3.1 Materials

All chemicals used throughout this work were of analytical grade and used as supplied. Deionized water was used to prepare all the solutions.

A total of 15 composite surface rock phosphate samples were collected from Uro phosphate deposits. Chisel and hammer were used to break off the phosphate chips, which were then collected in cloth bags. Samples were crushed in a jaw crusher and then ground to very fine powder.

3.2 Methods

3.3 Instrumentation

3.4 Statistical analysis

Multivariate correlation coefficient, cluster analysis of uranium, and major and trace elements were carried out in order to observe the elemental associations in the uranium concentrates; yellow cake and uranium trioxide. To achieve this purpose, the statistical analysis “Statistical Program for the Social Science (SPSS/PC)” was used (SPSS Inc., 1998). The applied statistical analyses are:

1. Single variate statistics: Single variate statistics was used to describe the statistical characteristics of the data. These include: range, arithmetical mean, measures of dispersion of the data, and standard deviation (σ).

2. Bivariate statistics: Bivariate correlation analysis is also carried out, as bivariation statistics in order to examine the mutual relations and strength of association between pairs of elements through calculation of the linear Pearson correlation coefficient “r”.

3. Hierarchical cluster analysis: Cluster analysis comprises a series of multivariate methods, which are used to find true groups of data. In clustering, the objects are grouped such that similar objects fall into the same class. Hierarchical clustering joins the most similar observations, and then successively the next most similar observations. The levels of similarity at which observations are merged are used to construct a dendrogram. Some measure of similarity must be computed between every pair of objects. In this work, a cluster analysis using Ward’s method to calculate the Euclidean distance between the variables average is used.

4.1 Phosphate ore analysis

Table 2 shows the uranium content and the percentage of the major elements as oxides in phosphate ore samples. Fig. 2 shows the mean values of the percentage of the major elements as oxides in phosphate ore samples. Table 2 shows that uranium content is very high in all of the collected samples relative to the average uranium content in the earth crust (2.7 μg/g) and shale rocks (3 μg/g) (Turekian and Wedepohl, 1961). Moreover, the phosphate samples show very high content of uranium relative to the world average (120 μg/g) (Adams et al, 1959). The maximum uranium in Uro phosphate is 1055 μg/g, with an average of 310.3 μg/g and a standard deviation of 281.8. These values indicate that there is a significant uranium abundance in Uro phosphate, which is 2.6 times higher than the average of world phosphate. The estimated total phosphorus as P₂O_5% is in the range of 16.4–25% with an average of 20.3%, which generally indicates that the Uro phosphate ores are typical phosphate ores. Table 2 also shows that phosphorous, silicon and aluminum are the most abundant elements; this is clearly represented in Fig. 2.

Close

Figure 2

The mean values of the major elements in phosphate ore samples.

Export to PPT

Table 3 shows that titanium, strontium and barium are the most abundant trace elements in the samples. Moreover, Fig. 3 shows that all samples contain a significant concentration of titanium with a mean value of 7300 μg/g. The mean values of concentration of the trace elements show that they exist in the decreasing order as: Ti > Sr > Ba > Zn > Mn > U > Cu > Mo > Ni > Co > Pb > Zr > Cr.

Close

Figure 3

The mean values of the trace elements in phosphate ore samples.

Export to PPT

The Pearson correlation coefficients for uranium and eight major element oxides and trace elements are shown in Tables 4 and 5, respectively. Table 4 shows that no correlation exists between uranium and any major element oxide, while Table 5 shows that uranium has strong positive correlation with Sr (+0.57), and a moderate positive correlation with Zr (+0.37). This indicates the close relationship between uranium and these two elements in Uro phosphate minerals. The variation of uranium content appears to be generally controlled by the variations of Sr and is not related to the abundance of any major element.

Table 6 shows uranium content and P₂O₅% for phosphoric acid solutions prepared from the corresponding phosphate ore samples. The obtained green phosphoric acid was found to contain uranium in the range of 185.8–2048.8 μg/l. The specifications of phosphoric acid are given in Table 7. This table indicates that during the manufacturing of phosphoric acid by digestion with sulphuric acid at 75–85 °C, around 98% of uranium content in the phosphate ore samples recovers in phosphoric acid.

