Origin of the gypsum-rich silica nodules, Moghra Formation, Northwest Qattara depression, Western Desert, Egypt

Sedimentary Geology 177 (2005) 41 – 55 www.elsevier.com/locate/sedgeo

Research paper

Origin of the gypsum-rich silica nodules, Moghra Formation, Northwest Qattara depression, Western Desert, Egypt Essam M. El Khoriby Geology Department, Faculty of Science, Mansoura University-35516, Mansoura University, Mansoura, Egypt Received 5 February 2004; received in revised form 21 September 2004; accepted 28 January 2005

Abstract Gypsum rich-silica nodules appear in two shale horizons of the Moghra Formation (early Miocene) northwestern Qattara Depression, Western Desert, Egypt. These nodules are gray to milky white in colour, mostly botroidal and rose-like in shape and range in diameter from 2 to 7.5 cm. The silica nodule-bearing shale is composed mainly of smectite with a little minor kaolinite. The silica nodules consist mainly of quartz and are composed of gypsum-free matrix and gypsum-rich megacrystalline quartz. The matrix consists of microflamboyant quartz (less than 36 Am in diameter) and chalcedony. The megacrystalline quartz occurs as lenticular and prismatic forms (length: 90–250 Am; width: 30–90 Am). The microprobe, petrographic and SEM examinations confirmed the occurrence of gypsum relics (diameter; 2–16 Am) within the megacrystalline quartz. The chalcedony and mosaic microcrystalline quartz occurs as pore-lining and pore-filling cements. The structure of the silica nodules begins with quartzine in its outer rim, then gypsum-free microcrystalline quartz in the middle part and ends with gypsum-rich lenticular to prismatic megaquartz in the center. Field study, petrographic examination and microprobe analysis reveal that the silica nodules were formed by silicification of precursor gypsum nodules deposited in a marginal sabkha environment under an arid climate. The silicification selectively affected the gypsum nodules rather than the surrounding shale and occurred both through gypsum replacement and void filling. Transformation of isopachous chalcedony into mosaic microcrystalline quartz also occurred. The texture of the silica minerals reflects the different physico-chemical conditions under which they crystallized. Spherical nodules grew chiefly by the diffusive supply of the silica, and elongated ones grew by pore water advection. The integrated effect of climate, pH, salinity, crack systems within the sediment and oscillation in the groundwater level and its chemical composition contributed to the formation of the nodules. D 2005 Elsevier B.V. All rights reserved. Keywords: Nodules; Paragenesis; Silicification; Miocene; Qattara Depression

1. Introduction

E-mail address: [email protected]. 0037-0738/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.sedgeo.2005.01.014

Nodules are defined as concretionary bodies in which the authigenic mineral does not incorporate

42

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

clastic materials during growth (Selles-Martinez, 1996). Nodules are sometimes monomineralic and homogeneous, but polymineralic nodules displaying concentric zonation, may also develop. Different mineralogical compositions have been reported for nodule bodies, chert (James et al., 2000; Bustillo, 2001), barite (Pepper et al., 1985), evaporite (Milliken, 1979; Orti et al., 1997), phosphate (Morad and Al-Aasm, 1994) and manganiferous nodules (Mc Kelevy, 1986). There are fewer studies of silica occurrences for continental than marine environments. Silicification is a common diagenetic phenomenon in marine sediments. The source of silica required for development of chert nodules in marine sediments is generally thought to be intraformational redistribution of biogenic silica during diagenesis. Silicification of host sediments may happen even under shallow burial diagenesis (Thiry and Ribet, 1999). Silicification associated with evaporites is complex and the time needed for silicification can be difficult to determine (Krainer and Spotl, 1998). In the sulfate facies of ancient marine evaporites chert nodules are scarce. These nodules are frequently pseudomorphs of precursor anhydrite nodules and may appear as replacement products of stromatolitic carbonates as well as laminated facies developed in pre-evaporitic stages (Orti et al., 1997). The occurrence of modern and ancient Magaditype silica (abiogenic silica) is generally rare in the geologic record. Its recognition, however, has important palaeoenvironmental implications. This is because formation of Magadi type chert requires a delicate balance of silica source (s), hydrologic setting and semiarid climate (Eugster, 1986; Krainer and Spotl, 1998). Intervals of chert nodules can form condensed sections that can be correlated regionally using sequence stratigraphic concepts, and more precisely, a maximum flooding surface of a thirdorder sequence (Haq et al., 1987; James et al., 2000). Although silicification of clastic deposits in the continental environments was described (Eugster, 1969, 1986; Murata and Larson, 1975; Bustillo et al., 1991; Bustillo and Bustillo, 2000; Bustillo, 2001; Simon-Coincon et al., 1996; Krainer and Spotl, 1998), its genesis and mode of formation are not clear. Up till now there are several disputable unanswered questions concerning silica nodules in continental depos-

its. These questions include (1) What is the silica source? (2) What factors affect the growth of the nodules? (3) What is their mechanism of formation? This paper is an attempt to find the answers for the above-mentioned questions via sedimentological study of silica nodules developed within the lower Miocene Moghra Formation cropping out in Talkh ElFawakheir area, northwestern Qattara Depression, Western Desert, Egypt (Fig. 1).

