Previous studies have identified U₃O₈ at 0.016 wt% tenor according to surface field radioactivity measurements and chemical data around the Kasar-Köprübaşı area in western Turkey [1] - [3] . Yılmaz et al. identified two different types of uranium mineralization (oxidized and unoxidized) in the sandstone and conglomerate formations in the Köprübaşı area, and suggested that leaching of tuff and weathering and erosion of the metamorphic rocks could have supplied this uranium. Studies of the Miocene basin in Turkey have found low tenor and small areas of uranium mineralization close to the studied area [4] . In recent years, Banning et al. examined the relationship between uranium and arsenic in the volcano-sedimentary basins in north central Mexico, and demonstrated that these two trace elements are strongly correlated in volcano-sedimentary rocks [5] . Al-Eshaikh et al. attempted to determine the uranium content of phosphate ores using different measurement techniques, and their results showed a high uranium concentration in the Saudi phosphate ores [6] . Wang et al. clarified the mechanisms of uranium distribution in river sediments and its speciation in river water near a phosphate mine of the Mianyuan River in China that the effective control mechanism is pH values and total phosphate concentration [7] . The main speciation of uranium in the river water near the phosphate mine was when the pH was lower than 7.6. Romaniello et al. examined the relationship between uranium concentrations and ²³⁸U/²³⁵U isotope ratios in modern carbonates from the Bahamas. From the results of this study, it was determined that uranium precipitates in carbonate rocks [8] . The influence of acidic and alkaline fluids properties on uranium migration in subsurface sediments was examined by Szecsody et al., who determined that the combination of acidic and alkaline solutions aids the transport of uranium [9] . These solutions affect uranium dissolution/precipita- tion reactions and uranium sorption, but the effect of these solutions may vary with depth and over time. The distribution of uranium and thorium in dolomitic gravel fill and shale saprolite was examined by Philiphs and Watson [10] . They determined that these rocks contain high concentrations of uranium in the gravel below the water table and high concentration of thorium above the water table.

The main aim of this study is to determine the geological features and geochemical composition of the Rağıllar uranium mineralization area, with a particular focus on the behavior of major, trace and rare earth elements (REEs). The results are discussed in terms of the source area for the uranium. This study provides an opportunity to examine the origin of uranium using by whole rock geochemical data and prospect new uranium mineralization due to determining uranium-rich lithologic units.

The study area is located the geographic coordinates of 38˚50'48.13 N latitude and 28˚26'52.90 E longitude (UTM/50ED) in the Menderes Massif in the Alpide orogenic system (Figure 1(a)). North-trending basins, including the Demirci basin, are located to the north of the East-West trending Alaşehir graben. The presence of these basins was first put forward by Kaya [11] . The Demirci basin is surrounded by the Gördes basin to the west, the Selendi basin to the east, the Simav graben to the north, and the Alaşehir graben to the south in western Turkey (Figure 1(b)). The Menderes Massif rocks in the study area have been studied by many researchers [1] [12] [13] . The floor of the basin in the study area is formed by the Menderes Massif rocks. The metamorphic rocks of the Menderes Massif in the study area can be categorized into two different units in the study area: 1) gneissic metagranites and 2) metasedimentary rocks such as schists. The Menderes Massif rocks in the eastern part and western part (observed only around the Kale crest) of the Köprübaşı-Demirci basin are overlain by Mesozoic limestones. Limestones in the western part are described as dolomite, replaced by calcite [12] . The Menderes Massif rocks and limestones are overlain unconformably by the Neogene sediments. The infill of the Köprübaşı-Demirci basin is dominated by continental clastic sediments, lava flows, and volcaniclastic deposits. The Neogene sediments in the study area can be categorized into three different units: The base of the sedimentary sequence is formed by conglomerates with intercalated sand bodies and volcanics (N1). This formation is also exposed at the southern end of the study area, overlying the Kale crest dolomitic limestones and the Menderes Massif rocks. During active volcanism the basin formed and subsided, creating space for fluvial, lacustrine (N2), and lacustrine limestone deposits (N3) (Figure 2). The sedimentary succession (N1 to N2) is intercalated by alternating rhyodacites, trachyandesites, and agglomerates.

