This item is only available as the following downloads:
UNITED STATES ATOMIC ENERGY COMMISSION
STUDIES OF THE DELAYED NEUTRONS
I. CHEMICAL ISOLATION OF THE 56-SECOND AND THE 23-SECOND ACTIVITIES
Arthur H. Snell J. S. Levinger
E. P. Meiners, Jr. M. B. Sampson
R. G. Wilkinson
University of Chicago
This document is reproduced as a project report and
is withouteditorial preparation. The manuscript has
been submitted to The Physical Review for possible
May 9, 1947
Issuance of this document does not constitute
authority for declassification of classified
copies of the same or similar content and title
and by the same authors.
Technical Information Branch, Oak Ridge, Tennessee
AEC, Oak Ridge, Tenn., 5-2-49--850-A4417 _
Printed in U.S.A.
PRICE 10 CENTS
Digitized by Ihe Internei Archive
in 2011 with lunding from
University ol Florida, George A. Smathers Libraries with support trom LYRASIS and the Sloan Foundation
htlp: www.archive.org details studiesofdelayedOOuniv
STUDIES OF THE DELAYED NEUTRONS
IT. CHEMICAL ISOLATION OF THE 56-SECOND AND THE 23-SECOND ACTIVITIES
By Arthur H. Snell, J. S. Levinger, E. F. Meiners, Jr.,
M. B. Sampson, and R. G. Wilkinson
The 23-second delayed neutron activity is found to follow the chemistry of iodine, and the 56-
second delayed neutron activity is found to follow the chemistry of bromine. Comparison with known
beta emitters of like half-lives suggests that the neutron-emitting nuclei may be Xe137 and Kr87.
In the preceding paper, resolution of the deLJy curve of 'he delayed neutrons resulting from the
fission of uranium is described. Activities with half-lives of 0.4, 1.8. 4.4, 23, and 53 seconds were
found. The fact that discrete decay periods are present is good evidence n favor of the interpre-
tation put forward by Bohr and Wheeler' of the existence of delayed neutrons, namely that they
originate in the fission products, and are emitted when the beta-decay of a fragment leaves the
nucleus in a state of excitation higher than the binding energy of a neutron in that nucleus. The
neutron is then immediately emitted, and the rate of decay of the neutron-emitting activity observed
is just that of the preceding beta-activity. In this paper we shall describe successful chemical iso-
lation of the 56-second and the 23-second activities in the fission products.
EXPERIMENTAL PROCEDURE AND RESULTS
Most of the irradiations were made with the University of Chicago cyclotron, using 7.3 Mev
deuterons on beryllium as a neutron source. Beam currents ranged up to 100 microamperes, giving
in the paraffin-surrounded sample a slow neutron flux of about 109 neutrons per cm2 per sec. The
samples were usually a few hundred milliliters of aqueous solution of uranyl nitrate. The counting
was done with a boron trifluoride proportional counter having a total boron cross section of about
2 cm2, surrounded with several inches of paraffin. The counting was done by three workers; one
called the time, the second read the scaler, and the third recorded the readings. During the early
part of the decay the accumulating count was recorded every 5 seconds, but as the activity became
weaker this interval was lengthened to 10 seconds and finally to 30 seconds.
The problem on first approach looked like a difficult one, namely, to identify within the time
limit for chemistry of about one minute, one or two of the thirty-odd fission product elements. With
the idea of at least narrowing down on the possibilities, we made a number of experiments which led
to negative results. It seems, nevertheless, worth while to mention them briefly because some of
them rule out possibilities for the still unidentified shorter activities, and others give evidence
against the odd chance of the presence in the elements concerned of activities vnth about the same
decay periods as those which we have been able to extract.
1) Sulphate precipitation: We irradiated 250 cc aqueous solution containing 40 g uranyl nitrate,
fission product carrier, and barium nitrate. After irradiation we added sulphuric acid, filtered, and
examined the precipitate. Only a very weak neutron-emitting activity was found in the precipitate 30
seconds after the stop of the activation.
2) Water extraction from ether solution: We irradiated 80 g of uranyl nitrate dissolved in 500
cc ethyl ether in a separating funnel, with 10 cc water and fission product carrier. After irradiation,
the liquid was shaken, allowed to settle, and the water was drained off. Most of the neutron-emitting
activity remained in the ether subsequent to 40 seconds after stop of activation.
3) Barium: A barium chloride precipitate was taken from the water layer following an ether
extraction. Only very weak activity was found in the precipitate 60 seconds after stop of activation.
