Determination of Small Concent rat ions of Indium by Radioactivation J. E. HUDGENS AND L. C. NELSON U . S. .4tomic Energy Commission, New Brunswick, N. J . This work was undertaken to develop a radiochemical procedure for the determination of microgram or submicrogram quantities of indium in materials that might be used in nuclear reactors. A method was developed in which neutron activation with a nuclear reactor and with a radium beryllium source was used to determine indium concentrations of a few parts per million. The ultimate sensitivity of the method is lo-” gram when the thermal neutron flux of 4 X 1012 neutrons per second per square centimeter obtainable in a reactor is used. A rapid solvent extraction method for separating the indium from other sample components was developed. The method is applicable to the determination of indium in a w-ide variety of sample types.
I
A previous paper the use of radioactivation as a method of analysis was reviewed and the method was applied in the determination of small concentrations of antimony ( I S ) . I n the present work the generalized technique has been extended to the determination of very small concentrations of indium. Isotopes of mass 113 and 115 are found in naturally occurring indium in abundances 4.16 and 95.84%, respectively (IO). Indium 113 is naturally radioactive and has a half-life of approximately 1014 years ($55). The calculated specific activity for indium 113 is approximately 60 disintegrations per minutts per gram and the radioactivity contributed by this isotope is negligible in comparison with the activities which can be produced within a short time by pile activation or with a radiumberyllium neutron source. Seutron irradiation of indium 113 produces 50-day indium 114 and of indium 115 produces 54minute indium 116. The slow neutron cross section for the production of indium 114 is 5.6 X sq. em. ( 2 0 ) . An average of 1.08 gamma rays and 0.97 beta rays are released per disintegration ( 2 ) . The total counting rate from indium 114 is approximately twice the beta ray counting rate of the 50-day isomer because the 0.192 m.e.v. gamma ray is nearly completely converted to photoelectrons in the K and L electron shells ( 2 ) . Indium 115 has a cross section of 1.45 X lo-** sq. cm. for the production of 54-minute indium 116 ( 2 0 ) .
Moellc~~ (18)and Lacrolx (16) extracted the Shydroxyquinolate of indium in the pH range of 3.2 to 4.5 by shaking an aqueous iolution of the indium with a chloroform solution of 8-quinolinol. HOT? ever, interference by gallium, aluminum, and thallium as well as by tin, bismuth, iron, cobalt, sodium, and copper prevented the use of this separation. Knox and Spinks (16) extracted the chloride of indium with various solvents including ethyl ether but the distribution ratios were low and the concentrations of indium were far above those contemplated for this de termination. Wada and Ishii (22, 23) and Vanossi (21) investigated the extiaction of milligram quantities of the bromides of many elementh including indium, gallium, and thallium, from solutions rontaining various concentrations of hydrobromic acid using ethyl ether. Thallium(II1) and gold(II1) are quantitatively rutracted from 1 M hydrobromic acid solution and negligible
CHEMICAL METHOD
Although the necessity for performing a chemical separation of indium from other sample constituents can be obviated for certain types of samples by combining decay curve and energy measurements, the accuracy of the determination is often limited by the accuracy with which the different components of the curves can be isolated and a simplei procedure is to isolate the desired element cheniically before making the radioactivity determination. Glendenin ( 7 ) separated indium from a mixture of the fission products by precipitation of the sulfide. Dennis and Bridgman ( 3 ) separated indium from gallium by precipitating the hydroxide in the presence of ammonium nitrate. Each of these compounds is a suitable weighing form, but the usefulness of precipitationtype separation methods is often limited by coprecipitation and absorption phenomena (12). Jacobi ( 1 4 ) separated carrier-free indium from a cadmium target by using a hydroxide precipitation and magnesium as a nonisotopic carrier. Because of the short half-life of indium 116 (54 minutes), a rapid and specific method of separation was desired and for this reason two extraction procedures for the separation of indium from the usual sample constituents and particularly from aluminum, gallium, and thallium were considered.
/ 3
Figure 1.
