Exhumation during oblique transpression: The Feiran–Solaf region, Egypt

Abstract

The Feiran–Solaf metamorphic complex of Sinai, Egypt, is one of the highest grade metamorphic complexes of a series of basement domes that crop out throughout the Arabian-Nubian Shield. In the Eastern Desert of Egypt these basement domes have been interpreted as metamorphic core complexes exhumed in extensional settings. For the Feiran–Solaf complex an interpretation of the exhumation mechanism is difficult to obtain with structural arguments as all of its margins are obliterated by post-tectonic granites. Here, metamorphic methods are used to investigate its tectonic history and show that the complex was characterized by a single metamorphic cycle experiencing peak metamorphism at ∼700–750 °C and 7–8 kbar and subsequent isothermal decompression to ∼4–5 kbar, followed by near isobaric cooling to 450 °C. Correlation of this metamorphic evolution with the deformation history shows that peak metamorphism occurred prior to the compressive deformation phase D2, while the compressive D2 and D3 deformation occurred during the near isothermal decompression phase of the P–T loop. We interpret the concurrence of decompression of the P–T path and compression by structural shortening as evidence for the Najd fault system exhuming the complex in an oblique transpressive regime. However, final exhumation from ∼15 km depth must have occurred due to an unrelated mechanism.

Full text

Exhumation during oblique transpression: The Feiran–Solaf
region, Egypt
T. S. ABU-ALAM AND K. STU
¨WE
Institut fu
¨r Erdwissenschaften, Universita
¨t Graz, Universita
¨tsplatz 2, A-8010 Graz, Austria (tamer.abu-alam@uni-graz.at)
ABSTRACT The Feiran–Solaf metamorphic complex of Sinai, Egypt, is one of the highest grade metamorphic
complexes of a series of basement domes that crop out throughout the Arabian-Nubian Shield. In the
Eastern Desert of Egypt these basement domes have been interpreted as metamorphic core complexes
exhumed in extensional settings. For the Feiran–Solaf complex an interpretation of the exhumation
mechanism is difficult to obtain with structural arguments as all of its margins are obliterated by post-
tectonic granites. Here, metamorphic methods are used to investigate its tectonic history and show that
the complex was characterized by a single metamorphic cycle experiencing peak metamorphism at 700–
750 C and 7–8 kbar and subsequent isothermal decompression to 4–5 kbar, followed by near isobaric
cooling to 450 C. Correlation of this metamorphic evolution with the deformation history shows that
peak metamorphism occurred prior to the compressive deformation phase D
2
, while the compressive D
2
and D
3
deformation occurred during the near isothermal decompression phase of the P–T loop. We
interpret the concurrence of decompression of the P–T path and compression by structural shortening as
evidence for the Najd fault system exhuming the complex in an oblique transpressive regime. However,
final exhumation from 15 km depth must have occurred due to an unrelated mechanism.
Key words: deformation phases; exhumation process; Pan-African Orogeny; P–T–D path; thermo-
dynamic modelling.
INTRODUCTION
The Sinai Peninsula is the northern segment of the
Arabian-Nubian Shield, which comprises the north-
ern part of the East African Orogen. The East
African Orogen was assembled by accretion of an
intraoceanic island arc system associated with the
closing of the Mozambique Ocean during the Neo-
proterozoic (Engel et al., 1980; Gass, 1982; Kro
¨ner,
1984; Kro
¨ner et al., 1994; Fritz et al., 1996, 2002;
Katz et al., 2004). The entire arc assembly was then
thrust during the Pan-African orogeny over a cra-
tonic, gneissic basement, which is exposed in a series
of tectonic windows throughout northern Africa and
Arabian Peninsula (Fig. 1). This basement is often
considered to be of much older age than the arc
assembly (El-Gaby et al., 1990). A lack of knowledge
regarding the relationship between the low-grade arc
assembly and the much higher grade basement units
underneath severely limits our ability to reconstruct
the tectonic evolution of this region (Stern &
Manton, 1987).
One way to constrain this relationship is by studying
the metamorphic evolution of the gneissic basement
and correlating this with the structural events that
covered and later exhumed the gneisses from under-
neath the arc assembly. Such studies have been per-
formed in several of the gneissic windows in the
Eastern Desert of Egypt (e.g. Meatiq and Sibai core
complexes, Fritz et al., 2002) and also in Sinai (e.g. Kid
area: Abu El-Enen et al., 2003; Blasband et al., 1997,
2000; Brooijmans et al., 2003; Eliwa et al., 2008; Taba
area: Abu El-Enen et al., 2004; Eliwa et al., 2008).
These studies have shown that the basement domes
may be high grade equivalents of the low-grade arc
assembly, and that some of these domes are core
complexes in the classic sense. Several of these studies
have also shown that the exhumation of the gneiss
domes is associated with the activity of a major (more
than 1000 km long and hundreds of km wide) NW–SE
striking sinistral shear zone system called the Najd
fault system (Fig. 1). However, one of the highest
grade complexes in Egypt and the largest metamorphic
one in Sinai has not been studied metamorphically and
its exhumation history has not been explained: the
Feiran–Solaf metamorphic complex (FSMC) (Fig. 2).
This complex constitutes a narrow strip of gneisses in
the northwestern part of the exposed igneous and
metamorphic rocks in Sinai.
El-Gaby & Ahmed (1980) recognized that the
FSMC is made up of two asymmetric doubly-plunging
antiforms, namely the Feiran antiform and the Solaf
antiform, separated by synformal trough. This geo-
metry suggests that the FSMC may be a core complex
similar to those of the Eastern Desert of Egypt (e.g.
Fritz et al., 1996, 2002; Loizenbauer et al., 2001; Abd
El-Naby et al., 2008; Khudeir et al., 2008), but we will
show here that this is not the case.
J. metamorphic Geol., 2009, 27, 439–459 doi:10.1111/j.1525-1314.2009.00827.x
2009 Blackwell Publishing Ltd 439

This study investigates the metamorphic evolution
of Feiran–Solaf complex to constrain the tectonic
evolution of the area. A mineral equilibria approach is
used with petrogenetic pseudosections and relate the
derived metamorphic history to the detailed structural
map and structural evolution derived by El-Shafei &
Kusky (2003). Our tectonic model is then correlated
with independent field evidence, existing geochrono-
logical ages from the region and with other gneiss
domes elsewhere in the Arabian-Nubian Shield.
GEOLOGICAL SETTING
The FSMC is a NW trending elongate folded belt
40 km long and 5–11 km wide (Fig. 2) parallel to the
orientation of the Najd fault system. It is made up of
migmatitic biotite and hornblende gneisses, quartzo-
feldspathic gneisses and hornblende gneisses with
subordinate schists, amphibolite and calcsilicate rocks.
The complex was intruded and surrounded by a
number of granitic plutons, including pre-, syn- to
post-tectonic granitoids, diorite and large volumes of
dykes with mafic to felsic composition. These intru-
sions obliterate all structural relationships along the
margins of the complex. The northwestern part of the
complex is covered by Phanerozoic sedimentary rocks.
The complex can be divided into two zones: the Feiran
zone in the north-west and the Solaf zone in the south-
east. These two zones are separated by a diorite
intrusion (Fig. 2). The Feiran zone contains migmatitic
biotite and hornblende gneisses, hornblende gneisses,
quartzofeldspathic gneisses and locally some dolerites.
In contrast, the Solaf zone is mostly composed of
quartzofeldspathic gneisses (with very minor bands of
metagraywackes but with mappable portions of calc-
silicate rocks; Fig. 2). There are also less abundant
granitic gneisses and amphibolites as small lenses and
pockets.
The rock units in the entire complex are systemati-
cally arranged around two doubly-plunging antiformal
structures: The Feiran antiform in the north-west and
the Solaf antiform in the south-east (Fig. 2). Stern &
Manton (1988) suggested that a thrust fault separates
the two antiforms, but El-Shafei & Kusky (2003) found
no evidence for such a fault. Migmatitic biotite and
hornblende gneisses occur in the core of the structure,
the hornblende gneisses structurally above and the
quartzofeldspathic gneisses above those. The calcsili-
cates generally appear to be located in the structurally
highest levels. This arrangement has been interpreted
as a stratigraphic succession by Ahmed (1981).
El-Gaby & Ahmed (1980) interpreted the Feiran
36 42
24
28
Red Sea
Meatiq
Sibai
Kid
Taba
Sa'al
Feiran-Solaf
Qazaz
Wajiyah
An Nakhil
Habariyah
Kirsh
Tin
24
28
N
Pan-African arc assembly
Gneiss complexes
Phanerozoic sequence
Strike-Slip
Marsa
Hamadat
Alam
Hafafit
Hurghada
Quseir
Allaqi-Heiani
suture
Jiddah
Halaban-Zarghat
fault zone
Ar Rika
fault zone
Yanbu
suture
Ad Damm
fault zone
Bir Umq-Nakasib
suture
Onib-Sol-Hamed
suture
Extensional
exhumation
Transpressive
exhumation
3
1
Fig. 1. Metamorphic gneiss complexes, Najd fault system and sutures of the northern part of the Arabian-Nubian Shield. Black
arrows show principal stress directions for the Najd fault system as a whole. Small white arrows show strain in individual gneiss
complexes. Note that this is extensional in the Eastern desert, but transpressive in the Feiran–Solaf metamorphic complex.
(Complied from various sources, including: Abdelsalam & Stern, 1996; Fritz et al., 1996; Johnson, 1998).
440 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

gneisses to represent a thick (>5000 m) sedimentary
succession with minor mafic intercalations. Several
authors have observed that metamorphism in the
FSMC appears to increase towards the core of the
structure, i.e. towards deeper structural levels. In
the migmatitic biotite and hornblende gneisses, the
metamorphic grade is reported to have reached upper
amphibolite facies conditions from 640 to 700 Cat
pressure from 4 to 5 kbar during the Pan-African
Orogeny (e.g. Eliwa et al., 2008). A zircon age derived
by Stern & Manton (1987) suggests 632 ± 3 Ma for
the high-grade migmatitic event. A K–Ar biotite
cooling age (Eliwa et al., 2008) suggests 617–594 Ma
as the time of cooling and exhumation.
Structure and kinematics
Most authors agree that the history of the Pan-African
deformation in Sinai can be described in terms of four
Proterozoic deformation phases D
1
–D
4
(e.g. Eliwa
et al., 2008), where D
1
is the primary foliation forming
event interpreted to be related to an early Pan-African
extensional environment. D
2
and D
3
are shortening
phases forming the large scale geometry of the meta-
morphic complexes and D
4
is a gentle warping event.
For the FSMC, Fowler & Hassan (2008) suggested
that the major doubly-plunging antiforms building the
map scale geometry of the complex were formed
during D
3
(Fig. 2). As this is a somewhat different
interpretation from that of El-Shafei & Kusky (2003),
a summary of the structural evolution is presented
based on the literature and our own data below.
The earliest structure observed in the FSMC is the
primary lithological layering termed S
0
. Lithological
bands are from <1 cm to several metres in thickness and
are defined mainly by variation in the amounts and type
of mafic minerals. Good examples of this primary lith-
ological layering (S
0
) are thin continuous layers of
amphibolite (Fig. 3a) and calcsilicate rock within
quartzofeldspathic gneisses (Fowler & Hassan, 2008).
Fowler & Hassan (2008) as well as El-Shafei &
Kusky (2003) suggested that the dominant metamor-
phic foliation in the FSMC is S
1.
While this nomen-
clature is retained here, we note that Abu El-Enen
′
′
Fig. 2. Location map (MC, metamorphic complex) and simplified geological map of the Feiran–Solaf metamorphic complex. Contacts
are modified from Ahmed (1981); El-Shafei & Kusky (2003) and Fowler & Hassan (2008). The white squares are the sample locations
(F1, F63, F93d & F39). Structural cross-sections, A–A¢through the Solaf zone and B–B¢are drawn sub-parallel to the main fold axes
of the Feiran and the Solaf antiforms. Lithological changes are often gradational and the mapped contacts are therefore approximate.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 441
2009 Blackwell Publishing Ltd

