You are on page 1of 5


KESIN 2777

amounts. Specific separations of the fission prod- quantitative separation of large masses of rare-
uct radioisotopes, including the individual rare earth elements are outlined. The principles in-
earth species, have been demonstrated. The ap- volved can be applied to the determination of
plication of the results to the problem of curie- physico-chemical constants and to the develop-
level fission product separation is shown and the ment of new types of analyses.
necessary development steps required for the P. 0. BOX w, OAK RIDGE, TENN. RECEIVED JULY 7,1947


The Separation of Rare Earths by Ion Exchange.lP2 I. Cerium and Yttrium


1. Introduction trict, of the use of columns of Amberlite type res-

For many years one of the most difficult proc- ins for the separation of fission products, both with
esses i n the field of chemistry has been the separa- and without the use of citric acid-ammonium ci-
tion of the rare earths from each other into their trate eluants a t controlled PH has been described
pure states. Their chemical and physical prop- elsewhere and will not be discussed here.?
erties are so similar that in general a single opera- The present paper is the first of a series, from
tion leads only to a partial separation or enrich- this laboratory dealing with the successful separa-
ment. tion of macro quantities of rare earths of spectro-
Ever since the beginning of the Manhattan grHphic purity, by adsorption on Amberlite type
Project there has been a constant demand for resins and subsequent elution with complexing
samples of rare earths of exceptional purity in agents such as citric acid-ammonium citrate solu-
gram amounts or greater. This demand arose for tions a t controlled pH. This paper establishes
numerous reasons, but mainly because some of that cerium and yttrium can be separated rela-
the rare earths ;are formed as fission fragments tively rapidly by these methods on any desired
during fission of the heavy elements. It was scale.
highly desirable, therefore, to have a means of The marked success of the process described
preparing pure rare earths so that their nuclear depends on the fact that the rare earths form com-
properties could be studied and also to allow ;L plexes with the citrate ions. If the PH is suitably
more thorough Consideration of their chemical adjusted, competition is set up for the rare earth
behavior. Their radioisotopes are less well under- ions between the citrate complexes and the active
stood than those d any other group of elements. centers of the resin. Therefore, as the citrate
In general, the best means of separating these solution washes the rare earths down the column,
elements has been the well known but laborious each rare earth ion is adsorbed and desorbed
method of fractional crystallization a s used by many times. Since the equilibrium constants for
James and further developed in many laborato- the rare earth citrate complexes vary slightly
ries. Exceptions are cerium with its quadrivalent among the different rare earths, their rates of
state, and samarium, europium and ytterbium travel down the column differ sufficiently to lead
with their di-valent states which do permit a to their separation. The repeated cycles in the
means of separation from the normal trivalent columns effectively replace the thousands of in-
rare earth ions. dividual operations required by the older methods
A number of workers have reported studies on for separating the rare earths and lead to a highly
the application of chromatographic and ion ex- effective process analogous to the use of distilla-
change methods to the separation of the ran: tion columns.
earthsa.*J'6 While they obtained considerable
enrichment their results were not sufficiently 11. Materials, Apparatus and General Proce-
promising to lead to further intensive investiga- dure
tion or to the quantity production of pure rare 1. Materials.-The cerium stock solution
earths. The history within the Manhattan Dis- was prepared from pure (NH&Ce(NO& which
(1) This work was supported, in part, by a grant from The Man. was dissolved in dilute nitric acid, feduced with
hattan District, U. S. C!orps of Engineers. sulfuric dioxide and precipitated as Ce(OH)s.
(2) This work was done during the period of January to March, The hydroxide was filtered off, dissolved in sul-
1945, and was presented before the Manhattan Project Couacil,
Chicago, Illinois, on April 15, 1945, and before the Chicago Section furic acid and stored. The yttrium stock solution
of the American Chemical Society in November, 1945. was commercial yttrium chloride having a purity
(3) E. Lange and K. Nagel. 2. Elcktmchctn., 41, 210 (1936). greater than 80%.
(4) G. Sahama and V. Kamula, A n n . Acad. Sci. Pcnnicac. AW8 The citric acid was of commercial grade but
no. 3, 5 (1941).
( 5 ) U . Croatto, Riccrca sci., 11, 157 (1941). (7) F. Daniels. W. C. Johnson and L. L. Quill. Chcm. Erg. News.
(6) R . G. Russell and D. W.Pearce. THISJOURNAL. 66.595 (1943) 9 1 2494 11941)
2778 F. H. SPEDDING,