1472
4
5
MOLESILITER HB:
7
8
Effect of Hydrobromic Acid Concentration on Extraction of Indium
V O L U M E 24, NO. 9, S E P T E M B E R 1 9 5 2
1413
amounts of indium are extracted a t this acid concentration. At a hydrobromic acid concentration of 4.5 to 6.5 M , indium is extracted along with some gallium( 111), iron( 111), antimony(Ir),zinc, tellurium, molybdenum, rhenium, and iridium (IS). Ato ( 1 ) reported that indium can be recovered from the ether phase by equilibrating with a 4 to 7 M solution of hydrochloric, acid. Under these conditions such element's as iron, antimony, gallium, thallium, and tellurium remain nearly quantitatively in the ether phase ( 4 , 8 , 8, $3). Of the elements which might be extracted with indium from a hydrobromic acid solution by et,her and re-extracted from the ether into a hydrochloric acid solution, only zinc and tellurium have half-lives that are ~uffiviently similar to that of indium to interfere seriously with the radioactivity determination (5,9, 11, 17,19). Extraction of Indium. The use of isopropyl ether for thr, extraction of indium has not been reported previously, although it is known t h a t several elements are efficiently ext,racted R-ith isopropyl ether and that it has seveml advantages over ethyl cther
to the oxide and weighing. shown in Table I.
The data for these experiments are
PROCEDURE
The irradiated sample was dissolved in about 10 ml. of concentrated hydrobromic acid, 10 mg. of indium as a solution of the bromide and several milligrams of iron as the nitrate were added, and the solution was evaporated just to dryness. The residue was then taken u in 1 hi' hydrobromic acid, washed into a separatory funnel, anxmade up to about 8 ml. with l M hydrobromic acid. Two successive 30-ml. portions of is0 ropy1 ether were equilibrated with the acid phape and discard)ed. Enough concentrated hydrobromic acid (48'%) was added to make the hydrobromic acid concentration 4.5 .If. The total volume was about 20 ml. A fresh 30-ml. portion of isopropyl ether was added and *haken with the aqueous solution for a minute, and the acid phase was separated. This step was repeated and after the second equilibration with ether the acid phase was discarded. The two ether phases were combined and n-ashed with three succefisive 5-ml. portions of 4.5 111 hydrobromic acid. The washings
(41. The authors have determined the distribution ratio (Co/Lya, where C, refers to t,he concentrat'ion in the organic phase and C, to t h a t in the aqueous phase) for indium as a function of the hydrobromic acid concentration of the aqueous phase before equilibration wit,h isopropyl ether. The isopropyl ether (Eastm:m Kodak &. alcohol-free reagent) used in these experiments :tnd t,hose cited below was purified by refluxing with sodium m f h l and by distillation. Catechol (.io mg. per liter of ether) was added to keep the ether peroxide free. The data shown in Figure 1 ivere obtained by equilibrating the aqueous hydrobromic acid .solution of the indium with an equal volume of ether. The poiiits were corrected for the small solubility of the ether in the aqueou!: phases. The data (uncorrected for solubilit'y of the ether in the aqueous phases) obtained by JVada and Ishii are also plotted in the figure to facilitate conipnriPon of the two ethers as extract:inti; for indium. The ext,raction of the iridium is nearly complrte over a range of hydrobromic acid concentrations and, thercfore, the adjustment of the wid concentration is not critical. Preliminary experiment L; were performed using indium 114 its :I tracer to determine the dependenc~of the extraction of indium by isopropyl ether 011 the indium concentration. At (,onstant hydrobromic acid (.oncentration (4.5 M ) the quaiitit!. of iiidium extracted is only slightly dependent on the indium concentration. At the lowest indium concentration studied (0.9 mi(-rogramof indium per nil.), 81% of the indium ~ v a 5extracted by an equal volume of isopropyl ether. The extent of coseparation of zinc, iron, or tellurium with the indium was determined by making an aqueous solution containing either zinc, tellurium, or iron tracer activities along rvith bveighable amounts of indium, 4.5 JI in hydrobromic acid, and extrxting the indium into isopropyl cther as described in the procedure belon-. The washing step and the step involving the estraction of the indium from isopropyl ether into aqueous hydrochloric acid solution \\-ere also included. The quantity of zinr, tellurium, or iron which W M coseparated with the indium WLS determined by measuring the radioactivity with an endwindow Geiger-~Iullertube. The indium yield through the (lheniical separations \vas determined by converting the hydroxide
Table 1. Degree of Coseparation of F'arious Elements with Indium Active Elrmentb _ Added _ ~ _ _ Activity, Element counts/min. ~
.
