds of ﬂuid inclusions may be in-
timately associated inside a same host-mineral,though
the vapour-dominated inclusions are restricted to the
Description of ﬂuid inclusions in epidote
Several epidote-hosted inclusions, >15 µm in size,
could be studied in the margins of hydrothermalveins.
They are elongate, negative crystal-shaped, parallel
to epidote crystallographic planes. These ﬂuid inclu-
sions are mainly 2-phase, liquid-dominated inclusions
(gas ﬁlling <30%). Epidote also contains monophase
inclusions, stretched along microcracks.
About 270 ﬂuid inclusions were observed and stud-
ied in quartz and in epidote from hydrothermal veins
in plagiogranites, from high-level gabbros, and from
the lower sheeted dyke complex, of which 130 in pla-
giogranite MU 94–45. Tables 4 to 7, and Figures 9 to
12, summarise the results of ∼520 microthermomet-
ric measurements performed on these ﬂuid inclusions
(170 microthermometric measurements in plagiogran-
ite MU 94–45).
Three kinds of measurementswere done: ice melt-
ing temperatures, liquid-vapour homogenisation tem-
peratures, and dissolution temperatures of halite cubes
Nb of measurements
Homogenization temperature (°C)
0 50 100 150 200 250 300 350 400 450
-4 -3 -2 -1 0
Ice melting temperature (°C)
Salinity (weight %NaCl eq.)
Figure 9. Microthermometric measurements done in epi-
dote-hosted ﬂuid inclusions (diabases, plagiogranites). A.
Homogenization temperature measurements. B. Ice-melting
temperature measurements. C. Corresponding calculated salinities
(weight % NaCl equivalent).
Ice-melting temperatures and corresponding salinities
a) Liquid-dominated, 2-phase ﬂuid inclusions. All in-
clusions were frozen between −35 ◦Cand−65 ◦C.
No late liquid nor solid phase appeared during these
cryometric runs. Ice melting temperatures range be-
tween −0.8◦Cand−7.1◦C (Table 4), conﬁrming the
aqueous nature of these hydrothermal ﬂuids. Corre-
sponding salinities range between 1.4 to 10.6 wt%
eq. NaCl. However, if we discard one exceptional ice
melting temperature measured at −7.1◦C(Table4),
the rest of the measurements rangeon averagebetween
−1.7◦Cand−3.7◦C, corresponding to an average
Table 4. Microthermometric measurements done on two-phase, liquid-dominated ﬂuid incusions. Sample: nature and number of sample.
FI/HM: ﬂuid inclusion/host-mineral. I, II: primary and secondary, respectively. Lithology: Q: quartz. E: epidote. S: disseminated sulﬁdes. Vein:
hydrothermal vein. Plagiogr: plagiogranite. Thermometry: N: number of measurements. Th: homogenization temperatures intervals (◦C). Av:
mean homogenization temperature (◦C). Dev: standard deviation (◦C). Salinity: Sal: range of calculated ﬂuid salinities based on freezing point
depression (in wt% NaCl eq.). A. mean salinities (wt% NaCl eq.). Dev: standard deviation (wt% NaCl eq.).
Sample FI/HM Lithology Thermometry Cryometry Salinity
N Th Av Dev. N Tmi Av Dev. Sal Av Dev.
• Diabase dikes
AND-94-07 IQII Vein Q/E/S 9
35.3 3 -2.8/-3.2 -2.9 0.2 4.6/5.2 4.8 0.3
MU-94-13 IEII Vein Q/E 7
12.6 5 -1.8/-3.5 -2.3 0.6 3.1/5.7 3.8 1.0
MU-94-13 IQII Vein Q/E 7
42.1 4 -0.9/-2.1 -1.5 0.5 1.6/3.5 2.6 0.8
DA-94-23 IIQI Plagiogr. 5
37.5 3 -2.8/-3.2 -3.0 0.2 4.6/5.2 4.9 0.3
DA-94-09 IQII Vein Q/E 7
36.0 3 -1.1/-3.7 -2.6 1.1 1.9/6.0 4.3 1.8
DA-94-09 IEII Vein Q/E 5
38.9 3 -2.3/-3.2 -2.8 0.4 3.9/5.2 4.6 0.6
• Isotropic gabbros
MU-94-46 IQII VeinQ/E 11
49.3 9 -0.8/-7.1 -3.2 1.9 1.4/10.6 5.1 2.9
DA-94-13 IQII VeinQ/E 7
31.9 5 -1.2/-3.7 -2.5 1.0 2.1/6.0 4.2 1.5
salinity of 4.2 wt% eq. NaCl, slightly higher than that
of present-day seawater.
In epidote, the measurements done in two sam-
ples (quartz-epidote veins from the Muwaylah sheeted
dyke complex and from the plagiogranite intrusions of
wadi Falah, respectively), gave ice melting tempera-
tures ranging from −1.8◦Cto−3.2◦C, indicating a
ﬂuid salinity close to that of seawater (Table 4 and
In quartz from plagiogranites or diabase dykes, ice
melting temperatures are clustered around −2◦Con
average, with a standard deviation of 0.6 ◦C(Table4).
In quartz from high-level gabbros, ice melting temper-
atures are more variable and may reach −7.1◦C, with
a standard deviation of 2–3◦C (Figure 10).
In plagiogranite MU94-45, liquid-dominated 2-
phase ﬂuid inclusions exhibit relatively homogeneous
ice melting temperatures (Table 5), ranging from −4.1
to −2.2◦C (average −3.3◦C). Calculations (following
Bodnar, 1993) indicate that these temperatures corre-
spond to salinities of 3.7 to 6.5 wt% eq. NaCl, with a
mean of 5.1 wt% eq. NaCl (Table 5 and Figure 12),
clearly above the average salinity of present day sea-
water (3.2 wt % eq. NaCl). Figure 12 shows that ice
melting temperatures cluster around an average value
b) Vapour-dominated, 2-phase ﬂuid inclusions. Eight
ﬂuid inclusions gave an average ice melting temper-
ature slightly lower than the liquid-dominated ones,
ranging from −1.5◦Cto−2.7◦C . Corresponding
salinities range from 2.6 to 4.5 wt% eq. NaCl, with
an average of 3.8 wt% eq.NaCl.
In plagiogranite MU 94-45, vapour-dominated, 2-
phase ﬂuid inclusions exhibit similar ranges of ice
melting temperatures and salinities, except for one
value at −7.8◦C, corresponding to 11.5 wt% eq. NaCl
(Table 5; Figure 12A).
In total, the salinities of the 2-phase ﬂuids in this
plagiogranite are higher than in other plagiogranites.
Eutectic temperatures Because of the small size of
the 2-phase ﬂuid inclusions, we obtained only three
reliable measurements of eutectic temperatures in pla-
giogranite MU 94–45: −25.5◦,−26◦and −34.5◦C
(Table 5). Eutectic temperatures of the order of
−25/−26◦C are compatible with a system dominated
by NaCl, as the metastable eutectic temperature of
the H2O-NaCl system is −20.8◦C, whereas temper-
0 50 100 150 200 250 300 350 400
N = 140
Homogenization temperatures (°C)
Number of measurements
-8 -7 -6 -5 -4 -3 -2 -1 0
N = 60
Ice melting Temperatures (°C)
Number of measurements
N = 60
Salinity (weight %NaCl equ.)
Number of measurements
30 35 40 45 50 55
200 250 300 350 500400
N = 13
Halite melting temperatures (°C)
N = 13
Salinity (weight %NaCl eq.)
0 50 100 150 200 250 300 350 400 450
Homogenization temperatures (°C)
Figure 10. Microthermometric measurements done in quartz-hosted ﬂuid inclusions. N indicates the number of measurements. A. Ice-melting
temperature measurements, and corresponding calculated salinities. B. Homogenization temperature measurements. C. Temperatures of halite
dissolution in brine-rich ﬂuid inclusions. D. Homogenization temperature measurements in sets of inclusions showing important variations of
salinity. L: Liquid. V: Vapor. H: Halite (solid phase).
Table 5. Plagiogranite sample MU 94-95. Microthermometric measurements done on two-phase, liquid-dominated ﬂuid inclusions. Te eutectic:
measured eutectic temperatures, when possible. Te corrected: corrected eutectic temperatures. Tmi: measured ice melting temperatures. Tli
corrected: corrected ice melting temperatures. Salinity: calculated ﬂuid salinity, based on freezing point depression Tmi (in wt% NaCl eq.). Th:
measured homogenization temperature (◦C). Thcorrected: corrected homogenization temperature.
eutectic corrected corrected (wt %) corrected
-4.3 -4.0 6.4 320 308
-26.5 -26.0 -8.2 -7.8 11.5
-3.5 -3.2 5.2 358 344
-4.4 -4.1 6.5 360 346
-3.6 -3.3 5.4 361 347
-3.7 -3.4 5.5
-26 -25.5 420 404
-3.4 -3.1 5.1 380 365
-3 -2.7 4.5 366 352
-3.5 -3.2 5.2 365 351
-3.7 -3.4 5.5 348 334
-3.4 -3.1 5.1 368 354
-3.4 -3.1 5.1 383 368
-3.7 -3.4 5.5 372 358
-3.6 -3.3 5.4 320 308
-3.2 -2.9 4.8 414 398
-2.9 -2.6 4.3 306 294
-4.2 -3.9 6.3 325.5 313
-3.2 -2.9 4.8 265 255
-3.8 -3.5 5.7 432 415
-3.8 -3.5 5.7 441 424
-3.8 -3.5 5.7 420 404
-3.4 -3.1 5.1
-35 -34.5 -2.5 -2.2 3.7
-3.1 -2.8 4.6 195 187
-2.9 -2.6 4.3 240 231
-3.2 -2.9 4.8 262 252
-2.9 -2.6 4.3 374 360
-2.8 -2.5 4.2 234 225
-3.6 -3.3 5.4
-3.4 -3.1 5.1 267 257
atures of −35◦C imply that other salts like MgCl2are
probably present in the ﬂuids.
and 3-phase ﬂuid inclusions were heated to measure
liquid-vapour homogenisation temperatures, which
represent minimal trapping temperatures for two-
phase ﬂuid inclusions. All the homogenisation tem-
peratures cluster around a mode of 380◦C . Those
recorded in epidote range between 290 ◦C and 390◦C
(average around 350◦C,Figure 9). In quartz-hosted
liquid-dominated ﬂuid inclusions, homogenisation to
the liquid occurs between 220◦C and 430 ◦C, with a
mode at 345◦C (Figure 10B). Vapour-dominated in-
clusions homogenise to the vapour between 330◦C
In plagiogranite MU-94-45:
– all 2-phase ﬂuid inclusions, including the vapour-
rich ones, homogenised to the liquid phase. The
homogenisation temperatures range from 187◦Cto
415◦C (Figure 12C).
