fractionated and the various cuts from the distillation were analyzed by gas chromatography. Large samples (100 pl.) of the cuts rich in the minor constituents were rechromatographed, the desired peaks collected in carbon disulfide traps, and their infrared spectra obtained. The gas chromatography and infrared identification are compared in Table VI. The peak numbers correspond to those shown in Figure 3. Agreement between the two methods was good for all of the constituents, with the infrared spectra revealing the presence of several additional trace components. Camphene (peak 4) was present in the starting material and was not a product of the reaction. Analysis of Wood Turpentine. Figuie 4 was obtained by chromatographing a sample of crude wood turpentine on didecyl phthalate. The area percentages and the identifications made on the basis of relative retention ratios are given in Table VII.
Infrared analysis of this sample showed the presence of 79% a-pinene, 14% of a compound containing an exocyclic terminal methylene group (calculated as camphene), and 4% tricyclene. The absence of the tricyclene peak in the chromatogram is probably due to the fact that it is obscured by the large a-pinene peak. The (Ypinene values by both methods show good agreement, and the 14y0 of a compound or compounds containing an exocyclic terminal methylene group is accounted for by the presence of 8.4% a-fenchene and 6.2% camphene. Three of the four additional minor constituents were also tentatively identified by gas chromatography.
LITERATURE CITED
(1) Bernhard, R. A., Food Research 23,
ACKNOWLEDGMENT
213 (1958). (2) Bernhard, R. A., J . Assoc. 0 8 c . Agr. Chemists 40, 915 (1957). (3) Groth, A. B., Svensk Papperstidn. 61, 311 (1958). (4) Johnson, H. W., Stross, F. H., ANAL. CHEM.30, 1586 (1958). (5) Liberti, A., Cartoni, G. P., “Gas Chromatography 1958,” D. H. Desty, ed., p. 321, Academic Press, New York, 1958. (6) Naves, Y. R., CGmpt. rend. 246, 1734 (1958). ( 7 ) Eaves, Y . R., -4drizi0, P., Favre, C., Bull. SOC. chim. France 1958, 566. (8) Stanley, R. G., hIirov, N. T., Division of Agricultural and Food Chemistry, Abstract 16, 133rd Meeting, .4CS, San Francisco, Calif., April 1958.
The authors acknowledge the technical assistance of W. A. Smith in obtaining some of the data, and of W. C. Kenyon for helpful discussion throughout this work.
RECEIVEDfor review April 27, 1959. Accepted April 21, 1960. Symposium on Terpene Chemistry, ACS Southeastern Meeting, Gainesville, Fla., December 12, 1958.
Cation Exchange Separation and Spectrophotometric Determination of Microgram Amounts of Rhodium in Uranium-Base Fissium Alloys J. 0.KARTTUNEN and H. B. EVANS Argonne National laboratory, lemont, 111. )Uranium is extracted with 30% tributyl phosphate in carbon tetrachloride from a nitric acid solution of the fissium alloy which contains uranium, molybdenum, ruthenium, palladium, and zirconium, and the raffinate i s strongly fumed with perchloric acid. The solution is then passed through a Dowex 50W-X8 cation exchange resin. Palladium is eluted with 0.3M hydrochloric acid, and the rhodium is then eluted with 6M hydrochloric acid. Rhodium is determined spectrophotometrically with tin(ll) chloride.