Zn
5
x
106
TP
3
x
105
I'r
14.756
_
Inactive Elements Added Element Mg. Zn In Ill I;? In
5 9.1
n: 1
.50
$1 1
c /o
Extracted 0.001 98 0.02 48 0.02 48
.92C M . @ q
SAMPLE
'2h MCDERATOR
SOURCE
Figure 2. Paraffin Geometry for Radium-Beryllium Neutron Source
n-ere discarded. The ether phase was then equilibrated with three successive 5-ml. portions of 5 rlf hydrochloric acid. T h r aqueous phases were combined and washed with another fresh 15-ml. portion of isopropyl ether. The hydrochloric acid solution was then diluted to 50 ml., 5 grams of ammonium nitrate added, and the solution just, neutralized n-ith ammonium hydroxide. The mixture was centrifuged, and the precipitate was washed with water and quantitatively transferred to a 1-inch glass dish. The precipitate was dried under an infrared lamp M o r e the radioactivity determination was made n-it,h an endwindo%- Geiger-1Iiiller counter and a commercial scaler. The indium precipitate mas then quantitatively transferred to a crucible, ignited a t 800" to 850" C., and weighed as In&. A correction for chemical yield through the procedure was applied where the losses were significant. Total time required to c-omplete thr chemical procedure WRP usually a little more than 1 hour. NEUTRON IRR-IUI \TIOY O F THE S A M P L E
Keutrons sources of two types may be used to activate iiidiuni. A source composed of a radium salt mised with beryllium metal powder can he used conveniently for the determination of indium in concentrations of a few parts per million or more. The nuclear reactor is a very intense source of neutrons and has made possible the determination of very small concentrations of most of tht. elcniente. The quantity of indiumwhichcan bedetermined using a radiumberylliuni neutron source is a direct. function of the quantity of radium incorporated in the source, and the efficiency a-ith which the nc,utroiis van be used. By using 0.254gram of radium mixed v-ith npprosimately 0.75 grani of lm-yllium in the arrangement shown in Figure 2, 10-5 part of iiidium per part of sample can be determined by using a IO-gram sample. By improving the geometrical :trrangement, decreasing the neutron loss due to causes other than capture by the sample, and increasing the amount of
ANALYTICAL CHEMISTRY
1494
radium (and beryllium) in the source, even smaller concentrations of indium can be detected. The cross section of indium varies with the neutron energy and the amount of radioactive indium produced by irradiation of a sample with the neutrons from a radium-beryllium source is therefore also a function of the number of collisions between the neutrons and moderating material before they are captured by the sample material.
The d a b in Figure 3 were obtained by placing different thicknesses of paraffin between the source and a piece of indium foil. The foil was placed directly on a disk of paraffin, and thus the distance from the neutron source to the foil increased with increasing thicknesses of the paraffin disk. The foil was irradiated for 3 hours and the decay of 54-minute indium 116 observed as a function of time after removal from the neutron flux. The counting rate due to long-lived isotopes was subtracted from the curve, and the activity of 54-minute indium 116 obtained was plotted in the figure. I t is evident that the increase in the solid angle subtended by the foil as it was moved closer to the source was sufficient to compensate for any decrease in the capture cross section of indium due to the slight increase in average neutron energy. In subsequent irradiations nrith this source a paraffin disk thickness of 9.2 mm. was used. The samples used to test the procedure were repared for irradiation by the radium-beryllium source by pelreting 10 grams of aluminum sulfate containing known amounts of indium in a 1inch diameter split-type pellet mold with a hydraulic press. The samples were irradiated for 3 hours or more. After this time the rate of decay of indium 116 is approximately equal to its rate of formation, and any increase in the amount of indium activity is due to the formation of indium 114. h thallium salt could also have been used in similar experiments as a matrix. Because a nearly quantitative separation of indium from thallium is obtained in each of the extraction steps of the procedure, demonstration of the absence of interference by thallium was unnecessary (62, 23). The samples which were irradiated in the nuclear reactor were sealed in 3-cm. lengths of 1-mm. inside diameter quartz tubing. Sample weights of 19 and 22 mg. were used. The quartz tubes were placed in a suitable hole drilled in a graphite cylinder as shown in Figure 4. An aluminum cap and screw held the quartz tubes in place. All four samples plus the monitors for the neutron flux were thus irradiated under conditions of equal neutron intensity. RESULTS
