January, 19581 URANIUM BY SOLVENT EXTRACTION. PART I 13 The Determination of Uranium Solvent Extraction Part 11. The Separation of Uranium-233 from Irradiated Thorium as the Oxine Complex in the Presence of Ethylenediaminetetra-acetic Acid BY R. F. CLAYTON, W. H. HARDWICK, M. MORETON-SMITH AND R. TODD (Chemical Engineering Division, A .E.R.E., Harwell, nr. Didcot, Berks.) The development of solvent-extraction methods for determining trace amounts of uranium-233 in irradiated thorium is described. Thorium and its alpha-emitting daughters are complexed with EDTA, and, when uranium- 233 is extracted as its diethyldithiocarbamate complex, only bismuth-212 accompanies it. This is immaterial for colorimetric or fluorimetric finishes, but, for determination of the uranium-233 by alpha counting, the bismuth-212 must first be allowed to decay.If, however, the uranium-233 is extracted as its 8-hydroxyquinoline complex, no alpha emitter accompanies it and concentrations of uranium-233 ranging from 100 pg per ml down to 0.01 pg per ml in 0-7 M thorium solution have been determined in this way. IN Part I of this series1 a method of determining uranium-233 in irradiated thorium was described, in which uranium-233 was separated from thorium daughters by solvent extraction with isobutyl methyl ketone (hexone). Separation from thorium itself was completed by forming the uranium - diethyldithiocarbamate complex, which was retained in the solvent while the thorium was washed into the aqueous phase. It was thought that the method might be improved by the introduction of ethylenediaminetetra-acetic acid (EDTA) , since this does not prevent the solvent extraction of uranium as its diethyldithiocarbamate complex, whereas thorium and interfering daughter activities are preferentially complexed by EDTA.Unfortunately, as Bode2 has since reported, the bismuth - diethyldithiocarbamate complex is stronger than the bismuth- EDTA complex and, under the conditions for extracting uranium, the separated uranium-233 was found to be contaminated with alpha activity, owing to the presence of bismuth-212. The problem was to find a complexing agent that would form a strong solvent-soluble complex with uranyl ions, but any complexes formed with thorium and daughter alpha emitters must be weaker than their respective EDTA complexes.Di- and tri-92-butyl phosphates were tried, but were found to be unsatisfactory. Pfibil and Mal&t3 had reported that only uranium, titanium, vanadium, molybdenum and tungsten are precipitated by 8-hydroxyquinoline (oxine) in the presence of EDTA from an acetic acid - ammonium acetate buffer. Solvent extraction of the uranium complex14 CLAYTON, HARDWICK, MORETON-SMITH AND TODD THE DETERMINATION OF [VOl. 83 appeared promising under these conditions, and experiments were carried out to establish the optimum conditions for determining small concentrations (0.01 to 1OOpg per ml) of uranium-233 in thorium nitrate solutions, after the addition of EDTA, by extraction with oxine dissolved in either hexone or chloroform. These experiments are described and the analytical methods that were evolved are given inL detail, together with results obtained on a number of solutions during routine use of the methods.EXPERIMENTAL All equilibrations were carried out in 40-nil centrifuge tubes, the two phases being mixed by stirring for 5 minutes. After centrifuging, samples of the solvent phase were evaporated on stainless-steel counting trays by using the technique described by Bode.2 Counting was then carried out for 10 minutes or for a total of 10,000 counts by using an alpha-scintillation counter. EXTRACTION WITH SODIUM DIETHYLDITHIOCARBAMATE DISSOLVED IN HEXONE IN PRESENCE The experiments carried out to test the efficiency of EDTA as a “masking” agent for thorium-chain alpha activities during the extraction of uranium as its diethyldithiocarbamate complex showed that the extraction of uranium is quantitative even in the presence of thorium and excess of EDTA.Unfortunately, the thorium daughter, bismuth-212, is also extracted and must be allowed to decay before determination of uranium-233 by counting; this time lag will depend on the relative amounts of thorium and uranium originally present and on the accuracy required. When the necessary time lag is permissible, this method is very satisfactory; it has all the advantages of the original