200 250 300 350 400 450
Homgenization temperatures (°C)
Salinity (% NacCl eq.)
L + V + H
Figure 11. Homogenization temperature – salinity diagramme (all
ﬂuid incusions), showing two main groups of ﬂuid inclusions: those
having salinities close to seawater salinity,andbrine-rich ﬂuid
inclusions. L: Liquid. V: Vapor. H: Halite.
– the studied 3-phase ﬂuid inclusions were 10- to
15-µm wide and with rounded shapes. Larger, ﬂat,
3-phase inclusions, with irregular contours and often
showing evidence for necking, were carefully avoided.
On heating, the disappearance of the vapour bubble
was observed before halite dissolution, with vapour-
disappearance temperatures ranging from 178◦Cto
342◦C, and a mean of 275 ◦C (Figure 12E).
Salt dissolution temperatures The dissolution of
halite cubes, ranging from 292 ◦C to 441◦C(Table6
and Figure 11), occurs systematically at temperatures
higher than the homogenisation of the ﬂuid phases,
occurring between 230 ◦C and 403◦C. They therefore
are minimal trapping temperatures. The 13 measure-
ments show a scatter, probably due at least in part to
the difﬁculty of determining the ﬁnal melting point of
a solid phase at high temperature.
In plagiogranite MU 94-45, the dissolution of
halite cubes (Tm>ThL+V(L)) occurs at temperatures
ranging from 358◦C to 496◦C (Table 7), correspond-
ing to very high salinities, from 43.1 to 59.2 wt % eq.
NaCl (Figure 12D).
The total homogenisation of halite-bearing inclusions
by halite dissolution indicates that these inclusions
were trapped in the absence of a vapour phase (Ram-
boz, 1979; Cloke and Kesler, 1979; Roedder, 1984;
Kelley, 1990), but only if heterogeneous trapping
of halite did not occur. The latter process can be
discarded in plagiogranites in the absence of halite
casts in the quartz host (Kelley and Robinson, 1990).
Halite-saturated inclusions from plagiogranite MU 94-
45, particularly, show halite-dissolution temperatures
which exceed vapour disappearance temperatures by
≈70 to 200◦C. Given that the P-T path of such
inclusion ﬂuids in the (L+H) ﬁeld has a slope of
≈20 bar/◦C (Bodnar, 1994), minimal trapping pres-
sures ranging from 1.5 to 4 kb are implied. Even
the trapping pressure of 1.5 kb interpreted in terms
of lithostatic conditions yields emplacement depths
≈5 km which are outside the subsurface depth-range
estimated for the HTRZ (1.5 to 2.5 km). At least two
factors should be considered for a further interpreta-
tion of these puzzling microthermometric results. (1)
A more rigorous thermobarometric interpretation of
halite-saturated brines requires reference to the multi-
component Na-K-Ca-Fe-Mn-Cl system (see below).
(2) There are clear textural evidences that some halite-
saturated inclusions have been partially annealed. The
possibility that many such inclusions have changed
their volume-composition properties after trapping,
for instance by necking, could possibly account for
the inconsistancy between present microthermometric
properties of the ﬂuid inclusions and their subsurface
Raman microspectroscopy in ﬂuid inclusions of
plagiogranite MU 94-45
Laser Raman analysis of the gas-part of vapour-rich
or halite-rich inclusions conﬁrmed that they are domi-
nated by water, without any detectable volatile species
2and SO2). Similar results
were already obtained for the ﬂuid inclusions in pla-
giogranites and gabbros from the Troodos ophiolite
in Cyprus (Kelley, 1990), Many brine-rich inclusions
contain other solid phases coexisting with halite:
(1) Hematite, a red-brown, translucent, hexagonal
mineral, was determined optically, and conﬁrmed by
Raman molecular microspectrometry analyses.
(2) Anhydrite was determined also by Raman mi-
crospectrometry in some ﬂuid inclusions.
(3) Another unknown rounded, isotropic, clear min-
eral was observed, smaller than the halite cubes,
and dissolving at about 350◦C before halite on heat-
ing, (Figures 7D, 7E). Raman analyses have shown
the presence of OH-groups in its structure. In Troo-
dos ﬂuid inclusions, Kelley (1990) also described
an unidentiﬁed OH-bearing“blade mineral” (chloride
hydrate or amphibole?).
Biphased inclusion : type L
Biphased inclusion : type V
0-8 -7 -6 -5 -4 -3 -2 -1 0
Ice melting temperature (°C)
Biphased inclusion : type L
Biphased inclusion : type V
00 1.5 3 4.5 6 7.5 9 10.5 12
Salinity, weight %NaCleq.
Triphased inclusions : type S
Biphased inclusions : type L and V
0 100 200 300 400 500
Homogenization temperature (°C)
0100 200 300 400 450
150 250 350
Homogenization temperature (°C)
200 300 400 500250 350 450
Homogenization temperature (°C)
Temperature of halite dissolution (°C)
TmH >Th L+V (L)
Tm<Th L+V (L)
Figure 12. Microthermometric measurements in quartz-hosted ﬂuid inclusions of plagiogranite sample MU 94-45 (Moreau, 1998). A.
Ice-melting temperature measurements. Type L: liquid-dominated inclusions. Type V: vapor-dominated inclusions. B. Corresponding cal-
culated salinities. C. Homogenization temperature measurements. Type S: inclusions showing a solid phase (halite cube). D. Homogenization
temperature – salinity diagramme. E. Homogenization temperature – Halite dissolution temperature diagramme. TmH: temperature of melting
of Halite cube. Th L+V (L): temperature of homogenization of liquid and vapor phases (in the liquid phase).
Table 6. Microthermometric measurements done on three-phase, brine-rich inclusions; Upper part: this study. Sample: number of sample.
Lithology nature of sample (Plag: plagiogranite. Plag.Pust: ‘pustule’ in plagiogranite). FI/HM: ﬂuid inclusions/host-mineral. I, II: primary and
secondary, respectively. Phases: L: liquid. V: vapor. H: halite (solid phase). Thermometry: N: number of measurements. Th: homogenization
temperatures intervals (◦C). Av: mean homogenization temperature (◦C). Dev: standard deviation (◦C). TmH: temperatures intervals of melting
of Halite cube (◦C). Cryometry: N: number of measurements. Tmi: temperatures intervals of ice melting (◦C). Av: mean ice melting temperature
(◦C). Dev: standard deviation (◦C). Salinity: Sal: range of calculated ﬂuid salinities based on freezing point depression (in wt% NaCl eq.).
A: mean salinities (wt% NaCl eq.). Dev: standard deviation (wt% NaCl eq.). Lower part: measurements published by Nehlig (1991), for
comparison. PN 81 and 82: Brine-rich 3-phase secondary ﬂuid inclusions in primary quartz from plagiogranite dikes intruding the sheeted
dike complex of wadi Haymiliyah (Haylayn massif), and associated low-salinity inclusions. Pn 403: low-salinity inclusions with evidences of
Sample Lithology Fi/HM Phases Thermometry Cryometry Salinity
N Th Av Dev. N TmH Av Dev. N Tmi Av Dev. Sal Av Dev.
