Separation of Uranium from Thorium, Bismuth, and Ores with Tributyl Phosphate Spectrophotometric Determination with 8-Quinolinol A. R. EBERLE and M. W. LERNER
U S. Afomic Energy Commission, New Brunswick Laboratory, New Brunswick, N. J. b Uranium in thorium, bismuth, and a variety of ores can b e separated b y extraction with tributyl phosphate in such purity as to permit its determination with 8-quinolinol. Uranium is separated from thorium b y extraction from a hydrochloric acid solution, and from bismuth and other elements from a nitric acid solution. Optimum conditions for the extraction o f uranium from nitric acid solution and the subsequent color development with 8quinolinol are presented. Results are reported on the recovery of uranium from synthetic solutions containing many diverse ions. Thorium, bismuth, and low-grade ores were analyzed.
A
and reliable spectrophotometric method of determining microgram amounts of uranium in bism u t h was needed, which could be adapted, with little modification, to the analysis of thorium and a wide variety of complex materials. Uranium in bismuth has been determined by a polarographic procedure after a partial separation of the bismuth by an oxychloride precipitation (28). However, in addition to the danger of coprecipitation when the lower microgram range of uranium must be determined, this procedure suffers both from a lack of sensitivity of the polarographic methods for uranium and from its limited applicability. While there are many chromogenic reagents for uranium, all are nonspecific. T h e development of a spectrophotometric method, then, must involve the prior separation of the uranium in high purity. This separation problem has been met in the case of thorium by an oxalate precipitation (3, 29), by a cellulose column procedure (6, 6), and by solvent extraction with diethyl ether (4, dibutoxytetraethylene glycol (SS), thenoyltrifluoroacetone (4,and methyl isobutyl ketone (4). For generally complex materials, a similar list of solvents has been used: diethyl ether (SO, S I ) , ethyl acetate ( 7 , 11, 1?2, 1 4 ) , dibutoxytetraethylene RAPID
‘1 134
ANALYTICAL CHEMISTRY
glycol (16, 18, 19, 38). diethyl Cellosolve ( 2 7 ) , and other polyethers (18). Solvent extraction appeared to be the only technique that would satisfy the requirements of the proposed general method. However, in the extraction procedures mentioned above, thorium. bismuth, cerium(It’), vanadium, iron, molybdenum, zirconium, phosphate, fluoride, chloride, or sulfate may cause difficulties. Recently, great interest has been shown in the use of tributyl phosphate as a n extracting agent for uranium. Fundamental studies on the use of this solvent mere made by Karf (%), Moore ( d S ) , Il‘right (S?’),Bartlett ( I ) , and others (2, 10). A study of the relative merits of dibutoxytetraethylene glycol and tributyl phosphate was made by Wright (38). LeStrange, Lerner, and Petretic (22) determined the uranium directly in the solvent spectrophotometrically or polarographically after extraction with tributyl phosphate. Paige, Elliott, and Rein (24) determined uranium spectrophotometrically in the ultraviolet region after a similar extraction. Guest ( I S ) extracted uranium with tributyl phosphate and ethyl acetate and determined the uranium in an aqueous wash of the solvent by the usual peroxide or thiocyanate procedure. Kimball and Rein (17’) recently determined the uranium in the organic phase with thiocyanate after extraction. I n these extraction procedures, the aqueous phase is either a 4 to 6hr nitric acid solution, or a less concentrated nitric acid solution containing a salting agent such as ammonium, sodium, calcium, or aluminum nitrate. Under these conditions, iron(III), copper, nickel, bismuth, the rare earths, and chromium(II1) are extracted in trace amounts, thorium, zirconium, and cerium(1V) are extracted in larger quantities. I n the proposed method, all of these impurities except thorium and zirconium are removed by repeated washing of the organic phase with nitric acid solutions. Traces of thorium and zir-