spectroscopic analysis showed only traces of metal I n the experiments designed to determine the
ions. The term “570 citric acid solution,” used nature of the elution curves for each ionic species,
throughout this paper, refers to 50 g. of the citric when only tracer amounts were employed (see
acid monohydrate per liter of solution. The oxa- Figs. 1 and 2), a very small aliquot of the original
lic acid used as precipitating agent was commercial solution was evaporated on a watch glass and
grade and contained no salts capable of introduc- measured on the electroscope. This permitted a
ing metallic ions into the precipitate. calculation of the total amount of activity used
Cerium and/or yttrium tracers were used in all in the experiment. During the course of the elu-
the experiments; their use in analysis will be dis- tion 1-ml. samples of eluate were collected a t vari-
cussed later in this paper. The tracers were ob- ous volume intervals and measured similarly. A
tained by chemically separating the true rare graph showing activity/ml. as a function of vol-
earth group and yttrium from the other fission ume was plotted. A simple integration under
products in neutron irradiated uranium. The ce- this curve permitted the calculation of the per
rium was then oxidized to the quadrivalent state cent. of the ionic species eluted a t any desired
and separated, as the insoluble iodate, from the point (see Fig. 3).
other active rare earths. After several months A somewhat similar method was employed in
the activities, other than Y , had all decayed leav- those experiments in which macro amounts of ma-
ing radio-pure Y g lwith its 57 day period and beta terial were used. A known amount of inactive ma-
activity of 1.4 MeV. maximum energy. The Ce terial was weighed out and dissolved. Tracer ac-
isotope was Ce144with a 275 day half-life. The tivity was added and the solution made up to
other cerium isotope, 28 day Ce141,had decayed known volume; a small aliquot was dried and
when the measurements were made; measure- measured on the electroscope. This gave the ac-
ments were always made after the Ce144had tivity per unit weight of carrier or, conversely, the
established equilibrium with its 17.5 min. ‘Pr weight of carrier corresponding to unit activity.
daughter. This pair of isotopes emits beta rays Samples were collected as above during the elu-
of 0.31 MeV. and 3.1 Mev., respectively. While tion and their radioactivity measured. Using
the tracer solutions were not entirely carrier free the conversion factor, the weight of material
the amount of inactive carrier was negligible. eluted was plotted as a function of volume of elu-
Individual stock solutions of these tracers were ate. Integration under the curve to any point
eniployed in the subsequent experiments.
2. Calculation and Presentation of Data.-
Two methods have been employed in presenting
the data, each one depending upon a measure-
ment of the activity in a 1-ml. sample of eluate

as the means of measuring the amount of ma-
Ce, pH 2.74
terial eluted.




100 300 500 100 300 500

M I . eluate 50 150 250 50 150 250
Fig. 1.--The elution of tracer amounts of Ce and Y using M1. eluate.
5 and 2057, citric acid: PH 2.75, flow rate 3 crn./min.. Fig. 2.-The elution of Ce and Y tracers in 5% citrate
column dimensions 1 cm. X 60 rm. Tbe curves for Ce at various PH values: flow rate 3 cm./min.; column di-
drv nti iht- lrft. Y on the riaht mensions, 1 rm X 60 cm
RESIN 2779