The samples used to test the method were prepared by adding known amounts of indium, as the oxide, to pure aluminum oxide or aluminum sulfate powder. The aluminum oxide samples shown in Figure 5 were given a 9-day irradiation a t the Brookhaven National Laboratory pile. Although the indium was isolated from small samples (-20 mg.), the counting rates were sufficiently high (10 to 50 counts per
second) to give good accuracy. The counting rate could have been increased by using larger sample sizes or by irradiating for a longer time. Alternatively, samples containing smaller amount? of indium could be analyzed. The amount of indium detectable with the conditions used in this experiment is 3 X lo-* gram. If one count per second is assumed to be the minimum determinable counting rate, and the samples are irradiated a t the maximum Brookhaven National Laboratory flux of 4 X 10l2 neutrons per second per sq: cm. ( 2 4 ) for a time sufficiently long to achieve a saturation radioactivity of the mass 114 or 116 isotope, and the counting rate determination is made before more than 75% of the radioactivity present a t saturation has decayed away, then lo-" gram quantitier; of indium can be determined. The proximity of the individual points to the straight line shown in Figure 5 illustrates the accuracy of the method, and the added indium serves as an internal standard. Each point on the curve represents a single chemical separation and counting rate determination, and the precision which can be easily attained is thus indicated. The results shovm in the figure indicate that 15 p.p.m. of indium was present in the original aluminum oxide. Quantitative recovery of the indium from the sample was not necessary because a correction was applied for any losses incurred in the performance of the chemical separations. The energy of the radiations from each of the isolated indium oxide residues was determined by an absorpHOLE FOR tion curve. The curves 10-32 NF BOLT for each of the samples were i d e n t i c a l t o t h e -03 DIA. wrves of the monitor. The aluminum sulfate A-A samples listed in Table I1 xere given a 3 to 4 hour irradiation with t,he radium-beryllium source (Figure 2). The samples weighed about 10 gram6 and the counting rate6 due to the 54-minut,e indium 116 were from 1 to 8 counts per second, 2 hours after cessation of the neutron irradiation. The amount, of indium in each of the samples shown in Table I1 was calculated by using the average of the BODY specific activities (in Figure 4. Irradiation Capsule counts per minute per Dimensions in om. microgram) calculflted from t h e observed counting rate and the weight of indium added. The sensitivity of this method could have been increased by using a larger neu-
@
Table 11. Determination of Indium i n Aluminum Sulfate by Radioactivation with Radium-Beryllium Neutron Source Sample
8 9
10 11
12 13
Indium Added,
P.P.M. 507 741 102 102 102 248
Indium Found
P.P.M: 501 752
101 102 100 253
Error, % -1.2 +1.5 -0.98 0 -2.0 f2.0
Av.
1.28
V O L U M E 2 4 , NO. 9, S E P T E M B E R 1 9 5 2 7
bx
1475
1
3.
z
---
0
4
1
T - 3 0 - 7 P P M ADDED
50
60
70
EO
INDIUM
Figure 5. Indium Activity Formed as a Function of Inacti\e Indium 4dded to .4luminuni Oxide
tron ciource and under these conditions indium concentrations of 10 p.p.m. or less could have been detected. The results given in Table I1 indicate that the amount of indium in the original aluminum sulfate samples was prohal)ly less than a few parts per million, since this is approsimate1~equivalent to the accuracy of the determination of the indium concentration in samples 10, 11, and 12. The activity of t h e 54-minute indium 116 was determined by plotting the radioactivity as a function of t,ime. Because of the short half-life, this technique served not only to increase the counb ing statistics but also iis a check on the efficiency of the chemical separations. T h t measured half-life of the chemically isolated indium samples varied from 52 to 58 minutes and the purity of the indium was t.hur sh0u.n to he good. LITERATURE CITED