conium, however, cannot be removed by this procedure. Peppard and Gergel (26) determined the distribution coefficients of uranium and thorium in tributyl phosphatehydrochloric acid systems. With equal volumes of tributyl phosphate and 5.2N hydrochloric acid, the distribution coefficients (organic-aqueous) of uranium and thorium were found to be 40 and 0.006; nith 11.81: acid, 52.000 and 6, respectively. The use of this type of system, therefore, seemed promising for the analytical separation of uranium from thorium-containing materials. Indeed, preliminary tests on the application of a 6 S hydrochloric acid-tributyl phosphate system to the separation of trace amounts of uranium in thorium metal were successful (20). I n more recent exploratory work, it was found that the concentration of hydrochloric acid could be increased to 7 N . With this concentration, the extraction loss of uranium m s kept at a minimum, and the thorium and zirconium could still be washed out of the tributyl phosphate. This concentration was used in the procedure reported here. The available chromogenic reagents for uranium have been summarized by Rodden ( S I ) and M7are (55). From considerations of availability, sensitivity, and stability, 8-quinolinol was selected. [Recently Pribil and Jelinek (26) and Yoe, Will, and Black (39) have proposed the use of dibenzoylmethane. Although this reagent is much more sensitive than 8-quinolinol, it was not used in the present study, as it was being used in a companion investigation ( d l ) . ] The applications of 8-quinolinol to the determination of uranium have been discussed by Rodden (31). More recently, Silverman, hloudy, and Hawley (54)determined uranium in the presence of iron with 8-quinolinol and summarized the separation procedures for interferingelements. Dryssen and Dahlberg (8) studied the extraction of uranium by 8-quinolinol in chloroform. Rulfs, De, Lakritz. and Elving (52) also investigated the extraction of uranium
by 8-quinolinol and some of its derivatives. I n the present procedure, the uranium is extracted from a 7N hydrochloric acid solution of thorium metal Kith an equal volume of 50% (v./v.) tributyl phosphate-methyl isobutyl ketone. The thorium extracted along with the uranium is removed by washing the organic phase with 7 N hydrochloric acid. Other impurities, such as iron(111), which are also extracted from the 7N hydrochloric acid solution, are then removed by washing the organic phase with 4.7N nitric acid. The uranium is removed from the tributyl phosphate by washing with an ammonium acetate solution, and then is extracted by a n 8-quinolinol-chloroform solution and determined spectrophotometrically. The same procedure is used for the analysis of monazite sands. For materials containing no significant amounts of thorium or zirconiumsuch as carnotite, phosphate rock, lowgrade pitchblende, and bismuth metalextraction from a hydrochloric acid solution is not necessary. The uranium is extracted from a 4.7N nitric acid solution with 50% tributyl phosphatediethyl ether. The elements in the organic phase other than uranium are removed by washing with 4.7N nitric acid. The interference of phosphate and sulfate ions is eliminated by the addition of aluminum nitrate t o the acid solution before extraction. Some loss of uranium can be expected during the extraction and washing steps. From approximate interpolations of the data of Peppard and Gergel ( 2 5 ) , the loss during the 7N hydrochloric acid extraction and washing can be assunied to be small, less than 5yG, and constant. On the other hand, because the final removal of uranium from the tributyl phosphate and the subsequent color development are carried out in each case after a nitric acid extraction and washing procedure, the concentration of nitric acid has a direct effect on the over-all recovery of uranium in these three steps. This effect was investigated. Likewise, the optimum concentration of the ammonium acctate solution used t o remove uranium from the organic phase, and the most favorable pH for the extraction of the uranium 8-quinolinolate from the ammonium acetate solution T\ere determined. The absorption spectra of uranium-, thorium-, bismuth-, and zirconium 8-quinolinolates in chloroform were compared to obtain the optimuni wave length for the dctermination of uranium. Finally, recovery tests were carried out with synthetic solutions containing diverse metal ions, and analyses were made of thorium, bismuth, monazite, phosphate rock, pitchblende, and carnotite samples.
