Determination of Thallium by Radioactivation Application to Potassium Iodide C. J. DELBECQ, L. E. GLENDENIN, AND P. H. YUSTER Chemistry Department, Argonne Wational Laboratory, Chicago, Ill. CONNECTION with an investigation ( 7 ) of the optical I properties of single crystals of potassium iodide containing traces of thallous ion, it was of interest to know the content of thallium imphrity over a considerable range of concentration. The few methods for the determination of small quantities of thallium that have been reported (6) are insufficiently sensitive or are not feasible in the presence of large quantities of potassium iodide. It was therefore decided to employ the method of radioactivation analysis, which has the advantages of simplicity, high sensitivity, and reasonable accuracy. The principles of this method have been reviewed by Boyd (1). I n the present work a neutron activation method for the determination of very small quantities of thallium in potassium iodide has been developed n-hich should be applicable, with modifications in the chemical procedure where necessary, to the determination of thallium in othei materials. CHEMICAL SEPARATION OF THALLIUM
The radiochemical procedure for the separation of thallium from an irradiated sample vias devised to provide good separation from as many of the other elements as possible. Of the many methods of separation available, a combination of thallous iodide precipitation and ether extraction of thallium(II1) from hydrochloric acid appeared most promising. Although several elements are extracted in varying amounts by ether from hydrochloric acid solution (41,only tellurium might be expected to coseparate to some extent in the precipitation of thallous iodide because of reduction to the elementary state. For samples containing small amounts of tellurium thcl procedure described below should provide adequate separation, since the extraction of tellurium is of the order of only a few per cent and the cross section for radioactivation is low. I n the presence of large quantities of tellurium, however, it' would be advisable to modify the procedure to include a specific removal of tellurium. This is accomplished by evaporating the ether phase containing the thallium with 3 S hydrochloric acid, adding a few milligrams of tellurium carrier, and saturating the hot solution with sulfur dioxide to precipitate elementary tellurium. After removal of the tellurium, the thallium is precipitated as thallous iodide in the usual n a y . RADIOACTIV4TION OF THALLIUM
Xaturally occurring thallium consists of two isotopes, T 1 2 0 3 (29.5%) and TlZo5(70.5%),which give rise to two radioactive isotopes when subjected to thermal neutron irradiation. The production of these radioactivities is represented by the following reactions (6): TI203 (n,y ) T I 2 0 1
0.77 m.e.v. 8 3 years
p
Pb20'
were sufficiently thin, so that the neutron flux could be considered homogeneous throughout the sample. EXPERIMENTAL PROCEDURE
Standard thallium samples were prepared for irradiation by boring holes into large (1-gram) single crystals of pure potassium iodide and pipetting into the holes 1O-fi1. aliquots of standard thallous sulfate solutions prepared to contain varying quantities of thallium by appropriate dilutions of a standard stock solution. The stock solution was standardized gravimetrically by weighing thallous iodide. The crystals containing the added thallium together with several crystals of unknown thallium content and a crystal to which no thallium was added were wrapped in aluminum foil, placed in an aluminum can, and irradiated for 3 days in the Argonne heavy water reactor. After irradiation the samples were set aside to "cool" for a few days and then analyzed for T 1 2 0 4 by the following radiochemical procedure: The potassium iodide sample is dissolved in 10 ml. of 1 Asulfuric acid, and 2 m]. of standardized thallous sulfate solution, containing 10 mg. of thallium per ml., is added. The precipitate of thallous iodide is digested on a steam bath for a few minutes, and sulfur dioxide is bubbled through the solution to reduce any iodine present. The precipitate is separated by centrifugation, washed once with 1 A; sulfuric acid, and dissolved bv gentle heating with 10 ml. of 4 N hydrochloric acid containing 6 drops of 30% hydrogen peroxide. The addition of hydrogen peroxide plus hydrochloric acid also oxidizes thallium(1) to thallium( 111). The thallium is extracted by shaking with 10 ml. of ethyl ether saturated xith 4 N hydrochloric acid. The ether phase is washed three times with IO-ml. portions of 4 N hydrochloric acid saturated with ether, then mixed with 10 ml. of 1N sulfuric acid, and placed on a steam bath until the ether is completely evaporated. The thallium(111) is reduced to thallium(1) by bubbling sulfur dioxide for a few moments through the hot solution, and thallous iodide is precipitated by adding a slight excess of potassium iodide solution. The precipitate is digested for a few minutes, collected by filtration on a weighed disk of filter paper, washed with 1 N sulfuric acid, ethyl alcohol, and ether, then dried a t 110" C. for 10 minutes, and weighed to determine the chemical yield. The sample is prepared for counting by mounting on a card and covering v i t h a thin (0.6 mg. per sq. cm.) square of "rubber hydrochloride" (available from Reed Laboratories, Akron, Ohio). The time required for this procedure is about 2 hours, and chemical yields are usually about 85%.
