Introduction

Textural and compositional maturity of siliciclastic sediments depends on many factors, the most important of which are provenance, chemical weathering and alteration, duration and distance of sediment transport, sediment recycling, and diagenesis (e.g., Nesbitt and Markovics 1980; Nesbitt and Young 1982; Middelburg et al. 1988; Haughton et al. 1991; Nesbitt et al. 1996; Wanas and Abdel-Maguid 2006; Akarish and El-Gohary 2008; Zaid 2012; Zaid et al. 2015; Sallam and Wanas 2019; Wanas and Assal 2021). Mineralogical and geochemical compositions of siliciclastic sediments have been widely used to recognize sediment provenance, tectonic setting, and palaeoweathering conditions, thus providing significant information useful in assessing palaeogeographical reconstructions and crustal evolution (e.g., Basu et al. 1975; Dickinson and Suczek 1979; Ingersoll and Suczek 1979; Bhatia 1983; Dickinson et al. 1983; Bhatia and Crook 1986; Roser and Korsch 1986, 1988; Haughton et al. 1991; Morton and Hallsworth 1994; Armstrong-Altrin et al. 2013; Löwen et al. 2018; Garzanti 2019; Garzanti and Andò 2019; Mohammedyasin and Wudie 2019; Dinis et al. 2020; He et al. 2020). Whole-rock geochemical analyses have long been applied to characterize sediment provenance linking geochemical signatures to source rocks and tectonic settings (e.g., Nesbitt and Young 1982; Dickinson et al. 1983; Haughton et al. 1991; McLennan et al. 1993; Verma and Armstrong-Altrin 2013, 2016). Numerous sophisticated methods (petrological and geochemical) and discriminant functions with their relevant diagrams were introduced by many authors interested in provenance analysis (e.g., Bhatia 1983; Dickinson et al. 1983; Bhatia and Crook 1986; Roser and Korsch 1986, 1988). A new portion of multi-dimensional diagrams for tectonic setting discrimination of siliciclastic sediments and their applications were proposed by Verma and Armstrong-Altrin (2013, 2016)).

The Red Sea Hills in eastern Egypt boast rich geological record that is crucial for deciphering the Mesozoic–Cenozoic evolution of northeastern Africa before, during, and after the separation of Arabia together with the opening of the Red Sea. Moreover, this domain is similarly crucial for understanding the sedimentary processes at the transition between interiors and periphery of a large continent. Particularly, the Cretaceous rocks in the Quseir area consist predominantly of siliciclastics, with the lowermost package comprising mainly of paralic and fluvial cross-bedded sandstones of the Nubia Formation (Ward and McDonald 1979; Ward et al. 1979; Van Houten et al. 1984; Ruban et al. 2021). This formation unconformably overlies the Precambrian basement rocks and is conformably overlain by the Quseir Formation. The thickness of the Nubia Formation is highly variable, and its age was generally assigned as Late Cretaceous (Said 1962; Guiraud et al. 2001), although some other studies indicate Early Cretaceous age (Bosworth et al. in Hamimi et al. 2020). In the present work, the Early Cretaceous Age of the Nubia Formation is followed.

The previous studies of the Nubia Formation in Egypt dealt mainly with stratigraphy and sedimentology (e.g., Ward and McDonald 1979; Van Houten et al. 1984), whereas a minority of studies focused on provenance analysis of similar sandstones in neighboring countries (e.g., Weissbrod and Nachmias 1986; Amireh 1991; Kolodner et al. 2009; Ahfaf et al. 2021). Therefore, the results of petrographic and whole-rock geochemistry (major and trace elements) analyses of the Nubian sandstones at Gebel Duwi (Fig. 1) have been interpreted to clarify their provenance, tectonic setting, and source area weathering contribution, in addition to better understanding the paleogeography of northeastern Africa and Arabia in the Early Cretaceous. However, the spatial distribution of the Nubia Formation is much wider, and, thus, similar attention should be paid to composition and provenance of the relevant sandstones in many other areas of northeastern Africa and Arabia. Of special interest is how their deposition was controlled by major shoreline shifts and palaeoclimate changes on the regional scale.

Fig. 1
figure 1

A Landsat image showing the location of the investigated section at Gebel Duwi on the western side of the Red Sea, B geological map of the study area (after Youssef 1957; Khalil and McClay 2009), and C composite lithostratigraphic columnar section of the Mesozoic-Paleogene platform sediments in the Quseir area (after Youssef 1957)

Geological setting

Geographically, the study area belongs to the Central Eastern Desert of Egypt where the chain of Red Sea hills separates the internal part of northeastern Africa from the periphery of the Red Sea. More specifically, the study area corresponds to Gebel Duwi. Tectonically, the study area is related to two main extensional fault systems and a number of half-graben basins (Khalil and McClay 2009). The junction of fault segments forms a zigzag fault pattern and rhomboidal blocks (Khalil and McClay 2009). The tectonic re-organizations linked to the Red Sea rifting led to rotation of some blocks; as a result, both pre- and syn-rift sedimentary sequences are exposed (Khalil and McClay 2009) (Fig. 1).