(a)
F2 fold axe
F3axial plane
N
(c)
(e)
(g)
(b)
(d)
(f)
(h)
Granites
Metamorphic foliatio n
F2 axial plane
442 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

et al. (2003, 2004) suggested that the main foliation
forming event in the Taba and Kid belts of Sinai is D
2
.
Indeed, we found remnants of an earlier fabric as
inclusion trails in garnet in the FSMC. Nevertheless, as
these inclusion trails are rare, the main foliation event
is termed D
1
, keeping in mind that earlier phases may
have occurred.
D
1
formed a strong stretching lineation and
intrafolial F
1
-folds generally trending NW–SE. Shear
sense indicators indicate west and north-west directed
transport but Fowler & Hassan (2008) showed that the
principal strain during this event was vertical flattening
with stretching in both NW–SE and NE–SW direc-
tions. They also suggested that this extensional D
1
reflected the larger-scale extension related to the
breakup of Rodinia. In contrast, El-Shafei & Kusky
(2003) inferred that D
1
was responsible for the map
scale geometry of the FSMC. However, because the S
1
foliation is wrapped around the hinges of the map scale
antiforms, we follow here Fowler & Hassan (2008) and
suggest that the map scale geometry must be related to
later deformation phases.
The D
2
phase formed tight folds (F
2
) (Fig. 3b)
occasionally associated with a crenulation cleavage
(S
2
). The F
2
-folds commonly refold F
1
-folds. These
structures are well preserved in the Solaf zone along
Wadi Solaf. The F
2
-folds trend parallel to the F
1
trends
(NW–SE and NNW–SSE) and clearly formed due to
shortening in NE–SW direction. Because F
1
and F
2
axes are largely parallel, El-Shafei & Kusky (2003)
interpreted F
1
and F
2
to be both related to the same
deformation phase D
1
.
The third deformation phase D
3
formed inclined
open folds (F
3
) causing the map scale structure of the
complex and the Feiran and the Solaf antiforms
(Fowler & Hassan, 2008). El-Shafei & Kusky (2003)
interpreted this phase to be related to the Najd fault
system and associated it with NW–SE sinistral strike–
slip shear zones. These sinistral movements are well
recorded along the entire study area (Fig. 3c) and well
preserved in the calcsilicates. The F
3
hinges are roughly
parallel to the F
1
and the F
2
hinges trending shallowly
(NW–SE and NNW–SSE) parallel to the regional
trend of the map scale antiforms. The F
3
-folds
commonly refold F
2
-folds (Fig. 3b).
The D
4
deformation phase is an open warping event
that domed up the entire area and is associated with
NNW–trending quartz veins and gashes. F
4
-folds are
extremely rare and trend NE–SW, ENE–WSW and
E–W (Fig. 3d). The best example of an F
4
-fold is the
NE–SW synformal trough which separates the Feiran
and the Solaf antiforms (Fig. 2). The presence of
NNW-trending quartz veins and the gashes indicates
extension in NE–SW direction. It is possible that D
3
(shear zones along the margins of the complex) and D
4
are responsible for the exhumation of the region.
However, this is difficult to infer in the field as the
post- and syn-tectonic D
4
granitic bodies surround the
entire complex and obliterate all of its margins. Some
strong deformation features are developed along the
margins of the metamorphic complex, but these are
apparently related to the intrusion of the syn-tectonic
granites. For example, horizontal flattening shows
both sinistral and dextral movements and is especially
found at the end of Wadi Um-Takha and in the
calcsilicate rocks at the end of Wadi Dehest Abu-Talb
(Fig. 2 SE end of study region). During the Red Sea
and the Suez Gulf rifting, the entire study area was
deformed by brittle extensional structural features.
These features appear clearly as dyke swarms with
Phanerozoic age (e.g. Ahmed & Youssef, 1976;
El-Shafei & Kusky, 2003) and dextral strike–slip
faults. The best example of these faults is the fault
which cuts Gabal El-Banat and Gabal Serbal (post-
tectonic granites) (Fig. 2).
Lithologies
Most rocks of the FSMC are quartzofeldspathic
paragneisses with or without hornblende and with or
without migmatitic features. Aluminous metasedi-
ments are rare. Stern & Manton (1987) suggested that
the protoliths of the Feiran gneisses were probably
very immature, volcanogenic wackes or perhaps lithic
arenites inter-bedded with felsic tuffs.
The migmatitic biotite and hornblende gneisses
(henceforth called ÔmigmatitesÕ) (Fig. 3e) in the core of
Feiran zone are metatexites formed due to partial
melting (El-Shafei & Kusky, 2003). They are composed
of quartzofeldspathic leucosomes containing minor
biotite and hornblende and hornblende-rich meso-
somes containing minor biotite. In general, the
migmatites are pyroxene-free rocks, but rare pyroxene-
bearing migmatites occur. Complex structural patterns
are observed; include stromatic, folded stromatic,
ptygmatic and schlieren structures. The characteris-
tic feature of the migmatites is the abundance of
leucosomes that occur both layer parallel and cross-
cutting to the principal foliation (S
1
). The migma-
tites are intercalated with hornblende gneisses and
Fig. 3. Field photographs. (a) Two conformably inter-bedded amphibolite bands parallel to the sedimentary relict bedding of the
paragneisses in the Solaf zone. (b) Close up view showing F
2
isoclinal fold superimposed by F
3
open fold, looking NW. (c) Close
up view showing the D
3
sinistral strike–slip movement. (d) Two small-scale domes in the same orientation of the Feiran and the
Solaf doubly-plunging antiforms. The small-scale synform (F
4
), which separates the two domes, has the orientation of the NE–SW.
(e) The migmatites in the core of the Feiran antiform crop out in the more deeply eroded part along Wadi Feiran. (f) Close up view
showing the exfoliation appearance of the dolerite plugs at the northeastern part of the Feiran zone (Wadi Rummana). (g) Deformed
granitic dyke, the paragneisses around this dyke are rich in K-feldspar. (h) Syn-tectonic granites intruded perpendicular to the
metamorphic foliations (S
1
) at the end of Wadi Um-Takha.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 443
2009 Blackwell Publishing Ltd

quartzofeldspathic gneisses. These gneisses lack mig-
matitic features. There is no contact between the mi-
gmatites and the granite intrusions surrounding the
FSMC suggesting that the migmatization process is
not related to these granites.
The most abundant rock units in the Feiran zone are
the hornblende gneiss and the quartzofeldspathic
gneiss (Fig. 2). Rare metagraywacke bands occur.
There is intercalation and repetition between these two
rock types, but the proportion of quartzofeldspathic
gneisses increases toward the outer-rim of the Feiran
antiform. In the Solaf zone, the quartzofeldspathic
gneisses are the predominant rock unit. The horn-
blende gneiss is fine- to medium-grained, strongly
foliated and dark in colour. The quartzofeldspathic
gneisses are brownish grey colour, fine- to medium-
grained and strongly foliated. The hornblende gneisses
consist of hornblende, plagioclase, minor orthoclase
and quartz with or without biotite. They also contain
subordinate amounts of chlorite, iron oxides, zircon,
apatite, titanite and epidote. The quartzofeldspathic
gneisses are free of hornblende but otherwise very
similar. In the hornblende and the quartzofeldspathic
gneisses, the presence of relict sedimentary bedding
(e.g. El-Gaby & Ahmed, 1980; El-Gaby et al., 1990;
El-Shafei & Kusky, 2003) suggests that they had a
sedimentary origin.
The calcsilicates running parallel to and in contact
with the granitoid rocks at the outer most rim of the
Solaf antiform (Fig. 2) are characterized by well-
developed garnet crystals up to several centimetres in
size. The matrix is fine-grained, massive and light to
dark-green in colour containing calcite, wollastonite,
diopside, Ca-amphibole, scapolite, plagioclase and
quartz with subordinate amounts of zoisite ⁄clinozoi-
site.
Minor lithologies (with smaller occurrence than is
mappable on Fig. 2) include amphibolites, doleri te plugs
and granitic as well as mafic dykes. The amphibolites
occur as conformable inter-bedded bands, linear
bodies, and irregular lenses in the paragneisses of the
Solaf zone and in the syntectonic granitoids. They are
composed essentially of hornblende and plagioclase
with minor biotite and quartz. Titanite, apatite,
epidote and opaques are common accessory minerals.
El-Tokhi (1992) suggested an igneous origin for the
amphibolites, based on chemical evidence. But,
according to Fowler & Hassan (2008) and our field
evidence (Fig. 3a), the amphibolites are interpreted to
be of sedimentary origin. The dolerites occur as plugs
within the hornblende gneisses and the quartzofelds-
pathic gneisses at the very northeastern part of the
Feiran zone (Wadi Rummana) and consist of plagio-
clase, amphibole, pyroxene relics and iron oxides. The
amphibole and the pyroxene occur interstitially between
plagioclase crystals. Granitic dykes and pegmatitic veins
also cut the region. These are deformed and contain
inclusions from the surrounding rocks. The paragneisses
around these granitic dykes are enriched in K-feldspar.
The amount of K-feldspar in the country rocks
increasing toward these dykes (Fig. 3g).
The plutonic rocks in the study area are represented
by pre-, syn- and post-tectonic granites and diorite.
The pre- and the syn-tectonic granitoid rocks intruded
along the southern and eastern borders of the Solaf
zone. They extend further east forming numerous low-
relief hills characterized by intense shearing and they
contain many xenoliths of amphibolite, hornblende
gneisses and quartzofeldspathic gneisses (El-Shafei &
Kusky, 2003). In the eastern part of the study area,
these granites intruded along pre-existing oblique-slip
faults and parallel to the metamorphic foliation. In the
southern part of the study area, they intruded more or
less perpendicular to the metamorphic foliation
(Fig. 3h) and along the NE–SW normal faults. The
post-tectonic granites are characterized by the absence
of the dyke swarms and form large mountainous out-
crops bordering the complex along the eastern and the
southern parts of the Feiran zone (Gabal El-Banat and
Gabal Serbal). A large undeformed diorite body forms
an elongated NE-striking intrusion. El-Shafei & Kusky
(2003) suggested that this diorite intrusion may have
intruded along a pre-existing crustal weakness located
between the Solaf and Feiran zones. The entire area is
cut by a large number of dyke swarms ranging from
the Precambrian to the Cenozoic in age (El-Shafei &
Kusky, 2003).
PETROGRAPHY AND MINERAL CHEMISTRY
The rock types of the Feiran–Solaf complex can be
grouped into four major types: (i) migmatitic biotite
and hornblende gneisses (migmatites); (ii) hornblende
gneisses; (iii) quartzofeldspathic gneisses and (iv)
calcsilicates. The quartzofeldspathic gneisses have a
very simple mineralogy (Fig. 4a) consisting of biotite,
quartz, plagioclase and Fe–Ti oxides and will not be
discussed further. The calcsilicates have a complicated
fluid history and are studied separately (Abu-Alam &
Stu
¨we, Unpublished data). Here, we concentrate on
the migmatites, the hornblende gneisses, and the minor
but nevertheless important metagraywacke inclusions
therein. The minerals were analysed at the Institute of
Earth Science, Karl-Franzens-Universita
¨t Graz, Aus-
tria, using a JEOL JSM-6310 scanning electron
microscope following standard procedures, operating in
EDS ⁄WDS mode at 5 nA beam current, accelerating
voltage 15 kV and duration time is 100 s. The chemical
formulae were calculated using the PET1.1 (Dachs,
2004). The mineral abbreviations which will be used in
the following sections are from Holland & Powell
(1998).
Migmatitic biotite and hornblende gneisses (Migmatites)
The leucosomes are composed of quartz, plagioclase
and K-feldspar with or without biotite and the meso-
somes contain plagioclase, amphibole, quartz and
444 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

