Spatial distribution patterns of molybdenum (Mo) concentrations in potable groundwater in Northern Jordan

Abstract

Two hundred and three groundwater samples were collected during March 2011 to June 2012 from the B2/A7 aquifer water supply wells of northern part of Jordan. The physicochemical properties were analyzed in situ for the major cations, anions, while certain heavy metals were analyzed in the laboratory. Some oilshale rock samples were geochemically analyzed. The Upper Cretaceous aquifer (B2/A7) is used as water supply for most of the communities in the study area. It consists of limestone, marly limestone, bedded chert, and minor phosphorite. Hydrochemical results from the B2/A7 aquifer indicate two main water types: alkaline-earth water (CaHCO₃) and alkaline-earth water with high alkaline component (NaHCO₃ ^–, Na₂SO₄). Standard column leaching experiments on oilshale rock samples and the R-mode factor analysis suggest that the sources for elevated Mo concentrations in the groundwater of certain parts of northern Jordan are attributed to water-oilshale interaction, mobility of Mo down to the groundwater and the extensive use of fertilizers within these areas. Molybdenum (Mo) concentrations in the groundwater water range from 0.07 to 1.44 mg/L with an average value of 98 μg/L. They are found to exceed the JISM and WHO guidelines in two areas in northern part of Jordan. Spatial distribution of Mo, using ordinary kriging techniques and the resulting map, shows high Mo concentration in the northwestern part near Wadi Al Arab area reaching concentrations of 650 μg/L and in the southeastern corner of the investigated area, south of Al Ukaydir village, with an average concentration of 468 μg/L. Both areas are characterized by extensive oilshale exposures with average concentration of 11.7 mg/kg Mo and intensive agricultural activities. These two areas represent approximately 33 % of the groundwater in the northern part of Jordan. Mobility of Mo to the groundwater in northern part of Jordan is attributed to two mechanisms. First, there is reductive dissolution of Fe-oxide, which releases substantial adsorbed Mo concentrations. Secondly, there is oxidation of Mo into dissolved forms in sulfide organic-rich system.

Introduction

Molybdenum (Mo) is a transition, naturally occurring element; where the average abundance of Mo indifferent rock types ranges from zero to 40 mg/L (Das et al. 2007). It is primarily found in different mineral forms as molybdenite (MoS₂) and jordisite (amorphous MoS₂), often in association with vanadium, arsenic, or copper (Barceloux 1999). The most common oxidation states of Mo are ⁺6, ⁺5 and ⁺4 (Das et al. 2007; Barceloux 1999) and the predominant states are Mo (IV) and Mo (VI). Its behavior is closely linked to sulfur, and has properties similar to tungsten and vanadium. Molybdenum is important for plant growth and is added to some fertilizers in trace amounts to enhance crop production. However, the main anthropogenic sources of Mo into the environment are coal and porphyry copper mining, oil refining operations, oil sands development, and combustion of fossil fuels (Eisler 2000).

Groundwater in the northern part of Jordan has become an important water resource in recent years, which makes it important to investigate Mo occurrence and distribution in this area. During the past 10 years, water quality problems have increased as more new wells are being drilled and demands on groundwater continue to increase. The population density in the study area is growing and many farms have been developed there which is expected to increase the demand for water resources for irrigation. Studies have shown that increased water demands have lowered the water table in the study area (Al Basha 2012). This will allow oxygen to get into the bedrock aquifers, creating chemical reactions that release molybdenum into the water, thus making the northern Jordan area one case where this issue is a potential problem.

On the other hand, natural and chemical fertilizers use may be an additional source of pollutants to the groundwater. Fertilizers used in northern part of Jordan are mainly manufactured from the famous Jordanian phosphate deposits. Moreover, oilshale in many places contains potential concentrations of heavy metals such as As, Se, Cr, Zn, Mo, V, Cd, and Ni, besides U. This is the case of the oilshale in Jordan (Abed and Amireh 1983; Abed et al. 2009). Consequently, Mo, tied up in bituminous limestone and phosphate deposits common in the rock units of the study area, can be released from soil and rock into the groundwater and drawn into wells.

As far as the authors are aware, the only published data on the mobility of metals from Jordanian’s oilshale are those of Awawdeh and Jaradat (2010), which evaluated aquifers vulnerability to contamination in the Yarmouk River basin, northern Jordan, based on DRASTIC method. Batayneh (2010) studied the heavy metals in water springs of the Yarmouk basin and their potentiality in health risk assessment. Al Qudah and Abu-Jaber (2009) used a GIS database for sustainable management of shallow water resources in the Tulul al Ashaqif Region, NE Jordan. Alomary (2012) did a study for the determination of trace metals in drinking water in Irbid City in north Jordan within the Yarmouk University area only. The samples were collected from three different water types: tap water (TW), home-purified water (HPW), and plant-purified water (PPW). The results showed that HPW samples have the lowest level of trace metals and the concentrations of some essential trace metals in these samples are less than the recommended amounts. Al Basha (2012) had investigated the groundwater vulnerability to pollution with molybdenum in Yarmouk Basin. It has been found that around 6 % of the studied samples have higher values above the permissible limit of WHO and Jordanian standards. However, none of these studies investigated the spatial distribution of Mo in the groundwater resources of north Jordan and the geochemical processes that lead to high Mo concentrations.

Molybdenum is considered an essential trace element and is routinely found in human metabolism. However, high concentrations of Mo in drinking water will lead to toxicity for humans depending on its chemical form (Vyskočil and Viau 1999). Effects of acute Mo toxicity in humans include diarrhea, anemia, and gout; in addition, chronic occupational exposure has been linked to weakness, fatigue, lack of appetite, anorexia, joint pain, and tremor (Smedley et al. 2014). Moreover, the monitoring program from the authorized agencies must continue and the water blending method for reducing the high Mo concentrations in the groundwater to ensure that the drinking water is safe for all people.

