Introduction

Uranium is ubiquitously present in the environment and forms an important part of the nuclear fuel cycle. The uranium content in the various matrices of the environment has to be determined for many purposes right from prospecting to waste management and environmental protection. Uranium estimation by the common techniques like alpha- spectrometry, laser fluorometry, anodic stripping voltammetry, involve interferences due to dissolved solids and iron, manganese, calcium and other elements [1–4]. Iron is a major component of most of the environmental matrices and accompanies uranium in all the environmental compartments. In most of the radiochemical separation procedures iron follows the same chemistry as uranium. Determination of uranium at environmental levels often involves pre-concentration steps, due to low natural concentrations present. Most of the pre-concentration steps use ferric hydroxide for collecting uranium and other actinides. But in the further stages of the radiochemical estimation of uranium by alpha spectrometry Fe interferes by reducing the efficiency of electro-deposition of U, degrading the quality of the deposit on the stainless steel planchet and making it less adherent. This leads to errors in estimation of uranium by alpha spectrometry as the spectrum generated is also degraded and sometimes difficult to differentiate the peaks from each other. Application of neutron activation analysis for estimation of uranium is also faced with problems of Compton contribution of 59Fe gammas to 239Np peak. Large quantities of iron and manganese are reported to reduce efficiency of laser flurometric determination of uranium due to quenching effects [1–3]. Separation of iron from uranium hence becomes essential for interference free quantitative estimation. A number of methods are reported in the literature for separation of iron from uranium based on solvent extraction, ion exchange, co-precipitation etc. Uranium is reported [5] to be isolated from iron by precipitating iron as Fe2O3·nH2O with ammonium carbonate or bicarbonate and uranium remains in solution as (NH4)4UO2(CO3)3. But some uranium can remain occluded on the iron precipitate, leading to variable efficiency and a recovery of about 70–75% of the uranium in the solution in a single precipitation. The supernatant has to be subjected to repeated precipitations and proper washings of the precipitate have to be carried out to get to get quantitative recovery of uranium. This reduces the reproducibility of the experiment and recovery factor has to be accounted, in every experiment. Methods based on ion exchange resins using sulphuric acid and ammonium sulphate, controlled redox potential methodology using hydrogen peroxide for separation of traces of uranium from iron have been reported [6–8] As per the reported [6] method uranium and iron mixture was loaded on the anion exchanger in 0.2N H2SO4 and 0.6N (NH4)2SO4. Uranium was then attempted to be eluted with 2N H2SO4, as reported in [8] but no separation of iron and uranium could be obtained in the first step. Studies on K d of uranium on Dowex-1 in H2SO4 medium [6] have shown a high K d up to 590 in 0.025N H2SO4 which reduces with increasing concentration of H2SO4. Based on the information gained from these experiments and references a number of experiments were performed with different concentrations of H2SO4 (2, 0.2, 0.025N) for separation of iron and uranium. A method using 0.025N H2SO4 was standardized for removal of iron from uranium.

Experimental

All the reagents hydrochloric acid (8N), nitric acid (0.1N), sulphuric acid (0.025N)used during the entire experimental work were of analar grade. Uranium standard was prepared from high purity uranium metal and standard stock solution of iron from ferric chloride. Suitable aliquots were used to carry out the preliminary experiments. Anion exchange resin Dowex-1X8, having mesh size 100–200 μm packed in glass column of inner diameter 1 cm and a average height of 15 cm, allowing a free flow rate of about 1–2 mL/min was used for the separation. A MCA coupled to a Si surface barrier detector with a resolution of 60 keV and 17% efficiency was used for alpha spectrometric determination of uranium. U source for alpha spectrometry was prepared by electroplating, on stainless steel planchet, in aqueous medium using (NH4)2SO4 at pH 2.3 and at an applied voltage of 6 V and 0.3 A current for 3 h. Estimation of Fe was carried out by using GBC atomic absorption spectrophotometer.

Materials and methods

The radiochemical procedure adopted in the present study was standardized as follows. Five liter of distilled water was acidified with about 25 mL of conc. HNO3 to adjust the pH to 1, spiked with known amounts of uranium in form of uranyl nitrate solution (5–200 μg/mL U). Fe carrier was added in varying amounts from 20 to 50 mg and the solution was heated with stirring. Aqueous concentrated ammonia was added to the boiling solution until the pH reached around 9, leading to the formation of a hydroxide precipitate and continued heating until the precipitate had coagulated. This precipitate was then allowed to settle overnight. The supernatant was removed by centrifugation at 2,500 rpm for 15 min. The process was repeated on the supernatant to ensure complete precipitation of the radionuclide. The precipitate was dissolved in 5 mL of 8N HCl and loaded on an anion exchange resin Dowex-1X8 for separation of U and Fe. 15 mL (5 mL, 3 times) of 8N HCl was passed through the column. In the second step 15 mL (5 mL, 3 times) of 0.025N H2SO4 was added to the column. Finally 0.1N HNO3 solution was used to elute out U from the column which was estimated by alpha spectrometry.