4.2 Uranium recovery from phosphoric acid

The recovery of uranium(VI) from the prepared green phosphoric acid solution is based on liquid–liquid solvent extraction, which involves two processes: extraction and stripping process. The extraction process was performed by the extraction of uranium(VI) with 25% TBP solution in kerosene in three stages using an A/O of 2:1. Uranium is extracted with this organic solvent as uranyl nitrate in nitric medium according to the following equation: $${{\text{UO}}_2^{2+}}_{(\text{aq}.)}+2{{\text{NO}}_3^{-}}_{(\text{aq}.)}+2{\text{TBP}}_{(\text{org}.)}\rightleftarrows {\text{UO}}_2({\text{NO}}_3)·{(\text{TBP})}_{2(\text{org}.)}$$ The organic solvent TBP is a neutral organophosphorus extractant exhibiting satisfactory extractive properties towards uranium in nitrate solution. It was diluted with kerosene as an inert diluent to induce quick and clear phase separation and high extraction performance. This extraction using TBP requires phosphoric acid with special specifications and high quality. For example, the solid content in acid should be reduced from 1–4 wt.% to less than 0.01 wt.%. Also most of the organic and humic matter should be removed and uranium should be oxidized to the hexavalent state (Khaled and Ali, 2000). If humic matter is still present in the phosphoric acid, it will contact the kerosene-based organic solvent, and the resulting mixture is allowed to settle; an oily, viscous emulsion will form at the interface between the organic and aqueous layers. If this interfacial emulsion (also, named “gunk”) was allowed to collect at the interface of the two layers; it would soon interfere with the normal operation; uranium extraction. TBP is not selective for uranium; so many elements are co-extracted with uranium(VI) by TBP, especially the ions of Sc(III), V(V), Fe(III), Co(II), Ni(II), and Zr(IV). Therefore, these ions will represent serious interference in the uranium(VI) extraction with TBP. Stripping the loaded uranium organic phase with 0.5 M sodium carbonate will remove the interference of Sc(III), V(V), Co(II) and Ni(II), but not that of Fe(III) and Zr(IV). To eliminate the interferences due to Zr(IV), sodium fluoride solution was added as a masking agent for Zr(IV). This agent masks zirconium and prevents its extraction from aqueous phase by TBP in kerosene. The interference due to Fe(III) was removed by extraction as iron(III) oxinate complex from the acidic aqueous solution using 8-hydroxyquinoline in chloroform (Korkish and Hubuer, 1976). After the uranium extraction process, the P₂O₅ entrained by the pregnant organic solvent from the extraction was completely removed by scrubbing the loaded extract with water in one stage. P₂O₅ must be eliminated because: it is a contaminate of the yellow cake concentrate, and striping process is carried out with a carbonate solution, which would be neutralized by P₂O₅ (IAEA, 1990).

The stripping process is the opposite of extraction in the sense that uranium is now brought back to the aqueous phase using suitable chemical reagents. Several reagents can be used to strip uranium from the loaded solvent. A trial was made to determine the stripping coefficient (S) from the loaded organic solvent (TBP) using distilled water, acidic and alkaline solutions. In this test, the A/O phase ratio and the mixing time were fixed at unity and 5 min, respectively, in a single stage. Table 8 shows the stripping coefficient for different strip solutions. From this table it is clearly seen that the complete stripping recovery of uranium content (more than 99.8%) was obtained with 0.5 M sodium carbonate solution in which uranium reports as uranyl tricarbonate complex ion [UO₂(CO₃)₃]⁴⁻. Therefore, sodium carbonate solution is a very suitable strip solution for the quantitative back-extraction (stripping) of uranium from an organic phase, and the uranyl tricarbonate complex ion is soluble and extremely stable (Harvinderpal et al., 2004). In this work, uranium was stripped from the organic scrubbed phase in two stages using an aqueous alkaline solution of 0.5 M Na₂CO₃ at an A/O phase ratio of 1:4 as sodium uranyl tricarbonate according to the following equation: $${\text{UO}}_2{({\text{NO}}_3)}_2·{(\text{TBP})}_{2(\text{org}.)}+3{\text{Na}}_2{\text{CO}}_{3(\text{aq}.)}\to {\text{Na}}_4{[{\text{UO}}_2{({\text{CO}}_3)}_3]}_{(\text{aq}.)}+2{\text{TBP}}_{(\text{org}.)}+2{\text{NaNO}}_{3(\text{aq}.)}$$ Table 9 shows uranium content in the strip solution samples recovered from phosphoric acid, with mean recovery of 97.8%.