Polished thin sections were studied by transmitted and reflected optical microscopy. The radiometric measurements for U in the study field were obtained by γ-ray spectrometer (Gamma-Scout, Online Model, radiation detector, US standard). Whole rock samples were analyzed by inductively coupled plasma optic emission spectrometry (ICP-OES) for major, trace and rare earth elements (REEs) and major oxides by X-ray fluorescence (XRF) for 16 trace elements, and for total sulfur, organic carbon, and total carbon dioxide at the ACME analytical laboratories (Canada). The sixteen rock samples were rushed, split, and pulverized (250 gr to 200 mesh), then digested with LF following a multi-acid digestion (1:1:1 (HCl-HNO₃-HClO₄)) and analyzed by ICP-OES. CO₂ total, Corganic and Stotal were measured after leaching by HNO₃ with Leco 744. SO₄ contents of the studied samples were measured by calcination at 550˚C (Leco 744). Corg for the studied samples was measured by Leco 832. All samples were carried out from dolomitic limestomes in M1, conglomerate, sandstone and dacitic tuff in N1 and sandstone, limestone and marl in N2 zones where are shown on Figure 2.

Thinly scattered phosphorite (collophanite) grains and nodules are relatively common in many carbonate units deposited on stable shelves or in warm epicontinental seas [14] . Cathcart and Gulbrandsen listed numerous phosphate-containing limestones, marls, and carbonatic shales in North America [15] . These rocks, having the usual high trace U content (of the order of 60 - 100 ppm in the phosphate), are normally considered uneconomic unless they grade into bedded diagenetic phosphorites (as in Morocco) or unless scattered nodules have been residually enriched (as in Florida). In phosphorite deposits, values of 0.01 to 0.03 wt% U₃O₈ are generally regarded as potential sources of uranium [12] . The measured radioactivity values are range from 900 to 7600 cps, 400 to 1000 cps and 200 to 400 in M1, N1 and N2 units, respectively (Figure 3). If the Kale crest dolomitic limestones of the Demirci basin exceed this range, then these dolomitic limestones can be regarded as potential sources of uranium (Figure 4). The uranium mineralizations are in the form of thinly scattered phosphorite (collophanite) concentrated in breccia zones and fractures in the dolomitic limestone (Figure 4). The formation origin proposed for these mineralized zones is leaching of uranium and phosphate from the overlying low-grade Neogene cover (dacitic tuff-marl claystone), concentrated in the Kale crest dolomitic limestone. Mineralization has occurred in breccia zones and fractures in the Kale crest dolomitic limestone. The solutions were probably leached from overlying dacitic tuff-marl-claystone strata that have generally now been removed by erosion. The fractures in the dolomitic limestones do not exceed 50 cm in width and are generally discontinuous in length, rarely exceeding 15 m. The highest-grade ore is in a very fractured and brecciated zone, associated with the Kale crest dolomitic limestones of the Demirci basin. The uranium mineralizations, from a few to 40 cm, are observed depending on supergene phosphate-filled cracks and fractures. The supergene enrichment in the dolomitic limestones is overlain by an intensely altered volcanic rocks, and leached from overlying Neogene strata (dacitic tuff-marl-claystones) that have generally now been removed by erosion. The geometry of the supergene enrichment zone is irregular and is in the form of cracks, fractures, and breccias, associated with dolomitic limestones. The mineralogical features show that the uranium mineralizations associated with flourapatite-rich rocks have been formed within calcite matrix fractures; cracks systems; and limestone, claystone, marl, and silicified breccia zones in the lacustrine sediments that are observed in the Kale crest dolomitic limestones. Brecciated fragments of the Kale crest dolomitic limestones occur in a matrix of uraniferous collophanite (thinly scattered phosphorite). Needle-like apatite crystals are observed in a matrix of uraniferous collophanite (thinly scattered phosphorite). Opaque uranium minerals are observed in the carbonate matrix and calcite-filled fractures and fracture systems of studied samples taken from different units of the Neogene sediments (marl, claystone, and volcanic levels) and from limestones outcropping around Rağıllar in western Turkey.

The analyses results for the studied samples are shown in Table 1. The main oxide content of the studied rock samples suggests the presence of three different rock groups: 1) limestone (LF1, LF6, LF7, LF8, LF10, LF 11, LF 12, LF 13, LF 14, LF 15, and LF 16), 2) marl (LF 3 and LF 4), and 3) claystone (LF 2, LF 5, and LF 9). Figure 5(a), Figure 5(b), Figure 5(c), and Figure 5(d) shows the relationships between uranium and total sulphur TOT/S total carbon TOT/C, organic carbon (ORGC), and SO₄, respectively. According to Figure 5(a) and Figure 5(d) showing the relationship between uranium and total S, uranium and SO₄ were more closely correlated than uranium and C_tot, or uranium and C_org (Figure 5(b) and Figure 5(c)). In addition, the distribution relationship between thorium and sulfate shows a poorer correlation than the distribution relationship with uranium. This indicates that sulfate concentrations in solution do not affect the thorium concentration. Figure 5(e) and Figure 5(f) shows the associations between uranium, and P₂O₅ and SiO₂, respectively. Figure 5(g) shows the associations between Al₂O₃ and Fe₂O₃. Excellent distribution relationships are observed for P₂O₅ and uranium, regardless of the lithological features. However, except for some sample points (LF 2, LF 4, LF 5, LF 12, and LF 13), the Fe₂O₃ has watched that it offers same distribution relation with uranium. When the distribution relationships between uranium and SiO₂ and Al₂O₃ are examined, no distribution relationship was observed between the main oxides and uranium.