DA-94-06 Plag. pust. IIQI L+V->L 11 345/415 386.0 23.1 - - - - 7 -0.9/-2.3 -1.7 0.5 1.6/3.8 2.9 0.9
L+V+H->L+H->L 6 251/386 316.5 48.4 2 333/441 387.0 54.0 - - - - 40/48 44.0 4.0
DA-94-08 Plag.pust IIQII L+V->L 9 322/421 393.9 36.6 - - - - 5 -0.6/-1.8 -1.1 0.5 1.0/3.0 1.9 0.8
L+V->V 7 334/402 381.6 22.9 - - - - 4 -1.5/-2.7 -2.3 0.5 2.6/4.5 3.8 0.8
L+V+H->L+H->L 6 230/401 352.7 68.1 4 297/415 364.5 42.6 - - - - 37/48 43.1 3.9
DA-94-11 Plag. IIQII L+V->L 10 230/417 284.0 65.1 - - - - 7 -1.4/-2.5 -2.3 0.6 2.4/5.5 3.8 1.0
L+V->V 4 375/405 401.1 17.8 - - - - 2 -2.4/-2.5 -2.5 0.1 4.0/4.2 4.1 0.1
L+V+H->L+H->L 6 278/398 345.2 43.0 4 338/406 368.1 25.1 - - - - 40.3/47 43.3 2.5
DA-94-21 Gabbro IQII L+V->L 11 256/425 331.9 61.0 - - - - 6 -1.6/-5.3 -3.2 1.4 2.7/8.3 5.2 2.1
L+V->V 4 352/386 366.5 13.5 - - - - 2 -2.1/-2.5 -2.3 0.2 3.5/4.2 3.9 0.3
L+V+H->L+H->L 8 230/403 300.3 51.4 3 292/348 320.3 22.9 - - - - 37/41 39.0 1.6
PN 82 Plag. IIQI L+V->L 21 213/429 280 72 7 -1.4/-2.5 2.4/4.2 3.2
L+V->C 1 452 1 -2.6 4.3
L+V+H->L+H 6 230/404 295 54 5
L+H->L 5 374/446 411 27 5 44/52 48
PN 81 Plag. IIQI L+V->L 28 243/410 293 48 14 -1.1/-5.1 1.9/8.0 4.0
L+V->C 1 402 1 -2.5 4.2
L+V->V 4 390/419 402 13 1 -2.2 3.7
L+V+S->L+H 5 288/314 302 10 5
L+H->L 5 333->415 389 28 5 40/48 44.7
PN403 Plag. IIQI L+V->V 24 346/429 393 23 4 -1.0/-2.6 1.7/4.3 3.6
L+V->C 7 399/425 411 9
L+V->V 5 350/420 403 27
PIXE analyses of brine-rich ﬂuid inclusions
Selection of samples
Six quartz-hosted brine-rich ﬂuid inclusions, with a
rounded shape, were analyzed in sample MU 94-45,
with the purpose of identifying the dissolved elements
in the ﬂuids, and of measuring their concentrations.
Seven elements were detected in the ﬂuid inclusions:
Cl, K, Ca, Mn, Fe, Zn and Br (Table 8). These data
conﬁrm the presence of cations other than Na in the
ﬂuids, as indicated mainly by the presence of many
solids other than NaCl in the inclusions. The fact
that Ca, K and Br were not systematically detected
in each inclusion marks the heterogenous composi-
tion of the analysed ﬂuids rather than results from the
variable depth of the analysed inclusions in quartz.
K- and Ca-contents can reach 34000 and 16000 ppm,
Iron, abundant in all the analysed inclusions, ex-
hibits by far the highest concentrations, ranging from
Table 7. Plagiogranite sample MU 94-95. Microthermometric measurements done on three-phase, brine-rich ﬂuid inclusions. Th: measured
homogenization temperature (◦C). T unknown phase: temperature of melting of a second solid phase of unknown nature (see text). T phase
corrected: corrected temperature of melting of the unknown phase (◦C). Tm (halite): measured temperature of melting of halite cube (◦C). Tm
(halite) corrected: corrected temperature of melting of halite (◦C). Tmh/100: Tm (halite divided by 100. Salinity: calculated ﬂuid salinity, based
on salt dissolution at high temperature (Potter et al., 1978) (in wt% NaCl eq.).
corrected phase corrected corrected wt%NaCl eq.
190 182 408 392 46.6
406 390 46.4
210 202 450 433 51.2
185 178 372 358 43.1
250 240 460 442 52.3
336 323 487 468 55.6
516 496 59.2
480 462 54.7
292 281 504 485 57.7
270 259 395 380 45.3
268 257 455 437 51.7
445 428 50.6
280 269 370 356 468 450 53.3
320 308 406 390 46.4
461 443 52.4
471 453 53.6
443 426 50.4
438 421 49.8
402 386 46.0
491 472 56.1
493 474 56.3
480 462 54.7
505 486 57.8
472 454 53.7
374 360 43.3
437 420 49.7
433 416 49.3
398 383 45.6
438 421 49.8
380 365 43.9
475 457 54.1
462 444 52.6
421 405 48.0
514 494 59.0
406 390 46.4
460 442 52.3
451 432 51.3
475 457 4.1
400 385 502 483 57.4
284 273 425 409 48.4
343 330 430 413 48.9
356 342 431 414 49.0
334 321 470 452 53.5
460 442 52.3
335 322 310 298 461 443 52.4
296 284 442 425 50.3
334 321 468 450 53.3
323 310 466 448 53.0
425 409 48.4
410 394 46.8
406 390 46.4
402 386 46.0
187 180 424 408 48.3
370 356 456 451 53.4
460 442 52.3
270 259 450 433 51.2
347 334 365 351 480 462 54.7
416 400 47.4
336 323 486 467 55.4
450 433 51.2
450 433 51.2
370 356 480 462 54.7
495 476 56.6
450 433 51.2
486 467 55.4
475 457 54.1
460 442 52.3
380 365 43.9
462 444 52.6
473 455 53.9
332 319 470 452 53.5
Table 8. PIXE analyses, giving the contents of some light elements andmetals in six quartz-hosted, brine-rich ﬂuid inclusions of plagiogranite
sample MU 94-45. Z: atomic number. LOD: limit of detection.
Fluid Inclusion IF4
Elément (Z) Content ppm (LOD)
Cl (17) non measured
K (19) non detected
Ca (20) 6660±2244 ; (4000)
Mn (25) 7393±688 ; (947)
Fe (26) 33802±710 ; (657)
Zn (30) 907±84 ; (105)
Br (35) 102±57 ; (80)
Fluid Inclusion IF7
Elément (Z) Content ppm (LOD)
Cl (17) non measured
K (19) non detected
Ca (20) id.
Mn (25) 5026±1114 ; (1696)
Fe (26) 26223±997 ; (668)
Zn (30) id.
Br (35) 598±123 ; (175)
Fluid Inclusion IF5
Elément (Z) Content ppm (LOD)
Cl (17) 372140±54700 ; (77810)
K (19) 33998±5133 ; (5860)
Ca (20) non detected
Mn (25) 20763±870 ; (1030)
Fe (26) 86882±1042 ; (753)
Zn (30) 2236±125 ; (94)
Br (35) 1408±98 ; (57)
Fluid Inclusion IF8
Elément (Z) Content ppm (LOD)
Cl (17) non measured
K (19) non detected
Ca (20) id.
Mn (25) 8510±1362 ; (2150)
Fe (26) 86162±1551 ; (730)
Zn (30) 891±174 ; (248)
Br (35) 669±140 ; (104)
Fluid Inclusion IF6
Elément (Z) Content ppm (LOD)
Cl (17) 336020±84676 ; (136125)
K (19) non detected
Ca (20) 16174±3744 ; (6244)
Mn (25) 16193±1311 ; (1852)
Fe (26) 92000±1564 ; (1430)
Zn (30) 2003±184 ; (236)
Br (35) 1398±154 ; (122)
Fluid Inclusion IF9
Elément (Z) Content ppm (LOD)
Cl (17) 375583±28920 ; (33538)
K (19) 18137±4425 ; (6778)
Ca (20) non detected
Mn (25) id.
Fe (26) 53412±1282 ; (1121)
Zn (30) id.
Br (35) 351±114 ; (128)
Elément (Z) Content ppm (LOD)
361247 ± 56099 ; ( 82491)
26067 ± 4779 ; ( 6319)
11417 ± 2994 ; ( 5122)
11577 ± 1069 ; ( 1535)
63080 ± 1191 ; ( 873)
Zn (30) 1509
± 141 ; ( 170)
754 ± 114 ; ( 111)
26223 ppm to 92000 ppm, with an average content of
63000 ppm (Table 8). Manganese is present in lesser
amounts than iron in all the analysed ﬂuid inclusions,
with an average Mn-content of about 11500 ppm (Ta-
ble 8). Finally, zinc-contents are low, ranging from
900 to 2200 ppm, with an average value of about
1500 ppm (Table 8).
Discussion of PIXE analyses
(1) These preliminarydata conﬁrm that the hydrother-
mal ﬂuids circulating at the base of the oceanic crust
are far more complex in composition than a simple
(2) The presence of ﬂuid inclusions with K-contents
up to >3 wt% and Mn-contents up to 2 wt% in low
K2O and low MnO granitic-rocks is problematic.
(3) Only three metals were analysed in these in-
clusions: iron, manganese and zinc. Many authors
previously concluded that plagiogranite-hosted halite-
saturated ﬂuid inclusions are Fe-rich, based on the
presence of hematite in the inclusions (Kelley and
Robinson, 1990), or of an iron chloride solid phase
such as FeCl2(Dubois, 1984). Present data show
that all the 6 halite-saturated inclusions analysed in
the Oman plagiogranites have high and variable Fe-
contents ranging from 2.6 to 9.2 wt%. Copper was
never detected in any of the analysed ﬂuid inclusions
(detection limit for Cu around a few hundred ppm),
which contrasts with the fact that both copper and zinc
were leached at the base of the diabase sheeted dyke
complex in the Oman ophiolite (Nehlig, 1989). Also
note that high temperature modiﬁed seawater venting
at mid-ocean ridges contain 1000 to 2000 ppm Cu
and that magmatic brines exsolved from Cu-rich por-
phyries contain several thousands ppm Cu (Ramboz,
1979; Ulrich et al., 1999).