procedure,l e.g., thin sources are obtained, but is much simpler. For fluorimetric or colorimetric finishes, z.e., for microgram or milligram amounts of natural uranium, the presence of bismuth-212 is immaterial. EXTRACTION WITH OXINE DISSOLVED IN CHLOROFORM OR HEXONE- The use of oxine was next investigated, since uranium is precipitated by oxine in the presence of EDTA, but bismuth and thorium are According to Wel~her,~ quoting results by Goto, Fleck and Ward, in the absence of EDTA, uranium is completely precipitated by oxine in the pH range 4.1 to 8.8, whereas bismuth is completely precipitated in the pH range 4.5 to 10.5.Below pH 4-5, complexing oE bismuth by oxine is incomplete and its retention in the aqueous phase as the EDTA complex is more likely. So experiments were carried out at pH 4.1 to see if thorium-chain activities were extracted by oxine solutions in the presence of EDTA, and, if not, whether uranium itself were quantitatively extracted. PRELIMINARY EXPERIMENTS- A solution containing 200 mg of thorium as nit rate and 400 mg of the dihydrated disodium salt of EDTA, a molar ratio of 1 to 1.25, was adjusted to pH 4-1 and stirred with a 2.5 per cent.w/v solution of oxine in chloroform. Portions of the solvent phase were evaporated and no alpha activity was detected. The experiment was repeated with added uranium-233 and showed that about 91 per cent. of the uranium was extracted. The results were similar when a 2.5 per cent. solution of oxine in hexone was used. With an initial aqueous phase containing 18 mg of thorium at pH 4, recoveries of uranium-233 were 100 and 102 per cent. by using oxine dissolved in hexone and in chloroform. THE EFFECT OF pH ON THE EXTRACTION OF THORIUM-CHAIN ALPHA ACTIVITY- To assess the effect of pH on the extraction of thorium-chain alpha emitters, 50 ml of a solution containing 180 mg per ml of thorium in nitric acid was treated with a solution of the tetrasodium salt of EDTA, the molar ratio of thorium to EDTA being 1 to 1.25.The thorium - EDTA solution was then adjusted to pH 10 and stirred with 20 ml of a 2.5 per cent. solution of oxine in chloroform. The solvent phase was sampled as the pH was reduced in unit stages. At pH 10 and pH 9 the solvent activities were 24 and 39 disintegrations per minute per ml, respectively, but from pH 8 down to pH 2 the solvent activity was 6 disintegrations per minute per ml, equivalent to only 3 x 10-4 pg of uranium-233 per ml. The aqueous activity was approximately 2 x 106 disintegrations per minute per ml. OF EDTA- THE EFFECT OF pH ON THE EXTRACTION OF URANILfM-233- The preliminary extraction of uranium-233 with oxine had indicated that a pH of 4.1 was A sample of dissolved irradiated thorium rod solution designated No.3977 satisfactory.January, 19581 URANIUM BY SOLVENT EXTRACTION. PART I1 16 was analysed. Portions of 0.25 ml were diluted with a solution of the tetrasodium salt of EDTA and water and the pH was adjusted to 4.1, the final volume of the solution being about 2 ml. Five millilitres of a 2-5 per cent. solution of oxine in hexone or in chloroform were used as the solvent. The results of duplicate experiments are given in Table I. This sample was also analysed by extraction with a solution of sodium diethyldithiocarbamate in hexone and by extraction with a solution of sodium diethyldithiocarbamate in hexone in the presence of EDTA, the bismuth-212 being allowed to decay.TABLE I ANALYSIS OF SAMPLE No. 3977 BY DIFFERENT METHODS The aqueous phase contained 354 mg of thorium per ml and 56 pg of uranium-233 per ml; 0.25-ml portions of the solvent phase were taken for each determination Tiraniurn-233 found, Solvent phase CLg Per ml Oxine in hexone . , .. .. .. . . 55.5 t 1.7 Oxine in chloroform .. .. .. . . 56-9 & 2.2 Sodium diethyldithiocarbamate in hexone . . 53.5 & 1.5 Sodium diethyldithiocarbamate in hexone in . . 55.0 1.5 presence of EDTA . . .. .. These results were encouraging. To obtain an indication of the precision and accuracy of the procedure, ten individual analyses were carried out on one sample made up to contain 100 pg of uranium-233 per ml and 150 mg of thorium per ml; 100-p1 portions were taken and 80mg of EDTA were added to each aqueous phase, the molar ratio of thorium to EDTA being 1 to 3-3.The mean recovery was 99.5 per cent., with a coefficient of variation of 1.6 per cent. Four portions of each solvent phase were counted. 