The aims of this paper are to investigate (a) the molybdenum quantities in the groundwater resources and its spatial distribution, and (b) the hydrogeochemical processes responsible for the release of molybdenum in the groundwater.

Topographical and geological setting

The study area extends from the highlands east of the Jordan Valley almost to the city of Mafraq in the east, and from the Yarmouk River in the north to the area north of Ajlun and Jarash in the south (Fig. 1). The topographical elevations drops from >1100 m in the Ajlun mountains (Ibillin) to <200 m below sea level in the lower Yarmouk valley. Towards the Yarmouk River and the Jordan valley, the wadis are steeply incised and slope inclinations of >30 % are common.

Fig. 1

Location of the study area

In the area east of Ar Ramtha and north of Nuaymeh, the topography is less steep and the inclination of slopes usually does not exceed 5 %. Average annual rainfall (year) in the area varies from <200 mm/year, in the easternmost part north of Mafraq, to >500 mm/year in a narrow strip stretching from Rihaba in the south to Malka in the north.

The lithostratigraphical sequence and the hydrogeological classification of the different units are summarized in Table 1. The geological map (Fig. 2) shows the distribution of the outcrop areas of the different hydrogeological units and the general geological setting. The Upper Cretaceous, lower Ajlun Group Hummar Formation (A4) forms the oldest exposed geological unit in the study area. Shu’ayb (A5/6) Formation consists of a sequence of marl and marly limestone with intercalations of limestone, dolomite, and shale. The uppermost part of the Ajlun Group and the lower part of the overlying Belqa Group are considered as one hydrogeological unit, the B2/A7 aquifer. It comprises the predominantly limestone Wadi Es Sir Formation (A7), the chalky Wadi Umm Ghudran Formation (B1) the interbedded chert limestone Amman Formation (B2a), and the Al-Hisa Phosphorite Formation (B2b). The B2/A7 unit (aquifer) consists of thin to massive limestone, dolomitic limestone, and dolomite at its base, overlain by a chalk horizon then bedded chert alternating with limestone with phosphorites towards its top. Outcrops of the B2/A7 unit occur in the south and southwestern parts of the study area. It also crops out in many localities in the northwestern part of the study area as a consequence of folding and erosion. It should be emphasized that around 10 m of high-grade phosphorites (up to 35 % P₂O₅), making the topmost part of the B2/A7 aquifer immediately below the oilshale deposits, are present throughout the northwestern part of the study area.

Table 1 Nomenclature of the stratigraphic formations (Masri 1963; El Hiyari 1985)

Fig. 2

Geological map for the study area

The B2/A7 is overlain by the predominantly marly chalky Muwaqqar Formation (B3) which crops out in a strip reaching from the area south of Ar Ramtha via the city of Irbid to the slopes of the Jordan Rift Valley in the west. It also crops out along the Yarmouk valley. The total thickness of the B3 ranges from ∼125 m in the eastern part of the study area to >500 m towards the Jordan Valley in the west and along the Yarmouk River. The B3 can be divided into two parts: lower or oilshale-bearing with a thickness exceeding 100 m in the Yarmouk river basin. This oilshale horizon is black due to the presence of abundant organic matter averaging 14 % (Abed and Amireh 1983). The upper part of the B3 consists of normal yellowish chalk marl with virtually no organic matter. The contact between the two parts is gradational.

The northern Jordan oilshale deposits are kerogen-rich bituminous argillaceous limestone. They were deposited in a shallow marine, open-shelf environment with upwelling currents from the Neo-Tethys Ocean during the Maastrichtain-Paleocene times (Abed and Amireh 1983; Powell 1989; Abed 2013). The origin of the kerogen is dominantly from the organic matter of the marine phyto- and zooplanktons remains that were accumulated in Tethys Ocean that covered most of Jordan during the Upper Cretaceous and Tertiary (Abed and Amireh 1983).

The Eocene Umm Rijam Chert Limestone Formation (B4) overlies the B3 Formation close to the entire northern boundary of the Wadi Al Arab basin. It consists of chalk alternating with bedded chert. Wadi Shallala Formations (B5) overlies the B4 and consists of limestone, chalk, and chert with a maximum thickness of >200 m (Fig. 2). In the plateau area, adjacent to the Yarmouk River, local outcrops of Neogene alkali basalt are present with varying thickness. Alluvial wadi deposits consisting of poorly sorted gravel, sand, and silt are common in the wadis.

Abed (2000) considered the study area as part of the fault block mountain east of the rift. North Jordan area is characterized by the presence of faults, which are grouped into two sets. The first set consists of normal fault strikes NW–SE to WNW–ESE, while the second are strike-slip faults striking E–W. Faults are of Late Tertiary in age (Atallah and Mikbel 1992). The fold belts occur as gentle parallel anticlines and synclines. Another prominent structure in northern Jordan is the Ajlun structure. Folds are obvious in the central and northwestern parts of the area. The fold axes are of two trends; the first is NNE–SSW, and the other is ENE–WSW. Generally, they are plunging to the north. The first trend is formed due to WNW–ESE compression related to the Syrian Arc system, while the second trend is younger than the first; and is caused by NNW–SSE compression; and is related to Dead Sea system (Al-Taj 2008). Fracturing and folding result in a high degree of inhomogeneity in the hydrogeological characteristics of different aquifers. This inhomogeneous character causes aquifer yields and ground water flow direction to vary over the whole area (Mulwa et al. 2005). A characteristic feature of aquifers tapped through boreholes located along or close to faults is that all of them have water yield in excess of 46 m³/h (cubic meters per hour), and boreholes sited on such fault zones are quite deep with an average total depth of 300 m (Al-Taj 2008).