(4) The Cl/Br ratios calculated from PIXE data in three
of the analysed inclusions are 240, 264 and 1070 (Ta-
ble 8). The two former values are close to the Cl/Br
ratio of seawater or of evolved seawater circulating in
the oceanic crust (Campbell and Edmond, 1989). In
contrast, the third ratio is more than three times that
of seawater, and was measured in a Br-poor, Mn-free
inclusion. The ratios of elements measured by PIXE
in ﬂuid inclusions are accurate, therefore these Cl/Br
data may suggest that the chloride-rich ﬂuids might
be derived from variable mixtures of evolved seawater
with other (magmatic?) ﬂuids. In support of that in-
terpretation, note that measured Fe/Mn mass ratios in
4 of the 5 analysed Mn-bearing halite-saturated inclu-
sions are comprised between 4.1 and 5.7, close to the
Fe/Mn ratios of some high-temperature hydrothermal
vent ﬂuids from sediment-starved mid-ocean ridges
(11◦N, Southern Juan de Fuca Ridge and TAG MAR:
Oxygen isotopes of water inclusions trapped in
Selection of samples
In order to obtain meaningful measurements of the
oxygen isotopic ratios of ﬂuids extracted from quartz
or epidote minerals, we looked for host-minerals con-
taining a single population of ﬂuid inclusions, as
homogeneous as possible. In most studied samples,
the host-minerals contain several generations of ﬂuid
inclusions, displayed along microcracks crosscutting
each other. This criterion limited severely our choice
to two samples from the Haylayn massif:
– Sample MU 94-46 is a ﬁne-grained high-level gab-
bro from the Muwaylah area (see location on Fig-
ure 3B), crosscut by centimetric quartz-epidote veins.
Most inclusions in quartz are primary, with a small
amount of microﬁssures ﬁlled with secondary in-
clusions. The primary inclusions are two-phase and
homogenise to the liquid phase. Their gas-ﬁlling is
relatively constant, but their salinity is variable, rang-
ing from 1.4 to 10.6 wt% eq.NaCl, with an mean
of 5.1 wt% eq.NaCl. The average homogenisation
temperature is about 330 ◦C.
– Sample DA 94-08 is a plagiogranitecollected along
wadi Falah (see location on Figure 3D), with abun-
dant quartz-epidote ‘blebs’ (Plate I, Figure 3). The
modal composition of the ‘pustules’ is: quartz =43%;
epidote =49%; amphibole =8% (cf. Table 3). The
quartz crystals in the ‘pustules’ present two types of
ﬂuid inclusions: a brine-rich type with halite cubes,
concentrated in the core of the ‘pustules’, and a salt-
poor, biphased, vapour-dominated type, concentrated
in the margins of the ‘pustules’.
(1) MU 94-46. In this sample, the quartz host and
the extracted hydrothermal ﬂuids both exhibit high
positive δ18O-values (6.8hfor quartz, and 4.9hfor
the ﬂuids, Table 9). Assuming equilibrium conditions,
the oxygen isotopic compositions of both quartz and
water inclusions measured in this sample yield a cal-
culated temperature of 630◦C using the quartz-water
fractionation equation given by Zheng (1993a). This
high temperature combined with a high δ18O-value for
the ﬂuid are compatible with a magmatic origin for this
quartz occurring in veins.
(2) DA 94-08. Oxygen isotopic compositions of
quartz along with that of water in associated inclu-
sions have been measured in an epidositic ‘pustule’
of plagiogranite sample DA 94-08. Average water-
content in quartz is about 1000 ppm. Quartz from
the epidositic ‘pustule’ DA 94-08 has a δ18O-value
of +7.0h(Table 9). Fluid inclusions from the rim of
the epidosite have homogenisation temperatures that
range from 322◦C to 421 ◦C, with variable salinities
ranging from 1.0% to 4.5 wt% eq. NaCl. The oxygen
isotope composition of their water (δ18O=−0.4h)
is close to the isotopic composition of pure seawa-
ter (Table 9). Quartz from the core of the epidositic
‘pustule’ contains brine-rich ﬂuid inclusions that are
characterised by highly variablehomogenisation tem-
peratures from 230 ◦C to 401◦C, and an average salin-
ity of 43 wt% eq. NaCl associated with a δ18O value of
+3.4h(Table 9). Coexisting epidote has a δ18O-value
of −0.3h(Table 9).
Discussion of isotopic results
Isotopic exchange may occur between water in the
ﬂuid inclusions and their quartz host (e.g., King et
al., 1997). On the other hand, original oxygen isotopic
compositions of trapped ﬂuids may be preserved dur-
ing millions of years (e.g., Vityk et al., 1993), depend-
ing on the thermal history of the rocks. Homogeni-
sation temperatures of liquid-rich aqueous inclusions
are proxies of trapping temperatures, given the subsur-
face environment of the HTRZ. They can be used to
Plate 1. Typical outcrops in the hydrothermal reaction zone of the Oman ophiolite. 1. Epidosite veins and concretions developed at the base
of the sheeted dike complex. Most of the epidosite veins, about 2–3 cm wide, are parallel to the diabase dikes margins, with a spacing here of
about 2 veins/m. 2. Epidote dike, 30 cm in width. This former diabase dike has been used as a drain by the hot, ascending hydrothermal ﬂuids,
and is totally transformed to massive epidosite (Samail block, Wadi Andam section). 3. Hydrothermal ‘pustules’ developed in a plagiogranite
intrusion. These spherical hydrothermal ‘pockets’ are made of massive epidote (80%), quartz (15%) and green actinolite (5%). Epidote is
typically cryptocrystalline at the margins, and form large idiomorphic crystals at the center of the ‘pustules’ (wadi Falah, Haylayn massif). 4.
Hydrothermal ‘pockets’ of hydrated pegmatitic gabbro, composed of albitic plagioclase, actinolite, chlorite and epidote, in a high-level gabbro.
5. Network of hydrothermal veins in a small plagiogranite intrusion. This picture taken at sunset gives an idea of the intensity of hydrothermal
percolation in the plagiogranites of the reaction zone (Haylayn block, Wadi Haymiliyah). 6. Spherical concentrations of hydrothermal epidote
and quartz in a ﬁne-grained, isotropic gabbro (Haylayn block, Muwaylah area).
Table 9. Oxygen isotope compositions of minerals and their water inclusions from the Oman ophiolite.
Sample Mineral δ18O mineral H2O inclusion δ18O(H2O)
‰SMOW wt% ‰SMOW
MU-94-46 quartz +6.8 0.09 +4.9
DA-94-08(*) quartz +7.0 0.07 -0.4
DA-94-08(#) quartz +7.1 0.12 -3.4
DA-94-08 epidote -0.3 0.35 +0.9
* quartz from the rim of the epidosite
# quartz from the core of the epidosite
test whether measured oxygen isotopic compositions
of ﬂuid inclusions are in equilibrium with the miner-
als. In the absence of sub-solidus isotopic exchange,
temperatures calculated on the basis of mineral-water
equilibrium fractionations should lie in the range of
trapping temperatures given by ﬂuid inclusion stud-
ies. If oxygen isotope exchange occurred between
quartz and water in the inclusions during cooling of
the sample, the δ18O-value of the aqueous ﬂuid would
decrease, thus leading to calculate temperatures that
are lower than trapping temperatures.
Isotopic compositions of mineral pairs may also
be used to estimate closure temperatures of isotopic
exchange (Dodson, 1973). The quartz-epidote associ-
ation in sample DA-94-08provides a closure temper-
ature of 230◦C according to the set of fractionation
equations proposed by Zheng (1993a, b). This temper-
ature is signiﬁcantly lower than the temperature range
of322–421 ◦C given by the two-phaseﬂuid inclusions.
This discrepancy may be explained either by retro-
grade isotopic exchange between the solid and liquid
phases, or by the different isotopic compositions of the
aqueous ﬂuids that deposited quartz and epidote. Oxy-
gen isotopic analyses of both quartz and their trapped-
water yield calculated temperatures of at least 290 ◦C
in the rim of the epidosite, and of 450◦Cinthecore.
These ‘isotopic’ temperatures are more consistent with
trapping temperatures inferred from ﬂuid inclusions
(Table 4). The core of the epidosite has been likely
deposited by a high-temperature, highly saline and
18O-enriched hydrothermal ﬂuid, whereas the later-
formed epidosite derived from colder hydrothermal
ﬂuids with marked seawater afﬁnities. According to
the ﬂuid inclusion thermometric study, epidote from
the epidosite was likely formed at temperaturesnot ex-
ceeding 450◦C (Table 4). This result implies that the
δ18O-value of the epidote-depositing ﬂuid was at least
equal to −0.5h, indicating an isotopic composition
similar or slightly 18O-enriched relative to Cretaceous
seawater, which likely had a δ18O-value between 1h
and −0.5h, in a context of continental ice-free world
Concerning sample MU 94-46, oxygen isotope ra-
tios of the host-quartz and of the extracted ﬂuids seem
to be in equilibrium with a δ18O close to that of the
primary magmatic crust, suggesting that the analysed
hydrothermal veins may have formed from a residual
ﬂuid of magmatic origin.
Discussion and conclusion
Origin of brine-rich ﬂuids in the oceanic crust
Brine-rich ﬂuid inclusions (>16 wt% eq. NaCl)
have long been described in gabbroic samples of the
present-day oceanic crust (Jehl et al., 1977; Kelley
and Delaney, 1987; Butterﬁeld et al., 1990), and at the
base of the fossil oceanic crust in ophiolite complexes
(Nehlig, 1989; Lécuyer, 1990; Kelley, 1990; Kelley
et al., 1992). Kelley and Robinson (1990) ﬁrst un-
derlined that saline ﬂuids from the Troodos ophiolite
exhibit similar V-X properties to the ﬂuids found in
inclusions from porphyry copper felsic intrusions and
Four kinds of processes were proposed to explain
such extreme salinity enrichmentsrelative to seawater:
Hydration of oceanic crust
This process may be efﬁcient when the water/rock ra-
tio is low, under rock-dominated conditions. In this
case, the salinity of the residual hydrothermal ﬂuids
may increase by simple hydration of the host-rocks,
with crystallisation of secondary hydrated minerals
(smectites, chlorites, amphiboles, etc.). Calculations
performed on variousstudy sites have shownthat:
– the hydration of 50% of the gabbros and pla-
giogranites is necessary to double the salinity of a
– the hydration of 93% of the whole plutonic sequence
is necessary to obtain brine-rich ﬂuids from seawater
(Cathles, 1983; Vanko, 1986; Kelley, 1990).