.ill sampleb (standards and unknon ns) prepared by the above proredure were counted under identical conditions with a thin end-window beta proportional counter and standard scaling circuit a t a geometry of about 30%. The observed counting rates were corrected for chemical yield, but no correction for self-absorption of beta radiation in the sample was necessary, as all samples were kept nearly the same in weight. Decay and absorption characteristics of all samples were checked to make sure that the observed activities were due to Tl*0d4. The weight of thallium in an unknown sample is obtained from its counting rate by comparison LTith a standard curve of counting rate zs. weight of thallium. RESULTS
The thermal neutron capture cross sections of T P 3 and T P are 11 and 0.16 barn, respectively. Owing to the short halflife of Tl206,the only thallium activity remaining a day or so after irradiation is due to 3-year T1204. I n determining the concentration of an element by a radioactivation method it is essential that the effect of self-shielding be either calculable or negligible ( I ) . In this case the crystals
The counting rates of the irradiated standard samples are presented in Table I. Column 2 gives the n-eights (in micrograms) of thallium added to the 1-gram crystals of potassium iodide (obtained from the Harshaiv Chemical Co), and column 3 gives the observed counting rates of the samples (in counts per minute) corrected for chemical yield. I t is seen that a small but easily measurable thallium activity is present in sample 1, the pure
350
351
V O L U M E 2 5 , NO. 2, F E B R U A R Y 1 9 5 3 potassium iodide to which no thallium was added. This activity is subtracted from the observed counting rates of the other samples to give the net counting rates due to the added thallium shown in column 4. The average specific activity from the data in column 4 is 668 3~ 24 counts per minute per microgram of thallium in the sample. The data show that the counting rate is proportional to the amount of thallium added within esperimental error. The thallium content of an unknown sample is obtained by use of either a standard curve or the specific activit?. value given abow.
served was less than that espected for 0.1 p.p.m. of thallium in the zinc. illuminum absorption measurements were made on all samples as a check on the purity of the observed radioactivity. The range of the beta radiation was found to be 280 mg. per sq. em. (corresponding to 0.75 m.e.v.), in agreement with the known energy of T1204. In addition, the counting rates of all samples n’ere found to be constant over a period of 2 weeks, showing the absence of short-lived contaminants. DISCUSSION
Table I.
Counting Rates of Standard Thallium Samples Observed
TI Added, Sample
Y
1 2 3 4
0 1.23 12.3 123
Activity T1 Due t o Activity .4dded T’ Counts per .AxinUte 218 997 8,668 84.418
...
