Lithos 388–389 (2021) 106051

Contents lists available at ScienceDirect

Lithos

journal homepage: www.elsevier.com/locate/lithos

Research Article

Post-collisional volcanism with adakitic signatures

in the Arabian-Nubian Shield: A case study of calc-alkaline
Dokhan volcanics in the Eastern Desert of Egypt
Bassam A. Abuamarah a,⁎, Mokhles K. Azer b, Paul D. Asimow c, Qingshang Shi d
a
Department of Geology and Geophysics, King Saud University, Riyadh 11451, Saudi Arabia
b
Geological Sciences Department, National Research Centre, 12622-Dokki, Cairo, Egypt
c
Division of Geological & Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USA
d
State Key Laboratory of Geological Processes and Mineral Resources, School of Earth Science and Resources, China University of Geosciences, Beijing 100083, China

a r t i c l e i n f o a b s t r a c t

Article history: A volcano-sedimentary succession, including the Guruf volcanic series, is exposed at Gabal Um Guruf in the north
Received 21 December 2020 Eastern Desert of Egypt. A series of intermediate to silicic volcanic flows and associated pyroclastic deposits un-
Received in revised form 9 February 2021 conformably overlie quartz-diorite and granodiorite plutons and are intruded by high-level alkali feldspar gran-
Accepted 11 February 2021
ites. The metavolcanic xenoliths in the granodiorite are shown to be unrelated to the Guruf volcanics. Two
Available online 15 February 2021
volcanic cycles can be recognized. The lower unit is a high-K calc-alkaline series with minor basaltic andesite
Keywords:
and mostly andesite through dacite. The upper unit consists of rhyolite and rhyodacite that is transitional towards
Arabian–Nubian Shield alkaline, A-type magma. Both units are assigned to the Dokhan volcanics and interpreted as post-collisional. REE
Post-collisional volcanics patterns are characterized by slow progressive enrichment throughout the suite, with heavy REE apparently act-
Lithospheric delamination ing more incompatibly than light REE; concave-upwards REE patterns in basalt andesite suggest residual amphi-
Adakite bole in the source. Only the upper rhyolite displays prominent negative Eu anomalies (Eu/Eu* = 0.35–0.73).
Pyrophanite Apparent crystallization temperatures of primary pyroxene are 800–1000 °C and of primary amphiboles
781–944 °C. The major and trace element chemistry of the suite can be modeled as a single liquid line of descent,
with pressure decreasing as the magma cools and evolves, and a prominent role for apatite in sequestering CaO,
P2O5, and light REE. Although upper crustal contamination and magma mixing cannot be ruled out, the data do
not require them. The suite includes pyrophanite, an unusual Mn-rich ilmenite-group solid solution, not previ-
ously recorded in volcanic rocks in the Arabian-Nubian Shield (ANS). The Guruf volcanic series is undeformed
and unmetamorphosed, unlike the typical subduction-related volcanic units of the ANS. The geochemical charac-
teristics are consistent with those of many post-collisional ANS rocks. The mostly calc-alkaline character and
other traits previously interpreted to indicate an active arc setting instead more likely reflect remelting of earlier
arc-related material from the pre-collisional stage (850–740 Ma). The lava flows of the lower succession have
adakitic characteristics, including high Sr, low Y, low Yb, high Sr/Y, and high (La/Yb)n. This likely reflects the in-
fluence of residual or fractionated amphibole in the ANS crust, rather than slab melting during active subduction.
A post-collisional episode of delamination and heating of tonalitic lower crustal material, followed by mixing
with asthenospheric mantle, created a primary magma with the characteristics of the Guruf basaltic andesite, in-
cluding residual amphibole. There is no need for either late subduction in the area or long-term persistence of
subduction-influenced material in the local convecting mantle.
© 2021 Elsevier B.V. All rights reserved.

1. Introduction north to Eritrea and Ethiopia in the south. The ANS is a collage of
Neoproterozoic ophiolites and juvenile arc terranes, younger sedimen-
The Arabian-Nubian Shield (ANS) (Fig. 1a) is part of the Pan-African tary and volcanic basins, voluminous granitoid intrusions, and enclaves
belt, the largest known system of orogenies on Earth (Johnson and of pre-Neoproterozoic crust (Hargrove et al. 2006). Basement exposures
Woldehaimanot 2003). The ANS extends from Egypt in the west to of the ANS were exhumed in the Neogene by uplift associated with Red
Saudi Arabia in the east and from Jordan and southern Israel in the Sea rifting (Azer and Asimow, 2021).
Post-collisional magmatism in the northernmost ANS includes both
⁎ Corresponding author. intermediate to silicic plutonic bodies and numerous volcanic series.
E-mail address: [email protected] (B.A. Abuamarah). This vast magmatic phase commenced at ca. 635–625 Ma and

https://doi.org/10.1016/j.lithos.2021.106051
0024-4937/© 2021 Elsevier B.V. All rights reserved.
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 1. (a) Index map of the Arabian-Nubian Shield showing the location of Fig. 1b; (b) Regional geological map of Egypt showing distribution of the Dokhan volcanics and Hammamat
sediments in the Eastern Desert and Sinai, Egypt (modified after Abdel Rahman, 1996; Wilde and Youssef 2002). The location of the study area displayed in Fig. 2 is marked and the ap-
proximate boundaries between the north Eastern Desert (NED), central Eastern Desert (CED) and south Eastern Desert (SED) are shown as red dashed lines.

culminated at 610–600 Ma. There was a transition (with some overlap the Guruf volcanics of the lower unit are calc-alkaline and show many
in time) from batholithic calc-alkaline to alkaline A-type granite typical features of subduction-related magmas, they do not show any
magmatism at ca. 600 Ma (Be'eri-Shlevin et al. 2011). The early post- evidence of metamorphism, which is a universal feature of the
collisional volcano-sedimentary sequences dominate in the northern subduction-related volcanic phase in the ANS. We seek to resolve this
part of the ANS, especially in the Eastern Desert and Sinai, where they issue and to interpret our new discovery of adakitic members of the vol-
are associated with clastic sedimentary deposits. In the Eastern Desert canic series through detailed field work, mineralogical and geochemical
of Egypt, they are known as the Dokhan volcanics and dated at investigations of the Guruf volcanics, and studies of the metavolcanic
630–590 Ma (Abu El-Enen et al. 2018; Breitkreuz et al. 2010; Wilde xenoliths in the adjacent granodiorite that clarify the relationship of
and Youssef 2000). They differ from the preceding arc metavolcanic se- the volcanic series to its country rocks. Based on the present results,
quences in the abundance of felsic varieties, their high potassium con- we assign the Dokhan volcanics that outcrop at G. Um Guruf to the
tents, the presence of ignimbrites, and the absence of metamorphism. post-collisional phase of the ANS and their subduction-related charac-
The Dokhan volcanics include intermediate to felsic subaerial lava teristics to inheritance from earlier arc episodes rather than to active
flows and their pyroclastic equivalents, as well as subvolcanic bodies subduction, for which there is no other evidence in the region at
and minor basalt and basaltic andesite. The Dokhan volcanics have this time.
been separated in many areas into lower and upper sequences
(Breitkreuz et al. 2010; Obeid and Azer 2015). 2. Geologic setting
In this work, we document a late Neoproterozoic volcano-
sedimentary succession exposed at Gabal (G.) Um Guruf, in the north The Guruf volcano-sedimentary succession is exposed in the north-
Eastern Desert of Egypt. The calc-alkaline volcanic members of this suc- ern sector of the Eastern Desert, to the west of Hurghada town
cession (here named Guruf volcanics in order to accurately indicate (Fig. 1b). The study area is bounded by latitudes 27° 06′ and 27° 16′ E
their location) have been previously interpreted as subduction-related and by longitudes 32° 57′ and 33°10′ N. The mapped area includes
volcanism despite their young age compared to other evidence of sub- Neoproterozoic granitoids, calc-alkaline volcanic rocks and molasse
duction in the region (by Maurice et al. 2018, who named the section (Hammamat) sedimentary units (Fig. 2). The area is dissected by
they studied after the nearby Wadi Hamad). Although, as we will see, many structurally controlled Wadis (W.). The granitoid rocks include

2
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

quartz-diorite, granodiorite and alkali feldspar granite. Published works mafic microgranular enclaves (MMEs) and xenoliths of gneiss and por-
on the study area have left unsettled substantial controversy about the phyritic metavolcanic rocks. The MMEs are fine-grained and their
geologic settings of the various rock units (Ali 2015; Basta et al. 2017; shapes are ellipsoidal to irregular due to embayment of their margins.
Maurice et al. 2018; Moussa et al. 2008). In places, their margins show evidence of hybridization or gneissic tex-
Basta et al. (2017) concluded that all the granitoid rocks in the G. Um ture due to shear. The granodiorite is cut by many dikes of various com-
Guruf area are younger than the Dokhan volcanics and that all the gran- positions that abut against the Guruf volcanics and are cut by alkali
itoids were emplaced in a subduction-related tectonic setting, implying feldspar granite. The previous studies concur with our observations of
that the local Dokhan volcanics also date from an era of active subduc- the intrusive relations, concluding that the quartz-diorite and granodio-
tion. However, our field observations of intrusive relations demonstrate rite in the area are older than the Guruf volcanics (Hossny 2008). Basta
that the granitoids in the study area can be divided into two different et al. (2017) were compelled to adopt the opposite sequence on the
phases, separated by eruption of the local sequence of Dokhan volcanics. basis of their assumption that the metavolcanic xenoliths in the
The early phase includes quartz diorite and granodiorite, while the later quartz-diorite and granodiorite are derived from the Guruf volcanics,
phase consists of alkali feldspar granite. The quartz-diorite and granodi- but they presented no evidence to support this interpretation. We sam-
orite are both medium-to-coarse-grained and form low hills with dark pled and characterized the gneiss and metavolcanic xenoliths in the
grey, grey, or whitish-grey color. They are extensively weathered, exfo- granodiorite in order to test the conclusion of Basta et al. (2017) and re-
liated, and exhibit cavernous weathering. The granodiorite contains solve the question of relative chronology.

Fig. 2. Geologic map of Gabal Um Guruf area, North Eastern Desert, Egypt (modified after Basta et al. 2017).

3
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

The Guruf volcanic rocks are exposed mainly around G. Um Guruf dacite, while the upper succession consists of rhyodacite, rhyolite and
and to the northwest, across W. Hamad. Although outcrop is discontin- ignimbrite. Outcrops of the upper succession are limited and are well
uous, they apparently represent a western extension of the Dokhan vol- exposed only east of the G. Um Guruf mass; there are few exposures
canics from their type locality at Gabal Abu Dokhan, which lies in the western part of the study area along W. Hamad. There is a clear
immediately east of the study area. They extruded onto and unconform- angular unconformity between the upper and lower successions
ably overlie the quartz-diorite and granodiorite with sharp contacts (Fig. 3c). Pyroclastic deposits (agglomerates, ignimbrites and tuffs) al-
(Fig. 3a), whereas they are truncated and intruded by alkali feldspar ternate with the lava flows in both the upper and lower successions.
granite (Fig. 3b). Again, a well-defined angular unconformity can be observed where
The Guruf volcanic rocks show a range of colors in the field: dark tilted pyroclastic deposits of the upper sequence rest on nearly flat-
grey to grey, greenish-grey, and reddish-brown. They form part of an flying lava flows of the lower sequence (Fig. 3d). Maurice et al. (2018)
interlayered volcano-sedimentary series that can be divided into focused their discussion of the Dokhan volcanic rocks of the W.
upper and lower successions of lava flows intercalated with pyroclastic Hamad area exclusively on andesites and dacites, despite reporting nu-
rocks. The lower succession consists of basaltic andesite, andesite and merous rhyolitic and rhyodacitic analyses.

Fig. 3. Field photographs: (a) lower Guruf volcanics unconformably emplaced on a granodiorite surface, (b) alkali feldspar granite intruding the lower succession of Guruf volcanics,
(c) upper Guruf volcanics unconformably resting on the lower succession, (d) titled pyroclastic beds of the upper cycle unconformably resting on the lower lava flows, (e) Hammamat
sediments deposited nonconformably on a granodiorite surface, and (f) a roof-pendant of the Guruf volcanics above alkali feldspar granite.

4
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

The Hammamat molasse-type sedimentary rocks are exposed in the Andesite is massive, fine-grained and greenish to reddish-grey. It is
southwestern part of the mapped area. They vary in color from commonly porphyritic and consists of abundant plagioclase and subor-
greenish-grey to brick red. They are deposited directly on a paleosurface dinate amphibole and clinopyroxene phenocrysts set in a fine-grained
of granodiorite with a basal conglomerate (Fig. 3e) or intercalated with groundmass. Some andesite samples contain amygdales filled with sec-
the Guruf volcanics. They consist of alternating beds of mostly conglom- ondary minerals. Accessory phases include opaques and apatite;
erate and greywacke, with less common siltstone and mudstone. sericite, calcite and chlorite are secondary. Some plagioclase crystals
The alkali feldspar granites are the most abundant rocks in the are saussuritized; others show prominent oscillatory zoning (Fig. 4c).
mapped area; they include more than one intrusive phase (Basta et al. Equant microphenocrysts of augite may display irregular edges suggest-
2017). These granites are the youngest igneous activity in the mapped ing partial resorption or replacement around their margins by tremolite,
area; field observations demonstrate that they intruded into both chlorite, and iron oxides. Primary amphibole forms subhedral to
granodiorite and Guruf volcanics; off-shoots of the alkali feldspar gran- euhedral, elongated phenocrysts that are simply twinned and rarely
ite are observed in the Guruf volcanics and xenoliths of both granodio- zoned. The groundmass displays shape-preferred alignment and is com-
rite and Guruf volcanics are enclosed in alkali feldspar granite near posed of plagioclase laths, amphibole, biotite, augite and opaques.
their contacts. The alkali feldspar granites are massive, medium- to Dacite is massive, fine-grained and pinkish-grey. It carries abundant
coarse-grained, and pink to red in color. Unlike the quartz diorite subhedral to euhedral plagioclase phenocrysts up to 3.0 mm long, to-
and granodiorite, the alkali feldspar granites are devoid of mafic gether with scarce biotite, amphibole and quartz microphenocrysts,
microgranular enclaves or dikes. Some roof-pendants of Guruf volcanics set in a microcrystalline felsic groundmass. Apatite, zircon and Fe\\Ti
above alkali feldspar granite were noted (Fig. 3f). In some places, peg- oxides are the common accessory phases, while epidote, sericite and
matite and fluorite veins occur along contacts between the alkali feld- chlorite are present as secondary minerals. The plagioclase phenocrysts
spar granites and their country rocks. are subhedral to euhedral tabular crystals showing complex zoning
Two generation of dikes are recognized in the study area. The first (Fig. 4d). Some plagioclase phenocrysts are slightly to moderately al-
group (calc-alkaline dikes) cuts quartz-diorite and granodiorite but is tered to sericite and epidote. Subhedral tabular microphenocrysts of bi-
truncated by both Guruf volcanics and alkali feldspar granites. The sec- otite are slightly altered to chlorite and opaques. A few small crystals of
ond group (alkaline dikes) cuts quartz-diorite, granodiorite and Guruf altered amphibole and of quartz are observed in some samples.
volcanics, but is intruded by alkali feldspar granites.
There are some age constraints for the granitoids. The U\\Pb zircon 3.1.2. Upper Guruf volcanics
age of alkali feldspar granites just to the north of the study area at G. Abu The lava flows of the upper Guruf volcanic series include rhyolite
Marwa is 595 ± 8.5 Ma (Ali 2015), consistent with a Rb\\Sr whole-rock and rhyodacite. Rhyolite is fine-grained with 10–25% phenocrysts set
age of 590 ± 4 Ma for the same granites (Moussa, 1998). Likewise, in microcrystalline groundmass. The phenocrysts mostly occur as dis-
Moussa et al. (2008) obtained an age of 595.3 ± 3.3 Ma for alkali feld- crete crystals but rare glomerophyric or cumulophyric aggregates are
spar granite at G. Abu Harba, just at the NE border of the present present. The phenocrysts include plagioclase, K-feldspar and quartz
mapped area. The U\\Pb zircon age of alkali feldspar granite at G. El- with rare microphenocrysts of biotite and pyrophanite. Biotite occurs
Glouf (the SW corner of the mapped area) is 596.4 ± 11.2 Ma (Ali as tabular corroded crystals (Fig. 4e). Pyrophanite is recorded for the
2015). These dated granitoids intrude the Guruf volcanics, which are first time in the post collisional calc-alkaline volcanics in Egypt. It is an
therefore older than ~595 Ma. There is no formal upper bound on the opaque member of the ilmenite group with formula MnTiO3 and crys-
maximum age of the Guruf volcanics, as the local granodiorite and tallizes in the trigonal system. It occurs as anhedral to subhedral
quartz diorite have not been dated. Although U\\Pb zircon dates are microphenocrysts associated with apatite and zircon (Fig. 4f). The
available for a number of ignimbrites associated with Dokhan volcanics groundmass is microcrystalline and shows microgranular and spheru-
in the Eastern Desert (e.g., Breitkreuz et al. 2010), none of them are litic textures. The textures of these rhyolite samples definitively indicate
within the current study area and we do not assume any correlation emplacement as a lava flow; there are not pyroclastic rocks.
to other dated sections. Rhyodacite is observed only in the upper series of the Guruf volca-
nics. It is rather similar to the dacite in the lower series but contains
more K-feldspar and quartz phenocrysts and a lower abundance of
3. Petrography
mafic minerals. Phenocrysts in rhyodacite, either as discrete crystals or
glomerophyric aggregates, vary significantly in modal abundance and
This section will describe in detail the petrography of the Guruf lava
are set in microcrystalline groundmass. Phenocryst phases include pla-
flows (basaltic andesite, andesite, dacite, rhyodacite and rhyolite), their
gioclase, K-feldspar, and quartz with rare microphenocrysts of biotite.
associated pyroclastic deposits (ignimbrite, agglomerates, lapilli tuffs,
Again, these rhyodacite samples were emplaced as flows; they are not
crystal tuffs and banded tuffs), the two types of dikes, and the
pyroclastic.
metavolcanic xenoliths in the granodiorite.
3.2. Pyroclastic deposits
3.1. Lava flows
The pyroclastic rocks are recognized and distinguished from the
3.1.1. Lower Guruf volcanics lavas by their textures. Their nomenclature and classification is purely
The lower lava flows of the Guruf volcanic series include basaltic an- descriptive and based primarily on their granulometric state. Pyroclastic
desite, andesite and dacite. Basaltic andesite is porphyritic, with abun- rocks occur in both lower and upper units; they include ignimbrite, ag-
dant plagioclase and clinopyroxene phenocrysts set in a fine-grained glomerate, and several varieties of tuff (lapilli tuff, crystal tuff, and
groundmass of plagioclase, clinopyroxene, and abundant (~8 vol%) banded tuff).
opaque minerals. Accessory phases include apatite and titanite; second- Ignimbrites, both welded and unwelded, are abundant in the upper
ary minerals include chlorite, actinolite, sericite and veinlets of quartz succession, associated with rhyolite. They are rhyolitic in composition
and epidote. Plagioclase phenocrysts are generally twinned, rarely and show a foliation defined by flattened lenticular glass particles
zoned, and slightly altered to sericite; small plagioclase laths in the (fiamme) in a groundmass of felsic volcanic glass, pumice, and irregular
groundmass show a clear preferred orientation (Fig. 4a). Clinopyroxene glass shards. The fiamme may be sinuously curved or flattened; their
forms anhedral microphenocrysts of augite (Fig. 4b), partly replaced by preferred orientation is attributed to compaction and welding. The
secondary amphibole and chlorite. Some augite phenocrysts show oscil- groundmass shows fluidal structure and encloses crystals, crystal frag-
latory or reverse zoning, confirmed by microprobe analyses. ments, and scarce felsic rock fragments with various degrees of welding.