Both conditions are very unrealistic for the Oman
ophiolite. Nehlig (1991) has convincingly argued that
in the open hydrothermal system of Oman, with wa-
ter/rock ratios typically higher than 10 within the
sheeted dyke complex, the salinity may increase by
no more than 0.01wt% eq. NaCl by hydration of the
Retention/dissolution of ephemeral Cl-bearing
minerals in the crust
This process is certainly operating at a large scale in
the present-day and fossil oceanic crust, as attested by
the secondary phases observed in the oceanic crust,
and it has certainly contributed to modify seawater-
derived hydrothermal ﬂuids. It can particularly be
invoked to explain the composition of numerous 2-
phase ﬂuid inclusions with average salinities around
5 wt% eq. NaCl. This process alone, however, is
unlikely to yield the very high salinities observed in
halite-saturated ﬂuid inclusions (48 wt% eq. NaCl on
average), since the plutonic section is far from being
This process, allowing the temporary retention of
chlorine in the secondary phases of the oceanic crust,
was invoked to explain the variations of salinity of
the ﬂuids exiting on the seaﬂoor. For instance, the
study of secondary amphiboles in the crust has shown
an increase in chlorine in these phases (Vanko, 1986.
Mevel, 1984). Apatite from plagiogranites may also
concentrate chlorine signiﬁcantly (0.49% on average
in the apatite of plagiogranite MU 94-45). Reversely,
a retromorphic recrystallisation of these phases may
liberate chlorine in the ﬂuids, increasing their salinity.
However, taking into account the Cl-content of amphi-
boles or apatite of the crustal rocksin the hydrothermal
‘reaction zone’ (Alt, 1995),this process can only very
slightly modify the salinity of hydrothermal ﬂuids.
Bischoff (1980) was the ﬁrst to evoke this process,
which was later conﬁrmedby numerous detailed stud-
ies on oceanic and ophiolitic samples (Ramboz et al.,
1988; Von Damm, 1988; Bischoff and Rosenbauer,
1989; Nehlig, 1989; Berndt and Seyfried, 1990; Kel-
ley, 1990; Manac’h, 1997). During the upward migra-
tion of a seawater-derived ﬂuid in the oceanic crust,
ﬂuid unmixing may occur, leading to the formation of
a NaCl-enriched brine coexisting with a vapour (Kel-
ley and Delaney, 1987; Nehlig, 1989; Kelley, 1990;
Kelley et al., 1992; Nehlig, 1993), which both may
eventually vent at the seaﬂoor (Ramboz et al., 1988;
Lécuyer et al., 1999).
In the Oman ophiolite, most of the studied pla-
giogranites and some isotropic gabbros exhibit the
close spatial association of brine-rich, 3-phase ﬂuid
inclusions, with 2-phase, low-salinity ﬂuid inclusions
(vapour or liquid dominated). Our data bring new
evidences for such a coexistencein several plagiogran-
ites of the Haylayn block and in one gabbro sample.
Nehlig (1989, 1991) has argued in favour of a phase
separation process in the Reaction Zone of Oman
ophiolite on the basis of two plagiogranite samples
showing coexisting high-salinity and low-salinity in-
clusions (samples PN 81 and PN 82, see Table 6).
However, the V-X properties of many NaCl-rich in-
clusions are incompatible with boiling conditions (see
sample MU 94-45 particularly).
Magmatic differentiation processes
A fourth hypothesis to explain the formation of these
brine-rich ﬂuid inclusions is linked to the magmatic
differentiation process. This process, consisting in the
formation and concentration of residual aqueous ﬂu-
ids of magmatic origin during the emplacement and
fractionation of an evolved magma, was proposed for
the formation of porphyry copper ore-deposits (Hen-
ley and Mc Nabb, 1978). Recent theoretical modelling
on ﬂuid evolution in silicate magmas shows that brines
may be exsolved during the latest magmatic stages in
low-pressure systems (Cline and Bodnar, 1991). On
the other hand, the fact that the high-temperature Cl-
enriched-chalcopyrite-pyritebrine inclusions found in
porphyry copper intrusions are bearing hematite and
anhydrite is consistent with their exsolution from a
magma with characteristic fO2-conditions ﬁxed by the
H2S/SO4equilibrium (Burnham and Ohmoto, 1980).
More recently, the possibility of magma-derived
hydrothermal ﬂuids in the oceanic crust was dis-
cussed (Cowan and Cann, 1988; Kelley and Delaney,
1987; Roedder, 1992). Effectively, brine-rich ﬂuid
inclusions are observed mainly in evolved magmatic
rocks such as plagiogranites in Oman (Nehlig 1991;
Manac’h, 1997) or in Cyprus (Dubois, 1984; Kel-
ley, 1990). These highly fractionated rocks not only
were enriched in residual magmatic water during mag-
matic differentiation (Cline and Bodnar, 1991), but are
also rich in hydrated diabase xenoliths, resulting in a
contamination in ﬂuids.
Considering an initial water-content of 0.2% in the
basaltic glasses, and a Cl/H2O ratio of 0.06 (Muenow
et al., 1990), the residual liquid after 95% crystallisa-
tion may contain 5% water. At one kb pressure, the
exsolved ﬂuids have a salinity of 80-90 wt% eq. NaCl
in the extreme stages of differentiation (Nehlig, 1991;
Cline and Bodnar, 1991).
In plagiogranite sample MU 94-45, the bulk V-
X properties of NaCl-rich inclusions are compatible
with their derivation by exsolution from a magma.
This interpretation however, is not supported by PIXE
data, except for one inclusion which is Mn-free (pla-
giogranites are low MnO-bearing rocks: Dixon and
Rutherford, 1979) and has a Cl/Br mass ratio of 1070.
Oxygen isotopic composition of quartz supports the
presence of magmatic water in the core of epidosite
sample DA-94-08 in Oman. On the other hand, the
presence of hematite and/or anhydrite in many pla-
giogranite halite-saturated inclusions (Dubois, 1984;
Kelley and Robinson, 1990; this work) preclude that
these brines have been exsolved from the plagiogranite
magma. This is because plagiogranites are derived by
partial melting and crystal fractionation of primitive
tholeites at low fO2-conditions (W-M: Spulber and
Rutherford, 1983). Finally, 5 halite-bearing inclusions
analysed in sample MU 94-45yield Fe/Mn ratios com-
patible with that of evolved seawater circulated in the
oceanic crust, consistent with the Cl/Br ratios ≈300
measured in 2 of them.
The source of metals in the ﬂuid inclusions
To explain the presence of Fe, Mn and Zn in the brine-
rich ﬂuid inclusions of plagiogranite sample MU 94-
45, we have to consider three possibilities:
(1) The metals were leached by the hydrothermal ﬂu-
ids in the reaction zone, and particularly in the lower
sheeted dyke complex.
Starting from chemical analyses of the altered
rocks of the Oman crust, Nehlig (1989) has shown
that copper, zinc and iron were leached in the diabase
sheeted dyke complex and precipitated either in the
hydrothermal veins, or in the shallow stockworks and
massive sulphide deposits. In the Troodos ophiolite
also, Richardson et al. (1987) have shown convinc-
ingly that metals in the lower sheeted dyke complex
have been leached extensively. Copper is leached from
clinopyroxene (abundant in diabases), whereas iron,
manganese and zinc are leached mainly from Fe-Ti
oxides. Part of the iron and zinc detected in the brine-
rich inclusions of sample MU 94-45 could likely result
from the leaching of the diabases.
(2) The metals were leached from the plagiogran-
ites of the reaction zone. Plagiogranites also are rich
in Fe-Ti oxides, and could be an alternative source
for the metals detected by PIXE analyses. Processes
of hydrothermal alteration and leaching are clearly
demonstrated in sample MU 94-45 oxides, resulting
in the presence of abundant ilmenite exsolution lamel-
lae in the magnetite crystals, and in the development
of many rutile crystals at the margins of these same
oxides. However, three analyses of treillis magnetite
revealed Zn-contents below the detection level (≈50
ppm) and Cu-contentsof 150–200 ppm in two of them.
These preliminary results do not support the interpre-
tation that metals in the analysed brines were directly
leached from plagiogranites.
(3) The metals were concentrated in magmatic resid-
ual ﬂuids. A third hypothesis for the source of the
metals may be the contribution of Fe-rich magmatic
residual ﬂuids in the deep hydrothermalsystem (Kel-
ley, 1996; Yang and Scott, 1996). Such ﬂuids may
present high contents in NaCl (Cline and Bodnar,
1991), and are potentially able to form ore deposits
(Constantinou, 1987; Yang and Scott, 1996). How-
ever, the Cu-free nature of the brines analysed in
sample MU 94-45 remains to be explained. A possible
explanation for the absence of Cu in the plagiogranite
brine-rich ﬂuid inclusions could be its strong parti-
tioning toward a coexisting vapour phase generated by
ﬂuid unmixing. Published thermodynamic estimates,
experimental data and PIXE analyses on coexisting
brine and vapour inclusions have shown that Cu is
highly volatile as gaseous Cu-Cl complexes (Henley
and McNabb, 1978; Ryan et al., 1993; Ulrich et al.,
1999). PIXE-analyses of gas-rich inclusions in Oman
plagiogranites are not available. However, the fact that
many Cl-enriched brine inclusions from Oman ophi-
olites are both oxidised and pyrite-chalcopyrite-free
also favours their interpretation in terms of residual
liquids that have boiled, because H2SandHClare
also partitioned towards the vapour (Drummond and
Ohmoto, 1985) strongly.