200 300 400 500 3 5 7 9 11

M1. eluate. Liters of eluate.
Fig. 3.-Individual elution curves for Ce and Y tracers: Fig. 4.-The elution of a small amount of Y in the pres-
PH 2.74; flow rate, 3 cm./min.; column dimensions, I ence of a large amount of Ce: PH 2.77; flow rate, 5 cm./
cm. X 160cm. min.; column dimensions, 1.6 X 190 cm.; composition of
starting material, 10 mg. Y and 250 mg. Ce.
along the abscissa 'gave the weight of material chamber and the highly active samples were read from a
eluted to that point (see Fig. 4). stage somewhat removed from the ionization champer.
In general, in studies of mixtures,of Ce and Y , 4. Preparation of Columns.-The columns were con-
only one species of tracer was added. This per- structed by drawing out one end of the glass tube and
sealing on a small outlet tube, finally resembling a buret.
mitted a better determination of the intermediate A plug of glass wool was inserted a t the bottom of the
region between the two peaks. The total elution column to prevent washing out of the resin; the tube was
curve was obtained by precipitating the mixture dented directly above the glass wool to prevent it from
in any volume interval as oxalate, igniting to ox.- pushing up the resin bed during the process of backwashing.
The resin, Amberlite IR-1, was first screened to remove
ide and weighing. The weight per unit volume the large particles (>20 mesh) and then soaked in water.
was plotted as a function of the volume of eluate. The resin slurry was added t o the column and backwashed
The difference between the total weight curve and with water to remove fine resin particles which tended to
the one obtained by using activity measurements plug the column and thus cut down the flow rate. The
resin was conditioned as follows. A solution of 5y0
is due to the species without the tracer. hydrochloric acid was passed through the column con-
Not only does this method give a means of a s verting the resin to the acid cycle. The contact time was
saying the samples which is faster and more ac- at least thirty minutes to insure complete conversion.
curate than simple weighing but i t also allows the This process was followed with a 2% sodium chloride solu-
tion to convert the resin t o the sodium cycle. Treatment
detection of small amounts of material a t or near with acid and salt was repeated two or more times, the resin
the point of break-through. The break-through being finally left in the acid cycle.
represents the total volume of eluant a t which the
presence of the given element is lirst evident. At; 111. Experimental
low concentrations of rare earths the oxalates 1. Elution of Ce and Y Singly
may have sufficient solubility or tendency to su- a. Effect of Concentration of Citrate in Eluant-Pre-
persaturate so that no precipitate is formed. Sim- liminary experiments were undertaken to determine the
effect of variation in original concentration of citric acid
ple weighing of the ignited precipitate thus has a in the eluant. The concentrations of citric acid tested
practical lower limit of detection of 5 mg. per wcre 1, 5 and 20% adjusted to pH 2.74 by the addition of
liter. However, tracer measurements can be concentrated ammonium hydroxide; the elution was
made much more sensitive by increasing thc spe- followed with both Ce and Y tracers. The 1% solution
was ineffective a t the pH employed. Thc change in citrate
cific activity of the original sample. concentration had L: large cffect on the rate of elution of
3 . Description of Apparatus.-Thc principal picccs of both Cc and Y tracers. At the highcr concentration they,
apparatus were long vertical glass tubcs filled with the especially Ce, were eluted iii a much smaller volume of