5
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 4. Photomicrographs showing petrographic textures in cross-polarized transmitted light (a, c, d and e) or back-scattered electrons (b and f): (a) large plagioclase phenocryst
embedded in groundmass showing flow texture in basaltic andesite, (b) phenocryst of clinopyroxene marginally altered to chlorite in basaltic andesite, (c) plagioclase phenocryst
show oscillatory zoning in andesite, (d) plagioclase phenocryst showing complex zoning in dacite, (e) microphenocryst of biotite embedded in felsic groundmass in rhyolite, and
(f) microphenocrysts of pyrophanite associated with apatite and zircon.

Rarely, groundmass shows spherulitic texture and devitrification into with pyroclast sizes extending from 3 to 45 mm. The lithic fragments in-
equigranular mosaic crystal aggregates. clude andesite, dacite and rhyolite, all with porphyritic textures. The
Agglomerates are poorly sorted rocks assembled from rounded to crystal fragments are mainly cracked crystals of plagioclase, altered
subrounded volcanic 3–6 cm rock fragments embedded in a glassy to mafic minerals, and quartz with lesser opaque minerals and rare K-
tuffaceous groundmass. The rock fragments are volcanic, including por- feldspars. The matrix may be glassy or may consist of fine plagioclase
phyritic andesite and dacite with rare basaltic andesite. Some coarse and quartz fragments with minor opaque minerals.
crystal fragments of altered mafic minerals, plagioclase, K-feldspars, Crystal tuffs are composed predominantly of crystal fragments of
and quartz are observed in the groundmass. The groundmass consists variable sizes, embedded along with minor rock fragments in a
of glassy tuff, small lithic and crystal fragments, and abundant stretched glassy or fine tuffaceous groundmass. The crystal fragments include
vesicles filled with secondary minerals such as quartz, carbonates, adu- plagioclase, quartz and mafic minerals. The plagioclase and mafic
laria and epidote. minerals crystal fragments are highly altered. Rare angular lithic
Lapilli tuffs are composed of angular to subangular crystal and lithic (andesite to rhyolite) and glass fragments of variable sizes are
fragments set in fine tuffaceous groundmass. They are poorly sorted, present.

6
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Banded tuffs are very fine-grained. They are composed of alternating in the relict fresh plagioclase. Some plagioclase phenocrysts show bent
laminae (< 2.0 mm thick) and thin beds (< 10 mm thick) of variable twin planes due to deformation. Amphibole is highly altered to actino-
color, reflecting changes in grain size and composition. The fine laminae lite, chlorite, and opaques. Augite occurs as subhedral prismatic crystals
consist of very fine-grained, vitreous-ashy material, with tiny particles that occasionally show simple twinning.
of quartz and feldspars. The thin beds contain coarser crystals and crys- Meta-andesite is porphyritic and fine-grained, massive, and grey to
tal fragments of plagioclase and quartz with rare altered mafic minerals greenish-grey. The phenocrysts include plagioclase, pyroxene, amphi-
set in a sericite-rich groundmass. bole and opaques in fine-grained groundmass of the same minerals as
well as accessory zircon and titanite. Actinolite, epidote, calcite, and
3.3. Dikes chlorite are secondary. Plagioclase phenocrysts occur as large subhedral
to euhedral tabular crystals highly altered in the cores to epidote and
3.3.1. Calc-alkaline dikes sericite. Augite is found as individual large phenocrysts and as fine crys-
The calc-alkaline dikes in the study area include basalt, andesite and tals in the groundmass.
rhyolite. Basalt dikes are fine-grained and include porphyritic,
subophitic and rarely doleritic varieties. Phenocrysts of plagioclase,
4. Analytical methods
hornblende and clinopyroxene are embedded in a fine-grained to mi-
crocrystalline groundmass of plagioclase, augite and magnetite with
Based on the petrographic studies, 42 rock samples (29 lava flows,
minor secondary epidote and chlorite. Accessory phases include
10 dikes and 3 volcanic xenoliths) were analyzed for major, trace and
opaques and apatite; chlorite, actinolite, sericite and titanite are second-
rare earth elements at Activation Laboratories Ltd. (Actlabs, Canada)
ary. Subhedral zoned plagioclase is the dominant phenocryst. Horn-
and the California Institute of Technology (Caltech, USA). In the Actlabs
blende forms euhedral to subhedral phenocrysts and small prisms in
(Canada) analyses, the major oxides were measured by lithium
the groundmass; it variably altered to chlorite. Clinopyroxene displays
metaborate/tetraborate fusion ICP-AES. Trace and rare earth elements
different habits, from phenocrysts to small interstitial grains; it is some-
were measured by ICP-MS following lithium borate fusion and acid di-
times replaced by amphibole.
gestion. Loss on ignition (LOI) is determined by weight difference after
Andesite dikes are massive, fine-grained and reddish-grey. They
ignition at 1000 °C. Precision and accuracy were controlled by analysis
consist essentially of plagioclase and hornblende phenocrysts with
of international reference materials and replicate analyses and are 1%
micro-phenocrysts of augite in a fine-grained groundmass. Accessory
for major elements and 2% to 5% for trace elements. Full details are on
opaque minerals and apatite are present, as well as secondary sericite,
the laboratory website (actlabs.com).
calcite and chlorite. The groundmass is composed of subhedral to
At Caltech, whole-rock compositions of major and minor elements
anhedral plagioclase laths, up to 0.3 mm long, along with hornblende,
were determined using a Panalytical Zetium wavelength-dispersive
quartz, and opaques.
XRF spectrometer. Sample powders were milled in an agate ball mill
Rhyolite dikes are fine-grained, with variable phenocryst abun-
and loss on ignition (LOI) determined by weighing before and after fir-
dances (<30%) and microcrystalline groundmass. The phenocrysts
ing in air for 1 h at 1050 °C for mafic samples and at 1000 °C for felsic
occur as discrete crystals and as glomerophyric or cumulophyric aggre-
samples. Fused-glass beads were prepared from fired powders by
gates. Phenocrysts include plagioclase, K-feldspar and quartz with rare
mixing with 10 times their weight in 66.67% Li2B4O7–32.83% LiBO2-
microphenocrysts of biotite. The groundmass is microcrystalline and
0.50% LiI flux and fusing at 1200 °C. The XRF protocol was calibrated
shows microgranular and spherulitic textures.
against 21 USGS rock standards. USGS standards BCR-2, AGV-2, and
RGM-2 (a basalt, an andesite, and a rhyolite) were routinely analyzed
3.3.2. Alkaline dikes
along with unknowns for quality control and drift correction. After
Alkaline dikes in the study area include trachyandesite and alkali
XRF measurement, 25 ± 1 mg chips of the fused-glass beads were dis-
rhyolite. Trachyandesite dikes are porphyritic, with phenocrysts of pla-
solved in new 50 mL polypropylene tubes by refluxing in 2 mL of hot
gioclase, clinopyroxene and amphibole in a fine-grained groundmass of
(100 °C) 3:1 nitric and hydrofluoric acid for at least 8 h. After dilution
the same minerals along with accessory opaques and apatite. Plagio-
with milli-Q distilled water to 30 mL total volume, elemental concentra-
clase forms both tabular phenocrysts and fine-grained laths in the
tions of rare earth elements (REE) and selected major and trace ele-
groundmass. Hornblende forms both large prismatic crystals and gran-
ments were determined with an Agilent Technologies 8800 triple
ular aggregates. Clinopyroxene, mainly augite, is present as euhedral to
quadrupole ICP-MS. A working curve for instrument sensitivity was de-
subhedral prismatic phenocrysts and granular aggregates in the
veloped using a blank fused bead from the same batch of flux along with
groundmass.
USGS standards AGV-2 and RGM-2. To control for quality, additional
Alkali rhyolite dikes are porphyritic and composed mainly of alkali
USGS standards (DTS-2, BCR-1, G-2) were included as unknowns.
feldspar phenocrysts in a fine-grained groundmass. Alkali feldspar is
Mineral chemical analyses, electron backscatter images, and X-ray
mostly turbid and stained reddish-brown. Quartz occurs as small
mapping were obtained with a five-spectrometer JEOL JXA-8200
anhedral microphenocrysts and in the groundmass. The mafic minerals
electron microprobe at Caltech. Surfaces were coated with ~15 nm of
are difficult to distinguish as they are highly altered to chlorite and iron
carbon. Operating conditions were 15 kV, 25 nA, 1 μm beam, 20 s on-
oxides. The groundmass shows trachytic texture defined by aligned al-
peak counting times, the mean atomic number (MAN) background-
kali feldspar laths and altered mafic microlites. Apatite and zircon are
correction method, and the CITZAF matrix correction routine. The ana-
the main accessory minerals; chlorite, epidote and calcite are the main
lytical standards used for analyses were synthetic forsterite, fayalite,
secondary minerals.
Mn-olivine, anorthite, TiO2, and Cr2O3; Amelia albite, Asbestos micro-
cline, and Durango apatite. Quartz, MgO, and Al2O3 provided additional
3.4. Metavolcanic xenoliths
control points for the MAN background subtraction.
Metavolcanic xenoliths in granodiorite include metabasalt and
meta-andesite. Metabasalt is porphyritic and fine-grained, massive, 5. Mineral chemistry
and dark greenish-grey to black. It is composed essentially of plagio-
clase, augite and amphibole phenocrysts in fine-grained groundmass Compositions of the essential minerals (pyroxene, amphiboles, feld-
of plagioclase and augite with accessory titanite, apatite, and opaque spars, chlorite, pyrophanite, and Fe\\Ti oxides) were analyzed in basal-
minerals. Subhedral to euhedral plagioclase crystals are partly to tic andesite, andesite, dacite and rhyolite. Most analyses were
completely altered to epidote and sericite but zoning can be recognized performed on phenocrysts or microphenocrysts large enough for

7
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

single-phase analyses. All the microprobe data are presented as an on- phenocrysts from basaltic andesite, andesite, and dacite show CaO con-
line supplement (Tables 1SD-9SD). tents decreasing as expected with petrographic type: andesine to labra-
Chemical compositions, structural formulae and end-member com- dorite (An41-62) in basalt andesite, oligoclase to labradorite (An14-53) in
ponents of pyroxene phenocrysts in basaltic andesite and andesite are andesite, and oligoclase to andesine (An11-36) in dacite. Where zoning
given in Supplementary Table 1SD. All pyroxene analyses are rich in was identified in basaltic andesite and andesite, it may be normal (de-
CaO and classify mainly as augite according to Morimoto et al. (1988) creasing An content towards rim) or oscillatory. Albite, An<8.4, was an-
(Fig. 5a). Compositional variation, whether among crystals or within alyzed only in dacite and rhyolite. K-Feldspars were analyzed in dacite
zoned phenocrysts, extends parallel to isotherms in the pyroxene quad- and rhyolite; K2O contents are high (14.6–16.1 wt% in dacite and
rilateral projection. On the SiO2 versus Al2O3 diagram of Le Bas (1962), 15.7–16.3 wt% in rhyolite) and CaO contents are low (<0.6 wt%).
all clinopyroxene analyses plot well into the subalkaline field (Fig. 5b). Perthitic exsolution was not observed in any of the samples.
Amphiboles, both primary and secondary (replacing clinopyroxene), Chlorite formed by replacement of pyroxene, biotite and amphibole;
were analyzed in basaltic andesite, andesite and dacite. Oxide analyses compositions and structural formulae of chlorite on the basis of 28 oxy-
and structural formulae are given in Supplementary Table 2SD. Primary gens are given in Supplementary Table 6SD. Adopting the nomenclature
amphibole is high in TiO2 (>1.0 wt%), whereas secondary amphibole of Hey (1954), most analyses are pycnochlorite, with minor extension
has low TiO2 (< 0.5 wt%). All the amphibole analyses are calcic; primary into the neighboring fields of diabantite, ripidolite and brunsvigite
amphibole ranges from magnesio-hornblende to tschermakitic horn- (Fig. 5d). No particular correlation was noted between chlorite compo-
blende, whereas secondary amphibole ranges from actinolite to sition and which mafic mineral it replaced.
actinolitic hornblende. The discrimination diagram of Keeditse et al. Pyrophanite, a member of the ilmenite group with end-member
(2016) supports the petrographic interpretation that both primary formula MnTiO3, was identified in the rhyolite. The compositions,
and secondary amphibole is present in all rock types (Fig. 5c). structural formulae and end-member components are given in
The analyzed feldspars include plagioclase, albite and K-feldspars. Supplementary Table 7SD. Pyrophanite analyses are principally TiO2
Their oxide compositions, structural formulae and end-member compo- (45.4–51.7 wt%), FeO* (14.0–22.5 wt%) and MnO (28.8–33.4 wt%), cor-
nents are given in Supplementary Tables 3SD, 4SD and 5SD. Plagioclase responding to 62.8–72.0 mol% pyrophanite component. A single

Fig. 5. Mineral chemistry: (a) pyroxene quadrilateral with classification of Morimoto et al. (1988) and isotherms after Lindsley (1983), (b) SiO2-Al2O3 discrimination diagram for
clinopyroxene (after Le Bas 1962), (c) discrimination diagram for primary and secondary amphiboles (after Keeditse et al. 2016), and (d) chlorite nomenclature (after Hey, 1954).