2. Geological setting The Qattara Depression forms one of the most significant features in the north Western Desert of Egypt. It is closed and lies at 134 m below sea level. The Depression is bounded from the north and west by steep escarpments, with an average elevation of about 200 m above sea level. The exposed stratigraphic section around the study area is composed of a sedimentary sequence ranging in age from the middle Eocene to the Quaternary (Fig. 1). The middle Eocene calcareous sediments of the Mokattam Formation form the southern scarp of the depression. The upper Eocene Qasr El Sagha Formation is composed of black shales with coquina and oyster intercalations (Said, 1990). The Miocene rocks in the study area are represented by two formations: a lower Miocene fluvio-marine Moghra Formation and a middle Miocene shallow marine Marmarica Formation (Said, 1962). Sandy and clayey beds of the lower Miocene Moghra Formation form the bottom and the surroundings of the Qattara Depression, where the ground level reaches 50 to 80 m below sea level. In parts of the study area, the Moghra sediments occur as small plateaux and residual hills within the Quaternary sabkhas. The Quaternary deposits are represented by unconsolidated eolian sands, sabkhas and wadi filling unconformably overlying the Miocene rocks (Ball, 1933; El Bassyony, 1995; Aref et al., 2002). The eolian sands occur as seif dunes that are composed of very fine sands with few detrital carbonates. The sabkha sediments cover large areas of the floor and lower slopes of the Qattara Depression, occurring at or below the elevation of 50 m below sea level (Ball, 1933). In the Qattara Depression Aref et al. (2002) distinguished three types of evaporites: 1, 2 and 3

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

43

Fig. 1. Map showing location and geology of the study area, modified after Conoco Coral, 1986.

based on their relative age and ground elevation, in relation to the groundwater level. Type 1 evaporite sediment is the oldest, representing the earliest record of the Quaternary aridity in the Qattara Depression. The evaporite sediments are present as random, isolated or dense, 2–15 cm evaporite nodules that grow displacively within the top of the Moghra clastics. Type 1 evaporite sediment is regarded as an erosional remnant of a sabkha deposit that was formed at time when the floor of the depression stood 5–30 m higher than at present. This type also represents the former level of the groundwater table which may coincide with the arid episodes of the Quaternary (Haynes, 1980) or may have controlled by the groundwater discharge pattern during the Quaternary. Type 2 evaporite sediment forms an indurated rough sabkha surface. It consists of either gypsum/ anhydrite or halite crusts. Type 3 evaporite sediment is recorded as levels lower than types 1 and 2 evaporite sediments, as wet, rough sabkha surface. It represents the last stages of a lowering groundwater table, where

the sabkha surface is recorded in the capillary evaporation zone of near surface groundwater.

3. Methodology The silica nodules were collected from the Moghra Formation exposed at the Talkh El Fawakheir area, northwest of the Qattara Depression (Fig. 1). The dimensions and geometry of the nodules are described according to the scheme of Selles-Martinez (1996). The texture and composition of the nodules and their host sediments were studied using conventional petrographic polarizing microscope, XRD, SEM and electron microprobe. The mineralogical analysis was carried out on 15 powdered bulk samples using a Siemens diffractometer model D5000 with Ni filter, Cu KB radiation. The Siemens Diffractometer was operated at 40 am, 45 Kv, step size of 0.02, and counting time of 1 s. Oriented three clay samples from the silica nodulebearing shales were examined through XRD analysis

44

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

of air-dried, glycolated and heated (550 8C for 2 h) (Hardy and Tucker, 1988). The identification of the clay minerals followed the method of Thorez (1976). Relative proportions of clay minerals were determined using the methods suggested by Moore and Reynolds (1989). A JEOL-JSM-T330 Scanning electron microscope (SEM) was used to investigate the crystal habit and textural characteristics of the silica nodules and their host shales. The samples were covered with a thin layer of gold. Operation conditions were an acceleration voltage of 15 Kv and a beam current of 0.4 nA. Polished thin sections of the nodules were coated with a thin layer of carbon to determine the elemental composition of the inclusions within the nodules. The elemental composition was determined using the Cameca Camebax BX50 instrument equipped with three mass spectrometers and a backscattered electron detector (BSE). Operation conditions for electron microprobe (EMP) analyses were accelerated voltage of 20 kv and a beam current of 10 nA.

4. Facies and the depositional environments The lithostratigraphic sequence exposed at the study area consists of the Moghra Formation (7.3 m thick) at the bottom and the Marmarica Formation

(2.5 m thick) at the top (Figs. 2 and 3A). The Moghra Formation is composed of sandstone, shale and marl facies. The shale beds vary in thickness from 0.5 to 2.7 m, have a green colour, and are composed essentially of silty quartz grains, clay minerals and gypsum veins (1 mm to 6 cm thick). The clay minerals include smectite (86%) and kaolinite (14%) (XRD analysis). Fish and shark teeth, silicified wood fragments and vertebrate remains are present locally in some shale beds (Fig. 2). The silica nodules occur as discrete and random bodies within the upper parts of the T1 and T2 shale beds (Fig. 2). Some silica nodules have been removed from their host shale beds and rest on an aeolian deflation surface (Fig. 3B). The marl facies (2.5 m thick) is greenish yellow and composed of dolomite, clays and quartz (XRD analysis). The Marmarica Formation occupies the uppermost part of the studied sequence. It is composed of quartz silty dolostone facies (2.5 m thick). The dolostone facies is composed of silty detrital quartz (size; 18–54 Am) and well-rounded glauconite grains embedded in well-crystalline dolomitic matrix (Fig. 3C). The dolomite crystals are sachroidal and vary in size from 5 Am to 10 Am (Fig. 3D). Foraminifera and fragments of echinoids and bryozoa are also recorded. The facies and faunal contents indicate that the Moghra Formation was deposited under fluvio-marine

M. Miocene

Marnarica

Fm. Age Bed Lithology No.