APPARATUS AND REAGENTS
All spectrophotometric measurements were made with a Beckman Xodel DU spectrophotometer with 2- or 5-cm. cells. The measurements of pH were made with a Beckman RIodel h l pH meter. Reagent-grade chemicals were used unless otherwise stated. 8-Quinolinol-Chloroform Solution. A 1% solution was made from 8-quinolinol obtained from Eastman Organic Chemicals. This solution is stable for a t least 5 days. Ammonium Acetate Solution. A 2501, solution was aremred bv diluting 369 ml. of glacial ice& acid k i t h 1000 k l . of water, neutralizing with 430 ml. of concentrated ammonium hydroxide, adjusting the pH after cooling to 7.6 with ammonium hydroxide, and diluting to 2 liters. Interfering impurities were removed by extracting 500-ml. portions of the solution with 100 ml. of the 1% 8quinolinol-chloroform solution followed by washing with two 100-ml. portions of chloroform. Aluminum Nitrate. Some lots of reagent-grade aluminum nitrate contain traces of uranium. The material can be analyzed for uranium by carrying a saturated solution through the procedure outlined below for the analysis of bismuth. If uranium is present, it can be removed by extracting a saturated solution iyith an equal volume of tributyl phosphate. Dilute hydrofluoric acid was prepared by adding 10 ml. of concentrated (48%) acid to 200 ml. of water in a polyethylene bottle. Tributyl phosphate, obtained from Commercial Solvents Corp., was used as received. Standard Uranium Solution. A stock solution of 1 mg. per ml. was prepared by dissolving 1.1702 grams of pure black oxide, U308, in nitric acid and diluting the solution to 1 liter. A standard solution of 100 y per ml. was made by diluting 100 ml. of this stock solution to 1 liter. PROCEDURE
Preparation of Standard Curves. No BISMUTHA N D ORES COXTAIKING THORIUM OR ZIRCOKIUM. Add 0, 100, 300, and 500 y of uranium t o 50 ml. of 4.7hT nitric acid in a 250-ml., shortstemmed, separatory funnel. Add 25 ml. of tributyl phosphate and 25 ml. of diethyl ether. Shake the cantents vigoroiisly for about 30 seconds. Discard the aqueous (lower) phase. K a s h the organic phase with five 50nil. portions of 4 . 7 s nitric acid. Discard the washings. Remove the uranium from the organic phase by extracting successively with 20-, 15-, and 15-ml. portions of the 25% ammonium acetate solution, collecting the ammonium acetate solution in a 150-ml. beaker. Adjust the pH nith concentrated ammonium hydroxide to 7.6. If this value is exceeded, use dilute nitric acid t o obtain the correct pH. Transfer the solution to a 250-ml. separatory funnel. Wash the beaker with no more than 15 ml. of water and
add the washing to the funnel. Extract the solution with 25 ml. of the 1% 8-quinolinol-chloroform. Filter the chloroform phase through a folded R h a t m a n No. 41 paper, 12.5-cm., placed in the mouth of a dry 150-ml. beaker. Measure the absorbance a t 470 and 500 mp in 5-cm. cells against the standard containing no uranium as a reference and plot the absorbance against uranium concentration. THORIUM AND MONAZITE ORES. Add 100, 300, and 500 y of uranium to 50 ml. of 7” hydrochloric acid in a 250-ml., short-stemmed, separatory funnel. Add 25 ml. of tributyl phosphate and 25 ml. of methyl isobutyl ketone. Shake the contents vigorously for 30 seconds. Discard the aqueous (lower) phase. Wash the organic phase with fire 50-nil. portions of 7N hydrochloric acid. Discard the washings. Wash the organic phase with four 50-ml. portions of 4.7K nitric acid. Discard the washings. Remove the uranium from the organic phase and re-extract as previously described. Measure the absorbance of the filtered extract in 5-em. cells a t 470 and 500 mp, against a reference solution prepared as described above, and plot the absorbance against uranium concentration. Analysis of Samples. T H O R I U M METAL. For thorium metal containing less than 40 p.p.m. of uranium, dissolve a 15-gram sample in a 400-ml. beaker with 100 ml. of concentrated nitric acid and 5 ml. of the dilute hydrofluoric acid. Place the beaker on the steam bath until dissolution is complete, then transfer it to a hot plate. Evaporate the solution to dryness and continue heating the residue until it becomes powdery. Ignite the residue in a muffle furnace a t 400’ C. for 30 minutes. Add 50 ml. of concentrated hydrochloric acid and 5 ml. of the dilute hydrofluoric acid t o the beaker and boil the mixture until the oxide is dissolved. Transfer the solution, TI hile still warm, t o a 250-ml., short-stemmed separatory funnel. Add 25 ml. of tributyl phosphate and 25 ml. of methyl isobutyl ketone, and proceed as described above under the procedure for the standard curve preparation for thorium. &leasure the absorbance a t 470 and 500 my. ~IONAZIT ORES. E. To a finely ground 0.5-gram sample in a 15O-ml. beaker add 10 ml. of concentrated sulfuric acid. Heat the mixture on a hot plate to near dryness. Cool the mass, add 100 ml. of ice-water, and stir until the salts dissolve. Transfer the solution t o a 250nil. volumetric flask, add 25 ml. of concentrated nitric acid, and dilute t o volume with water. Samples resistant to this attack may require fusion v, ith sodium peroxide. With samples rich in zirconium and other elements that yield hydrolytic precipitates after decomposition, filtration of the precipitate may be carried out before dilution to volume, as occlusion of uranium by these precipitates is generally negligible ( 1 2 ) . Transfer an aliquot of the solution containing 50 t o 500 y t o a 250-ml. beaker. Dilute the solution to about 200 ml. with water and precipitate the VOL. 29, NO. 8, AUGUST 1957
1135