8
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

analysis of mixed ilmenite-series solid solution has subequal propor- Three samples of metavolcanic xenoliths in the granodiorite were ana-
tions of pyrophanite, ilmenite, and hematite components; it is probably lyzed to test whether their protoliths were associated with the calc-
altered material. alkaline Guruf volcanics (Basta et al. 2017) or with the older island-arc
Fe\\Ti oxides were analyzed in basaltic andesite, andesite, dacite volcanic phase in the region (Shadli metavolcanics). Moreover, 10
and rhyolite. Magnetite and ilmenite are most common, with rare he- dike samples, associated with the studied volcanic series or cutting the
matite and rutile considered to be alteration products. Chemical compo- country rocks, were analyzed to clarify the genetic relationships be-
sitions, structural formulae and end-member components of magnetite tween the dikes and the volcanic units.
and ilmenite are given in Supplementary Tables 8SD and 9SD. The mag-
netite in dacite is nearly homogenous, with low TiO2 contents
(0.1–0.8 wt%; ≤ 2.2 mol% ulvöspinel). Magnetite in andesite and basaltic 6.1. Geochemical characteristics
andesite mostly contains elevated TiO2 contents (2.0–24.1 wt%,
6–73 mol% ulvöspinel) together with a small population of Ti-poor mag- 6.1.1. Lava flows
netite (<1.5 wt% TiO2, ≤4 mol% ulvöspinel). Reflected light observations Major and trace element analyses of 29 new lava flow samples
show coexisting homogeneous or exsolved titanomagnetite together (Tables 1 and 2) are considered alongside 65 lava flow and pyroclastic
with Ti-poor magnetite in andesite and basaltic andesite. Analyzed il- analyses from other outcrops of the same volcanic sequences published
menite is close to nominal end-member ilmenite (89–94 mol% by Maurice et al. (2018). The new analyses span a wide range of compo-
FeTiO3). It is notably poor in Al2O3, MgO and CaO. Ilmenite in dacite is sitions (52.0–74.7 wt% SiO2). On the total alkali-silica (TAS) diagram
enriched in MnO (3.3–3.4 wt%) compared to that in andesite (Cox et al. 1979), the lower unit ranges from basaltic-andesite through
(2.8–3.1 wt%) and basaltic andesite (1.3–2.9 wt%), as expected from andesite to dacite, while the upper unit is restricted to rhyolite (the cat-
progressive differentiation of calk-alkaline rocks under relatively oxi- egory rhyodacite is not used on this diagram) (Fig. 6a). The Guruf volca-
dizing conditions. nics straddle the dividing line between calc-alkaline and alkaline fields.
The SiO2 vs. Zr/TiO2 diagram of Winchester and Floyd (1977) gives con-
sistent results, with the new Guruf volcanic data spanning the
6. Geochemistry subalkaline andesite, dacite, rhyodacite and rhyolite fields (Fig. 6b).
Maurice et al. (2018) applied the term Dokhan volcanics only to andes-
Twenty-nine lava samples (4 basaltic-andesite, 8 andesite, 5 dacite, ite and dacite (plus minor basalt and basaltic andesite), despite publish-
4 rhyodacite and 8 rhyolite) were selected for chemical analyses. ing numerous analyses of rhyodacite and rhyolite (Fig. 6 a,b). The Guruf

Table 1
Chemical composition of the lower lava flows of Guruf volcanics, North Eastern Desert, Egypt.

Rock type Basaltic andesite Andesite Dacite

Sample GH22 GH23 GH26 GH27 GH28 GH30 GH31 GH32 GH33 GH34 GH35 GH36 GH2a GH3a GH13a GH24 GH25 GH29

SiO2 52.05 54.19 54.88 53.75 56.48 56.16 58.7 57.7 60.93 57.44 61.36 57.22 65.23 62.454 66.07 64.36 63.73 63.92
TiO2 1.48 1.35 1.51 1.4 1.35 1.39 1.39 1.13 1.05 1.16 0.99 1.2 0.61 0.624 0.48 0.98 0.87 0.88
Al2O3 17.82 17.33 15.53 17.3 16.63 16.92 16.03 16.56 15.3 16.38 15.86 15.76 15.72 16.380 14.20 14.12 15.95 16.02
Fe2O3 8.04 8.09 8.29 7.74 6.94 7.38 6.1 7.86 6.71 6.99 5.97 7.65 3.11 4.291 3.63 6.24 4.7 4.67
MnO 0.13 0.12 0.16 0.15 0.11 0.11 0.11 0.11 0.08 0.09 0.06 0.12 0.12 0.122 0.06 0.09 0.07 0.06
MgO 4.93 4.88 4.75 4.27 3.51 3.69 3.29 3.39 2.92 3.03 3.04 3.74 0.97 1.647 1.97 2.63 2.58 2.25
CaO 6.29 6.51 5.59 6.7 4.87 4.81 5.14 4.75 4.66 4.91 3.77 5.76 1.83 3.464 2.27 2.8 3.85 3.72
Na2O 3.81 3.72 5.23 4.12 4.92 4.91 4.27 3.96 3.53 4.23 4.1 3.27 6.35 4.474 3.69 4.15 4.32 4.21
K2O 1.55 1.38 1.6 1.56 2.14 2.07 2.4 2.41 2.52 2.45 2.73 2.02 2.77 3.116 3.95 2.87 2.9 3.07
P2O5 0.53 0.49 0.45 0.53 0.39 0.37 0.39 0.32 0.3 0.34 0.29 0.39 0.15 0.203 0.13 0.19 0.24 0.33
LOI 2.92 1.23 1.22 2.21 1.83 1.55 1.61 1.81 0.98 2.31 1.6 2.19 0.39 0.891 1.03 1.18 1.07 1.1
Total 99.55 99.29 99.21 99.73 99.17 99.36 99.43 100 98.98 99.33 99.77 99.32 97.24 97.666 97.50 99.61 100.28 100.23
Sc 16.99 20.48 16.96 17.35 14.54 14.96 11.09 14.49 9.54 13.07 8.62 13.20 3.28 5.16 1.99 8.72 16.67 17.62
Rb 31.19 25.96 41.93 30.86 35.98 34.22 40.24 58.51 45.10 41.94 54.41 23.41 80.82 118.75 101.14 64.64 52.50 46.65
Ba 630.93 519.52 648.63 649.56 744.21 666.50 809.39 701.25 1113.69 811.77 1098.16 978.24 1393.33 768.45 856.74 989.83 674.08 605.21
Sr 971.66 901.66 855.72 825.17 862.56 769.41 889.33 852.15 826.44 872.36 660.71 827.36 295.33 527.31 360.17 606.04 675.46 732.35
Nb 6.63 5.08 7.51 7.45 7.25 7.95 8.10 6.10 6.44 6.05 6.68 9.67 13.24 9.30 7.22 6.20 9.24 8.91
Zr 161.10 117.74 182.07 164.94 165.25 177.71 226.45 151.99 168.53 158.77 205.88 271.84 318.63 215.61 153.97 175.92 175.54 234.52
Hf 3.65 3.09 4.42 4.33 4.20 4.53 5.22 3.88 5.83 3.95 6.23 4.84 8.37 5.19 4.17 7.68 5.40 5.38
Y 14.18 9.88 13.46 16.69 16.83 18.26 17.46 15.85 13.81 13.49 13.70 16.21 29.55 20.59 15.06 16.35 15.81 16.28
Zn 76.25 75.50 92.04 74.73 66.70 70.28 66.02 62.05 53.79 66.70 61.65 113.47 61.99 61.10 44.77 51.73 86.67 88.08
Cu 30.72 48.75 7.71 13.59 6.37 7.70 29.99 18.49 57.48 18.15 44.03 17.48 <d.l. 13.50 6.55 17.96 20.18
Ni 39.07 89.72 63.07 38.46 51.00 17.05 29.45 42.76 56.86 38.64 71.44 517.48 <d.l. 53.40 21.16 74.60 31.33 52.73
Co 25.43 37.02 24.93 21.00 19.84 23.12 15.12 23.30 19.81 27.12 19.92 21.22 26.53 43.05 32.08 19.06 18.50 18.75
Cr 95.52 317.47 145.40 125.75 90.68 62.15 160.10 129.16 179.72 176.55 310.28 68.22 19.02 119.07 84.00 312.43 54.33 105.08
V 161.66 181.31 161.94 164.17 150.14 173.29 110.25 145.43 118.67 131.67 113.80 54.08 20.39 67.84 65.78 118.10 52.83 55.23
Pb 9.53 5.32 8.93 5.08 8.31 7.73 15.32 7.54 8.73 6.34 9.38 14.32 10.33 14.58 16.65 7.27 19.83 21.55
Th 2.51 2.07 2.27 2.94 2.72 3.07 3.24 4.18 4.53 3.32 4.60 4.33 10.16 8.77 12.20 4.24 6.01 5.19
Ta 0.44 0.23 0.38 0.43 0.40 0.47 0.53 0.42 0.41 0.43 0.36 0.44 0.78 0.66 0.73 0.41 0.73 0.64
U 1.04 0.72 0.81 1.00 0.93 1.05 1.03 1.36 1.53 1.07 1.72 1.08 2.47 3.09 4.01 1.59 2.16 1.44
Mo 2.04 7.39 10.12 5.28 6.89 9.40 5.83 6.41 3.91 10.42 9.21 4.21 0.56 2.24 2.00 7.52 7.81 6.84
Ga 18.16 16.84 17.68 17.12 19.25 18.94 17.81 16.95 16.25 17.79 18.46 20.22 – – – 16.72 19.17 18.84
As 3.65 16.89 2.02 2.92 7.69 4.71 2.19 2.41 2.46 7.23 4.28 2.89 – – – 4.36 4.27 3.58
Cs 1.56 2.58 2.29 1.99 1.34 0.90 2.35 1600.90 1.69 1.93 1.93 1.38 – – – 1.83 0.96 1.81
Mg# 55.86 55.46 54.18 53.24 51.07 50.79 52.68 47.09 47.32 47.22 51.24 50.22 38.25 44.20 52.82 46.52 53.12 49.86
AI 0.45 0.44 0.67 0.49 0.63 0.61 0.60 0.55 0.56 0.59 0.61 0.48 0.86 0.66 0.73 0.70 0.64 0.64
ASI 0.92 0.89 0.76 0.84 0.86 0.89 0.85 0.93 0.90 0.88 0.96 0.87 0.94 0.96 0.98 0.94 0.92 0.94
Sr/Y 68.52 91.26 63.58 49.44 51.25 42.14 50.94 53.76 59.84 64.67 48.23 51.04 9.99 25.61 23.91 37.07 42.72 44.98
a
Samples analyzed in Caltech (USA).

9
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Table 2
Chemical composition of the upper lava flows of Guruf volcanics, North Eastern Desert, Egypt.

Rock type Rhyodacite Rhyolite

Sample No. GH1a GH8a GH12a GH18a Gh4a GH10a GH15a GH39 GH40 GH42 GH45

SiO2 68.63 67.25 69.49 67.56 74.71 74.19 73.68 72.26 71.51 70.75 70.05
TiO2 0.55 0.44 0.46 0.39 0.26 0.19 0.16 0.40 0.42 0.51 0.54
Al2O3 14.18 14.95 15.69 15.02 14.04 11.71 13.06 14.00 14.09 14.71 14.90
Fe2O3 2.87 2.95 3.06 2.64 1.44 2.61 1.57 2.16 2.47 2.55 2.64
MnO 0.09 0.09 0.09 0.08 0.07 0.01 0.04 0.06 0.06 0.10 0.09
MgO 0.69 0.76 0.83 0.61 0.18 0.10 0.11 0.28 0.35 0.53 0.49
CaO 1.56 1.54 1.64 1.77 0.35 0.14 0.70 0.59 0.96 1.35 1.29
Na2O 6.43 4.64 4.67 4.52 4.72 4.18 3.73 5.48 5.22 5.50 5.78
K2O 1.81 3.85 3.82 3.89 4.47 3.94 4.66 3.60 3.72 3.50 3.29
P2O5 0.13 0.14 0.14 0.12 0.04 0.01 0.02 0.05 0.07 0.09 0.09
LOI 0.28 0.76 0.80 0.41 0.38 0.28 0.50 0.71 0.97 0.47 0.46
Total 97.23 97.36 100.70 97.00 100.66 97.38 98.24 99.59 99.84 100.06 99.62
Sc 2.08 2.68 2.96 1.77 0.76 2.17 2.44 7.99 7.76 8.23 10.09
Rb 48.09 97.05 99.40 115.83 88.17 95.06 116.96 68.33 75.53 67.31 78.18
Ba 928.76 878.17 866.47 826.09 721.07 828.85 693.59 1007.42 1001.49 1032.20 1077.16
Sr 197.33 310.89 324.41 287.15 57.42 29.75 115.51 121.89 167.02 212.84 261.28
Nb 14.50 12.22 12.77 12.06 15.08 15.12 14.04 16.94 15.06 16.26 18.27
Zr 272.64 274.38 254.13 276.27 222.64 272.98 186.51 432.01 366.23 390.06 442.69
Hf 7.59 7.57 6.93 7.70 7.12 8.11 6.51 11.32 9.70 9.75 11.55
Y 30.24 26.38 25.02 25.87 32.33 37.06 29.79 40.76 33.89 36.13 42.26
Zn 46.14 64.99 70.51 59.66 58.11 18.43 44.13 61.92 56.29 63.44 75.44
Cu <d.l. <d.l. <d.l. <d.l. <d.l. 3.61 32.10 4.99 5.78 5.68 5.76
Ni <d.l. <d.l. 5.99 2.02 17.17 <d.l. <d.l. 4.70 4.89 5.89 5.77
Co 47.58 22.67 11.02 25.26 35.43 22.24 12.70 0.68 1.76 1.54 2.49
Cr 32.46 4.57 17.26 17.69 58.11 8.67 5.39 151.46 136.13 190.48 132.43
V 17.85 27.03 30.65 15.99 6.79 3.84 7.27 15.10 22.90 23.05 23.04
Pb 5.04 19.02 13.44 20.16 17.06 13.54 22.83 15.65 17.83 15.29 15.67
Th 7.93 8.15 8.39 9.44 7.27 6.65 8.77 8.98 9.12 7.48 7.78
Ta 0.71 0.85 0.86 0.89 0.93 0.86 1.07 1.10 1.01 1.02 1.19
U 2.49 4.15 4.03 4.16 3.92 3.17 4.52 2.78 2.96 2.53 2.54
Mo 0.65 1.15 1.49 1.53 1.13 0.61 13.32 10.34 9.04 11.24 8.48
Ga – – – – – – – 20.58 20.74 18.53 22.89
As – – – – – – – 1.13 1.13 2.18 2.81
Cs – – – – – – – 0.55 0.53 1.56 1.83
Mg# 32.30 33.68 34.86 31.47 19.84 7.04 12.51 20.43 21.92 29.17 26.89
AI 0.88 0.79 0.75 0.78 0.90 0.95 0.86 0.92 0.90 0.87 0.88
ASI 0.92 1.02 1.06 1.01 1.06 1.03 1.05 1.00 0.98 0.96 0.97
Sr/Y 6.53 11.79 12.96 11.10 1.78 0.80 3.88 2.99 4.93 5.89 6.18
a
Samples analyzed in Caltech (USA).