I
Experimental Breeding Reactor 11, the fuel plates contain zirconium, molybdenum, ruthenium, rhodium, niobium, and palladium in addition to uranium. This reactor is initially fueled with enriched uranium fissium, but later will operate on plutonium fissium. I n the pyrometallurgical processing of irradiated or spent fuel plates, studies have indicated that a certain group of the fission products cannot be removed and that these elements would bend to build up N THE
to an equilibrium value after several successive recycles of the fuel. The elements that tend to build up upon repeated pyrometallurgical processing and their equilibrium percentages are present>ed in the simulated 5 weight % fissium alloy (Table I). Fortunately, these elements are harmless and even beneficial because of the markedly increased radiation stability and improved metallurgical properties of the fuel plates. I n this may they also aid in achieving higher burn-up. They all have small capture cross sections for fast neutrons and hence do not interfere with the breeding properties of the reactor. To avoid dealing with a fuel of changing composition in the reactor, this fuel alloy (called fissium) is made t o contain originally the equilibrium concentration of these fissium elements (except technetium). The rhodium, once separated, could be determined spectrophotometrically in hydrochloric acid solution with tin(I1) chloride (1, 4). Uranium, molybdenum, palladium, and ruthenium interfered with the spectrophotometric procedure. Ruthenium could be easily
and completely evolved as the volatile tetraoxide, so this interference was easily resolved. Uranium is known to extract with 30% tributyl phosphate in carbon tetrachloride, so this source of error n-as also easily eliminated. Stevenson et al. (6) reported the separation of certain platinum group metals with cation exchange resins. This basic approach by Stevenson’s group was elaborated upon to develop a quantitative method for the separation of rhodium from palladium.
Table I. Calculated Equilibrium Fission Product Concentration (5) (5 weight % of new fissium metal is added
per cycle)
Element Zr Nb 110 Tc Ru Rh Pd U
Weight yo 0.10 0.01
3.42 0.99 2.63 0.47 0.30 92.08
VOL. 32, NO. 8, JULY 1960
917
11.
Table
Comparison of Recoveries
Rhodium
pg. Rh/A. Rhodium, pg. Variation, ~ n i t / 5 0 Taken Found % cc. Controls with rhodium alone through entire procedure -4.2 274 134.8 129.2 137.9 -0.4 264 138.4 263 133.0 0.0 133.0 -3.4 272 133.9 129.2 128.4 128.9 +0.4 262 121.2 119.8 -1.1 266 134.1 132.6 -1.1 266 271 134 5 -3 0 138 7 144.7 141 5 -2 3 269 137 9 132 7 -3 8 273 270 126 9 -2 7 130 5 131.8 128 2 -2 7 270 -2.7 270 125.8 122.3 138.2 135.1 -2 3 269 135.0 -1.5 267 137.1
Standard simulated fissium alloy 266 130.1 -1.1 131.6 0.0 263 127.9 127.9 119 5 115.9 -3 0 271 127.0 121.7 -4.2 274 118.9 119.3 1-0.4 262 108.4 105.1 -3 0 271 116.6 115.3 -1.1 266 136.6 131.4 -3 8 273 119.9 116 3 -3 0 271 -0 4 264 125.0 124.5 112 9 110 8 -1 9 268 126.0 -0.8 i65 127.0 127.6 128.1 +0.4 262 120.0 -2.3 269 122.8 134.4 131.9 -1.9 268
Rhodium was also separated from molybdenum by the resin. A chromatographic separation procedure used by Evans and Patterson ( 2 ) in the determination of rhodium in uranium fissium alloys gave good results, but was very time-consuming, especially nhen large numbers of samples had to be analyzed. K i t h the procedure outlined in this paper, as many as 20 resin columns were run simultaneously. Statistics on rhodium recovery are given in Table 11. EXPERIMENTAL
Apparatus. Absorbance measurements were made n-ith a Beckman Model D U quartz spectrophotometer, using matched 5.00-em. cells. T h e instrument was operated at constant sensitivity. T h e cation exchange resin columns contained 1.5 X 17 cm. of resin. Reagents. A solution of rhodium chloride was standardized by taking a weighed aliquot from t h e rhodium chloride solution, transferring it t o a Rose crucible, and taking t o dryness. T h e d r y salt was then roasted and t h e oxide reduced t o metallic rhodium in a stream of hydrogen gas. A simulated standard fissium solution was made from standard solutions of t h e other elements. All other chemicals n-ere analytical reagent grade. 918
0
ANALYTICAL CHEMISTRY
Tin(I1) chloride solution, prepared from the dihydrate, was lOy0 tin(I1) chloride in 10% hydrochloric acid. Dowex 5OW-X8 cation exchange resin, hydrogen form, styrene-type sulfonic acid, 50 to 100 mesh, with a n exchange capacity of 5.2 meq. per dry gram was used. Characteristics of Cation Exchange Resins. The degree of cross linking controls t h e swelling properties of exchangers. Resins containing a small degree of cross linking may s\vell t o many times their d r y volume if immersed in water or dilute electrolyte solutions, nhile those n i t h a large degree of cross linking usually change little in volume. The degree of si\ elling also varies vrith the electrolyte concentration of the surrounding medium; it is largest in n a t e r and decreases with increasing osmotic pressure of the