The soils of the area are young and the moisture regime of this area is classified as xeric (rainfall > 200 mm/year) (millimeter yearly). According to the USDA soil taxonomy, the soils in the area are classified as 29 % aridisols, 12 % entisols, 51 % inceptisols, 3 % mollisols, and 5 % vertisols (HTS and SSLRC, 1993). The land use types (Fig. 3) of the study area contain 29 % bare rock with thin soils and urbanization, 23 % natural vegetation, 4 % forest, 17 % irrigated agriculture (cereals, vegetables, fruit trees, olives, bananas and citrus), and 37 % rainfed agriculture (cereals, vegetables, fruit trees, olives, and citrus) (Al Basha 2012).

Fig. 3

Land use of the study area

Hydrogeological setting

The different hydrogeological units, aquifers, and water table contour lines and depth to the aquifers for the study area are present in Fig. 4. The B4 aquifer is the uppermost aquifer in the northern part of the study area. The aquifer materials consist of chalk, chert, and limestone, which are jointed and fractured with solution channels and cavities in the carbonates portions. The aquifer is highly anisotropic and heterogeneous.

Fig. 4

Hydrogeological map and cross section A-A′ (modified after Margana 2006)

Groundwater levels are relatively shallow in the plateau area east of ArRamtha and are sometimes only about 10 m below ground level. The hydraulic conductivities range between 1E⁻⁰⁴ and 1E⁻⁰⁶ m/s, with an average of 5E⁻⁰⁵ m/s. Groundwater recharge to the B4 aquifer is believed to be around 8–10 % of the rainfall (BGR and WAJ 1997). Salameh et al. (2014) concluded, “The groundwater in the different parts of the country is found in a variety of aquifers, overlying each other and separated from each other by aquicludes but nonetheless these different aquifers are hydraulically interconnected. Therefore, extracting water from the deep aquifers overlain by shallower ones is practically quasi an extraction from the shallower aquifers because the water in the shallow aquifer due to its hydraulic interconnection with the lower aquifer will increasingly leak downwards into the lower aquifer via joints, faults, fractures and other weakness zone within the aquicludes separating both aquifers”. Therefore, due to the highly faulting, fracturing and jointing the hydraulic head of the B2/A7 is lower than in the overlying B4 aquifer in areas where the aquifer becomes confined. In addition, due to the fractured solution channels and cavities in the B3 and B4 formations downwards leakage from the B4 aquifer through the B3 oilshale formation has therefore to be expected in the area. This phenomenon was described and discussed in detail by Mulwa et al. 2005; it was reflected by the high degree of inhomogeneity in the hydrogeological characteristics of the aquifers. This inhomogeneous character causes aquifer yields and ground water flow direction to vary over the whole area. In the middle and southern part of the study area, the B2/A7 aquifer forms the uppermost aquifer. The strongly karstified carbonates of the Wadi Es Sir Formation (A7) are partially overlain by the heavily fractured bedded chert/carbonates of the Amman Formation (B2), and all together form the B2/A7 aquifer in Jordan (Table 1). It is covered by a bituminous aquiclude (B3) containing phosphorites, cherts, and chalk, followed by the locally exploited limy B4 aquifer. Thin Plio-Pleistocene basalts cover small areas in the SE′ part of the study area (Fig. 2). Due to the anticlinal structure of the Ajlun area, the strata dip NW-wards resulting in a groundwater flow following the inclination (Siebert et al. 2014). The B2/A7 aquifer is the most important aquifer in the mapped area and is used for the water supply for most of the communities. The complete thickness of the aquifer increases from ∼300 m in the area of the Ajlun Dome in the south to >500 m in the northern and western parts of the study area (BGR and WAJ 1997). The hydraulic conductivities range between 1E⁻⁰³ m/s and 1E⁻⁰⁷ m/s. As a regional mean value, 2E⁻⁰⁵ m/s was assumed (BGR and WAJ 1997). The B2/A7 aquifer is mainly recharged in the Ajlun Dome area in the south. Because B2/A7 is covered by the impervious B3, the groundwater is confined towards the Yarmouk as observable in the Mukheibeh area the piezometry of the B2/A7 aquifer and its limit of confinement are shown in Fig. 4. The annual groundwater withdrawal from the aquifer increased to reach 80 Million cubic meters. The depth to the saturated B2/A7 aquifer is shown in Fig. 4. In the investigated area, there are many wells drilled either for drinking or irrigation purposes. The study area was exploited by >500 wells, ranging from 100 to >1000 m in depth. These wells have different diameters ranging from 6 to 12 in.

Methodology

The water sampling was conducted during the period of March 2011 to June 2012. Two hundred and three groundwater samples were collected (Fig. 5) in cooperation with the Ministry of Water and Irrigation at each sampling site. All samples were collected from the B2/A7 aquifer, which is the main water resource for drinking water in north Jordan. The water samples were collected and stored in 1000 mL polyethylene bottles with zero headspace for inorganic chemical analysis. Water was examined for major cations and anions. Analyses of the cations and anions were performed with an ion chromatograph (Shimadzu) and a flame emission photometer at the laboratories of the University of Jordan, using the standard methods recommended for the required analysis (Arnold et al. 1998). Estimated detection limits for each constituent are shown in Table 2. In addition, 30 randomly selected water samples from the sampled wells were analyzed two to three times to estimate the analytical precision. An overall precision, expressed as percent relative standard deviation (RSD), was computed for all samples. Analytical precision for cations and anions is within 5 % and the charge balance error (CBC) were calculated and is found to be within the permissible limit of ±5 %. For heavy metal analysis, the water samples were placed without filtering in 100-mL polyethylene bottles that were previously treated with grade nitric acid diluted to 50 % with double-deionized water, for a period of 3 days. Water samples were acidified using concentrated analytical-grade HNO₃ for analysis of trace elements to prevent chemical precipitation (0.5 in 500 mL bottle to achieve pH 2). The analyses, including As, Cr, Fe, Mn, Mo, Ni, Pb, Sr, Ce, U, and V, was performed using an inductively coupled plasma-mass spectrometry (ICP-MS) at Acme Laboratories in Canada.