To conclude, the only explanationwe see to recon-
cile the oxygen isotope data on minerals and inclusion
ﬂuids related, and the V-X properties of the inclu-
sion ﬂuids in the Oman plagiogranites, taking also
in account the plausible mechanisms of generation of
NaCl- and metal-rich brines in the oceanic crust is as
follows : brine inclusions are a mixture of exsolved
magmatic water and evolved seawater, that have been
oxidized, Cl-enriched and depleted in Cu and S by a
process of phase separation (Drummond and Ohmoto,
1985; Ryan et al., 1993; Ulrich et al., 1999). The
primary V-X properties of most halite-saturated and
two-phase ﬂuid inclusions do not presently keep the
microthermometric record of the boiling process be-
cause, probably, they have been altered by various
post-trapping changes, such as necking.
Summary and main conclusions
In the Oman ophiolite, about 25 occurrences of poly-
metallic massive sulﬁde deposits displayed in the
extrusive sequence attest the former intensity of a
high-temperature hydrothermal circulation above the
accretion magma chambers. Elongated parallel to the
average trend of the sheeted dyke complex over hun-
dreds of km, this ophiolite is particularly suitable for
studying the along-strike variations of the distribution
and intensity of the hydrothermal circulation (whereas
the Troodos ophiolite in Cyprus, for instance, bet-
ter exposes the variations in width of the epidotised
The results of the present study are summarised in
the following points:
(1) The distribution of diabases, chloritised dykes,
spilitized dykes and epidosites (in the order of increas-
ing intensity of hydrothermal alteration) was studied
along continuous outcropsin three selected areas.
(2) Epidosites, the most intensely altered zones, are
conﬁned to the lower part of the sheeted dyke com-
plex. They appear as narrow vertical bands of pale
green-yellowish colour, always parallel to the dykes
margins. Their distribution and spacing is highly vari-
able along the strike of the Oman ophiolite:
– The Muwaylah section, in the Haylayn massif,
representing a fossil axial discontinuity, is the most
massively epidotised area, with epidosite zones of 3–
15 m wide with an average spacing of 12 m. In this
area, there are two directions of dyking, and at least
one known massive sulﬁde deposit (Daris prospect).
– The wadi Andam area (Samail massif), representing
a much less tectonised site, close to a mantle diapir
and probably located midway between the tip and the
centre of an accretion segment, is also well (but less)
epidotised, with epidosites spacing of 10–25 m.
– The wadi Salahi area represents probably the cen-
tral part of a fossil accretion segment and is by far the
less altered site, with very few epidosites, in spite of
being located quite beneath the Zuha sulﬁde prospect.
This conﬁrms that the hot ascending ﬂuids in major
discharge zones are strongly focused.
(3) Halite-bearing ﬂuid inclusions are common among
the ﬂuid inclusions from about50 samples collected in
the plagiogranites, gabbros and sheeted dykes of the
Muwaylah and wadi Falah areas (SE Haylayn massif),
which also include: (a) monophased (liquid) inclu-
sions, generally stretched and deformed; (b) vapour-
dominated, low-salinity, 2-phase inclusions, with an
average salinity of 3.8 wt% eq. NaCl and average Th
of 370◦C; (c) liquid-dominated, low-salinity, 2-phase
inclusions, with an average salinity of 4.2 wt% eq.
NaCl and an average Th of 325◦C, and (d) liquid-
dominated, high-salinity inclusions, always contain-
ing a solid halite daughter phase, dissolving at higher
temperature (292–441◦C) than homogenisationof the
ﬂuid phases (230–403◦C).
(4) Plagiogranite sample MU 94-45 collected in the
Muwaylah area (Haylayn massif) is particularly rich
in quartz-epidote hydrothermal veins. The quartz is
rich in high-salinity, brine-rich inclusions, which con-
tain halite cubes dissolving at very high temperatures
(358–496◦C), corresponding to very high salinities
(43 to 59.2 wt% eq. NaCl ), and liquid-vapour homog-
enization around 275◦C on average. Besides:
– Molecular Raman Spectroscopy analyses have con-
ﬁrmed the aqueous nature of these inclusions and
the absence of detectable CO2,NH
Among the solid phases other than NaCl, hematite and
anhydrite could be determined.
– PIXE analyses on six halite-bearing inclusions in
plagiogranite MU 94-45 allowed to detect signiﬁcant
contents in Cl, K, Ca, Fe, Mn, Zn and/or Br. These
preliminary data conﬁrm that elements such as K, Ca,
Fe are present in these hydrothermalﬂuids, as deduced
previously from the presence of hematite, anhydrite
and/or an Fe-chloride identiﬁed in plagiogranitic in-
clusions. Scattered measured Cl/Br ratios range be-
tween that of seawater (close to 300), and values typ-
ical of magmatic felsic ﬂuids (close to 1000). The re-
markable absence of detectable Cu in these brine-rich
inclusions probably indicates that they have degassed.
(5) The measurement of the oxygen isotopic compo-
sition of the ﬂuids extracted from some ﬂuid inclu-
sions and of the associated host-minerals(quartz, epi-
dote) suggests that both seawater-derivedand magma-
derived ﬂuids have mixed in the High Temperature
Reaction Zone, in good agreement with previous
interpretations derived from PIXE-data.
(6) Many NaCl-rich brines, saturated in anhydrite
and/or hematite, are too oxidized to have been at equi-
librium with the plagiogranite magma, as the latter
has been generated by partial melting and Fe-Ti ox-
ide fractionation at low fO2-conditions (Spulber and
Rutherford, 1983). This, plus the Cu-depleted na-
ture of the brine inclusions (e.g., Ryan et al., 1993),
strongly supports the fact that the Cl-rich brines are
residual liquids that have degassed. However, the V-X
properties of most NaCl-rich brines in plagiogranite,
particularly the data from sample MU 94-45, do not
support at all the phase separation process. In order to
explain the inconsistency between microthermometric
data and oxygen isotopic or PIXE data, we tentatively
suggest that most brine-rich inclusions in plagiogran-
ites from Oman have had their shape modiﬁed after
trapping, for instance by necking.
We are grateful to the scientiﬁc team of the LPS-
CEA Laboratory at Saclay for providing us with an
easy access to the nuclear microprobe. J.-P. Gallien,
L. Daudin and H. Khojda are particularly acknowl-
edged for their technical assistance on the PIXE data
acquisition. Also , M. Volﬁnger and S. Gama are
acknowledged for their help in the processing of X-
ray spectra. Many thanks also to Craig Manning and
Christian Marignac for their careful comments and
reviewing of this paper.
Alabaster, T. and Pearce, J.A., 1985, The interrelationship between
magmatic and ore forming hydrothermal processes in the Oman
ophiolite, Econ. Geol. 80: 1–16.
Alabaster, T., Pearce, J.A., Mallick, D.I.J. and El Boushi, I.M.,
1980, The volcanic stratigraphy and location of massive sulphide
deposits in the Oman ophiolite, In A. Panayiotou (ed.) Ophi-
olites, Proceedings of the International Ophiolite Symposium,
Cyprus 1979, Geologocal Survey Dept, Nicosia, pp. 751–757.
Alabaster, P., Pearce, J.A. and Malpas, J., 1982, The volcanic
stratigraphy and petrogenesis of the Oman ophiolite complex,
Contrib. Mineral. Petrol. 81: 168–183.
Alt, J.C., 1995, Subseaﬂoor processes in mid-ocean ridge hy-
drothermal systems, In Susan Humphris et al. (eds), Seaﬂoor
hydrothermal systems: physical, chemical, biological and geo-
logical interactions, AGU, Geophys. Monograph 91: 85–114.
Alt, J.C., Zuleger, E. and Erzinger J., 1995, Mineralogy and sta-
ble isotopic compositions of the hydrothermally altered lower
sheeted dyke complex, Hole 504B, Leg 140, Proc. Ocean Drill.
Program, Sci. Results,137/140: 155–166.
Alt, J.C., Kinoshita, H., Stokking, L.B. and Michael, J.P., 1996,
Proceedings of the Ocean Drilling Programme, College Station,
Anderson, A.J., Clark, A.H., Ma, X.P., Palmer, G.R., MacArthur,
J.D. and Roedder, E., 1989, Proton-Induced X-Ray and Gamma-
Ray Emission Analysis of Unopened Fluid Inclusions, Econ.
Geol. 84: 924–939.
Berndt, M.E. and Seyfried, W.E. Jr, 1990, Boron, bromine, and
other trace elements as clues to the fate of chlorine in mid-ocean
ridge vent ﬂuids, Geochim. Cosmochim. Acta 54: 2235–2245.
Bischoff, J.L., 1980, Geothermal system at 21◦N, East Paciﬁc
Rise: physical limits on geothermal ﬂuid and role of adiabatic
expansion, Science 27: 1465–1469.
Bischoff, J.L. and Rosenbauer, R.J., 1989, Salinity variation in
submarine hydrothermal systems by layered double-diffusive
convection, J. Geol. 97: 613–623.
Bodnar, R.J., 1993, Revised equation and table for determining
the freezing point depression of H2O-NaCl solutions, Geochim.
Cosmochim. Acta 57: 683–684.
Bodnar, R.J., 1994, Synthetic ﬂuid inclusions: XII. The system
H2O-NaCl. Experimental determination of the halite liquidus
and isochores for a 40 wt% NaCl solutions, Geochim. Cos-
mochim. Acta 58: 1053–1063.