volcanics are dominantly high-K, although some medium-K varieties granite fields (Fig. 9a). On the same diagram all the samples of
exist in the lower sequence (Fig. 6c). The analyzed lava flows define a Maurice et al. (2018) with SiO2 > 68 wt% (19 samples) plot in the al-
typical calc-alkaline trend on the AFM diagram (Fig. 6d) of Irvine and kaline field. On the Ga/Al vs. FeO*/MgO diagram of Whalen et al.
Baragar (1971). (1987), the upper unit rhyolite samples have modestly elevated
The new whole-rock major oxide and trace element compositions Ga/Al and plot in the A-type granite field, whereas the other new
are shown in variation diagrams in Figs. 7 and 8. The whole suite lava flow analyses plot in the I- and S-type granite field (Fig. 9b). It
shows continuous trends with restricted scatter in most elements. should be noted that highly fractionated felsic I-type granites can
With increasing silica, compositions progressively decrease in TiO2, have Ga/Al ratios and some major and trace element values that
Al2O3, Fe2O3*, MgO, CaO and P2O5 while increasing in K2O (Fig. 7). overlap those of typical A-type granites (Whalen et al. 1987).
Na2O concentrations are scattered and do not correlate well with silica. Most of the lava flows (basaltic andesite, andesite and dacite) of the
Rb, Y, Ta, Nb, Hf and Th increase whereas Sr decreases with increasing lower succession have low Y and Yb contents and high (Sr/Y) and (La/
silica content; Ba is essentially uncorrelated with SiO2 (Fig. 8). Yb)n ratios. Based on the geochemical definition of “adakite” proposed
The alumina saturation index [ASI = A/CNK, where A, C, N, and K by Defant and Drummond (1990), these Guruf samples have modestly
are respectively the whole-rock molar abundances of Al 2O 3, CaO, to moderately adakitic character (Fig. 9c, d). The more primitive
Na2O, and K2O] of the lower lavas are in the metaluminous range, whole-rock analyses of Maurice et al. (2018) likewise extend into the
from 0.76 to 0.98 (Table 1), while the upper unit spans the transition adakitic field. Similar adakitic volcanic suites have been noted in many
to slightly peraluminous, with ASI from 0.92 to 1.06 (Table 2). The localities in the ANS, where they are strongly associated with post-
agpaitic index (AI = NK/A) of most lava samples is less than 0.87, collisional calc-alkaline activity (Abdelfadil et al. 2018; Be'eri-Shlevin
characteristic of calc-alkaline rocks (Liégeois et al. 1998), except for et al. 2011; Eliwa et al. 2006; Obeid and Azer 2015).
some more alkaline samples from the rhyolite of upper unit, with Primitive mantle-normalized trace-element concentrations of the
0.92 < AI <1.0. Fractionation of alkali feldspar and quartz do not lava samples define a series of related, nearly parallel, patterns with
change AI, but fractionation of plagioclase or other aluminous phases continuous enrichment in most elements from basaltic andesite
can drive increases in AI towards apparent alkaline character within through andesite to rhyolite, coupled with deepening negative anoma-
a single cogenetic suite. On the diagram of Sylvester (1989) for rocks lies in Sr, P, and Ti (Fig. 10a). The data of Maurice et al. (2018) are sim-
with SiO2 > 68 wt%, the rhyodacite of the upper unit straddles the ilar. All the lavas are enriched in large ion lithophile elements (LILE)
boundary between the calc-alkaline and alkaline fields, whereas relative to high field strength elements (HFSE), expressed by troughs
the rhyolite samples lie in the alkaline and highly fractionated in the patterns at Ta and Nb relative to Th and Ce.

10
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 6. Whole-rock chemical classification diagrams: (a) total alkalis vs. silica (TAS) diagram (Cox et al. 1979); the field of Dokhan volcanics is adopted from Azer and Farahat (2011);
(b) SiO2 versus Zr/TiO2 diagram (after Winchester and Floyd 1977); (c) SiO2 versus K2O diagram (after Rickwood 1989); (d) AFM diagram (after Irvine and Baragar 1971).

Concentrations of rare earth elements (REE) and key ratios of REE primitive (50.8–56.0 wt% SiO2) basalt, basaltic andesite, or andesite
concentrations are given in Tables 3 and 4. Chondrite-normalized REE (Fig. 6a, b). They plainly belong to the tholeiitic low-K series (Fig. 6c,
patterns are presented in Figure 10b. Data given by Maurice et al. d). They resemble the island-arc metavolcanic sequences of the ANS
(2018) overlap the new results for the most part; the inclusion of rhyo- (Shadli metavolcanics; Abdel-Karim et al. 2019) and are composition-
litic ignimbrite extends to greater REE enrichment than any of the new ally unrelated to the Dokhan volcanics (contrary to Basta et al. 2017). Al-
lava samples and the Maurice et al. data includes one dacite with anom- though some of the most primitive volcanic analyses of Maurice et al.
alously low LREE. The Guruf volcanic rocks are enriched in LREE relative (2018) resemble the metavolcanic xenolith compositions in selected
to HREE [(La/Yb)n = 5.6–14.5). The basaltic andesite samples have plots, they do not overlap in other plots (e.g., Fig. 6b, c) and their trace
slightly concave-upwards HREE patterns, which evolve to flat HREE pro- element character is totally different.
files in the other lava types. Eu concentrations relative to neighboring The primitive mantle-normalized trace element (Fig. 11a) and
REE decrease through the volcanic suite, from positive anomalies [Eu/ chondrite-normalized REE (Fig. 11b) patterns of the calc-alkaline
Eu* = 1.07–1.33] in basaltic andesite, to slightly negative or small pos- dikes are fully encompassed by the range of Guruf volcanic data,
itive anomalies [Eu/Eu*= 0.98–1.21] in andesite, consistently negative whereas the patterns of the alkaline dikes and metavolcanic xenoliths
anomalies [Eu/Eu* = 0.71–0.96] in dacite and rhyodacite, and pro- are plainly different. The alkaline dikes are highly enriched in ∑REE,
nounced negative anomalies [Eu/Eu* = 0.35–0.73] in rhyolite. while the metavolcanic xenoliths have unfractionated REE and much
lower LREE and LILE concentrations than any Guruf volcanic sample
6.1.2. Dikes and volcanic xenoliths (Fig. 11b, Table 6). The REE patterns of the metavolcanic xenoliths are
Major and trace element analyses of 10 dike samples and 3 samples similar to those of the Shadli metavolcanics (Abdel-Karim et al. 2019).
of metavolcanic xenoliths in granodiorite (Table 5) are presented to
constrain their relationship to the Guruf volcanics. The dikes vary 6.2. Tectonic setting
widely in composition (53.0–76.7 wt% SiO2). The older dikes include
trachyandesite and rhyolite, while the strongly alkaline younger dikes The present and previous studies confirm that the Guruf volcanic
include mugearite, benmoreite, trachyte and rhyolite (Fig. 6a). The suite is a member of the larger calc-alkaline Dokhan volcanic assem-
calc-alkaline dikes belong to the high-K series, whereas the alkaline blage. Hence this work can contribute to resolving the long-running dis-
dikes are shoshonitic (Fig. 6c). The calc-alkaline dikes resemble the crepancy in the literature regarding the tectonic setting of this volcanic
Guruf volcanics in major elements (Fig. 6a-d), whereas the alkaline phase, which has been variously assigned to a subduction environment
dikes are plainly distinct. The metavolcanic xenoliths are all relatively (Abdel Rahman, 1996; Maurice et al. 2018), an extensional setting that

11
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 7. Silica variation diagrams of some major oxides. Model curves, discussed in section 7.4 of the text, are shown as well. The dashed curves in the CaO and P2O5 panels show the
predictions of rhyoliteMELTS 1.0.2 for fractional crystallization along a linear pressure-temperature path from the liquidus at 1155 °C and 300 MPa to 722 °C at 125 MPa, with 2.0 wt%
initial H2O and along the NNO oxygen buffer. The solid curves show a modified model in which 0.0004 wt% apatite is fractionated for each 1 °C of cooling.

12
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 8. Silica variation diagrams of some trace elements. Symbols as in Fig. 7. Model curves are discussed in section 7.4.

post-dates crustal thickening (Mohammed et al., 2000; Abu El-Enen One source of controversy about the tectonic assignment of the
et al. 2018), or a post-collisional transition from subduction to extension Dokhan volcanics is the geochemical overlap in many cases between
(Ressetar and Monard, 1983; Eliwa et al. 2006, 2014; Azer and Farahat subduction-related and post-collisional magmatism. Although Pearce
2011; Obeid and Azer 2015). et al. (1984) proposed a volcanic discrimination diagram, Rb vs.

13
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 9. More whole-rock discrimination diagrams: (a) 100*(MgO + Fe2O3 + TiO2)/SiO2 vs. (Al2O3 + CaO)/(Fe2O3 + Na2O + K2O) (after Sylvester, 1989); (b) Ga/Al vs. FeO*/MgO (after
Whalen et al. 1987), (c) Y vs. Sr/Y (after Defant and Drummond 1990), and (d) (Yb)n vs (La/Yb)n classification diagram (after Defant and Drummond 1990). Symbols as in Fig. 6.

Y + Nb, that separates volcanic-arc granite (VAG) from within-plate elsewhere around the world. Consequently, the persistence of a subduc-
granite (WPG), Pearce (1996) later added a post-collisional granite tion signature in some potential magmatic source region from the end
field that fully overlaps the boundary between VAG and WPG. In the of island arc activity until the early phase of Dokhan volcanism seems
Guruf case, the metavolcanic xenoliths plot exclusively in the VAG re- reasonable. This is supported by the lull in magmatic activity
gion, the lower unit of Guruf volcanics crosses from the upper part of (740–640 Ma) between the end of the subduction-related island-arc
the VAG field into the area of post-collisional granites, and the upper stage and the beginning of post-collisional volcanic activity in most of
unit rhyolites plot within the post-collisional field at the boundary sep- the northernmost ANS (Samuel et al. 2011), although ages spanning
arating VAG from WPG (Fig. 12a). The SiO2-Nb discrimination diagram this gap have been reported from the southern ANS (Ali et al. 2010;
of Pearce and Gale (1977) also provides no clear decision about the Johnson and Kattan 2007). Field relations plainly show that the Guruf
Guruf suite; again, the metavolcanic xenoliths plot as clearly arc- volcanics are younger than the subjacent, subduction-related calc-
related but the volcanic samples straddle the boundary between the alkaline quartz-diorite and granodiorite (Moghazi et al. 2004) and
volcanic arc field and the area of overlap between orogenic and older than the voluminous within-plate alkali feldspar granites. The
anorogenic rocks (Fig. 12b). On the Rb-Hf-Ta diagram of Harris et al. use of metavolcanic xenoliths in the granodiorite to suggest that the
(1986), the Guruf volcanics plot in the volcanic arc field but lie close pluton postdates the Dokhan volcanics (Basta et al. 2017) turns out to
to the boundary of the within-plate field (Fig. 12c). As noted above, be incorrect; these xenoliths plot completely in the volcanic arc field
samples with SiO2 > 68 wt% in the present study and in Maurice et al. are not related to even the most primitive of the Dokhan lavas. Indeed,
(2018) have high AI (>0.87) and plot in Sylvester's (1989) alkaline enough time passed after emplacement of the subduction-related plu-
field (Fig. 9a). These features are not observed in subduction-related tonic rocks for them to be unroofed, exposed, and covered unconform-
volcanic sequences in the ANS; they are characteristic instead of the ably by surface lava flows of the Guruf series.
post-collisional calc-alkaline volcanic units (Be'eri-Shlevin et al. 2011). Maurice et al. (2018) argued that the ages of younger Dokhan felsic
Authors that have adopted a subduction-related tectonic setting for rocks (Abu El-Enen et al. 2018; Breitkreuz et al. 2010) do not support di-
the Dokhan volcanics exposed around G. Um Guruf have reasoned viding the Dokhan volcanic rocks in the North Eastern Desert into two
mainly from their chemical compositions, which are indeed in many successive but related cycles. They focus on a 5 m.y. gap between the
regards similar to those of typical Andean calc-alkaline volcanic rocks youngest dated members of the lower cycle and the oldest dated mem-
(Maurice et al. 2018). However, the high-K calc-alkaline character and ber of the upper cycle as evidence that the cycles are unrelated. Instead,
other traits previously interpreted to indicate arc magmatism for the they group the lower cycle with a tectonic setting that is thought to have
Dokhan volcanics may simply reflect remelting of arc material or ended tens of m.y. earlier. Thus, they propose to restrict use of the term
enriched lithospheric mantle from the earlier subduction period of “Dokhan volcanics” to the calc-alkaline andesitic to dacitic volcanic
ANS evolution. Similar inheritance is observed in post-collisional plu- rocks, excluding the later rhyolites. Their conclusion that there was per-
tonic rocks in the ANS (Azer et al. 2012; Gahlan et al. 2016) and sistent subduction at this location until the time of emplacement of the

14
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 10. Normalized trace-element diagrams for lava flow: (a) primitive-mantle normalized extended trace element patterns and (b) CI-Chondrite normalized REE plots. Primitive mantle
and CI-Chondrite normalization values are from Pearce (1983) and Sun and McDonough (1989), respectively. Symbols as in Fig. 6. (c) Results of the rhyoliteMELTS fractionation model for
REE patterns. The dashed lines show the nominal model results, whereas the solid lines show the result of added apatite fractionation.

Dokhan volcanics is then based only on the lower unit samples. This is Yb)n proxy, 50 km, is significantly greater than the modern depth to
contrary to the general consensus among Egyptian scientists studying the seismic Moho in the ANS (23–39 km; Al-Damegh et al. 2005). It is
the Dokhan volcanics since 1934 that the sequence can be subdivided likely that the depletion of Yb due to the adakitic character of the
into upper and lower units, with the upper one consisting essentially lower Dokhan sequence in the area (Abdelfadil et al. 2018; Be'eri-
of rhyolite (Abdelfadil et al. 2018; El-Gaby et al. 1991; Obeid and Azer Shlevin et al. 2011; Eliwa et al. 2006; Obeid and Azer 2015) leads to
2015). Moreover, rhyolite is present in the lower Dokhan volcanic series an exaggerated estimate of crustal thickness. Finally, Maurice et al.
(Be'eri-Shlevin et al. 2011). Rhyolite with A-type character is observed (2018) applied some coarse tectonic discrimination diagrams based
in many of the calc-alkaline Dokhan volcanic exposures in the Eastern on the chemistry of clinopyroxene (Leterrier et al. 1982) that are not ad-
Desert and Sinai, especially in the upper unit (Be'eri-Shlevin et al. 2011). equate for distinguishing active subduction island arc activity from
The data of Maurice et al. (2018) extend from the island arc field into post-collisional continental margin activity.
the continental margin arc field in several trace element ratio diagrams High-precision U\\Pb dating for subduction-related volcanic rocks
(their Fig. 13), which they attributed to temporal variation in island arc in the Eastern Desert shows that their ages range from 720 Ma to
crust thickness over the lifetime of Dokhan volcanic activity, dismissing 770 Ma (Ali et al. 2010). Recent U\\Pb dating of the post-collisional
on isotopic grounds the possibility of a continental margin (Andean phase in the ANS includes both plutonic rocks and their volcanic equiv-
type) arc setting. However, the references invoked to support the ab- alents and ranges between 630 and 590 Ma (Be'eri-Shlevin et al. 2011;
sence of old continental crust in the Eastern Desert and the low initial Moussa et al. 2008). U\\Pb zircon dating of the Dokhan Volcanics indi-
87
Sr/86Sr ratios of the Dokhan volcanic rocks (Ali et al. 2012; Liégeois cates that the early pulse began at 630 Ma, whereas the later pulse
and Stern 2010) are incomplete. Other authors report abundant isotopic ended around 590 Ma (Abu El-Enen et al. 2018; Be'eri-Shlevin et al.
evidence for involvement of old continental crust underlying the juve- 2011; Breitkreuz et al. 2010; Wilde and Youssef 2000). Many authors
nile crust of the northern ANS (e.g., Be'eri-Shlevin et al. 2011; have emphasized that there is no evidence for subduction or collisional
Hargrove et al. 2006; Li et al. 2018; Samuel et al. 2011; Wilde and related processes in the ANS later than ~635–620 Ma (e.g., Avigad and
Youssef 2002). Moreover, the arc crust thickness at the end of the Gvirtzman 2009; Eyal et al. 2010), which suggests that the Dokhan vol-
lower Dokhan cycle inferred by Maurice et al. (2018) using the (La/ canics post-date the subduction period. Therefore, there is a general

15
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 10 (continued).

consensus that the Dokhan volcanics formed in a post-collisional tec- the field of Dokhan volcanics (Fig. 6a). Possibly they represent a lower
tonic setting (Be'eri-Shlevin et al. 2011; Azer and Farahat 2011, and stratigraphic unit unrelated to the Guruf volcanics; not enough field
many others). Also, the two pulses of volcanic activity in the Dokhan context is presented to be sure. In any case, they are highly altered;
volcanics are comparable with the two magmatic phases of post- their low apparent SiO2 is the result of their high LOI. The intermediate
collisional plutonic activity: an early magmatic phase at ca. and silicic magmas of the Guruf successions display continuous trends
635–620 Ma and a later one at 610–590 Ma (Be'eri-Shlevin et al. without any compositional gaps (Figs. 7, 8), suggesting a common
2011; Eyal et al. 2010). All these criteria support a post-collisional set- source. Also, they have closely related and progressively evolving trace
ting for the studied volcanics rather than an active subduction setting. elements and REE patterns (Fig. 10a, b).
Maurice et al. (2018) distinguished the Dokhan volcanic rocks of the
7. Discussion W. Hamad area, including the Guruf volcanics, into an early pulse dom-
inated by basalt and andesite and a later pulse dominated by dacite. On
7.1. Petrogenesis the basis of the basaltic samples, Maurice et al. (2018) judged that the
primary magmas of the entire sequence must be mantle-derived and
The Guruf volcanics include intermediate to silicic volcanics and lack described it as unlikely that subduction-influenced mantle astheno-
basalts, a feature common to sequences assigned to the Dokhan volcanic sphere could persist beneath the ANS for tens of m.y. beyond the end
series in the Eastern Desert; all the Guruf samples plot within the of subduction. The evolving trace element character of the suite was
Dokhan field in the TAS diagram (Fig. 6a). The handful of basalts re- taken to indicate an evolution in arc maturity. However, as the earlier is-
corded by Maurice et al. (2018) are puzzling; some of them plot outside land arc stage in the Eastern Desert of Egypt evolved from immature to

Table 3
REE concentrations of the lower lava flows of Guruf volcanics, North Eastern Desert, Egypt.