hydroxides by adding concentrated ammonium hydroxide until the solution is strongly basic. Add a small amount I $ of macerated filter pulp and filter the 0.4 k, precipitate by means of a Fisher Filtrator and a 60-ml. sintered-glass Buchner funnel of medium porosity. Dissolve the hydroxides with 50 ml. of warm 7 S hydrochloric acid. Extract and determine the uranium as described above for thorium metal. BISMUTHMETAL. Dissolve a 1- to 10gram sample to contain from 50 to 500 y '. '. .' of uranium in the minimum quantity of '. concentrated nitric acid and evaporate '. --.--. the solution to near dryness on the steam -.__* - - - _ _ _ _ bath. Dissolve the salt in 50 ml. of -. > I I--. t---\ I 4 . 7 s nitric acid and transfer the soluz30 450 470 490 610 530 550 tion to a 250-ml., short-stemmed separatory funnel. Add 25 ml. of tributyl phosphate and 25 ml. of diethyl ether. Proceed as described for the preparation of the standard curve for uranium in bismuth and ores containing no thorium. Uranium 8-quinolinolate Measure the absorbance a t 500 mp. _ _ _ _ _ _ Bismuth 8-quinolinolate _ _ _ Zirconium 8-quinolinolate PHOSPHATE ROCK.T o a finely ground - _ - _ - Thorium 8-quinolinolate 5-gram sample in a 250-ml. beaker, add ~8-Quinolinol 50 nil of 8M nitric acid, boil the mixture for 5 minutes, and filter through a 11cm. Whatman No. 42 paper. Wash the 100 beaker and transfer the insolubles to the monium acetate solutions a t pH 7.6 paper with a small amount of nater. containing either no added element or Reserve the filtrate. 100 y of the element concerned. Ignite the paper in platinum a t 900' A variety of wave lengths have been C. Add a few drops of concentrated used for the spectrophotometric meassulfuric acid and about 10 nil. of concenurement of uranium 8-quinolinolate trated (48%) hydrofluoric acid to the (31). Recently, wave lengths of 400, residue and evaporate the mixture t o 420, or 440 mfi (34), 430 mp (Sd), and dryness on a hot plate. Add 5 grams of potassium bisulfate and fuse the mixture 425 mp ( 1 5 ) have been proposed. over a Meker burner until a clear melt Figure 1 shows that bismuth 8-quino0 1 2 3 4 5 6 7 8 is obtained. Dissolve the cooled melt in linolate begins to absorb strongly N O R M A L I T Y OF NITRIC A C I D 4.7N nitric acid and add the solution t o I n extracting uranium below 490 mp. the filtrate. Dissolve 10 grams of aluFigure 2. Effect of nitric acid confrom large amounts of bismuth, traces minum nitrate nonahydrate in the solucentration in initial extraction and of bismuth may not be removed by the tion and dilute to 250 ml. with 4.75 washings on over-all recovery of mashes. To eliminate the interference nitric acid. uranium of these traces of bismuth, the uranium Transfer an aliauot of less than 100 0 50 ml. of 25% ammonium acetate 8-quinolinolate is measured a t 500 mp. mll containing 50 'to 500 y of uranium A 80 ml. of 25% ammonium acetate to a 2X-ml. separatory funnel. Add 25 I n the determination of uranium in ml. of tributyl phosphate and 25 ml. of thorium and in ores, a wave length of diethyl ether and continue as described tributyl phosphate extraction, re470 mp is used to obtain more sensiabove for the preparation of the standextraction with ammonium acetate, tivity, although traces of thorium and ard curve for bismuth and ores containand the uranium 8-quinolinolate eyzirconium may interfere. At lower ing no thorium. Measure the absorbtraction step was studied. I n these wave lengths, these elements interfere ance a t 470 and 500 mp. tests, 50-ml. portions of nitric acid of CARNOTITE AND LOW-GRADE PITCH- seriously. At 500 mp the interference 2 to 8 5 concentration containing is negligible. BLEXDE. Dissolve samples of suitable 200 y of uranium were extracted with weight in the manner described for The standard curves, therefore, were 50 ml. of 50% (v./v,) of tributyl phosphosphate rock, but omit the addition prepared a t both 470 and 500 mp. of the aluminum nitrate. Determine the phate-methyl isobutyl ketone preI n analyzing a sample of thorium or uranium by the procedure used for phosviously equilibrated with an equal thorium-containing ore, the measurephate rock. portion of nitric acid of the same ment of the absorbance a t both wave concentration. The organic phase was lengths was found to be an excellent washed with 50-ml. portions of nitric criterion for the purity of the uranium EXPERIMENTAL acid of the same concentration and 8-quinolinolate-chloroform extract. A then the uranium was removed and designificant difference in the results calAbsorption Curves. The spectral termined in the usual manner. culated from the standard curves a t absorption curves of 8-quinolinol and As shown in Figure 2, the recovery 470 and 500 mp indicates an imperfect the 8-quinolinolates of uranium, thoof uranium decreases rapidly when the separation of the uranium. rium, bismuth, and zirconium in chloronitric acid concentration is lowered The color of the uranium 8-quinoform were determined over the range from 4-Y,because of incomplete extraclinolate in chloroform is completely of 430 to 550 mp. These curves, tion of uranium by the tributyl phosstable for at least 40 hours. shown in Figure 1, were obtained by phate, Above 5N nitric acid concenEffect of Nitric Acid Concentration. measuring the absorbance with 2-cm. tration, the recovery again falls off, beThe effect of the nitric acid concentracells against chloroform, of 25-ml. cause of incomplete removal of the tion of the aqueous phase during the portions of the 1% 8-quinolinoluranium from the organic phase by the extraction and washing procedure on chloroform shaken for 2 to 3 minutes ammonium acetate. By increasing the the over-all recovery of uranium in the with 50-ml. portions of the 25% am-
1
.--___ -.
