Introduction

During growth of population and expansion of urbanization, especially in developing countries without formal planning, residential household areas are overlapping with industrial zones. As a result, millions of people living in close vicinity to industrial areas are facing different health diseases particularly from pollutant emission. Assessment of such pollutant emission and their distribution through the urban dust considered significance in respect to the human health impacts. Contribution of the such particulate on the human health impacts can be physically by inhalation the small sized particles (diameter <10 µm) or chemically by carrying pollutants such as heavy metals or polycyclic aromatic hydrocarbons (Dockery and Pope 1994; Aunan 1996; Smith et al. 1998; Pope and Dockery 1999; Pope et al. 2002; Bernabé et al. 2005; Blaha et al. 2008; Wang et al. 2011; Rai 2013). Moreover, the surface soil is more definitive aspect for tracing such particles (Li et al. 2004; Škrbic and Ðurišic-Mladenovic 2007) where the pedosphere represents archive for the input aerial deposition particulates whether of geogenic/anthropogenic origin. The accumulation of heavy metals in soils is of increasing concern due to its impacts on soil ecosystems and potential health risk as well. The urban and industrial emissions contribute with valuable amount to the anthropogenic heavy metal input (Jones and Olsen 1990: Banerjee 2003; Sharma and Tripathi 2008) Since most of heavy metals are emitted and transported in form of particulates, studies about the urban dust and its particulates are important for defining the origin, dispersion and concentration of heavy metal contamination in surface environment (Banerjee 2003; Duzgoren-Aydin et al. 2006; Shinggu et al. 2007). Among that particulate matters are the magnetic particles. Magnetic particles (spherules and spheroidal particles) are of great importance because they document the influence of specific industrial activities (Lougheed 1966; Dudas and Warren 1987; Jordanova et al. 2004; Maier and Scholger 2004; Bernabé et al. 2005; Shoumkova 2006; Yang et al. 2007; Coz et al. 2009; Magiera et al. 2011; Wang et al. 2014). They can be called technogenic particles (Magiera et al. 2011; Catinon et al. 2014) and as defined by Vassilev and Vassileva (1996) they are produced as a result of softening, partial and complete melting and vitrification of coal minerals, such as clay minerals, chlorite, mica, feldspars, quartz and other fluxing minerals and phases with lower melting points. Consequently, the magnetic susceptibility of the uppermost part of the soils (topsoil) can be enhanced due to the deposition of such magnetic particles that are emitted from industrial plants. Such phenomenon can be used as a proxy to determine the pollution of surface soil by the input of magnetic particles (Hunt et al. 1984; Hay et al. 1997; Petrovský et al. 2000; Wang and Qin, 2006; Kapicka et al. 2008; Kim et al. 2009; D’Emilio et al. 2010; El Baghdadi et al. 2012; Jordanova et al. 2013; El-Hasan and Lataifeh 2013; Xia et al. 2014).

In Egypt, most of the areas consist of dry desert lands; therefore, settlements have concentrated along the banks of the Nile River and its delta where fertile soils have been formed during annual spring floods. The Nile Delta forms a triangle between Cairo and the shore line of the Mediterranean Sea (Fig. 1). Ancient settlements have now grown into big cities. Although the area of the Nile Delta represents only 2.3 % of the whole terrain of Egypt, the Nile Delta bears 46 % of the total cultivated area and accommodates approximately 45 % of inhabitants. Major cities in the central part of the Nile delta are Tanta and Kafr El Zayat, commercial and industrial spots in a region of farmland.

Fig. 1
figure 1

a Location of the study area in the Nile Delta of Egypt; b sampling sites around main industrial and commercial areas; c brick kilns at Kafr El Zyat city. The predominant direction of wind in Egypt (Rose diagram based on data monitored by the Egyptian Meteorological Authority) is shown in the inset

Full size image

Obviously, more than one source is responsible for the air pollution in the study area. Industrial sources include the emission of fumes and gases (e.g., SO2, NOX, and CO2) from chimneys of chemical factories, brick kilns, especially those using diesel fuel, and petroleum refineries, in addition to the dust from phosphate fertilizer factory in Kafr El Zayat. Urban sources include car exhaust and generation of methane resulting from anaerobic decomposition of organic matter present in landfill sites. Burning of open landfill sites can emit smoke and unpleasant gases that pollute the air. What is called the ‘Black Cloud, resulting from the burning of fields by farmers, especially burning rice straw in the fields to avoid the spread of pests and rodents is an agricultural air pollution source. As a result, it is assumed that traces of the various components (particles, dust, and elements of solutions) are fixed to the soils in different amounts and ways. Until now, the distribution characteristics of pollutant emission in the Nile Delta have not been investigated in detail.

This study presents a first evaluation of soil pollution around Tanta and Kafr El in the Nile Delta soil, whereby the major aim is to map the polluted areas and to characterize origin and pathways of anthropogenic particles. This will be a first step towards a database to evolve strategies for remedial measures to minimize environmental degradation.

