Jan. 5, 1952
THESZILARD-CHALMERS ENRICHMENT OF 35.9 h Br8*
237
[CONTRIBUTION FROM THE OAKRIDGENATIONAL LABORATORY ]
Recoil Reactions with High Intensity Slow Neutron Sources. I.’ The SzilardChalmers Enrichment of 35.9 h Br82 BY G. E. BOYD,J. W. COBBLE A N D SOLWEXLER~ Enriched 35.9 b J W samples from a Szilard-Chalmers reaction in crystallinc KBrG (in the Oak Ridge graphite pile) s h o d specific activities up t0 750 rd./mg., corresponding t o a 22,000-foM enrichment. The variation of the specific activity with irradiation time and neutron source intensity agreed with a n elementary theory based on the activation equatiao and the assumptiaa that the rate of the non-activating radiation decomposition was proportional t o the mass irradiated a d to. the intensity. The prediction that the s p t q f k activity will he independent of the source intensity and will decrease with increasing time was centinned as was the predicted inverse dependence of the enrichment on source strength and time. Theretention of r a d b k o m i n e as bromate was found to increase with time suggesting a radiation recombination may occur.
Introduction tive data on other compounds with this theory has The recoil effect following the capture of slow been published e1sewherea4 neutrons by atomic nuclei has been known almost Experimental Procedures from the time of the diseovery of artificial radioactivity. S i n e its demonstration by Szilard and Commercial C.P. potassium bromate which ordinarily Chalmers,3 numerous investigators have employed contains 200-1oOO p.p.m. of bromide was unsatisfactory since many of the specific activity determinations in this the reaction to prepare sources of undetermined work invohwd only a few hundred p.p.m. Approximately high specific activity, or to produce recoil atoms 400 g. of bromate was purified in batches as follows: About possessing unusual chemical reactivities. There 40 g. of Baker Analyzed KBrOj was dissolved in 800 ml. of have been, however, virtually no quantitative distilled water, cooled t o loo, and gaseous chlorine passed for two hours. The treated solution was filtered studies of the total changes in compounds in which through and extracted wlth reagent grade carbon t e t r a c h k i d e until the recoiling radioatoms are formed. Gamma rays it was colorless and odorless. After three more extractions, and fast neutrons always present in slow neutron the aqueous phase was evaporated to about one-half its sources regardless of how they are sustained also v o b m e and cmled slowly to ca. 10’. The crystallized act to rupture chemical bonds. Non-radioactive KBrOs was separated, washed with 95% alcohol and vacuum Chemical analyses of one gram aliquots showed 24 isotopes are released which dilute the active species dried. p.p.m. bromide and a water content of 0.41%. The puriand lower the specific activity. The radiation de- fied bromate, which was stored over Gael,, appeared t o be composition fragments which are many orders of stable for at least six months as shown by periodic analyses. The analysis for total bromine (bromide + bromine + magnitude more numerous than the rewiled radiohypobromite) produced in the K B r G by radiation decomatoms may also react with the latter and reduce position was conducted first using a cobrimetric procedure the yield of separable radioactivity appreciably. based on the bromination of phenol red in alkaline solution,6 Very little attention has been given by workers in and subsequently by a micro-potentiometric titration with the Geld of “hot atom” chemistry to this possible 0.01 N A g N q . This latter proceduree involved dissolving a few hundred inilligrams of the irradiated salt in water concomplication. taining 150 X of a 270 sodium arsenite solution which served The past neglect of radiation decomposition to reduce all the f o r m of bromine present in the crystal exeffects has resulted from the limitations of con- cept bromate t o bromide. The potentiometric method is veutional analytical methods for the estimation believed reliable dowii to a total bromine content of about p.p.m. A blarik determination was always performed at of the ultramicro ammnts formed by weak neutron 20 the same time. sources. The availability of modern high intensity Radioactivity measurements were made with a calibrated slow neutron sources, such as the chain reacting 4 T geometry ionization chamber of 20-liters volume filled pile, has permitted a re-examination of this aspect, with 40 atmospheres of argon.’ The ion currents were amplified by means of a dynamic condenser electrometer whose since there has been a proportionately increased output was fed into a continuous recording Brown potenradiation decomposition to levels accurately deter- tiometer which indicated the instantaneous value of the acminable by microchemical techniques. A further tivity as a function of time. The chamber was calibrated advantage over radium-beryllium and cyclotron to correct for non-linearity in scale response owing to ionrecombination, for resistance ratios between scales, and for sustained neutron wurces is that the intensity may the dependence of the scale reading on the position of the be varied aver wide ranges, making it convenient sample in the measuring thimble. The estimation of absoto determine the effects of source strength. lute disintegration rates for radionuclides whose decay Carefully purified crystalline potassium bromate schemes were known was made possible by a careful measureof the detection efficiency of the chamber as a function was employed in this s t d y of the formation of ment of gamma ray energies between 0.177 and 2.07 MeV. The chemically separable 35.9 h E P 2 . Measurements decay scheme of Deutsch and Siegbahns for radio-BrSP of the inactive bromine simultaneously produced was used in the conversion of the observed ionization chamby radiation ckcompssitwn were made so that the ber readings to rutherfords (106 d / s ) in the specific activity calculations given below. Because of its great stability and variation of the specific activity and degree of en- wide range for activity detection, this ionization chamber richment of the recoiled radiobromine could be electrometer arrangement was well suited for an accurate compared with the predictions from a simple theory. half-life measurement of Brazwhich was found to be 35.87 f A comparison of a limited amount of semi-quantita( 4 ) R . R U’illiams. J . P h y s . Colloid Chrm., 62, 603 (1048). (1) Presented before Division of Physical and Inorganic Chemistry. 118th Meeting, American Chemical Society, September 4-8, 1950, Chicago, I11 (2) Department of Physics, Argonne National Laboratory, P 0 Box 5207, Chicago 80. Illinois (3) I, Szilard and T . A Chalmers, Nofuvc, 134, 462 (1934).
(5) Snell, “ODlorimetric Methods of Analysis,” Vol. I , 11. Van Nostrand CQ,In.., New York, N. Y . , 1036. (6) P. F. Thomasson, private communication. (7) C.J. Borkowski, Anal. Chcm.. 21, 348 (1949). (8) M. Deutsch, Phys. Rev., 61, 672 (1942); K. Siegbahn, A . Hedgran and M. Deutsch, i b i d . , 76, 1263 (1949).
238 0.05 hours.0 A measurement of the slow neutron capture isotopic cross-section for stable Brn1gave 2.8 barns. Radioactive bromine was separated as follows: A few hundred milligrams of irradiated KBrOa were dissolved in water and the solution made up to 25 ml. Aliquots were pipetted into actinic glass separatory funnels, bromine water was added and the mixture extracted with an equal volume of carbon tetrachloride.'O Two extractions of the aqueous phase were conducted to insure a quantitative separation. I n some experiments the bromine activity was re-extracted back into water with bisulfite in order that no foreign activities be measured. It was observed, however, that the recovery from the organic phase was not quantitative since even after repeated washing 2-3% of the activity remained in the carbon tetrachloride, possibly because of a bromination reaction. Since accurate determinations of the decay rates of the organic layers gave a 35.9 h half-life, the re-extraction step was not continued. Activity measurements were also performed on the extracted aqueous KBrOc solutions. The observed activities were corrected to a common zero taken as the end of the bombardment. A t the end of the irradiations the KBrOs samples contained appreciable amounts of the shorter lived 4.4 h and 17.5 ni Br" as well as 12.8 h IC4*activities. Consequently, the separation and measurement of the 35.9 h B*Z activity was performed after a sufficient time elapsed (usually one week) to allow for the decay of these shorter periods. The ratio of the bromine