The major oxides and trace element concentrations were obtained from the study area. These values were examined by comparison with average values for claystone and carbonate rocks from Clarke and Washington [17] ; Vogt [18] . When these data are taken into account, the average P₂O₅ content of the samples is estimated (17.16%) is higher than 101 and 429 times the average claystone and carbonate rocks content, respectively. That is why uranium may be adsorbed in dolomitic limestones that the adsorbtion of uranium can be increased the raltive amounds of precipitated due to metal-ligand complex. Al-Eshaikh et al. (2013) determined the concentration of the major oxides, such as MgO, Al₂O₃, and Fe₂O₃, present in the phosphate ore of the Jalamed regions [6] . Nevertheless, the concentration of phosphorus-pentoxide was determined to be greater (30%) than the content of the other three main oxides. The concentration of calcium oxide varies from about 49% - 55% depending on the location. Consequently, Al-Eshaikh et al. observed that the uranium concentration seems to follow the concentration of phosphorus-pentoxide; where the P₂O₅ concentration increases, the uranium concentration also increases [6] . Many studies have determined that hydraulic and geochemical processes that cause contamination (U⁶⁺) in the groundwater-river mixing zone [19] [20] . These studies have determined that the distribution and concentration of calcite can increase the pH of groundwater, and that uranium-bearing solutions prefer solid phases to aqueous systems. This condition is more effective in metal precipitation that may occur from infiltration and reactions in the vadose zone than in metal precipitation that occurs from seasonal water table fluctuations. Philips and Watson showed that dolomitic gravels can increase groundwater pH from 3.2 to 6.5 [10] . Therefore, precipitated metals in the dolomite gravel have been transported from sandy and shaly sediments.

Positive correlations between major oxides, trace elements, and rare earth elements are shown in Table 2 and negative correlations between major oxides, trace elements and rare earth elements are shown in Table 3. Positive correlations between major oxides and rare earth elements are shown in Table 4 and negative correlations between trace elements and rare earth elements are shown in Table 5. When these data are taken into account, a positive correlation (r ≥ 0.90) was determined between SiO₂, and Al₂O₃, K₂O, and Ta. A moderate positive correlation was determined between P₂O₅ and CaO. The presence of strong positive correlations (r ≥ 0.90) was determined between TiO₂ and Ba, Co, Nb, Cu, La, Ce, Pr, and Nd.

A strong negative correlation (r ≥ −0.90) was determined between CaO and SiO₂, Al₂O_3, and K₂O. A moderate negative correlation (r > −0.70 to −0.90) was determined between CaO and La, Ce, Pr, and Nd. A strong positive correlation (r ≥ 0.90) was determined between the following trace elements: Ba-Co, Ba-Cs, Ba-Cu, Ba-Ni, Be-U, Be-Zr, Co-Nb, Co-Cu, Co-Ni, Ga-Rb, Nb-Ta, Nb-Cu, Th-U, Th-Y, Th-Pb, Th-Sb, U-Zr, U-Y, U-Pb, U-As, U-Sb, Y-Pb, Mo-As, Pb-Sb, Pb-As, Zn-Hg, and As-Sb. A strong positive correlation (r ≥ 0.90) was determined between the following trace elements and rare earth elements: Ba-Ce, Co-La, Co-Ce, Co-Pr, Co-Nd, Nb-La, Nb-Ce, Nb-Pr, Nb-Nd, Th-Dy, Th-Ho, Th-Er, Th-Tm, Th-Yb, Th-Lu, U-Ho, U-Er, U-Tm, U-Yb, U-Lu, Y-Tb, Y-Dy, Y-Ho, Y-Er, Y-Tm, Y-Yb, Y-Lu, Cu-La, Cu-Ce, Cu-Pr, Cu-Nd, Pb-Dy, Pb-Ho, Pb-Er, Pb-Tm, Pb-Yb, Pb-Lu, Sb-Yb, and Sb-Lu (Table 5). A moderate negative correlation (from r = −0.70 to −0.90) was determined between the following trace elements: Be-Ta and Nb-U. Thorium generally shows an inverse relationship with U, indicating that it precipitates from the surface down (above the water table) to the deeper zones (below the water table) as indicated by Philiphs and Watson [10] . In this study, it was determined that uranium has been much more deposed in dolomitic limestones due to increase pH in alkalin environment. Also, the other important factor on uranium deposit is uranium sulphate compounds in the zone of oxidation due to sulfur-rich fluids in the zone of reduction (e.g. pyrite) in fine grained sedimentary rocks such as clay. Figure 4 shows that uranium may be transported by acidic groundwater; however, in carbonate environments and in the groundwater-river mixing zone it is likely to be enriched through occurrence in CO₃, PO₄, and SO₄ compounds (Figure 5(c); Figure 5(d); Figure 5(e)).