cation exchanger, Amberlitc IR-1 resin. A cylindrical, eluant. The data for the higher concentrations of citrate
silvered glass Geiger-Mueller tube of the type made by the are presented in Fig. 1. Since the 5% solution gave satis-
Eck & k c b s Company was mounted in a copper housing factory results, it was chosen as standard for the subse-
having a stiiall window opening. This unit was conncctetl quent experiments.
t o a counting circuit and scaler by a long lead which made b. Effect of pH Using Tracer Amounts.-The resin bed
the G M tube fairly portable. By moving the counter was 60 cm. in height and 1 cm. in diameter. The 5y0
tube up and down the column, the position of the material citrate solution was used as eluant a t pH values of 3.04,
could be dcteriniried at any tiiiie during those experiments 2.91, 2.81, 2.74 and 2.62. The flow rate was maintained
iii which radioactive tracers were used. a t 3 ml. per minute per square cm. of cross-section area.
The smiplcs of eluted Cc and Y trcrc rnc:isurctl O I I :L This unit (~nl.jmin./sq. cni.) is equivalent to cm. of
Lauritscn qti:irt7. lilwr clcctro>copc having :i mtiltiijlc >t:igtt coluinn length pcr minute and is so designatcd in subse-
i~ioulitillgi n which c011vcrhioIIfxctors il:icl becll workcd out qucnt discussioii. San~plcsfor analysis consisted of l'nil.
bet\vccn the stages to f;icilit.rtc the i~~ca~iireiiient of ex- of eluate on watch glasses. They were dricd undcr infra-
tremes of activity. The <;imides containing but l i t t l v red lamps and the activity read on thc electroscope. Thr
activitv wrrv r w r l from _ I i t n g r rlow to t h e ionizntiori cerium zamples werr wxl tllrnugh a 70 mg./sq. cm.
aluminum absorber to remove the soft beta rays; the was equivalent to 250 mg. of Ce,Oa and 250 mg. of YzOr.
yttrium samples were determined without an absorber. I n one set of experiments Ce tracer and, in another set, Y
The data obtained are shown in Fig. 2. The elution of tracer was added to the inert material. At PH values
both Ce and Y increased with increase in PH. Yttrium higher than 2.87 both Ce and Y were eluted so rapidlythat
was satisfactorily eluted at all the PH values tested while there was little separation. The best PH value for separa-
the elution of Ce was relatively poor a t PH values less than tion is in the neighborhood of 2.60. The summary of the
2.81 ; the change in shape of the curve with varying PH is results of these experiments is given in Table 11.
especially striking with this element. However, a t lower b. Effect of Original Y/Ce Ratio.-In the previous
PH values yttrium would undoubtedly have shown similar experiment the Y/Ce ratio was unity. Since, in actual
changes in shape of curve. practice, the ratio may be far from unity, data were
c. Effect of Column Length Using Tracer Amounts.- obtained for a mixture equivalent to 10 mg. of YZOS and
The preceding experiment showed that the Ce tracer 250 mg. of CerOa, a ratio of 1 to 25. The bed length of the
moved down the column a t a slower rate than did the Y resin was 190 cm., the diameter 16 mm. and the pH
tracer. 'These results indicated that a longer column 2.77. The elution curves are shown in Fig. 4. When all
should improve the separation. Experiments were per- of the Y had been eluted, about 60% of the Ce was still in
formed by the same procedure as in the previous experi- the column and was subsequently eluted pure. Even
ment except that the bed length was increased to 162 cm. when the Y was present in such a small relative amount,
and the pH was 2.75. The superimposed elution curves are more than 60% of it was obtained in a pure form before the