activity in the organic layer to the total bromine activity initially in the irradiated potassium bromate serves t o define the recovery factor, +t. Repeated determinations showed & to be reproducible usually t o better than 2%. The neutron irradiations were carried out on one-gram samples sealed in air in quartz ampoules which were transferred t o the pneumatic tube terminal near the center of the Oak Ridge pile. These bombardments were monitored using weighed cobalt wires whose induced activities subsequently were measured in the ion chamber described. The thermal neutron flux ( L e . , full source intensity) a t this position so determined was nv = 6.7 X 1011 neutrons/ cm z/sec. A preliminary investigation was made to find out if an appreciable thermal decomposition of potassium bromate could occur during the irradiation by estimating the temperature at the point where the samples were bombarded to be ca. 8O0, and then heating weighed aliquots at this temperature in an air oven. Chemical analyses showed the thermal decomposition was negligible. Several experimental studies were made to determine if a significant thermal exchange tending to lower the recovery of radiobromine occurred either: (1) in the crystalline bromate during irradiation or upon standing at room temperature afterwards; (2) during the dissolving of the irradiated crystals because of a surface-catalyzed reaction; or, ( 3 ) because of a reaction in homogeneous aqueous solution between the various oxidation states accessible to the recoiled Br** and bromate. Repeated determinations of the recovery from the irradiated salt allowed to stand up to 150 hours at room temperature showed no changes in &. At 80", however, a slight decrease was observed which amounted t o 5% after 64 hours of heating after irradiation. The occurrence of a possible surface reaction seemed to be excluded by the following observations: (a) The recovery was measured in an experiment in which crystals previously irradiated for 24 hours were dissolved directly in bromine water. A value (0.70) within the experimental error of that obtained (0.69) upon dissolving in pure water was observed. (b) Irradiated bromate was dissolved in a KBrOa solution, bromine water was added after a brief period and the mixture extracted. A value of 0.69 was observed. (c) No change in & was found when the irradiated bromate was dissolved in an aqueous solution made from strongly irradiated KBrOa whose radioactivity had disappeared by decay. (d) An aqueous solution containing small amounts of tagged bromate ( L e . , KBrlOs) was prepared in which, in separate experiments, pure and "dead" irradiated KBrOa, respectively, were dissolved. Subsequently bromine water
was added and an extraction was made. KO radiobromine was detected in the organic phase. The preliminary observation11 that no exchange occurs between Br-, Brz and BrOa- in neutral aqueous solutions was carefully checked under the conditions of the foregoing radiochemical separation procedure, since the slight hydrolysis of the bromine water added lowered the PH to about 4. Radiobromine as bromide bromine was mixed with inactive potassium bromate in the same relative proportion as in irradiated salt. Aliquots from this mixture were dissolved in water, allowed to stand for periods up to two hours and then carried through the separation procedure. In every case all the 35.9 h B P was extracted into carbon tetrachloride. A study of the exchange over a wide PH range confirmed its existence in solutions more acid than 10-3 M. I t was concluded from these and other experiments which will not be described that no exchange occurs in the activity separation procedure employed, and that the extraction of separable radiobromine was quantitative.
+
Experimental Results and Discussion Quantitative data on the amounts of inactive bromide formed in the radiation decomposition of KBrOa and on the recovery and specific activity of 35.9 h BrS2also formed in the neutron irradiations conducted in this study are presented in Tables I and 11. An inspection of the second column of Table I shows that the decomposition increased TABLE I SPECIFICACTIVITYAND ENRICHMENT OF 35.9 h Br82 P R O DUCED BY A SZILARD-CHALMERS REACTIONWITH SOLID KBrOl IN THE OAKRIDGEPILE (IRRADIATIONSAT SP, = 6.7 X 1011 NEUTRONS/CY.~/SEC.)
.