solution. The excessive swelling of low cross-linked resins introduces considerable complexity into column operations, as the resin beds tend to pack n ith change of electrolyte concentration. I n this work, Doweu ;iOW-X8 n i t h its 870 cross linkage gal-e good results and higher cross linkage is not necessary. The more highly cross-linked resins tend to reach exchange equilibrium more slonly than those of lower cross linking. This disadvantage can be compensated, to some extent, by use of repins of small mesh size. A 50- to 100niesh size mas found to be ideal and also gravity flow of the solution through the column was a t an optimum rate of approximately 8 to 10 ml. per minute. Furthermore, the generally higher selectivity of more highly cross-linked resins also often permits use of shorter columns (ferer theoretical plates) than is possible n i t h low cross-linked exchangers. Too high a degree of cross linkage can lead to exaggerated specificity n i t h resultant elution problems. Proper selection of mesh size and cross linkage will keep these problems to a minimum. Treatment of Resin. Comniercial resins almost always need some purification or pretreatment before they can be used in analytical applications. Undesirable ionic impurities may be removed by treatment n i t h excess of t h e proper electrolyte solutionsfor example, hydrochloric acid if t h e chloride form of t h e exchanger is desired. Often treatment with both concentrated and dilute electrolyte solutions is necessary because certain metallic impurities may remain strongly absorbed if only one reagent is used. The Dowex 50W-X8 cation exchange resin should be stored in concentrated or 6 M hydrochloric acid when not in use. It can be left in the glass column and prior t o use the resin should be washed with concentrated hydrochloric acid, distilled water, 6M hydrochloric
Mg. u R A N I U M / 5 0 m ~
Figure 1 . Uranium interference in rhodium(Ill)-tin(ll) chloride complex
9 C '
'
'
'
3
'
' '
I
' i C 34
P E R CENT
Figure 2. Dependence of rhodium adsorption on acid concentration
i
/
, , / , I ,
85
C
50
/
,
100
V,I IlItelS
Of
6 V
I
I
50
200
*c
Figure 3. Dependence of rhodium recovery on elutrient volume
acid, distilled IT-ater, 3 X hydrochloric acid, and finally again n i t h distilled water until all trace of acid is gone as indicated by p H paper or litmus paper. Dissolution of Fissium Alloy. The alloy is cautiously treated n i t h a solution of aqua regia t h a t is 0.3M in hydrofluoric acid and boiled gently until all of t h e alloy is in solution. Extraction of Uranium. Uranium is retained appreciably on t h e cation exchange resin, and even after the primary 0.3M hydrochloric acid wash, causes significant interference. Because of the additional absorbances caused by uranium as shown in Figure 1, i t was n e c e s m y t o extract t h e uranium nith 3Oy0 tributyl phosphate in carbon tetrachloride. One extraction is sufficient. .A negligible amount of rhodium is extracted with the uranium. A recent report by Gardner and Hues (3) determines rhodium in uraniunirhodium alloys nithout separation by measuring the absorbance of the red chloro complex of rhodium at 520 nib. There is some loss of sensitivity in measuring a t 520 mp, but precision w-as
good when the uranium was treated according to this procedure. Effect of Perchloric Acid Concentration. T h e perchloric acid concentration was tested from 1 t o 10% as shorn in Figure 2. Although t h e variations in recovery are slight, approximately 5 t o 7% solutions a r e Iecommended for t h e best precision. Optimum Eluting Volume. As +own in Figure 3, t h e minimum volume of 6 M hydrochloric acid for eluting was approximately 150 ml., b u t because t h e flow rate of t h e elutrient through t h e column is relatively fast, 175 t o 200 ml. are recommended. Heating t h e 631 acid t o 70" C. does not lessen t h e amount of elutrient needed, a n d therefore serves n o purpose. Adherence of Other Ions to Resin Column. Palladium adheres t o t h e cation exchange resin also, b u t i t was quantitatively eluted with 7 5 ml. of 0 . 3 X hydrochloric acid. This acid concentration should n o t be above 0.5M nor less t h a n 0.2M. If i t is greater t h a n 0.5M, t h e possibility of eluting rhodium becomes greater. Uranium also adheres t o t h e column, and it is not completely eluted with the 0.3M acid, so the extraction mentioned previously is necessary. Spectrochemical analysis and a series of experiments showed that molybdenum liken-ise adheres, but i t is nearly completely removed with the 0.3X acid and does not interfere if the molybdeiiuni content per aliquot is less than 2.3 mg. Kormally the amount of molybdenum per aliquot would be approximately 0.95 mg., assuming that an aliquot containing 0.130 nig. of rhodium n as chosen. Therefore the molybdenum content in the alloy could easily be doubled from 3.42 to 6.84Yc without interference. When the molybdenum content is more than 20 times that of rhodiiim, small errors arise. Some zirconium adheres, but causes no interference.