Materials and methods

Study area

The study area in the Middle Nile Delta is located between 30º42′ to 30º56′N and 30º42′ to 31º05′ (Fig. 1). Its topographic slope is very gentle, dipping northward, changing from about 18 m (+MSL) in the south to about 8 m (+MSL) in the north, with an average value of approximately 1.2 m/10 km. The surficial view of the study area is representative of the main cultivated lands in Egypt, including channels and drains, railways and highway networks, urban and suburban areas and industrial areas. Stratigraphically, the sedimentary sequence of the Nile Delta basin is subdivided into three sedimentary cycles (Rizzini et al. 1978; Stanley and Warne 1993; Atwia et al. 2006). The oldest cycle (Miocene) begins with clayey sediments of the Sidi Salem Fm. and ends with the deposition of a fluvial, littoral, and lagoonal series of the Qawasim Fm. and evaporites of the Rosetta Fm. The second cycle begins with the deposition of the sandy, basal Abu Madi Fm. of Lower Pliocene followed by neritic marine environment of the Kafr El Sheikh Fm. and ends with sediments of fluvial–deltaic and inner-neritic marine environments represented by El Wastani and Mit Ghamr Fms. of Pliocene-Quaternary. The third cycle (Holocene) is represented by the Bilqas Formation of thickness ranges from 10 to 30 m in the Middle part of the Nile Delta region (Atwia et al. 2006). Bilqas formation represents the uppermost exposed subaerial deltaic unit in the Nile Delta region and its valley and constitutes the most fertile agricultural alluvial agriculture soils in Egypt. The mineralogical composition of the Nile Delta alluvial soil can be arranged in the following order: quartz > plagioclase > hornblende > pyroxene > epidote > garnet > K-feldspars, and a clay size fraction predominantly consisting of smectite > kaolinite > illite (Melegy and El-Agami 2004).

The Nile Delta has a Mediterranean climate characterized by scarce rainfall with an annual average rainfall of about 37 mm3, which usually occurs from October to March. Wind speeds range from 6.9 kT in autumn to 8.3 kT in spring, but sometimes exceed this range during the year (Egyptian Meteorological Authority (EMA). The wind direction is variable; however, the predominant direction is north–west (Fig. 1).

The Middle Nile Delta region is well known for agricultural activity over thousands of years, especially the cultivation of rice and cotton. In addition to agricultural and residential areas, this area contains the oldest and most important industrial and commercial cities in Egypt: Tanta and Kafr El Zayat. Tanta is the principal commercial centre in the Middle Nile Delta including an urban area of 33 km2 with a population of 966,000 (2007). Kafr El Zayat is located directly on the right bank of the Rosetta branch of the River Nile. The urban area of Kafr El Zayat is about 17 km2 with a population of 387,555. It has a long history of industrial activity, particularly in the field of chemical and agrochemical production and this region includes the largest fertilizer, pesticide and insecticide factories in Egypt.

Soil sampling and analysis

Since more than one source is responsible for dispersion of the pollutants in the soil as described above, there should be differences in the enrichment of these parameters in relation to the regional differences. Therefore, the study area was differentiated into three main types, e.g., agricultural, urban (residential and commercial), and industrial sites. The anthropogenic components of the soils have been separated and physically and chemically identified.

A total of 88 representative samples (Fig. 1) of top soil (0–3 cm) have been taken from areas of different land uses (industrial, commercial, residential and cultivated sites). Each sample represents an area of about 5 × 5 m and was mixed thoroughly. Sampling close to the main motorway and railways has been avoided as much as possible. Only the uppermost 3 cm has been sampled for this study. These samples were collected using a small plastic shovel. About 0.75–1.25 kg of soil was collected from each sampling site and stored in plastic containers. Cultivated soil samples were collected from the agricultural lands that grow seasonal crops. Soil samples were collected from gardens, graves, playgrounds and parks from commercial zones and different sites near villages and residential areas. During this sampling stage we focused on the residential communities located downwind from the industrial zone to investigate the effect of industrial plants on the surroundings. Soil samples were collected from the cities of Kafr El Zayat and Tanta close to the main industrial plants. For further comparison, burnt material from pills accumulated at the foot of chimneys in industrial sites was also obtained. Moreover, to distinguish the anthropogenic input from natural background, six manual push cores (50 ± 5 cm length and 6 cm diameter) were collected from different land use sites (2 cores from industrial, 2 cores from residential and 2 cores from cultivated sites).

A total of 12 thin sections of different soil types were prepared and investigated under the polarizing microscope. The magnetic susceptibility of collected surface and downcore soil samples was determined using the kappabridge KLY-4S (AGICO, Brno) at the Geozentrum Nordbayern (GZN), Erlangen for monitoring the distribution of industrial emission in the area under investigation. The individual sample was placed in a cylindrical plastic container (ca. 8 cm3) and measured in low field (300 A/m). This procedure was repeated three times for each sample and the average value gives an accuracy of about 0.1 %. The kappabridge operates with a frequency of 875 Hz and a sensitivity of 3 × 10−8 SI. For calibration and comparison with other results, the magnetic susceptibility is divided by the sample weight (mass susceptibility χ), which was determined for air-dried samples before the measuring procedure.