778 8,450 B4,200
The counting rate of sample 1 corresponds to 0.33 microgram or 0.33 p.p.m. of thallium in the pure potassium iodide. At this low concentration spectrophotometric measurements on single crystals of the pure potassium iodide are not sufficiently sensitive to determine the thallium content, as only a trace of the characteristic absorption bands of thallium ( 2 ) was observed. A sample of Merck reagent grade potassium iodide was also analyzed for thallium impurity. In this case, holvever, an activity barely detectable above the background of the counter was observed, corresponding to an upper limit of about 0.01 p.p.m. -4s a test of the application of the method to materials other than potassium iodide, a 1-gram sample of pure zinc (reagent, Baker and Adamson) Tas analyzed for thallium impurity. [It is reported (3) that zinc often contains traces of thallium.] ,4slight activity (approyimately equal to the background of the counter) was observed. The energy of the beta radiation as determined by absorption, however, was about 2 ni.e.v., which is not characteristic of T 1 2 0 4 , and the activity ob-
The data of Table I shoTT that with an exposure to a flux of 2 x 1012 neutrons per sq. em. per second for 3 days in the Argonne reactor, 1microgram of thallium gives rise to about 670 counts per minute of T1204. Since the background of the counter is about 20 counts per minute, the indicated sensitivity of the radioactivation method under these conditions is about 0.1 microgram of thallium. The sensitivity can be increased further by lengthening the time of irradiation. From the data obtained in this work it is estimated that the method is accurate to about 5% down to a t least the l-microgram level. I n order to check the accuracy of the method, two similar sets of samples (one standard and two unknowns) were given separate neutron irradiations and analyzed for thallium. The data obtained on the unknowns agreed to within 5%. LITERATURE CITED
(1) Boyd, G. E., A N . ~ LCHEM., . 21, 336 (1949). (2) Hilsch, R., P h y s . Z . , 38, 1031 (1937). (3) Hopkins, B. S.,“Chapters in the Chemistry of the Less Familiar
Elements,” Champaign, Ill., Stipes Publishing Co., 1939. (4) Lundell, G. E. F., and Hoffman, J. I., “Outlines of Methods of Chemical Analysis,” New York, John Wiley 8- Sons, 1938. ( 5 ) Sandell, E. B., “Colorimetric Determination of Traces of Metals,” 2nd ed., New York, Interscience Publishers, 1950. (6) Seaborg, G. T., and Perlman, I., Rers. Modern Phys., 20, S o . 4, 585 11948). (7) Yuster, P. H., and Delbecq. C. J., J . Chem. Phys. (to be published). RECEIVED for review July 10, 1952.
Accepted October 13, 1952.
Separation of Antimony by Solvent Extraction CHARLES E. WHITE AND HARRY J. ROSE‘ University of Maryland, College Park, M d .
l H E difficulty of separating antimony from tin and arsenic for the quantitative determination of antimony is well known. The extraction of the antimony-rhodamine B complex with benzene ( 1 ) is a long established procedure and recently West and Hamilton (3)have shown that the antimony iodide complex can be extracted with benzene. I n order to find a more suitwble method for the separation of antimony from other similar elements, a study was initiated in which various complexing agents were used in conjunction with extracting solvents. REAGENTS
Intimony Solution. iintimony trichloride solution in 8 ,V hydrochloric acid about 1 gram per liter standardized with iodine. Dilute to 10 mg. per liter in 8 N hydrochloric acid. Rhodamine B, 0.2% aqueous solution. Ceric sulfate, 0.1 N ceric sulfate in 2 N sulfuric acid. Hydroxylamine hvdrochloride, 0.1 % aqueous solution. Benzene, oxalic acid, citric acid, 85% phosphoric acid; reagent grade. Ethyl acetate, reagent grade, redistilled. 1
Present address, C.S. Geologiral S u r i e y , Washington 26, D. C.
EXPERIRZENTA L
The following complesing agents were tried: tartaric acid, citric acid, oxalic acid, dibenzoyl methane, acetylacetone, catechol and resorcinol. Extracting solvents employed were: methyl acetate, ethyl acetate, butyl acetate, acetoacetic ester, acetylacetone, ethyl acrylate, chloroform, and carbon tetrachloride. Preliminary tests were conducted with combinations of the complexing agents and the various solvents. Both trivalent and pentavalent antimony salts were used and the residual antimony was estimated by sulfide precipitation or by rhodamine B. The results of a series of experiments showed that oxalic acid and citric acid were both good complexing agents for extraction purposes and that a mixture of these tlvo was better than either used separately. Pentavalent antimony seemed to extract slightly easier than the trivalent with these reagents. The increasing orders of extraction with the other complexing agents used were as follows: dibenzoyl methane, catechol, resorcinol, acetylacetone, tartaric acid, osalic acid, citric acid. Ethyl acetate proved to be the best of the extracting liquids for the citrate-oxalate complex. Methyl acetate was too soluble in water for use as an extracting agent. The increasing order