The pre-rift sequences (Precambrian–early Eocene) start with highly deformed volcaniclastics that were metamorphosed into the greenschist facies (Akaad and Noweir 1980; Stern 1981, 1994; Stoeser and Camp 1985; Kröner 1984, 1993). These basement rocks are unconformably overlain by a 500–700-m-thick sedimentary succession of Cretaceous–early Eocene strata. These are visible in outcrops in the Safaga–Quseir area (e.g., Gebel Duwi, Gebel Atshan, Gebel Hamadat, Gebel Mohamed Rabah, and Gebel Wasif). Particularly, the ~180-m-thick sandstones of the Nubia Formation unconformably overlie basement rocks, and these sandstones are overlain by ~70-m-thick variegated shales. The latter were separated by Youssef (1957) as a distinct stratigraphical unit of Campanian age. The upper strata include the Duwi Formation overlain by ~300 m of interbedded shales and limestones of the Upper Cretaceous–Paleocene Dakhla, Tarawan, and Esna formations (Youssef 1957; Abd El-Razik 1967; Said 1990, 2017). The youngest pre-rift unit comprises ~200-m-thick cherty limestones of the Thebes Formation.

The oldest syn-rift sequences include chert-cobble conglomerates and grits of the late Oligocene Nakheil Formation (Said 1990). It is overlain by reefal limestones, clastics, and evaporites of the Ranga, Um Mahara, Abu Dabbab, and Um Gheig formations (Philobbos et al. 1993; Said 1990). The Late Miocene–Quaternary sequences reach 1000 m in thickness offshore (Orszag-Sperber et al. 1998; Heath et al. 1999). These are marine sandstones, conglomerates, and reefal limestones of the Mersa Alam Formation and conglomerates, reefal limestones, and coarse sands of the Samadi Formation.

Sampling and methods

In this study, a multi-method approach of modal analysis, petrography, and whole-rock geochemistry has been performed to infer the provenance, depositional tectonic setting, and source area palaeoweathering of the Cretaceous sandstones in the Red Sea Hills. One representative section of the Nubia Formation was measured, described, and sampled. The locality is Gebel Duwi—a conspicuous mountain consisting of a long sharp ridge, which drops precipitously to the southwest and slopes gently to the northeast. Its geological setting is demonstrated in Fig. 1.

A total of 65 sandstone samples were collected from the studied Nubian sandstones at Gebel Duwi (Quseir area, Eastern Desert). For the purposes of petrographic analysis, 45 thin sections of the sandstones were investigated under the standard polarizing microscope. Description of sandstone followed the classification scheme of Garzanti (2016) and Garzanti et al. (2018). Approximately 250–300 grains were counted per thin section following the traditional methodology of Gazzi-Dickinson (Ingersoll et al. 1984) for the detrital constituents (quartz, feldspars, and lithic fragments) of each studied sample. The results of modal composition analysis were tabulated and plotted in QFL ternary diagrams of Dickinson et al. (1983).

Fine to medium-grained sandstones (250–65 μm) were subjected to heavy mineral separation using bromoform (specific gravity 2.89 gm/cm3). Then, the heavy mineral fractions were dried and mounted on a glass slide for microscopic investigation. Some of opaque and non-opaque heavy mineral fractions have been identified in the studied samples.

The samples from the Nubian sandstones were analyzed for major and trace element geochemistry using the X-ray fluorescence (XRF) spectrophotometry technique on fused and pressed beads, respectively. Analytical precision was better than 5% for major oxides and trace elements. Loss of ignition (LOI) was estimated by heating the dried sample at 1000 °C for 2 h. Major element data were recalculated to an anhydrous (LOI-free) basis and adjusted to 100% before using them in various diagrams. Major and trace element analyses were carried out at the Central Laboratories of the Egyptian Mineral Resources Authority (EMRA) in Cairo, Egypt.

Results

Field observation and sedimentology

In the Gebel Duwi section, the Nubia Formation unconformably overlies Precambrian basement rocks and is succeeded by the Quseir Formation (Fig. 2). The Nubia Formation consists mainly of varicolored whitish-gray, pale yellow, and reddish sandstones with few interbeds of siltstones and claystones (Fig. 2). It attains thickness of about 180 m at the studied section. Bedding is medium and small scale. The sandstones exhibit diverse sedimentary structures such as ripples, load casts, flat- and cross-bedding, and both planar and trough types on small and medium scale. They show a fining-upward cyclic pattern (each cycle is 0.5 to 1.0 m in thickness), coarser at bottom and becomes finer upward. No marine fossils were found in these sandstones.