For anthropogenic particles separation, air-dried subsamples (50 g) were wet sieved to different grain size (>125, 63–125, 36–63 µm) after ultrasonic desegregation. Every grain size fraction was transferred to 500-ml glass beaker. In the aqueous bath, the magnetic particles were separated from the bulk sample using a magnetic stirring bar. The wet sample with embedded magnet was lifted to magnetic plate stirrer for 3 min with stirrer rate 100–150 rpm. The magnetic bar is extracted and washed by water in cleaned ceramic plate. This step was repeated until the magnetic bar was extracted apparently without any black film (magnetic particles) attached to its both poles. Both the ceramic plate with wet magnetic particles and the remaining apparently magnetic particle-free sample were dried (40 °C) and subjected afterward to optical microscope for investigation, counting and picking up the industrial and anthropogenic magnetic and non-magnetic particles. To evaluate this method, each run was repeated to ensure that the semiquantitative measurements can be compared between different samples without any partiality. The total numbers of magnetic spherule particles of different grain size fractions were counted into size groups of 10 µm intervals using an eyepiece graticule, a routine described in McLean (1991).

The morphological and chemical characteristics of the selected anthropogenic particles were made at Zentrum für Werkstoffanalytik Lauf, using a LEO 1525 Gemini field emission scanning electron microscope (SEM) operating at 15 kV equipped with X-ray Dispersive Energy Spectrometry (EDX). The separated particles were coated by carbon.

The density of the various magnetic spherules occurring in different grain size fractions was determined based on their mass and volume. 10 samples were weighted for every fraction. Every sample contained 10–15 grains in the finest fraction (<36–63 µm), 5–10 grains in the medium grain size fraction (63–125 µm) and 1–3 grains in the coarse grain size fraction), using a very sensitive microanalytical balance (Sartorius MC5). To calculate the volume, the diameter of the individual spherules was measured using a digitized optical microscope (Leica M420). Then, the density of every fraction was calculated based on the average mass and volume of the spherules. For accuracy, the density of different fractions of pure minerals quartz, apatite and magnetite was calculated. The standard deviation (SD) of these measurements lies between 7 and 17 %.

For mapping of the polluted area, two different methods have been tested: (1) tracing aerial distribution of industrial magnetic particles by semiquantitative measurements (spherules per gram) counted under the optical microscope and (2) magnetic susceptibility measurement “χ” (m3g−1) of the same bulk samples. As pronounced formerly both the methods are based on the evidence that industrial and urban dust contains relatively high amounts of different grain size fractions of magnetic particles and often correlated to heavy metal concentration. The regional distribution of the amount of industrial magnetic particles is considered to reflect the distribution pattern of pollutants as a function of the wind direction in general. Contour map and 3D wireframe maps “as approximative indicators” of semiquantitative data (spherules per gram) and magnetic susceptibility data are shown to visualize the industrial emission and its influence on neighboring areas for the Middle Nile Delta. For documentation of the natural background we have also included examples of vertical magnetic susceptibility logs from shallow downcores and the average magnetic susceptibility values of lowermost parts of the downcores are assumed to represent the natural local background magnetic signal.

Results

Soil characteristics from downcores

To distinguish anthropogenic input from the natural (local) background of soils (geogenic/lithogenic and pedogenic contributions; Hanesch and Scholger 2002; El Baghdadi et al. 2012), the magnetic susceptibility has been recorded in downcores to a depth of 50 cm (Fig. 2). Downcores from cultivated soils show homogeneity with massive, slightly hard, and of medium to coarse subangular blocky structure. They show an upper tillage layer which in general is about 25-cm thick but locally can become up to 40-cm thick. The lowermost parts are characterized by a thin to very thin platy layer structure. By contrast, urban and industrial soils reveal a vertical and spatial heterogeneity in their structure, at least in the upper surface layer (0–15 cm) relative to the cultivated soil. In all soil types, the upper horizons are characterized by medium-sized subangular coagulated particles and single sedimentary grains. Open cracks (0.3–0.7 mm in size) formed by drying (shrinkage cracks) and anthropogenic materials (soot, single grains, e.g., gravel, minerals and rock fragments) are common. The middle part of the downcore soils is more massive, very hard, dewatered, and has a medium-sized subangular blocky structure. The lower part is very fine grained and exhibits a fine to medium platy structure consisting of single plates of about 3–7 mm thick.

Fig. 2
figure 2

Vertical distribution of the magnetic susceptibility within the downcore soil samples of the different land use types; a cultivated; b residential and c industrial lands

Full size image

All profiles display a variability and general enhancement of magnetic susceptibility in the upper part (uppermost 30 cm) but with differences within and between the different soil types. In the lower parts (30–50 cm) the values are quite constant; the variation of the average value of magnetic susceptibility of all soil downcores is 35 × 10−5 m3g−1 (standard deviation of 3.8 × 10−5). This value can be used as an approximate value for the natural local background magnetic signal of the Nile Delta soil.

Soil mineralogy

According to the SEM/EDX analysis, the main primary mineral components are quartz, K-feldspars and biotite with occurrence of amphiboles, pyroxenes, garnet iron oxide minerals and glauconite grains imbedded in very fine grained groundmass of high organic matter content. The urban and industrial soil samples are characterized by occurrence of grain fragments and dark particles most probably of anthropogenic sources while these components are missing in the cultivated areas. These samples contain, black spheroidal particles with different grain size scattered in a fine-grained groundmass (Fig. 3). These non-geogenic particles were separated, estimated and subjected to intensive analysis by SEM/EDX. The separated non-geogenic particles are classified and described in the following.