T5

L. Miocene

Moghra

T4

. . .. ... . .. . ... . . . . . .. ... . .. . .. .. .. . . . ..

(Facies)

Yellow quartz sandy dolostone.

Green shale with gypsum veinlets (1-2 mm thick). 2m

T3

T2 T1

Greenish yellow marl.

1

0 Quatrz silty green shale with abundant silica nodules, shark tooth and vertebrate bones.

Green shale with gypsum veins (5-6 cm thick). Silica nodules and plant remains are abundant.

Fig. 2. The lithologic log exposed at Talkh El Fawakheir area showing occurrence of the silica nodules at T1 and T2 beds.

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

45

Fig. 3. (A) Photograph showing distribution of the silica nodules in the upper parts of T1 and T2 shale beds of the Moghra Formation, Talkh El Fawakheir area, Qattara Depression. (B) Photograph showing erosional remnants of the silica nodules. (C) Photomicrograph showing the quartz silty dolostone facies composed of scattered subhedral quartz and rounded glauconite grains (G) embedded in well-crystalline dolomite, plane polarized light. (D) SEM photomicrograph showing the dolomite groundmass of the quartz siIty dolomite facies composed of well-crystalline rhombic crystals varying in size from 5 Am to 10 Am.

environment, whereas the Marmarica Formation was deposited under open shallow marine environment (Said, 1962; Sharaf, 1995). The fluvio-marine environment may include coastal lagoons, estuary, delta and swamps. The evaporite nodules are widely distributed within the Moghra Formation exposed in the western part of the Qattara Depression (Aref et al., 2002).

5. Nodules 5.1. Size and shape The classification of Selles-Martinez (1996) has been used in the description and classification of the silica nodules. Based on their shape, the silica nodules under investigation are classified into dominant compound (more than 95%) and rare single (Fig. 4). The compound nodules include botroidal (90%, Fig. 4A, B), crystalline rose-like (7%, Fig. 4C) and platy irregular (3%, Fig. 4D) nodules. The platy irregular

nodules are dominant in the upper part of the T2 shale bed (Fig. 2). The crystalline rose-like nodules vary between 2 and 2.9 cm in length, whereas their width ranges between 1.1 and 2.6 cm. The botroidal nodules can also be classified according to their colour into waxy (86%) and milky white (14%) types. The length of waxy botroidal nodules ranges between 0.5 and 7 cm, whereas their width ranges between 0.5 and 4 cm. The white botroidal nodules vary in length from 0.1 to 4.0 cm and in width from 1.0 to 3.0 cm. There are two types of compound botroidal nodules; those composed of spheres of different diameters (5 mm to 1.5 cm, Fig. 4A), and those of spheres of a nearly similar diameter (0.9 cm, Fig. 4B). The silica nodules sometimes occur as single spheres of diameter 1.5 cm (Fig. 4E) or as single elongated nodules (5 cm length and 2.5 cm width). The nodules were measured to characterize their shape according to Blatt et al. (1972). Length and width measurements were made for nodules that showed only minor abrasional shape modifications. Data obtained are summarized in Fig. 5. The waxy

46

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

Fig. 4. The morphological characteristics of the silica nodules. (A) A compound elongated silica nodule composed of several spherules varying in diameter from 5 mm to 1.5 cm. (B) A compound spherical botroidal silica nodule composed of spherules of a nearly similar diameter (9 mm). (C) A well-crystalline rose-like silica nodule (3.5 cm long) composed of well-crystalline bipyramidal quartz. (D) A platy irregular silica nodule composed of well-crystalline quartz crystals of nearly similar size (3–5 mm in diameter). Note the pore space (length; 0.7 cm to 1.5 cm) developed between the quartz crystals. (E) A single spherical silica nodule of diameter of 1.5 cm.

botroidal nodules are gypsum rich (Fig. 5A) while the milky white and the crystalline rose like-nodules are gypsum free (Fig. 5B, C). In plan view, most of the nodules studied are circular constituting 56% of the total in waxy botroidal nodules and 83% in crystalline rose like nodules. The subcircular nodules vary between 8% in the milky botroidal nodules to 40% in the waxy botroidal nodules. The elongated nodules range between zero in the crystalline rose like nodules to 17% in the waxy botroidal nodules. 5

Width of nodules (cm)

7 6 5

1.5:1

4 2.5:1

3 2 1

A

0 0

1

2 3 4 5 6 7 Length of nodules (cm)

4 3 2

2.5:1

1

B

0 8

3

1:1 Milky white botroidal nodules n = 12 Circular = 75 % Subcircular = 8 % 1.5:1 Elongate = 17 %

1:1 Width of nodules (cm)

Waxy botroidal nodules n = 71 Circular = 56 % Subcircular = 40 % Elongate = 4 %

X-ray diffraction analysis indicates that the silica nodules consist mainly of quartz. The SEM examination shows that the silica nodule-bearing shale is composed of flaky detrital smectite. Based on their colour and crystal size, the silica nodules are characterized by three concentric layers: outer (rind), middle and inner. Each layer has its own textural and silica mineral characteristics.