Fig. 5

Wells sampling points

Table 2 Analytical methods used for measuring parameters

In addition, 12 oilshale rock samples were collected from the different occurrences in the study area and analyzed for major and trace elements by Acme Analytical Laboratories in Vancouver, Canada. To understand the chemical composition of minerals containing Mo in the studied area, six samples of oilshale were studied by an ESEM FEI Quanta 600 FEG scanning electron microscope, operated in low-vacuum mode (0.6 mbar), such that gold- or carbon-sputtering was not necessary. A Genesis 4000 EDAX was used for chemical characterization.

To study the mobility of the molybdenum in the oilshale resulting from leaching by infiltrated rainfall in the study area, 500 g of four different 4-mm ground oilshale samples, were loaded in a column 50 mm in diameter and leached by 1 L rainwater collected during the winter months at a slow flow rate (Fig. 6). In addition, an acidic solution (pH = 4) was used as a leaching fluid. The column effluent then collected volume with time. The first 100 mL from each column were collected as ten samples, 10 mL each in 24 h; then 20 samples from each column were collected during 30 days, 10 mL each. The heavy metals were measured using the ICP-Ms.

Fig. 6

Leaching column experiments

In order to interpret the groundwater chemical characteristics, all analyzed groundwater samples were subjected to statistical analysis. All analyzed parameters were subjected to product linear correlation analysis in order to identify the correlations between the parameters. Correlation matrices and factor analysis were constructed by using the computer program Statistica (8). Finally, a hydrogeochemical modeling for the water samples were elaborated on the analyzed samples. This type of study is aided with specialized software. In this study, the geochemical software Visual Minteq (Allison et al. 1991) has been used to detect the species and minerals formation of molybdenum and calculate the saturation index of the different mineral phases present in the aquifer.

Results and discussion

General hydrochemical results

The hydrochemical data of groundwater from the B2/A7 aquifer is presented in Table 3 and a univariate statistical overview of the data is present in Table 4. The chemical composition of the groundwater is rather heterogeneous indicating that the groundwater in the study area is not uniform. The salinity of the groundwater was ranged from 496 to 3460 μS/cm with an average value of 944 μS/cm (Table 4). (TDS ∼ 604 mg/l). The dominant cation is sodium with mean concentration of ∼78 mg/L, followed by calcium, which has a mean concentration of ∼69 mg/L and then by magnesium and potassium, which have mean concentrations of around ∼35 and ∼5 mg/L, respectively. On the other hand, HCO₃ is the major anion, having a mean concentration, ∼282 mg/L, followed by Cl (mean concentration, 126.41 mg/L) and then SO₄ ²⁻ and NO₃ ⁻ with mean concentrations of 93.16 and 16.27 mg/l, respectively. Moreover, the different ionic ratios for Ca, Mg, Na and Cl where calculated in milliequivalents per liter and their averages are presented in Table 4. By correlating the Na/Cl ratio to the Ca/Cl and “sum of equivalents” (Fig. 7a, b) yield vertical trends of calculated Na/Cl ratios indicating dilution by fresh water mainly rainfall (Siebert et al. 2014).

Table 3 Water chemistry for samples from the B2/A7 aquifer in North Jordan

Table 4 Univariate statistical overview of data set of water analysis

Fig. 7

Correlations of a total dissolved ion equivalents and b Ca/Cl ratios versus Na/Cl ratios in groundwater

The pH of the groundwater ranges between 6.41 and 8.09, with a mean of 6.94, indicating that the dissolved carbonates are predominantly in the form of HCO₃ ⁻. To understand the similarities between groundwater in the study area and identification of hydrochemical processes, they have been classified hydrochemically using major cations and anions with conventional Piper trilinear diagram (Piper 1944) and Chadha’s diagram (Chadha 1999). According to Chadha (1999), hydrochemical diagram for classification of natural waters, the water can be classified into two groups: firstly, where alkaline-earth water and weak acidic anions exceed both alkali metals and strong acidic anions, respectively. Such water has temporary hardness and can be classified as Ca, Mg-HCO₃ type and Ca, Na-HCO₃ type as present in Fig. 8. This type is mostly present in the areas of Wadi Al Arab and Irbid. Secondly, where alkali metals exceed alkaline earths and strong acidic anions exceed weak acidic anions. Such water is characterized by high salinity and generally creates salinity problems both in irrigation and in drinking uses. This water can be classified as Na₂SO₄ water. This water type is mainly found in the areas of Al Ukaydir and Ar Ramtha. On the other hand, detailed classification of water types using Piper plot illustrates similar water types (Ca, Mg-HCO₃ type) in Irbid and Wadi Al Arab area. While the southeastern part of the study area is of Na₂SO₄-type (Fig. 8). The water types in the B2/A7 aquifer is influenced by Ca–HCO₃-type indicating recharge from Ajlun and Irbid regions. These two areas receive high amounts of rainfall (>500 mm) and due to the naturally structural features, it has a great influence on the aquifer recharge. The area between Al Ukaydir and Mafraq shows dominance of Na₂SO₄ in groundwater, which might be due to the influence of rock–water interaction processes and high abstraction volumes from the aquifer.