Rock type Basaltic andesite Andesite Dacite

Sample GH22 GH23 GH26 GH27 GH28 GH30 GH31 GH32 GH33 GH34 GH35 GH36 GH2a GH3a GH13a GH24 GH25 GH29

La 16.72 19.26 21.28 21.89 22.94 22.54 26.63 26.42 28.29 23.94 23.32 24.20 37.34 29.26 26.20 25.70 27.73 30.77
Ce 35.46 40.49 45.04 47.77 53.20 46.99 59.10 58.90 62.43 52.20 51.56 47.94 77.47 57.21 47.54 56.86 54.64 55.73
Pr 4.59 5.15 5.57 6.21 6.98 6.04 7.92 7.89 8.10 6.76 6.44 6.04 9.16 7.05 5.58 7.25 6.61 7.21
Nd 18.81 21.89 22.91 24.16 31.11 25.11 34.19 34.09 34.79 29.66 26.49 23.62 34.98 25.77 19.46 28.56 25.06 27.70
Sm 4.31 4.81 4.85 5.34 6.64 4.93 6.97 6.95 6.87 6.27 4.94 4.57 6.03 4.58 3.67 6.31 5.82 5.34
Eu 1.56 1.46 1.43 1.71 1.70 1.61 1.77 1.65 1.80 1.69 1.54 1.49 1.77 1.29 0.86 1.49 1.36 1.26
Gd 2.97 3.29 3.46 3.64 3.56 3.37 4.13 3.83 3.94 3.72 3.74 3.29 5.22 4.09 2.95 4.18 3.98 3.85
Tb 0.34 0.39 0.42 0.45 0.50 0.41 0.56 0.52 0.53 0.53 0.56 0.49 0.77 0.59 0.46 0.58 0.64 0.59
Dy 1.76 2.10 2.18 2.11 2.65 2.48 3.34 3.14 2.94 2.82 2.91 2.80 4.58 3.43 2.65 3.34 3.45 3.65
Ho 0.34 0.39 0.38 0.41 0.49 0.47 0.64 0.59 0.54 0.49 0.59 0.49 0.91 0.66 0.52 0.68 0.69 0.69
Er 0.91 0.99 1.00 1.03 1.31 1.18 1.61 1.50 1.40 1.29 1.39 1.31 2.64 1.95 1.51 1.67 1.67 1.70
Tm 0.13 0.14 0.14 0.14 0.20 0.16 0.23 0.22 0.22 0.20 0.21 0.19 0.39 0.28 0.23 0.26 0.24 0.23
Yb 0.90 0.98 0.99 1.02 1.37 1.11 1.58 1.47 1.40 1.25 1.29 1.31 2.61 1.90 1.58 1.65 1.66 1.65
Lu 0.14 0.15 0.15 0.16 0.20 0.16 0.23 0.21 0.23 0.20 0.21 0.18 0.42 0.29 0.24 0.25 0.24 0.23
Eu/Eu* 1.33 1.12 1.07 1.19 1.07 1.21 1.01 0.98 1.06 1.07 1.10 1.17 0.96 0.91 0.80 0.89 0.86 0.85
(La/Yb)n 12.56 13.29 14.54 14.51 11.34 13.77 11.43 12.18 13.71 12.94 12.25 12.45 9.67 10.44 11.18 10.55 11.32 12.63
(La/Sm)n 2.45 2.53 2.77 2.59 2.18 2.89 2.41 2.40 2.60 2.41 2.98 3.34 3.91 4.03 4.51 2.57 3.01 3.64
(Gd/Lu)n 2.60 2.69 2.83 2.79 2.21 2.55 2.25 2.27 2.15 2.30 2.21 2.24 1.54 1.72 1.52 2.03 2.01 2.02
(La/Lu)n 12.24 13.16 14.54 14.02 11.87 14.26 12.13 13.08 12.88 12.39 11.54 13.78 9.19 10.32 11.26 10.45 11.69 13.47
a
Samples analyzed in Caltech (USA).

16
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Table 4
REE concentrations of the upper lava flows of Guruf volcanics, North Eastern Desert, Egypt.

Rock type Rhyodacite Rhyolite

Sample No. GH1a GH8a GH12a GH18a Gh4a GH10a GH15a GH39 GH40 GH42 GH45

La 34.28 30.21 30.08 31.83 29.66 26.39 25.21 38.52 34.95 35.09 40.75
Ce 75.32 69.18 68.29 74.80 73.56 68.68 72.57 82.35 74.40 75.63 84.18
Pr 8.34 7.50 7.40 7.89 7.97 7.13 6.67 10.41 9.48 9.67 10.67
Nd 32.10 27.62 27.81 28.71 30.58 27.24 25.05 37.20 35.36 39.89 37.12
Sm 5.67 4.90 5.17 5.58 5.86 5.22 5.11 7.75 7.43 7.90 7.49
Eu 1.62 1.11 1.13 1.16 0.88 0.88 0.55 1.09 0.99 1.24 1.27
Gd 4.87 4.28 4.39 4.47 5.02 4.99 4.55 5.85 5.40 6.29 5.79
Tb 0.71 0.66 0.66 0.69 0.79 0.81 0.72 0.89 0.82 0.95 0.91
Dy 4.32 3.95 3.86 4.05 4.72 5.21 4.25 5.39 5.59 5.26 5.49
Ho 0.83 0.77 0.75 0.82 0.96 1.06 0.88 1.09 1.05 1.03 1.13
Er 2.38 2.25 2.30 2.45 2.96 3.03 2.58 3.12 2.97 2.82 3.18
Tm 0.36 0.35 0.35 0.37 0.45 0.47 0.40 0.47 0.45 0.42 0.48
Yb 2.40 2.38 2.27 2.48 3.00 3.19 2.69 3.08 2.94 2.82 3.16
Lu 0.37 0.38 0.36 0.38 0.49 0.49 0.42 0.47 0.46 0.41 0.48
Eu/Eu* 0.94 0.74 0.73 0.71 0.50 0.52 0.35 0.49 0.48 0.54 0.59
(La/Yb)n 9.67 8.60 8.94 8.68 6.68 5.59 6.35 8.46 8.04 8.41 8.72
(La/Sm)n 3.82 3.89 3.67 3.60 3.20 3.19 3.12 3.14 2.97 2.80 3.43
(Gd/Lu)n 1.60 1.39 1.50 1.45 1.27 1.26 1.34 1.53 1.44 1.88 1.48
(La/Lu)n 9.40 8.19 8.58 8.62 6.26 5.55 6.21 8.40 7.78 8.77 8.70
a
Samples analyzed in Caltech (USA).

Table 5
Chemical composition of the dykes and volcanic xenoliths associated with Guruf volcanics, North Eastern Desert, Egypt.

Rock type Dykes Xenoliths

Alkaline dykes Calc-alkaline dykes

Sample No. GH9a GH14a Gh20a GH11a GH19a GH7a Gh17a Gh5a GH16a GH6a Xen.1 Xen.6 Xen.9

SiO2 61.26 52.99 63.62 76.73 69.33 72.65 60.13 69.87 70.51 68.02 56.02 50.78 52.81
TiO2 0.95 1.92 0.82 0.12 0.32 0.35 0.94 0.33 0.21 0.41 1.03 1.18 1.12
Al2O3 17.18 17.80 17.02 11.92 11.67 14.11 16.75 14.36 13.71 14.54 16.06 16.14 16.09
Fe2O3 4.29 7.73 3.75 1.00 3.86 1.93 5.35 1.82 1.51 2.65 11.31 12.24 12.11
MnO 0.13 0.10 0.12 0.03 0.09 0.07 0.09 0.05 0.07 0.07 0.21 0.28 0.25
MgO 0.86 1.35 0.76 0.04 0.42 0.44 1.28 0.47 0.06 0.84 4.59 5.54 4.98
CaO 2.69 4.72 2.35 0.26 0.41 0.71 4.10 0.82 1.28 1.70 4.93 8.48 6.17
Na2O 5.05 4.68 4.99 4.01 5.10 4.63 5.30 4.48 4.01 4.29 3.26 2.71 3.01
K2O 4.91 4.32 5.05 4.21 4.53 4.41 2.28 4.72 4.51 3.73 0.36 0.27 0.31
P2O5 0.39 1.04 0.29 0.01 0.03 0.08 0.31 0.08 0.03 0.12 0.19 0.27 0.21
LOI 0.30 0.91 0.31 0.45 0.52 0.43 1.08 0.82 1.49 0.92 1.63 2.06 1.97
Total 98.01 97.55 99.07 98.77 96.27 99.80 97.63 97.80 97.40 97.29 99.59 99.95 99.03
Sc 0.96 3.04 1.23 1.84 0.74 0.91 2.81 1.76 1.61 1.80 39.40 42.70 41.30
Rb 51.29 98.42 61.02 231.56 118.17 116.51 51.30 156.68 134.95 88.60 10.10 4.40 6.20
Ba 3140.09 1896.32 2467.76 232.41 225.94 744.66 665.15 831.32 965.00 808.78 166.60 107.30 125.80
Sr 443.92 1648.45 361.44 89.88 29.78 175.88 651.23 237.80 115.04 390.88 281.40 176.50 211.10
Nb 20.95 22.12 22.74 48.77 102.87 12.13 10.60 11.12 13.07 8.73 3.80 2.40 3.20
Zr 272.19 449.70 294.43 207.20 657.47 228.57 296.04 215.90 240.17 195.74 60.30 52.50 56.40
Hf 5.89 7.63 6.61 10.01 17.93 6.56 7.33 5.99 7.18 5.49 2.20 1.70 2.10
Y 30.45 27.96 32.03 58.79 81.44 22.11 24.81 20.58 26.85 15.28 29.60 24.10 25.40
Zn 81.92 84.48 84.44 49.45 141.54 37.31 80.17 31.85 29.96 110.93 72.40 86.40 77.40
Cu <d.l. 3.40 <d.l. <d.l. <d.l. <d.l. <d.l. <d.l. 1.52 <d.l. 399.50 99.40 84.00
Ni <d.l. 23.34 <d.l. 1.89 21.21 56.82 7.24 4.48 <d.l. 3.32 19.40 22.50 20.50
Co 23.31 34.14 28.00 27.95 2.37 22.41 61.78 33.19 20.27 50.17 42.20 55.30 44.20
Cr 8.74 15.48 19.04 16.67 46.00 121.58 34.36 30.88 10.06 36.37 14.60 50.20 29.60
V 22.60 141.87 16.04 4.76 15.86 21.37 58.94 17.27 <d.l. 38.87 338.90 372.10 363.70
Pb 13.54 34.10 13.86 14.92 7.68 7.39 20.61 5.16 42.02 113.74 14.70 13.80 10.20
Th 4.18 3.32 5.65 36.44 5.56 11.87 7.39 10.24 6.70 10.82 0.90 1.70 1.20
Ta 1.00 1.57 1.12 3.43 6.42 0.88 0.55 0.82 0.91 0.90 0.20 0.10 0.20
U 2.00 1.56 2.45 7.88 3.25 4.18 2.22 4.33 4.81 4.61 0.40 0.30 0.20
Mo 1.51 1.16 2.17 0.54 1.36 1.89 1.24 1.27 1.74 3.78 1.80 1.40 1.50
Ga – – – – – – – – – – 14.20 11.70 13.60
As – – – – – – – – – – 0.86 0.93 1.01
Cs – – – – – – – – – – 0.48 0.52 0.64
Mg# 28.46 25.69 28.64 7.68 17.68 31.17 32.17 33.97 7.84 38.61 44.57 47.28 44.89
AI 0.79 0.70 0.80 0.94 1.14 0.88 0.67 0.87 0.84 0.76 0.36 0.29 0.33
ASI 0.93 0.85 0.95 1.03 0.83 1.03 0.90 1.03 0.99 1.02 1.09 0.80 0.97
Sr/Y 14.58 58.96 11.28 1.53 0.37 7.95 26.25 11.56 4.29 25.59 9.51 7.32 8.31
a
Samples analyzed in Caltech (USA).

17
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Fig. 11. Normalized trace-element diagrams for the dike and xenolith samples from the study area: (a) primitive-mantle normalized trace element patterns and (b) CI-Chondrite
normalized REE plots. Symbols as in Fig. 6.

mature, it remained stubbornly low-K (Abdel-Karim et al. 2019), unlike The most primitive samples (basaltic andesite) in the Guruf volca-
the medium- to high-K character of the Guruf volcanics. Maurice et al. nics have Mg# ≤ 56, Ni ≤90 μg/g, and Cr ≤ 317 μg/g. Primary basaltic
(2018) reasoned from inconsistencies between the observed evolution melts in equilibrium with residual mantle are expected to have high
of Mg# and REE abundances and those expected for fractional crystalli- Ni and Cr contents (> 250 and > 1000 μg/g, respectively) and Mg# in
zation to infer an evolving source within the suite. The Mg# results may the range 68–75 (Roeder and Emslie, 1970). Hence, there are no candi-
have been affected by strong alteration of the basalts, by magma mixing, dates for a primary mantle-derived melt among the Guruf rocks.
or by combining unrelated stratigraphic units. The slowly evolving REE In the present work and many published reports (Be'eri-Shlevin
contents during differentiation are common in all post-collisional plu- et al. 2011; Obeid and Azer 2015), REE patterns and other features sug-
tonic and volcanic rocks in the ANS (Be'eri-Shlevin et al. 2011; Eyal gestive of subduction-related magmatism have been interpreted to re-
et al. 2010); we model this feature below by appealing to apatite frac- flect inheritance due to remelting of earlier arc-related material from
tionation. Based on the present field work and geochemical data, we the pre-collisional stage (850–740 Ma). Therefore, the proposed
find that the Guruf volcanic suite lacks basaltic members and was model for the genesis of the Guruf volcanics is partial melting of lower
emplaced in two cycles of eruption of medium- to high-K calc-alkaline continental crust followed by ascent and fractionation at upper crustal
volcanic rocks derived from a common source levels. The most primitive members of the Guruf volcanics represented

18
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

Table 6
REE concentrations of the dykes and volcanic xenoliths associated with Guruf volcanics, North Eastern Desert, Egypt.