4 5 6 7 8 9 pH of aqueous phase before extraction Fig. 1. Extraction of uranium-233 from thorium nitrate solutions by means of a 2.5 per cent. solution of oxine in hexone (0.17M) as a function of pH. The concentrations of uranium and thorium were M and 1.3 x 34, respectively However, further analyses showed much wider variations and it was thought advisable to 'investigate the effect of pH on the extraction. A solution was made up to contain about 2 pg of uranium-233 per ml, 3 mg of thorium per ml, 0.06 N nitric acid and 19 mg of EDTA per ml, the molar ratio of thorium to EDTA being 1 to 4.The experiment was carried out by first adjusting the solution to pH 4-1, and then two 6-ml portions were removed. These were extracted with 5 ml of a 2.5 per cent. solution of oxine in hexone. The pH of the solution was then raised to 5.0, when two more 5-ml portions were removed and extracted with the solution of oxine in hexone. The pH was raised in steps to 8.5 and samples were taken for extraction at each stage. The results are shown graphically in Fig. 1 and they indicate that the extraction reaches a maximum at pH 6 and is constant above this value. No extractions16 CLAYTON, HARDWICK, MORETON-SMITH AND TODD : THE DETERMINATION OF [Vol.83 were carried out above pH 8.5. It should be noiced that the addition of the solvent causes a drop in pH, and the aqueous phase, initially at pH 8.5, fell to pH 7.9 during extraction. The fall in recovery that begins as the pH falls below 6 is surprising in view of many satisfactory results obtained at pH 4.1. The recoveries from high concentrations of thorium of sub-microgram amounts of uranium-233 are variable at lower pH values, but are apparently consistent above pH 6. The low recoveries as the pH falls below 7 are discussed later in relation to the complexing of uranium by EDTA. The application of the method t o the determination of sub-microgram amounts of uranium- 233 in 0.7 M thorium solution had likewise given inconsistent results at pH 4.At pH 6.5, however, complete recovery could be effected provided that adequate volumes of solvent and concentrations of oxine were used. Results of a number of experiments are given in Table 11. By using 2 ml of a 10 per cent. solution of oxine in chloroform, recoveries from 3 ml of a solution containing 0.05 pg of uranium-233 per ml and 180 mg of thorium per ml were satisfactory . Hexone is a more convenient solvent than chloroform, since it is lighter than the aqueous phase and the extracted uranium-233 is separated in the upper layer. After centrifuging, portions of the solvent can be measured by pipette directly without fear of contamina- tion by the aqueous phase. A number of analyses were accordingly carried out with hexone as the solvent.During the adjustment of pH, hexamethylenetetramine was added to the slightly acid solution to buffer the pH at 645. This was not really necessary, however, since at these concentrations of thorium and :EDTA, the solutions themselves are quite strongly buffered. Two 0.25-ml portions of each solvent phase were taken for counting, but were only counted for 10 minutes and gave about 1300 counts each. In view of this, the results that are summarised in Table I show that the accuracy and precision obtained are satisfactory . TABLE I1 EFFECT OF VARIATIONS OF THE CONCENTRATIONS OF OXINE AND THORIUM ON THE DETERMINATION OF SUB-MICROGRAM AMOUNTS OF URANIUM-233 Volume of solution Volume Concentration containing 180 mg Thorium of solvent of oxine Extraction of thorium per ml present, Solvent used, in solvent, Recovery, No.used, ml mg ml % % 4 a 1 180 chloroform 2 2.5 68, 74 4b 1-5 270 chloroform 2 2.5 50, 36 4 C 2 360 chloroform 2 2-5 24, 24 5 2-5 104, 96 4 d 2 360 chloroform 4f 2 360 hexone 5 2-5 60, 72 4e 2 360 chloroform 2 10.0 108,108 Single analysis, volume of solvent 2 ml, oxine concentration 10 per cent.