gives highest extraction rfficiency at 0.351 indium, while double extraction with the same solvent gives essentially complete removal of indium over the range of 0.3 to 0.6N indium. I n all indium concentrations investigated 2pentanone is a bettrr rstraetion solvent than diethyl ethrr. Figure 2 shows the resulbs of estractions of indium with 2-pentanone at various hydrobromic arid normalities from 3.0 to 6.5. For the single extra?tions, t'lie most complete one is accoinplished at 6.ON hydrobromic acid, but it was found that with the second extraca0 I! I tion no phase separation occurred wit>h 36 oa V 3 . A q I i Y O r NDlUM the hydrobromic acid solutions a b o w 5 .OL\~. Figure 1 . Indium extraction vs. indium The extraction losses for the various concentration impurity elenirnts upon extraction with A O n e extraction 2-pentanonc are shown in Table I. 0 Two extractions Thc values given represent the per cent of the original impurity cniicentration line and RindoR ijidth set bo that only cstrarted into the organic phasr. .-i the 47-k.e.v. gamma radiation from the few of t h c b elements which shon-rd high lead-210 Jtas measured. The measureext'raction losses in 4.5-Yhydrobromic3 m m t s of the beta emitters \%eremade acid ncw run again in 3.,?S hydronith a windoLi1eqs flon counter on 0.1bromic acid t o see if there n-orild lie less nil. portions of the tcst .olritinns dried loss at, the l o w r normality. Thallium on copper plancliets. \vas not tried because the data of Irving As the extraction cfhciciiry \I as being and Rossot'ti (1) shotvet! 100% exaluated by comparing a solution 0.6;M in indium \\it11 a qolution from tracat'ion of thallium from 1.0 to 4.0.Y I\ hich most of the indium had been hydrohromic acid. The agreement tmtractcd, it was nect-ary to determine hetween the values from duplicatr the effect t h a t thiq vririation in indium determinations in Table I shows that I oncentration R ould have on the selfthe extraction l o s m arc quite renbsorption of the iolutions. In exproducible. p t ~ i m e n t sdesigned t o determine this cffect. it was fount! that in every case DISCUSSION the change in vlf-ah-orption was iwgligible, and that no correction for 'Thi, separation of indium from a thi.: effect need Iw applied numhcr of its impurities can be effected Jvith varying efficiency by holvent exRESULTS traction. Indium can be extracted 'The extractibilit? of iiitlium 111 confrom hydrobromic acid solutions con(witrations of 0.1 to 0.7.11 fiom 4.5,17 taining relatively high concmtrations Lydiobromic acid with single and (up to 0.6V) of indium leaving only double extractions with 2-pentanonc trace quantities (6 X 10-5M) of indium (methyl propyl ketone) and with and much of the impuritirs in the acid (liethy1 ether is shown in Figure 1. l a y i . Consequently, the rxtraction .1 single extraction n ith 2-pentanonr caan 1 usrd with some w c o w in the 0 2
01
(%
6 0
c
-
-2 1 2
4:
% '
k 1 R t I C L TY ?FHFr
Figure 2. Indium extraction with 2pentanone vs. acid concentration
A O n e extraction 0 Two extractions
production of high purity indium. The extraction can also find application in an analytical scheme which requires separation and preconcentration of the impurities from indium. The reproducibility of the extraction losses of the impurities makes i t possible to make the appr opriate corrections for t h t w I n s w in t h r analytical rtwilts. ACKNOWLEDGMENT
'The author wishes to acknowledgi, the assistance of Frank Dunnington iii performing the extractions and radioactivity measurements, and of W.R. Haivrv in ieviewing the maniisci ipt LITERATURE CITED
(1) Irving, H. M., Rossotti, I?. J . C., Bnulyst 77, 801 (1952). (2) Irving, H. M., Rossott>i. F. J. ('.) J . Chem. SOC.1955, 1027. ( 3 ) Zbid., p. 1938. (4) Zbid., p. 1946. ( 5 ) Knox, K. L., Spinks, J. \V. T., Cun. Chem. ProcessZnd. 30,KO.11, 85 (19463. ( 6 ) Sunderman, D. S., Ackermann, I. B., Meinke, W. W., A 4 ~ aCHEM. ~ . 31, 40 (1959). EDWARI)
B.
OWENS
Lincoln Laboratory
?*lassachusetts Institute of Technology Lexington 73, Mass. Operated with support from the U. S Army. Snvy. a n d Air Force.
Colorimetric Determination of Palladium in Uranium-Fission E le me nt AI loys drh. In tht *tiidit)* cwiinc,c.ted nith the processing of fucxl- for thr Second I .\perirnental Breeder Rractor. it has I x m necessary t o determine lrsb than lc;C palladiuni ill uranium alloys contEiining ruthenium, molybdenum, zirconium. rhodium, and palladium. Although the spectrophotom(,tric detcsrmination of palladium with 2-nitrosoI-naphthol reported by Alvarez ( 1 ) :ind modified by C'hcng ( 2 ) hac the neces-
sary sensitivity, it could not be used Fsitli these alloys because of the large miount of uranium present. I n Cheng's procedure, the palladium iiaphtholate complex is formed a t a pI-1 of 1.5 to 3.5 in the presence of (ethy1enedinitrilo)tetraacetic acid (EDTA) to mask interfering elements, thc solution is made basic with ammonium hydroxide, and thr complc.; i i r\tractcd with toluene. Vntlt~rt h r v
conditione, uranium precipitatw as the diuranate and seriously interferes with the extraction. Alvarez used a somewhat similar procedure, but suggested that the extraction be made from a n acid solution. The excess reagent was then scrubbed out of the toluene with dilutr base. It appeared that the advantages of each method might be rombined to achirve a method which would be uscful VOL. 32, NO. 10, SEPTEMBER 1960
1367
-
U
Ru
€3 spectrophotometer in I-cni. cells
I.