(9) J W. Cobble and R . W. Atteberry, Phys. Reo., 80, 917 (1950) (10) The addition of carrier bromine i s not necessary t o extract the radio-bromine: however, the procedure i s shortened and the radioactivity can be concentrated into a few cc. if small amounts are used. Further, the presence of the carrier serves t o effect the exchange of all the radio-bromine into an extractable form (except that in RrOs-)
Irradiatinn time (hours)
0.25 .50 .75
1.00 1.33 2.00 3.00 4.00 6.on
8.00 10.0 12.0 16.0 22.0 24.0 46.0 48.0 64.0
Decompnsition" (Av. p.p.m. bromide)
RecovCry factorb
(1) 17 31 f 2 (3) (1) 45 71 f 7 ( 3 ) (1) 142 197 Jr 20 (4) 293 (1) 403 i 32 (3) 528 rt 67 (3) 730 i 50 (2) 863 f 40 (2) 1140 i 10 (2) 1563 f 100 (2) 2198 f 56 (2) 2267 (1) 4400 (1) 4911 (1) 6800 (1)
0.80 .77 .76 .75 ( .74) .73 .72
(#a)
Ratio of totdl Bra atoms by de- gnSt compn. to richMeasured radioment Bra* facspecific tor activity atoms (rd./mg.) X 10-4 ( E )
746 818 20 824 677 * 68 500 500 f 40 472 . 7 3 459 f 40 . 7 z 508 f 67 ( .52) 501 f 30 .72 500 rt 23 ( .71) 439 f 5 .69 400 * 30 .69 371 .69 385 378 f 10 .70 3191 ( .69) 286 300 & 20 .67 238
*
22000 11900 8100 5000 2600 1780 1180
5.4 4.9 4.9 6.0 8.1 7.5 8 6 8 8 7 9 8 1
1170 fi.50
8.1 9.0 10.1 10.9 10.5 12.7 14 2 17.0
490 400 310 210 148
71 47
Number of independent determinations in parentheses.
* Value in parentheses interpolated from a smooth curve.
TABLEI1 SPECIFIC ACTIVITYOF 35.9 h B F PRODUCED IN SOLID KBrO, BY R E c o n FOLLOWING SLOW NEUTRONCAPTURE, AND THE VARIATION OF ITSENRICHMENT WITH THE SOURCE INTENSITY (TWO-HOUR IRRADIATIONS) Rela ti ve source intensity
0.20
.40 .60
.so
1.00
Decomposition (Av. p.p.m. bromide)
Recovery factor
Measured specific activit (rd./mg.y
49 89 130 170 22 1
0.73 .73 .72 .73 ,72
425 516 511 547 530
(11) W F. Libhy, TRISJOURNAL,61, 1930 (1940).
Enrichment factor ( E )
7760 4310 2860 2220 1700
THESZILARD-CHALMERS ENRICHMENT OF 35.9 h Brs2
Jan. 5, 1952
linearly with the time of irradiation. Similarly, the second column of Table I1 shows the direct proportionality of the decomposition to the source intensity. These findings confirm the hitherto assumed decomposition equation for the rate of formation of inactive atmns, N ,
-
dNt/dt = IkT( No Nt) m I kThTo Nt = NO [I-exp ( - I k ~ t ) a ] I kTNd
(la) (Ib)
where ( N o - N , ) is the number of undecomposed molecules containing one atom of activating element per molecule; I , the neutron source intensity; and, kT, the total radiation decomposition coefficient, taken as a measure of the stability of the compound to the total radiations emitted by the neutron source. This coefficient will be independent of the source intensity and irradiation time; it will depend, however, upon the nature of the compound and upon the character of the neutron source. A least squares treatment of the decomposition data (Table I) down to one hour gave a value of 97 p.p.m./hr. for k T when the source intensity, measured by the slow neutron flux, was 6.7 X 10" neutrons/cm. z/sec. Since the total decomposition even after 64 hours was only 0.68'% the approximation indicated in Equations 1 will be fairly good. The linear time dependence of the decomposition suggests that probably a radiation induced back reaction was absent in this instance. However, much longer irradiations would be needed to establish this point. The rate of formation of active atoms, Nt*, is given by dlV*t/dt
=
IULVOO- AN$*
239
considerations could be found for the variation of Na*,the semi-empirical relation
was employed together with Equation (lb) to obtain an expression for the specific activity, St, of the separated atoms
and for the limiting specific activity a t zero time,
so
so
(4b)
= +ou%A/kT
A value oi 615 rd./mg. Br for So was estimated using the extrapolated value of q50 = 0.91 and u = 2.78 X cm.2/atom Br81, 0 = 0.494, X = 1.93 X hr.-l and k~ = 97 p.p.m./hr. A comparison between the specific activities computed from Equation (4a) and those observed is afforded by Fig. 1. Excepting for irradiations shorter than one hour in duration, the agreement is within the experimental errors involved. According to Equation (4a) the specific activity should be independent of the intensity of the slow neutron source. This was confirmed by irradiations in which the exposure time was kept constant a t two hours (Table 11). I