- Volume of solution containing 180 mg Thorium Extraction of thorium per ml present, Recovery, No. used, ml mg % 5 900 42 5 900 50 4 720 80 4i 3 540 98 3 540 102 4 k 41 3 540 94 4 m 3 540 78 4 n 3 540 102 4 0 3 540 100 3 4j 4 P 3 540 102 EXTRACTION OF SUB-MICROGRAM AMOUNTS OF URANIUM-233 FROM CONCENTRATED THORIUM Uranium-233 was determined by extraction with either a solution of oxine in chloroform or in hexone from an aqueous phase containing initially 180 mg of thorium per ml, 0.05 pg of uranium-233 per ml and 3 N nitric acid. An excess of a solution of the tetrasodium salt of EDTA was added and the pH of the solution was adjusted to 6.5, the final volume of the aqueous phase being about 5ml.The results are shown in Table 11. The accuracy and SOLUTIONS-January, 19581 URANIUM BY SOLVENT EXTRACTION. PART I1 17 precision of the method for sub-microgram amounts of uranium-233 by extraction with 2 ml of a 10 per cent. solution of oxine in hexone were determined by a further series of tests in which the conditions were the same as for extractions 4 k to 4 p in Table 11, except that hexone was used as the solvent. The mean counting rate of the solvent phase was 516 counts per minute per ml (10-minute counts), the coefficient of variation was 6-5 per cent.and the mean recovery was 99.2 per cent. EFFECT OF THE CONCENTRATION OF OXINE AND THE VOLUME OF THE SOLVENT PHASE ON THE As can be seen from Table 11, the recovery of small amounts of uranium-233 from thorium nitrate solutions depends upon both the volume of the solvent used and the concentrations of oxine. Tables IIIA, IIIB, IIIc and IIID give the results of analyses carried out on solutions of four irradiated thorium rods under a variety of conditions, together with the results of extractions with a solution of sodium diethyldithiocarbamate in hexone. TABLE IIIA EXTRACTION OF URANIUM-233- RESULTS OF ANALYSES OF SOLUTION OF IRRADIATED THORIUM METAL IN NITRIC ACID, SAMPLE NO. 3977 Ratio of EDTA to thorium was 1 to 5.Final volume of aqueous phase was 2 ml, Analysis NO. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Volume of sample solution taken, ml 0.1 0.1 0.1 0.1 0.1 0.1 0.5 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.25 0.25 0.25 0.25 0.25 0.1 0.1 0-1 0.1 containing 354 mg of thorium Solvent chloroform chloroform chloroform chloroform chloroform chloroform chloroform hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone Volume of solvent, ml 2 2 2 2 2 2 5 2 5 5 5 5 5 6 5 5 5 5 5 5 6 5 5 5 2 2 2 2 Concentra- tion of oxine in solvent, 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 2.5 10.0 10.0 10.0 10-0 % w/v per m i Number portions prepared and counted 3 2 3 3 4 4 2 2 3 3 3 3 3 3 3 3 3 3 2 3 2 2 3 3 5 3 3 3 of 250-p1 Spread of counts, 1 1 4 2 5 7 8 1 1 3 4 1-5 1 1 1 1 1 2 1 4 4.5 1.25 1 5 5 1 1 1 % Mean of results 9 to 24 = 56-1 pg per ml.Standard deviation = & 0.7 pg per ml = & 1.2 per cent. * Precipitate formed in aqueous phase. I7ranium- 233 found, 67.4 63.1 59.8 61.2 69.1 59.5 55.5 56.0 56.4 56-7 55.6 57-2 56.5 56.0 54.8 56.7 65-5 55-7 56.6 57.1 55.5 55.5 55.5 55.7 57.5* 59*8* 59.4* 57*1* l"g Per ml The results, including those for sample No. 5354, are plotted in Fig. 2 and show that recovery is complete when 5 ml of a 2.5 per cent. solution of oxine in hexone or chloroform are used. With 10 per cent. solutions of oxine in hexone, precipitation occurs at these low concentrations of thorium, and, with 10 per cent.solutions in chloroform, the recoveries apparently in excess of 100 per cent. are probably due to evaporation of the solvent.18 CLAYTON, HARDWICK, MORETON-SMITH AND TODD: THE DETERMINATION OF [VOl. # TABLE IIIB RESULTS OF ANALYSES OF A SOLUTION OF IRRADIATED THORIUM METAL IN NITRIC ACID, SAMPLE NO. 5352 Ratio of EDTA to thorium was 1 to 5. Fi:naI volume of aqueous phase was 2 ml, containing 304mg of thorium per ml Analysis No. 1 2 3 4 6 6 7 8 9 10 11 12 13 14 16 16 17 18 19 20 21 22 23 24 Volume of sample solution taken, ml 0.1 0- 1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0- 1 0.1 0.1 0.1 0.1 Solvent chloroform chloroform hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone hexone chloroform chloroform chloroform chloroform chloroform chloroform chloroform chloroform chloroform Volume of solvent, ml 2 2 2 2 2 2 2 2 2 6 6 6 6 6 6 6 6 5 5 6 6 6 6 6 Concentra- tion of oxine in solvent, 2-5 2.6 2.6 2.5 2-5 2.5 2.6 2.5 110-0 2.5 2-6 2-6 2.5 2.5 2.6 2.6 2.6 2.5 2-6 2.5 2.6 2.6 2.6 2.6 x, w/v Number portions prepared and counted 3 3 of 250-pl - - - - - - 3 3 3 3 3 3 3 3 3 3 3 3 2 3 3 3 Spread of counts, 6 2 % - - - - - - 1 1 3 1 1 1 1 10 4 12 3 7 1 8 8.6 6 Uranium- 233 found, 56.9 67-4 49*1* 61.5* 48-38 49.4. 63.2.62-2* 57-0t 63.2 52.9 64.2 64.2 61.6 66.4 53.1 66.5 51.2 62.6 65.9 63.1 53.3 63.2 64.6 Per ml Mean of results 10 to 24 = 63.6 pg per ml. Standard deivation = -f 1.3 pg per ml = & 2.4 per cent.