I:.
0.4
0.2 0.0 400
425
450
500
550
600 650
W I V E LENGTH ( m ~ l l ~ m ~ C r O o l l
Figure 4. Absorption spectra of rhodium(ll1) chloride complex
Spectral Characteristics. Figure 4 shows t h e spectral curves for t h e rhodium(II1) chloride complex developed Tvith tin(1I) chloride b y t h e
standardized procedure described below. The molar absorptivity for the complex used in this procedure is 3840, which is equivalent t o 1.3 fig. of rhodium per ml. to give about 0.25 absorption unit in 5-em. cells. Rhodium Chloro Complex. The spectrophotometric determination of rhodium is based upon t h e pink t o red color with a n absorbance maximum a t 475 mp, formed by the addition of tin(I1) chloride in hydrochloric acid solution. At room temperature, about 12 hours are required for complete color development; a t temperatures near the boiling point, the full color intensity is developed in 3 to 5 minutes ( 1 ) . The amount of hydrochloric acid can be varied from 0.5 to 20 ml. of concentrated acid per 50 ml. of final solution. Absorbance readings were constant within the limit of accuracy of making the measurement. This was also noted by hyres, Tuffly, and Forrester ( I ) . Standardized Procedure. A sample of fissium alloy is accurately weighed, transferred t o a beaker, and cautiously treated with a solution of aqua regia t h a t is 0.3M in hydrofluoric acid. K h e n t h e reaction has subsided, boil gently, until all of t h e alloy is in solution, a n d make u p t o a suitable volume. An aliquot containing 100 to 160 pg. of rhodium is placed in a 30nil. beaker, approximately 1 ml. of perchloric acid is added, and the solution is evaporated to heavy fumes of perchloric acid. Do not take to dryness! After cooling, the sides of the beaker are nashed down with distilled water and the solution is again taken to fumes. Fluoride, chloride, and ruthenium are expelled during this fuming. -4 high chloride or fluoride content will lessen the efficiency of the uranium extraction. Approximately 10 ml. of 1 to 1 nitric acid are added to the solution and the solution is heated to a gentle boil and then alloned to cool. The solution is n o x transferred to a 60-ml. separatory funnel. The beaker is n-ashed n ith approximately 3 nil. of 1 to 1 nitric acid, the washings are added to the solution in the funnel, and a n equal volume of 307, tributyl phosphate in carbon tetrachloride is added. One extraction is made and the aqueous solution is washed once with an equal volume of carbon tetrachloride to reniore droplets of the phosphate. The raffinate containing the rhodium is returned t o the original beaker, to which are added approximately 3 ml. of concentrated perchloric acid. The separatory funnel is nashed twice with 3-nd portions of distilled water and the washings are added to the raffinate. The solution is taken to dense fumes of perchloric acid, cooled, and the sides are washed don-n with distilled n-ater. The fuming process is repeated two to three times to remove all nitrate, chloride, fluoride, and ruthenium. This
fuming is important, and, if insufficient., is one of the greatest sources of error. Rhodium mill form a complex with chloride, fluoride, and nitrate, causing incomplete retention of the rhodium on the resin. The perchloric acid is finally fumed t o a volume of approximately 0.5 t o 0.7 ml. and allowed to cool. Do not let the acid fume to dryness, or spurious results \vi11 arise. The volume of acid after fuming is tested with a 500-p1. pipet. The volume of perchloric acid is increased until there are definitely more than 500 p1. of acid if the acid has fumed too long. The acid volume is not' critical (Figure 2 ) , but a volume of 500 t o 700 pl. is recommended. The