Determination of Palladium in Synthetic Solutions Foreign Ion Added, Mg. Palladium,
Table
Mo
Zr
Rh
for solutions containing large amounts of alkali-insoluble materials. Such was found to be the case. Palladium is quantitatively extracted along with the excess reagent in the p H range of 1.0 to 2.0, no uxanium precipitation occurs, and E D T A serves to buffer the solution as well as to mask interfering elements. The excess reagent is scrubbed from the organic phase with dilute sodium hydroxide. I n making the initial p H adjustment, ammonium hydroxide is preferable to sodium hydroxide. With the latter, results are low and erratic. Similarly, low results were observed by Cheng when he used sodium hydroxide to make the solution basic. He attributed this to the adsorption of palladium on the precipitate formed. Since no precipitation occurs under the conditions described here, the reason for the low results is not apparent. Both sodium and ammonium hydroxide were tested as scrub solutions to remove excess reagent from the tol-
Zn
Cd
Taken
y
Found
uene. Although no large differences in absorbance were noted, the use of 1N sodium hydroxide gave lower and more reproducible blanks. Excessively high blank values were found to be due t o decomposition of the reagent. Purification was effected by dissolving the reagent in ethyl alcohol and precipitating i t with water.
at 370 mp. Determine the palladiiirn content from a calibration curve. The molar absorptivity found by using the above procedure is 22,300. From absorbance data reported by Cheng, a molar absorptivity of 20,800 has been calculated. The color is stable for a t least 48 hours. The procedure has been tested on synthetic solutions containing known amounts of uranium, palladium, ruthenium, zirconium, and rhodium and was found to be satisfactory. The method has also been applied to the determinxtion of palladium in zinc and cadmium. Typical data on solutions containing knowii amounts of palladium are ~ h o w n in Tnblc 1.
RECOMMENDED PROCEDURE
LITERATURE CITED
Pipet a sample containing 5 t o 25 y of palladium into a 60-ml. separatory funnel and dilute to about 30 ml. with distilled water. Add 2 ml. of 301, EDTA solution. Adjust the p H to 1.0 to 2.0 with ammonium hydroxide or hydrochlorir acid. Add 0.1 ml. of 1% 2nitroso-I-naphthol and let stand for 10 minutes. Add 10.0 ml. of toluene. Extract for 1 minute, then discard the aqueoup phase. Add 10 ml. of 1N sodium hydroxide and shake. Transfer the toluene phase to a 15-ml. stoppered centrifuge cone and centrifuge. Measure the absorbance us. a reagent blank with a Beckman Model
(1) Alvarez, E. R., Anales direc. ycit. ofic. quim. nacl. (Buenos Aires) 2, 88
(1949).
(2) Cheng, K. I,., ANAL. CHEM.26, 1894 (1954).
L. E. Ross G . KESSER
E. T.KTTrERA Chemical Engineering Division Brgonne National Laboratory 9700 S. Cass Ave. Argonne, Ill.
Operated by The University of Chicago under Contract No. W-31-109-eng-38. Work performed under the auspicee of the U. S.Atomic Energy Commission.
Determination of Fluorine by Neutron Activation SIR: The method of shorb activation analysis has recently gained some very deserving attention. An activation procedure developed in this laboratory is thus presented for the determination of fluorine. An earlier procedure for fluorine employing the P-rays of F20 for counting 11-as reported by Atchison and Beamer [ANAL.C H E W 28, 237 (1956)]. The heart of the new method, which uses the F13 (n, CY) K16reaction, is a fast-transfer shuttle rabbit system permitting multiactivation runs with cumulative counting to improve both srnsitivity and statistical certainty of the analysis. The operation of the system is illustrated in Figure 1. The activations are carried out with the epicadmium neutron flux of a beryllium target irradiated with 2m.e.v. deuterons from a Van de Graaff accelerator (High Voltage Engineering Corp., Burlington, Mass.). The sample, packaged in a n aerodynamically-designed polyethylene rabbit running inside a 35 foot long polyethylene tube, inch in diameter, is inserted into the 1368
ANALYTICAL CHEMISTRY
system via the air lock located at the 3 X 3 inch NaI(T1) scintillation counter. Counter and air lock are inside a cubical lead shield of 3-inch wall thickness and 14-inch inside edge length. A camdriven timer activates simultaneously the solenoid air valves 1 and 2 and
Table 1.
transfers the rabbit into the irradiation position. The arrestor pin positions the rabbit reproducibly in front of the cadmium-covered neutron source. The solenoid valves are then deactivated and the airflow is stopped. A small air-leak prevents hack pressure that
Experimental Data of Some Fluorine Analyses by Fast Neutron Activation Fluorine, Mg. No. of
Samplr. Countsi 2162 LiF Standard $1 LiF 4140 3179 #2 LiF 1038 LiF Standard 1511 $3 SaCl(F), 517.1 mg. 3286 #4 Organic compound, 22 mg. 268 i 176 Background #5 H20,600 mg. 4 i 17 #6 NH4NOs,650 mg. 10 f 17 = Corrected for background. b Only statistical error considered. c Actual count of an empty rabbit. * Calculated for pure compound.
Cvclri
Foundb
10
IO 10
5 5 10 8 8 8
104*02 8 0 f 0 2
Present 5 45 105 8 0
5.45 T.9 f 0 . 2 8.610.1
..
..
..
8.0 8 7d
...
0
n