Fig. 2
figure 2

Detailed lithostratigraphy of the Nubia Formation at Gebel Duwi, Quseir area

The Nubia Formation in the studied section can be sub-divided into three lithofacies types: lower, middle, and upper lithofacies. The lower lithofacies overly unconformably the basement rocks with an erosional surface in between (Fig. 3A) and is made up of about 70-m-thick non-fossiliferous, pale yellow, medium to coarse-grained, lenticular, and planar and trough cross-bedded pebbly sandstones (Fig. 3B, C). The middle lithofacies is approximately 85 m thick and is composed of reddish-brown, fine- to medium-grained sandstones. These sandstones are characterized by the common occurrence of small-scale flat-bedded and planar cross-bedding. The latter exhibits a poly-directional pattern of palaeoflow, generally towards NNW and NNE. The upper lithofacies are about 25 m thick, and it consists of dark brown to reddish brown, fine-grained sandstones, and sandy siltstones that commonly show well-developed ripple marks (Fig. 3D, E).

Fig. 3
figure 3

Field photographs of the Nubia Formation showing A non-conformable contact between the Precambrian basement rocks and the overlying Nubia Formation at Gebel Duwi. Person for scale 165 cm long, B trough cross-bedded sandstones at the lower lithofacies of the Nubia Formation. The hammer handle for scale is 26 cm long, C tabular planar cross-bedded sandstones at the middle lithofacies of the Nubia Formation with a paleoflow direction to the southwest, and D, E rippled-sandstones characterizing of the upper lithofacies type of the Nubia Formation. The hammer handle for scale is 26 cm long, and coin is 2.4 cm in diameter. All photographs are made at Gebel Duwi, Quseir area

Overall, the Nubia Formation was deposited predominantly in fluvial and estuarine near-shore environments interrupted by flood plains (cf. Shawa and Issawi 1978; Ward and McDonald 1979; Van Houten et al. 1984). The common occurrence of planar cross-bedded sandstones suggests a downstream migration of transverse bars and sand waves in shallow water stream channels under lower flow regime conditions (Miall 1988; Wanas et al. 2015; Sallam et al. 2018; Ruban et al. 2019; Sallam and Ruban 2020). The siltstones and claystone interbeds represent overbank (floodplain) sedimentation. Color mottling, desiccation cracks, and red coloration indicate that these floodplain deposits were subaerially exposed and underwent pedogenesis and soil formation (Issawi et al. 2005; Sallam et al. 2015; Wanas et al. 2015). The upper part of the Nubia Formation consists mainly of paralic and deltaic sediments, which gradually passed up into shallow-marine sediments (variegated shales and glauconitic sandstones) of the overlying Quseir Formation (Ward and McDonald 1979). Gradational contact between the Nubia Formation and the overlying Quseir Formation is noticed.

Petrography, heavy mineral fractions, and modal composition

The petrographic investigation of the studied sandstones revealed that the framework composition of sediments is largely quartz-dominated (average 92.26% of the rock volume), subordinate lithic fragments (average 4.6% of the rock volume and are represented mainly by siltstones and sandy siltstones), and minor proportions of feldspars (mainly microcline) and heavy fractions (Fig. 4A–F). Feldspars and heavy fractions constitute more or less 2% of the rock volume. The sandstones are fine to medium-grained and moderately well-sorted. The matrix is chiefly composed of silt-grade quartz. The cement material is commonly ferruginous and quartz overgrowths. The pores are intergranular and irregular in shape, and they are relatively connected. Modal composition analysis (quartz–feldspars–lithic fragments, QFL) of the analyzed samples showed that the Nubian sandstones have average proportions of Q92.2F1.1L4.6 (Table 1). In the QFL ternary diagrams of Garzanti (2016) and Garzanti et al. (2018), all examined sandstone samples can be classified as quartzose and pure quartzose to litho-quartzose (Fig. 5A–C). The recognized heavy minerals mainly comprise zircon, tourmaline, rutile (high ZTR index), garnet, kyanite, in addition to minor epidote, ilmenite, and leucoxene (Fig. 5D and Fig. S-1, see supplement).