pipet is washed three times d h distilled wat'er. -4pproximately 9.5 ml. of distilled water are added t o the perchloric acid and the solution is heated to a gentle boil. This solut'ion is allowed to cool and is then transferred wit'h a transfer pipet to the activated and cleaned cation exchange resin column. The solution is drained off at the rate of approximately 3 nil. per minute. The beaker is washed with freshly made 57, perchloric acid, and this wash is also pipetted onto the column and allowed to drain a t t'he same rate. Then 75 ml. of 0.3M hydrochloric acid are added and allon-ed to run through the column a t the rate of approximately 8 to 10 ml. per minute (ordinary gravity flow) Elut'rient,sare discarded. Now 175 to 200 ml. of 6 M hydrochloric acid are run through the column a t the same rate and collected in a clean 250-ml. beaker. The solution is evaporated to appro% metely 50 ml. and then transferred to a 150-ml. beaker. The 250-ml. beaker is nashed twice with distilled v-ater and the washings are transferred to the 150-ml. beaker also. Approximately 5 ml. of nitric acid, 1.5 ml. of sulfuric acid, and 1 ml. of perchloric: acid are added to the solution and evaporated to sulfur trioxide fumes. The nitric and perchloric acids are used to destroy any small amounts of organic material eluted from the resin. The solution is allon-ed to cool, The sides of the container are n-ashed down n-ith distilled water and the sulfuric acid solution is then evaporated to a ~ o l u m eof -0.25 to 0.50 ml. Then 5 nil. of hydrochloric acid and 5 ml. of water are added to the solution and heated gently. To the warm solution 10 ml. of 10% stannous chloride in 10% hydrochloric acid are added and heated t o boiling. The solution is placed on a steam bath for 1 hour. The solution is cooled and t,ransferred to a 30-ml. volumetric flask and made up to volume with 2 X hydrochloric acid. The absorbance is read at 475 mp, using a reagent blank for comparison, The measurement is made in a 3.00-em. cell with the sensitivity setting a t 3. RESULTS AND DISCUSSION
Effect of Diverse Ions on Ion Exchange. Platinum, rhodium, paladium, and iridium are retained on t h e VOL. 32, NO. 8, JULY 1960
919
cation exchange resin. These elements may be eluted separately as described by Stevenson et al. (6). Sulfate, nitrate, fluoride, and chloride form complexes with rhodium of a type which are not quantitatively retained on the resin. Nitrate, fluoride, and chloride are fumed off in the perchloric acid treatment phase and, therefor?, are easily eliminated. If sulfate is present, the removal of the sulfate by either fuming to dryness or by funiing t o dryness and then roasting with a n open flame gives nonquantitative results. The removal of sulfate by precipitation with barium chloride gave poor results. It is unnecessary to use sulfuric acid in the dissolution of the fissium alloy. Effect of Diverse Ions on Rhodium (111) Chloride Complex. T h e effect
of diverse ions has been tabulated
pg.
Bromide and sulfate ions have little efiect on the rhodium(II1) chloride complex while the nitrate ion tolerance is 200 p.p.m., and the iodide ion tolerance is 75 p.p.m. The tolerances in parts per million for some of the more common ions are as follows: platinum(IV), 1; ruthenium(III), 1; palladium(II), 2; gold(III), 5 ; chromium(VI), 5 ; iron(III), 50; cobalt(II), 50; and copper(II), 50. The molybdenum tolerance from these experiments was 5 p.p.m., and the uranium tolerance mas 12 pap.m. Precision a n d Accuracy. Using a standard simulated fissium solution, the standard deviation for 15 analyses \vas 1.5%, with a 95% confidence limit of 3.0%. The reciprocal of t h e slope of the standard curve is 268
50 ml.