Width of nodules (cm)

8

5.2. Mineralogy and petrography

0

1 2 3 4 Length of nodules (cm)

Crystalline rose-like nodules n=6 Circular = 83 % Subcircular = 17 %

2.5 2

1:1

1.5:1

1.5 2.5:1

1 0.5

C

0 5

0

0.5 1 1.5 2 2.5 Length of nodules (cm)

3

Fig. 5. Cross-plot of silica nodule dimensions. (A) Length versus width of waxy botroidal gypsum rich nodules. (B) Length versus width of milky white botroidal gypsum free nodules. (C) Length versus width of well-crystalline rose like gypsum free nodules. The 1:1 line is for reference; the 1.5:1 line separates circular from subcircular shapes; the 2.5:1 line. Separates subcircular from elongate shapes.

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

Under the microscope several quartz forms can be distinguished in the silica nodules. These include megaquartz, microcrystalline quartz, chalcedony, quartzine and lutecite. These quartz forms occur as discrete layers and strings, and/or as void-lining and void-fill cement. The abundance of the different forms of silica varies from sample to sample. The quartzine (small fibers of length-slow chalcedony) occupies the outer rim of the nodules or occurs as successive very thin layers (6–15 Am thick). It has a brown colour and is composed of small fibrous quartz crystals with parallel extinction (diameter; 5–10 Am). The quartzine contains no inclusions. Some length-slow small fibrous quartz with undulose or inclined extinction is present. This may be lutecite as described by Millken (1979) and Maliva and Siever (1988). Lutecite occurs as a partial transformation of isopachous chalcedony grown on megaquartz crystals. Maliva and Siever (1988) attributed the formation of the luteicite to recrystallization of quartz chalcedony. Microcrystalline quartz crystals occupy the middle part of the silica nodule. They occur as lenticular and prismatic quartz crystals of different dimensions (length, 3–36 Am; width, 1–14 Am; Fig. 6A) and as microflamboyant quartz. Milliken (1979) described microflamboyant quartz as an interlocking fabric of quartz crystals showing undulatory extinction and highly irregular crystal boundaries. She also noted that microflamboyant quartz is considered a variety intermediate between quartzine and megaquartz. The microcrystalline quartz rarely appears as granular quartz (diameter, 9–15 Am) and does not contain inclusions. Megaquartz crystals occupy the inner layer of the silica nodules. These crystals are the most dominant quartz form developed in the silica nodules. The megaquartz appear as lenticular (length, 140–180 Am; width 80–90 Am) and prismatic (length, 90–250 Am; width, 30–43 Am) crystals (Fig. 6B, C). Megaquartz crystals are selective to the replacement of lenticular and prismatic gypsum crystals. Inclusions of gypsum (diameter, 2–16 Am) are dominant within the megaquartz (Fig. 6C, D). Thin sections and SEM examinations and microprobe analysis confirmed the presence of the gypsum relics in the megaquartz crystals (Fig. 6D, E). Most of the gypsum relics occur as prismatic laths throughout the central part of megaquartz crystals (Fig. 6C, F). Microprobe analysis

47

of the gypsum relics shows that the CaO content ranges between 35 and 39 wt.%, whereas the SO3 content ranges between 42 and 43 wt.% (Fig. 6G). The gypsum relics are generally rare or absent at the margins of megaquartz. Gypsum inclusions in some individual megaquartz crystals have uniform optical orientations, whereas gypsum inclusions in others have random orientations. The former pattern indicates that the inclusions are a remnant of the precursor gypsum crystals. Several cracks were observed within the megaquartz. The shape of these cracks is variable: rectilinear, irregular, branched and V-shaped (Fig. 6F). The isopachous fibrous chalcedony occurs as pore-lining cement (Fig. 6H). Some granular quartz crystals (diameter, 5–30 Am) appear as pore filling cement (Fig. 6I). Some granular quartz crystals also appear to result from the partial transformation of isopachous chalcedony (Fig. 6H). Euhedral bipyramidal gypsum-free megaquartz crystals (diameter, 100–300 Am) also occur as pore-filling cement (Fig. 6I).

6. Discussion 6.1. Shape of the silica nodules Two main types of nodules were recognized; gypsum-rich silica nodules (Fig. 5A) and gypsumfree silica nodules (Fig. 5B, C). The gypsum-rich waxy botroidal nodules (Fig. 5A) are attributed to the shape of the precursor gypsum nodules. The different shapes of the precursor gypsum nodules may be related partly to the strength of crystallization of the precursor evaporites when gypsum generates crystallization pressure of 282 to 1900 kbar/cm3 (Winkler and Stringer, 1972). The gypsum-free milky white botroidal nodules (Fig. 5B) occur as spherical, subspherical and elongated. The spherical gypsum-free nodules may have developed where the silica-bearing solution is supplied by diffusion mechanism, whereas the elongate-shaped nodules most likely formed where the silica-bearing solution was supplied by advection mechanism (McBride et al., 1994, 1999). However, some degree of elongation of the nodules can be formed during growth by diffusion if pore tortuosity is

