Citation
Gamma and neutron irradiation of pure fluorocarbons

Material Information

Title:
Gamma and neutron irradiation of pure fluorocarbons
Creator:
Mailen, J. C
Place of Publication:
[Gainesville]
Publisher:
[s.n.]
Publication Date:
Language:
English
Physical Description:
xiii, 187 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Carbon ( jstor )
Electrons ( jstor )
Energy absorption ( jstor )
Fluorine ( jstor )
Fluorocarbons ( jstor )
Gas density ( jstor )
Irradiation ( jstor )
Liquids ( jstor )
Molecules ( jstor )
Sand sheets ( jstor )
Fluorocarbons ( fast )
Irradiation ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 184-186).
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
11013689 ( OCLC )
ocm11013689
30084865 ( Aleph number )
Classification:
547.8 M219g ( ddc )

Downloads

This item has the following downloads:


Full Text













GAMMA AND NEUTRON IRRADIATION

OF PURE FLUOROCARBONS













By
JAMES CLIFFORD MAILEN


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY













UNIVERSITY OF FLORIDA
August, 1964














ACICNTOWL1EDGE1121T


This work was done under contract to the United States Atomic Energy Commission. The author also received financial support in the form of a fellowship from the National Science Foundation. The financial support of these two agencies is greatly appreciated.

The author is indebted to many people for assistance in the

completion of these studies but especially to Dr. T. 11. Reed, III, for his beneficial direction of the project, Mr. Shashi Dat6 for his assistance in the early stages of the work and Dr. J. H. Simons for many helpful suggestions in the course of the research. Thanks are due Dr. Merril Wilcox for obtaining the infrared absorption spectra of the solid fluorocarbon. The other members of the Ph.D. committee,

Dr. Mack Tyner, Dr. R. J. Hanrahan, Dr. T. F. Parkinson, and Dr. H. A. Meyer (deceased) have also lent their kind support.













TABLE OF CONTENTS


INTRODUCTION . . . . . . . . . . . . . . . . . . . ...

IAEOHANISM . . . . . . . * * - * * : - * I - - * 9 9 *
Free Radical Formation . .
Reactions and Mass Spectrograph Sensitivities . ....
Types of Reactidns in Bulk Phase . ..........
Reactions with Wall . . . . . * . . . . . . . . .

MATERIALS IRRADIATED . . . . . . . . . . . . . . . . . ....

IRRADIATION OF SAMPLES . . . . . . . . . . . . . . . . ....
0A/qNN G . *.. . . . .e . . . . . .. . . . . . . . ..0 0


ANALYSIS . . . . . . . . . . . . . . . .. .. .. .


RESULTS AND DISCUSSION ........... . . . . . . .
Irradiations in the Oak Ridge LITR Reactor . . .
Solid Fluorocarbon from LITR Irradiations ....
Irradiations in the Oak Ridge Graphite Reactor
Irradiations in Cobalt-60 and Detailed Discussion


Perfluoromethane . . . .... Perfluoroethane . . . .... Perfluoropropane . . . . Perfluoro-n-Butane . . . ... Perfluoro-n-Pentane . . ... Perfluorocyclopentane . ... Perfluoro-n-Hexane . . .... Perfluoro-2-M-ethylpentane . . Perfluoro-2,5-Dimethylbutane .


* . . 9 9 9 9
* . 9 9 9 * 9



* . 9 . . 9 9

* 9 * * * * *


SU-11RY AND FUTURE WORK ......... . ....

Appendices . . . . . . . . . * * . * . . .. ..

APPENDIX 1. ENERGY ABSORPTION . . . . . . ...
Flux Determination ............
Flux Exposure of Samples . . * * Variation of Flux Across Sample Zone . ...
Absorbed Dose . . . . . . . . . . . . . . .
Energy Absorption in Samples .....
Energy Absorbed in Reactor Irradiated Samples


iii


.*... 105


9 9 . 0


lO6

107 107 108
110
120 127 132


0 * . 0









APPENDIX 2. SAMLE CLEANUP AND CANNING . ......... 155

APPENDIX 5. ANALYTICAL . .. . ........ ...... 156
Physical Description of Analytical Equipment . . . . 156 Calibration of Molecular Weight System . . . . . . . 157 Sample Treatment . ..... 159 Switching Columns from Series to*Parallel i 140 Description of Analytical Columns .......... 142 Urea and Thiourea Columns .. .... ....... 142
Mole-Area. Calibration . . .. ........... 146
Fluorine Balance . . . . . . ............ 146

APPENDIX 4. DECANNING OF PILE IRRADIATED SAMPLES . . . . . 151

APPENDIX5- RADICAL TRANSPORT ... ............ 155
Diffusion Coefficient of Radicals in Perfluoropropane 153 Statistical Model for Calculating Mean Free Path . . 156 Mean Free Paths of Radicals in Perfluoropropane . . . 158

APPENDIX 6. STATISTICAL RECOMBINATION DATA CORRELATION � � 165
CHEMiICAL NOMENCLATURE .. .. .. .. .. .. .. .. . .. 182

ABBREVIATIONS AND DEFINITIONS ............... 185

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . 184














LIST OF TABLES


Tab le

1 COMPARISON OF MASS SPECTROMETRIC SENSITIVITIES AND
G VALUES . . . . . . . . . .. .. .. .. . . .

2 IMPURITIES IN IRRADIATED SAMPLES . . . . . . . ..

5 WEIGHT PERCENT CF4 IN SAMPLES IRRADIATED IN THE LITR

4 RESULTS OF MOLE FRACTION DETERMINATION ON SOLID
FLUOROCARBON . . . . . . . . . . . . . . . ....

5 GASEOUS PRODUCTS FROM THE THERMAL DECOMPOSITION OF
THE LITR SOLID' FLUOROCARBON . . . . . . . . ....

6 LITR IRRADIATIONS--ANALYSIS OF GASEOUS PRODUCTS . . .

7 RESULTS OF OAK RIDGE GRAPHITE REACTOR IRRADIATIONS .


8 MASS SPECTROGRAPH DATA FOR CF4(5)


MECHANISM FOR MECHANISM FOR MECHANISM FOR MECHANISM FOR MECHANISM FOR MECHANISM FOR MECHANISM FOR MECHANISM FOR MECHANISM FOR


OF4 . . . 02F6 -. . 03F8 � . . n-C4Flo

n-O 5F12 � cy-C5Fo10 n-C6F14 � 2-OF3)C F 2,5-(CF3) 2c


0 . f . .






* . 0 0 0 . 0 0 * 0 * 0






04F 0 0 . . . 0 4F8 . .0 - 0o


CF4 IRRADIATED BY COBALT-60 GAM,24A RAYS . O2F6 IRRADIATED BY COBALT-60 GAMMA RAYS


5



22 26 31 35

4O


CHEMICAL CHEMICAL

CHEMICAL CHEMICAL OHIEiICAL CHEMICAL CHEICAL CHEMICAL CHEMICAL PART 1. PART 2.









Tabl e Page 18 PART 5. C5F' IRRADIATED BY COBALT-60 GAM!4A RAYS, 8
T 99oC . ...... . .. .. .. .. 89

PART 4. c3F8 IRRADIATED BY COBALT-60 GA1lzu RAYS,
T = 5000 . ....... ....* *. . * . . . 92

PART 5. C F WITH POWDERED ALUMINUM IRRADIATED BY
coBALT-6 AMA RAYS . . . . . . ... ... . . . . . 97

PART 6. n-CF10 IRRADIATED BY COBALT-6o GA=A RAYS,
T = 500 . . . . . . . . . . . . . . . . . . . . 98

PART 7. n-CFl2 IRRADIATED BY COBALT-60 GAMvIA RAYS . 99

PART 8. CYCLO-C5F1o IRRADIATED BY COBALT-60 GAI'.A
RAYS . . . . *. . . . . . . . . . . . . . 100

PART 9. n-06F14 IRRADIATED BY COBALT-60 GA$M !A RAYS . 101

PART 10. 2-CF 0 F AND 2,5-(CF)204F8 IRRADIATED
BY COBALT-60/-9 RAYS ..T.-.6. .......... 102

19 GAriMA ABSORPTION CROSS-SECTIONS PER ELECTRON . . . . 121 20 NIUlTlBER OF ELECTRONS PER GRA . . . . . . . . . . . . . 125

21 RADS ABSORBED PER ROEITGE NEGLECTING WALL AND DEITSI TY EFFEOTS .................. 124

22 ELECTRON STOPPING POWER RATIOS RELATIVE TO AIR FROM FIGURE 51 FOR MATERIALS OF INTEREST . . . . . . . . 129 25 ENERGY ABSORPTION BY BRAGG-GRAY CAVITY THEORY . . . . 150 24 TEMPERATURE PROGRAM AND APPEARANOE TIMES ON THE SILICA GEL COLU L. .... . ........... 145

25 STANDARD RETENTION VOLUMES ON THE SQUALANE COLUMN . . 144 26 STANDARD RETENTION VOLUMES ON THE n-HEXADECANE COLUMN . .9. . ... . . .........oo. 145

27 STANDARD RETENTION VOLUMES ON THE THIOUREA O0LUlN . . 147 28 COMPARISON OF MEAN FREE PATHS CALCULATED FROM GAS KINETICS AID FROM THE STATISTICAL MODEL . . . . . . 159 29 MEAN FREE PATHS FOR RADICALS IN PERFLUOROPROPANE . . 162 50 RANDOM RECOMBINATION CALCULATION AND RESULTS FOR C2F6 166







Table LaM 51 RESULTS OF RANDOM RECOMBiNATION CALCULATION FOR 0F8 168 52 RESULTS OF RANDOM RECOM ITATION CALCULATION FOR n-04Fl0 .. .. .. .. .. . . . .... .... . .. 170

55 RESULTS OF RANDOM RECOMINATION CALCULATION FOR n-C 12 .. . . . . . . . . . . . . ... .. . . .. 172
34 RESULTS OF RANDOM RECOMINATION CALCULATIONS FOR n-06714 . .. .. .. .. .. .. .. .. .. . .. 174

55 RESULTS OF RANDOM RECOBINATION CALCULATIONS FOR 2-CF5CsFI, . . . . . . . . . . . . . . . . . . . . 176
36 RESULTS OF RANDOM RECOMBINTATION CALCULATIONS FOR 2, 5-(CF)2O4F8 . . . . . . . . . . . . . . . . . . 179


vii












LIST OF FIGURES


iarege
1 Electrically heated sample holder . . . . . . . . . . 14 2 LITR sample holder . . . . . . ... . . . . . . . 16
5 Weight percent OF) versus fluorine to carbon ratio of
original material for samples irradiated in the
LITR ... . ..... . . .. .. 21

4 Isoteniscope used for determination of the molecular
weight of the solid fluorocarbon produced in the
LITR . *. * .. .. . . . . . . . . . . . . . . 25

5 Indicated mole fraction vs. solution vapor pressure
of solid fluorocarbon in n-O7F16 . . . . . . . . . 27
6 Infrared spectrogram of solid fluorocarbon from LITR
irradiations . . . . . . . . . . . ....... 29
7 G(OF4) versus density for 9900 05F8 samples ..... 52

8 V(0F4) versus bulk density for 500C CF8 samples
(.obalt-60) ...... .. .. * * * . 53
9 G(C F ) versus density for 990 03 5F8 samples
(9oalt-6o) ....... ............ 54
10 C(02F6) versus bulk density for 5000 05F8 samples
(cobalt-60) . . . . . . . . . . . . . . ...... 55
11 G(Fp) lost versus density for 990C 05F8 samples
(cobalt-60) . . .. .. .. . . . . . . . . . . . . 57
12 G(F2) lost versus bulk density for 5000 05F8 samples
(cobalt-60) . 0. . .. .... ..... 58

15 G(C4F 0) versus density for 9900 05F8 samples
(oal6o) .. ..%. . . .. .. . . . ..... 59

14 G(04P 0) versus bulk density for 300 C5F8 samples
(coiat-6o) .. .. * . . . .....*. .... 6o

15 G(n-C F ) versus density for 9900 C5F8 samples
(0o 6o) .. . . . . . . . . .. . 61
viii








IF--, r e Pagqe
16 G(n-C F ) versus bulk density for 3000 C3F8 samples (coKi-60) .... ........ . * . .. .. . . . 62

17 G(i-C F ) versus density for 9900 CF8 samples (c0o i 6o) . .. .. .. .. . . . . . ..... 63

18 G(i-C F 0) versus bulk density for 5000 C5F8 samples (c ; 6) .. .. .. .. .. .. '.. ... 64

19 G(n-C F 4) versus density for 990C CF8 samples (c at60) .............. 65

20 -(n-C F4) versus bulk density for 5000 0 F8 samples

21 G(2-CF0 F ) versus density for 9900 0 F samples (bob al t -0) . . . . ........... . . . . . 67
22 G(2-CF C F ,) versus bulk density for 5000 0 3F8

25 G(2,5-(0Fs)2C4Fs) versus density for 9900 CsF8 samples (cobalt-60) . o . . . ...8 69 24 G(2,5-(CF5)204F) versus bulk density for 500C C3F8 samples (cob alt-60) . . . . . . . . . * . . . . 70

25 Sum of G values versus density for 9900 03F8 samples (cobalt-60) . . . . .. . . . . . 71
26 Sum of G values versus bulk density for 5000 C0F8 samples (cobalt-60) .. . . . .. . . .. . ... 72

27 Flux in tube 11 of the engineering cobalt-60 source . 109 28a Diagram for relative flux across sample zone calculation-planar . . . . . . . . . . . . . . . . 112
28b Diagram for calculation of relative flux across sample zone-volume .. .......... . . . . *. 114
28c Diagram for calculation of relative flux across sample zone by infinite shell model . . . . . o o o 116 29 Arrangement of cobalt-60 rods around tube number 11 of the engineering cobalt-60 source . o . o . . . . 117 50 Relative flux across sample zone by infinite wire and infinite shell models . * . . .*. . * . # . . # 118









Figure

51 Electron stopping power relative to air versus
atomic number . . . . . . . . . . . . . . . .... 128

52 Sample purification and canning system . . ... . . . 154 55 Decanning and storage system ............ 158

54 Series and parallel connections of chromatographic
columns . . . . . . . . . . . . . . . 141

55 Log(moles) versus log(area) for chromatographed
fluorocarbons . . . . . . ............. 143

56 System used for decanning pile irradiated samples . . 152














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

GAMMA AND NEUTRON IRRADIATION
OF PURE FLUOROCARBONS

By

James Clifford Mailen

August', 1964'


Chairman: Thomas M. Reed, III

Major Department: Chemical Engineering

Samples of perfluoromethane, perfluoroethane, perfluoropropane, perfluoro-n-butane, perfluoro-n-pentane, perfluoro-cyclopentane, perfluoro-n-hexane, perfluoro-2-methylpentane, and perfluoro2,-dimethylbutane have been exposed to cobalt-60 gamma rays and the mixed radiation of the Oak Ridge Graphite Reactor and the Oak Ridge LITR reactor. Analysis of the resulting mixtures was by a three column gas chromatograph allowing essentially complete analysis of all compounds and isomers from C1 to C6 and analysis by number of carbon atoms from 07 to 012.

In the irradiation of pure saturated fluorocarbons by gamma

rays, the important chemical reactions in the sample are recombination of radicals, disproportionation of small radicals, addition of fluorine and radicals to unsaturated molecules, and the reaction of radicals with molecular fluorine. These reactions differ from those found in hydrocarbon irradiations by the absence of abstraction reactions and








the presence of reactions between radicals and molecular fluorine. These two differences are explained by bond energy considerations.

Molecular fluorine, fluorine radicals, and other radicals can be removed from the reaction system by wall capture. The reaction between

radicals and molecular fluorine depends on the density, total energy absorbed, and the distance to the sample tube wall. One of the consequences of this type of reaction is that the G values for molecules smaller than the parent increase with both sample density and total energy absorption. Capture of radicals other than fluorine by the sample tube wall becomes important at low density (below about

0.2 gm/cm3).

When fluorocarbons are irradiated to high energy absorption, as in a nuclear reactor, chemical equilibrium considerations are

important. The equilibrium depends on the fluorine to carbon ratio of the sample with a ratio of four resulting in nearly pure perfluoromethane. (Smaller ratios yield less perfluoromethane6) For this reason, for long irradiations, perfluoromethane appears to be extremely stable.

The initial G values for disappearance of the parent molecules are related to each other in much the same way that the sensitivities to electron bombardment in the mass spectrograph are related. This is not unexpected since, in both cases, the species causing bond rupture is electrons. Thus, given the sensitivities for a known and an unknown fluorocarbon and the G value for the known fluorocarbon the G value for the unknown can be estimated as the ratio of the sensitivities times the known G value.








The G values for disappearance of the parent compound are approximately; CF4, 1.5; 02F6, 5.25; 05F8, 4.5; n-C4F10, 5.4; n-O5F12, 4.9; cyclo-C5Flo, 5.0; n-C6F14, 6.0; 2-OFSC5FiI, 6.0; and 2,5-(CF )204F8, 9.6. These values are only approximate due to the dependence of certain reactions on density and amount of energy absorbed.

The important products in perfluoromethane are perfluoroethane and perfluoroacetylene produced in about equal amounts. For the other materials for short irradiations the major products are saturated and are those predicted by simple bond rupture and radical recombination reactions. The results for perfluoroethane and larger molecules, excluding cyclo-compounds, can be correlated by statistical recombination calculations using empirically determined radical effectiveness numbers. Since these effectiveness numbers are the same for radicals formed in the same way this method can be used to predict product distributions for other fluorocarbons. In conjunction with prediction

of overall G values from mass spectrometer sensitivities the G values for the products can be estimated.


xiii













INTRODUCTION


Samples of fluorocarbons irradiated in a cobalt-60 source, in the Graphite reactor at Oak Ridge, and in the LITR reactor at Oak

Ridge have been analyzed by three-column analytical gas chromatography. The three columns were one meter of silica gel temperatureprogrammed for compounds with from one to four carbons, 15.5 moters of squalane on Ohromosorb-P at 9200 for compounds with six to about fourteen carbons, and 16 meters of n-hexadecane on Chromosorb-P at 2500 for compounds with three to seven carbon atoms per molecule.

The results of these analyses for which mole fractions and G values have been calculated are given in Tables 6, 7, and 18. When no values are given for high molecular weight materials it indicates that definite peaks could not be distinguished. From 012 on up, this is partially due to the overlapping of the large number of isomers which are formed by radiation processes. This makes it impossible to distinguish between minor base line instability and small amounts of the compounds.

Previous to this work two irradiations of fluorocarbons have been reported. J. H. Simons and E. H. Taylor(l) irradiated C6F14 in the Oak Ridge Graphite Reactor and Florin, Wall, and Brown(2) exposed 07F16 to gamma rays. Both these studies were hampered by the use of impure starting materials and the use of relatively poor analytical methods. In this study the compounds were purified by standard gas

1





2

chromatographic methods allowing study of pure single isomers. Analysis was also by gas chromatography allowing separation of most of the materials up through six carbons.

Other studies of interest are those of Mastrangelo(7) where

fluorocarbon radicals were generated by electric discharge and those of Pritchard, Hsia, and Miller, (8) Seeger and Oalvert,(3) and Ayscough and Steacie(4) who generated fluorocarbon radicals by photolysis. Mass spectrograph results(5'6) are of some value in determining the important species in irradiation. These types of data are valuable in the determination of mechanisms.













MECHANISM


Free Radical Formation


When a fluorocarbon is bombarded by electrons a probable first

step is fracture of the molecule into ionic and non-ionic fragments. The positively charged ionic fragments may then be converted to free radicals as is shown below.


M +i e* io R+ + R + 3e
R+ + --Rl R+ + e- R2


The formation of positively charged ions is known to be important from mass spectrometer work. The neutralization scheme is supported by the data of Mastrangelo(7) who passed perfluorocyclobutane and perfluoroethane through an electric discharge and froze out the excited species on a liquid nitrogen finger. He failed to detect

any space charge at the finger, indicating that the ionic species are quickly neutralized by electron capture or some other equivalent

mechanism.


Reactions and Mass Spectrograph Sensitivities


When irradiating pure saturated fluorocarbons a number of deductions about the expected reactions can be made.

3





4

First of all, there should not be an unusual stability. This is in contrast to the observed stability of these materials to heat and chemical attack where stability is attributed to the shielding effect of the fluorine atoms. In saturated, unstrained fluorocarbons the fluorine atoms form a protective shield against attack on the bonds by chemical agents and free radicals. In irradiation there is no shielding effect since the attack is by electrons and not relatively bulky species. This expectation of ordinary stability toward radiation is also indicated by the results of mass spectrograph studies.(5'6) While the electron energies in the mass spectrograph are quite low and cannot be expected to give the same distribution of excited species it is obvious that a material experiencing considerable fragmentation in the mass spectrograph will not show exceptional stability to the higher energy electrons produced by gamma rays.

In fact, for fluorocarbons, the sensitivity in the mass spectrograph appears to be closely related to the stability of the materials toward gamma irradiation. In Table 1 are listed the sensitivities for the major mass spectrograph peak relative to n-butane for the compounds for which data are available and which were irradiated in this study. The ratios of sensitivities to that for CF4 are compared to the ratios of total G values for these compounds relative to OF4. The average difference of these two ratios is 25.5 percent and the trends of sensitivity are similar.

The G values listed are for fairly low conversion. From this

we conclude that the differences in initial G values for the fluorocarbons are due to their ease of fragmentation by electrons and not to any secondary reactions.





5










TABLE 1. COMPARISON OF MASS SPET5OMETRIC
SENSITIVITIES A1%TD G VALUES a)


Ratio to Total Ratio to
Compound Sensitivity for major peak CF4 G Value CF4
Sensitivity for n-butane

CF4 0.575 1.00 1.47 1.00 C2F6 0.955 1.66 5.24 2.2 C0F8 1.71 2.97 4.5 5.o6 n-04F10 1.78 5.10 5.42 5.69 n-C5F12 2.59 4.51 4.91 5.54 Cy-05F10 1.75 5.01 2.98 2.05


(a) Sensitivities are from reference 6.







Types of Reactions in Bulk Phase


If the excited species in fluorocarbon and hydrocarbon irradiation are of the same type, we can make the following general predictions of the types of reactions to be expected in fluorocarbon irradiation.

1. Recombination of fragments. This will be of major importsance as has been demonstrated by Florin, Wall, and Brown(2) and by Mastrangelo(7) in their work with fluorocarbon systems.

2. Disproportionation. Disproportionation reactions between fluorocarbon fragments has been observed by Mastrangelo(7) for the following reactions:


2CF2--CF + CF5

2CF-3 CF4 + 0F2

205F7-" 05F6 + 03F8

2C4F9 --*C4F8 + C4F10.


However, Pritchard, Hsia, and Miller(8) found no evidence for the disproportionation;


203F7-03F8 + C3F6


Probably disproportionation is not competitive with recombination with the exception of some small fragments.

3. Removal of fluorine by molecular processes resulting in an unsaturated molecule. Dewhurst(9) has observed molecular processes in cyclohexane where about 15 percent of the excited cyclohexane molecules decomposed by molecular processes to give products. In the work of Nevitt and Remsberg(lO) about 60 percent of the hydrogen formed in the





7

irradiation of cyclohexane was found to originate from the removal of molecular hydrogen or other non-radical processes. In fluorocarbons it appears from mass spectrograph data that similar results are possible by removing more than one F radical from a single molecule.

4. Abstraction of F from a neutral molecule by an F radical.

Although this type of reaction has been extensively observed in hydrocarbon work, it is not to be expected in fluorocarbon irradiation for the following reason. The bond energy for the H-H bond in hydrogen is 104.2 k. cal./g. mole(11) and that for the C-H bond of a hydrocarbon is about 98.8 k. cal./g. mole("I) leading to a bond energy difference for hydrogen abstraction of about +5.4 k. cal./g. mole. For fluorene the latest value for the F-F bond energy is 57.5 k. cal./g. mole(12) while that for a F-C bond is about 105.4 k. cal./g. mole.(") This leads to a bond energy difference for fluorine abstraction of -67.9 k. cal./g. mole. Unless the F atom performing the abstraction is very energetic the abstraction cannot occur.

5. Abstraction of F from a neutral molecule by a radical other than fluorine. In hydrocarbon work abstraction of hydrogen atoms by the small radicals H(,15314s15) F,,(16) CD5 (17) and CH (18) has been observed. Energetically it appears possible for a fluorocarbon radical to abstract fluorine from a neutral molecule. However, for the same reason given for the chemical and temperature stability of fluorocarbons it does not appear that this reaction is important. That is, the fluorine atoms will act as a protective barrier preventing attack on the fluorocarbon bonds by free radicals. This conclusion is confirmed by photolysis experiments. Seeger and Calvert(5) found no CF4 in the photolysis of trifluoroacetone and concluded from this that





8

fluorine abstraction by OF5 is not important. In the photolysis of

hexafluoroacetone, Ayscough and Steacie(4) also found no CF4 leading to the same conclusion.

6. Addition of radicals across double bonds in unsaturated molecules. This type of reaction has been observed in hydrocarbons by various authors.(19' 20) Mastrangolo(7) has observed the addition of CF5 to ethylene and perfluoroethylene, the addition of OF2 to perfluoroethylene, and the addition of OF5 to perfluoropropene.

7. The reaction of fluorocarbon radicals with molecular fluorine. This reaction will become more important in longer irradiations where the fluorine concentration is not depleted by reaction with the wall and at high density where diffusion of fluorine to the wall is restricted. This reaction will lead to the formation of molecules of size smaller than the parent as is shown below;


R + F2- RF + F.


Thus one may expect a small increase in the G values of these

smaller radicals with time. Larger molecules than the parent will not result from this reaction.

In summary, the processes which are expected in the bulk phase of irradiated fluorocarbons are; recombination of radicals, disproportionation of small radicals, addition of fluorine and radicals to unsaturates, and the reaction of radicals with elementary fluorine.


Reactions with Wall


In addition to reactions in the bulk phase, additional reactions can occur at the sample tube wall. These are particularly important





19

in a geometry such as was used in this work. Since the inside diameter

of the sample tubes was only 0.191 centimeters, the surface to volume ratio is relatively high and the surface is not too distant from the bulk phase to allow some migration of fluorine and radicals to the surface. The probability of migration of radicals to the surface is dependent on the size of the radical, the size of the molecules

of the medium, and the density of the medium among other things. Thus one would expect fluorine radicals to migrate most easily followed by elementary fluorine, the perfluoromethyl radical, and so on. The

results of this migration should also be most easily observed at low density. As a qualitative estimate of the case of migration one can calculate the mean free path for the various radicals in the medium of interest. In Appendix 5 this has been done for perfluoropropane.

At a specific density of the medium a radical's mean free path becomes less than the average distance between molecules because this average

distance is not great enough to allow passage of the radical. For

the perfluoropropane, perfluoroothane, perfluoromethane, and fluorine radicals, these are respectively about 0.5 g./cm.5, o.3 g./cm.3,

0.55 g./cm.3, and 0.55 g./cm.3. Since free radicals have relatively ... short lifetimes, the radicals are effectively trapped in the bulk phase at high density.

Reactions of perfluoromethane radicals with tellurium, lead, and bisinath metals(21) to form rather unstable compounds have been observed.

With fluorine and F atoms, one would expect displacement of

the oxygen from the oxide layer. For this reason, until a protective fluoride layer is built up, the wall will act as a fluorine sink and reactions of radicals with elementary fluorine will be suppressed.'











MATERIALS IRRADIATED


The materials which were irradiated in this work are perfluoromethane (0F4) obtained from the Matheson Company, perfluoroethane (02F6) obtained from Minnesota Mining and Manufacturing Company, perfluoropropane (03T8) obtained from Minnesota Mining and Manufacturing Company, perfluoro-n-butane (n-C4Flo) obtained from Minnesota Mining and Manufacturing Company, perfluoro-n-pentane (n-0sF12) prepared by J. H. Simons,(22) perfluoro-n-hexane (n-06F14) prepared by R. D. Dresdner, T. M. Reed, III, T. E. Taylor, and J. A. Young,(25) perfluoro-2-methylpentane (2-OF305FlI) prepared by J. A. Young,(24) perfluoro-2,3-dimethylbutane Z2,,-(cF)2C4Fa7 prepared by J. A. Young, (24) and perfluorooyclopentane (cy-C5Flo) prepared by J. H. Simons. (2)
The OF4, 02F6, 0CF8, n-C4Flo, n-05F12, cy-C5FI0, n-C6F14,
2-CFCFI, and 2,3-(CF3)204F8 were purified by standard gas chromatographic methods(25) using either the silica gel or the n-hexadecane packings described in Appendix 3. The 2-OF 0 F1 and 2,3-(CF3)204F8 were further purified to remove unsaturated material using the thiourea packing described in Appendix 3.
After purification the compounds were found to contain the following amounts of impurities, identified where known. (See Table 2.)
Of these compounds only the perfluoromethane, ethane, and
propane were available in reasonably large quantities. The perfluoro10

















TABLE 2. IMPURITIES IN IRRADIATED SAMPLES


Mole % Impurity


0F4 02F6 0F8

n-04Flo

n- 5F12

cy-0,F12

n-C6F14 2-CF31CFll 2, -(oF3)2o4'8


small traces of 002, 02F6 small traces of 002, unknown

0.3% cy-C3F6

0.18% 002, less than. 9.-% i-C4Flo none detected none detected none detected

0.24% 06F12 none detected


Compound





12

propane was selected for extensive study because it could be studied in both the liquid and gaseous states with ease (critical temperature

- 71.9�0). In addition, it was the largest molecule available in quantity and was expected to give a greater range of reaction complexity.













IRRADIATION OF SAMPLES


Of the materials listed in Table 2, all except the perfluoron-butane and perfluoro-2,5-dimethylbutane were irradiated in the engineering cobalt-60 source, the Oak Ridge Graphite Reactor, and the Oak Ridge Low Intensity Training Reactor (LITR). These two materials were not available until after the reactor irradiation phase of the

experiments had been completed and received irradiation only in the engineering cobalt-60 source.

The flux distribution in the engineering cobalt-60 source

(tube 11) was determined by irradiation of samples of water saturated with benzene sealed in polyethylene bottles. The energy absorbed is determined by analyzing for phenol by ultraviolet spectroscopy. A further discussion of the method and a presentation of the results is given in Appendix 1.

In the engineering cobalt-60, runs were made at both ambient temperature (about 5000) and at 9900 in an electrically heated container. For details of the electrically heated container see Figure 1. The reasons for operating at these two temperatures were two-fold. First, this was necessary to see if there are any large effects of temperature on yield and secondly, by raising the temperature above the critical point of 05F8 it is possible to study the reactions of C3F8 in a single phase.

The samples to be irradiated in the Oak Ridge Graphite Reactor were placed in hole #41 and irradiated for 5 days.--The thermal" flux 13






14
Cork sealI Thermocouple Heater wire Heater wire


161/a"








% Asbestos
tape


Double glass tubes L.1 " Asbestos base plate Figure 1. Electrically heated sample holder.


Asbestos tape







in this position is 1012 n/cm.2-sec. and the temperature is about 10000. The energy absorption in this position can be estimated from

the data of D. M. Richardson, A. 0. Allen, and J. A Boyle(26) as about 8.5 x 1010 erg/gram for graphite. Assuming the value for fluorocarbons is not greatly different from this approximate G values can be calculated (see Appendix 1).

The samples irradiated in the LITR were in hole #A and were irradiated for a period of 28 days at a thermal flux of 1014 zVcm.2seo. In the LITR the samples are placed in a flow of coolant water and have a temperature of approximately 10000. See Figure 2 for a sketch of the sample holder. No estimate of the energy absorbed in the LITR samples can be given; however, since the samples appear to have reached equilibrium distributions, this would be of little value.
























Seven 5/32i bored holes equally spaced on 1/4" radius circle


Figure. 2.


Mushroom handL.ng post


-Nut LITR sample holder.













CANNING


Before irradiation, the samples were treated in vacuum to

remove air, carbon dioxide, and water by passage through absorbents and alternate freezing, pumping, and vaporization. After cleanup, the samples were sealed in aluminum tubes 11.5 inches long by 0.0752 inches inside diameter by heliarc following collection in the sample tubes by freezing in liquid nitrogen. One sample of 03F8, number 51, was sealed in an identical fashion in a copper tube 12-1/8 inches long by 0.0664 inches inside diameter. A further description of this procedure and the apparatus is given in Appendix 2.














ANALYSIS


All the samples except numbers 35, 7, 31, and 29 were analyzed on a three-column gas chromatograph consisting of 16 meters of nhexadecane at room temperature, 15.3 meters of squalane at 92�C, and one meter of silica gel. The silica gel column was temperature programmed. For details on the columns see Appendix 3. The samples mentioned above were analyzed without the high temperature squalane column. The silica gel column was used to separate air, perfluoromethane, perfluoroethane, perfluoroethylene, carbon dioxide (perfluoroethylene and carbon dioxide appear together), perfluoroaeetylene,

perfluoropropane, perfluorocyclopropane, perfluoro-n-propene, and perfluorobutane (no separation of perfluorobutane isomers). The nhexadecane column was used to separate perfluoropropane, perfluorobutane (partial separation of isomers), perfluoro-n-propene, the perfluoropentane isomers, the perfluorohexane isomers, and the perfluoroheptane isomers (most unidentified). The squalane column separated the perfluorohexane isomers with the exception of the 2and 3-methylpentane isomers and was used to determine the total amounts of material in each carbon number range up through 012 and occasionally, where amounts were sufficient for detection, up through Ol.

* Before being introduced into the chromatographic train, the

samples were decanned directly into an adjacent vacuum system and stored in bulbs of sufficient size to assure complete vaporization. After storage for at least 24 hours to assure mixing of the gases, 18





19

they were admitted to the volume calibrated vacuum system where their molecular weight was determined and then into the chromatographic train by gas sampling valves.

The chromatographs were supplied with thermal conductivity detectors and the output was recorded on Honeywell Electronik recorders. The areas under the curves were determined by use of a polar planimeter. Area was found experimentally to be proportional to number of moles from 01 to 09 and was assumed proportional over the entire range (Appendix 5).

The amount of fluorine lost to the wall was calculated from the fluorine to carbon ratio of the products (Appendix 5).













RESULTS AND DISCUSSION


Irradiations in the Oak Ridge LITRReactor


In the LITR irradiations the fluorocarbons were col.letely

degraded to an apparent equilibrium mixture. Of the gaseous products the major constituent is CF4. Complete analyses are given in Table 6. In addition to gaseous products solid products were formed from thc. materials above 04F10. These will be discussed further. The apparent stability of OF4 is not surprising since the thermodynamic equilibrium for carbon with sufficient fluorine is CF4. In Figure 5, the weight percent CF4 in the LITR samples is plotted versus the fluorine to carbon ratio of the original material. The data for the plot is given in Table 3. This is seen to give a smooth curve starting with zero percent CF4 for pure carbon and approaching 100 percent OF4 at a ratio of four. It should be noted that the structure of the original material is not a factor. A complete discussion of the mechanism responsible for this equilibrium is given in the section on the OF4 results.

An interesting further study would be the irradiation of a

mixture of perfluorocyclopentane with sufficient elementary fluorine, to give a four-to-one ratio to see if this mixture behaves like CF4 in the reactor; that is, gives a resulting mixture containing about 98 percent CF4. An equally interesting experiment would be irradiation of a mixture of graphite powder and fluorine with an F/C ratio of about two to see if the same solid fluorocarbon would result.

20














100


80 -


/
/


20 -


0
C


-0


1 2 3 F/C ratio (original compound)


Figure 3. Weight percent CF4 versus fluorine to carbon ratio of

original material for samples irradiated in the' LIR.


40 --


|













TABLE 5. WEIGHT PERCENT CF4 IN SAMPLES
IBRADIATZD IN THE LITR


Tube No. Compound F/C ratio Wt. Percent CF4

8 CF4 4 98.4 10 OF4 4 97.5 4 C2F6 5 54.0
3 02F6 3 43 28 C3F8 2.67 5 12 C5F12 2.4o 65
4 0 5F12 2.40 32.1 19 C5Fo 2.00 31.6 20 5Fo10 2.00 54.2 24 n-C6F14 2.55 57.0 33 n-C6 14 2.53 39.0 15 2-CF CY11 2.3 42.0 16 2-oF 3oCF1 2.53 29.8





23

Solid Fluorocarbon from LITR Irradiations


Solids were found to be present in the materials having five or six carbons exposed in the LITR. The solids were recovered by dissolving the sample tubes (after removal of gases) in hydrochloric acid, filtering the residue, and washing. The solids thus recovered were found to have negligible radioactivity. A total of 0.55 grams of solid material was recovered from the eight tubes (two each of n-05F12, cyC5FI0, n-C6F14, and 2-CFC5Fll). Since only a small amount was available from each original material and since the gaseous products of all these were virtually identical the solids were combined for study.

Solubil ity

The solids were soluble in fluorocarbon solvents including Fluorochemical 102 (Minnesota Mining and Manufacturing Co.) with the exception of about 0.02 grams which was probably a residue of the sample tube. (Fluorochemical 102 is a mixture of cyclic ethers of formula 07F140 containing a six membered ring.) They are apparently miscible with these solvents since upon evaporation no crystallization was observed; instead, a gradual increase in viscosity to solidity was noted. The solids are nearly insoluble in acetone, benzene, and carbon tetrachloride. Boiling point

A portion of the solid which had been held at 1000C for several

days to remove the solvent was observed in a Thomas-Hoover melting point apparatus. The material was seen to begin noticeable softening at about 14500 with the viscosity decreasing thereafter as the temperature was increased. This is the expected behavior of such a wide mixture of molecular species. At 19400 and above a very slow bubbling of the liquid was observed. After reaching 24100 the temperature was gradually





24

reduced. No bubbling was noted during cooling. This probably means the bubbling was due to distillation of relatively low boiling materials from the mixture.

Molecular weight

The molecular weight of the solid fluorocarbon was determined by its effect on the vapor pressure of n-07F16. In Figure 4 is a sketch of the isoteniscope constructed for this determination. The sample side of the isoteniscope is made of capillary tubing to minimize the loss of solvent to the vapor space.

To. operate the isoteniscope the solution of solid fluorocarbon in n-07F16 is injected into the sample bulb by a hypodermic. The sample is cooled in ice and a vacuum pulled on the two openings. Vacuum is maintained until a portion of the sample (about 1/3) has boiled off to dispell air from the sample. The sample is then frozen in liquid nitrogen and the capillary pulled off at the arrows. While the sample is still frozen the mercury is dropped into the U tube and the apparatus is ready for use. The isoteniscope is mounted in a water bath and the vapor pressure determined as a function of temperature. After the vapor pressure has been determined at several temperatures, the sample bulb is cooled and broken off. The sample is removed by hypodermic and the concentration determined.

The mole fraction is then the vapor pressure lowering of the

solvent divided by the vapor pressure of the pure solvent. The vapor pressure of the pure solvent has been previously measured(27) and is given by;


log(P,mm) - 6.07134
1.9,52 + t0
-0






-f 25


Mercury reservoir


Figure 4. Isoteniscope used for determination of the molecular

weight of the solid fluorocarbon produced in the

LITR.


Sample bulb







In Table 4 are listed the results of the experiment and the calculated mole fraction.



TABLE 4. RESULTS OF MOLE FRACTION DETERMINATION ON SOLID FLUOROOARBON



Too Solution Pmm Solvent P,mm A (uole fraction)


24 %o.l !6 73.79 -7.37 -0.0998 29.9 103.06 99.74 -3.02 -0.0303 32.6 115.91 113.8 -2.11 -O.0185 34.7 123.13 125.8 +2.7 ..0215 39.65 153.85 157.9 +4.1 +0.026 45.75 195.27 2o6.0 +10.7 +0.0519


The negative results at low temperature indicate some air must have remained in the solution. This is possible due to the high viscosity of the solution. As the pressure of the solvent becomes greater, the air present becomes of less importance and the indicated mole fraction approaches the true value. In Figure 5 is plotted the indicated mole fraction versus the vapor pressure of the solution. The curve approaches a value of approximately 0.055. The solution

was found to be 40.8 weight percent solid. From this data we may now determine the average molecular weight of the solid fluorocarbon.

Let x - molecular weight of solute









0.21 0.20 0.19 0.18 0.17 0o.16 0.15

0.14 013

0.12 0o1 0.10
80


120

Vapor


140 160 pressure, rm. Hg


180


200


Figure 5. Indicated mole fraction vs. solution vapor pressure of solid

fluorocarbon in n-C7F16.


100










x
.52. .4o8 - 0.05
388 x

x - 46o


Some information concerning the structure of the solids can be obtained from infrared absorption and examination of the shorter products determinable by gas chromatography.

Figure 6 is the infrared spectrogram of the solid material run

in a solid pellet containing a mixture of 0.00575 grams of fluorocarbon solid and 0.95816 grams of sodium chloride. The instrument was a Perkin-Elmer 237 Spectrophotometer. All that can be determined from this spectrogram is that there is some unsaturation as indicated by the peak at 6 microns. The large peak at 8 microns corresponds to various carbon-fluorine bonds and is indicative of the complexity of the mixture.

The fluorocarbons formed in the LITR which can be analyzed on the gas chromatograph show a preponderance of the most branched isomer excluding neo types. For 04F1o the iso is the most prevalent. For C5F12

-05F12 predominates, and for 06F14, 2,,-(0F3)204F8 is in largest concentration. From this it is probable that the solid fluorocarbon is highly branched but essentially lacking in neo type structures.

We can summarize the data known about the solid fluorocarbon below;

(1) molecular weight - 4600

(2) contains some unsaturation

(3) has a large amount of branching in its structure.














'0.0




0.10

0 0.20
C)
oas
00
0
0.40


0.o .0.70


5.0 6.0 7-0 8.0 9.o lo.o ll.o 12.0 13,0 14.0 15.0 3.60

Wave length, microns


Figure 6. Infrared spectrogram of solid fluorocarbon from LITR irradiations.







From the above data we should not expect the solid fluorocarbon to have exceptional thermal stability. As has. been mentioned before the thermal stability of fluorocarbons is due to the shielding effect of the fluorine atoms. When one introduces double bonds and strained bonds (due to the highly branched structure) this stability is reduced. Thermal decomposition products

It was thought that some additional information concerning the

solid would be determined by thermally decomposing some of it. A sample weighing 0.2089 grams was sealed in a Vicor tube with a volume of four cubic centimeters and heated to 47000 for a period of about 18 hours. At the end of this time, the gaseous portion of the sample was removed for analysis on the gas chromatograph. The material appeared to decompose almost completely to the gaseous products and carbon. The gaseous products accounted for 0.1688 grams or about 81 wt. percent of the sample,

In Table 5 the mole fractions of the gaseous products are tabulated. Of especial interest is the fact that, like teflon, this material decomposes largely into perfluoroethylene.


Irradiations in the Oak Ridge Graphite Reactor


The irradiations in the Oak Ridge Graphite Reactor represent energy absorptions intermediate between those obtainable in the engineering cobalt-60 source and the energy absorbed in the LITR. In the latter, as has been pointed out, an apparent equilibrium condition was reached.

The energy absorbed by the samples in the Oak Ridge Graphite

Reactor can be estimated from the data of Richardson, Allen, and Boyle(26) for energy deposition in graphite with an appropriate flux









TABLE 5. GASEOUS PRODUCTS FROM THE THEML DECOMPOSITION
OF THE LITR SOLID FLUOROCARBON

Initial Wt., gms. - 0.2089 Wt. Gaseous Products, gms. - 0.1688 Remainder (0.0401 gims.) is carbon with small amount of soluble solid


Decomposition
'I U U


Temperature - 47000 Time = 18 hrs. Tube material = Vicor Tube dimensions:


Length I.D.
Volume


11.2 cm. 1:0.675 cm.
4.01 cm.3


Compound in assay Mole fraction


CF4 0.0006 02F6 0.0025 C2F4 0.7723 o2F 0.0019 C F 0.0154
CY-5F6 oo
cy-OsF60.0006
n-oF6 0.0012 C04F10 0.0517
25.0 min. on silica gel 0.0025 26.2 ain. on silica gel 0.0012
n-C5F12 0.0027 i-03F12 0.0151
VR� - 198, 20 cc. H (n-hex) 0.0811
VR - 269 cc. H2(n-hex) 0.0151 n C F0.0081 2-CF3C5 0. oo4
5-oFCF1 �o.ooi4 2,5- (C 04F8 0.0005
Total O716 0.013 (5)(a)
Total CaF 0.0059 (5 Total C9F " 0.0019 (2) Total " K2 0.0017 (2)
Total C1 24 0.0011
Molecular wt., gas density 95.4


(a) Number of peaks in group





52

correction. From this the G values of the materials can be estimated. These are listed in Table 7. It is seen that these G values are considerably less, by a factor of about three, than those found in the cobalt-60 irradiations. This cannot be due to assuming neutron energy deposition in fluorocarbons is the same as in graphite since the authors state that about 82 percent of the energy results from gauma ray absorption. This G value decrease may represent the approach toward equilibrium. Aside from the G value decrease, the relative amounts of products are close to those encountered in cobalt-6OiQirradiations with the exception of fluorine lost which is about one-third of that expected in 05F8 samples if its ratio to CF4 was maintained. This indicates that in these samples a protective fluoride coating has been established on the wall.


Irradiations in Cobalt-60 and Detailed Discussion


The data from all cobalt-60 irradiations is compiled in Table 18. Perfluoromethane

The most remarkable finding in the irradiation of OF4 is its

great resistance to degradation for high total energy absorption. At low total energy absorption it is more stable than the other fluorocarbons examined, but not unusually so and this stability is due to a greater bond stability as discussed earlier. In the LITR where the other fluorocarbons were completely degraded to an apparent equilibrium

mixture, CF4 remained essentially unchanged. This exceptional stability is explained as being a result of kinetic equilibrium in the LITR irradiation section. In Figure 5, the weight percent OF4 in the samples irradiated in the LITR is plotted versus the fluorine to carbon'ratio








TABLE 6. LITR IRRADIATIONS-ANALYSIS OF GASEOUS PRODUCTS
Position - 0-41 Flux = 1014 n/cm.2-sec.
No. days - 28 Irradiation Temperature 1 i00c

Tube No. 15 16
Compound 2-CF50511 2-CF,05F1i

Compound in assay or standard Mole Fraction(a)
retention volume VR�, Cc. H2

CF0.8546 0.7593 02'6 o.o4 0.1426
02F4 nil nil
C2F2 0.0062 0.0026 CF8 0.0289 O.04108
cy- 8F 0.0010 0.0005
CO (mos ly iso) 0.0159 0.0111
V0 1r cc. H2(n-hex) nil 0.0003 VR- 100 cc. H2(n-hex) nil 0.0019 VR- - 121 cc. H2(n-hex) nil 0.0004
n-C5F12 0.0008 o.0024 i-C5F12 0.0065 0.0217
VR� = 224 cc. H2(n-hex) nil 0.0004 VR0 .(n-O6e14 trace 0.0007
VR - 508 cc. H,2(n-hex) nil 0.0009
2-0F35510FI 2 + 3 0M .0015 2-OF OFll 0.0008 0.0015
2,5-(C;)C4F8 0.0019, 0.0084
Total C7F16 0.0005 (2) 0.0047 (2) Total 08F18 0.0005 (2 0.0060 (2) Total C F 0 .o0001 (1) 0.0048 (4) Total C0 22 nil 0.o060 (5) Total C11F24 0.0012 (4) Total 012F26 0.0009 (1)
Sample wt., Gins. (Initial) 0.1947 0.619
Molecular wt., Gas Density (of gas) 108.5 117.5 Weight percent recovery as gas 57.7 55.8 Moles F2. lost/mole mixture -- Molecular wt. of gas from analysis 100.5 117.8
(a) Number in brackets is number of peaks in group.
Note: In second samples, more care taken to recover higher boilershence the high m.w. analyses are more nearly correct.







TABLE 6 (continued)


Tube No. 19 20 Compound CY-05F10 CY-5Fo10


Compound in assay or standard Mole Fraction(a)
retention volume VR�, cc. H2


oF4
02F6 0 JA 02F2 05F8
0 cy-83F6
- 49 cc. H2(n-hex) "" 04F1o(mostly iso)
VR�= 100 cc. H2(n-hex)
n-C5,F12 i-cFl2
= 165, 177 cc. H(n-hex)
0 - 224 cc. H2(n-hex)
n-06Fl4
VR0 - 8 cc. H (n-hex)
2-oF c Fl1

29,5- (0;X5 0Fa
Total a7f16
Total C8F8 Total 09?,
Total 0190-22 Total O-oF24 Total 012F26
Total 0 F
Sample wt., Gmns. (Initial)


Molecular wt., Gas Density (of gas) Weight percent recovery as gas Molecular wt. of gas from analysis


0.8497 0.0912
nil
0.0079 0.0258 o.o=6
0.o016
nil
0.0128
nil
0.0009
0.0076




nil
Ii





U
n Ii

0.0002 (1) 0.0o02 (1)
nil
U
n

0.263

118 49.8
:99. 4


(a) Number in brackets is number of peaks in group.


0.7920 0.0921
nil
0.0029
0.0502 0.0005 0.0003 0.0191
0.0052
0.0007 0.0021 0.240
0.0007
o.=o4 0.0017

0.0010 0.*0010
0.o94 0.0052
0.0054 0.0026
0.0022 0.0011
0.ooo6 0.0007
0.214

129.3 85.5 109.8







TABLE 6 (continued)


Tube No. 8 10 Compound OF4 OF4


Compound in assay or standard (a)
retention volume VR I co H2i


CF4 02F6
C2F4
02F2
0 F8
cy-0 F6
C4Fl (mos ly iso)
'R 0 o cc. H2(n-hex)
n-05F12
i-C 5F lknhx
VR0 224 cc.'(nhx
n-C6Fl4
VR0 - 308 cc. H2(n-hex)
2-0F 0 F
2-CF F11

2,5-(C 324F8
Total 07F16
Total C8F8

Total C1F24 Total 012F26 Total %-F26
Total 1, Total 016 Total 01 Total 016 Total C17
Toa 18
Total 019
Sample wt., Gms. (Initial)

Nblecular wt., Gas Density (of gas) Weight percent recovery as gas Moles F2 lost/mole mixture Molecular wt. of gas from analysis

(a) .Number in brackets is number of


0.9866 0.0077
nil
0.0055
0.0002
nil
a n

It
.
U













100 II U U



02 8.
n U U a
U



lO
0.,4 88U
pe U ngop


(b) This average molecular weight is obviously grossly in error.


0.9809
0.0077
nil 0.0091
0.0001 0.0001
0.0021 nil
I,
U

n


n n

n

n
11
ti







o.o II


It U U U
u U
n

0.1804

143.2(b)
100
o.o414 88.46







TABLE 6 (continued)


Tube No. 4 5
Compound 02F6 02F6

Compound in assay or standard Mole Fraotion(a)
retention volume VR�, cc. H2

OF] 0.7467 0.7244 COF6 0.1005 O.lO48
02F4 nil nil
02F2 0.0047 0.oo47 C F8 0.0594 0.0520
Cy-O8F 0.0007 0.0009
V -1(m0s Y iso) 0.0271 0.0184
VR� 100 cc. H2(n-hex) nil nil
n-05Fl2 0.0050 0.0014
�i-O .5F 0.0382 0.0o18
VR- 224 cc.142(n-hex) nil nil
n-06F14 0.0009 0.0006
vRO- 08 cc. 2 (n-hex) nil nil
2-CF5C5FI- 0.0022 0.0014 5-C3C5 F 0.0015 0.0012
2, ()5o08 0.0180 0.0122
Total 07F16 0.0055 (2) 0.0048 (2) Total 08F18 0.0062 (2) 0.0042 (1) Total C9FP0 0.0022 (5 0.0129 (4) Total 0OF 0.0059 (5) 0.0157 (5) Total 011F24 0.0018 (5) 0.0095 (5) Total 012F26 0.0005 (2) 0.0092 (7) Total 0 Fj8 nil 0.0055 (4) Total 3 -" 0"0017 (4 Total 015 0.0007 (25 Total 016 0.0015 () Total 017 " 0.0054 (2 Total lB0o0o61 (4) Total 018 " 0.0025 (I
Sample wt., Gms. (Initial) 0.2887 0.505

Molecular wt., Gas Density (of gas) 122.7 153 Weight percent recovery as gas 100 100 Moles F2 lost/mole mixture -- -Molecular wt. of gas from analysis 121.9 148.8


(a) Number in brackets is number of peaks in group.






TABLE 6 (continued)


Tube No. 28 12
Compound 03% n-05Fi2

Compound in assay or standard (a)
retention volume VR�, Co. H2 Mol Fraction

OF4 0.7293 0.8338 02F6 0.1358 0.0979
CA nil nil
02F2 0.0032 0.0050 C F8 0.0491 0.0314
cy-; F6 0.0006 0.0006
0F (mosly iso ) 0.0194 0.0150
V n4 00 cc. H2(n-hex) nil nil
n-O5F12 0.0018 0.0018 i-C5F12 O.0169 0.0082
VR0 - 224 cc. H2(n-hex) trace 0.0003
n-C6Fl4 0.0004 trace
VRO - 308 cc. H2(n-hex) nil nil
2-CF 0] 1 0.0017 2 + = 3-OF'O' 0.0010 0.0005
2To-(OK3,F8 0.0080 0.0039
Total 0.0044 (2 0.0004 (2 Total 7 18 0.0078 (2 0.0006 (2 Total CF 0.0042 (3 0.0001 (1
Total C0 2 0.0080 (5 nil
Total CllF 4 0.0053 (4)
Total 012F26 0-0039 (5 Total 0 F28 o.0015 (4
Total 14 nil
N N
Total 015 n
Total 016 a
Total 017 . n
Total 018
Total 019 " "
Sample wt., Gms. (Initial) 0.5110 0.2110
Molecular wt., Gas Density (of gas) 120.5 114.5 Weight percent recovery as gas 67.6 90.2 Moles F2 lost/mole mixture -Molecular wt. of gas from analysis 123.9 101.6


(a) Number in brackets is number of peaks in group.







TABLE 6 (continued)


Tube No. 14 24

Compound n-C5F12 n-C6F11


Compound in assay or standard Mole Fraction(a)
retention volume VR, cc. H2


CF4 0.76.57 0.8251
02F6 0.1194 0.1080
02F4 nil nil
C2F2 0.0055 0.0017
OsF8 0.0519 0.0526
cy-F8 0.oo6 0.0005

C4FlO(oosly iso) 0.0185 0.0175
VRO 1 100 cc. H2(n-hex) 0.0001 nil
n-C5F12 0.0019 0.0012 i-C5F19 0l4o0 0.0082
V0 - 224 cc. 2(n-hex) trace nil
R T-C6F14 0.0004 0.0001
V - 508 cc. H2(n-hex) nil nil
2-F0 F 0.0010 0.0005 5-00FlI 0.0005 0.0005
2,5-(C3 6F24F8 0.0065 0.0028
Total C F 6 0.0097 (2) 0.0007 (2) Total 08F1 0.0072 (2 0.0007 (2 Total C9F20 0.0025 (5) 0.0002 (5 Total 010 22 0.0027 (5 0.000 (5 Total ClF24 0.0019 (4) 0.0001 (1)
Total C12F26 0.0104 2 nil
Total 0 Fj80=
Total CF2
Total 4 nil
Total C1 "
Total C16
Total 017 "
C1 7
Total u18 Total 019
Sample wt., Gins. (Initial) 0.4210 0.2556

Molecular wt., Gas Density (of gas) 112 115 Weight percent recovery as gas 56.8 52.5 Moles F2 lost/mole mixture - Molecular wt. of gas from analysis 119.6 102.8


-(a) Number in brackets is number of peaks in group.







TABLE 6 (continued)


Tube No. 35
Compound n-C 6F14

Compound in assay or standard Mole Fraction(a)
retention volume VR�P cc. H2

OF4 0.8017 C2F6 0m1162
C2F4 nil
C 2F2 0.0036
C-F 0.0574
cy-8 0.0005
04F10(mos ly iso) 0.0124
VR�- 100 cc. H2(n-hex) 0.0001
n-C5F12 0.0013 i-C F1 0.0103
S- 224 2. 2(n-hex) 0.00004
n-C6Fla 0.0005
VRO 508 cc. I2(n-hex) nil
2-CF 0 Fll 0.0008 5-OF'CO'FI 0.0004
2,5- (C5201F8 0.0052
Total 07F16 0.0022 (2) Total CF, 0.0050 (2 Total CF P 0.0015? (2)
Total 0 -6o.oo (5) Total 010 22 0.0021 (5 .1Total 0-F24 0.0005 (2 Total C2F26 0.0002 (1)
Total C-F-6 nil
Total 628 It
Total 014 Total 0-5 Total C16 Total C17 Total C18 Total 01e
Sample wt.,, Gms. (Initial) 0.575
Molecular wt., Gas Density (of gas) 107 Weight percent recovery as gas 59.2 Moles F2 lost/mole mixture -Molecular wt. of gas from analysis 107

(a) Number in brackets is number of peaks in group.







TABLE 7. RESULTS OF OAK RIDGE GRAPHITE REACTOR IRRADIATIONS


Position - Hole #A


Flux - 1012 n/cm.2-sec.


No. Days - 55


Approximate energy absorption(26) - 8.5 x 1010 erg/gram

Irradiation temperature - 10000 or greater


Compound in assay or standard Tubelo. F70
0 yclo-O F retention volume VROp ccH2 Mole Frac iona)


OF4 C0F6
CAF C2F2 0 F8
04;i0
79 on 06H,4
117 on C16514
n-C5F1 2 i-C5F12
184 on C63
Oy-05 10
n-6F14
2-C0F_0F1


Total ;7 16 Total C8F18 Total 09F2
Total ClOF22 Total 011F24 Total Cl 2F26 Total 013 F28 Total C14F30
Sample wt. gns.
Molecular wt., Gas Density Percent recovery as gas
,9


0.0098" 0.0088
0.0oo6
0.0015
0.0055 o.0oo6
0.0033 0.0007
0.0023 0.0049 0.0007
0.0202 0.8866
0.0013 0.0005
0.0008 0.0099
0.0109 4 0.0136 4 0.0090 4
0.0018
0.0031 4 o.0o6 4 nil
0.0001 1
0.2359
264. 100. 0503


(a) Number in brackets is number of peaks in group.







TABLE 7 (continued)


Tube No. 82 67 Original Material OF4 C2F6 E"Ompound in ;' "or'sadr
assy Vstand.ard
retention Volume o ,o.t2Ml rcina


Moles 2





91
2










I I





Wt. percent c


lost/mole mixture
OF4 02F6
C2F w22 03F8
Oy- F6

4.5 on 19H5 L17 on 016H54
n-OFl2 i-CF12
.6 on C16HZIL .95 on C161 54 30 on C16 4 a6O on C,< z


recovered as


gas
Sample wt., gins.
.Molecular wt., Gas Density (of gas)
7-G, gaseous products


0.0208 0.9905 0.005
0.0054 o.oo=4
0.0001 nil



U

U U U Is






U a




n 3,





10.
01









8.
00.3
0.17 87. o.


0.020 0,0800
o.8491 0.0012 o.o452
0.0007
0.0136
nil
U
0.0017 0.0031
nil



0.0005
0.0002 0.0020
0.005 (2) 0.0011 (2) o0oo (2)

0.0005 (4) 0.0001 (1)
nil
aI


100.
0.3639
141.5 1.38


(a Number in brackets is number of peaks in groups.







TABLE 7 (continued)


Tube No. 77 Original Material 05F8 n-O5F12 Compound in assay or standard
retention Volume VR, c H2 Mole Fractiona)


Moles F2 lost/mole mixture
OF4
02F6 C 2F2
0 F


94.5 18H 4
117 on

i-C5F12
163 on C6H54 195 on C16p4 250 on 016H54 260 on C16H54
n-06FI4
2-0F 0 F
5 0 11Cl

2$3-( ;PC
Total 07F16 Total 08F18 Total C9F20
Total 010 22
Total 01F24Total C12F26 Total 013F28
Wt. percent of sample recovered as gas
Sample wt., gins.
Molecular wt., Gas Density (of gas)
.. G, gaseous products


o.o482 0.05559 0.0470 0.0005 0.7701
Not determined
0.0500
nil
n
0. 0117 0.0204
nil
a It IT
0.0042 0.0087
0.0017 0.0093
0.0059 (2) o.o44 (2) 0.0038 (2) 0.o8 (6)


nil
100.
o.6180 198.
1.55


(a) Number in brackets is number of peaks in groups.


0. 0455 o.o5 1 0.0011 0.0298 0. 0004 0.0309
nil

0.6787
nil
0.0031
nil
0.0055
nil
0.0114 0.0107
0.0148 O.0016 o.o594 0.0362 0.0088 0.0363
o.096
0.00 8
nil
92.2
o.4454 293.
1.27


(5)


(5)
(2)

(5)







TABLE 7 (continued)


Tube No. 79 70(b) Original Material n-06F14 .2-0F505F11
Compound in assay or standard Mole oa retention Volume VR�, co. H2


Moles F2 lost/mole mixture
CF4
CF 2 22 C38
Cy-C3F6

94.5 on 016H54 117 on 016H 4
n-C0912
i-C F
163 og 6H5
16653
195 on 016 54
250 onCH
26o on 016 h

To
n-O6F1 63 2-oFzC5 1
TOFtal CF
2o K-( i

Total 0F
Total C 7F1

Total 10 F22
Total 011 Total C9 FP6 1022
Total C 1F28
Wt. percent of sampl6 recovered as gas
Sample wt., gns.
Molecular wt., Gas Density (of gas)
S. G, gaseous products


o.o468
0.0354 o. 0006 0.0501
o.=4o 0.0004 0.0511 0.0005
0.0002 o.0268 0.0051
nil
N

nil
0.7079 0-0038

0.0009 0.0542 0.0247
0.0220 0.*0131 0.0103
0.0032
nil
80.2
0. 5595
551.
1.05 -


0.1087 0.0366
0.0008 0,0502

o.=a6
nil
0.0006 0.0120 0.0087
nil
0.0003
nil
0.0005 0.0008
0.5978
Not determined
0. 0041,
o.o436 (2) 0.0238 (2
0.0514
o.o554 (4 oxi82 (4) 0.011 (5

0.0030 (1)
68.
0.5193
520.
1.47


(a Number in brackets is number of peaks in groups. (b This sample contained unsaturate.







of the original sample. This is seen to give a smooth curve starting with zero percent OF4 for pure carbon and approaching 100 percent OF4 for ratios of four and greater. It should be noted that the structure of the original material is not a factor.

To obtain an idea of the radicals likely to be formed in the irradiation, we will first examine the results of mass spectrograph work. In Table 8 are given the species produced in the mass spectrographC5) for three electron energies; 50, 70 and 100 volts. From this data the ratios of the various ions are seen to vary considerably over even this small energy range. This indicates the fallacy in prediction of gamma yields from mass spectrograph data. However, from this data we see that the two fluorocarbon species which are increasing relative to OF3 are OF and OF2. From this we may expect that these species are even more prevalent when the electrons have the much higher energies encountered in gamma ray irradiation and for this reason they are included in the proposed mechanism for OF4.

In Table 9 is given the chemical mechanism for perfluoromethane with the understanding that ions are probably the initial products followed by neutralization.

As can be seen, the possible reactions for even this simple

system are very great. Reactions A-4 to A-9 and reaction A-15 result in the regeneration of OF4. Reactions A-7 to A-9 will be important for long irradiations where a significant amount of fluorine has built up. These reactions (A7 to A-9) account for the equilibrium for OF4 being nearly pure OF4. Since no O2F4 was detected from OF4 irradiations, reaction A-13 must predominate over reaction A-12. In most cases the













TABLE 8. MASS SPECTROGRAPH DATA FOR OF4(5)


Species Relative Intensity at 50v 70v 100v


6.80 3.09

3.67

14.o4 1o1.16


10.21 5.95

4.91 21.50 123.35


11.12 8.14 6.32

26.77 138.38


Ratio to CF3


0.0672 0.0305 0.0363 0.1388


0.0828 0.0483 0.0598 0.174


0.0803 0.0587 0.o46 0.193


C

F

OF

OF2 OF3


0

F OF

OF2








TABLE 9. OHEAIOAL MBOHAWISM FOR OF4



Important species in mass spectrogram(6) OF5, OF2, C, OF Radical Generation,
OF4 V CF3 + F (A-1) oF2 + 2-F (A-2)
-"--4 F + 37. (A-3) Radical Recombination
OF + F --) OF2 (A-4) 0F2 + F - OF (A-5) oF5 + F -* CF4 (A-6) OF + F2 --OF (A-7) CF2 + F2 - O CF4 (A-8) CF3 + F2 -+oF4 + F (A-9)
OF + OF - 0 C2F2 (A-10) OF + OF -CF2 + 0 (A-i) OF2 + oF2 -- 02F4 (A-12) oF2 + oF2--- OF + OF3 (A-13) CF3 + OF3 --* 2F6 (A-14) oF3 + OF, -3 oF2 + OF4 (A-15) unsaturates + F, F2, radicals --i saturated molecules (A-16)





47

amounts of 02F2 and 02F6 were nearly equal which could be accounted for by the simple mechanism below:


OF4 .-V OF2 + 2F

OF2 + OF2- OF + OF3

OF + OF- 02F2

CF3 + OF 36


However at elevated temperature (9900) only a very small amount of C2F2 was produced and no CF was detected. From this we conclude that the simple mechanism above is not adequate and that the true mechanism is considerably more complex.


Perfluoro ethane


In the mass spectrograph of 02(6) the most prevalent species in decreasing importance are OF 3 2F OF, and OF2. The stability of perfluoroethane to fragmentation by electron bombardment is not as great as that of OF4 as is shown by the mass spectrograph studies (see Table 8). In long term irradiations it will be degraded a great deal since by Figure 3 at equilibrium the sample will be about 50 weight percent OF4.

Judging from the mass spectrograph studies the initial radical forming reactions must be those given in Table 10.

Other reactions are possible, but the ones in Table 10 are

sufficient to explain the results. At room temperature, no C2F2 is found, indicating that reaction B-11 is not important or that reactions of the type of B-15 are important. Evidence for the latter is that at elevated temperature where the amount of 02F2 is increased, the amount












TABLE 10. CHEMICAL MECHANISM FOR 02F6




Important mass spectrograph species (6) CF3 25 CFi CF2 Radical Generation

02F%-.A-- 2CF3 (B-1)

-W-- C2F5 + F (B-2) -V---OF, CF2,1 F (B-3) Radical Recombination

F + F --wF2 (B-4) CF3 + F 0 CF4 (B-5) 20F3 1,0 (B-6) 02F5 + F -0.CAF (B-7) CF9 CF29, - v- C2F6 (B-8) C73F 2 :v i. F (B-9) C2Fg + F2- .0 CA + F (B-10) CF + CF - 0 2F2 (B-11) CF 2 + 0F2- CF + CF5 (B-12) 02F5 + CF3 o~ C3F8 (B-135) 02F5 + C2F5 -+ n-00F10 (B-14) unsaturated molecules + F, F2, radicals :o saturated
molecules (B-15)






An

of larger molecules is decreased. A simplified mechanism which does a fairly good job of correlating the data at room temperature and for low energy absorption consists of equations B-i, B-2, B-4 to B-7, B-13,

and B-14.

This mechanism has been used in Appendix 6 to correlate the

ambient temperature data by a statistical method. In essence this method involves assignment of an empirical effectiveness to each type of radical. This effectiveness is related to the ease of rupture of the bond to form the radical and possibly to various geometric factors. Multiplication of the effectiveness times the number of ways the radical can

be formed results in an effective concentration of the radical. By random recombination calculations, the relative concentrations of products can be calculated. In this case, the reverse of this method is used to arrive at the effectiveness. These effectiveness numbers are of value in two ways. First, using them ,.an.- estimate of the fraction of the radicals which recombine to form the original compound can be obtained and second, they give an estimate of which type of bonds are preferentially broken in gamma irradiation. A further discussion of this approach and the calculation are given in Appendix 6.

For 2 F6 approximately one third of the radicals recombine to yield 02F6. From the effectiveness numbers we see that OF radicals are more than twice as effective as 02F5 radicals. This probably arises from preferential formation of OF as indicated by the mass spectrograph results.


Perfluoropropane

A more detailed examination of perfluoropropane was made than

for any of the other materials studied. Perfluoropropane was irradiated





50

over a range of energy absorptions and densities and at two temperatures in the engineering cobalt-60 source. At the lower temperature, about 5000, the samples were partially liquid and these allow examination of the mechanism in the liquid phase. At the upper temperature, about 10000, the samples were in a single phase allowing examination of the effects of density. Samples were also irradiated in the Oak Ridge Graphite Reactor and the LITR reactor.

In the LITR, where an equilibrium distribution was obtained, we see that at equilibrium about 40 weight percent of the sample is perfluoromethane. (See Table 5). Thus OF4 will show a tendency to higher concentrations for higher energy absorptions.

The proposed mechanism for perfluoropropane is given in Table 11. During the irradiation of perfluoropropane fluorine is produced by reaction o-4 and may either react with the wall by reaction 0-25, with radicals as in reactions 0-20, 0-21, and 0-22, or with unsaturated materials as in 0-19. Thus the wall competes with the fluorocarbon system for the fluorine. Reaction at the wall will be expected to decline over a period of time due to the formation of a protective fluoride coating. Reaction of fluorine at the wall will also decline with increasing density since diffusion of fluorine to the wall will be retarded. These two effects are seen in Figures 7 and 9 for the products OF4 and C2F6 in the single phase cobalt-60 irradiations, and in Figure 8 for 0F4 in the partially liquid cobalt-60 irradiation. Figure 10 for the product 02F6 in the partially liquid runs does not show such a striking trend although some difference may be present for low density. In general, these plots show an increase in the G values for CF4 and 02F6 with increasing density and energy absorption. This




51
TABLE 11. CHEM1IOAL MHANISM FOR 098

Important mass spectrograph species(6) OF3, OF, 03F7 Radical Formation
03F8 - 0937+ F c-') C-F8 .F -+- F. OF3 (0-2) 03F8 -Vv--+ unsaturated. radicals + F (0-5) Radical Recombination
F + F --- F2 (-4) CF3 + F 01 OF4 (0-5) 02F5 + F -- O2F6 (C-6) OF3 + OF 3-- bCA (0-7) 05F7 + F - F8 (0-8) 02F5+ CF 3 --- 3F8 (0-9) unsaturated radicals + F -*0F8 (0-10) 02F5 + o2F5 -n-C4F1o (0-1) n-0F7 + OF5----*n-O4Flo (0-12) i-0F7 + CF5--- '-00F0 (0-13) n-C3F7 + 02F5 -on-05F12.- (0-14) i-0F7 + 02F5 ---*"-05F2 (0-15) n-C3F7 + n-C F7 -3 n-06F14 (0-16) n-C F7 + i-c3 F 7-- 2-CF305Fll (0-17) i-C5F7 + i-C3F -+-- 2,5-(OF3)204F8 (0-18) unsaturates + radicals or F2- saturates (0-19)
OF3 + F2 - OF4 + F (-20)

02F5--2F6 + F (0-21) O3F7 + F-.., o8 + F (0-22) 2A + 3F2 --A A2F6 (0-23)














0.8 0.7

o.6 0.5


0.4


0.3 16.5
0 6.5
0.2


0.1


00
0 .1 0.2 0.3 0.4

Density, g Figure 7. G(CF4) versus density for

(a) Numbers beside points are ergs


0. 0.6 0.7

/cm.3


990C C3F8 .pl.(a). absorbed per gram xi 10".


0
C,













1.0 80.2 "



0. 870 20
0 -7 64..7 36.2 9 0 115

000.6 o -X 8 03. 6600Q61 0 3640.5
0 061111111


*0 0,.1 0.2 0.3 0o 0.$ 0,.6 0.,7 0.8 0.9
Bulk density, g�/czn�3.

Figure 8. G(CF ) versus bulk density for 3000 C3F8 samples (0obal- )(a).

(a) Nubers beside points are. ergs absorbed per gram I 0"8.


















1.0

0.9
0.8.
0.8 2.5

0, 66 '.5 Do16
0.6 325Liquid~FYT

.16.5
o e0

0o.3 --1O6-5

Aluminum
0Oe2 --powder

0.1
0 1 I I I *1 I

0 0-1 0.2 03 04 0.5 0.6 0.7 0.8 Density, g./cm.3


Figure 9. G(C2F6) versus density for 9900 3F8 samplesi
(cobalt-60).

(a) Numbers beside points are ergs absorbed per gram X 108.

(b) This is the G value for liquid C3F8 found by extrapolation

of the plot for partially liquid samples to the liquid
density.





















Q) 59.2 (976.4


. 64.7 - l1.2 Q 188
87--V . aeq 80.2 .0. 87.0 120 34.2
X0 37.7 ( 36.7
5 /









I I I I I I II I I .


0 Oo. 0.2 0.3

Bulk


o.4 0.5

density,


0.6 0.7

gm./cm.3


0.8 0.9 1.0


Figure 10.


G(C2F6) versus bulk density for
(cobalt-60)(a-).


300C C3F8 samples


Numbers beside points are ergs absorbed per gram X 10"8. The line indicates the average G value for the samples.


14o 0.9

0.8 0.7

0.6 0.5

0.4

0.3


NO Ir%4 CoJ C,
C,


'4


0.2

0.


(a)
(b)







is to be expected since they can be formed by reaction of radicals with fluorine (equations 0-20, 0-21). Molecules larger than 05F8 cannot be formed in this way and no corresponding increase of G value is noted (Figures 15-26). Correspondingly, the plots of G values for fluorine lost to the wall show a decrease with increasing density (Figures 11, 12). No effect of energy absorption is seen indicating that the equilibrium concentration of fluorine had not been built up over the time period since such an effect was seen for CF4 and O2F6. Thus for liquid fluorocarbons where the density is about 2 grams per cubic centimeter

diffusion of fluorine to the wall should be less important, especially with a less favorable geometry. This is confirmed by the work of J. H. Simons(I) with 06F14 irradiated in the Oak Ridge Graphite Reactor where no fluorine attack on the aluminum container was detected.

The capture of radicals at the wall is indicated by the very low density sample (number 92) irradiated at 5000. In this sample the overall G value was only about 0.8 as compared with about 4.8 for higher densities. From this we conclude that at low density the radicals produced can be scavenged by the aluminum wall. Another experiment was run with the sample tube partially filled with powdered aluminum (number 85). The overall G value for this tube was not significantly different from the G values of tubes not containing aluminum powder. However, the product distribution was considerably different with smaller amounts of OF4, C2F6, and 04Fl0 being formed indicating a decrease in the species F2, F, CF5, and 02F5, presumable by capture by the solid surface. Larger molecules show no such decrease. The largest individual peak in this sample has not been identified, but it is not a fluorocarbon and must contain oxygen derived from the aluminum














2.2
2.0 C 16.5

1.8



1.4
1, 016.5
i.2 1 8932.5
*j 1.0 O 32.5. FZz i. _ 3. - 0.,6.5

0.8 32, 0 1

0.6 0.4

0.2


0 01 0.2 0.3 o0. 0.5 0.6 0.7 Qo8 0.9 Density, g./cm.3 Figure 11. G(F2) lost versus density for 990G.C3F8 samples
(cobalt-0) ab dm
(a) Numbers beside points are ergs absorbed per gram X 10"8.














2.0 1.8 1.6



1.2

r.0
C
0.8 0.6 0.2 0,2
0


064.
---- t 26.6




37






_ Q225.6
15.1
-1 I


0 0.1 0.2 0.3

Bulk


7

0

036.2
87.0 115
087.8


120.0
034.2



0 108.8 36.1

Q 80.2

( 36.-7


0 76.4


0.4 0.5 0.6 0.7 density, g./cm..3


85.1
0

!


0.8 0.9 1.0


Figure 12. G(F2) lost versus bulk density for 300C C3F8
samples (cobalt-60)(a).

(a) Numbers beside points are ergs absorbed per gram X 108.












































0 0.1 0.2 0.3 o.4 0.5 0.6

Density, gm.//cm.3


0.7 0.8 0.9 1.0


Figure 13. G(C4Flo) versus density for 990C C3F8 samples

(cobalt-60) (a).

(a) Numbers beside points'are ergs absorbed per gram X 10

(b) This is the G value for liquid C3F8 found by extrapolation

of the plot for partially liquid samples to the liquid

density.

(c) The line indicated the average G value for the-samples.


16.5
16.5)
16.5

(c)


Liquid (b


1.3

1.2 1.1 1.0

0.9 0.8 0.7 0.6


0 Aluminum
powder


- 016.5


0.5

o.h 0.3

0.2

0.1

0


| |


3 2 . . 16 .5 -













1.3

1.2 1..

1.0, 0.9

0.8 0.7 0.6 0.5


Oh

0.3 0.2 0.1

0


()59.o2

108.8
36,2 O0
0 76.4 ) 034.2
37.7 0 C115.0 36i7 (b)


(64.7 66.0 0 26.6


o 87.8 o9 87.0


080.2

0 36..1 020.0


8S .1
0


~1


( 25.6


15.1
I


0 0.1 0.2 0.3

Bulk


0.4 o.5 density,


I I


0.6 0.7 0.8

gm./cm.3


0.9 1.0


Figure 14. G(ChFIo) versus bulk density for 300C C3F8
(a)
samples (cobalt-60).
(a) Numbers beside points are ergs absorbed per gram X 108
(b) The line indicated the average G value for the samples.


I I




















Q.3


0.2 1-


0.1


O 0.1 0.2 0.3 O4 0.5 0.6 0.7 0.8 0.9 1.0

Density, gm./cm.3


Figure 15. G(n-C5FI2) versus density for 990C C3F8 samples
(cobalt-60) (a)).,

(a) Numbers beside points are ergs absorbed per gram X 10-8.
(b) This is the G value for liquid C3F8 found by extrapolation
of the plot for partially liquid samples to the liquid

density.
(c) The line indicates the average G value for the samples.


032.5

16 .S .

3.5 l Aluminum
32 powder

- 16. 32.5

I I 11111 II





































0 0.3 0.2 0.3

Bulk


0.4 O.5

density,


0.6 7 0.8 0.9 1.0 gm.*/Cm.3


Figure 16. G(n-C5FI2) versus bulk density for 300C C3F8
samples (cobalt-60)(a).
(a) Numbers beside points are ergs absorbed per gram X 10-P.

(b) The line indicates the average G value for the samples.


0.4



0.3



0.2


r-4
C?
0


36.2 0 59.
64.7 O LQ520
0 034.2
37.7 6. 3.
-- 587.8 80.2
025.6' .%,


0 26.6 0 120.0
36.1

15.1
A 111111I I


8~ .1


0.1



0















0.5.


o l.S0 16.5 � 16.5 Q1.
0.4 126 , 165 16. (c).
.5 16..1c 5

0"216., Liquid (b)
S32.5 32.5

0.3
rAluminum
powder
0 0.2



0.1



0 , "i 1 1 1 ,1 i 1
O 0.1 0.2 0.3 0.l 0.5 0.6 0.7 0.8 0.9 1.0 Density, gm./cm.3

Figure 17. G(i-C5F12) versus density for 990C C3F8 samples
(cobalt-60) (a).

(a) Numbers beside points are ergs absorbed per gram X 10-8.
.(b) This is the G value for liquid C3F8 found by extrapolation.

of the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average. G value for the samples.















36.259.2
0 0

.37.7 0 0 36.7
64.7 115.0 108.8
0 0 .00, (b) 66.o0 87.8 Q 000 80.2
42 .6.5 0.
( 26.6 87.0 \ I 76..4 85 o120,0 8.
87.8




15 .1


0 o.. 0.2 o.3o.4 o.5

Bulk density,


0.6 0.7 0.8 gm./cm.3


0.9 1.0


Figure 18. G(i-C5FI2) versus bulk density for 300C C3F8
samples (cobalt-60)(a).
(a) Numbers beside points are ergs absorbed per gram X 10"8.

(b) The line indicates the average G value for the samples.


CI
,,
U


0.3



0.2 0.1



0


0





. 65


Aluminum powder

32.5
0


pwe*rl5 16 ' L.


0 0 0


16.5,
- 32.5
0
- I


0 16.5 0 16.5.

I I


0.1 0.2 0.3 0.4
Density,


0.5 0.6 0.7 gm.//cm.3


0.8 0.9


Figure 199 G(n-C6F4) versus bulk density for 990G C3F8
samples (cobalt-60)(a&).
(a) Numbers beside points are ergs absorbed per gram X 10-.

(b) This is the G value for liquid C3F8 found by extrapolation
of the plot for partially liquid samples to the liquid density.

(c) The line indicates the average G value for the samples.


0


0.10 0.08 o..o6



0.02
0





.66


0.18

0.16 0.14 0.12

0.10 26.
-0
0.08 C

0.06

0.02

0 15.1 0 01i


36.1
0


11.0


108.8


37.7 U
0 59.2D 34.2 6 36.2 36. '-0
5.6 76.4 80.2 87.0(Q 0) �'�

6.o 8 8.8


(b)


).2 0.3

Bulk


0.4 o.5 0.6 0.7 density, gm./cm.3


S * 1 85.1
0


0.8 0.9


Figure 20. G(n-C6F,4) versus bulk density for 300C C3F8
samples (cobalt-60)(&).

(a) Numbers beside points are ergs absorbed per gram X 10"8.

(b) The line indicates the average G value for the samples.


'0
%


7




67


0.3
Aluminum
powder

0.2 16.5 16.5 U guid (b)

0.2 32 S32.5 ()16.5
c.. 0.1



0 1 1 I I I I I I I
0 0.1 0.2 Oed o. 0.5 0.6 0.7 0.8 0.9

Density, gm./cm.3

Figure 21. G(2-CF3C5F? ) versus density for 9900 C3F8
samples (cobalt-60) aa.

(a) Numbers beside points are ergs absorbed per gram X 10-.

(b)' This is the G value for liquid C3F8 found by extrapolation
of. the plot for partially liquid samples to the liquid
density.
(c)- The line indicates the average G value for the samples.




































0 0.1 0.2


0.3

Bulk


0.4 0.5
density,


0.6 0.7
gm./cm.3


0.8 0.9 1.0


Figure 22. G(2-CF3C5Fl) versus bulk density for 3000 C3F8
samples (cobalt-60)(a).
(a) Numbers beside points are ergs absorbed per gram X 10-.

(b) The line indicates the average G value for the samples.


Q.14 0.3



0.2 0.1


C


6.2


Q) o 0 36.7, 34.2
026.6 36.1 0 59.2 (b)

"-'L.70 37.7 Q,76.40 -0 88
6.00 87.7 j 80.2 C )...
87.0 85 0
120.0Q

15.1
1111111I I


-- 25.6


115.0

































0 0.1 0.2 0.3


0.4 0.5 0.6 o.7 Density, gm./cm.3


-.-Figure 23. G(2,3-(CF3)2ChF8) versus density for 990C C3F8
samples (cobalt-60)(a)�

(a) Numbers beside points are ergs absorbed per gram X 108.

(b) This is the G value for liquid C3F8 found by extrapolation

of the plot for partially liquid samples to the liquid

density.

(c) The line indicates the average G value for the samples.


9.3


0.2 0.1


co

Nz4

Cj


iJ

o 32.5

16.5 16.5

32.5 16.5 '9 32.5
0O 0Al.j Power
16.5 16.5 16.5

1 1 1 1 111 I


0.8 0.9





70


.0.3 1-


0 0.1


0.2 0.3

Bulk


0.4 0.5 o.6 0.7 o.8

density, gm./cm.3


0.9 1.0


Figure 24. G(2,3-(CF3)2ChF8) versus bulk density for 30�0

C98 samples (cobalt-60)(a).

(a) Numbers beside points are ergs absorbed per gram X 108.

(b) The line indicates the average G value for the samples.


o CV
1-S

o7~


0.2 0.1


26.6 0 2


15I25.6 3610. 85.1 .1 36.1 120.0


r~I


59.2





71


6,o 5.0


I4.0 0 3.0

2.0

1.0
A%


0.1 0.2 0.3 004 0.5 0.6 0.7 0.8 0.9. 1.o

Density, gm./cm.3"


Figure 25. Sum of G values versus density for 99�C C3F8

samples (cobalt-60)(a).
(a) Numbers beside points are erg& absorbed per gram X 10-8.

(b) This is the G value for liquid C3F8 found by extrapolation

of the plot for partially liquid samples to the liquid

density.
(c) The line indicates the average G value for the samples.


Aluminum powder 16 32.5

- 2" 32.5 Liquid (b) 7) 0325 35 io,
___ 16.5 (9 16.5
3Z.P 0P16.5

__ 016.5





72


0 0o 0.2 0.3

Bulk


0..4 0.5 0.6 0.7 density, gm./cm.3


0.8 0.9 1.0


Figure 26. Sum of G values versus bulk density for 30�C 03F8'
samples. (cobalt-60)(a).
(a) Numbers beside points are ergs absorbed per gram X 10-8.

(b) The. line indicates the average G value for the samples.


6.0 5.0

4.0

3.0

.2.0

1.0

0


:
64.7 36.2 66.60 26.60 D


025.6 15.1


59.2


I .





73

oxide. Both these tubes showed an increased amount of unsaturated material leading to the conclusion that, in perfluoropropane irradiation, unsaturated molecules are present as intermediates. Relative amounts of unsaturated materials were determined from the amounts of 022, o24, and 036 detected.

G values in the partially liquid samples and in the single

phase samples are not significantly different so the mechanism is probably the same.

As in the case of 02F6, a random recombination calculation has been made to determine the effective concentrations of the radicals. For this calculation, a-simplified mechanism consisting of reactions C-1, 0-2, 0-4 to 0-9, and C-l to 0-18 has been used. The results indicate that approximately 26 percent of the radicals formed recombine to form 03F8. Of the radicals formed, the i-C3F7 radical is the most effective leading to the conclusion that the fluorine atoms attached to a carbon atom next to a OF3 group are relatively easy to remove.

The n-C3F7 radical is only 0.515 times as effective so removal of a fluorine atom from an attached OF5 group is rather difficult. The 02F5 radical is slightly more effective than the n-03F7 radical either from geometric effects or because the bond is slightly easier to rupture. The OF effectiveness is roughly twice that of CZ5 probably resulting from some complete rupture of 03F8 into single carbon entities. This is to be expected from the mass spectrograph data.

In summary we can say for 03F8, and by analogy for the other fluorocarbons, both energy absorption and density affect the G values of the products of irradiation if the geometry of the sample container is such that the wall can act as a competitor for fluorine and radicals.







Perfluoro-n-Butane

The purification of this material by gas chromatography was hampered by the inability to separate the two isomers. To. obtain a reasonable separation the material on the first part of the peak, where n-04F10 appears, was cut. Direct analysis by chromatography can give no indication of the purity, but an estimate can be obtained by examination of the radiation products. Of the 06 isomers the only ones which can be produced from n-C4F10 are n-06F14 and 3.-CF C5FII. The only isomer which can be produced from i-04F10 is 2,5-(CF5)2C4F8. From mixtures of the two isomers 2-CF3C5F11 can be produced. In the analysis of the radiation products relatively large amounts of n-C6Fl4 and 5-CF 05 F1 are found along with relatively small amounts of 2-CF C Fll and 2,5-(CF3)204F8. If we assume that the ratios of the sums of products from each initial isomer are the same as the ratios of the isomers in the original material and credit the 2-CF C F to both isomers then;

i- 4F10 . .02 + .02 + .01 + .01 w .06 .09
n-04Flo .11 + .12 + .19 + .21 + .01 + .02 .39


or the n-04F10 is about 91.7 mole percent pure. It may be more pure than this since some of the two minor peaks may be second generation

products.

The samples of n-C4Flo were not available until after the elevated temperature cobalt-60 and the reactor irradiations were completed so the only data available is for ambient temperature in the engineering cobalt-60 source. It should be noted that approximately one-third of sample number 95 leaked out between canning and decanning for analysis. This does not appear to have affected the G values





75

with the possible exception of a small decrease in the amount of OF4 formed.

No unsaturated molecules were detected, but these must have been present as intermediates to account for the relatively large amount of material found between 09 and 0 12* The mechanism for n-C4F10 is given in Table 12.

The random recombination calculation has been made for n-C4Flo using a simplified mechanism consisting of equations D-1 to D-4, D-6 to D-9, and D-12 to D-26. From this calculation we see that about 52 percent of the radicals formed recombine to yield n-C F . These data,
4 10
as for the case of 09F8, state that the carbon-fluorine bond next to a CF5 group is relatively weak compared to the carbon-fluorine bond on an intact OF group. CF5 is more important than the other radicals formed by carbon-carbon bond rupture indicating, once again, that in some cases the molecule is ruptured into several fragments. The radicals C2F5 and n-C3F7 which are formed by similar processes have identical effectiveness numbers as would be expected if these numbers are due to ease of formation and not dependent a great deal on geometric factors.


Perfluoro-n-Pentane


In the cobalt-60 irradiation of perfluoro-n-pentane at ambient and elevated temperature, there is no discernable difference between the G values for the products at the two temperatures. For low energy absorption no unsaturated materials are found and after prolonged irradiation only small amounts of 02F2 and cy- A are found. This indicates that any unsaturates formed as initial products are saturated










TABLE 12. CHEMICAL MEOHANISM FOR n-C4Fl0


Important mass spectrometer peaks(6)
oF3, 02F,, OF, 0F59
Radical Generation
n-C4F10 - 3 n-CF 7 + OF3 (D-1)
2 2 F (D-2)
---- n-0c4F + F (D-)3
.-y---~ 2-C4 F + F
unsaturated radicals + F (D-5)
Radical Recombination

F + F -F2 (D-6) CF9 + F --* C4F10 (D-7) CF 9+ n-0 F7 !p n-O4Flo (D-8)
a 82F5 -4F10 (D-9) unsaturated radicals + F -BIn-OFio (D-lO) unsaturated radicals - unsaturated molecules (D-lI) OF; + F -CF4 (D-12) n-3F7 + F IIC 3 F8 (D-13 C2F5 + F - .26 (D-14) CF3 + CF3 -C2F6 (D-15) CF C 9 C- 2F--OF (D -16) C2F q 035F8- -0 F12 (D-17) n-C4Fq + OF---n-d)F 1 (D-18) i-C4F .+ 0P - 2i-C (F19) 2 +(D-20) i-C4"9 + 02EV5 2 -OFs sFlI (D-21) n-C4F + n-C F7---?n-C F 6 (D-22) i-o4F * n-C5F7 ----0 P F (D- 25)
49 (D.. r.
n-C4F n-49- n-818 (D-24) n-04F9 + i-C49 --- 0FC7F15 (D-25) i-04F9 + i-04F 0 3--"* 4-1:5)�CF12 (D-26) OF3 +F2 --O4 + F ( "-27 02;5 + F--C2F6 + F (D-28) n-CF7 +"F2--CF8 + F (D-29) 04F9 + F2 -'04F~o + F (D-30) unsaturated mlecules + F2, F, radicals
saturated molecules (D-1)





77

by radical or fluorine capture. Small amounts of unsaturates are necessary to account for the significant amounts of Oi1 and 012 formed in the irradiation.

Since the possible number of reactions has become so large, a

generalized reaction mechanism has been written from which the individual reaction equations are easily derived (Table 13).

For the statistical recombination calculation a simplified mechanism consisting of equations E-1 to E-5 and E-7 has been used. From this calculation we conclude that, on a statistical basis, about 25 percent of the radicals formed will recombine to give n-C5FI2. Examination of the calculated effectiveness numbers leads to the conclusion that the further a OF2 group is from a OF3 group the easier its 0-F bonds are to split. This follows from the effectiveness of the radical formed by removal of fluorine in the three position being greater than that formed by removal of fluorine in the two position. In all other respects this data confirms that for the smaller molecules. That is, radicals formed by carbon-carbon bond breakage, other than OF are equally effective indicating that geometric effects are not important, and OF must be formed partially by extensive fragmentation of the original molecules to yield more than one single-carbon group. This is also expebted from mass spectrograph data where OF is in great excess.


Perfluorocycl opentane

Little can be said about the mechanism of radiolytic change in cy-05F10 since very few of the products have been successfully identified. The probable mechanism is given in Table 14.










TABLE 13. OHEMICAL MECHANISM FOR n-05F12




Important species from mass spectrograph(6)
OF3, C2F5, 09F7, OF, 2F, 0 F5

Radical Generation

n- F2N ---onCF,+ F (E-1)
- - 2-05Fl1 + F (E-2)
-, 3-05F11 + F (E-5)

-,V;- OF3 o n-009 (E-4)

n2F5 + -O3 F7 (E-5) unsaturated radicals + F (E-6) Radical Recombination

R1 + R2 R RR2 (E-7) R + F2-- R F + F (E-8) unsaturated radicals -- unsaturated molecules (E-9) unsaturated molecules + F, F2, radicals -4
saturated molecules (E-10)








TABLE 14. CHE4ICAL MECHANISM FOR cy-C5F10




Important species in mass spectrograph(5)
o.51 02F4 OF, OF 04F71 0 3, OF2, 0.F9 Radical Generation


+ F (F-1) 1 I, 5-n-C 5F10 (F-2) j > various smaller fragments (F-5) Recombination
F + F ---F2 (F-4) ~"+ R K iliR (F-5)




+ (F- 6) R 1 + R 2,-*RR2 (F-7) R, + F2--9 RF + F (F-8) unsaturated radicals a, unsaturated molecules (F-9) unsaturated molecules + F, F2,R m .saturated molecules (F-10)





80

Two types of initial radical generation steps appear to occur; splitting of the ring as in equations F-2 and F-5 and removal of a fluorine atom from an otherwise intact ring as in F-4. Production of fragments by F-2 and F-3 does not appear to be as important in gamma irradiation as it is in the mass spectrometer where fragmentation of the ring is highly dominant over removal of a fluorine atom. In gamma radiolysis a large fraction of the products are larger than the original material leading to the conclusion that reactions F-3, F-6, and F-7 are important.

Fluorine lost to the wall cannot be calculated for these samples since the products are not identified. Again since the products are unknown a statistical recombination calculation is not possible.

In the LITR irradiations the cyclic structure of this material did not affect the equilibrium distribution of the products. For its

fluorine to carbon ratio of two the final sample contains about 25 weight percent OF4.


Perfluoro-n-Hexane


Normal perfluorohexane shows considerable difference between

the total G values for the two samples. The reason for this is uncertain since it could be due to either the large difference in the total energy absorbed in the two cases or to the difference between ambient temperature and 9900. Since no temperature effect has been noted in any other case this is probably due to the difference in energy absorption.

The chemical mechanism in Table 15 has been written in general terms since the total number of possible reactions is very high.













TABLE 15. CHEMICAL MOHANISM FOR n-06F14




Mass spectrograph not available


Radical Generation

r-06F14. n-06F13 + F (G-1)

2-C6F13 + F (G-2) 3-06F13 + F (G-3)

SCF5+ n-O5F1 (G-4)

~ C9F5 + ri-04F (G-5) 2--C3F 7 (G--6)

.-T unsaturated radicals + F (G-7)


Radical Recombination

F + F "-,- F2 (G-8) R, + R2 RR2 (G-9) R+ + F2 -"R F + F (G-o) unsaturated radicals -. unsaturated molecules (G-1i) unsaturated molecules + F, F2, radicals
saturated molecules (G-12)





82

A simplified mechanism consisting of equations G-1 to G-6,

G-8, and G-9 has been used in the statistical recombination calculation. From this calculation the recombination of fragments to yield the original n-C6F14 is estimated to be 22.3 percent. As has been the case for the smaller molecules the further a OF2 group is from a OF5 group the easier its carbon-fluorine bonds are to split. Also as in the. other molecules the radicals formed by carbon-carbon bond splitting, excepting OF5, have the same effectiveness number indicating the rupture of any carbon-carbon bond is equally likely. The larger effectiveness of OF must be due to greater fragmentation of some molecules to yield more than one single-carbon radical.

In the LITR at equilibrium about 50 weight percent of the sample is converted to OF4.


Perfluoro- 2-Methyl pentane


A number of samples of this material were irradiated early in the program. However, in testing the thiourea column (see Appendix 3) it was found that about 7 mole percent unsaturates were present. These impurities gave fairly large amounts of unwanted peaks probably resulting from radical capture. While this confirms the hypothesis that relatively small amounts of unsaturated material can contribute significantly to the final product distribution by radical capture the results of these tubes are too complex for interpretation at the present time.

For this reason the cobalt-60 irradiations of the impure material are not reported, and instead a sample irradiated after further purification is reported. The irradiations of the impure material in the Oak Ridge Graphite Reactor and in the LITR are reported since after repurification





83

there was insufficient time for additional reactor runs. In the LITR, due to the formation of an equilibrium mixture, the small amount of unsaturate has little effect.

The mechanism for 2-OF3 05FI is given in Table 16. A simplified mechanism consisting of reactions H-I to H-9, H-I, and H-12 has been used in the random recombination calculations. These calculations predict a 22.8 percent recombination of the fragments to form the original 2-0F30 FI1. As in the other molecules the further a C-F bond is from a CF5 group the more easily it is ruptured. All radicals formed by rupturing carbon-carbon bonds, excepting CF5, are equally effective indicating that for radiolytic rupture all carbon-carbon bonds are equivalent. Hence, as in the other cases, some molecules must rupture extensively to yield more than one single-carbon group.


Perfluoro-2, 3-Dimethyl butane


A sample of this material was not available until late in the

program so only one sample was irradiated in the engineering cobalt-60 source at ambient temperature.

A general mechanism is written in Table 17 for this material

since the possible reactions are numerous. Although no mass spectrograph data is available for this material it is probable that owing to the large number of OF5 groups OF3 would be very predominant. A possible extensive shattering of the molecule is given below.


C-C-C-C
i -v%-- - 4F3 + 2CF
CO 0





84

TABLE 16. CHEMICAL MECHANISM FOR 2-CFA0911



No mass spectrogram available Radical Generation

I I
0 C

-c---o (H-a)
C
0

oC-o-o-o-o (H-43)


I
o
o-- +----c (H-4)
C

-F--+ c-C-c-C-c (H-5)
C
o
�- ,----. CF5 + c-c-c-c. (H-6)

I
--- CF, c-c-c-c- (u- 7)

oaF + c-c-c. (H-8)
I
C

-- n-C3F7 + i-03F7 (H-9) - unsaturated radicals + F (H-10) Radical Recombination

F + F- F2 (H-Il) R+ + R2--, lR2 (H-12) R, + F2--*R1F + F (H-i3) unsaturated radicals--- unsaturated molecules (H-14) unsaturated molecules + F, F2, radicals--saturated molecules (H-I5)











TABLE 17. OHEI-aCAL MECHANISM FOR 2,5-(CF3)204F8


No mass spectroram available Radical Generation
0-0-0o-0 C-0.-0..0. + F(-)
o0 o 0
CC-C-.-C + F (1-2)
f, I

0
-C C OFC3 + --o (1-3) II"

0

2 i-5F7 (1-4)

- ,- unsaturated radicals + F (1-5)

Radical Recombination

F + F -- F2 (1-6) R + R2--. RIR2 (1-7) R, +2"- " R1F + F (1-8) unsaturated radicals -- unsaturated molecules (1-9) unsaturated molecules + F, F2, radicals-saturated molecules (1-10)





86

Evidence for this reaction is the presence of 02F2 in the products from this irradiation. Since this is the only case for a high density parent where an unsaturated material is present in significant amounts for low energy absorption this reaction probably is of considerable importance.

A simplified mechanism consisting of Equations 1-i to 1-4, 1-6, and 1-7 was used to perform the random recombination calculations. The results indicate that about 24 percent of the radicals recombine to give the parent. The effectiveness numbers confirm the conclusions from the other molecules. Carbon-fluorine bonds are more easily ruptured the further they are from OF3 groups. Removal of a fluorine from an attached OF3 group is rather difficult. All carbon-carbon bonds are

equivalent for radiation damage. Some OF 3 radicals are formed by extensive rupture of the molecule into more than one single-carbon fragment.






TABLE 18 PART 1. CF4 IRRADIATED BY COBALT-60 GAMA RAYS


Tube No.
Sample Wt., gms. Density, gm./cm.3 No. Days
Integrated Flux, 107 Roent. Energy Absorbed, 108 ergs
Temperature, OC Compound in assay or as listed


89
o.1681
0.202 9.92 1.705 2.54
30


90
0.1718
0.2o6
9.92 1.705
2.60
30
Mole fraction--G


83
0.1646 0.197
9.9
1.795 2.62

Value a)


Moles F2/mole mixture--G Value


0.00153-1.10


0.99927


C2F6

C2F4
C2F2


3F8

C4Flo C 5F 12 C6 F14


0.0002-0.14

nil

0.00027-0.20

0.00026-0.19

nil nil nil 1.63


0.00145-1.06

0.9991

0.0005-0.36

nil

0.00015-0.11 0.00025-0.19

nil nil nil 1.85


o.oo08-o.55

0.9994 o.ooo4-o.27

nil


trace


o.ooo16-o.11

nil nil nil 0.93


0.001-1.73

0.9948

0.0025-0.44

nil

0.0025-0.44 0.0002-0.04

nil nil nil 2.65


Molecular wt., Gas Density


88.3


(a) Number in brackets is number of peaks in group.


9
0.1852 0.222
33.09 6.65
10.8
30


87.5


88.5


93.4




Full Text
G(CF. )
55
Bulk density, g./cm.3
Figure 8* G(CF^) versus bulk density for 30C C^Fg samples
(cobalt-60)
(a) Numbers beside points are ergs absorbed per gram X 10


108
in the stainless steel and polyethylene are applied Vie find that the
radiation exposure of the solution should be 93*4 percent of that
in the empty radiation tube. This tube is designated as number 11 of
the engineering dobalt-60 source.
One irradiation of 20 minutes was made to determine the flux
in the high radiation zone and one run of 60 minutes was made to
obtain points in the lower portion. This gives a total of six useful
points over the distance of 18 cm. where a significant amount of flux
is encountered. A plot of the flux curve is given in Figure 27 where
the zero point is the bottom of the sample space. This corresponds
to a point 2.4 inches below the center line of the cobalt-60 rods.
Flux Exposure of Samples
The flux to which the samples are exposed was somewhat less
than that in the unoccupied cavity.
In the ambient temperature experiments the flux is absorbed
in the metal of the stainless steel baskets and the wall of the
aluminum sample container. The metal basket has a wall thickness of
about 1.1 millimeter of steel and the aluminum wall thickness is 0.655
millimeters.- This leads to a flux exposure of the sample equal to
94.4 percent of that in the unperturbed radiation zone.
At elevated temperature, in addition to the absorbing layers
present at ambient temperature, there was present in the heating
device J.h millimeters of glass and the equivalent of about one milli
meter of compacted asbestos (density 2.5)* Since a large number of
tubes were irradiated at the same time, the average thickness of
aluminum encountered by a gamma ray is estimated to be 1.2 millimeters.


2
chromatographic methods allowing study of pure single isomers. Analy
sis was also by gas chromatography allowing separation of most of the
materials up through six carbons.
(7)
Other studies of interest are those of Mastrangelo'' where
fluorocarbon radicals were generated by electric discharge and those
of Pritchard, Hsia, and Miller, ^ Seeger and Calvort,^ and Ayscough
and Steacie'* who generated fluorocarbon radicals by photolysis. Mass
spectrograph results^*^ are of some value in determining the impor
tant species in irradiation. These types of data are valuable in the
determination of mechanisms.


5
TABLE 1. COMPARISON OF MASS SPECTROMETRIC
SENSITIVITIES AND G VALUES(a)
Ratio to
Compound Sensitivity for major peale CF
Sensitivity for n-butane
Total
G Value
Ratio to
OF4
cf4
0.575
1.00
1.47
1.00
c2f6
0.955
1.66
5.24
2.2
5f8
1.71
2.97
4.5
5.06
n"c4Fl0
1.78
5.10
5.42
5.89
n"5F12
2.59
4.51
4.91
5.54
Cy-O^F10
1.75
5.01
2.98
2.05
(a) Sensitivities are from reference 6.


175
Table >4 (continued)
Statistical
Experimental
Ratio
Pred.
#52(25uc)"
' ^81(99 6)
#79(0RGR)
f/cf4
1.05
2.28
7.5 8
c2f6/cf4
.688
.719
.91
.725
c5pf/cp4
.664
.875
1.455
.645
n-C4P10/CF4
.712
.750
1.455
.665
n-05P12/0F4
.760
.625
1.045
.575
Total Cy/0F4
5.12
1.625
5.75
.750
Total Cq/GF^
1.465
1.28
2.59
.528
Total C^/CF^
1.417
1.22
2.82
.470
Total C10/OF4
I.570-
1.06
1.68
.280
Total 0 /0F4
1.520
1.155
1.82
.220
Total C^^/CF^
9.02
1.095
5.05
.068


11
TABLE 2.
IMPURITIES IN IRRADIATED SAMPLES
Compound
Mole % Impurity
CF4
small traces of C02, C2F6
2F6
small traces of C02, unknown
C5F8
0.3% cy-CjF
nC4P10
0.18^5 C02, less than 9.5^ ^^4^10
n-5F12
none detected
cy"5F12
none detected
n-C6Fi4
none detected
2-CF5C5F11
0.2k% C6F12
2,>-(0F5)2C4F8
none detected


7
irradiation of cyclohexane was found to originate from the removal of
molecular hydrogen or other non-radical processes. In fluorocarbons
it appears from mass spectrograph data that similar results aro pos
sible by removing moro than one F radical from a single molecule.
4. Abstraction of F from a neutral molecule by an F radical.
Although this type of reaction has been extensively observed in hydro
carbon work, it is not to be expected in fluorocarbon irradiation for
the following reason. The bond energy for the K-H bond in hydrogen
is 104.2 k. cal./g. mole^^ and that for the C-H bond of a hydrocarbon
is about 98.8 k. cal./g. mole^^ leading to a bond energy difference
for hydrogen abstraction of about +5.4 k. cal./g. mole. For fluorene
the latest value for the F-F bond energy is 57.5 k. cal./g. mole^2^
while that for a F-C bond is about 105.4 k. cal./g. mole.^) This
leads to a bond energy difference for fluorine abstraction of -67.9
k. cal./g. mole. Unless the F atom performing the abstraction is very
energetic the abstraction cannot occur.
5. Abstraction of F from a neutral molecule by a radical other
than fluorine. In hydrocarbon work abstraction of hydrogen atoms by
the small radicals and has been
observed. Energetically it appears possible for a fluorocarbon radi
cal to abstract fluorine from a neutral molecule. However, for the
same reason given for the chemical and temperature stability of fluoro
carbons it does not appear that this reaction is important. That is,
the fluorine atoms will act as a protective barrier preventing attack
on the fluorocarbon bonds by free radicals. This conclusion is con
i'
firmed by photolysis experiments. Seeger and Calvert' found no CF4
in the photolysis of trifluoroacetone and concluded from this that


G(2,3-(CF3)2CUF8)
69
Density, gm./cm.^
Figure 23. G(2,3''(CF3)2G[iFg) versus density for 99C C^Fg
samples (cobalt-60)^a\
Q
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average G value for the samples.
!


172
TABLE 55- RESULTS OF RANDOM RECOMBINATION CALCULATION FOR n-C F^
Fragment
N
E
N x E
F
12
.50
5-8 '
OF*
5
2
1.0
2.0
c2f5
2
.45
.90
n-C^Fy
2
ir\
a
.90
n-4F9
2
.45
.90
n'5Pll
6
.54
2.04
2-C5Fn
4
1.065
4.26
5c5fii
2
1.50
5.00
Molecule
Relative .Amount
F2
12.98
cp4
l4.4o
2F6
10.48
C5F8
10.08
n-Vio
IO.89
n_C5F12
72.2
nc6Fi4
10.59
2-0F5c5Fii
17.05
12.0
Total Cy
IS.5
8
17.5
9
16.7
c10
86.5
£. 509.87
percent recombination 25*5


INTRODUCTION
Samples of fluorocarbons irradiated in a cobalt^O source, in
the Graphite reactor at Oak Ridge, and in the LITR reactor at Oak
Ridge have been analyzed by three-column analytical gas chromatog
raphy. The three columns were one meter of silica gel temperature-
programmed for compounds with from one to four carbons, 15*5 meters
of squalane on Chromosorb-P at 92C for compounds with six to about
fourteen carbons, and 16 meters of n-hexadecane on Chromosorb-P at
25C for compounds with three to seven carbon atoms per molecule.
The results of these analyses for which mole fractions and G
values have been calculated are given in Tables 6, 7 and 18. When
no values are given for high molecular weight materials it indicates
that definite peaks could not be distinguished. From C^ on UP> "this
is partially due to the overlapping of the large number of isomers
which are formed by radiation processes. This makes it impossible
to distinguish between minor base line instability and small amounts
of the compounds.
Previous to this work two irradiations of fluorocarbons have
been reported. J. H. Simons and E. H. Taylor^) irradiated C^F-^ in
the Oak Ridge Graphite Reactor and Florin, Wall, and Brown^ exposed
C7F16 SaEma rays. Both these studies were hampered by the use of
impure starting materiels and the use of relatively poor analytical
methods. In this study the compounds were purified by standard gas
1


Weight % CF, in sample
21
Figure 3.
Weight percent CF^ versus fluorine to carbon ratio of
original material for samples irradiated in the LITR


186
(55) G. J. Hie and G. L. Brovmell, Radiation Dosimetry, Academic
Press, Hew York (1956) pp. 25-100.
(56) E. H. Cart, Thesis, Ohio State University.
(57) J. A. Brown, Jnl. of Chem. and Eng. Data .8, 106 (1965).
(58) J. F. Brown, Jr., Scientific American 82, 207 (July, 1962).
(59) Shashi Dat, Personal communication.
(40) R. C. Reid and T. IC. Sherwood, The Properties of Gses and
Liquids, McGraw-Hill Book Co., Inc., Hew York (1956) p. 267.
(41) C. D. Hodgman, ed. in chief, Handbook of Chemistry and Physics,
57th Edition, Chemical Rubber Publishing Co., Cleveland, Ohio
(1955) P- 5078.
(42) R. B. Bird, W. E. Stewart and E. H. Lightfoot, Hotes on Transport
Phenomena, John Wiley & Sons, Inc., Hew York (1958) p. 11.
(4p) M. D. McKinley, Personal communication.
(44) Burton, M., J. Chem. Ed. 28, 412 (1951).


TABLE 18 PART 4 (continued)
Tube No.
Sample Wt. Gms.
Bulk Density, gm./cm.
No. Days
Integrated Dose, 10 Roent.
Energy Absorbed, 10 ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
40
0.5446
0.652
10.0
4.13
20.0
36. T
43
0.1514
0.1815
23.0
7.42
9.99
66.0
Mole Fraction-
39
0.1300
0.1558
10.0
2.99
3.46
2 6.6
G Value
49
0.3190
0.382
23.0
9.9
28.0
87.8
Moles F?/mole mixtureG Value
0.0058-0.78
0.0196-1.46
0.0080-1.52
0.0150-0.85
CFk
0.0040-0.5U
O.OO78-O.58
0.0022-0.42
0.0119-0.67
C2F6
0.0041-0.56
0.0084-0.62
0.0018-0.34
O.OH6-O.65
c2f4
nil
nil
nil
nil
C2Fo
0.00014-0.02
nil
0.0009-0.17
0.0001-0.006
C3F8
0.9728
0.9577
0.9861
0.9401
n-c^F?
nil
nil
nil
nil
cfrio
0.0070-0.95
0.0105-0.78
0.0035-0.66
0.0152-0.86
n-C5F12
0.0018-0.24
0.0024-0.18
0.0007-0.13
0.0037-0.20
i-C^F£2
0.0032-0.44
0.0045-0.34
0.0014-0.27
0.0059-0.34
n-C6Fl4
0.00077-0.10
0.0006-0.05
0.0005-0.09
0.0014-0.07
2-CF3CF
0.0019-0.26
0.0021-0.16
0.0012-0.23
0.0029-0.17
3-CF^C^F
nil
nil
nil
nil
2 ,3-(CFf)pCFo
0.0015-0.20
0.0021-0.16
0.0012-0.23
0.0026-0.15
TotalW,8
0.00072-0.09(2)
0.0014-0.10(2)
0.0003-0.06(2)
0.0016-0.09(2)
Total CgFI'n
0.00075-0.10(2)
0.0013-0.09(2)
0.0004-0.07(2)
0.0012-0.07(2)
Total CqF
0.00056-0.07(2)
0.0009-0.07(2)
0.0001-0.02(1)
0.00085-0.05(2)
Total f
0.00064-0.08(4)
0.0004-0.03(2)
nil
0.00075-0.05(5)
Total C Fpk
nil
nil
nil
0.00009-0.005(2;
Total cJF£g
nil
nil
nil
nil
Total C1oF2Q
£ G
4.43
4.62
4.21
4.231
Molecular Wt., Gas Density
194.3
194.3
189.
192.3
(a) These three samples were j
analyzed before the
addition of the high temperature
squalane column and
do not have complete analyses of the high molecular weight compounds. (b) This tube was copper,
(c) This sample was a single phase. (d) Number in brackets is number of peaks in group.


Rotameter 1
Rot ame te ir 2
B
,-cIj
X-
2
w
Sq.ualane
column
A
Silica
column
B
Figure 3U Series and parallel connections of chromatographic
columns


178
Table 55 (continued)
Ratio
Statistical
Pred.
Experimental
V0F4
.70
5.24
2 V0F4
.507
oil
5VOF4
.407
.289
n-c4Fic/CF4
115')
1
7 572
f 0 289 (mostly
i-C,F /OF.
4 10 4
.257 J
J
n_C5Fl/CF4
.518
.40
i-cf5c4f9/cf4
.212
.178
Total C7/CF4
2.28
1.445
Total Cd/0F.
o 4
.555
.755
Total C /OF,
9 4
.710
1.155
Total C10/CF4
.401
1.51
Total Cn/CF4
.995
1.58
Total C /OF,
12 4
5.59
.778


2$
In Table 4 are listed the results of the experiment and the
calculated mole fraction.
TABLE 4. RESULTS OF MOLE FRACTION DETERMINATION
ON SOLID FLUOROCARBON
TC
Solution P,ram
Solvent P,mm
A
a/p0
(mole fraction)
24
81116
75-79
-7.57
-O.O998
29.9
105.06
99.74
-5.02
-O.O505
52.6
115.91
115.8
-2.11
-0.0185
5^.7
125.15
125.8
+2.7
+0.0215
59.65
155.85
157.9
+4.1
+0.026
45.75
195.27
206.0
+10.7
+0.0519
The negative results at low temperature Indicate some air must
have remained in the solution. This is possible due to the high
viscosity of the solution. As the pressure of the solvent becomes
greater, the air present becomes of less importance and the indicated
mole fraction approaches the true value. In Figure 5 is plotted the
indicated mole fraction versus the vapor pressure of the solution.
The curve approaches a value of approximately 0.055 The solution
was found to be 40.8 weight percent solid. From this data we may
now determine the average molecular weight of the solid fluorocarbon.
Let x molecular weight of solute


58
TABLE 6 (continued)
Tube Ho.
14
24
Compound
n-C5F12
nc6Fl4
Compound in assay or standard
retention volume V^0, cc. Hg
Mole
Fraction^
CF4
2F6
2F4
2F2
c,f8
cy-C^Fg
C4F10(mostly iso)
V 0
VR
v o
VR
i-C^F^
224 cc.
100 cc. Hp(n-hex)
5F12
2 (11-hex)
n_c6Fl4 ,
508 cc. ^(n-hex)
2CF tOcFi
5-CF0Fn
5)204
4f8
2 5 (CFz; 2'
Total C7Fl6
Total CgFl8
Total -
CnF
Total C
9 2
10F22
Total 0^2f24
Total C^2F26
' ' i?f28
Total
Total
Total
Total
Total
Total
Total
14
15
c16
18
19,
Sample wt., Gms. (initial)
0.7657
0.1194
nil
0.0055
0.0519
0.0006
0.0185
0.0001
0.0019
0.0140
trace
0.0004
nil
0.0010
0.0005
O.OO65
0.0097 (2)
0.0072 (2)
0.0025 (5)
0.0027 (5)
0.0019 (4)
0.0104 (2)
0.0004 (1)
nil
11
it
11
11
11
0.4210
0.8251
0.1030
nil
0.0017
0.0526
0.0005
0.0175
nil
0.0012
0.0082
nil
0.0001
nil
0.0005
0.0005
0.0028
0.0007 (2)
0.0007 (2)
0.0002 (5)
0.0005 (5)
0.0001 (1)
nil
n
it
11
I!
11
U
n
0.2556
Molecular wt., Gas Density (of gas) 112
Weight percent recovery as gas 56.8
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis 119.6
115
52.5
102.8
(a) Humber in brackets is number of peaks in group.


147
TABLE 27. STANDARD RETENTION VOLUMES ON THE THIOUREA COLUMN
Pi = 54.7 psi. Po = l6.8 psi. Temp, of column 25C
Carrier Gas = 29*9 cc./min. He at 1 atm. and 25C
Column Length 15*2 meters containing 155*8 grams of 0.2 grams
thiourea/gm. Chromosorb-P.
Compound cc. He at 1 atm. and 25 C
. A
2,5-(cf5)2c4f8
2-CF5C5Fh
5-cf5c5f11
n-C6Fi4
trans 0 C = C -
C
/
C
\
c
544
568
585
4l0
525
c
/
cis c C a C C 577
\
c


24
reduced. No bubbling was noted during cooling. This probably means
the bubbling was due to distillation of relatively low boiling materials
from the mixture.
Molecular weight
The molecular weight of the solid fluorocarbon was determined by
its effeot on the vapor pressure of c-C^F^. In Figure 4 Is a sketch
of the isoteniscope constructed for this determination. The sample
side of the isoteniscope is made of capillary tubing to minimize the
loss of solvent to the vapor space.
To operate the isoteniscope the solution of solid fluorocarbon
in is injected into the sample bulb by a hypodermic. The
sample is cooled in ice and a vacuum pulled on the two openings. Vacuum
is maintained until a portion of the sample (about 1/5) has boiled off
to dispell air from the sample. The sample is then frozen in liquid
nitrogen and the capillary pulled off at the arrows. While the sample
is still frozen the mercury is dropped into the U tube and the apparatus
is ready for use. The isoteniscope is mounted in a water bath and the
vapor pressure determined as a function of temperature. After the vapor
pressure has been determined at several temperatures, the sample bulb
is cooled and broken off. The sample is removed by hypodermic and the
concentration determined.
The mole fraction is then the vapor pressure lowering of the
solvent divided by the vapor pressure of the pure solvent. The vapor
pressure of the pure solvent has been previously measured''and is
given byj
VogCP.mo) 6.07154 r^\ t .
- 0
t


GAMMA AND NEUTRON IRRADIATION
OF PURE FLUOROCARBONS
By
JAMES CLIFFORD MAILEN
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
August, 1964


IRRADIATION OF SAMPLES
Of the materials listed in Table 2, all except the perfluoro-
n-butane and perfluoro-2,5-dimethylbutane were irradiated in the
engineering cobalt-60 source, the Oak Ridge Graphite Reactor, and the
Oak Ridge Low Intensity Training Reactor (LITR). These two materials
were not available until after the reactor irradiation phase of the
experiments had been completed and received irradiation only in the
engineering cobalt-60 source.
The flux distribution in the engineering cobalt~60 source
(tube 11) was determined by irradiation of samples of water saturated
with benzene sealed in polyethylene bottles. The energy absorbed is
determined by analyzing for phenol by ultraviolet spectroscopy. A
further discussion of the method and a presentation of the results is
given in Appendix 1.
In the engineering cobalt-60, runs were made at both ambient
temperature (about 500) and at 99C in an electrically heated con
tainer. For details of the electrically heated container see Figure 1.
The reasons for operating at these two temperatures were two-fold.
First, this was necessary to see if there are any large effects of
temperature on yield and secondly, by raising the temperature above
the critical point of C^Fg it is possible to study the reactions of
C^Fg in a single phase.
The samples to be irradiated in the Oak Ridge Graphite Reactor
were placed in hole //4l and irradiated for 55 days. The thermal flux
15


l6l
The (T values for the radicals are found similarly and are
tabulated below.
Species
si.
5F8
6.42
F
2.9
CF,
4.65
c2f5
5-55
5F7
6.22
These values are used in the statistical model to calculate
the mean free paths of these radicals for several densities. The
results are given in Table 29.
Values of zero for Dn mean the average distance between mole
cules is such that the radical cannot pass between them. This may be
interpreted as a possible cage11 mechanism. At densities above
about 0.5 grams per cubic centimeter, ChFis "trapped," C^F^. is also
trapped fairly close to this density, CF^ is trapped slightly above
55 and- F is trapped slightly above 0.55* This means that at low
density all the radicals can escape to the wall, but above a density
of about 0.55 g./cm. only F should reach the wall in significant
quantities.
At low density g./cm.the ratios of the mean free paths
are:
Dm(DF;>
Dm(F)
^(02?g)
.89
.8 %


8
fluorine abstraction by CFj is not important. In the photolysis of
hexafluoroacetone, Ayscough and Steacie' also found no CF^ leading
to the same conclusion.
6. Addition of radicals across double bonds in unsaturated
moleoules. This type of reaction has been observed in hydrocarbons
by various authors. (^920) Mastrangolo^y has observed the addition
of CFj to ethylene and perfluoroethylene, the addition of CFg to per
il uoroethyl ene, and the addition of CFj to perfluoropropene.
7. The reaction of fluorocarbon radicals with molecular fluor
ine. This reaction will become more important in longer irradiations
where the fluorine concentration is not depleted by reaction with the
wall and at high density where diffusion of fluorine to the wall is
restricted. This reaction will lead to the formation of molecules of
size smaller than the parent as is shown- below;
R + ?2 RF + F.
Thus one may expect a small increase in the G values of these
smaller radicals with time. Larger molecules than the parent will not
result from this reaction.
In summary, the processes which are expected in the bulk phase
of irradiated fluorocarbons are; recombination of radicals, dispro
portionation of small radicals, addition of fluorine and radicals to
unsaturates, and the reaction of radicals with elementary fluorine.
Reactions with Wall
In addition to reactions in the bulk phase, additional reactions
can occur at the sample tube wall. These are particularly important


119
In using the infinite shell approximation a similar development
to the more complex case is required.
The distances required are defined in Figure 28c.
L2 = r^ + D2 2rD cos &
Flux at A = D
Flux at Center L
The average flux at A is given by
2 'if
(D)(D/2)d?
Rr
7?=oVr2 + D2 2rD cos &
71 D
D_
217
60
+/r2 + d2 2rD cos 0
If we let P r/D and simplify we obtain:
2# 7
/ 2 1
V 1 + P 2P cos 0
Where P = 0.5 with graphical integration we obtain:
Rp 1.068
In Figure JO the points corresponding to P = 0.5 for the shell
at the near edge of the rods, at the center of the rods, and at the
far edge of the rods are plotted.


125
TABLE
20. NUMBER
OF ELECTRONS
PER GRAM
Material
Mi
i
NA X 10 25
Air
l4.4l
7.21
5.01
AL
26.97
15
2.90
cf4
88
42
2.87
c2f6
158
66
2.88
c5f8
188
90
2.88
C4F10
258
114
2.88
c5f12
288
158
2.88
c5fio
250
120
2.89
0Pl4
558
162
2.88
Cu
65.54
29
2.75


67
Density,, gm./cm.3
Figure 21. G(2-CF-jC^F-q) versus density for 99C C^Fg
samples (cobalt-60)^.
CD
(a) Numbers beside points are ergs absorbed per gram X 10
(b)' This is: the G value for liquid C^Fg found by extrapolation
of.the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average G value for the samples.


ACKNOWLEDGEMENT
This work was done under contract to the United States Atomic
Energy Commission. The author also received financial support in the
form of a fellowship from the National Science Foundation. The
financial support of these two agencies is greatly appreciated.
The author is indebted to many people for assistance in the
completion of these studies but especially to Dr. T. M. Reed, III, for
his beneficial direction of the project, Mr. Shashi Dat for his
assistance in the early stages of the work and Dr. J. H. Simons for
many helpful suggestions in the course of the research. Thanks are
due Dr. Merril Y/ilcox for obtaining the infrared absorption spectra
of the solid fluorocarbon. The other members of the Ph.D. committee,
Dr. Mack Tyner, Dr. R. J. Hanrahan, Dr. T. F. Parkinson, and Dr. H. A.
Meyer (deceased) have also lent their kind support.
ii


APPENDIX 1. ENERGY ABSORPTION
Flux Determination
The variation of the flux over the length of the sample tube
was determined using benzene in water solutions as described by-
Johnson and Martin. In brief, this method requires the irradia
tion of water saturated with benzene. Subsequently the solution is
divided and one portion diluted with distilled water and the other
with approximately 0.08 N sodium hydroxide solution. The amount of
phenol formed is determined from the ultraviolet optical densities
at 2900l of the neutral and alkaline solutions. This work vas done
with a Beckman DK-2 Spectrophotometer. The yield of phenol is
2.14 + 0.02 molecules per 100 electron volts. This yield is insen
sitive to dose up to about 70,000 to 80,000 rep. Over the range
100-7000 rep./minute it is independent of dose rate, and above 0.007 M
it is insensitive to benzene concentration. For this reason it pro
vides a simple method of measuring gamma flux.
In this work solutions were sealed in polyethylene bottles
3.20 cm. high and about 2.5 cm. in diameter. There were empty spaces
0.34 cm. at the bottom and 0.22 cm. at the top. The average thickness
of the side of the container was 0.293 cm. Ten of these bottles were
stacked and retained in the center of the sample area. They were
enclosed in a stainless steel irradiation basket with a wall thick
ness of approximately 0.11 cm. When the correction for absorption
107


4l
TABLE 7 (continued)
Tube No.
82
Original Material
CF4
c2F6
Compound in assay or standard
retention Volume V^, co. H2
Mole Fraction^
Moles lost/mole mixture
OF4
C2F6
n2?2
5F8
Cy^c5F6
94.5 on Cl6H34
117 on Cl6K54
n-C5F12
i-05P12
165 on C-¡/cH*4
195 on O^hS
250 on C^H54
260 on G15K24
a-6pl4
2-CF5C5F11
5-CF, ^
2,5-(ci
Total G7F15
Total C8F18
Total CcjFpO
Total C]_qF22
^5Fn
5/ 2c4f8
Total
Total
Total
C11F24
12f26
15F28
lo re
recovered as
Wt. percent of sampl
gas
Sample wt., gms.
Molecular wt., Gas Density (of gas)
2 G gaseous products
0.0208
0.9905
O.OO55
0.0054
0.0004
0.0001
nil
it
u
11
11
11
11
n
11
it
11
is
n
n
n
11
n
11
it
11
100.
0.178
87.8
O.589
0.020
0.0800
0.8491
0.0012
0.0452
0.0007
O.OI56
nil
u
0.0017
0.0051
nil
11
11
11
0.0005
0.0001
0.0002
0.0020
0.005 (2)
0.0011 (2)
0.0004 (2)
0.0005 (4)
0.0001 (1)
nil
100.
O.5659
141.5
1.58
t*y
Number in brackets is number of peaks in groups.


51
TABLE 5. GASEOUS PRODUCTS FROM THE THERMAL DECOMPOSITION
OF THE LITR SOLID FLUOROCARBON
Initial Wt., gms. = 0.2089
Wt. Gaseous Products, gms. = 0.1688
Remainder (0.0401 gms.) is carbon with small amount
of soluble solid
Decomposition Temperature = 470C
" Time = 18 hrs.
" Tube material = Vicor
" Tube dimensions;
Length =11.2 cm.
I.D. = '0.675 cm.
Volume = 4.01 cm.5
Compound in assay
Mole fraction
CF4
2f6
C2F4
c2f2
C.Fq
cy-c5F
n-C5F6
4F10
25.0 min. on silica gel
26.2 min. on silica gel
n-C5Fi2
_
198, 203 cc
- 269
V
cc
Ho(n-hex)
. H2(nhex)
20F^n
11
^"7 5 5
2,5-(cF^
Total
Total
Total
Total
,c4f8
c7*l6
C9F20
Cif22
Total C^3_F24
Molecular wt., gas density
0.0006
0.0025
0.7723
0.0019
0.0154
0.0006
0.0012
0.0517
0.0025
0.0012
0.0027
0.0151
0.0811
0.0151
0.0081
0.0043
0.0014
0.0005 ( \
0.0131 (3)u;
0.0039 (3)
0.0019 (2)
0.0017 (2)
0.0011 (2)
95-4
(a) Number of peaks in group


APPENDIX 2. SAMPLE CLEANUP AND CANNING
The major cleanup of all materials is accomplished by standard
gas chromatographic methods. After this treatment the only major
impurities present are air, carbon dioxide, and water. The sample
is distilled through calcium sulfate to remove most of the water,
caroxite to remove the and finally through magnesium perchlorate
to remove the last traces of water. Air is removed by alternately
thawing, freezing, and pumping on the sample until no residual un
condensable gas is observed. The apparatus for this treatment is
shown in Figure J2.
The first time a sample is to be canned the desired amount is
condensed in the calibrated tube and allowed to thaw to observe to
what manometer pressure this corresponds. Thereafter the amount of
sample can be determined by the pressure reading. The aluminum sample
tube, which has been previously baked out under vacuum using a cool
torch flame, is connected to the system by means of Flex fittings and
a copper tube force fitted into a section of rubber hose. The sample
is condensed into the aluminum sample tube with liquid nitrogen and
the tube pinched shut near the top after evacuation of any residual
sample. The tube is then heliarced at the pinch. After sealing the
tube is removed and weighed. In most cases they were then heated to
100C for several days to check for leaks.
155


121
TABLE 19. GAMMA ABSORPTION CROSS-SECTIONS PER ELECTRON
Atom
A
/*
hl
H
1
.0927
.0927
Be
4
.572
.0950
C
6
557
.0928
N
7
.650
.0928
0
8
.745
.0952
Na
11
1.021
.0929
Mg
12
1.111
.0927
A1
15
1.210
.0952


159
TABLE 28. COMPARISON OF MEAN.FREE PATHS CALCULATED FROM GAS
KINETICSAND FROM THE STATISTICAL MODEL
Molecule
o
T K
Gas
Kinetics
Statistical
Percent Error
Based on Gas
Kinetics
He 2.18
100
645
654
+1.595
500
1956
I960
+2.27
500
5226
5010
-O.496
Ar
100
251
229
-O.865
500
694
702
+1.152
500
1157
1185
2.25
N2 5.75
100
218
215
-1.577
500
664
702
+5.82
500
1090
1112
-2.02
Average
error =
I.96 percent
O
_ 0
T K
d, A
100
26.7
500
5 8.6
500
45.8
(a) Gas kinetics mean free paths are taken from reference 4l.


56
is to be expected since they can be formed by reaction of radicals
with fluorine (equations C-20, C-21). Molecules larger than C^Fg cannot
bo formed in this way and no corresponding increase of G value is noted
(Figures 15-26). Correspondingly, the plots of G values for fluorine
lost to the wall show a decrease with increasing density (Figures 11,
12). No effect of energy absorption is seen indicating that the equilib
rium concentration of fluorine had not been built up over the time
period since such an effect vas seen for CF^ and 0^^. Thus for liquid
fluorocarbons where the density is about 2 grams per cubic centimeter
diffusion of fluorine to the wall should be loss important, especially
with a less favorable geometry. This is confirmed by the work of
J. H. Simons^ with CgF-¡^ irradiated in the Oak Ridge Graphite Reactor
where no fluorine attack on the aluminum container was detected.
The capture of radicals at the wall is indicated by the very low
density sample (number 92) irradiated at ydG. In this sample the
overall G value vas only about 0.8 as compared with about 4.8 for
higher densities. From this we conclude that at low density the
radicals produced can be scavenged by the aluminum wall. Another
experiment was run with the sample tube partially filled with powdered
aluminum (number 85). The overall G value for this tube was not signif
icantly different from the G values of tubes not containing aluminum
powder. However, the product distribution vas considerably different
with smaller amounts of CF^, C^F^, and C^.F-^q being formed indicating a
decrease in the species Fg, F, CF^, and 0^^, presumable by capture by
the solid surface. Larger molecules show no such decrease. The larg
est individual peak in this sample has not been identified, but it is
not a fluorocarbon and must contain oxygen derived from the aluminum


145
TABLE 26. STANDARD RETENTION VOLUMES ON THE N-HEXADECANE COLUMN
Pi *= 21.7 psi. Po = 14.7 psi. Temp, of Column 25 C
Carrier Gas = 22 co./min. at 1 atm. and 25 C
Column Length = 16 meters containing 197.1 grams of 0.428 grams
n-hexadecane/gm. Chromosorb-P.
Compound
Vo cc. H0 at 1 atm.
K c.
C2F6
8.77
c5F8
29.8
n-C4Fio
87.3
^lO
74.5
n-C5F12
135
i-C5Fi2
168
n-6Fl4
284
2-CF505Fh
325
5-OF55Fll
351
2,5-(cf5)2c4f8
412
n-7Pl6
554
2-CF500Fii'
573
Probable
5-CF5C6Fn
626
Identifications
J-VsVuj
693
-jP6
103.5
cy-C^Fpo
218
co2
52.5


45
TABLE 7 (continued)
Tube No.
Original Material
79
n-C6Fi4
70W
Z-CFjCjFn
Compound in assay or standard
retention Volume VR, cc. H2
Mole
Fraction^
Moles Fr
lost/mole mixture
cf4
2F6
2F2
c5f8
Cy-C5F6
94.5 on
117 on C^H*4
n"5P12
i5F12
165 oil l5H34
195 on 015H24
250 on Cl6H34
260 on C1H££
2>"G6Fl4
,Fi
2-CF5C5ill
^CF5C5Fn
2,5-(C?5}2C4F8
Total d7Fl6
Total CnF,g
Total CoFq
1^ ufe
Total C-qF24
Total C,2F2
Total Ci5F28
Wt. percent of sample recovered as gas
Sample wt., gms.
Molecular wt., Gas Density (of gas)
2 G, gaseous products
0.0458
0.0554
0.0006
0.0501
0.0004
0.0511
0.0005
0.0002
0.0268
. 0.0051
nil
n
o.ooo4
nil
0.7079
0.0058
o.ooAo
0.0009
0.0542 (2)
0.0247 (5)
0.0220 (2)
0.0151 (4)
0.0105 (5)
0.0052 (5)
nil
80.2
0.5595
551.
1.05 .
0.1087
O.O566
0.0008
0.0502
0.0004
0.0206
nil
0.0006
0.0120
0.0087
nil
0.0005
nil
0.0005
0.0008
0.5978
Not determined
0.0041
O.G456
(2)
0.0258
(2)
0.0514
(2)
0.0554
(A)
0.0182
W
0.0110
(5)
0.0050
(1)
68.
0.5195
520.
1.47
(a) Number in brackets is number of peaks in groups.
(b) This sample contained unsaturate.


45
TABLE 8.
MASS SPECTROGRAPH DATA FOR
Species
Relative Intensity at
50v
70v
lOOv
0
6.80
10.21
11.12
F
5.09
5.95
8.14
CF
5*67
4.91
6.52
of2
].4.o4
21.50
26.77
CF5
101.16
125.55
158.58
Ratio to QFj
C
0.0672
0.0828
0.0805
F
O.0505
0.0465
0.0587
CF
0.0565
0.0598
0.0456
cf2
O.1588
0.174
0.195


Ill
In Figure 28a are shown the various distance and angles
required for calculating the relative flux for a slice of element.
We wish to calculate the flux received at A from B and integrate
for all values of B over the element.
The value of D, the distance between points A and B can be
found by vector addition.
~R Ri
~~q q cos

r = r cos G i + r sin 3
R + q + D- r =
D -~T R ~q =
0
i(r cos & R q cos ty) + j(r sin 9 q sin 9)
The distance D is then found to be
|d|-V7T cos R q cos + (r sin ~0 q sin
-V777T7T 2 q(cos 9 cos

+ 2R(q cos f> r cos 9)
Since there are six cobalt-60 rods in the array in integrating
over subsequent rods Q is increased by JO0.
Since this method can only give relative values we will now
compute the ratio of flux contributions at point A from point B as
compared to that at the center of the radiation zone.
The distance from point B to the center of the radiation zone
is obtained by vector additions


150
TABLE 25. ENERGY ABSORPTION BY BRAGG-GRAY CAVITY THEORY
Molecule
ergs/gm. roent.
cf4
88.7
C2F6
89.0
5f8
89.0
G4F10
89.2
c5f12
89.5
5fio
89.4
6Fl4
89.5
C^Fq in Cu
99.8


APPENDIX 2.
SAMPLE CLEANUP AND CANNING
Page
APPENDIX 5. ANALYTICAL ...
Physical Description of Analytical Equipment . .
Calibration of Molecular Weight System
Sample Treatment
Switching Columns from Series to Parallel
Description of Analytical Columns .....
Urea and Thiourea Columns
Mole-Area Calibration
Fluorine Balance
135
156
156
137
139
140
142
142
146
146
APPENDIX 4. DECANNING OF PILE IRRADIATED SAMPLES 151
APPENDIX 5- RADICAL TRANSPORT 153
Diffusion Coefficient of Radicals in Perfluoropropane 155
Statistical Model for Calculating Mean Free Path . 156
Mean Free Paths of Radicals in Perfluoropropane ... 158
APPENDIX 6. STATISTICAL RECOMBINATION DATA CORRELATION . I65
CHEMICAL NOMENCLATURE 182
ABBREVIATIONS AND DEFINITIONS 18J
BIBLIOGRAPHY 184


TABLE 18 PART 4 (continued)
Tube No.
44
4l
92(c
Sample Wt., Gms.
0.5281
0.5419
0.0295
Bulk Density, gm./cm.^
0.633
0.649
0.0353
No. Days
23.0
10.0
9.92
Integrated Dose, 10 Roent.
Energy Absorbed, 10 g ergs
Energy Abs./gm., 10 ergs
9.02
3.83
1.705
42.3
18.5
0.446
80.2
34.2
-G Value'
Compound in Assay or as Listed
Mole Fraction-
Moles Fp/mole mixtureG Value
0.0160-0.98
0.0120-1.74
^ zero
CFj,
0.0148-0.91
0.0043-0.62
0.0013-0.
C2F6
C2F4
C2F2
C3F8
n-C3F6
C4F10
D"n5>
1-C5F12
n_c6Fl4
2-CF C Fn
3-CF3C5Fll
2,3-(CF ) C^Fg
Total C7Fi8
Total C8F7q
Tota! C F
Total C£0F22
Total C,,F2,
Total C12F26
Tota^ Ci3F28
Molecular Wt., Gas Density
0.0105-0.64
nil
0.0001-0.007
0.9395
nil
0.0139-0.
0.0032-0.
0.0053-0.
0.0014-0.
0.0028-0.
0.0004-0.
0.0024-0.
0.0016-0.
0.0013-0.
0.00093-0
0.0008-0.
0.0003-0.
0.0002-0.
4.677
193.
0.0043-0.62
nil
nil
0.9729
nil
86 0.0071-1.03
20 0.0017-0.25
33 0.0027-0.39
08 0.0008-0.11
17 0.0018-0.26
03 nil
15 0.0011-0.16
10(2) 0.0010-0.15(2)
08(2) 0.0009-0.13(2)
.06(2) 0.0004-0.06(2)
05(6) 0.0004-0.06(4)
02(2) 0.0001-0.02(1)
01(1) 0.0007-0.10(2)
5.70
191.4
0.0006-0.20
nil
nil
0.9976
0.0005-0.16
trace
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
0.78
188
(a)These three samples were analyzed before the addition of the high temperature squalane column and
do not have complete analyses of the high molecular weight compounds, (b) This tube was copper,
(c) This sample was a single phase. (d) Number in brackets is number of peaks in group.


124
TABLE 21.
RADS ABSORBED PER ROENTGEN NEGLECTING
WALL AND DENSITY EFFECTS
Material
Rads/Roentgen
Aluminum
0.8p2
Copper
0.78 9
cf4
0.824
g2f6
0.826
c5f8
0.826
4f10
0.828
5f12
0.82 6
c6fi4
0.82 6
c5fio
0.828


The G values for disappearance of the parent compound are
approximately; CF4, 1.5; C2Fg, 5.25; CjFg, 4.5; n-C^F.^, 5*4;
n-C5Fi2, 4.9; cyclo-C^F10, 5.0; n-CgF^, 6.0; 2-CF^F^, 6.0;
and 2,5-(CF^)^C^Fg, 9*6. These values are only approximate due to
the dependence of certain reactions on density and amount of energy
absorbed.
The important products in perfluoromethane are perfluoroethane
and perfluoroacetylene produced in about equal amounts. For the
other materials for short irradiations the major products are saturated
and are those predicted by simple bond rupture and radical recombina
tion reactions. The results for perfluoroethane and larger molecules,
excluding cyclo-compounds, can be correlated by statistical recombina
tion calculations using empirically determined radical effectiveness
numbers. Since these effectiveness numbers are the same for radicals
formed in the same way this method can be used to predict product
distributions for other fluorocarbons. In conjunction with prediction
of overall G values from mass spectrometer sensitivities the G values
for the products can be estimated.
xiii


Stopping power relative to air
128
Atomic number
Figure 31o Electron stopping power relative to air versus
atomic number


50
over a range of energy absorptions and densities and at two tempera
tures in the engineering cobalt-60 source. At the lower temperature,
about 50C, the samples were partially liquid and these allow examina
tion of the mechanism in the liquid phase. At the upper temperature,
about 100C, the samples were in a single phase allowing examination of
the effects of density. Samples were also irradiated in the Oak Ridge
Graphite Reactor and the LITR reactor.
In the LITR, where an equilibrium distribution was obtained,
we see that at equilibrium about 40 weight percent of the sample is
porfluoromethane. (See Table 5). Thus CF^ will show a tendency to
higher concentrations for higher energy absorptions.
The proposed mechanism for perfluoropropane is given in Table
11. Daring the irradiation of perfluoropropane fluorine is produced
by reaction 0-4 and may either react with the wall by reaction C-25,
with radicals as in reactions C-20, C-21, and 0-22, or with unsaturated
materials as in 0-19* Thus the wall competes with the fluorocarbon
system for the fluorine. Reaction at the wall will be expected to
decline over a period of time due to the formation of a protective
fluoride coating. Reaction of fluorine at the wall will also decline
with increasing density since diffusion of fluorine to the wall will
be retarded. These two effects are seen in Figures 7 and 9 for the
products CF^ and C£F£ in the single phase cobalt-60 irradiations, and
in Figure 8 for CF4. in the partially liquid cobalt-60 irradiation.
Figure 10 for the product *n Par'tially liquid runs does not
show such a striking trend although some difference may be present for
low density. In general, these plots show an increase in the G values
for CFj. and CgF^ with increasing density and energy absorption. This


12
propane was selected for extensive study because it could be studied
in both the liquid and gaseous states with ease (critical temperature
7190) In addition, it was the largest molecule available in
quantity and was expected to give a greater range of reaction com
plexity.


131
effect is not unexpected since the addition of fluorine atoms to
a carbon atom tends to strengthen other C-F bonds already present.^J
2. OF-* is more effective than other fluorocarbon radicals
P
formed by carbon-carbon bond splitting. This must result from ex
tensive rupture of molecules to give more than one single-carbon
species.
5. Since the other radicals formed by carbon-carbon bond
breakage did not differ greatly in their effectiveness values the
stearic factors for recombination must be small.
Using this statistical recombination method and the empirically
determined effectiveness numbers in conjunction with the sensitivities
in the mass spectrometer an estimate can be made of the products and
their G values for other saturated fluorocarbons. In this calculation
the ratios of the products are determined by the statistical recombina
tion method and the overall G from the mass spectrometer sensitivity.


MECHANISM
Free Radical Formation
Y/hen a fluorocarbon is bombarded by electrons a probable first
step is fracture of the molecule into ionic and non-ionic fragments.
The positively charged ionic fragments may then be converted to free
radicals as is shown below.
M + e* R£ + R+ + 3e
R^ + e > Rj.
R2 + e > R2
The formation of positively charged ions is known to be impor-
/pr £\
taut from mass spectrometer work.^ The neutralization scheme is
(7)
supported by the data of Mastrangelo' who passed perfluorocyclo
butane and perfluoroethane through an electric discharge and froze out
the excited species on a liquid nitrogen finger. He failed to detect
any space charge at the finger, indicating that the ionic species are
quickly neutralized by electron capture or some other equivalent
mechanism.
Reactions and Mass Spectrograph Sensitivities
Y/hen irradiating pure saturated fluorocarbons a number of
deductions about the expected reactions can be made.
5


are:
At a density of 0.17 ga./cm. the ratios of the mean free paths
.614
BW .761
qn(0?F7)
Dm(C2F5)
.795
Prom these numbers we see there is little difference in the
mean free paths of the molecules CF^, C^F,-, and C-F-,. From this we
conclude that a difference in products due to radicals reaching the
wall is not to be expected except in reactions where the total amount
of radicals is important end not their relative concentrations or where
F or Fg is one of the radicals. An example of this is the reaction of
radicals with unsaturated molecules. Apparent G values for the total
reactions would be expected to be lower at low density since a portion
of the radicals produced may be removed at the wall.
In conclusion, whether diffusion or mean free path mechanisms
determine the removal of radicals at the wall the results are similar:
1. Fluorine radicals will reach the wall in significantly
greater amounts than other radicals


TABLE OF CONTENTS
Page
INTRODUCTION 1
MECHANISM 5
Free Radical Formation 5
Reactions and Mass Spectrograph Sensitivities ..... 5
Types of Reactions in Bulk Phase 6
Reactions with Wall 8
MATERIALS IRRADIATED 10
IRRADIATION OF SAMPLES 15
CANNING 17
ANALYSIS 18
RESULTS AND DISCUSSION 20
Irradiations in the Oak Ridge LITR Reactor 20
Solid Fluorocarbon from LITR Irradiations 25
Irradiations in the Oak Ridge Graphite Reactor .... 50
Irradiations in Cobalt-60 and Detailed Discussion ... 52
Perfluoromethane 52
Perfluoro ethane h~¡
Perfluoropropane ..... 4-9
Perfluoro-n-Butane 7^
Perfluoro-n-Pentane 75
Perfluorocyclopentane 77
Perfluoro-n-Hexane 80
Perfluoro-2-Methyl pentane 82
Perfluoro-2,5-Dimethylbutane 85
SUMMARY AND FUTURE WORK .......... 105
Appendices 106
APPENDIX 1. ENERGY ABSORPTION 107
Flux Determination 107
Flux Exposure of Samples 108
Variation of Flux Across Sample Zone 110
Absorbed Dose 120
Energy Absorption in Samples ..... 127
Energy Absorbed in Reactor Irradiated Samples 152
iii


50
From the above data Vie should not expect the solid fluorocarbon
to have exceptional thermal stability. Aa has been mentioned before
the thermal stability of fluorocarbons is due to the shielding effect
of the fluorine atoms. When one introduces double bonds and strained
bonds (due to the highly branched struoture) this stability is reduced.
Thermal decomposition products
It was thought that some additional information concerning the
solid would be determined by thermally decomposing some of it. A sample
weighing 0.2089 grams was sealed in a Vicor tube with a volume of four
cubic centimeters and heated to 470C for a period of about 18 hours.
At the end of this time, the gaseous portion of the sample was removed
for analysis on the gas chromatograph. The material appeared to de
compose almost completely to the gaseous products and carbon. The
gaseous products accounted for 0.1688 grams or about 81 wt. percent of
the sample.
In Table 5 the mole fractions of the gaseous products are tabu
lated. Of especial interest is the fact that, like teflon, this
material decomposes largely into perfluoroethylene.
Irradiations in the Oak Ridge Graphite Reactor
The irradiations in the Oak Ridge Graphite Reactor represent
energy absorptions intermediate between those obtainable in the engi
neering cobal-b-60 source and the energy absorbed in the LITR. In the
latter, as has been pointed out, an apparent equilibrium condition was
reached.
The energy absorbed by the samples in the Oak Ridge Graphite
Reactor can be estimated rrom the data of Richardson, Allen, and
Boylefor energy deposition in graphite with an appropriate flux


55
TABLE 6. LITR IRRADIATIONSANALYSIS OF GASEOUS PRODUCTS
Position => C-4l Flux = 10^ n/cm.^-sec.
No. days 28 Irradiation Temperature = 100C
Tube No.
15
16
Compound
2-CFjC^F-^^
2-CF5C5Fn
Compound in assay or standard
Mole
Fraction^
retention volume V^, cc. H2
vRo
cf4
C2F6
G2F4
2?2
CjFg
cy-CzF8
C4F1Q(mostly iso)
?1 cc. H2(n-hex)
100 cc. H2(n-hex)
121 cc. Hp(n-hex)
VR
'R
n-C5F12
-C5F12
224 cc. H2(n-hex)
n"G6Fl4
** 508 cc. Hp(n-hex)
2-CF,C5F11
5-CF3C5F11
2,5-(o^2C4F8
Total
Total
Total
Total
Total
Total
Wi
8F18
9Fg0
10f22
c11f24
Cl oFo6
ivotu. ^12r26
Sample wt., Gms. (Initial)
0.8546
0.1045
nil
0.0062
0.0289
0.0010
0.0159
nil
nil
nil
0.0008
0.0065
nil
trace
nil
2 + 5 =
0.0008
O.OOI9-
0.0005 (2)
0.0005 (2)
0.0001 (1)
nil
n
ti
0.1947
Molecular wt., Gas Density (of gas) 108.5
Weight percent recovery as gas 57.7
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis 100.5
(a] Number in brackets is number of peaks in group.
0.7595
0.1426
nil
0.0026
0.0408
0.0005
0.0111
0.0005
0.0019
0.0004
0.0024
0.0217
0.0004
0.0007
0.0009
0.0015
0.0015
0.0084
0.0047
(2)
0.0060
(2)
0.0048
(4)
0.0060
(5)
0.0012
(4)
0.0009
(1)
0.619
117.5
55.8
117.8
Notej In second samples, more care taken to recover higher boilers
hence the high m.w. analyses are more nearly correct.


122
The numbers in the fourth column, the cross-section per electron,
ore seen to be essentially constant for all atoms over the range of
interest. For this reason the last term can be deleted from equation
K-l and we have:
Dabs 0-875 (K-2)
N air Jk
Tire number of electrons per gram can be calculated from the
following equation
N. ** ^av ^ ^ (K5)
Mi
where
25
Nav Avogadro's number = 6.02 x 10 ^
= molecular or atomic weight of ith species
= number of electrons per atom or molecule
These values are listed in Table 20.
From these values of number of electrons per gram and equation
K32, we may now calculate the energy absorbed in the materials of
interest neglecting the wall effect and any density effect (Table
21). These two considerations will be examined later.
When a sample is contained in a relatively small cavity a
/
significant portion of the energy absorbed in it can originate as
energetic electrons in the wall material. Let us examine what occurs
in a sample being irradiated.
When a gamma ray is absorbed in either the tube wall or the
sample secondary electrons are set free which actually cause the
radiolytic chemical changes. Some of the electrons originating in


Flux relative to center
118
Inches from center
Figure 30. Relative flux across sample zone by infinite wire and
infinite shell models.
(a) Infinite shell at inner edge of rods.
(b) Inf initef'shell at center of rods.
(c) Infinite shell at outer edge of rods.


115
In Equation J-l we now malee the substitutions:
2 2
N for M
2 2
P for D
and integrate over the height of the element.
q r2^rH
['R2 + q^ + 2Rq cos + (Z h)^7'lqd^ dZ
When this equation is summed over the six cobalt-60 elements
the result is the relative flux at the point specified by r, 6, and
h to the center of the irradiation space.
Unfortunately this equation is too complex for ordinary solution
and would require use of a computer. The equation can be integrated
once by an appropriate substitution of variables but this form is still
too cumbersome.
In order to obtain an estimate of the percent variation across
the irradiation zone without recourse to a computer it is necessary
to study simpler cases. Two of these will be considered: the infinite
shell and the infinite wire. For the case of the infinite wire the
flux can be shown to be inversely proportional to the distance from
the wire. A graphical method has been used to obtain the relative
flux across the radiation zone for infinite wires at the center of the
cobalt-60 rods and for infinite wires at the projection of the rods
defined by tangents to the central tube and the cobalt-60 rods as is
shown in Figure 29. The results of these infinite wire calculations
are shown in Figure JO.


Ip4
Using the Le Bas numbers we find the following molar volumes.
Vb(CjFg) *=> ll4 cm.^/mole
Vb(CF^) = 40.9 cm.Vnole
Vb(C2F5) = 75-1 cm.'Vmole
Vb(C7,Fy)=> 105.5 cm. 2/mole
Vb(F) 8.7 cm.Vaiole
Substituting these values in equation (L-l) yields'the following
diffusion coefficients for the various radicals in C^Fg.
D(F)
5/2
2.17 x 10~v cm./sec.
D(CF5) 8.79 x 10'
-6 T^/2 __ 2
cm
. /sec.
A rp5/ 2 n
D(C2F5) o 6.18 x 10" -4_ cm. /sec.
, ^5/2
D(C5F7) 4.98 x 106 cm.2/sec.
Using these values we find the following ratios of diffusivities
at a given temperature and pressure
D(CF-)
- 0.405
D(F)
d(c2f5)
0.705
d(cf5)
d(c5f?)
d(02f5)
0.805


185
(17) J. R. McNesby, J. Phys. Chora. 4, 1671 (i960).
(18) C. D. Bass and G. C. Pimental, J. -Am. Chem. Soc. 82, 5754 (191).
(19) J. Said and M. Szware, J. Am. Chem. Soc. J2, 1554 (1957).
(20) J. K. Futrell, J. Phys. Chem. 65, 5^5 (1961).
(21) T. N. Bell, B. J. Pullman, and B. 0. V/est. Australian Jnl.
Chem. 16, 722 (1965).
(22) J, K. Simons, Personal communication.
(2p) R. D. Dresdner, T. M. Reed III, T. E. Taylor, and J. A. Young,
J. Org. Chem. 2£, 1464 (i960).
(24) J. A. Young, Personal communication.
(2p) T. M. Reed III, J. F. V/alter, R. R. Cecil, and R. D. Dresdner,
Ind. Eng. Chem. 1, 271 (1959).
(26) D. M. Richardson, A. 0. Allen, and J. W. Boyle, USAEC unclassified
report, ORNL-129 (Dec., 1948).
(27) T. M. Reed III and T. E. Taylor, J. Phys. Chem. 62, 58 (1959).
(28) T. R. Johnson and J. J. Martin, Nucleonics 20, 85 (1962).
(29) L. H. Gray, Br. Jnl. of Radiology 22, 555 (1958).
(50) J. W. Boag, pp. 100-115 of Quantities, Units, and Measuring
Methods of Ionizing Radiation, ed. by Franco Fossati, Ulrico
Hoepli, Publisher, Milan (1958).
(51) L. G. Alexander, USAEC unclassified report, OKNL 58-12-9.
(52) G. Friedlander and J. H. Kennedy, Nuclear and Radiochemistry,
John Wiley & Sons, Inc., New York (1958) P* 202.
(55) Report of the International Commission on Radiological Units
and Measurements (ICRU) 1958; NBS Handbook 62, 10-17 (1958).
(54) L. V. S. Spencer, NBS Monograph 1 (Sept. 10, 1959).


.ANALYSIS
All the samples except numbers 55 7 51 and 29 were analyzed
on a three-column gas chromatograph consisting of 16 meters of n-
hexadecane at room temperature, 15.5 meters of squalane at 92C, and
one meter of silica gel. The silica gel column was temperature
programmed. For details on the columns see Appendix 5 The samples
mentioned above were analyzed without the high temperature squalane
column. The silica gel column was used to separate air, perfluoro-
methane, perfluoroethane, perfluoroethylene, carbon dioxide (perfluoro-
ethylene and carbon dioxide appear together), perfluoroacetylene,
perfluoropropane, perfluorocyclopropane, perfluoro-n-propene, and
perfluorobutane (no separation of perfluorobutane Isomers). The n-
hexadecane column was used to separate perfluoropropane, perfluoro
butane (partial separation of isomers), perfluoro-n-propene, the
perfluoropentane isomers, the perfluorohexane isomers, and the per-
fluoroheptane isomers (most unidentified). The squalane column
separated the perfluorohexane isomers with the exception of the 2-
and 5-methylpentane isomers and was used to determine the total amounts
of material in each carbon number range up through occasionally,
where amounts were sufficient for detection, up through C^.
Before being introduced into the chromatographic train, the
samples were decanned directly into an adjacent vacuum system and
stored in bulbs of sufficient size to assure complete vaporization.
After storage for at least 24 hours to assure mixing of the gases,
18


170
TABLE 52. RESULTS OF RANDOM RECOMBINATION CALCULATION FOR n-C4F
Fragment
N
E
N x E
F
10
1.0
10.0
CF-
2
1.0
2.0
c2f5
2
.45
.90
n-05F7
2
.45
.90
n-C4F9
6
.54
2.04
2-4F9
4
1.065
4.26
Molecule
Relative Amount
F2
100.
cf4
40.0
C2F6
22.0
C5F8
21.6
n-4Fio
150.4
n-C5P12
9.8
2-CF5C4F9
17.0
n-6F14
' 4.5
5-GF55Pu
7.7
Total Cy
11.5
Total Cg
4o4.0
percent recombination 52.5


BIOGRAPHICAL SKETCH
James Clifford Mailen was born August 19, 1957, at Colorado
Springs, Colorado. In June, 1955, lie was graduated from Hi chita
Jest High School in Wichita, Kansas. In June, 1959, he received the
degree of Bachelor of Science from Kansas State University. In 1959
he enrolled in the Graduate School of the University of Florida. Until
June, I960, he was supported by a Graduate School fellowship and from
then until June, 1965, by NSF Cooperative fellowships.
James Clifford Mailen is married to the former Jean Hester
Bolick and has one stepchild. He is a member of the American Institute
of Chemical Engineers, American Chemical Society, Phi ¡Cappa Phi, and
Sigma Tau.
187


54
vG
*A
CM
O
>'
O
Density, g./cm.^
Figure 9 GCCgF^) versus density for 99C CgFg sampler
(cobalt-60).
ft
() Numbers beside points are ergs absorbed per gram X 10".
(b) This is the G value for liquid C3F3 found by extrapolation
of the plot for partially liquid samples to the liquid
density.


G(CF,,)
52
Figure 7. G(CF^) versus density for 99C C^Fg samples^3).
(a) Numbers beside points are ergs absorbed per gram X 10^.


19
they were admitted to the volume calibrated vacuum system where their
molecular weight was determined and then into the chromatographic train
by gas sampling valves.
The chromatographs were supplied with thermal conductivity de
tectors and the output was recorded on Honeywell Electronik recorders.
The areas under the curves were determined by use of a polar planim-
eter. Area was found experimentally to be proportional to number of
moles from C-^ to and was assumed proportional over the entire range
(Appendix 3).
The amount of fluorine lost to the wadi was calculated from the
fluorine to carbon ratio of the products (Appendix 3).


g(f2)
57
Density, g./cm.^
Figure 11* G(F2) lost versus density for 99C C^Fg samples
(cobalt-60)(a).
(a) Numbers beside points are ergs absorbed per gram X 10"*


126
of those requirements is exactly fulfilled under the irradiation con
ditions. The thickness of the aluminum is not sufficient to stop
electrons having an energy of 0.5 mev. or greater^2) and traveling
perpendicular to the wall. And except in the samples containing no
liquid there will be a significant amount of electron absorption in
the sample.
From the above discussion Vie see no clear-cut indication as to
whether to include wall effects. However, the authors cited state
that deviations from the conditions for rigorous application are often
acceptable. For this reason we will use the Bragg-Gray values realiz
ing this may introduce an error of at most 8 percent in the G values
calculated.
The Bragg-Gray cavity equation, for this case, may be written:
(50)
ef eas
(K-4)
where:
E^ = energy deposited in the aluminum
E <= energy deposited in the fluorocarbon
F
S = average electron stopping power ratio of the fluorocarbon
to aluminum
Electron stopping power ratios relative to air are tabulated in
reference 55 and. stopping powers are tabulated in reference 54. In
(55)
general these numbers are correlatable to atomic number.'" Although
the stopping power of a material varies considerably with the energy
of the electron, the ratios are nearly constant except for very low
(50)
energies. For this reason the average stopping power ratio will
be nearly equal to the stopping power ratio at the mean electron energy


A8
TABLE 10. CHEMICAL MECHANISM FOR CgFg
(6)
Important mass spectrograph species CF:
5*
C2F5*
CFj CF2
Radical Generation
C2F6 ^ 2CFj
-Mr C2F5 + F
vv CFS CF2 F
Radical Recombination
F + F *F2
CF^ + F >CF^
2CF5> C2F6
C2F5 + F C2F6
CF, CF2, F C2F6
CF^ + F2 > CF^ + F
C2F5 + F2> C2F + F
CF + CF > C9F2
CF2 + CF2-* CF + CF^
c2f5 + cf5 * c5f8
C2f5 + C2F5 n~Cij.F10
unsaturated molecules + F, F2, radicals > saturated
molecules
(B-l)
(B-2)
(B-5)
(B-4)
(b-5)
(B-6)
(B-7)
(B-8)
(B-9)
(B-10)
(B-ll)
(B-12)
(B-15)
(B-lA)
(B-15)


BIBLIOGRAPHY
(1) J. H. Simons and E. H. Taylor, J. Phys. Chem. 6j5, 656 (1959)
(2) R. E. Florin, L. A. Wall, and D. W. Brown, Jnl. Res. Nat. Bur.
Stds. 64A, 269 (I960).
(5) R. A. Seeger and J. G. Calvert, J. Am. Chem. Soc. J6, 5197 (19>4).
(4) P. B. Ayscough and E. W. R. Steacie, Proc. Roy. Soc. A2f,
476 (1956).
(5) 7. H. Dibeler and F. L. Mohler, J. Res. Nat. Bur. Stds. 40,
25 (1948).
(6) API Project 44 Mass Spectral Data.
(7) S. V. R. Mastrangelo, J. Am. Chem. Soc. 84, 1122 (1962).
(8) G. 0. Pritchard, Y. P. Hsia, and G. H. Miller, J. Am. Chem.
Soc. 85, 1565 (I9t>5).
(9) H. A. Dewhurst, J. Phys. Chem. 6^, 815 (1959).
(10) T. D. Nevitt and L. P. Remsberg, J. Phys. Chem. 64, 969 (i960).
(11) L. Pauling, The Nature of the Chemical Bond, Cornell University
Press, Ithaca, N. Y. (i960) p. 85.
(12) J. H. Simons, Fluorine Chemistry, Volume 5. Academic Press,
New York, N. Y. (to be published) p. 81.
(15) G. J. Mains and A. S. Newton, J. Phys. Chem. 6, 212 (l9l).
(14) G. R. Freeman, J. Phys. Chem. 64, 1576 (i960).
(15) D. Urch and R. Wolfgang, J. Am. Chem. Soc. 8, 2982 (1961).
(16) P. B. Ayscough, J. G. Polangi, and E. W. R. Steacie, Can. J.
Chem. 22, 745 (1955).
184


Table
18
19
20
21
22
25
24
25
26
27
28
29
50
PART 5. C7Fa IRRADIATED BY COBALT-60 GAMMA RAYS,
T = 990;> 89
PART 4. C,F, IRRADIATED BY COBALT-60 GAMMA RAYS,
T = 50CP ? 92
PART 5- C^Fq WITH POWDERED ALUMINUM IRRADIATED BY
COBALT-60 GAMMA. RAYS
PART 6. n-CI.F. n IRRADIATED BY COBALT-60 GAMMA RAYS,
T = 50C 7 ^
PART 7. n-C_F12 IRRADIATED BY COBALT-60 GAMMA RAYS .
PART 8. CYCLO-C^F, n IRRADIATED BY COBALT-60 GAMMA
EATS .
PART 9. IRRADIATED BY COBALT-60 GAMMA RAYS .
PART 10. 2-CF,05F11 AND 2,5-(CF5)2C4F3 IRRADIATED
BY COBALT-60'uAMMA RAYS
GAMMA ABSORPTION CROSS-SECTIONS PER ELECTRON ....
HUMBER OF ELECTRONS PER GRAM
97
98
99
100
101
102
121
125
RADS ABSORBED PER ROENTGEN NEGLECTING WALL AND
DENSITY EFFECTS 124
ELECTRON STOPPING POWER RATIOS RELATIVE TO AIR FROM
FIGURE 51 FOR MATERIALS OF INTEREST 129
ENERGY ABSORPTION BY BRAGG-GRAY CAVITY THEORY .... 150
TEMPERATURE PROGRAM AND APPEARANCE TIMES ON THE
SILICA GEL COLUMN l4j
STANDARD RETENTION VOLUMES ON THE SQUALANE COLUMN . l44
STANDARD RETENTION VOLUMES ON THE n-HEXADECANE
COLUMN 145
STANDARD RETENTION VOLUMES ON THE THIOUREA COLUMN . 147
COMPARISON OF MEAN FREE PATHS CALCULATED FROM GAS
KINETICS AND FROM THE STATISTICAL MODEL 159
MEAN FREE PATHS FOR RADICALS IN PERFLUOROPROPANE . lo2
RANDOM RECOMBINATION CALCULATION AND RESULTS FOR C2F 166
vi


2.0
1..8
1.6
lJi
1.2
1.0
0.8
0.6
0.1;
0.2
0
O 64-7
(7)26,6
'-Q.O
120.0
O
(7)36,2
37.7 O 36~
Q3U.2
Q 108.8
- o25-6
15 j.
o o G 59-2
315 0 80.2
087.8
O 36.7
o 76.1*
0 0.1 0.2 0.3 0.4 0,5 0,6 .7 o,
Bulk density, g./cm.3
Figure 12. G(Fg) lost versus bulk density for
samples (cobalt-60)
(a)
85.1
o
8 0.9 1.0
30C C3F8
8
(a) Numbers beside points are ergs absorbed per gram X 10'


TABLE 18 PART 8.
Tube No.
Sample Wt.} Gms.
No. Days
Integrated Dose,
Energy Absorbed,
Temperature, C
Compound in Assay or as Listed
cyclo-c5f10 IRRADIATED BY COBALT-60 GAMMA RAYS
10^ Roent.
10 ergs
TU
0.4916
9.83
3.52
15.5
99
21
0.2102
18.9
8.95
16.9
30
Mole FractionG Value
(c)
x(a)
0.6277
32.2
13.0
72.9
30
O.OOO9-0". 04
0.0012-0.06
nil
0.0003-0.01
0.0006-0.03
nil
0.0011-0.05
0.0006-0.03
0.0008-0.04
nil
0.0058-0.26
0.9386
nil
nil
nil
0.0023-0.10
nil
0.0009-0.04(1)
0.0053-0.24(2)
0.0045-0.20(2)
0.0017-0.07(2)
0.0138-0.62(5)
0.0155-0.71(4)
0.0018-0.08(3)
0.0025-0.11(4)
0.0008-0.04(1)
2.73
264.5
CF4
C2F6
c2f4
C2F2
C3F8
n-C3F6
c4f10
n-c5F12
i-CrFi2
VR = 167 cc. H2(n-hex)
VR = l88 cc. Hp(n-hex)
Cy-C5F10
VR = 281 cc. H2(n-hex)
Vr = 321 cc. Hp(n-hex)
VR = 351 cc. H2(n-hex)
VD = 403 cc. Ho(n-hex)
D w> \~\~4 '
VR = 456 cc. Hp(n-hex)
Total CyF^lb)
Total CgFjg
Total C^F2q
Total C^qF22
Total C11F2i|
Total C12F26
Total C-^Fgg
Total C^FgQ
Total C-, £-F^0
S.G i
Molecular Wt, Gas Density
0.0001-0.01
0.0001-0.01
nil
nil
0.0001-0.01
0.0001-0.01
0.0001-0.01
0.0003-0.04
0.0002-0.03
nil
0.0024-0.28
0.9719
nil
nil
nil
0.0013-0.15
nil
0.0005-0.06(1)
0.0031-0.36(2)
0.0021-0.24(2)
0.0009-0.10(2)
0.0048-0.56(5)
0.0086-1.01(4)
0.0010-0.12(3)
0.0011-0.13(4)
0.0009-0.10(4)
3.23
262.
of air.(b)Theserefer
appearance correlation. (
0.0025-0.08
0.0019-0.06
nil
0.0011-0.03
0.0010-0.03
nil
0.0010-0.03
0.0009-0.03
0.0032-0.10
0.00008-0.003
0.0086-0.26
0.9032
0.0012-0.04
0.0005-0.01
0.0012-0.04
0.0059-0.18
0.0024-0.07
0.0010-0.03(1)
0.0123-0.37(2)
0.0089-0.27(2)
0.0073-0.22(3)
0.0148-0.45(4)
0.0180-0.54(3)
0.0016-0.05(4)
0.0011-0.04(2)
0.00025-0.01(1)
2.943
276.
(a) This sample contained a few mi
appearance times corresponding
is number of peaks in group.
llimeters pressure
to the non-cyclic
only to materials having
c) Number in brackets
100


. 66
Figure 20. GCn-C^F-^) versus bulk density for 30C C3F5
samples (cobalt-60)(ai)*
O
(a) Numbers beside points are ergs absorbed per gram X 10.
(b) The line indicates the average G value for the samples.


151
the vapor and liquid densities of the materials as a function of
temperature. Then by use of the two following equations the weight
and volume of liquid and gas can be determined.
VLfL + Vg c ^
where:
= volume of liquid
Vq = volume of gas
L = density of liquid
jOq ~ density of gas
wt = weight of sample
Vrp = volume of sample tube
The liquid and gas densities are from references 12 and 57*
The weight average flux is then determined by
_ flwl + fgwg
wt
where:
F = weight average flux
F^ = flux over the volume occupied by the liquid
F flux over the volume occupied by the gas
G
This equation neglects the density effect, but as has already
been indicated this should introduce errors of less than 1 percent.
A more serious error may be the assumption that all the liquid is in
the bottom of the tube. It is possible that small drops of liquid
clinging to the walls above the liquid level caused some of the large
scatter of the data for these two phase samples. ^


86
Evidence for this reaction is the presence of C2F2 "t^ie Pro ucts from this irradiation. Since this is the only case for a high
density parent where an unsaturated material is present in significant
amounts for low energy absorption this reaction probably is of consider
able importance.
A simplified mechanism consisting of Equations 1-1 to 1-4, 1-6,
and 1-7 was used to perform the random recombination calculations. The
results indicate that about 24 percent of the radicals recombine to
give the parent. The effectiveness numbers confirm the conclusions
from the other molecules. Carbon-fluorine bonds are more easily rup
tured the further they are from CFj, groups. Removal of a fluorine from
an attached CF^ group is rather difficult. All carbon-carbon bonds are
equivalent for radiation damage. Some CF^ radicals are formed by ex
tensive rupture of the molecule into more than one single-carbon
fragment.


116
i
Figure 28c. Diagram for calculation of relative flux
across sample zone by infinite shell model.


16
Figure 2. LITR sample holder


77
by radical or fluorine capture. Small amounts of unsaturatee are
necessary to account for the significant amounts of C-q and formed
in the irradiation.
Since the possible number of reactions has become so large, a
generalized reaction mechanism has been written from which the individual
reaction equations are easily derived (Table 15).
For the statistical recombination calculation a simplified
mechanism consisting of equations E-l to E-5 and E-7 has been used.
From this calculation we conclude that, on a statistical basis, about
25 percent of the radicals formed will recombine to give n-C^F-^. Ex
amination of the calculated effectiveness numbers leads to the conclusion
that the further a 0Fo group is from a CF-. group the easier its C-F
2 ?
bonds are to split. This follows from the effectiveness of the radical
formed by removal of fluorine in the three position being greater than
that formed by removal of fluorine in the two position. In all other
respects this data confirms that for the smaller molecules. That is,
radicals formed by carbon-carbon bond breakage, other than CF, are
5
equally effective indicating that geometric effects are not important,
and CF^ must be formed partially by extensive fragmentation of the
original molecules to yield more than one single-carbon group. This
is also expected from mass spectrograph data where GF^ is in great
excess.
Perfluorocyclopentane
Little can be said about the mechanism of radiolytic change in
cy-CjpF-LQ since very few of the products have been successfully identi
fied. The probable mechanism is given in Table l4.


TABLE 18 PART 3. C^Fg IRRADIATED BY COBALT-60 GAMMA RAYS, T = 99C
Tube No.
53
50
63
52
Sample Wt., gms.
0.1767
0.1700
0.561+8
0.1767
Density gm./cm.^
0.212
0.2035
0.677
0.2115
No. Days
1 8
10ft
108
20.17
10.25
20.17
10.25
Integrated Dose,
Roent.
3.66
1.86
3.66
1.86
Energy Absorbed,
ergs
5.1b
2.8l
18.1+
2.92
Energy Abs./gm.,
ergs
32.5
16.5
32.5,
Value'a'
16.5
Compound in Assay
Mole FractionG
Moles F2/moleG Value
cfh
C2F6
c2fi+
C2F2
c3f8
ChllO
n-c5Fi2
1-C5F12
n-C6Fll+
2-CF C F
3"CF3C5F11
2,3-(CF3)^Cu8
Total C-vF-jg
Total CgF^g
Total CgFgQ
Total 0
Total
Total
Z
Molecular wt. Gas Density
Z^IO 22
11^
C12f26
G
O.OOTT5-1.19
0.0037-0.57
0.001+1-0.62
nil
nil
0.9973
0.0061+-0.98
0.0010-0.15
0.0025-0.38
0.0005-0.07
0.0008-0.12
nil
0.0015-0.23
0.0007-0.11(2)
0.0007-0.11(2)
0.0006-0.09(2)
0.000l+-0.07(2)
nil
nil
1+.69
19U.5
0.0032-1.00
0.0008-0.25
0.0010-0.31
nil
nil
0.9926
0.0022-0.68
0.0005-0.16
0.0011-0.3!+
0.00008-0.03
0.0005-0.16
nil
0.0003-0.09
0.0003-0.09(2)
0.0003-0.09(2)
0.0001-0.03(1)
0.00009-0.03(2)
nil
nil
3.26
188.
0.0055-0.85
0.001+5-0.69
0.0055-0.85
nil
nil
0.971+9
0.0071+-1.11+
0.0018-0.28
0.0022-0.3l+
0.0001+-0.07
0.0009-0. ll+
nil
1.0007-0.11
0.0006-0.09(2)
0.000l+-0.07(2)
0.00035-0.06(2)
0.0003-0.05(2)
nil
nil
1+.71*
191.3
0.001+2-1.29
0.0008-0.2l+
0.0013-0.1+0
nil
nil
0.9908
0.0031-0.96
0.0005-0.16
0.0011+-0.1+3
trace
0.0006-0.19
nil
0.0003-0.09
0.0006-0.19(2)
0.000lt-0.12(2)
0.0002-0.07(1)
nil
nil
nil
l+.lU
189.3
CD
'O
(a) Number in brackets is number of peaks in group.


82
A simplified mechanism consisting of equations G-l to G-6,
G-8, and G9 has been used in the statistical recombination calculation.
From this calculation the recombination of fragments to yield the
original n-C^F^ is estimated to bo 22.5 percent. As has been the
case for the smaller molecules the further a 0F2 group is from a CFj
group the easier its carbon-fluorine bonds are to split. Also as in the
other molecules the radicals formed by carbon-carbon bond splitting,
excepting CF^ have the same effectiveness number indicating the rupture
of any carbon-carbon bond is equally likely. The larger effectiveness
of CF^ must be due to greater fragmentation of some molecules to yield
more than one single-carbon radical.
In the LITR at equilibrium about 50 weight percent of the sample
is converted to CF^.
Perfluoro-2-Methylpentane
A number of samples of this material were irradiated early in
the program. However, in testing the thiourea column (see Appendix 5)
it was found that about 7 mole percent unsaturates were present. These
impurities gave fairly large amounts of unwanted peaks probably result
ing from radical capture. While this confirms the hypothesis that
relatively small amounts of unsaturated material can contribute signifi
cantly to the final product distribution by radical capture the results
of these tubes are too complex for interpretation at the present time.
For this reason the cobalt-60 irradiations of the impure material axe
not reported, and instead a sample irradiated after further purification
is reported. The irradiations of the impure material in the Oak Ridge
Graphite Reactor and in the LITR axe reported since after repurification


56
TABLE 6 (continued)
Tube No.
4
5
Compound
g2F6
c2f6
Compound in assay or standard
retention volume VR, cc. H2
Mole
Fraction^
cf4
02f6
0.7467
0.1005'
0.7244
0.1048
V
V
C 2F4
2F2
c3f8
cy-6*F6
O4P10(mostly iso)
V^0 100 cc. ^(n-hex)
n-C5Fi2
i-CcFi2
224 cc. Ln-hex)
n-C6Pl4
508 cc. Kp(n-hex)
2-CF5C5F1f
5-c?iciFu
2,MCF)2C4F8
Total CyF^g
Total CgF^g
Total CpFpo
Total Ci f)Fpp
Total C^^F24
Total 0^2^26
i5f28
cl4
15
Total C-j_g
Total C]_y
Toted C^q
Total
Sample wt., Gms. (initial)
Total
Total
Total
nil
0.0047
0.0594
0.0007
0.0271
nil
0.0050
0.0582
nil
0.0009
nil
0.0022
0.0015
0.0180
0.0055 (2)
0.0062 (2)
0.0022 (5)
0.0059 (5)
0.0018 (5)
0.0005 (2)
nil
11
11
11
11
11
11
0.2887
nil
0.0047
0.0520
0.0009
0.0184
nil
0.0014
0.0218
nil
0.0006
nil
0.0014
0.0012
0.0122
0.0048
(2)
0.0042
(1)
0.0129
W
0.0157
(5)
0.0095
(5)
0.0092
(7
O.OO55
W
0.0017
W
0.0007
(2)
0.0015
(1)
0.0054
(2)
0.0061
W
0.0025
(1)
0.505
Molecular vrfc., Gas Density (of gas)
122.7
155
Weight percent recovery as gas
100
100
Moles ?2 lost/mole mixture

-
Moleculeir wt. of gas from analysis
121.9
148.8
(a) Number in brackets is number of peaks in group


105
More work is needed in further investigation of other fluoro
carbons to study the density effect and energy absorption effects of
the type carried out on perfluoropropane.
A number of interesting experiments could be done to investi
gate the equilibrium encountered in the LITR irradiations. Two of
these have already been mentioned; irradiation of cyclo-C^F^Q with
additional fluorine to give a fluorine-to-carbon ratio of four to see
if this results in nearly pure OF^., and irradiation of mixtures of
powdered graphite and fluorine.


TABLE 18 PART 4 (continued)
Tube No.
Sample Wt., Gms.
Bulk Density, gm./cm.
No, Days
Integrated Dose, 10 ^ Roent.
Energy Absorbed, 10^ ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
45
0.5297
0.634
23.0
6.66
31.3
59.2
26
0.4903
0.587
33.0
13.5
58.8
120.0
Mole Fraction-
6
0.3627
0.434
30.0
12.92
41.7
115.0
-G Value'd)
48
0.3081
0.369
23.0
9.80
26.8
87.0
Moles Fp/mole mixtureG Value
0.0134-1.12
0.041*5-1.78
0.0255-1.09
0.0210-1.18
CFjj
0.0108-0.90
0.0199-0.80
0.0170-0.73
.'.0.0134-0.75
C2F6
0.0107-0.89
0.0155-0.62
0.0156-0.67
0.0113-0.63
c2f4
nil
nil
nil
nil
C2F2
0.0002-0.02
0.0002-0.01
0.0004-0.02
0.0001-0.006
C3F8
0.9436
0.9221
0.9097
0.9378
n-C^F?
nil
nil
nil
nil
cplo
0.0145-1.21
0.0177-0.71
0.0223-0.95
0.0134-0.75
n-CjFg
0.0037-0.31
0.0036-0.14
0.0064-0.27
0.0031-0.18
i-C^F^
0.0058-0.48
0.0059-0.24
0.0090-0.39
0.0054-0.31
nc6fi4
2-CFoC5F1]l
0.0013-0.11
0.0018-0.07
0.0032-0.14
0.0014-0.07
0.0027-0.22
0.0030-0.12
0.0064-0.27
0.0026-0.15
3_CF3C5F]1
2,3-(CF|)|cJg
Total C?Fi6
Total CoFT'o
Total CqF^q
0.0003-0.03
0.0003-0.01
trace
0.0003-0.02
0.0025-0.20
0.0023-0.09
0.0038-0.16
0.0023-0.13
0.0014-0.12(2)
0.0025-0.10(2)
0.0023-0.10(2)
0.0024-0.14(2)
0.0010-0.08(2)
0.0019-0.08(2)
0.0017-0.07(2)
0.0019-0.11(2)
0.00062-0.06(2)
0.0012-0.05(2)
0.0017-0.07(3)
0.0015-0.08(2)
Total C10F22
0.00053-0.05(4)
0.0014-0.05(2)
0.0003-0.01(1)
0.0017-0.09(5)
Total CllF2,
nil
0.00033-0.01(3)
nil
0.0003-0.02(3)
Total C12F2g
nil
0.00029-0.01(2)
nil
0.0005-0.03(3)
Total C-. oF0q
£G13 28
5.80
4.89
4.94
0.0003-0.02(3)
4.666
Molecular Wt., Gas Density
195.3
213.
193.
191.1
(a) These three samples were analyzed before the addition of the high temperature squalane column and
do not have complete analyses of the high molecular weight compounds. (b) This tube was copper,
(c) This sample was a single phase. (d) Number in brackets is number of peaks in group.


TABLE 9. CHEMICAL MECHANISM FOR CF^
Important species in mass spectrogram^ CFj, CF2, C, CF
Radical Generation
CF,
CFg + 2F
CF + 3F
(A-l)
(A-2)
(A-5)
Radical Recombination
CF + F CF2 (A-4)
cf2 + F > cf5 (A-5)
CF^ + F * CF4 (A-6)
CF + F2 CF5 (A-7)
CF2 + F2 > CF4 (A8)
CFj + F2 *CF4 + F (A-9)
CF + CF > C2F2 (A-10)
CF + CF ~^CF2 + C (A-ll)
CF2 + CF2 C2F4 (A-12)
CF2 + CF2 i CF + CF^ (A-15)
cf5 + cf5 c2F0 (a-i4)
CF^ + CF? CF2 + 0F4 (A-15)
unsaturates + F* F2, radicals saturated molecules (A-l£>)


9
in a geometry such as was used in this work. Since the inside diameter
of the sample tubes was only 0.191 centimeters, the surface to volume
ratio is relatively high and the surface is not too distant from
the bulk phase to allow some migration of fluorine and radicals to
l,
the surface. The probability of migration of radicals to the surface
is dependent on the size of the radical, the Bize of the molecules
of the medium, and the density of the medium among other'things. Thus
one would expect fluorine radicals to migrate most easily followed by
elementary fluorine, the perfluoromethyl radical, and so on. The
results of this migration should also be most easily observed at low
density. As a qualitative estimate of the ease of migration one can
calculate the mean free path for the various radicals in the medium
of interest. In Appendix 5 this has been done for perfluoropropane.
At a specific density of the medium a radical1s mean free path becomes
less than the average distance between molecules because this average
distance is not great enough to allow passage of the radical. For
the perfluoropropane, perfluoroethane, perfluoromethane, and fluorine
radicals, these are respectively about O.J g./cm.^, 0.5 g./cm.5,
0.55 g./cm.^, and 0.55 g./cm.^. Since free radicals have relatively
short lifetimes, the radicals are effectively trapped in the bulk
phase at high density.
Reactions of perfluoromethane radicals with tellurium, lead, and
( 21 ^
bismuth metals' to form rather unstable compounds have been observed.
With fluorine and F atoms, one would expect displacement of
the oxygen from the oxide layer. For this reason, until a protective
fluoride layer is built up, the wall will act as a fluorine sink and
reactions of radicals with elementary fluorine will be suppressed.


129
TABLE 22. ELECTRON STOPPING POWER RATIOS RELATIVE TO AIR
FROM FIGURE 51 FOR MATERIALS OF INTEREST
Material
At. No.
Stopping Power
Relative to Air
Stopping Power
Relative to A1
A1
15
0.922
1.0
cf4
8.4
0.9&5
1.067
2F6
8.25
0.986
1.07
5F8
8.18
0.986
1.07^
4F10
8.18
0.987
1.071
5f12
8.12
0.988
1.072
c5fio
8.0
0.990
1.075
6F14
8.10
0.988
1.072
Cu
29
0.780

a. Stopping power of
C^Fq relative to
copper = 1.264


ABBREVIATIONS AND DEFINITIONS
LITR
G
Roentgen
rad
rep
Oak Ridge Low Intensity Testing Reactor
number of molecules produced per 100 e.v. absorbed
that quantity of x or (f radiation such that the
associated corpuscular emission per 0.001295 £?n. f
air produces, in air, ions carrying 1 esu of quantity
of electricity of either sign
100 ergs absorbed/gram
roentgen equivalent physical, 95 ergs absorbed/gram.
185


84
TABLE l6. CHEMICAL MECHANISM FOR 2-CFjC^F-^
No mass spectrogram available
Radical Generation
C-C-O-C-C -AV c-c-c-c-c*
I I
c c
YV 1 ^ CCCCC
c
> C-C-C-C-C
I
c
C-C-C-C-C
-^ CCCCc
c
\\ CF^ + C-C-C-C*
c
CFj + C-C-C-C-C
""> C2F5 + CC C*
c
> nC^Fy + i-C^Fy
unsaturated radicals + F
Radical Recombination
F + F F2
R1 + R2 >R1R2
R-^ + F2 R-^F + F
unsaturated radicals* unsaturated molecules
unsaturated molecules + F, F2, radicals >
saturated molecules
(H-l)
(H-2)
(H-5)
(H-4)
(H-5)
(H-6)
(H-7)
(H-8)
(H-9)
(H-10)
(H-ll)
(H-l2)
(K-15)
(H-14)
(H-15)


155
This method of sealing is quite satisfactory and has the dis
tinct advantage that the sample tubes are not removed from the vacuum-
system until they are sealed. This prevents recontamination with air
and water vapor.


81
TABLE 15. CHEMICAL MECHANISM FOR
Mass spectrograph not available
Radical Generation
n"c6Fl4JV'' n"6Fi5 + F (G-1)
w
-* 2-6f15 + F
(6-2)
i-Vv* F
(G-?)
~-w
- CF^ + n0^F11
(g-4)
C2F5 + n-C^Fp
(g-5)

> 2CjFy
(g6)
> unsaturated radicals + F
(G7)
Radical Recombination
F + F *F2
(6-8)
R1 + R2 R1
R2
(6-9)
R1 + F2 R1
F + F
(G-10)
unsaturated radicals unsaturated molecules
(G11)
unsaturated molecules + F, F2, radicsds
saturated molecules
(6-12)


To vacuum
system
Figure 32, Sample purification and canning system,


Appendices


104
this reason for long irradiations perfluoromethane appears to be
extremely stable.
The initial G values for disappearance of the parent molecules
are related to each other in much the same way that the sensitivities
to electron bombardment in the mass spectrograph are related. This
is not unexpected since, in both cases, the species causing bond rup
ture is electrons. Thus, given the sensitivities of a known and an
unknown fluorocarbon and the G value for the known fluorocarbon the
G value for the unknown can be estimated as the ratio of the sensitiv
ities times the known G value.
The G values for disappearance of the parent compound are ap
proximately; cp^, 1.5; c2?6, 5.25; c^Fg, 4.5; n-c^Q, 5A; n-5F12
4.9; cyclo-C^F10, 5*0; n-C^F^, 6.0; S-CF^C^F^, 6.0; and 2,5-(CF^)2C^FQ,
9.6. These values are only approximate due to the dependence of cer
tain reactions on density and amount of energy absorbed.
The important products in perfluoromethane irradiation are per-
fluoroethane and perfluoroacetylene produced in about equal amounts.
For the other materials for short irradiations the major products are
saturated and are those predicted by simple bond rupture and radical
recombinations. The results for C2F£ and larger molecules, excluding
cyclo compounds can be correlated by statistical recombination calcula
tions, using empirically determined radical effectiveness numbers.
Since these effectiveness numbers are the same for radicals formed in
the same way this method can be used to predict product distributions
for other fluorocarbons. In conjunction with prediction of the overall
G values from mass spectrometer sensitivities the G values for the
products can be estimated.


CANNING
Before irradiation, the samples were treated in vacuum to
remove air, carbon dioxide, and water by passage through absorbents
and alternate freezing, pimping, and vaporization. After cleanup,
the samples were sealed in aluminum tubes 11.5 inches long by 0.0752
inches inside diameter by heliarc following collection in the sample
tubes by freezing in liquid nitrogen. One sample of C^Fg, number 51
was sealed in an identical fashion in a copper tube 12-1/8 inches
long by 0.0664 inches inside diameter. A further description of this
procedure and the apparatus is given in Appendix 2.
17


16
TABLE 12. CHEMICAL MECHANISM FOR n-C^Q
Important mass spectrometer peaks
(6)
CF3 C2F5, CF, C^Fc;
Radical Generation
n-C^F-, n nC^F-y + CFv (Dl)
VV > 2 C^F' ^ (D-2)
^-AA > n-C^Fo + F (D5)
AAi & 2 C4F9 + F (D-4)
VV > unsaturated radicals + F (D-5)
Radical Recombination
F + F -F2
C4F9 + P > C4F10
CF3 + n-C^Fy > n-C^F-^Q
2 C2F^ ^ n-C4F^Q
unsaturated radicals + F ^n-C^F^Q
unsaturated radicals'> unsaturated molecules
CF3 + F > CF4
n-CzFy + F >CzFq
C2F5 + F * C2F§
CFj + CF3 >C2F6
CF* + C2F3<>CzFq
c2f5 y8 >n-CkF12
n-C4F + CF* n-Cf-Fn 0
i~C4F9 + CF5 >iC5F12
n-C4F3 + C2% > n-c^F14
-C4F9 + 02F3 * 5-CFzCgFjj
n-C4F9 + n-CsFy^n-CyF,^
-C4FU + n-C^Fy +}-CFzpF1z
nC4P9 + nC4F9 nr-C3F]_3
nC4P9 + -C4F0 5~CFzCyFic
-C4F0 + -C4F0.>54-(cf5;2C6F12
CFz + F2 CF4 + F
C2Fc + F2 > C2F6 + F
nC3F7 + F21 >OzFg + F
C4P9 + P2>C4F10 + F
vmsaturated molecules + f2, p, radicals *
saturated molecules
(I>*)
(D~7)
(I>3)
Cd-9)
(D-10)
(D-ll)
(D-12)
(D-15)
(D-l4)
(D-15)
(D-16)
(D-17)
(D-18)
(D-19)
(D-20)
(D-21)
(D-22)
(3>25)
(D-24)
(D-25)
(D-26)
(D-27)
(D-28)
(D-29)
CD-50)
(D-51)


125
the aluminum wall will be absorbed in the wall, some will be absorbed
in the sample, and some will pass through the sample to the opposite
wall. If the gamma ray is absorbed in the sample similar things can
occur. The electrons may be absorbed in the sample or may pass through
the sample and be absorbed in the wall. Thus there is an "atmosphere"
of electrons present in the sample space the density of which is
dependent upon the gamma absorption in the two media and the electron
stopping powers of the two media. If the sample space is large the
electron "atmosphere" will be characteristic of the sample material
and independent of the wall material. If the sample space is small
the electron "atmosphere" will be characteristic-of the wall material
and essentially independent of the sample. This is the basis of the
Bragg-Gray cavity law.
In order for the Bragg-Gray cavity law to apply rigorously two
conditions must be met:
(1) The intensity of the primary radiation must be essentially
uniform over the region from which the secondary electrons can reach
the cavity;
(2) The secondary electrons should be nearly uniform in the
region of the cavity.
The first condition is met in this case due to the geometry
of the source. The intensity will not vary much over 10 percent across
the area used for irradiation as is shown in Flux Exposure of Samples"
in this appendix.
The second condition requires that the range of the secondary
electrons be less than the wall thickness and that an insignificant
portion of them are absorbed in passing through the cavity. Neither


TABLE l8 PART 4 (continued)
Tube No.
30
42
46
47
Sample Wt., Gms.
0.4808
0.1452
0.2995
0.2792
Bulk Density, gm./cm.
0.577
0.1738
0.359
0.334
No. Days
10.0
23.0
10.0
10.0
Integrated Dose, 10 7 Roent.
Energy Absorbed, 10 ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
4,o6
7.28
4.24
4.17
17.35
9.39
11.3
10.1
36.1
64.7
Mole Fraction-
3T*7fd)
-G Valueva;
36.2
Moles F0/mole mixtureG Value
0.0080-1.11
0.0230-1.74
0.0086-1.i4
0.0093-1.;
CFU
'2F6
C2F4
^2F2
c3F8
n~C3F6
C4F10
n_c f12
i-C5Fi2
nc6FlU
2-CF3C5Fn
2,3-7Si?1c^8
Total
Total
Total
Total
Total
Total C]L2^26
ctf16
C8Fl8
C9F20
C10^22
Total C-,oFoA
G13 20
Molecular Wt., Gas Density
0.0033-0.46
0.0047-0.65
nil
trace
0.9773
nil
0.0056-0.78
0.0015-0.20
0.0024-0.34
0.0012-0.17
0.0016-0.22
0.0001-0.01
0.0007-0.09
0.0004-0.06(1)
0.0006-0.08(2)
0.0004-0.06(2)
0.0004-0.06(2)
nil
nil
4.29
192.2
(a) These three samples were analyzed before the
. and do not have complete analyses of the high
copper. (c) This sample was a single phase
group.
0.0089-0.67
0.0087-0.66
nil
nil
0.9528
nil
0.0108-0.82
0.0034-0.26
0.0051-0.39
0.00085-0.06
0.0025-0.19
nil
0.0019-0.14
0.0014-0.11(2)
0.0015-0.11(2)
0.0009-0.07(2)
0.0010-0.08(5)
0.0002-0.01(1)
nil
5.31
191.8
0.0043-0.57
0.0042-0.56
nil
nil
0.9841
nil
0.0071-0.94
0.0016-0.21
0.0034-0.45
O.OOO82-O.II
0.0014-0.19
nil
0.0011-0.
0.00082-0
O.OOO58-O
0.00042-0
0.00026-0
0.00005-0
nil
4.60
191.0
15
.11(2)
.07(2)
.06(2)
.03(3)
.01(1)
0.0049-0.67
0.0054-0.74
nil
nil.
O.9702
nil
0.0075-1.02
0.0021-0.29
O.OO36-O.49
0.00068-0.09
0.0022-0.30
nil
0.0014-0.20
0.00077-0.10(2)
0.00049-0.07(2)
0.00044-0.06(2)
0.00037-0.05(2)
nil
nil
5.36
197.5
addition of the high temperature squalane column
molecular weight compounds. (b) This tube was
(d) Humber in brackets is number of peaks in
vo
\>i


LIST Ox* FIGURES
Figure Page
1 Electrically heated sanple holder l4
2 LITR sample holder l6
5Weight percent CF4 versus fluorine to carbon ratio of
original material for samples irradiated in the
LITR 21
4 Isoteniscope used for determination of the molecular
weight of the solid fluorocarbon produced in the
LITR 25
5 Indicated mole fraction vs. solution vapor pressure
of solid fluorocarbon in n-CyF-^g 27
6 Infrared spectrogram of solid fluorocarbon from LITR
irradiations 2 9
7 G(CF^) versus density for 99G C^Fg samples 52
8 G(CF4) versus bulk density for 50C Cj-Fo samples
(cobalt-60) 55
9 G(C0F) versus density for 99C C^Fn samples
(§ooalt-60) A 54
10 C(C0Fg) versus bulk density for 30C CkFP samples
(cobalt-60) ? 55
11 G(Fo) lost versus density for 99C C-Fo samples
(cobalt-60) ? 57
12 G(F0) lost versus bulk density for 50C C_FQ samples
(cobalt-60) J.8 58
G(Cz.Fin) versus density for 99C C^Fo samples
(cooalt-60) V 59
14 G(C4Fin) versus bulk density for pOG C*Fo samples
(coBaL-t-60) . 60
15 G(n-CcF-, 0) versus density for 99G CrFo samples
(co?al£-6o) < 61
viii


Log(n, moles)
145
Figure 35 Log(moles) versus log(area) for chromatographed
fluorocarbons


25
Figure U Isoteniscope used for determination of the molecular
weight of the solid fluorocarbon produced in the
LITfi .


APPENDIX 5. ANALYTICAL
Physical Description of Analytical Equipment
In the analysis of the irradiated samples a series' of three
gas chromatographs was used. Two Perkin-Elmer Yapor Fractometers
equipped with gas sampling valves and one rack-mounted detector-column
assembly were used. Thermal conductivity detectors with thermistors
were used in all three chromatographs. To allow easy gas sample
introduction the two Vapor Fractometers were equipped with Precision
Gas Sampling valves (Perkin-Elmer). For analyses a 25 cc. sample was
introduced through each valve. The pressure of these samples varied
from a few millimeters to about 80 millimeters of mercury; temperature
vas about 25C.
The columns used in the analyses were; 75*5 meters of squalane
operated at 92C, 16 meters of n-hexadecane operated at 25C, and one
meter of silica gel which was temperature programmed. The squalane
and silica gel columns were arranged to operate either in series or
parallel. Initially they were in series to allow the materials from
air to C^F^q to enter the silica gel column and were then separated
by a valve arrangement and run in parallel. Thus the silica gel
column was used to analyze for air, CF^, CgF^, C2F2*
n-C^F^, Cy-C^F^, and C^F^Q. The squalane column analyzed those
materials above C5> but was unable to separate all isomers. The
156


85
TABLE 17. CHEMICAL MECHANISM FOR 2,5-(CF5)2C^Fq
No mass spectrogram available
Radical Generation
0-C-C-C -W C-C-C-C* + F
M II
C C c c
> C-C-C-C + F
I I
c c
> CFt + CCCC
5 I
(1-1)
(1-2)
(1-5)
*YV-
j~'VV*
2 -C3F7
unsaturated radicals + F
(1-4)
(1-5)
Radical Recombination
F + F
R + R,
R1 + V
R1R2
Rj^F + F
unsaturated radicals
unsaturated molecules
unsaturated molecules + F, F2 radicals'
saturated molecules
(1-6)
(1-7)
(1-8)
(1-9)
(1-10)


65
>
0.10
_ 0.08
% 0.06
V
5 CfwOUi
0.02
O
a ojl o.2 o.3 o.ii o.5 o.6 0.7 0.8 0.9
Density, gm./cm.^
Figure 19. GCn-CgF^) versus bulk density for 99C CgFg
samples (cobalt-O)^3'^.
(a) Numbers beside points are ergs absorbed per gram X 10.
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid density.
(c) The line indicates the average G value for the samples.


110
From this data the flux for the samples can be shown to be about 88
percent of that in the unperturbed radiation zone.
Ambient temperature
^d (.11)(7.7) + (.0655)(2.7)
steel + aluminum
- 1.019
= .058
Id ^-(.058)(1.019)
Io "
9^4
Elevated temperature
2yd = (,54)(2.25) + (.ll)(7.7) + (.12)(2.7) + (.10)(2.5)
= glass + steel + aluminum + asbestos
- 2.19
Id _-(.058)(2.18)
.881 = .88
Variation of Flux Across Sample Zone
The variation in the flux across the radiation zone is not
easily measured since accurate measurements would require either a
very small probe or small amounts of a chemical dosimeter. Because
of the difficulties involved in either of these approaches it was
decided to approach the problem by a simple mathematical model.
If we neglect scattering and absorption of the photons by
material in their paths the radiation from a point source will follow
the inverse square law. Using this assumption it is possible to
develop an equation allowing one to calculate the relative intensity
of radiation throughout a volume given the geometry of the source.


TABLE 18 PART 2.
C2F6
IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.
Sample wt., gras.
Density, gm./cm.^
No. Days
Integrated Dose, 10^ Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in Assay or as Listed
88
0.2134
0.255
9.92
1.80
3.41
99
68
0.3723
0.447
17.17
3.44
11.4
25
Mole Fraction-
35(a)
0.2931
0.351
33.09
6.65
17.3
25,,^
-G Value''
69
0.7075
0.846
17.17
3.44
21.7
25
Moles F2/mole mixtureG Value
0.0035-1.53
0.0050-1.21
0.0090-1.01
0.0053-1.17
' CF.
0.0010-0.44
0.0029-0.70
0.0072-0.80
0.0038-0.84
C2F6
0.99605
0.9901
0.9789
0.9877
C2Fk
nil
nil
nil
nil
^2F2
0.0012-0.52
0.0002-0.05
0.0003-0.04
0.0003-0.07
C3F8
Cy-c3F6
0.0012-0.52
0.0045-1.08
0.0083-0.93
0.0055-1.22
nil
0.0002-0.05
0.0005-0.06
0.0002-0.05
C^io (mostly normal)
n-<"'5F12
Not Detected
0,0011-0.26
0.0028-0.32
0.0017-0.37
Total C^F^
0.0001-0.03
0.0005-0.06
0.0003-0.07
i-C5F12
n_c6Fl4
0.0002-0.08
0.0002-0.05
0.0007-0.07
0.0001-0.02
0.0001-0.05
0.0001-0.03
0.0003-0.04
0.0001-0.02
2-CF3C5Fll
3CF3C5F11
Total 2 + 3 =
0.0001-0.03
nil
Total 2 + 3 =
O.OOO2-O.O8
0.0001-0.03
trace '
0.0001-0.02
2,3-(CF ) C4F8
\ Total C7Fl6
Total CqF^q
Tta^9F20
0.00005-0.02
nil
nil
nil
0.0001-0.03
0.00013-0.03(2)
0.00013-0.03(2)
0.00003-0.01(1)
0.0005-0.06
0.0001-0.02
0.00004-0.0i(2)
0.00014-0.03(2)
0.00004-0.01(1)
3.24
3.62
3.39
3.92
Molecular wt., Gas Density
138
143.3
145
141.5
(a) This tube analyzed before 92C squalane column installed so amount of compounds above Cg is
. not known.
(b) Number in brackets is number of peaks in group.
03
CD


Figure Page
16 G(n-C^F. ) versus bulk density for 50C C-,FQ samples
(coSal u-60) 62
17 G(i-C-F10) versus density for 99C C^Fn samples
(cobalu-60) i 6?
18 G(i-C,-F, 0) versus bulk density for JOC CAFo samples
(cooait-60) ^ 64
19 G(n-C/'F, j,) versus density for 99C C-Fo samples
(cobalt-60) ? ? 65
20 G(n-0yF,;,) versus bulk density for ^0C C-Fo samles
(coDait-60) 66
21 G(2-CF-Cc-F-,, ) versus density for 99C C,FA samples
(cobalt-50) 67
22 G(2-CF^C^F1. ) versus bulk density for 5C C^Fo
sampfei; \cobalt-60) ; . 68
25 G(2,5(GiP5)2C4F8^ versus density for 99C CF
samples'^ cobalt-60) '. ... 69
24 G(2,5-(CFz)pC,Fq) versus bulk density for 5>0C CkFn
samples^(cobalt-60) V 70
25 Sum of G values versus density for 99C 0,Fo samples
(cobalt-60) t8. .... 71
26 Sum of G values versus bulk density for 50C C^Fq
samples (cobalt-60) ; . 72
27Flux in tube 11 of the engineering cobalt-60 source 109
28a Diagram for relative flux across sample zone
calculation-planar 112
2ob Diagram for calculation of relative flux across
sample zone-volume Il4
28c Diagram for calculation of relative flux across
sample zone by infinite shell model Il6
29 Arrangement of cobalt-60 rods around tube number 11
of the engineering cobalt-60 source 117
50. Relative flux across sample zone by infinite vre
and infinite shell models 118
ix


164
2. The other radicals will reach the wall in similar amounts
leading to the conclusion
5. The distribution of products will be similar at all densities
except for materials formed by reaction with F or and materials
formed by unsaturates reacting with radicals.


59
TABLE 6 (continued)
Tube No.
55
Compound
n-6Pl4
Compound in assay or standard
retention volume VR, cc. H2
fa"}
Mole Fractionv '
cf4
C2F6
2F4
c2f2
cy-Cf6
C4F10(mostly iso)
Vr,0 = 100 cc. H0(n-hex)
R
V-
R
n-C5F12
i~CF12
= 224 cc. H2(n-hex)
508 cc.^pL(n-hex)
2CFzGcF-i n
5-CFgc§Fn
2,5-(of512c4f8
Total^
7l6
efe
Total
Total
Total
Total
V2
Total Cn zF0o
Total XV8
Total C15
Total
Total -
Total
Total
Sample wt.,
£7
as
Gms^( Initial)
0.8017
0.1162
nil
O.OO56
0.0574
0.0005
0.0124
0.0001
0.0015
0.0105
0.00004
0.0005
nil
0.0008
o.ooo4
0.0052
0.0022 (2)
0.0050 (2)
0.0015 (2)
0.0021 (5)
0.0005 (2)
0.0002 (1)
nil
it
tt
It
It
U
II
0.575
Molecular wt., Gas Density (of gas) 107
Weight percent recovery as gas 592
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis 107
(a)Number in brackets is number of peaks in group


150
n
2W +
i=l
jS
A3
2
i=l
i^s
KiCi
Solving for W we get:
W
Xs
yT NiCi
i-1
NiCiXi
2
It should be noted here that the total composition of these
mixtures is not exactly known. In particular there is a possibility
of unsaturated compounds of four or more carbon atoms since these
could not be detected with the analytical system used. There were
probably not significant amounts of these present for original
materials of two carbons and greater since very small amounts of
unsaturated two and three carbon material.s were found. If unsaturated
materials were present and undetected this would make the fluorine
loss values higher.


166
TABLE 50. RANDOM RECOMBINATION CALCULATION AND RESULTS FOR C2F
Ruptura
2 Reactions
Recombination Reactions
F F
1 1
F + F F2
1 1
- C C F
| (
> F + C 2 F^
F + CF_ *
2 ?
2r6
F F
-A/ > CP, + CF,
3 3
OF, + CF,
CF* + F
3
* 2F6
cf4
C0F¡- + CF,
2 3 3
>0 F
58
2F5 + 2P5^ 4f10
Fr affluent
N(number ways formed) E(effectiveness)
N x E
F
6
0.50
1.80
CF?
2
1.00
2.00
C2F5
6
.34
2.04
Relative Yields of Products

f2 (1.80)2 = 5.24
C?4 (2)(2.00)(1.80) 7.20
C2F6 (2)(2.04)(1.80) + (2.00)
2 = 11.54
C-Fg (2)(2.04)(2.00) = 8.17
nG4F10 = (2*4)2 = 4.16
£ = 54.11
Yields of products relative
to CF4
Product
Yield/CF^ yield(correlation)
Yield/CF^ tube 68
F
2
.45
1.73
cf4
1.0
1.0
C/8
1.135
1.305
n4FlO
578
.467


This dissertation was prepared under the direction of the
chairman of the candidate's supervisory committee and has been
approvod by all members of that committee. It was submitted to the
Dean of the College of Engineering and to the Graduate Council, and
\ras approved as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
August 8, 1904
Dean, Graduate School
Supervisory Committee:


71
*3
Density, gm./cm.
Figure 25. Sum of G values versus density for 99C CgFg
samples (cobalt-60)^a^.
Q
(a) Numbers beside points are ergs, absorbed per gram X 10.
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average G value for the samples.


144
TABLE 25. STANDARD RETENTION VOLUMES ON THE SQUALANE COLUMN
Pi =* 29.7 psi. Po = l6.8 psi. Temp, of Column = 92C
Carrier gas = 5^*9 cc./min. Hg at 1 atm. and 250
Column length = 15*5 meters of 168.7 grams of 0.p26 gm.
squalane/gm. Chromosorb-P.
Compound
V^0 cc. at 1 atm. and 25C
C2F6
9.17
c4f10
25
n-C^Fj g
28.9
n-C6Fi4
58.4
2 and J-CF^C^F-^
2,5-(0Pj)20j,P8
6Q.6
87.7
n-7Pl6
90.4
n8Fl8
159
n"9F20
195


G(2-CF3C^F11)
68
Bulk density, gm./cm.^
Figure 22. G(2-CF3C£F-q.) versus bulk density for 30C C3Fq
samples (cobalt-60)^a^
(a) Numbers, beside points are ergs absorbed per gram X 10"^.
(b) The line indicates the average G value for the samples.


55
TABLE 6 (continued)
Tube No.
Compound
8
cf4
0
i-1
4=- O
Compound in assay or standard
retention volume V^, cc. H2
Mole
Fraction^
CF4
0.9866
0.9809
C2F6
0.0077
0.0077
C
nil
nil
C2f2
0.0055
0.0091
5f8
0.0002
0.0001
cy-CjF6
nil
0.0001
c4fio(mos^iy is0)
V_ = lOO cc. H2(n-hex)
it
0.0021
n
nil
VR
(n-hex)
n-CcF-i o
i_C5F12
224 cc. H2
n-c6Fl4",
508 cc. Hp(n-hex)
2-CF5C5F1i
5-CF^C^F11
2,5-(c#3?2C4F
TotaTCyF^
Total CF^q
Total GgF^Q
Total 0]_oF22
Total C*|t Fpii
Total C12F26
Total
Total
Total
Total
Total
Total
Total
5F28
C14
p15
16
c17
c18
19/
Sample wt., Gma. (initial)
Molecular wt., Gas Density (of gas)
Weight percent recovery as gas
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis
0.1752
95.6
100
0.0245
88.24
0.1804
145.2^
100
0.04i4
08.46
(a)Number in brackets is number of peaks in group.
(b) This average molecular weight is obviously grossly in error.
r


72
Bulk density, gm./cm.^
Figure 26 Sum of G values versus bulk density for 30C CgFg
samples (cobalt-6G)^a^e
ft
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) The line indicates the average G value for the samples.


28
.408
x
.392 .408 0.035
388 x
x 4600
Some information concerning the structure of the solids can be
obtained from infrared absorption and examination of the shorter prod
ucts determinable by gas chromatography.
Figure 6 is the infrared spectrogram of the solid material run
in a solid pellet containing a mixture of 0.00373 grams of fluorocarbon
solid and 0.95816 grams of sodium chloride. The instrument was a
Perkin-Elmer 237 Spectrophotometer. All that can be determined from
this spectrogram is that there is some unsaturation as indicated by the
peak at 6 microns. The large peak at 8 microns corresponds to various
carbon-fluorine bonds and is indicative of the complexity of the mixture.
The fluorocarbons formed in the LITR which can be analyzed on the
gas chromatograph show a preponderance of the most branched isomer ex
cluding neo types. For C^F^Q the iso is the most prevalent. For C^F^2
i-C5Fi2 predominates, and for C^F-^, 2,3-(CF^)2C^Fg is in largest con
centration. From this it is probable that the solid fluorocarbon is
highly branched but essentially lacking in neo type structures.
We can summarize the data known about the solid fluorocarbon
below;
(1) molecular weight = 4600
(2) contains some unsaturation
(3) has a large amount of branching in its structure.


145
TABLE 24. TEMPERATURE PROGRAM AND APPEARANCE TIMES
ON THE SILICA GEL COLUMN
Gas H2
Flow rate = 5^.9 cc./min.
Column = 1 meter of silica gel
Program
Time
Initied
5 min. after air
12 min. after air
Temperature
2p C
increase to 70C
increase to 90C
Appearance Times
Compound
t-t min.
air
cf4
C2F6
0^2 C2F4
2F2
c5f8
cyclo-C^F^
n"5F6
4F10
1.1
7.0
8.7
5-4
14.4
17.0
18.5
21.7


47
amounts of C2F2 ^2?6 were nearly equal which could be accounted
for by the simple mechanism below:
CF^ ^ CF2 + 2F
CF2 + OFg OF + CF^
CF + CF ^2^2
cf5 + cf5 c2f6.
However at elevated temperature (99C) only a very small amount
of C2F2 was produced and no C2F^ was detected. From this we conclude
that the simple mechanism above is not adequate and that the true
mechanism is considerably more complex.
Perfluoroethane
In the mass spectrograph of O^F^^ the mos'<' Prevalent species
in decreasing importance are CF^, CgF^, CF, and CFg. The stability
of perfluoroethane to fragmentation by electron bombardment is not as
great as that of CF^ as is shown by the mass spectrograph studies
(see Table 8). In long term Irradiations it will be degraded a great
deal since by Figure 5 at equilibrium the sample will be about 50
weight percent CF^.
Judging from the mass spectrograph studies the initial radical
forming reactions must be those given in Table 10.
Other reactions are possible, but the ones in Table 10 are
sufficient to explain the results. At room temperature, no CgFg is
found, indicating that reaction B-ll is not important or that reactions
of the type of B-15 are important. Evidence for the latter is that at
elevated temperature where the amount of C2F2 is increased, the amount


Pare
Figure
51 Electron stopping power relative to air versus
atomic number 128
52 Sample purification and canning system ....... 154
55 Decanning and storage system 158
54 Series and parallel connections of chromatographic
columns l4l
55 Log(moles) versus log(area) for chromatographed
fluorocarbons l48
56 System used for decanning pile irradiated samples . 152
x


1
2
5
4
5
6
7
8
9
10
11
12
15
l4
15
16
17
18
LIST OF TABLES
Page
COMPARISON OF MASS SPECTROMETRIC SENSITIVITIES AND
C- VALUES
IMPURITIES IN IRRADIATED SAMPLES '
WEIGHT PERCENT CF4 IN SAMPLES IRRADIATED IN THE LITR
RESULTS OF MOLE FRACTION DETERMINATION ON SOLID
FLUOROCARBON .
GASEOUS PRODUCTS FROM THE THERMAL DECOMPOSITION OF
THE LITR SOLID FLUOROCARBON
LITR IRRADIATIONSANALYSIS OF GASEOUS PRODUCTS . .
RESULTS OF OAK RIDGE GRAPHITE REACTOR IRRADIATIONS .
MASS SPECTROGRAPH DATA FOR CFi^)
CHEMICAL MECHANISM FOR CF4
CHEMICAL MECHANISM FOR
CHEMICAL MECHANISM FOR C^Fq
CHEMICAL MECHANISM FOR n-C^Q
CHEMICAL MECHANISM FOR n-C^F^
CHEMICAL MECHANISM FOR cy-C^Q
CHEMICAL MECHANISM FOR n-C^F^
CHEMICAL MECHANISM FOR 2-CFzCcF..
CHEMICAL MECHANISM FOR 2,^(0?-)
PART 1. CF4 IRRADIATED BY COBALT-60 GAMMA RAYS . .
PART 2. C2F<5 IRRADIATED 3Y COBAXT-6O GAMMA RAYS .
5
11
22
26
51
55
4o
45
k6
48
51
16
78
79
81
84
85
87
88
v


158
To test the application of this theory to the ideal case the
mean free paths of argon, helium, and nitrogen were calculated at
one atmosphere and 100, $00, and 500 1C and compared with those found
from gas kinetics calculations. This comparison is found in Table 28.
From the tabulated values we see that the two models ore in
very good agreement. There appear to bo several distinct advantages
to the statistical model. These are listed below.
1. The statistical model points up the fact that mean free
path does not depend directly on the temperature of the medium or the
energy of the particle. The dependence is only through the density
factor.
2. The statistical model allows calculation of the probability
of other path lengths than the mean free path.
5. The statistical model allows the calculation of the mean
free path of a foreign molecule in a bulk material consisting of
another species. This last point is very important for the present
application.
Mean Free Paths of Radicals in Perfluoropropane
The values of d are calculated for several densities of C^Fo
5 8
^3 gm./crn.^
.0554
.17
55
.4
55
d, A
25.5-
15.73
10.87
9.55
9.52
7.62
1.0


142
flow rates. A finer adjustment is made by balancing for zero shift
of the base lines of the recorders when shifting from series to
parallel.
Description of Analytical Columns
As has been previously mentioned whereas the n-hexadecane and
squalane columns were operated at constant temperature it was necessary
to program the temperature of the silica gel column to achieve satis
factory appearance times. The program and appearance times are given
in Table 24.
The conditions of operation and the standard retention volumes
of various materials are given for the squalane column in Table 25.
The conditions of operation and the standard retention volumes
of various materials on the n-hexadecane column are reported in Table
26.
Urea and Thiourea Columns
In the course of development of chromatographic columns, a
number of columns were prepared and tested, but were not used directly
in the analytical procedure. The most interesting of these were the
urea and thiourea columns.
These two columns were devised in an attempt to make use of
(38
the canal complexes they are known to make with numerous compounds.
If this occurred in a chromatographic column, a separation of isomers
would be possible resulting in the elution of the most highly branched
isomers first and the normal isomers last. This is indeed the case
with both these columns with approximately equal separation being


Calculated mole fraction
Figure 5 Indicated mole fraction vs. solution vapor pressure of solid
fluorocarbon in n-C^F^


15
in this position is 10 n/cm. -sec. and the temperature is about
100C. The energy absorption in this position can be estimated from
the data of D. M. Richardson, A. 0. Allen, and J. V*'. Boyle^2<^ as
about 8.5 x 10* erg/gram for graphite. Assuming the value for
fluorocarbons is not greatly different from this approximate G values
can be calculated (see Appendix l).
The samples irradiated in the LITR were in hole $A and were
irradiated for a period of 28 days at a thermal flux of 10-^ n/cm.2-
sec. In the LITR the samples are placed in a flow of coolant water and
have a temperature of approximately 1000. See Figure 2 for a sketch
of the sample holder. No estimate of the energy absorbed in the LITR
samples can be given; however, since the samples appear to have reached
equilibrium distributions, this would be of little value.


159
The pressure in the system is read from the manometer and
recorded as A. Then assuming the ideal gas law holds the volume of
the system is calculated by:
P(atmosphere) x V(burette) = P(system) x V(system)
Additional volumes of air are admitted by the same procedure. This
yields a volume which shows a slight increase with increasing pressure
due to the increasing volume in the left leg of the manometer. Thus
the slope is known from the size of the tubing used in the manometer.
This is found to be .044 cc./mm. Fitting this slope to the four sets
of bulb calibration data yields the following volume equations:
Bulb 1
V 427.5 + .044 A
Ain millimeters of mercury
Bulb 2
V = 728 + .044 A
Bulb 5
V = 729 + .044 A
Bulb 4
V = 1289 + .044 A
Sample Treatment
The sample is introduced into the system by scratching the
aluminum tube and introducing it through a Flex fitting force-fitted
into a rubber tube. After the system is pimped down the aluminum
tube is snapped off and the sample distilled into the system. It is


49
of larger molecules is decreased. A simplified mechanism which does
a fairly good job of correlating the data at room temperature and for
low energy absorption consists of equations B-l, B-2, B-4 to B-7> B-13
and B-l4.
This mechanism has been used in Appendix 6 to correlate the
ambient temperature data by a statistical method. In essence this method
involves assignment of an empirical effectiveness to each type of radi
cal. This effectiveness is related to the ease of rupture of the bond
to form the radical and possibly to various geometric factors. Multi
plication of the effectiveness times the number of ways the radical can
be formed resulta in an effective concentration of the radical. By
random recombination calculations, the relative concentrations of prod
ucts can be calculated. In this case, the reverse of this method is
used to arrive at the effectiveness. These effectiveness numbers are
of value in two ways. First, using them,r an estimate of the fraction
of the radicals which recombine to form the original compound can be
obtained and second, they give an estimate of which type of bonds are
preferentially broken in gamma irradiation. A further discussion of
this approach and the calculation are given in Appendix 6.
For C^F^. approximately one third of the radicals recombine to
yield CgF^. From the effectiveness numbers we see that CF^ radicals
are more than twice as effective as C2F5 radicals. This probably arises
from preferential formation of CF^ as indicated by the mass spectro
graph results.
Perfluoropropane
A more detailed examination of perfluoropropane was made than
for any of the other materials studied. Perfluoropropane was irradiated


75
oxide. Both these tubes showed an increased amount of unsaturated
material leading to the conclusion that, in perfluoropropane irradia
tion, unsaturated molecules are present as intermediates. Relative
amounts of unsaturated materials were determined from the amounts of
2F2* 2F4* and 5F6 dotocte G values in the partially liquid samples and in the single >
phase samples are not significantly different so the mechanism is
probably the same.
As in the case of CgFg, a ran(*0In recombination calculation has
been made to determine the effective concentrations of the radicals.
For this calculation, a.simplified mechanism consisting of reactions
C-l, 0-2, C-4 to 0-9 and 0-11 to 0-18 has been used. The results
indicate that approximately 26 percent of the radicals formed recombine
to form C^Fg. Of the radicals formed, the i-C^Fy radical is the most
effective leading to the conclusion that the fluorine atoms attached
to a carbon atom next to a CF~ group are relatively easy to remove.
The n-CjFy radical is only O.515 times as effective so removal of a
fluorine atom from an attached CF^ group is rather difficult. The
^2F5 Radical is slightly more effective than the n-C^Fy radical either
from geometric effects or because the bond is slightly easier to rup
ture. The CF^ effectiveness is roughly twice that of probably
resulting from some complete rupture of C^Fg into single carbon entities.
This is to be expected from the mass spectrograph data.
In summary we can say for C^Fg, and by analogy for the other
fluorocarbons, both energy absorption and density affect the G values
of the products of irradiation if the geometry of the sample container
is such that the wall can act as a competitor for fluorine and radicals.


Absorbance
Wave length, microns
Figure 6* Infrared spectrogram of solid fluorocarbon from LITR irradiations


75
with the possible exception of a small decrease in the amount of CF^
formed.
No unsaturated molecules were detected, but these must have
been present as intermediates to account for the relatively large
amount of material found between 0^ and The mechanism for n-C^ F10
is given in Table 12.
The random recombination calculation has been made for n-C^ F10
using a simplified mechanism consisting of equations D-l to D-4, D-6
to D-9, and D-12 to D-26. From this calculation we see that about 52
percent of the radicals formed recombine to yield n-C^F^. These data,
as for the case of CjFq, state that the carbon-fluorine bond next to a
CF group is relatively weak compared to the carbon-fluorine bond on
5
an intact CF^ group. CF^ is more important than the other radicals
formed by carbon-carbon bond rupture indicating, once again, that in
some cases the molecule is ruptured into several fragments. The
radicals 02F5 n-C^Fy which are formed by similar processes have
identical effectiveness numbers as would be expected if these numbers
are due to ease of formation and not dependent a great deal on geometric
factors.
Perfluoro-n-Pentane
In the cobalt-00 irradiation of perfluoro-n-pentane at ambient
and elevated temperature, there is no discernable difference between
the G values for the products at the two temperatures. For low energy
absorption no unsaturated materials are found and after prolonged
irradiation only small amounts of C^F^ 021(1 are i>oun<** This
indicates that any unsaturates formed as initial products are saturated


157
Hence the probability of the particle passing successfully
through one shell is:
^ ,,2
VT1 £
4
P 1 -
17
cr+ cr' \2
2 YT
The probability of passing through each successive shell is
equal and hence the probability of successful passage through N
shells is
N
(4-2)
from.
The mean free path is that for which P
0.5 and is calculated
- d V~T /2 (N)
(L-5)
The value of d is found from the value of atoms/cm. for the
medium and setting this equal to
4/24
ZJ
.11785d
or
atoms m 1.415
cm. d


158
Figure 33 Decanning and storage system.


168
TABLE 51. RESULTS OF RANDOM RECOMBINATION CALCULATION FOR C-Fg
Fraspnent
N
E
N x E
F
8
oo
2.-40
cf5
2
1.00
2.00
2p5
2
.45
.90
n-C,F7
5 7
6
.54
2.04
2
I.065
2.15
Molecule _..c; ......c. Relative Amount
P2
5.76
of4
9.60
2F6
8.55
C5F8
25.61
n"4F10
8.97
i-^lO
8.55
n_G5F12
5.68
2~CF5C4F9
5.84
a-6pl4
4.16
2-cf55fh
8.70
5-(cf5)2c4fq
4.54
2 89.72
percent recombination = 26.5


TABLE 7. RESULTS OF OAK RIDGE GRAPHITE REACTOR IRRADIATIONS
Position * Hole Flux 10^ n/cm.^-sec. No. Days 55
Approximate energy absorption 8.J x 10^ erg/gram
Irradiation temperature ** 100C or greater
Compound in assay or standard
retention volume V^, CC.H2
Tube No. 75
Cyclo-C5F10 .
Mole Fraction'a'
CF4
2F6
2f4
2F2
3F§
Cy-CjF^
4Fio
79 on Cl6K54
117 on C1(5Hj4
n"5F12
i-C$F12
184 on Ci^Hz4
Cy-CcFio
n-C6Fl4
2-CFiCcFt n
5-F5C5Fll
2,5-(of*5)264f8
Total CyF-,/-
Total CgFl8
Total CgF20
Total C10F22
Total C^1F24
Total Cj_ 2f2
Total C-,zF?8
Total ct^F^o
Sample wt. gms.
Molecular vrt., Gas Density
Percent recovery as gas
£o
0.0098
0.0088
0.0006
0.0015
0.0055
0.0006
0.0055
0.0007
0.0025
0.0049
0.0007
0.0202
0.8866
0.0015
0.0005
0.0008
0.0099
0.0109 (2)
0.0156 (2)
0.0090 (2)
0.0018 (4)
0.0051 (5)
0.0056 (1)
nil
0.0001 (1)
0.2559
264.
100.
0.505
(a) Number in brackets is number of peaks in group


TABLE l8 PART 9. a-CgF,^ IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.
Sample Wt. Gms.
No. Days
Integrated Dose, lol Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in Assay or as Listed
32
0.481+5
30.0
14.05
6o. 8
30 ( )
Mole FractionG Value'a'
Moles F2 lost/mole mixtureG Value
CFh
C2F6
c2F2
C fq
C4F10.
n-CcF!2
i-C5Fi2
nc6FlU
VR = 338 cc. H2(n-hex)
Total CjF^g
Total CgF^g
Total C^Fgo
C10F22
Total
Total C-QF24
Total C1?F?
Total C33F2q
Total Cll+F30
2 0
Molecular Wt., Gas Density
0.0350-0.T3 0
0.0155-0.32 0
0.0113-0.23 0,
trace
0.013^-0.28 0
0.0116-0.24 0
0.0093-0.20 0,
0.0007-0.02 0,
0.8211 0
nil 0
0.0250-0.52(2) 0,
0.0198-0.41(2) 0,
0.0190-0.39(2) 0
O.OI65-O.3^(4) 0
0.0179-0.37(3) 0
0.0170-0.35(4) 0
0.0017-0.04(2) 0
nil 0
4.44 6
362. 369
(a) Number in brackets is number of peaks in group.
81
0.4605
9.83
3.91
16.03
99
0220-1.62
0031-0.22
0026-0.20
nil
001*3-0.32
001*3-0.32
0032-0.23
0002-0.02
9313
0005-0.01*
0111-0.83(2)
0077-0.57(2)
0085-0.62(2)
0051-0.37(3)
0055-0.1*0(1*)
0091-0.67(5)
0038-0.28(5)
0004-0.03(1)
T3
101


Abstract of Dissertation Presented to the Graduate Council in Partial
Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
GAMMA AND NEUTRON IRRADIATION
OF PURE FLUOROCARBONS
By
James Clifford Mailen
August, 1964
Chairman: Thomas M. Reed, III
Major Department: Chemical Engineering
Samples of perfluoromethane, perfluoroethane, perfluoropro-
pane, perfluoro-n-butane, perfluoro-n-pntane, perfluoro-cyclo-
pentane, perfluoro-n-hexane, perfluoro-2-methylpentane, and perfluoro-
2,5-dimethylbutane have been exposed to cobalt-60 gamma rays and the
mixed radiation of the Oak Ridge Graphite Reactor and the Oak Ridge
LITR reactor. Analysis of the resulting mixtures was by a three
column gas chromatograph allowing essentially complete analysis of
all compounds and isomers from C^ to C^ and analysis by number of
carbon atoms from Cy to C-^
In the irradiation of pure saturated fluorocarbons by gamma
rays, the important chemical reactions in the sample are recombination
of radicals, disproportionation of small radicals, addition of fluorine
and radicals to unsaturated molecules, and the reaction of radicals
with mole evil ar fluorine. These reactions differ from those found in
hydrocarbon irradiations by the absence of abstraction reactions and
xi


TABLE 18 PART 10. 2-CF3C5F1]L AND 2,3-( CF^C^Fg IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.Material
Sample Wt., Gms.
No. Days
Integrated Dose, 10 7 Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in assay or as listed
93 2-CFoCc-F-] -i
0.42&T
9.83
4.10
9h 2,3-(CF ) C.F
0.^133 2 4 8
9.83
4.13
15.T
30
Mole FractionG Value
(a)
15.3
30
Moles F2 lost/mole mixtureG Value
CFh
n2l2
c3f8
C^Fio (Mostly iso)
= 100 cc. H2(n-hex)
ll4 cc. H2(n-hex)
Vo
R
V
VR
nc5F12
i_C5F12
228 cc. H2(n-hex)
nc6fiU
2-Cf3c5F11
3"CF3C5Fll
2,3-(CF^)^CI|P8
Total ~ "
Total
Total
Total
Total C^1F24
Total
Total C^3F25
Total
z
C7Fl6
c8f18
9 20
C10F22
Cl4F30
Molecular Wt., Gas Density
(a) Number in brackets is number of peaks in group.
0.0195-1.16
0.0059-0.45
0.0019-0.14
nil
0.0018-0.13
0.0018-0.13
nil
trace
0.0024-0.18
0.0011-0.08
nil
0.0002-0.02
0.9386
nil
0.0005-0.04
0.0087-0.65(2)
0.0043-0.33(3)
0.0068-0.51(2)
0.0078-0.59(3)
0.0094-0.71(3)
0.0047-0.35(4)
0.0041-0.31(2)
nil
6.08
345.
0.0370-2.73
0.0098-0.73
0.0004-0.03
0.0006-0.05
0.0025-0.19
0.0004-0.03
0.0015-0.11
nil
0.0004-0.03
0.0021-0.16
0.0003-0.02
trace
0.0010-0.07
0.0009-0.07
0.9063
0.0160-1.18(4)
0.0128-0.94(2)
0.0065-0.47(2)
0.0082-0.60(4)
0.0048-0.35(3)
0.0159-1.17(5)
0.0032-0.23(4)
0.0063-0.47(2)
9.63
342.
102


then condensed in one of the four storage bulbs and allowed to mix by-
diffusion for at least 24 hours. Experiments performed indicate that
a representative sample is not obtained with less than about 6 hours
mixing. After the mixing period the sample is introduced into the
system with all valves closed except the one to the bulb and the one
to the manometer. Knowing the weight of the sample (determined by
weighing the sample tube with and without the sample), the pressure
read by the manometer, and the temperature it is possible to determine
the average molecular weight of the sample. After these measurements
have been made the sample is allowed to expand into the two evacuated
sampling valves (one not shown) preparatory to introduction to the
chromatographs.
Switching Columns from Series to Parallel
A schematic of the valves required for operation of the squalane
and silica gel columns in series and parallel is shown in Figure 24.
Valves A' and B1 are variable resistance and are used to replace
the resistance of columns A and B respectively when operating in paral
lel. These are necessary to maintain equal flows in the two modes
of operation. With valves 1 and 4 closed and valves 2 and 3 open the
two columns are in series and the two resistances are in series. Thus
if the resistances are correctly adjusted the two rotameters will read
equal flows. With valves 2 and 3 closed and valves 1 and 4 open the
columns are in parallel with columns A and resistance B1 connected
and column B and resistance A1 connected. The two rotameters should
now read equal flows and these should be the same as the series flows.
A rough adjustment of the resistances is made by equating these four


0(0.*,)
55
Figure 10, G^F^) versus bulk density for 30C C^g samples
(cobalt-60)(a).
Q
(a) Numbers beside points are ergs absorbed per gram X 10.
(b) The line indicates the average G value for the samples.


112
Radiation zone
Figure 28a.
Diagram for relative flux across sample zone
calculation-planar.


TABLE 18 PART 3 (continued)
Tube No.
Sample Wt., gms.
Density gm./cm.3
No. Days
Integrated Dose, 10^ Roent.
Energy Absorbed, 10 ergs
Energy Abs./gm., 10 ergs
Compound in Assay
51
0.1681
0.2015
20.17
3.66
5.47
32.5
Mole Fraction-
Moles F2/moleG Value
CFU
C2F6
C2F4
C2F2
C3F8
VlO
n-C^ig
i-c5f12
2-CF^C^
3-CF^C^F"
23-(CF3)|C#8
Total C7Fi6
Total CqFi8
Total C F
Total
Total C-| 1 F?k
Total C-,0F0ir
ZG12 26
Molecular wt., Gas Density
0.0058-0.89
0.0030-0.1+7
0.0042-0.64
nil
nil
0.9796
0.0062-0.95
0.0010-0.16
0.0022-0.34
0.0002-0.03
0.0009-0.l4
nil
0.0010-0.16
0.0007-0.11(2)
0.0007-0.11(2)
0.0003-0.05(1)
nil
nil
nil
4.05
191.9
55
0.3424
0.410
20.17
3.66
11.11
0.0070-1.07
0.0039-0.60
0.0051-0.78
nil
0.0002-0.03
0.9752
0.0064-0.98
0.0016-0.24
0.0025-0.38
0.0004-0.07
0.0009-0.14
trace
0.0009-0.14
0.0009-0.14(2)
0.0007-0.10(2)
0.0006-0.09(2)
0.0004-0.07(2)
nil
nil
4.83
191.
(a) Number in brackets is number of peaks in group.


127
of approximately 0,5 mev. (55)* This value is plotted in Figure pi
versus atomic number.
In Table 22 are tabulated the materials of interest, their
average atomic numbers, and the stopping power ratios taken from
Figure 51.
From the figures in the last column of Table 22 and equation
(K-4) we may now calculate the energy absorption per roentgen as given
by the Bragg-Gray theory. These values are summarized in Table 25.
The values are about 8 percent higher than those calculated from
simple gamma absorptions.
A final consideration in the energy absorption calculation is
the effect of density. In general the absorption of energy from a
charged particle in a dense medium is less than that in a diffuse
medium due to polarization effects. This correction amounts to about
2 percent in water exposed to 1 mev. electrons.
(50)
This correction
(35)
is related to the dielectric constant of the medium and since
(5 6)
fluorocarbons have low dielectric constants relative to water
this correction will be less than 1 percent $nd can be neglected.
Energy Absorption in Samples
The calculation of the energy absorbed in the single phase
samples is straightforward, involving only integration to find the
average flux over the tube and multiplication by the exposure time.
For the samples containing two phases the average flux must be weighted
according to the weight fraction of the flux in the liquid and gas
phases. In this calculation the liquid is assumed to occupy the
bottom portion of the tube. The calculation requires knowledge of


114
i
Radiation zone
Z-H
A
Z-h
A
h
'r
Z=Q)
Figure 28b. Diagram for calculation of relative flux
across sample zone-volume.


51
TABLE 11. CHEMICAL MECHANISM FOR C^FQ
Important mass spectrograph species 05) 0Fy CF, C3P7
Radical Formation
OsjFg Aa* C^Fy+ F
CjFg /VV ' C2F^ OF^
C^Fq VV~ * unsaturated radicals + F
Radical Recombination
F + F
CF^ + F i
c2f^ + F
CF^ + CFj
Cj-Fy + F
c2f5 + cf5-
cf4
c2f6
"*C2F6
*c5f8
>05P8
unsaturated radicals + F
5f8
c2f5 + c2f5-
->n-C4Fio
nC^Fy + CF^-
i-C5F7 + CF5-
+ n-C4F1Q
* w4fio
unsaturates + radicals or F > saturates
cf5 + ?2
c2f5-
"CF4 + F
2F6 F
O^Fg + F
2A1 + 5F2>A12F6
(0-1)
(0-2)
(o-5)
(0-4)
(o-5)
(0-6)
(o-7)
(0-8)
(0-9)
(c-io)
(0-11)
(0-12)
(0-15)
n-C-Fy + 02F^
^C5F12
(0-14)
i-C5Fy + C2F5
i5pl2
(0-15)
n-C^Fy + n-C^Fy-
n"6Fl4
(C-16)
n-CzF-. + i-C,F-,
5 7 5 7
^ 2-CF5C5F11
(0-17)
i-CjFy + i-C^Fy
-* 2,5-(cf5)2c4f8
(C-18)
(0-19)
(C-20)
(0-21)
(C-22)
(0-25)


80
Two types of initial radical generation steps appear to occur;
splitting of the ring as in equations F-2 and F-5 and renoval of a
fluorine atom from an otherwise intact ring as in F-l. Production of
fragments by F-2 and F-5 does not appear to be as important in gamma
irradiation as it is in the mass spectrometer where fragmentation of
the ring is highly dominant over removal of a fluorine atom. In gamma
radiolysis a large fraction of the products are larger than the original
material leading to the conclusion that reactions F-5 F-6, and F-7
are important.
Fluorine lost to the wall cannot be calculated for these samples
since the products are not identified. Again since the products are
unknown a statistical recombination calculation is not possible.
In the LITR irradiations the cyclic structure of this material
did not affect the equilibrium distribution of the products. For its
fluorine to carbon ratio of two the final sample contains about 25
weight percent CF^.
Perfluoro-n-Hexane
Normal perfluorohexane shows considerable difference between
the total G values for the two samples. The reason for this is uncer
tain since it could be due to either the large difference in the total
energy absorbed in the two cases or to the difference between ambient
temperature and 99C. Since no temperature effect has been noted in
any other case this is probably due to the difference in energy ab
sorption.
The chemical mechanism in Table 15 has been written in general
terms since the total number of possible reactions is very high.


the presence of reactions between radicals and molecular fluorine.
These two differences are explained by bond energy considerations.
Molecular fluorine, fluorine radicals, and other radicals can be re
moved from the reaction system by wall capture. The reaction between
radicals and molecular fluorine depends on the density, total energy
absorbed, and the distance to the sample tube wall. One of the
consequences of this type of reaction is that the G values for
molecules smaller than the parent increase with both sample density
and total energy absorption. Capture of radicals other than fluorine
by the sample tube wall becomes important at low density (below about
0.2 gm/cm^).
Y/hen fluorocarbons are irradiated to high energy absorption,
as in a nuclear reactor, chemical equilibrium considerations are
important. The equilibrium depends on the fluorine to carbon ratio
of the sample with a ratio of four resulting in nearly pure perfluoro-
methane. (Smaller ratios yield less perfluoromethane;) For this
reason, for long irradiations, perfluoromethane appears to be
extremely stable.
The initial G values for disappearance of the parent molecules
are related to each other in much the same way that the sensitivities
to electron bombardment in the mass spectrograph are related. This
is not unexpected since, in both cases, the species causing bond
rupture is electrons. Thus, given the sensitivities for a known and
an unknown fluorocarbon and the G value for the known fluorocarbon
the G value for the unknown can be estimated as the ratio of the
sensitivities times the known G value.
Xll


117
Figure 29. Arrangement of cobalt-60 rods around tube number
11 of the engineering cobalt-60 source.


115
M R + q = i(R + q cos ty) + j (q sin
|m I Vr2 + q2 cos2 9+2 Rq cos Q> + q2 sin2
'n/r.2 + q2 + 2Rq cos ^
The fraction of the center contribution received by point A
from point source is:
Jjvj|2 R + q + 2Rq cos (p
Id|2 R2 + q2 + r' 2rq(cos & cos <£> + sin (p sin &) + 2R(q cos (p r cos
The average contribution from the Area of the element is found
by integration over the area.
(J-1)
where qA = radius of element
Rp relative flux at A
The integral must also be extended over the length of the
cobalt-60 element. In Figure 28b the distances required are defined.
The relative intensity at A referred to the center is:
N2 (Z h)2 + M2
N = V7z h)2 + M2 '
P2 D2 + (Z h)2
P Vd2 + (Z h)2


APPENDIX 6. STATISTICAL RECOMBINATION DATA CORRELATION
Since the most important reactions in irradiation of fluoro
carbons appear to be recombination of fragments, on obvious method of
correlation of the data is to assume random recombination. The
statistical nature of radiation chemistry reactions has been previously
(44)
noted elsewhere.' For any given radical there are a given number
of bonds which, when broken, will yield it. This number of bonds
should be roughly proportional to the yield of this radical from
different types of molecules. It is recognized that certain bonds
will rupture more easily and certain radicals may be more reactive
toward other radicals due to stearic factors. For this reason the number
of possible ways a radical can be formed must be multiplied by an
empirical effectiveness factor. The method will be illustrated with
C2F5 as an example. This is shown in Table 50
From this correlation the fraction of the radicals which react
to give the original products can be estimated.
* A 11.5^
2F6 54.ll
555
Hence the actual G value for molecules giving radicals is
approximately the value found by analysis divided by O.667. For tube
number 68 this is
G = a 5 43
C2F6 .667
165


1 46
attained in each. Of practical interest was the accidental discovery
that thiourea (but not urea) gives good separations of the saturated
fluorocarbons and the unsaturated analogs with six carbons. The
n-hexadecane columns are not capable of this separation. Testing of
the thiourea column led to the discovery that the perfluoro-2-methylpentane
irradiated up to that time had been contaminated with a considerable
amount of unsaturated material. Following this discovery a large
thiourea column was constructed and used' to purify this isomer.
Listed in Table 27 are the operating conditions and standard
retention volumes on the thiourea column. The last unsaturate listed
was the contaminate in the 2~CF2C,-F,. .
P P 11
Mole-Area Calibration
Mole-area calibrations were run^^; using C~Fg, C^F^, C^F^g,
and In all cases the data is found to show that the area under
the curve on the chromatogram is proportional to the number of moles
of the material. For this to be true, a plot of log (no. moles)
versus log (area) should have a slope of unity and the various materials
should define a single line. This is illustrated in Figure 55 where
the data for C^Fg, C^F^ and C^F^^ is plotted. A least squares fit
of this data gives a slope of 1.004 which is within experimental
accuracy of unity slope.
Fluorine Balance
A material balance method for determination of fluorine lost
to the wall was used in this study,


179
TABLE 56. RESULTS OF RANDOM RECOMBINATION CALCULATIONS
FOR 2,5-(0F5)2C4F8
Fr ament
1'T
E
N x E
F
14
1.57
19.2
cf5
4
1.0
4.0
2-5F7
2
.45
.90
O O
1
.O
1
O
C
4
.45
1.80
c c c -
1 1
c c
c
12
.54
4.08
c 0 c -
1 1
c c
c
2
I.O65
2.15
6 Molecule
Relative
Amount
Molecule
Relati'
Amovin
F2
568.
Total.
7
49.7
cf4
155.5.
Total
8
5.2
2p6
16.0
Total
c
9
11.2
5V
54.6
Total
10
5.2
i-C,F_
4 10
7.2
Total
11
22.4
2-CF^C4Fp
69.1
Total
12
58.6
2,;H0F5)2C4F8
245.2
^ 1018.7
percent recombination = 25.9


APPENDIX 4. DECANNING OF PILE IRRADIATED SAMPLES
Both the LITR and Graphite reactor samples were too radioactive
to handle in Reed Laboratory (the site of this investigation) and
had to be decanned in the nuclear engineering building. Figure 56
is a sketch of the apparatus used in this decanning. The sample to
be decanned is inserted in a greased section of l/8 inch tubing and
this is in turn greased and slipped into a piece of tight fitting
larger tubing as is shown. After the system is evacuated and closed
off, the sample is introduced by breaking off the end of the tube
inserted into the larger tubing. The spun glass filter serves to
slow down the escaping gases and thus minimize blowout of the solids
and it also catches any solids which may be blov/n out. The sample
is then condensed in the glass sample receiver tube and sealed off
at the constriction unless air is seen to be present (by residual
pressure at the manometer). If air is present, the sample is trans
ferred to the sample holder tube and deaeriated by alternately thawing,
freezing, and pumping. The weight of the gas sample is determined
by weighing the glass sample tube before and after the sample is
removed with a correction for the air contained in the opened tube.
The weight of the original sample is determined by weighing the
aluminum tube before and after filling with the sample.
151


152
Figure 36. System used for decanning pile irradiated samples


MATERIALS IRRADIATED
The materials which were irradiated in this work are periluoro-
methane (CF4) obtained from the Matheson Company, porfluoroethane
(C2F5) obtained from Minnesota Mining and Manufacturing Company,
perfluoropropane (C^Fg) obtained from Minnesota1 Mining and Manufactur
ing Company, perfluoro-n-butane (n-C^F^Q) obtained from Minnesota
Mining and Manufacturing Company, perfluoro-n-pentane (n-C^F^2)
prepared by J. H. Simons, ^22^ perfluoro-n-hexane (n-C^F,^) prepared
(23)
by R. D. Dresdner, T. M. Reed, III, T. E. Taylor, and J. A. Young,'
(24)
perfluoro-2-methylpentane (2-CF^C^F^^) prepared by J. A. Young,
perfluoro-2,3dimethylbutane prepared by J. A.
(24)
Young,' and perfluorocyolopentane (cy-C^F^Q) prepared by J. H.
Simons/22^
The CF4, C2Fg, C^Fg, n-C^Q, n-C^F^, cy-C^F-^, n-CgF^,
2-CFjC^Fjj, and 2,3-(CFj)2C4Fg were purified by standard gas chromato
graphic methods^2^ using either the silica gel or the n-hexadecane
packings described in Appendix 3 The 2-CF^C^F^^ and 2,3-(CFj)2C4Fg
were further purified to remove unsaturated material using the
thiourea packing described in Appendix 3
After purification the compounds were found to contain the
following amounts of impurities, identified where known. (See
Table 2.)
Of these compounds only the perfluoromethane, ethane, and
propane were available in reasonably large quantities. The perfluoro-
10


120
The plot indicates the flux at the edge of the sample container
was probably about 10 percent higher than at the center of the con
tainer. However, since the samples were generally confined near the
center of the sample space the flux should not vary more than + 5 per
cent from the value determined by the benzene-water dosimeter.
Absorbed Dose
Doses in radiological work were formerly expressed in roentgens
exposure. A preferable method is the absorbed dose. In order to
calculate this quantity one must know what fraction of the exposed
dose is absorbed in the material under study.
If wall effects are neglected the absorbed dose may be found
(29)
from the following equation,
abs
0.875 O^f) x N medium x e+ 6 9 ^medium
(X-l)
N air + eY* e/(air
where:
W average energy expended by the ionizing particles per
ion pair produced in air
(50)
= approximately 55*55 +1*5 percent
N = number of electrons per gram
eira> e'T'f e/(=< true gamma absorption coefficients per electron
Dabs = rads absorbed per roentgen exposure
The values for the absorption cross sections per atom for
twenty-four elements are tabulated in reference 51* The sum of these
is listed as/i in Table 19 where is the number of electrons per atom,


TABLE l8 PART 5. C^Fg WITH POWDERED ALUMINUM IRRADIATED BY COBALT-60 GAMMA PAYS
Tube number = 85
Sample Wt. = 0.4237
Weight of Aluminum Powder = 0.1327 gms.
Sample Density Corrected for Aluminum Powder = 0.584 gm./cm.^
Bulk Density of Aluminum Powder = 1.22 gm./cm.3
Temperature = 99C
No. Days = 10.25
Integrated Dose = 1.52 x 10^ Roent.
Energy Absorbed in Fluorocarbon, 10 ergs = 5.74
Compound in Assay or as Listed
Mole FractionG Value
(a)
1
R
C16H34
CFU
C2F6
C2F4
C2F2
c3F8
n_c3F6
c4fio
0 = 54.5 cc. H2 on
n-C5F12
1^5^
n-C6Fl4
27CF3?5F11
2,3-(CF3) C4Fq
IT] ^16
Total CoF,o
so818
Molecular Wt., Gas Density
(b)
0.0006-0.22
0.0004-0.15
0.0010-0.37
0.0005-0.19
0.9877
0.0007-0.26
0.0013-0.48
0.0050-1.88
0.0004-0.15
0.0010-0.37
0.0002-0.07
0.0005-0.19
0.0003-0.11
0.00037-0.14
0.00013-0.05
4.63
192.5
vo
-4
"(a) Number in brackets is number of peaks in group,
(b) This material comes out with CgFg on silica gel.


171
Table 52 (continued)
Ratio
Stat. Pred.
Experimental
$95(leaking tube)
. M.
V5pe
4.65
4.58
5.65
GFi/5F8
1.85
1.58
1.958
2V5F8
1.02
.705
.844
"-05Fi/5F8
.455
.649
.594
2-0F504y05F8
.757
1.054
1.28
H-O^ii/OjFg
.208
.524
.406
.568
.719
Total Cy/C^Fg
.525
1.568
2.16
Total Cg/C^Fg
1.84
1.595
2.25


G(2,3-(CF3)2CiiF8)
70
Bulk density, gm./cm.^
Figure 2U. G(2,3(CF3)2C^Fq) versus bulk density for 30C
C^Fg samples (cobalt-O)^s
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) The line indicates the average G value for the samples.


RESULTS OF RANDOM RECOMBINATION CALCULATION FOR C^Fq l68
RESULTS OF RANDOM RECOMBINATION CALCULATION FOR
nC^F-^o 170
RESULTS OF RANDOM RECOMBINATION CALCULATION FOR
n-C5F12 172
RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR
n_C6Fl4 174
RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR
2-CF5C5Fli 176
RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR
2,5-(cf5)2c4f8
179


169
Table pi (continued)
Ratio
Stat. Pred.
Experimental
(typical high T
high density)
VCF4
.60
1.0
c2f6/0F4
.868
1.22
Totai 04F10/OF4
1.825
1.80
n5Pl/0F4
p84
.415
2-C?5C4F8/CF4
.bo
.65
n_G6Fl4/F4
.455
.12
2-CF5C5Fu/CF4
.906
.51
2,5-(cf,)c4f8/cf4
.475
.21


16o
Values of are known for the following molecules:
Molecule
2P6
f
5.0^5)
5.8
Values of O''for other molecules and radicals of interest can
be estimated from these data.
Assume the volumes of the various atoms and groups are additive.
Volume of sphere = (l/6)7/D^
vp = (1 /6)(/7)(5.655)5 = 25.5 P
2
vcf4 (1/6)(^)(5)5 5.5
Vc2F6 (1/6)(tf)(5.8)5 = 102.4
The differences between V and V
and that between Vpp
2 F6 4
and
Vq are both the volume of a CF2 group. The values are respectively
40 and 57. It is assumed that the value of 57 will more nearly apply
for addition of volumes.
Using this value, the volume for a C^Fg molecule is:
v0 p 159.4 P
5 8
6.42 A
Then o'for C^Fg is calculated,


4
First of all, there should not be an unusual stability. This
is in contrast to the observed stability of these materials to heat
and chemical attack where stability is attributed to the shielding
effect of the fluorine atoms. In saturated, unstrained fluorocarbons
the fluorine atoms form a protective shield against attack on the bonds
by chemical agents and .free radicals. In irradiation there is no
shielding effect since the attack is by electrons and not relatively
bulky species. This expectation of ordinary stability toward radiation
is also indicated by the results of mass spectrograph studies.
V/hile the electron energies in the mass spectrograph are quite low and
cannot be expected to give the same distribution of excited species
it is obvious that a material experiencing considerable fragmentation
in the mass spectrograph will not show exceptional stability to the
higher energy electrons produced by gamma rays.
In fact, for fluorocarbons, the sensitivity in the mass spectro
graph appears to be closely related to the stability of the materials
toward gamma irradiation. In Table 1 are listed the sensitivities for
the major mass spectrograph peak relative to n-butane for the compounds
for which data are available and which were irradiated in this study.
The ratios of sensitivities to that for CF^ ane compared to the ratios
of total G values for these compounds relative to CF^. The average
difference of these two ratios is 25*5 percent and the trends of sen
sitivity are similar.
The G values listed are for fairly low conversion. From this
we conclude that the differences in initial G values for the fluorocar
bons are due to their ease of fragmentation by electrons and not to
any secondary reactions.


Density, gm./cm.^
Figure 13. G(ci|Fio) versua density for 99 C CgFg samples
(cobalt-60)
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) This is the G value for liquid C-^Fg found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c)The line indicated the average G value for the"" samples.


TABLE l8 PART 6.
Tube No.
Sample Wt., Gms.
No. Days
Integrated Dose, 10^ Roent.
Energy Absorbed, 10 ergs
Compound in Assay or as Listed
n-C4F10 IRRADIATED BY COBALT-60 gamma RAYS, T = 30c
95 (a)
0.384 to 0.213
9.83
4.0
13*73
Mole FractionG Value'0'
96
0.3642
9.83
4.02
13.07
Moles F^/mole mixtureG Value
CFh
C3F8
n-c4Fl0
n-C5F12
i-C5F12
n-C6Fi4
2-CF C Fn
3"CF3C5F
Total
Total
Total
Total
C7Fl6
C8Fl8
c9F20
C10F22
Total 02^24
Total
C12F26
Molecular Wt., Gas Density
O.Ol45-l.2
0.0046-0.51
0.0023-0.26
0.0033-0.37
0.9626
0.0022-0.24
0.0035-0.39
0.0011-0.12
0.0001-0.01
0.0019-0.21
0.0001-0.01
0.0052-0.58(2)
0.0053-0.59(3)
0.0031-0.34(4)
0.00145-0.16(4)
0.0013-0.15(4)
0.00205-0.23(2)
5.79
259.
0.0105-1.l6
0.0056-0.62
0.0024-0.27
0.0029-0.32
0.9644
0.0017-0.19
0.0037-0.41
0.0012-0.13
0.0002-0.02
0.0021-0.23
0.0002-0.02
0.0061-0.67(3)
0.0065-0.72(3)
0.0011-0.12(4)
0.0010-0.11(4)
0.00055-0.06(4)
nil
5.05
239.
VO
CD
(a) Approximately 1/3 of this sample leaked out of the sample tube prior to analysis.
(b) Number in brackets is number of peaks in group.


176
TABLE 55* RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR 2-CF-C^F.^
C
C
C
c
c
Fragment
N
E
N :< E
F
14
.30
4.20
cf5
5
1.0
5.00
c2f5
1
.45
.45
n-CjF7
l
.45
.45
2-05F7
1
.45
.45
O
1
O
1
O

l
.45
.45
c
2~5fii
2
.45 '
.90
0
t
0
1
0
i
0
0
I
1
.45
.45
c
1
0
s
0
i
0
j
c
3
.34
1.02
c
- c c c -
1
c
6
.34
2.04
c
0
1
0
0
1
0.
!
c
2
1.065
2.150
c
1
0
1
Q
0
!
c
2
1.50
5.00
c
-c-c-c-c
1.065
c
1
1.065


TABLE 6 (continued)
Tube No.
19
20
Compound
Cy-05Fio
Cy-C^F10
Compound in assay or standard
retention volume VR, cc. H2
Mol e
Fraction^
49
OF4
2F6
2f4
c2f2
3f8
cy-c5F5
^ T7 cc. ^(n-hex)
C4F10(mstly iso)
V 0
VR
100 cc. H2(n-hex)
n5F12
V -
V
v
1>5, 177 cc. Hp(n-hex)
>(n-hex)
224 cc. Hk
n6Fl4^
308 cc. H2(n-hex)
2-OF ^CcF-^2_
3-0FC5Fu
2,3-(0142G3F8
Total
Total
Total
Total
Total
Total
Total
Sample wt,
Ml '6
c8?18
C9*20
C10F22
11F24
12F26
C13F28 .
Gms. (Initial)
0.8497
0.0912
nil
0.0079
0.0258
0.0016
nil
0.0128
nil
0.0009
0.0076
nil
it
11
11
11
11
0.0023
0.0002 (1)
0.0002 (1)
nil
II
n
n
11
0.263
0.7920
0.0921
nil
0.0029
0.0302
0.0003
0.0003
0.0191
0.0032
0.0007
0.0021
0.0240
0.0007
0.0014
0.0017
0.0010
0.0010
0.0094
0.0052 (2)
0.0054 (2)
0.0026 (3)
0.0022 (5)
0.0011 (3)
0.0006 (1)
0.0007 (1)
0.214
Molecular wt., Gas Density (of gas) 118
Weight percent recovery as gas 49.8
Molecular wt. of gas from analysis '99-4
129.3
85.3
109.8
(a) Number in brackets is number of peaks in group


74
Perfluoro-n-Butane
The purification of this material by gas chromatography was
hampered by the inability to separate the two isomers. To. obtain a
reasonable separation the material on the first part of the peak, where
n-C^F^0 appears, was cut. Direct analysis by chromatography can give
no indication of the purity, but an estimate can be obtained by examina
tion of the radiation products. Of the isomers the only ones which
can be produced from n-C^F-^Q are n-C^F^ and 5CF^C^ Fir The only
isomer which can be produced from i-C^F^ is 2,5-(CF^)2C^Fg. From
mixtures of the two isomers 2-CF^C^F^ can be produced. In the analysis
of the radiation products relatively large amounts of n-C^F^ and
3-CFjO^Fjj are found along with relatively small amounts of 2-CF^C^F-^
and 2,5-(CF^)2C4Fq. If we assume that the ratios of the sums of prod
ucts from each initial isomer are the same as the ratios of the isomers
in the original material and credit the 2-CF,C_F,, to both isomers then;
P 5 11
1~4F10 .02 + .02 + .01 + .01 .06 09
n- 04F10 .11 + .12 + .19 + .21 + .01 + .02 735 *
or the n-C^F^Q is about 91*7 mole percent pure. It may be more pure
than this since some of the two minor peaks may be second generation
products.
The samples of n-C^F-^Q were not available until after the
elevated temperature cobalt-60 and the reactor irradiations were
completed so the only data available is for ambient temperature in the
engineering cobalt-60 source. It should be noted that approximately
one-third of sample number 95 leaked out between canning and decanning
for analysis. This does not appear to have affected the G values


TABLE 18 PART 3 (continued)
Tube No.
56
61
62-
Sample Wt., gms.
Density gm./cm.
0.3496
0.5322
0.4977
0.418
0.637
0.597
No. Days
10.25
10.25
10.25
Integrated Dose, 10 Roent.
Energy Absorbed, 10^ ergs
Energy Abs./gm., 10 ergs
1.86
1.86
1.86
5.78
8.80
8.25
16.5
16.5
16.5
Value'a)
Compound in Assay
Mole FractionG
57
0.36l8
0.U33
10.25
1.86
5.97
16.5
Moles F2/moleG Value
CFjj
C2F6
C2F4
C2F2
C3F8
C4f10
n_C5F12
i_C5F12
n_c6Fl4
2-OF 0 F
3-CFC'F^
2,3-(CF)'cj^8
Total C^F^g
Total CqF^q
Total CqF20
Total C^qF22
Total CllF24
Total Cj2F26
£g
Molecular vt., Gas Density
0.0068-2.08
0.0013-0.1*0
0.0017-0.52
nil
nil
0.9897
0.0036-1.11
0.0006-0.19
0.0013-0.1*0
0.0001-0.03
0.0006-0.19
nil
0.0004-0.15
0.0004-0.12(2)
0.0002-0.07(2)
0.0001-0.03(1)
nil
nil
nil
5.29
190.7
0.0034-1.04
0.0019-0.59
0.0023-0.75
nil
nil
0.9873
0.0038-1.16
0.0008-0.24
0.0015-0.46
0.0003-0.09
0.0008-0.24
nil
0.0005-0.15
0.0003-0.09(2)
0.0003-0.09(2)
0.0002-0.07(1)
nil
nil
nil
4.97
192.3
0.0025-0.76
0.0018-0.55
0.0024-0.74
nil
0.00001-0.003
0.9887
0.0037-1.14
0.0006-0.19
0.0012-0.37
0.0003-0.09
0.0005-0.15
nil
0.0004-0.12
0.00017-0.06(2)
0.00015-0.05(2)
0.00008-0.03(2)
0.00006-0.02(2)
nil
nil
4.273
191.3
0.0030-0.93
0.0010-0.31
O.OOI8-O.56
nil
0.00004-0.01
0.9908
0.0030-0.93
0.0005-0.16
0.0013-0.40
0.00012-0.04
0.00004-0.12
nil
0.0003-0.09
0.0003-0.09(2)
0.00017-0.06(2)
0.0001-0.03(2)
0.00013-0.04(2)
nil
nil
3.77
192.
(a) Number in brackets is number of peaks in group.


6
Types of Reactions in Bulk Phase
If the excited species in fluorocarbon and hydrocarbon irradia
tion are of the same type, we can malee the following general predic
tions of the types of reactions to be expected in fluorocarbon irradia
tion.
1. Recombination of fragments. This will be of major impor
tance as has been demonstrated by Florin, Wall, and Brown^2) and by
Mastrangelo('7) in their work with fluorocarbon systems.
2. Disproportionation. Disproportionation reactions between
fluorocarbon fragments has been observed by Mastrangelo^) for the
following reactions:
2CF2 i>CF + CF~
2CF^ CF¡. + CF0
2CjFy*> CjF + CjFq
2CjFp >C^Fq + C4Fi0.
However, Pritchard, Hsia, and Miller^ found no evidence for
the disproportionation;
2C5F7- >CjFg + CjF^
Probably disproportionation is not competitive with recombination with
the exception of some small fragments.
5. Removal of fluorine by molecular processes resulting in an
unsaturated molecule. Dewhurst^) has observed molecular processes in
cyclohexane where about 15 percent of the excited cyclohexane molecules
decomposed by molecular processes to give products. In the work of
Nevitt and Remsberg^^)
about 60 percent of the hydrogen formed in the


42
TABLE 7 (continued)
Tube No.
Original Material
Compound in assay or standard
retention Volume V^, cc. H2
58 77
CjFq n-C5F12
Mole Fraction
(a)
Moles Fr
lost/mole mixture
cf4
2F6
c2f2
5F8
Cy-c^F6
94.5 on~5,6H,4
117 El6H54
n-CcFp
i-cjFia
165 on Ci^Hz4
195 on Ci6h4
250 on Ci6h4
260 on C1(5Hz4
n-C6Fl4 '
2-CFxCrF-
5-cf^c;
2,5-(c
Total
Total OgF^g
Total
Total
5*11
*11
>4F8
7*16
O10*22
Total 0UF24 ,
Total @2.2^2.6
Tot al Gi ^F2g
Wt. percent of sample recovered as gas
Sample wt., gms.
Molecular vrt., Gas Density (of gas)
^G, gaseous products
0.0482
0.0555
0.0470
0.0005
0.7701
Not determined
0.0500
nil
11
0.0117
0.0204
nil
11
11
I!
0.0042
0.0087
0.0017
0.0095
0.0059 (2)
0.0044 (2)
0.0058 (2)
0.0048 (6)
0.0017 (4)
0.0006 (5)
nil
100.
0.6180
198.
1.55
0.0455
0.0546
0.0011
0.0298
0.0004
0.0509
nil
11
0.6787
nil
0.0051
nil
0.0055
nil
0.0114
0.0107
0.0148
0.0016
0.0594
(5)
O.O562
(5
0.0088
(2)
O.O565
W
0.0096
W
0.0058
(5)
nil
92.2
0.4454
295-
1.27
la) Number in brackets is number of peales in groups.


RESULTS AND DISCUSSION
Irradiations in the Oak Ridge LITR Reactor
In the LITR irradiations the fluorocarbons were completely
degraded to an apparent equilibrium mixture. Of the gaseous products
the major constituent is CF^. Complete analyses are given in Table 6.
In addition to gaseous products solid products were formed from th¡;
materials above C^F-^q. These will be discussed further. The apparent
stability of CF4 is not surprising since the thermodynamic equilibrium
for carbon with sufficient fluorine is CF^. In Figure 5 the weight
percent CF^ in the LITR samples is plotted versus the fluorine to
carbon ratio of the original material. The data for the plot is given
in Table 5* This is seen to give a smooth curve starting with zero
percent CF^ for pure carbon and approaching 100 percent CF^ at a ratio
of four. It should be noted that the structure of the original material
is not a factor. A complete discussion of the mechanism responsible
for this equilibrium is given in the section on the GF^, results.
An interesting further study would be the irradiation of a
mixture of perfluorocyclopentane with sufficient elementary fluorine
to give a four-to-one ratio to see if this mixture behaves like CF^
in the reactor; that is, gives a resulting mixture containing about
98 percent CFj. An equally interesting experiment would be irradia
tion of a mixture of graphite powder and fluorine with an F/C ratio
of about two to see if the same solid fluorocarbon would result.
20


174
TABLE 54. RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR n-C.F .
6 14
Fragment
N
s
N x E
F
14
.50
4.2
cf5
2
1.0
2.0
2P5
2
.45
.90
5P7
2
.45
.90
n~c4F9
2
.45
.90
n-05Fu
2
.45
.90
n-0P15
6
.54
2.04
2-6p15
4
1.065
4.27
^6F15
4
1.50
6.00
Molecule
Relative
Amount
Molecule
Relative
Amount
P2
17.65
Total C
52.5
cf4
16.8
Total C
8
24.6
C2F6
11.56
Total C^
25.8
5P8
11.16
Total 0
10
25.O
n-4Fio
11.97
Total C
22.2
n-5Fl2
12.78
Total C, _
12
151.5
n-6Fi4
108.7
483.22
percent recombination = 22.5


149
Since all the fluorine present in the original material must
either be present in the products or on the wall, we may perform the
balance as follows assuming each molecule of the original yields one
molecule of product.
Ni mole p /'cent of product i in final mixture
Si ratio of fluorine to carbon in i
Ci number of carbon atoms in i
Xs ratio of fluorine to carbon ir original material
Cs number of carbon atoms in original material
n number of compounds being summed
number formed
V/ = moles Fg on wall per 100 moles mixture
which simplifies to


52
correction. From this the G values of the materials can be estimated.
These are listed in Table 7 It is seen that these G values are con
siderably less, by a factor of about three, than those found in the
cobalt-60 irradiations. This cannot be due to assuming neutron energy
deposition in fluorocarbons is the same as in graphite since the authors
state that about 82 percent of the energy results from gamma ray ab
sorption. This G value decrease may represent the approach toward
equilibrium. Aside from the G value decrease, the relative amounts of
products are close to those encountered in cobalt-60uirradiations with
the exception of fluorine lost which is about one-third of that ex
pected in C^Fq samples if its ratio to CF^ was maintained. This indi
cates that in these samples a protective fluoride coating has been
established on the wall.
Irradiations in Cobalt-60 and Detailed Discussion
The data from all cobalt-60 irradiations is compiled in Table 18.
Perfluoromethane
The most remarkable finding in the irradiation of CF^ is its
great resistance to degradation for high total energy absorption. At
low total energy absorption it is more stable than the other fluoro
carbons examined, but not unusually so and this stability is due to a
greater bond stability as discussed earlier. In the LITR where the
other fluorocarbons were completely degraded to an apparent equilibrium
mixture, OF^ remained essentially unchanged. This exceptional stabil
ity is explained as being a result of kinetic equilibrium in the LITR
irradiation section. In Figure 5 the weight percent CF^ in the samples
irradiated in the LITR is plotted versus the fluorine to carbon ratio


TABLE 18 PART 1. CF^ IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.
Sample Wt., gms.
Density, gm./cm.3
No. Days
Integrated Flux, 10 7 Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in assay or as listed
89
0.1681
0.202
9.92
1.705
2.5h
30
90
0.1718
0.206
9.92
1.705
2.60
30
Mole fractionG
83
0.161+6
0.197
9.9
1.795
2.62
Valued
9
0.1852
0.222
33.09
6.65
10.8
30
Moles F2/mole mixtureG Value
0.00153-1.10
0.0011+5-1.06
0.0008-0.55
0.001-1.73
cf4
0.99927
0.9991
0.999b
0.991+8
c2f6
0.0002-0.lb
0.0005-0.36
0.0001+-0.27
0.0025-0.1+1+
nil
nil
nil
nil
c2f2
0.00027-0.20
0.00015-0.11
trace
0.0025-0.1+1+
C3F8
0.00026-0.19
0.00025-0.19
0.00016^-0.11
0.0002-0.0l+
Cl+F10
nil
nil
nil
nil
SF12
nil
nil
nil
nil
c6fiU
nil
nil
nil
nil
£G
1.63
1.85
0.93
2.65
Molecular wt., Gas Density
88.3
87.5
88.5
93.1+
(a) Number in brackets is number of peaks in group


44
of the original sample. This is seen to give a smooth curve starting
with zero percent CF^ for pure carbon and approaching 100 percent CFj^
for ratios of four and greater. It should be noted that the structure
of the original material is not a factor.
To obtain an idea of the radicals likely to be formed in the
irradiation, we will first examine the results of mass spectrograph
work. In Table 8 are given the species produced in the mass spectro-
(pj)
graphw/ for three electron energies; 50 70 and 100 volts. From this
data the ratios of the various ions are seen to vary considerably over
even this small energy range. This indicates the fallacy in prediction
of gamma yields from mass spectrograph data. However, from this data
we see that the two fluorocarbon species which are increasing relative
to 0F^ axe OF and CFg. From this we may expect that these species are
even more prevalent when the electrons have the much higher energies
encountered in gamma ray irradiation and for this reason they are in
cluded in the proposed mechanism for CF^.
In Table 9 is given the chemical mechanism for perfluoromethane
with the understanding that ions are probably the initial products
followed by neutralization.
As can be seen, the possible reactions for even this simple
system are very great. Reactions A-4 to A-9 and reaction A-15 result
in the regeneration of CF^. Reactions A-7 to A-9 will be important
for long irradiations where a significant amount of fluorine has built
up. These reactions (A-7 to A-9) account for the equilibrium for CF^
being nearly pure CF^. Since no C2F4 was detected from CF^ irradiations,
reaction A-15 must predominate over reaction A-12. In most cases the ,


TABLE 5. WEIGHT PERCENT CF, IN SAMPLES
IRRADIATED IN THE LITR
Tube No.
Compound
F/C ratio
Wt. Percent
CF4
8
CF4
4
98.4
10
CF4
4
97.5
4
C2f6
5
54.0
5
2F6
5
45
28
5f8
2.67
55
12
c5f12
2.40
65
14
5f12
2.40
52.1
19
5fio
2.00
51.6
20
5fio
2.00
54.2
24
-06f14
2.55
57.0
55
n-6f'l4
2.55
59.0
15
2.55
42.0
16
2-CF^C^F-^^
2.55
29.8


G(i-C5F12)
64
Bulk density, grn./cm.-^
Figure 18. G(i-C^F]_2) versus bulk density for 30C C^Fg
samples (cobalt-60)^a)'.
Q
(a) Numbers beside points are ergs absorbed per gram X 10
(b) The line indicates the average G value for the samples.


Solid Fluorocarbon from LITR Irradiations
Solids were found to be present in the materials having five or
six carbons exposed in the LITR. The solids were recovered by dis
solving the sample tubes (after removal of gases) in hydrochloric acid,
filtering the residue, and washing. The solids thus recovered were
found to have negligible radioactivity. A total of 0.55 grams of solid
material was recovered from the eight tubes (two each of n-C^F^ cy-
C^Fiq, n-CF^, and 2-CFjC^F-jj ). Since only a small amount was avail
able from each original material and since the gaseous products of all
these were virtually identical the solids were combined for study.
Solubillty
The solids were soluble in fluorocarbon solvents including Fluoro-
chemical 102 (Minnesota Mining and Manufacturing Co.) with the exception
of about 0.02 grams which was probably a residue of the sample tube.
(Fluorochemical 102 is a mixture of cyclic ethers of formula CyF^O con
taining a six membered ring.) They are apparently miscible with these
solvents since upon evaporation no crystallization was observed; instead,
a gradual increase in viscosity to solidity was noted. The solids are
nearly insoluble in acetone, benzene, and carbon tetrachloride.
Boiling point
A portion of the solid which had been held at 100C for several
days to remove the solvent was observed in a Thomas-Hoover melting point
apparatus. The material was seen to begin noticeable softening at about
l45C with the viscosity decreasing thereafter as the temperature was
increased. This is the expected behavior of such a wide mixture of
molecular species. At 194C and above a very slow bubbling of the liquid
was observed. After reaching 24l0 the temperature was gradually


62
Bulk density, gm./cm.
Figure 16 Gin-C^F^) versus bulk density for 30C C^Fg
samples (cobalt-60)(a^.
(a) Numbers beside points are ergs absorbed per gram X 10"
(b) The line indicates the average G value for the samples.


157
additional n-hexadecane column was used to analyze OjFg, C02, n-CjF6,
04^10 C^F12(isomers), isomers), C^F^(isomers). For C^F^q a
partial separation was accomplished, but only sufficient to give a
rough idea of the distribution between the two isomers.
Before introduction into the chromatographs the sample was
evaporated into a calibrated gas handling system in which it was
mixed and its average molecular weight determined. This system is
shown in Figure 55
Calibration of Molecular Weight System
In order to determine the molecular weight of a sample the
volumes of the various bulbs plus the volume of the connecting tubes
must be determined.
Figure 55 is a diagram of the decanning and storage system.
To calibrate a given bulb all other bulbs are cut off from
the system by closing their valves. The system plus the bulb is then
pumped down with valve 8 closed. Valves 8, 10, and 11 serve to define
the calibrated volumes and remain closed except as described.
Valve 7 is opened and the mercury level is read and recorded
as A. Valve 7 is closed and a small amount of air let into the
evacuated system. The levels of mercury in the leveling bulb and in
the gas burette are equalized as nearly as possible by eye and this
level is recorded as B. Valve 7 is then opened and the level attained
by the mercury in the gas burette is recorded as C. Since the surface
area in the leveling bulb is much greater than that in the burette
the corrected gas burette reading, D, is found by:
D=B+(B-C)=2B-C


TABLE l8 PART 4. C^Fg IRRADIATED BY COBALT-60 GAMMA RAYS, T = 30C
Tube No.
7(a)
2^( a)
37(a
Sample Wt., Gms.
0.4239
0.6654
0.4916
0.1311
Bulk Density, gm./cm.3
0.508
0.898
0.589
0.157
No. Days
18.94
28.9
28.9
10.0
Integrated Dose, 10 J Roent.
Energy Absorbed, 10;: ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
8.6
8.53
12.23
2.88
32.4
56.6
53.5
3.36
76.4
85.1
Mole FractionG
108.8
Valued)
25.6
Moles F2/mole mixtureG Value
CFk
C9FA
44
C3F8
n-CoF/r
cMo
n-C5F12
i-c5Fi2
n~c6Fl4
2-GF G F
3"CF3C5F11
2,3-(CF
TotalJC7Fl6
Total CqF-^q
Tota! C F
Total C^qF22
Total
Total C^2F2g
Total
£G
Molecular Wt.
C13F28
Gas Density
0.0080-0.52
0.0126-0.82
0.0126-0.82
nil
nil
0.9^16
nil
0.0159-1.03
0.0035-0.22
0.0053-0.34
0.0012-0.07
0.0029-0.19
trace
0.0026-0.IT
0.0018-0.12(3)
Incomplete
Analysis
4.30
188.6
(a) These three samples were
and do not have complete
(b) This tube was copper. (
of peaks in group.
0.0062-0.37
0.0112-0.67
0.0112-0.67
nil
nil
0.9501
nil
0.0125-0.75
0.0028-0.17
0.0043-0.26
0.0012-0.07
0.0025-0.15
0.0001-0.006
0.0016-0.10
0.0019-0.12(8)
0.0005-0.o4(i)
Incomplete
Analysis
3.376
190.6
0.0186-0.83
0.0160-0.72
nil
nil
0.9102
nil
0.0239-1.07
0.0042-0.19
0.0084-0.37
0.0028-0.12
0.0048-0.21
trace
0.0037-0.17
0.0043-0.20(8)
0.0029-0.13(10)
trace
Incomplete
Analysis
5.35
190.5
0.0020-0.40
0.0018-0.34
0.0019-0.37
nil
nil
0.9885
0.00085-0.17
0.0023-0.46
0.0085-0.17
0.0015-0.30
0.00043-0.08
0.0014-0.27
trace
0.00053-0.10
trace
2.66
187.9
VO
IO
analyzed before the addition of the high temperature squalane column
analyses of the high molecular weight compounds.
c) This sample was a single phase. (d) Number in brackets is number


180
Table 56 (continued)
Statistical
Ratio
Pred.
Experimental
f/cf4
2.4
5.74
Wop4
.1044
.1095
05V0P4
.226
.26
i-O.F /OF,
4 10 4
.047
.0412
R-CFjC^Fp/OFi^
.451
.219
Total Cy'CF^
.524
1.6l6
Total Co/CF^
.0209
1.287
Total C^/CF,
.075
.644
Total C1q/CF4
.0209
.822
Total Cu/CFa
.146
.480
Total C /OF.
12 4
.252
1.60p


60
Bulk density, gra./cm.^
Figure Hi.. GCC^F-^q) versus bulk density for 30C CgFg
samples (cobalt-6o/~ \
(a) Numbers beside points are ergs absorbed per gram X 10
(b) The line indicated the average G value for the sample


175
Table 55 (continued)
Experimental
Ratio
Stat. Pred.
#11(?0"C)
#78(99~C)
VOP4
.902
5.71
5.00
2V0F4
.728
.615
1.00
5fS/cf4
.700
.615
.800
nc4Fio/',CF4
.756
.615
.715
n-6Fl4/CP4
.755
.587
.685
2-CF5C5Fn/CF4
1.185
.452
.715
5CP5C5F11/CF4
.854
.615
.828
Total Cy/CF^
1.27
1.68
2.03
Total Cg/CF4
1.215
1.097
1.286
Total Cp/CF^
1.16
1.42
1.545
Total C10/CF4
6.01
2.00
1.60
Vc5p8
1.288
6.05
5.75
V5F8
1.45
1.65
1.25
2F6/05F8
1.04
1.00
1.25
n"4Flo/o5F8
1.08
1.00
.895
n-6F14/05P8
1.05
.652
.858
2-F55Pll/05F8
1.69
.757
.895
5-0F55f11/0;f8
1.19
1.00
1.057
Total Cy/C^Fg
1.815
2.74
2.61
Total Cg/C^Fg
1.757
1.79
I.61
Total C^/C^Fg
1.655
2.51
1.68
Total C10/C^q
8.57
5.26
2.00


177
Table 55 (continued)
Relative
Molecule
Amount
F2
17.65
0F4
25.2
2F6
12.78
5F8
10.26
n-C, F
b 10
2.9
WlO
6.48
n-5P12
7.97
2-0F5C4F9
5.55
Molecule
Relative
Amount
2-0F55Fll
87.0
Total Q-j
57.6
Total CQ
8.9
Total
17.9
Total 010
10.1
Total C
25.
Total C
85.5
^ 580.66
percent recombination of original fragments to
yield original material = 22.8


6l
A
o
Density, gra./cm.
3
Figure 1$. G(n-C^F]_2) versus density for 99C C3F3 samples
(cobalt-60)(a).
(a) Numbers beside point3 are ergs absorbed per gram X 10".
(b) This is the G value for liquid CjFq found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c)The line indicates the average G value for the samples.


14
Figure- 1. Electrically heated sample holder


78
TABLE 15. CHEMICAL MECHANISM FOR n-C^F12
Important species from mass spectrograph* '
OFy c2f^, C^Fy, CF, c2F4, C^
Radical Generation
*-C5Fl2 -W *ri"C5Fll + F (2~1)
2-5Fll + F (B-2)
-A/V--5- 5-c5f1]l + f (e-3)
VV CFj + n-C^Fp (E-4)
-4/V > C2F5 + n-C^Fy (E-5)
7> unsaturated radicals + F (E-6)
Radical Recombination
R1 + R2 ** R1R2 (E-7)
R! + F2* RiF + F (B-8)
unsaturated radicals unsaturated molecules (E-9)
unsaturated molecules + F, F radicals
saturated molecules (E-10)


SUMMARY AND FUTURE WORK
In summary we can say that in the irradiation of pure saturated
fluorocarbons the important chemical reactions in the sample are re
combination of radicals, disproportionation of small radicals, addi
tion of fluorine and radicals to unsaturated molecules, and the reaction
of radicals with molecular fluorine. These reactions differ from those
found in hydrocarbon work by the absence of abstraction reactions and
the presence of reactions between radicals and molecular fluorine.
These two differences are the result of bond energy differences between
hydrocarbons and hydrogen, and fluorocarbons and fluorine. Molecular
fluorine, fluorine radicals, and other radicals can be removed from the
reaction system by wall capture. The reaction between radicals and
molecular fluorine depends on the density, total energy absorbed, and
the distance to the tube wall. One of the consequences of this type of
reaction is that the G values for molecules smaller than the parent
increase with both sample density and total energy absorption. Capture
of radicals other than fluorine by the sample tube wall becomes impor-
tant at low density (below about 0.2 gm./cm.).
When fluorocarbons are irradiated to high energy absorption,
as in a nuclear reactor, chemical equilibrium considerations are
important. The equilibrium depends on the florine-to-carbon ratio
of the sample with a ratio of four resulting in nearly pure per-
fluoromethane. Smaller ratios yield less perfluoromethane. For
105


CHEMICAL NOMENCLATURE
The follov/ing are equivalent chemical nomenclature and are
used interchangeably:
perfluoromethane, CF^
perfluoroethano, C^Fg
perfluoropropane, C_F0
perfluoro-n-butane, n-C^F^0
perfluoroisobutane, i-C^F^
perfluoro-n-pentane, n-C_F^
perfluoroisopentane, i-C-F^
perfluorocyclopentane, cy-C^.F^o
perfluoro-n-hexane, n-C.F .
6 14
perfluoro-2-methylpentane, 2-CF^C_F^
perfluoro-3-methylpentane, J-CF^C^F-^
perf luoro-2,5-dimethyl butane, 2,3- (CF,) ^C^Fg
perfluoroethylene, C^F^
perfluoroacetylene, C2F2
perfluoro-n-propene, n-C^F^
perfluorocyclopropane, cy-C^F^
182


85
thero was insufficient time for additional reactor runs. In the LITR,
due to the formation of an equilibrium mixture, the small amount of
unsaturate has little effect.
The mechanism for 2-CF^C^F^^ is given in Table l6. A simplified
mechanism consisting of reactions H-l to H-9, H-ll, and H-l 2 has been
used in the random recombination calculations. These cal evil ations pre
dict a 22.8 percent recombination of the fragments to form the original
2-CF^OcjF-Q As in the other molecules the further a C-F bond is from
a CF^ group the more easily it is ruptured. All radicals formed by
rupturing carbon-carbon bonds, excepting CF^, are equally effective
indicating that for radiolytic rupture all carbon-carbon bonds are
equivalent. Hence, as in the other cases, some molecules must rupture
extensively to yield more than one single-carbon group.
Perfluoro-2,3-Dimethylbutane
A sample of this material was not available until late in the
program so only one sample was irradiated in the engineering cobalt-00
source at ambient temperature.
A general mechanism is written in Table 17 for this material
since the possible reactions are numerous. Although no mass spectrograph
data is available for this material it is probable that owing to the
large number of CF^ groups CF^ would be very predominant. A possible
extensive shattering of the molecule is given below.
C-C-C-C
I | -0 > 4CF5 + 2CF
CO y


APPENDIX 5. RADICAL TRANSPORT
Diffusion Coefficient of Radicals in .Perfluoronropane
The escape of radicals to the wall may be dependent on either
the mean free path of the radicals in the bulk or on the diffusion
coefficient of the radicals in the bulk medium. If the half-life of
the radicals is long enough for actual diffusion to take place the
diffusion coefficient will apply; if escape to the wall is essentially
escape from the cage of surrounding molecules with subsequent flight
to the wall with few molecular collisions, the mean free path will
give an estimate of the relative probability of escape to the wall.
It must be admitted that neither of these two quantities can be
calculated with any great precision.
The diffusion coefficient can be estimated by Gilliland's
(Mo)
equation using the data of Le Bas to estimate the values of Vh.
D12 "
(L-l)
where
T absolute temperature, K
M => molecular weight
P pressure, atmospheres
Vb * molar volume
Dir = diffusion coefficient, cm. /sec.
155


109
Figure 27. Flux in tube 11 of the engineering cobalt-60
source.


15 6
Statistical Model for Calculating Mean Free Path
The following symbols which will be used in the derivation
are defined.
Pj. =* probability of passing through the Nth shell
collision diameter of the particle for which the mean
free path is being calculated
d = average distance between molecules in the medium
N = number of shells of medium molecules passed through
by the particle
D = mean free path
m A
Consider the particle in the average distribution of molecules.
This distribution is a series of equilateral tetrahedrons of side
length d.
Starting from a point of one tetrahedron the free area of the
first shell of molecules is (\/$1/2)d away.
Let us examine one triangle of one of the average tetrahedrons.
The total area of the triangle is:
The area occupied by the molecules plus that excluded to pas
sage by the diameter of the particle is
since each molecule belongs to 6 triangles in a planar projection.


TABLE 18 PART 7* n-c5F12 IRRADIATED
Tube No. 13
Sample Wt., Gms. 0.3762
No. Days 33.2
Integrated Dose, 10' Roent. 157
Energy Absorbed, 10 ergs 52.7
Temperature, C 30
Compound in Assay or as Listed
Moles F2/mole mixtureG Value
CFU
C2F6
C2F2
C3F8
Cy-a3F6
n"c4F10
n"C5F12
i-C^F12
nc6FlU
2_cf3C5F11
3-Cf3C5f11
2,3-( CF^oCi F_
Total"-2-1' 8
Total
Total
Total
Total C^F^
Total C22F2g
Total C^2?28
2 G
Molecular Wt., Gas Density
^7Fl6
C8Fl8
C9F20
C10F22
0.0305-0.68
0.0163-0.36
0.0099-0.22
0.0004-0.01
0.0106-0.23
trace
0.011*6-0.33
0.8758
0.0015-0.04
0.0066-0.15
0.0062-0.14
0.0075-0.17
nil
0.0091-0.20(3)
0.0102-0.22(3)
0.0123-0.27(2)
0.0152-0.33(2)
0.0026-0.06(4)
0.0012-0.03(4)
0.00026-0.006(2)
3.446
303.
(a) Number in brackets is number of peaks in group.
COBALT-60 GAMMA RAYS
11
0.3797
9.9
4.8
16.3
30
Mole FractionG Value
78
0.4384
9.9
3.65
14.35
0.0160-1.15
0.0043-0.31
0.0026-0.19
nil
0.0026-0.19
nil
0.0026-0.19
0.9539
nil
0.0016-0.12
0.0019-0.14
0.0027-0.19
nil
0.0072-0.52(3)
0.0047-0.34(3)
0.0061-0.44(2)
0.00864-0.62(2)
0.00091-0.06(4)
0.00021-0.01(2)
nil
4.47
301.
0.0105-1.05
0.0035-0.35
0.0036-0.35
nil
0.0028-0.28
nil
0.0025-0.25
0.9568
nil
0.0024-0.24
0.0025-0.25
0.0029-0.29
nil
0.0073-0.73(3)
0.0045-0.45(3)
0.0047-0.47(2)
0.0056-0.56(4)
0.0006-0.06(4)
0.0002-0.02(1)
nil
5.35
294.
V)
VO


79
TABLE 14. CHEMICAL MECHANISM FOR cy-C^F^
(.5)
Important species in mass spectrographx '
o5f5, c2f4, cf, cf5, c4f7, c5f?, cf2, c5f9
Radical Generation
(F-l)
(F-2)
(F-5)
(F-4)
(f-5)
(F-6)
R1 + R2~*R1R2 (F-7)
R1 + f2~RF + F (F-8)
unsaturated radicals unsaturated molecules (F9)
i
unsaturated molecules + F, F2,R saturated molecules (F-10)


g(-csf12)
65
Density, gm./cmj
Figure 17. G(i-C£F}_2) versus density for 99C C3F3 samples
(cobalt-60)(a). .
O
(a) Numbers besid points are ergs absorbed per gram X 10",
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid
density,
(c) The line indicates the average G value for the samples.


155
From these numbers we observe a large difference in the dif-
fusivities of the large radicals compared to fluorine, but not very
large differences among the large radicals. From this we conclude that,
if diffusion controls the transport of the radicals to the wall, more
fluorine radicals than any other will reach the wall and the amounts of
other radicals will not differ greatly among themselves if the average
radical concentrations are nearly equal.
If we assume the fragment effective concentrations given in
Appendix 6 for C^Fg reflect the actual average fragment concentrations
we may approximate the relative amounts of each radical that diffuses
to the wall as the product of the diffusivity and this effective con
centration. These products are listed below.
Radical
CF^
c2f5
c5F7
Product of D and effective cone,
relative to F
1.0
0.559
0.107
0.4o
Hence it appears that fluorine radicals definitely have the
advantage so far as wall capture is concerned.
To calculate the mean free path, it was necessary to develop
an equation allowing calculation from molecular diameters and densities
of the bulk. This development follows in the next section.


The procedure for calculating these numbers for the other
fluorocarbons is similar. The resulting data is listed in Tables 51
through 56.
In this treatment the values of the empirically determined
effectiveness factor were found to be about the same for most radical;
formed by similar processes. These are listed below:
Type of Radical and
Method of Formation
F from C-F bond breakage
except C4F1Q, 2,3-(CFj)2C^Fq
F from C-F bond breakage
in C4F10
F from C-F bond breakage
in 2,5-(CF5)2C4F8
CF, from C-C bond
P
Other radicals from C-C bond
breakage
Fluorocarbon radical from rupture
of CF bond on CF^ group
Fluorocarbon radical from removal
of F from carbon next to
CF_ group
P
Fluorocarbon radical from removal of
F from carbon two down from
CF^ group
E (Effectiveness)
0.50
1.00
' 1.57
1.0
0.45
0.54
1.065
1.50
Althoughthe correlation with experimental data leaves much to be
desired, the fit is close enough to draw fairly good conclusions from
the empirical effectiveness numbers.
1. The fluorine atoms on the CF^ groups are much more diffi
cult to remove than those on CF^ groups and those on CF^ groups
furthest removed from CF groups are the easiest to remove. This
P


GAMMA AND NEUTRON IRRADIATION
OF PURE FLUOROCARBONS
By
JAMES CLIFFORD MAILEN
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
August, 1964

ACKNOWLEDGEMENT
This work was done under contract to the United States Atomic
Energy Commission. The author also received financial support in the
form of a fellowship from the National Science Foundation. The
financial support of these two agencies is greatly appreciated.
The author is indebted to many people for assistance in the
completion of these studies but especially to Dr. T. M. Reed, III, for
his beneficial direction of the project, Mr. Shashi Dat for his
assistance in the early stages of the work and Dr. J. H. Simons for
many helpful suggestions in the course of the research. Thanks are
due Dr. Merril Y/ilcox for obtaining the infrared absorption spectra
of the solid fluorocarbon. The other members of the Ph.D. committee,
Dr. Mack Tyner, Dr. R. J. Hanrahan, Dr. T. F. Parkinson, and Dr. H. A.
Meyer (deceased) have also lent their kind support.
ii

TABLE OF CONTENTS
Page
INTRODUCTION 1
MECHANISM 5
Free Radical Formation 5
Reactions and Mass Spectrograph Sensitivities ..... 5
Types of Reactions in Bulk Phase 6
Reactions with Wall 8
MATERIALS IRRADIATED 10
IRRADIATION OF SAMPLES 15
CANNING 17
ANALYSIS 18
RESULTS AND DISCUSSION 20
Irradiations in the Oak Ridge LITR Reactor 20
Solid Fluorocarbon from LITR Irradiations 25
Irradiations in the Oak Ridge Graphite Reactor .... 50
Irradiations in Cobalt-60 and Detailed Discussion ... 52
Perfluoromethane 52
Perfluoro ethane h~¡
Perfluoropropane ..... 4-9
Perfluoro-n-Butane 7^
Perfluoro-n-Pentane 75
Perfluorocyclopentane 77
Perfluoro-n-Hexane 80
Perfluoro-2-Methyl pentane 82
Perfluoro-2,5-Dimethylbutane 85
SUMMARY AND FUTURE WORK .......... 105
Appendices 106
APPENDIX 1. ENERGY ABSORPTION 107
Flux Determination 107
Flux Exposure of Samples 108
Variation of Flux Across Sample Zone 110
Absorbed Dose 120
Energy Absorption in Samples ..... 127
Energy Absorbed in Reactor Irradiated Samples 152
iii

APPENDIX 2.
SAMPLE CLEANUP AND CANNING
Page
APPENDIX 5. ANALYTICAL ...
Physical Description of Analytical Equipment . .
Calibration of Molecular Weight System
Sample Treatment
Switching Columns from Series to Parallel
Description of Analytical Columns .....
Urea and Thiourea Columns
Mole-Area Calibration
Fluorine Balance
135
156
156
137
139
140
142
142
146
146
APPENDIX 4. DECANNING OF PILE IRRADIATED SAMPLES 151
APPENDIX 5- RADICAL TRANSPORT 153
Diffusion Coefficient of Radicals in Perfluoropropane 155
Statistical Model for Calculating Mean Free Path . 156
Mean Free Paths of Radicals in Perfluoropropane ... 158
APPENDIX 6. STATISTICAL RECOMBINATION DATA CORRELATION . I65
CHEMICAL NOMENCLATURE 182
ABBREVIATIONS AND DEFINITIONS 18J
BIBLIOGRAPHY 184

1
2
5
4
5
6
7
8
9
10
11
12
15
l4
15
16
17
18
LIST OF TABLES
Page
COMPARISON OF MASS SPECTROMETRIC SENSITIVITIES AND
C- VALUES
IMPURITIES IN IRRADIATED SAMPLES '
WEIGHT PERCENT CF4 IN SAMPLES IRRADIATED IN THE LITR
RESULTS OF MOLE FRACTION DETERMINATION ON SOLID
FLUOROCARBON .
GASEOUS PRODUCTS FROM THE THERMAL DECOMPOSITION OF
THE LITR SOLID FLUOROCARBON
LITR IRRADIATIONSANALYSIS OF GASEOUS PRODUCTS . .
RESULTS OF OAK RIDGE GRAPHITE REACTOR IRRADIATIONS .
MASS SPECTROGRAPH DATA FOR CFi^)
CHEMICAL MECHANISM FOR CF4
CHEMICAL MECHANISM FOR
CHEMICAL MECHANISM FOR C^Fq
CHEMICAL MECHANISM FOR n-C^Q
CHEMICAL MECHANISM FOR n-C^F^
CHEMICAL MECHANISM FOR cy-C^Q
CHEMICAL MECHANISM FOR n-C^F^
CHEMICAL MECHANISM FOR 2-CFzCcF..
CHEMICAL MECHANISM FOR 2,^(0?-)
PART 1. CF4 IRRADIATED BY COBALT-60 GAMMA RAYS . .
PART 2. C2F<5 IRRADIATED 3Y COBAXT-6O GAMMA RAYS .
5
11
22
26
51
55
4o
45
k6
48
51
16
78
79
81
84
85
87
88
v

Table
18
19
20
21
22
25
24
25
26
27
28
29
50
PART 5. C7Fa IRRADIATED BY COBALT-60 GAMMA RAYS,
T = 990;> 89
PART 4. C,F, IRRADIATED BY COBALT-60 GAMMA RAYS,
T = 50CP ? 92
PART 5- C^Fq WITH POWDERED ALUMINUM IRRADIATED BY
COBALT-60 GAMMA. RAYS
PART 6. n-CI.F. n IRRADIATED BY COBALT-60 GAMMA RAYS,
T = 50C 7 ^
PART 7. n-C_F12 IRRADIATED BY COBALT-60 GAMMA RAYS .
PART 8. CYCLO-C^F, n IRRADIATED BY COBALT-60 GAMMA
EATS .
PART 9. IRRADIATED BY COBALT-60 GAMMA RAYS .
PART 10. 2-CF,05F11 AND 2,5-(CF5)2C4F3 IRRADIATED
BY COBALT-60'uAMMA RAYS
GAMMA ABSORPTION CROSS-SECTIONS PER ELECTRON ....
HUMBER OF ELECTRONS PER GRAM
97
98
99
100
101
102
121
125
RADS ABSORBED PER ROENTGEN NEGLECTING WALL AND
DENSITY EFFECTS 124
ELECTRON STOPPING POWER RATIOS RELATIVE TO AIR FROM
FIGURE 51 FOR MATERIALS OF INTEREST 129
ENERGY ABSORPTION BY BRAGG-GRAY CAVITY THEORY .... 150
TEMPERATURE PROGRAM AND APPEARANCE TIMES ON THE
SILICA GEL COLUMN l4j
STANDARD RETENTION VOLUMES ON THE SQUALANE COLUMN . l44
STANDARD RETENTION VOLUMES ON THE n-HEXADECANE
COLUMN 145
STANDARD RETENTION VOLUMES ON THE THIOUREA COLUMN . 147
COMPARISON OF MEAN FREE PATHS CALCULATED FROM GAS
KINETICS AND FROM THE STATISTICAL MODEL 159
MEAN FREE PATHS FOR RADICALS IN PERFLUOROPROPANE . lo2
RANDOM RECOMBINATION CALCULATION AND RESULTS FOR C2F 166
vi

RESULTS OF RANDOM RECOMBINATION CALCULATION FOR C^Fq l68
RESULTS OF RANDOM RECOMBINATION CALCULATION FOR
nC^F-^o 170
RESULTS OF RANDOM RECOMBINATION CALCULATION FOR
n-C5F12 172
RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR
n_C6Fl4 174
RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR
2-CF5C5Fli 176
RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR
2,5-(cf5)2c4f8
179

LIST Ox* FIGURES
Figure Page
1 Electrically heated sanple holder l4
2 LITR sample holder l6
5Weight percent CF4 versus fluorine to carbon ratio of
original material for samples irradiated in the
LITR 21
4 Isoteniscope used for determination of the molecular
weight of the solid fluorocarbon produced in the
LITR 25
5 Indicated mole fraction vs. solution vapor pressure
of solid fluorocarbon in n-CyF-^g 27
6 Infrared spectrogram of solid fluorocarbon from LITR
irradiations 2 9
7 G(CF^) versus density for 99G C^Fg samples 52
8 G(CF4) versus bulk density for 50C Cj-Fo samples
(cobalt-60) 55
9 G(C0F) versus density for 99C C^Fn samples
(§ooalt-60) A 54
10 C(C0Fg) versus bulk density for 30C CkFP samples
(cobalt-60) ? 55
11 G(Fo) lost versus density for 99C C-Fo samples
(cobalt-60) ? 57
12 G(F0) lost versus bulk density for 50C C_FQ samples
(cobalt-60) J.8 58
G(Cz.Fin) versus density for 99C C^Fo samples
(cooalt-60) V 59
14 G(C4Fin) versus bulk density for pOG C*Fo samples
(coBaL-t-60) . 60
15 G(n-CcF-, 0) versus density for 99G CrFo samples
(co?al£-6o) < 61
viii

Figure Page
16 G(n-C^F. ) versus bulk density for 50C C-,FQ samples
(coSal u-60) 62
17 G(i-C-F10) versus density for 99C C^Fn samples
(cobalu-60) i 6?
18 G(i-C,-F, 0) versus bulk density for JOC CAFo samples
(cooait-60) ^ 64
19 G(n-C/'F, j,) versus density for 99C C-Fo samples
(cobalt-60) ? ? 65
20 G(n-0yF,;,) versus bulk density for ^0C C-Fo samles
(coDait-60) 66
21 G(2-CF-Cc-F-,, ) versus density for 99C C,FA samples
(cobalt-50) 67
22 G(2-CF^C^F1. ) versus bulk density for 5C C^Fo
sampfei; \cobalt-60) ; . 68
25 G(2,5(GiP5)2C4F8^ versus density for 99C CF
samples'^ cobalt-60) '. ... 69
24 G(2,5-(CFz)pC,Fq) versus bulk density for 5>0C CkFn
samples^(cobalt-60) V 70
25 Sum of G values versus density for 99C 0,Fo samples
(cobalt-60) t8. .... 71
26 Sum of G values versus bulk density for 50C C^Fq
samples (cobalt-60) ; . 72
27Flux in tube 11 of the engineering cobalt-60 source 109
28a Diagram for relative flux across sample zone
calculation-planar 112
2ob Diagram for calculation of relative flux across
sample zone-volume Il4
28c Diagram for calculation of relative flux across
sample zone by infinite shell model Il6
29 Arrangement of cobalt-60 rods around tube number 11
of the engineering cobalt-60 source 117
50. Relative flux across sample zone by infinite vre
and infinite shell models 118
ix

Pare
Figure
51 Electron stopping power relative to air versus
atomic number 128
52 Sample purification and canning system ....... 154
55 Decanning and storage system 158
54 Series and parallel connections of chromatographic
columns l4l
55 Log(moles) versus log(area) for chromatographed
fluorocarbons l48
56 System used for decanning pile irradiated samples . 152
x

Abstract of Dissertation Presented to the Graduate Council in Partial
Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
GAMMA AND NEUTRON IRRADIATION
OF PURE FLUOROCARBONS
By
James Clifford Mailen
August, 1964
Chairman: Thomas M. Reed, III
Major Department: Chemical Engineering
Samples of perfluoromethane, perfluoroethane, perfluoropro-
pane, perfluoro-n-butane, perfluoro-n-pntane, perfluoro-cyclo-
pentane, perfluoro-n-hexane, perfluoro-2-methylpentane, and perfluoro-
2,5-dimethylbutane have been exposed to cobalt-60 gamma rays and the
mixed radiation of the Oak Ridge Graphite Reactor and the Oak Ridge
LITR reactor. Analysis of the resulting mixtures was by a three
column gas chromatograph allowing essentially complete analysis of
all compounds and isomers from C^ to C^ and analysis by number of
carbon atoms from Cy to C-^
In the irradiation of pure saturated fluorocarbons by gamma
rays, the important chemical reactions in the sample are recombination
of radicals, disproportionation of small radicals, addition of fluorine
and radicals to unsaturated molecules, and the reaction of radicals
with mole evil ar fluorine. These reactions differ from those found in
hydrocarbon irradiations by the absence of abstraction reactions and
xi

the presence of reactions between radicals and molecular fluorine.
These two differences are explained by bond energy considerations.
Molecular fluorine, fluorine radicals, and other radicals can be re
moved from the reaction system by wall capture. The reaction between
radicals and molecular fluorine depends on the density, total energy
absorbed, and the distance to the sample tube wall. One of the
consequences of this type of reaction is that the G values for
molecules smaller than the parent increase with both sample density
and total energy absorption. Capture of radicals other than fluorine
by the sample tube wall becomes important at low density (below about
0.2 gm/cm^).
Y/hen fluorocarbons are irradiated to high energy absorption,
as in a nuclear reactor, chemical equilibrium considerations are
important. The equilibrium depends on the fluorine to carbon ratio
of the sample with a ratio of four resulting in nearly pure perfluoro-
methane. (Smaller ratios yield less perfluoromethane;) For this
reason, for long irradiations, perfluoromethane appears to be
extremely stable.
The initial G values for disappearance of the parent molecules
are related to each other in much the same way that the sensitivities
to electron bombardment in the mass spectrograph are related. This
is not unexpected since, in both cases, the species causing bond
rupture is electrons. Thus, given the sensitivities for a known and
an unknown fluorocarbon and the G value for the known fluorocarbon
the G value for the unknown can be estimated as the ratio of the
sensitivities times the known G value.
Xll

The G values for disappearance of the parent compound are
approximately; CF4, 1.5; C2Fg, 5.25; CjFg, 4.5; n-C^F.^, 5*4;
n-C5Fi2, 4.9; cyclo-C^F10, 5.0; n-CgF^, 6.0; 2-CF^F^, 6.0;
and 2,5-(CF^)^C^Fg, 9*6. These values are only approximate due to
the dependence of certain reactions on density and amount of energy
absorbed.
The important products in perfluoromethane are perfluoroethane
and perfluoroacetylene produced in about equal amounts. For the
other materials for short irradiations the major products are saturated
and are those predicted by simple bond rupture and radical recombina
tion reactions. The results for perfluoroethane and larger molecules,
excluding cyclo-compounds, can be correlated by statistical recombina
tion calculations using empirically determined radical effectiveness
numbers. Since these effectiveness numbers are the same for radicals
formed in the same way this method can be used to predict product
distributions for other fluorocarbons. In conjunction with prediction
of overall G values from mass spectrometer sensitivities the G values
for the products can be estimated.
xiii

INTRODUCTION
Samples of fluorocarbons irradiated in a cobalt^O source, in
the Graphite reactor at Oak Ridge, and in the LITR reactor at Oak
Ridge have been analyzed by three-column analytical gas chromatog
raphy. The three columns were one meter of silica gel temperature-
programmed for compounds with from one to four carbons, 15*5 meters
of squalane on Chromosorb-P at 92C for compounds with six to about
fourteen carbons, and 16 meters of n-hexadecane on Chromosorb-P at
25C for compounds with three to seven carbon atoms per molecule.
The results of these analyses for which mole fractions and G
values have been calculated are given in Tables 6, 7 and 18. When
no values are given for high molecular weight materials it indicates
that definite peaks could not be distinguished. From C^ on UP> "this
is partially due to the overlapping of the large number of isomers
which are formed by radiation processes. This makes it impossible
to distinguish between minor base line instability and small amounts
of the compounds.
Previous to this work two irradiations of fluorocarbons have
been reported. J. H. Simons and E. H. Taylor^) irradiated C^F-^ in
the Oak Ridge Graphite Reactor and Florin, Wall, and Brown^ exposed
C7F16 SaEma rays. Both these studies were hampered by the use of
impure starting materiels and the use of relatively poor analytical
methods. In this study the compounds were purified by standard gas
1

2
chromatographic methods allowing study of pure single isomers. Analy
sis was also by gas chromatography allowing separation of most of the
materials up through six carbons.
(7)
Other studies of interest are those of Mastrangelo'' where
fluorocarbon radicals were generated by electric discharge and those
of Pritchard, Hsia, and Miller, ^ Seeger and Calvort,^ and Ayscough
and Steacie'* who generated fluorocarbon radicals by photolysis. Mass
spectrograph results^*^ are of some value in determining the impor
tant species in irradiation. These types of data are valuable in the
determination of mechanisms.

MECHANISM
Free Radical Formation
Y/hen a fluorocarbon is bombarded by electrons a probable first
step is fracture of the molecule into ionic and non-ionic fragments.
The positively charged ionic fragments may then be converted to free
radicals as is shown below.
M + e* R£ + R+ + 3e
R^ + e > Rj.
R2 + e > R2
The formation of positively charged ions is known to be impor-
/pr £\
taut from mass spectrometer work.^ The neutralization scheme is
(7)
supported by the data of Mastrangelo' who passed perfluorocyclo
butane and perfluoroethane through an electric discharge and froze out
the excited species on a liquid nitrogen finger. He failed to detect
any space charge at the finger, indicating that the ionic species are
quickly neutralized by electron capture or some other equivalent
mechanism.
Reactions and Mass Spectrograph Sensitivities
Y/hen irradiating pure saturated fluorocarbons a number of
deductions about the expected reactions can be made.
5

4
First of all, there should not be an unusual stability. This
is in contrast to the observed stability of these materials to heat
and chemical attack where stability is attributed to the shielding
effect of the fluorine atoms. In saturated, unstrained fluorocarbons
the fluorine atoms form a protective shield against attack on the bonds
by chemical agents and .free radicals. In irradiation there is no
shielding effect since the attack is by electrons and not relatively
bulky species. This expectation of ordinary stability toward radiation
is also indicated by the results of mass spectrograph studies.
V/hile the electron energies in the mass spectrograph are quite low and
cannot be expected to give the same distribution of excited species
it is obvious that a material experiencing considerable fragmentation
in the mass spectrograph will not show exceptional stability to the
higher energy electrons produced by gamma rays.
In fact, for fluorocarbons, the sensitivity in the mass spectro
graph appears to be closely related to the stability of the materials
toward gamma irradiation. In Table 1 are listed the sensitivities for
the major mass spectrograph peak relative to n-butane for the compounds
for which data are available and which were irradiated in this study.
The ratios of sensitivities to that for CF^ ane compared to the ratios
of total G values for these compounds relative to CF^. The average
difference of these two ratios is 25*5 percent and the trends of sen
sitivity are similar.
The G values listed are for fairly low conversion. From this
we conclude that the differences in initial G values for the fluorocar
bons are due to their ease of fragmentation by electrons and not to
any secondary reactions.

5
TABLE 1. COMPARISON OF MASS SPECTROMETRIC
SENSITIVITIES AND G VALUES(a)
Ratio to
Compound Sensitivity for major peale CF
Sensitivity for n-butane
Total
G Value
Ratio to
OF4
cf4
0.575
1.00
1.47
1.00
c2f6
0.955
1.66
5.24
2.2
5f8
1.71
2.97
4.5
5.06
n"c4Fl0
1.78
5.10
5.42
5.89
n"5F12
2.59
4.51
4.91
5.54
Cy-O^F10
1.75
5.01
2.98
2.05
(a) Sensitivities are from reference 6.

6
Types of Reactions in Bulk Phase
If the excited species in fluorocarbon and hydrocarbon irradia
tion are of the same type, we can malee the following general predic
tions of the types of reactions to be expected in fluorocarbon irradia
tion.
1. Recombination of fragments. This will be of major impor
tance as has been demonstrated by Florin, Wall, and Brown^2) and by
Mastrangelo('7) in their work with fluorocarbon systems.
2. Disproportionation. Disproportionation reactions between
fluorocarbon fragments has been observed by Mastrangelo^) for the
following reactions:
2CF2 i>CF + CF~
2CF^ CF¡. + CF0
2CjFy*> CjF + CjFq
2CjFp >C^Fq + C4Fi0.
However, Pritchard, Hsia, and Miller^ found no evidence for
the disproportionation;
2C5F7- >CjFg + CjF^
Probably disproportionation is not competitive with recombination with
the exception of some small fragments.
5. Removal of fluorine by molecular processes resulting in an
unsaturated molecule. Dewhurst^) has observed molecular processes in
cyclohexane where about 15 percent of the excited cyclohexane molecules
decomposed by molecular processes to give products. In the work of
Nevitt and Remsberg^^)
about 60 percent of the hydrogen formed in the

7
irradiation of cyclohexane was found to originate from the removal of
molecular hydrogen or other non-radical processes. In fluorocarbons
it appears from mass spectrograph data that similar results aro pos
sible by removing moro than one F radical from a single molecule.
4. Abstraction of F from a neutral molecule by an F radical.
Although this type of reaction has been extensively observed in hydro
carbon work, it is not to be expected in fluorocarbon irradiation for
the following reason. The bond energy for the K-H bond in hydrogen
is 104.2 k. cal./g. mole^^ and that for the C-H bond of a hydrocarbon
is about 98.8 k. cal./g. mole^^ leading to a bond energy difference
for hydrogen abstraction of about +5.4 k. cal./g. mole. For fluorene
the latest value for the F-F bond energy is 57.5 k. cal./g. mole^2^
while that for a F-C bond is about 105.4 k. cal./g. mole.^) This
leads to a bond energy difference for fluorine abstraction of -67.9
k. cal./g. mole. Unless the F atom performing the abstraction is very
energetic the abstraction cannot occur.
5. Abstraction of F from a neutral molecule by a radical other
than fluorine. In hydrocarbon work abstraction of hydrogen atoms by
the small radicals and has been
observed. Energetically it appears possible for a fluorocarbon radi
cal to abstract fluorine from a neutral molecule. However, for the
same reason given for the chemical and temperature stability of fluoro
carbons it does not appear that this reaction is important. That is,
the fluorine atoms will act as a protective barrier preventing attack
on the fluorocarbon bonds by free radicals. This conclusion is con
i'
firmed by photolysis experiments. Seeger and Calvert' found no CF4
in the photolysis of trifluoroacetone and concluded from this that

8
fluorine abstraction by CFj is not important. In the photolysis of
hexafluoroacetone, Ayscough and Steacie' also found no CF^ leading
to the same conclusion.
6. Addition of radicals across double bonds in unsaturated
moleoules. This type of reaction has been observed in hydrocarbons
by various authors. (^920) Mastrangolo^y has observed the addition
of CFj to ethylene and perfluoroethylene, the addition of CFg to per
il uoroethyl ene, and the addition of CFj to perfluoropropene.
7. The reaction of fluorocarbon radicals with molecular fluor
ine. This reaction will become more important in longer irradiations
where the fluorine concentration is not depleted by reaction with the
wall and at high density where diffusion of fluorine to the wall is
restricted. This reaction will lead to the formation of molecules of
size smaller than the parent as is shown- below;
R + ?2 RF + F.
Thus one may expect a small increase in the G values of these
smaller radicals with time. Larger molecules than the parent will not
result from this reaction.
In summary, the processes which are expected in the bulk phase
of irradiated fluorocarbons are; recombination of radicals, dispro
portionation of small radicals, addition of fluorine and radicals to
unsaturates, and the reaction of radicals with elementary fluorine.
Reactions with Wall
In addition to reactions in the bulk phase, additional reactions
can occur at the sample tube wall. These are particularly important

9
in a geometry such as was used in this work. Since the inside diameter
of the sample tubes was only 0.191 centimeters, the surface to volume
ratio is relatively high and the surface is not too distant from
the bulk phase to allow some migration of fluorine and radicals to
l,
the surface. The probability of migration of radicals to the surface
is dependent on the size of the radical, the Bize of the molecules
of the medium, and the density of the medium among other'things. Thus
one would expect fluorine radicals to migrate most easily followed by
elementary fluorine, the perfluoromethyl radical, and so on. The
results of this migration should also be most easily observed at low
density. As a qualitative estimate of the ease of migration one can
calculate the mean free path for the various radicals in the medium
of interest. In Appendix 5 this has been done for perfluoropropane.
At a specific density of the medium a radical1s mean free path becomes
less than the average distance between molecules because this average
distance is not great enough to allow passage of the radical. For
the perfluoropropane, perfluoroethane, perfluoromethane, and fluorine
radicals, these are respectively about O.J g./cm.^, 0.5 g./cm.5,
0.55 g./cm.^, and 0.55 g./cm.^. Since free radicals have relatively
short lifetimes, the radicals are effectively trapped in the bulk
phase at high density.
Reactions of perfluoromethane radicals with tellurium, lead, and
( 21 ^
bismuth metals' to form rather unstable compounds have been observed.
With fluorine and F atoms, one would expect displacement of
the oxygen from the oxide layer. For this reason, until a protective
fluoride layer is built up, the wall will act as a fluorine sink and
reactions of radicals with elementary fluorine will be suppressed.

MATERIALS IRRADIATED
The materials which were irradiated in this work are periluoro-
methane (CF4) obtained from the Matheson Company, porfluoroethane
(C2F5) obtained from Minnesota Mining and Manufacturing Company,
perfluoropropane (C^Fg) obtained from Minnesota1 Mining and Manufactur
ing Company, perfluoro-n-butane (n-C^F^Q) obtained from Minnesota
Mining and Manufacturing Company, perfluoro-n-pentane (n-C^F^2)
prepared by J. H. Simons, ^22^ perfluoro-n-hexane (n-C^F,^) prepared
(23)
by R. D. Dresdner, T. M. Reed, III, T. E. Taylor, and J. A. Young,'
(24)
perfluoro-2-methylpentane (2-CF^C^F^^) prepared by J. A. Young,
perfluoro-2,3dimethylbutane prepared by J. A.
(24)
Young,' and perfluorocyolopentane (cy-C^F^Q) prepared by J. H.
Simons/22^
The CF4, C2Fg, C^Fg, n-C^Q, n-C^F^, cy-C^F-^, n-CgF^,
2-CFjC^Fjj, and 2,3-(CFj)2C4Fg were purified by standard gas chromato
graphic methods^2^ using either the silica gel or the n-hexadecane
packings described in Appendix 3 The 2-CF^C^F^^ and 2,3-(CFj)2C4Fg
were further purified to remove unsaturated material using the
thiourea packing described in Appendix 3
After purification the compounds were found to contain the
following amounts of impurities, identified where known. (See
Table 2.)
Of these compounds only the perfluoromethane, ethane, and
propane were available in reasonably large quantities. The perfluoro-
10

11
TABLE 2.
IMPURITIES IN IRRADIATED SAMPLES
Compound
Mole % Impurity
CF4
small traces of C02, C2F6
2F6
small traces of C02, unknown
C5F8
0.3% cy-CjF
nC4P10
0.18^5 C02, less than 9.5^ ^^4^10
n-5F12
none detected
cy"5F12
none detected
n-C6Fi4
none detected
2-CF5C5F11
0.2k% C6F12
2,>-(0F5)2C4F8
none detected

12
propane was selected for extensive study because it could be studied
in both the liquid and gaseous states with ease (critical temperature
7190) In addition, it was the largest molecule available in
quantity and was expected to give a greater range of reaction com
plexity.

IRRADIATION OF SAMPLES
Of the materials listed in Table 2, all except the perfluoro-
n-butane and perfluoro-2,5-dimethylbutane were irradiated in the
engineering cobalt-60 source, the Oak Ridge Graphite Reactor, and the
Oak Ridge Low Intensity Training Reactor (LITR). These two materials
were not available until after the reactor irradiation phase of the
experiments had been completed and received irradiation only in the
engineering cobalt-60 source.
The flux distribution in the engineering cobalt~60 source
(tube 11) was determined by irradiation of samples of water saturated
with benzene sealed in polyethylene bottles. The energy absorbed is
determined by analyzing for phenol by ultraviolet spectroscopy. A
further discussion of the method and a presentation of the results is
given in Appendix 1.
In the engineering cobalt-60, runs were made at both ambient
temperature (about 500) and at 99C in an electrically heated con
tainer. For details of the electrically heated container see Figure 1.
The reasons for operating at these two temperatures were two-fold.
First, this was necessary to see if there are any large effects of
temperature on yield and secondly, by raising the temperature above
the critical point of C^Fg it is possible to study the reactions of
C^Fg in a single phase.
The samples to be irradiated in the Oak Ridge Graphite Reactor
were placed in hole //4l and irradiated for 55 days. The thermal flux
15

14
Figure- 1. Electrically heated sample holder

15
in this position is 10 n/cm. -sec. and the temperature is about
100C. The energy absorption in this position can be estimated from
the data of D. M. Richardson, A. 0. Allen, and J. V*'. Boyle^2<^ as
about 8.5 x 10* erg/gram for graphite. Assuming the value for
fluorocarbons is not greatly different from this approximate G values
can be calculated (see Appendix l).
The samples irradiated in the LITR were in hole $A and were
irradiated for a period of 28 days at a thermal flux of 10-^ n/cm.2-
sec. In the LITR the samples are placed in a flow of coolant water and
have a temperature of approximately 1000. See Figure 2 for a sketch
of the sample holder. No estimate of the energy absorbed in the LITR
samples can be given; however, since the samples appear to have reached
equilibrium distributions, this would be of little value.

16
Figure 2. LITR sample holder

CANNING
Before irradiation, the samples were treated in vacuum to
remove air, carbon dioxide, and water by passage through absorbents
and alternate freezing, pimping, and vaporization. After cleanup,
the samples were sealed in aluminum tubes 11.5 inches long by 0.0752
inches inside diameter by heliarc following collection in the sample
tubes by freezing in liquid nitrogen. One sample of C^Fg, number 51
was sealed in an identical fashion in a copper tube 12-1/8 inches
long by 0.0664 inches inside diameter. A further description of this
procedure and the apparatus is given in Appendix 2.
17

.ANALYSIS
All the samples except numbers 55 7 51 and 29 were analyzed
on a three-column gas chromatograph consisting of 16 meters of n-
hexadecane at room temperature, 15.5 meters of squalane at 92C, and
one meter of silica gel. The silica gel column was temperature
programmed. For details on the columns see Appendix 5 The samples
mentioned above were analyzed without the high temperature squalane
column. The silica gel column was used to separate air, perfluoro-
methane, perfluoroethane, perfluoroethylene, carbon dioxide (perfluoro-
ethylene and carbon dioxide appear together), perfluoroacetylene,
perfluoropropane, perfluorocyclopropane, perfluoro-n-propene, and
perfluorobutane (no separation of perfluorobutane Isomers). The n-
hexadecane column was used to separate perfluoropropane, perfluoro
butane (partial separation of isomers), perfluoro-n-propene, the
perfluoropentane isomers, the perfluorohexane isomers, and the per-
fluoroheptane isomers (most unidentified). The squalane column
separated the perfluorohexane isomers with the exception of the 2-
and 5-methylpentane isomers and was used to determine the total amounts
of material in each carbon number range up through occasionally,
where amounts were sufficient for detection, up through C^.
Before being introduced into the chromatographic train, the
samples were decanned directly into an adjacent vacuum system and
stored in bulbs of sufficient size to assure complete vaporization.
After storage for at least 24 hours to assure mixing of the gases,
18

19
they were admitted to the volume calibrated vacuum system where their
molecular weight was determined and then into the chromatographic train
by gas sampling valves.
The chromatographs were supplied with thermal conductivity de
tectors and the output was recorded on Honeywell Electronik recorders.
The areas under the curves were determined by use of a polar planim-
eter. Area was found experimentally to be proportional to number of
moles from C-^ to and was assumed proportional over the entire range
(Appendix 3).
The amount of fluorine lost to the wadi was calculated from the
fluorine to carbon ratio of the products (Appendix 3).

RESULTS AND DISCUSSION
Irradiations in the Oak Ridge LITR Reactor
In the LITR irradiations the fluorocarbons were completely
degraded to an apparent equilibrium mixture. Of the gaseous products
the major constituent is CF^. Complete analyses are given in Table 6.
In addition to gaseous products solid products were formed from th¡;
materials above C^F-^q. These will be discussed further. The apparent
stability of CF4 is not surprising since the thermodynamic equilibrium
for carbon with sufficient fluorine is CF^. In Figure 5 the weight
percent CF^ in the LITR samples is plotted versus the fluorine to
carbon ratio of the original material. The data for the plot is given
in Table 5* This is seen to give a smooth curve starting with zero
percent CF^ for pure carbon and approaching 100 percent CF^ at a ratio
of four. It should be noted that the structure of the original material
is not a factor. A complete discussion of the mechanism responsible
for this equilibrium is given in the section on the GF^, results.
An interesting further study would be the irradiation of a
mixture of perfluorocyclopentane with sufficient elementary fluorine
to give a four-to-one ratio to see if this mixture behaves like CF^
in the reactor; that is, gives a resulting mixture containing about
98 percent CFj. An equally interesting experiment would be irradia
tion of a mixture of graphite powder and fluorine with an F/C ratio
of about two to see if the same solid fluorocarbon would result.
20

Weight % CF, in sample
21
Figure 3.
Weight percent CF^ versus fluorine to carbon ratio of
original material for samples irradiated in the LITR

TABLE 5. WEIGHT PERCENT CF, IN SAMPLES
IRRADIATED IN THE LITR
Tube No.
Compound
F/C ratio
Wt. Percent
CF4
8
CF4
4
98.4
10
CF4
4
97.5
4
C2f6
5
54.0
5
2F6
5
45
28
5f8
2.67
55
12
c5f12
2.40
65
14
5f12
2.40
52.1
19
5fio
2.00
51.6
20
5fio
2.00
54.2
24
-06f14
2.55
57.0
55
n-6f'l4
2.55
59.0
15
2.55
42.0
16
2-CF^C^F-^^
2.55
29.8

Solid Fluorocarbon from LITR Irradiations
Solids were found to be present in the materials having five or
six carbons exposed in the LITR. The solids were recovered by dis
solving the sample tubes (after removal of gases) in hydrochloric acid,
filtering the residue, and washing. The solids thus recovered were
found to have negligible radioactivity. A total of 0.55 grams of solid
material was recovered from the eight tubes (two each of n-C^F^ cy-
C^Fiq, n-CF^, and 2-CFjC^F-jj ). Since only a small amount was avail
able from each original material and since the gaseous products of all
these were virtually identical the solids were combined for study.
Solubillty
The solids were soluble in fluorocarbon solvents including Fluoro-
chemical 102 (Minnesota Mining and Manufacturing Co.) with the exception
of about 0.02 grams which was probably a residue of the sample tube.
(Fluorochemical 102 is a mixture of cyclic ethers of formula CyF^O con
taining a six membered ring.) They are apparently miscible with these
solvents since upon evaporation no crystallization was observed; instead,
a gradual increase in viscosity to solidity was noted. The solids are
nearly insoluble in acetone, benzene, and carbon tetrachloride.
Boiling point
A portion of the solid which had been held at 100C for several
days to remove the solvent was observed in a Thomas-Hoover melting point
apparatus. The material was seen to begin noticeable softening at about
l45C with the viscosity decreasing thereafter as the temperature was
increased. This is the expected behavior of such a wide mixture of
molecular species. At 194C and above a very slow bubbling of the liquid
was observed. After reaching 24l0 the temperature was gradually

24
reduced. No bubbling was noted during cooling. This probably means
the bubbling was due to distillation of relatively low boiling materials
from the mixture.
Molecular weight
The molecular weight of the solid fluorocarbon was determined by
its effeot on the vapor pressure of c-C^F^. In Figure 4 Is a sketch
of the isoteniscope constructed for this determination. The sample
side of the isoteniscope is made of capillary tubing to minimize the
loss of solvent to the vapor space.
To operate the isoteniscope the solution of solid fluorocarbon
in is injected into the sample bulb by a hypodermic. The
sample is cooled in ice and a vacuum pulled on the two openings. Vacuum
is maintained until a portion of the sample (about 1/5) has boiled off
to dispell air from the sample. The sample is then frozen in liquid
nitrogen and the capillary pulled off at the arrows. While the sample
is still frozen the mercury is dropped into the U tube and the apparatus
is ready for use. The isoteniscope is mounted in a water bath and the
vapor pressure determined as a function of temperature. After the vapor
pressure has been determined at several temperatures, the sample bulb
is cooled and broken off. The sample is removed by hypodermic and the
concentration determined.
The mole fraction is then the vapor pressure lowering of the
solvent divided by the vapor pressure of the pure solvent. The vapor
pressure of the pure solvent has been previously measured''and is
given byj
VogCP.mo) 6.07154 r^\ t .
- 0
t

25
Figure U Isoteniscope used for determination of the molecular
weight of the solid fluorocarbon produced in the
LITfi .

2$
In Table 4 are listed the results of the experiment and the
calculated mole fraction.
TABLE 4. RESULTS OF MOLE FRACTION DETERMINATION
ON SOLID FLUOROCARBON
TC
Solution P,ram
Solvent P,mm
A
a/p0
(mole fraction)
24
81116
75-79
-7.57
-O.O998
29.9
105.06
99.74
-5.02
-O.O505
52.6
115.91
115.8
-2.11
-0.0185
5^.7
125.15
125.8
+2.7
+0.0215
59.65
155.85
157.9
+4.1
+0.026
45.75
195.27
206.0
+10.7
+0.0519
The negative results at low temperature Indicate some air must
have remained in the solution. This is possible due to the high
viscosity of the solution. As the pressure of the solvent becomes
greater, the air present becomes of less importance and the indicated
mole fraction approaches the true value. In Figure 5 is plotted the
indicated mole fraction versus the vapor pressure of the solution.
The curve approaches a value of approximately 0.055 The solution
was found to be 40.8 weight percent solid. From this data we may
now determine the average molecular weight of the solid fluorocarbon.
Let x molecular weight of solute

Calculated mole fraction
Figure 5 Indicated mole fraction vs. solution vapor pressure of solid
fluorocarbon in n-C^F^

28
.408
x
.392 .408 0.035
388 x
x 4600
Some information concerning the structure of the solids can be
obtained from infrared absorption and examination of the shorter prod
ucts determinable by gas chromatography.
Figure 6 is the infrared spectrogram of the solid material run
in a solid pellet containing a mixture of 0.00373 grams of fluorocarbon
solid and 0.95816 grams of sodium chloride. The instrument was a
Perkin-Elmer 237 Spectrophotometer. All that can be determined from
this spectrogram is that there is some unsaturation as indicated by the
peak at 6 microns. The large peak at 8 microns corresponds to various
carbon-fluorine bonds and is indicative of the complexity of the mixture.
The fluorocarbons formed in the LITR which can be analyzed on the
gas chromatograph show a preponderance of the most branched isomer ex
cluding neo types. For C^F^Q the iso is the most prevalent. For C^F^2
i-C5Fi2 predominates, and for C^F-^, 2,3-(CF^)2C^Fg is in largest con
centration. From this it is probable that the solid fluorocarbon is
highly branched but essentially lacking in neo type structures.
We can summarize the data known about the solid fluorocarbon
below;
(1) molecular weight = 4600
(2) contains some unsaturation
(3) has a large amount of branching in its structure.

Absorbance
Wave length, microns
Figure 6* Infrared spectrogram of solid fluorocarbon from LITR irradiations

50
From the above data Vie should not expect the solid fluorocarbon
to have exceptional thermal stability. Aa has been mentioned before
the thermal stability of fluorocarbons is due to the shielding effect
of the fluorine atoms. When one introduces double bonds and strained
bonds (due to the highly branched struoture) this stability is reduced.
Thermal decomposition products
It was thought that some additional information concerning the
solid would be determined by thermally decomposing some of it. A sample
weighing 0.2089 grams was sealed in a Vicor tube with a volume of four
cubic centimeters and heated to 470C for a period of about 18 hours.
At the end of this time, the gaseous portion of the sample was removed
for analysis on the gas chromatograph. The material appeared to de
compose almost completely to the gaseous products and carbon. The
gaseous products accounted for 0.1688 grams or about 81 wt. percent of
the sample.
In Table 5 the mole fractions of the gaseous products are tabu
lated. Of especial interest is the fact that, like teflon, this
material decomposes largely into perfluoroethylene.
Irradiations in the Oak Ridge Graphite Reactor
The irradiations in the Oak Ridge Graphite Reactor represent
energy absorptions intermediate between those obtainable in the engi
neering cobal-b-60 source and the energy absorbed in the LITR. In the
latter, as has been pointed out, an apparent equilibrium condition was
reached.
The energy absorbed by the samples in the Oak Ridge Graphite
Reactor can be estimated rrom the data of Richardson, Allen, and
Boylefor energy deposition in graphite with an appropriate flux

51
TABLE 5. GASEOUS PRODUCTS FROM THE THERMAL DECOMPOSITION
OF THE LITR SOLID FLUOROCARBON
Initial Wt., gms. = 0.2089
Wt. Gaseous Products, gms. = 0.1688
Remainder (0.0401 gms.) is carbon with small amount
of soluble solid
Decomposition Temperature = 470C
" Time = 18 hrs.
" Tube material = Vicor
" Tube dimensions;
Length =11.2 cm.
I.D. = '0.675 cm.
Volume = 4.01 cm.5
Compound in assay
Mole fraction
CF4
2f6
C2F4
c2f2
C.Fq
cy-c5F
n-C5F6
4F10
25.0 min. on silica gel
26.2 min. on silica gel
n-C5Fi2
_
198, 203 cc
- 269
V
cc
Ho(n-hex)
. H2(nhex)
20F^n
11
^"7 5 5
2,5-(cF^
Total
Total
Total
Total
,c4f8
c7*l6
C9F20
Cif22
Total C^3_F24
Molecular wt., gas density
0.0006
0.0025
0.7723
0.0019
0.0154
0.0006
0.0012
0.0517
0.0025
0.0012
0.0027
0.0151
0.0811
0.0151
0.0081
0.0043
0.0014
0.0005 ( \
0.0131 (3)u;
0.0039 (3)
0.0019 (2)
0.0017 (2)
0.0011 (2)
95-4
(a) Number of peaks in group

52
correction. From this the G values of the materials can be estimated.
These are listed in Table 7 It is seen that these G values are con
siderably less, by a factor of about three, than those found in the
cobalt-60 irradiations. This cannot be due to assuming neutron energy
deposition in fluorocarbons is the same as in graphite since the authors
state that about 82 percent of the energy results from gamma ray ab
sorption. This G value decrease may represent the approach toward
equilibrium. Aside from the G value decrease, the relative amounts of
products are close to those encountered in cobalt-60uirradiations with
the exception of fluorine lost which is about one-third of that ex
pected in C^Fq samples if its ratio to CF^ was maintained. This indi
cates that in these samples a protective fluoride coating has been
established on the wall.
Irradiations in Cobalt-60 and Detailed Discussion
The data from all cobalt-60 irradiations is compiled in Table 18.
Perfluoromethane
The most remarkable finding in the irradiation of CF^ is its
great resistance to degradation for high total energy absorption. At
low total energy absorption it is more stable than the other fluoro
carbons examined, but not unusually so and this stability is due to a
greater bond stability as discussed earlier. In the LITR where the
other fluorocarbons were completely degraded to an apparent equilibrium
mixture, OF^ remained essentially unchanged. This exceptional stabil
ity is explained as being a result of kinetic equilibrium in the LITR
irradiation section. In Figure 5 the weight percent CF^ in the samples
irradiated in the LITR is plotted versus the fluorine to carbon ratio

55
TABLE 6. LITR IRRADIATIONSANALYSIS OF GASEOUS PRODUCTS
Position => C-4l Flux = 10^ n/cm.^-sec.
No. days 28 Irradiation Temperature = 100C
Tube No.
15
16
Compound
2-CFjC^F-^^
2-CF5C5Fn
Compound in assay or standard
Mole
Fraction^
retention volume V^, cc. H2
vRo
cf4
C2F6
G2F4
2?2
CjFg
cy-CzF8
C4F1Q(mostly iso)
?1 cc. H2(n-hex)
100 cc. H2(n-hex)
121 cc. Hp(n-hex)
VR
'R
n-C5F12
-C5F12
224 cc. H2(n-hex)
n"G6Fl4
** 508 cc. Hp(n-hex)
2-CF,C5F11
5-CF3C5F11
2,5-(o^2C4F8
Total
Total
Total
Total
Total
Total
Wi
8F18
9Fg0
10f22
c11f24
Cl oFo6
ivotu. ^12r26
Sample wt., Gms. (Initial)
0.8546
0.1045
nil
0.0062
0.0289
0.0010
0.0159
nil
nil
nil
0.0008
0.0065
nil
trace
nil
2 + 5 =
0.0008
O.OOI9-
0.0005 (2)
0.0005 (2)
0.0001 (1)
nil
n
ti
0.1947
Molecular wt., Gas Density (of gas) 108.5
Weight percent recovery as gas 57.7
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis 100.5
(a] Number in brackets is number of peaks in group.
0.7595
0.1426
nil
0.0026
0.0408
0.0005
0.0111
0.0005
0.0019
0.0004
0.0024
0.0217
0.0004
0.0007
0.0009
0.0015
0.0015
0.0084
0.0047
(2)
0.0060
(2)
0.0048
(4)
0.0060
(5)
0.0012
(4)
0.0009
(1)
0.619
117.5
55.8
117.8
Notej In second samples, more care taken to recover higher boilers
hence the high m.w. analyses are more nearly correct.

TABLE 6 (continued)
Tube No.
19
20
Compound
Cy-05Fio
Cy-C^F10
Compound in assay or standard
retention volume VR, cc. H2
Mol e
Fraction^
49
OF4
2F6
2f4
c2f2
3f8
cy-c5F5
^ T7 cc. ^(n-hex)
C4F10(mstly iso)
V 0
VR
100 cc. H2(n-hex)
n5F12
V -
V
v
1>5, 177 cc. Hp(n-hex)
>(n-hex)
224 cc. Hk
n6Fl4^
308 cc. H2(n-hex)
2-OF ^CcF-^2_
3-0FC5Fu
2,3-(0142G3F8
Total
Total
Total
Total
Total
Total
Total
Sample wt,
Ml '6
c8?18
C9*20
C10F22
11F24
12F26
C13F28 .
Gms. (Initial)
0.8497
0.0912
nil
0.0079
0.0258
0.0016
nil
0.0128
nil
0.0009
0.0076
nil
it
11
11
11
11
0.0023
0.0002 (1)
0.0002 (1)
nil
II
n
n
11
0.263
0.7920
0.0921
nil
0.0029
0.0302
0.0003
0.0003
0.0191
0.0032
0.0007
0.0021
0.0240
0.0007
0.0014
0.0017
0.0010
0.0010
0.0094
0.0052 (2)
0.0054 (2)
0.0026 (3)
0.0022 (5)
0.0011 (3)
0.0006 (1)
0.0007 (1)
0.214
Molecular wt., Gas Density (of gas) 118
Weight percent recovery as gas 49.8
Molecular wt. of gas from analysis '99-4
129.3
85.3
109.8
(a) Number in brackets is number of peaks in group

55
TABLE 6 (continued)
Tube No.
Compound
8
cf4
0
i-1
4=- O
Compound in assay or standard
retention volume V^, cc. H2
Mole
Fraction^
CF4
0.9866
0.9809
C2F6
0.0077
0.0077
C
nil
nil
C2f2
0.0055
0.0091
5f8
0.0002
0.0001
cy-CjF6
nil
0.0001
c4fio(mos^iy is0)
V_ = lOO cc. H2(n-hex)
it
0.0021
n
nil
VR
(n-hex)
n-CcF-i o
i_C5F12
224 cc. H2
n-c6Fl4",
508 cc. Hp(n-hex)
2-CF5C5F1i
5-CF^C^F11
2,5-(c#3?2C4F
TotaTCyF^
Total CF^q
Total GgF^Q
Total 0]_oF22
Total C*|t Fpii
Total C12F26
Total
Total
Total
Total
Total
Total
Total
5F28
C14
p15
16
c17
c18
19/
Sample wt., Gma. (initial)
Molecular wt., Gas Density (of gas)
Weight percent recovery as gas
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis
0.1752
95.6
100
0.0245
88.24
0.1804
145.2^
100
0.04i4
08.46
(a)Number in brackets is number of peaks in group.
(b) This average molecular weight is obviously grossly in error.
r

56
TABLE 6 (continued)
Tube No.
4
5
Compound
g2F6
c2f6
Compound in assay or standard
retention volume VR, cc. H2
Mole
Fraction^
cf4
02f6
0.7467
0.1005'
0.7244
0.1048
V
V
C 2F4
2F2
c3f8
cy-6*F6
O4P10(mostly iso)
V^0 100 cc. ^(n-hex)
n-C5Fi2
i-CcFi2
224 cc. Ln-hex)
n-C6Pl4
508 cc. Kp(n-hex)
2-CF5C5F1f
5-c?iciFu
2,MCF)2C4F8
Total CyF^g
Total CgF^g
Total CpFpo
Total Ci f)Fpp
Total C^^F24
Total 0^2^26
i5f28
cl4
15
Total C-j_g
Total C]_y
Toted C^q
Total
Sample wt., Gms. (initial)
Total
Total
Total
nil
0.0047
0.0594
0.0007
0.0271
nil
0.0050
0.0582
nil
0.0009
nil
0.0022
0.0015
0.0180
0.0055 (2)
0.0062 (2)
0.0022 (5)
0.0059 (5)
0.0018 (5)
0.0005 (2)
nil
11
11
11
11
11
11
0.2887
nil
0.0047
0.0520
0.0009
0.0184
nil
0.0014
0.0218
nil
0.0006
nil
0.0014
0.0012
0.0122
0.0048
(2)
0.0042
(1)
0.0129
W
0.0157
(5)
0.0095
(5)
0.0092
(7
O.OO55
W
0.0017
W
0.0007
(2)
0.0015
(1)
0.0054
(2)
0.0061
W
0.0025
(1)
0.505
Molecular vrfc., Gas Density (of gas)
122.7
155
Weight percent recovery as gas
100
100
Moles ?2 lost/mole mixture

-
Moleculeir wt. of gas from analysis
121.9
148.8
(a) Number in brackets is number of peaks in group

57
TABLE 6 (continued)
1
Tube No.
28
12
Compound
5f8
n~C5Pl2
Compound in assay or standard
retention volume VR, cc. H2
Mole
Fraction^
cf4
C2F6
2f4
C2F2
c5f8
cy-CzF^
C4F10(mostly iso)
Vg 100 cc. ^(n-hex)
n-CcF-i p
i-Ccpi2
VR 224 cc. K2(n-hex)
n-CFl4
VR cc. ^(n-hex)
2-CF,C5F11
5-CF£tFn
2,5-(CF5j2C4F8
Total^C7Fl6
Total C8F18
Total
Total
Total CTtFpA
Total C12F2
Total C17F2q
Total 0^4
Total C. c
Total oti
Total C17
Total C-¡_g
Total C-^q
Sample wt., Gms. (initial)
9F20
10F22
0.7295
O.1558
nil
0.0052
0.0491
0.0006
0.0194
nil
0.0018
0.0169
trace
0.0004
nil
0.0017
0.0010
0.0080
0.0044 (2)
0.0073 (2J
0.0042 (5)
0.0080 (5)
0.0055 (4)
0.0059 (5)
0.0015 (4)
nil
n
n
11
11
11
0.5110
0.8558
0.0979
nil
0.0050
0.0514
0.0006
0.0150
nil
0.0018
0.0082
0.0005
trace
nil
2 + 5 =
0.0005
0.0059
0.0004 (2)
0.0006 (2)
0.0001 (1)
nil
11
n
it
ti
u
u
11
it
11
0.2110
Molecular wt., Gas Density (of gas)
120.5
114.5
Weight percent recovery as gas
67.6
90.2
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis
125.9
101.6
(a) Numoer in brackets is number of peaks in group

58
TABLE 6 (continued)
Tube Ho.
14
24
Compound
n-C5F12
nc6Fl4
Compound in assay or standard
retention volume V^0, cc. Hg
Mole
Fraction^
CF4
2F6
2F4
2F2
c,f8
cy-C^Fg
C4F10(mostly iso)
V 0
VR
v o
VR
i-C^F^
224 cc.
100 cc. Hp(n-hex)
5F12
2 (11-hex)
n_c6Fl4 ,
508 cc. ^(n-hex)
2CF tOcFi
5-CF0Fn
5)204
4f8
2 5 (CFz; 2'
Total C7Fl6
Total CgFl8
Total -
CnF
Total C
9 2
10F22
Total 0^2f24
Total C^2F26
' ' i?f28
Total
Total
Total
Total
Total
Total
Total
14
15
c16
18
19,
Sample wt., Gms. (initial)
0.7657
0.1194
nil
0.0055
0.0519
0.0006
0.0185
0.0001
0.0019
0.0140
trace
0.0004
nil
0.0010
0.0005
O.OO65
0.0097 (2)
0.0072 (2)
0.0025 (5)
0.0027 (5)
0.0019 (4)
0.0104 (2)
0.0004 (1)
nil
11
it
11
11
11
0.4210
0.8251
0.1030
nil
0.0017
0.0526
0.0005
0.0175
nil
0.0012
0.0082
nil
0.0001
nil
0.0005
0.0005
0.0028
0.0007 (2)
0.0007 (2)
0.0002 (5)
0.0005 (5)
0.0001 (1)
nil
n
it
11
I!
11
U
n
0.2556
Molecular wt., Gas Density (of gas) 112
Weight percent recovery as gas 56.8
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis 119.6
115
52.5
102.8
(a) Humber in brackets is number of peaks in group.

59
TABLE 6 (continued)
Tube No.
55
Compound
n-6Pl4
Compound in assay or standard
retention volume VR, cc. H2
fa"}
Mole Fractionv '
cf4
C2F6
2F4
c2f2
cy-Cf6
C4F10(mostly iso)
Vr,0 = 100 cc. H0(n-hex)
R
V-
R
n-C5F12
i~CF12
= 224 cc. H2(n-hex)
508 cc.^pL(n-hex)
2CFzGcF-i n
5-CFgc§Fn
2,5-(of512c4f8
Total^
7l6
efe
Total
Total
Total
Total
V2
Total Cn zF0o
Total XV8
Total C15
Total
Total -
Total
Total
Sample wt.,
£7
as
Gms^( Initial)
0.8017
0.1162
nil
O.OO56
0.0574
0.0005
0.0124
0.0001
0.0015
0.0105
0.00004
0.0005
nil
0.0008
o.ooo4
0.0052
0.0022 (2)
0.0050 (2)
0.0015 (2)
0.0021 (5)
0.0005 (2)
0.0002 (1)
nil
it
tt
It
It
U
II
0.575
Molecular wt., Gas Density (of gas) 107
Weight percent recovery as gas 592
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis 107
(a)Number in brackets is number of peaks in group

TABLE 7. RESULTS OF OAK RIDGE GRAPHITE REACTOR IRRADIATIONS
Position * Hole Flux 10^ n/cm.^-sec. No. Days 55
Approximate energy absorption 8.J x 10^ erg/gram
Irradiation temperature ** 100C or greater
Compound in assay or standard
retention volume V^, CC.H2
Tube No. 75
Cyclo-C5F10 .
Mole Fraction'a'
CF4
2F6
2f4
2F2
3F§
Cy-CjF^
4Fio
79 on Cl6K54
117 on C1(5Hj4
n"5F12
i-C$F12
184 on Ci^Hz4
Cy-CcFio
n-C6Fl4
2-CFiCcFt n
5-F5C5Fll
2,5-(of*5)264f8
Total CyF-,/-
Total CgFl8
Total CgF20
Total C10F22
Total C^1F24
Total Cj_ 2f2
Total C-,zF?8
Total ct^F^o
Sample wt. gms.
Molecular vrt., Gas Density
Percent recovery as gas
£o
0.0098
0.0088
0.0006
0.0015
0.0055
0.0006
0.0055
0.0007
0.0025
0.0049
0.0007
0.0202
0.8866
0.0015
0.0005
0.0008
0.0099
0.0109 (2)
0.0156 (2)
0.0090 (2)
0.0018 (4)
0.0051 (5)
0.0056 (1)
nil
0.0001 (1)
0.2559
264.
100.
0.505
(a) Number in brackets is number of peaks in group

4l
TABLE 7 (continued)
Tube No.
82
Original Material
CF4
c2F6
Compound in assay or standard
retention Volume V^, co. H2
Mole Fraction^
Moles lost/mole mixture
OF4
C2F6
n2?2
5F8
Cy^c5F6
94.5 on Cl6H34
117 on Cl6K54
n-C5F12
i-05P12
165 on C-¡/cH*4
195 on O^hS
250 on C^H54
260 on G15K24
a-6pl4
2-CF5C5F11
5-CF, ^
2,5-(ci
Total G7F15
Total C8F18
Total CcjFpO
Total C]_qF22
^5Fn
5/ 2c4f8
Total
Total
Total
C11F24
12f26
15F28
lo re
recovered as
Wt. percent of sampl
gas
Sample wt., gms.
Molecular wt., Gas Density (of gas)
2 G gaseous products
0.0208
0.9905
O.OO55
0.0054
0.0004
0.0001
nil
it
u
11
11
11
11
n
11
it
11
is
n
n
n
11
n
11
it
11
100.
0.178
87.8
O.589
0.020
0.0800
0.8491
0.0012
0.0452
0.0007
O.OI56
nil
u
0.0017
0.0051
nil
11
11
11
0.0005
0.0001
0.0002
0.0020
0.005 (2)
0.0011 (2)
0.0004 (2)
0.0005 (4)
0.0001 (1)
nil
100.
O.5659
141.5
1.58
t*y
Number in brackets is number of peaks in groups.

42
TABLE 7 (continued)
Tube No.
Original Material
Compound in assay or standard
retention Volume V^, cc. H2
58 77
CjFq n-C5F12
Mole Fraction
(a)
Moles Fr
lost/mole mixture
cf4
2F6
c2f2
5F8
Cy-c^F6
94.5 on~5,6H,4
117 El6H54
n-CcFp
i-cjFia
165 on Ci^Hz4
195 on Ci6h4
250 on Ci6h4
260 on C1(5Hz4
n-C6Fl4 '
2-CFxCrF-
5-cf^c;
2,5-(c
Total
Total OgF^g
Total
Total
5*11
*11
>4F8
7*16
O10*22
Total 0UF24 ,
Total @2.2^2.6
Tot al Gi ^F2g
Wt. percent of sample recovered as gas
Sample wt., gms.
Molecular vrt., Gas Density (of gas)
^G, gaseous products
0.0482
0.0555
0.0470
0.0005
0.7701
Not determined
0.0500
nil
11
0.0117
0.0204
nil
11
11
I!
0.0042
0.0087
0.0017
0.0095
0.0059 (2)
0.0044 (2)
0.0058 (2)
0.0048 (6)
0.0017 (4)
0.0006 (5)
nil
100.
0.6180
198.
1.55
0.0455
0.0546
0.0011
0.0298
0.0004
0.0509
nil
11
0.6787
nil
0.0051
nil
0.0055
nil
0.0114
0.0107
0.0148
0.0016
0.0594
(5)
O.O562
(5
0.0088
(2)
O.O565
W
0.0096
W
0.0058
(5)
nil
92.2
0.4454
295-
1.27
la) Number in brackets is number of peales in groups.

45
TABLE 7 (continued)
Tube No.
Original Material
79
n-C6Fi4
70W
Z-CFjCjFn
Compound in assay or standard
retention Volume VR, cc. H2
Mole
Fraction^
Moles Fr
lost/mole mixture
cf4
2F6
2F2
c5f8
Cy-C5F6
94.5 on
117 on C^H*4
n"5P12
i5F12
165 oil l5H34
195 on 015H24
250 on Cl6H34
260 on C1H££
2>"G6Fl4
,Fi
2-CF5C5ill
^CF5C5Fn
2,5-(C?5}2C4F8
Total d7Fl6
Total CnF,g
Total CoFq
1^ ufe
Total C-qF24
Total C,2F2
Total Ci5F28
Wt. percent of sample recovered as gas
Sample wt., gms.
Molecular wt., Gas Density (of gas)
2 G, gaseous products
0.0458
0.0554
0.0006
0.0501
0.0004
0.0511
0.0005
0.0002
0.0268
. 0.0051
nil
n
o.ooo4
nil
0.7079
0.0058
o.ooAo
0.0009
0.0542 (2)
0.0247 (5)
0.0220 (2)
0.0151 (4)
0.0105 (5)
0.0052 (5)
nil
80.2
0.5595
551.
1.05 .
0.1087
O.O566
0.0008
0.0502
0.0004
0.0206
nil
0.0006
0.0120
0.0087
nil
0.0005
nil
0.0005
0.0008
0.5978
Not determined
0.0041
O.G456
(2)
0.0258
(2)
0.0514
(2)
0.0554
(A)
0.0182
W
0.0110
(5)
0.0050
(1)
68.
0.5195
520.
1.47
(a) Number in brackets is number of peaks in groups.
(b) This sample contained unsaturate.

44
of the original sample. This is seen to give a smooth curve starting
with zero percent CF^ for pure carbon and approaching 100 percent CFj^
for ratios of four and greater. It should be noted that the structure
of the original material is not a factor.
To obtain an idea of the radicals likely to be formed in the
irradiation, we will first examine the results of mass spectrograph
work. In Table 8 are given the species produced in the mass spectro-
(pj)
graphw/ for three electron energies; 50 70 and 100 volts. From this
data the ratios of the various ions are seen to vary considerably over
even this small energy range. This indicates the fallacy in prediction
of gamma yields from mass spectrograph data. However, from this data
we see that the two fluorocarbon species which are increasing relative
to 0F^ axe OF and CFg. From this we may expect that these species are
even more prevalent when the electrons have the much higher energies
encountered in gamma ray irradiation and for this reason they are in
cluded in the proposed mechanism for CF^.
In Table 9 is given the chemical mechanism for perfluoromethane
with the understanding that ions are probably the initial products
followed by neutralization.
As can be seen, the possible reactions for even this simple
system are very great. Reactions A-4 to A-9 and reaction A-15 result
in the regeneration of CF^. Reactions A-7 to A-9 will be important
for long irradiations where a significant amount of fluorine has built
up. These reactions (A-7 to A-9) account for the equilibrium for CF^
being nearly pure CF^. Since no C2F4 was detected from CF^ irradiations,
reaction A-15 must predominate over reaction A-12. In most cases the ,

45
TABLE 8.
MASS SPECTROGRAPH DATA FOR
Species
Relative Intensity at
50v
70v
lOOv
0
6.80
10.21
11.12
F
5.09
5.95
8.14
CF
5*67
4.91
6.52
of2
].4.o4
21.50
26.77
CF5
101.16
125.55
158.58
Ratio to QFj
C
0.0672
0.0828
0.0805
F
O.0505
0.0465
0.0587
CF
0.0565
0.0598
0.0456
cf2
O.1588
0.174
0.195

TABLE 9. CHEMICAL MECHANISM FOR CF^
Important species in mass spectrogram^ CFj, CF2, C, CF
Radical Generation
CF,
CFg + 2F
CF + 3F
(A-l)
(A-2)
(A-5)
Radical Recombination
CF + F CF2 (A-4)
cf2 + F > cf5 (A-5)
CF^ + F * CF4 (A-6)
CF + F2 CF5 (A-7)
CF2 + F2 > CF4 (A8)
CFj + F2 *CF4 + F (A-9)
CF + CF > C2F2 (A-10)
CF + CF ~^CF2 + C (A-ll)
CF2 + CF2 C2F4 (A-12)
CF2 + CF2 i CF + CF^ (A-15)
cf5 + cf5 c2F0 (a-i4)
CF^ + CF? CF2 + 0F4 (A-15)
unsaturates + F* F2, radicals saturated molecules (A-l£>)

47
amounts of C2F2 ^2?6 were nearly equal which could be accounted
for by the simple mechanism below:
CF^ ^ CF2 + 2F
CF2 + OFg OF + CF^
CF + CF ^2^2
cf5 + cf5 c2f6.
However at elevated temperature (99C) only a very small amount
of C2F2 was produced and no C2F^ was detected. From this we conclude
that the simple mechanism above is not adequate and that the true
mechanism is considerably more complex.
Perfluoroethane
In the mass spectrograph of O^F^^ the mos'<' Prevalent species
in decreasing importance are CF^, CgF^, CF, and CFg. The stability
of perfluoroethane to fragmentation by electron bombardment is not as
great as that of CF^ as is shown by the mass spectrograph studies
(see Table 8). In long term Irradiations it will be degraded a great
deal since by Figure 5 at equilibrium the sample will be about 50
weight percent CF^.
Judging from the mass spectrograph studies the initial radical
forming reactions must be those given in Table 10.
Other reactions are possible, but the ones in Table 10 are
sufficient to explain the results. At room temperature, no CgFg is
found, indicating that reaction B-ll is not important or that reactions
of the type of B-15 are important. Evidence for the latter is that at
elevated temperature where the amount of C2F2 is increased, the amount

A8
TABLE 10. CHEMICAL MECHANISM FOR CgFg
(6)
Important mass spectrograph species CF:
5*
C2F5*
CFj CF2
Radical Generation
C2F6 ^ 2CFj
-Mr C2F5 + F
vv CFS CF2 F
Radical Recombination
F + F *F2
CF^ + F >CF^
2CF5> C2F6
C2F5 + F C2F6
CF, CF2, F C2F6
CF^ + F2 > CF^ + F
C2F5 + F2> C2F + F
CF + CF > C9F2
CF2 + CF2-* CF + CF^
c2f5 + cf5 * c5f8
C2f5 + C2F5 n~Cij.F10
unsaturated molecules + F, F2, radicals > saturated
molecules
(B-l)
(B-2)
(B-5)
(B-4)
(b-5)
(B-6)
(B-7)
(B-8)
(B-9)
(B-10)
(B-ll)
(B-12)
(B-15)
(B-lA)
(B-15)

49
of larger molecules is decreased. A simplified mechanism which does
a fairly good job of correlating the data at room temperature and for
low energy absorption consists of equations B-l, B-2, B-4 to B-7> B-13
and B-l4.
This mechanism has been used in Appendix 6 to correlate the
ambient temperature data by a statistical method. In essence this method
involves assignment of an empirical effectiveness to each type of radi
cal. This effectiveness is related to the ease of rupture of the bond
to form the radical and possibly to various geometric factors. Multi
plication of the effectiveness times the number of ways the radical can
be formed resulta in an effective concentration of the radical. By
random recombination calculations, the relative concentrations of prod
ucts can be calculated. In this case, the reverse of this method is
used to arrive at the effectiveness. These effectiveness numbers are
of value in two ways. First, using them,r an estimate of the fraction
of the radicals which recombine to form the original compound can be
obtained and second, they give an estimate of which type of bonds are
preferentially broken in gamma irradiation. A further discussion of
this approach and the calculation are given in Appendix 6.
For C^F^. approximately one third of the radicals recombine to
yield CgF^. From the effectiveness numbers we see that CF^ radicals
are more than twice as effective as C2F5 radicals. This probably arises
from preferential formation of CF^ as indicated by the mass spectro
graph results.
Perfluoropropane
A more detailed examination of perfluoropropane was made than
for any of the other materials studied. Perfluoropropane was irradiated

50
over a range of energy absorptions and densities and at two tempera
tures in the engineering cobalt-60 source. At the lower temperature,
about 50C, the samples were partially liquid and these allow examina
tion of the mechanism in the liquid phase. At the upper temperature,
about 100C, the samples were in a single phase allowing examination of
the effects of density. Samples were also irradiated in the Oak Ridge
Graphite Reactor and the LITR reactor.
In the LITR, where an equilibrium distribution was obtained,
we see that at equilibrium about 40 weight percent of the sample is
porfluoromethane. (See Table 5). Thus CF^ will show a tendency to
higher concentrations for higher energy absorptions.
The proposed mechanism for perfluoropropane is given in Table
11. Daring the irradiation of perfluoropropane fluorine is produced
by reaction 0-4 and may either react with the wall by reaction C-25,
with radicals as in reactions C-20, C-21, and 0-22, or with unsaturated
materials as in 0-19* Thus the wall competes with the fluorocarbon
system for the fluorine. Reaction at the wall will be expected to
decline over a period of time due to the formation of a protective
fluoride coating. Reaction of fluorine at the wall will also decline
with increasing density since diffusion of fluorine to the wall will
be retarded. These two effects are seen in Figures 7 and 9 for the
products CF^ and C£F£ in the single phase cobalt-60 irradiations, and
in Figure 8 for CF4. in the partially liquid cobalt-60 irradiation.
Figure 10 for the product *n Par'tially liquid runs does not
show such a striking trend although some difference may be present for
low density. In general, these plots show an increase in the G values
for CFj. and CgF^ with increasing density and energy absorption. This

51
TABLE 11. CHEMICAL MECHANISM FOR C^FQ
Important mass spectrograph species 05) 0Fy CF, C3P7
Radical Formation
OsjFg Aa* C^Fy+ F
CjFg /VV ' C2F^ OF^
C^Fq VV~ * unsaturated radicals + F
Radical Recombination
F + F
CF^ + F i
c2f^ + F
CF^ + CFj
Cj-Fy + F
c2f5 + cf5-
cf4
c2f6
"*C2F6
*c5f8
>05P8
unsaturated radicals + F
5f8
c2f5 + c2f5-
->n-C4Fio
nC^Fy + CF^-
i-C5F7 + CF5-
+ n-C4F1Q
* w4fio
unsaturates + radicals or F > saturates
cf5 + ?2
c2f5-
"CF4 + F
2F6 F
O^Fg + F
2A1 + 5F2>A12F6
(0-1)
(0-2)
(o-5)
(0-4)
(o-5)
(0-6)
(o-7)
(0-8)
(0-9)
(c-io)
(0-11)
(0-12)
(0-15)
n-C-Fy + 02F^
^C5F12
(0-14)
i-C5Fy + C2F5
i5pl2
(0-15)
n-C^Fy + n-C^Fy-
n"6Fl4
(C-16)
n-CzF-. + i-C,F-,
5 7 5 7
^ 2-CF5C5F11
(0-17)
i-CjFy + i-C^Fy
-* 2,5-(cf5)2c4f8
(C-18)
(0-19)
(C-20)
(0-21)
(C-22)
(0-25)

G(CF,,)
52
Figure 7. G(CF^) versus density for 99C C^Fg samples^3).
(a) Numbers beside points are ergs absorbed per gram X 10^.

G(CF. )
55
Bulk density, g./cm.3
Figure 8* G(CF^) versus bulk density for 30C C^Fg samples
(cobalt-60)
(a) Numbers beside points are ergs absorbed per gram X 10

54
vG
*A
CM
O
>'
O
Density, g./cm.^
Figure 9 GCCgF^) versus density for 99C CgFg sampler
(cobalt-60).
ft
() Numbers beside points are ergs absorbed per gram X 10".
(b) This is the G value for liquid C3F3 found by extrapolation
of the plot for partially liquid samples to the liquid
density.

0(0.*,)
55
Figure 10, G^F^) versus bulk density for 30C C^g samples
(cobalt-60)(a).
Q
(a) Numbers beside points are ergs absorbed per gram X 10.
(b) The line indicates the average G value for the samples.

56
is to be expected since they can be formed by reaction of radicals
with fluorine (equations C-20, C-21). Molecules larger than C^Fg cannot
bo formed in this way and no corresponding increase of G value is noted
(Figures 15-26). Correspondingly, the plots of G values for fluorine
lost to the wall show a decrease with increasing density (Figures 11,
12). No effect of energy absorption is seen indicating that the equilib
rium concentration of fluorine had not been built up over the time
period since such an effect vas seen for CF^ and 0^^. Thus for liquid
fluorocarbons where the density is about 2 grams per cubic centimeter
diffusion of fluorine to the wall should be loss important, especially
with a less favorable geometry. This is confirmed by the work of
J. H. Simons^ with CgF-¡^ irradiated in the Oak Ridge Graphite Reactor
where no fluorine attack on the aluminum container was detected.
The capture of radicals at the wall is indicated by the very low
density sample (number 92) irradiated at ydG. In this sample the
overall G value vas only about 0.8 as compared with about 4.8 for
higher densities. From this we conclude that at low density the
radicals produced can be scavenged by the aluminum wall. Another
experiment was run with the sample tube partially filled with powdered
aluminum (number 85). The overall G value for this tube was not signif
icantly different from the G values of tubes not containing aluminum
powder. However, the product distribution vas considerably different
with smaller amounts of CF^, C^F^, and C^.F-^q being formed indicating a
decrease in the species Fg, F, CF^, and 0^^, presumable by capture by
the solid surface. Larger molecules show no such decrease. The larg
est individual peak in this sample has not been identified, but it is
not a fluorocarbon and must contain oxygen derived from the aluminum

g(f2)
57
Density, g./cm.^
Figure 11* G(F2) lost versus density for 99C C^Fg samples
(cobalt-60)(a).
(a) Numbers beside points are ergs absorbed per gram X 10"*

2.0
1..8
1.6
lJi
1.2
1.0
0.8
0.6
0.1;
0.2
0
O 64-7
(7)26,6
'-Q.O
120.0
O
(7)36,2
37.7 O 36~
Q3U.2
Q 108.8
- o25-6
15 j.
o o G 59-2
315 0 80.2
087.8
O 36.7
o 76.1*
0 0.1 0.2 0.3 0.4 0,5 0,6 .7 o,
Bulk density, g./cm.3
Figure 12. G(Fg) lost versus bulk density for
samples (cobalt-60)
(a)
85.1
o
8 0.9 1.0
30C C3F8
8
(a) Numbers beside points are ergs absorbed per gram X 10'

Density, gm./cm.^
Figure 13. G(ci|Fio) versua density for 99 C CgFg samples
(cobalt-60)
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) This is the G value for liquid C-^Fg found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c)The line indicated the average G value for the"" samples.

60
Bulk density, gra./cm.^
Figure Hi.. GCC^F-^q) versus bulk density for 30C CgFg
samples (cobalt-6o/~ \
(a) Numbers beside points are ergs absorbed per gram X 10
(b) The line indicated the average G value for the sample

6l
A
o
Density, gra./cm.
3
Figure 1$. G(n-C^F]_2) versus density for 99C C3F3 samples
(cobalt-60)(a).
(a) Numbers beside point3 are ergs absorbed per gram X 10".
(b) This is the G value for liquid CjFq found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c)The line indicates the average G value for the samples.

62
Bulk density, gm./cm.
Figure 16 Gin-C^F^) versus bulk density for 30C C^Fg
samples (cobalt-60)(a^.
(a) Numbers beside points are ergs absorbed per gram X 10"
(b) The line indicates the average G value for the samples.

g(-csf12)
65
Density, gm./cmj
Figure 17. G(i-C£F}_2) versus density for 99C C3F3 samples
(cobalt-60)(a). .
O
(a) Numbers besid points are ergs absorbed per gram X 10",
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid
density,
(c) The line indicates the average G value for the samples.

G(i-C5F12)
64
Bulk density, grn./cm.-^
Figure 18. G(i-C^F]_2) versus bulk density for 30C C^Fg
samples (cobalt-60)^a)'.
Q
(a) Numbers beside points are ergs absorbed per gram X 10
(b) The line indicates the average G value for the samples.

65
>
0.10
_ 0.08
% 0.06
V
5 CfwOUi
0.02
O
a ojl o.2 o.3 o.ii o.5 o.6 0.7 0.8 0.9
Density, gm./cm.^
Figure 19. GCn-CgF^) versus bulk density for 99C CgFg
samples (cobalt-O)^3'^.
(a) Numbers beside points are ergs absorbed per gram X 10.
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid density.
(c) The line indicates the average G value for the samples.

. 66
Figure 20. GCn-C^F-^) versus bulk density for 30C C3F5
samples (cobalt-60)(ai)*
O
(a) Numbers beside points are ergs absorbed per gram X 10.
(b) The line indicates the average G value for the samples.

67
Density,, gm./cm.3
Figure 21. G(2-CF-jC^F-q) versus density for 99C C^Fg
samples (cobalt-60)^.
CD
(a) Numbers beside points are ergs absorbed per gram X 10
(b)' This is: the G value for liquid C^Fg found by extrapolation
of.the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average G value for the samples.

G(2-CF3C^F11)
68
Bulk density, gm./cm.^
Figure 22. G(2-CF3C£F-q.) versus bulk density for 30C C3Fq
samples (cobalt-60)^a^
(a) Numbers, beside points are ergs absorbed per gram X 10"^.
(b) The line indicates the average G value for the samples.

G(2,3-(CF3)2CUF8)
69
Density, gm./cm.^
Figure 23. G(2,3''(CF3)2G[iFg) versus density for 99C C^Fg
samples (cobalt-60)^a\
Q
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average G value for the samples.
!

G(2,3-(CF3)2CiiF8)
70
Bulk density, gm./cm.^
Figure 2U. G(2,3(CF3)2C^Fq) versus bulk density for 30C
C^Fg samples (cobalt-O)^s
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) The line indicates the average G value for the samples.

71
*3
Density, gm./cm.
Figure 25. Sum of G values versus density for 99C CgFg
samples (cobalt-60)^a^.
Q
(a) Numbers beside points are ergs, absorbed per gram X 10.
(b) This is the G value for liquid CgFg found by extrapolation
of the plot for partially liquid samples to the liquid
density.
(c) The line indicates the average G value for the samples.

72
Bulk density, gm./cm.^
Figure 26 Sum of G values versus bulk density for 30C CgFg
samples (cobalt-6G)^a^e
ft
(a) Numbers beside points are ergs absorbed per gram X 10 .
(b) The line indicates the average G value for the samples.

75
oxide. Both these tubes showed an increased amount of unsaturated
material leading to the conclusion that, in perfluoropropane irradia
tion, unsaturated molecules are present as intermediates. Relative
amounts of unsaturated materials were determined from the amounts of
2F2* 2F4* and 5F6 dotocte G values in the partially liquid samples and in the single >
phase samples are not significantly different so the mechanism is
probably the same.
As in the case of CgFg, a ran(*0In recombination calculation has
been made to determine the effective concentrations of the radicals.
For this calculation, a.simplified mechanism consisting of reactions
C-l, 0-2, C-4 to 0-9 and 0-11 to 0-18 has been used. The results
indicate that approximately 26 percent of the radicals formed recombine
to form C^Fg. Of the radicals formed, the i-C^Fy radical is the most
effective leading to the conclusion that the fluorine atoms attached
to a carbon atom next to a CF~ group are relatively easy to remove.
The n-CjFy radical is only O.515 times as effective so removal of a
fluorine atom from an attached CF^ group is rather difficult. The
^2F5 Radical is slightly more effective than the n-C^Fy radical either
from geometric effects or because the bond is slightly easier to rup
ture. The CF^ effectiveness is roughly twice that of probably
resulting from some complete rupture of C^Fg into single carbon entities.
This is to be expected from the mass spectrograph data.
In summary we can say for C^Fg, and by analogy for the other
fluorocarbons, both energy absorption and density affect the G values
of the products of irradiation if the geometry of the sample container
is such that the wall can act as a competitor for fluorine and radicals.

74
Perfluoro-n-Butane
The purification of this material by gas chromatography was
hampered by the inability to separate the two isomers. To. obtain a
reasonable separation the material on the first part of the peak, where
n-C^F^0 appears, was cut. Direct analysis by chromatography can give
no indication of the purity, but an estimate can be obtained by examina
tion of the radiation products. Of the isomers the only ones which
can be produced from n-C^F-^Q are n-C^F^ and 5CF^C^ Fir The only
isomer which can be produced from i-C^F^ is 2,5-(CF^)2C^Fg. From
mixtures of the two isomers 2-CF^C^F^ can be produced. In the analysis
of the radiation products relatively large amounts of n-C^F^ and
3-CFjO^Fjj are found along with relatively small amounts of 2-CF^C^F-^
and 2,5-(CF^)2C4Fq. If we assume that the ratios of the sums of prod
ucts from each initial isomer are the same as the ratios of the isomers
in the original material and credit the 2-CF,C_F,, to both isomers then;
P 5 11
1~4F10 .02 + .02 + .01 + .01 .06 09
n- 04F10 .11 + .12 + .19 + .21 + .01 + .02 735 *
or the n-C^F^Q is about 91*7 mole percent pure. It may be more pure
than this since some of the two minor peaks may be second generation
products.
The samples of n-C^F-^Q were not available until after the
elevated temperature cobalt-60 and the reactor irradiations were
completed so the only data available is for ambient temperature in the
engineering cobalt-60 source. It should be noted that approximately
one-third of sample number 95 leaked out between canning and decanning
for analysis. This does not appear to have affected the G values

75
with the possible exception of a small decrease in the amount of CF^
formed.
No unsaturated molecules were detected, but these must have
been present as intermediates to account for the relatively large
amount of material found between 0^ and The mechanism for n-C^ F10
is given in Table 12.
The random recombination calculation has been made for n-C^ F10
using a simplified mechanism consisting of equations D-l to D-4, D-6
to D-9, and D-12 to D-26. From this calculation we see that about 52
percent of the radicals formed recombine to yield n-C^F^. These data,
as for the case of CjFq, state that the carbon-fluorine bond next to a
CF group is relatively weak compared to the carbon-fluorine bond on
5
an intact CF^ group. CF^ is more important than the other radicals
formed by carbon-carbon bond rupture indicating, once again, that in
some cases the molecule is ruptured into several fragments. The
radicals 02F5 n-C^Fy which are formed by similar processes have
identical effectiveness numbers as would be expected if these numbers
are due to ease of formation and not dependent a great deal on geometric
factors.
Perfluoro-n-Pentane
In the cobalt-00 irradiation of perfluoro-n-pentane at ambient
and elevated temperature, there is no discernable difference between
the G values for the products at the two temperatures. For low energy
absorption no unsaturated materials are found and after prolonged
irradiation only small amounts of C^F^ 021(1 are i>oun<** This
indicates that any unsaturates formed as initial products are saturated

16
TABLE 12. CHEMICAL MECHANISM FOR n-C^Q
Important mass spectrometer peaks
(6)
CF3 C2F5, CF, C^Fc;
Radical Generation
n-C^F-, n nC^F-y + CFv (Dl)
VV > 2 C^F' ^ (D-2)
^-AA > n-C^Fo + F (D5)
AAi & 2 C4F9 + F (D-4)
VV > unsaturated radicals + F (D-5)
Radical Recombination
F + F -F2
C4F9 + P > C4F10
CF3 + n-C^Fy > n-C^F-^Q
2 C2F^ ^ n-C4F^Q
unsaturated radicals + F ^n-C^F^Q
unsaturated radicals'> unsaturated molecules
CF3 + F > CF4
n-CzFy + F >CzFq
C2F5 + F * C2F§
CFj + CF3 >C2F6
CF* + C2F3<>CzFq
c2f5 y8 >n-CkF12
n-C4F + CF* n-Cf-Fn 0
i~C4F9 + CF5 >iC5F12
n-C4F3 + C2% > n-c^F14
-C4F9 + 02F3 * 5-CFzCgFjj
n-C4F9 + n-CsFy^n-CyF,^
-C4FU + n-C^Fy +}-CFzpF1z
nC4P9 + nC4F9 nr-C3F]_3
nC4P9 + -C4F0 5~CFzCyFic
-C4F0 + -C4F0.>54-(cf5;2C6F12
CFz + F2 CF4 + F
C2Fc + F2 > C2F6 + F
nC3F7 + F21 >OzFg + F
C4P9 + P2>C4F10 + F
vmsaturated molecules + f2, p, radicals *
saturated molecules
(I>*)
(D~7)
(I>3)
Cd-9)
(D-10)
(D-ll)
(D-12)
(D-15)
(D-l4)
(D-15)
(D-16)
(D-17)
(D-18)
(D-19)
(D-20)
(D-21)
(D-22)
(3>25)
(D-24)
(D-25)
(D-26)
(D-27)
(D-28)
(D-29)
CD-50)
(D-51)

77
by radical or fluorine capture. Small amounts of unsaturatee are
necessary to account for the significant amounts of C-q and formed
in the irradiation.
Since the possible number of reactions has become so large, a
generalized reaction mechanism has been written from which the individual
reaction equations are easily derived (Table 15).
For the statistical recombination calculation a simplified
mechanism consisting of equations E-l to E-5 and E-7 has been used.
From this calculation we conclude that, on a statistical basis, about
25 percent of the radicals formed will recombine to give n-C^F-^. Ex
amination of the calculated effectiveness numbers leads to the conclusion
that the further a 0Fo group is from a CF-. group the easier its C-F
2 ?
bonds are to split. This follows from the effectiveness of the radical
formed by removal of fluorine in the three position being greater than
that formed by removal of fluorine in the two position. In all other
respects this data confirms that for the smaller molecules. That is,
radicals formed by carbon-carbon bond breakage, other than CF, are
5
equally effective indicating that geometric effects are not important,
and CF^ must be formed partially by extensive fragmentation of the
original molecules to yield more than one single-carbon group. This
is also expected from mass spectrograph data where GF^ is in great
excess.
Perfluorocyclopentane
Little can be said about the mechanism of radiolytic change in
cy-CjpF-LQ since very few of the products have been successfully identi
fied. The probable mechanism is given in Table l4.

78
TABLE 15. CHEMICAL MECHANISM FOR n-C^F12
Important species from mass spectrograph* '
OFy c2f^, C^Fy, CF, c2F4, C^
Radical Generation
*-C5Fl2 -W *ri"C5Fll + F (2~1)
2-5Fll + F (B-2)
-A/V--5- 5-c5f1]l + f (e-3)
VV CFj + n-C^Fp (E-4)
-4/V > C2F5 + n-C^Fy (E-5)
7> unsaturated radicals + F (E-6)
Radical Recombination
R1 + R2 ** R1R2 (E-7)
R! + F2* RiF + F (B-8)
unsaturated radicals unsaturated molecules (E-9)
unsaturated molecules + F, F radicals
saturated molecules (E-10)

79
TABLE 14. CHEMICAL MECHANISM FOR cy-C^F^
(.5)
Important species in mass spectrographx '
o5f5, c2f4, cf, cf5, c4f7, c5f?, cf2, c5f9
Radical Generation
(F-l)
(F-2)
(F-5)
(F-4)
(f-5)
(F-6)
R1 + R2~*R1R2 (F-7)
R1 + f2~RF + F (F-8)
unsaturated radicals unsaturated molecules (F9)
i
unsaturated molecules + F, F2,R saturated molecules (F-10)

80
Two types of initial radical generation steps appear to occur;
splitting of the ring as in equations F-2 and F-5 and renoval of a
fluorine atom from an otherwise intact ring as in F-l. Production of
fragments by F-2 and F-5 does not appear to be as important in gamma
irradiation as it is in the mass spectrometer where fragmentation of
the ring is highly dominant over removal of a fluorine atom. In gamma
radiolysis a large fraction of the products are larger than the original
material leading to the conclusion that reactions F-5 F-6, and F-7
are important.
Fluorine lost to the wall cannot be calculated for these samples
since the products are not identified. Again since the products are
unknown a statistical recombination calculation is not possible.
In the LITR irradiations the cyclic structure of this material
did not affect the equilibrium distribution of the products. For its
fluorine to carbon ratio of two the final sample contains about 25
weight percent CF^.
Perfluoro-n-Hexane
Normal perfluorohexane shows considerable difference between
the total G values for the two samples. The reason for this is uncer
tain since it could be due to either the large difference in the total
energy absorbed in the two cases or to the difference between ambient
temperature and 99C. Since no temperature effect has been noted in
any other case this is probably due to the difference in energy ab
sorption.
The chemical mechanism in Table 15 has been written in general
terms since the total number of possible reactions is very high.

81
TABLE 15. CHEMICAL MECHANISM FOR
Mass spectrograph not available
Radical Generation
n"c6Fl4JV'' n"6Fi5 + F (G-1)
w
-* 2-6f15 + F
(6-2)
i-Vv* F
(G-?)
~-w
- CF^ + n0^F11
(g-4)
C2F5 + n-C^Fp
(g-5)

> 2CjFy
(g6)
> unsaturated radicals + F
(G7)
Radical Recombination
F + F *F2
(6-8)
R1 + R2 R1
R2
(6-9)
R1 + F2 R1
F + F
(G-10)
unsaturated radicals unsaturated molecules
(G11)
unsaturated molecules + F, F2, radicsds
saturated molecules
(6-12)

82
A simplified mechanism consisting of equations G-l to G-6,
G-8, and G9 has been used in the statistical recombination calculation.
From this calculation the recombination of fragments to yield the
original n-C^F^ is estimated to bo 22.5 percent. As has been the
case for the smaller molecules the further a 0F2 group is from a CFj
group the easier its carbon-fluorine bonds are to split. Also as in the
other molecules the radicals formed by carbon-carbon bond splitting,
excepting CF^ have the same effectiveness number indicating the rupture
of any carbon-carbon bond is equally likely. The larger effectiveness
of CF^ must be due to greater fragmentation of some molecules to yield
more than one single-carbon radical.
In the LITR at equilibrium about 50 weight percent of the sample
is converted to CF^.
Perfluoro-2-Methylpentane
A number of samples of this material were irradiated early in
the program. However, in testing the thiourea column (see Appendix 5)
it was found that about 7 mole percent unsaturates were present. These
impurities gave fairly large amounts of unwanted peaks probably result
ing from radical capture. While this confirms the hypothesis that
relatively small amounts of unsaturated material can contribute signifi
cantly to the final product distribution by radical capture the results
of these tubes are too complex for interpretation at the present time.
For this reason the cobalt-60 irradiations of the impure material axe
not reported, and instead a sample irradiated after further purification
is reported. The irradiations of the impure material in the Oak Ridge
Graphite Reactor and in the LITR axe reported since after repurification

85
thero was insufficient time for additional reactor runs. In the LITR,
due to the formation of an equilibrium mixture, the small amount of
unsaturate has little effect.
The mechanism for 2-CF^C^F^^ is given in Table l6. A simplified
mechanism consisting of reactions H-l to H-9, H-ll, and H-l 2 has been
used in the random recombination calculations. These cal evil ations pre
dict a 22.8 percent recombination of the fragments to form the original
2-CF^OcjF-Q As in the other molecules the further a C-F bond is from
a CF^ group the more easily it is ruptured. All radicals formed by
rupturing carbon-carbon bonds, excepting CF^, are equally effective
indicating that for radiolytic rupture all carbon-carbon bonds are
equivalent. Hence, as in the other cases, some molecules must rupture
extensively to yield more than one single-carbon group.
Perfluoro-2,3-Dimethylbutane
A sample of this material was not available until late in the
program so only one sample was irradiated in the engineering cobalt-00
source at ambient temperature.
A general mechanism is written in Table 17 for this material
since the possible reactions are numerous. Although no mass spectrograph
data is available for this material it is probable that owing to the
large number of CF^ groups CF^ would be very predominant. A possible
extensive shattering of the molecule is given below.
C-C-C-C
I | -0 > 4CF5 + 2CF
CO y

84
TABLE l6. CHEMICAL MECHANISM FOR 2-CFjC^F-^
No mass spectrogram available
Radical Generation
C-C-O-C-C -AV c-c-c-c-c*
I I
c c
YV 1 ^ CCCCC
c
> C-C-C-C-C
I
c
C-C-C-C-C
-^ CCCCc
c
\\ CF^ + C-C-C-C*
c
CFj + C-C-C-C-C
""> C2F5 + CC C*
c
> nC^Fy + i-C^Fy
unsaturated radicals + F
Radical Recombination
F + F F2
R1 + R2 >R1R2
R-^ + F2 R-^F + F
unsaturated radicals* unsaturated molecules
unsaturated molecules + F, F2, radicals >
saturated molecules
(H-l)
(H-2)
(H-5)
(H-4)
(H-5)
(H-6)
(H-7)
(H-8)
(H-9)
(H-10)
(H-ll)
(H-l2)
(K-15)
(H-14)
(H-15)

85
TABLE 17. CHEMICAL MECHANISM FOR 2,5-(CF5)2C^Fq
No mass spectrogram available
Radical Generation
0-C-C-C -W C-C-C-C* + F
M II
C C c c
> C-C-C-C + F
I I
c c
> CFt + CCCC
5 I
(1-1)
(1-2)
(1-5)
*YV-
j~'VV*
2 -C3F7
unsaturated radicals + F
(1-4)
(1-5)
Radical Recombination
F + F
R + R,
R1 + V
R1R2
Rj^F + F
unsaturated radicals
unsaturated molecules
unsaturated molecules + F, F2 radicals'
saturated molecules
(1-6)
(1-7)
(1-8)
(1-9)
(1-10)

86
Evidence for this reaction is the presence of C2F2 "t^ie Pro ucts from this irradiation. Since this is the only case for a high
density parent where an unsaturated material is present in significant
amounts for low energy absorption this reaction probably is of consider
able importance.
A simplified mechanism consisting of Equations 1-1 to 1-4, 1-6,
and 1-7 was used to perform the random recombination calculations. The
results indicate that about 24 percent of the radicals recombine to
give the parent. The effectiveness numbers confirm the conclusions
from the other molecules. Carbon-fluorine bonds are more easily rup
tured the further they are from CFj, groups. Removal of a fluorine from
an attached CF^ group is rather difficult. All carbon-carbon bonds are
equivalent for radiation damage. Some CF^ radicals are formed by ex
tensive rupture of the molecule into more than one single-carbon
fragment.

TABLE 18 PART 1. CF^ IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.
Sample Wt., gms.
Density, gm./cm.3
No. Days
Integrated Flux, 10 7 Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in assay or as listed
89
0.1681
0.202
9.92
1.705
2.5h
30
90
0.1718
0.206
9.92
1.705
2.60
30
Mole fractionG
83
0.161+6
0.197
9.9
1.795
2.62
Valued
9
0.1852
0.222
33.09
6.65
10.8
30
Moles F2/mole mixtureG Value
0.00153-1.10
0.0011+5-1.06
0.0008-0.55
0.001-1.73
cf4
0.99927
0.9991
0.999b
0.991+8
c2f6
0.0002-0.lb
0.0005-0.36
0.0001+-0.27
0.0025-0.1+1+
nil
nil
nil
nil
c2f2
0.00027-0.20
0.00015-0.11
trace
0.0025-0.1+1+
C3F8
0.00026-0.19
0.00025-0.19
0.00016^-0.11
0.0002-0.0l+
Cl+F10
nil
nil
nil
nil
SF12
nil
nil
nil
nil
c6fiU
nil
nil
nil
nil
£G
1.63
1.85
0.93
2.65
Molecular wt., Gas Density
88.3
87.5
88.5
93.1+
(a) Number in brackets is number of peaks in group

TABLE 18 PART 2.
C2F6
IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.
Sample wt., gras.
Density, gm./cm.^
No. Days
Integrated Dose, 10^ Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in Assay or as Listed
88
0.2134
0.255
9.92
1.80
3.41
99
68
0.3723
0.447
17.17
3.44
11.4
25
Mole Fraction-
35(a)
0.2931
0.351
33.09
6.65
17.3
25,,^
-G Value''
69
0.7075
0.846
17.17
3.44
21.7
25
Moles F2/mole mixtureG Value
0.0035-1.53
0.0050-1.21
0.0090-1.01
0.0053-1.17
' CF.
0.0010-0.44
0.0029-0.70
0.0072-0.80
0.0038-0.84
C2F6
0.99605
0.9901
0.9789
0.9877
C2Fk
nil
nil
nil
nil
^2F2
0.0012-0.52
0.0002-0.05
0.0003-0.04
0.0003-0.07
C3F8
Cy-c3F6
0.0012-0.52
0.0045-1.08
0.0083-0.93
0.0055-1.22
nil
0.0002-0.05
0.0005-0.06
0.0002-0.05
C^io (mostly normal)
n-<"'5F12
Not Detected
0,0011-0.26
0.0028-0.32
0.0017-0.37
Total C^F^
0.0001-0.03
0.0005-0.06
0.0003-0.07
i-C5F12
n_c6Fl4
0.0002-0.08
0.0002-0.05
0.0007-0.07
0.0001-0.02
0.0001-0.05
0.0001-0.03
0.0003-0.04
0.0001-0.02
2-CF3C5Fll
3CF3C5F11
Total 2 + 3 =
0.0001-0.03
nil
Total 2 + 3 =
O.OOO2-O.O8
0.0001-0.03
trace '
0.0001-0.02
2,3-(CF ) C4F8
\ Total C7Fl6
Total CqF^q
Tta^9F20
0.00005-0.02
nil
nil
nil
0.0001-0.03
0.00013-0.03(2)
0.00013-0.03(2)
0.00003-0.01(1)
0.0005-0.06
0.0001-0.02
0.00004-0.0i(2)
0.00014-0.03(2)
0.00004-0.01(1)
3.24
3.62
3.39
3.92
Molecular wt., Gas Density
138
143.3
145
141.5
(a) This tube analyzed before 92C squalane column installed so amount of compounds above Cg is
. not known.
(b) Number in brackets is number of peaks in group.
03
CD

TABLE 18 PART 3. C^Fg IRRADIATED BY COBALT-60 GAMMA RAYS, T = 99C
Tube No.
53
50
63
52
Sample Wt., gms.
0.1767
0.1700
0.561+8
0.1767
Density gm./cm.^
0.212
0.2035
0.677
0.2115
No. Days
1 8
10ft
108
20.17
10.25
20.17
10.25
Integrated Dose,
Roent.
3.66
1.86
3.66
1.86
Energy Absorbed,
ergs
5.1b
2.8l
18.1+
2.92
Energy Abs./gm.,
ergs
32.5
16.5
32.5,
Value'a'
16.5
Compound in Assay
Mole FractionG
Moles F2/moleG Value
cfh
C2F6
c2fi+
C2F2
c3f8
ChllO
n-c5Fi2
1-C5F12
n-C6Fll+
2-CF C F
3"CF3C5F11
2,3-(CF3)^Cu8
Total C-vF-jg
Total CgF^g
Total CgFgQ
Total 0
Total
Total
Z
Molecular wt. Gas Density
Z^IO 22
11^
C12f26
G
O.OOTT5-1.19
0.0037-0.57
0.001+1-0.62
nil
nil
0.9973
0.0061+-0.98
0.0010-0.15
0.0025-0.38
0.0005-0.07
0.0008-0.12
nil
0.0015-0.23
0.0007-0.11(2)
0.0007-0.11(2)
0.0006-0.09(2)
0.000l+-0.07(2)
nil
nil
1+.69
19U.5
0.0032-1.00
0.0008-0.25
0.0010-0.31
nil
nil
0.9926
0.0022-0.68
0.0005-0.16
0.0011-0.3!+
0.00008-0.03
0.0005-0.16
nil
0.0003-0.09
0.0003-0.09(2)
0.0003-0.09(2)
0.0001-0.03(1)
0.00009-0.03(2)
nil
nil
3.26
188.
0.0055-0.85
0.001+5-0.69
0.0055-0.85
nil
nil
0.971+9
0.0071+-1.11+
0.0018-0.28
0.0022-0.3l+
0.0001+-0.07
0.0009-0. ll+
nil
1.0007-0.11
0.0006-0.09(2)
0.000l+-0.07(2)
0.00035-0.06(2)
0.0003-0.05(2)
nil
nil
1+.71*
191.3
0.001+2-1.29
0.0008-0.2l+
0.0013-0.1+0
nil
nil
0.9908
0.0031-0.96
0.0005-0.16
0.0011+-0.1+3
trace
0.0006-0.19
nil
0.0003-0.09
0.0006-0.19(2)
0.000lt-0.12(2)
0.0002-0.07(1)
nil
nil
nil
l+.lU
189.3
CD
'O
(a) Number in brackets is number of peaks in group.

TABLE 18 PART 3 (continued)
Tube No.
56
61
62-
Sample Wt., gms.
Density gm./cm.
0.3496
0.5322
0.4977
0.418
0.637
0.597
No. Days
10.25
10.25
10.25
Integrated Dose, 10 Roent.
Energy Absorbed, 10^ ergs
Energy Abs./gm., 10 ergs
1.86
1.86
1.86
5.78
8.80
8.25
16.5
16.5
16.5
Value'a)
Compound in Assay
Mole FractionG
57
0.36l8
0.U33
10.25
1.86
5.97
16.5
Moles F2/moleG Value
CFjj
C2F6
C2F4
C2F2
C3F8
C4f10
n_C5F12
i_C5F12
n_c6Fl4
2-OF 0 F
3-CFC'F^
2,3-(CF)'cj^8
Total C^F^g
Total CqF^q
Total CqF20
Total C^qF22
Total CllF24
Total Cj2F26
£g
Molecular vt., Gas Density
0.0068-2.08
0.0013-0.1*0
0.0017-0.52
nil
nil
0.9897
0.0036-1.11
0.0006-0.19
0.0013-0.1*0
0.0001-0.03
0.0006-0.19
nil
0.0004-0.15
0.0004-0.12(2)
0.0002-0.07(2)
0.0001-0.03(1)
nil
nil
nil
5.29
190.7
0.0034-1.04
0.0019-0.59
0.0023-0.75
nil
nil
0.9873
0.0038-1.16
0.0008-0.24
0.0015-0.46
0.0003-0.09
0.0008-0.24
nil
0.0005-0.15
0.0003-0.09(2)
0.0003-0.09(2)
0.0002-0.07(1)
nil
nil
nil
4.97
192.3
0.0025-0.76
0.0018-0.55
0.0024-0.74
nil
0.00001-0.003
0.9887
0.0037-1.14
0.0006-0.19
0.0012-0.37
0.0003-0.09
0.0005-0.15
nil
0.0004-0.12
0.00017-0.06(2)
0.00015-0.05(2)
0.00008-0.03(2)
0.00006-0.02(2)
nil
nil
4.273
191.3
0.0030-0.93
0.0010-0.31
O.OOI8-O.56
nil
0.00004-0.01
0.9908
0.0030-0.93
0.0005-0.16
0.0013-0.40
0.00012-0.04
0.00004-0.12
nil
0.0003-0.09
0.0003-0.09(2)
0.00017-0.06(2)
0.0001-0.03(2)
0.00013-0.04(2)
nil
nil
3.77
192.
(a) Number in brackets is number of peaks in group.

TABLE 18 PART 3 (continued)
Tube No.
Sample Wt., gms.
Density gm./cm.3
No. Days
Integrated Dose, 10^ Roent.
Energy Absorbed, 10 ergs
Energy Abs./gm., 10 ergs
Compound in Assay
51
0.1681
0.2015
20.17
3.66
5.47
32.5
Mole Fraction-
Moles F2/moleG Value
CFU
C2F6
C2F4
C2F2
C3F8
VlO
n-C^ig
i-c5f12
2-CF^C^
3-CF^C^F"
23-(CF3)|C#8
Total C7Fi6
Total CqFi8
Total C F
Total
Total C-| 1 F?k
Total C-,0F0ir
ZG12 26
Molecular wt., Gas Density
0.0058-0.89
0.0030-0.1+7
0.0042-0.64
nil
nil
0.9796
0.0062-0.95
0.0010-0.16
0.0022-0.34
0.0002-0.03
0.0009-0.l4
nil
0.0010-0.16
0.0007-0.11(2)
0.0007-0.11(2)
0.0003-0.05(1)
nil
nil
nil
4.05
191.9
55
0.3424
0.410
20.17
3.66
11.11
0.0070-1.07
0.0039-0.60
0.0051-0.78
nil
0.0002-0.03
0.9752
0.0064-0.98
0.0016-0.24
0.0025-0.38
0.0004-0.07
0.0009-0.14
trace
0.0009-0.14
0.0009-0.14(2)
0.0007-0.10(2)
0.0006-0.09(2)
0.0004-0.07(2)
nil
nil
4.83
191.
(a) Number in brackets is number of peaks in group.

TABLE l8 PART 4. C^Fg IRRADIATED BY COBALT-60 GAMMA RAYS, T = 30C
Tube No.
7(a)
2^( a)
37(a
Sample Wt., Gms.
0.4239
0.6654
0.4916
0.1311
Bulk Density, gm./cm.3
0.508
0.898
0.589
0.157
No. Days
18.94
28.9
28.9
10.0
Integrated Dose, 10 J Roent.
Energy Absorbed, 10;: ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
8.6
8.53
12.23
2.88
32.4
56.6
53.5
3.36
76.4
85.1
Mole FractionG
108.8
Valued)
25.6
Moles F2/mole mixtureG Value
CFk
C9FA
44
C3F8
n-CoF/r
cMo
n-C5F12
i-c5Fi2
n~c6Fl4
2-GF G F
3"CF3C5F11
2,3-(CF
TotalJC7Fl6
Total CqF-^q
Tota! C F
Total C^qF22
Total
Total C^2F2g
Total
£G
Molecular Wt.
C13F28
Gas Density
0.0080-0.52
0.0126-0.82
0.0126-0.82
nil
nil
0.9^16
nil
0.0159-1.03
0.0035-0.22
0.0053-0.34
0.0012-0.07
0.0029-0.19
trace
0.0026-0.IT
0.0018-0.12(3)
Incomplete
Analysis
4.30
188.6
(a) These three samples were
and do not have complete
(b) This tube was copper. (
of peaks in group.
0.0062-0.37
0.0112-0.67
0.0112-0.67
nil
nil
0.9501
nil
0.0125-0.75
0.0028-0.17
0.0043-0.26
0.0012-0.07
0.0025-0.15
0.0001-0.006
0.0016-0.10
0.0019-0.12(8)
0.0005-0.o4(i)
Incomplete
Analysis
3.376
190.6
0.0186-0.83
0.0160-0.72
nil
nil
0.9102
nil
0.0239-1.07
0.0042-0.19
0.0084-0.37
0.0028-0.12
0.0048-0.21
trace
0.0037-0.17
0.0043-0.20(8)
0.0029-0.13(10)
trace
Incomplete
Analysis
5.35
190.5
0.0020-0.40
0.0018-0.34
0.0019-0.37
nil
nil
0.9885
0.00085-0.17
0.0023-0.46
0.0085-0.17
0.0015-0.30
0.00043-0.08
0.0014-0.27
trace
0.00053-0.10
trace
2.66
187.9
VO
IO
analyzed before the addition of the high temperature squalane column
analyses of the high molecular weight compounds.
c) This sample was a single phase. (d) Number in brackets is number

TABLE l8 PART 4 (continued)
Tube No.
30
42
46
47
Sample Wt., Gms.
0.4808
0.1452
0.2995
0.2792
Bulk Density, gm./cm.
0.577
0.1738
0.359
0.334
No. Days
10.0
23.0
10.0
10.0
Integrated Dose, 10 7 Roent.
Energy Absorbed, 10 ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
4,o6
7.28
4.24
4.17
17.35
9.39
11.3
10.1
36.1
64.7
Mole Fraction-
3T*7fd)
-G Valueva;
36.2
Moles F0/mole mixtureG Value
0.0080-1.11
0.0230-1.74
0.0086-1.i4
0.0093-1.;
CFU
'2F6
C2F4
^2F2
c3F8
n~C3F6
C4F10
n_c f12
i-C5Fi2
nc6FlU
2-CF3C5Fn
2,3-7Si?1c^8
Total
Total
Total
Total
Total
Total C]L2^26
ctf16
C8Fl8
C9F20
C10^22
Total C-,oFoA
G13 20
Molecular Wt., Gas Density
0.0033-0.46
0.0047-0.65
nil
trace
0.9773
nil
0.0056-0.78
0.0015-0.20
0.0024-0.34
0.0012-0.17
0.0016-0.22
0.0001-0.01
0.0007-0.09
0.0004-0.06(1)
0.0006-0.08(2)
0.0004-0.06(2)
0.0004-0.06(2)
nil
nil
4.29
192.2
(a) These three samples were analyzed before the
. and do not have complete analyses of the high
copper. (c) This sample was a single phase
group.
0.0089-0.67
0.0087-0.66
nil
nil
0.9528
nil
0.0108-0.82
0.0034-0.26
0.0051-0.39
0.00085-0.06
0.0025-0.19
nil
0.0019-0.14
0.0014-0.11(2)
0.0015-0.11(2)
0.0009-0.07(2)
0.0010-0.08(5)
0.0002-0.01(1)
nil
5.31
191.8
0.0043-0.57
0.0042-0.56
nil
nil
0.9841
nil
0.0071-0.94
0.0016-0.21
0.0034-0.45
O.OOO82-O.II
0.0014-0.19
nil
0.0011-0.
0.00082-0
O.OOO58-O
0.00042-0
0.00026-0
0.00005-0
nil
4.60
191.0
15
.11(2)
.07(2)
.06(2)
.03(3)
.01(1)
0.0049-0.67
0.0054-0.74
nil
nil.
O.9702
nil
0.0075-1.02
0.0021-0.29
O.OO36-O.49
0.00068-0.09
0.0022-0.30
nil
0.0014-0.20
0.00077-0.10(2)
0.00049-0.07(2)
0.00044-0.06(2)
0.00037-0.05(2)
nil
nil
5.36
197.5
addition of the high temperature squalane column
molecular weight compounds. (b) This tube was
(d) Humber in brackets is number of peaks in
vo
\>i

TABLE 18 PART 4 (continued)
Tube No.
Sample Wt. Gms.
Bulk Density, gm./cm.
No. Days
Integrated Dose, 10 Roent.
Energy Absorbed, 10 ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
40
0.5446
0.652
10.0
4.13
20.0
36. T
43
0.1514
0.1815
23.0
7.42
9.99
66.0
Mole Fraction-
39
0.1300
0.1558
10.0
2.99
3.46
2 6.6
G Value
49
0.3190
0.382
23.0
9.9
28.0
87.8
Moles F?/mole mixtureG Value
0.0058-0.78
0.0196-1.46
0.0080-1.52
0.0150-0.85
CFk
0.0040-0.5U
O.OO78-O.58
0.0022-0.42
0.0119-0.67
C2F6
0.0041-0.56
0.0084-0.62
0.0018-0.34
O.OH6-O.65
c2f4
nil
nil
nil
nil
C2Fo
0.00014-0.02
nil
0.0009-0.17
0.0001-0.006
C3F8
0.9728
0.9577
0.9861
0.9401
n-c^F?
nil
nil
nil
nil
cfrio
0.0070-0.95
0.0105-0.78
0.0035-0.66
0.0152-0.86
n-C5F12
0.0018-0.24
0.0024-0.18
0.0007-0.13
0.0037-0.20
i-C^F£2
0.0032-0.44
0.0045-0.34
0.0014-0.27
0.0059-0.34
n-C6Fl4
0.00077-0.10
0.0006-0.05
0.0005-0.09
0.0014-0.07
2-CF3CF
0.0019-0.26
0.0021-0.16
0.0012-0.23
0.0029-0.17
3-CF^C^F
nil
nil
nil
nil
2 ,3-(CFf)pCFo
0.0015-0.20
0.0021-0.16
0.0012-0.23
0.0026-0.15
TotalW,8
0.00072-0.09(2)
0.0014-0.10(2)
0.0003-0.06(2)
0.0016-0.09(2)
Total CgFI'n
0.00075-0.10(2)
0.0013-0.09(2)
0.0004-0.07(2)
0.0012-0.07(2)
Total CqF
0.00056-0.07(2)
0.0009-0.07(2)
0.0001-0.02(1)
0.00085-0.05(2)
Total f
0.00064-0.08(4)
0.0004-0.03(2)
nil
0.00075-0.05(5)
Total C Fpk
nil
nil
nil
0.00009-0.005(2;
Total cJF£g
nil
nil
nil
nil
Total C1oF2Q
£ G
4.43
4.62
4.21
4.231
Molecular Wt., Gas Density
194.3
194.3
189.
192.3
(a) These three samples were j
analyzed before the
addition of the high temperature
squalane column and
do not have complete analyses of the high molecular weight compounds. (b) This tube was copper,
(c) This sample was a single phase. (d) Number in brackets is number of peaks in group.

TABLE 18 PART 4 (continued)
Tube No.
Sample Wt., Gms.
Bulk Density, gm./cm.
No, Days
Integrated Dose, 10 ^ Roent.
Energy Absorbed, 10^ ergs
Energy Abs./gm., 10 ergs
Compound in Assay or as Listed
45
0.5297
0.634
23.0
6.66
31.3
59.2
26
0.4903
0.587
33.0
13.5
58.8
120.0
Mole Fraction-
6
0.3627
0.434
30.0
12.92
41.7
115.0
-G Value'd)
48
0.3081
0.369
23.0
9.80
26.8
87.0
Moles Fp/mole mixtureG Value
0.0134-1.12
0.041*5-1.78
0.0255-1.09
0.0210-1.18
CFjj
0.0108-0.90
0.0199-0.80
0.0170-0.73
.'.0.0134-0.75
C2F6
0.0107-0.89
0.0155-0.62
0.0156-0.67
0.0113-0.63
c2f4
nil
nil
nil
nil
C2F2
0.0002-0.02
0.0002-0.01
0.0004-0.02
0.0001-0.006
C3F8
0.9436
0.9221
0.9097
0.9378
n-C^F?
nil
nil
nil
nil
cplo
0.0145-1.21
0.0177-0.71
0.0223-0.95
0.0134-0.75
n-CjFg
0.0037-0.31
0.0036-0.14
0.0064-0.27
0.0031-0.18
i-C^F^
0.0058-0.48
0.0059-0.24
0.0090-0.39
0.0054-0.31
nc6fi4
2-CFoC5F1]l
0.0013-0.11
0.0018-0.07
0.0032-0.14
0.0014-0.07
0.0027-0.22
0.0030-0.12
0.0064-0.27
0.0026-0.15
3_CF3C5F]1
2,3-(CF|)|cJg
Total C?Fi6
Total CoFT'o
Total CqF^q
0.0003-0.03
0.0003-0.01
trace
0.0003-0.02
0.0025-0.20
0.0023-0.09
0.0038-0.16
0.0023-0.13
0.0014-0.12(2)
0.0025-0.10(2)
0.0023-0.10(2)
0.0024-0.14(2)
0.0010-0.08(2)
0.0019-0.08(2)
0.0017-0.07(2)
0.0019-0.11(2)
0.00062-0.06(2)
0.0012-0.05(2)
0.0017-0.07(3)
0.0015-0.08(2)
Total C10F22
0.00053-0.05(4)
0.0014-0.05(2)
0.0003-0.01(1)
0.0017-0.09(5)
Total CllF2,
nil
0.00033-0.01(3)
nil
0.0003-0.02(3)
Total C12F2g
nil
0.00029-0.01(2)
nil
0.0005-0.03(3)
Total C-. oF0q
£G13 28
5.80
4.89
4.94
0.0003-0.02(3)
4.666
Molecular Wt., Gas Density
195.3
213.
193.
191.1
(a) These three samples were analyzed before the addition of the high temperature squalane column and
do not have complete analyses of the high molecular weight compounds. (b) This tube was copper,
(c) This sample was a single phase. (d) Number in brackets is number of peaks in group.

TABLE 18 PART 4 (continued)
Tube No.
44
4l
92(c
Sample Wt., Gms.
0.5281
0.5419
0.0295
Bulk Density, gm./cm.^
0.633
0.649
0.0353
No. Days
23.0
10.0
9.92
Integrated Dose, 10 Roent.
Energy Absorbed, 10 g ergs
Energy Abs./gm., 10 ergs
9.02
3.83
1.705
42.3
18.5
0.446
80.2
34.2
-G Value'
Compound in Assay or as Listed
Mole Fraction-
Moles Fp/mole mixtureG Value
0.0160-0.98
0.0120-1.74
^ zero
CFj,
0.0148-0.91
0.0043-0.62
0.0013-0.
C2F6
C2F4
C2F2
C3F8
n-C3F6
C4F10
D"n5>
1-C5F12
n_c6Fl4
2-CF C Fn
3-CF3C5Fll
2,3-(CF ) C^Fg
Total C7Fi8
Total C8F7q
Tota! C F
Total C£0F22
Total C,,F2,
Total C12F26
Tota^ Ci3F28
Molecular Wt., Gas Density
0.0105-0.64
nil
0.0001-0.007
0.9395
nil
0.0139-0.
0.0032-0.
0.0053-0.
0.0014-0.
0.0028-0.
0.0004-0.
0.0024-0.
0.0016-0.
0.0013-0.
0.00093-0
0.0008-0.
0.0003-0.
0.0002-0.
4.677
193.
0.0043-0.62
nil
nil
0.9729
nil
86 0.0071-1.03
20 0.0017-0.25
33 0.0027-0.39
08 0.0008-0.11
17 0.0018-0.26
03 nil
15 0.0011-0.16
10(2) 0.0010-0.15(2)
08(2) 0.0009-0.13(2)
.06(2) 0.0004-0.06(2)
05(6) 0.0004-0.06(4)
02(2) 0.0001-0.02(1)
01(1) 0.0007-0.10(2)
5.70
191.4
0.0006-0.20
nil
nil
0.9976
0.0005-0.16
trace
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
nil
0.78
188
(a)These three samples were analyzed before the addition of the high temperature squalane column and
do not have complete analyses of the high molecular weight compounds, (b) This tube was copper,
(c) This sample was a single phase. (d) Number in brackets is number of peaks in group.

TABLE l8 PART 5. C^Fg WITH POWDERED ALUMINUM IRRADIATED BY COBALT-60 GAMMA PAYS
Tube number = 85
Sample Wt. = 0.4237
Weight of Aluminum Powder = 0.1327 gms.
Sample Density Corrected for Aluminum Powder = 0.584 gm./cm.^
Bulk Density of Aluminum Powder = 1.22 gm./cm.3
Temperature = 99C
No. Days = 10.25
Integrated Dose = 1.52 x 10^ Roent.
Energy Absorbed in Fluorocarbon, 10 ergs = 5.74
Compound in Assay or as Listed
Mole FractionG Value
(a)
1
R
C16H34
CFU
C2F6
C2F4
C2F2
c3F8
n_c3F6
c4fio
0 = 54.5 cc. H2 on
n-C5F12
1^5^
n-C6Fl4
27CF3?5F11
2,3-(CF3) C4Fq
IT] ^16
Total CoF,o
so818
Molecular Wt., Gas Density
(b)
0.0006-0.22
0.0004-0.15
0.0010-0.37
0.0005-0.19
0.9877
0.0007-0.26
0.0013-0.48
0.0050-1.88
0.0004-0.15
0.0010-0.37
0.0002-0.07
0.0005-0.19
0.0003-0.11
0.00037-0.14
0.00013-0.05
4.63
192.5
vo
-4
"(a) Number in brackets is number of peaks in group,
(b) This material comes out with CgFg on silica gel.

TABLE l8 PART 6.
Tube No.
Sample Wt., Gms.
No. Days
Integrated Dose, 10^ Roent.
Energy Absorbed, 10 ergs
Compound in Assay or as Listed
n-C4F10 IRRADIATED BY COBALT-60 gamma RAYS, T = 30c
95 (a)
0.384 to 0.213
9.83
4.0
13*73
Mole FractionG Value'0'
96
0.3642
9.83
4.02
13.07
Moles F^/mole mixtureG Value
CFh
C3F8
n-c4Fl0
n-C5F12
i-C5F12
n-C6Fi4
2-CF C Fn
3"CF3C5F
Total
Total
Total
Total
C7Fl6
C8Fl8
c9F20
C10F22
Total 02^24
Total
C12F26
Molecular Wt., Gas Density
O.Ol45-l.2
0.0046-0.51
0.0023-0.26
0.0033-0.37
0.9626
0.0022-0.24
0.0035-0.39
0.0011-0.12
0.0001-0.01
0.0019-0.21
0.0001-0.01
0.0052-0.58(2)
0.0053-0.59(3)
0.0031-0.34(4)
0.00145-0.16(4)
0.0013-0.15(4)
0.00205-0.23(2)
5.79
259.
0.0105-1.l6
0.0056-0.62
0.0024-0.27
0.0029-0.32
0.9644
0.0017-0.19
0.0037-0.41
0.0012-0.13
0.0002-0.02
0.0021-0.23
0.0002-0.02
0.0061-0.67(3)
0.0065-0.72(3)
0.0011-0.12(4)
0.0010-0.11(4)
0.00055-0.06(4)
nil
5.05
239.
VO
CD
(a) Approximately 1/3 of this sample leaked out of the sample tube prior to analysis.
(b) Number in brackets is number of peaks in group.

TABLE 18 PART 7* n-c5F12 IRRADIATED
Tube No. 13
Sample Wt., Gms. 0.3762
No. Days 33.2
Integrated Dose, 10' Roent. 157
Energy Absorbed, 10 ergs 52.7
Temperature, C 30
Compound in Assay or as Listed
Moles F2/mole mixtureG Value
CFU
C2F6
C2F2
C3F8
Cy-a3F6
n"c4F10
n"C5F12
i-C^F12
nc6FlU
2_cf3C5F11
3-Cf3C5f11
2,3-( CF^oCi F_
Total"-2-1' 8
Total
Total
Total
Total C^F^
Total C22F2g
Total C^2?28
2 G
Molecular Wt., Gas Density
^7Fl6
C8Fl8
C9F20
C10F22
0.0305-0.68
0.0163-0.36
0.0099-0.22
0.0004-0.01
0.0106-0.23
trace
0.011*6-0.33
0.8758
0.0015-0.04
0.0066-0.15
0.0062-0.14
0.0075-0.17
nil
0.0091-0.20(3)
0.0102-0.22(3)
0.0123-0.27(2)
0.0152-0.33(2)
0.0026-0.06(4)
0.0012-0.03(4)
0.00026-0.006(2)
3.446
303.
(a) Number in brackets is number of peaks in group.
COBALT-60 GAMMA RAYS
11
0.3797
9.9
4.8
16.3
30
Mole FractionG Value
78
0.4384
9.9
3.65
14.35
0.0160-1.15
0.0043-0.31
0.0026-0.19
nil
0.0026-0.19
nil
0.0026-0.19
0.9539
nil
0.0016-0.12
0.0019-0.14
0.0027-0.19
nil
0.0072-0.52(3)
0.0047-0.34(3)
0.0061-0.44(2)
0.00864-0.62(2)
0.00091-0.06(4)
0.00021-0.01(2)
nil
4.47
301.
0.0105-1.05
0.0035-0.35
0.0036-0.35
nil
0.0028-0.28
nil
0.0025-0.25
0.9568
nil
0.0024-0.24
0.0025-0.25
0.0029-0.29
nil
0.0073-0.73(3)
0.0045-0.45(3)
0.0047-0.47(2)
0.0056-0.56(4)
0.0006-0.06(4)
0.0002-0.02(1)
nil
5.35
294.
V)
VO

TABLE 18 PART 8.
Tube No.
Sample Wt.} Gms.
No. Days
Integrated Dose,
Energy Absorbed,
Temperature, C
Compound in Assay or as Listed
cyclo-c5f10 IRRADIATED BY COBALT-60 GAMMA RAYS
10^ Roent.
10 ergs
TU
0.4916
9.83
3.52
15.5
99
21
0.2102
18.9
8.95
16.9
30
Mole FractionG Value
(c)
x(a)
0.6277
32.2
13.0
72.9
30
O.OOO9-0". 04
0.0012-0.06
nil
0.0003-0.01
0.0006-0.03
nil
0.0011-0.05
0.0006-0.03
0.0008-0.04
nil
0.0058-0.26
0.9386
nil
nil
nil
0.0023-0.10
nil
0.0009-0.04(1)
0.0053-0.24(2)
0.0045-0.20(2)
0.0017-0.07(2)
0.0138-0.62(5)
0.0155-0.71(4)
0.0018-0.08(3)
0.0025-0.11(4)
0.0008-0.04(1)
2.73
264.5
CF4
C2F6
c2f4
C2F2
C3F8
n-C3F6
c4f10
n-c5F12
i-CrFi2
VR = 167 cc. H2(n-hex)
VR = l88 cc. Hp(n-hex)
Cy-C5F10
VR = 281 cc. H2(n-hex)
Vr = 321 cc. Hp(n-hex)
VR = 351 cc. H2(n-hex)
VD = 403 cc. Ho(n-hex)
D w> \~\~4 '
VR = 456 cc. Hp(n-hex)
Total CyF^lb)
Total CgFjg
Total C^F2q
Total C^qF22
Total C11F2i|
Total C12F26
Total C-^Fgg
Total C^FgQ
Total C-, £-F^0
S.G i
Molecular Wt, Gas Density
0.0001-0.01
0.0001-0.01
nil
nil
0.0001-0.01
0.0001-0.01
0.0001-0.01
0.0003-0.04
0.0002-0.03
nil
0.0024-0.28
0.9719
nil
nil
nil
0.0013-0.15
nil
0.0005-0.06(1)
0.0031-0.36(2)
0.0021-0.24(2)
0.0009-0.10(2)
0.0048-0.56(5)
0.0086-1.01(4)
0.0010-0.12(3)
0.0011-0.13(4)
0.0009-0.10(4)
3.23
262.
of air.(b)Theserefer
appearance correlation. (
0.0025-0.08
0.0019-0.06
nil
0.0011-0.03
0.0010-0.03
nil
0.0010-0.03
0.0009-0.03
0.0032-0.10
0.00008-0.003
0.0086-0.26
0.9032
0.0012-0.04
0.0005-0.01
0.0012-0.04
0.0059-0.18
0.0024-0.07
0.0010-0.03(1)
0.0123-0.37(2)
0.0089-0.27(2)
0.0073-0.22(3)
0.0148-0.45(4)
0.0180-0.54(3)
0.0016-0.05(4)
0.0011-0.04(2)
0.00025-0.01(1)
2.943
276.
(a) This sample contained a few mi
appearance times corresponding
is number of peaks in group.
llimeters pressure
to the non-cyclic
only to materials having
c) Number in brackets
100

TABLE l8 PART 9. a-CgF,^ IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.
Sample Wt. Gms.
No. Days
Integrated Dose, lol Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in Assay or as Listed
32
0.481+5
30.0
14.05
6o. 8
30 ( )
Mole FractionG Value'a'
Moles F2 lost/mole mixtureG Value
CFh
C2F6
c2F2
C fq
C4F10.
n-CcF!2
i-C5Fi2
nc6FlU
VR = 338 cc. H2(n-hex)
Total CjF^g
Total CgF^g
Total C^Fgo
C10F22
Total
Total C-QF24
Total C1?F?
Total C33F2q
Total Cll+F30
2 0
Molecular Wt., Gas Density
0.0350-0.T3 0
0.0155-0.32 0
0.0113-0.23 0,
trace
0.013^-0.28 0
0.0116-0.24 0
0.0093-0.20 0,
0.0007-0.02 0,
0.8211 0
nil 0
0.0250-0.52(2) 0,
0.0198-0.41(2) 0,
0.0190-0.39(2) 0
O.OI65-O.3^(4) 0
0.0179-0.37(3) 0
0.0170-0.35(4) 0
0.0017-0.04(2) 0
nil 0
4.44 6
362. 369
(a) Number in brackets is number of peaks in group.
81
0.4605
9.83
3.91
16.03
99
0220-1.62
0031-0.22
0026-0.20
nil
001*3-0.32
001*3-0.32
0032-0.23
0002-0.02
9313
0005-0.01*
0111-0.83(2)
0077-0.57(2)
0085-0.62(2)
0051-0.37(3)
0055-0.1*0(1*)
0091-0.67(5)
0038-0.28(5)
0004-0.03(1)
T3
101

TABLE 18 PART 10. 2-CF3C5F1]L AND 2,3-( CF^C^Fg IRRADIATED BY COBALT-60 GAMMA RAYS
Tube No.Material
Sample Wt., Gms.
No. Days
Integrated Dose, 10 7 Roent.
Energy Absorbed, 10 ergs
Temperature, C
Compound in assay or as listed
93 2-CFoCc-F-] -i
0.42&T
9.83
4.10
9h 2,3-(CF ) C.F
0.^133 2 4 8
9.83
4.13
15.T
30
Mole FractionG Value
(a)
15.3
30
Moles F2 lost/mole mixtureG Value
CFh
n2l2
c3f8
C^Fio (Mostly iso)
= 100 cc. H2(n-hex)
ll4 cc. H2(n-hex)
Vo
R
V
VR
nc5F12
i_C5F12
228 cc. H2(n-hex)
nc6fiU
2-Cf3c5F11
3"CF3C5Fll
2,3-(CF^)^CI|P8
Total ~ "
Total
Total
Total
Total C^1F24
Total
Total C^3F25
Total
z
C7Fl6
c8f18
9 20
C10F22
Cl4F30
Molecular Wt., Gas Density
(a) Number in brackets is number of peaks in group.
0.0195-1.16
0.0059-0.45
0.0019-0.14
nil
0.0018-0.13
0.0018-0.13
nil
trace
0.0024-0.18
0.0011-0.08
nil
0.0002-0.02
0.9386
nil
0.0005-0.04
0.0087-0.65(2)
0.0043-0.33(3)
0.0068-0.51(2)
0.0078-0.59(3)
0.0094-0.71(3)
0.0047-0.35(4)
0.0041-0.31(2)
nil
6.08
345.
0.0370-2.73
0.0098-0.73
0.0004-0.03
0.0006-0.05
0.0025-0.19
0.0004-0.03
0.0015-0.11
nil
0.0004-0.03
0.0021-0.16
0.0003-0.02
trace
0.0010-0.07
0.0009-0.07
0.9063
0.0160-1.18(4)
0.0128-0.94(2)
0.0065-0.47(2)
0.0082-0.60(4)
0.0048-0.35(3)
0.0159-1.17(5)
0.0032-0.23(4)
0.0063-0.47(2)
9.63
342.
102

SUMMARY AND FUTURE WORK
In summary we can say that in the irradiation of pure saturated
fluorocarbons the important chemical reactions in the sample are re
combination of radicals, disproportionation of small radicals, addi
tion of fluorine and radicals to unsaturated molecules, and the reaction
of radicals with molecular fluorine. These reactions differ from those
found in hydrocarbon work by the absence of abstraction reactions and
the presence of reactions between radicals and molecular fluorine.
These two differences are the result of bond energy differences between
hydrocarbons and hydrogen, and fluorocarbons and fluorine. Molecular
fluorine, fluorine radicals, and other radicals can be removed from the
reaction system by wall capture. The reaction between radicals and
molecular fluorine depends on the density, total energy absorbed, and
the distance to the tube wall. One of the consequences of this type of
reaction is that the G values for molecules smaller than the parent
increase with both sample density and total energy absorption. Capture
of radicals other than fluorine by the sample tube wall becomes impor-
tant at low density (below about 0.2 gm./cm.).
When fluorocarbons are irradiated to high energy absorption,
as in a nuclear reactor, chemical equilibrium considerations are
important. The equilibrium depends on the florine-to-carbon ratio
of the sample with a ratio of four resulting in nearly pure per-
fluoromethane. Smaller ratios yield less perfluoromethane. For
105

104
this reason for long irradiations perfluoromethane appears to be
extremely stable.
The initial G values for disappearance of the parent molecules
are related to each other in much the same way that the sensitivities
to electron bombardment in the mass spectrograph are related. This
is not unexpected since, in both cases, the species causing bond rup
ture is electrons. Thus, given the sensitivities of a known and an
unknown fluorocarbon and the G value for the known fluorocarbon the
G value for the unknown can be estimated as the ratio of the sensitiv
ities times the known G value.
The G values for disappearance of the parent compound are ap
proximately; cp^, 1.5; c2?6, 5.25; c^Fg, 4.5; n-c^Q, 5A; n-5F12
4.9; cyclo-C^F10, 5*0; n-C^F^, 6.0; S-CF^C^F^, 6.0; and 2,5-(CF^)2C^FQ,
9.6. These values are only approximate due to the dependence of cer
tain reactions on density and amount of energy absorbed.
The important products in perfluoromethane irradiation are per-
fluoroethane and perfluoroacetylene produced in about equal amounts.
For the other materials for short irradiations the major products are
saturated and are those predicted by simple bond rupture and radical
recombinations. The results for C2F£ and larger molecules, excluding
cyclo compounds can be correlated by statistical recombination calcula
tions, using empirically determined radical effectiveness numbers.
Since these effectiveness numbers are the same for radicals formed in
the same way this method can be used to predict product distributions
for other fluorocarbons. In conjunction with prediction of the overall
G values from mass spectrometer sensitivities the G values for the
products can be estimated.

105
More work is needed in further investigation of other fluoro
carbons to study the density effect and energy absorption effects of
the type carried out on perfluoropropane.
A number of interesting experiments could be done to investi
gate the equilibrium encountered in the LITR irradiations. Two of
these have already been mentioned; irradiation of cyclo-C^F^Q with
additional fluorine to give a fluorine-to-carbon ratio of four to see
if this results in nearly pure OF^., and irradiation of mixtures of
powdered graphite and fluorine.

Appendices

APPENDIX 1. ENERGY ABSORPTION
Flux Determination
The variation of the flux over the length of the sample tube
was determined using benzene in water solutions as described by-
Johnson and Martin. In brief, this method requires the irradia
tion of water saturated with benzene. Subsequently the solution is
divided and one portion diluted with distilled water and the other
with approximately 0.08 N sodium hydroxide solution. The amount of
phenol formed is determined from the ultraviolet optical densities
at 2900l of the neutral and alkaline solutions. This work vas done
with a Beckman DK-2 Spectrophotometer. The yield of phenol is
2.14 + 0.02 molecules per 100 electron volts. This yield is insen
sitive to dose up to about 70,000 to 80,000 rep. Over the range
100-7000 rep./minute it is independent of dose rate, and above 0.007 M
it is insensitive to benzene concentration. For this reason it pro
vides a simple method of measuring gamma flux.
In this work solutions were sealed in polyethylene bottles
3.20 cm. high and about 2.5 cm. in diameter. There were empty spaces
0.34 cm. at the bottom and 0.22 cm. at the top. The average thickness
of the side of the container was 0.293 cm. Ten of these bottles were
stacked and retained in the center of the sample area. They were
enclosed in a stainless steel irradiation basket with a wall thick
ness of approximately 0.11 cm. When the correction for absorption
107

108
in the stainless steel and polyethylene are applied Vie find that the
radiation exposure of the solution should be 93*4 percent of that
in the empty radiation tube. This tube is designated as number 11 of
the engineering dobalt-60 source.
One irradiation of 20 minutes was made to determine the flux
in the high radiation zone and one run of 60 minutes was made to
obtain points in the lower portion. This gives a total of six useful
points over the distance of 18 cm. where a significant amount of flux
is encountered. A plot of the flux curve is given in Figure 27 where
the zero point is the bottom of the sample space. This corresponds
to a point 2.4 inches below the center line of the cobalt-60 rods.
Flux Exposure of Samples
The flux to which the samples are exposed was somewhat less
than that in the unoccupied cavity.
In the ambient temperature experiments the flux is absorbed
in the metal of the stainless steel baskets and the wall of the
aluminum sample container. The metal basket has a wall thickness of
about 1.1 millimeter of steel and the aluminum wall thickness is 0.655
millimeters.- This leads to a flux exposure of the sample equal to
94.4 percent of that in the unperturbed radiation zone.
At elevated temperature, in addition to the absorbing layers
present at ambient temperature, there was present in the heating
device J.h millimeters of glass and the equivalent of about one milli
meter of compacted asbestos (density 2.5)* Since a large number of
tubes were irradiated at the same time, the average thickness of
aluminum encountered by a gamma ray is estimated to be 1.2 millimeters.

109
Figure 27. Flux in tube 11 of the engineering cobalt-60
source.

110
From this data the flux for the samples can be shown to be about 88
percent of that in the unperturbed radiation zone.
Ambient temperature
^d (.11)(7.7) + (.0655)(2.7)
steel + aluminum
- 1.019
= .058
Id ^-(.058)(1.019)
Io "
9^4
Elevated temperature
2yd = (,54)(2.25) + (.ll)(7.7) + (.12)(2.7) + (.10)(2.5)
= glass + steel + aluminum + asbestos
- 2.19
Id _-(.058)(2.18)
.881 = .88
Variation of Flux Across Sample Zone
The variation in the flux across the radiation zone is not
easily measured since accurate measurements would require either a
very small probe or small amounts of a chemical dosimeter. Because
of the difficulties involved in either of these approaches it was
decided to approach the problem by a simple mathematical model.
If we neglect scattering and absorption of the photons by
material in their paths the radiation from a point source will follow
the inverse square law. Using this assumption it is possible to
develop an equation allowing one to calculate the relative intensity
of radiation throughout a volume given the geometry of the source.

Ill
In Figure 28a are shown the various distance and angles
required for calculating the relative flux for a slice of element.
We wish to calculate the flux received at A from B and integrate
for all values of B over the element.
The value of D, the distance between points A and B can be
found by vector addition.
~R Ri
~~q q cos

r = r cos G i + r sin 3
R + q + D- r =
D -~T R ~q =
0
i(r cos & R q cos ty) + j(r sin 9 q sin 9)
The distance D is then found to be
|d|-V7T cos R q cos + (r sin ~0 q sin
-V777T7T 2 q(cos 9 cos

+ 2R(q cos f> r cos 9)
Since there are six cobalt-60 rods in the array in integrating
over subsequent rods Q is increased by JO0.
Since this method can only give relative values we will now
compute the ratio of flux contributions at point A from point B as
compared to that at the center of the radiation zone.
The distance from point B to the center of the radiation zone
is obtained by vector additions

112
Radiation zone
Figure 28a.
Diagram for relative flux across sample zone
calculation-planar.

115
M R + q = i(R + q cos ty) + j (q sin
|m I Vr2 + q2 cos2 9+2 Rq cos Q> + q2 sin2
'n/r.2 + q2 + 2Rq cos ^
The fraction of the center contribution received by point A
from point source is:
Jjvj|2 R + q + 2Rq cos (p
Id|2 R2 + q2 + r' 2rq(cos & cos <£> + sin (p sin &) + 2R(q cos (p r cos
The average contribution from the Area of the element is found
by integration over the area.
(J-1)
where qA = radius of element
Rp relative flux at A
The integral must also be extended over the length of the
cobalt-60 element. In Figure 28b the distances required are defined.
The relative intensity at A referred to the center is:
N2 (Z h)2 + M2
N = V7z h)2 + M2 '
P2 D2 + (Z h)2
P Vd2 + (Z h)2

114
i
Radiation zone
Z-H
A
Z-h
A
h
'r
Z=Q)
Figure 28b. Diagram for calculation of relative flux
across sample zone-volume.

115
In Equation J-l we now malee the substitutions:
2 2
N for M
2 2
P for D
and integrate over the height of the element.
q r2^rH
['R2 + q^ + 2Rq cos + (Z h)^7'lqd^ dZ
When this equation is summed over the six cobalt-60 elements
the result is the relative flux at the point specified by r, 6, and
h to the center of the irradiation space.
Unfortunately this equation is too complex for ordinary solution
and would require use of a computer. The equation can be integrated
once by an appropriate substitution of variables but this form is still
too cumbersome.
In order to obtain an estimate of the percent variation across
the irradiation zone without recourse to a computer it is necessary
to study simpler cases. Two of these will be considered: the infinite
shell and the infinite wire. For the case of the infinite wire the
flux can be shown to be inversely proportional to the distance from
the wire. A graphical method has been used to obtain the relative
flux across the radiation zone for infinite wires at the center of the
cobalt-60 rods and for infinite wires at the projection of the rods
defined by tangents to the central tube and the cobalt-60 rods as is
shown in Figure 29. The results of these infinite wire calculations
are shown in Figure JO.

116
i
Figure 28c. Diagram for calculation of relative flux
across sample zone by infinite shell model.

117
Figure 29. Arrangement of cobalt-60 rods around tube number
11 of the engineering cobalt-60 source.

Flux relative to center
118
Inches from center
Figure 30. Relative flux across sample zone by infinite wire and
infinite shell models.
(a) Infinite shell at inner edge of rods.
(b) Inf initef'shell at center of rods.
(c) Infinite shell at outer edge of rods.

119
In using the infinite shell approximation a similar development
to the more complex case is required.
The distances required are defined in Figure 28c.
L2 = r^ + D2 2rD cos &
Flux at A = D
Flux at Center L
The average flux at A is given by
2 'if
(D)(D/2)d?
Rr
7?=oVr2 + D2 2rD cos &
71 D
D_
217
60
+/r2 + d2 2rD cos 0
If we let P r/D and simplify we obtain:
2# 7
/ 2 1
V 1 + P 2P cos 0
Where P = 0.5 with graphical integration we obtain:
Rp 1.068
In Figure JO the points corresponding to P = 0.5 for the shell
at the near edge of the rods, at the center of the rods, and at the
far edge of the rods are plotted.

120
The plot indicates the flux at the edge of the sample container
was probably about 10 percent higher than at the center of the con
tainer. However, since the samples were generally confined near the
center of the sample space the flux should not vary more than + 5 per
cent from the value determined by the benzene-water dosimeter.
Absorbed Dose
Doses in radiological work were formerly expressed in roentgens
exposure. A preferable method is the absorbed dose. In order to
calculate this quantity one must know what fraction of the exposed
dose is absorbed in the material under study.
If wall effects are neglected the absorbed dose may be found
(29)
from the following equation,
abs
0.875 O^f) x N medium x e+ 6 9 ^medium
(X-l)
N air + eY* e/(air
where:
W average energy expended by the ionizing particles per
ion pair produced in air
(50)
= approximately 55*55 +1*5 percent
N = number of electrons per gram
eira> e'T'f e/(=< true gamma absorption coefficients per electron
Dabs = rads absorbed per roentgen exposure
The values for the absorption cross sections per atom for
twenty-four elements are tabulated in reference 51* The sum of these
is listed as/i in Table 19 where is the number of electrons per atom,

121
TABLE 19. GAMMA ABSORPTION CROSS-SECTIONS PER ELECTRON
Atom
A
/*
hl
H
1
.0927
.0927
Be
4
.572
.0950
C
6
557
.0928
N
7
.650
.0928
0
8
.745
.0952
Na
11
1.021
.0929
Mg
12
1.111
.0927
A1
15
1.210
.0952

122
The numbers in the fourth column, the cross-section per electron,
ore seen to be essentially constant for all atoms over the range of
interest. For this reason the last term can be deleted from equation
K-l and we have:
Dabs 0-875 (K-2)
N air Jk
Tire number of electrons per gram can be calculated from the
following equation
N. ** ^av ^ ^ (K5)
Mi
where
25
Nav Avogadro's number = 6.02 x 10 ^
= molecular or atomic weight of ith species
= number of electrons per atom or molecule
These values are listed in Table 20.
From these values of number of electrons per gram and equation
K32, we may now calculate the energy absorbed in the materials of
interest neglecting the wall effect and any density effect (Table
21). These two considerations will be examined later.
When a sample is contained in a relatively small cavity a
/
significant portion of the energy absorbed in it can originate as
energetic electrons in the wall material. Let us examine what occurs
in a sample being irradiated.
When a gamma ray is absorbed in either the tube wall or the
sample secondary electrons are set free which actually cause the
radiolytic chemical changes. Some of the electrons originating in

125
TABLE
20. NUMBER
OF ELECTRONS
PER GRAM
Material
Mi
i
NA X 10 25
Air
l4.4l
7.21
5.01
AL
26.97
15
2.90
cf4
88
42
2.87
c2f6
158
66
2.88
c5f8
188
90
2.88
C4F10
258
114
2.88
c5f12
288
158
2.88
c5fio
250
120
2.89
0Pl4
558
162
2.88
Cu
65.54
29
2.75

124
TABLE 21.
RADS ABSORBED PER ROENTGEN NEGLECTING
WALL AND DENSITY EFFECTS
Material
Rads/Roentgen
Aluminum
0.8p2
Copper
0.78 9
cf4
0.824
g2f6
0.826
c5f8
0.826
4f10
0.828
5f12
0.82 6
c6fi4
0.82 6
c5fio
0.828

125
the aluminum wall will be absorbed in the wall, some will be absorbed
in the sample, and some will pass through the sample to the opposite
wall. If the gamma ray is absorbed in the sample similar things can
occur. The electrons may be absorbed in the sample or may pass through
the sample and be absorbed in the wall. Thus there is an "atmosphere"
of electrons present in the sample space the density of which is
dependent upon the gamma absorption in the two media and the electron
stopping powers of the two media. If the sample space is large the
electron "atmosphere" will be characteristic of the sample material
and independent of the wall material. If the sample space is small
the electron "atmosphere" will be characteristic-of the wall material
and essentially independent of the sample. This is the basis of the
Bragg-Gray cavity law.
In order for the Bragg-Gray cavity law to apply rigorously two
conditions must be met:
(1) The intensity of the primary radiation must be essentially
uniform over the region from which the secondary electrons can reach
the cavity;
(2) The secondary electrons should be nearly uniform in the
region of the cavity.
The first condition is met in this case due to the geometry
of the source. The intensity will not vary much over 10 percent across
the area used for irradiation as is shown in Flux Exposure of Samples"
in this appendix.
The second condition requires that the range of the secondary
electrons be less than the wall thickness and that an insignificant
portion of them are absorbed in passing through the cavity. Neither

126
of those requirements is exactly fulfilled under the irradiation con
ditions. The thickness of the aluminum is not sufficient to stop
electrons having an energy of 0.5 mev. or greater^2) and traveling
perpendicular to the wall. And except in the samples containing no
liquid there will be a significant amount of electron absorption in
the sample.
From the above discussion Vie see no clear-cut indication as to
whether to include wall effects. However, the authors cited state
that deviations from the conditions for rigorous application are often
acceptable. For this reason we will use the Bragg-Gray values realiz
ing this may introduce an error of at most 8 percent in the G values
calculated.
The Bragg-Gray cavity equation, for this case, may be written:
(50)
ef eas
(K-4)
where:
E^ = energy deposited in the aluminum
E <= energy deposited in the fluorocarbon
F
S = average electron stopping power ratio of the fluorocarbon
to aluminum
Electron stopping power ratios relative to air are tabulated in
reference 55 and. stopping powers are tabulated in reference 54. In
(55)
general these numbers are correlatable to atomic number.'" Although
the stopping power of a material varies considerably with the energy
of the electron, the ratios are nearly constant except for very low
(50)
energies. For this reason the average stopping power ratio will
be nearly equal to the stopping power ratio at the mean electron energy

127
of approximately 0,5 mev. (55)* This value is plotted in Figure pi
versus atomic number.
In Table 22 are tabulated the materials of interest, their
average atomic numbers, and the stopping power ratios taken from
Figure 51.
From the figures in the last column of Table 22 and equation
(K-4) we may now calculate the energy absorption per roentgen as given
by the Bragg-Gray theory. These values are summarized in Table 25.
The values are about 8 percent higher than those calculated from
simple gamma absorptions.
A final consideration in the energy absorption calculation is
the effect of density. In general the absorption of energy from a
charged particle in a dense medium is less than that in a diffuse
medium due to polarization effects. This correction amounts to about
2 percent in water exposed to 1 mev. electrons.
(50)
This correction
(35)
is related to the dielectric constant of the medium and since
(5 6)
fluorocarbons have low dielectric constants relative to water
this correction will be less than 1 percent $nd can be neglected.
Energy Absorption in Samples
The calculation of the energy absorbed in the single phase
samples is straightforward, involving only integration to find the
average flux over the tube and multiplication by the exposure time.
For the samples containing two phases the average flux must be weighted
according to the weight fraction of the flux in the liquid and gas
phases. In this calculation the liquid is assumed to occupy the
bottom portion of the tube. The calculation requires knowledge of

Stopping power relative to air
128
Atomic number
Figure 31o Electron stopping power relative to air versus
atomic number

129
TABLE 22. ELECTRON STOPPING POWER RATIOS RELATIVE TO AIR
FROM FIGURE 51 FOR MATERIALS OF INTEREST
Material
At. No.
Stopping Power
Relative to Air
Stopping Power
Relative to A1
A1
15
0.922
1.0
cf4
8.4
0.9&5
1.067
2F6
8.25
0.986
1.07
5F8
8.18
0.986
1.07^
4F10
8.18
0.987
1.071
5f12
8.12
0.988
1.072
c5fio
8.0
0.990
1.075
6F14
8.10
0.988
1.072
Cu
29
0.780

a. Stopping power of
C^Fq relative to
copper = 1.264

150
TABLE 25. ENERGY ABSORPTION BY BRAGG-GRAY CAVITY THEORY
Molecule
ergs/gm. roent.
cf4
88.7
C2F6
89.0
5f8
89.0
G4F10
89.2
c5f12
89.5
5fio
89.4
6Fl4
89.5
C^Fq in Cu
99.8

151
the vapor and liquid densities of the materials as a function of
temperature. Then by use of the two following equations the weight
and volume of liquid and gas can be determined.
VLfL + Vg c ^
where:
= volume of liquid
Vq = volume of gas
L = density of liquid
jOq ~ density of gas
wt = weight of sample
Vrp = volume of sample tube
The liquid and gas densities are from references 12 and 57*
The weight average flux is then determined by
_ flwl + fgwg
wt
where:
F = weight average flux
F^ = flux over the volume occupied by the liquid
F flux over the volume occupied by the gas
G
This equation neglects the density effect, but as has already
been indicated this should introduce errors of less than 1 percent.
A more serious error may be the assumption that all the liquid is in
the bottom of the tube. It is possible that small drops of liquid
clinging to the walls above the liquid level caused some of the large
scatter of the data for these two phase samples. ^

152
One the energy absorption is known the G values (molecules/100
e.v. absorbed) are calculated. These are listed in Table 18.
Energy Absorbed in Reactor Irradiated Samples
Dosimetry in a nuclear reactor is much more complex than that
of a gamma source due to the mixture of radiation types having a
broad spectrum of energies.
The energy absorption in the Oak Ridge Graphite Reactor can
(
be estimated from the data of Richardson, Allen, and Boyle' for
energy absorption in graphite. Since 82 percent of the energy ab
sorption is from gamma rays the energy absorbed in a fluorocarbon per
gram will be essentially the same as is absorbed in graphite per gram.
With an appropriate flux correction from their position to ours the
rate of energy absorption is found to be approximately 2.7^ x 10^
erg/g.sec. Over the entire irradiation the energy absorbed will be
8.5 x 10^ erg/g. From this number the G values in Table 6 are cal
culated.
For the LITR reactor no estimate of the energy absorption is
available. However, such information is not necessary in this case
since the samples were degraded to equilibrium mixtures.

APPENDIX 2. SAMPLE CLEANUP AND CANNING
The major cleanup of all materials is accomplished by standard
gas chromatographic methods. After this treatment the only major
impurities present are air, carbon dioxide, and water. The sample
is distilled through calcium sulfate to remove most of the water,
caroxite to remove the and finally through magnesium perchlorate
to remove the last traces of water. Air is removed by alternately
thawing, freezing, and pumping on the sample until no residual un
condensable gas is observed. The apparatus for this treatment is
shown in Figure J2.
The first time a sample is to be canned the desired amount is
condensed in the calibrated tube and allowed to thaw to observe to
what manometer pressure this corresponds. Thereafter the amount of
sample can be determined by the pressure reading. The aluminum sample
tube, which has been previously baked out under vacuum using a cool
torch flame, is connected to the system by means of Flex fittings and
a copper tube force fitted into a section of rubber hose. The sample
is condensed into the aluminum sample tube with liquid nitrogen and
the tube pinched shut near the top after evacuation of any residual
sample. The tube is then heliarced at the pinch. After sealing the
tube is removed and weighed. In most cases they were then heated to
100C for several days to check for leaks.
155

To vacuum
system
Figure 32, Sample purification and canning system,

155
This method of sealing is quite satisfactory and has the dis
tinct advantage that the sample tubes are not removed from the vacuum-
system until they are sealed. This prevents recontamination with air
and water vapor.

APPENDIX 5. ANALYTICAL
Physical Description of Analytical Equipment
In the analysis of the irradiated samples a series' of three
gas chromatographs was used. Two Perkin-Elmer Yapor Fractometers
equipped with gas sampling valves and one rack-mounted detector-column
assembly were used. Thermal conductivity detectors with thermistors
were used in all three chromatographs. To allow easy gas sample
introduction the two Vapor Fractometers were equipped with Precision
Gas Sampling valves (Perkin-Elmer). For analyses a 25 cc. sample was
introduced through each valve. The pressure of these samples varied
from a few millimeters to about 80 millimeters of mercury; temperature
vas about 25C.
The columns used in the analyses were; 75*5 meters of squalane
operated at 92C, 16 meters of n-hexadecane operated at 25C, and one
meter of silica gel which was temperature programmed. The squalane
and silica gel columns were arranged to operate either in series or
parallel. Initially they were in series to allow the materials from
air to C^F^q to enter the silica gel column and were then separated
by a valve arrangement and run in parallel. Thus the silica gel
column was used to analyze for air, CF^, CgF^, C2F2*
n-C^F^, Cy-C^F^, and C^F^Q. The squalane column analyzed those
materials above C5> but was unable to separate all isomers. The
156

157
additional n-hexadecane column was used to analyze OjFg, C02, n-CjF6,
04^10 C^F12(isomers), isomers), C^F^(isomers). For C^F^q a
partial separation was accomplished, but only sufficient to give a
rough idea of the distribution between the two isomers.
Before introduction into the chromatographs the sample was
evaporated into a calibrated gas handling system in which it was
mixed and its average molecular weight determined. This system is
shown in Figure 55
Calibration of Molecular Weight System
In order to determine the molecular weight of a sample the
volumes of the various bulbs plus the volume of the connecting tubes
must be determined.
Figure 55 is a diagram of the decanning and storage system.
To calibrate a given bulb all other bulbs are cut off from
the system by closing their valves. The system plus the bulb is then
pumped down with valve 8 closed. Valves 8, 10, and 11 serve to define
the calibrated volumes and remain closed except as described.
Valve 7 is opened and the mercury level is read and recorded
as A. Valve 7 is closed and a small amount of air let into the
evacuated system. The levels of mercury in the leveling bulb and in
the gas burette are equalized as nearly as possible by eye and this
level is recorded as B. Valve 7 is then opened and the level attained
by the mercury in the gas burette is recorded as C. Since the surface
area in the leveling bulb is much greater than that in the burette
the corrected gas burette reading, D, is found by:
D=B+(B-C)=2B-C

158
Figure 33 Decanning and storage system.

159
The pressure in the system is read from the manometer and
recorded as A. Then assuming the ideal gas law holds the volume of
the system is calculated by:
P(atmosphere) x V(burette) = P(system) x V(system)
Additional volumes of air are admitted by the same procedure. This
yields a volume which shows a slight increase with increasing pressure
due to the increasing volume in the left leg of the manometer. Thus
the slope is known from the size of the tubing used in the manometer.
This is found to be .044 cc./mm. Fitting this slope to the four sets
of bulb calibration data yields the following volume equations:
Bulb 1
V 427.5 + .044 A
Ain millimeters of mercury
Bulb 2
V = 728 + .044 A
Bulb 5
V = 729 + .044 A
Bulb 4
V = 1289 + .044 A
Sample Treatment
The sample is introduced into the system by scratching the
aluminum tube and introducing it through a Flex fitting force-fitted
into a rubber tube. After the system is pimped down the aluminum
tube is snapped off and the sample distilled into the system. It is

then condensed in one of the four storage bulbs and allowed to mix by-
diffusion for at least 24 hours. Experiments performed indicate that
a representative sample is not obtained with less than about 6 hours
mixing. After the mixing period the sample is introduced into the
system with all valves closed except the one to the bulb and the one
to the manometer. Knowing the weight of the sample (determined by
weighing the sample tube with and without the sample), the pressure
read by the manometer, and the temperature it is possible to determine
the average molecular weight of the sample. After these measurements
have been made the sample is allowed to expand into the two evacuated
sampling valves (one not shown) preparatory to introduction to the
chromatographs.
Switching Columns from Series to Parallel
A schematic of the valves required for operation of the squalane
and silica gel columns in series and parallel is shown in Figure 24.
Valves A' and B1 are variable resistance and are used to replace
the resistance of columns A and B respectively when operating in paral
lel. These are necessary to maintain equal flows in the two modes
of operation. With valves 1 and 4 closed and valves 2 and 3 open the
two columns are in series and the two resistances are in series. Thus
if the resistances are correctly adjusted the two rotameters will read
equal flows. With valves 2 and 3 closed and valves 1 and 4 open the
columns are in parallel with columns A and resistance B1 connected
and column B and resistance A1 connected. The two rotameters should
now read equal flows and these should be the same as the series flows.
A rough adjustment of the resistances is made by equating these four

Rotameter 1
Rot ame te ir 2
B
,-cIj
X-
2
w
Sq.ualane
column
A
Silica
column
B
Figure 3U Series and parallel connections of chromatographic
columns

142
flow rates. A finer adjustment is made by balancing for zero shift
of the base lines of the recorders when shifting from series to
parallel.
Description of Analytical Columns
As has been previously mentioned whereas the n-hexadecane and
squalane columns were operated at constant temperature it was necessary
to program the temperature of the silica gel column to achieve satis
factory appearance times. The program and appearance times are given
in Table 24.
The conditions of operation and the standard retention volumes
of various materials are given for the squalane column in Table 25.
The conditions of operation and the standard retention volumes
of various materials on the n-hexadecane column are reported in Table
26.
Urea and Thiourea Columns
In the course of development of chromatographic columns, a
number of columns were prepared and tested, but were not used directly
in the analytical procedure. The most interesting of these were the
urea and thiourea columns.
These two columns were devised in an attempt to make use of
(38
the canal complexes they are known to make with numerous compounds.
If this occurred in a chromatographic column, a separation of isomers
would be possible resulting in the elution of the most highly branched
isomers first and the normal isomers last. This is indeed the case
with both these columns with approximately equal separation being

145
TABLE 24. TEMPERATURE PROGRAM AND APPEARANCE TIMES
ON THE SILICA GEL COLUMN
Gas H2
Flow rate = 5^.9 cc./min.
Column = 1 meter of silica gel
Program
Time
Initied
5 min. after air
12 min. after air
Temperature
2p C
increase to 70C
increase to 90C
Appearance Times
Compound
t-t min.
air
cf4
C2F6
0^2 C2F4
2F2
c5f8
cyclo-C^F^
n"5F6
4F10
1.1
7.0
8.7
5-4
14.4
17.0
18.5
21.7

144
TABLE 25. STANDARD RETENTION VOLUMES ON THE SQUALANE COLUMN
Pi =* 29.7 psi. Po = l6.8 psi. Temp, of Column = 92C
Carrier gas = 5^*9 cc./min. Hg at 1 atm. and 250
Column length = 15*5 meters of 168.7 grams of 0.p26 gm.
squalane/gm. Chromosorb-P.
Compound
V^0 cc. at 1 atm. and 25C
C2F6
9.17
c4f10
25
n-C^Fj g
28.9
n-C6Fi4
58.4
2 and J-CF^C^F-^
2,5-(0Pj)20j,P8
6Q.6
87.7
n-7Pl6
90.4
n8Fl8
159
n"9F20
195

145
TABLE 26. STANDARD RETENTION VOLUMES ON THE N-HEXADECANE COLUMN
Pi *= 21.7 psi. Po = 14.7 psi. Temp, of Column 25 C
Carrier Gas = 22 co./min. at 1 atm. and 25 C
Column Length = 16 meters containing 197.1 grams of 0.428 grams
n-hexadecane/gm. Chromosorb-P.
Compound
Vo cc. H0 at 1 atm.
K c.
C2F6
8.77
c5F8
29.8
n-C4Fio
87.3
^lO
74.5
n-C5F12
135
i-C5Fi2
168
n-6Fl4
284
2-CF505Fh
325
5-OF55Fll
351
2,5-(cf5)2c4f8
412
n-7Pl6
554
2-CF500Fii'
573
Probable
5-CF5C6Fn
626
Identifications
J-VsVuj
693
-jP6
103.5
cy-C^Fpo
218
co2
52.5

1 46
attained in each. Of practical interest was the accidental discovery
that thiourea (but not urea) gives good separations of the saturated
fluorocarbons and the unsaturated analogs with six carbons. The
n-hexadecane columns are not capable of this separation. Testing of
the thiourea column led to the discovery that the perfluoro-2-methylpentane
irradiated up to that time had been contaminated with a considerable
amount of unsaturated material. Following this discovery a large
thiourea column was constructed and used' to purify this isomer.
Listed in Table 27 are the operating conditions and standard
retention volumes on the thiourea column. The last unsaturate listed
was the contaminate in the 2~CF2C,-F,. .
P P 11
Mole-Area Calibration
Mole-area calibrations were run^^; using C~Fg, C^F^, C^F^g,
and In all cases the data is found to show that the area under
the curve on the chromatogram is proportional to the number of moles
of the material. For this to be true, a plot of log (no. moles)
versus log (area) should have a slope of unity and the various materials
should define a single line. This is illustrated in Figure 55 where
the data for C^Fg, C^F^ and C^F^^ is plotted. A least squares fit
of this data gives a slope of 1.004 which is within experimental
accuracy of unity slope.
Fluorine Balance
A material balance method for determination of fluorine lost
to the wall was used in this study,

147
TABLE 27. STANDARD RETENTION VOLUMES ON THE THIOUREA COLUMN
Pi = 54.7 psi. Po = l6.8 psi. Temp, of column 25C
Carrier Gas = 29*9 cc./min. He at 1 atm. and 25C
Column Length 15*2 meters containing 155*8 grams of 0.2 grams
thiourea/gm. Chromosorb-P.
Compound cc. He at 1 atm. and 25 C
. A
2,5-(cf5)2c4f8
2-CF5C5Fh
5-cf5c5f11
n-C6Fi4
trans 0 C = C -
C
/
C
\
c
544
568
585
4l0
525
c
/
cis c C a C C 577
\
c

Log(n, moles)
145
Figure 35 Log(moles) versus log(area) for chromatographed
fluorocarbons

149
Since all the fluorine present in the original material must
either be present in the products or on the wall, we may perform the
balance as follows assuming each molecule of the original yields one
molecule of product.
Ni mole p /'cent of product i in final mixture
Si ratio of fluorine to carbon in i
Ci number of carbon atoms in i
Xs ratio of fluorine to carbon ir original material
Cs number of carbon atoms in original material
n number of compounds being summed
number formed
V/ = moles Fg on wall per 100 moles mixture
which simplifies to

150
n
2W +
i=l
jS
A3
2
i=l
i^s
KiCi
Solving for W we get:
W
Xs
yT NiCi
i-1
NiCiXi
2
It should be noted here that the total composition of these
mixtures is not exactly known. In particular there is a possibility
of unsaturated compounds of four or more carbon atoms since these
could not be detected with the analytical system used. There were
probably not significant amounts of these present for original
materials of two carbons and greater since very small amounts of
unsaturated two and three carbon material.s were found. If unsaturated
materials were present and undetected this would make the fluorine
loss values higher.

APPENDIX 4. DECANNING OF PILE IRRADIATED SAMPLES
Both the LITR and Graphite reactor samples were too radioactive
to handle in Reed Laboratory (the site of this investigation) and
had to be decanned in the nuclear engineering building. Figure 56
is a sketch of the apparatus used in this decanning. The sample to
be decanned is inserted in a greased section of l/8 inch tubing and
this is in turn greased and slipped into a piece of tight fitting
larger tubing as is shown. After the system is evacuated and closed
off, the sample is introduced by breaking off the end of the tube
inserted into the larger tubing. The spun glass filter serves to
slow down the escaping gases and thus minimize blowout of the solids
and it also catches any solids which may be blov/n out. The sample
is then condensed in the glass sample receiver tube and sealed off
at the constriction unless air is seen to be present (by residual
pressure at the manometer). If air is present, the sample is trans
ferred to the sample holder tube and deaeriated by alternately thawing,
freezing, and pumping. The weight of the gas sample is determined
by weighing the glass sample tube before and after the sample is
removed with a correction for the air contained in the opened tube.
The weight of the original sample is determined by weighing the
aluminum tube before and after filling with the sample.
151

152
Figure 36. System used for decanning pile irradiated samples

APPENDIX 5. RADICAL TRANSPORT
Diffusion Coefficient of Radicals in .Perfluoronropane
The escape of radicals to the wall may be dependent on either
the mean free path of the radicals in the bulk or on the diffusion
coefficient of the radicals in the bulk medium. If the half-life of
the radicals is long enough for actual diffusion to take place the
diffusion coefficient will apply; if escape to the wall is essentially
escape from the cage of surrounding molecules with subsequent flight
to the wall with few molecular collisions, the mean free path will
give an estimate of the relative probability of escape to the wall.
It must be admitted that neither of these two quantities can be
calculated with any great precision.
The diffusion coefficient can be estimated by Gilliland's
(Mo)
equation using the data of Le Bas to estimate the values of Vh.
D12 "
(L-l)
where
T absolute temperature, K
M => molecular weight
P pressure, atmospheres
Vb * molar volume
Dir = diffusion coefficient, cm. /sec.
155

Ip4
Using the Le Bas numbers we find the following molar volumes.
Vb(CjFg) *=> ll4 cm.^/mole
Vb(CF^) = 40.9 cm.Vnole
Vb(C2F5) = 75-1 cm.'Vmole
Vb(C7,Fy)=> 105.5 cm. 2/mole
Vb(F) 8.7 cm.Vaiole
Substituting these values in equation (L-l) yields'the following
diffusion coefficients for the various radicals in C^Fg.
D(F)
5/2
2.17 x 10~v cm./sec.
D(CF5) 8.79 x 10'
-6 T^/2 __ 2
cm
. /sec.
A rp5/ 2 n
D(C2F5) o 6.18 x 10" -4_ cm. /sec.
, ^5/2
D(C5F7) 4.98 x 106 cm.2/sec.
Using these values we find the following ratios of diffusivities
at a given temperature and pressure
D(CF-)
- 0.405
D(F)
d(c2f5)
0.705
d(cf5)
d(c5f?)
d(02f5)
0.805

155
From these numbers we observe a large difference in the dif-
fusivities of the large radicals compared to fluorine, but not very
large differences among the large radicals. From this we conclude that,
if diffusion controls the transport of the radicals to the wall, more
fluorine radicals than any other will reach the wall and the amounts of
other radicals will not differ greatly among themselves if the average
radical concentrations are nearly equal.
If we assume the fragment effective concentrations given in
Appendix 6 for C^Fg reflect the actual average fragment concentrations
we may approximate the relative amounts of each radical that diffuses
to the wall as the product of the diffusivity and this effective con
centration. These products are listed below.
Radical
CF^
c2f5
c5F7
Product of D and effective cone,
relative to F
1.0
0.559
0.107
0.4o
Hence it appears that fluorine radicals definitely have the
advantage so far as wall capture is concerned.
To calculate the mean free path, it was necessary to develop
an equation allowing calculation from molecular diameters and densities
of the bulk. This development follows in the next section.

15 6
Statistical Model for Calculating Mean Free Path
The following symbols which will be used in the derivation
are defined.
Pj. =* probability of passing through the Nth shell
collision diameter of the particle for which the mean
free path is being calculated
d = average distance between molecules in the medium
N = number of shells of medium molecules passed through
by the particle
D = mean free path
m A
Consider the particle in the average distribution of molecules.
This distribution is a series of equilateral tetrahedrons of side
length d.
Starting from a point of one tetrahedron the free area of the
first shell of molecules is (\/$1/2)d away.
Let us examine one triangle of one of the average tetrahedrons.
The total area of the triangle is:
The area occupied by the molecules plus that excluded to pas
sage by the diameter of the particle is
since each molecule belongs to 6 triangles in a planar projection.

157
Hence the probability of the particle passing successfully
through one shell is:
^ ,,2
VT1 £
4
P 1 -
17
cr+ cr' \2
2 YT
The probability of passing through each successive shell is
equal and hence the probability of successful passage through N
shells is
N
(4-2)
from.
The mean free path is that for which P
0.5 and is calculated
- d V~T /2 (N)
(L-5)
The value of d is found from the value of atoms/cm. for the
medium and setting this equal to
4/24
ZJ
.11785d
or
atoms m 1.415
cm. d

158
To test the application of this theory to the ideal case the
mean free paths of argon, helium, and nitrogen were calculated at
one atmosphere and 100, $00, and 500 1C and compared with those found
from gas kinetics calculations. This comparison is found in Table 28.
From the tabulated values we see that the two models ore in
very good agreement. There appear to bo several distinct advantages
to the statistical model. These are listed below.
1. The statistical model points up the fact that mean free
path does not depend directly on the temperature of the medium or the
energy of the particle. The dependence is only through the density
factor.
2. The statistical model allows calculation of the probability
of other path lengths than the mean free path.
5. The statistical model allows the calculation of the mean
free path of a foreign molecule in a bulk material consisting of
another species. This last point is very important for the present
application.
Mean Free Paths of Radicals in Perfluoropropane
The values of d are calculated for several densities of C^Fo
5 8
^3 gm./crn.^
.0554
.17
55
.4
55
d, A
25.5-
15.73
10.87
9.55
9.52
7.62
1.0

159
TABLE 28. COMPARISON OF MEAN.FREE PATHS CALCULATED FROM GAS
KINETICSAND FROM THE STATISTICAL MODEL
Molecule
o
T K
Gas
Kinetics
Statistical
Percent Error
Based on Gas
Kinetics
He 2.18
100
645
654
+1.595
500
1956
I960
+2.27
500
5226
5010
-O.496
Ar
100
251
229
-O.865
500
694
702
+1.152
500
1157
1185
2.25
N2 5.75
100
218
215
-1.577
500
664
702
+5.82
500
1090
1112
-2.02
Average
error =
I.96 percent
O
_ 0
T K
d, A
100
26.7
500
5 8.6
500
45.8
(a) Gas kinetics mean free paths are taken from reference 4l.

16o
Values of are known for the following molecules:
Molecule
2P6
f
5.0^5)
5.8
Values of O''for other molecules and radicals of interest can
be estimated from these data.
Assume the volumes of the various atoms and groups are additive.
Volume of sphere = (l/6)7/D^
vp = (1 /6)(/7)(5.655)5 = 25.5 P
2
vcf4 (1/6)(^)(5)5 5.5
Vc2F6 (1/6)(tf)(5.8)5 = 102.4
The differences between V and V
and that between Vpp
2 F6 4
and
Vq are both the volume of a CF2 group. The values are respectively
40 and 57. It is assumed that the value of 57 will more nearly apply
for addition of volumes.
Using this value, the volume for a C^Fg molecule is:
v0 p 159.4 P
5 8
6.42 A
Then o'for C^Fg is calculated,

l6l
The (T values for the radicals are found similarly and are
tabulated below.
Species
si.
5F8
6.42
F
2.9
CF,
4.65
c2f5
5-55
5F7
6.22
These values are used in the statistical model to calculate
the mean free paths of these radicals for several densities. The
results are given in Table 29.
Values of zero for Dn mean the average distance between mole
cules is such that the radical cannot pass between them. This may be
interpreted as a possible cage11 mechanism. At densities above
about 0.5 grams per cubic centimeter, ChFis "trapped," C^F^. is also
trapped fairly close to this density, CF^ is trapped slightly above
55 and- F is trapped slightly above 0.55* This means that at low
density all the radicals can escape to the wall, but above a density
of about 0.55 g./cm. only F should reach the wall in significant
quantities.
At low density g./cm.the ratios of the mean free paths
are:
Dm(DF;>
Dm(F)
^(02?g)
.89
.8 %

12
TABLE 29. MEAN FREE PATHS FOR RADICALS IN PERFLUOROPROPANE
Species
/o,gm./cm.^
N
F
.0554
4.42
89.2
.17
1.29
15.4
.625
5.88
.40


.55
.295
2.57
1.0
0
0
cf5
.0554
5.o4
61.4
.17
.795
9.48
55
.259
2.44
.4o
0
0
.55
0
0
C2P5
.0554
2.54
51.5
.17
.<504
7.2
0
0
5F7
.0554
2.25
45.0
.17
.479
5.71
55
0
0

are:
At a density of 0.17 ga./cm. the ratios of the mean free paths
.614
BW .761
qn(0?F7)
Dm(C2F5)
.795
Prom these numbers we see there is little difference in the
mean free paths of the molecules CF^, C^F,-, and C-F-,. From this we
conclude that a difference in products due to radicals reaching the
wall is not to be expected except in reactions where the total amount
of radicals is important end not their relative concentrations or where
F or Fg is one of the radicals. An example of this is the reaction of
radicals with unsaturated molecules. Apparent G values for the total
reactions would be expected to be lower at low density since a portion
of the radicals produced may be removed at the wall.
In conclusion, whether diffusion or mean free path mechanisms
determine the removal of radicals at the wall the results are similar:
1. Fluorine radicals will reach the wall in significantly
greater amounts than other radicals

164
2. The other radicals will reach the wall in similar amounts
leading to the conclusion
5. The distribution of products will be similar at all densities
except for materials formed by reaction with F or and materials
formed by unsaturates reacting with radicals.

APPENDIX 6. STATISTICAL RECOMBINATION DATA CORRELATION
Since the most important reactions in irradiation of fluoro
carbons appear to be recombination of fragments, on obvious method of
correlation of the data is to assume random recombination. The
statistical nature of radiation chemistry reactions has been previously
(44)
noted elsewhere.' For any given radical there are a given number
of bonds which, when broken, will yield it. This number of bonds
should be roughly proportional to the yield of this radical from
different types of molecules. It is recognized that certain bonds
will rupture more easily and certain radicals may be more reactive
toward other radicals due to stearic factors. For this reason the number
of possible ways a radical can be formed must be multiplied by an
empirical effectiveness factor. The method will be illustrated with
C2F5 as an example. This is shown in Table 50
From this correlation the fraction of the radicals which react
to give the original products can be estimated.
* A 11.5^
2F6 54.ll
555
Hence the actual G value for molecules giving radicals is
approximately the value found by analysis divided by O.667. For tube
number 68 this is
G = a 5 43
C2F6 .667
165

166
TABLE 50. RANDOM RECOMBINATION CALCULATION AND RESULTS FOR C2F
Ruptura
2 Reactions
Recombination Reactions
F F
1 1
F + F F2
1 1
- C C F
| (
> F + C 2 F^
F + CF_ *
2 ?
2r6
F F
-A/ > CP, + CF,
3 3
OF, + CF,
CF* + F
3
* 2F6
cf4
C0F¡- + CF,
2 3 3
>0 F
58
2F5 + 2P5^ 4f10
Fr affluent
N(number ways formed) E(effectiveness)
N x E
F
6
0.50
1.80
CF?
2
1.00
2.00
C2F5
6
.34
2.04
Relative Yields of Products

f2 (1.80)2 = 5.24
C?4 (2)(2.00)(1.80) 7.20
C2F6 (2)(2.04)(1.80) + (2.00)
2 = 11.54
C-Fg (2)(2.04)(2.00) = 8.17
nG4F10 = (2*4)2 = 4.16
£ = 54.11
Yields of products relative
to CF4
Product
Yield/CF^ yield(correlation)
Yield/CF^ tube 68
F
2
.45
1.73
cf4
1.0
1.0
C/8
1.135
1.305
n4FlO
578
.467

The procedure for calculating these numbers for the other
fluorocarbons is similar. The resulting data is listed in Tables 51
through 56.
In this treatment the values of the empirically determined
effectiveness factor were found to be about the same for most radical;
formed by similar processes. These are listed below:
Type of Radical and
Method of Formation
F from C-F bond breakage
except C4F1Q, 2,3-(CFj)2C^Fq
F from C-F bond breakage
in C4F10
F from C-F bond breakage
in 2,5-(CF5)2C4F8
CF, from C-C bond
P
Other radicals from C-C bond
breakage
Fluorocarbon radical from rupture
of CF bond on CF^ group
Fluorocarbon radical from removal
of F from carbon next to
CF_ group
P
Fluorocarbon radical from removal of
F from carbon two down from
CF^ group
E (Effectiveness)
0.50
1.00
' 1.57
1.0
0.45
0.54
1.065
1.50
Althoughthe correlation with experimental data leaves much to be
desired, the fit is close enough to draw fairly good conclusions from
the empirical effectiveness numbers.
1. The fluorine atoms on the CF^ groups are much more diffi
cult to remove than those on CF^ groups and those on CF^ groups
furthest removed from CF groups are the easiest to remove. This
P

168
TABLE 51. RESULTS OF RANDOM RECOMBINATION CALCULATION FOR C-Fg
Fraspnent
N
E
N x E
F
8
oo
2.-40
cf5
2
1.00
2.00
2p5
2
.45
.90
n-C,F7
5 7
6
.54
2.04
2
I.065
2.15
Molecule _..c; ......c. Relative Amount
P2
5.76
of4
9.60
2F6
8.55
C5F8
25.61
n"4F10
8.97
i-^lO
8.55
n_G5F12
5.68
2~CF5C4F9
5.84
a-6pl4
4.16
2-cf55fh
8.70
5-(cf5)2c4fq
4.54
2 89.72
percent recombination = 26.5

169
Table pi (continued)
Ratio
Stat. Pred.
Experimental
(typical high T
high density)
VCF4
.60
1.0
c2f6/0F4
.868
1.22
Totai 04F10/OF4
1.825
1.80
n5Pl/0F4
p84
.415
2-C?5C4F8/CF4
.bo
.65
n_G6Fl4/F4
.455
.12
2-CF5C5Fu/CF4
.906
.51
2,5-(cf,)c4f8/cf4
.475
.21

170
TABLE 52. RESULTS OF RANDOM RECOMBINATION CALCULATION FOR n-C4F
Fragment
N
E
N x E
F
10
1.0
10.0
CF-
2
1.0
2.0
c2f5
2
.45
.90
n-05F7
2
.45
.90
n-C4F9
6
.54
2.04
2-4F9
4
1.065
4.26
Molecule
Relative Amount
F2
100.
cf4
40.0
C2F6
22.0
C5F8
21.6
n-4Fio
150.4
n-C5P12
9.8
2-CF5C4F9
17.0
n-6F14
' 4.5
5-GF55Pu
7.7
Total Cy
11.5
Total Cg
4o4.0
percent recombination 52.5

171
Table 52 (continued)
Ratio
Stat. Pred.
Experimental
$95(leaking tube)
. M.
V5pe
4.65
4.58
5.65
GFi/5F8
1.85
1.58
1.958
2V5F8
1.02
.705
.844
"-05Fi/5F8
.455
.649
.594
2-0F504y05F8
.757
1.054
1.28
H-O^ii/OjFg
.208
.524
.406
.568
.719
Total Cy/C^Fg
.525
1.568
2.16
Total Cg/C^Fg
1.84
1.595
2.25

172
TABLE 55- RESULTS OF RANDOM RECOMBINATION CALCULATION FOR n-C F^
Fragment
N
E
N x E
F
12
.50
5-8 '
OF*
5
2
1.0
2.0
c2f5
2
.45
.90
n-C^Fy
2
ir\
a
.90
n-4F9
2
.45
.90
n'5Pll
6
.54
2.04
2-C5Fn
4
1.065
4.26
5c5fii
2
1.50
5.00
Molecule
Relative .Amount
F2
12.98
cp4
l4.4o
2F6
10.48
C5F8
10.08
n-Vio
IO.89
n_C5F12
72.2
nc6Fi4
10.59
2-0F5c5Fii
17.05
12.0
Total Cy
IS.5
8
17.5
9
16.7
c10
86.5
£. 509.87
percent recombination 25*5

175
Table 55 (continued)
Experimental
Ratio
Stat. Pred.
#11(?0"C)
#78(99~C)
VOP4
.902
5.71
5.00
2V0F4
.728
.615
1.00
5fS/cf4
.700
.615
.800
nc4Fio/',CF4
.756
.615
.715
n-6Fl4/CP4
.755
.587
.685
2-CF5C5Fn/CF4
1.185
.452
.715
5CP5C5F11/CF4
.854
.615
.828
Total Cy/CF^
1.27
1.68
2.03
Total Cg/CF4
1.215
1.097
1.286
Total Cp/CF^
1.16
1.42
1.545
Total C10/CF4
6.01
2.00
1.60
Vc5p8
1.288
6.05
5.75
V5F8
1.45
1.65
1.25
2F6/05F8
1.04
1.00
1.25
n"4Flo/o5F8
1.08
1.00
.895
n-6F14/05P8
1.05
.652
.858
2-F55Pll/05F8
1.69
.757
.895
5-0F55f11/0;f8
1.19
1.00
1.057
Total Cy/C^Fg
1.815
2.74
2.61
Total Cg/C^Fg
1.757
1.79
I.61
Total C^/C^Fg
1.655
2.51
1.68
Total C10/C^q
8.57
5.26
2.00

174
TABLE 54. RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR n-C.F .
6 14
Fragment
N
s
N x E
F
14
.50
4.2
cf5
2
1.0
2.0
2P5
2
.45
.90
5P7
2
.45
.90
n~c4F9
2
.45
.90
n-05Fu
2
.45
.90
n-0P15
6
.54
2.04
2-6p15
4
1.065
4.27
^6F15
4
1.50
6.00
Molecule
Relative
Amount
Molecule
Relative
Amount
P2
17.65
Total C
52.5
cf4
16.8
Total C
8
24.6
C2F6
11.56
Total C^
25.8
5P8
11.16
Total 0
10
25.O
n-4Fio
11.97
Total C
22.2
n-5Fl2
12.78
Total C, _
12
151.5
n-6Fi4
108.7
483.22
percent recombination = 22.5

175
Table >4 (continued)
Statistical
Experimental
Ratio
Pred.
#52(25uc)"
' ^81(99 6)
#79(0RGR)
f/cf4
1.05
2.28
7.5 8
c2f6/cf4
.688
.719
.91
.725
c5pf/cp4
.664
.875
1.455
.645
n-C4P10/CF4
.712
.750
1.455
.665
n-05P12/0F4
.760
.625
1.045
.575
Total Cy/0F4
5.12
1.625
5.75
.750
Total Cq/GF^
1.465
1.28
2.59
.528
Total C^/CF^
1.417
1.22
2.82
.470
Total C10/OF4
I.570-
1.06
1.68
.280
Total 0 /0F4
1.520
1.155
1.82
.220
Total C^^/CF^
9.02
1.095
5.05
.068

176
TABLE 55* RESULTS OF RANDOM RECOMBINATION CALCULATIONS FOR 2-CF-C^F.^
C
C
C
c
c
Fragment
N
E
N :< E
F
14
.30
4.20
cf5
5
1.0
5.00
c2f5
1
.45
.45
n-CjF7
l
.45
.45
2-05F7
1
.45
.45
O
1
O
1
O

l
.45
.45
c
2~5fii
2
.45 '
.90
0
t
0
1
0
i
0
0
I
1
.45
.45
c
1
0
s
0
i
0
j
c
3
.34
1.02
c
- c c c -
1
c
6
.34
2.04
c
0
1
0
0
1
0.
!
c
2
1.065
2.150
c
1
0
1
Q
0
!
c
2
1.50
5.00
c
-c-c-c-c
1.065
c
1
1.065

177
Table 55 (continued)
Relative
Molecule
Amount
F2
17.65
0F4
25.2
2F6
12.78
5F8
10.26
n-C, F
b 10
2.9
WlO
6.48
n-5P12
7.97
2-0F5C4F9
5.55
Molecule
Relative
Amount
2-0F55Fll
87.0
Total Q-j
57.6
Total CQ
8.9
Total
17.9
Total 010
10.1
Total C
25.
Total C
85.5
^ 580.66
percent recombination of original fragments to
yield original material = 22.8

178
Table 55 (continued)
Ratio
Statistical
Pred.
Experimental
V0F4
.70
5.24
2 V0F4
.507
oil
5VOF4
.407
.289
n-c4Fic/CF4
115')
1
7 572
f 0 289 (mostly
i-C,F /OF.
4 10 4
.257 J
J
n_C5Fl/CF4
.518
.40
i-cf5c4f9/cf4
.212
.178
Total C7/CF4
2.28
1.445
Total Cd/0F.
o 4
.555
.755
Total C /OF,
9 4
.710
1.155
Total C10/CF4
.401
1.51
Total Cn/CF4
.995
1.58
Total C /OF,
12 4
5.59
.778

179
TABLE 56. RESULTS OF RANDOM RECOMBINATION CALCULATIONS
FOR 2,5-(0F5)2C4F8
Fr ament
1'T
E
N x E
F
14
1.57
19.2
cf5
4
1.0
4.0
2-5F7
2
.45
.90
O O
1
.O
1
O
C
4
.45
1.80
c c c -
1 1
c c
c
12
.54
4.08
c 0 c -
1 1
c c
c
2
I.O65
2.15
6 Molecule
Relative
Amount
Molecule
Relati'
Amovin
F2
568.
Total.
7
49.7
cf4
155.5.
Total
8
5.2
2p6
16.0
Total
c
9
11.2
5V
54.6
Total
10
5.2
i-C,F_
4 10
7.2
Total
11
22.4
2-CF^C4Fp
69.1
Total
12
58.6
2,;H0F5)2C4F8
245.2
^ 1018.7
percent recombination = 25.9

180
Table 56 (continued)
Statistical
Ratio
Pred.
Experimental
f/cf4
2.4
5.74
Wop4
.1044
.1095
05V0P4
.226
.26
i-O.F /OF,
4 10 4
.047
.0412
R-CFjC^Fp/OFi^
.451
.219
Total Cy'CF^
.524
1.6l6
Total Co/CF^
.0209
1.287
Total C^/CF,
.075
.644
Total C1q/CF4
.0209
.822
Total Cu/CFa
.146
.480
Total C /OF.
12 4
.252
1.60p

131
effect is not unexpected since the addition of fluorine atoms to
a carbon atom tends to strengthen other C-F bonds already present.^J
2. OF-* is more effective than other fluorocarbon radicals
P
formed by carbon-carbon bond splitting. This must result from ex
tensive rupture of molecules to give more than one single-carbon
species.
5. Since the other radicals formed by carbon-carbon bond
breakage did not differ greatly in their effectiveness values the
stearic factors for recombination must be small.
Using this statistical recombination method and the empirically
determined effectiveness numbers in conjunction with the sensitivities
in the mass spectrometer an estimate can be made of the products and
their G values for other saturated fluorocarbons. In this calculation
the ratios of the products are determined by the statistical recombina
tion method and the overall G from the mass spectrometer sensitivity.

CHEMICAL NOMENCLATURE
The follov/ing are equivalent chemical nomenclature and are
used interchangeably:
perfluoromethane, CF^
perfluoroethano, C^Fg
perfluoropropane, C_F0
perfluoro-n-butane, n-C^F^0
perfluoroisobutane, i-C^F^
perfluoro-n-pentane, n-C_F^
perfluoroisopentane, i-C-F^
perfluorocyclopentane, cy-C^.F^o
perfluoro-n-hexane, n-C.F .
6 14
perfluoro-2-methylpentane, 2-CF^C_F^
perfluoro-3-methylpentane, J-CF^C^F-^
perf luoro-2,5-dimethyl butane, 2,3- (CF,) ^C^Fg
perfluoroethylene, C^F^
perfluoroacetylene, C2F2
perfluoro-n-propene, n-C^F^
perfluorocyclopropane, cy-C^F^
182

ABBREVIATIONS AND DEFINITIONS
LITR
G
Roentgen
rad
rep
Oak Ridge Low Intensity Testing Reactor
number of molecules produced per 100 e.v. absorbed
that quantity of x or (f radiation such that the
associated corpuscular emission per 0.001295 £?n. f
air produces, in air, ions carrying 1 esu of quantity
of electricity of either sign
100 ergs absorbed/gram
roentgen equivalent physical, 95 ergs absorbed/gram.
185

BIBLIOGRAPHY
(1) J. H. Simons and E. H. Taylor, J. Phys. Chem. 6j5, 656 (1959)
(2) R. E. Florin, L. A. Wall, and D. W. Brown, Jnl. Res. Nat. Bur.
Stds. 64A, 269 (I960).
(5) R. A. Seeger and J. G. Calvert, J. Am. Chem. Soc. J6, 5197 (19>4).
(4) P. B. Ayscough and E. W. R. Steacie, Proc. Roy. Soc. A2f,
476 (1956).
(5) 7. H. Dibeler and F. L. Mohler, J. Res. Nat. Bur. Stds. 40,
25 (1948).
(6) API Project 44 Mass Spectral Data.
(7) S. V. R. Mastrangelo, J. Am. Chem. Soc. 84, 1122 (1962).
(8) G. 0. Pritchard, Y. P. Hsia, and G. H. Miller, J. Am. Chem.
Soc. 85, 1565 (I9t>5).
(9) H. A. Dewhurst, J. Phys. Chem. 6^, 815 (1959).
(10) T. D. Nevitt and L. P. Remsberg, J. Phys. Chem. 64, 969 (i960).
(11) L. Pauling, The Nature of the Chemical Bond, Cornell University
Press, Ithaca, N. Y. (i960) p. 85.
(12) J. H. Simons, Fluorine Chemistry, Volume 5. Academic Press,
New York, N. Y. (to be published) p. 81.
(15) G. J. Mains and A. S. Newton, J. Phys. Chem. 6, 212 (l9l).
(14) G. R. Freeman, J. Phys. Chem. 64, 1576 (i960).
(15) D. Urch and R. Wolfgang, J. Am. Chem. Soc. 8, 2982 (1961).
(16) P. B. Ayscough, J. G. Polangi, and E. W. R. Steacie, Can. J.
Chem. 22, 745 (1955).
184

185
(17) J. R. McNesby, J. Phys. Chora. 4, 1671 (i960).
(18) C. D. Bass and G. C. Pimental, J. -Am. Chem. Soc. 82, 5754 (191).
(19) J. Said and M. Szware, J. Am. Chem. Soc. J2, 1554 (1957).
(20) J. K. Futrell, J. Phys. Chem. 65, 5^5 (1961).
(21) T. N. Bell, B. J. Pullman, and B. 0. V/est. Australian Jnl.
Chem. 16, 722 (1965).
(22) J, K. Simons, Personal communication.
(2p) R. D. Dresdner, T. M. Reed III, T. E. Taylor, and J. A. Young,
J. Org. Chem. 2£, 1464 (i960).
(24) J. A. Young, Personal communication.
(2p) T. M. Reed III, J. F. V/alter, R. R. Cecil, and R. D. Dresdner,
Ind. Eng. Chem. 1, 271 (1959).
(26) D. M. Richardson, A. 0. Allen, and J. W. Boyle, USAEC unclassified
report, ORNL-129 (Dec., 1948).
(27) T. M. Reed III and T. E. Taylor, J. Phys. Chem. 62, 58 (1959).
(28) T. R. Johnson and J. J. Martin, Nucleonics 20, 85 (1962).
(29) L. H. Gray, Br. Jnl. of Radiology 22, 555 (1958).
(50) J. W. Boag, pp. 100-115 of Quantities, Units, and Measuring
Methods of Ionizing Radiation, ed. by Franco Fossati, Ulrico
Hoepli, Publisher, Milan (1958).
(51) L. G. Alexander, USAEC unclassified report, OKNL 58-12-9.
(52) G. Friedlander and J. H. Kennedy, Nuclear and Radiochemistry,
John Wiley & Sons, Inc., New York (1958) P* 202.
(55) Report of the International Commission on Radiological Units
and Measurements (ICRU) 1958; NBS Handbook 62, 10-17 (1958).
(54) L. V. S. Spencer, NBS Monograph 1 (Sept. 10, 1959).

186
(55) G. J. Hie and G. L. Brovmell, Radiation Dosimetry, Academic
Press, Hew York (1956) pp. 25-100.
(56) E. H. Cart, Thesis, Ohio State University.
(57) J. A. Brown, Jnl. of Chem. and Eng. Data .8, 106 (1965).
(58) J. F. Brown, Jr., Scientific American 82, 207 (July, 1962).
(59) Shashi Dat, Personal communication.
(40) R. C. Reid and T. IC. Sherwood, The Properties of Gses and
Liquids, McGraw-Hill Book Co., Inc., Hew York (1956) p. 267.
(41) C. D. Hodgman, ed. in chief, Handbook of Chemistry and Physics,
57th Edition, Chemical Rubber Publishing Co., Cleveland, Ohio
(1955) P- 5078.
(42) R. B. Bird, W. E. Stewart and E. H. Lightfoot, Hotes on Transport
Phenomena, John Wiley & Sons, Inc., Hew York (1958) p. 11.
(4p) M. D. McKinley, Personal communication.
(44) Burton, M., J. Chem. Ed. 28, 412 (1951).

BIOGRAPHICAL SKETCH
James Clifford Mailen was born August 19, 1957, at Colorado
Springs, Colorado. In June, 1955, lie was graduated from Hi chita
Jest High School in Wichita, Kansas. In June, 1959, he received the
degree of Bachelor of Science from Kansas State University. In 1959
he enrolled in the Graduate School of the University of Florida. Until
June, I960, he was supported by a Graduate School fellowship and from
then until June, 1965, by NSF Cooperative fellowships.
James Clifford Mailen is married to the former Jean Hester
Bolick and has one stepchild. He is a member of the American Institute
of Chemical Engineers, American Chemical Society, Phi ¡Cappa Phi, and
Sigma Tau.
187

This dissertation was prepared under the direction of the
chairman of the candidate's supervisory committee and has been
approvod by all members of that committee. It was submitted to the
Dean of the College of Engineering and to the Graduate Council, and
\ras approved as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
August 8, 1904
Dean, Graduate School
Supervisory Committee:



12
TABLE 29. MEAN FREE PATHS FOR RADICALS IN PERFLUOROPROPANE
Species
/o,gm./cm.^
N
F
.0554
4.42
89.2
.17
1.29
15.4
.625
5.88
.40


.55
.295
2.57
1.0
0
0
cf5
.0554
5.o4
61.4
.17
.795
9.48
55
.259
2.44
.4o
0
0
.55
0
0
C2P5
.0554
2.54
51.5
.17
.<504
7.2
0
0
5F7
.0554
2.25
45.0
.17
.479
5.71
55
0
0


57
TABLE 6 (continued)
1
Tube No.
28
12
Compound
5f8
n~C5Pl2
Compound in assay or standard
retention volume VR, cc. H2
Mole
Fraction^
cf4
C2F6
2f4
C2F2
c5f8
cy-CzF^
C4F10(mostly iso)
Vg 100 cc. ^(n-hex)
n-CcF-i p
i-Ccpi2
VR 224 cc. K2(n-hex)
n-CFl4
VR cc. ^(n-hex)
2-CF,C5F11
5-CF£tFn
2,5-(CF5j2C4F8
Total^C7Fl6
Total C8F18
Total
Total
Total CTtFpA
Total C12F2
Total C17F2q
Total 0^4
Total C. c
Total oti
Total C17
Total C-¡_g
Total C-^q
Sample wt., Gms. (initial)
9F20
10F22
0.7295
O.1558
nil
0.0052
0.0491
0.0006
0.0194
nil
0.0018
0.0169
trace
0.0004
nil
0.0017
0.0010
0.0080
0.0044 (2)
0.0073 (2J
0.0042 (5)
0.0080 (5)
0.0055 (4)
0.0059 (5)
0.0015 (4)
nil
n
n
11
11
11
0.5110
0.8558
0.0979
nil
0.0050
0.0514
0.0006
0.0150
nil
0.0018
0.0082
0.0005
trace
nil
2 + 5 =
0.0005
0.0059
0.0004 (2)
0.0006 (2)
0.0001 (1)
nil
11
n
it
ti
u
u
11
it
11
0.2110
Molecular wt., Gas Density (of gas)
120.5
114.5
Weight percent recovery as gas
67.6
90.2
Moles F2 lost/mole mixture
Molecular wt. of gas from analysis
125.9
101.6
(a) Numoer in brackets is number of peaks in group


152
One the energy absorption is known the G values (molecules/100
e.v. absorbed) are calculated. These are listed in Table 18.
Energy Absorbed in Reactor Irradiated Samples
Dosimetry in a nuclear reactor is much more complex than that
of a gamma source due to the mixture of radiation types having a
broad spectrum of energies.
The energy absorption in the Oak Ridge Graphite Reactor can
(
be estimated from the data of Richardson, Allen, and Boyle' for
energy absorption in graphite. Since 82 percent of the energy ab
sorption is from gamma rays the energy absorbed in a fluorocarbon per
gram will be essentially the same as is absorbed in graphite per gram.
With an appropriate flux correction from their position to ours the
rate of energy absorption is found to be approximately 2.7^ x 10^
erg/g.sec. Over the entire irradiation the energy absorbed will be
8.5 x 10^ erg/g. From this number the G values in Table 6 are cal
culated.
For the LITR reactor no estimate of the energy absorption is
available. However, such information is not necessary in this case
since the samples were degraded to equilibrium mixtures.