Fig. 8

Hydrochemical diagram for classification of natural water

Water chemistry is mainly influenced by water rock interaction taking place from the recharge area to location of sampling. Hence, it is very important to understand the ongoing geochemical processes in the study area. Gibbs (1970, 1971) differentiated water quality based on evaporation, water–rock interaction and precipitation using ratio plots of Cl⁻/(Cl⁻ + HCO₃ ⁻) and Na⁺/(Na⁺ + Ca²⁺) vs. TDS, is widely used to trace functional sources of dissolved chemical constituents. Gibbs plot (Fig. 9a, b) shows that all samples plot at the center of Gibbs plot, indicating the dominance of weathering and dissolution of rocks in groundwater chemistry. Gibbs plot for cations (Fig. 9b) also supports the dominance of rock–water interaction, but the higher concentration of sodium shows influence of overexploitation of the aquifer and influence of evaporation of percolating rainwater and irrigation return flow.

Fig. 9

a Log ratio plots of Cl⁻/(Cl⁻+ HCO₃ ⁻) and b Na⁺/(Na⁺ + Ca^{2 +})vs. TDS plot showing dominant source of groundwater chemistry

Groundwater temperatures range between 21 and 38.7 °C (mean 30.59 °C). These relatively high temperatures, in the studied samples, can be achieved by a normal to slightly elevated geothermal gradients in the sampled aquifers. However, circulation to 2.5–3 km could produce the estimated temperature of about 100 °C (Siebert, et al. 2014). Similar temperatures were also reported by Mazor et al. (1980) for many thermal springs in the western side of the Dead Sea Rift and appear to reflect a regional similar geothermal regime along this major structural zone.

Field redox potential values (oxidation-reduction potential, ORP) are higher in the shallow wells (close to 150 mV) (Table 4) due to the existence of dissolved oxygen coming from infiltrating water and from both karst and tectonic structures, mainly faults and joints. In the deep boreholes, ORP and dissolved oxygen values decrease, in some cases reaching very reducing conditions (up to −174 mV) most likely in Wadi Al Arab well field. The concentrations of the measured heavy elements like As, Cr, Fe, Li, Mo, Ni, Se, Sr, U, V, and Zn are present in Table 3. The maximum concentration of As, Fe, Li, Mo, Ni, Sr, and V exceeds the maximum permissible limits of the Jordan Standards for drinking purposes (Table 4).

Mo concentration is studied spatially using ordinary kriging techniques as shown in Fig. 10. High Mo concentrations present in the northwestern part near Wadi Al Arab area reach of 650 μg/L and in the southeastern corner of the investigated area south of Al Ukaydir village the average concentration is 468 μg/L. These two areas host approximately 33 % of the groundwater in the northern part of Jordan and contain groundwater with Mo concentrations greater than the JISM (2008) and WHO (2011) drinking water guideline.

Fig. 10

Spatial distribution of Mo in groundwater from B2/A7 Aquifer

Molybdenum mobility and geochemistry

Geochemical background

Molybdenum is a redox-sensitive metal and metal-oxygen bond is easily formed and/or broken (McEwan et al. 2001). Therefore, the Mo-oxyanion complex (MoO₄ ²⁻) requires very little energy and Mo has a high reducing potential. Reductive dissolution of Mn and Fe in suboxic waters is responsible for releasing Mo (IV) in pore water (Schlieker et al. 2001). However, Mo concentration in water is controlled by redox conditions and pH. In oxic waters at pH > 5, Mo occurs as the molybdate oxyanion (MoO₄ ⁻²), although it adsorbs readily to Fe oxides at low to neutral pH (Xu et al. 2006b) as well as to pyrite (Xu et al. 2006a; Goldberg et al. 1996). Under acidic conditions, molybdate adsorbs to organic matter and clays (Wichard et al. 2009). In oxic alkaline conditions, molybdate is stable and is evidenced by conservative behavior in open ocean water (Bruland and Lohan 2003). In addition, Mo can accumulate in aqueous solution by its release via reductive dissolution of Fe and Mn oxides under mildly reducing conditions (Schlieker et al. 2001; Bennett and Dudas 2003). In strongly reducing sulfidic conditions, the formation of FeS or FeS₂ are involved in the immobilization of Mo and the precipitation of MoS₂ (Helz et al. 2004, Amrhein et al. 1993). Although the kinetics of MoS₂ precipitation is slow, the mineral is difficult to produce in laboratory conditions and it is never seen in sedimentary environments (Helz et al. 2011). The conversion of Mo to MoS₂ ⁻⁴ in very reducing conditions has a great influence on its mobility due to its strong affinity to adsorb to pyrite over a range of pH values (Helz et al. 1996, 2011 and Bostick et al. 2003). Moreover, the reduction to dissolved Mo (IV) or transient Mo (V) forms was observed in reducing organic-rich sediments and generation of Mo (IV) or Mo (V) polysulfide anions (S _n ²⁻) by reaction with S was observed in polysulfide-rich waters (Wang et al. 2011; Vorlicek et al. 2004). The reduced mobility of Mo in sulfidic environments is consistent with its observed abundance in oilshale and other sulfide organic-rich sediments.

Outlines in oilshale geochemistry

In the study area, the chemical composition of the oilshale is studied in 12 samples and the results showed that the analyzed samples are not shales but carbonates. The oilshale is studied by the X-ray diffractometry and the scanning electron microscope, and the results indicated that it is made of carbonates (calcite and dolomite), silica as quartz, phosphates as apatite, clay minerals, and pyrite (Fig. 11). The inorganic constituents (and their averages) are carbonates (dominantly calcite, 66.8 %), clays mostly as kaolinite (4.4 %), biogenic free silica (6.2 %), phosphates (5.1 %), and sulfur (0.58 %) in addition to 17.02 % total organic carbon (see also Abed and Amireh 1983; Abed et al. 2009). The oilshale contains high concentration of Mo (ranging from 0.93 to 34.72 mg/kg (Table 5)). The mean value for the 12 samples analyzed is 11.77 ppm. For a better understanding of the natural fluxes responsible for the characterization of Mo in the groundwater aquifer, correlation coefficient matrix analyses were carried out for the geochemical rock data sets (Table 6). Mo has a significant positive correlation with Cu (R ² = 0.78), Zn (R ² = 0.94), Ni (R ² = 0.94), U (R ² = 0.95), Cr (R ² = 0.79), S (R ² = 0.93), and Se (R ² = 0.95) indicating the same source of origin. The negative correlation between Mo and S with TOC (R ² = −0.77 and −0.63, respectively) indicate that Mo is not present as sulfide. Furthermore, S has a negative correlation coefficient with Fe (R ² = −0.57), indicating that most Fe is not associated with the pyrite shown in Fig. 11 and the pyrite, most probably, is only minor.