Boudier, F. and Coleman, R.G., 1981, Cross section through the
peridotite in the Semail ophiolite, southeastern Oman Mountains,
J. Geophys. Res. 86: 2573–2592.
Bougault, H., Charlou, J.L., Fouquet, Y. and Needham, H.D.,
1990, Activité hydrothermale et structure axiale des dorsales
est-paciﬁque et médio-atlantique, Oceanol. Acta Vol. spéc. 10:
Burnham, C.W. and Ohmoto, H., 1980, Late-stage processes of
felsic magmatism, Min. Geol. Spec. Issue 8: 1–11.
Butterﬁeld, D.A., Massoth, G.J., Mc Duff, R.E., Lupton, J.E. and
Lilley, M.D., 1990, Geochemistry of hydrothermal ﬂuids from
axial seamount emissions study vent ﬁeld: Juan de Fuca Ridge,
the Cleft Segment: sub-seaﬂoor boiling and subsequent ﬂuid
interaction. J. Geophys. Res. 95: 12895–12921.
Campbell, A.C. and Edmond, J.M., 1989, Halide systematics of
submarine hydrothermal vents, Nature 342: 168–170.
Cathless, L.M., 1983, An analysis of hydrothermal system respon-
sible for massive sulﬁde deposition in the Hokuroku Basin of
Japan, Econ. Geol. 5: 439–487.
Ceuleneer, G., Nicolas, A. and Boudier, F., 1988, Mantle ﬂow pat-
tern at an oceanic spreading center: the Oman peridotites record,
Tectonophysics 151: 1–26.
Clayton, R.N. and Mayeda, T.K., 1963, The use of bromine
pentaﬂuoride in the extraction of oxygen from oxides and sil-
icates for isotopic analyses, Geochim. Cosmochim. Acta 27:
Cline, J.S. and Bodnar, R.J., 1991, Can economic porphyry copper
mineralization be generated by a typical calk-alcaline melt? J.
Geophys. Res. 96, 8113–8126.
Cloke, P.L.and Kesler, S.E., 1979, The halite trend in hydrothermal
solutions, Econ. Geol. 74: 1823–1831.
Constantinou, G., 1987, Massive sulﬁde deposits, In C. Xenophon-
tos and J.G. Malpas (ed.), Troodos 87, Ophiolites and Oceanic
Lithosphere Symposium, Field excursion Guidebook, Geologi-
cal Survey Dept of Cyprus, Nicosia, pp. 92–113.
Cowan, J. and Cann, J., 1988, Supercritical two-phase separation of
hydrothermal ﬂuids in the Troodos ophiolite. Nature 333: 259–
Craig, H., 1965, The measurement of oxygen isotope paleotempera-
tures, in Spoleto conference on stable isotopes in oceanographic
studies and paleotemperatures, Proc., 3, Pisa, Italy, Consiglio
Nazionale delle Recherche, Laboratorio di Geologia Nucleare,
Dhamelincourt, P., Bény, J.M., Dubessy J. and Poty, B, 1979,
Analyse d’inclusions ﬂuides à la microsonde MOLE à effet
Raman, Bull. Miner. 102: 600–610.
Dick, H.J.B., Meyer, P.S., Bloomer, S., Kirby, S., Stakes, D.
and Mawer, C., 1991, Lithostratigraphic evolution of an in-situ
section of oceanic layer 3, In R.P. Von Herzen, P.T. Robin-
son et al. (eds), Proc. ODP Sci. Results, 118, College Station,
Dixon, S. and Rutherford, M., 1979, Plagiogranites as late-stage
immiscible liquids in ophiolite and mid-ocean ridge suites: an
experimental study, Earth and Planet. Sci. Letters 45: 45–60.
Dodson, M.H., 1973, Closure temperature in cooling geochrono-
logical and petrological systems, Contrib. Mineral. Petrol. 40:
Drummond, S.E. and Ohmoto, H., 1985, Chemical evolution and
mineral deposition in boiling hydrothermal systems, Econ. Geol.
Dubois, M., 1984, Plagiogranites et hydrothermalisme: une ap-
proche à partir des complexes ophiolitiques de Chypre et
d’oman. Ph.D. Thesis, Univ. Nancy I, France, pp. 135.
Engelmann, C. and Revel, G., 1991, Design and possibilities of
the new ‘Pierre Sue’ laboratory microprobe. Nuclear Instr. Meth.
Fehn, U., Green, K.E., Von Herzn, R.P. and Cathles, L.M., 1983,
Numerical models for the hydrothermal ﬁeld at the Galapagos
Spreading Center, J. Geophys. Res. 88: 1033–1048.
Gunter, W.D., Chou I-Ming and Girsperger S., 1983, Phase relation
in the system NaCl-KCl-H2O. II: Differential thermal analysis
of the halite liquidus in the NaCl-H2O binary above 450 ◦C,
Geochim. Cosmochim. Acta 47: 843–873.
Haymon, R.M., Koski, R.A. and Sinclair, C., 1984, Fossils of hy-
drothermal vent worms discovered in Cretaceous sulﬁde ores of
the Semail ophiolite, Oman, Science 223: 1407–1409.
Haymon, R.M., Koski, R.A. and Abrams, M.J., 1989, Hydrothermal
discharge zones beneath massive sulﬁde deposits mapped in the
Oman ophiolite, Geology 17: 531–535.
Henley, R.W. and McNabb, A., 1978, Magmatic vapour plumes and
ground water interaction in porphyry copper emplacement. Econ.
Geology 73: 1–20.
Ixer, R.A., Alabaster, T. and Pearce, J.A., 1984, Ore petrography
and geochemistry of massive-sulphide deposits within the Samail
ophiolite, Oman, Trans. Inst. Min. Metall. B93: 114–124.
Jehl, V., Poty, B. and Weisbrod, A., 1977, Hydrothermal metamor-
phism of the oceanic crust in North Atlantic Ocean, Bull. Soc.
Géol. France XIX: 1213–1221.
Juteau, T., Ernewein, M., Reuber, I., Whitechurch, H. and Dahl,
R., 1988a, Duality of magmatism in the plutonic sequence of the
Sumail nappe, Oman, Tectonophysics 151: 107–135.
Juteau, T., Beurrier, M., Dahl, R. and Nehlig, P., 1988b, Segmen-
tation at a fossil spreading axis: the plutonic sequence of the
Wadi Haymiliyah area (Haylayn Block, Samail Nappe, Oman),
Tectonophysics 151: 167–197.
Kelley, D.S., 1990, Fluid circulation in a submarine paleohydrother-
mal system, Troodos ophiolite, Cyprus: ﬂuid inclusion evidence
for deep-seated circulation of brines in the oceanic crust. Ph.D.
thesis, Dalhousie University, Halifax, Canada, pp. 168.
Kelley, D.S. and Delaney, J.R., 1987, Two-phase separation and
fracturing in mid-ocean gabbros at temperatures greater than
700◦C, Earth Planet. Sci. Lett. 83: 53–66.
Kelley, D.S. and Robinson, P.T., 1990, Development of a brine-
dominated hydrothermal system at temperatures of 400–500 ◦C
in the upper level plutonic sequence, Troodos ophiolite, Cyprus,
Geochim. Cosmochim. Acta 54: 653–661.
Kelley, D.S., Robinson, P.T. and Malpas, J.G., 1992, Processes of
brine generation and circulation in the oceanic crust: ﬂuid inclu-
sion evidence from the Troodos ophiolite, Cyprus, J. Geophys.
Res. 97: 9307–9322.
Kelley, D.S., Vanko, D.A. and Gu, C., 1995, Fluid evolution
in crustal layer 2: Fluid inclusion evidence from the sheeted
dyke complex, Hole 504B, Costa Rica Rift, Proc. Ocean Drill
Program, Sci. Results 140: 191–198.
King, E.M., Barrie, C.T. and Valley, J.W., 1997. Hydrothermal alter-
ation of oxygen isotope ratios in quartz phenocrysts, Kidd Creek
mine, Ontario: Magmatic values are preserved in zircon. Geology
Kishima, N. and Sakai, H., 1980. Oxygen-18 and deuterium deter-
mination on a single water sample of a few milligrams, Anal.
Chem. 52: 356–358.
Laverne, C., Vanko, D.A., Tartarotti, P. and Alt, J.C., 1995, Chem-
istry and geothermometry of secondary minerals from the deep
sheeted dyke complex, Hole 504B, Proc. Ocean Drill. Program,
Sci. Results 137/140, 167–189.
Lécuyer, C., 1990, Hydrothermalisme fossile dans une paléo-croûte
associée à un centre d’expansion lent: le complexe ophiolitique
de Trinity (N. Californie, USA), Ph.D. thesis, Univ. Rennes I
(France), Mémoires et Documents du CAESS, 33, pp. 279.
Lécuyer C., Dubois M., Marignac Ch., Gruau G., Fouquet Y. and
Ramboz C., 1999. Phase separation and ﬂuid mixing in sub-
seaﬂoor back arc hydrothermal systems: A microthermometric
and oxygen isotope study of ﬂuid inclusions in the baryte-sulﬁde
chimneys of the Lau basin, J. Geophys. Res. 104: 17 911–17 927.
Lescuyer, J.L., Oudin, E. and Beurrier, M., 1988, Review of the
different types of mineralization related to the Oman ophiolitic
volcanism, in 7th IAGODD Symposium Lullea 1986 (Sweden),
Schwetzerbart’sche, Stuttgart, Germany, pp. 489–500.
Lippard, S.J., Shelton, A.W. and Gass, I.G., 1986, The ophiolite of
northern Oman, The Geol Soc. Memoir No 11, Blackwell Sci.
Publ., Oxford, pp. 178.
Manac’h, G., 1997, Caractérisation des circulations ﬂuides dans la
croûte océanique actuelle (zone de fracture Blanco) et fossile
(ophiolite d’Oman), Ph.D. thesis, Univ. Bretagne Occidentale,
Brest, France, pp. 341.