Fig. 3
figure 3

Internal structures and particles of surface soil samples; a surface soil material with fine-grained, mixed porphyroskelic fabric,//Nic.; b black spherules of industrial source in clayey sediment,//Nic

Full size image

Morphology, chemical characteristic and origin of particles

Anthropogenic particles originating from various sources of human and industrial activities can be classified according to their magnetic characteristics (magnetic, magnetically attracted, and non-magnetic), their morphology (spherules; non-spherules) as seen under the light microscope and their chemical characteristics based on SEM/EDX analysis. Their relative abundance is compiled in Table 1, and their size distribution throughout different soil types is illustrated in Fig. 4.

Table 1 Relative abundance of particles collected in the different surface soil types
Full size table
Fig. 4
figure 4

Anthropogenic particle size ranges within the different soil types

Full size image

Non-magnetic particles are represented by glassy, transparent to opaque spherules of various colors, by mineral and rock fragments, by soot aggregates, glass, plastic fragments, and biogenic particles. Based on chemical and morphological parameters, the non-magnetic particles can be subdivided into siliceous spherules, calcareous particles, Ca-rich particles, Ca–P-rich particles, carbonaceous particles. Furthermore few irregular fragments of different origin are seen.

Siliceous spherules vary from a glassy, transparent to opaque appearance, with diameters ranging from 5 to 500 µm. They are composed predominantly of silica with varying amounts of aluminum, iron, calcium sodium, potassium, magnesium, manganese and carbon. The concentration of Al, Fe, Ca, Na, K, Mg and Mn of spherule decreases with increasing transparency. Siliceous spherules collected from cultivated areas and from both urban and industrial areas can be differentiated by size, color, transparency, surface morphology and element contents occurring together with silica (Fig. 5). In cultivated areas, siliceous spherules are transparent to semitransparent with various colors (white gray, yellow and red). They show a predominantly glassy surface and are characterized by a relatively small particle size (diameter 5–80 µm). The chemical characteristics as inferred from EDX analysis indicate that impurities (Fe, Ca, Na, K, Mg and Mn) do not exceed more than 17 wt% and the Fe contents range from 8 to 12 wt%.

Fig. 5
figure 5

Binocular microscope and SEM photomicrographs with corresponding EDX analysis of; a yellow and gray and b transparent siliceous spherules, collected from cultivated soils

Full size image

In urban and industrial areas the siliceous spherules are commonly spheroidal, frequently opaque and characterized by uneven surfaces and coarser sizes (50–500 µm). The level of impurities as documented by the amounts of Fe, Ca, Na, K, Mg and Mn is relatively high (27–48 wt%). The Fe content can reach 38 wt% in cases where siliceous spherules appear to be mixed with iron-rich materials. Siliceous spherules from industrial and urban areas can be differentiated just by their mean size. In industrial areas the spherules are coarser (100 µm–0.5 mm) than in urban areas (40–250 µm). Furthermore, the most collected urban siliceous spherules were collected from windblown of industrial dust. Along with Al and Fe and other elements (Ca, Na, K, Mg and Mn) analyzed in very low amounts in siliceous spherules from cultivated areas, Cu, Zn and Ti are also found as traces in siliceous spherules from urban and industrial areas (Fig. 6).

Fig. 6
figure 6

Binocular microscope and SEM photomicrograph with corresponding EDX analysis of a mixed siliceous, Fe-bearing spherules collected from urban soil (SE industrial zone). The EDX analysis at the left side shows predominantly the siliceous yellowish part while the composition of the iron part, at the right side. b Represents clusters of spheroidal particles of siliceous and Fe-bearing spherules collected from dust piles near a brick chimney arising from brick kilns

Full size image

Calcareous particles are particles in which oxygen and calcium make up between 73 and 97 wt% and in general contain more than 25 wt% of calcium with minor amounts of other elements (e.g., P, Mg, S, Si, and Na). These particles are further subdivided into calcium-rich particles (Ca content more than 37 wt%) and Ca–P-rich particles (phosphorus content is more than 10 wt%).

Ca-rich particles occur as spherules, tablets, and capsules or have a bar-like shape. They are relatively small in size (30–80 µm). The spherule-shaped particles are transparent, exhibit a yellow color and reveal a granulated surface. They are composed predominantly of Ca with minor amounts (<2.5 wt%) of P, Mg and Na. These particles were found only in cultivated soil samples (Fig. 7a). The other forms (spheroidal, tablet, capsules or bar like) of Ca-rich particles have a pale white color and a size range of 0.8 mm–150 µm. Along with Ca, other elements (Fe, K, Mg, P and S) are present in small amounts (<5 wt%). These types of particles occur in all investigated soil types in the southeastern section of the study area. Their abundance can be ordered according to industrial > urban ≫ cultivated soil. In industrial areas the bar-like forms prevail and become platy and translucent with increasing distance from the potential industrial source (Fig. 7b).

Fig. 7
figure 7

Varieties of non-magnetic anthropogenic particles collected from different soil types: binocular microscope (left) and SEM photomicrograph (right) with a corresponding EDX analysis. a Ca-rich calcareous spherule collected from cultivated soil; b P-calcareous non-spherule particles collected from industrial soil and c Ca-rich calcareous particles

Full size image

Ca–P-rich particles are observed in small amounts forming larger spherules. They are characterized by a white color with a shiny smooth surface and sizes range from 100 to 250 µm. They are composed predominantly of Ca and P with small amounts of K, Mg, Na, Si and Al. Under a binocular microscope, some individual spherules have marbled surfaces with dark brown spots (Fig. 7c). EDX analysis reveals that these brown spots are composed of Si and Al with minor amounts of K, Mg and Na and are, therefore, interpreted as relics of clay minerals from the soil fixed to the surface of the spherules.