48

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

highly asymmetric (BjØrkum and Walderhaug, 1990). By analogy with models for nodule growth, the dominance of spherical and subspherical free-gypsum nodules over prolate or elongated nodules in the Moghra Formation indicates that most gypsum-free silica nodules were supplied by diffusion. The elongate nodules suggest that there was asymmetric pore tortuosity that is likely to develop in terrigenous muds dominated by platy clays. 6.2. Paragenetic sequence Most of the silica nodules consist of several spherules. Each spherule is composed essentially of three main concentric layers: the outer (rind), middle and inner (Fig. 7). The paragenetic sequence was determined according to the petrographic observations (Fig. 6). The silica minerals were formed through several successive stages of silicification (Fig. 8) as follows: 1) Volume-for-volume replacement of Ca-sulfate of the gypsum nodules mainly by small fibrous quartzine, microquartz and megaquartz respectively. 2) Deposition of the pore-lining isopachous chalcedony cement around the megaquartz crystals (Fig. 6H). 3) Partial transformation of the chalcedony rim into granular microquartz and finally. 4) Accumulation of pore-filling, gypsum-free megaquartz crystals. The variation in textures of the silica nodules is attributed to the changing chemical composition of the solutions from which they formed. That is the silica nodules formed from water of varying composition, ranging from intermediate between sea water and meteoric water to from meteoric water. This is evidenced by the source of underground water west the Qattara Depression

derived from marine and meteoric water sources (Aref et al., 2002). Milliken (1979) also mentioned that temperatures of silicification of evaporites vary from near-surface temperatures to temperatures no higher than 40 8C. 6.3. Source of the silica The source of the silica required for formation of the silica nodules developed in mudstones is still a controversial subject. Potential sources of silica according to Meyer and Pena dos Reis (1985), McBride (1989), Thiry and Milnes (1991) and Loi and Dabard (2002) include the following: 1) Precipitation from descending meteoric water after dissolution of clayey silicate minerals in the zone of weathering through leaching of the other cations; 2) Dissolution of quartz silt grains due to shale compaction; 3) Silica released from opaline marine skeletal grains (diatoms, radiolaria, and sponge spicules); 4) Released from hydration of volcanic grains forming opal cement; 5) Silica released from silicates replaced by carbonates; 6) Clay mineral transformation including illitization of kaolinite and smectite during burial diagenesis; 7) Dissolution of detrital quartz grains occurring in sand without pressure solution. Most of the groundwater that evaporates in the western part of Qattara Depression comes from the Moghra aquifer which is recharged from Nubian and upper Cretaceous–Eocene aquifer systems (El Bassyony, 1995). Most of the groundwater samples in the Qattara Depression are of chloride type (Mg Cl2 and Ca Cl2) of marine origin (Aref et al., 2002). A few samples are of Na HCO3 and Na2SO4 types of

Fig. 6. Petrographical characteristics of the silica nodules. (A) Photomicrograph showing the middle layer of the silica nodule composed of parallel, lenticular microcrystalline quartz crystals of different dimensions. Crossed Nichols. (B) Photomicrograph showing a sharp transition in crystal size from microcrystalline quartz (left) to prismatic and lenticular megaquartz (right). Crossed Nichols. (C) Close up of (B) showing abundance of the parallel gypsum inclusions (arrows) developed within the lenticular megaquartz crystals. Note also development of the granular quartz crystals (G) (diameter; 9–15 cm) at the contact between the lenticular quartz crystals. Crossed Nichols. (D) SEM photomicrograph showing remnants of the prismatic gypsum crystals (18 cm–125 cm) (arrows) developed in the megaquartz crystals. (E) BSE image showing presence of relicts of gypsum crystals (white) within the quartz crystals (gray) forming the silica nodules. (F) Photomicrograph showing the occurrence of the irregular empty cracks within the megaquartz crystals (arrows). Crossed Nichols. (G) Microprobe analysis chart showing the chemical analysis of gypsum relicts mentioned in Fig. (6E). (H) Photomicrograph showing growth of pore-lining, free-gypsum isopachous fibrous chalcedony (I). Note also the partial transformation of fibrous chalcedony crystals into granular quartz crystals (G). Crossed Nichols. (I) SEM photomicrograph showing pore-filling euhedral bipyramidal quartz crystals. Note an increase in the quartz size toward the center of the cavity. SEM photomicrograph showing development of empty V-shaped pore space between the mega quartz crystals.

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

49

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

Gypsum relics

lo thr ngit ou udi gh na on l se e s ct ph ion er u le

50

Cavity filling megacrystalline quartz free in gypsum relics

Quartzine (Outer rind)

Microcrystalline quartz (middle layer)

Megacrystalline quartz rich in gypsum relics (inner layer)

Fig. 7. Sketch showing the general composition of the of the silica nodule. Note that each silica nodule is composed of three layers: outer, middle and inner.

meteoric origin. This shows either the large effect of original sea water invasion, or the dissolution of the Moghra aquifer water of salts from the host rocks or preexisting salts. Based on the above-mentioned criteria the groundwater in the Qattara Depression played a great role in formation of the gypsum nodules under arid climate. The groundwater also liberated the silica from host sediments of the Moghra Formation. This is evidenced by the following: 1) Absence of opaline marine skeletal grains; 2) Formation of silica nodules near or above the surface of the Moghra clastics. This indicates that the silica nodule-bearing shale has not undergone deep burial; 3) The silica nodule-bearing shale is composed of detrital smectite and kaolinite. This explains the absence of illitization; 4) During the wet periods of the Quaternary, rainfall resulted in the rise of the level of the groundwater in the Qattara Depression (Haynes, 1980). Accordingly, the groundwater partially dissolved gypsum veins and plant remains present in the Moghra Formation with production of sulfuric and organic acids (Bennett and Siegel 1989). These acids resulted in an increase in the acidity of the groundwater and in turn increased dissolution of the clastics of the Moghra Formation and liberation of the silica required for the silicification process. This agrees with the ideas of Thiry and Millot (1987),