* Solvent removed with washings and made up to 25 ml. t Precipitate formed in aqueous phase. TABLE IIIc RESULTS OF ANALYSES OF A SOLUTION OF IRRADIATED THORIUM METAL IN NITRIC ACID, SAMPLE NO. 5354 Ratio of EDTA to thorium was 1 to 5. Final volume of aqueous phase was 2 ml, Analysis NO. 1 2 3 4 6 6 7 8 9 10 11 Volume of sample solution taken, ml 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.26 containing 350 mg of thorium per mf Number portions Conc entra- of solvent Volume tion of prepared of oxine in and Solvent solvent, solvent, counted hexone 2 12.5 3 hexone 2 !1.6 2 hexone 2 !2-5 3 hexone 2 !Mi 3 hexone 3 21.5 3 hexone 4 2.5 33 hexone 5 2.5 3 hexone 6 2.5 2 hexone 7 2.5 2 hexone 10 2.6 2 of 250-pl ml % w/v sodium diethyldithiocarbamate in hexone in presence of EDTA - Mean of results 7 to 11 = 63.4 pg per ml.Spread of counts, 7 1 1 3 2 2.6 1 3 2 3 % - Uranium- 233 found, 50.7 68.9 52.4 55.5 68.8 61.7 64.6 63.1 61.7 64- 1 634 PLg Per mlJanuary, 19581 URANIUM BY SOLVENT EXTRACTION. PART I1 TABLE IIID RESULTS OF ANALYSES OF A SOLUTION OF IRRADIATED THORIUM METAL IN NITRIC ACID, SAMPLE NO. 5355 Ratio of EDTA to thorium was 1 to 5. Final volume of aqueous phase was 2 ml, 19 Analysis NO. 1 2 3 4 5 Volume of sample solution taken, ml 0.1 0.1 0.1 0.1 0.25 containing 321 mg of thorium per d- Number portions of 260-pl Concentra- of solvent Volume tion of prepared Spread of oxine in and of Solvent solvent, solvent, counted counts, ml % w/v % hexone 2 2.5 2 2.5 hexone 2 2.5 3 1.5 hexone 2 2-5 3 2.0 hexone 5 2.5 2 hexone in presence of EDTA - - - sodium diethyldithiocarbamate in Mean of results 4 and 5 = 50.2 pg per ml.Uranium- 233 found, 46.1 48-8 47.2 50.3 60.2 Pg Per ml Volume of 2 5 % oxine solution in hexone, ml Fig. 2. Recovery of uranium-233 from sample No. 6354, which contained 63.4pg per ml, as a function of volume of solvent From these experiments, it appears that, for high ratios of uranium-233 to thorium, when an aliquot containing less than 50 mg of thorium is adequate, 5 ml of a 2.5 per cent. solution of oxine in hexone is the optimum volume of solvent. With very low concentra- tions of uranium-233, 2 ml of a 10 per cent. solution of oxine in hexoiie is preferable, since this gives a solvent phase of higher specific activity.Sufficient counts can then be obtained from the evaporation of two or more 0.25-ml portions directly on counting trays, rather than by the destruction of much larger portions by the procedure described in Part I of this series1 This saves much time and the possibly lower accuracy that results from working with small volumes of concentrated oxine solution is negligible at these levels. EXTRACTION OF URANIUM-233 IN THE PRESENCE OF FLUORIDE IONS- It is usual to add fluoride ions to about 0.05 M to catalyse the solution of thorium or thoria in nitric acid.6 Many of the analyses described earlier, such as those in Table 111, were carried out on such solutions. The possible interference of fluoride ions was tested by carrying out duplicate determinations of uranium-233 in the presence of various concentra- tions of fluoride ions.The solvent phase was 5 ml of a 2-5 per cent. solution of oxine in hexone and the aqueous phase was 0.1 ml of thorium nitrate solution, diluted to 5 ml and containing 101 pg of uranium-233 per ml and EDTA, the results being as follows- Initial concentration of fluoride ions, M 0 0.05 0.1 0.5 Uranium-233 found, pg per ml . . . . 100,104 101,102 102,104 103.10320 CLAYTON, HARDWICK, MORETON-SMITH AND TODD : THE DETERMINATION OF [VOl. 83 With 3 ml of a solution containing 180 mg of thorium per ml and 0.05 pg of uranium-233 per ml and which was made 0.5 M in fluoride ions in the presence of excess of EDTA, 84 and 86 per cent. of the total of 0-15 pg of uranium-2331 was recovered in duplicate determinations.Since this concentration of fluoride is ten to twenty-times greater than is encountered in practice, and since no effect has been observed with fluoride ions at 0-03 to 0-05 M , it may be concluded that fluoride ions do not interfere at the higher concentration levels and that at very low concentrations of uranium-233 their effect is negligible. EFFECT O F EDTA ON THE EXTRACTION OF URANIUM-233- It was found that, with solutions of diethyldithiocarbamate in hexone at pH 2.5, EDTA did not interfere with the extraction of microgram amounts of uranium. Similar experiments with oxine in hexone were carried out, in which 2-ml portions of a solution containing only uranium-233 and different amounts of EDTA were extracted with 2-ml portions of a 2.5 per cent. solution of oxine in hexone at pH 7 . The amount of uranium-233 present was 20 pg and the results, which show that there was no interference by EDTA up to a total of 80 mg (0.11 M ) , were as follows- Concentration of disodium salt of EDTA in aqueous phase, mg per ml .. .. 