(1).
of rhodium per absorbance unit per LITERATURE CITED
(1) Ayres, G. H., Tuffly, B. L., Forrester, J. S., ANAL.CHEM.27,1742 (1955). (2) Evans, H. B., Patterson, J. H., Re-
actor Fuel Measurement Technique8
Symposium, TID-7560,189 (June 1958). (3) Gardner, R. D., Hues, A. D., ANAL. CHEM.31 ,‘1488 (1959). . (4) Ivanov. V. N.. J. Russ. Phus. Chem. ‘ &oc. 50,460 (1913). (5) Schraidt, J. H., Rodger, W. A., Levenson, bI., Symposium on the Reprocessing of Irradiated Fuels, TID-7534, Book 2, 748 (May 1957). (6) Stevenson, P. C., Franke, A. A., Borg, R., Nervik, W., J . $7n. Chem. SOC. 75,4876 (1953). RECEIVEDfor review October 26, 1959. Accepted April 21, 1960. Ba.sed on work performed under the auspices of the United States Atomic Energy Commission.
Quantitative Radiochemical Analysis by Ion Exchange Uranium and Tellurium LEON WISH U. S. Naval Radiological Defense laboratory, Sun Francisco 24, Calif.
,Tellurium radionuclides from fission product mixtures are eluted with uranium from Dowex 2 anion resin in the quantitative radiochemical procedure previously reported. The separation of these two elements from each other i s necessary for their determination. The adsorption of tellurium(1V) from phosphoric acid solutions was investigated. From similar data on uranium(V1) a procedure for the separation of tellurium from uranium was obtained. Column runs indicated that the yields were quantitative and the purities greater than 99%. The procedure was then expanded to permit inclusion in the sequential radiochemical analytical scheme for mixed fission product and actinide elements.
T
ELLURIUM and uranium radionuclides are eluted together from fission product mixtures in the quantitative ion exchange separation scheme employing Dowex 2 resin, as previously reported (4). These two elements must be separated for quantitative determinations, although it is possible in some cases to obtain a satisfactory analysis of a combined tellurium-uranium sample with a y-ray spectrometric
920
ANALYTICAL CHEMISTRY
method developed by Lee (3). The spectrum of the mixture is obtained with a multichannel analyzer and, by complementing, the pure spectrum of one of the isotopes can be subtracted until only that of the other remains. Uranium-237 could be determined with a standard deviation of 1 to 2% when the samples contained at least 10% of the isotope. However, when the uranium237 content was less, the errors increased so t h a t with 2.5% uranium-237 the standard deviation was 5%. A pure sample of the isotope to be subtracted is required and the end point which is determined visually must be estimated somewhat subjectively. An eluting agent which would be specific for either tellurium or uranium is most desirable so that no further treatment of the eluted fractions would be necessary. Equilibrium distribution coefficient data between Dowex 2 anion resin and hydrochloric, nitric, sulfuric, and hydrochloric-hydrofluoric acid mixtures (1, 4 ) have shown that tellurium(1V) and uranium(P1) will not separate satisfactorily in these acids. More recently Freiling, Pascual, and Delucchi ( 2 ) have found that when a fission product mixture in 0.1N phosphoric acid is loaded on a Dowex 2 phosphate resin column, a large frac-
tion of the activity in the loading eluate is tellurium. This column run indicates that tellurium may be separated from uranium and neptunium using phosphoric acid solutions. Therefore, an investigation has been macle to determine the optimum conditions for the anion exchange separation of tellurium from uranium in phosphoric acid, with the additional objective of its inclusion in the sequential radiocheniical analytical procedure EXPERIMENTAL
The distribution coefficients (Kd’S) of tellurium(1V) in phosphoric acid solutions were obtained by the batch process of shaking the resin with the solution of the radionuclide for 16 t o 20 hours ( 2 ) . The K d is the ratio of the tellurium activity per gram of resin t o the tellurium activity per milliliter of solution at equilibrium. The exchange column separation experiments were performed on a resin column 0.2 cm. in diameter and 15 cm. long, which has been described in detail (4). The flow rate was about 1 ml. per 4 minutes. All of the radionuclides were measured in a y-ray scintillation counter with a 3 X 3 inch sodium iodide-thallium crystal with a we11 inches in diameter and 2 inches