Rayot et al. (1992) and Bariteau and Thiry (2001) who mentioned that silica may come from the alteration of clay minerals. 6.4. Mechanism of the silicification process Deposition of silica is probably a complex process related to variations in the physico-chemical characteristics of the groundwater. The main types of continental silicification include pedogenic, groundwater and silicification associated with evaporites (Thiry, 1999). The silicified evaporite nodules are known from rocks ranging in age from Precambrian to the Quaternary, in sediments of widely differing lithologic character, and in rocks with greatly different diagenetic histories (West, 1964; Chowns and Elkins, 1974; Siedlecka, 1976; Tucker, 1976; Milliken, 1979, Orti et al., 1997; Krainer and Spotl, 1998; Thiry, 1999). The silica nodules in the study area were formed as a result of silicification of the gypsum nodules deposited in a marginal continental sabkha (Fig. 9). This is evidenced by the following: 1) the presence of quartzine and lutecite indicating evaporite mineral replacements (Folk and Pittman, 1971; Siedlecka, 1972; Milliken 1979 and Thiry, 1999). 2) The presence of lenticular and prismatic gypsum pseudomorphs in the Quaternary sabkha deposits in the

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

Level of ground water rich in sulfate

W W

1

2

Arid climate

W W W W

51

Cavity resulted from dissolution of gypsum nodule

3

Humid climate

5

St ag e

ilic

ific

Quartzine

of silicification

Isopachous fibrous chalcedony cement of S

4

Second stage

Transformation of isopachous chacedony

Th ird

Rose-like well-crystalline quartz

+

Arid climate

Lenticular and prismatic gypsum crystals

6

Partial transformation of isopachous chalcedony to microquartz crystals

First stage of silicification due to

Gypsum nodules

Clastic of Moghra Fm. (host rock)

Silica rich groundwater

Gypsum vein

Lenticular and prismatic quartz crystals rich in gypsum relics

7

ati

on

Cavity filling megaquartz crystals

Fig. 8. A flow chart showing the paragenetic sequence of the silica minerals and their textures developed in the silica nodules.

north Western Desert (El Bassyony, 1995; ElKhoriby and Issa, 1998 and El-Khoriby, 2000). 3) Abundance of the parallel gypsum relics developed within the megaquartz crystals. 4) The absence of stromatolitic laminated and arrow-head gypsum and absence of the free growth of chevron, cornet and rafted halite characteristic of the salina environment. This also indicates deposition of gypsum nodules in a marginal sabkha with low salinity (Orti et al., 1997). 4) The abundance of random evaporite nodules within the top part of the Moghra clastics located at the western part of the Qattara Depression. The evaporite nodules were formed displacively within the mud flats deposited in the sabkha environment at 25–100 below sea level (Aref et al., 2002).

The occurrence of the silica nodules in two horizons of the upper part of the Moghra clastics is due to the oscillation of the groundwater level (Fig. 9). This means that the silicification process occurred primarily through changes in pH and salt concentration of the groundwater. These changes are due to mixing of fresh continental water that was dominant during humid periods with brines deposited during the dry periods. This is evidenced by existence of three alternating long wet-dry phases in the Qattara Depression during the Quaternary period (Haynes, 1980). The arid climate resulted in evaporation of the groundwater rich in sulfates and calcium and in turn, formation of the gypsum nodules in the marginal sabkhas (Fig. 9A). The humid climate made the

52

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

A

Arid climate (Evaporation)

Marginal sabkha with gypsum nodules

. . . . .. . Marmarica Fm. . .. . . . . . .

.. . . .. ..

.

Qattara Depression Moghra Fm

Fall of the groundwater level

B

Humid climate (Rain fall)

. . . . . . .. . . . . . .. .. . . .

. . . . .. . .

. Plant remains

Gypsum vein

Rise of the groundwater level Partial Dissolution of clastics of Moghra Fm.

C

Aridclimate (Evaporation) Silica nodules

. . . . .. . .

. ..

.

. .. . .. . . . . . .

Fall of the groundwater level

Fig. 9. Model showing the mechanism of formation of the silica nodules (not to scale).

groundwater of low salinity and in turn, partial dissolution of the gypsum nodules scattered in the sabkhas deposits and gypsum veins and plant remains encountered in the sediments of Moghra Formation (Fig. 9B). This led to an increase in acidity of the groundwater and in turn, led to partial dissolution of the siliciclastic sediments of the Moghra Formation. Accordingly, the groundwater became rich in silica. Evaporation of the silica rich solution during the arid climate led to concentration

of the silica and in turn, silicification of the gypsum nodules (Fig. 9C). The silica nodules are irregularly and discordantly distributed in relation to the stratigraphy. This suggests a post-depositional silicification. During the winter period (humid climate), the rainfall led to an increase in the level of the groundwater. During the summer period (arid climate), an increase in the evaporation rate, and lowering of the water table led to increase of silica concentration

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

near or at ground surface. The upward increase in the concentration of silica indicates that the evaporative pumping mechanism played a significant role in the upward movement of the siliceous rich solutions. Groundwater silicification is linked to absolute silica accumulation processes. This may develop in superimposed layers as a result of groundwater fluctuations and/or downcutting of the landscape.