(2)
where u is the isotopic activation cross-section ;
No, the total number of atoms of element being activated; 0, the relative abundance of the target isotope; and A, the decay constant. The rate of formation of chemically separable radiobroinine atoms NZ does not follow Equation (Z), however. Further, there was a clear indication that sonic sort of radiation back-reaction involving them did occur, for (Column 3, Table I) the separable activity was not constant but diminished with increasing time. Several equations, including one assuming that the radiation induced back-reaction of recoiled BF2 was first order, were derived and tested with the recovery data (Table I) after correcting for the small loss owing to thermal exchange mentioned earlier. Although all of these equations predicted a constant recovery after long times, none of them could be made to fit the data over the whole time interval. The observed approach of $.t to an apparent limiting value was much more rapid than could be accounted for by a first order back-reaction. These theoretical considerations did, however, indicate that the apparent independence of & on source intensity found in the two hour irradiations (Column 3, Table 11) was not real. The relative contribution of the back-reaction after so brief an irradiation was smaller than the experimental errors in the recovery determinations. A demonstrable dependence of 6 on source strength may be expected for a 100-hour bombardment. Since no simple equation based on eIementary
'I "~----2a-*.-
8Fs.D
.,,OW
7,-
/*O"S,,
-~
__+r-----
Fig. 1.-Variation of the specific activity of separable 35.9 h BrSnproduced by neutron irradiation of crystalline KBr03 with time. Solid line according t o Eq.4a using observed values of & in Table I.
It is also useful in discussing recoil reactions to define an enrichment factor, E , by the ratio of the Szilard-Chalmers specific activity to the specific activity attainable with a non-enriching radioisotope producing method
Equation (5) shows that for a given compound the enrichment will diminish with increasing time and neutron source strength. These predictions are borne out by the entries in the last columns of Tables I and 11, respectively, as well as by Fig. 2. Insofar as Equations (4) and ( 5 ) , which have seemed to describe the Szilard-Chalmers process in solid KBrOa, may be applied to other compounds, certain generalizations may be made. Evidently, (Fig. 2) quite high enrichments can be obtained
N.HARRYS~MANT AND CLAREMCGINNIS
240
Vol. 74
of 5 X lo5 might be expected for an intensity of 8.7 X 109 (typical of cyclotran saurces) and of 5 X 108 for an intensity of 6.7 X 1V n/em.2/stc. which is of the order of magnitude of fluxes sustained by Ra Be neutron soureek Interestingly, although very few data can he hand in the literature, the values of lo7 for the enrichment of P3? produced by a recoil reaction with triphenyl phosphatei2 and of lo8 for I i a from ethyl iodidei3 are of this latter order of magnitude. h soniewhat smaller enrichment (los) than W Q U be ~ ~ expected has been reported in the preparation of 2.59 h Mnsb from KXInO4.I4 Despite the €ugh enrichment with weak neutron sources it is emphasized that the specific activity of the separated radioisotope will be independent of the strength of the source. From the foregoing considerations the advantage of the Szilard-Chalmers effect should disappear as progressive1 m r e intense slow neutron sources ~ ~ C Q I W availa le. However, at resent, a quite worthwhde 90-fold enrichment of g5.9 h Brs2 may be achieved by irradiating pure crystalline KBr03 for one half-life in a flux of 6.7 X 10" n., crn.?,sec. Acknowledgments.--The autbors wish to express their sincere thanks to Messrs. P.F. Thomasson and Stanley Rasmussen of the Analytical Division for total bromide analyses on numerous irradiated bromate samples.