amph
cpx
Migmatites
Sphene
(c)
~ 750 °C
~ 5 kbar
old Mg-hb
phase
~ 450-550 °C
young act
phase
Chlorite
Sphene
200 µm
Hornblende gneisses
(e)
Hornblende gneisses
(d)
g
bi
q
bi, q, chl
inclusions
(pre- ) S
1
Metagraywackes
(f)
S
1 foliation
bi
Iron oxide
q+pl
(a)
Quartzofeldspathic gneisses
cpx
amph
Migmatites
(b)
Metamorphic foliation
(h)
Metagraywackes 1000 µm
(g)
Metagraywackes
g
bi
cd
1000 µm
1 mm 1 mm
0.02 µm1 mm
500 µm
Fig. 4. Photomicrographs of the main rock types. (a) The poor mineral assemblage of the quartzofeldspathic gneisses, (b) Clinopyroxene
of sample F39 is highly deformed and aligned along the S
1
foliation. (c) Hornblende porphyroblast of sample F39 has inclusions of
clinopyroxene. (d) Metamorphic foliation (parallel crystals of the hornblende) of sample F1. (e) Backscattered-electron image showing
two different phases of amphibole, where the high temperature hornblende was replaced by low temperature actinolite. (f) Garnet
porphyroblast of sample F93d. This porphyroblast has biotite, plagioclase, quartz and chlorite inclusions. (g) Deformed garnet crystal of
sample F93d with long axes parallel to the matrix foliation. (h) Cordierite porphyroblast of sample F93d with sigmoidal shape.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 445
2009 Blackwell Publishing Ltd

clinopyroxene (in order of abundance) along with
minor titanite, magnetite and epidote. Clinopyroxene
occurs as anhedral elongated crystals up to
0.3 ·0.7 mm which are highly deformed and dyna-
mically recrystallized in the S
1
foliation (Fig. 4b).
Hornblende occurs as both small grains (<0.5 mm)
and porphyroblasts (>1.5 mm). The small crystals are
highly deformed, parallel to S
1
foliation and in equi-
librium with the clinopyroxene. The porphyroblasts
are subhedral to anhedral weakly deformed grains.
These porphyroblasts have inclusions of clinopyroxene
(Fig. 4c). The long axes of the hornblende porphyro-
blasts are sub-parallel to the S
1
foliation. The horn-
blende crystals are overgrown by actinolite. Plagioclase
occurs as subhedral to anhedral equant crystals
attaining 0.75 mm in diameter and highly altered to
epidote. Quartz with undulose extinction occurs
interstitially between the plagioclase, hornblende and
the clinopyroxene crystals. The magnetite occurs as
euhedral crystals (0.3 mm) that overgrow the S
1
foli-
ation. Titanite occurs as small euhedral inclusion
crystals (<0.01 lm) in hornblende porphyroblasts
within S
1
foliation (Fig. 4c).
Based on the observations above we interpret the
metamorphic peak assemblage as hornblende, clino-
pyroxene, plagioclase, magnetite and titanite and sug-
gest that it grew during D
1
or just thereafter. The
hornblende crystals are overgrown by actinolite and
epidote overgrows plagioclase. Therefore, the assem-
blage hornblende, actinolite, plagioclase, albite,
epidote and titanite is interpreted as a later assemblage
(post-peak).
The mineral chemistry of the amphibole reveals
that they are magnesio-hornblende, edenite, edenitic-
hornblende, actinolitic-hornblende and actinolite
(Fig. 5a,b). The X
Fehb
=Fe
+2
⁄(Fe
+2
+Mg
+2
) ran-
ges (0.385–0.429), while the X
Feact
ranges around the
value 0.384. The Y(hb) = X(Al, M
2
) = (0.121–0.265)
and the Y(act) = (0.206–0.207) (Tables 1 & S1). The
cores of the plagioclase (pre-peak assemblage) have
X(an) = Ca ⁄(Ca + Na + K) ranges between 0.297
and 0.30, the plagioclase of the peak assemblage has
X(an) = (0.278–0.281), while plagioclase associated
with the late epidote and actinolite (post-peak assem-
blage), has X(an) = (0.252–0.278) (Table S2). Clino-
pyroxene has diopside–augite composition (Fig. 5c)
with X
Fe
=Fe
+2
⁄(Fe
+2
+Mg
+2
) = (0.401–0.431)
(Tables 1 & S3).
Hornblende gneisses
The mineralogy of the hornblende gneisses includes
hornblende, actinolite, plagioclase, albite, K-feldspar,
quartz, chlorite as well as accessory amounts of
epidote, magnetite, ilmenite, titanite, apatite and
zircon. Plagioclase and amphibole are the most abun-
dant minerals, amounting to 80 vol.%. Plagioclase
occurs as anhedral to subhedral elongated crystals
attaining 0.35 ·0.7 mm. The plagioclase crystals show
albite and albite-Carlsbad twinning, but untwined
crystals are observed too. The plagioclase crystals are
highly retrogressed to epidote. Hornblende is a green
colour and strongly pleochroic. Hornblende occurs as
anhedral elongated crystals (1.63 ·0.68 mm)
(Fig. 4d). Hornblende is retrogressed to actinolite,
chlorite and titanite (Fig. 4e). Quartz occurs as anhe-
dral to subhedral crystals that lie interstitially between
the plagioclase and the hornblende crystals. It fre-
quently exhibits undulose extinction. The long axes of
the plagioclase and the hornblende crystals are parallel
to the metamorphic foliation (Fig. 4d). K-feldspar
occurs as very small (apparent only under the scanning
electron microscope) anhedral equant crystals, which
lie interstitially between the plagioclase, the quartz and
the hornblende crystals. Ilmenite grows over magnetite
as lamellar intergrowths.
The peak assemblage is identified by a continuous
metamorphic foliation, which is defined by parallel
crystals of hornblende, plagioclase, quartz and iron
oxides (Fig. 4d). The retrogression of plagioclase to
epidote and hornblende to actinolite, chlorite and
titanite as well as the presence of albite identify the
retrograde assemblage. The large difference between
the crystal size of the K-feldspar and the surrounding
minerals, the anhedral shape of the K-feldspar
crystals as well as the presence of K-feldspar-rich
metasomatic zones around the granitic dykes
(Fig. 3g) suggests that the K-feldspar are in disequi-
librium with the other phases.
Chemical analyses reveal that the studied amphiboles
are magnesio-hornblende, edenite, actinolitic-
hornblende and actinolite (Fig. 5a,b). The Al
2
O
3
value
ranges between 8.11 and 7.79 wt.% for the magnesio-
hornblende and the edenite, while this value ranges
between 2.69 and 4.34 wt.% for the actinolitic-
hornblende and the actinolite (Tables 1 & S4). The
X
Fehb
ranges (0.475–0.477), while the X
Feact
ranges
(0.214–0.226). The Y(hb) = (0.063–0.066) and the
Y(act) = (0.16–0.22). The decrease in the Al
IV
and
Al
VI
from earlier to later amphiboles (from hornblende
to actinolite) reveals that the amphibole goes from
relatively high temperature–high pressure field to low
temperature–low pressure field (Rasse, 1974; Blundy &
Holland, 1990) (Fig. 5d). The relationship between
Al
VI
and Si (Fig. 5e) indicates that the hornblende
crystallized under pressure of £5 kbar while the
pressure of the actinolite is most probably <5 kbar
(Fig. 5d). According to the graphical geothermometer
of Blundy & Holland (1990), the hornblende crystal-
lized at a temperature below 750 C, while actinolite
formed between 450–600 C (Fig. 5f).
Metagraywackes
In Wadi Aleiyat (Fig. 2) a few rock bands were found
that contain cordierite and garnet. These rocks have a
metagraywacke bulk composition similar to sample
ES356 of Sawyer (1986) which was modelled by
446 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

Johnson et al. (2008, e.g. their fig. 3a). Because these
rocks prove to be petrologically useful (see below),
they are described in detail here although they are
quite rare in the FSMC. The rocks contain biotite,
garnet, cordierite, plagioclase, quartz, chlorite, ilmen-
ite and rutile. Muscovite and K-feldspar are notably
absent. Garnet and cordierite occur as porphyroblasts
(Fig. 4f). Garnet porphyroblasts are subhedral to
anhedral equant grains up to 5 mm in diameter with
inclusions of quartz, plagioclase, biotite, chlorite and
ilmenite. The inclusions define linear trails that are
oblique to the matrix foliation (S
1
) giving relic evi-
dence to an earlier deformation phase prior to D
1
.
Some garnet is idiomorphic and has straight crystal
boundaries against biotite in the matrix. Other garnet
crystals are highly deformed with long axes parallel to
the matrix foliation (Fig. 4g). The cordierite por-
phyroblasts are anhedral, highly altered grains 0.5 mm
and 3 mm in diameter. The cordierite crystals are
equidimensional porphyroblasts with sigmoidal shape
Fig. 5. Amphibole chemistry. (a) and (b) Amphibole composition classified after Leake et al. (1997). (c) Ca–Mg–Fe composition
ranges and nomenclature of clinopyroxene (after Morimoto, 1988). (d) Shows the relation between the Al
IV
and Al
VI
in the
amphibole of samples F1 and F63. (e) Relation between Al
VI
and Si of hornblende (after Rasse, 1974). (f) Al
IV
v. temperature for
amphiboles after Blundy & Holland (1990). The open circle is the hornblende of the hornblende gneisses (the old phase), the filled
circle is actinolite of the hornblende gneisses (the young phase) and the rhomb is the amphibole of the migmatite. (g) Zoning
profile of XMg = Mg
+2
⁄(Fe
+2
+Mg
+2
+Mn
+2
) through garnet grain.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 447
2009 Blackwell Publishing Ltd