Carbonaceous particles occur in varying amounts as spherules in all the soil types analyzed. But, they are not very common in residential and cultivated areas (less than 3–6 %), whereas they are abundant (30–50 %) in high traffic sites of commercial and industrial areas. Carbonaceous particles (Fig. 8) are soft, friable and have diameters ranging from 50 to 150 µm. The surface of spherules can be smooth, rough, porous, spongiform, cellular and sometimes reveal an etched convoluted texture with a dark gray to black color. Carbonaceous spherules are composed nearly exclusively of carbon. Due to the analysis by EDX, for which the samples were coated with carbon, the carbon content of the spherules was not analyzed. The spherules contain minor amounts of S, Si, Fe Mg, Al and sometimes Mg and Ti.

Fig. 8
figure 8

SEM photomicrographs with corresponding EDX analysis of carbonaceous particles collected from urban soil. These articles most likely arise from fuel–oil composition

Full size image

Irregular fragments include fragments of phosphatic rocks, limestone, and other crystalline and sedimentary rocks. Furthermore, aggregates and single grains of minerals (e.g., halite (Na–Cl rich), sulfur (mainly S), apatite (P–Ca rich), gypsum (Ca–S) and barite (Ba–S rich) as shown by SEM/EDX analysis also occur. Other fragments are glass, plastic, rubber, non-magnetic metals, building materials and paint particles, organic and biological elements (pollen grains, vegetables and animal particle grains) and soot aggregates.

Magnetic particles

The magnetic particles represent 65–92 % of the total separated anthropogenic particles in the different studied soil types. Magnetic particles of anthropogenic origin consist predominantly of spherules; but non-spherule shapes also occur. The spherules occur as cenospheres and plerospheres. The terms cenosphere and plerosphere are used according to Fisher et al. (1976) and describe “empty” and “capsulated with smaller particles” spherules, respectively. Hollow spheres are interpreted as a result of gas escaping from plastic spherules bursting during burning processes, as documented by Vassilev and Vassileva (1996). The sizes of spherules vary between 10 and 200 μm where their size decreases with increasing distance from the industrial zone (Fig. 9). The color ranges from black, dark gray, gray to even pale brown, with a shining metallic, submetallic to dull luster (Fig. 10a). Most magnetic particles occur as single grains; others form aggregates embedded in different cement materials. Agglomerates are rare and restricted to industrial zones. Spherules and spheroidal particles are most common. They exhibit different surface structure patterns ranging from smooth, rutted, creased or a marcasitic “brain-like” (Fig. 10b) as described by Matzka (1997) to a crystalline network surface. A network of rhombic, rhomboid forms within the crystalline surface can well be recognized (Fig. 10c). The average bulk density of spherules varies from 4.4 ± 0.7, 4.8 ± 0.9 and 5.4 ± 0.6 g/cm3 for the >125, 63–125, 36–63 µm fractions, respectively. Within the same grain size fraction, shiny luster-surface Ti-bearing spherules have half the density (average 2.8 g/cm3) of the dull rustier surface spherules (average 4.9 g/cm3). No relation between the grain size of the magnetic particles and their chemical characteristics has been detected.

Fig. 9
figure 9

Grain size distribution of the magnetic spherules and the magnetic susceptibility (Mag. Sus., and its natural local background value) of the samples with distance to the industrial area, A–A′ is cross section on the sampling location map (Fig. 1)

Full size image
Fig. 10
figure 10

Characteristic magnetic particles: a overview of magnetic spherules and nonspherical particles, b and c magnetic spherules with marcasitic and crystal pattern surface structure, respectively, d large pores, surface magnetic fragment filled with mixed material, e titanium-bearing magnetic spherules with smooth porous surface casts filled with different material, f nodule-type spherule connected by tails of melted material, g magnetic spherules of different grain sizes embedded in cement matrix; a, f and g are binocular while b, c, d and e are SEM photomicrographs with corresponding EDX analysis

Full size image

Small differences occur in morphological features of magnetic spherules which probably related to the aerial sorting process during transport by wind. In urban areas, magnetic spherules and their fragments commonly exhibit homogeneous cenospheres, characterized by relatively small sizes, a black to dark gray color, and a smooth porous surface. Fragments of spherule walls are frequently observed and sometimes contain numerous surface geometric voids packed with different fine grains (Fig. 10d). The spherules with smooth surfaces commonly contain high amounts of titanium (<48 wt%) as well as iron and oxygen (Fig. 10e). Spherules with a tail-like shape, which are coated by carbon (Fig. 10f) may indicate that they are produced by melting material of relatively high viscosity. Many magnetic spherules collected from residential samples sited within in the windblown (SE) direction of industrial zone contains attached dark red grains resembling that collected from dust piles near a brick chimney (Fig. 6).

In industrial sites, spherules have various shapes which are related to compositional variations in the burning fossil fuels used by different industrial activities. The sizes of these spherules range from 20 to 200 µm. They vary in color from black, dark gray, gray to even pale brown, with a submetallic to dull luster. Most industrial-type spherules occur as plerospheres with a rough and mottled surface. Only minor spherules reveal a smooth surface with abundant holes. Different spheroidal shapes such as subspherical, cask-like or drop-like spherules are observed packed together with fine to very fine magnetic spherules or other mineral grains (quartz, calcite, sulfur and apatite).