In other experiments, aqueous solutions were activated and barium chloride precipitates were taken
out after irradiation. The precipitates had no neutron-emitting activity 40 seconds after the stop of
4) Rare gases: (a) A flask containing an aqueous solution of uranyl nitrate was boiled under
reduced pressure while under irradiation. An air stream was led through a flask to a NaOH trap,
and thence to a charcoal trap. A weak neutron-emitting activity built up in the NaOH trap, but none
in the charcoal, although the latter became beta active. (b) 1400 cc of uranyl nitrate solution con-
taining 100 grams of the nitrate were irradiated, and after irraoiation allowed to pour through a pipe
into another vessel in which the boron trifluoride counter was set. A decay curve was taken. Then
the experiment was repeated except that the solution was kept boiling during activation. The boiling
should have greatly weakened any rare gas activities, but the decay curves were of the same shape
and the activities were of about the same intensity subsequent to 10 seconds after stop of activation.
Experiments which gave positive results started after we tried a silver halide precipitation,
and found that both the 23-second and the 56-second activities came down very strongly. Since silver
selenite and presumably tellurite also would have precipitated from the solution, we made separa-
tions from solutions which had been made strongly acid with HNO3, under conditions such that (as
we verified by trial) silver selenite would not precipitate. The 23-second and 56-second activities
still came down with the silver halide. This evidence that these activities were distributed between
bromine and iodine seemed not to be in disagreement with the results of experiments described in
the preceding paragraph.
THE IODINE ACTIVITY
About 150 cc of an aqueous solution of uranyl nitrate were irradiated for 2 minutes in a sepa-
rating funnel. A few milligrams of KI and KBr were present to act as carrier, and a few cc of
carbon tetrachloride were present. The concentration of the uranyl nitrate solution was held down
so that the carbon tetrachloride would settle promptly after the funnel had been shaken. One cc of
concentrated HCI was present in the solution, and after irradiation, 10 cc of 5" sodium nitrite
solution were added. Following shaking and settling, the carbon tetrachloride layer (colored violet
by the iodime was drawn off and counted. The result was a single exponential decay with a half-
life of 24 seconds, covering the time interval 30 seconds to 200 seconds after stop of irradiation.
Later we found that the sodium nitrite could just as well be present during the activation, and
that repeated extractions could be made from one batch of solution. One added new carrier and new
carbon tetrachloride before each activation, and did not shake until after the activation. By standard-
izing the procedure so that counts were always recorded at the same time after the stop of activa-
tion, we ran through eight extractions and by averaging the resulting readings we obtained the curve
reproduced in Figure 1 and labeled "Iodine". It will be noticed that the first point comes at 28
seconds after the stop of activation, and that the curve-is a simple exponential over an intensity
factor of about 100. and that the half-life is 23.8 0.7 seconds. The limits of error for this figure
are those allowed by the scatter of the points.
THE BROMINE ACTIVITY
Since the 23-second delayed neutron activity appeared to follow the chemistry of iodine, it
seemed very probable that the 56-second activity would follow that of bromine. A positive experiment
required a specific separation for bromine in an attempt to get the longer-lived delayed neutron
emitter clear of the others. This was accomplished at the Clinton Laboratories, and the experiment
was as follows:
About 1 cc of uranyl nitrate solution was irradiated for 2 minutes in the Clinton pile. The trans-
fer in and out of the pile was accomplished with the pneumatic tube arrangement. After irradiation,
the sample was allowed to stand for 16 seconds to permit decay to a moderately safe level of a
strong 8-second nitrogen 16 activity which was induced in the lucite container of the solution. Then
the uranyl nitrate was poured into a separating funnel containing 30 ml of 8N nitric acid which was
saturated with potassium chlorate. Br- and I-carriers were added and Br2 and 12 extracted with
carbon tetrachloride. The carbon tetrachloride was run into a second separating funnel which con-
tained a potassium nitrite solution made slightly acid with nitric acid. Here the bromine was re-
duced to bromide, the iodine being at the same time kept oxidized. The carbon tetrachloride was
drained off, and the water layer counted. As in the case of iodine, we standardized times and pro-
cedures so that five good runs could be averaged. The resulting decay curve is given in Figure 1,
For comparison, we took readings on samples of uranium nitrate irradiated for 2 min but
followed by no chemical extractions. These gave the curve labeled unseparatedd". The presence
in the bromine of a small amount of shorter activity as indicated by the earliest two points on the
bromine curve is explicable on the basis of imperfect chemical separation in the somewhat hasty
manipulations involved. The isolated bromine activity seems to have a half-life of 54 1 seconds,
where again the limits of error are those permitted by the scatter of the points.