Rock type Dykes Xenoliths

Alkaline dykes Calc-alkaline dykes

Sample GH9a GH14a Gh20a GH11a GH19a GH7a Gh17a Gh5a GH16a GH6a Xen.1 Xen.6 Xen.9

La 39.46 43.29 42.86 21.96 32.30 29.83 36.09 26.96 21.82 29.58 1.66 2.25 1.69
Ce 107.52 121.52 118.01 61.45 134.61 70.60 74.88 66.08 65.16 61.36 4.64 6.03 4.98
Pr 10.70 12.09 11.08 7.85 8.27 7.22 9.29 6.54 5.69 6.64 0.73 0.95 0.81
Nd 42.75 47.25 42.29 29.64 32.05 25.90 36.11 23.64 21.57 23.51 3.79 4.61 4.11
Sm 7.40 8.24 7.80 8.17 7.34 4.44 6.78 4.31 4.31 3.77 1.21 1.48 1.32
Eu 3.58 2.54 3.04 0.09 1.10 0.73 1.73 0.75 0.76 0.95 0.56 0.56 0.61
Gd 6.34 6.37 6.29 8.15 7.78 3.62 5.73 3.45 3.82 3.12 1.77 1.6 1.83
Tb 0.87 0.80 0.89 1.51 1.31 0.56 0.80 0.52 0.60 0.45 0.31 0.28 0.32
Dy 4.81 4.21 4.82 9.42 8.44 3.49 4.37 2.99 3.67 2.56 2.1 1.91 2.17
Ho 0.89 0.79 0.92 1.90 1.70 0.69 0.83 0.59 0.74 0.51 0.46 0.46 0.47
Er 2.36 2.14 2.46 5.80 5.14 1.94 2.38 1.79 2.17 1.37 1.42 1.26 1.46
Tm 0.33 0.29 0.35 0.89 0.77 0.29 0.33 0.27 0.34 0.21 0.21 0.19 0.21
Yb 2.06 1.97 2.28 5.99 5.08 2.00 2.13 1.80 2.26 1.44 1.47 1.25 1.47
Lu 0.32 0.29 0.34 0.92 0.78 0.33 0.30 0.27 0.36 0.23 0.21 0.16 0.2
Eu/Eu* 1.60 1.07 1.33 0.03 0.44 0.55 0.85 0.60 0.58 0.84 1.17 1.11 1.20
(La/Yb)n 12.93 14.86 12.69 2.48 4.30 10.07 11.48 10.12 6.52 13.90 0.76 1.22 0.78
(La/Sm)n 3.36 3.32 3.47 1.70 2.78 4.24 3.36 3.95 3.20 4.95 0.87 0.96 0.81
(Gd/Lu)n 2.39 2.71 2.26 1.09 1.23 1.36 2.31 1.56 1.29 1.67 1.03 1.23 1.12
(La/Lu)n 12.46 15.41 12.89 2.45 4.27 9.38 12.18 10.19 6.16 13.19 0.81 1.44 0.87
a
Samples analyzed in Caltech (USA).

in our sample suite are basaltic andesites with ~54 wt% SiO2 and ~ 5 wt% formation have been suggested (for review see Castillo 2012). Gener-
MgO; a plausible model for the origin of the suite must define conditions ally, such models divide into two main groups: (1) partial melting of
where such a liquid can coexist with residual amphibole. Such a condi- an oceanic slab in an active subduction-related environment (e.g.
tion was identified in the experimental study of Carroll and Wyllie Defant and Drummond 1990; Zhu et al. 2009) and (2) non-slab origins,
(1989), who found that a Sierra Nevada (USA) tonalite (a continental wherein the adakite signature may be derived in garnet- or amphibole-
arc rock) mixed with 10 wt% peridotite, in the presence of 10 wt% bearing upper plate crust, by differentiation processes (such as frac-
H2O, has hornblende near the liquidus at 1.5 GPa and up to ~950 °C. tional crystallization, magma mixing, or crustal contamination) or by
Below this temperature, the MgO in the liquid rapidly decreases; anatexis of pre-existing crustal materials (Castillo 2012; Macpherson
above this temperature, hornblende melts out. At higher pressure, gar- et al. 2006; Obeid and Azer 2015).
net occurs; at lower pressure, plagioclase becomes dominant. The TP10 Many of the post-collisional calc-alkaline magmas in the ANS have
composition studied by Carroll and Wyllie is a remarkably close match adakitic affinity (Abdelfadil et al. 2018; Be'eri-Shlevin et al. 2011;
to the average Guruf basaltic andesite (Fig. 7). We suggest that it offers Eliwa et al. 2006; Hargrove et al. 2006; Obeid and Azer 2015). Mirroring
a guide to the conditions of formation of the primary magma for the the global debate, many fundamentally different models for the origins
Guruf volcanics: beneath the ~30 km ANS crust, a lower crustal delam- of the ANS adakitic rocks have been suggested (Abdelfadil et al. 2018;
ination event led to sinking arc-affinity tonalite being mixed with as- Be'eri-Shlevin et al. 2011; Eliwa et al. 2006). Models differ in the ex-
thenospheric peridotite and experiencing heating as it sank. Near pected time of the events that triggered melting: ≥ 630 Ma for the
950 °C and 1.5 GPa, a liquid of basaltic andesite composition separated subducted ridge model (Avigad and Gvirtzman 2009) versus
from an amphibole-bearing but plagioclase- and garnet-free residue 610–600 Ma for delamination (Eyal et al. 2010; Moghazi 2003). They
and ascended back into the crust. differ also in the source components; e.g., a broken oceanic slab
The extrusion of the Guruf volcanics was associated with exten- (Moghazi 2003) versus mafic lower continental arc crust (Avigad and
sive denudation of the pre-630 Ma orogenic edifice as well as the Gvirtzman 2009; Eyal et al. 2010).
630–600 Ma post-collisional products. The vertical displacements that Eliwa et al. (2006) suggested that a hot oceanic ridge was subducted
drove such erosion may have been surface expressions of isostatic re- (pre- or syn-collision) and then melted some 20 m.y. later, in a
bound due to thinning (and likely delamination) of the previously post-collisional regime, yielding the post-collisional adakitic volcanic
thickened lithosphere, facilitating the upwelling of hot asthenospheric rocks of southern Israel and the Eastern Desert. The lack of evidence
material (e.g., Avigad and Gvirtzman 2009) and the rapid elevation of for oceanic crust younger than ca. 690 Ma in the northern ANS (Stacey
the lower crustal geotherm. This may have been the trigger for the tran- et al. 1984) does not support this model, even if the various adakitic
sition into dominantly alkaline A-type magmatism at 600 Ma. post-collisional volcanic series are all assumed to be as old as ca.
630 Ma, which is probably not the case.
7.2. Geodynamic significance of adakitic lavas The Guruf adakitic volcanics have geochemical characteristics associ-
ated with upper plate crustal sources or interactions, which is con-
Adakites were originally interpreted as low-degree partial melts of sistent with a post-collisional setting: Mg# (<56), Cr (<10 μg/g), and
subducted oceanic crust that interacted with mantle peridotite at high Ni (<90 μg/g) contents are lower than those expected for primary adakite
pressure, leaving a garnet-rich but plagioclase-poor residue that yields magma derived from partial melting of subducted oceanic crust that has
high Sr/Y and La/Yb ratios (Defant and Drummond 1990). However, only interacted with mantle wedge peridotite (e.g. Zhu et al. 2009). The
the definition of adakite is geochemical rather than petrogenetic need for crustal residence and evolution favors a model whereby the
(Richards and Kerrich 2007) and volcanic rocks meeting the definition adakitic characteristics were acquired in the upper plate crust and
of adakite have been identified in a range of tectonic environments as- makes it quite challenging to trace them back to primary slab or mantle
sociated with a corresponding range of magmatic sources and pro- derivation. Furthermore, the present volcanic suite, when plotted as ei-
cesses. Therefore, the sources of individual suites of adakitic rocks ther MgO (Condie 2005) or Mg# (Zhu et al. 2009) versus SiO2, does not
have been debated extensively and a number of models for their plot in the field of slab melting-related adakites (Fig. 13a, b).

19
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

delamination (Gao et al. 2004). This mechanism is not feasible in situ

where crust is less than ~30 km thick, typically the minimum depth of
garnet amphibolite or eclogite-facies metamorphism (Macpherson
et al. 2006). Garnet is increasingly abundant in mafic magmas at
≥40 km depth (Richards and Kerrich 2007) and increasingly likely to re-
main in the residue and retain HREE and Y after significant degrees of
partial melting. The crust beneath the ANS is essentially mafic and has
(today) an average thickness of 39 km in Saudi Arabia, rapidly thinning
to 23–25 km along the Red sea coast and the margin of the Gulf of Aqaba
(Al-Damegh et al., 2005). That is, the ANS crust is at the present day only
marginally thick enough for garnet to be a significant residual phase at
the base of the crust and there is limited evidence for major crustal thin-
ning associated with post-Neoproterozoic activity. Moreover, the mod-
erately fractionated REE patterns of the Guruf volcanics (Fig. 10b), with
undepleted and nearly flat HREE slopes, are inconsistent with batch
melting of an eclogite source (e.g., Stern and Gottfried 1986). Therefore,
the most primitive observed basaltic andesite is unlikely to have been
derived directly from partial melting of garnet amphibolite or eclogite
facies rocks in-place in stable thickened lower crust.
A simpler model that would explain the post-collisional adakitic
rocks of the ANS in general, and the Guruf volcanism in particular, is
melting of delaminated lower crust (Kay and Kay 1993). Heating of
delaminated lower crust in the mantle would explain the high degrees
of melting needed to produce basaltic andesites (requiring up to 25%
melting of eclogite; e.g., Stern and Gottfried 1986). Independent
geodynamic arguments in favor of delamination of lower crust in the
ANS at this time (630–590 Ma) have been recently discussed by many
authors (Abdelfadil et al. 2018; Avigad and Gvirtzman 2009; Farahat
and Azer 2011; Moghazi 2003; Obeid and Azer 2015). The melting of
delaminated lower crust is more consistent with the tectonic setting
of the post-collisional adakitic volcanic sequences than the oceanic
ridge model of Eliwa et al. (2006). The geochemical signatures of
adakitic magma require sufficiently high pressures and water contents
to suppress residual plagioclase (Macpherson et al. 2006).
Delamination and the ensuing isostatic rebound would trigger an
extensional tectonic regime that would open conduits for small-
degree alkaline magmas and, at the surface, cause development of
basins to provide accommodation space for volcano-sedimentary accu-
mulation. Indeed, there is a record of peak extensional tectonics in the
northernmost ANS at ca. 600 Ma (Stern et al., 1984). The proposed
geodynamic model for generation of the studied volcanic sequences in-
volved the following steps, beginning some tens of m.y. after the end of
subduction in the northern ANS: (1) delamination of lithospheric man-
tle and tonalitic lower arc crust, triggering (2) partial melting of the
delaminated lower crust as it undergoes amphibolite-grade metamor-
phism and heating by intimate mixing with ascending asthenosphere,
followed by (3) the onset of extension and subsidence, (4) ascent of ba-
saltic andesite together with fractional crystallization to form the andes-
ite and dacite of the lower Guruf volcanics, and (5) after a hiatus and
further fractionation, renewed eruption of the upper felsic volcanic
cycle of rhyodacite and rhyolite.

7.3. Physicochemical conditions of evolution and alteration

Mafic igneous minerals are sensitive to variations in both the chem-

istry and the P-T-aH2O-fO2 conditions of crystallization and so can be
Fig. 12. Tectonomagmatic diagrams: (a) Yb + Nb vs. Rb (Pearce et al. 1984), with post- used in some cases to estimate the P-T conditions early in magmatic
collisional granite field after Pearce (1996); (b) SiO2 versus Nb (after Pearce and Gale evolution (e.g., Lindsley 1983; Ridolfi et al. 2010). The Wo-En-Gs
1977); and (d) Hf-Rb/30-3*Ta ternary (after Harris et al. 1986). Symbols as in Fig. 6. clinopyroxene quadrilateral was contoured for temperature estimation
by Lindsley (1983). The present data, drawn from augite in the basaltic
andesite and andesite samples, plot mainly between the 800 and
According to Be'eri-Shlevin et al. (2011), post-collisional adakitic 1000 °C contours (Fig. 5a).
volcanic rocks of the ANS simply require derivation from a source The compositions of primary hornblende have been used by
with garnet or amphibole in the residue. Thickened, mafic, lower conti- many authors for geothermometry, oxybarometry, and geobaro-
nental crust may be partially melted (e.g., Moyen and Martin 2012) in metry of igneous rocks (e.g., Ridolfi et al. 2010). Application of the
the eclogite or garnet amphibolite stability fields, perhaps during temperature-sensitive Si-index for amphibole from Ridolfi et al.

20
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

make in this work between the minor low-temperature alteration that

has affected the Guruf volcanics and the greenschist-facies or higher
metamorphism that is characteristic of the subduction-related rocks of
the ANS.

7.4. Fractional crystallization

The systematic variation of major and trace element contents of the

Guruf volcanics (Figs. 7, 8) can be interpreted in terms of magmatic dif-
ferentiation processes. Decreasing Fe2O3*, MgO, CaO, TiO2, P2O5 and Sr
coupled with increasing Zr, Rb, Y, Nb contents are all consistent with
fractionation of plagioclase, apatite and Fe\\Ti oxides. Likewise, the in-
creasing development of strong negative P, Ti, and Sr anomalies in the
normalized trace element patterns of dacite, rhyodacite and rhyolite
samples (Fig. 10a) indicates fractionation of these phases. The progres-
sion from positive Eu anomalies in basaltic andesite (suggesting the ab-
sence of residual plagioclase in the source), to small negative Eu
anomalies in dacite and onward to large negative Eu anomalies in the
rhyolite also indicates steady removal of Ca-bearing phases such as pla-
gioclase and apatite.
Middle and heavy-REE patterns are good indicators of the roles of
amphibole and garnet, respectively, in a magmatic system (Castillo
2012; Defant and Drummond 1990). The concave-up REE patterns of
basaltic andesite (Fig. 10b) imply a role for amphibole, likely as a resid-
ual phase in the source (Be'eri-Shlevin et al. 2011). The increases in
HREE with evolution and the persistence of low MREE/HREE requires
the absence of garnet fractionation and allows the possibility of ongoing
amphibole fractionation from a hydrous parental magma.
Inspired by the presence of magmatic amphibole, by the evidence of
pressures decreasing from ~300 MPa to ~125 MPa during evolution, and
by the high fO2 recorded by the amphibole, we constructed a quantita-
tive model of evolution using the rhyoliteMELTS 1.0.2 model (Ghiorso
and Gualda, 2013) and the easyMelts 0.2.4 and alphaMELTS 1.9 inter-
faces. The model is assumed to begin from the average composition of
the basaltic andesite samples on its liquidus (1155 °C) at 300 MPa,
with 2 wt% dissolved H2O on the nickel‑nickel oxide (NNO) oxygen
buffer. Fractionation proceeds in 1 °C cooling steps with 0.4 MPa of de-
compression at each step, terminating at 725 °C and 125 MPa. The liquid
Fig. 13. Discrimination diagrams for different adakite source models: (a) MgO vs. SiO2, line of descent generated by this simple model captures many signifi-
(b) Mg# vs. SiO2. Crustal AFC curve according to Stern and Kilian (1996). Fields cant features of the major and trace element evolution of the suite
indicating mantle melt, adakites related to slab melting, lower-crustal melting, and (Figs. 7, 8). In particular, the steepness of enrichment of incompatible el-
delaminated lower crust-derived adakites are according to Gao et al. (2004), Condie
ements is matched or exceeded by the model with no need for assimila-
(2005), and Zhu et al. (2009). Symbols as in Fig. 6.
tion of an enriched component. However, there are four features that

(2010) gives crystallization temperature of primary hornblende

807–944 °C in basaltic andesite, 814–855 °C in andesite, and
781–835 °C in dacite, with typical uncertainty of ±22 °C. The estimated
pressures are 120–345 MPa in basaltic andesite, 108–346 MPa in andes-
ite, and 91–145 MPa in dacite. Despite significant uncertainty, the am-
phibole compositions appear to record progressive cooling and ascent
of the magma as it evolves. The oxygen-sensitive Fe/(Fe + Mg) versus
AlIV amphibole discrimination diagram (Anderson and Smith 1995)
shows that the primary amphiboles in the Guruf volcanics crystallized
under high fO2 conditions (Fig. 14).
The observed compositions of chlorite depend, most probably, on
the nature of their precursor ferromagnesian minerals and the
conditions of alteration (temperature, composition, and redox state of
aqueous solutions), as discussed by several investigators (e.g., Abdel-
Rahman 1995; Cathelineau 1988; Kranidiotis and MacLean 1987).
Kranidiotis and MacLean (1987) calibrated an empirical chlorite
geothermometer for estimating the temperature of chlorite formation
by replacement of mafic minerals. Calculated results from this formula-
tion are less than 306 °C (supplementary table 6SD), consistent with the
stability limits of chlorite and the petrographic evidence that chlorite
formation happened at a late stage. This supports the distinction we Fig. 14. Fe/(Fe + Mg) vs. AlIV diagram for primary amphibole (Anderson and Smith 1995).