Fig. 4
figure 4

Photomicrographs showing the detrital constituents of the studied sandstones: AE quartz-arenites from the lower and middle lithofacies of the Nubia Formation (samples # 1, 4, 6, 7, 11, respectively), F fine-grained silty sandstones from the upper lithofacies of the Nubia Formation (sample # 13). Abbreviations: Qm, monocrystalline quartz; Qp, polycrystalline quartz; F, feldspars (microcline); L, lithic fragments; op, opaque minerals; u, undulose extinction, point contacts (red arrows), straight contacts (yellow arrows), concavo-convex contacts (blue arrows), and sutured contacts (green arrows)

Table 1 Modal composition (%) of the analyzed sandstone samples
Fig. 5
figure 5

AC The QFL ternary diagram of detrital components of the studied sandstones (after Garzanti 2016 and Garzanti et al. 2018). D ZTR (zircon + tourmaline + rutile) versus garnet and epidote minerals. Abbreviations: Q, total quartz; F, feldspars; L, lithic fragments. Deep yellow circles represent the studied samples

The quartz grains are sub-angular to sub-rounded and show point to straight and concavo-convex grain-to-grain contacts. Monocrystalline quartz dominates (91.8–97.2%) and shows unit (non-undulose) extinction (81.5–94.4%). A few quartz grains display slightly undulose extinction (2.6–11.5%). Polycrystalline (composite) quartz is less frequent (2.8–8.1%). The modal composition data are shown in Table 1. The main diagenetic features recognized in the investigated sandstone samples include compaction, cementation, hematitization, and quartz overgrowths. In general, the studied samples of the Nubia sandstones are chemically (compositionally) and texturally mature.

Geochemistry

Major element geochemistry

The results of major oxides of our studied samples are compared to those of the upper continental crust (UCC) (Table 2). The studied sandstone samples are enriched in SiO2 (average 96.61%) (Table 2). In contrast, they have very low concentrations of TiO2 (average 0.078%) and MnO (average 0.075%), and the content of CaO, MgO, K2O, Na2O, and P2O5 is also very low (Table 2). The very low percentages of K2O and Na2O are consistent with the scarcity of feldspar minerals. Besides, Al2O3 shows strong positive linear correlation with TiO2 (r2 = 0.78) and slight positive correlation with CaO (r2 = 0.47), whereas it displays a strong negative linear correlation with SiO2 (r2 = –0.81) and with NaO2 (r2 = –0.03) (Fig. 6A–D).

Table 2 Major element geochemical data (wt. %), CIA and CIW values of the analyzed sandstone samples
Fig. 6
figure 6

AD Correlation diagrams of Al2O3 with other major oxides

Trace element geochemistry

The results of trace elements of our studied sandstone samples are compared to those of the upper continental crust (UCC) (Table 3). Trace element data shows a relatively high concentrations of Ba (148–244 ppm), Th (8.4–11.4 ppm), Zr (27–117 ppm), and Sr (27–117 ppm); moderate concentrations of Co (44–48 ppm) and Cu (12–18 ppm); and low concentrations of La (14–17 ppm), Ni (4–14 ppm), Zn (3–10 ppm), Nb (3–10 ppm), and Pb (5–12 ppm) (Table 3). Apparently, the enrichment in Ba, Zr, and Sr concentrations most probably reflect zircon and possibly monazite in the heavy mineral fractions. In addition, Al2O3 has slight positive correlations with Co (r2 = 0.16), Ni (r2 = 0.21), and Cu (r2 = 0.2), while displaying relatively high positive correlations with Zn (r2 = 0.5) (Fig. 7A–D). Plots of Fe2O3tot versus Co, Ni, Cu, and Zn showed slight positive correlations, with correlation coefficient (r2) values of 0.06, 0.06, 0.3, and 0.19, respectively (Fig. 8A–D).

Table 3 Trace element geochemical data (given in ppm) of the analyzed sandstone samples
Fig. 7
figure 7

AD Correlation diagrams of Al2O3 with Co, Ni, Cu, and Zn

Fig. 8
figure 8

AD Correlation diagrams of Fe2O3tot with Co, Ni, Cu, and Zn

Concentrations of selected trace elements of the studied Nubia samples have been normalized to average upper continental crust (UCC; values from Rudnick and Gao 2003). These include the large-ion lithophile easily mobilized elements (e.g., Ba, Rb, and Sr) and high-field strength immobile elements (e.g., Nb and Zr) (Fig. 9), which can reveal significant clues into sedimentary provenance (Taylor and McLennan 1985).

Fig. 9
figure 9

A UCC-normalized multi-element diagram for the studied sandstones. Average normalized values derived from Rudnick and Gao (2003)

Discussion

Interpretation procedures

In this study, multidisciplinary methods of petrography, modal analysis, heavy mineralogy, and whole-rock geochemistry of the Nubian sandstones at Gebel Duwi have been performed to constrain their provenance, tectonic setting, source area palaeoweathering, and palaeoclimatic conditions. The present study followed the algorithm of provenance analysis based on the following established interpretation procedures.