Maxwell, J.A., Teesdale, W.J. and Campbell, J.L. 1995, The Guelph
PIXE software package II, Nuclear Instr. Meth. B49: 407-421.
Mevel, C., 1984, Le métamorphisme dans la croûte océanique:
apport de la pétrologie à la compréhension des phénomènes
de circulation hydrothermale et de déformation (exemples dans
l’Atlantique), Thesis Doctorat d’Etat, Univ. Paris VI, pp. 434.
Mevel, C. and Cannat, M., 1991, Lithospheric stretching and hy-
drothermal processes in oceanic gabbros from slow-spreading
ridges. In Tj. Peters et al. (eds), Ophiolite genesis and evolution
of the oceanic lithosphere, Ministry of petroleum and Minerals,
Sultanate of Oman, pp. 293–312.
Moreau, O., 1998, Etude du système hydrothermal fossile
de l’ophiolite d’Oman: étude microthermométrique d’un pla-
giogranite situé dans la zone de réaction au toit des chambres
magmatiques, DEA Report, Univ. Bretagne Occidentale, Brest,
France, pp. 71.
Muenow, D.W., Garcia, M.O., Aggrey, K.E., Bednarz, U. and
Schminke, H.U., 1990, Volatiles in submarine glasses as a dis-
criminant of tectonic origin: application to the Troodosophiolite,
Nature 343: 159–161.
Nehlig, P., 1989, Etude d’un système hydrothermal océanique fos-
sile: l’ophiolite de Semail (Oman). Ph.D. thesis, Univ. Bretagne
Occidentale, Brest, France, pp. 600.
Nehlig, P., 1991, Salinity of oceanic hydrothermal ﬂuids: a ﬂuid
inclusion study, Earth Planet. Sci. Lett. 102: 310–325.
Nehlig, P., 1993, Interactions between magma chambers and
hydrothermal systems: oceanic and ophiolitic constraints, J.
Geophys. Res. 98: 19621–19633.
Nehlig, P. and Juteau, T., 1988a, Flow porosities, permeabilities
and preliminary data on ﬂuid inclusions and fossil thermal gra-
dients in the crustal sequence of the Semail ophiolite (Oman),
Tectonophysics 151: 199–222.
Nehlig, P. and Juteau, T., 1988b, Deep crustal seawater penetra-
tion and circulation at oceanic ridges: evidence from the Oman
ophiolite, Marine Geology 84: 209–228.
Nehlig, P., Watremez, P. and Juteau, T., 1990, Geometry of the
ocean ridges hydrothermal circulation: the example of the Se-
mail Ophiolite, Oman, In C. Xenophontos and J.G. Malpas
(ed.), Proceedings of theTroodos 87, Ophiolites and Oceanic
Lithosphere Symposium, Geological Survey Dept of Cyprus,
Nicosia, pp. 415-431.
Nehlig, P., Juteau, T., Bendel, V. and Cotten, J., 1994, The root
zones of oceanic hydrothermal systems: constraints from the
Semail ophiolite (Oman), J. Geophys. Res. 99: 4703–4713.
Nicolas, A., 1989, Structures of ophiolites and dynamics of oceanic
lithosphere, Kluwer Acad. Publ., pp. 367.
Nicolas, A. and Boudier, F., 1991, Rooting of the sheeted dyke com-
plex in the Oman ophiolite, In Ophiolite genesis and evolution
of the oceanic lithosphere, Peters, Tj. et al. (eds), Ministry of
petroleum and Minerals, Sultanate of Oman, pp. 39–54.
Nicolas, A., Reuber, I. and Benn, K., 1988, A new magma cham-
ber model based on structural studies in the Oman ophiolite,
Tectonophysics 151: 87–105.
Nicolas, A., Boudier, F., Ildefonse, B. and Ball, E., 2000, Accretion
of Oman and United Arab Emirates ophiolite – Discussion of a
new structural map 000: 000–000 (this issue).
O’Neil, J.R. and Adami, L.H., 1969, The oxygen isotope partition
function ratio of water and the structure of liquid water, J. Phys.
Chem. 73: 1553–1558.
Pallister, J.S. and Hopson, C.A., 1981, Samail ophiolite plutonic
suite: ﬁeld relations, phase variation, cryptic variation and lay-
ering, and a model of a spreading ridge magma chamber, J.
Geophys. Res. 86: 2593–2644.
Pﬂumio, C., 1991, Evidences for polyphased oceanic alteration of
the extrusive sequence of the Semail ophiolite from the Salahi
Block (Northern Oman), In Tj. Peters et al. (eds) Ophiolite
genesis and evolution of the oceanic lithosphere, Ministry of
petroleum and Minerals, Sultanate of Oman, pp. 313–351.
Ramboz, C., 1979, A ﬂuid inclusion study of the copper miner-
alization in the south-west Tintic District (Utah), Bull. Soc. Fr.
Mineral. 102: 622–632.
Ramboz, C., Oudin, E. and Thisse Y., 1988, Geyser-type discharge
at the bottom of the Atlantis II Deep (Red Sea): evidence from
ﬂuid inclusions in epigenetic anhydrite, Canad. Mineral. 26:
Regba, M., Agrinier, P., Pﬂumio, C. and Loubet, M., 1991, A geo-
chemical study of a fossil oceanic hydrothermal discharge zone
in the Oman ophiolite (Zuha sulphide prospect): evidence for a
polyphased hydrothermal history, In Tj Peters et al. (eds), Ophi-
olite genesis and evolution of the oceanic lithosphere, Ministry
of petroleum and Minerals, Sultanate of Oman, pp. 353–384.
Reuber, I., Nehlig, P. and Juteau, T., 1991, Evidence of a fossil
off-axis mantle diapir, deduced from crustal structures in the
Haylayn Block (Samail Nappe, Oman), J. Geodyn. 13: 253–278.
Richardson, C.J., Cann, J.R., Richards, H.G. and Cowan J.G.,
1987, Metal-depleted root zones of the Troodos ore-forming
hydrothermal systems, Cyprus, Earth Planet. Sci. Lett. 84:
Roedder, E., 1984, Fluid inclusions. Reviews in Mineralogy, 12,
Mineralogical Association of America, pp. 644.
Roedder, E., 1992, Fluid inclusion evidence for immiscibility in
magmatic differentiation, Geochim. Cosmochim Acta 56: 5–20.
Rosencrantz, E., 1983, The structure of sheeted dykes and as-
sociated rocks in North Arm Massif, Bay of Islands ophiolite
complex, and the extrusive process at oceanic spreading centers,
Can. J. Earth Sci. 20: 787–801.
Rothery, D.A., 1983, The base of the sheeted dyke complex,
Oman Ophiolite: implications for magma chambers at oceanic
spreading axes, J. geol. Soc. London 140: 287–296.
Ryan, C.G., Cousens, D.R., Heinrich, C.A., Grifﬁn, W.L., Sie, S.H.
and Mernagh, T.P., 1991, Quantitative PIXE microanalysis of
ﬂuid inclusions based on a layered yield model, Nucl. Instr. and
Meth. B54: 292–297.
Ryan, C.G., Heinrich, C.A. and Mernagh, T.P., 1993, PIXE micro-
analysis of ﬂuid inclusions and its application to study ore metal
sgregation between magmatic brine and vapour, Nucl. Instr. and
Meth. B77: 463–471.
Scott, S.D., 1997, Chemistry and physics of submarine hydrother-
mal systems. In H.L. Barnes (ed.), Geochemistry of hydrother-
mal ore deposits, 3rd ed., Chapt. 16, pp. 797–875.
Sleep, N.H., 1991, Hydrothermal circulation, anhydrite precipita-
tion, and thermal structure at Ridge axes, J. Geophys. Res. 96:
Spulber, S.D., and Rutherford, M.J., 1983, The origin of rhyolite
and plagiogranite in oceanic crust: An experimental study, J.
Petrol. 24: 1–25.
Sterner, S.M. and Bodnar, R.J., 1984, Synthetic ﬂuid inclusions in
natural quartz. I: Compositional types synthetized and applica-
tions to experimental geochemistry, Geochim. Cosmochim. Acta
Ulrich, T., Günther, D. and Heinrich, C.A., 1999, Gold concen-
trations of magmatic brines and the metal-budget of porphyry
copper deposits, Nature 399: 676–679.
Vanko, D.A., 1986, High chlorine amphiboles from oceanic rocks:
product of highly saline hydothermal ﬂuids? Am. Mineral. 71:
Vityk, M.O., Krouse, H.R. and Demihov, Y.N., 1993. Preservation
of δ18O values of ﬂuid inclusion water in quartz over geo-
logical time in an epithermal environment: Beregovo deposit,
Transcarpathia, Ukraine, Earth Planet. Sci. Lett. 119: 561–568.
Volﬁnger, M., Choï, C.G., Ramboz, C., Aïssa, M. and Edon, M.,
1997, Tracing fossil and present day ﬂuids in rocks: application
of the nuclear microprobe, Nucl. Instr. And Meth. B130: 692–
Von Damm, K.L., 1988, Systematics and postulated controls on
submarine hydrothermal solution chemistry, J. Geophys. Res. 93:
Yang, K. and Scott, S.D., 1996, Possible contribution of a metal-rich
magmatic ﬂuid to a sea-ﬂoor hydrothermal system, Nature 383:
Zheng, Y.F., 1993a. Calculation of oxygen isotope fractionation
in anhydrous silicate minerals, Geochim. Cosmochim. Acta 57:
Zheng, Y.F., 1993b. Calculation of oxygen isotope fractionation in
hydroxyl-bearing silicates, Earth Planet. Sci. Lett. 120: 247–263.