21
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

are not captured by the model: CaO does not decrease steeply enough the Dokhan volcanic rocks and various deformed granite plutons, rather
with increasing SiO2, P2O5 has altogether the wrong trend, the LREE than on structural, stratigraphic, or metamorphic evidence indicating
act more incompatibly than the HREE in the model (whereas the that the Dokhan lavas themselves were caught up in collisional defor-
data show the opposite, Fig. 10), and the Eu anomalies predicted in mation after their eruption. Andresen et al. (2009), for example, studied
the rhyolite are too small. All of these features can be simultaneously a certain area (Meatiq) in the Eastern Desert of Egypt and found that de-
remedied simply by adding a larger quantity of fractionating apatite formation of I-type granite occurred at 605–600 Ma. From this evidence,
than is predicted by the model. It has been frequently noted since the they reasoned that collision between East and West Gondwana
rhyoliteMELTS calibration came out that it underpredicts the stability occurred at this late time, younger than the Dokhan volcanics and
of apatite in felsic systems. Hence, we artificially added more apatite supporting their assignment to a pre-collisional phase. Without
to the fractionating assemblage; a constant rate of apatite fractionation questioning the accuracy of the age of deformation of this granite, we
(0.0004 wt%/°C) simultaneously fits the CaO, P2O5, LREE, and Eu anom- must still ask whether it supports a broad regional interpretation. Defor-
aly trends (Figs. 7, 8, 10c). mation of various rock types may occur after collision and post-
With the apatite adjustment, the simple decompression and frac- collisional plutons may inherit geochemical characteristics from earlier
tional crystallization model is successful at explaining most features of island arc stages. Hence, the measured age may not date the terminal
the major and trace element chemistry, so we conclude that the collision locally, the deformed pluton may not be an island-arc stage
whole-rock data do not require assimilation, recharge, or other com- pluton, and a local collision may not constrain the end of collisional ac-
plexity during magmatic evolution of the Guruf volcanic suite. However, tivity elsewhere. We find the absence of deformation or metamorphism
oscillatory zoning in plagioclase in the andesite and dacite (Fig. 4c, of the Dokhan sequence itself to be the most compelling evidence that
d) does suggest the occurrence of fluctuations in magmatic conditions Dokhan volcanics are post-collisional.
(pressure, temperature and H2O content), which may indicate that re-
charge or mixing events did occur. On several variation diagrams, in- 7.6. Existence of old continental crust beneath the ANS
cluding those for Al2O3, Fe2O3*, MgO, and Na2O, the data trend is
straighter than the somewhat curved model liquid line of descent; the The ANS is a product of the Neoproterozoic East African orogeny
intermediate compositions could therefore be explained by a process (870–550 Ma; Stern, 1994) and several authors have argued that the
of recharge of the fractionated magma by more primitive material ANS was entirely juvenile material, with no incorporation of pre-
entering the magma chamber. Studies of Dokhan volcanics at other lo- Neoproterozoic crust (Stern, 1994; Maurice et al. 2018). However, a
calities in the Eastern Desert and Sinai have frequently called for a role large number of pre-870 Ma inherited zircons (0.9–3.0 Ga) have been
for crustal contamination accompanying fractional crystallization found in ANS magmas (e.g., Hargrove et al. 2006; Breitkreuz et al.
(e.g., Abdel Rahman, 1996; Mohamed et al., 2000; Moghazi 2003; 2010; Stern et al., 2010; Be'eri-Shlevin et al. 2011) and sediments (Li
Eliwa et al. 2006; Obeid and Azer 2015). There is no evidence locally et al. 2018; Samuel et al. 2011; Wilde and Youssef 2002). Thus, other au-
of over-enrichment in incompatible elements, so we consider the pres- thors (e.g. El-Gaby et al. 1988; Khudeir et al., 2008; Li et al. 2018) have
ent suite to be permissive of a magmatic evolution without contamina- judged that extensive tracts of pre-Neoproterozoic continental crust
tion, but we also do not rule out some amount of contamination. exist beneath the Pan-African rocks of the ANS, at least in the Eastern
Desert. Some authors (e.g., Liégeois and Stern 2010) have attempted
7.5. Collision of East and West Gondwana to reconcile these views by stressing the durability and inheritability
of zircons, such that pre-Neoproterozoic zircons need not require the
The ANS developed over a protracted period of ~300 Ma, passing presence of extensive tracts of older crust. However, the oxygen isotope
through a series of distinct stages (Stern, 1994; Stern and Johnson compositions of zircons are relatively easy to modify (Cherniak
2010; Fritz et al. 2013): rifting and break-up of Rodinia into East and and Watson 2003); old zircons that retain crustal oxygen isotope signa-
West Gondwana fragments sometime in Tonian time, a prolonged epi- tures likely were never recycled into the mantle and transported
sode of formation of island arc terranes and their accretion from magmatically into juvenile crust. Thus Stern et al. (2010) stated that
~800–670 Ma, and reassembly of East and West Gondwana and inter- “it seems clear that ~1 Ga crust can be found near the Sa'al-Zaghra re-
vening fragments by a continent-continent collision at ~630 Ma. Several gion of the Sinai Peninsula”.
intraoceanic subduction zones now incorporated in the ANS were active The high-grade gneisses and migmatites in the Eastern Desert oc-
from 870 to 630 Ma. In Sinai, Eyal et al. (2014) concluded that the last cupy the lowest stratigraphic position, making them the substratum of
(Kid) island arc in the northern ANS existed from ~640 to ~620 Ma, im- all other units (El-Gaby et al. 1988; Li et al. 2018). Some authors have ar-
plying that the Mozambique Ocean persisted (at least locally) until gued that these high-grade gneisses and migmatites represent an east-
~620 Ma. erly extension of the Sahara Metacraton, remobilized and exposed in
The collision between East and West Gondwana was a complex and tectonic windows during the East African orogeny (e.g., Khudeir et al.,
lengthy process and it is challenging to interpret the age of any particu- 2008). On the other hand, late Neoproterozoic ages for high-grade
lar sample as the time of collision or even to argue that there was a sin- gneisses in the northwestern ANS (e.g., Andresen et al. 2009; Liégeois
gle collision time (Andresen et al. 2009; Fritz et al. 2013; Stern and and Stern 2010) have supported the contrary idea that they are part of
Johnson 2010). At the largest scale, there was a collision of two main the juvenile ANS crust, in spite of the older zircons (Be'eri-Shlevin
fragments, East and West Gondwana, but along the whole length of et al., 2011). Azer and Asimow (2021) preferred the classical interpreta-
the suture this event was diachronous and, given the irregular shapes tion of these gneisses and sheared granites as cratonic, arguing that the
of the continental margins, the age of “final collision” in one area may Pan-African ages, especially in the northwestern part of the ANS, date
not preclude the continuing existence of open ocean elsewhere along episodes of metamorphism rather than protolith ages.
the front (Fritz et al. 2013). There was a lengthy series of small Maurice et al. (2018) claimed that the Dokhan volcanic rocks of the
and large terrane collisions throughout the whole period from Eastern Desert developed in a subduction zone environment and pre-
~850–620 Ma (Fritz et al. 2013). ferred an intraoceanic arc setting over an active continental margin set-
However, many published studies have attempted to define the gen- ting due to the absence of isotopic evidence for involvement of old
eral chronology of the collision on the basis of local or regional studies. continental crust. However, there is in fact significant evidence for the
This kind of reasoning has led several authors to assign the high-K calc- involvement of old continental crust in the generation of Dokhan volca-
alkaline Dokhan volcanics series in the Eastern Desert of Egypt to the nics (Breitkreuz et al. 2010; Stern et al. 2010; Be'eri-Shlevin et al., 2011).
pre-collisional phase of ANS development (Maurice et al. 2018). This in- Li et al. (2018) studied, for example, a ~ 755 Ma rhyolite cobble from the
terpretation is generally based on geochronological correlation between Atud Formation in the Eastern Desert, whose exclusively negative

22
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

zircon εHf(t) values indicate a heterogeneous source enriched in Hf and NSF award 1947616. We thank Einari Suikkanen for creating and shar-
not a juvenile oceanic arc source. Melting of a Paleoproterozoic-Archean ing the easyMelts interface and Paula Antoshechkina for supporting
protolith at ca. 755 Ma (Li et al. 2018) shows that the subduction stage the alphaMELTS trace element modeling capability. Finally, thorough
of the ANS mobilized material of older continental crust origin and con- reviews by the two reviewers as well as comments by Prof. Michael
firms the presence of such crust beneath the ANS. Roden (editor) greatly improved the present version of the manuscript.

8. Summary Appendix A. Supplementary data

The Guruf volcanics represent post-collisional magmatic activity fol- Supplementary data to this article can be found online at https://doi.
lowing the assembly of the northernmost segment of the ANS. The two org/10.1016/j.lithos.2021.106051.
genetically related pulses together span a wide range of compositions
(52.0–74.7 wt% SiO2) without any compositional gaps. They evolved References
from adakitic medium-K calc-alkaline rocks towards alkaline high-K
Abdel Rahman, A.M., 1996. Pan-African volcanism: petrology and geochemistry of the
character with progressive ascent and fractional crystallization. The Dokhan Volcanic suite in the northern Nubian Shield. Geol. Mag. 133, 17–31.
metavolcanic xenoliths within the granodiorite are tholeiitic, with Abdelfadil, K.M., Obeid, M.A., Azer, M.K., Asimow, P.D., 2018. Late Neoproterozoic adakitic
very low K2O and unfractionated REE and trace element patterns; they lavas in the Arabian-Nubian shield, Sinai Peninsula, Egypt. J. Asian Earth Sci. 158,
301–323.
are not related to the Guruf volcanics and instead appear similar to
Abdel-Karim, A.M., Ali, S., El-Awady, A., Elwan, W., Khedr, M.Z., Tamura, A., 2019. Mineral
the island-arc stage Shadli metavolcanics. Hence, they do not show and bulk–rock chemistry of Shadli bimodal metavolcanics from Eastern Desert of
that the granodiorite postdates the Guruf volcanics. The field relations Egypt: Implication for tectonomagmatic setting and Neoproterozoic continental
indicate the opposite. growth in the Arabian–Nubian Shield. Lithos 338, 204–217.
Abdel-Rahman, A.M., 1995. Chlorites in a spectrum of igneous rocks: mineral chemistry
Primitive mantle-normalized multi-element diagrams show that the and paragenesis. Mineral. Mag. 59, 129–141.
Guruf volcanics are enriched in large ion lithophile elements (LILE) rel- Abu El-Enen, M.M., Lorenz, J., Ali, K.A., von Seckendorff, V., Okrusch, M., Schüssler, U.,
ative to high field strength elements (HFSE), with significant troughs in Brätz, H., Schmitt, R.T., 2018. A new look on Imperial Porphyry: a famous ancient di-
mension stone from the Eastern Desert of Egypt—petrogenesis and cultural rele-
Sr, P and Ti in the rhyodacite and rhyolite. All the Guruf volcanic rocks vance. Int. J. Earth Sci. 7, 2393–2408.
are LREE-enriched relative to HREE, but the patterns become less Al-Damegh, K., Sandvol, E., Barazangi, M., 2005. Crustal structure of the Arabian plate:
fractionated with progressive evolution, contrary to the standard expec- new constraints from the analysis of teleseismic receiver functions. Earth Planet.
Sci. Lett. 231 (3–4), 177–196.
tations of simple fractional crystallization. The patterns are well-
Ali, B.H., 2015. SHRIMP U–Pb zircon geochronology: evidence for emplacement time of
explained by the quantity of apatite fractionation required to match some granitoids north Eastern Desert, Egypt. Arab. J. Geosci. 8, 5465–5474.
the Ca and P trends in the data. The consistently low MREE/HREE and Ali, K.A., Azer, M.K., Gahlan, H.A., Wilde, S.A., Samuel, M.D., Stern, R.J., 2010. Age con-
flat HREE patterns indicate residual amphibole in the source of the ba- straints on the formation and emplacement of Neoproterozoic ophiolites along the
Allaqi–Heiani Suture, South Eastern Desert of Egypt. Gondwana Res. 8 (4), 583–595.
saltic andesites and rule out garnet fractionation. Ali, K.A., Andresen, A., Stern, R.J., Manton, W.I., Omar, S.A., Maurice, A.E., 2012. U-Pb zircon
On most tectonic setting discrimination diagrams, the Guruf volca- and Sr–Nd–Hf isotopic evidence for a juvenile origin of the ~634 Ma El-Shalul Granite,
nics lie between volcanic arc and within-plate fields. Throughout the Central Eastern Desert, Egypt. Geol. Mag. 149, 783–797.
Anderson, J.L., Smith, D.R., 1995. The effects of temperature and fO2 on the Al-in-
ANS, this feature is characteristic of the post-collisional tectonic setting; hornblende barometer. Am. Mineral. 80, 549–559.
true island-arc stage subduction-related magmas plot clearly within the Andresen, A., Abu El-Rus, M.A.A., Myhre, P.I., Boghdady, G.Y., Corfu, F., 2009. U–Pb TIMS
volcanic arc fields. The evidence of subduction influence in the geo- age constraints on the evolution of the Neoproterozoic Meatiq Gneiss dome, Eastern
Desert, Egypt. Int. J. Earth Sci. 98 (3), 481–497.
chemistry of these lavas, despite their post-collisional setting, is
Avigad, D., Gvirtzman, Z., 2009. Late Neoproterozoic rise and fall of the northern Arabian–
interpreted to reflect remelting of previously formed arc material ca. Nubian shield: the role of lithospheric mantle delamination and subsequent thermal
750–650 Ma in age. subsidence. Tectonophysics 477, 217–228.
The most likely source and setting for generation of the primitive ba- Azer, M.K., Asimow, P.D., 2021. Petrogenetic Evolution of the Neoproterozoic Igneous
Rocks of Egypt. The Geology of the Egyptian Nubian Shield. Springer, Cham,
saltic andesite magma was lower crust undergoing high-P metamor- pp. 343–382.
phism and heating during delamination. Experimental studies show a Azer, M.K., Farahat, E.S., 2011. Late Neoproterozoic volcano-sedimentary succession of
narrow window near 950 °C, 1.5 GPa, 10% mixing with peridotite, and Wadi Rufaiyil, southern Sinai, Egypt: a case of transition from late- to post-
collisional magmatism. J. Asian Earth Sci. 42, 1187–1203.
10 wt% H2O, where tonalite could melt to produce basaltic andesite
Azer, M.K., Abu El-Ela, F.A., Ren, M., 2012. The petrogenesis of late Neoproterozoic mafic
with residual amphibole. Such conditions could occur if lower arc- dike-like intrusion in South Sinai, Egypt. J. Asian Earth Sci. 54, 91–109.
crust material were heated as it delaminated and sank into the astheno- Basta, F.F., Maurice, A.E., Bakhit, B.R., Azer, M.K., El-Sobky, A.F., 2017. Intrusive rocks of the
sphere to a depth of ~50 km. Cooling during ascent and simple fractional Wadi Hamad Area, North Eastern Desert, Egypt: Change of magma composition with
maturity of Neoproterozoic continental island arc and the role of collisional
crystallization provide a working model for evolution through the an- plutonism in the differentiation of arc crust. Lithos 288-289, 248–263.
desite and dacite of the lower Guruf volcanics. After a hiatus and contin- Be’eri-Shlevin, Y., Samuel, M.D., Azer, M.K., Rämö, O.T., Whitehouse, M.J., Moussa, H.E.,
ued fractionation, renewed eruption of the upper volcanic cycle 2011. The late Neoproterozoic Ferani and Rutig volcano-sedimentary successions of
the northernmost Arabian-Nubian Shield (ANS): New insights from zircon U-Pb geo-
emplaced the rhyodacite and rhyolite. chronology, geochemistry and O-Nd isotope ratios. Precambrian Res. 188, 21–44.
Breitkreuz, C., Eliwa, H., Khalaf, I., El Gameel, K., Bühler, B., Sergeev, S., Larionov, A.,
Murata, M., 2010. Neoproterozoic SHRIMP U-Pb zircon ages of silica-rich Dokhan Vol-
canics in the North Eastern Desert, Egypt. Precambrian Res. 182, 163–174.
Declaration of Competing Interest Carroll, M.J., Wyllie, P.J., 1989. Experimental phase relations in the system tonalite-
peridotite-H2O at 15 kb; Implications for assimilation and differentiation processes
The authors declare that they have no known competing financial near the crust-mantle boundary. J. Petrol. 30 (6), 1351–1382.
Castillo, P.R., 2012. Adakite petrogenesis. Lithos 134–135, 304–316.
interests or personal relationships that could have appeared to influ-
Cathelineau, M., 1988. Cation site occupancy in chlorites and illites as a function of tem-
ence the work reported in this paper. perature. Clay Miner. 23, 471–485.
Cherniak, D.J., Watson, E.B., 2003. Diffusion in zircon. Rev. Mineral. Geochem. 53 (1),
Acknowledgements 113–143.
Condie, K.C., 2005. TTGs and adakites: are they both slab melts? Lithos 80, 33–44.
Cox, K.G., Bell, J.D., Pankhurst, R.J., 1979. The Interpretation of Igneous Rocks. Allen &
The authors would like to extend their appreciation and gratitude to Unwin, London, p. 450.
the King Saud University for funding and supporting this work through Defant, M.J., Drummond, M.S., 1990. Derivation of some modern arc magmas by melting
of young subducted lithosphere. Nature 347, 662–665.
Researchers Supporting Project number (RSP-2020/151), King Saud
El-Gaby, S., List, F.K., Tehrani, R., 1988. Geology, evolution and metallogenesis of the Pan-
University, Riyadh, Saudi Arabia. PDA's participation is supported by African belt in Egypt. The Pan-African belt of Northeast Africa and Adjacent Areas:
Tectonic Evolution and Economic Aspects of a Late Proterozoic Oregon, pp. 17–68.