shown in Fig. 3 ; it is evident that the Y is eluted much Ce break-through. It can be concluded that the pro-
more rapidly, under these conditions, than is the Ce. The cedure described can yield the products of spectroscopic
Y/Ce ratios, as a function of the per cent. of Y eluted are purity.
given in the second column of Table I . c. Effect of Length of Column.-It has been shown
d. Elution of Macro Amounts of Ce and Y.-Since the that a separation of Ce and Y of greater than 96% can be
results with tracers, on a 162 cm. column, showed a effected, from a 1: 1 mixture of these two components, on a
separation of Y and Ce, the same procedure was applied to column 190 cm. long and 16 mm. in diameter a t a PH
macro amounts of the materials. The Y/Ce ratios, as a range of 2.55-2.65. It was also shown that when Ce and
furiction of the amounts of Y eluted, are given in the third Y are eluted independently, they move a t different rates,
column of Table I. It is evident that a much better indicating that the band fronts might move farther apart
separation was obtained with the macro amounts than as the length of the column is increased. I t , therefore,
with the tracer amounts. seemed of interest to see whether the separation of the
bands of the two elements continually increases or whether
2. Elution of Mixtures of Ce and Y in Macro they move apart a certain distance and then proceed at
much the same rate.
Amounts A column was prepared which was 610 cm. in bed length
The previous discussion has dealt with elution and 25 mm. in diameter. The mixture contained 250 mg.
of each component employing tracers of Ce and Y. The
in which the primary purpose was to determine pH of the eluant was 2.66 and the flow rate 5 cm. of column
the nature of the elution curve for each ionic spe- length per minute. The band fronts were determined by
cies, under identical conditions, and in the ab- use of the portable counting apparatus. The data ob-
sence of interfering ions. Data are now presented tained are given in Table 111.
on the elution of macro amounts of mi.xtures of TABLE
Eluate, Separation of band
Y C e R A T I OA S A FUSCTIONOF THE PER CENT.OF Y liters fronts, Em.
ELUTEL, 16 173 * 1.5
Y/Ce Y.ICe
(tracer) (macro) 20 193 * 15
5.0 24 218 * 15
4.0 42 28 236 f 20
3.7 30 32 297 * 20
2.9 13 Analysis of the eluates, after the break-through, showed
2.6 8.3 that only 90-95% of the Y was obtained free from Ce
2.3 5.2 even though the band fronts were moving further apart.
This phenomenon is probably associated with channeling
1.9 and possibly somewhat with back diffusion. This experi-
. . I 2.7 ment indicates that there is a critical bed length of column
beyond which no improvement in separation can be gained
a. Effect of pH.--A series of experiments was con- by increasing the length of the column.
ducted with the pH values of the eluant varying in the
ratige 2.50 to 2.87. The column beds were 190 cm. in IV. Summary
length and the diameter 16 mm.; each sample adsorbed
Conditions have been described for the separa-
TABLEI1 tion of cerium and yttrium by adsorption on Am-
EFFECTOF pII ON PER CENT. Y ELUTEDREPORE Ce hxlite IR-1 columns and elution with 57, citric
BREAK-THROUGH acid-ammonium citrate solutions. The separa-
Per cent. Y eluted before tions were studied both for tracer and macro
91% Ce break-through amounts, using radioactive isotopes of the ele-
2.87 70.0 ments for the determination of the degree of sepa-
2.77 82.0 ration. 'The factors studied included concentra-
2 titi 96.0 tion ~i el iant, PH o f eluant, the ratio of cerium
2.61 98.5 m t l yttriiitn i n the starting sample and length of
2 *io 91 . o columns. Conditions were established by which