SEM/EDX analysis reveal that the spherules are composed predominantly of Fe, O, and C (sometimes contain low amounts of Ti), and minor amounts of Si, Mn, Mg, Al, Ca, Cr and Cu. Commonly attached grains consist of (1) quartz (EDX; predominantly Si), (2) silicates (EDX; predominantly Al and Si, with small amounts of Ca, Mg and S), and (3) apatite (EDX; P and Ca). The common cement matrix material of a red to dark brown color (Fig. 10g) is composed predominantly of Si, Al, Fe, and Ca as well as minor amounts of P and S. Other materials commonly attached to the adhesive material of the grains are usually a pale gray color (Fig. 11) and composed predominantly of P, Ca, S, Zn, and Ba with minor amounts of Fe, Si Na and Ti.

Fig. 11
figure 11

Magnetic particles embedded in matrix collected from Kafr El Zayat industrial area, (a) binocular microscope and (b) SEM photomicrograph with corresponding EDX analysis (c and d)

Full size image

In cultivated soil samples the abundance of spherules is very low. The magnetic particles are usually represented by disturbed, dull gray rough-surfaced cenospheres. Thin angular platy magnetic particles are mainly related to the annual plowing practices.

Frequency and distribution of contaminants

All downcore samples collected near industrial zones (even the cultivated lands) are characterized by the distinctly enhanced magnetic susceptibility in the upper parts down to ca. 25 cm (Fig. 2) compared to the lower parts are recorded in the Kafr El Zayat industrial zone (Fig. 1b) close to the brick kilns in industrial lands (core 72: up to 69 × 10−5 m3g−1), even the cultivated land near the factories show strongly enhanced values (core 49: 56 × 10−5 m3g−1) compared to industrial areas in larger distance (core 87).

A distinct higher magnetic susceptibility in the upper part is also seen in the core from the Tanta commercial zone (up to 49 × 10−5 m3g−1) which is located nearby a factory of rubber, battery cases and molded products.

Spatial distribution of contaminants

The distribution of magnetic particles is quite variable within and between the different soil types. A significant difference in average concentration for cultivated (0. 9 sph. g−1), urban (39 sph. g−1) and industrial areas (66 sph. g−1) is recognized (Table 2) with different levels of contamination. Surprisingly, this is not reflected in the mean magnetic susceptibility values of the respective soil types (industrial: 48 × 10−5 m3g−1 > cultivated: 43 × 10−5 m3g−1 > urban soil: 38 × 10−5 m3g−1) which in the cultivated soils show mean values almost as high as in the industrial areas. This failure in correlation is also seen in a cross-plot of both the data sets showing the correlation for individual sites (Fig. 12).

Table 2 Magnetic susceptibility of bulk soil samples and quantitative analyses of magnetic particle concentration (Sph.g−1) and their fraction sizes in different soil types
Full size table
Fig. 12
figure 12

Plot of magnetic spherule per gram vs. magnetic susceptibility showing the distinctive fields of different soil samples

Full size image

Discussion

Soil contaminants in the study area show a wide variability in morphology and chemical characteristics. Petrographically, identification of the non-geogenic particles is easy because of their shape and appearance which facilitates the separation from the geogenic background material. In SEM/EDX studies, siliceous spherules are well recognized among the non-magnetic anthropogenic particles and more than one source has to be considered to explain the occurrence of such particle types. In industrial zones the brick kilns are most probably the main producer of most of the siliceous spherules. This observation is supported by the analysis of samples collected from burnt material at the foot of chimneys, which contains the same siliceous spherules as found in the surface soil sediments. These samples also commonly contain individual spherules and clusters of red siliceous spherules. Siliceous particles can be produced by the reduction of SiO2 in coal ash by carbon and vaporization of SiO in furnaces followed by atmospheric oxidation and nucleation (Li and Winchester 1993; Xie et al. 2005). Siliceous spherules found in cultivated soil are produced by burning wood and straw in farmlands (Ausset et al. 1998; Olson and Jones 2001; Humphreys et al. 2003). Moreover, aluminosilicates such as clay can be transformed by agricultural fire and anthropogenic heating to sillimanite or mullite (Catinon et al. 2011; Campos-Ramos et al. 2010), forming new minerals primarily not present in the soils.

The other type of the non-magnetic anthropogenic particles which is identified in the study area is calcareous particles. The chemical characteristics and their spatial distribution indicate that they are probably derived from fertilizer additives. There is a close correlation of all these calcareous particles of various shapes and morphological features to the fertilizer factories. They can be dispersed and resuspended as particle dust around the industrial area from the pile, during the grinding processes of raw materials or during the production operations. Such calcareous particles along with other morphological properties were found near phosphatic fertilizer factories as documented in other studies (e.g., Bernabé et al. 2005; Aoun et al. 2010; Silva et al. 2010).

Carbonaceous particles compose principally of carbon with minor amounts of other elements (S, Si, Fe Mg, Al, Mg, Ti) and dispersed predominantly close to busy traffic sites. These types of spherules generally originate from fuel–oil combustion (Griffin and Goldberg 1981; Wik and Renberg 1987; Powell et al. 2002; Umbría et al. 2004; Rodríguez et al. 2009). The varying morphological features depend on the operating parameters of the combustion process as carried out by Goldstein and Siegmunce (1976).