The QFL schemes of Garzanti (2016) and Garzanti et al. (2018) and the bipyramidal diagram of Tortosa et al. (1991) are followed for provenance discrimination. The discriminant diagrams of Basu et al. (1975) and Roser and Korsch (1988) are followed for provenance characterization, whereas the tectonic setting is distinguished following the discriminant diagrams of Bhatia (1983), Dickinson et al. (1983), Roser and Korsch (1986), Kroonenberg (1994), and Verma and Armstrong-Altrin (2013, 2016)). The chemical index of weathering (CIW) and the chemical index of alteration (CIA) are used to measure the degree of source area weathering. The CIA is calculated following the equation of Nesbitt and Young (1982) as CIA = [Al2O3/(Al2O3 + K2O + CaO* + Na2O)] × 100. The CIW is calculated based on the equation of Harnois (1988) as CIW = [Al2O3/(Al2O3 + CaO* + Na2O)] ×100 (where CaO* is CaO content in silicate fraction: CaO*= CaO-(10/3) × P2O5).

Provenance

In this study, the provenance analysis of the sandstones is based mainly on both petrography and whole-rock geochemistry (major and trace elements). Petrographic characteristics of sandstones can provide significant clues for reconstructing their depositional settings especially when fossil datasets are unobtainable or inadequate (Akinlotan et al. 2021). Generally, quartzose and litho-quartzose are the most common type of sandstones in all studied samples. Due to the scarcity of feldspars in the studied sandstones, the provenance was chiefly inferred by relying only on crystallinity and extinction of quartz grains. The high quartz contents (average 92.26%) may reflect that the samples were subjected to intense periods of chemical and physical weathering and possible derivation from pre-existing sandstones (most probably multi-recycling of Paleozoic sandstones). The quartzose sandstones are generally characteristic of fluviatile sediments (cf. Kolodner et al. 2009; Ahfaf et al. 2021). Based on the bipyramidal plot diagram after Basu et al. (1975) and modified by Tortosa et al. (1991), the ratio of polycrystalline quartz grains versus undulose (strained) and non-undulose (unstrained) monocrystalline quartz grains (Table 1) indicates that quartz grains are possibly derived from plutonic igneous rocks (Fig. 10). In addition, the relative abundance of the unstrained quartz grains in the investigated sandstone samples (Table 1) points to a plutonic igneous provenance (Basu et al. 1975; Hindrix 2000).

Fig. 10
figure 10

Point-count data of the studied Nubian sandstones plotted on the diagram of Tortosa et al. (1991) for provenance discrimination

The heavy mineral assemblage recognized from the studied sandstone samples comprising zircon and rutile might be derived from acidic igneous rocks and/or recycled older sediments (Folk 1974; Morton 1985; Kolodner et al. 2009), whereas garnet and tourmaline grains were derived likely from metamorphosed igneous rocks, especially gneisses and schists (Morton 1985; Morton and Hallsworth 1994). The ZTR-dominated assemblages beside the low plagioclase/feldspars ratio in the examined sandstone samples indicate their derivation from recycled older siliciclastic rocks. The compositional and textural maturity of the sandstones confirms the predominance of recycling processes within the basin (Wanas and Assal 2021).

The discrimination function diagram after Roser and Korsch (1988) indicates that the studied sandstones were sourced from polycyclic older quartzose sandstones (Fig. 11). Different proxies such as Al2O3 versus TiO2 and Zr versus TiO2 cross plots after Hayashi et al. (1997), Y/Ni versus Cr/V cross plot after Hiscott (1984), and Ni versus Cr cross plot after Garver et al. (1996) indicate mixed felsic and mafic source rocks of the upper continental crust (UCC) for the studied sandstone samples (Fig. 12A–D). Plotting of the studied samples on the ternary Ni-V-Th*10 diagram of Bracciali et al. (2007) and on the bivariate Ni versus TiO2 diagram of Floyd et al. (1989) indicates a felsic source for the studied sandstones (Fig. 13A, B).