7. Conclusions The conclusions of this study can be summarized in the following: 1) The silica nodules were formed as a result of the silicification of the precursor gypsum nodules formed in marginal sabkha environment. 2) The presence of both the plant remains and gypsum veins encountered in the sediments of the Moghra Formation acted as a catalyst in the liberation of the silica from the clastics of Moghra Formation. 3) The source of silica required for the silicification processes is derived from the dissolution of the clastics of the Moghra Formation during the stability of the sediment–groundwater interface. 4) The silicification process is more selective to the gypsum nodules rather than the silicabearing shale. 5) The occurrence of the silica nodules in two horizons reflects the oscillation in the level of the groundwater. 6) The climatic changes, pH and pore water chemistry played the major role in the development of the silica nodules. 7) There were several stages of silicification leading to the formation of the different silica textures and minerals that characterized the silica nodules. 8) The silicification process was diagenetic and occurred near or at the ground surface. 9) The pores and cracks developed within the silica nodules and their host shales played an important role in passing the silica-rich solutions responsible for the occurrence of the silicification process.

53

10) The silica nodules can provide us with important information about the palaeoclimate, palaeosalinity and palaeotemperature prevailed during their formation. Acknowledgements The Department of Earth Sciences, Uppsala University, Sweden is acknowledged for offering the research facilities including X-ray diffraction, Scanning electron microscope and microprobe analysis. Many thanks are expressed to the reviewers of the present paper, P. Mozley, S. Lugli and B. Sellwood, for their constructive remarks that improved the manuscript. I thank also Prof. Z. Zaghloul and Prof. S El-Beialy for the critical reading of the early draft of the manuscript that improved the language of the article. The author is also grateful to Prof. Medard Thiry who provided him with the relevant articles.

References Aref M, El-Khoriby E, Hamadan M. The role of salt weathering in the origin of the Qattara Depression, Western Desert, Egypt. J Geomorphol 2002;45:181 – 95. Ball J. The Qattara Depression of the Libyan Desert and the possibility of its utilization for power production. Geogr J 1933;82:289 – 314. Bariteau A, Thiry M. Analysis and simulation of the geochemical transfers within the aquifer: the Beauce ground-water and the alteration of the Fontainebleau Sands. Bull Soc Geol Fr 2001; 172:367 – 81. Bennett P, Siegel D. Silica organic complexes and enhanced quartz dissolution in water by organic acids. In: Miles D, editor. Proceedings 6th International Symposium Interaction. Rotterdam7 Balkema; 1989. p. 69 – 72. BjØrkum P, Walderhaug O. Geochemical arrangement of calcite cementation within shallow marine sandstones. Earth-Sci Rev 1990;29:145 – 61. Blatt H, Middleton G, Murry R. Origin of Sedimentary Rocks. New York7 Pentice-Hall; 1972. 634 pp. Bustillo M. Cherts with moganite in continental Mg-clay deposits: an example of bfalseQ Magadi-type cherts, Madrid basin, Spain. J Sediment Res 2001;71:436 – 43. Bustillo M, Bustillo M. Miocene silcretes in argillaceous playa deposits, Madrid Basin Spain: petrological and geochemical features. Sedimentology 2000;47:1023 – 37. Bustillo M, Armenteros I, Blanco J. Silcretes of arid environments in closed continental basins (Miocene, Duero Basin, Spain). In: Elorza J, Bustillo M, editors. Marine and

54

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55

Continental Facies with Siliceous Sedimentary Rocks VI, International Flint Symposium, Excursion Guidebook, Bilbao; 1991. p. 95 – 139. Chowns T, Elkins J. The origin of quartz geodes and cauliflower cherts through the silicification of anhydrite nodules. J Sediment Petrol 1974;44:885 – 903. Conoco Coral. Geological map of Egypt, NH 35 Siwa Oasis; 1986. 1:500,000. El Bassyony A. Sabkhas of Qattara Depression, Western Desert, Egypt—A survey. Sedimentol Egypt 1995;3:13 – 26. El-Khoriby EM. An example of Quaternary sabkha; north western Egyptian border. Sedimentol Egypt 2000;8:75 – 87. El-Khoriby EM, Issa GhI. Sedimentology and mineralogy of recent continental sabkhas, Siwa Depression, Western Desert, Egypt. Egypt J Geol 1998;42:621 – 40. Eugster H. Inorganic bedded cherts from the Magadi area, Kenya. Contrib Mineral Petrol 1969;22:1 – 31. Eugster H. Lake Magadi, Kenya; a model for rift valley hydrochemistry and sedimentation? In: Frostick L, editor. Sedimentation of the African Rifts. Geol Soc Spec Publ, vol. 25; 1986. p. 177 – 89. Folk L, Pittman J. Length-slow chalcedony: a new testament for vanished evaporites. J Sediment Petrol 1971;41:1045 – 58. Haq B, Hardenbol J, Vail P. Chronology of fluctuation of sea level since the Triassic. Science 1987;235(6);1156 – 67. Hardy R, Tucker M. X-ray Powder Diffraction of Sediments. In: Tucker M, editor. Techniques in Sedimentology. Blackwell Scientific Pub; 1988. p. 191 – 228. Haynes C. Geological evidence of pluvial climates in the Nabta area of the Western Desert, Egypt. In: Wendrof F, Schild R, editors. Prehistory of the Eastern Sahara. Academic Press; 1980. p. 353 – 71. James V, Canerot J, Meyer A, Biteau J. Growth and destruction of Bathonian silica nodules in the Western Pyrenees (France). Sediment Geol 2000;132:5 – 23. Krainer K, Spotl C. A biogenic silica layers within a fluviolacustrine succession, Bolzano Vocanic Complex, northern Italy: a Permian analogue for Magadi-type cherts? Sedimentology 1998;45:489 – 505. Loi A, Dabard M. Controls of sea level fluctuations of Ordovician siliceous nodules in terrigenous offshore environments. Sediment Geol 2002;153:65 – 84. Maliva R, Siever R. Mechanism and controls of silicification of fossils in limestone. J Geol 1988;96:387 – 98. McBride E. Quartz cement in sandstones—a review. Earth-Sci Rev 1989;26:26 – 112. McBride E, Picard M, Folk R. Oriented concretions, Ionian coast, Italy: evidence of groundwater flow direction. J Sediment Res A 1994;64:344 – 58. McBride E, Abdel-Wahab A, El-Younsy A. Origin of spheroidal chert nodules, Drunka Formation (Lower Eocene), Egypt. Sedimentology 1999;46:733 – 55. McKelvey V. Subsea mineral resources. US Geol Surv Bull 1986;1689-A [106 pp]. Meyer R, Pena dos Reis R. Paleosols and alunite silcretes in continental Cenozoic of western Portugal. J Sediment Petrol 1985;55:76 – 85.

Milliken K. The silicified evaporite syndrome—two aspects of silicification history of former evaporite nodules from southern Kentucky and northern Tennessee. J Sediment Petrol 1979; 49:245 – 56. Moore D, Reynolds R. X-ray Diffraction and the Identification and Analysis of Clay Mineral. Oxford University Press; 1989. 33 pp. Morad S, Al-Aasm I. Conditions of formation and diagenetic evolution of Upper Proterozoic phosphate nodules from southern Sweden. Sediment Geol 1994;88:267 – 82. Murata K, Larson R. Diagenesis of Miocene siliceous shales, Temblor Range, California. J Res US Geol Surv 1975;3: 553 – 66. Orti F, Rosell L, Salvany J, Ingles M. Chert in continental evaporites of the Ebro and Calatayud basins (Spain): distribution and significance. In: Ruinos-Millan A, Bustillo M, editors. Siliceous Rocks and Culture, Universidad de Granada (Spain), Monografica, Arte y Arqueologia, vol. 42; 1997. p. 78 – 89. Pepper J, Clark S, Witt W. Nodules of diagenetic barite in Upper Devonian shales of western New York. US Geol Surv Bull 1985;1653. 11 pp. Rayot V, Self P, Thiry M. Transition of clay minerals to opal CT during groundwater silicification. Mem Sci Terre 1992; 18:47 – 59. Said R. The geology of Egypt. Amsterdam7 Elsevier; 1962. 377 pp. Said R. Cenozoic. In: Said R, editor. The Geology of Egypt. AA. Balkema Publ; 1990. p. 451 – 86. Selles-Martinez J. Concretions morphology, classification and genesis. Earth-Sci Rev 1996;41:177 – 210. Sharaf, E. Sedimentlogy and mineralogy of some Miocene exposures around Qattara Depression, Western Desert, Egypt. Unpublished MSc Thesis, Mansoura Univ.; 1995. 160 pp. Siedlecka A. Length-slow chalcedony and relicts of sulphates— evidence of evaporitic environments in the Upper Carboniferous and Permian beds of Bear Island, Svalbard. J Sediment Petrol 1972;42:812 – 6. Siedlecka A. Silicified Precambrian evaporite nodules from northern Norway: a preliminary report. Sediment Geol 1976; 16:161 – 75. Simon-Coincon R, Milnes A, Thiry M, Wright M. Evolution of landscapes in northern to south Australia in relation to the distribution and formation of silcretes. J Geol Soc (Lond) 1996;153:467 – 80. Thiry M. Diversity of continental silicification features: example from Cenozoic deposits in the Paris Basin and neighbouring basement. In: Thiry M, Simon-Coincon R, editors. Palaeoweathering, Palaeosurfaces and Related Continental Deposits. Inter Ass Sedimentol, Spec Pub, vol. 27; 1999. p. 87 – 127. Thiry M, Millot G. Mineralogical forms of silica and their sequence of formation in silcrete. J Sediment Petrol 1987;57:343 – 52. Thiry M, Milnes A. Pedogenic and groundwater silcretes at Stuart Creek Opal Field, South Australia. J Sediment Petrol 1991; 61:111 – 7. Thiry M, Ribet I. Groundwater silicification in Paris basin limestones: fabrics, mechanisms and modeling. J Sediment Petrol 1999;60:171 – 83.

E.M. El Khoriby / Sedimentary Geology 177 (2005) 41–55 Thorez J. Practical identification of clay minerals. In: Lelotte G., editor. A Handbook for Teachers and Students in Clay Mineralogy; 1976. Belgium, 90 pp. Tucker M. Quartz-replaced anhydrite nodules (Bristol Diamonds) from the Triassic of the Bristol District. Geol Mag 1976;113: 569 – 74.

55

West I. Evaporite diagenesis in the Lower Purbeck beds of Dorset: Yorkshire. Geol Soc Proc 1964;34:315 – 30. Winkler EM, Stringer PC. Crystallization pressure in stones and concrete. Geol Soc Amer Bull 1972;83:3509 – 14.