in S
1
(Fig. 4h). Cordierite porphyroblasts have quartz,
plagioclase, biotite, chlorite and ilmenite inclusions.
Garnet and cordierite porphyroblasts are surrounded
by a matrix of biotite, plagioclase, chlorite, quartz,
ilmenite and rutile that define a continuous S
1
folia-
tion. The plagioclase occurs as euhedral to anhedral
elongate crystals attaining 0.15 ·0.3 mm. Quartz
occurs as anhedral to subhedral crystals that lie inter-
stitially between the plagioclase and the biotite crys-
tals. Biotite occurs as anhedral to subhedral flaky and
stumpy crystals attaining 0.25 ·0.75 mm and is
rarely overgrown by chlorite.
Garnet is weakly zoned (Fig. 5g). The cores of these
garnet crystals have Mg contents between 0.71 and
0.75 and Fe
+2
contents between 1.91 and 1.92, with
X
Feg
=Fe
+2
⁄(Fe
+2
+Mg
+2
) = (0.72–0.73) and Z(g) =
Ca ⁄(Fe + Mg + Ca) = (0.04–0.044). The rims of
these garnet crystals have Mg cation range between
0.61 and 0.69 and Fe
+2
cation range between 1.83 and
1.9, with X
Feg
= (0.734–0.752) and Z(g) = (0.04–
0.048). The cordierite is Mg-rich, with Mg and Fe
+2
cations (1.37–1.49) and (0.49–0.53) respectively. The
X
Fecd
=Fe
+2
⁄(Fe
+2
+Mg
+2
) ranges (0.250–0.261).
The biotite within the matrix (S
1
biotite), has X
Febi
=
Fe
+2
⁄(Fe
+2
+Mg
+2
) ranges (0.468–0.483) and
Y(bi) = X(Al, M
2
) = (0.27–0.29) (Table S5). The
chlorite minerals have X
Fechl
=Fe
+2
⁄(Fe
+2
+
Mg
+2
) = (0.381–0.4) Table S6.
Biotite–plagioclase–garnet geothermometery of
Bhattacharya et al. (1992) gives a temperature of
668–707 C for the peak assemblage and 676–722 C
using the thermometer of Ganguly & Saxena (1984).
The same minerals gave a pressure range of 6.2–8 kbar
according to Hoisch (1990) (Tables 2 & S7). The cor-
dierite–garnet (garnet-rim) geothermometer according
to Thompson (1976) gave a temperature range between
654 and 649 C and a range between 635 and 638 C
according to Perchuk & LavrentÕeva (1983) Table S8.
The presence of highly deformed garnet crystals
which have long axes parallel to the matrix foliation S
1
indicates that it grew early during D
1.
The sigmoidal
shapes of the cordierite crystals indicate that these
porphyroblasts grew later – possibly during D
2
. The
peak assemblage was identified by the presence of
garnet porphyroblasts. Based on these observations,
the peak metamorphism was interpreted as an event
between D
1
and D
2
. Biotite, chlorite, plagioclase,
quartz and ilmenite inclusions inside the garnet por-
phyroblasts were interpreted as the pre-peak assem-
blage while biotite, chlorite, plagioclase, quartz,
ilmenite, rutile (matrix mineral) with or without cor-
dierite were interpreted as the post-peak assemblage.
The presence of idiomorphic garnet crystals (Vernon,
2004) as well as the temperatures given by the biotite–
plagioclase–garnet geothermometer possibly indicates
that the rocks reached peak conditions above the
Table 1. Representative mineral analyses of samples F39, F1 and F63; amphibole (normalized to 23 O and ignoring H
2
O), plagioclase
(normalized to 8 O) and pyroxene (normalized to 6 O).
Sample no. F39 F1 F63
Assemblage hb di pl mt sph (peak assemblage) hb di pl sph (pre-peak ass.) hb act pl ep sph
(post-peak ass.)
hb act ab ep sph
(late post-peak
ass.)
Mineral hb6 hb14 pl28 pl31 hb13 pl16 pl18 py2 py30 act2 pl3 pl15 ab3 ab8 hb3 act2 hb8 act3
SiO
2
47.28 48.78 61.35 61.61 44.89 60.9 61.04 50.06 48.4 49.44 61.96 61.65 68.59 69.03 42.56 54.09 42.72 52.35
TiO
2
0.58 0.47 0.03 b.d.l. 0.94 0.05 b.d.l. 0.12 0.96 0.57 b.d.l. 0.03 b.d.l. b.d.l 0.92 0.13 0.98 0.23
Al
2
O
3
6.07 6.18 23.84 23.98 8.95 24.56 24.48 5.37 7.49 5.05 24.03 23.73 19.55 19.57 9.1 2.69 8.58 4.34
Cr
2
O
3
0.02 0.01 b.d.l. b.d.l. 0.05 0.03 b.d.l. 0.02 b.d.l. b.d.l. b.d.l. 0.03 0.02 b.d.l b.d.l 0.09 b.d.l. 0.06
FeO 16.85 16.1 0.27 0.15 18.45 0.29 0.26 14.54 16.81 15.27 0.24 0.42 b.d.l b.d.l 20.22 8.89 18.43 9.49
MnO 0.62 0.64 0.1 b.d.l 0.58 b.d.l 0.03 0.73 0.63 0.56 0.04 b.d.l. b.d.l b.d.l 0.69 0.56 0.45 0.49
MgO 12.66 13 0.04 b.d.l. 11.31 b.d.l. b.d.l. 12.18 11.66 13.5 b.d.l. b.d.l. 0.02 b.d.l 9.89 18.26 11.33 17.81
CaO 11.34 11.65 5.9 5.82 11.51 6.14 6.11 15.82 11.52 11.37 5.15 5.74 0.4 0.3 11.74 11.69 12.52 11.37
Na
2
O 0.87 0.81 8.13 8.18 1.17 7.8 7.86 0.65 1.19 0.75 8.00 8.03 11.22 11.38 1.64 0.39 1.4 0.57
K
2
O 0.61 0.54 0.32 0.24 0.98 0.19 0.19 0.07 0.82 0.62 0.68 0.32 0.01 0.01 0.89 0.17 0.8 0.22
Total 96.94 98.22 99.97 99.98 98.87 99.96 99.96 99.56 99.48 96.98 100.12 99.95 99.8 100.29 97.65 96.96 97.21 96.93
O 2323 8 823 8 8 6 623 8 8 8 823232323
Si 7.05 7.15 2.73 2.73 6.64 2.7 2.71 1.74 1.83 7.32 2.74 2.74 2.99 3.00 6.44 7.71 6.52 7.49
Ti 0.06 0.05 0.00 – 0.1 0.00 – 0.31 0.02 0.06 – 0.00 – – 0.14 0.01 0.15 0.02
Al 1.06 1.06 1.25 1.25 1.56 1.28 1.28 0.22 0.33 0.88 1.25 1.24 1.01 1.00 1.62 0.45 1.54 0.73
Cr 0.00 0.00 – – 0.01 0.00 – 0.00 – – – 0.00 0.00 – – 0.01 – 0.01
Fe
+3
0.32 0.18 0.01 0.01 0.4 0.01 0.01 0.00 0.00 0.02 0.01 0.01 – – 0.53 0.00 0.00 0.02
Fe
+2
1.77 1.78 0.00 0.00 1.87 0.00 0.00 0.42 0.51 1.86 0.00 0.00 – – 2.03 1.06 2.35 1.11
Mn 0.07 0.08 0.00 – 0.07 – 0.00 0.02 0.02 0.07 0.00 – – – 0.09 0.06 0.06 0.05
Mg 2.81 2.84 0.00 – 2.49 – – 0.63 0.67 2.98 – – 0.00 – 2.23 3.88 2.58 3.8
Ca 1.81 1.83 0.28 0.27 1.82 0.29 0.29 0.59 0.46 1.8 0.24 0.27 0.02 0.01 1.90 1.78 2.05 1.74
Na 0.25 0.23 0.7 0.7 0.33 0.67 0.67 0.04 0.08 0.21 0.68 0.69 0.95 0.96 0.48 0.1 0.41 0.15
K 0.11 0.1 0.02 0.01 0.18 0.01 0.01 0.00 0.04 0.11 0.04 0.02 0.00 0.00 0.17 0.03 0.16 0.04
X
Fehb
or
act
0.387 0.385 0.429 0.384 0.475 0.214 0.477 0.226
Y(hb) or (act) 0.121 0.221 0.207 0.207 0.063 0.16 0.066 0.22
X(an) 0.281 0.278 0.300 0.297 0.252 0.278 0.020 0.010
X
Fedi
0.401 0.431
Additional data are in the online data repository.
448 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

solidus. However, as the metagraywackes do not have
any partial melting indications in the field, we suggest
that the liquid mode in the rock during the peak con-
ditions was below 5% (Sawyer, 1999).
PSEUDOSECTIONS
Pseudosections were constructed for several bulk
compositions to constrain the P,Tand Xo paths.
Four samples were chosen to represent the rock types
discussed above. Sample F39 was chosen to represent
the migmatites. It is from one of the pyroxene-bearing
mesosomes in the central part of the study area
(2842¢13¢¢N and 3339¢30¢¢E; Fig. 2). Samples F1 and
F63 are hornblende gneisses from the next higher
unit. They were collected at 2844¢28¢¢N and
3334¢44¢¢E (F1) and at 2843¢39¢¢N and 3332¢41¢¢E
(F63). Sample F93d is characterized by the presence
of garnet and cordierite porphyroblasts. This sample
was collected from one of the rare metagraywacke
bands which are intercalated with the hornblende
gneisses and the quartzofeldspathic gneisses at the
end of Wadi Aleiyat (2840¢52¢¢N and 3339¢37¢¢E;
Fig. 2).
For bulk rock chemical analysis a Bruker Pioneer S4
X-ray fluorescence spectrometer was used at the
Institute of Earth Science, Karl-Franzens-Universita
¨t,
Graz, Austria. Samples were prepared as fused pellets
using Li
2
B
4
O
7
flux. The pseudosections were con-
structed using THERMOCALCTHERMOCALC tc330 (Powell & Holland,
1988) and the internally consistent dataset of Holland
& Powell (1998). The following a–x models were used:
amphibole (Diener et al., 2007); muscovite ⁄paragonite
(Coggon & Holland, 2002); biotite (White et al., 2007);
orthopyroxene (White et al., 2002); clinopyroxene
(Green et al., 2007); plagioclase ⁄K-feldspar (Holland
& Powell, 2003); garnet (White et al., 2007); cordierite
(Holland & Powell, 1998); chlorite (Mahar et al., 1997;
Holland et al., 1998); epidote (Holland & Powell,
1998); melt (White et al., 2007); ilmenite ⁄hematite
(White et al., 2000) and magnetite ⁄spinel (White et al.,
2002).
Migmatites P–T pseudosection
The XRF whole-rock analysis of sample F39 showed
that the K
2
O and the MnO constitute <1% of the
bulk rock (Table 3). Therefore, the P–T pseudosection
for this sample was calculated in the system
NCFMASHTO for the phases listed in the figure
caption of Fig. 6a and for quartz in excess. As there is
no melt model for this system, it is only meaningful to
interpret the sub-solidus parts of the P–T path for this
sample where water is likely to have been present.
Table 2. Representative mineral analyses and conventional thermobarometers of sample F93d; biotite (normalized to 11 O and
ignoring H
2
O), chlorite (normalized to 14 O and ignoring H
2
O), cordierite (normalized to 18 O and ignoring H
2
O), plagioclase
(normalized to 8 O) and garnet (normalized to 12 O).
Assemblage bi chl pl ru ilm H
2
O (latest post–assemblage) bi g pl ilm H
2
O (peak assemblage) bi cd pl ilm H
2
O
(post–peak assemblage)
Mineral bi11 bi20 chl4 chl5 chl31 chl33 g16 bi18 pl4 g17 bi17 pl5 cd3 cd8 bi3 bi13
SiO
2
35.33 25.20 25.76 25.34 28.13 27.93 37.37 35.46 59.93 38.05 36.02 63.68 48.91 47.47 34.63 34.13
TiO
2
1.78 2.14 0.38 0.38 0.12 0.26 0.17 2.49 b.d.l 0.07 1.61 0.09 b.d.l b.d.l 2.94 1.77
Al
2
O
3
18.39 19.27 21.45 21.50 19.85 18.43 20.42 18.85 23.13 20.69 19.69 21.19 32.59 32.21 18.5 18.67
Cr
2
O
3
0.07 0.17 0.10 0.10 0.01 0.12 0.02 0.07 0.06 0.08 0.17 b.d.l b.d.l 0.07 0.26 0.04
FeO 19.20 18.64 20.35 20.17 21.96 21.92 30.07 19.36 0.94 30.08 19.20 0.59 5.79 5.92 18.12 19.19
MnO 0.16 0.16 0.11 0.11 0.23 0.23 3.14 0.13 0.03 3.32 0.16 0.05 0.29 0.11 0.22 0.48
MgO 12.21 11.18 17.99 17.83 18.60 18.54 6.29 10.32 1.66 6.08 10.68 0.46 9.7 9.38 11.41 12.21
CaO 0.12 b.d.l b.d.l b.d.l 0.01 0.09 1.30 0.12 4.42 1.43 0.13 7.7 0.06 0.04 0.13 0.16
Na
2
O 0.41 0.42 b.d.l b.d.l b.d.l b.d.l b.d.l 0.41 9.17 b.d.l 0.42 7.20 b.d.l b.d.l 0.41 0.44
K
2
O 8.71 9.90 0.05 0.04 0.07 0.15 0.05 9.55 0.12 b.d.l 8.57 0.18 0.03 b.d.l 9.12 7.97
O 111114141414121181211818181111
Si 2.64 2.62 2.68 2.66 2.85 2.88 2.95 2.65 2.66 2.98 2.67 2.78 5.03 4.99 2.62 2.60
Ti 0.1 0.12 0.03 0.03 0.01 0.02 0.01 0.14 – 0.00 0.09 0.00 – – 0.16 0.10
Al 1.62 1.69 2.63 2.66 2.37 2.24 1.9 1.66 1.21 1.91 1.72 1.09 3.95 3.99 1.65 1.67
Cr 0.00 0.01 0.01 0.01 0.00 0.01 0.00 0.00 0.00 0.01 0.01 – – 0.01 0.01 0.002
Fe
+3
0.00 0.00 0.00 0.00 0.00 0.00 0.13 0.00 0.07 0.1 0.00 0.04 0.00 0.00 0.00 0.00
Fe
+2
1.2 1.16 1.77 1.77 1.86 1.89 1.9 1.21 0.00 1.9 1.19 0.00 0.49 0.52 1.14 1.22
Mn 0.01 0.01 0.01 0.01 0.02 0.02 0.21 0.01 0.00 0.22 0.01 0.00 0.03 0.01 0.01 0.03
Mg 1.36 1.24 2.79 2.79 2.81 2.85 0.75 1.15 0.11 0.73 1.18 0.03 1.49 1.47 1.28 1.38
Ca 0.01 – – – 0.00 0.01 0.11 0.01 0.21 0.12 0.01 0.36 0.01 0.01 0.01 0.01
Na 0.06 0.06 – ––––0.06 0.79 – 0.06 0.61 – – 0.06 0.06
K 0.83 0.94 0.01 0.01 0.01 0.02 0.01 0.91 0.01 – 0.81 0.01 0.00 – 0.88 0.77
X
Febi
0.467 0.483 0.471 0.468
X
Fechl
0.388 0.388 0.397 0.399
X
Feg
0.711 0.722
Z(g) 0.040 0.044
Y(bi) 0.26 0.31 0.31 0.39 0.27 0.29
Pkbar (Hoisch, 1990) 8.1 6.2
TC (Bhattacharya et al., 1992) 706 675
TC (Ganguly & Saxena, 1984) 723 697
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 449
2009 Blackwell Publishing Ltd