Fig. 11
figure 11

Provenance of the studied sandstones based on discriminant function diagram of Roser and Korsch (1988)

Fig. 12
figure 12

Bivariate plot diagrams showing the provenance of the studied Nubian sandstones: A Al2O3 versus TiO2 diagram (after Hayashi et al. 1997), B Zr versus TiO2 cross plot diagram (after Hayashi et al. 1997), C Cr/V versus Y/Ni cross plot (after Hiscott 1984), and D Ni versus Cr diagram (after Garver et al. 1996)

Fig. 13
figure 13

A Plots of the studied sandstone samples on the Ni–V–Th*10 ternary diagram after Bracciali et al. (2007) and B Ni/TiO2 bivariate diagram of Floyd et al. (1989) used for provenance determination of the analyzed sandstone samples

The studied sandstones show high Ba and Co concentrations, and the high Ba/Co ratio confirms the derivation from a weathered intermediate igneous provenance (Cullers et al. 1988; McLennan et al. 1993; Cullers 2000; Akarish and El-Gohary 2008; Löwen et al. 2018). Trace element values given in Table 3 show that the studied sandstones are rich in Ba, Th, and Sr, while exhibit depletion in transitional elements (e.g., Cr, Ni, and V). Notable negative anomalies of Cr, Rb, Sr, and Ba (see Fig. 9) are most probably attributed to the remarkable paucity of feldspars. On the other hand, high positive anomalies of Co, Zr, and Nb suggest prevailing felsic than mafic provenance. The low concentrations of Cr and Ni may suggest a mafic provenance, and the minerals containing these elements (e.g., pyroxenes) were broken down over time as a result of chemical and physical weathering (Zhang et al. 1998; Osae et al. 2006; Akarish and El-Gohary 2008). Relatively high concentration of Zr indicates increasing of heavy minerals, especially zircon and garnet.

Tectonic control of deposition

The results of the undertaken petrographic examination imply that the studied sandstones are distinguished by high content of quartz, predominance of monocrystalline quartz, and scarcity of feldspars and lithic fragments. These characteristics are consistent largely with deposition in the continental craton interior basins (Crook 1974; Potter 1978; Taylor and McLennan 1985). Additionally, plotting detrital mode data of the analyzed samples on the QFL ternary diagram of Dickinson et al. (1983) suggests that the Nubian sandstones are mainly of cratonic interior and quartzose recycled orogenic tectonic provenances (Fig. 14) (cf. Taylor and McLennan 1985; Ahfaf et al. 2021).

Fig. 14
figure 14

Ternary plot diagram of detrital components of the studied sandstones on the tectonic provenance discrimination diagram of Dickinson et al. (1983). Abbreviations: Q, total quartz; F, feldspars; L, lithic fragments.

Percentages of quartz in the sandstones can also be used to recognize their tectonic settings. According to Crook’s classification (1974), the studied sandstones are quartz-rich (average 92.26%); thus, they are consistent chiefly with the sandstones of the Atlantic-type (i.e., passive margin).

Plotting of the major element values on the ternary and bivariate diagrams of Bhatia (1983), Roser and Korsch (1986), Kroonenberg (1994), and Verma and Armstrong-Altrin (2016) (Figs. 15A–C and 16A, B) ascertains that the studied sandstones were deposited mainly in the passive continental margin. Plotting trace element values on the logarithmic function diagrams after Verma and Armstrong-Altrin (2016) (Fig. 17A, B) implies a passive continental margin setting of the studied sandstones. The employment of the function diagram of Verma and Armstrong-Altrin et al. (2013) implies that the studied sandstones are attributed to continental rift setting (Fig. 17C).

Fig. 15
figure 15

Plots of the major element composition of the studied sandstones on the discrimination diagrams for tectonic setting (after Bhatia (1983)).

Fig. 16
figure 16

Plots of the major element composition of the studied sandstones on the discrimination diagram for tectonic setting: A after Roser and Korsch (1986) and B after Kroonenberg (1994).

Fig. 17
figure 17

Tectonic setting discrimination of the Nubian sandstones. A Multi-dimensional discriminant function diagram for the discrimination of tectonic settings after Verma and Armstrong-Altrin et al. (2013). Multi-dimensional discriminant function diagrams for the discrimination of active and passive margin settings after Verma and Armstrong-Altrin (2016): B using major elements and C using both major and trace elements. Abbreviations: Arc, island or continental arc; Rift, continental rift; Col, collision

Source area palaeoweathering and palaeoclimate

In the studied sandstones, the high proportions of monocrystalline quartz (average 94.85% of the total quartz), less amount of polycrystalline quartz grains (average 5.41% of the total quartz), as well as the very scarcity of feldspar minerals (average 1.17%) (Table 1) indicate a prolonged chemical palaeoweathering of the source rocks and several phases of recycling (Dabbagh and Rogers 1983; Wanas and Abdel-Maguid 2006; Al-Habri and Khan 2008).