Fig. 11

Scanning electron micrographs of Framboidal pyrite in oil shale sample

Table 5 Average chemical composition of oilshale

In this study, the hypothesis is that oilshale is responsible for the high concentrations of Mo in some areas especially where the oilshale is highly exposed to the surface. Several studies showed that oilshales are of a great importance in controlling the heavy metals as they act as a reservoir of these metals, which after temporary storage of metals, can act as a source and a sink for metal contamination. Many workers have reported data concerning enriched values of trace elements in oilshales throughout the world (Bopeng et al. (2006), Fouad and El-Rakaiby (2009)). Abed and Amireh (1983) reported Mo values up to 129 ppm in the oilshales of the Yarmouk basin, Jordan.

Furthermore, another potential source for Mo is the high-grade phosphorite (B2b) immediately underlying the oilshale. The Mo average, minimum, and maximum of 65 samples of the phosphorite from the northwestern part of the study area are 10, 0.5, and 60 ppm, respectively (Abed et al. 2015). The samples were taken from several outcrops between Wadi Al Arab and Ziqlab (Fig. 1). The phosphorites are slightly pH dependent and rain water as well as groundwater can leach them and mobilize their content of elements.

These insights into Mo speciation and behavior can explain the distribution patterns in potable ground water from north Jordan. Ambient pH, redox conditions, and proximity to enriched (sulfide mineral or pollution) sources appear to be the dominant controls. The highest concentrations of Mo were found in the groundwater samples with low concentrations of dissolved O₂. This indicates that concentrations are generally higher in anoxic and suboxic conditions.

MO distribution in groundwater

Anoxic groundwater from the B2/A7 aquifer in Wadi Al Arab area has the highest observed concentrations (330 to 650 μg/L mean 427 μg/L), although high values are represented in reducing groundwater from the Al Ukaydir area especially where mineralized with metal sulfides. Lowest concentrations overall are found in most cases from the shallow water depths where it is abstracted from highly fractured and karstificated zones, typically oxic with pH range from 6.3 to 7.15, mean 6.87. It should be mentioned that the low Mo values in this area, the southern or Ajlun area, could also be explained by removal of the oilshale (B3) and the phosphorite (B2b), the two potential sources of Mo. In addition, concentrations of Mo increase down the groundwater flow gradient as indicated by Fig. 10. It is clear from Fig. 10 that the Mo concentrations increased along the ground water flow in the following manner. The B2/A7 aquifer is unsaturated near Ibilin village, moving to the north, the aquifer becomes saturated and first sample was collected from the well (AB1441), it is 255 m deep and the water level is 430 m above sea level (ASL). Next sample was taken from well (AB1375), with a depth of 284 m and water level of about 489 m ASL. Both wells are unconfined and the oilshale horizon had long been eroded (Figs. 4 and 5). Molybdenum concentrations in these two wells are 12 and 23 μg/L respectively. Moving downward, the aquifer becomes deep and confined by the oilshale beds, a sample from the well (AE1000), at a depth of 895 m below ground surface, water level about 185 m ASL and Mo concentration increase dramatically and reach 183 μg/L. Furthermore, many samples were collected from Wadi Al Arab area dominated by outcrops of the oilshale horizon, wells (AE3008 and AE3001), where the depth of the two wells are 460 and 400 m, respectively, below ground surface and the water level is about 75 m ASL. The highest observed concentration of Mo was found in this area reaching 650 μg/L. It should be mentioned that this anomalous Mo area is the same area where the best phosphorite sections are present (Abed et al. Abed et al. 2015). Thus, the two potential sources of Mo, the oilshale and the phosphorites are present within this area.

Along the ground water flow, few redox-controlled parameters were monitored like pH, Eh (ORP) and DO. These parameters are highly sensitive to the presence of oxygen; it is measured along the groundwater flow as present in Table 7. Down gradient redox changes are indicated by a sudden reduction in Eh and concentrations of dissolved O₂, which mark the location of a redox boundary and confined conditions. In addition, down gradient increase in SO₄ concentration occurs in response to oxidation of organic matter and common sulfide minerals pyrite (FeS₂). In the deep part of the confined aquifer, conditions become more strongly reducing and the groundwater at the western part of the aquifer shows evidence of SO₄ oxidation as evidenced by the H₂S smell of water. In addition, in the southeastern corner of the study area (Fig. 2), molybdenum concentrations increase from typically 10 μg/L in the unconfined section of the aquifer, which is occupied by young basaltic layer, to typically 133 μg/L under confined conditions beyond the redox boundary, where Fe- and Mn-reducing conditions occur. Concentrations of SO₄ increase in the confined aquifer (up to 271 mg/L; Table 6) as groundwater has been influenced by fresh and saline influxes of varying recharge water (rainfall, wastewater, and irrigation return flow). The area between Al Ukaydir and Sumayya is highly cultivated by different kinds of vegetables and trees. Siebert et al. 2014 studied the isotope signatures for Wadi Al Arab well field; he concludes that the δ²H and δ¹⁸O of the B2/A7 aquifer has signatures similar to meteoric water line MWL trend with an evaporation process on the percolating water. In addition, the δ¹⁸O and δ³⁴S signatures of sulfate ranges between −5.75 and −5.08 with an average of −5.31 for δ¹⁸O and between −27.80 and −23.40 with an average values of −24.71 for δ³⁴S (Siebert et al. 2014). These values indicate that the origin of sulfate in groundwater is originating from sulfide oxidation process of the marine deposits.