0 8 16 40 Uranium-233 found, pg .. .. .. 20.3 20.0 20.9 20.1 Cabell6 showed that a weak complex was formed between uranyl ions and EDTA in acid solution, but he did not determine the stability constant of the complex. Further-, over the pH range in which we are interested, Le., 4-1 to 6.5, his curve for the titration of the complex with alkali showed a steadily increasing rate of consumption of hydroxyl ions, which indicates considerably less hydrolysis of the complex in the lower part of this range.Thus in this lower pH region, the stability of the complex may be sufficiently great for complexing by EDTA to compete with complexing by oxine. Recoveries will be expected to decrease with decreasing pH and increasing concentration of E:DTA. To test this hypothesis, experiments were carried out in which both pH and concentration of EDTA were varied. A series of solutions containing 2 pg of uranium-233 per ml was treated with increasing amounts of the disodium salt of EDTA and the solutions were adjusted to pH 4.1 ; another series was treated likewise but adjusted to pH 7 . In each series the amount of uranium-233 extracted by a solution of oxine -in hexone decreased with increasing EDTA concentration. For 90 mg of EDTA per ml at pH 7, 78 per cent. of the uranium-233 was extracted, but only 20 per cent.was extracted at pH 4-1. These very low recoveries were in marked contrast with the results given in Table I-[, from which it can be seen that extraction was 100 per cent. at pH 6.5 in the presence of 100mg of thorium per ml and an excess of EDTA of as much as 140 mg per ml. An explanation is readily found; according to Cabell,6 the thorium-mono EDTA complex will bind a further molecule of EDTA (Y)- Th4+ + Y4- FA ThY ThY + Y4- +Thy$- Provided that the second molecule of EDT,4 (Y) is sufficiently strongly bound under the prevailing pH conditions, complete extraction of the uranium should then be realised at all concentrations of EDTA up to that corresponding to a molar ratio with thorium of 2 to 1. ACCURACY AND PRECISION OF THE METHOD The results given in Table 111, which were obtained under conditions that are accepted as being reliable, show standard deviations of 1.2 and 2.4 per cent.An earlier test on sample No. 3977, although at pH 4, had shown a standa.rd deviation of 1.6 per cent. No other tests of precision were carried out, but, in the use of the method for plant control, agreement between duplicate analyses was better than 2 per cent. in 80 per cent. of the analyses carried out at the 100 pg per ml level, while at sub-microgram levels, agreement was better than 10 per cent. The method is rapid, a duplicate analysis at. sub-microgram levels taking only 2 hours, of which 1 hour is occupied in alpha counting. The latter operation is done automatically. No systematic error has revealed itself.DISCUSSION OF RESULTS It has been shown that a solution of oxine in chloroform or hexone can quantitatively extract sub-microgram to milligram amounts of uranium from a solution containing thoriumJanuary, 19581 URANIUM BY SOLVENT EXTRACTION. PART I1 21 and its daughters, while the latter elements are held in the aqueous phase as their EDTA complexes. Drysen and Dahlberg' have studied the extraction of uranium from perchlorate solutions by solutions of oxine in hexone and in chloroform, and it may be estimated from their data that extraction into the solvent phase when a 2.5 per cent. solution of oxine is used as the solvent should be complete above pH 3.5. In all experiments, recovery was complete at pH 4.1 in the absence of EDTA, and often recovery was complete at pH 4.1 from thorium solutions containing EDTA.I t is probable that low results at pH 4.1 were due to interference by EDTA, when the amount of EDTA added was in excess of that which could form complexes with the thorium present. In the experiment to determine the effect of pH on the extraction