The calculated CIA and CIW values for the studied sandstone samples range from 86.23 to 99.46 (average 96.30) and 87.31 to 99.73 (average 97.62), respectively (Table 2). These high values of the CIA and CIW indicate intensive weathering and reflect warm humid climatic conditions in the source rock area. The CIA values for the studied sandstones are plotted in the Al2O3 - (CaO* + Na2O) - K2O ternary diagram after Nesbitt and Young (1982) (Fig. 18A). The highest degree of alteration (CIA) points to felsic-granitic source. As indicated in Table 2, the high values of the CIA and CIW are attributed mainly to the depletion of mobile elements (e.g., Ca, Na, and K) relative to the less mobile and residual elements (Al and Ti) (Nesbitt and Young 1982). On the ternary Al2O3-Zr-TiO2 diagram of Garcia et al. (1991), the analyzed sandstone samples display high Zr and relatively low Al2O3 and TiO2 (Fig. 18B) indicating an increasing of sorting (Nesbitt et al. 1996). The bivariate plot of SiO2 versus total Al2O3 + K2O + Na2O proposed by Suttner and Dutta (1986) was used to identify the chemical maturity of the Nubia sandstones as a function of climate. This plot reveals the increasing of chemical maturity of the studied sandstones under humid and semi-humid climatic conditions in the source area (Fig. 18C).

Fig. 18
figure 18

A Ternary Al2O3–CaO*+Na2O–K2O plot diagram for the studied sandstones (after Nesbitt and Young 1982). CaO* is the molar proportion of CaO in the silicate fraction; and CIA is the Chemical Index of Alteration; B Al2O3–Zr–TiO2 ternary diagram showing the influence of sorting process (after Garcia et al. 1991) and C binary plot diagram of SiO2 versus (Al2O3+K2O+Na2O) for the determination of the climatic conditions during the deposition of the studied sandstones (after Suttner and Dutta (1986))

Palaeogeographical interpretation

The usage of various interpretative approaches based on geochemical and petrographic data permits making a series of the relevant interpretations. Additionally, the position of the samples of the sandstones is illustrated on several provenance-related diagrams (Figs. 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17 and 18). Generally, these interpretations imply that the sandstones were sourced mainly from an older quartzose sedimentary provenance.

In southern Egypt, thick Paleozoic successions are well developed in three sedimentary basins, namely, Uweinat–Gilf, South Nile Valley, and Etbai (Issawi et al. 2016; Issawi and Sallam 2018). The Paleozoic rocks in these basins consist mainly of fluvial and glacial sediments (mostly sandstones and conglomerates). These sediments were classified into the Araba Fm. (Cambrian) at the base, followed upward by the Gabgaba Fm. (Ordovician), Naqus Fm. (Silurian), Wadi Malik Fm. (Devonian), Gilf Fm. (Carboniferous), and Abu Ras Fm. (Permo-Triassic) (Osman et al. 2003, 2005; Issawi and Sallam 2018). These lithostratigraphic units unconformably overlie the Precambrian basement rocks (Klitzsch 1981; Garfunkel 1999, 2002). It is apparently that the Paleozoic sediments in these basins were derived mostly from intense denudation, disintegration, and erosion of the surrounding Precambrian rocks of the Arabo–Nubian Shield and Gebel Uweinat located in southwestern Egypt (Fig. 19). During the Early Cretaceous, multiple cycles of erosion and sedimentation took place (Kolodner et al. 2006), and the Paleozoic rocks were eroded and recycled repeatedly by ancient fluvial streams flowed in poly-directional patterns. This recycling process led to the formation of thick sheets of younger matured cross-bedded sandstones in low-lying basinal areas in the passive continental margin in southern and southeastern Egypt (Fig. 19). In Libya, the country bordering Egypt from the west, Ahfaf et al. (2021) documented a predominant plutonic provenance with some metamorphic supply for the Upper Sarir quartzose sandstones in the Sirt Basin, which are coeval with the Nubian sandstones in southern Egypt and northern Sudan. They also indicated that the Sarir sandstones were formed from a cratonic basement in humid palaeoclimatic environments accompanied by a rigorous chemical palaeoweathering in the source area. These conditions probably produced similar Lower Cretaceous quartzose sandstones in other territories in north and north-eastern Gondwana (Ahfaf et al. 2021). On the other side, in Palestine/Israel which border Egypt from the northeast, Kolodner et al. (2009) attributed the origin of the Lower Cretaceous Nubian sandstones cropping out in the Negev Desert to recycling of relatively proximal Paleozoic sandstones, using U-Pb Sensitive High Resolution Ion Micro-Probe (SHRIMP) dating of detrital zircon from these sandstones. From the foregoing discussion, it is concluded that the studied Nubian sandstones were partly produced by fluvial recycling from Paleozoic sandstones in the proximal surrounding basins, with a considerable contribution of basement denudation. This interpretation is consistent and in agreement with the conclusion of Shawa and Issawi (1978) that attributed the source area of the Nubian sandstones to an exposed landmass located in southeastern and southern Egypt. This provenance patterns need to be brought in correspondence to the palaeogeographical reconstructions of northeastern Africa in the Early Cretaceous in regard to the palaeolocation of the study area.