Table 6 Correlation coefficient matrix for rock samples from oilshale (Bold numbers are those significant at 99 % confidence level)

As for groundwater chemistry, the water analysis has been statistically analyzed using linear regression analysis (Fig. 12) and R-mode factor analysis (principal component analysis) (Table 8) by means of STATISTICA (8) software. Figure 12 shows a strong positive linear correlation between Mo and Ni (R ² = 0.67) indicating the same origin for these elements. In addition, Fe and SO₄ ²⁻ have a strong positive correlation with Mo (R ² = 0.53 and 0.38, respectively). Note that, although the correlation between Mo and SO₄ ²⁻ in the water samples in positively significant (R ² = 0.38), it is much lower than the correlation between Mo and S in the oilshale rock samples (R ² = 0.93). Thus indicating that Mo and S are not proportionally released from the rock to the groundwater and other constituent/s are interfering in the distribution of both elements in the rock samples, i.e., the organic matter. This is also suggested by the release of substantial concentration of TOC during the leaching experiments. This result is in agreement with previous findings about the depositional environment of the oilshales in the open shelf of the eastern Mediterranean. Abed and Amireh (1983) demonstrated using chemistry and trace fossils that the sediment/water interface was oxygenated at the time of deposition and the oxygen minimum zone (MOZ) was higher in the water column; i.e., the water column was stratified with oxygen prevailing at the lower water mass. Almogi‐Labin et al. (1993) arrived at the same conclusion using distribution of planktonic and benthic microfossils in the same shelf but further northwest. Al Hasan (2008), Berner and Raiswell (1983) and ALGEO et al. (2004) also arrived to the same conclusion. They found that the negative correlation might indicate that accumulation proceeded under anoxic to euxinic conditions. On the other hand, Al Hasan (2008) found that sulfur has positive correlations almost with all trace elements values and is negatively correlated with Fe and Se (−0.64 and −0.62, respectively). Although strong correlations between total organic carbon (TOC) and trace elements (TE) values might be expected, the results show a weak TOC-TE covariance. This might indicate that TOC content depends only weakly on the metal enrichment process and thus may indicate different sources of TE and the organic matter. Mo occurs principally as the molybdate oxyanion (MoO₄ ²⁻) in anoxic waters (pH > 5) and adsorbs to Fe oxyhydroxides at low to neutral pH (Kaback and Runnells 1980; Dzombak and Morel 1990; Morrison and Spangler 1992; Xu et al. 2006b) and pyrite (Bostick et al. 2003; Xu et al. 2006a). Therefore, this significant positive correlation indicates that the abundance high concentration of Mo in some wells could be attributed to the reductive dissolution of Fe oxides (Schlieker et al. 2001; Bennett and Dudas 2003) releasing adsorbed Mo. In addition to the reduction of Mo to dissolved Mo (IV) or transient Mo (V) forms in sulfide organic-rich system (Wang et al. 2011) influencing the chemical composition of the groundwater.

Fig. 12

Scatterplots of Mo with a Fe, b Ni, c SO₄ ²⁻ and d pH with significant correlation

Table 7 Water chemistry variations of selected wells along groundwater flow path

Factor analysis

The R-mode factor analysis, of the water samples, distinguished different probable sources of contamination in the study area. The five most significant extracted factors (with eigenvalues ≥1), whose variability exceed 79.69 % of the variance in the dataset, can be named as follows (Table 8):

Table 8 Factor analysis for the water samples

• Factor 1: Salinity factor. It embraces EC, Ca, Mg, Na, K, Cl, SO₄ ²⁻, Sr ions. It accounts for around 33 % of the total variability inherited by the analyzed variables. Therefore, any increase in the concentration of these variables will increase the salinity due the dissolution of aquifer matrix or due to the influence of other sources like irrigation return flow or overexploitation of the aquifer.

• Factor 2: Carbonate factor. It is dominated by carbonate variables, which are Ca, Mg, HCO₃ ⁻, and Sr. It accounts for around 19 % of the total variability inherited by the analyzed variables. Factor two clearly shows the proportional dissolution of the carbonate elements from the oilshales overlying the B2/A7 aquifer or from the aquifer itself.

• Factor 3: Anthropogenic factor. It contains seven elements, which are K, NO₃ ⁻, As, Mo, Se, U, and V. It accounts for around 12 % of the total variability inherited by the analyzed variables. It could be explained as an agricultural and phosphorite signature. The relatively high loading of K and NO₃ are the motif of the interpretation of this factor as due to agricultural signature. Uranium and V are rather high in the Jordanian phosphorites as well as in the fertilizer’s industry; up to 234 and 590 ppm respectively (Abed and Sadaqah 2013). Abed et al. (2008) clearly demonstrated that most of the heavy metals including U, V, Mo, and As follow PO₄ in the DAP (diammonium phosphate) fertilizer used in agriculture in Jordan. In addition, nitrate concentrations ranged between 12 and 77 mg/L in this area. It is thus plausible to explain the presence of the associated elements on this factor as partially due to agricultural return flow. This is also emphasized by the high values and an increase trends for these elements observed in the aquifer beneath the agricultural lands. So this factor reflects the diffused contamination of groundwater by fertilizers and anthropogenic sources. Nitrate can also be sourced from liquid waste water sites (Al Akaydir liquid waste site). However, such sites are very limited in the study area and more relevant to the southern and unconfined aquifer.