( I ) Ato, S.,Z;ci. P a p e r s I n s t . Phys. Chem. Research (Tokyo), 24, 162
(3) Dennis, L. S . ,and Bridgman. J. .A,, J . A m . C h m . Sou.. 40, 1552 (1918). (4) Dodson, R. W., Forney, G. T., and Swift, E. H.. Ibid.. 5 8 , 2573 (1936). (5) Edwards, J. E., and Pool, M . L., Phys. Rev., 69, 140 (1946). (6) Edwards, F. C., and Voigt, A. F., - 4 ~ 4CHEM., ~. 21, 1204 (1949). (7) Glendenin, L. E., “Radiochemical Studies. The Fission Products,” C. D. Coryell and K.Sugarman, ed., NNES Div. IV, Val. 9, p. 1575, Kew York, ,\IcGran~-HillBook Co., 1951. (8) Grahame, D. C., and Seaborg, G. T., J . Am. Chem. Soc., 60, 2524 (1938). (9) Hanson, A. O., Duffield, R. B., Knight, J. D., Diver), B. C., and Palevsky, H., Phys. Rev., 76, 578 (1949). (10) Hibbs, R. F., U. S. Atomic Energy Commission, Repl. AECU556 (1949). (11) Hill, R. D., Scharff-Goldhaber, G.. and Friedlander, G., Phljs. Rec., 75, 324 (1949). (12) Hope, H. B., Ross, M., and Skelly, J. F., IND.ENG.CHEM., AKAL. ED.,8, 51 (1936). 113) Hudgene, J. E., Jr., and Cali, P. J.. ANAL.CHEM.,24. 171 (1952). (14) Jacobi, E., Helv. Phys. Acta, 22, 66 (1949) (15) Knox. K. L., and Spinks, J. W.T., Can. Chem. Process Inds.. 30, No. 11, 85 (1946). (16) Lacroix, S., Anal. Chim. Acta, 1, 260 (1947). (17) Livingood. J. J., and Seaborg, G. T., Phys. Rea., 5 5 , 457 (1939). (18) Moeller, T., IND. ENG.CHEM.,ANAL.ED., 1 5 , 270 (1943). (19) Seaborg, C. T., Livingood, J. J., and Kennedy, J. W., Phys. Rev..57, 363 (1940). (20) Seren, L., Friedlander, H. N., and Turkel, S.,Ibid., 72, 888 (1947). (21) Vanossi, R., Anales (LSOC. Q U ~ Argentina, . 38, 363 (1950). (22) Wada, I., and Ishii, R., Sci. Paper Inst. Phys. C h m . Research (Tokyo),24, 135 (1934). (23) Ibid., 34, 787 (1938). (24) U. S. Atomic Energy Commission, “Catalog and Price List No. 4, Isotopes Division,” 1951. (25) Way, K., Fano, L., Scott, hl. R., and Thew, K., Katl. Bur. Standards, Circ. 499 (1950). RECEIVED for revieir March 6, 1952. Accepted July 3, 1952. Presented at the Pittsburgh Conference on Analytical Chemistry and Applied Speotroscopy, Pittshiirgh. Pa., March 5 t o 7, 1952.
(1934). ,‘2) Boehm. F.,and Preiswwk, P..H e h . P h y s . Acta, 22, 331 (1949).
Determination of Macro and Micro Quantities of 3-(p=Chlorophenyl)-l,l -dimet hylurea 5..K. LOFEN
AND
H. AI. BAKER
Grasselli Chemicals Department, E. I . d u Pont de .Yemours & Co., Inc., It’ilmington, Del.
T
HE herbicidal properties of the new chemical compound
mately 247 mp, was employed in the earlier assays of CMU weed 3-(p-chlorophenyl)-l, I-dimethylurea, which is the active killer However, the application of this technique to relatively ingredient of 80% CMU weed killer, have been described by pure materials gave a standard deviation of 1.67%, and the search Bucha and Todd ( 2 ) . The development of CMU weed killer for a more precise method led to a study of hydrolytic reactions aa a soil sterilant and evaluations of this material for pre-emeras the basis for quantitative analysis. The 3-(p-chlorophenyl)gence treatment of croplands have required methods for deter1,l-dimeth) lurea may be hydrolyzed to yield pchloroaniline, mining macro and micro amounts of the active ingredient. Condimethvlaminc, and rarhon dioui le according to the follouing qideration of various possibilities for analyzing 3-(p-chlorophenyI)reartions: 1,l-dimethylurea included a study of 2H + C1 /’ “-SH (CH3hNH + reactions to obtain hydrolysis prodCOP ucts which might be readily deterrH mined. Both acid and basic condiH O tions have been established for comN(CH3h HzO plete hydrolysis of this compound, OHand the reaction products have served L+C’U--NH2 (CH,),NH HCOaas a basis for the development of precise analytical techniques. The original compound may be assayed by determining the MACROMETHODS dimethylamine liberated under conditions which have been The ultraviolet absorption spectrum of 3-(p-chlorophenyl)established for obtaining complete hydrolysis. 1,l-dimethylurea, using the major absorption band of approxiQuantitative hydrolysis mag he achieved by refluxing thp mate-
+
clo-N-&+ 1 7 -
+
+
[
+