Thus, for Fig. 6a, water was chosen to be in excess. As
a reference for the onset of melting, we have super-
imposed the initial melting reactions in the system
NCKFMASHTO on Fig. 6a keeping in mind that all
equilibria at temperatures above this line must be
interpreted with care (as the water saturated assump-
tion does not hold there).
The P–T pseudosection for sample F39 is charac-
terized by a series of mineral assemblage fields with
steep boundaries (Fig. 6a). Only one small divariant
field (hb+act+di+pl+ab+ep+sph) appears at 7.4–
8.3 kbar and 520–535 C. The hb+di+pl+sph
pentivariant field is the largest field and is stable at
4.3–10 kbar and 525–795 C. Magnetite-in reactions
appear between 3 kbar – 485 C and 9 kbar – 800 C.
The proportion of magnetite increases with increasing
temperature. The amphibolite to granulite facies
boundary identified by the first appearance of ortho-
pyroxene, occurs at 760 C. The melting reactions
appear as near isothermal reactions between (625 C–
10 kbar) and (675 C – 3 kbar). The stability fields of
albite appear at low temperature conditions (430 C–
3 kbar) while this temperature increases with the
increase of pressure (560 C – 10 kbar).
The migmatite peak assemblage of hb+di+pl+
mt+sph appears on Fig. 6a in a quadrivariant field in
the region (580–760 C at 3 kbar), and narrows to
higher pressure where it terminates at 785 Cat
8.65 kbar. Within this field, the peak conditions can be
further constrained using the X
Fehb
and the Y(hb)
isopleths. However, much of this field is within the
partial melting region for which Fig. 6a is not well
constrained. Thus, while isopleths indicate peak con-
ditions between 5.3 and 7.8 kbar we suggest that these
cannot be used for further interpretation. A lower limit
of the temperature at the metamorphic peak is identi-
fied as the first appearance of the melt at 660 C while
an upper temperature limit is identified by the absence
of orthopyroxene, which appears above 775 C.
Conditions for the pre-peak assemblage hb+di+
pl+sph are mainly defined on the basis of the different
chemistry of texturally earlier amphibole and plagio-
clase crystals. The estimated temperature for these
minerals is 555–655 C. Unfortunately, because of the
parallelism of the mineral isopleths (Fig. 6a), there are
no pressure constraints. The post-peak conditions are
well preserved by the presence of the hb+act+
pl+ep+sph assemblage. The hb+act+pl+ep+sph
assemblage was contoured for the X
Feact
and the X(an)
isopleths (Table 1), to constrain the retrograde pressure
and the temperature conditions. The isopleths reveal
that this mineral assemblage was stable at temperature
range of (455–520 C). Because of the presence of the
albite, the migmatites were re-equilibrated at low tem-
perature (<450 C). The pressure range at the low
temperature condition cannot be closely constrained,
but the actinolite chemistry (Fig. 5d) is consistent with
cooling most probably <5 kbar.
Hornblende gneiss P–T pseudosections
The P–T pseudosection for the hornblende gneisses
(samples F1 & F63) was calculated in the system
NCFMASHTO (Fig. 6b) for a bulk composition from
XRF neglecting 0.95 wt% K
2
O (Table 3). Calculations
were performed assuming water to be in excess which is
likely to be justified in the light of the fact that these
rocks show no evidence for partial melting.
The P–T pseudosection is characterized by a series
of mineral assemblage fields with steep boundaries,
except the garnet-in reactions (Fig. 6b). The garnet-
bearing assemblages appear at high temperature and
high pressure conditions (>600 C and >8.8 kbar).
Only one and small isothermal divariant field
(hb+chl+pl+ab+ep+sph+ilm) occurs at 3–8 kbar
and 457–520 C. All the assemblages below 565 C–
3 kbar and 620 C – 10 kbar are free of magnetite. The
first appearance of magnetite is in the trivariant
hb+chl+pl+g+ilm+mt and in the quadrivariant
hb+chl+pl+ilm+mt fields. The largest field in the
P–T pseudosection is the pentivariant field hb+pl+
ilm+mt. This assemblage is stable at the range of 572–
755 C at 3 kbar, 625 C – 8.8 kbar and 791 C–
9.65 kbar. The albite-bearing assemblages appear at
low temperature conditions 460 C – 3 kbar and
550 C – 10 kbar.
Table 3. Representative XRF analyses of different rock groups; major elements in wt%, trace elements (ppm).
Group Sample LOI SiO
2
Al
2
O
3
Fe
2
O
3tot
MnO MgO CaO Na
2
OK
2
O TiO
2
P
2
O
5
Sum Ba Ce Cr Rb Sr) V Y Zn Zr
Hornblende gneisses F59 1.39 67.16 15.47 4.45 0.08 1.84 3.25 4.46 1.43 0.71 0.22 100.60 354 33 <20 39 483 86 22 86 160
F1 0.87 64.67 14.43 6.21 0.13 2.96 3.44 4.93 0.95 0.85 0.21 99.81 465 42 97 38 332 128 27 94 171
F63f 1.20 61.81 16.78 6.54 0.11 2.13 4.50 5.34 0.93 0.95 0.31 100.73 188 77 <20 <15 526 129 39 66 208
F4 0.54 62.47 15.34 6.68 0.15 3.07 5.00 4.96 0.94 0.97 0.22 100.47 253 54 105 <15 370 146 34 86 182
F7 0.27 66.93 15.11 6.57 0.07 1.80 2.40 4.71 1.70 0.89 0.17 100.81 679 57 61 39 377 116 32 59 227
Quartzofeldspathic gneisses F77 0.56 77.58 12.10 1.22 0.06 0.47 1.39 1.52 5.73 0.13 b.d.l 100.93 839 27 <20 195 31 <20 <20 48 108
F77a 0.65 81.62 10.21 1.12 0.13 0.43 2.20 2.30 1.77 0.11 b.d.l 100.67 720 33 <20 61 34 <20 24 24 103
F35 0.58 64.37 13.39 3.69 0.67 2.17 7.79 2.63 4.75 0.49 0.20 100.92 773 68 22 138 140 31 32 377 252
Calcsilicate F51b 13.18 40.59 4.20 0.93 0.22 1.36 37.23 1.26 0.64 0.17 0.07 99.90 115 <30 25 <15 371 50 <20 26 49
F33 1.33 53.83 9.91 4.09 0.28 2.07 23.87 2.63 1.92 0.40 0.16 100.58 349 43 30 42 133 66 26 205 108
Migmatite F39 0.75 59.91 14.59 6.20 0.24 4.08 8.13 4.21 0.73 0.78 0.21 99.96 167 60 96 <15 409 118 38 91 166
Metagraywackes F93d 1.57 60.82 14.60 8.91 0.11 5.00 2.01 2.34 3.15 0.97 0.15 99.83 483 84 429 90 162 143 25 96 202
F93a 1.00 46.95 2.17 34.93 0.07 2.21 7.44 b.d.l b.d.l 0.15 5.17 100.22 24 <30 26 <15 544 36 <20 61 88
F93c 1.00 47.72 4.26 27.53 0.09 2.31 10.05 0.25 b.d.l 0.22 6.83 100.41 83 <30 37 <15 574 43 <20 53 170
F94a 1.50 55.26 20.64 6.25 0.10 2.68 7.79 3.03 1.54 0.87 0.22 100.05 433 22 31 64 661 137 28 75 134
450 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

hb act pl sph
400 450 500 550 600 650 700 750
3
4
5
6
7
8
9
hb gl ab
ep sph
hb ab
ep sph
hb di ab
ep sph hb di pl
ab ep sph
hb di pl
ep sph
hb act ab
ep sph
hb act pl
ab ep sph
hb act pl
ep sph
hb act pl
mt sph
hb act di
ab ep sph
hb act di
pl ep sph
0.28
0.26
0.22
0.2
hb act di
pl mt sph
hb act di
pl sph
0.24
P
kbar
P
kbar
T (°C)
400 450 500 550 600 650 700 750
3
4
5
6
7
8
9
hb pl ilm mt
hb chl pl
g ilm
chl ab ilm
ep sp
h
ch l
ep sp h
ab
chl pl ab ilm ep sp
h
chl pl
ab ep
sp h
Hb act
chl pl
ilm
hb act chl pl
ab ep sph
hb chl pl ab
ep sph ilm
hb chl ab
ep sph ilm
hb chl
ab ep
ilm
hb chl pl
ab ep ilm
hb chl pl
ep ilm
hb chl pl
ep sph ilm
NCFMASHT O system (+ q, H2O)
hb chl pl ilm mt
Hb pl
ilm
opx
mt
hb pl g ilm mt
hb pl
ilm
opx
gm
t
hb chl pl
g ilm mt
hb chl pl ilm
opx out
g out
act out
hb act ch
l
pl sph ilm
T (°C)
NCFMASHT O system (+ q, H2O)
Hornblende gneisses
(a)
(b)
ab out
0. 3
0.29
0.41
act chl pl ab ep sp
h
Hb chl pl ab ilm sp h
hb chl pl sph ilm
hb chl pl ab ep sp
h
0.473
0.475
0.477
0.479
0.062
0.064
0.066
X
Fe
hb
Y (hb)
Migmatites
hb di pl
mt sph
hb di opx
pl sph
0.14
0.16
0.18
0.24
0.2
0.22
0.38
0.12
0.44
0.43
0.42
0.39
0.4
0.41
X
Fe
hb
Y (hb)
X (an)
Y (act)
X
Fe
di
Transparen tr egio no f
pseudosection metastable
with respect to melt
hb di pl sph
Fig. 6. P–T pseudosections for migmatites
and hornblende gneisses. For both rock
types, the oxidation state was calculated
using Le Maitre (1976) equation for the
partitioning of FeO and Fe
2
O
3
from
Fe
2
O
3tot
. (a) P–T pseudosection of migma-
tite sample F39. The bulk composition is
(mol.%): SiO
2
: 64.12, Al
2
O
3
: 9.20, CaO:
9.38, MgO: 6.55, FeO: 5.41, Na
2
O: 4.38,
TiO
2
: 0.63, O: 0.32. The thick dashed line
shows the first appearance of the melt and
was calculatedin the system NCKFMASHTO
using K
2
O: 0.73%. As this line was not
calculated in the system for which the
pseudosection is drawn, the region above
it cannot be interpreted (shaded region).
(b) P–T pseudosection of average composi-
tion of samples F1 and F63. The bulk com-
position is (mol. %) SiO
2
: 69.56, Al
2
O
3
: 9.98,
CaO: 4.36, MgO: 4.22, FeO: 5.34, Na
2
O:
5.53, TiO
2
: 0.75, O: 0.27. The vertical and
the horizontal bars are the average Pand T
according to Fig. 5. The white polygons
indicate conditions constrained by mineral
isopleths.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 451
2009 Blackwell Publishing Ltd