the elements, of spectrographic purity, were sep- for the large scale separation of adjacent rare
arated in macro quantities. The principles estab- earths; the details will be published in subsequent
lished for the separation of cerium and yttrium papers.
are being applied to the development of methods AYES, IOWA RECEIVED
JULY 7, 1947



The Chemical Identification of Radioisotopes of Neodymium and of Element 611

Introduction Earlier work on the'Manhattan Project had
In the comprehensive studies of the radioactive revealed the presence of two unidentified fission
species produced in the fission of uranium' i t has products in the rare earth region. A soft beta
been found that over thiity are members of the emitting activity of -4y half-life was discovered
rare earth family (isotopes of yttrium and the by Balloue and independently by Goldschmidt
group lanthanum thrbugh europium), The chem- and Morgan.Io This activity was later studied by
ical and physical identification of these was an Seiler and Winsberg6 who set the half-life a t
important part of the research program of the -3.7~. These investigators considered the activ-
Manhattan Project. Standard oxidative separa- ity to be an isotope of praseodymium, neodymium,
tions and fractional precipitations4 and the use of or element 61 on the basis of detailed qualitative
radiochemical methods based on chain relations chemical separations between lanthanum and
served to distinguish the activities of yttrium, praseodymium. Another unidentified rare earth
lanthanum, cerium, and some of praseodymium, activity was later discovered by Davies" through
and those6of samarium and europium. The char- its characteristic gamma radiations. This was
acterization of the sequence praseodymium, neo- confirmed by Hume and Martens12 who deter-
dymium, and element 61 presented very difficult mined its half-life as -1ld and showed that the
problems6 that were solved only with the intensi- activity was not cerium. We have been able to
fication of ion exchange methods originally devel- identify these two activities as isotopes of element
oped by Boyd and co-workers' and applied to the 61 and of neodymium, respectively, both of mass
rare earth field by Cohn and co-workers.* In this number 147. In addition the 47h 61149 has been
paper is reported the successful separation of these identified among the fission products. All three
three elements, the first to have been achieved with of these activities, together with a previously uni-
radioisotopes of neodymium and of element 61, dentified 1.7h isotope of neodymium, are also pro-
(1) This document is based on work performed a t the Clinton
duced as the resiilt of the activation of neody-
Laboratories operated under Contract No. W-35-058-eog-71 for the mium with slow neutrons.
Manhattan Project. classified reports of which are referred to here
as "Manh. Proj. Rep." The information covered herein will appear Chemical Identification
io the Plutonium Project Record (Plut. Proj. Rec., Vol. 9B, papers The l l d activity was first subjected to standard
7.54.3, 7.54.4, 7.54.5, 7.54.6. and l l . l ) , io Division I V of the Mrrn-
hattan Project Technical Series.
qualitative serpations4q6in order to h i t the iden-
(2) Present address: Department of Chemistry, Massachusetts tification problem to a definite group of rare earth
Institute of Technology, Cambridge, Mass. elements. Separations by carbonate digestion
(3) THE PLUTONIUM PROJECT,Nuclei Formcd io Fission:' Decay definitely eliminated yttrium and all rare earth
Characteristics, Fission Yields and Chain Relationships, THISJ O U R -
NAL,68,2411-2442 (1946). Reprints are obtainable from the offices
elements of atomic number greater than 61. Oxi-
of The American Chemical Society, dative fusion with sodium nitrate4f6demonstrated
(4) A. A. Noyes and W.C. Bray, "A System of Qualitative Analy- that the activity was not an isotope of lanthanum
sis for the Rare Elements,"The Macmillan Co., New York, N. Y., and must therefore be identified with praseody-
(5) L. Winsberg, Manh. Proj. Rep. CC-2310, pp, 231-244 (Jan.
mium, neodymium or element 61. Finally, the
1945) and CC-2966 (Apr. 1945); Plut. Proj. Rec., Vol. 9B, 7.55.1, activity was limited to an isotope of neodymium
7.55.2, 7.56.1, 7.56.2, and 7.56.3 (1946). or element 61 by demonstration of separation from
(6) (a) N. E. Ballou, Manh. Proj. Rep. CC-389B (Dec. 1942), praseodymium with potassium hydroxide fusions.
CC-920 (Sept. 1943), CC-1204 (Jan. 19441, CN-1312 (May 1945),
Plut. Proj. Rec., Vol. 9B, 8.24.7 and 8.24.12 (1946); N. E. Ballou Positive identification of'the 3 . 7 ~
and the l l d
and J. A. Marinsky, ibid., 8.24.11; (b) J. A. Seiler and L. Winsberg, radioisotopes was then achieved by the recently
Manh. Proj. Rep. CC-2310, pp. 227-230 (Jan. 1945); Plut. Proj. developed ion-exchange method using a synthetic
Rec., Vol. 9B. 7.54.2 (1946). organic cation exchanger (Amberlite IR-I) of the
(7) (a) G. E. Boyd, J. A. Swartout, et at., Manh. Proj. Rep. CN-
3346 (Dec. 1945); J. A. Scbubert, Maoh. Proj. Rep. CN-1873 (Jan.
sulfonated phenol-formaldehyde type. This new
1945); G . E. Boyd. ,J. Schubert. and A. W . Adamson, THIS J O C R - (9) N. E. Ballou, Manh. Proj. Rep. CC-680, p. 22 (May 1943) and
NAL. W, 2818 (1947); (b) A . W. Adamson. Manh. Proj. Rep. CN- CC-3418 (Feb. 1946); Plut. Proj. Rec., Vol. 9B, 7.64.1 (1946).
1859 (Apr. 1944); G. E. Boyd, A. W. Adamson and L. S. Myers. (10) B. Goldschmidt and F. Morgan, Canadian Pruj. Rep. MC-II
THIS JOURNAL, 69, 2836 (1947). (Aug. 1943).
(8) E. R. Tompkins, and J. X. Khym, W. E. Cohn. Manh. Proj (11) T. H. Davies, Maoh. Proj. Rep. M-CN-1424 (Apr. 1944).
Rep. CL-WEC-10 (Dec. 1046); THISJ O U R N A L , 69, 2769 (1947). (12) D. N Hume and R . I. hlartens, ibid., CN-1311 (June 1944).