Occurrence other fragments, e.g., halite, sulfur, apatite, gypsum and barite, most probably originated from raw materials used in different factories. It is worth mentioning that you can smell the sulfur odor at all of the sites in the Kafr el Zayat industrial area.

No doubt that the most important type among anthropogenic particles is magnetic spherules. These particles are deposited and accumulated on the surface soil layer in considerable concentration and can represent about 75 % of the total fly ash in the urban surface soil layer (Jones and Olsen 1990). Because of the nature of their origin (e.g., industrial and power plants, metallurgy, vehicular traffic, etc.) and their large specific surface areas, these particles can be an imperative carrier for many heavy metals (Strzyszcz and Magiera 1998; Karimi et al. 2011; Jordanova et al. 2014). Previous studies have shown that their concentration generally corresponds with the spatial distribution of heavy metal (Hanesch and Scholger 2002; Gautam et al. 2005; Sharma and Tripathi 2008). In the study area the magnetic spherules display great surface morphological variance. The variations in the surface morphology of the spherules are interpreted as a result of phase heterogeneity (Vassilev 1992), cooling rates (Blaha et al. 2008), variations in chemical properties (Kim et al. 2009) and/or physicochemical conditions (Shoumkova 2006; Magiera et al. 2011). Moreover, the density measurements reveal that the composition of coarser individual spherules is more porous, hollow and heterogeneous than the finer ones. The origin of different spherules can be concluded on the basis of their chemical composition, their inclusions, attached grains and cement matrix based on SEM/EDX analysis. The occurrence of higher amounts of Si, Al, Ca and Fe is most probably attributed to specific raw materials (iron-rich clay) from brick kilns and spherules produced from brick chimneys as a melting product of the iron-rich clay minerals (Vassilev 1992). By contrast, high contents of Ca, Zn, P, Mn, Ba and Fe are most probably attributed to chemical and fertilizer industries (Abu Khatita 2011). The results indicated that the aerial sorting processes play an important role in the distribution of these particles.

Except for the cultivated soil, there is a relationship between the magnetic susceptibility values and the concentration of the magnetic spherules. While for urban and industrial soils a general increase of particle concentration is also reflected in higher magnetic susceptibility values, the cultivated soils show generally low concentration and high variability in magnetic susceptibility of the individual sampling sites. Reducing the number of magnetic spherules in the cultivated surface soil samples is due to disturbance of the upper (20–30 cm) tillage arable layer. Regardless, agricultural processes and agrochemical applications are also responsible for increasing the magnetic susceptibility values in the cultivated soil than the other soil types. It has been observed that the lower parts of the soil downcores collected from cultivated lands display higher average susceptibility value (39 × 10−5 m3g−1) than the other lower parts of the soil downcores collected from residential and industrial lands (33 × 10−5 m3g−1). Likewise, the core (86) which is collected from the cultivated land is characterized by high susceptibility values in the upper parts than the upper parts of the core collected from adjacent residential site (core 19) in the middle part of the study area. This most probably attributed to different deposition and migration mechanisms through the cultivated lands (Magiera et al. 2006) which reflect the effect of land use on the physical properties of the soil (Maher 1986; Kapicka et al. 2008). Such enhancing of magnetic signal in cultivated lands can attribute to fermentation processes and farming practices such as plowing, application of fertilizer and organic manure and firing in fields by farmers, especially burning rice straw in the fields. Anaerobic microbial Fe in organic-rich soil can promote higher magnetic susceptibility (Schmidt 2009) by a so-called fermentation mechanism. Microaerophilic assimilatory bacteria (magnetotactic bacteria) and anaerobic dissimilatory bacteria (GS-15 type) contribute to the magnetic properties of soils by producing magnetite (Fassbinder et al. 1990; Lovley 1991) which may oxidize to maghemite (Dearing et al. 1996). Meena et al. (2011) documented that the fertile alluvial agriculture soils usually exhibit higher paramagnetic/diamagnetic contribution and weak ferrimagnetic concentration than the surrounding soil of higher ferrimagnetic concentration values in the contaminated area. Otherwise, utilization of pesticides and fertilizer during agricultural practices (Duan et al. 2010; D’Emilio et al. 2010), burning the waste fields after the annual plow (strictly rice straw in studied area) in the presence of a reducing agent, such as organic matter, can lead to increases in magnetic susceptibility by thermal transformation of weakly magnetic iron oxides (e.g., haematite) to secondary (e.g., magnetite or maghemite), very fine (0.05–0.1 µm) more magnetic forms (Le Borgne 1955, 1960; Oldfield et al. 1981; Humphreys et al. 2003). The susceptibility of cultivated soil can also be enhanced by fallen and broken steel fragments from modern farming machinery (Schmidt 2009). As a result the magnetic signal of cultivated soil can affect former processes in addition to the atmospheric emissions. Moreover, the uppermost part of core 19 was characterized by low magnetic susceptibility value than the lower part of the core and lower than the assumed geogenic magnetic signal. This pattern of distribution is most probably due to admixture of the upper parts with various allochthonous grains of urban sources (e.g., quartz and concrete of fine sand grain size) of diamagnetic character.