The absence of orthopyroxene and garnet in the
peak assemblage hb+pl+ilm+mt constrains the peak
conditions to below 760 C and 9 kbar. X
Fehb
and
Y(hb) contours (Table 1) were added to the peak
assemblage field in order to constrain the P–T condi-
tions of the equilibrium assemblage in more detail.
These conditions are agreement with the metamorphic
conditions derived from the relationship between
Al
VI
and the Si and the graphical geothermometer
(Fig. 5e,f). The post-peak conditions are difficult to
determine. Some constraints however are provided by
the late stage minerals chlorite, albite, actinolite
and epidote. The presence of these phases implies that
the hornblende gneisses re-equilibrated below the
albite-in boundary and above the actinolite-out
boundary (465–480 C and <<5 kbar according to
Fig. 4d,e).
Metagraywacke P–T pseudosection
In contrast to the migmatites and the hornblende
gneisses, the metagraywackes contain a series of Fe-
bearing phases (garnet, cordierite and biotite). For the
metagraywackes, we did not rely on Le Maitre (1976)
equation to constrain the oxidation state, but con-
strained it in more detail using a T–Xo pseudosection.
Indeed, the mineralogy of this sample does allow a
much closer constraint using this method than the
mineralogy of the other rock types would. A T–Xo
pseudosection (Fig. 7a) of the metagraywackes (sam-
ple F93d) was constructed in the system
NCKFMASHTO at 8 kbar (the pressure given by
biotite–plagioclase–garnet equilibria, Table 2). The
loss of ignition (LOI = 1.57) was used to evaluate the
H
2
O content of this sample. Z(g) isopleths and liquid
modes were added to the peak assemblage bi+g+
pl+ilm+liq in order to constrain the correct T–Xo
conditions. The Z(g) isopleths (0.04–0.044) intersected
with 0.05 liquid mode below 735 C and in a range of
Xo of 0.09–0.17. From the T–Xo pseudosection, the
best Xo estimate is 0.13, which is equivalent to an
oxidation state with a free oxygen content of 0.43.
Using this bulk composition, a P–T pseudosection
of the metagraywackes was constructed in the system
NCKFMASHTO. This pseudosection shows that
gedrite is stable over a wide P–T range – which is not
observed in the metagraywackes of the FSMC. We
suggest that this can be explained through a fraction-
ation in bulk composition during the retrograde path
(Stu
¨we, 1997): garnet crystals are only zoned along
20% of their margins (Fig. 5g) indicating that most
of their volume was effectively removed from the bulk
composition at the peak. We have therefore used the
bulk composition described above only to interpret
the peak conditions (Fig. 7b). For the interpretation of
the retrograde evolution a new effective bulk compo-
sition was calculated by removing 80% of the garnet
from the bulk of Fig. 7b (Fig. 7c).
Both pseudosections (Fig. 7b,c) are characterized by
the absence of univariant and divariant assemblages.
There is one trivariant field; the bi+hl+pl+ep+
ab+ru+ilm+H
2
O at low temperature conditions.
The garnet-in reactions appear at high pressure con-
ditions (>6.5 kbar) in the temperature range 525–
720 C. Gedrite-bearing fields appear in the pressure
range 5–9.6 kbar and temperature range 600–790 C.
The cordierite-bearing assemblages appear at low
pressure conditions and high temperature (>600 C)
conditions. The orthopyroxene-in reactions appear as
isothermal boundaries at 765 C. At high pressure
conditions (>6.5 kbar), the orthopyroxene-in reac-
tions appears at >800 C. The melting reactions
appear as isothermal reactions (700–720 C). The
chlorite-bearing fields are stable at different pressure
conditions but at low temperature (<590 C). On the
other hand, albite is stable at the same pressure as
chlorite but at much lower temperature (<410–
530 C).
The metamorphic peak is constrained by the
assemblage bi+g+pl+ilm+liq on Fig. 7b. The
X
Feg
,Z(g) isopleths and liquid modes were added to
the field of the peak assemblage to determine the
peak conditions. The Z(g) is 0.04–0.0445 implying
that the peak pressure is 7.7–8.2 kbar. The X
Feg
ranges between 0.71 and 0.72, this indicates that
sample F93d reached peak conditions at 690–715 C.
This is consistent with P–T estimates obtained from
conventional thermobarometers as discussed above.
The intersections of the X
Feg
and Z(g) isopleths
support the assumption that the liquid mode of the
rock is not more than 5%. The presence of cordierite
and the absence of orthopyroxene and gedrite indi-
cate that sample F93d re-equilibrated at low pres-
sure–high temperature conditions (<6 kbar, 600–
760 C). Unfortunately, mineral isopleths of the
assemblage bi+cd+pl+ilm+H
2
O are more or less
isothermal which means that the temperature can be
determined. Y(bi) reveals that sample F93d re-equili-
brated at a temperature range 650–700 C. Low
temperature retrograde conditions could be identified by
the presence of late chlorite and rutile and the absence of
epidote and albite. The compositions of the late chlorite
and late biotite constrains the retrograde P–T condi-
tions to 3.5–4.8 kbar and 465–500 C (Fig. 7c).
DISCUSSION
The pseudosections discussed above constrain different
metamorphic conditions in different rock types (Figs 6
& 7). In this section, we discuss if these different
conditions reflect separate events or whether they can
be connected to a path for a single orogenic cycle. This
path is then related to the deformation history and an
integrated P–T–D evolution for the FSMC is derived.
It is also discussed as to why different rock types
record different stages of the evolution.
452 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

ged bi g
pl ilm liq
ged bi g
pl ilm
H2O
bi g pl
ilm
H2O
ged bi pl
ilm
H2O
ged bi chl
pl ilm H2O
ged bi chl g
pl ilm H2O
bi chl pl
ilm
H2O
bi chl ep ab
ru ilm
H2O
bi chl pl ep
ru ilm
H
2
O
bi chl g pl
ep ilm H2O
bi chl g pl
ilm
H
2
O
ged bi cd opx
pl ilm liq
bi cd pl
ilm liq
bi cd pl ilm
H2O
ged bi pl
ilm H O
2
bi cd opx pl ilm liq
ged bi cd
pl ilm liq
bi pl ru ilm
H
2
O
ged bi pl ilm liq
bi pl ilm
H
2
O
bi chl pl ru ilm
H2O
bi chl ep ab
ru ilm
H2O
bi chl pl ep
ru ilm H
2
O
0.28
0.29
0.3
400 450 500 550 600 650 700 750
3
4
5
6
7
8
9
NCKFMASHT O system (+ q)
P
kbar
T (°C)
Metagraywackes
(b)
600
650
700
750
NCKFMASHT O system (+ q) at
kb
ar P8
X o
T (°C)
0.06 0.16 0.26 0.36
Metagraywackes
ged bi ilm
chl pl
H2O
ged bi ilm pl
H2O
ged bi g
ilm
pl
H2O
ged bi g
ilm
pl
liq ged bi ilm pl liq
bi g ilm pl liq
0.045 0.039
0.041
0.042
0.043
0.05
ged bi ilm liq pl H O
2
Z(g)
Liq mode
0.04
0.044
bi ilm chl pl
H
2
O
(a)
(c)
ged bi
pl ilm liq
bi g pl
ilm liq
0.04 0.05 0.06
0.7 0.71 0.72
0.039
0.041
0.043
0.045
Z (g)
liq mode
X
Fe
g
ged bi cd
pl ilm H O
2
X
Fe
chl
X
Fe
bi
Y (bi)
0.48
0.478
0.482
0.396
0.398
0.394
bi chl pl ep
ilm
H
2
O
bi chl pl ru
ilm H
2
O
0.045
Fig. 7. P–T and T–Xo pseudosections for
the metagraywackes (sample F93d).
(a) T–Xo pseudosection. The bulk com-
position is that derived from XRF analysis
(mol.%): SiO
2
: 66.41, Al
2
O
3
: 9.4, CaO: 2.35,
MgO: 8.14, FeO: 8.14, Na
2
O: 2.48, K
2
O:
2.19, TiO
2
: 0.79, H
2
O: 5.41 (Table 3). (b)
P–T pseudosection at high pressure condi-
tions using the bulk composition derived
from (a) with Xo: 0.13. (c) P–T pseudosec-
tion at low pressure conditions using the
bulk composition of (b) from which 80% of
the garnet was removed (mol.%; SiO
2
: 64.35,
Al
2
O
3
: 8.48, CaO: 2.25, MgO: 7.5, FeO:
5.04, Na
2
O: 2.53, K
2
O: 2.23, TiO
2
: 0.81,
H
2
O: 5.51). The vertical and the horizontal
bars are the average Pand Taccording to
the garnet–biotite–plagioclase thermo-
barometers (Table 2) and the garnet–
cordierite thermometer.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 453
2009 Blackwell Publishing Ltd