1 136
ANALYTICAL CHEMISTRY
0.1
1
6.0
I
I
I
7.0
6.5
I 8.0
7.5
With a single 50-ml. wash or two 25-ml. washes, only 95% of the uranium was removed. Rather than increase the volume of the ammonium acetate solution, it was felt advisable to use the three washes indicated. It can be seen from Figure 4 that with a concentration of ammonium acetate below 20% the uranium is not completely removed from the organic phase with the three washes. Above 30% the absorbance again decreases, presumably because of the cornplexing action of large concentrations of acetate ion. Between 20 and 30% the absorbance is constant. Accordingly, a 25% solution \vas selected for the removal of the uranium from the tributyl phosphate. Diluents. The use of a diluent for the tributyl phosphate possessing a lower specific gravity than tributyl phosphate materially assists the extraction and mashing steps because of the resulting rapid phase disengagements. Diethyl ether and methyl isobutyl ketone were selected for the extractions with nitric acid and hydrochloric acid solutions, respectively. While both diluents are equally satisfactory for extractions from nitric acid, the solubility of diethyl ether in the hydrochloric acid solutions causes a fairly large loss of the ether in the extraction and washing steps. The loss of methyl isobutyl ketone in these manipulations is less. All work with this latter solvent should be carried out in an efficient hood, as the vapors are toxic (9). The tributyl phosphate-diethyl ether solution is much less hazardous to handle than ether alone, because of drastically reduced vapor pressure of the ether in the mixture. Warm nitric acid solutions were extracted without any pressure build-up in the separatory funnels. hloreover, concentrations of nitric acid up to 8N were extracted without any noticeable decomposition of the ether.
8.5
PH
Figure 3. Effect of pH on extraction of uranium with 8-quinolinol from ammonium acetate wash
1
0
eo
10
X
30
AMMONIUM
40 ACETATE
50
Figure 4. Effect o f ammonium acetate concentration on removal of uranium from tributyl phosphate and extraction with 8-quinolinol
total quantity of the 25% ammonium acetate solution to 80 ml. and using two additional 15-ml. washings, the recovery of the uranium is found to be cssriitially constant, 85 t o 8695, with 4 to 8N nitric acid concentration. With a nitric acid concentration of 4 . 7 5 (30% v./v.), as recommended in the final procedure, the 50-ml. (total) wash of the 25y0 ammonium acetate solution can be used. It can be seen from Figure 2 that approximately 2.7% of the uranium present is lost with each contact of the organic phase with the 4 to 8.1' acid solutions. This makes it necessary to use the same number of washings for the analysis of samples as for preparation of the standard curve. Effect of pH on Extraction of Uranium 8-Quinolinolate. I n this study, 50-ml. portions of 4.7')- nitric acid containing 200 y of uranium xere extracted with 50 ml. of 50% (v./v.) tributyl phosphate-methyl isobutyl ketone. After washing the organic phase and removing the uranium with the 25y0 ammonium acetate, the combined ammonium acetate solution mas adjusted to a pH value between 6 and 8.5 with concentrated ammonium hydroxide and the uranium was extracted with 25 ml. of the 1% 8-quinolinolchloroform solution. According to the results shown in Figure 3, a pH range of 7.2 to 8.1 may be used. A pH of 7.6, the mid-point of the range, was arbitrarily selected for the final procedure. Effect of Ammonium Acetate Concentration. The effect of the ammo-
nium acetate concentration on the removal of the uranium from the tributyl phosphate-methyl isobutyl ketone together with the subsequent extraction of the uranium with the 8quinolinol-chloroform solution is shorn-n in Figure 4. I n these tests. 250 y of uranium in 50-ml. portions of 4.7N nitric acid nere extracted with 50 ml. of 50% (v./v.) of tributyl phosphate-diethyl ether. After the usual mashing of the organic phase, the uranium was removed by washing n-ith 20-, 15-, and 15-ml. portions of ammonium acetate solutions of concentrations ranging from 10 to 50'%',. The combined wash solution was adjusted to a pH of 7.6 and the uranium was extracted and measured as before. The use of the three washes was found necessary to remove the uranium from the tributyl phosphate in preliminary tests with 25% ammonium acetate.