4.3 Yellow cake precipitation

Uranium in the loaded sodium carbonate stripping solution was precipitated as sodium diuranate (Na₂U₂O₇) by applying sodic decomposition using 50% sodium hydroxide solution at pH of 12.0, according to the following equation: $$2{\text{Na}}_4[{\text{UO}}_2{({\text{CO}}_3)}_3]+6\text{NaOH}\to {\text{Na}}_2{\text{U}}_2{\text{O}}_7+6{\text{Na}}_2{\text{CO}}_3+3{\text{H}}_2\text{O}$$ The slurry was then subjected to filtration where the filtered product; sodium diuranate (Na₂U₂O₇), represents an intermediate product called “yellow cake” (Gupta et al., 2004). During the precipitation of sodium diuranate concentrate, sodium carbonate was produced and it was recycled to use as strip solution. The precipitation efficiency of the yellow cake was calculated by the following equation: $$E\%=[(F-R)/F]\times 100$$ where F is uranium concentration in the strip (feed) solution. R is uranium concentration in the raffinate solution.

The result is shown in Table 9, which indicates that by applying sodic decomposition up on stripping carbonate solution, generally, about 98.0% of uranium in the strip solution was precipitated as sodium diuranate concentrate (Na₂U₂O₇).

In order to analyse crude yellow cake, firstly, sodium diuranate samples were dissolved in nitric acid according to the following equation: $${\text{Na}}_2{\text{U}}_2{\text{O}}_7+6{\text{HNO}}_3\to 2{\text{UO}}_2{({\text{NO}}_3)}_2+2{\text{NaNO}}_3+3{\text{H}}_2\text{O}$$

Secondly, uranium was quantitatively separated from the yellow cake nitrate solution using 25% TBP extractant in kerosene, then uranium was stripped from the loaded organic phase and determined. Finally, trace elements of P, K, Ca, Mg, Mn, Cr, Fe, Co, Ni, Cu, Zn, Cd, Mo, and Pb had been properly determined in raffinate aqueous phase solution. Table 10 indicates that mean uranium concentration in the sodium diuranate concentrate is 70.35%. This concentration of uranium and that of the comparative concentrations of impurities on the obtained yellow cake agree well with the standard commercial sodium diuranate illustrated in Table 14.

The major metallic impurities in the obtained yellow cake concentrate are phosphorus, calcium, potassium and magnesium (Fig. 4). The decreasing order of mean concentrations of metallic impurities is: P > Ca > K > Mg > Ni > Zn > Cu > Pb > Fe > Mn > Co > Cd > Mo > Cr. Uranium has a strong positive correlation coefficient with Ca (+0.74). High positive correlations have also been found between other metallic impurities. Zinc correlates positively with U (+0.52), P (+0.56) and Ca (+0.52). Manganese correlates positively with Co (+0.57), Ni (+0.53) and Cu (+0.5) (Table 11).

Close

Figure 4

Metallic impurities concentrations in yellow cake samples.