To demonstrate the distribution of pollutant emission in the studied area, contour maps and 3D grids were established based on the spherule frequency distribution (Fig. 13) and the results of the magnetic susceptibility mapping (Fig. 14). Since the cultivated soils have very low magnetic spherule concentrations and relatively high magnetic susceptibility, the data of cultivated soil samples were excluded in the evaluation of particle distribution. The distribution pattern indicates a main anomaly located around the industrial zone (Kafr El-Zayat city) which systematically decreases in a NW to SE direction corresponding with the main wind direction. In the wind rose (according to EMA) diagram, NW represents the prevailing (26 %) wind direction of the study area beside NWW (14.8 %) and N (14 %) and other imperceptible (2.7–7.5 %) directions. Therefore, it is obvious that the amount and the size of particles correlate strongly with the main direction of the wind. The particles are blown (southeasterly) away from industrial and commercial centers in the main direction of the wind. Industrial areas are characterized by the highest concentration of unsorted spherules, whereas in surrounding urban (commercial and residential) areas, the spherules are lower in concentration and relatively uniform in size. The small discontinuity of the distribution pattern is due to the occurrence of windbreak trees. Compared with the magnetic spherules distribution pattern, a slight shifting of the magnetic susceptibility distribution pattern of the anomaly arising from the industrial zone has been occurred to the south direction. This slight shifting may due to affecting the susceptibility by fin burned clay fragments emitted from the brick kilns and dispersed southerly by the other wind directions (particularly N) than the prevailing one. Another anomaly occurs around the commercial zone (Tanta city), which does not display a systematic decrease as documented in the industrial zone. The abundance of high chimneys (especially of brick factories which have more than 80 chimneys) in the industrial zone results in more homogeneous distribution patterns owing to undisturbed winds at a higher level. Lower chimneys in the commercial zone result in a more inhomogeneous distribution, probably due to local wind disturbances and circulation in lower wind zones induced by buildings, and trees.

Fig. 13
figure 13

Contour (a) and 3D wireframe map (b) illustrating the aerial distribution of industrial magnetic particles

Full size image
Fig. 14
figure 14

Contour map (a) and 3D wireframe map (b) illustrating the aerial distribution of the bulk magnetic susceptibility

Full size image

The occurrence of the anthropogenic pollutant particles within the urban dust of the Nile Delta will definitely influence the air quality of such highly condensed population area (with a population density of up to 1600 inhabitants per km2). Particularly those inhalable particles pose superlative impacts on the human health. It is associated with adverse effects on respiratory system, e.g., allergic airway disease, lung cancer and increased the rate of cardiovascular mortality (Heinrich and Wichmann 2004; Brook et al. 2004; Vermylen et al. 2005; Schwarze et al. 2006; Mills et al. 2009) and immune systems (Leonardi et al. 2000; Bateson and Schwartz 2007; Williams et al. 2011).

Conclusions

Soil around industrial and commercial areas in the Middle Delta contaminated with different anthropogenic pollutant particles show wide variability in morphology and chemical characteristics. The magnetic phases are the major prevailing anthropogenic particles over the investigated area, whereas, other non-magnetic particle types are less dispersed and mostly restricted to specific soil types than the others due to the land use pattern.

  • Cultivated soils are characterized by siliceous spherules originating from wood and straw combustion in farmlands. Calcareous and Ca–P-bearing particles along with non-spheroidal magnetic particles arise from fertilizers and farming practices, respectively.

  • Urban area sediments are characterized by the occurrence of carboniferous and magnetic spherules emitted from fuel–oil combustion due to vehicle traffic. Calcareous non-spheroidal particles arise from the deposition of industrial dust originating from the Kafr El Zayat industrial zone.

  • The industrial area with the chemical and fertilizer plants was responsible for the considerable amounts of magnetic spherules along with other particles, e.g., calcareous and Ca–P-bearing particles. The brick kilns contribute large amounts of Fe-bearing siliceous spheroidal particles and magnetic spherules.

This work represents the first study on the alluvial soil of the Nile Delta (and its valley) and has documented the anthropogenic contamination of the topsoil in the Middle Nile area in the magnetic susceptibility pattern. The pollution by magnetic particles is seen in enhanced magnetic susceptibility values in the upper parts of the downcores compared to the lower parts representing the local natural background magnetic signal. Particle analysis by SEM has given evidence that magnetic particles deposited from atmospheric dust emitted from anthropogenic sources.

It is obvious from mapping of pollutants that the anthropogenic particles in the industrial zone are transported by wind in a SE direction up to a distance of 16.2 km, whereas in contrast the maximum transport distance in the commercial zone is just 4.8 km.

The occurrence of the anthropogenic pollutant particles within the urban dust of the Nile Delta will definitely influence the air quality of such highly condensed population. The concentration of such particles can be detected using environmental magnetic measurements and contaminated areas can be mapped to get constraints on the pollution pattern. However, in the context of this study only a general geochemical characterization of the anthropogenic particles, as deduced from the EDX analyses, is given. The evaluation and spatial distribution of potentially toxic elements and their accumulation in different soil types and fractions needs a more thorough geochemical characterisation which is under work. However, many previous studies in heavily polluted areas all over the world have substantiated the direct link between magnetic susceptibility intensity, concentration of anthropogenic ferromagnetic minerals and heavy metal concentration (Strzyszcz and Magiera 1998; Hanesch and Scholger 2002; Gautam et al. 2005; Sharma and Tripathi 2008) which underlines the relevance of our study for quantification of the pollution in the middle Nile Delta.