Further experiments were performed to standardize the volume of eluant (0.025N H2SO4) required to completely remove iron without affecting the recovery of uranium. Different quantities of iron i.e. 20, 50 and 100 mg were spiked along with 200 mg uranium in distilled water and the same procedure as above was followed for separation. A number of fractions of eluant 0.025NH2SO4 (2 mL) were collected and analyzed for iron and uranium.

The above standardized procedure was applied to tap water by similarly spiking uranium activity. Ground water samples were also analyzed for their uranium content using the same procedure using U-233 as tracer for monitoring the efficiency of the chemical procedure and compared with laser fluorometric analysis.

Results and discussion

This method provides a simple two step method with good efficiency for removing iron from uranium. The ferric hydroxide precipitated at pH 9–10 with ammonia carries uranium, calcium, magnesium and thorium. This precipitate is dissolved in 8N HCl and passed through an anion exchange resin. Both iron and uranium get adsorbed on the resin in 8N HCl, but most of the divalents like Ca, Mg etc. and Th are not exchanged on the resin and get separated from uranium. In the second step 15 mL of 0.025N H2SO4 selectively and completely elutes Fe adsorbed on the column. Figure 1 shows the result of the experiments performed to standardize the amount of eluant (0.025N H2SO4) required for separating all the iron (50 mg) without affecting the recovery of uranium (200 mg). Concentration of iron and uranium in eluant fractions collected after every 2 mL elution is plotted. The figure shows that most of the iron is eluted with 15 mL of the eluant. Uranium is not present in the iron fraction up to the first 20 mL of elution. After that some uranium gets washed with 0.025N H2SO4 because of mass effect. Hence iron elution should be carried out with 15 mL of the eluant. Samples spiked with 20 and 100 mg of Fe showed similar elution pattern and even 100 mg of Fe is completely removed with 15 mL of 0.025N H2SO4. But higher amount of Fe reduced the exchange of uranium on the anion resin and more uranium was eluted along with iron. In this case the separation can be possibly improved with more bed volume of the resin. 50 mg of iron was considered as efficient for collection of uranium from the water sample and it was also confirmed from other reports [9].

Fig. 1
figure 1

Ratio of concentration of Fe/U eluted in different fractions of 0.025N H2SO4

Full size image

Table 1 gives the recovery of uranium spiked (5–200 μg/mL U) along with 50 mg of Fe from 5 L distilled water using the above method. The recovery varied from 73 to 95%. Similar experiment in tap water gave a recovery of uranium in the range 62.5–75%. These experiments show that almost 95% of the spiked uranium can be recovered and separation from iron is also obtained. The tap water samples similarly spiked gave lower recovery than distilled water spiked with uranium. Silica present in the tap water and ground water sample interfered and caused slowing down of the elution from ion exchange column and also affected the separation efficiency. The variable and lower recovery was also attributed to silica interference. Hence the procedure was modified for silica destruction before loading on to the anion column. The ferric hydroxide precipitate was transferred to platinum dish and treated twice with 2–3 mL HF and HNO3 mixture, heated strongly to remove any silica if present. The residue after this treatment was taken up in 8N HCl for loading in the anion column.

Table 1 Comparative recovery of spiked uranium from distilled water and tap water and effect of silica removal
Full size table

The standardized method including the silica destruction was then applied to increase the recovery of uranium to 95–97%. The method was also checked for reproducibility by carrying out the analysis in triplicate for a few spiked solutions and the average values of recovery by this method are given in Table 1. The same samples when subjected to silica destruction gave improved recoveries and good efficiency was obtained for uranium content in the samples.

Table 2 gives the comparison of analysis of uranium contents in ground water samples using the method of iron removal standardized in this study followed by alpha spectrometry. The levels were in good agreement with results of estimation by laser fluorometry. Some ground water samples with TDS content higher than 2,500 mg/L did not show good agreement between the alpha spectrometry measurement and laser fluorometry as compared to those with low TDS < 500 mg/L. Earlier studies have also reported that chemical methods such as fluorometry or laser-excited fluorescence are ideally suited for waters with low total-dissolved solids [10]. Uranium analysis in ground water by conventional neutron activation analysis, laser induced fluorometry and cyclic neutron activation analysis were observed to be in good agreement especially when the total dissolved solids (TDS) were low [11, 12]. Studies investigating the effect of TDS on direct analysis of uranium in water by laser fluorometry are required to quantitatively establish this interference.

Table 2 Results of some ground water samples analyzed by alpha spectrometry and laser fluorometry
Full size table

Conclusions

The determination of uranium in water samples especially at low levels essentially requires pre-concentration. But this includes the problems of iron and dissolved solid interferences, which introduce errors in direct analysis methods. Hence chemical treatments to separate iron and destroy dissolved solids like silica become essential. The method standardized in this study is recommended for estimation of U in surface and ground waters around nuclear facilities. This method can be applied to other environmental matrices like soil and vegetation where the interference due to iron is much more due to its natural abundance.