Table
I.
Mo(V1) Y , Ce(III), La, Hg(II), Ag
Ga
sc
Ti Zr
Synthetic solutions containing ura-
Recovery of Uranium in the Presence of Diverse Ions
Ion Cu, Co, Ni, Zn, Pb Al, Fe(III), Be, Cr(II1) Ca, Sr, Mg, Bi Mn(II), Cd
WV)
RESULTS
Uranium, y Added Found 500 500
500 500 150 100 10 10 1 10
10
200 200 200
200 207
200
203
200 200 200
200 200
200 200
VOL. 29, NO. 8, AUGUST 1957
199
200 198
200
1137
nium and many diverse ions were analyzed with the results s h o m in Table I. The quantities of elements used were selected arbitrarily. The procedure for bismuth was used for solutions containing no thorium or zirconium. The solution containing zirconium was analyzed by the thorium procedure Nost of the elements listed were in the form of the nitrates or chlorides, depending on the type of final solution required. The vanadium was added as vanadate, the molybdenum as molybdate, the zirconium as zirconyl nitrate, and the titanium as titanium sulfate. Samples of the same thorium metal weighing either 10 or 20 grams were analyzed with and without the addition of uranium before dissolution of the metal (Table 11).
alyzed by the appropriate method (Table IV). Replicate analyses of a standard sample of phosphate rock, NBL phosphate rock, No. 1, were carried out to determine the precision of the method for ores of this type. Based upon nine determinations, the standard deviation for a single determination was &0.0002% and the 9570 confidence limit of the average, 0.02807,, n-as 3= 0.0002%. DISCUSSION
The method was not compared to the other spectrophotometric methods involving solvent extraction separations (7, 11, I S , 1‘7). These methods are not applicable to the general type of sam-
mogenic reagent in the above method and also in the methods involviiig ethyl acetate extraction (7, 11, I S ) is in the authors’ opinion a disadrantngc because of the instability of the r w gent and the color produced despite its greater sensitivity than 8-quinolinol. Bismuth and zirconium also are extracted with ethyl acetate and the sample size again is limited. The same general comnients can be made of the method of Guest ( I S ) . If peroxide is used to develop the color with the latter method, thorium is bound to interfere badly. The method proposed, in spite of the multiple washes involved, is not time.consuming. Once the. sample is i n solution, a result can usually be obtained within 45 minutes. LITERATURE CITED
Table 11.
Sample Weight, G.
a
Analysis of Thorium Metal and Recovery of Uranium
Uranium, y Present” Added Total Found 10 .. .. .. 39 20 .. .. 80 10 40 50 90 84 10 40 100 140 142 10 40 200 240 239 20 80 200 280 277 10 40 300 340 340 10 * 40 400 440 441 10 40 500 540 550 Based on mean of duplicate analyses on “unspiked” samples. -
Bismuth nitrate pentahydrate, 25gram samples equivalent to 10.8-gram samples of metal, was analyzed similarly (Table 111).
Table 111. Analysis of Bismuth Nitrate and Recovery of Uranium
Uranium, y Added Found 0
50 100 200 300 500
0
Recovery,
YO
50 102
100 102
495
98 99
197 2i4
9_18_
Samples of monazite, phosphate rock, carnotite, and pitchblende were an-
Table IV.
1 138
ANALYTICAL CHEMISTRY
7% .. ..