of uranium-233 in the presence of thorium and EDTA, the molar ratio of EDTA to thorium was 3 to 1 and extraction was incomplete below pH 6. However, in the ten extractions to test the precision of the method there was a molar ratio of EDTA to thorium of 3 to 1, but the volume of the aqueous phase was made up to 5 ml, which gave a concentration of EDTA of 20 mg per ml. Interference by EDTA is reduced by raising the pH and complete extraction is readily achieved at higher pH values provided that the amount of EDTA added is not too great.A 10 per cent. excess above equimolar proportions appears to be quite satisfactory and is recommended, although much larger excesses can be tolerated. TABLE IV EFFECT OF EDTA ON THE EXTRACTION OF URANIUM IN THE ABSENCE OF THORIUM Five-millilitre portions of aqueous stock solution were extracted with 5ml of a 2.5 per cent. solution of oxine in hexone Volume of aqueous Weight of Concentration of pH of stock Uranium-233 stock solution, EDTA EDTA, solution before found, ml added, mg mg per ml sampling PI3 Per ml Experiment A - 50 40 30 20 50 40 30 20 Expeviment B- - 270 2500 300 1300 1300 1300 - 6.8 90 9 0 6 38-5 82 137 4.1 4- 1 4.1 7 7 7 7 7 2.0 1-95 0.39 1-56 2.07 2.0 1-85 1.45 Extraction with oxine is obviously applicable to other uranium isotopes. When the isotope is of low specific activity, e g ., uranium-238, it may be determined after separation by other methods, such as Auorimetry or colorimetry. For a fluorimetric finish, a portion of the solvent should be evaporated on a clean platinum dish before fusion with sodium fluoride. It is hoped to describe the use of extraction with oxine for the determination of uranium in the presence of other metals, such as bismuth, in Part I11 of this series. The use of EDTA in the extraction with a solution of diethyldithiocarbamate in hexone does not prevent bismuth-212 entering the solvent phase, but when the uranium in the solvent is to be determined by methods other than counting this is no disadvantage.Also, when the delay necessary for bismuth-212 to decay is allowable, this represents a very simple and accurate method for determining uranium-233 in thorium solutions. DISCUSSION OF THE DETERMINATION OF URANIUM-233 BY COUNTING King and JacksonS have shown that uranium-232 is likely to be present in uranium-233 produced by neutron irradiation of thorium. They point out, for example, that if a thorium ore has contained uranium-238, then its daughter thorium-230 (ionium) will be present in the separated thorium and thorium-230 produces uranium-232 by successive neutron capture as follows- fl,Y P %,Y P 230Th -+ 231Th ----f 231Pa -> 232pa -> 232U Uranium-232 is an alpha emitter and decays with a half-life of 70 years to give thorium-228, a member of the thorium-232 decay chain.Hence a process that has separated uranium-233 from thorium containing uranium-232 will give a product whose apparent specific activity22 CLAYTON, HARDWICK, MORETON-SMITH AND TODD: THE DETERMINATION OF [VOl. 83 will slowly increase because of the growth of t:horium-228 and its daughters. King and Jackson have also demonstrated that a specimen containing as little as 0.002 atom per cent. of uranium-232 may show as much as 2 per cent. contamination 1 year after separation. The analytical methods described separate uranium from thorium and its daughters and the alpha activity of the oxine in hexone phase is due only to uranium isotopes. This has been confirmed by alpha pulse analysis carried out by Mrs. K. M. Glover.However, in experiments in which uranium-233 is used as a tracer and when concentrations are deter- mined by direct alpha counting, without separation from possible thorium contamination, errors may arise. METHOD FOR DETERMINING URANIUM-233 IN THORIUM NITRATE SOLUTIONS BY EXTRACTION WITH OXINE REAGENTS- ketone. ketone. acid in 500 ml of water containing 80 g of sodium hydroxide and make up to 1 litre. Oxine solution A--4 10 per cent. w/v solution of 8-hydroxyquinoline in isobutyl methyl Oxine solution B-A 2.5 per cent. w/v solution of 8-hydroxyquinoline in isobutyl methyl EDTA solution-Dissolve 372.9 g of the disodium salt of ethylenediaminetetra-acetic 1 ml = 232 mg of thorium. Nitric acid, N. Ammonia solution, sp.gr. 0.880. Ammonia solution, 0.2 N. Bromothymol blue indicator solution.Anti-creeping solution-A 20 per cent. soluticln of ammonium chloride