Export to PPT

The correlation found between metallic impurities and uranium in yellow cake samples can be observed qualitatively in Fig. 5. Cluster analysis to show elemental association in the yellow cake produced a dendrogram that formed two-groups : the first group includes: Na, Co, Cd, Cr, Mo, Mn, Fe, Cu, Pb, Ni, Zn, K, and Mg, and the second group is formed by the elements U, P, and Ca. The first group is divided into two sub-groups: (a) Na, Co, Cd, Cr, Mo, Mn, Fe, Cu, Pb, Ni, and Zn; (b) K and Mg. The sub-group (a) consisting of 10 metallic impurities has elements of the transition elements (except Na) that have chemical similarity. The sub-group (b) consisting of K and Mg, which are coherent because they are reactive metals from group I and II in the periodic table that have chemical similarities. The second group is divided into two sub-groups: (a) U and P; (b) Ca. The second group shows the close relationship between U and P on the one hand and their association with Ca on the other.

Close

Figure 5

Dendrogram showing metallic impurities and uranium association in yellow cake samples.

Export to PPT

4.4 Refining of uranium yellow cake concentrate

Chemical analysis of the obtained yellow cake intermediate product; sodium diuranate, indicates the presence of many metallic impurities. Most of these are substantially harmful and may cause serious problems during fabrication and performance of the nuclear fuel (IAEA, 1990). Therefore, uranium must be refined to meet rigid international standard specification of nuclear purity. In this approach, solvent extraction technique is almost exclusively adopted for uranium refining and purification, and TBP is the most common commercial process applied for uranium extraction from nitrate solution (IAEA, 1993). In the present work, TBP was used to refine crude yellow cake in order to obtain nuclear pure uranium trioxide. Performing of the refining of uranium yellow cake concentrate was carried out through five steps. Firstly, the intermediate product was dissolved in concentrated nitric acid to obtain the liquid uranyl nitrate, which is used as feed liquor for extraction process. Secondly, uranium was quantitatively extracted from aqueous nitrate feed solution using 25% TBP in three batches using the extraction phase ratio A/O of 2:1. Thirdly, uranium loaded organic phase was scrubbed using 1 M nitric acid solution at A/O phase ratio of 1:5, through single contact. Scrubbing is extremely needed for further removal of the co-extracted impurities. Then, uranium was stripped from the loaded scrubbed organic phase by using 0.5 M Na₂CO₃ in two batches applying stripping phase ratio A/O of 1:4. Fourthly, uranium was precipitated from the acidified carbonate solution using 30% hydrogen peroxide. Hydrogen peroxide has been reported as a very selective precipitant for uranium from acidic solutions (IAEA, 1993). It is reasonable to precipitate uranium peroxide (UO₄·2H₂O) from the stripped sodium uranyl tricarbonate liquor at pH 2. Advantage of such precipitation came from the great minimization of impurities in the obtained uranium precipitate. Finally, the orange powder of nuclear grade uranium trioxide (UO₃) was obtained by calcination of uranium peroxide in controlled furnace, in two thermal stages: $${\text{UO}}_4·2{\text{H}}_2\text{O}\stackrel{\text{de-hydration,}300°\text{C}}{⟶}{\text{UO}}_4\stackrel{\text{de-oxygenation,}450°\text{C}}{⟶}{\text{UO}}_3$$ The chemical analysis of the metallic impurities in the obtained uranium trioxide (UO₃) product summarized in Table 12 is in great agreement with that of nuclear purity specification (Table 14).

The major metallic impurities in the obtained uranium trioxide are Na, Fe, Ni and K (Fig. 6). The metallic impurities in the decreasing order of mean abundance are: Na > Fe > Ni > K > Ca > Pb > P > Mn > Mg > Zn > Mo > Cd > Co > Cr. Table 13 shows that uranium correlates positively with Cd (+0.36), and iron correlates with P (+0.54) and Cr (+0.52). In addition, chromium has a strong positive correlation with Ca (+0.70). Cluster analysis of metallic impurities and uranium in uranium trioxide samples produced two-group dendrogram of element association (Fig. 7). The first group includes all of the estimated metallic impurities, while the second group includes only uranium. This indicates that uranium in uranium trioxide concentrate has no association with any metallic impurities.

Close

Figure 6

Metallic impurities concentrations in uranium trioxide samples.

Export to PPT

Close

Figure 7

Dendrogram showing metallic impurities and uranium association in uranium trioxide samples.

Export to PPT