93 101 99 99 100 100 102
ples for which the present method was developed. I n particular, the method of Kimball and Rein (17), reported after this work was completed, cannot be used for phosphate rock samples or for solutions containing sulfate or fluoride, as these anions interfere with the extraction of uranium into tributyl phosphate in the absence of sufficient aluminum ions. Bismuth samples cannot be analyzed by the method, since bismuth extracts and gives a color with thiocyanate. Zirconium also extracts and in the analysis of monazite should interfere. The sample size of the method is also limited, so that large samples of bismuth, thorium. or low-grade ores cannot be used. The use of thiocyanate as the chro-
Analysis of Standard Ores
Sample h-atl. Bur. Standards monazite, No. 2601 New Brunswick Laboratory Phosphate rock, ATo. 1 Carnotite, KO. 4 Carnotite, No. 5 Pitchblende, No. 3-A Florida phosphate rock, KO.15 ~~
Recovery,
Accepted LTranium Content, u308,
%
0 38 0 0 0 4 0
029 18 11 29
009
Uranium Found, U308,
Mean TO 0.40 0 028 0.17 0 11
4 36 0 0089
No. of
Detns. 5 9
(1) Bartlett, T. \V.,U.S. Atomic Eneip\Commission, K-706 (1951). (2) Bearse, A. E., Bohlmann, E. ( + , Calkins, G. D., Filbert, R. H , Fishel. J. B.. Zbid.. BMI-TDS-169 ( 1949j. (3) Brown, E. -4., Ibid., FMPC-505, 201 (\ -195.5). ---*
(4) Burelbach, J. P., March, R. J , I b / ( i ,
ANL-5240 (1953). (5) Burstall, F. H., Wells, R. A , .t>,(rlust 76, 396 (1951). (6) Center, E. J., Henry, W. AI., Hal(,. R. W.,U. S. Atomic Energy Conimission. BMI-TDS-193 f l’&l). DeSesa, XI. A., Rietzel, 0’. A,, Zbid , ACCO-54 (1954). Dryssen, D., Dahlberg, V., Llcicc Chem. Scand. 7, 1186 (1953). Elkins. H. B.. “Chemistrv of 1ndii.trial Toxicology,” p. i19, n’ilc! . S e w York. 1950. ~ - - Ellison, C. V., Ferguson, D. E., Runion, T. C., LT. S. AtomicEnergy Commission, ORNL-258 (19491. 11) Feinstein. H. I., Ibid., TEI-555 I
(1955):’
Grimaldi, F. S., May, I., Fletclhei, hl. H., Titcomb, J., U. S. Ckol Survey, Bull. 1006 (1954), Guest, R. J., Dept. of hlines 2nd Tech. Surveys, hlines Branch, Canada, TR-128/55 (1955). Guest, R. J., Zimmerman, J. H , Zbid., Tech. Paper 8 (1954). Hok. 13.. Si’ensk Kern. T i d s k r . 65. 1O(i (1953). Jones, A . G., U. S. Atomic Energ) Commission, C-4-360-3 (1945). Kimball, R. B., Rein, J. E., I b i d , IDO-14380 (1956). ICraus C. A.. Ibid.. A-1099 (1944 1 Zbid.,A-2301. Lerner, 11. IT., Zbid., NBL-103, 89 (1955). Lerner, U.W.,Eberle, A. R., I b z i l , NBL-127, 8 (1956). LeStrange, R. J., Lerner, RI. K.. Petretic, G. J., Ibid., NYO-2047 (1954). Moore, R. L., Ibid., HW-15230 (1949). Paige, B. E., J. E., I i ’ .
3 2
2
5
47,’lOlf (1953). (27) Price, T. D., Ernsberger, F. hI., Ballard, A. E., U. S. Atomic Energy
Commission, CD-B-S-518 (1944). Raseman, C., Keisman, J., Nucleonics 12, ( 7 ) , 20 (1954). Read, E. B., private communication, 1956. Rodden, C. J., ASAL. CHEX 25, 1598 (1953). Rodden, C. J., “Analytical Chemistry of the Manhattan Project,” RIcGraw-Hill, New York, 1950.
(32) Rulfs, C. L., De, A. K., Lakritz, J., Elving, P. J., AKAL. CHEM. 27, 1802 (1955). Silverman, L., Moudy, L., Sucleonics 12, (9), 60 (1954). Silverman, L., Moudy, L., Hamley, D. W.,AKAL. CHEM. 25, 1369 (1953). \Tare, E., U. S.Atomic Energy Commission, MDDC-1432 (1947).
(36) Warf, J. C., J . Am. Chem. SOC.71, 3257 (1949). (37) Wright, W. B., C . S. Atomic Energy Commission, Y-838 (1952). (38) Ibid., Y-884. (39) Yoe, J. H., Will, F., Black, R. A . , AKAL.CHEST.25, 1200 (1953). RECEIVEDfor review January 25, 1957. Accepted April 3, 1957.