To compare the decay of the separated bromine with that of an unseparated sample in which
other activities have fully decayed, we activated several grams of uranyl nitrate, waiting 6 minutes
before starting the count. The resulting curve is labeled "Larger Sample, Unseparated" in the
figure. Between 6 and 12 minutes after stop of irradiation, the slope corresponds to a half-life of
57 + 1 seconds. This is a little longer than that of the extracted bromine, but the difference is
probably attributable to experimental effects such as the presence of a little iodine in the bromine
Probably the bromine and iodine activities account for all of the 23-second and 56-second
activity observed in the total delayed neutron decay curve. We are not able to set very low experi-
mental limits on the presence in other fission products of activities having about the same decay
periods, but in one set of experiments we compared the intensities of silver halide precipitates with
those of their filtrates. The filtrates were 5 to 10 times weaker, and this residual activity is at-
tributable to the imperfection of the fast filtering, and to the probable presence of some activity
in the form of bromates and lodates which did not precipitate. The filtrate decay curves were rough
but did not require new decay periods for their description.
It is tempting to identify these two delayed neutron activities respectively with the 30 1 6-second
iodine beta activity and the 50 10-second bromine beta activity found by Strassmann and Hahn.2
Seelmann-Eggebert and Borns subsequently found that a 3.8-minute xenon grows from the 30 1 6-
second iodine, and a 75-minute krypton grows from the 50 1 10-second bromine. In other work,
the 75-minute krypton was identified as krypton87 and the 3.8-minute xenon identified as probably
xenonl37. If the possible existence of conflicting bromine and iodine beta activities had been ruled
out, it would seem that the mass assignments of the delayed neutron emitters would be settled.
ON033S h3d SINRO NOyin3N U3AV-13 : AI1AII)V
u u c
. U .0
As it is, coincidences may exist whereby another isotope of bromine has a period close to that of
bromine87, or another isotope of iodine a period close to that of lodine137. We did some work on
the short-lived bromine and iodine beta and gamma emitters, but our results were not of direct as-
sistance. They did indicate, however, that the 50-second bromine and the 30-second iodine are not
the only short-lived halogen fission products; the situation is more complicated, both in bromine and
A comparison of the fission product yield data with the intensity of the bromine and iodine de-
layed neutrons gives some information about their emission. The yields of the short-lived 87 and
137 chains have not been measured explicitly, but according to the fission product yields as deter-
mined chemically,5 2.5-c of the fissions should lead to products of mass 87, and 6.21 should lead
to products of mass 137.
On the other hand, the intensity of the delayed neutrons relative to the instantaneous neutrons'
tells us that only about 0.2c of the fissions produce the 56-second delayed neutron activity, and
only about 1C of the missions produce the 23-second delayed neutron activity. Increasing the as-
signed masses does not bring closer agreement, and one must conclude that one or both of the
following factors must be coming in: (1) a sharp drop in yield for the early members of the chain;
(2) a branching decay process. Such a branching might be as indicated m Figure 2, where the mass
numbers 87 and 137 are provisionally written in. Here it is assumed that the decay of the bromine
(iodine) leaves a few of the krypton (xenoni nuclei in one or more states which are excited highly
enough to permit neutron evaporation, but that most of the decay passes through a lower state and
eventually to strontium (barium) m the conventional manner.
Bohr and Wheeler give general theoretical arguments indicating how the energy of beta trans-
itions in some fission fragments can exceed the neutron binding energy in the product nucleus and
thus lead to delayed neutron emission. The identification of the neutron-emitting nuclei as isotopes
of krypton and xenon now permits closer comparison with their theory, and one can see what the
Bohr-Wheeler considerations predict with regard to the mass assignments of the activities. In
Table 1, we give the energies available for beta transformation and the neutron binding energies
as calculated according to the method of Bohr and Wheeler for isotopes m the region with which we
are concerned. It will be noticed that according to these figures delayed neutron emission would be
energetically possible for bromine-krypton transitions of mass 88 or higher, and for the iodine-
xenon transitions it would be possible for mass 140 or higher. Although the statistical theory prob-
ably cannot be forced to detail for individual isotopes, the figures illustrate the a prior probability
that the delayed neutrons should come from nuclei heavier than 87 and 137.
Table 1. Bohr-Wheeler beta decay energies and neutron binding energies.