+
6
I
I
IO'
I
IO2
mt
1 IO"
Fig. 2.-Variation of the enrichment of 356 h Brd' separable from neutron irradiated crystalline KBrO, with exposure (time expressed as hours).
either by short irradiations, or by the employment of low intensity sources. If an enrichment of 5000 for Bra* is found after a one-hour exposure in a s b w neutron flux of 6.7 X 10" n,/crn.?/sec., then a value
(12) 0. Erbachtr and K (I:))
(.'hem., !19A, 263
o. 1:rbauher and hi. Ueck, z. anora. c h m , too,357 ( i ~ i q
(1.1) t!
OAK
Phitlip, %. j + m k .
1)rehninii. i! p h y u b . C h r w . , 631). 027 (1943).
RIDGE,~ 1 ; ~ s .
iCOSTRlBUTION FROM T I f E D E P A R T M E S T OF CIIEMISTRY,
IlIcr;~venJ i x i I l i , 3951
DUQUESNE UNIVERSITY]
The Absorption Spectra of Some Diary1 Ketones, Hydrazones and Azines I3v I i . H.\RRS SL>i.\sr .\XI)C'i..\HE A ~ C ( ; I S S I S ~ The comparison of t b t m n of a nurnkr of diaryl ketorics. t h e w bydrarorus and azimtx p r d , t b chPlJirnrion d t k o k i m r w d o ~ l l g nR*JOinl$thrfdhwrrrgcategorrr: (1) I L~ a k s r b c h r r e r r r kfnwn~kcconjlrptmrcdtkuyi group aitb (he douklz bwA of the c r b y l group, aqd s h r h are also fouod io the h y d r a z w s a d azhm d the kecows, (21t hr peaks which resat from the resomqce which traverses the whole length of the arinc rnolccule; (3) (he p c a b which corrc. spond to the r m n a n c r d the a*partirms of the molccak, or which arc traceable I O the contributions by inchidual \true'. tural features ( I c , t h e phrriylmcrcapto. or tbe phenylsulfonyl Rroaps)
In order t h a t ;I better untkrstanding may be 01). talneti of the ultraviolet absorption spectra of 3 s r rics o f diaryl kctazines thwe were also investignted the spectra of thr coneqmrithng kctories 3 1 ~h1y d n ~ ( J I I C ' S . Of the diiiryl ketazinm only t h a t derived f r o m t)ciizoplicrioIir seems t o have t x e n studied spectrophotometric~~l~v.* T h e substitutccl berizo. plienorir azines itre closely related to the rccciitly described azines of 3wbstitated awtophcnoiies.'
E e e a t d Rasulto 'Tht hflramnes a d azines ernplo-wtl I
I ) A p o r c m a of thil
r a n d b y C L( to &h -7 i m Jum. )9Q1 R h r r r u r 01 T c 4 . e l o g . 12) E.R B I O ~ I . v w E ~ ~
Vrw.ar)4
III
this
I, rrll.s horn the M S I h n i s ruh. t h 4 h d 01 D y r u o c W -a ot C. M , The h f u c h u
wits
C I H sf Corrtrlo.T I I I S J u ~ , R w A ~u. ., iga? ( 1 9 4 ~ ( 3 ) H if S i f u n t i n 4 H J Mbnbnwk. i b d 71. 4981 ( 1 3 i o )
study were preparetl by tlic rnetliods t1evelol)t.d i i i t h i s 1,al~iratory.~T h e physicil properties and ana. lytical data f o r thr new coiiipovnds are listed i i i 'Fithk I . The absorption spectra wre drterrriinetl by rricaiis of a Rcckrnan I3C spcctmphotometer c m . pk)yirig solutioris i r i 957; cthariol. T h e absorptiori ciinvs are reproduced m Figs. 1-1, and the spectral data ;ire siirnriiari7ed i r i T;it)lc I I