P–T path and its relationship to deformation
In summary from above, pre-metamorphic peak
conditions are constrained by the chemistry of
clinopyroxene and plagioclase (and the absence of
magnetite) in the migmatites. These pre-peak condi-
tions are 550–620 C and 7.2–8.2 kbar. Peak meta-
morphic conditions derived for the metagraywackes
are 690–715 C and 7.7–8.2 kbar (Fig. 7). In the
migmatites, the peak is constrained at a similar pres-
sure, but temperatures 660–775 C (Fig. 6a). The
somewhat higher peak temperatures may simply reflect
that the migmatites (located in the core of the FSMC)
were located at slightly deeper crustal levels at the time
of peak metamorphism. However, it is also important
to note that the peak conditions of the migmatites were
derived without considering a melt model for this bulk
composition so that they are subject to some uncertainly.
Peak assemblages in both the migmatites and the
metagraywackes are overprinted by later parageneses
that equilibrated at 645–700 C and 4.7–5.3 kbar.
Finally, all studied samples give petrographical and
mineral chemical indications for partial re-equilibra-
tion at low pressure–low temperature conditions of
460–530 C and 4–5 kbar.
Interestingly, the metamorphic peak recorded by
the migmatites and metagraywackes is not preserved
in the hornblende gneisses. The coarse-grained peak
assemblage in the hornblende gneisses (hb+pl+
ilm+mt) formed at 630–770 C and <5.3 kbar.
These conditions correspond to the post-peak
assemblage in the other rock types. The final isobaric
cooling is recorded in all rock types. We suggest that
the migmatites and hornblende gneisses record dif-
ferent stages of the P–T path and interpret this as a
reflection of the water content of the rocks: in the
migmatites, water was partitioned into the melt at
the metamorphic peak, causing a relative dry residue
assemblage and thus preservation of the peak
assemblage (e.g. White & Powell, 2002). Conversely,
the paragneisses around the granitic dykes are en-
riched in K-feldspar which indicate metasomatic
infiltration. In addition all the metamorphic rocks
around the migmatites were affected by complex
fluid-infiltration processes after the peak metamor-
phism (Abu-Alam & Stu
¨we, Unpublished data). We
interpret therefore that the hornblende gneisses had
sufficient water during their entire evolution so that
later equilibration along the same P–T path of the
migmatites and metagraywackes was possible (e.g.
Guiraud et al., 2001).
The different metamorphic conditions described
here were interpreted by El-Shafei & Kusky (2003) as
evidence for two discrete metamorphic events: a
medium grade event M
1
(recorded by the hornblende
gneisses) and a high-grade M
2
event associated with
the formation of migmatites. However, Fowler &
Hassan (2008) suggested only one metamorphic event.
Textural, microstructural and thermobarometric evi-
dence presented here is agreement with Fowler &
Hassan (2008) and suggests that minor grade varia-
tions can be interpreted in terms of structural level and
different rock types. We infer therefore that the pre-
peak-, peak- and retrograde conditions reflect the
partial record of a single continuous ÔclockwiseÕP–T
path. This interpretation is consistent with geochro-
nological and tectonic evidence from elsewhere in the
Arabian-Nubian shield (Abu El-Enen et al., 2004;
Eliwa et al., 2008). Then, the inferred P–T path is
characterized by heating with moderate burial at mid
crustal levels (for an overburden density 2850 kg m
)3
and assuming lithostatic conditions the peak meta-
morphic conditions corresponds to a depth of
25.7–29.3 km) followed by isothermal decompression
because of rapid exhumation to 5 kbar and final
cooling to stable geothermal conditions at this depth
(14.3–17.9 km). The final isobaric cooling path is
consistent with observations in Taba metamorphic
complex (the nearest high-grade metamorphic com-
plex) where final cooling is also near isobaric at 4.5–
5 kbar (Abu El-Enen et al., 2004). Exhumation from
this depth (14.3–17.9 km) to the surface is not
recorded by the metamorphic parageneses and will be
interpreted below.
The P–T evolution can now be correlated with the
four deformation phases that affected the Feiran–Solaf
region (Fig. 8). All peak parageneses grew in the S
1
foliation as well as the presence of leucosomes which
are parallel and cross-cutting to the principal foliation
suggesting that peak metamorphism was syn- or just
post-D
1
(Fig. 8). This interpretation is in agreement
with the observations of Fowler & Hassan (2008) in the
Feiran complex, but we note that it is different from
other metamorphic complexes in Sinai: in Taba and
Elat metamorphic complexes, Cosca et al. (1999)
suggested that the peak metamorphism was associated
with three ductile deformation phases D
1
–D
3
, while
Abu El-Enen et al. (2004) suggested that the peak
metamorphism associated with D
2
.
However, we note that D
2
in Taba may correspond
to D
1
in the FSMC as we have found sparse evidence
for a deformation phase prior to peak metamorphism.
Microstructural evidence and field evidence shows that
D
2
and D
3
are compressive deformation phases. Thus,
these deformation phases must have occurred during
the isothermal decompression and ⁄or subsequent
cooling. As D
2
and D
3
structures are in places cut by
leucosomes, these phases are likely to have occurred
during the decompression phase above the solidus
(Fig. 8). In the Taba, Kid and Elat metamorphic
complexes, the D
2
and D
3
phases are also recorded as
ductile phases (e.g. Cosca et al., 1999; Abu El-Enen
et al., 2003, 2004; Eliwa et al., 2008) but their
relationship to the decompression path has not been
documented. On D
4
there are no direct constraints. We
suggest that it either occurred during cooling or during
the exhumation post-date to the 4 kbar cooling docu-
mented here.
454 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

The interpreted P–T–D evolution raises a series of
problems of both conceptual and regional relevance
that are now discussed. In particular, we discuss (i) the
counter-intuitive synchronous occurrence of isother-
mal decompression (recorded by the parageneses and
suggested exhumation) and convergent deformation
phases (recorded by field evidence and suggested bur-
ial) and (ii) the tectonic relevance of the documented
evolution.
Exhumation during compressive deformation
The observation of synchronous decompression and
convergent deformation phases is not unique to the
FSMC. Carson et al. (1997) observed a similar rela-
tionship in the Larsemann Hills in Antarctica and in
several other high-grade complexes convergence is
observed synchronously with the decompression phase
of the P–T path. For the Larsemann Hills, Carson
et al. (1997) suggested that it reflects rapid upper
crustal extension at the time of lower crustal conver-
gence. However, for the FSMC the model of Carson
et al. (1997) is unsatisfactory, as the region is a rela-
tively small area for which orogen scale considerations
may not apply. In a more simple, one-dimensional
interpretation, Stu
¨we & Barr (1998) showed that
simultaneous erosion and convergence will result in a
deformation field characterized by decompression
during convergence above a critical point in the crust
and burial below that. This critical point separates an
upper part of the crust where upwards advection by
erosion outweighs convergence from a lower part
where deformation outweighs erosion. The depth of
this critical point depends on the relative rates of
erosion and deformation but can be as deep as
30 km. Because the FSMC records decompression
from 29 km, this model may, in principal, be a pos-
sible explanation. However, here we suggest that the
synchronous occurrence of convergent deformation
and decompression is simply a reflection of a compli-
cated obliquely convergent geometry of the D
2
and D
3
events which caused exhumation in a wrench system.
Within this wrench system overall convergence was
accompanied by extrusion of the complex to the sur-
face. Such a wrench system is well known from the
Pan-African basement in Sinai and the Eastern Desert:
the Najd fault system.
Tectonic implications in the context of the Najd fault system
The lithosphere-scale Najd fault system (Fig. 1) is
known to be related to the exhumation history of a
series of high-grade metamorphic terranes throughout
the Eastern Desert of Egypt and Sinai that are of
similar age and grade to the FSMC (615–610 Ma;
Stern & Manton, 1987): Pan-African metamorphic
areas are known for example in Wadi Hafafit (620–
580 Ma, Abd El-Naby et al., 2008; Loizenbauer et al.,
2001; Stern & Hedge, 1985; Hashad et al., 1981), Taba
complex (605 Ma, Eliwa et al., 2008), Meatiq meta-
morphic core complex (640–580 Ma, Fritz et al., 1996,
2002; Loizenbauer et al., 2001; Stern & Hedge, 1985)
and Sibai dome (614 Ma, Fritz et al., 1996, 2002). All
of these regions are characterized by low pressure–high
temperature peak conditions 6–8.5 kbar and 600–
750 C comparable with those described here (e.g.
Loizenbauer et al., 2001; Abd El-Naby et al., 2008;
Eliwa et al., 2008). Interestingly, most of the terranes
in the Eastern Desert were exhumed as core complexes
in extensional environments, while we interpret the
FSMC to have exhumed in an overall regime of
compression.
Here, it is argued that both mechanisms are possible
within the overall setting of the Najd fault system: this
fault system has a general NW–SE orientation (with
substantial local variation, see e.g. Shalaby et al., 2006)
and a sinistral motion, so that the maximum com-
pressive stress is oriented roughly W–E. We suggest
that minor variation in the orientation of the various
metamorphic complexes with respect to the Najd fault
system boundaries cause the different exhumation
mechanisms: in the Eastern Desert, the intermediate
principal stress may be oriented vertically causing
overall extension, while in the FSMC the minimum
principal stress is vertical causing exhumation in an
overall horizontal transpressive regime (Fig. 1). We
therefore suggest that the Najd fault system (D
3
in the
FSMC) is the principal phase causing the exhumation
of the complex in an oblique transpressive regime
(Fig. 9). Within this setting, major volumes of post-
tectonic granites that surround the complex were
produced by decompression-enhanced melting and are
transported through the Najd fault system, thereby
400 450 500 550 600 650 700 750
3
4
5
6
7
8
9
P
(kbar)
Depth
(km)
T (°C)
D1deformation phase and foliation.S1
DD D D
F
D
23 4 3
3
3
, (compressional phases) and (open warping event),
formed open folds ( ) causing the map scale structure. At the end
of , the study area affected by the Najd Fault system.
Metagraywackes
Migmatites
Hornblende gneisses
10.7
14.2
17.7
21.2
24.7
28.2
31.7
Migmatite solidous
(Fig. 5)
M
1
610 – 615 Ma
Fig. 8. Summary of the P–T paths of metagraywackes,
hornblende gneisses and migmatites. The thick arrows mark
the P–T paths. The lines and the polygons are the stability fields
of the mineral assemblages from Figs 6 & 7.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 455
2009 Blackwell Publishing Ltd

obliterating the margins of the complex. Our inter-
pretation is nicely consistent with interpretations from
the Eastern Desert where the D
3
phase is also inter-
preted as the Najd Fault system and related to ascent
of the post-tectonic granitic magma (e.g. Shalaby
et al., 2005; Farahat et al., 2007).
Final exhumation to the surface
An interesting and as yet unresolved aspect of the
tectonic evolution of the FSMC is its final exhumation
to the surface. This study has inferred a metamorphic
evolution that terminates at 450 C and 15 km
depth suggesting that the final exhumation must be
related to an independent event. Interestingly, this
depth is roughly the peak metamorphic depth of the
low-grade metamorphic belts in Sinai (Ayalon et al.,
1987; Eliwa et al., 2004) suggesting that these belts
were not affected by the Najd fault system. Zircon and
apatite fission track thermochronometers indicate that
the study area was subject to a complex history of
exhumation and reburial from the Cambrian to the
Early Tertiary period (e.g. Kohn et al., 1992). How-
ever, Vermeesch et al. (2009) suggested that the
calcalkaline rocks (at the same crustal level of the high-
grade gneisses) were exposed by c. 590 Ma suggesting
that the final exhumation occurred already at the end
of the Pan-African. In fact, in the FSMC, the presence
of an unconformity between gneisses and a Cretaceous
succession and the absence of the pre-Cretaceous
sedimentary cover (Fig. 2) suggest that the study area
was exhumed and subjected to intense pre-Cretaceous
erosion prior to being reburied under the Cretaceous
succession. This implies that the final exhumation of
the FSMC from 15 km occurred in an event fol-
lowing the Pan-African high-grade evolution. Such an
event has not been described, but there is sedimentary
evidence from the Hammamat molasse that subsidence
occurred and basins formed at c. 585 ± 15 Ma (Willis
et al., 1988). Within this event the lower-grade meta-
morphic belts of Sinai, like the SaÕal belt, may also
have exhumed.
CONCLUSIONS
In conclusion, the following tectonic evolution is
inferred from our study (Fig. 9). The rocks of the
Wadi Feiran belt include a series of metasedimentary
and metavolcanic rock types with minor mafic inter-
calations and calc-silicates that are consistent with an
interpretation as a sedimentary succession in a
marginal basin between Pan-African volcanic arcs
(El-Gaby & Ahmed, 1980). Vertical flattening (Fowler
& Hassan, 2008) was associated with horizontal
High grade
rocks (FSMB) Low grade
rocks
Granitic
intrusions
F2
DD
23
&
F3
(b)
(c)
(d)
Volcanic arc
Mozambique
Ocean D
1
(a)
Najd system
(Extensional phase)
(Transpressive exhumation)
(Obliteration of contacts)
(Final exhumation)
Fig. 9. Tectono-metamorphic model of Wadi Feiran Solaf metamorphic complex during the Pan-African event. (a) Igneous
activity formed volcanic arcs and related intrusive rocks (632 ± 3) Ma. Immature sediments were deposited in a marginal basin,
followed by D
1
deformation due to vertical flattening. (b) Due to the arc–arc accretions and the closing of the Mozambique Ocean,
the D
2
and D
3
compressional phases occurred. (c) At the end of D
3
, the study area was exhumed by oblique compressive deformation
associated with sinistral strike–slip shear zone to the depth of 14.3–17.9 km. (d) The present day situation. The solid circles are the
position of the FSMC during the evolution.
456 T. S. ABU-ALAM & K. STU
¨WE
2009 Blackwell Publishing Ltd