23
B.A. Abuamarah, M.K. Azer, P.D. Asimow et al. Lithos 388–389 (2021) 106051

El-Gaby, S., Khudier, A.A., Abdel Tawab, M., Atalla, R.F., 1991. The metamorphosed Moghazi, A.M., 2003. Geochemistry and petrogenesis of a high-K calc-alkaline Dokhan
volcano-sedimentary succession of Wadi Kid, southeastern Sinai, Egypt. Annals of Volcanic suite, south Safaga area, Egypt: the role of late Neoproterozoic crustal exten-
Geological Survey of Egypt XVII, pp. 19–35. sion. Precambrian Res. 125, 161–178.
Eliwa, H.A., Kimura, J.I., Itaya, T., 2006. Late Neoproterozoic Dokhan Volcanics, North East- Moghazi, A.M., El-Sayed, M.M., Arslan, A.I., Hosny, H., 2004. Geochemistry and petrogen-
ern Desert, Egypt: geochemistry and petrogenesis. Precambrian Res. 151, 31–52. esis of the late Neoproterozoic granitoids at Wadi Hamad, Northern Eastern Desert,
Eliwa, H.A., El-Bialy, M.Z., Murata, M., 2014. Ediacaran post-collisional volcanism in the Egypt. 6th International Conference on Geochemistry, Faculty of Science, Alexandria
Arabian-Nubian Shield: the high-K calc-alkaline Dokhan Volcanics of Gabal Samr University, Egypt, Abstract, V. 2, p. 144.
El-Qaa (592 ± 5 Ma), North Eastern Desert, Egypt. Precambrian Res. 246, 180–207. Mohammed, F.H., Moghazi, A.M., Hassanen, M.A., 2000. Geochemistry, petrogenesis and
Eyal, M., Litvinovsky, B., Jahn, B.M., Zanvilevich, A., Katzir, Y., 2010. Origin and evolution of tectonic setting of late Neoproterozoic Dokhan-type volcanic rocks in the Fatira
post-collisional magmatism: coeval Neoproterozoic calc-alkaline and alkaline suites area, Eastern Egypt. Internat. J. Earth Sci. 88, 764–777.
of the Sinai Peninsula. Chem. Geol. 269, 153–179. Morimoto, N., Fabries, J., Ferguson, A., Ginzburg, I., Ross, M., Seifert, F., Zussman, J., Aoki, K.,
Eyal, M., Be’eri-Shlevin, Y., Eyal, Y., Whitehouse, M.J., Litvinovsky, B., 2014. Three succes- Gottardi, G., 1988. Nomenclature of pyroxenes. Mineralo. Magazine 52, 535–550.
sive Proterozoic island arcs in the northern Arabian-Nubian Shield: evidence from Moussa, E.H., 1998. Geochronological Studies of Some Granitoids; Application to Geo-
SIMS U-Pb dating of zircon. Gondwana Res. 25, 338–357. chemical Evolution and Tectonic History of the Northern Eastern Desert, Egypt. Ph.
Farahat, E.S., Azer, M.K., 2011. Post-collisional magmatism in the northern Arabian- D. Thesis. Ain Shams University, Egypt, p. 284.
Nubian Shield: the geotectonic evolution of the alkaline suite at Gebel Tarbush Moussa, E.M.M., Stern, R.J., Manton, W.I., Ali, K.A., 2008. SHRIMP zircon dating and Sm/Nd
area, South Sinai, Egypt. Geochemistry 71 (3), 247–266. isotopic investigations of Neoproterozoic granitoids, Eastern Desert, Egypt. Precam-
Fritz, H., Abdelsalam, M., Ali, K.A., Bingen, B., Collins, A.S., Fowler, A.R., Ghebreab, W., brian Res. 160, 341–356.
Hauzenberger, C.A., Johnson, P.R., Kusky, T.M., Macey, P., Muhongo, S., Stern, R.J., Moyen, J.F., Martin, H., 2012. Forty years of TTG research. Lithos 148, 312–336.
Viola, G., 2013. Orogen styles in the East African Orogen: a review of the Obeid, M.A., Azer, M.K., 2015. Pan-African adakitic rocks of the north Arabian-Nubian
Neoproterozoic to Cambrian tectonic evolution. J. Afr. Earth Sci. 86, 65–106. Shield: petrological and geochemical constraints on the evolution of the Dokhan vol-
Gahlan, H., Azer, M.K., Asimow, P., Al-Kahtany, K., 2016. Late Ediacaran post-collisional A- canics in the north Eastern Desert of Egypt. Int. J. Earth Sci. 104, 541–563.
type syenites with shoshonitic affinities, northern Arabian-Nubian Shield: a possible Pearce, J.A., 1983. Role of the sub-continental lithosphere in magma genesis at active con-
mantle-derived A-type magma. Arab. J. Geosci. 9 (12), 603. tinental margins. In: Hawkesworth, C.J., Norry, M.J. (Eds.), Continental Basalts and
Mantle Xenoliths. Shiva, Nantwich, pp. 230–249.
Gao, S., Rudnick, R.L., Yuan, H.L., Liu, X.M., Xu, W.L., Lin, W.L., Ayers, J., Wang, X.C., Wang,
Pearce, J.A., 1996. Sources and settings of granitic rocks. Episodes 19, 120–125.
Q.H., 2004. Recycling lower continental crust in the North China Craton. Nature 432,
Pearce, J.A., Gale, G.P., 1977. Identification of ore deposition environment from trace ele-
892–897.
ments geochemistry of associated igneous host rocks. Volcanic Processes in Ore Gen-
Ghiorso, M.S., Gualda, G.A., 2013. A method for estimating the activity of titania in mag-
esis. Inst. Geological Society of London, Special Publication 7, pp. 14–24.
matic liquids from the compositions of coexisting rhombohedral and cubic iron–
Pearce, J.A., Harris, N.B.W., Tindle, A.G., 1984. Trace element discrimination diagrams for
titanium oxides. Contribut. Mineral. Petrol. 165 (1), 73–81.
the tectonic interpretation of granitic rocks. J. Petrol. 25, 956–983.
Hargrove, U., Stern, R., Kimura, J-.I., Manton, W., Johnson, P., 2006. How juvenile is the
Ressetar, R., Monard, J.R., 1983. Chemical composition and tectonic setting of the Dokhan
Arabian–Nubian Shield? Evidence from Nd isotopes and pre-Neoproterozoic
Volcanic formation, Eastern Desert, Egypt. J. Afr. Earth Sci. 1, 103–112.
inherited zircon in the Bi’r Umq suture zone, Saudi Arabia. Earth Planet. Sci. Lett.
Richards, J., Kerrich, R., 2007. Adakite-like rocks: their diverse origins and questionable
252 (3), 308–326.
role in metallogenesis. Econ. Geol. 102 (4), 537–576.
Harris, N.B., Pearce, J.A., Tindle, A.G., 1986. Geochemical characteristics of collision-zone
Rickwood, P.C., 1989. Boundary lines within petrologic diagrams, which use oxides of
magmatism. In: Coward, M.P., Ries, A.C. (Eds.), Collision Tectonics. Geological Society
major and minor elements. Lithos 22, 247–263.
of London, Special Publications 19, pp. 67–81.
Ridolfi, F., Renzulli, A., Puerini, M., 2010. Stability and chemical equilibrium of amphibole
Hey, M.H., 1954. A new review of the chlorites. Mineralo. Magazine 30, 277–292. in calc-alkaline magmas: an overview, new thermobarometric formulations and ap-
Hossny, H.A.H., 2008. Geochemistry, Petrogenesis and Geochemical Exploration of the plication to subduction-related volcanoes. Contrib. Mineral. Petrol. 160, 45–66.
Rock Association at Wadi Hamad District, North Eastern Desert, Egypt. Ph. D. Thesis. Roeder, P.L., Emslie, R.F., 1970. Olivine-liquid equilibrium. Contrib. Mineral. Petrol. 29,
Alexandria University, Egypt, p. 246. 275–289.
Irvine, T.N., Baragar, W.R.A., 1971. A guide to the chemical classification of the common Samuel, M.D., Be’eri-Shlevin, Y., Azer, M.K., Whitehouse, M.J., Moussa, H.E., 2011. Prove-
volcanic rocks. Can. J. Earth Sci. 8, 523–548. nance of conglomerate clasts from the volcano-sedimentary sequence at Wadi
Johnson, P.R., Kattan, F., 2007. Geochronologic Dataset for Precambrian Rocks in the Ara- Rutig in southern Sinai, Egypt as revealed by SIMS U-Pb dating of zircon. Gondwana
bian Peninsula. A Catalogue of U–Pb, Rb–Sr, Ar–Ar, and Sm–Nd Ages: Saudi Geological Res. 20, 450–464.
Survey Open-File report SGS-OF-2007-3. p. 21. Stacey, J.S., Stoeser, D.B., Greenwood, W.R., Fischer, L.B., 1984. U–Pb zircon geochronology
Johnson, P.R., Woldehaimanot, B., 2003. Development of the Arabian-Nubian Shield: Per- and geological evolution of the Halaban–Al Amar region of the Eastern Arabian
spectives on accretion and deformation in the East African Orogen and the assembly Shield, Kingdom of Saudi Arabia. J. Geol. Soc. Lond. 141, 1043–1055.
of Gondwana. In: Yoshida, M., Windley, B.F., Dasgupta, S. (Eds.), Proterozoic East Stern, R.J., 1994. Arc assembly and continental collision in the Neoproterozoic East African
Gondwana: Supercontinent Assembly and Breakup. Geological Society of London, Orogen: implications for the consolidation of Gondwanaland. Ann. Rev. Earth Planet.
Special Publication, pp. 289–325. Sci. 22 (1), 319–351.
Kay, R.W., Kay, S.M., 1993. Delamination and delamination magmatism. Tectonophysics Stern, R.J., Gottfried, D., 1986. Petrogenesis of a late Precambrian (575–600 Ma) bimodal
219, 177–189. suite in northeast Africa. Contrib. Mineral. Petrol. 92 (4), 492–501.
Keeditse, M., Rajesh, H.M., Belyanin, G.A., Fukuyama, M., Tsunogae, T., 2016. Primary mag- Stern, R.J., Johnson, P., 2010. Continental lithosphere of the Arabian Plate: a geologic, pet-
matic amphibole in Archaean meta-pyroxenite from the central zone of the Limpopo rologic, and geophysical synthesis. Earth-Sci. Rev. 101 (1–2), 29–67.
complex, South Africa. South Afr. J. Geol. 119 (4), 607–622. Stern, C.R., Kilian, R., 1996. Role of the subducted slab, mantle wedge and continental
Khudeir, A.A., Abu El-Rus, M.A., El-Gaby, S., El-Nady, O., Bishara, W.W., 2008. Sr–Nd iso- crust in the generation of adakites from the Andean Austral volcanic zone. Contrib.
topes and geochemistry of the infrastructural rocks in the Meatiq and Hafafit core Mineral. Petrol. 123, 263–281.
complexes, Eastern Desert, Egypt: Evidence for involvement of pre-eoproterozoic Stern, R.J., Gottfried, D., Hedge, C.E., 1984. Late Precambrian rifting and crustal evolution in
crust in the growth of Arabian–Nubian Shield. Island Arc 17 (1), 90–108. the Northeastern Desert of Egypt. Geology 12 (3), 168–172.
Kranidiotis, P., MacLean, W.H., 1987. Systematics of chlorite alteration at the Phelps Stern, R.J., Ali, K.A., Liégeois, J.P., Johnson, P.R., Kozdroj, W., Kattan, F.H., 2010. Distribution
Dodge massive sulfide deposit, Matagami, Quebec. Econ. Geol. 82, 1898–1911. and significance of pre-Neoproterozoic zircons in juvenile Neoproterozoic igneous
Le Bas, M.J., 1962. The role of aluminum in igneous clinopyroxenes with relation to their rocks of the Arabian-Nubian Shield. Am. J. Sci. 310 (9), 791–811.
parentage. Am. J. Sci. 260, 267–288. Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematics of oceanic basalts:
Leterrier, J., Maury, R.C., Thonon, P., Girard, D., Marchal, M., 1982. Clinopyroxene compo- Implications for mantle composition and processes. In: Saunders, A.D., Norry, M.
sition as a method of identification of the magmatic affinities of paleo-volcanic series. (Eds.), Magmatism in Ocean Basins. Geological Society of London Special Publication
Earth Planet. Sci. Lett. 59 (1), 139–154. 42, pp. 313–345.
Li, X.H., Abd El-Rahman, Y., Abu Anbar, M., Li, J., Ling, X.X., Wu, L.G., Masoud, A.E., 2018. Sylvester, P.J., 1989. Post-collisional alkaline granites. Journal of Geology 97, 261–280.
Old continental crust underlying juvenile oceanic arc: evidence from northern Whalen, J.B., Currie, K.L., Chappell, B.W., 1987. A-type granites: geochemical characteris-
Arabian-Nubian Shield, Egypt. Geophys. Res. Lett. 45 (7), 3001–3008. tics, discrimination and petrogenesis. Contrib. Mineral. Petrol. 95 (4), 407–419.
Liégeois, J.P., Stern, R.J., 2010. Sr–Nd isotopes and geochemistry of granite-gneiss com- Wilde, S.A., Youssef, K., 2000. Significance of SHRIMP U-Pb dating of the imperial por-
plexes from the Meatiq and Hafafit domes, Eastern Desert, Egypt: no evidence for phyry and associated Dokhan Volcanics, Gebel Dokhan, north Eastern Desert, Egypt.
pre-Neoproterozoic crust. J. Afr. Earth Sci. 57 (1–2), 31–40. J. Afr. Earth Sci. 31 (2), 403–413.
Liégeois, J.P., Navez, J., Black, R., Hertogen, J., 1998. Contrasting origin of post-collision Wilde, S., Youssef, K., 2002. A re-evaluation of the origin and setting of the Late Precam-
high-K calc-alkaline and shoshonitic versus alkaline and peralkaline granitoids. The brian Hammamat Group based on SHRIMP U–Pb dating of detrital zircons from Gebel
use of sliding normalization. Lithos 45, 1–28. Umm Tawat, North Eastern Desert, Egypt. J. Geol. Soc. 159 (5), 595–604.
Lindsley, D.H., 1983. Pyroxene thermometry. Am. Mineral. 68, 477–493. Winchester, J.A., Floyd, P.A., 1977. Geochemical discrimination of different magma series
and their differentiation products using immobile elements. Chem. Geol. 20,
Macpherson, C.G., Dreher, S.T., Thirlwall, M.F., 2006. Adakites without slab melting: High
325–343.
pressure differentiation of island arc magma, Mindanao, the Philippines. Earth Planet.
Zhu, D.C., Zhao, Z.D., Pan, G.T., Lee, H.Y., Kang, Z.Q., Liao, Z.L., Wang, L.Q., Li, G.M., Dong,
Sci. Lett. 243, 581–593.
G.C., Liu, B., 2009. Early cretaceous subduction-related adakite-like rocks of the
Maurice, A.E., Bakhit, B.R., Basta, F.F., Asimow, P.D., Wälle, M., Azer, M.K., El-Sobky, A.F.,
Gangdese Belt, southern Tibet: products of slab melting and subsequent melt-
2018. The last subduction-related volcanism in the northern tip of the Arabian- Nu-
peridotite interaction? J. Asian Earth Sci. 34, 298–309.
bian Shield: a Neoproterozoic arc preceding the terminal collision of East and West
Gondwana. Precambrian Res. 310, 256–277.

24