Photometric Determination of Silicon in Steel WILLIAM F. SANDERS and CHARLES H. CRAMER United Engineering and Foundry Co. , Vandergriff, Pa.
A . photometric method has been developed for determining the small percentages of silicon present during the refining period in the melting of steel. During the investigation, the method was found to b e adaptable to silicon in the regular range for plain carbon and low alloy steels. The photometric molybdenum blue method i s used with the ferrous ions of the sample as the reducing agent and with a compensating blank for reference. The method i s accurate, relatively simple, and well suited for the simultaneous determination o f silicon in a large number of samples.
S
ILICON is
usually determined photometrically by measuring the color of the yellow or blue complex formed with silicon and molybdate ions. The yellow color of the silicomolybdate complex is the basis for the work in various procedures (1, 2, 6, 7 , 10). The sensitivity of the method is increased by reduction of the yellow complex to molybdenum blue with a suitable reducing agent (4, 6, 8, 9, 11). The work presented here was undertaken to provide a rapid and accurate photometric method for the routine determination of silicon in steel by measuring the blue color of the reduced complex. The method comprises an rqansion and improvement of the original work briefly outlined by Hill ( 2 ) ) and provides a n analytical procedure which mill be of considerable value to practical analytical chemists. APPARATUS
Filter Photometer. T h e Fisher electrophotometer with t h e standard 23-ml. Fisher cells a n d the 650A red filter furnished with the instrument. Spectrophotometer. T h e Beckman Model B with t h e Beckman test tube holder and standard 23-ml. Fisher cells. T h e instrument was also used
with matched 1-em. borosilicate glass cells in the regular cell holder. EXPERIMENTAL
The measurement of the yelloiy color was not considered in the present work because of its low sensitivity. The far more sensitive reaction of the blue complex n-as tried and the best conditions for measurement of the blue color were established. Most of the work described was done with the Fisher electrophotometer; however, a number of absorbance values were checked on the Beckman spectrophotometer fitted with the test tube holder and the same 23-ml. cells which were used in the electrophotometer. Also, with the 1-em. cells in the spectrophotometer, the upper limit of the permissible silicon concentration could be greatly increased owing to the use of the shorter cell length and the narrow and more selective wave length of obtainable light, as compared with the filter instrument. Selection of Reducing Agent. Stanous chloride (6, Q), ferrous sulfate (8, 9), and sodium sulfite (4, 11) were tried as reducing agents hut the ferrous ion was found to offer the best possibilities for development of the blue color. According to Usatenko and Orlova (8), there is no reduction of free molybdate with the ferrous ion, and no interference from phosphorus and arsenic, as is possible with stannous chloride; also, there is no greenish hue and no erratic color development due to slight changes in temperature as there is when sulfite is used. Furthermore, instead of requiring a separate ferrous solution, the method has been simplified by partial oxidation with persulfate. thus allowing the ferrous ions of the sample itself to become the reducing agents. This procedure was described by Hill (2, S ) , who measured the blue color at 710 mp and presumably used the same sample weight, aliquots, and reagent concentrations as for the
develonment of his yellow silicomolybdate color. Concentration of Reagents and Sample Size. A careful study of the variables was made in order t o arrive a t a sufficiently stable color for accurate evaluation of silicon. Sulfuric acid, 8% and 1670, was established for acid concentration and 37, ammonium persulfate for partial osidation of the sample. Table I shows t h a t the acid concentration is not critical for the higher percentages of silicon, but should be kept within about lOyoof the amount specified in the procedure for silicon in the low range. The persulfate is stable for a t least a week and may be varied by as much as 207, of the given concentration without affecting bhe accuracy of the results. The greatest stability of color was produced when the molybdate concentration was reduced to l.GyOrather than the 5 to 10% called for in othcr procedures. As both the molybdatc and fluoride are additions to volume. they must be added from burets or pipets.
Table 1. influence of Acid Concentration on Color Formation
Acid ConcentraSamDle tion, % HZSOI Absorbanceo -b YBS 8h, 6.5 0.028 Si, Pro- 7 . 5 0.171 n,J..,.A A 8.0 0.171L ._ 8.5 0.169 9.5 0.159 SBS 72e, 10.0 0.205 0.2!12 Si, Pro- 15.0 0.210 cedure B 16.0 0.210 17.0 0,210 20.0 0.205 25.0 0.189 Absorbance values obtained on spectrophotometer at 620 mp with 23-ml. stmdard Fisher cc,lls. Precipitate formed on addition of molybdate. bGU