transport towards the west and north-west, forming
the penetrative S
1
foliation and causing burial of the
rocks. The rocks reached peak metamorphism 7.2–
8.2 kbar and 645–775 C at 610–615 Ma towards the
end of D
1
and prior to D
2
. The subsequent D
2
and D
3
deformation phases caused shortening in NE–SW
direction at the time of substantial near isothermal
decompression of the rocks. The D
3
, phase is corre-
lated with the sinistral NW–SE striking Najd fault
system and exhumed the complex in an oblique
transpressive regime. We suggest that the difference
between the transpressive exhumation interpreted
here and the extensional core complexes of the
Eastern Desert is a consequence of minor variation of
the orientation of the various belts with respect to the
Najd fault system. Because of the rapid exhumation,
large volumes of post-tectonic granitic magma was
formed and transported along the Najd fault system
into the complex thereby obliterating most of its
margins. The doming up of the entire complex and
the NE–SW F
4
were formed by the open warping
event (D
4
) which was possibly related to the final
exhumation from 15 km depth to the surface during
an independent event.
ACKNOWLEDGEMENTS
This project was supported by the Austria exchange
service (O
¨AD) scholarship. We thank F. Makroum,
A. Shalaby, M. El-Shafei and Y.M. Sultan for their
help during the field work. Help with the pseudosec-
tions by V. Tenczer is appreciated. C. Hauzenberger is
thanked for his help with the XRF analysis. R.J. Stern
is thanked for his help with inaccessible papers.
S. Boger and M. Abu El-Enen are thanked for their
constructive reviews. R. White is appreciated for his
constructive suggestions and efficient editorial
handling of the manuscript.
REFERENCES
Abd El-Naby, H., Frisch, W. & Siebel, W., 2008. Tectono-
metamorphic evolution of the Wadi Hafafit Culmination
(central Eastern Desert, Egypt). Implication for Neoprotero-
zoic core complex exhumation in NE Africa. Geologica Acta,
6, 293–312.
Abdelsalam, M. G. & Stern, R. J., 1996. Sutures and shear zones
in the Arabian-Nubian Shield. Journal of African Earth
Sciences,23, 289–310.
Abu El-Enen, M. M., Okrusch, M. & Will, T. M., 2003.
Metapelitic assemblages in the Umm Zariq schists, central
western Kid Belt, Sinai Peninsula, Egypt. Neues Jahrbuch
Mineralogie Abhandlungen,178, 277–306.
Abu El-Enen, M. M., Will, T. M. & Okrusch, M., 2004. P–T
evolution of the Pan-African Taba metamorphic belt, Sinai,
Egypt: constraints from metapelitic mineral assemblages.
Journal of African Earth Sciences,38, 59–78.
Ahmed, A. A., 1981. Reconsidered view on Feiran–Solaf
Gneisses southwest of Sinai, Egypt. Bulletin Faculty of
Sciences, Assiut University,10, 131–142.
Ahmed, A. A. & Youssef, M. M., 1976. Air photo interpretation
of dike swarms in the area around Feiran Oasis, southwest
Sinai, Egypt. Bulletin Faculty of Sciences, Assiut University,1,
1–17.
Ayalon, A., Steinitz, G. & Starinsky, A., 1987. K–Ar and Rb–Sr
whole-rock ages reset during Pan-African event in the
Sinai Peninsula (Ataqa Area). Precambrian Research,37, 191–
197.
Bhattacharya, A., Mohanty, L., Maji, A., Sen, S. K. & Raith,
M., 1992. Non-ideal mixing in the phlogopite–annite binary:
constraints from experimental data on Mg–Fe partitioning
and a reformulation of the biotite–garnet geothermometer.
Contributions to Mineralogy and Petrology,111, 87–93.
Blasband, B., Brooijmans, P., Dirks, P., Visser, W. & White, S.,
1997. A Pan-African core complex in the Sinai, Egypt. Geol-
ogie en Mijnbouw,76, 247–266.
Blasband, B., White, S., Brooijmans, P., De Brooder, H. &
Visser, W., 2000. Late Proterozoic extensional collapse in
Arabian-Nubian Shield. Journal of the Geological Society,
London,157, 615–628.
Blundy, J. & Holland, T., 1990. Calcic amphibole equilibria and
a new amphibole plagioclase geothermometer. Contributions
to Mineralogy and Petrology,104, 208–224.
Brooijmans, P., Blasband, B., White, S. H., Visser, W. J. &
Dirks, P., 2003. Geothermobarometric evidence for a meta-
morphic core complex in Sinai, Egypt. Precambrian Research,
123, 249–268.
Carson, C. J., Powell, R., Wilson, C. J. L. & Dirks, P. H. M. G.,
1997. Partial melting during tectonic exhumation of a granu-
lite terrane: an example from the Larsemann Hills, East
Antractica. Journal of Metamorphic Geology,15, 105–126.
Coggon, R. & Holland, T. J. B., 2002. Mixing properties of
phengitic micas and revised garnet–phengite thermobaro-
meters. Journal of Metamorphic Geology,20, 683–696.
Cosca, M. A., Shimron, A. & Caby, R., 1999. Late Precambrian
metamorphism and cooling in the Arabian-Nubian Shield:
petrology and
40
Ar ⁄
39
Ar geochronology of metamorphic
rocks of the Elat area (southern Israel). Precambrian Research,
98, 107–127.
Dachs, E., 2004. PET: petrological elementary tools for math-
ematica: an update. Computers and Geosciences,30, 173–182.
Diener, J. F. A., Powell, R., White, R. W. & Holland, T. J. B.,
2007. A new thermodynamic model for clino- and ortho-
amphiboles in the system Na
2
O–CaO–FeO–MgO–Al
2
O
3
–
SiO
2
–H
2
O–O. Journal of Metamorphic Geology,25, 631–656.
El-Gaby, S. & Ahmed, A. A., 1980. The Feiran–Solaf gneiss belt,
SW of Sinai, Egypt. In: Evolution and Mineralization of the
Arabian-Nubian Shield (eds Coory, P.G. & Tahoun, S.A.).
Institute of Applied Geology,4, 95–105, Jeddah.
El-Gaby, S., List, F. K. & Tehrani, R., 1990. The basement
complex of the Eastern Desert and Sinai. In: The Geology of
Egypt (ed. Rushdi, S.), pp. 175–184. Balkema, Rotterdam.
Eliwa, H. A., Abu El-Enen, M. M., Khalaf, I. & Itaya, T., 2004.
Metamorphic evolution of Sinai metapelites and gneisses:
constrains from petrology and K ⁄Ar age dating. Egyptian
Journal of Geology,48, 169–185.
Eliwa, H. A., Abu El-Enen, M. M., Khalaf, I. M., Itaya, T. &
Murata, M., 2008. Metamorphic evolution of Neoproterozoic
metapelites and gneisses in Sinai, Egypt: insights from
petrology, mineral chemistry and K–Ar age dating. Journal of
African Earth Sciences,51, 107–122.
El-Shafei, M. K. & Kusky, T. M., 2003. Structural and tectonic
evolution of the Neoproterozoic Feiran–Solaf metamorphic
belt, Sinai Peninsula: implications for the closure of the
Mozambique Ocean. Precambrian Research,123, 269–293.
El-Tokhi, M., 1992. Origin and tectonic implications of
Pan-African amphibolites of Wadi Feiran South Sinai. In:
Proceedings of the Third Conference on Geology of Sinai
Development (eds Abd El-Khalek, M. L., El-Ghawaby, M. A.
& Morsy, A. M.). Suez Canal University pp. 239–248. Ismailia.
Engel, A. E. J., Dixon, T. H. & Stern, R. J., 1980. Late
Precambrian evolution of Afro-Arabian crust from ocean to
craton. Geological Society America Bulletin,91, 699–706.
OBLIQUE TRANSPRESSION IN SINAI, EGYPT 457
2009 Blackwell Publishing Ltd

Farahat, E. S., Mohamed, H. A., Ahmed, A. F. & El Mahallawi,
M. M., 2007. Origin of I- and A-type granitoids from the
Eastern Desert of Egypt: implications for crustal growth in the
northern Arabian-Nubian Shield. Journal of African Earth
Sciences,49, 43–58.
Fowler, A. & Hassan, I., 2008. Extensional tectonic origin
of gneissosity and related structures of the Feiran–Solaf
metamorphic belt, Sinai, Egypt. Precambrian Research,164,
119–136.
Fritz, H., Wallbrecher, E., Khudeir, A. A., Abu El Ela, F. &
Dallmeyer, D. R., 1996. Formation of Neoproterozoic meta-
morphic core complexes during oblique convergence (Eastern
Desert, Egypt). Journal of African Earth Sciences,23,
311–323.
Fritz, H., Dallmeyer, D. R., Wallbrecher, E. et al., 2002.
Neoproterozoic tectonothermal evolution of the Central
Eastern Desert, Egypt: a slow velocity tectonic process of core
complex exhumation. Journal of African Earth Sciences,34,
137–155.
Ganguly, J. & Saxena, S. K., 1984. Mixing properties of
aluminosilicate garnets: constraints from natural and
experimental data, and applications to geothermo-barometry.
American Mineralogist,69, 88–97.
Gass, I. G., 1982. Upper Proterozoic (Pan-African) calcalkaline
magmatism in northeastern Africa and Arabia. In: Andesites
(ed. Thorpe, R.S.), pp. 591–609. Wiley, New York.
Green, E. C. R., Holland, T. J. B. & Powell, R., 2007. An order-
disorder model for omphacitic pyroxenes in the system
jadeite–diopside–hedenbergite–acmite, with applications to
eclogitic rocks. American Mineralogist,92, 1181–1189.
Guiraud, M., Powell, R. & Rebay, G., 2001. H
2
O in metamor-
phism and unexpected behaviour in the preservation of
metamorphic mineral assemblages. Journal of Metamorphic
Geology,19, 445–454.
Hashad, A. H., Sayyah, T. A. & El Manharawy, M. S., 1981.
Isotopic composition of strontium and origin of Wadi Kareim
volcanics, Eastern Desert. Egyptian Journal of Geology,25,
141–147.
Hoisch, T., 1990. Empirical calibration of six geobarometers for
the mineral assemblage quartz+muscovite+biotite+plagio-
clase+garnet. Contributions to Mineralogy and Petrology,104,
225–234.
Holland, T. J. B. & Powell, R., 1998. An internally consistent
thermodynamic dataset for phases of petrological interest.
Journal of Metamorphic Geology,16, 309–343.
Holland, T. J. B. & Powell, R., 2003. Activity–composition
relations for phases in petrological calculations: an asymmet-
ric multicomponent formulation. Contributions to Mineralogy
and Petrology,145, 492–501.
Holland, T. J. B., Baker, J. M. & Powell, R., 1998. Mixing
properties and activity–composition relationships of chlorites
in the system MgO–FeO–Al
2
O
3
–SiO
2
–H
2
O. European Journal
of Mineralogy,10, 395–406.
Johnson, P. R., 1998. Tectonic map of Saudi Arabia and
adjacents areas (scale: 1 ⁄4,000,000). In: Technical report
USGSOF-98-3 (ed. Saudi Arabian Deputy Ministry for
Mineral Resources), 2pp. Saudi Arabia, Jeddah.
Johnson, T. E., White, R. W. & Powell, R., 2008. Partial melting
of metagreywacke: a calculated mineral equilibria study.
Journal of Metamorphic Geology,