A Beta transition Neutron bmding A Beta transition Neutron binding
energy in bromine energy in krypton energy in iodine energy in xenon
Mev Mev Mev Mev
87 5.3 5.7 137 3.5 5.5
88 8.8 7.9 138 6.0 6.3
89 6.9 4.7 139 4.8 5.0
90 10.5 6.5 140 7.2 5.8
141 6.0 4.6
142 8.3 5.0
*Snell, A. H., et al., preceding paper.
Br87 ( 1 37)
Kr87 ( Xe137)
55 sec (23 sec)
RbT( Ca C '37'.\
Kr86 (Xe136) Stable
Figure 2. Possible branching mechanism which would account for the low yield of the iodine and
bromine delayed neutrons in comparison with the yields of fission products of mass 137 and 87. The
branching ratios have been adjusted on the assumption of the indicated mass assignments (which are
uncertain), and on the basis of an initial intensity of all of the delayed neutrons equal to 100 of the
intensity of the instantaneous fission neutrons.
Measurements have also been made in this laboratory upon the delayed neutrons from the fission
S of thorium.6 For the most part, the same decay periods were found, but the 56-second activity was
about three times more intense relative to the 23-second activity than was the case for fission of
uranium 235. This can be qualitatively understood from the fission yield curve5 if the mass numbers
87 and 137 are approximately correct. A shift of the lower mass peak toward still lower masses
would increase appreciably the relative yield at mass 87, while mass 137, being near the flattened
top of the other peak, would be almost unaffected. Disregarding the possibilities of changes in yield
along the chain and of changes m the branching ratio for the delayed neutron emission, the thorium
results may be said to imply that the bromine activity should be ascribed to a mass number less
than 90. The results of Jentschke7 for uranium 238 and thorium fission seem to indicate that the
lower mass peak is more sensitive than the higher mass peak with regard to changes in the mass
of the fissioning nucleus.
ATTEMPTS AT RECOIL COLLECTION
An experiment which looks attractive from the point of view of obtaining unambiguous mass
assignments and also of identifying the shorter-lived emitters is that of examining the radioactivity
of the nuclei which recoil because of the emission of the delayed neutrons. These recoil nuclei
might be expected according to present knowledge to have an energy of a few kilovolts. We attacked
the problem as follows:
Fission fragments emitted from the inner surface of a uranium cylinder about 3 inches in
diameter were collected electrostatically upon the surface of a metal rod arranged axially in the
cylinder. After irradiation, a paper or aluminum sleeve was slipped over (but not touching) the rod,
and the cylinder was evacuated to a few millimeters pressure. The sleeve was supposed to pick up
the delayed neutron recoils. Times of irradiation and times of waiting before placing the sleeve
could be arranged so as preferentially to emphasize collection from any desired delayed neutron
activity, and repeated collections could be made on one sleeve to budd up intensity. We found that
the sleeves always showed radioactivity-even when collection was started alter the delayed neu-
trons had all decayed. This meant that activity was evaporating from the surface of the rod and
blanketing the rather small effect which we sought. Variants of the experiment which we tried in
attempts to reduce this effect included the application of retarding electrostatic fields, variation in
pressure during the recoil collection, and cooling the rod bearing the fission fragments. Our re-
sults have been inconclusive, but the experiments might be worth pursuing under quite carefully
controled conditions. Incidental considerations m the case of the bromine and iodine activities are:
(1) the recoiling rare gas atom might not stay on the sleeve; and (2) if the mass assignments of 87
and 137 are correct, the recoiling nuclei are probably stable.
We are indebted to Dr. N. Sugarman for early advice on the chemical separations, and to Dr.
Katherine Way for consultations on the application of the Bohr-Wheeler theory.
1. Bohr, N., and J. A. Wheeler, Phys. Rev. 56:426 (1939).
2. Hahn, O., and F. Strassmann, Naturwiss. 31:59 (1943).
3. Born, H. J., and W. Seelmann-Eggebert, Naturwiss. 31:59 (1943).
4. Born, H. J., and W. Seelmann-Eggebert, Naturwiss. 31:86 (1943). (See also V. Riezler,
Naturwiss. 31:326 (1943).)
5. Rev. Mod. Phys. 18:513 (1946), or J. A. C. S. 68:2437 (1946).
6. Brolley, J. E., J. S. Levinger, M. B. Sampson, and R. G. Wilkinson, CP-787.
7. Jentschke, W., Zeits. Physik, 120:165 (1942).
END OF DOCUMENT
UNIVERSITY OF FLORIDA
111111111111111111111llll l 11111111111111111111111111111111
3 1262 08907 9726
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EB4ZS1D5D_5KQQZU INGEST_TIME 2012-02-29T19:01:55Z PACKAGE AA00009351_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC