Citation
Fission product degradation and neutron moderating properties of fluorocarbons

Material Information

Title:
Fission product degradation and neutron moderating properties of fluorocarbons
Creator:
Scott, Thomas Howard, 1932-
Place of Publication:
Gainesville
Publisher:
[s.n.]
Publication Date:
Copyright Date:
1966
Language:
English
Physical Description:
viii, 135 l. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Capsules ( jstor )
Dimers ( jstor )
Dosage ( jstor )
Fluorides ( jstor )
Fluorine ( jstor )
Fluorocarbons ( jstor )
Irradiation ( jstor )
Molecules ( jstor )
Neutrons ( jstor )
Polymers ( jstor )
Dissertations, Academic -- Nuclear Engineering Sciences -- UF ( lcsh )
Fluorocarbons ( lcsh )
Nuclear Engineering Sciences thesis Ph.D ( lcsh )
Radiochemistry ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: l. 132-134.
Additional Physical Form:
Also available on World Wide Web
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030085648 ( AlephBibNum )
11021108 ( OCLC )
ACG9888 ( NOTIS )

Downloads

This item has the following downloads:


Full Text







FISSION PRODUCT DEGRADATION AND

NEUTRON MORDERATING PROPERTIES

OF FLUOROCARBONS


















By
I H'i:AS H'-,:'.; R) SCOTT










I* ['::L .T'.T.:N intr.F:i TF T THE GRADUATE COUNCIL OF
TiHf F lNr LFP: T' OF FLORIDA
1.i f AI.TLLL Fi.LI LLi lENT OF THE REQUIREMENTS FOR THE
Lii.:PL *:-F DOCTOR OF PHILOSOPHY









LN I'. ERSITY' OF FLORIDA


April, 1966












ACi: :rO'JLEDGENr'IT S


The author wishes to express his appreciation to

Dr. John A Wethinmton, Jr.. chairman of the supervi-

sory committee, for his untirinr guidance, assistance,

and encoura.emrent without which this work would have

nor. been Cossible.

The assistance ,f Mr L. D. Butterfield in the

performance of the reactor irradi.aions and the as-

sisrance of .Lr. J A. r'acLean in the performance of th.

,amrma irradiitions are acknou.ledred.

Dr. T. V. Reed provided the distillation equipment,

and Dr. F. F. Parkinson assisted in the neur.orn pulsin<

experiment .

The imars sp.ectrometric analy3se3 were performed

by Dr. iM. P. F3allatter and Dr. R. J. Hanrahin.

The nuclear frma;neric resonance measurement s were

made b" Dr. ';. 3. Prey, and the infrared specrra were

measured -," Dr H C Prown.

The author is indebted tc the Oak Rid,!e Institute

of Nuclear Studies and to the Collece of En-ineerinz of

the University of Florida for fellowship assistance.

Finally, the patience, enccurscement, and persever-

ance of the author's wife, Janet, during, the preparation

of this rranuscript are rratefully acknowledged.

ii














TABLE OF CONTENTS


ACKNOWLEDGMENTS . .

LIST OF TABLES . .

LIST OF FIGURES . .

ABSTRACT . . . .


Chapter

I.

II.

III.

IV.

V.

VI.


INTRODUCTION . . . . . .. 1

PREVIOUS WORK . . . . . ... 4

EXPERIMENTAL PROCEDURE . . . .. 18

EXPERIMENTAL DATA AND RESULTS ... 41

DISCUSSION OF RESULTS . . . ... 56

CONCLUSIONS . . . . .... . . 95


Appendices

A. CALCULATIONS . . . . .

B. ANALYSIS FOR POLYMER IN IRRADIATED
SAMPLES . . . . . . .

LIST OF REFERENCES . . . . . . .

BIOGRAPHICAL SKETCH . . . . . . .


Page

. . . . . . ii

. . . . . . iv

. . . . . . v

. . . . . . vii









LIST OF TAPL[S


Table Fa' e


1. Suimirar,' of .onlu3iii-ns in Feferenc:es 1
Through 20 . . . . . . . . 9

2. Results of Anal'.'ses of CgF16 . . . . 2

3. Decomposition Percentages and Gaseous
Product Distribution for Gamma Irradiation
of C F16 . . . . . . . . .

4. G Values for Gamma Irradiation of CgF1 . .7

5. Decomposition Percentages and Gaseous
Product Distribution for Reactor
Irradiation of CF16 . . . . . . L9

6. Decomposition Percentages and Gaseous
Product Distribution for F3actor
Irradiation of C8F16 and UF6 . . ... 1

7. G Values for Reactor Irradiation of CgF-'
and UF6 . . . . . . . .. .

8. Geometric Buckling and Decay Con3tantz for
Pulsed Neutron Experiment. . . ... :.

9. Data for Figure 11 . . . . . . .

10. Macroscopic Absorption Cross Sections of
Moderators . . . . . . . . 9L

11. Supporting Data for Figure 20 . . .. 107

12. Supporting Data for Figure 21 . 108

13. Supportin C Data for Fi,:ure 2 . .. ... 109









LIST OF FIGURES


Figure Page

1. Chromatogram of Crude C8F16 . . . 19

2. Chromatogram of Distilled C8F16 . 22

3. Metal Vacuum System . . . . . 24

4. Metal and Glass Vacuum Systems . . .. 25

5. Capsule Fluorination System . . . 28

6. Glass Vacuum System . . . .... 37

7. Neutron Pulsing Experimental Equipment 38

8. C8F16 Decomposed to Gas in Gamma
Irradiation . . . . . . ... .60

9. C8F16 Converted to Polymer in Gamma
Irradiation . . . . . . . 61

10. Log (N x 100) Vs. Dose for Gamma


NO
Irradiation of CgF6 . . . . . 64

11. Log ( x 100) Vs. Dose for Other
Fluorocarbons . . . . . . .. 68

12. Dose-Decomposition Relationship for
Fluorocarbons Corrected for Effects of
Lighter Products . . . . .... .72

13. C8F16 Decomposed to Gas in Reactor
Irradiation . . . . . . ... .74

14. C8F16 Decomposed to Gas in Gamma and
Reactor Irradiation . . . .... 75

15. Analysis of Pulsed Neutron Experimental
Data . . . . . . . . . . 86









FiF ure


Fa e


16. Counts Per Channel Vs. Time for CF~1 .F 1

17. Counts Per Channel Vs. Time for Empty'
TanR . . . . . . . . Q

19. Counts Per Channel for CFP i'inus Counrs
Per Channel for Empty Tank Vs. Time . 91

19. Decay Constants Vs. B2

20. Chromatogram of Cases from Sample No. 36 10L

21. Chromatogram of Sample No. 36 in Vapor
Phase . . . . . . . . . 10

22. Chromatogram of Sample No. 36 in Liquid
Phase . . . . . . . . . 10-













Abstract of Dissertation Presented to the Graduate Council
in Partial Fulfillment of the Requirements for
the Degree of Doctor of Philosophy

FISSION PRODUCT DEGRADATION
AND NEUTRON MODERATING PROPERTIES OF FLUOROCARBONS

By


Thomas Howard Scott

April, 1966


Chairman: Dr. John A. Wethington, Jr.
Major Department: Nuclear Engineering

The radiolysis of perfluorodimethylcyclohexane

by gamma radiation and by reactor radiation was studied.

The primary product was a dimeric material with small

amounts of gaseous products including CF4, C2F6, C2F60,

C3Fg, and C3F0O. The production of dimer and gaseous
products was found to be a linear function of the total

dose.

An expression was derived which related the frac-

tion of sample remaining after exposure to a known dose.

This expression was modified to correlate all available

data for the irradiation of alicyclic and aliphatic

fluorocarbons. A separate expression was found for

vii









aromatic fluorocarbons. The expressions hold over a

wide ranre of total doses, dose rates, and for several

different types of radiations.

The decomposition of CjFlI in a solution of CjFlI

and UW. exposed to reactor irradiation was investigated.

Techniques were developed for the preparation of solu-

tions and for the subsequent separation of the CF1,'

from the JUFr and the fission fraj.ments. The primary

product were CF4, C-.FP, and C-F,0 in roughly equal

proportions. lio dimeric material was found. The kinetics

of this reaction supported the assumption that fiZsion

fragments shatter the ,FIr, into fragments which are then

saturated with fluorine present in the system.


viii













CHAPTER I


INTRODUCTION


During the past ten years considerable effort has

been devoted by the Atomic Energy Commission towards

the development of a nuclear reactor in which a hydro-

carbon liquid is used as the working fluid. During the

past ten years there has also been a large scale devel-

opment of the fluorocarbons. The proposal has been made

that fluorocarbons might be used as the working fluid

in a nuclear reactor (1).

This proposal is based on several factors. First

of all, the hydrocarbons tend to form complex compounds

with the iron in the walls of the colder regions of the

reactor system (2). These carbon-iron compounds then

tend to be deposited on the heat transfer surfaces of

the reactor fuel and seriously retard the rate of heat

transfer from these surfaces. The fluorocarbons are

generally considered to be much less reactive than

the hydrocarbons; therefore, this corrosion problem

might be eliminated. Secondly, uranium hexafluoride

is soluble in fluorocarbons. Thus a possible reactor

concept could be developed in which the uranium fuel

is dissolved in a fluorocarbon working fluid.

1








The uranium hexafluoride-fluorocarbon reactor

would have several distinct advantages. First of all,

uranium hexafluoride is the uranium compound used in

the gaseous diffusion process for producing enriched

uranium. Thus, the material could be used without fur-

ther processing,and the conversion and fuel fabrication

costs for solid fuels could be eliminated. These costs

are a significant portion of present fuel cycle costs.

Secondly, by means of a side stream, the fuel could be

continuously reprocessed and new fuel added. The need

for lengthy shutdowns for refueling would be eliminated

and the offsite reprocessing of solid fuel elements

would not be necessary. A final possible advantage of

such a reactor mieht be the radiation induced pro-

duction of fluorocarbons which can not be produced by

conventional chemical processes because of the nonre-

active chemical properties of the fluorocarbons. The

fissioning of uranium in a solution would produce fission

fragments of high energy. The possible decomposition of

the fluorocarbon and the subsequent recombination of the

various decomposition products might produce new and

useful fluorocarbons. In this consideration,the uranium

hexafluoride-fluorocarbon reactor would be closely akin

to the chemonuclear reactor which has been discussed for

several years. The chemonuclear reactor uses radiation-








induced chemical reactions or dissociations to produce

desired chemical compounds.

Obviously, much information regarding the per-

formance of fluorocarbons and uranium hexafluoride-

fluorocarbon solutions in the presence of radiation

must be obtained before any use is made of these

materials in reactors. Among the types of information

needed are the rates of decomposition, the products

formed, the possible effects of the evolution of

fluorine or hydrogen fluoride, and the compatibility of

fluorocarbons and uranium hexafluoride with regard to

possible side reactions. It is also desirable to know

the neutron moderating and absorbing properties of the

fluorocarbons.

The purpose of this work is to investigate the

effects of ionizing radiation on the fluorocarbon,

perfluorodimethylcyclohexane (C8F16), the performance

of CgF16-UF6 solutions in a reactor flux, and the
neutron moderating and absorbing properties of C8F16.












CHAPTER II


PREVIOUS WORK


Radiation Chemistry of Fluorocarbons


Fluorocarbon chemistry is a relatively new branch

of chemical science. Before 1937, perfluoromethane and

perfluoroethane were the only positively known fluoro-

carbons. In 1937 (3) and in 1939 (4), Simons and

Bloch reported the preparation and properties of

fluorocarbons containing more than three carbon atoms.

These compounds, which ranged from perfluoropropane to

perfluoroheptane, were reported to be stable both chemi-

cally and thermally. During the early stages of the

Manhattan Project the need arose for materials which

could be used as sealants and coolants in mechanical

systems containing corrosive uranium hexafluoride. Since

all other materials which had been tested had been found

to react with uranium hexafluoride, Simons (5) suggested

in July, 19LO, that the recently developed fluorocarbons

might be suitable for these applications. In December,

1940. Simons sent a two-cubic centimeter sample of liquid

fluorocarbon, virtually all of the material available,

4






5


to Columbia University for testing. These tests showed

that fluorocarbons had the desired properties (6), and an

extensive research program was then started to find the

most desirable fluorocarbons for the various applications

and to develop processes for producing these fluorocar-

bons. Thus by the end of the second World War,

fluorocarbon chemistry was established as a distinct

branch of chemical science.

The best known of the fluorocarbons commercially

produced in the post-war era was Teflon. A series of

papers published in the period between 1953 and 1956

(7, 8, 9) indicated that Teflon had very poor radiation
resistance. Indeed Teflon was found to be severely

damaged at a dose of 3 x 106 roentgens (9), and unfor-

tunately this property was generally extended to all

fluorocarbons without any real experimental evidence on

fluorocarbons other than Teflon. An additional negative

factor was the liberation of fluorine during the irradia-

tion of Teflon (5). The corrosion problems related to

the presence of fluorine gas in a mechanical system would

seriously restrict, if not prohibit, the use of fluoro-

carbons in ionizing radiation fields.

Slowly, experimentaldata were published which showed

that all fluorocarbons were not as sensitive to radiation

as Teflon. In 1958, Feng reported that the gamma irradi-

ation of mixtures of partially fluorinated hydrocarbons








and hydrocarbons and mixtures of fluorocarbons and hydro-

carbons indicated that the organic fluorine compounds

showed a much greater resistance to radiation than did
the analogous chlorine, bromine, and iodine compounds (10).

Also in 1660, Wall, Florin, and Brown (11) reported that

C6F6 had radiation stability approaching that of C6H6

and suggested that C6F> be used for the preparation of

radiation resistant polymers. Mixtures of CjF6 with

hydrocarbons were less stable than the parent material,

and the difference was attributed to the production of

hydrogen fluoride in the hydrocarbon-fluorocarbon

mixtures.

The first reactor irradiations of fluorocarbons

were performed by Simons and Taylor (12) who irradiated

the saturated fluorocarbons--C-Fl6, CGF160, and (CF9)-N--

in a reactor flux of 5.5 x 1011 neutrons/cm.2 sec. for

four weeks. The G value (number of molecules produced

or destroyed per 100 ev. of energy absorbed) was

calculated to be 2 or 3 molecules transformed per

100 ev. absorbed with a total energy absorption of 2.2

x 10 ev. per gram. No fluorine or CFL were observed,

and there was no evidence of corrosion, since the

inner walls of each capsule were bright and clean after

irradiation.

Concurrently, Wall and Florin (13) found that the

tensile strength of irradiated Teflon decreased very









rapidly in the presence of oxygen, but the loss of ten-

sile strength in the absence of oxygen was hardly

perceptible for long periods of irradiation. This find-

ing suggested that the reported radiation instability

of Teflon was due to oxygen impurities.

Bloch, MacKenzie, and Wiswall (14) used electron,

gamma, and reactor sources to irradiate C6F6, C12F10,

and C10F18. The two aromatic compounds had overall

stabilities of the same order as their hydrocarbon

analogs. Even though the fluorocarbons tended to undergo

more reaction to form high molecular weight compounds

than the comparable hydrocarbons, the fluorocarbons were

found to be more stable to fragmentation leading to

gaseous products. The gas yields for the alicyclic com-

pounds were much higher than the aromatics, but polymer

yields were of the same order or possibly lower. This

observation was in direct contrast to the hydrocarbons

where the alicyclics are much less stable to irradiation

than are the aromatics. Gas chromatography was used to

analyze the radiolysis products.

In 1962, Heine (15), using doses of 5 x 108 rads,

studied the gamma radiolysis of C8F18, (C4Fg) N,

and c-CeF160 ; the stability of these fluorocarbons was

shown to be greater than the radiation stability of the

analogous hydrocarbons. During 1964 and 1965,five papers

were published which added greatly to the understanding









of the effects of radiation on fluorocarbons. The

conclusions given in references (14) through (20) are

summarized in Table 1.

MacKenzie, Bloch, and Wiswall (16) performed radi-

olysis experiments on a series of highly purified cyclic

fluorocarbons--C6 F, C12F10, and C10 F-- and on their satu-

rated analocs--CF12, C12-F-, and ClF18,. The materials

were irradiated to doses of the order of 10 rads in

nickel cells usinc a 1.5 Mev. Van de Graff electron

beam. The irradiations were carried out at temperatures

chosen so that the materials were in the liquid phase.

It was concluded that the radiation stability of the

aromatic compounds was less than that of the correspond-

ing hydrocarbons; whereas the alicyclic fluorocarbons

were more stable than their hydrocarbon analogs. The

total C values for destruction of starting material

were very similar for aromatic and alicyclic compounds

and ranged from 1.3 to 2.4. Considerable amounts of

eases and products of molecular weight lower than the

startinE material were found in the radiolysis of

allcyclic compounds; whereas polymeric material was

almost exclusively yielded in the irradiation of aro-

matic fluorocarbons,with only traces of gaseous and

low molecular weight compounds being formed. No free

fluorine was detected in any of the irradiations.

In an extension of the aforementioned work, Bloch











4- Zc 4
9 4e) I
0*H *H 0)
M .0 1a *C 4
C o 0 C) H
.0 .0 V bO 0 -rl
C 0 C U 0o o
4 0 4. 0 0 C
) c S-, a) H- cd 4-

C) o0 '- U) )
C3 H H O (0 1 C O
C .0 0 . C O
4- O 0 CI *H $-
C o CO .0 4-' Cu
cO 0 O4-
o n o- n O o o
0 r *4-C 0 "C 0
S 0 C Cto C 0 i<

O 0 0 *T S O -H -
N I IL . I D 0) a)
Q O O H O C *
0 co 0 H a 0 4 O
. 4 0 0 C 4- 0 C .0
O 0 0 E + O 4-) 0 O:
4 )3 co $. -1 O r Uo
S O CA 4 O 0 C4 0 0
o U o r C 4-) t C .0 r-1
C) 0 m 0 Cu 0* 0 0 Cu

Cu i 0 0 (C 0 t O a 0
S4) -1 -H u S -r ) c
0 Cu to 0 -0 r Cu
CI A,- C Cu Cu Cal 4 Cu o 0 x0
0< 0 C K O < 0 4-'









m 'co
Cl)

O cT





S o 0 O
'co r


to 4-a r..
M E -E N
l OT O '0 0 H O N
0- > "A O C N
) 2 N H ) O N M
2 M- M H CL) I H O

O 0 U C 0 C
S O O0
Q C to N
h- E \O lB M C O M O
O 0 CC b-,
0 '0 O \0 0 )U

0
C 0

CO


6 0
K u -\\
M
fa

ca \
































D.l
co
0



U

0





(I
0
CC
t-a

g-l
C

0

!


-a

CL) 4.)

m



Cu





E*
<

















I,
01












L,
r
a0


to


rf C





LC- CL'


S.-- 0


0 0








O C
to
















t-a,
c

















E
u0 c








0 01
a La

mE
o a
o C














U 0


0I
O

0
r-l






C
ca
'-.



,L.

r.

I-=






C U
-f-I
-I





-
0r



CA-





























O

H
0
1-1




CO
0D



0



4:

I
0
C-)







c CO

CO



C)~


*a


,o




o
O

'
Ol *

4: 0


\D
o


O .


4)
C

ho 0
Li C
U. z
03 0

0 .
.0 .4
(t

t3 0
0* 0

o CU
u C
n)
cd 0
So
m -p
C (d

0 r












Oc to
S 0 --
r N I


<> c!




O C C








and MracKenzie (1?) used a Co(- source to irradiate a series

of fluorocarbons at elevate.d temperatures generally y ahout

4500 C.) to study the combined effects )f thermal decompo-

sition and radiolytic decomposition. The results indicated

that for the aromatic series there was no marked increase

in decomposition with temperature, however, a Jose of less

than 10- rads at L,5', almost completely destro',yed C1 F',

(perfluorobicyclohe:.--'l). ,rhe only alicyclic 'luorccarbon

studied. It was noted that C1F-,-) mi-ht not be typical of

alicyclic fluorocarbons (i.e., the bond joininri the rinrs

seemed especially vulnerable), but it was supposed that

the results indicated that, as a class, the alicyclic

fluorocarbon would not stanJ the combined effects of hith

temperature and hich radiation Jose.

The performance of C12F,, (perfluorobiphenvl) was

particularly rood in that its radiolvtic decomposition was

still reasonably snail at 5.0:' C. For comparison,the hy-

droarb.on terphenyls hecan to show drastic radiolytic

decorrmposition at temperatures over 4,50: C.,and ar. 500 C.

the thermal decomposition rates were of the order of 5.

per cent per day.

The radiolysis of C-F. usinr a gamma source has been

studied by h:evan and Hamlet (19) who made the following,

comparison between the radiolsis of saturated fluoro-

carbons and hydrocarbons.

(a) C-.,F was about one-fifth as sensitive to radia-

tion decomposition as was C,-H,. The G value for C-,Fe








was 1.9 while the G value for C2H6 was 9. The causes for
this apparent radiation stability included the absence of

F abstraction in fluorocarbons, the possibility of back

reactions of free radicals to reform C2F6, and the absence

of excited molecule decompositions which would have given

molecular fluorine. Molecular ejection of hydrogen from

excited molecules was shown to make important contribu-

tions to product formation in the alkanes. It was

suggested that similar excited molecule reactions in

perfluoroalkanes were unimportant.

(b) Hydrogen was unreactive in alkane radiolysis;

whereas fluorine was a reactive product in perfluoro-

alkane radiolysis. This dissimilarity may have been

caused by the great difference in bond strength and by

the exothermic reaction of fluorine with perfluoroalkyl

radicals.

(c) Linear alkanes were characterized by large yields

of hydrogen while perfluoroalkanes were characterized by

large yields of CF .

(d) The radiolysis of linear fluorocarbons may

produce cyclic compounds; however, no cyclic compounds

are produced in the radiolysis of linear hydrocarbons.

The radiation chemistry of C6F12 and c-C F8 was in-

vestigated by Fallgatter and Hanrahan (19). The gamma

doses used were from 0.2 Mrads to about 8 Krads. The

purpose of using doses lower than those used by previous








invest iLators was to study the initial products and the

initial product yields. At low doses dimeric fluorocar-

bons were found to te the predominant products of the

ra~nia radiolvis of CAF%2. The initial C value for CF 12

for products lighter than the parent compound was Eiven

as C0.', and the '3 value for products heavier than the

parent compound was civen as 3.1. The latter '3 value was

broken down into C = 2.2 for C12F2- (perfluorobicyclohexyl)

3 = 0.54. for C,1 F,2 (Ferfluoroc'clohe:-,lhexene), and C =

0.22 for ,-122 (perfluorocyclche:'lIhexane) These total

C values compared favorably with those given by MIacKenzie

et al. (1 ) of C0.. for lighter products and 2.1 for heav-

ier products. Portions of the analysis for this work

were performed on a F"as chromatocraph-mass spectrometer

combination. The nethod of separating a specific ras

front' the radiolysis products by means ,of the r'as chroma-

tograFh and then injecting the separated cas into the mass

spectromieter for analysis was discussed.

Although Fallcatter and Hanrahan felt that extended

speculation about mechanism was premature, it was

possible to make some generalizations. The treat strength

of the C-F bond compared to the C-H bond and the weak-

ness of the F-F bond as com-pared to the H-h bond were two

of the major differences between the fluorocarbons and hy-

drocarrons; however, since more than .?0 per cent. of the

observed hond rupture in the radiolysis of liquid CF12

involved C -F rather than C-C bonis, it was apparent








that the greater strength of the C-F bond was no deterrent

to its rupture. It was pointed out that C6F12 was

probably not a typical compound since C6H12 showed far

less C-C bond rupture than hydrocarbons in general.

Even so, the assumption that the radiation chemistry of

perfluorocarbons was centered in C-C bonds rather than

C-F bonds was to be avoided.

Abstraction from the fluorocarbon substrate by F

atoms to make F2 was essentially impossible because of

the weakness of the F-F bond. Consideration of this

factor alone showed that the radiation chemistry of fluoro-

carbons should be much different than that of hydrocarbons.

The lack of C6F10 in the radiolysis products tended to

confirm the lack of fluorine abstraction. Most proposed

mechanisms predicted the production of fluorine and
C 6Fo, and even though the fluorine might have reacted with

other products or the walls of the system, the C6F10

would not. A consideration of the reactions to produce

fluorine and C6F10 showed that the reactions were endo-

thermic and were quite unfavorable thermodynamically.

Reed, Mailen, and Askew (20) irradiated a series of

pure fluorocarbons using three different radiation

sources. Fluorocarbons ranging from CF4 to 2,3-(CF3)2C Fg

were irradiated in a Co60 gamma source, in the Low In-

tensity Test Reactor at Oak Ridge National Laboratory, and








in the Oak Ridge Graphite Reactor. In addition, the

radiolysis of mixtures of CFL and C2F6 was studied.
The gamma irradiations gave G values for the fluoro-

carbons which were considerably less than those of the
analogous hydrocarbons especially for the lower members
--CFL, C2F6, and C3F3. The highly branched structure

2,3-(CF3)2C F8 was the only exception, and it had a
total G value comparable to those of the hydrocarbon
alkanes.


Radiation Chemistry of Uranium Hexafluoride

The radiation stability of UF6 must also be consid-

ered. The use of UF6 in gaseous diffusion plants

necessitated studies on the critical sizes of various
configurations containing UF6. Snell and Rush (21)
determined the multiplication factor for 300 lb. product

drums containing 1.1 and 0.536 per cent U235. Grueling et al.
(22) studied the critical dimensions of water tamped

slabs and spheres of UF6, and Bull (23) calculated the
critical mass of a UF. core with reflectors of D20, Be,
and C. However, all of these efforts were concerned

with potential criticality incidents and not with the

use of UF6 in a reactor as fuel. Bernhardt, Davis, and
Shiflett (24) irradiated UF, with alpha particles and

found that intermediate solid fluorides and fluorine








were formed. It was also found that these products tended

to recombine,giving a net G value of one molecule of UF6
decomposed per 100 ev. of energy absorbed. Dmitrievskii
and Migachev (25) reported that the neutron irradiation
of UF6 yielded UF5 and free fluorine with a G value of

0.5 molecule of UF6 decomposed per 100 ev. of energy
absorbed in the gas. An equilibrium was found to exist
between the UFg and the UF5 and fluorine products with
the equilibrium concentrations being a function of the
dose rate. A further significant result was the fact that

the decomposition of the UF6 was completely prevented by
the addition of 25.5 per cent fluorine to the UF6. This

result had been previously predicted by Goodman (26).
Thus it appears that the radiation stability of UF6 may
not be good, but it can be brought up to acceptable

levels by the addition of fluorine to the UF6. In fact,
the operation of a UF6 reactor has been described by
Kikoin et al. (27) in which the decomposition of the
UF6 was controlled by the addition of chlorine tri-
fluoride to the fuel to serve as a source of fluorine.












CHAPTER III


EXPERIMENTAL PROCEDURE


Distillation

Approximately 50 liters of COF,6 were obtained

from the Oak Ridge rational Laboratory. No information

was available on the previous history of the CRFl6 or

on its purity. Chromatozraphic analysis of the crude

CgF16 indicated that there were several impurities as

shown in Figure 1. In addition, the material had a

yellowish cclor and a distinct odor which are not

characteristic of fluorocarbons. The material was also

analyzed using nuclear magnetic resonance measurements

and infrared spectra measurements. The results for

these analyses are shown in Table 2.

The impurities in the CpF16 were removed by a

batch-type distillation using a distillation apparatus

consisting of a 2.5 liter pot with a packed column 100

cm. hich. The column had a diameter of 1.5 cm. and was

packed with stainless steel helixes, and two indepen-

dently controlled electric heaters were used to maintain

the column temperature at the desired levels.


































0
4-) i
C.r
4 ~ ~
to


CL.




C-,
0





H a0


E
m0






too

N






0
\ E c.














0H
r-is






4-))
4-4











x0
I~' I I




r-1l

c4)-.

r(4

0 C..













TABLE 2


RESULTS OF ANALYSES OF COFI,


Material

Distilled C;?F


Crude CFl6


Infrared ;Measurements

Wave Length
(microns)


;. 7


C-H BonIs
(Relative units)

1.5

- ,'


Nuclear Magnetic Resonance Measurements

1o. of Hydrogen IHydrogen
Material Atoms per molecule (Wei-ht per cent)


Distilled CF'


0. 01


0.25


Material

Distilled CF16


Distilled CdF1,


Elemental Analysis

Hydrogen Fluorine
(ieicht per cent) (Weipht per cent)

2 .12 76T.16


2L.01 f7.94









A 1.5 liter batch was charged into the distilla-

tion pot at room temperature, and the pot temperature

was raised until the charge began to boil. Concurrent-

ly the column temperature was increased to 1000 C. for

the upper half of the column and to 1040 C. for the

lower half of the column. The product was taken off

at a rate of about 10 drops a second until the tem-

perature of the product reached 95.00 C., and the pot

temperature increased to about 1150 C. The product

taken in this temperature range consisted of two im-

miscible liquids. The lighter phase was yellowish and

had an odor somewhat like benzene while the heavier

phase was colorless and had an almost undetectable

odor.

The next cut, taken between product temperatures

of 100.70 C. and 102.1 C. at a flow rate of 4 drops

per second, was the purified C8F16 which was used in the

experimental work. The pot temperature increased from

1190 C. to 1220 C. during this cut,with approximately

1.0 liter of purified material being obtained from

each 1.5 liter of raw materials charged to the pot.

The distillation significantly decreased the amount of

impurities as shown by the chromatogram in Figure 2.

The analyses of the distilled material are given in

Table 2.




































































Pt





x I'


4, ~




.2 -


LL

C:












r,

t'



C,

0


F- L.
.C





33

LC.

I-










C









Fluorination


Since UF6 was to be used in some experiments, it

was necessary to fluorinate all surfaces which were to

be exposed to UF6 with fluorine gas in order to form a

protective fluoride film. This film prevents the reac-

tion of the UF6 with the metal surfaces to produce UF4

which would adhere to the surfaces and cause difficulties

in sample transfers and in material balance calculations.

The vacuum system shown in Figures 3 and 4 was fluo-

rinated using fluorine gas from a 1-pound fluorine gas

cylinder obtained from the Matheson Chemical Company. All

components except the bellows seal valves of the sys-

tem were carefully washed in carbon tetrachloride in

order to decrease all surfaces which were to be in

contact with fluorine. The bellows seal valves were

Nupro type B-4H bellows valves which were specially de-

signed for systems in which no leakage could be permitted.

These valves were cleaned by the manufacturer. The sys-

tem was pressure tested three times with a nitrogen

pressure of 60 psig. and vacuum tested five times for

periods from one to twelve hours. For fluorination, the

sodium fluoride traps were replaced with sodium chloride

traps. The outlet of the sodium chloride trap used for

capsule analysis was led to a glass bubbler containing

a solution of 20 per cent potassium hydroxide which was

vented to the hood.

















a i



L






I j I
Q-


0


12

E
0 L
s: L
'-i 2-


< ',
.-0


- I


3E

> .
cV) fj
- ^ -- tx -

r-'


- -

o
Cua ar





4rr,














co





L.
L ,


L



















am
T-
o -
2 G
a -
C r-
















L *l l r


L.'
E-C



o -'-4

U,
- -I
c?


CuL-
















- a a -- -f


E

4.)










cl







0g


4-)









The fluorine gas was introduced into the vacuum

system through a bellows seal valve on the vacuum rack.

A specially designed fluorine regulator (Matheson model

# 15F-670) was used to control the flow of fluorine.

This valve permitted the addition of small quantities of

fluorine to a carrier gas which in this case was nitro-

gen. A sodium fluoride trap was placed between the

regulator and the vacuum system to remove any hydrogen

fluoride which was present. The sodium fluoride trap

was heated under vacuum for one hour with a Eas-air

torch to activate the surfaces of the sodium fluoride

pellets. A tape heater was used to maintain the trap

at a temperature of 100 C. during the fluorination.

The fluorination operation was commenced with es-

tablishment cf a steady nitrogen flow. The rate of flow

was adjusted so that the flow of nitrogen in the bubbler

was as large as possible without causing the potassium

hydroxide solution to bubble out of the bubbler. The

nitrogen pressure on the system was about g psig.

The fluorine regulator valve was opened so that the

fluorine pressure between the fluorine regulator and

the fluorine control valve was about 12 psi. The

control valve was then opened for one-second intervals

approximately every two minutes for thirty minutes. The

interval for which the control valve was open was slowly









increased during the next hour so that at the end of

the hour the interval was fifteen seconds. The fluorine
control valve was then opened for intervals of one min-

ute every five minutes for thirty minutes. During the

operation, the sodium chloride trap temperature was mon-

itored with a slow increase in temperature being observed.

The system was purged with nitrogen and allowed to stand

overnight. Final purging was accomplished by removing

the glass bubbler, filling the entire vacuum system to

50 psig., and then allowing the nitrogen to blow off

quickly.

The vacuum system was then evacuated so that the

nitrogen pressure in the system was about 0.5 atmos-

phere, and the glass bubbler was reconnected. The

fluorine control valve was then opened, and the pressure

in the vacuum system was brought up to about 1.0 atmos-

pheres. The system was then completely isolated and

allowed to stand for thirty minutes. A very low nitro-

gen flow was established for thirty minutes after which

the pressurizing and blow off procedure described above

was repeated.

A system which could be used to fluorinate capsules,

valves, and other fittings was built as shown in Figure 5.

All items were carefully degreased with carbon

tetrachloride,and, in addition, the capsules were immersed

























CL


j. E

!-11


L






OJ
0E
Cl)









three times in boiling water to remove any remaining

silver solder flux. The capsules and fittings were

fluorinated in the reaction vessel while the valves were

fluorinated by being inserted into the nitrogen-fluorine

stream between the reaction vessel and the sodium

chloride trap.

The system was pressure tested to 50 psig. with

nitrogen for one hour. The pressure was quickly re-

leased with the glass bubbler disconnected to purge air

from the system. The pressurization and release operation

was repeated three times. A steady flow of nitrogen

was established with the bubbler in place, and the

fluorine pressure on the regulator outlet was raised

to 12 psig. Fluorine was introduced into the system

by opening the outlet valve for a period of one second

at intervals of one minute. After thirty minutes the

period during which the outlet valve was opened was

increased to five seconds with the intervals that the

valve was closed still being one minute. After another

thirty minutes fluorine was introduced for a one-minute

period, and the system was completely isolated and

allowed to stand for one hour. The purge was then re-

established for fifteen minutes. The glass bubbler was

disconnected, and the previously described pressuriza-

tion and release operation was carried out three times.










Capsule Fabrication


Capsules for Ganmma Irradiation3

The capsules for the pamma irradiations were fab-

ricated frru a ?-in. section of one-half in. copper

tubing. One end of the capsule was sealed h:, crimpin1t

and silver solderinr the copper tubing, and the other

end was sealed with a brass *'ite.: valve (1-VNA)

attached to the tube with a brass Swazelok fitting-.



Capsules for Reactor Irradiations

The capsules for the reactor irradiation were

fabricated from a 5-in. section of one-half in. alumi-

num tubing. One end of the capsule was sealed ,,by

crimping and heliarc welding the aluminum tubing,and

the other end was sealed with an alumiinum Whitey valve

(I-W'Al) attached to the tuting with an aluminum

Swagelok fitting. All reactor capsules were tested wi.h

a nitrogen pressure of 300 psic. All components f

the capsule were decreased with carbon tetrachloride

and fluorinated.








Sample Preparation

Preparation of C8F16 Samples

The C8F16 samples were prepared using the vacuum

system shown in Figure 3. The capsules were attached

to the system and both pressure and vacuum tested. The

calibrated burette was filled with C F16 which had been

degassed by vacuum transfer from one glass bulb to an-

other glass bulb for a total of three times. After each

transfer the dissolved gases were pumped to vacuum. The

C8F16 sample was vacuum transferred to the capsule, and
the size of the sample was determined by difference on

the calibrated burette.

A gold wire weighing 1 mg. was wrapped in tissue

paper and attached with tape to each capsule which was

to be exposed to reactor radiation. The gold wires had

previously been weighed to the nearest 0.001 mg.


Preparation of CgF16 UF6 Samples

Samples containing CgF16 and UF6 were also prepared

with the vacuum system. The UF6 was obtained from the

Allied Chemical Company and was found to contain some

dissolved air. Most of the air was removed by vacuum

transferring the UF6 from the storage tank to the cali-

brated tank and back to the storage tank. During each

transfer the UF6 was sublimed and recondensed,with the









dissolved air bein7 partially released. After each

transfer,the tank containing the UF6 was pumped to vacuum

while liquid nitrogen was still on the bottom of the

tank.

The volume of the calibrated tank was found by

filling the tank with water in a constant temperature

bath (250 C.) and then weighing the tank and water.

Subtraction of the empty tank weight from the full tank

weight gave the i.ei,:ht of the water in the tank. The

weight of the water was then converted to volume using

the density of water at 25' C., and the volume was found

to be 758 cc.

The calibrated tank was filled with UF5. by opening

valves as necessary, anJ allowing the system to equili-

brate for four minutes at room temperature. The valve

on the calibrated tank was then closed, and liquid

nitrogen was applied to the nipple on the storage tank

to recondense all UFP gas in the system other than that

in the calibrated tank. Using the vapor pressure of

UF. as given in Reference 28 for the observed room tem-

perature, the weight of the UF6 sample was calculated

by the ideal gas law. After five minutes the manometer

valve was cracked to determine that all of the UF, had

been condensed in the tank, and the valve on the storage

tank was then closed.









The UF6 was transferred to the capsule by opening

the capsule and the calibrated tank valves and applying

liquid nitrogen to the bottom one-half inch of the cap-

sule for two minutes. The manometer valve was opened,

and the residual gas pressure (generally about 3 mm.)

was checked to ascertain that the transfer had been

completed. The calibrated tank valve and the capsule

valves were closed, and the system was pumped to vacuum

with the liquid nitrogen still on the bottom of the

capsule. The vacuum valve was closed, and the cali-

brated burette valve was opened so that the C8F16 could

be transferred to the capsule. After the desired

amount of C8F16 had been transferred to the capsule, the

capsule valve was closed, and liquid nitrogen was applied

to the calibrated glass burette. After the manometer in-

dicated that the CsF16 had been condensed, the calibrated

burette valve was closed, and the burette was allowed to

come to room temperature. The amount of C8Fl6 trans-

ferred was then determined by difference.

A gold wire weighing about 1 mg. was attached to

the capsule as discussed in the previous section.








UFTR Irradiation

The reactor irradiations were performed in the

University of Florida Training Reactor which has been

described by Boynton (29).

The plastic handles on the capsule valves were re-

moved, and the capsules to be irradiated were inserted

into an aluminum cannister which was 10J in. long with

a 3-in. outside diameter. A 1-ft. section of a graphite

stringer with a L-in. x i-in. cross section was removed

from the thermal column of the UFTR. This stringer was

immediately adjacent to the core. The aluminum cannister

was inserted in the 1-ft. void, and the remainder of the

stringer and shielding blocks was reinserted.

Upon removal the cannister was monitored and stored

in a fuel storage pit until the radiation level had

dropped to levels which permitted transfer of the capsule

to the vacuum system.



Gamma Irradiation

The gamma irradiations were performed in the Univer-

sity of Florida pool-type food irradiator. The 30,000

curie Co60 source was in the form of 96 plates which

were placed so that a relatively flat gamma flux was

obtained. The capsules were lowered into the containers

within the irradiator and exposed to the desired dose.









The containers were filled with water. Only one capsule

of the entire series of gamma capsules was found to have

leaked water. The gamma dose was measured by the Fricke

dosimetry method (30).


Analysis

The analysis of the samples was performed with a

Model 700 F & M Chromatograph which has a thermal con-

ductivity detector. The conditions used for the various

analyses and the associated calculations are given in

the appendix.

The capsule containing the sample to be analyzed

was attached to the vacuum system. The vacuum system

was both pressure and vacuum tested up to the capsule

valve. The sample was then transferred to a glass bulb

in the vacuum system using the vacuum transfer technique.

The samples which contained UF6 were passed through the

sodium fluoride trap to remove the UF6 and fission frag-

ments. If the sample transferred to the glass bulb was

discolored, the sample was transferred back through the

sodium fluoride trap to the capsule. The sample was then

transferred once more through the sodium fluoride trap

to the glass bulb so that the sample was passed through

the trap for a total of three times. No sample required

more than three passes.









The bulb was surrounded by a mixture of ice and

water and allowed to stand for one hour so that equilib-

rium could be established. The gases over the sample
were then chromatographed several times using the

chromatofraph gas sampling valve with either a 1 milli-

liter, a 2 milliliter, or a 10 milliter gas sample loop.

A silica rel column was used for this analysis.

The volume of the sample was measured in the glass

burette on the vacuum system. The sample was then trans-

ferred to the 12 liter flask of the glass vacuum system

and vaporized. The lrass vacuum system is shown in Firure

6. Chromatographs were made of the completely vaporized

sample using a silicone gum rubber column. Finally, the

sample was recondensed, and liquid samples were chromato-

graphed with a silicone gum rubber column to determine

the amount of polymer.



Neutron Pulsing

The experimental equipment used for the neutron

pulsing is shown in Figure 7. The neutron generator

was a Cockcroft-'''alton type accelerator, Texas Nuclear

Corporation Model 150-1H. The neutrons were produced

by bombarding the tritium target with deuterons. The

enerr.' of the neutrons was about 14 Mev. The neutron

source was roughly isotropic and in this experiment was


























02


m



E e


El
0 u





O d

J ^



0
]P tfl
U1 0
























r4

t- L




L- C~
-c c
0-


c


C
L.



.&-
a,

ii













In
I-',
CT



L ..















Ct









placed under the tank containing the fluorocarbon. The

tank was lined with cadmium to prevent neutrons which

leave the system at higher energies from returning at

lower energies. The detector was placed just inside the

upper boundary of the system. The detector was a LiI(Eu)

scintillation detector which was designed to detect

thermal neutrons. The signal from the detector was fed

into an amplifier and then to a scaler where the total

counts were recorded. The signal was also fed into the

Technical Measurements Corporation, Model CN-110, 256

channel time analyzer.

The experimental procedure was as follows. A pulse

generator produced a signal which would initiate a steady

neutron beam. After the neutron beam had reached steady

state operation, the neutron beam was turned off, and the

decay of the neutron population in the tank was recorded

on the 256 channel time analyzer. The neutron pulse

width was 50 microseconds and the channel width for the

analyzer was 2 microseconds. Thus, the first 25 channels

showed the neutron pulse, and the remaining 231 channels

recorded the decay of the neutron population in the

fluorocarbon. After the 256 channel time analyzer had

finished counting, a new pulse was introduced, and the

cycle was repeated. The pulse repetition rate was 980






40


cycles per second. This procedure was followed until
sufficient counts for a reasonable statistical analysis

had been obtained. The moveable top was then changed to

a different height, and the volume of the fluorocarbon

was adjusted so that the fluorocarbon liquid level was

at the moveable top. This adjustment cave the system a

new geometricc buckling. The pulsing was then repeated

at the new liquid level.












CHAPTER IV


EXPERIMENTAL DATA AND RESULTS


The experimental data and results are tabulated

in Tables 3 through 8. Table 3 presents the data for

the gamma irradiation of CgF16 in the Food Irradiator.

The following explanation is given for the various side

headings. The doses, in terms of rads, were determined

by Fricke dosimetry. The per cent C8F16 converted to

gaseous products is the percentage of molecules in the

original sample converted to any gaseous product; like-

wise, the per cent CgF16 converted to dimer is the

percentage of molecules in the original sample converted

to any dimeric product. The calculations for these

percentages are shown in the appendix. The per cent

C8F16 converted to dimer chromatographh) gives the per-
centage of the original sample converted to dimer based

on the chromatographic analysis assuming each molecule

of dimer to be equivalent to two molecules of starting

material. The per cent CgF16 converted to dimer

(material balance) represents the percentage of the

original sample converted to dimer based on the differ-
ence in the volumes of the original and irradiated








samples.

Table L shows the G values for the gamma irradia-

tion of CFl The C value is defined as the number of

molecules of a given material produced per 100 ev. of

enerrm absorbed in the sample. For example, for Sample

Io. .3, G(CFI) is 0.129, and accordingly, 0.129 mole-

cules of CFP are produced per 100 ev. of energy absorbed

in the sample. The G dimerr) values are given for the

amount of dimer produced as indicated by both the

chromatographic analysis and the material balance loss.

The disappearance of CgF,l, is -iven as G(-CsF,) which

is the number of molecules of CF16 lost per 100 ev. of

enerrv absorbed. The calculations for the G values are

rPiven in the appendix.

Table c presents the data and results for the

reactor irradiation of CF1,' in the UFTR. The method

for calculating the integrated flux is given in the ap-

pendix. The per cent CFi6 converted to gaseous products

is the same as in Table 3. An aJditional sample, No.

11, was irradiated with an integrated flux of 59 x 101L

neutrons/cm.2. Liquid samples were taken directly from

this irradiated sample for dimer analysis,and the per

cent of the original sample converted to dimer was found

to be O.40 per cent. This sample was the only one in









this series that was checked for dimer. No gas analysis

was made.

The data and results for the reactor irradiation of

C8F16 and UF6 are presented in Tables 6 and 7. The cal-
culations for the data are given in the Appendix.

Table 8 presents the data and results for the pulsed

neutron experiment. The symbols used in Table 8 and the

method for calculating the decay constants are described

in Chapter V.

The range of error for the decomposition percentages

given in the tables is estimated to be + 10 per cent.

The range of error for the G values is estimated to be

+ 15 per cent.























0
O
?-I
E-

1-i rr-
c: O
E--4 ) 0

0 0,
l4 *J O 0

Cr -
U- r"f *. x 0


E-,^
SOU


O





4. C4
0






U0 rI

ii r-. O1 0 I 0

rrC 0 0 0









00 E3
09) O cu
O L- O 0 ,

O 0 O C
C ty i C ~0
m 0a






















Sr) U r S a
S 0 E r. X 0U
cc 0 o C
Lf>bz. r 1 -2 0



I-I















Ca -. ^ \- *-. ^















I,-
0

\o X Ot
to O O-
H *-
CV 0



0



-
Ni












0
rl









0 H 0
O










a oO
o I o
3u\ 0 r m
S.0 0 X 0 ?'

C tO 0 l
C0 \0
















H 0 0 0 o S4









H H H
W t t
iO- O














S0 0 0 ;4








O< o o o e
0a H '0 ^






K H -F '0 i a


CV0 0 0 C
OT ca 'e^ 's. U- '
















TO--1
to
C,








1-4











ro







0
-r ri











0- -a
rt










Wt


en

-





c:
CI
.-' CU
U C



r 0 t-
G .3 -'
'* .- .




Dl 0t'

0-


rj ~ U- L.

-0 ri (







N N en to

C-. '-4 -1


o rl





cl -4 N-




o -O~


I I
I I


C O
.*0 0 Ta to
Cr. C. ~ C4 C.
C.) 0 i C
0000U









TABLE 4


G VALUES FOR GANMA IRRADIATION OF C8716

SAMPLE NO. 33 38 39

Dose radss) 2.4 x 107 4.3 x 107 4.3 x 107


G Values

CF4 0.129

C2F6 0.020

C2F60 0.004


C3F ---

C3F0 --

G (Gas) Total 0.153



G (Dimer)

Chromatograph 0.40

Material 1.00
Balance


0.176

0.017

0.004

0.001

0.001

0.199






1.57

1.68


0.197

0.037

0.002

0.001

0.002

0.239






1.60

1.68


34

5.9 x 107




0.109

0.017

0.002

0.005

0.005

0.156






0.55

1.12


2.21 2.75


3.25 1.88


G (-c8F16)













SAT'LE IN0.


Dose radss)


'3 Values




CFn
C F,.
C2F0

CIF,

C 3FO

C (Cas) To~tal




'3 (Dimer)


Chromatoc-raph


TI.aterial Balance


G (-CFI,)


TABLE 4 --Continued

:. 12

h.P x 10 12.2 x 10'


O. 1O



0.001


0.001



0. 20,







1 -.L


1.99


0 1L


0.000.


0. O01

0.0007


0. 000L.

1'.1, ;










1 3


2. 2.. .F


13


1L.2 x 10


0.1"2


0.007


0.000 O


0.0006


0.000.?

0.133







1.12


0.9,7




1.7.


























H
E-






0
E-i




O
co o

t0
Ei-t


<0
HO
CO


0
a


















H



0
Ior

























0
03 M











a:










Q


u r-t
-l 0 uH




cv-



(\






100 0

r^j


x


NO -
~- 0
x O
0


--t

. H
CV
x


r-i -4t
r 0 to -0
S C \ 0


0 X


0\
oI -


to-t
f\ 0
0 O
C; x


S--t -.o
0 0 0'-A

00
0s


s o
0)

0 1 01


S0 0 o0

o H \o w0
4- X- -- co
























(1 0: C' I
tocI
















C- C C











1 ?:







U- 0 0
Ui 'I I I

'1


S-i

6)


'0
cr u
O c



if 0 It



en vC


U ---


c

C '4- '0
CjC,



















0
i O0
N


0
H







3H





0 QM
E-H




0









00


OH

CQ






M0
a
E<<





a \o
o





















O
00
C::4
^0 MS
H~ 0 1-










wC

MM
&
M&
E-
1-
oT
0,


0
o
r O











o O




0 0




0
H 0
C.-'


-~ 0
* '
H


*-r-
1-s
0


\0 x












0

,4-




,-

O

rr

0
0


0 -.4 C9-. O0
"a o

'o IV >
u 0c 0
00 0 0 r
0 M 0 4 0
S cp o o
0 O %
)O 4. 0 C 0o
)bf to r0t
Hc 3 e h 0 ) 0
0 U I) *0 0
co o f; u Jc c o
OT > ; ft; M *- *e U








5'i














C- .9 rr .r.






-,j C- 'C 2` C-





-2 'Q 4 ~ C













-- C- .-i r i -C
't 1 r


'-4f


I I I
I I


301-
4,

UC

"O *r-4
Lr)r-



. 3 C 0
dL. to
-a L. Cl. Cr. Cl.




0 'l to
tC Cc
u















r-4
0











0
,-I



















N -
tto

X
.-I







,.-t




0- -1


zx
0 0







00
n i--
W O-




0

E-' X

rlmc
c 0
0 r
0
< t


\0 N um 0 C- 0
\0 r\ M 0 0
0 0 0 0 0 0







N m \0 CO N
'r^ CT\ '0 i- UN '0
C- 0 0 0 0 '-4








0 ( 0 00 '0
O O O O H


tO O

c- 0


I 1 -0
O


0 0
0 o I'

f. 4L :3 \0C o to to 6
- hO -b. 3 O to
C r N Nr ,N m m 0
4 > C 0 0 0 0 U
< C -
CO H 0


N
















-2





'*'
I-I







S C
f-




C-






I-
r-



C C
-4






C
0



I

E--4



1'

C-
1-1


-3 CO 5 03 (
o o C c -I











=4 0 10 C. r
-4 -4 "I -- C











-4 r C* r


-- 0 0C 0 0C










- -4


7rC


-cc
U

C. t' 1
)c n A.


.- L m 4 L. CI. C. [t
4 l U C* U (2 '
< CC
U, (2'

































0.
0





rrr:
Co



o-



c4














~0
1HZ


u C

1-Io

04a
0
C-,


-k -d' -Zl
0 00
H- H H





0~ '.















tol 0 0 0








rm
m 0 0 0









C N)
C'.) oN





c'-l H H




C.) '. '0 '.
o o




('4 () (N












CHAPTER V


DISCUSSION OF RESULTS


Radiation Studies


Gamma Irradiations

The gamma irradiation of C vFo1 yielded two groups

of products. One group was composed of gaseous pro-

ducts. and the second croup was composed of heavy

polymeric products. The ratio of CAF1, converted to

polymer to CFi16 converted to gaseous products was

111 to 1. This value is the average of the data given

in Table 2 for per cent CgFI1 converted to dimer and

per cent CF1,6 converted to gaseous products.

The principal gaseous products of the gamma irra-

diation of CoF16 are seen from Tables 3 and 4 to he CFL

and C2F6 in the ratio of about L to 1. Smaller amounts

of C2F6,O and CF0 were also found,as well as a compound
tentatively identified as C3FgC'. Identification of

C3FCO was based on the chromatographic retention time of

a similar compound as reported in Reference 20.

The polymeric material was found by chromatoRraphic

analysis, and based on mass spectrometric analysis using








the equipment described in Reference 19, was identi-
fied as being dimeric. The dimeric material was
composed of at least two different molecules. The
first molecule had a molecular weight range of 750-
760, and one possible isomer is shown below. The
second molecule had a molecular weight range of 805-
815, and one possible isomer is also shown below.


CF3
CF
F2C CF--CF3
F2C CF
F2C CF


CF2
2


F3
CF
F2C F-CF3
I I


--- FC


/CF
CF2
2


Product, Molecular Weight, 750-760


CF-CF3
I3


CF3 CF
( 3


CF- CF --CF2- CF CF-CF----CF
S/2 2 2 3
CF2


Product, Molecular Weight, 805-815








The average C values for the production of gas-

eous products and of polymer are in good agreement with

the results of other investigators as shown below.

Compound C (Gas) G (Polymer) Investigators

C- .F0.3 .1 Fallrratter and
Hanrahan (19)
CAFA 0.3 1.1 Bloch, MacKenzie,
and Wiswall (i,)
C(F16 0.2 2.5 This study

A mechanism may be proposed which accounts for
the various products formed d-uring the gamma radiolysis

of CpF1A. A gamma photon strikes a COF,2 molecule and

produces a CF8F + ion and a F- ion. This reaction and

the subsequent reactions may be represented by the

following equations:

+
CgF16 4 CgF15 + F- (1)

C1 F + F- -> Cp F + F" (2)

CFo.lF + CyFli -1 > C1630

C F5 + F- -> CF + CF* ()

CF,- F" CFL (.)


CF1F + C F15- -)- C15F2


ClF,28 + 2F --> CF30









Primary chemical change occurs in reactions (1) and

(2). Subsequent products are then produced by the re-

maining reactions. Dimer is produced by the reactions

shown in equation (3) and in equations (4), (6), and

(7). The major gaseous product, CF4, is produced by
reactions (4) and (5). The other gases are produced
by various recombination reactions of the perfluoro-

methyl radical. The oxygen source for the production

of C2F60 and C3Fg0 was probably dissolved and ad-
sorbed air.

The production of gaseous products and dimer are

shown to be linear with respect to total dose by the

curves in Figures 8 and 9. The equation for the

curve in Figure 8 is

log -- Nx 100 = 0.025 + 0.53 x 10-9D



where NO N
-N-- x 100 = per cent of original sample con-
verted to gas, and
D = total dose in rads.


The equation for the

log P x 100

N N
where N0 N x 100
-N- 0


curve in Figure 9 is

= 0.29 + 0.91 x 10-7 D



per cent of original sample con-

verted to dimer, and



















N.







N








N


r- c



C-
L.










Cd
'-I0




cr-





0



L 0
0
0-
tc









-
Cr,
*- C


0 TO r.i
- 0 0 0 0


D
C 00

0r 0

*- o
i c% E3LI o
0cW J f


c,

C
0
C







61









4)
0 (



0 H 0


r:4
0 C
\rl (fl
4.() '









OU
\ 0
to
o (a)



00
l 0 H
\ cs

S


X H

I o\






0
t











o 0 0
O \ O


\ C,






C\)





\ I

*tO 'C -i- ('4 O TO -O (' 0




G -o



P-a C-) Q








D = total dose in rads.
The chemical structure of CpF1( shown below indi-
cates that there are sixteen possible ways for a CgF I

molecule to lose one fluorine atom.
CF
I 3

F2C CF--CF3

1 ]
FC, CF.
CF-
F'e rf 1luo rod imet h ylv.'c 1 ohe xane


Since two CFi molecules are required to form one

dimer molecule, there are 16 :. 16 or 256 possible ways
to form a molecule of dimer from two molecules of
CF1Fl. There are two ways for one CF-, group to be sep-

arated from a parent molecule, or,since two molecules

were considered in the dimer formation, there are 2 x 2
or 4 wayv in which a CF3 proup may be separated from
tw. CeF1, molecules. Thus the ratio of possible ways

to form dimer to the possible ways to form ;,as is 256
to I. or '-. to 1 as opposed to the eyperimentally found
ratio of 111 to 1.

The foregoinrn discussion has considered the decom-
position of CgF1. by ganjma radiation in terms of the

products yielded and in terms of the fraction of









original material converted to other materials. A
different analysis may be made by considering the

fraction of starting material remaining after a given

exposure to gamma radiation.

Let dN be the number of parent molecules decom-

posed by a dose dD on the sample which contained N

molecules. Since the number of molecules decomposed

is proportional to the number of molecules present and

to the dose, the proportionality can be given as

dNo. NdD,

and letting k be the proportionality constant, the

relationship becomes

dN = kNdD

where the minus sign indicates that dN is the number of

molecules lost. Rearrangement gives

dN = -kdD
J--

which can be integrated from N= NO to N = N and from

D = 0 to D = D. The result is

n N = -kD or


log N = -KD
0

k
where K =
2.303

The results for the gamma irradiation of C8F16 are

plotted in Figure 10 as log (N x 100) versus D. The
0











































/








//


,/


S






/

S




/


o *rr TC, Lr -7 ,^ *",j -i
C O' '- 0 o 0 0 0 0j


0



'I1


.-

1-



0
0




L




0
0
1o












0o -



X -I













L
sa'
n i













i-i
j
L.
t
C1.
0r x









curve is a straight line, and K is found to be 0.042

rad -. Therefore, it is possible to predict the num-

ber of molecules of CgF16 remaining in a sample with

a known number of starting molecules after the sample

has been given a dose of gamma radiation. Since the

dimensionless ratio N/NO is used, the units might be

grams or any other convenient unit which is related to

the number of molecules.

The question arises as to whether or not this re-

lationship holds only for CgF16 and only for the range

of doses used in this investigation. The data avail-

able from the literature at the present time for the

decomposition of fluorocarbons by gamma irradiation

are given in Table 9 and shown in Figure 11. Examina-

tion of Table 9 shows that the data used in Figure 11

cover a wide variety of compounds including alicyclic,

aromatic, and aliphatic fluorocarbons. Other variables

include type of radiation, dose rate, total dose, and

experimental conditions.

There are two distinct curves in Figure 11. One

curve relates data for aromatic fluorocarbons, and the

second curve correlates data for alicyclic and aliphatic

fluorocarbons. The decomposition of aromatic fluorocar-

bons is considerably less than the decomposition of

aliphatic fluorocarbons and alicyclic fluorocarbons,


























r







o,



a r
Cu
L 0 0-


t-



L.:
E -
t
E-





C r

+-'
4'-

> 'o
E- 'U
iX,


S- '


. C 0 0 00


c- *: )T I. .


C


' U


C C
c C
L L

ct c
Lt* (,
I t
-1 l


C C
C C.
L L
fJ ft
.' -'
t' O>
*- e--I
t* I.


C *'4 to
S ,O .' --i ''. ..
',. Er. L. Li' "--, a.
Ct Ci C r. C r, [r '.,










, .. .3 '
rl"-4 -, -7 C


2' C,
"


L
C, 03


C-

c.







67










0 l -. -t CI- Do
ON 0 0GN 0 0 0 0o







~to
0
4,- --t C Y QN to N N
07 N -rf -4 L' '0 N -4r
oX








0 X
4, L)
)N4)> -) 4J 4-) 4-) 4) 4)

0,

ONO
o B E E~c~la


40



E-4 4-1 dd t
Cd Cd o tv c




LJ, b.0 W bD hC O W3
co





\0 '0 \0 \0 '0 \-0 \0




0 rz t
0. t CC to
E 0 0 0UU
0









C-,
0 x
4, E
























































































0, 0 n: r- ,,



'I I-





rC-


0


-' ri ril -


J1
C
ri
.C
L

0







4,



0










Vl
L.








C,
r..





U)

C,
0










in



0:


C,-,
0








L


'1'
ci-









and the curve for aromatic fluorocarbons is a straight

line over the dose range for which data are available.

The curve for aliphatic fluorocarbons and alicy-

clic fluorocarbons is reasonably straight in the lower

dose range; however, the curve tends to slope downwards

at higher doses. This result is unexpected since it

might imply that the gamma radiation becomes more effec-

tive at higher doses. A better explanation is that the

number of small decomposition product molecules becomes

very large at high doses. These small molecules absorb

a large fraction of the gamma energy, but because of

their more stable structure, do not decompose. The

energy absorbed by the small molecules is transferred

to the heavier molecules, and the actual energy ab-

sorbed by the remaining parent molecules is greater

than that due to the absorption of gamma radiation

alone; therefore, the previously developed relation-

ship must be modified to take into consideration the

build-up of CF4 and other light gaseous products.

The number of molecules decomposed is proportional

to the number of parent molecules present plus the num-

ber of CFI molecules generated. The number of CF4

molecules present after some arbitrary exposure is

given by the following expression:









r!o. of CF, molecules present

no. of CgFI, molecules decomposeJ

1 molecule of CSFiP decomposed to ras
111 molecules COF1 decomposed
g molecules CF, produced
molecule ",F1 decomposed to ras

The number of molecules of 2FFl,. decomposed is simply

the number of molecules of CF1 originally present
minus the number cf C:F1, molecules remaining after

the exposure. If the number of molecules is iv'en in

terms of fractions with Ir-, = 1, then the following re-

lationship can be found for the number of CF, molecules
pre sent

r!o. of CF molecules present =

(1 N) < ~< x

.0;? (1 ) .

The original relati :nship becomes
jlu = .\ldD k [:.0-2(l I!]dD,

and it can be rewritten as


Sd = kdD
,.- + 0.0

Integration between the limits of II = I and N = !,-, and

of D = D and D = 0 ,iv es








1 0o.928N + 0.072
0.928 i 0.928NO + 0.072 = -
0.928k
By letting K = 02- k and NO = 1, the expression
2.303
becomes

log (0.928N + 0.072) = -KD .

Figure 12 is a plot of log (0.928N + 0.072) versus
D, and the constant K was found by a least squares fit
to have a value of 2.13 x 10- rad- for the aromatic
fluorocarbon curve and 6.64 x 10-10 rad-1 for the ali-
phatic and alicyclic fluorocarbon curve. Both curves

are straight lines and can be used to predict the
fraction of a sample which would be decomposed by a

known dose of radiation over the wide range of param-

eters given in Table 9.


Reactor Irradiations

Irradiation of C8F16.-- The reactor irradiations of

C8F16 gave results very similar to the results for the
gamma irradiation of C8F16. The principal gaseous pro-

ducts were CF4 and C2F6 in the ratio of about 4 to 1.

This finding is in agreement with data obtained by Davjs
for the irradiation of C7F16 in the Low Intensity Test

Reactor (31). Chromatographic analysis indicated that

there were also dimeric materials similar to those
found in gamma irradiated samples.





















/ -)
a / c




0 0
S/ L.




/ 00

/ -_

,/ ':*,
/' 0'L
C*







S / C.,/
'- f.- .r '


,, 'c. a :


SO- C C


,-. L 0

I,/ L i, C
l ~L,









", C-
I c









The per cent decomposition of the parent compound

to gaseous products is plotted versus integrated flux

in Figure 13. The curve is a straight line over the

range of integrated fluxes used. The equation for the

curve is

logg N x 107 = -13.2 + 1.1 log nvt


for the range 1012 < nvt < 1016

where NO N x 107 = per cent of original sample con-
NO
verted to gas x 105


The relationship between the neutron flux and the

gamma flux in a graphite moderated reactor was studied by

Richardson, Allen, and Boyle (32). Expressions were

derived which permit the conversion of integrated

neutron fluxes in terms of neutrons per cm.2 into ab-

sorbed gamma and neutron doses in terms of rads. These

conversion factors were used to convert the integrated

neutron fluxes for the reactor irradiations into rads.

(See appendix.)

The data for the decomposition of CgF16 to gaseous

products by reactor irradiation were combined with the

data for the decomposition of CgF16 to gaseous products

by gamma irradiation. The combined data are plotted in

Figure 14, and the continuity between the two sets of

data is very good. The equation for the curve is




































































101? 1014 1015 1016

Integrated Neutron Flux (neutrons/cm.2)


Figure 13 CSF1', Decomposed to Gas

for Reactor Irradiation.


V)

cc
C-1
0

10-

0
0.




00-

fu
U
L






C-)


IL)

cL.






75



100




10-1
io0





to
S10-3








CO







10-4
e Gamma Irradiation

E Reactor Irradiation



10-5 I I 1 1 I
o103 104 105 106 107 108

Dose radss)

Figure 14. C8F16 Decomposed to Gas
for Gamma And Reactor Irradiation.
U










*P Gam Irdto
















for Gamma And Reactor Irradiation .









I-j x lu00 = 0.007 x 0.72 / 10 -QD
Ti0

where = fraction of sample converted to gas, and


D = total dose in rads.

An analysis of the work by Richardson, Allen, and Boyle

reveals that the enerry deposited by the gamma flux in

a reactor of the type bein- used in this work greatly

exceeds the energy deposited by the neutrons which are

primarily thermal neutrons. Since most of the energy

deposited in the sample results from g amma photon

interactions, the reaction mechanism and rate equations

which can be applied rn this case are identical with

those developed in the previous section.

There was no evidence of any fluorine or hydrogen

fluoride. Some C-.FaC was found in the mixture. The

oxyr,en required to produce this compound was probably

either adsorbed on the walls of the :apsules or dis-

solved in the fluorocarbon.

A calculation was made to compare the decomposi-

tion of C1gFl, in a reactor radiation field with the

decomposition of the coolant used in the organic cooled

reactor. (See appendix.) This coolant is a mixture of

hydrocarbons including biphenyl and terphenyls.

The data for the Piqua Nuclear Power Facility were

used for comparison of the coolants (33). If CSFl6









were to be used as the coolant in the PNPF, the decom-

position rate for the CgFl6 would be about 0.50 per

cent per hour as compared with a calculated value of

3.0 per cent per hour for the organic coolant. In the
light of available data C8F16 would be a better coolant

for the PNPF than the currently used organic coolant.

Other important advantages of the C8F16 as a

coolant would be its higher thermal stability and its

lower corrosiveness as compared to the organic coolant.

Possible disadvantages of using CgFl6 would be its high-

er cost and the poorer heat transfer properties

generally ascribed to fluorocarbons (1 ).


Irradiation of C8Fl6 and UF6.--The results of the

reactor irradiation of CsF16 and UF6 differ significant-

ly from the reactor irradiation of CgF16 and from the

gamma irradiation of C8Fl6. The most significant differ-

ence is the absence of polymeric material in the reactor

irradiated solution. An extensive analytical effort

which is described in the appendix was directed towards

the confirmation of this finding. No evidence of poly-

meric material was found by any method. A second

significant difference was the different distribution

of the gaseous products. For the reactor irradiations

of C8F16 and UF6, roughly equal amounts of CF4, C2F6,









and C2F60 were found with lesser amounts of C Fg and

C3F~O; whereas in the gamma and reactor irradiations

of CFl., the major gaseous products were CF4 and C2F6

in the ratio of about four to one with only small amounts

of C2F60, C3F., and C F80.

A mechanism which accounts for these differences

can be proposed. In the reactor irradiation of CsF16

and UF, the U235 nuclei fission into two fragments.

These two fragments dissipate their kinetic oneriv--

about 195 Nev.-- into the surrounding medium. These two

fragments strike the fluorocarbon molecules and shatter

the ring at three or more bonds,producing three or

more fluorocarbon radicals. Since each fission also

releases six fluoride ions or fluorine atoms, an excess
of fluorine is available for recombination with the

fluorocarbon radicals. In addition it has been pre-

viously reported (25) that the neutron irradiation of

UF6 yields LFU5 and iree fluorine with a C value of 0.5

molecules of TUF, decomposed per 100 ev. of energy ab-

sorbed; consequently there is another source of fluorine

in the system. In addition to these sources of fluorine,

the compound UFP, is itself a potent fluorinating agent.

The processes can be represented as follows:









0n1 + U235F6 -> 2 fission fragments
+ 6 fluorine atoms or ions

+ neutrons gammas.

Collision of fission fragments with the surrounding

fluorocarbon molecules shatters the molecules and

produces various fluorocarbon radicals such as CF3*,

C2F5*, and C3F7' The excess of fluorine immediately

combines with the radicals to form small saturated

fluorocarbons such as CFi, C2F6, and C3F8. The oxygen

containing compounds are formed by the recombination

of necessary radicals with oxygen which is present in
the UF6; i.e., two CF3 radicals could combine with one

oxygen atom or ion to form C2F60, and likewise one CF3

radical and one C2F5 radical could recombine with one

oxygen atom or ion to form C3F0O.

The reactions which have been discussed are demon-

strative and do not necessarily include all possible

reactions; however, the "proposed mechanism" does

account for the lack of dimer because the parent mole-

cule is shattered into small fragments by the primary

act,and these small fragments are immediately saturated

with fluorine.

Consideration of the G values for the gamma irradi-

ated CgF16 and for the reactor irradiated CgF6 UF6

confirms the proposed shattering of the fluorocarbon









molecule. The G values for cas production from the

solution of CF16l and IJF, are higher by a factor of ten

than the C values for gas production from the CAFI6

alone when irradiated with gamma photons. In the reac-

tor irradiations of CF,16 and LWF6,the major products

are CF', C2F,, and C2FAG: and one molecule of CqFl6

thus produces eight CFL molecules, four C2F,t' molecules,

four CF,0 molecules, or some combination of these

molecules. In the pamma irradiation,a sizeable portion

of the energy deposited goes into the production of

pol'vmeric material. In this case one molecule of C1Fi6

produces only 0.5 molecule of dimer, and the number

of molecules produced per unit of energy absorbed is

less than when cases are formed. A comparison of the

C values for the decomposition of C1Fl6 indicates that

the energy required to decompose a molecule of C~F1

by gamma radiation is the same order of magnitude as

the energy required for decomposition by fission

fragments.

An expression may be derived which relates the

number of CaFI6 molecules converted to gas with the

integrated neutron flux. The number of molecules dN

converted to gas is proportional to the number of mole-

cules present II. the neutron flux nv. and the time

increment dt.









The proportionality can be written as

-dNo(Nnvdt ,

and, if k is the proportionality constant,

dN = kNnvdt.

Examination of the data from Table 6 shows that the

total decomposition is very small so that

N =- NO,


and the expression


Integration from N

gives


can be written as

-dN = knvNodt

= NO to N = N and


from t = 0 to t = t


-(N NO) = knvNt ,


N0 N
N knvt
NO


The term on the left is simply the fraction of molecules
converted to gas. The data are plotted in Figure 13

and the points fall along a straight line as expected.

The equation for the C8F16 + UF6 curve is

logN N -N x 107 = -9.80 + 0.92 log nvt


for the range 1012 < nvt < 1016


where NO N x 107 = per cent of original sample
0
converted to gas x 105, and









nvt = integrated neutron flux in neutrons/cm.2

Several other experimental results are worthy of

note. Solutions of IUF6 and CgFI were prepared and were

found to be homogeneous with no apparent side reactions.

The solubility of UF6 in CF16 was found to be about

0.65 gm./ml. The fluorocarbon was successfully sepa-

rated from the fission fragments and the IlF6 by passing

the samples through a sodium fluoride trap with the

samples beinr in the vapor phase. The largest number of

passes required was three with some separations being

complete after one pass. A light film of greenish

white powder remained in the capsule. This powder was

probably UF, which had been chemically reduced to UFL.

The radioactivity of the separated fluorocarbon was

barely detectable. Thus, separation of fluorocarbons

from UFF6 and fission products has been found to be a

simple process. There was no evidence of elemental

fluorine or hydrogen fluoride in the capsules.

A calculation was made to ascertain the decompo-

sition of a CjF16- UFT solution if the solution were to

be used in a reactor similar to the Homogeneous Reactor

Experiment I, HRE-1 (3L). Assuming the same concentra-

tion of U233 per unit volume and the same flux level

as the HKE-1, a typical decomposition rate for the

CFI16 solvent was found to be 93 per cent per hour.









This rate of decomposition and the cost of make-up

feed would counterbalance the fact that the solution is

electrically nonconducting and hence noncorrosive. The

decomposition rate was obtained by extrapolating the

decomposition rate an order of magnitude higher than the

highest data point. This assumption of linearity of

decomposition with dose may not be completely justified.

Another factor which should be considered before this

result is extended to other fluorocarbons is whether

or not C8F16 is a good representative material. In pre-

vious investigations, the alicyclic fluorocarbons, of

which C8F16 is a member, have been found to be consid-

erably less radiation resistant than have the aromatic

fluorocarbons.









Ileutron Phy'sics Parameters


In order to evaluate the fluorocarbons as worrini.

fluids in a nuclear reactor it is necessary that their

neutron absorbin( and scattering properties be known.

Conventional static experiments for measuring these

Properties require large amounts of material, but simi-

lar information can be obtain-d with a small sample by

m-easurin- neutron decay with the pulsed neutron tech-

nique. The basis of this technique can best be shown by

examinin!, the time dependent equation for the neutron

flux in a nonrultiplyin, mrediiume. The neutron flux as a

function of space anJd ime ) (r,t) is riven by


) (r,t) =I)(r)e-( a + Dv ,)L


where the neutron flux ) is the product of the number

of neutrons per unit rolume and the velocity,. of the

neutrons. The velocity of the neutrons is v, ani

S, is the macroscopic absorption cross section
for the medium. The macroscocic absorption cross section,

a property of the medium, is a measure of the medium's

tendercy to absorb neutrons by (n,f ) reactions and

thus remove the neutrons from the neutron population.

The diffusion coefficient D is an inverse measure of

the medium's ability to scatter neutrons by neutron









interactions with the nuclei of the medium. The mater-

ial buckling B2g for a cylinder is given by

B2 2.405 2 + TT 2
R H

where R = radius + 2 D

and H = height + 4 D
The spatial dependence of the neutron flux can be

removed if the flux is considered at one specific point

in the system. In this case ln((t) (v a + DvB2g)t

if the natural logarithms of both sides are taken. If

the neutrons are assumed to be monoenergetic, all of

the terms inside the parenthesis are constants for this

case, and these constants can be represented by a single
term X Substituting the decay constant in the previ-
ous relationship gives the simple expression In (t)'- -t,
and A the decay constant, can be determined by measur-

ing the neutron flux as a function of time.

If the physical dimensions of the system are

varied and the decay constant measured for each size

system, the macroscopic absorption cross section and the

diffusion coefficient can be determined as is shown by

Figure 15. The plot of \ versus B2 has as its inter-
cept with the ordinate, the quantity via, and by

knowing the neutron velocity, it is quite simple to

calculate Z a. Likewise, the slope is Dv which will





















S= v f. + vDB.
d '


B2s (cnr,2)


Figure 15. Analysis of Pulsed Neutror
Experimental DaLa.


X (sec-I)









yield the diffusion coefficient if the neutron velocity

is known. As was apparent from the mathematical defi-

nition of buckling, the buckling decreases as the

physical size increases. Therefore, in order to get

a good value of the intercept via, it is important to

have points on the curve near the ordinate, or in other

words,the system should be as large as possible. This

is an expected result since B2g = 0 actually represents

an infinite system which would lose no neutrons to the

surroundings.

The results of a typical run are shown in Figure

16 which is a plot of counts per channel versus channel

number or time. The curve has no region where the slope

is constant for more than 2 or 3 microseconds. This

result was characteristic of all of the runs and led to

the conclusion that there was a steady state contribu-

tion to the neutron decay which was independent of the

fluorocarbon. A run was made with all equipment

arranged as before with the exception that there was no

fluorocarbon in the tank. As shown in Figure 17, there

is a decay of the neutron population in the empty tank

which is significant when compared to the decay of the

neutron population in the fluorocarbon. This decay could

possibly be attributed to neutrons leaking from the sys-

tem, being moderated and reflected in the room, and

being returned at energies slightly above the cadmium















End of Neutron Fulse


V,


Y.


Data from Fun 1o. 12
H = 12.7 cm. C 8Fi, in Tank
E = 0.0'1, cm. -
6C, ,


S 12 1's .

Time (,-sec. )


Figure 1'. Counts Per Channel Vs. Time for CF1,. -


c. x 10,


Counts

Perl

Channel


2 10.'


0 .


-20 2 ." 2


I I










5 x 104


4


3



2





Counts

Per 104

Channel


2 x 103


0 4 8 12 16
Time (/ sec.)


20 24 28


Figure 17. Counts Per Channel Vs. Time for Empty
Tank.









cut-off. Another possible cause would be moderation of
the neutrons by the air in the tank.

The result of a channel by channel subtraction of

the decay of the neutron population in the empty tank

from the decay of the full tank is shown in Figure 18.

A reasonably straight region for which a decay constant

can be determined was civen by the subtraction of the

two curves. The decay constant for this curve was cal-

culated usint a least square fit of the data. This

subtraction procedure was repeated for other runs with

similar straight regions being found. It is to be noted

that the subtraction was performed channel by channel

rather than smooth curve from smooth curve. This channel

by channel subtraction was, of course, a much more severe

treatment of the data.

Figure 19 presents the plot of \ versus B'K. for

C8F16. The data were least square fitted by the lion

Linear Least Squares (TILLS) computer code. The points

are in a reasonable straight line: however, the point

for the lowest buckling, that is the maximum volume, is

not too close to the axis when it is realized that the

intercept is of the order of .O00 and the first point

has a value of about 3.1. The computer code gave a

value for the intercept of 0.0073 sec.-. A statistical

analysis showed that the deviation with a 95 per cent
























A= 6.651 x 104 sec.-1


Run No. 12 -Run No. 10
H = 12.7 cm.
Channel by Channel Subtraction
B2 = 0.0461 cm.-2
g


S I i I I


0 4 8 12 16 18 24 28

Time (q sec.)

Figure 18. Counts Per Channel for CgF16 Minus
Counts Per Channel for Empty Tank Vs. Time .


5 x 103


Counts

Per

Channel


103


5x102 F


SI I n


-




Full Text
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 EO5N6JM98_3RS826 INGEST_TIME 2017-07-19T20:34:07Z PACKAGE UF00097877_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

FISSION PRODUCT DEGRADATION AND NEUTRON MODERATING PROPERTIES [ OF FLUOROCARBONS By THOMAS HOWARD SCOTT A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTL^L FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA April, 1966

PAGE 3

ACKNOWLEDGMENTS The author wishes to express his appreciation to Dr. John A. Wethington, Jr., chairman of the supervisory committee, for his untiring: guidance, assistance, and encouragement without which this work would have not been possible . The assistance of Mr . L. D. Butterfield in the performance of the reactor irradiations and the assistance of Mir. J. A. Maclean in the performance of the gamma irradiations are acknowledged. Dr. T. M. Reed provided the distillation equipment, and Dr. T. F. Parkinson assisted in the neutron pulsing experiment . The mass spectrometric analyses were performed by Dr. M. B. Fallp:atter and Dr. R. J. Hanrahan. The nuclear magnetic resonance measurements were made by Dr. W. S. Brey, and the infrared spectra were measured by Dr . H . C . Brown . The author is indebted to the Oak Ridge Institute of Nuclear Studies and to the College of Engineering of the University of Florida for fellowship assistance. Finally, the patience, encouragement, and perseverance of the author's wife, Janet, during the preparation of this manuscript are gratefully acknowledged. ii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMENTS ii LIST OF TABLES iv LIST OF FIGURES v ABSTRACT vii Chapter I. INTRODUCTION 1 II. PREVIOUS VJORK 4 III. EXPERIMENTAL PROCEDURE l5 IV. EXPERIMENTAL DATA AND RESULTS 41 V. DISCUSSION OF RESULTS 56 VI. CONCLUSIONS 95 Appendices A. CALCULATIONS 99 B. ANALYSIS FOR POLYTffiR IN IRRADIATED SAT^LES 130 LIST OF REFERENCES 132 BIOGRAPHICAL SKETCH 13 5 iii

PAGE 5

LIST OF TABLES Table Paf^e 1. Summary of Conclusions in References Ik Through 20 9 2. Results of Analyses of CgFj^ 20 3. Decomposition Percentages and Gaseous Product Distribution for Gamma Irradiation of CgFj^^ U k. G Values for Gamma Irradiation of CgF-j^^ . . 47 5. Decomposition Percentar:es and Gaseous Product Distribution for Reactor Irradiation of CgF^^ k9 6. Decomposition Percentaf^es and Gaseous Product Distribution for Reactor Irradiation of CgF-L^ and UF^ 51 7. G Values for Reactor Irradiation of CgF-j^^ and UF^ '. . . 53 S. Geometric Buckling and Decay Constants for Pulsed Neutron Experiment 55 9. Data for Figure 11 66 10. Macroscopic Absorption Cross Sections of Moderators 94 11. Supporting Data for Figure 20 107 12. Supporting Data for Figure 21 lOS 13 . Supporting Data for Figure 22 109 IV

PAGE 6

LIST OF FIGURES Figure Page 1. Chromatogram of Crude C^F-^^^ 19 2. Chromatogram of Distilled CgF-j^^ 22 3 . Metal Vacuum System 24 k. Metal and Glass Vacuum Systems 25 5. Capsule Fluorination System 23 6. Glass Vacuum System 37 7. Neutron Pulsing Experimental Equipment . . 3^ B. CrtF-^^ Decomposed to Gas in Gamma Irradiation 60 9. CgF25 Converted to Polymer in Gamma Irradiation 61 10 N Irradiation of C^F.^ ok Log (SX 100) Vs. Dose for Gamma Nq 'g^l6 11. Log (~ X 100) Vs. Dose for Other Fluorocarbons oo 12. Dose-Decomposition Relationship for Fluorocarbons Corrected for Effects of Lighter Products 72 13 . CrtF-,^ Decomposed to Gas in Reactor Irradiation 74 14. '^g^i6 Decomposed to Gas in Gamma and Reactor Irradiation 75 15. Analysis of Pulsed Neutron Experimental Data S6

PAGE 7

Figure Page 16. Counts Per Channel Vs. Time for CgF^^ . . SB 17. Counts Per Channel Vs. Time for Empty Tank g9 1^Counts Per Channel for C^F-,^ T-linus Counts Per Channel for Empty Tank Vs. Time ... 91 19. Decay Constants Vs. B^ 92 20. Chromatogram of Gases from Sample No. 36 . 104 21. Chromatogram of Sample No. 36 in Vapor Phase 105 22. Chromatogram of Sample No. 36 in Liquid Phase 106 vi

PAGE 8

Abstract of Dissertation Presented to the Graduate Council in Partial Fulfillment of the Requirements for the Dep;ree of Doctor of Philosophy FISSION PRODUCT DEGRADATION AND NEUTRON MODERATING PROPERTIES OF FLUOROCARBONS By Thomas Howard Scott April, 1966 Chairman: Dr. John A. Wethington, Jr. Major Department: Nuclear Engineering The radiolysis of perf luorodimethylcyclohexane by gamma radiation and by reactor radiation was studied. The primary product was a dimeric material with small amounts of gaseous products including CF^, '-'2^6' ^2^(~>^' C^Fg, and C3FgO. The production of dimer and gaseous products was found to be a linear function of the total dose . An expression was derived which related the fraction of sample remaining after exposure to a knovm dose This expression was modified to correlate all available data for the irradiation of alicyclic and aliphatic fluorocarbons . A separate expression was found for vii

PAGE 9

aromatic f luorocarbons. The expressions hold over a wide range of total doses, dose rates, and for several different types of radiations. The decomposition of CgF-j^^ in a solution of C^F-j^^ and UF^ exposed to reactor irradiation was investigated. Techniques were developed for the preparation of solutions and for the subsequent separation of the C^F-j^^ from the UF^ and the fission fragments. The primaryproducts were OF, , C2F^, and ^2^(P in roughly equal proportions. No dimeric material v/as found. The kinetics of this reaction supported the assumption that fission fragments shatter the CgF-j^^ into fragments which are then saturated with fluorine present in the system. viii

PAGE 10

CHAPTER I INTRODUCTION During the past ten years considerable effort has been devoted by the Atomic Energy Commission towards the development of a nuclear reactor in which a hydrocarbon liquid is used as the working fluid. During the past ten years there has also been a large scale development of the f luorocarbons . The proposal has been made that f luorocarbons might be used as the working fluid in a nuclear reactor (1). This proposal is based on several factors. First of all, the hydrocarbons tend to form complex compounds with the iron in the walls of the colder regions of the reactor system (2). These carbon-iron compounds then tend to be deposited on the heat transfer surfaces of the reactor fuel and seriously retard the rate of heat transfer from these surfaces. The f luorocarbons are generally considered to be much less reactive than the hydrocarbons; therefore, this corrosion problem might be eliminated. Secondly, uranium hexafluoride is soluble in f luorocarbons . Thus a possible reactor concept could be developed in which the uranium fuel is dissolved in a fluorocarbon working fluid. 1

PAGE 11

2 The uranium hexaf luoride-f luorocarbon reactor would have several distinct advantages. First of all, uranium hexafluoride is the uranium compound used in the gaseous diffusion process for producing enriched uranium. Thus, the material could be used without further processing, and the conversion and fuel fabrication costs for solid fuels could be eliminated. These costs are a significant portion of present fuel cycle costs. Secondly, by means of a side stream, the fuel could be continuously reprocessed and new fuel added. The need for lengthy shutdowns for refueling would be eliminated and the offsite reprocessing of solid fuel elements would not be necessary. A final possible advantage of such a reactor might be the radiation induced production of f luorocarbons which can not be produced by conventional chemical processes because of the nonreactive chemical properties of the f luorocarbons. The fissioning of uranium in a solution would produce fission fragments of high energy. The possible decomposition of the fluorocarbon and the subsequent recombination of the various decomposition products might produce new and useful f luorocarbons. In this consideration, the uranium hexafluoride-fluorocarbon reactor would be closely akin to the chemonuclear reactor which has been discussed for several years. The chemonuclear reactor uses radiation-

PAGE 12

induced chemical reactions or dissociations to produce desired chemical compounds. Obviously, much information regarding the performance of f luorocarbons and uranium hexaf luoridefluorocarbon solutions in the presence of radiation must be obtained before any use is made of these materials in reactors. Among the types of information needed are the rates of decomposition, the products formed, the possible effects of the evolution of fluorine or hydrogen fluoride, and the compatibility of fluorocarbons and uranium hexafluoride with regard to possible side reactions. It is also desirable to know the neutron moderating and absorbing properties of the fluorocarbons . The purpose of this work is to investigate the effects of ionizing radiation on the fluorocarbon, perfluorodimethylcyclohexane (CgF25), the performance of CgF-L£,-UF^ solutions in a reactor flux, and the neutron moderating and absorbing properties of CgF-j^^^.

PAGE 13

CHAPTER II PREVIOUS WORK Radiation Chemistry of Fluorocarbons Fluorocarbon chemistry is a relatively new branch of chemical science. Before 1937, perfluorome thane and perf luoroethane were the only positively known fluorocarbons. In 1937 (3) and in 1939 (4), Simons and Bloch reported the preparation and properties of fluorocarbons containing more than three carbon atoms. These compounds, which ranged from perf luoropropane to perf luoroheptane , were reported to be stable both chemically and thermally. During the early stages of the Manhattan Project the need arose for materials which could be used as sealants and coolants in mechanical systems containing corrosive uranium hexaf luoride . Since all other materials which had been tested had been found to react with uranium hexaf luoride , Simons (5) suggested in July, 1940, that the recently developed fluorocarbons might be suitable for these applications. In December, 1940, Simons sent a two-cubic centimeter sample of liquid fluorocarbon, virtually all of the material available, 4

PAGE 14

to Columbia University for testing. These tests showed that f luorocarbons had the desired properties (6), and an extensive research program was then started to find the most desirable fluorocarbons for the various applications and to develop processes for producing these fluorocarbons. Thus by the end of the second World War, fluorocarbon chemistry was established as a distinct branch of chemical science. The best known of the fluorocarbons commercially produced in the post-war era was Teflon. A series of papers published in the period between 1953 and 1956 (7, ^, 9) indicated that Teflon had very poor radiation resistance. Indeed Teflon was found to be severely damaged at a dose of 3 x 10 roentgens (9), and unfortunately this property was generally extended to all fluorocarbons without any real experimental evidence on fluorocarbons other than Teflon. An additional negative factor was the liberation of fluorine during the irradiation of Teflon (5). The corrosion problems related to the presence of fluorine gas in a mechanical system would seriously restrict, if not prohibit, the use of fluorocarbons in ionizing radiation fields. Slowly, experimental data were published which showed that all fluorocarbons were not as sensitive to radiation as Teflon. In 195^, Fens reported that the gamma irradiation of mixtures of partially fluorinated hydrocarbons

PAGE 15

and hydrocarbons and mixtures of fluorocarbons and hydrocarbons indicated that the organic fluorine compounds showed a m.uch greater resistance to radiation than did the analogous chlorine, bromine, and iodine compounds (10). Also in i960. Wall, Florin, and Brown (11) reported that C^F^ had radiation stability approaching that of C^H^ and suggested that C^F^ be used for the preparation of radiation resistant polymers . Mixtures of C^F^ with hydrocarbons were less stable than the parent material, and the difference was attributed to the production of hydrogen fluoride in the hydrocarbon-f luorocarbon mixtures. The first reactor irradiations of fluorocarbons were performed by Simons and Taylor (12) who irradiated the saturated fluorocarbons — CyF^^, C^F-j^^O, and (CiFq)^N — in a reactor flux of 5.5 x 10 neutrons/cm. sec. for four weeks. The G value (number of molecules produced or destroyed per 100 ev. of energy absorbed) was calculated to be 2 or 3 molecules transformed per 100 ev. absorbed with a total energy absorption of 2.2 22 x 10 ev. per gram. No fluorine or CF, were observed, and there was no evidence of corrosion, since the inner walls of each capsule were bright and clean after irradiation. Concurrently, Wall and Florin (13) found that the tensile strength of irradiated Teflon decreased very

PAGE 16

7 rapidly in the presence of oxygen, but the loss of tensile strenp;th in the absence of oxygen was hardly perceptible for long periods of irradiation. This finding suggested that the reported radiation instability of Teflon was due to oxygen impurities. Bloch, MacKenzie, and Wiswall (14) used electron, gamma, and reactor sources to irradiate C^F^, C-jpF-iQ, and C^^qF-jh. The two aromatic compounds had overall stabilities of the same order as their hydrocarbon analogs. Even though the f luorocarbons tended to undergo more reaction to form high molecular weight compounds than the comparable hydrocarbons, the f luorocarbons were found to be more stable to fragmentation leading to gaseous products. The gas yields for the alicyclic compounds were much higher than the aromatics, but polymer yields were of the same order or possibly lower. This observation was in direct contrast to the hydrocarbons where the alicyclics are much less stable to irradiation than are the aromatics. Gas chromatography was used to analyze the radiolysis products. In 1962, Heine (15), using doses of 5 x 10° rads, studied the gamma radiolysis of CgF^g, (C^^Fgl^N, and c-CgF]^50 ; the stability of these f luorocarbons was shown to be greater than the radiation stability of the analogous hydrocarbons. During I964 and I965, five papers were published which added greatly to the understanding

PAGE 17

^ of the effects of radiation on f luorocarbons . The conclusions given in references (14) through (20) are summarized in Table 1. MacKenzie, Bloch, and Wiswall (16) performed radiolysis experiments on a series of highly purified cyclic f luorocarbons — CrF^, C^pF-iQ' ^^*^ '^lO^S — *"*^ °" their saturated analogs— C5Ftl2, ^12^22' ^"^ ^lO^lg* '^^® materials 9 were irradiated to doses of the order of 10 rads in nickel cells using a 1.5 Mev. Van de Graff electron beam. The irradiations were carried out at temperatures chosen so that the materials were in the liquid phase . It was concluded that the radiation stability of the aromatic compounds was less than that of the corresponding hydrocarbons; whereas the alicyclic f luorocarbons were more stable than their hydrocarbon analogs. The total G values for destruction of starting material were very similar for aromatic and alicyclic compounds and ranged from 1.3 to 2.4. Considerable amounts of gases and products of molecular weight lower than the starting material were found in the radiolysis of alicyclic compounds; whereas polymeric material was almost exclusively yielded in the irradiation of aromatic f luorocarbons, with only traces of gaseous and low molecular weight compounds being formed. No free fluorine was detected in any of the irradiations. In an extension of the aforementioned work, Bloch

PAGE 18

w CO t-1 o o o en o W CO o M CO tD t^ O o o o >-•

PAGE 19

10 n c o •H CO U C o o

PAGE 20

11

PAGE 21

12 and KacKenzie (17) used a Co source to irradiate a series of f luorocarbons at elevated temperatures (generally about /).50° C.) to study the combined effects of thermal decomposition and radiolytic decomposition. The results indicated that for the aromatic series there was no marked increase in decomposition with temperature; however, a dose of less than 10 rads at 450° almost completely destroyed C Fpp (perf luorobicyclohexyl ) , the only alicyclic fluorocarbon studied. It was noted that C,„Fp„ mi?;ht not be typical of alicyclic f luorocarbons (i..e . , the bond joininr the rings seemed especially vulnerable), but it was supposed that the results indicated that, as a class, the alicyclic fluorocarbon would not stand the combined effects of high temperature and high radiation dose . The performance of C, F (perf luorobiphenyl ) was particularly good in that its radiolytic decomposition was still reasonably small at 500° C. For comparison, the hydrocarbon terphenyls began to show drastic radiolytic decomposition at temperatures over 450° C . , and at 500 ° C. the thermal decomposition rates were of the order of 50 per cent per day. The radiolysis of ^2?^ using a gamma source has been studied by Kevan and Hamlet (l8) who made the following comparison between the radiolysis of saturated fluorocarbons and hydrocarbons: (a) C2F^ was about one-fifth as sensitive to radiation decomposition as was C2H£, . The G value for C2F^

PAGE 22

13 was 1.9 while the G value for C2H^ was 9. The causes for this apparent radiation stability included the absence of F abstraction in f luorocarbons, the possibility of back reactions of free radicals to reform C^F/^, and the absence of excited molecule decompositions which would have given molecular fluorine. Molecular ejection of hydrogen from excited molecules was shown to make important contributions to product formation in the alkanes. It was suggested that similar excited molecule reactions in perf luoroalkanes were unimportant. (b) Hydrogen was unreactive in alkane radiolysis; whereas fluorine was a reactive product in perfluoroalkane radiolysis. This dissimilarity may have been caused by the great difference in bond strength and by the exothermic reaction of fluorine with perfluoroalkyl radicals. (c) Linear alkanes were characterized by large yields of hydrogen while perfluoroalkanes were characterized by large yields of CFi . (d) The radiolysis of linear f luorocarbons may produce cyclic compounds; however, no cyclic compounds are produced in the radiolysis of linear hydrocarbons. The radiation chemistry of C^F,^ and c-C, Fg was investigated by Fallgatter and Hanrahan (19). The gamma doses used were from 0.2 Mrads to about & Mrads. The purpose of using doses lower than those used by previous

PAGE 23

14 investigators was to study the initial products and the initial product yields. At low doses dimeric fluorocarbons were found to be the predominant products of the ramma radiolysis of ^f,^i2' "^^^ initial G value for C^F^2 for products lighter than the parent compound was given as 0.3, and the G value for products heavier than the parent compound was given as 3.1. The latter G value was broken down into G = 2.2 for G-j^2^22 (p^^^l^orobicyclohexyl) , G = 0.66 for C^pFpp (perf luorocyclohexylhexene) , and G = 0.22 for C-j^2^24 (pe^'fl^o^ocyclohexylhexane ) . These total G values compared favorably with those given by MacKenzie et al. (16) of 0.3 for lighter products and 2.1 for heavier products. Portions of the analysis for this work were performed on a gas chromatograph-mass spectrometer combination. The m.ethod of separating a specific gas from the radiolysis products by means of the gas chromatogra^i and then injecting the separated gas into the mass spectrometer for analysis was discussed. Although Fallgatter and Hanrahan felt that extended speculation about mechanism was premature, it was possible to make some generalizations. The great strength of the C-F bond compared to the C-H bond and the weakness of the F-F bond as compared to the H-H bond were two of the major differences between the f luorocarbons and hydrocarbons; however, since more than ^0 per cent of the observed bond rupture in the radiolysis of liquid C^F-]l2 involved C-F rather than C-C bonds, it was apparent

PAGE 24

15 that the greater strength of the C-F bond was no deterrent to its r.upture . It was pointed out that C^F^p was probably not a typical compound since Cg^H^^ showed far less C-C bond rupture than hydrocarbons in general. Even so, the assumption that the radiation chemistry of perfluorocarbons was centered in C-C bonds rather than C-F bonds was to be avoided. Abstraction from the fluorocarbon substrate by F atoms to make Fp was essentially impossible because of the weakness of the F-F bond. Consideration of this factor alone showed that the radiation chemistry of fluorocarbons should be much different than that of hydrocarbons The lack of C^Fn-j in the radiolysis products tended to confirm the lack of fluorine abstraction. Most proposed mechanisms predicted the production of fluorine and ^6^10' *"^ even though the fluorine might have reacted with other products or the walls of the system, the C^F^^q would not . A consideration of the reactions to produce fluorine and C^F,q showed that the reactions were endothermic and were quite unfavorable thermodynamically . Reed, Mailen, and Askew (20) irradiated a series of pure fluorocarbons using three different radiation sources. Fluorocarbons ranging from CF/^ to 2,3-(CFt )2C||^Fg were irradiated in a Co gamma source, in the Low Intensity Test Reactor at Oak Ridge National Laboratory, and

PAGE 25

16 in the Oak Ridge Graphite Reactor. In addition, the radiolysis of mixtures of CF^ and ^2!^^ was studied. The gamma irradiations gave G values for the fluorocarbons which were considerably less than those of the analogous hydrocarbons especially for the lower members — CF, , C-F^, and C.F^. The highly branched structure 2,3-(CF3).C, F^ was the only exception, and it had a total G value comparable to those of the hydrocarbon alkanes. Radiation Chemistry of Uranium Hexafluoride The radiation stability of UF^ must also be considered. The use of UF^ in gaseous diffusion plants necessitated studies on the critical sizes of various configurations containing UF^. Snell and Rush (21) determined the multiplication factor for 3OO lb. product drums containing 1.1 and 0,536 per cent u235. Grueling et al. (22) studied the critical dimensions of water tamped slabs and spheres of UF^, and Bull (23) calculated the critical mass of a UF^ core with reflectors of D^O, Be, and C. However, all of these efforts were concerned with potential criticality incidents and not with the use of UFr in a reactor as fuel. Bernhardt, Davis, and Shiflett (24) irradiated UF^ with alpha particles and found that intermediate solid fluorides and fluorine

PAGE 26

17 were formed. It was also found that these products tended to recombine, giving a net G value of one molecule of UFr decomposed per 100 ev. of energy absorbed. Dmitrievskii and Migachev (25) reported that the neutron irradiation of UF^ yielded UFc and free fluorine with a G value of 0.5 molecule of UF^ decomposed per 100 ev. of energy absorbed in the gas. An equilibrium was found to exist between the UF^, and the UFc and fluorine products with the equilibrium concentrations being a function of the dose rate. A further significant result was the fact that the decomposition of the UF^ was completely prevented by the addition of 25-5 per cent fluorine to the UF^. This result had been previously predicted by Goodman (26). Thus it appears that the radiation stability of UF^ may not be good, but it can be brought up to acceptable levels by the addition of fluorine to the UF^. In fact, the operation of a UF^ reactor has been described by Kikoin et al. (27) in which the decomposition of the UFx was controlled by the addition of chlorine trifluoride to the fuel to serve as a source of fluorine.

PAGE 27

CHAPTER III EXPERIMENTAL PROCEDURE Distillation Approximately 50 liters of CgF^^g were obtained from the Oak Ridge National Laboratory. No information was available on the previous history of the Cc^F-./or on its purity. Chromatographic analysis of the crude CgFj5 indicated that there were several impurities as shown in Figure 1. In addition, the material had a yellowish color and a distinct odor which are not characteristic of fluorocarbons. The material was also analyzed using nuclear magnetic resonance measurements and infrared spectra measurements. The results for these analyses are shown in Table 2. The impurities in the CgF^^ were removed by a batch-type distillation using a distillation apparatus consisting of a 2.5 liter pot with a packed column 100 cm. high. The column had a diameter of 1.5 cm. and was packed with stainless steel helixes, and two independently controlled electric heaters were used to maintain the column temperature at the desired levels. IB

PAGE 28

19 to o 0) pi u o c t: o p E o o b.. •H o

PAGE 29

2.0 TABLE 2 RESULTS OF ANALYSES OF CgF^^^ Infrared Measurements Material

PAGE 30

21 A 1.5 liter batch was charged into the distillation pot at room temperature, and the pot temperature was raised until the charge began to boil. Concurrently the column temperature was increased to 100° C. for the upper half of the column and to 104° C. for the lower half of the column. The product was taken off at a rate of about 10 drops a second until the temperature of the product reached 95-0° C, and the pot temperature increased to about 115° C. The product taken in this temperature range consisted of two immiscible liquids. The lighter phase was yellowish and had an odor somewhat like benzene while the heavier phase was colorless and had an almost undetectable odor. The next cut, taken between product temperatures of 100.7° C. and 102.1° C . at a flow rate of 4 drops per second, was the purified O^F-^^ which was used in the experimental work. The pot temperature increased from 119° C. to 122° C. during this cut, with approximately 1.0 liter of purified material being obtained from each 1.5 liter of raw materials charged to the pot. The distillation significantly decreased the amount of impurities as shown by the chroraatogram in Figure 2. The analyses of the distilled material are given in Table 2.

PAGE 31

22 to O «) n •H Q tH o E CO U, to o 4-> (« E o O o

PAGE 32

23 Fluorination Since UF^ was to be used in some experiments, it was necessary to fluorinate all surfaces which were to be exposed to UF^ with fluorine Ras in order to form a protective fluoride film. This film prevents the reaction of the UF^ with the metal surfaces to produce UF, which would adhere to the surfaces and cause difficulties in sample transfers and in material balance calculations. The vacuum system shown in Figures 3 and U was fluorinated using fluorine gas from a 1-pound fluorine gas cylinder obtained from the Matheson Chemical Company. All components except the bellows seal valves of the system were carefully washed in carbon tetrachloride in order to decrease all surfaces which were to be in contact with fluorine. The bellows seal valves were Nupro type B-Z4.H bellows valves which were specially designed for systems in which no leakage could be permitted. These valves were cleaned by the manufacturer. The system was pressure tested three times with a nitrogen pressure of 60 psig . and vacuum tested five times for periods from one to twelve hours. For fluorination, the sodium fluoride traps were replaced with sodium chloride traps. The outlet of the sodium chloride trap used for capsule analysis was led to a glass bubbler containing a solution of 20 per cent potassium hydroxide which was vented to the hood.

PAGE 33

24 %AAA/^ a D. V Ci^ (1) *^ CO U CO S Eh a . a *) i rt > f CO iH b "J n> 3 a 3 e a, (1) O P cd E O Ih o E^%AAAA^-IX1 — I I -a 0) a, CO (B fn a:

PAGE 34

25 1st*!*!*!*!*!*!*!*! *!*l*l*l*t«t*t*'»*i*!***K*>*!*l*i»|«n CO e •p to >. CO > to to O TJ a 0) •H

PAGE 35

26 The fluorine gas was introduced into the vacuum system through a bellows seal valve on the vacuum rack. A specially designed fluorine regulator (Matheson model # I5F-67O) was used to control the flow of fluorine. This valve permitted the addition of small quantities of fluorine to a carrier gas which in this case was nitrogen. A sodium fluoride trap was placed between the regulator and the vacuum system to remove any hydrogen fluoride which was present. The sodium fluoride trap was heated under vacuum for one hour with a gas-air torch to activate the surfaces of the sodium fluoride pellets. A tape heater was used to maintain the trap at a temperature of 100° C. during the fluorination. The fluorination operation was commenced with establishment of a steady nitrogen flow. The rate of flow was adjusted so that the flow of nitrogen in the bubbler was as large as possible without causing the potassium hydroxide solution to bubble out of the bubbler. The nitrogen pressure on the system was about S psig. The fluorine regulator valve was opened so that the fluorine pressure between the fluorine regulator and the fluorine control valve was about 12 psig. The control valve was then opened for one-second intervals approximately every two minutes for thirty minutes. The interval for which the control valve was open was slowly

PAGE 36

27 increased during the next hour so that at the end of the hour the interval was fifteen seconds. The fluorine control valve was then opened for intervals of one minute every five minutes for thirty minutes. During the operation, the sodium chloride trap temperature was monitored with a slow increase in temperature being observed. The system was purged with nitrogen and allowed to stand overnight . Final purging was accomplished by removing the glass bubbler, filling the entire vacuum system to 50 psig., and then allowing the nitrogen to blow off quickly . The vacuum system was then evacuated so that the nitrogen pressure in the system was about 0.5 atmosphere, and the glass bubbler was reconnected. The fluorine control valve was then opened, and the pressure in the vacuum system was brought up to about 1.0 atmospheres. The system was then completely isolated and allowed to stand for thirty minutes. A very low nitrogen flow was established for thirty minutes after which the pressurizing and blow off procedure described above was repeated. A system which could be used to fluorinate capsules, valves, and other fittings was built as shown in Figure 5. All items were carefully degreased with carbon tetrachloride, and, in addition, the capsules were immersed

PAGE 37

23 c u p C 4J r-l s •) p 00 c o •H P « •H O 3 9 n a (f o «) 3

PAGE 38

29 three times in boiling water to remove any remaining silver solder flux. The capsules and fittings were fluorinated in the reaction vessel while the valves were fluorinated by being inserted into the nitrogen-fluorine stream between the reaction vessel and the sodium chloride trap. The system was pressure tested to 50 psig. with nitrogen for one hour. The pressure was quickly released with the glass bubbler disconnected to purge air from the system. The pressurization and release operation was repeated three times. A steady flow of nitrogen was established with the bubbler in place, and the fluorine pressure on the regulator outlet was raised to 12 psig. Fluorine was introduced into the system by opening the outlet valve for a period of one second at intervals of one minute . After thirty minutes the period during which the outlet valve was opened was increased to five seconds with the intervals that the valve was closed still being one minute. After another thirty minutes fluorine was introduced for a one-minute period, and the system was completely isolated and allowed to stand for one hour. The purge was then reestablished for fifteen minutes. The glass bubbler was disconnected, and the previously described pressurization and release operation was carried out three times.

PAGE 39

30 Capsule Fabrication Capsules for Gamma Irradiations The capsules for the gamma irradiations were fabricated from a 3-in. section of one-half in. copper tubing. One end of the capsule was sealed by crimping and silver soldering the copper tubing, and the other end was sealed with a brass VHiitey valve (1-VT^) attached to the tube with a brass Swagelok fitting. Capsules for Reactor Irradiations The capsules for the reactor irradiations were fabricated from a 5-in. section of one-half in. aluminum tubing . One end of the capsule was sealed by crimping and heliarc welding the altmiinum tubing, and the other end was sealed with an aluminum Whitey valve (1-VM/).A1) attached to the tubing with an aluminum Swagelok fitting. All reactor capsules were tested with a nitrogen pressure of 300 psig • All components of the capsule were degreased with carbon tetrachloride and fluorinated.

PAGE 40

31 Sample Preparation Preparation of CgFj^ Samples The CgFw samples were prepared using the vacuum system shovm in Figure 3 • The capsules were attached to the system and both pressure and vacuum tested. The calibrated burette was filled with C^^i() which had been degassed by vacuum transfer from one glass bulb to another glass bulb for a total of three times. After each transfer the dissolved gases were pumped to vacuum. The C^F, ^ sample was vacuum transferred to the capsule , and the size of the sample was determined by difference on the calibrated burette. A gold wire weighing 1 mg. was wrapped in tissue paper and attached with tape to each capsule which was to be exposed to reactor radiation. The gold wires had previously been weighed to the nearest 0.001 mg. Preparation of C^F,^ UF^ Samples Samples containing C^F^^ and UF^ were also prepared with the vacuum system. The UF^ was obtained from the Allied Chemical Company and was found to contain some dissolved air. Most of the air was removed by vacuum transferring the UF^ from the storage tank to the calibrated tank and back to the storage tank. During each transfer the UF^ was sublimed and recondensed, with the

PAGE 41

32 dissolved air being partially released. After each transfer, the tank containing the UF^ was pumped to vacuum while liquid nitrogen was still on the bottom of the tank. The volume of the calibrated tank was found by filling the tank with water in a constant temperature bath (25° C.) and then weighing the tank and water. Subtraction of the empty tank weight from the full tank weight gave the weight of the water in the tank. The weight of the water was then converted to volume using the density of water at 25° C, and the volume was found to be 75^ cc. The calibrated tank was filled with UF^ by opening valves as necessary and allowing the system to equilibrate for four minutes at room temperature. The valve on the calibrated tank was then closed, and liquid nitrogen was applied to the nipple on the storage tank to recondense all UF^ gas in the system other than that in the calibrated tank. Using the vapor pressure of UF(S^ as given in Reference 2^ for the observed room temperature , the weight of the UF^ sample was calculated by the ideal gas law. After five minutes the manometer valve was cracked to determine that all of the UF^ had been condensed in the tank, and the valve on the storage tank was then closed.

PAGE 42

33 The UF^ was transferred to the capsule by opening the capsule and the calibrated tank valves and applying liquid nitrofen to the bottom one-half inch of the capsule for two minutes. The manometer valve was opened, and the residual gas pressure (generally about 3 mm.) was checked to ascertain that the transfer had been completed. The calibrated tank valve and the capsule valves were closed, and the system was pumped to vacuum with the liquid nitrogen still on the bottom of the capsule. The vacuum valve was closed, and the calibrated burette valve was opened so that the CgF^^ could be transferred to the capsule. After the desired amount of C^jF-,^ had been transferred to the capsule, the capsule valve was closed, and liquid nitrogen was applied to the calibrated glass burette. After the manometer indicated that the CgF-,^ had been condensed, the calibrated burette valve was closed, and the burette was allowed to come to room temperature. The amount of C^F-j^^ transferred was then determined by difference . A gold wire weighing about 1 mg . was attached to the capsule as discussed in the previous section.

PAGE 43

34 UFTR Irradiation The reactor irradiations were performed in the University of Florida Training Reactor which has been described by Boynton (29). The plastic handles on the capsule valves were removed, and the capsules to be irradiated were inserted into an aluminum cannister which was 10^ in. long with a 3-in. outside diameter. A 1-ft. section of a graphite stringer with a 4in. x 4-in. cross section was removed from the thermal column of the UFTR. This stringer was immediately adjacent to the core. The aluminum cannister was inserted in the 1-ft. void, and the remainder of the stringer and shielding blocks was reinserted. Upon removal the cannister was monitored and stored in a fuel storage pit until the radiation level had dropped to levels which permitted transfer of the capsule to the vacuum system. Gamma Irradiation The gamma irradiations were performed in the University of Florida pool-type food irradiator. The 30,000 curie Co"*-* source was in the form of 96 plates which were placed so that a relatively flat gamma flux was obtained. The capsules were lowered into the containers within the irradiator and exposed to the desired dose.

PAGE 44

35 The containers were filled with water. Only one capsule of the entire series of gamma capsules was found to have leaked water. The gamma dose was measured by the Fricke dosimetry method (30). Analysis The analysis of the samples was performed with a Model 700 F & M Chromatograph which has a thermal conductivity detector. The conditions used for the various analyses and the associated calculations are given in the appendix. The capsule containing the sample to be analyzed was attached to the vacuum system. The vacuum system was both pressure and vacuum tested up to the capsule valve. The sample was then transferred to a glass bulb in the vacuim system using the vacuum transfer technique . The samples which contained UF^ were passed through the sodium fluoride trap to remove the UF^ and fission fragments. If the sample transferred to the glass bulb was discolored, the sample was transferred back through the sodium fluoride trap to the capsule. The sample was then transferred once more through the sodiimi fluoride trap to the glass bulb so that the sample was passed through the trap for a total of three times. No sample required more than three passes.

PAGE 45

36 The bulb was surrounded by a mixture of ice and water and allowed to stand for one hour so that equilibrium could be established. The gases over the sample were then chromatographed several times using the chromatograph gas sampling valve with either a 1 milliliter, a 2 milliliter, or a 10 milliter gas sample loop. A silica gel column was used for this analysis. The voliune of the sample was measured in the glass burette on the vacuum system. The sample was then transferred to the 12 liter flask of the glass vacuum system and vaporized. The glass vacuum system is shown in Figure 6. Chromatographs were made of the completely vaporized sample using a silicone gum rubber column. Finally, the sample was recondensed, and liquid samples were chromatographed with a silicone g\m rubber column to determine the amount of polymer. Neutron Pulsing The experimental equipment used for the neutron pulsing is shown in Figure 7. The neutron generator was a Cockcroft-Walton type accelerator, Texas Nuclear Corporation Model 150-lH. The neutrons were produced by bombarding the tritium target with deuterons. The energy of the neutrons was about 14 Mev. The neutron source was roughly isotropic and in this experiment was

PAGE 46

37 6 4) •P 0} CO o CO > m (0 o irN|_3

PAGE 47

0)

PAGE 48

39 placed under the tank containing the fluorocarbon. The tank was lined vdth cadmium to prevent neutrons which leave the system at higher energies from returning at lower energies. The detector was placed just inside the upper boundary of the system. The detector was a LiI{Eu) scintillation detector which was designed to detect thermal neutrons. The signal from the detector was fed into an amplifier and then to a scaler where the total counts were recorded. The signal was also fed into the Technical Measurements Corporation, Model CN-llC, 256 channel time analyzer. The experimental procedure was as follows. A pulse generator produced a signal which would initiate a steady neutron beam. After the neutron beam had reached steady state operation, the neutron beam was turned off, and the decay of the neutron population in the tank was recorded on the 256 channel time analyzer. The neutron pulse width was 50 microseconds and the channel width for the analyzer was 2 microseconds. Thus, the first 25 channels showed the neutron pulse, and the remaining 231 channels recorded the decay of the neutron population in the fluorocarbon. After the 256 channel time analyzer had finished counting, a new pulse was introduced, and the cycle was repeated. The pulse repetition rate was 9^0

PAGE 49

40 cycles per second. This procedure was followed until sufficient counts for a reasonable statistical analysis had been obtained. The moveable top was then changed to a different height, and the volume of the fluorocarbon was adjusted so that the fluorocarbon liquid level was at the moveable top. This adjustment gave the system a new geometric buckling. The pulsing was then repeated at the new liquid level.

PAGE 50

CHAPTER IV EXPERIMENTAL DATA AND RESULTS The experimental data and results are tabulated in Tables 3 through S. Table 3 presents the data for the gamma irradiation of CgF^^ in the Food Irradiator. The following explanation is given for the various side headings. The doses, in terms of rads, were determined by Fricke dosimetry. The per cent CgF-j^^ converted to gaseous products is the percentage of molecules in the original sample converted to any gaseous product ; likewise, the per cent C^F-^^ converted to dimer is the percentage of molecules in the original sample converted to any dimeric product. The calculations for these percentages are shown in the appendix. The per cent CgF]^5 converted to dimer (chromatograph) gives the percentage of the original sample converted to dimer based on the chromatographic analysis assuming each molecule of dimer to be equivalent to two molecules of starting material. The per cent C^F-j^^ converted to dimer (material balance) represents the percentage of the original sample converted to dimer based on the difference in the voliomes of the original and irradiated 41

PAGE 51

42 samples. Table Ashows the G values for the gamma irradiation of CrtF-,/;^. The G value is defined as the number of molecules of a given material produced per 100 ev. of energy absorbed in the sample. For example, for Sample No. 33, G(CF, ) is 0.129, and accordingly, 0.129 molecules of CF) are produced per 100 ev. of energy absorbed in the sample. The G (dimer) values are given for the amount of dimer produced as indicated by both the chromatographic analysis and the material balance loss. The disappearance of CgF-j^^ is given as G{-CgF-|^^) v/hich is the niimber of molecules of C^F-j^^ lost per 100 ev. of energy absorbed. The calculations for the G values are given in the appendix. Table 5 presents the data and results for the reactor irradiation of C^F-^S ^" ^^® UFTR. The method for calculating the integrated flux is given in the appendix. The per cent CgF-^^^ converted to gaseous products is the same as in Table 3An additional sample, No. 11, was irradiated with an integrated flux of 59 x lO-'-^ neutrons/cm. . Liquid samples were taken directly from this irradiated sample for dimer analysis, and the per cent of the original sample converted to dimer was found to be 0.40 per cent. This sample was the only one in

PAGE 52

43 this series that was checked for dimer. No gas analysis was made . The data and results for the reactor irradiation of C^Fig and UF^ are presented in Tables 6 and 7. The calculations for the data are given in the Appendix. Table B presents the data and results for the pulsed neutron experiment . The symbols used in Table S and the method for calculating the decay constants are described in Chapter V. The range of error for the decomposition percentages given in the tables is estimated to be + 10 per cent. The range of error for the G values is estimated to be + 15 per cent.

PAGE 53

44

PAGE 54

45 •00 O r-i X cv iH CO o H 00

PAGE 55

46 to ON to -dO

PAGE 56

47 TABLE U G VALUES FOR GAWik IRRADIATION OF CgFj^^, SAMPLE NO. 33 3^ 39 34 Dose (rads) 2.4 X lo'^ 4-3 x lo"^ 4-3 x 10*^ 5-9 x lo"^ G Values CF, 0.129 4 C^Fz; 0.020 2 o C2F^0 0.004 ^3^3 CoFcjO '3'S 0.176

PAGE 57

4d TABLE i+ — Continued SAMPLE NO. 36 12 l3 Dose (rads) 6.& x 10^ 12.2 x lo"^ 14.2 x lo'^ G Values ^2^6 C^F^O Vb C3FgO G (Gas) Total 0.014 0.001 0.001 0.001 0.205 0.140

PAGE 58

49 o o o H -*

PAGE 59

50 C\2 o to o

PAGE 60

51 O o o o H O r-\ X o o o M E-t :=> CQ M Cti E-i vO CO Ct, MED Q Q Q sO O r-{ Oh to o CO oo w COS O H Q < SH COD:^ W« OH oo WW 0-. cc: ^S oo M W f-< H CO O o o M O o o -4O O O O O vO --—

PAGE 61

52 vO -^

PAGE 62

r-i O 53 X vO

PAGE 63

54

PAGE 64

55 o r< -4— .

PAGE 65

CHAPTER V DISCUSSION OF RESULTS Radiation Studies Gamma Irradiations The gamma irradiation of C^F-^^ yielded two groups of products. One group was composed of gaseous products, and the second group was composed of heavypolymeric products. The ratio of CgF,^ converted to polymer to CgF^^^ converted to gaseous products was 111 to 1. This value is the average of the data given in Table 2 for per cent CgF^^ converted to dimer and per cent CgF-,^ converted to gaseous products. The principal gaseous products of the gamma irradiation of CgF]^^ are seen from Tables 3 and 4 to be CF^ and C2F5 in the ratio of about 4 to 1. Smaller amounts of C2F5O and CoFg were also found, as well as a compound tentatively identified as C-^FgO. Identification of C3F^0 was based on the chromatographic retention time of a similar compound as reported in Reference 20. The polymeric material was found by chromatographic analysis, and based on mass spectrometric analysis using 56

PAGE 66

57 the equipment described in Reference 19 , was identified as being dimeric . The dimeric material was composed of at least two different molecules. The first molecule had a molecular weip;ht range of 750760, and one possible isomer is shown below. The second molecule had a molecular weight range of 805815, and one possible isomer is also shown below. CF ^CF FgC CF — CF3 FgC^ XF — CF^ F2C CF FC CF^ \./ \„/ CFj CF^ Product, Molecular Weight, 750-760 f3 CF F2C CF — CF^ CF3 CF^ F C CF — CF2 CF2 CF^ CF — CF CF^ \..-^ CF2 Product, Molecular Weight, 805-815

PAGE 67

58 The average G values for the production of gaseous products and of polymer are in good agreement with the results of other investigators as shown below. Compound G (Gas) G (Polymer) Investigators ^2^6 0.3 3.1 Fallgatter and Hanrahan ( 19 ) ^6^*6 0-3 1.1 Bloch, MacKenzie, and Wiswall (16) CgF^^ 0.2 2.5 This study A mechanism may be proposed which accounts for the various products formed during the gamma radiolysis of CgF-|^^. A gamma photon strikes a GgF,^ molecule and produces a CgFn^^+ ion and a F" ion. This reaction and the subsequent reactions may be represented by the following equations: CgFi6 -^ CgFi^"" + F(1) ^^^is"" ^ ^' -^ Vl3* "^ ^^3* ^i^ CF3+ F-> CF^ (5) ^7^13^ Vl5-> ^15^23 ^6) ^15^23 "" 2F -^ C15F30 . (7)

PAGE 68

59 Primary chemical change occurs in reactions (_1) and (_2). Subsequent products are then produced by the remaining reactions. Dimer is produced by the reactions shown in equation Q) and in equations iU) , (6), and (7). The major gaseous product, CF, , is produced by reactions (4) and (5^) • The other gases are produced by various recombination reactions of the perfluoromethyl radical. The oxygen source for the production of CpF^O and C-^FgO was probably dissolved and adsorbed air. The production of gaseous products and dimer are shown to be linear with respect to total dose by the curves in Figures S and 9. The equation for the curve in Figure S ±3 log -" X 100 = 0.025 + 0.53 x 10 ^D ^ ^0 A Nn N where — KT X 100 = per cent of original sample con^0 verted to gas, and D = total dose in rads. The equation for the curve in Figure 9 is log "^0 ~ ^ X lool 0.29 + 0.91 X 10-7 d ^0 1 Nq N * — W^ — ^ = V^^ cent of original sample converted to dimer, and

PAGE 69

60 o iH •00

PAGE 70

61 \

PAGE 71

62 D = total dose in rads. The chemical structure of CgF-j,^ shown below indicates that there are sixteen possible ways for a C^^i^ molecule to lose one fluorine atom. I ^ CF FgC CF — ^CF F5C CFp CF2 Perf luorodimethylcyclohexane Since two ChF-,^ molecules are required to form one dimer molecule, there are 16 x I6 or 256 possible ways to form a molecule of dimer from two molecules of CmF-,|^. There are two ways for one CFo group to be separated from a parent molecule, or, since two molecules were considered in the dimer formation, there are 2x2 or U ways in which a CF-j p;roup may be separated from two CgF-j^^ molecules. Thus the ratio of possible ways to form dimer to the possible ways to form gas is 256 to 4 or 64 to 1 as opposed to the experimentally found ratio of 111 to 1. The foregoing discussion has considered the decomposition of C^F-j^^ by gamma radiation in terms of the products yielded and in terms of the fraction of

PAGE 72

63 original material converted to other materials. A different analysis may be made by considering the . fraction of starting material remaining after a given exposure to gamma radiation. Let dN be the niomber of parent molecules decomposed by a dose dD on the sample which contained N molecules. Since the number of molecules decomposed is proportional to the number of molecules present and to the dose, the proportionality can be given as d^oL NdD, and letting k be the proportionality constant , the relationship becomes dN = kNdD where the minus sign indicates that dN is the number of molecules lost. Rearrangement gives which can be integrated from N= Nq to N = N and from D to D = D. The result is In N = -kD , or log {{_ = -KD where K = 2.303 The results for the gamma irradiation of C^F-j^^ are plotted in Figure 10 as log {Ex 100) versus D. The

PAGE 73

64

PAGE 74

65 curve is a straight line, and K is found to be 0.042 rad ~ . Therefore, it is possible to predict the number of molecules of C^F^^ remaining in a sample with a known number of starting molecules after the sample has been given a dose of gamma radiation. Since the dimensionless ratio N/Nq is used, the units might be grams or any other convenient unit which is related to the number of molecules. The question arises as to whether or not this relationship holds only for CgF-j^^ and only for the range of doses used in this investigation. The data available from the literature at the present time for the decomposition of fluorocarbons by gamma irradiation are given in Table 9 and shown in Figure 11 . Examination of Table 9 shows that the data used in Figure 11 cover a wide variety of compounds including alicyclic, aromatic, and aliphatic fluorocarbons. Other variables include type of radiation, dose rate, total dose, and experimental conditions. There are two distinct curves in Figure 11. One curve relates data for aromatic fluorocarbons, and the second curve correlates data for alicyclic and aliphatic fluorocarbons. The decomposition of aromatic fluorocarbons is considerably less than the decomposition of aliphatic fluorocarbons and alicyclic fluorocarbons,

PAGE 75

66 o

PAGE 76

67

PAGE 77

63

PAGE 78

69 and the curve for aromatic f luorocarbons is a straight line over the dose range for which data are available. The curve for aliphatic fluorocarbons and alicyclic fluorocarbons is reasonably straight in the lower dose range; however, the curve tends to slope downwards at higher doses. This result is unexpected since it might imply that the gamma radiation becomes more effective at higher doses. A better explanation is that the number of small decomposition product molecules becomes very large at high doses. These small molecules absorb a large fraction of the gamma energy, but because of their more stable structure, do not decompose. The energy absorbed by the small molecules is transferred to the heavier molecules, and the actual energy absorbed by the remaining parent molecules is greater than that due to the absorption of gamma radiation alone; therefore, the previously developed relationship must be modified to take into consideration the build-up of CF, and other light gaseous products. The number of molecules decomposed is proportional to the number of parent molecules present plus the number of CFi^ molecules generated. The number of CFi molecules present after some arbitrary exposure is given by the following expression:

PAGE 79

70 No. of CF, molecules present = no. of CgF-j^^ molecules decomposed 1 molecule of CdF-,^ decomposed to gas jd _____^ , ___™_— — ^^— — — ^-^— — — — — 111 molecules CgF-j^^ decomposed 3 molecules CF^^^ produced ^ . — ___ — _ — , — • molecule ^^^2.6 decomposed to gas The number of molecules of CgF^^ decomposed is simply the number of molecules of CgF^^ originally present minus the number of ChF-i^ molecules remaining after the exposure. If the number of molecules is given in terms of fractions with Nq = 1 , then the following relationship can be found for the number of CF^ molecules present . No. of CF/^ molecules present = (1 N) X ^ X g = 0.072 (1 N) . The original relationship becomes dN = -kNdD k []o. 072(1 N)]dD, and it can be rewritten as dN = (-0,92SN 0.072)kdD, or dN _ . ,p, 0.923N + 0.072 ~ ^^" • Integration between the limits of N = N and N = Nq and of D = D and D gives

PAGE 80

71 072 __1 |r o.92gM +0.0 0.923 L°-923No + 0072 -kD. By lettin/^ K = ^ '928k ^^^ ^o = 1, the expression becomes log (O.92SN + 0.072) = -KD . Figure 12 is a plot of log (0.92SN + 0.072) versus D, and the constant K was found by a least squares fit to have a value of 2.13 x 10" rad" for the aromatic fluorocarbon curve and 6.64 x 10"-'-'-* rad"-^ for the aliphatic and alicyclic fluorocarbon curve . Both curves are straight lines and can be used to predict the fraction of a sample which would be decomposed by a known dose of radiation over the wide range of parameters given in Table 9. Reactor Irradiations Irradiation of CgF^^ — The reactor irradiations of C gF-|^^ gave results very similar to the results for the gamma irradiation of CgFj^ . The principal gaseous products were CF^ and C2F^ in the ratio of about 4 to 1. This finding is in agreement with data obtained by Davis for the irradiation of C^F,^ in the Low Intensity Test Reactor (31). Chromatographic analysis indicated that there were also diraeric materials similar to those found in gamma irradiated samples.

PAGE 81

72 0}

PAGE 82

73 The per cent decomposition of the parent compound to p;aseous products is plotted versus integrated flux in Figure 13 . The curve is a straight line over the range of integrated fluxes used. The equation for the curve is = -13 .2 + 1.1 log nvt log -0 X 10^ -Wo for the range 10^^ < ^vt < 10^^ where x lo'^ = per cent of original sample conNq verted to gas x 10^ The relationship between the neutron flux and the gamma flux in a graphite moderated reactor was studied byRichardson, Allen, and Boyle (32). Expressions were derived which permit the conversion of integrated neutron fluxes in terms of neutrons per cm.into absorbed gamma and neutron doses in terms of rads. These conversion factors were used to convert the integrated neutron fluxes for the reactor irradiations into rads. (See appendix. ) The data for the decomposition of ^g^i^ to gaseous products by reactor irradiation were combined with the data for the decomposition of CgF-^^ to gaseous products by gamma irradiation. The combined data are plotted in Figure lU , and the continuity between the two sets of data is very good. The equation for the curve is

PAGE 83

74 10^3 10^^ 10^5 Integrated Neutron Flux (neutrons/cm. 2) Figure 13 . Cg^ie, Decomposed to Gas for Reactor Irradiation . 10^

PAGE 84

75 10^ 10 -1 to o 5 10-2 •o «> n o a s o o « tTlO to o +J c «) o c (X. -3 10 -4 ® Gamma Irradiation H Realtor Irradiation 10 -5 103 10^ 105 10^ Dose (rads) 10 10^ Figure 14CgF^^^ Decomposed to Gas for Gamma And Reactor Irradiation

PAGE 85

76 ^Q ' ^ X 100 = 0.007 X 0.72 x 10~*^D No where ^ " — = fraction of sample converted to gas, and N D = total dose in rads. An analysis of the work by Richardson, Allen, and Boyle reveals that the enerp;y deposited by the gamma flux in a reactor of the type being used in this work greatly exceeds the energy deposited by the neutrons which are primarily thermal neutrons. Since most of the energy deposited in the sample results from gamma photon interactions, the reaction mechanism and rate equations which can be applied to this case are identical with those developed in the previous section. There was no evidence of any fluorine or hydrogen fluoride. Some C2F5O was found in the mixture. The oxygen required to produce this compound was probably either adsorbed on the walls of the capsules or dissolved in the f luorocarbon. A calculation was made to compare the decomposition of CgF-j^^ in a reactor radiation field with the decomposition of the coolant used in the organic cooled reactor. (See appendix.) This coolant is a mixture of hydrocarbons including biphenyl and terphenyls. The data for the Piqua Nuclear Power Facility were used for comparison of the coolants (33)* I^ ^S^16

PAGE 86

77 were to be used as the coolant in the PNPF, the decomposition rate for the CgF^^ would be about 0.50 per cent per hour as compared with a calculated value of 3.0 per cent per hour for the organic coolant. In the light of available data CgF-j^g would be a better coolant for the PNPF than the currently used organic coolant . Other important advantages of the CgF^^ as a coolant would be its higher thermal stability and its lower corrosiveness as compared to the organic coolant. Possible disadvantages of using CgF-j^^^ would be its higher cost and the poorer heat transfer properties generally ascribed to f luorocarbons ( 1 ) . Irradiation of CgF-j^^ and UF, . — The results of the reactor irradiation of CgF-j^^ and UF^ differ significantly from the reactor irradiation of CgF-]^^ and from the gamma irradiation of CgF]^^. The most significant difference is the absence of polymeric material in the reactor irradiated solution. An extensive analytical effort which is described in the appendix was directed towards the confirmation of this finding. No evidence of polymeric material was found by any method. A second significant difference was the different distribution of the gaseous products. For the reactor irradiations of CgF25 and UF^, roughly equal amounts of CF, , ^2?^,

PAGE 87

78 and C2F6O were found with lesser amounts of C-jFg and C^F^O; whereas in the gamma and reactor irradiations of CgF]^5, the major gaseous products were CF, and C2F5 in the ratio of about four to one with only small amounts of C2F5O, C^Fg, and C.FgO. A mechanism which accounts for these differences can be proposed. In the reactor irradiation of C^F-^^ and UF^, the U -^^ nuclei fission into two fragments. These two fragments dissipate their kinetic energy — about 185 Mev. — into the surrounding medium. These two fragments strike the fluorocarbon molecules and shatter the ring at three or more bonds, producing three or more fluorocarbon radicals. Since each fission also releases six fluoride ions or fluorine atoms, an excess of fluorine is available for recombination with the fluorocarbon radicals. In addition it has been previously reported (25) that the neutron irradiation of UF5 yields UF^ and free fluorine with a G value of 0.5 molecules of UF^ decomposed per 100 ev. of energy absorbed; consequently there is another source of fluorine in the system. In addition to these sources of fluorine, the compound UF^ is itself a potent fluorinating agent. The processes can be represented as follows:

PAGE 88

79 on-'+ U^^^F^ -> 2 fission fragments + 6 fluorine atoms or ions + neutrons + gammas. Collision of fission fragments with the surrounding fluorocarbon molecules shatters the molecules and produces various fluorocarbon radicals such as CF^*, ^2^5'' ^"^ '^3^7* * "^^^ excess of fluorine immediately combines with the radicals to form small saturated fluorocarbons such as CF. , ^2?^^, and C^Fg. The oxygen containing compounds are formed by the recombination of necessary radicals with oxygen which is present in the UF^; i.-©.. , two CFo radicals could combine with one oxygen atom or ion to form C2F^0, and likewise one CF^ radical and one C2Fr radical could recombine with one oxygen atom or ion to form CoF^O. The reactions which have been discussed are demonstrative and do not necessarily include all possible reactions; however, the "proposed mechanism" does account for the lack of dimer because the parent molecule is shattered into small fragments by the primary act, and these small fragments are immediately saturated with fluorine . Consideration of the G values for the gamma irradiated CgF-]^^ and for the reactor irradiated Cj^F^^ UF^^ confirms the proposed shattering of the fluorocarbon

PAGE 89

60 molecule. The G values for gas production from the solution of CgF-j^^ and UF^ are higher by a factor of ten than the G values for gas production from the CgF^^^ alone when irradiated with gamma photons. In the reactor irradiations of CgF-j^^ and UF^,the major products are CF^, C2F^, and C2F5O; and one molecule of CgF-j^^^ thus produces eight CF^^ molecules, four C2Fg molecules, four C2F5O molecules, or some combination of these molecules. In the gamma irradiation, a sizeable portion of the energy deposited goes into the production of polymeric material. In this case one molecule of CgF^^ produces only 0.5 molecule of dimer, and the number of molecules produced per unit of energy absorbed is less than when gases are formed. A comparison of the G values for the decomposition of O^F-^^ indicates that the energy required to decompose a molecule of CgF-i^ by gamma radiation is the same order of magnitude as the energy required for decomposition by fission fragments. An expression may be derived which relates the number of CgF^^ molecules converted to gas with the integrated neutron flux. The number of molecules dN converted to gas is proportional to the number of molecules present N, the neutron flux nv, and the time increment dt .

PAGE 90

di The proportionality can be written as -dNcKNnvdt , and, if k is the proportionality constant, dN = kNnvdt. Examination of the data from Table 6 shows that the total decomposition is very small so that N S„^, and the expression can be written as -dN = knvNgdt . Integration from N = Nq to N = N and from t ^ to t = t gives -(N Nq) = knvNot , or No N ^0 = knvt The term on the left is simply the fraction of molecules converted to gas. The data are plotted in Figure 13 and the points fall along a straight line as expected. The equation for the CgF,^ + UF^ curve is log 1!£_1-^ X lo"^ = -9.^0 + 0.92 log nvt for the range 10^^ < ^^^ < jq^^ where _2_I — x 10' = per cent of original sample ^0 r converted to gas x 10^, and

PAGE 91

d2 nvt = integrated neutron flux in neutrons/cm. Several other experimental results are worthy of note . Solutions of UF5 and C^Ti^S ^®^® prepared and were found to be homogeneous with no apparent side reactions. The solubility of UF^ in C^F^^^ was found to be about 0.65 gm./inl' The fluorocarbon was successfully separated from the fission fragments and the UF5 by passing the samples through a sodium fluoride trap with the samples being in the vapor phase. The largest number of passes required was three with some separations being complete after one pass. A light film of greenish white powder remained in the capsule. This powder was probably UF5 which had been chemically reduced to UF/^. The radioactivity of the separated fluorocarbon was barely detectable. Thus, separation of fluorocarbons from UF5 and fission products has been found to be a simple process. There was no evidence of elemental fluorine or hydrogen fluoride in the capsules. A calculation was made to ascertain the decomposition of a CgF-j^^UF^ solution if the solution were to be used in a reactor similar to the Homogeneous Reactor Experiment I, HRE-1 (34). Assuming the same concentration of \r^^ per unit volume and the same flux level as the HRE-1, a typical decomposition rate for the CgF25 solvent was found to be 93 per cent per hour.

PAGE 92

S3 This rate of decomposition and the cost of make-up feed would counterbalance the fact that the solution is electrically nonconducting and hence noncorrosive . The decomposition rate was obtained by extrapolating the decomposition rate an order of magnitude higher than the highest data point. This assumption of linearity of decomposition with dose may not be completely justified. Another factor which should be considered before this result is extended to other f luorocarbons is whether or not CgF-j^g is a good representative material. In previous investigations, the alicyclic f luorocarbons, of which CgFj^ is a member, have been found to be considerably less radiation resistant than have the aromatic f luorocarbons .

PAGE 93

34 Neutron Physics Parameters In order to evaluate the f luorocarbons as working fluids in a nuclear reactor it is necessary that their neutron absorbing and scattering properties be known. Conventional static experiments for measuring these properties require large amounts of material, but similar information can be obtained with a small sample by measuring neutron decay with the pulsed neutron technique. The basis of this technique can best be shown by examining the time dependent equation for the neutron flux in a nonmultiplying medium. The neutron flux as a function of space and time

{r,t) =(pjr)e-^''^a + ^^B^g^^ where the neutron flux
PAGE 94

S5 interactions with the nuclei of the medium. The material buckling B^ fo^ ^ cylinder is given by 52 = 2.405 ^ + JL 2 H or 'N^ /^ where R = radius + 2 D and H = height + 4 D . The spatial dependence of the neutron flux can be removed if the flux is considered at one specific point in the system. In this case ln(p(t)'~(vZ^^ + DvB^ )t if the natural logarithms of both sides are taken. If the neutrons are assumed to be monoenergetic , all of the terras inside the parenthesis are constants for this case, and these constants can be represented by a single term A . Substituting the decay constant in the previous relationship gives the simple expression In q5 ( t )'~-Xt , and A , the decay constant , can be determined by measuring the neutron flux as a function of time. If the physical dimensions of the system are varied and the decay constant measured for each size system, the macroscopic absorption cross section and the diffusion coefficient can be determined as is shown by Figure 15 . The plot of A versus B^ has as its intercept with the ordinate, the quantity vSla> ^""^ ^Y knowing the neutron velocity, it is quite simple to calculate L ^. Likewise, the slope is Dv which will

PAGE 95

66 X (aecT^) r vr. X = V I^ + vDB^g B^g (cm72: Figure 15. Analysis of Pulsed Neutron Experimental Data.

PAGE 96

57 yield the diffusion coefficient if the neutron velocity is known. As was apparent from the mathematical definition of buckling, the buckling decreases as the physical size increases. Therefore, in order to get a good value of the intercept v Z , it is iraoortant to have points on the curve near the ordinate, or in other words, the system should be as large as possible. This is an expected result since B^ = o actually represents an infinite system which would lose no neutrons to the surroundings . The results of a typical run are shown in Figure 16 which is a plot of counts per channel versus channel nxOTiber or time . The curve has no region where the slope is constant for more than 2 or 3 microseconds. This result was characteristic of all of the runs and led to the conclusion that there was a steady state contribution to the neutron decay which was independent of the f luorocarbon. A run was made with all equipment arranged as before with the exception that there was no f luorocarbon in the tank. As shown in Figure 17, there is a decay of the neutron population in the empty tank which is significant when compared to the decay of the neutron population in the f luorocarbon. This decay could possibly be attributed to neutrons leaking from the system, being moderated and reflected in the room, and being returned at energies slightly above the cadmium

PAGE 97

63 5 X 10^ 4 Counts Per ^o4 Channel 6 2 X 10End of Neutron Pulse Data from Run No. 12 H 12.7 cm. CgF^g in Tank B^ = 0.0461 cm. -2 -I I 1 I I I I I L. S 12 16 20 24 Time {/usee.) 2S Fifrure 16. Counts Per Channel Vs. Time for CgF-,^

PAGE 98

39 5 X 10^ 4 h 3 Counts Per 10^ Channel 2 X 10 End of Neutron Pulse Data from Run No. 10 H = 12.7 cm. Empty Tank b2 = 0.0/,6l cm. -2 E J I I I I 1 u 12 16 20 24 Time (/< sec. ) 28 Figure 17. Counts Per Channel Vs. Time for EmptyTank .

PAGE 99

90 cut-off. Another possible cause would be moderation of the neutrons by the air in the tank. The result of a channel by channel subtraction of the decay of the neutron population in the empty tank from the decay of the full tank is shown in Figure ig. A reasonably straight region for which a decay constant can be determined was given by the subtraction of the two curves. The decay constant for this curve was calculated using a least square fit of the data. This subtraction procedure was repeated for other runs with similar straight regions being found. It is to be noted that the subtraction was performed channel by channel rather than smooth curve from smooth curve. This channel by channel subtraction was, of course, a much more severe treatment of the data. Figure 19 presents the plot of X versus B^g for CgFjg. The data were least square fitted by the Non Linear Least Squares (NLLS) computer code. The points are in a reasonable straight line; however, the point for the lowest buckling, that is the maximum volume, is not too close to the axis when it is realized that the intercept is of the order of 0.007 and the first point has a value of about 3.1. The computer code gave a value for the intercept of 0.007S sec." . A statistical analysis showed that the deviation with a 95 per cent

PAGE 100

91 5 X 10 Counts Per Channel 105 X 10^ Run No . 12 Run No . 10 H = 12.7 cm. Channel by Channel Subtraction b2^ = 0.0461 cm."^ Z 2k 4 S 12 16 1^ Time (/{ ssc . ) Figure IS. Counts Per Channel for CgF25 Minus Counts Per Channel for Empty Tank Vs. Time . 2S

PAGE 101

92 7 X 10^

PAGE 102

93 confidence level is + 0.147S. Since there is not more ^3^16 available and accordingly it is not possible to pulse larger volumes to get points closer to the intercept, the intercept of the curve was taken to be less than 0.007^ + 0.147^ or less than 0.1556 sec.'-'-. Assuming a thermal neutron velocity of 2.2 x 10^ cm./ sec, the macroscopic absorption cross section for CgF-j^^ is less than 0.0067 cm."-^. The macroscopic cross section compares very favorably with the other materials shown in Table 10. Similarly, the slope was found to have a value of (1.396 + 0.524)x 10^ cm.^sec. which gives a diffusion coefficient of 6.3 5 ± 2. 33 cm. if it is assumed that the thermal neutron velocity was 2.2 X 10 cm. /sec. The calculated value of the diffusion coefficient for CgF-j^^ is 1.13 cm. so the experimentally determined value appears to be high by a factor of about six. Since the scattering cross section is inversely related to the diffusion coefficient, it seems that CgFjg is not as good a scatterer as would be expected; however, some of the apparently low scattering ability may actually be due to the high leakage associated with a small system.

PAGE 103

94 TABLE 10 MACROSCOPIC ABSORPTION CROSS SECTIONS OF MODERATORS Macroscopic Absorption Material

PAGE 104

CHAPTER VI CONCLUSIONS Radiation Studies Gamma Irradiation of CgF-j^^ The gaseous products formed in the gamma irradiation of CgF]^^ consisted primarily of CF^ and C2F^ in the ratio of four to one. Polymeric material was found and was identified as being dimeric. The average ratio of the CgF-j^^ converted to polymer to the G^'^iS converted to gaseous products was 111 to 1. The G values for gaseous products and for polymer compared very favorably with values for other f luorocarbons as obtained by other investigators. An expression was derived which relates the fraction of CgF^^ remaining after exposure to a .riven dose of gamma irradiation. The correlation was extended to other f luorocarbons with equations being found to describe the dosedecomposition relationship for aromatic fluorocarbons and for alicyclic and aliphatic fluorocarbons. At higher doses it was necessary to add a correction factor for the amount of gaseous products formed. 95

PAGE 105

96 No evidence of hydrogen fluoride or fluorine was observed in any of the experiments. Reactor Irradiation of ChF 3^16 The gaseous products formed in the reactor irradiation of CgF,^ corresponded to those formed in the gamma irradiation of CgF^^. Similarly, diraeric material was found. The data for the reactor irradiation were converted to the same energy units as used in the gamma irradiation. The data for the fraction of the CgF25 converted to gaseous products by reactor radiation and by gamma radiation were combined, and it was found that the combined data could be described with a single equation. It was noted that in the reactor irradiation the decomposition caused by the thermal neutrons was very slight . An analysis for dimeric material was made on one sample . The amount of dimer found agreed with the amount of dimer foiind in the gamma irradiated samples for the same dose . The correlation of data from gamma irradiation and reactor irradiation of C^F^^ indicates that a gamma source can be used to predict the behavior of fluorocarbons in a mixed gamma-thermal neutron radiation field.

PAGE 106

97 Reactor Irradiation of CgF^^^, and UF^ This work showed that a solution of a fluorocarbon and UF^ can be irradiated without any abnormal reactions. The separation of the irradiated fluorocarbon from the UFg and fission fragments can be accomplished by passing the vapor phase mixture through a simple sodium fluoride trap. The distribution of gaseous products was significantly different from the product distribution of gamma irradiated samples. The materials CF^, C2F5, and C2F^0 were found in roughly equal proportions . No polymeric material was found. These facts were combined with differences in the G values to show that the fission fragments decompose, or shatter, the large CgFj^^ molecule into smaller fragments. Excess UF(^ acts as a source of fluorine to saturate these fragments so that no polymeric material was formed. A calculation was made assuming that a CgF-j^^ and UF5 solution was used to fuel an HRE-1 type reactor. The decomposition of the C^'^iS was shown to be about 93 per cent per hour; however, the calculation cannot be considered to be exact since an assumption was made concerning the extrapolation of decomposition data. In addition, other investigators have shown that aromatic f luorocarbons are considerably more stable to

PAGE 107

9& gamma irradiation than alicyclic fluorocarbons such as ^3^16; however, no data exist about the effects of fission fragments on aromatic fluorocarbons. Neutron Physics Parameters The value obtained for the macroscopic absorption cross section for CgFjg is indeed low, and CgF^^^ could be satisfactorily used in reactors from the standpoint of low neutron absorption. The neutron pulsing experiment did point out that it was necessary to consider the decay of the neutron pulse in the empty tank. This consideration has not been necessary in previous experiments with more highly absorbing materials. In order to determine the precise values of the neutron physics parameters for ^S^l6» a system about four times as large as the available system must be used.

PAGE 108

APPENDIX A CALCULATIONS Calculation of Per Cent Gaseous and Polymeric Products Formed Durin,^ Irradiation The calculation will be performed for Sample No. 36 using the data obtained from the chromatograms for Sample No. 36 which are shown in Figures 20, 21, and 22. Three chromatographic analyses are required for the sample. In the first analysis, as shown in Figure 20, a silica gel column is used to determine the relative amounts of the different gaseous products which are present over the liquid sample. The second analysis uses a silicone gum rubber coliomn to separate the gaseous products from the CgFj^^ in the completely vaporized sample as shown in Figure 21. In the third analysis the dimer is separated from the parent compound with a silica gel column as shown in Figure 22. The data obtained from the three analyses are given in Tables 3 through 7. The supporting data for the figures are given in Tables 11, 12, and 13The per cent of the original sample decomposed to gas can be determined from information obtained from Figures 20 and 21. The area under each gaseous product peak 99

PAGE 109

< 100 is found by graphical integration using the extrapolated base line as shovm in Figure 20. The areas are given in relative units. The per cent of total area is the ratio of the area for a given product to the total area for all gaseous products multiplied by 100. Per Cent of Total Area for Gaseous Products Material

PAGE 110

101 For each eight CF. molecules created, one CgF-j^^ molecule was destroyed. Likewise, four ^2^6 °^ four CoF^O molecules were produced by the destruction of one CgF]^5 molecule, and eight C^Fg or eight C^Fj^O molecules were the yield from three molecules of CgF^^^^. Therefore, in terms of C^F^^ destroyed, area C^Fi^ area CFji^ 95^ converted to CF. = g = —3 — = II9.S, area CgF^^^ ^ area C^F^ _ 145 ^ ^ ^ converted to C2F^ 4 4 -^ • » area C^F^^ area C2F5O 11 converted to C2F^0 " 4 " X "2.8, area CgF-j^^ area C^Fg 20 converted to C3F = -^ >: 3 = 7.5, area C^F.. area CoFj*0 19 ^16 = g ^ ^ = -^ X 3 = 7.1, and 3 converted to C^FgO total area for gaseous products = 173 .4 • in terms of CgF^^^ From Figure 21, the area for the total gaseous products and the area for the remaining parent compound are found to be 41.0 and 7522.0, respectively. The chromatogram shown on the left side of Figure 21 is used to determine the air leaking into the gas sampling valve

PAGE 111

102 during the filling of the ,^as sample loop. The number given above for the area of the gaseous products has been corrected for the air leaking into the gas sampling valve. The air mentioned in the remainder of this calculation is air that was in the gases of the irradiated sample . Since the area caused by the siases represents the products, the product area must be converted to an equivalent area of CgF^^ using the ratios previously established from Figure 20; i.e., 173.4 to 117^. An additional correction must be made since there is air in the campletely vaporized sample. The area for the gaseous products can be converted to area in tenns of CgFj5 as follows: 41.0 (gaseous products + air) ^ 1153 (gaseous products) 1173 (gaseous products + air) 173-4 (gaseous products in terms of CgF,^) 1153 (gaseous products) = 6.1 (area of gaseous products in terms of CgF-,^). The amount of gases in the vaporized sample can now be determined. The area for the O^F-^f^ converted to gaseous products is 6.1, and the area for the unchanged CgF^^^ is 7522.0; therefore,

PAGE 112

103 6.1 X 100 = 0.080 per cent of the vaporized 7522.0 + 6.1 sample was gaseous products. A similar calculation can be made for the conversion of CgF-j^^ to polymer. From Figure 22 the area of the unchanged CgF-j^^ is seen to be 26, 53^ area units, and the area of the polym.er is 1101 units. The total is 27,639. Assuming that the polymeric material is a dimer (as indicated by the mass spectrometer), two molecules of CgF^g are required to make one molecule of polymer. Thus, in terms of OgF^^, the area caused by the dimer is 1101 x 2 or 2202 area units, and the per cent of the original sample converted to polymer is 2202 27639 X 100 =8.3 per cent.

PAGE 113

104

PAGE 114

105

PAGE 115

106 V to CT5 (X, -a •H cr

PAGE 116

107 TABLE 11 SUPPORTING DATA FOR FIGURE 20 SAMPLE NO. 36 Dose = 6.5 X 10*^ rads. Volume = 2.05 ml. CgF-,^^. Volume of gas sample chromatographed = 2.0 ml. Pressure (metal system) = 22.0 mm. Temperature (liquid) = 0° C. Column, 2 ft., silica gel. Temperatures, ° C. Injector port 125 Oven 40 Detector 112 Metal System 24 Initial oven variac setting =0.0. Oven variac setting upon injection =4.0. Oven variac setting after appearance of gaseous products ( 9 min.) = 5.1. Filament current = 225 ma. Gas flow (helium) = 50 ml. /min. at atmospheric pressure, Time to fill gas sampling valve = 1 min. Gas sample loop = 2 ml. Chart speed = 1 in. /min. (low x 4) . Attenuation Air and CF^ x B All others x 1

PAGE 117

TABLE 12 SUPPORTING DATA FOR FIGURE 21 SAl^lPLE NO. 36 Dose = 6.S X 10*^ rads. Volume = 2.05 ml. CgF]^^ . Voliime of sample chromatographed = 10.0 ml. Pressure (gas flask system) = 11.0 mm. Column, 5 f t . , 10^ silicone gum rubber. Temperature , ° C . Injector port 172 Oven 50 (constant temperature) Detector 205 Gas flask system 24 Filament current = 725 ma. Gas flow (helium) = 50 ml./min. at atmospheric pressure Time to dill gas sampling valve = 1 min. Gas sample loop = 10 ml . Chart speed = 1 in. /min. (low x 4) . Attenuation Gases x 1 C^Fi6 X 32

PAGE 118

109 TABLE 13 SUPPORTING DATA FOR FIGURE 22 SAMPLE NO. 36 Dose = 6.8 X 107 rads. Volume = 2.05 ml. C^F-^^ . Volume of sample chromatographed = .005 ml. (liquid injection) Column, 5 f t . , 10$^ silicone gum rubber . Temperatures, ° C. Injection port 1^0 Oven 115 (constant temperature) Detector 200 Filament current = 200 ma. Gas flow (helium) = 50 ml./min. Chart speed = 1 in./min. (low x U) . Attenuation ^she ^ 256 Polymer x 1

PAGE 119

110 Calculation of G Values for Gamma Irradiation The data for Sample No. 12 will be used. Volume = 1.99 ml. Dose = 1/f.l X lO'^ rads. Composition of product p-ases (relative per cent) CF4

PAGE 120

Ill The number of molecules of CF, formed (92.5 per cent of total gases) is Id 5.30 X 10 molecules CgF^^^ to f^as molecules CgF,^ to CF, 8 molecules CFi X 0.925 X molecule CgFx6 to gas molecule CgF]^^ to gas = 3.93 X lO-"-^ molecules CF^ . Similar calculations may be made for the other gases: for C2F^ 5.30 X 10^^ X 0.037 X 4 = 7.65 X 10^"^ molecules, for C2F5O 5.30 X 10^^ X 0.017 X 4 = 3.60 X 10^"^ molecules , for C^Fg 5.30 X 10^^ X 0.013 X 2.67 = I.S4 X 10^7 molecules , and for C3Fg0 5.30 X 10^^ X 0.008 X 2.67 = 1.13 X 10^"^ molecules . The dose in water is related to the dose in CgF-j^^ by the equation electron density (CgF^^^) dose (CdF-,/-) = dose (HpO) x — — . \ — . ^ -i-o <; electron density (H2O) The electron density (CgF^^^) = 6.023 X 10^-^ molecules ^ ^02 electrons mole molecule _ 400 p:m. mole

PAGE 121

112 2.83 X 1023 ^^^^trons ^ gm. The electron density (H2O) 6 023 X 10^3 molecules ^ -^q electrons mole molecule 18 -^ mole -) -)i Tn23 elect rons 3 .34 X 10 gm. The doses then are dose (H^O) = 14.1 X 10" rads, or dose (CmF,^) = 14.1 X 107 rads x 2.88 x 10 " ^^ 3.34 X 1023 12.2 X 10'^ rads. In terms of ev., the dose is 100 ^^ 12.2 X 107 rads x SHLt x L_ev^^ rad 1.6 X 10""'-'^ erg, X 3.67 -^^^^^ = 2.80 X 1022 —^^ . sample sample The G value is defined as the number of molecules of product obtained per 100 ev . of enerj^y absorbed so that 3 .93 X lO-^"^ molecules CF/ produced G (CFjj^) = -^ , and 2.80 X 10^2 ev. x -j^ G (CF^) = 0.140 molecules per 100 ev.