containing 2 per cent. of a water-soluble glue (Stephen’s Stefix was found to be suitable). PROCEDURE FOR 0.01 TO 1 pg OF U R A N I U M - ~ ~ ~ PER ml- With a pipette place a suitable volume of sample solution, containing not more than 600 mg of thorium, in a 40-ml centrifuge tube fitted with a glass stirrer. Add EDTA solution to give about a 10 per cent. excess over the thorium equivalent and then add 3 drops of bromothymol blue indicator solution. Return the colour of the indicator to yellow by adding N nitric acid and then add 0.2 N ammonia solution until the colour of the indicator just turns back to blue (pH 7). Add 2 ml of oxine solution A, stir for 5 minutes, spin in a centrifuge to separate the phases and then stopper the tube.Evaporate duplicate 0-25-ml portions of the solvent phase slowly on stainless-steel counting trays that have had 1 drop of anti-creeping solution evaporated at their centres1 Heat the trays to redness in the flame of a Meker burner, cool and count. Add ammonia solution, sp.gr. 0*880, until the indicator turns blue. PKOCEDUKE FOR 1 TO 1oopg OF URANIUM-233 PER ml- With a pipette place a suitable volume of sample solution, containing about 10 pg of uranium-233, in a 40-ml centrifuge tube and dilute to 3 ml with water. Add EDTA solution to give a 10 per cent. excess over the thorium equivalent. Add 2 drops of bromothymol blue indicator solution and adjust the pH to 7 as previously described. Add 5ml of oxine solution B, stir for 5 minutes, spin in a centrifuge to separate the phases and then stopper the tube.Evaporate duplicate 0.1 or 0.25-ml portions of the solvent phase for counting, as before. Note that for a fluorimetric finish to either procedure, suitable duplicate portions of the solvent phase should be evaporated in platinum fluorimeter dishes before fusion with sodium fluoride. METHOD FOR DETERMINING URANIUM-233 IN THORIUM NITRATE SOLUTIONS BY EXTRACTION WITH SODIUM DIETHYLDITHIOCARBAMATE REAGENTS- Hexone. Sodium diethyldithiocarbamate solution-A freshly prepared and filtered 20 per cent . w/v aqueous solution.January, 19581 URANIUM BY SOLVENT EXTRACTION. PART 11 23 EDTA solution-Prepared as described in reagents list, p. 22. Ammonium nitrate solution, 2 M, Ammonia solution, P.gr.04380. Nitric acid, concentrated and N. Screened methyl orange indicator solution. Anti-creeping solution-A 20 per cent. solution of ammonium chloride containing 1 per cent. of a water-soluble glue. PROCEDURE FOR 1 TO 100 pg OF U R A N I U M - ~ ~ ~ PER ml- With a pipette place a suitable volume of sample solution, containing about 10% of uranium-333, in a 40-ml centrifuge tube fitted with a glass stirrer. Dilute to 4-ml with 2 M ammonium nitrate and add EDTA solution to give a 10 per cent. excess over the thorium equivalent. Stir and make just alkaline to screened methyl orange by adding ammonia solution and then add 0.5 ml of sodium diethyldithiocarbamate solution. Stir and add N nitric acid until the solution is mauve (not red). Add 5 ml of hexone, stir for 5 minutes and add more acid to maintain the mauve colour if necessary. Spin in a centrifuge to separate the phases and then stopper the tube. Evaporate suitable duplicate portions of the solvent phase on stainless-steel counting trays that have had 1 drop of anti-creeping solution evaporated at their centres. Heat the trays to redness in the flame of a Meker burner, allow the bismuth-212 to decay, and then count. Alternatively, for a fluorimetric finish, evaporate duplicate portions of the solvent phase in platinum fluorimeter dishes for fusion with sodium fluoride. Note that greater sensitivity can be obtained by starting with a larger volume of sample or by evaporating larger portions of the solvent phase. REFERENCES 1. Hardwick, W. H., and Moreton-Smith, M., Analyst, 1958,83, 9. 2. Bode, H., 2. anal. Chem., 1955, 144, 165. 3. Pfibil, R., and MalAt, M., Coll. Czech. Chewz. Comm., 1950, 15, 120. 4. Welcher, F. J., “Organic Analytical Reagent,” D. Van Nostrand Co. Inc., Mew York, 1947, Volume 6. Seaborg, G. T., and Katz, J . J., “The Actinide Elements,” McGraw-Hill Book Co. Inc., New York, 6. Cabell, M. J., Analyst, 1952, 77, 859. 7. Drysen, D., and Dahlberg, V., Acta Chem. Scand., 1953, 7 , 1186. 8. King, D. T., and Jackson, N., Atomic Energy Research Establishment Report CE/R 1876, Harwell. Received May 28th, 1967 I, p. 266. 1954, p. 176.