Tang et al. indicated that uranium carbonate minerals such as liebigite (Ca₂UO₂(CO₃)₃∙10H₂O) might occur in

environments with pH > 5 [21] . Wang et al. examined the distribution of uranium in river sediments around the Tianyuan phosphate deposit in China, and stated that the distribution of uranium is controlled by the adsorption of organic compounds [7] . In this study, it has been determined that the relationship between organic carbon and uranium is not as well correlated as the relationship between U and P₂O₅. This is supported by the idea that the enrichment of uranium forms inorganic carbon compounds and inorganic phosphate complex. Therefore, uranium minerals are thought to be formed as UO₂(HPO₄)₂ compounds with phosphorus complexes derived from flourapatite, or may be formed as oxide compounds. The solubility coefficients of uranium compounds have been calculated. and using mineral transport equilibrium (MINTEQ Dat) [22] . These have been calculated using the following thermodynamic equations: logK = −5.25 (1), 9.94 (2), and 44.7 (3). Wang et al. [7] , identified that the dominant species of uranium minerals in river sediments within the pH range of 7.2 - 10.4 are as follows:

1)

2)

3)

U-Zr, U-Y, U-Pb, U-As, and U-Sb correlations exhibit high positive correlations (r > 0.90), indicating that uranium is enriched alongside these elements. It is suggested that source of uranium may be weathered metagranites from the Menderes Massif and moved by Pb, As and Sb rich hydrothermal fluids. The chondrite normalized and Ce/chondrite-Yb/chondrite distribution graphics indicate the presence of negative Eu anomalies (Fig- ure 6), as shown in the study by Graff [23] , indicative of the oxidation zone. However, Figure 7 shows an increase in light rare earth elements (LREEs) concentrations in samples LF 2, LF 4, and LF 9. These samples were taken from within the dacitic rocks, which are close to the fault zone. The presence of strong positive correlations between uranium and heavy earth elements (HREEs) are due to the stronger metallic properties of HREEs compared with LREEs. Moreover, the same relationship is also seen for Pb and Sb, suggesting the same origin for these metals. The presence of positive correlations between Cu, Ce, and La may suggest that copper is transported into shallow environments by moving groundwater. The negative Ce anomaly in the hydrothermal solution reveals the contributions of hydrothermal fluids with volcanogenic origin [23] ; a positive Ce anomaly is

observed in the moving groundwaters and river waters [24] . In additon, the uranium minerals were observed in the apatite-rich carbonate matrix and in calcite-filled fractures and fracture systems of the studied samples taken from different units of the Neogene sediments (marl, claystone, and volcanic levels) and from dolomitic limestone outcropping around Rağıllar village.

The distribution of uranium in rock samples and the factors that affect this distribution have been examined in samples taken from different units of the Neogene sediments (marl, claystone, and volcanic levels) and limestones around Rağılllar village, which show high levels of radioactivity. In this study, the distribution of uranium was determined to be dependent on the physicochemical properties of uranium-rich solutions. The presence of a strong positive correlation between uranium and P₂O₅ indicates that uranium can be deposited into phosphorus compounds. The negative distribution relationship between C_org and U shows that primary mineralization is formed by dissolving minerals in the form of inorganic compounds. The mineralization consists of uranium minerals filled by supergene redistribution in cavities (such as cracks, fractures, breccias) associated with dolomitic limestones. In addition, negative Eu anomalies indicate the oxidation zone, which is considered an appropriate zone for the enrichment of uranium in phosphate species. Appropriate zones for uranium enrichment were formed as a result of the movement of groundwater from acidic environments to alkaline environments along permeable, crack and fracture systems within the volcano-sedimentary deposits in the Rağıllar vicinity.

This work was supported by the Firat University Scientific Research Projects Unit (FUBAP).

Leyla Kalender,Ömer Nedim Alçiçek,Ezgi Gürgenç, (2016) Geological and Geochemical Features of Uranium Mineralization in Western Turkey. International Journal of Geosciences,07,409-424. doi: 10.4236/ijg.2016.73032