PAGE 122

113 The other G values may be calculated in the same manner: G (C,F,) = ^-^^ l^'''' , = '•'> ''II = 0.0026, 2.80 X 1022 X jg^ 2.80 X lO^O G (C2F6O) = ^^Q ^ ^Q ' = 0.0013 . 2.80 X 1020 G (C.Fg) = ^-^^ ^ ^^11 = 0.0006 , i ^ 2.80 X lO^O G (G3FgO) = ^'^3 X 10^7 ^ ^^^^^^ ^ 2.80 X 10' The calculation for the G value of the polymer is based on the assumption that the dimer is composed of two molecules of the parent compound. The chromatographic analysis of the irradiated sample indicates that 13 1 per cent of the parent compound is converted to dimer so that oT molecules C^jFta 5.52 X 1021 Li£_ sample molecules CdFT^: to polymer X 0.131 — 2j£ l__i molecule CgF^^ in sample jj Q r molecules polymer _, molecule CgF-j^^ to polymer 3.62 X 10^ molecules of polymer .

PAGE 123

lU 20 Then G (polymer) = 3.62 x 10 molecules = 2.30 X 10^2 1.30 molecules polymer per 100 ev. The G value for the disappearance of C^F^^ is found as follows: total molecules destroyed = 7.24 x 10'^'^ molecules to polymer a + 5.30 x 10° molecules to gas = 7-29 x 10^^ molecules lost, and 20 G(-CdFT^) =7-29 X 10 — ^ 2.60 molecules destroyed 2.^0 X 1022 per 100 ev. absorbed.

PAGE 124

115 Calculation of G Values for Reactor Irradiation Of CgFj^5 and UF^ The data for Sample No. 19 will be used. Size = 3.03 ml. CgF-^^ + I.46 gm. UF^ . Integrated flux = 3.74 x lO-"-^ neutrons/cm.^ Composition of product gases (relative per cent) CF4

PAGE 125

116 The number of molecules of CF, formed (27.3 per cent of total gases) is ,^ molecule to CFi 1.65 X 10l9 molecules to gas x 0.273 — ^ ^ — molecule to gas ^ molecules CFi V _k molecule converted to CF, = 3.61 X 10-^^ molecules CF^ . Similar calculations may be made for the other gases: for 02^^ 1.65 X 10^^ X 0.306 X 4 = 2.02 X lO"'-^ molecules , for CgF^O 1.65 X 10-'-^ X 0.277 x 4 = 1.^3 x 10^^ molecules , for C3Fg 1.65 X lO-"-^ X 0.129 X g = 0.56 x 10^^ molecules , J and for C3Fg0 1.65 X 10^9 X 0.015 X g = 0.066 x 10^^ molecules . The energy deposited in the sample is calculated assuming that all of the fission energy other than the energy of the emitted neutrons, neutrinos, and gamma photons is deposited in the sample.

PAGE 126

117 The number of atoms of uraniiim in the sample was gnuUFg 1 mole UF^ ^•^^-liE^^ 352 gm. ^^-^23 X 10^-^ molecule ,_ 2. 50 x lO^"'" atoms uranium . mole * sample The total number of fissions is given by N2!£>(bt, where N = the number of fissionable nuclei, 2.£. = fission cross section of fissionable material, cp = neutron flux, and t = time of radiation . For this sample N = 2.50 X 10^''" atoms, 2.f = 3.92 X 10" cm. for natural uranium, and 4>t = 14.9 X 10^^ neutrons ^ cm. The total number of fissions 2.50 X lO^-"X 3.92 X 10-2^ X 14.9 X 10^^ = 145.5 X lO-"--^ fissions . The energy transferred to the C^F^g per fission is 1^5 Mev. so that the total energy to sample is 145.5 X 10^^ fissions x 1^5 x 10^ -rr^ fission = 2.69 X lO^-"ev. The G values may next be determined as follows: molecules CF^ produced G(CF^) = 100 ev. '

PAGE 127

lid 3.61 X 10^9 molecules CF^ G (CF^^) = = , and 2.69 X 10^1 X ^ 1.34 molecules CFi ^ ^CF^) = WU-T7-. • The G values for the other gases may likewise be determined: G (CjFj) = 2-02 " 1°^^ = 0.75 , 2.69 X 10^9 G (OjP.O) = iJ2.JLi4! , 0.72 . 2.69 :

PAGE 128

119 Calculation of Enerp;y Absorbed During; Exposure to Reactor Radiation Richardson, Allen, and Boyle (32) used a calorimetric measurement to determine the total energy absorbed by a material when exposed to reactor radiation composed of neutrons and gamma photons. The following expression was developed to give the total heat produced by reactor irradiation per mole of material irradiated: QM^ = S(I) +G (Mm/Mn^o) (^m/^H20) where Q = total rate of energy absorption in cal./gm. sec, 11^ = molecular weight of material m, S = empirical scattering constant, which is the combined factor (actual flxix/assurned flux) X (cal./Mev.), I = scattering integral in Mev./mole, G = empirical gamma photon constant, which is the gamma heat absorbed in cal./s^c. by one mole of HpO, and /^ ? ^ ratio of mass absorption coefficients for the material and water .

PAGE 129

120 Values for the empirical constants S and G were foimd to be S = 1.749 X 10"^ TT^-^ ' and Mev. G = 7.416 X 10-3 cal: sec . mole of H2O The value for the neutron scattering integral for H2O, given as 0.045^ Mev. /sec., must be corrected for the difference in microscopic scattering cross sections for H2O and CgF^^. The scattering cross section for water is IO3 barns. The cross section for carbon is 4.^ barns and for fluorine is 3-9 barns. The cross section for C^F^^ is , g barns ^ ^ C atoms +39 ^^rns C atom 24 atoms in C^^F-j^^^ molecule F atom ^ 16 F atoms = L 2 barns 24 atoms in C^F^^ molecule The corrected neutron scattering integral is 4.2 barns (C^F^^) tX ' o V /„ M — X 0.045^ 103 barns (H2O) = 0.001^7 ^-^^for CgF^^ . The ratio of the mass absorption coefficients can be found by using information on the electron densities given in Reference 20' The electron density of water is 3.34 X 10^3 electrons per gm., and the electron

PAGE 130

121 density for C^F-j^^^ is estimated to be 2.SS x 10^^ electrons per gram. Finally, the molecular weight of CgF-,^ is 400 gm./inole and the molecular weight of water is iBgm./mole. Substitution of the values for the variables into the original expression gives Q X 400 SHf (CdF-iA) = 1.749 X 10-2 caK ^ o.OOlS? ^ISZ. mole ^ -^^ Mev. sec, 400 ^m^ (CgFi6) + 7.416 X 10-2 cal,(H20) ^ _^EL sec1& gm. (H?0) mole 2.^^ X 10^3 electrons g^j^^ ^ 3.34 X ]^o23 electrons Q = 0.000355 ^^^' gm.sec This heating rate would result from a thermal neutron flux of 0.772 X 10 neutrons/cm. sec. and 0-000355 giT^iFT^ ^-1^6 X 107 llff X 1 loQ^^^g/g^. = 1.49 X 10^ iad_ sec . so that 1.49 X 10^ rad./sec. is equivalent to 0.0772 TO / 2 X 10-^'^ neutrons/cm. 3ec. For Sample No. 25, the integrated neutron flux was 0.00^04 X 10"^^ neutrons/cm. 2 so that the absorbed energy

PAGE 131

122 in tenns of rads was S.OU X loll neutrons ^ 1./,? x 10 rads cm/ 0.00772 x lO 12 neutrons cm.' = 1.5 5 X 10^ rads. The doses for the remaining samples can be similarlyconverted. Integrated Flux Sample No. ( neutrons ) Dose (rads) ( cm.2 ) 25 S.04 X 10^2 1.55 X 10^ 26 5.5S X 10^^ 1.00 X 10^ 27 1.68 X 10^^ 2.95 x 10^ 2g 2.43 X 10^^ 4.31 x 10^ 22 4.29 X 10^^ 7.60 x 10^ 15 5.94 X 10^5 1,06 x lo"^

PAGE 132

123 Calculation of Intep^rated Thennal Neutron Flux The counting equipment was previously calibrated (35) and included the following: Baird Atomic Pulse-Height Analyzer Model 510, Precision Timer Model 930 , High Speed Scaler Model 134, Non-Overloading Amplifier Model 215, Super Stable High Voltage Power Supply Model 312A, Scintillation Detector Model SlO, and NaT (Tl) Well Type Crystal. The following settings were used. High voltage = ^95 volts , Coarse = 2 , Fine gain = 69 , Pulse height = 1.0 , Channel width = 10 volts , and Base line =3.0.

PAGE 133

124 Reference 3 5 gives the following relationship for calculating the thermal neutron flux: 1 AK 1 exp — ^ Ntre where (b = thermal neutron flux in "^^^^°"^ , cm.^sec. A = activity of foil in disintegrations ^^ ^^^ ^^ ^ sec . irradiation, M = atomic weight of trace element (Au^O in gm., N = Avagadro's number x weight of the sample in gm., (T" = activation cross section for Au "', Ti = half life of isotope produced, t = irradiation time, and = isotopic abundance of target element . The efficiency of the counter was taken to be 0.543 as given in Reference 35. The equipment drift which had occurred since the original calibration was found to be 0.923These two factors were combined to give a total correction factor of 1.701. The data for Sample No. 15 can be used for a calculation. The foils were counted at intervals of two

PAGE 134

125 days until the activity had dropped below 10,000 counts per minute. Three counts of one minute each were taken every time the foils were counted. The activity of the foil for Sample No. 15 was 26,S37 count s/min. after the foil had been out of the reactor for 319-0 hours. Thus the foil activity at reactor shutdown is 26,3^7 X exp 1 ^0.693 X gl9.o] L STTTg J or 29-54 x 10 disintegrations/sec. The foil wei,c:hed 0.99^0 mg . , and the irradiation time was 2f.0 hours. The half life of Au"""^ is 64S hours, 197 and the isotopic abundance of Au ^ is 100 per cent. The flux during the irradiation is then 29.54 X 10^ X 197 X 1.701 6.023 X 10^^ X 0.99^0 x 10^ x 96 x 10"^^ X 1.0 X exp 1-0.693 xTT ' °^ 1 — W:^ — r 2-42 X lO-'--^ neutrons/cm. sec. Finally, the integrated flux over the four-hour irradiation is 4.11 X 10^^ neutrons ^ 3^00 sec^ ^ ^ hr. cm. sec. "^* = 59.4 X 10^^ neutrons cm.^

PAGE 135

126 Calculation of the Decomposition Rate In an Organic Cooled Reactor The Piqua Nuclear Power Facility has an organic coolant decomposition rate of 50 lbs. per megawatt-day. The thermal power of the reactor is 45.5 megawatts, and the volume of coolant in the core is l600 1. At 13 full power the average neutron flux is 1.9 x 10 neutrons/cm. sec. Assume the reactor operates for one hour. The total amount of coolant in core is 1600 1. x 0.9 |'. or l.U X 10^ kg. The coolant decomposed in one hour is 50 ^^^^,,' . . X -^T — 1^ X /v5.5 megawatts megawatt-day ^4 nr. "^^ or 95.0 lbs. coolant decomposed per hour. The percentage decomposition can then be found to be 95.0 ^^^X I^p:— '— H^I 2-205 lb. ^ 100 = 3.00 per cent. 1.44 X 10^ kg. The integrated flux in the core during the one hour of operation would be 1.9 X lO""-^ neutrons/cm.^ sec. x 36OO -^^^^ hr.

PAGE 136

127 = 6.^4 X 10^^ neutrons ^^^^ Q^e hour) . cm. For an exposure of 6.S/4x 10^" neutrons/cm. , an extrapolation of the curve on Figure 13 for CgF-j^^ indicates that the decomposition rate would be 0.50 per cent for a one-hour period.

PAGE 137

128 Calculation of the Decomposition Rate in a Fluorocarbon Uranlum Hexafluoride Reactor The data for the Homogeneous Reactor Experiment 1 will be used as the basis for the calculation. The HRE-1 used a solution of water and uranyl sulfate with 235 a concentration of 40 gm. of U -^ ^ per kg. of water. The power of the reactor was one megawatt with an average flux of 1.9 X 10"^^ neutrons/cm.^ sec. The concentration of 1)23 5 in the irradiated samples can be determined as follows: t23 5 1.46 gm. natural uraniun, ^ r\ r\r\n gm. U" 5.^2 gmV of CgFi6 "" °*°°^ gra. Haturs ural uranium = o.ooia5 1^^ • The concentration of U^^^ i^ the HRE-1 is then found to be 40 gm. u23 5 gm. U^^^ looB gm. H20 0.040 Isrri^ Thus, the ratio of concentrations is 0.040 HRE-1 0.00185 sample 21.6 .

PAGE 138

129 During a one-hour period of operation the integrated flux in the HRE-1 would be 1 n 1^13 neutrons -./r^^ 1.9 X 10 -^ 2 ^ 3600 sec cm. sec. c dc inl'^ neutrons . = 6.«5 X 10 2 cm. sec. For an integrated flux of 6.S5 x 10 neutrons/ 2 cm. sec., the curve for ^^i(^ and UF^ in Figure 13 indicates that the decomposition rate per hour would be about Z4-.3 per cent. If the concentration factor is considered, the decomposition rate would be 4.3 per cent x 21.6 or about 93 per cent per hour.

PAGE 139

APPENDIX B ANALYSIS FOR POLYTffiR IN IRRADIATED SAMPLES The routine analysis for polymer is described in Chapter III; however, since considerable difficulty was encountered in the original efforts to find polymer, the background of the method will be discussed. Analysis of gamma irradiated samples of CgF^^ which had been vacuum transferred to a glass bulb in the vacuum system and then analyzed by liquid injection into the chromatograph showed no polymer present . Liquid samples of gamma irradiated C^F-^^ taken directly from the irradiated capsule were found to contain dimer. The two possible reasons for the lack of polymer were the removal of the polymer by the sodium fluoride trap and the failure of the polymer to vaporize because of its low vapor pressure and thus remain in the capsule. Experiments were performed which showed that the polymer was not removed by the sodium fluoride trap and did remain in the capsule. This finding made possible the analysis for polymer in a reactor irradiated sample of CgF-j^^ and UF^. The sample was vacuum transferred from the capsule through the sodium fluoride trap into a glass bulb in the vacuum system and then transferred back to the capsule. The 130

PAGE 140

131 two passes through the sodiiim fluoride trap removed the UF^ and fission products so that a liquid sample could be analyzed using the procedure that had been found to work successfully for analysis of the gamma irradiated material. Since no polymer was found, the size of the sample chromatographed was increased from 5 microliters to 100 microliters with the result still being negative .

PAGE 141

LIST OF REFERENCES 1. John A Wethington, Jr., "Research Pertaining to A Uranium (VI) Fluoride-Fluorocarbon Reactor," Puerto Rico Nuclear Center, Mayaguez, Puerto Rico (April 26, 1962). 2. P. Barbour and W. W. Davis, "Evaluation of Coolant Impurity Removal Equipment at the QMRE," NAA-SRSk7k (October 1964). 3. J. H. Simons and L. P. Block, Journal of the Amer ican Chemical Society . 59, 1407 (1937) . 4. J. H. Simons and L. P. Block, Journal of the American Chemical Society , 6I , 2962 (1939) . 5. T. J. Brice in "Fluorine Chemistry," Volume 1, page 423, edited by J. H. Simons, Academic Press, Inc., New York, N. Y. (1950). 6. A. V. Grosse and G. H. Cady. Industrial and Engi neering Chemistry , 22» 367 (1947) . 7. J. W. Ryan, Modern Plastics , 31, 152 (1953). 3. A. Charlesby, Nucleonics . 12, No. 6, 1^ (1954). 9. R. Harrington, "Plastics and Elastomers for Use in Radiation Fields," HW-44092 (1956). 10. P. Y. Feng, "Proceedings of the Second United Nations Conference on the Peaceful Uses of Atomic Energy, Volume 29, 166 (195^). 11. L. A. Wall, R. E. Florin, and D. W. Brown, Journal of Rese arch of the National Bureau of Standards, u^r7WTT9W)12. J. H. Simons and E. H. Taylor, Journal of Physical Chemistry . 63_, 636 (1959). 132

PAGE 142

133 13. L. A. Wall and R. E. Florin, Journal of Applied Polymer Science , 2, 251 (1959Tr^ 14. F. W. Bloch, D. R. MacKenzie , and R. H. Wiswall, Jr., "Preparation and Properties of Fluorocarbons of Interest in Reactor Technology," TID-7622, 109 (July 1962). 15. R. F. Heine, Journal of Physical Chemistry, 66, 2116 (1962). — 16. D. R. r>/[acKenzie , F. W. Bloch, and R. H. Wiswall, Jr., Journal of Physical Chemistry . 69, 2526 (1965). 17. F. W. Bloch and D. R. MacKenzie, "Radiation Stability of Some Liquid Fluorocarbon Systems at Elevated Temperatures," to be published. IS. L. Kevan and P, Hamlet, Journal of Chemical Physics, 42, 225^ (I965J. ^ 19. M. B. Fallp;atter and R. J. Hanrahan, Journal of Phys ical Chemistry , C9, 2059 (1965). 20. T. M. Reed, J. C. I^ilen, and W. C. Askew, "Experimental Effects of Pile Radiation on Pure Fluorocarbons," Final Report to the United States Atomic Energy Commission on Contract No. AT-(40-l)-2S46 (I965). 21. A. H. Snell and J. H. Rush, "The Multiplication Factor for Product Drums Containing Uranium Hexafluoride,' Mon P-4S (November 1945). 22. E. Grueling, H. Argo, G. Chew, M. E. Frunkel, E. J. Konopinski, C. Marvin, and E. Teller, "Critical Dimensions of Water-Tamped Slabs and Spheres of Active Material," LA-609 (August 1946). 23. G. I. Bull, "Calculations of the Critical Mass of UF5 as a Gaseous Core with Reflector of -OtO, Be, and C, " LA-lg74 (February 1955). 24. H. A. Bernhardt, W. Davis, Jr., and C. H. Shiflett, "Proceedings of the Second International Conference ,; on The Peaceful Uses of Atomic Energy," Volume 29, 62 (195^).

PAGE 143

134 25V. A. Dmitrievskii and A. I. Mifrachev, Atomnaya Ener.-^iya , 6, 533 (1959). Translated by G. Ryback, Journal of Nuclear Ener^, Part A: Reactor Science, 12, 185 TT9^0r 26. K. Goodiran , Editor, "'The Science and En/^ineerinp: of Nuclear Power,*' Addison-Wesley Press, Readinp, Mass. (1949). 27. I. K. Kikoin, V. A. Dmitrievskii , V. V. Gazkov, I. S. Grip;oriev, B. G. Bubovsky, and S. V. Dersnovsky, "Proceedinp-3 of the Second International Conference on the Peaceful Uses of Atomic Energy," Volume 9, 528 (1958). 2S. R. DeV/itt, "Uranium Hexaf luoride : A Survey of the Phvsico-Cherrical Properties." GAT 2^0, Goodyear Atomic Corporation, Portsmouth, Ohio. 29. A. R. Boynton, "Guide for Irradiations in the University of Florida Training Reactor," Volume XV, No. 7 (July 1961). EnpineerinF Progress at the University of Florida, Florida Engineering and Industrial Experiment Station, Gainesville, Florida. 30. American Society for Testing Materials, "Tentative Method of Measuring Absorbed Gamma Radiation Dose by Fricke Dosimetry," ASTM Designation D 1671-593, Philadelphia, Pa. (1959). 31. Wallace Davis, Jr.^to John A. Wethington, Jr., Private Communication (March 26, I962). 32. p. M. Richardson, A. 0. Allen, and J. W. Boyle, "Proceedings of the International Conference on the Peaceful Uses of Atomic Energy," Volume 14, 209, United Nations, New York (1956). 33. International Atomic Energy Agency, Directory of Nuclear Reactors , Volume t , page 177, Vienna, Austria (1958). 34. International Atomic Energy Agency, Directory of Nuclear Reactors , Volume 11, page 153, Vienna, Austria (1958) . 35. Louis G. Grant to William H. Ellis, Private Communication (September 12, 1962).

PAGE 144

BIOGRAPHICAL SKETCH Thomas Howard Scott was born October S, 1932, in Tuscumbia, Alabama. He attended Coffee High School in Florence, Alabama. In 1953, he was awarded a Bachelor of Science in Chemical Engineering by Auburn University. After serving in the United States Navy for three years, he returned to Auburn University and was awarded a Master of Science in Nuclear Science in 195^. Since 1963 he has attended the University of Florida. During this period he has held a College of Engineering Fellowship and a Special Fellowship from the Oak Ridge Institute of Nuclear Studies. He is married to the former Janet Cosby and has two children. He is an associate member of Sigma Xi and a member of the University of Florida Student Chapter of the American Nuclear Society. 135

PAGE 145

This dissertation was prepared under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Engineering and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. April, 1966 Dean, College of Engineering Dean, Graduate School Supervisory Committee: i/ IBh^ -^^^aJu^/

PAGE 146

4660