• Factor 4: Oilshale factor. This is dominated by ten variables, which are SO₄ ²⁻, As, Cr, Fe, Mo, Ni, Se, Sr, U, and V. It accounts for around 9 % of the total variability inherited by the analyzed variables. The elements loaded on this factor are typical for the trace metals, including Mo, normally adsorbed or complex with the organic matter. The role of oilshale in the Mo concentration in groundwater is discussed above.

• Factor 5: Sulfide factor. It contains four elements, which are Eh, SO₄ ²⁻, As, and Fe. It accounts for around 7 % of the total variability inherited by the analyzed variables.

Hydrogeochemical modeling

The recharge to the aquifer by rainfall or irrigated return water flow may dissolve or precipitate minerals and may be subjected to mixing or dilution. These processes depend on many factors like pH, dissolved CO₂, and mineral equilibrium that lead to predictable changes in the chemical composition in the infiltrated water or the groundwater flow in the aquifer itself. Chemical mass balance studies test whether such processes can explain changes seen along a flow path. The results are indicated that the groundwater of the B2/A7 aquifer is oversaturated with respect to carbonate minerals (aragonite, calcite, and dolomite) and goethite and hematite minerals, while it is undersaturated with respect to all other mineral phases. On the other hand, the groundwater is undersaturated to all molybdenum mineral phases with the exception of strontium molybdate. This explains the instability of Mo in this system and its mobility down to the aquifer.

Molybdenum analyses from the leaching test are present in Fig. 13. In the effluent solution (rainfall water pH = 7.5) during the early part of the test, Mo concentrations increased quickly from about zero to 40 μg/L (Fig. 13) by the end of the first day. After the first day, the Mo concentrations increased from about 40 μg/L and reached a maximum of about 90 μg/L after 7 days. The leaching test continued for another 21 days and reached Mo concentrations of 135 μg/L (Fig. 13). After 24 days, a constant value for the Mo concentration was reached (135 μg/L). This Mo concentration remained constant for over 1 week. The initial pH of the leaching water was about 7.5. During the first day, the pH value increased slightly, became more basic, and reached 8.09. The pH ranged from 7.5 to a maximum of 9.07 by the end of leaching test. On the other hand, the Mo concentrations increased and reached 215 μg/L over the same period when acidified distilled water of pH = 4 was used as percolating solution (Fig. 13). In the presently neutral to alkaline pH conditions of the deposit groundwater, Mo is mobile (Fig. 13).

Fig. 13

Changes in Mo concentrations during the leaching test

When acidic solution (pH = 4) percolated through the column, gas bubbles and H₂S smell were noticed due to the reaction with the carbonates of the oilshale present in the sample. In addition, the water became oily in color. Molybdenum is immobile under acidic oxidizing conditions (Boyle 2003) however, what happened here is what is called a geochemical switch where the H₂S/HS-transforms Mo from its conservative status to particle-reactive species (Helz et al. 1996). This switch depends on H₂S activity (Erickson and Helz 2000). These results explain the instability and mobility of Mo in this system and translocation of Mo down to the aquifer. Among the other results of this experiment, traces of organic compound were detected in the effluent (TOC <50 mg/L).

Conclusions

Two hundred and three groundwater samples taken between March 2011 and June 2012 are investigated for physicochemical properties and certain trace metals. Hydrochemical results for the B2/A7 aquifer indicates two main water types in northern Jordan, which are alkaline-earth water (CaHCO₃) and alkaline-earth water with high alkaline component (NaHCO₃–Na₂SO₄) water.

This investigation has attempted to understand the distribution pattern, geochemistry, and sources of Mo and some other heavy metals in the groundwater of northern Jordan. The analysis of 203 groundwater samples shows that Mo concentrations range from 0.71 to 1439 μg/L with an average of 98 μg/L. Spatial distribution, using kriging statistics, of Mo concentrations in the groundwater shows that it exceeds the JISM (2008) and the WHO (2011) guidelines for drinking water in two areas: Wadi Al Arab area in northwestern part of the study area (650 μg/L) and south of Al Ukaydir village in the southeastern part of the study area (468 μg/L).

Several lines of evidence suggest that the sources for elevated Mo concentrations in the groundwater of certain parts of northern Jordan are attributed to water-oilshale interaction, mobility of Mo down to the groundwater and the extensive use of fertilizers within these areas. Oilshale rock sample analysis confirms that the oilshale contains substantial Mo concentrations reaching 11,770 μg/kg Mo in average. Standard column leaching experiments have revealed detectable release of Mo and some other heavy metals to the percolating water with the increase of Mo concentrations over soaking time. Likewise, the R-mode factor analysis shows the association of K, NO₃ with Mo, U, V, and As, indicating an agricultural return flow, as there are no natural sources for NO₃ in northern Jordan and the above elements are relatively high in the DAP fertilizer produced from the phosphorite of Jordan.

Molybdenum could be easily released from oilshale deposits under alkaline and acidic conditions. Moreover, the presence of Mo is most probably associated with the organic matter, which explains its high mobility. Two factors control the high concentration of Mo in some wells of the study area. First, reductive dissolution of Fe-oxide, which releases substantial adsorbed Mo concentrations. Secondly, there is oxidation of Mo into dissolved forms in sulfide organic-rich system.

Factor analysis shows that Mo and certain other trace metals are distributed into five factors that explain about 80 % of the variance in the dataset. These factors are:

• Factor 1: Salinity factor (EC, Ca²⁺, Mg²⁺, Na, K, Cl, SO₄ ²⁻, and Sr)

• Factor 2: Carbonate factor (Ca²⁺, Mg²⁺, HCO₃ ⁻, and Sr)

• Factor 3: Anthropogenic/phosphorite factor (K, NO₃ ⁻, As, Mo, Se, U, and V)

• Factor 4: Oilshale factor (SO₄ ²⁻, As, Cr, Fe, Mo, Ni, Se, Sr, U, and V)

• Factor 5: Sulfide factor (Eh, SO₄ ²⁻, As, and Fe)