Fig. 19
figure 19

Early Cretaceous palaeogeographical map (modified after Guiraud et al. (2001)).

According to Guiraud et al. (2001, 2005), the northeastern Africa was embraced by a large epeiric sea in the Cretaceous (Fig. 19). Since the Early Cretaceous, this sea penetrated from the north, i.e., it was a marginal sea of the Eastern Mediterranean (Golonka 2004). The southern edge of the noted epeiric sea reached the southern part of modern Egypt, and it was controlled by a W–E-trending fault system with strike-slip displacements (Guiraud et al. 2005). The land corresponded to the African–Arabian continental landmass, with lowlands dominated by alluvial and lacustrine plains and denudated uplands located not so far from the coastline; the study area was located at the very land–sea transition (Fig. 19). More generally, the Nubia Formation exposed at Gebel Duwi represents a wide intracontinental zone marking a transition between the interiors (denudated land) and the periphery (epeiric sea) of a very large continent. The composition of the studied sandstones at Gebel Duwi reveals its fluvial origin, which corresponds well to the existence of the alluvial plain. Deltaic origin of the upper lithofacies of the Nubia Formation and its gradual transition to the shallow-marine deposits of the overlying Quseir Formation imply long-term transgression that is also well-documented by the regional-scale palaeogeographical reconstructions (Guiraud et al. 2005; An et al. 2017). Moreover, it is evident that the lower and middle lithofacies of the Nubia Formation were deposited on the alluvial plain, but not so far from the coastline, as the lithofacies change occurred already during the deposition of this formation.

The provenance patterns revealed by the present study are in agreement with the palaeogeographical reconstructions (Guiraud et al. 2001, 2005; Wanas and Assal 2021). The detrital material was denudated from the uplifts of the continental interiors. Most probably, these were the uplifts of modern southeastern part of Egypt. However, the spatial orientation of lowlands and uplifts (Fig. 19) allows hypothesizing delivery of some material also from the Arabo-Nubian Shield. Two questions are about the tectonic setting and the palaeoclimate. The results of the present study indicate on a stable tectonic regime. However, the Cretaceous fault activity, volcanism, and rather rapid changes in configuration of the principal palaeogeographical elements (Guiraud et al. 2005) imply that it was not so stable. This controversy should be explained in the terms of insufficient resolution of the provenance-related interpretation techniques for making clear distinction between fully stable and tectonically active regimes of the continental interiors. As for the climate, it appears that it was humid in the source rock areas, i.e., in the denudated uplifts. The relatively close position of the latter to the study area implies that humid conditions could be typical to the alluvial plain too. This is in agreement with some other interpretations (e.g., Selim 2017). However, Beauchamp et al. (1990) noted earlier the dry depositional environments of the sandstones, and Klitzsch (1990) suggested regional climate changes. So, the open question for further investigations is whether the Lower Cretaceous sandstones at Gebel Duwi reflects the accumulation of the entire Nubia Formation in humid conditions or these sandstones deposited during any humid episode.

Although the precise age of the Nubia Formation remains debatable (e.g., Bosworth et al. in Hamimi et al. 2020), this does not challenge the proposed palaeogeographical interpretations in regard to the persistence of more or less the same conditions through the Cretaceous (Guiraud et al. 2001; Wanas and Assal 2021). It is worth to add that the provenance-related interpretations for the Lower Cretaceous sandstones exposed at Gebel Duwi in the Quseir area resemble the outcomes of the previous study of the Middle Jurassic sandstones in the Khashm El-Galala area (Sallam and Wanas 2019). Probably, there were universal patterns of the mechanisms of crystalline basement denudation and fluvial sandstone formation at the long-existed continental interiors–periphery transition of northeastern Africa.

Conclusions

The studied Nubian sandstones at Gebel Duwi (Quseir area, Eastern Desert) are quartzose and litho-quartzose that are enriched in SiO2 and Ba, Co, Th, Zr, and Sr. These sandstones were deposited on an alluvial plain, and they were partly formed through multiple stages of fluvial recycling from Paleozoic siliciclastic rocks, with a considerable contribution of basement denudation in a humid environment, which certainly favored intense chemical and physical weathering. The enriched ZTR (zircon + tourmaline + rutile) content and paucity of feldspars confirm an increasing proportion of detritus recycled from older siliciclastic units. The compositional and provenance features of these sandstones characterize the very transition from the interiors of a large continent and its periphery.