Citation
Effects of radiation on mixtures containing fluorocarbons and the identification of the perfluoroheptane isomers

Material Information

Title:
Effects of radiation on mixtures containing fluorocarbons and the identification of the perfluoroheptane isomers
Added title page title:
The identificaion of the perfluoroheptane isomers
Creator:
Askew, William Crews, 1940-
Place of Publication:
[Gainesville]
Publisher:
[s.n.]
Publication Date:
Language:
English
Physical Description:
x, 111 l. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Aluminum ( jstor )
Boiling ( jstor )
Carbon ( jstor )
Cobalt ( jstor )
Dosage ( jstor )
Fluorocarbons ( jstor )
Irradiation ( jstor )
Isomers ( jstor )
Oxygen ( jstor )
Radiolysis ( jstor )
Chemical Engineering thesis Ph. D
Dissertations, Academic -- Chemical Engineering -- UF
Fluorocarbons ( lcsh )
Irradiation ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis - University of Florida.
Bibliography:
Bibliography: l. 109-110.
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:
030112165 ( ALEPH )
11046000 ( OCLC )

Downloads

This item has the following downloads:


Full Text
EFFECTS OF RADIATION ON MIXTURES CONTAINING FLUOROCARBONS AND THE
IDENTIFICATION OF THE
PERFLUOROHEPTANE ISOMERS
By
WILLIAM CREWS ASKEW
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 April, 1966




ACKNOWLEDGMENT
This work was performed under contract to the United States Atomic Energy Commission. The financial support of this agency is greatly appreciated.
The author is especially indebted to Dr. T.M. Reed III, chairman of the supervisory committee, whose patience, understanding, and direction made this work possible. The author is also grateful to Dr. M. B. Fallgatter and Dr. R. J. Hanrahan for the mass spectra identifications and to Dr. J. A. Wethington, Jr. for help in the spectrophotometric analysis. Appreciation is also extended to the other members of the supervisory committee, Dr. Mack Tyner and Dr. T. O. Moore, for their kind support. The encouragement and patience of the author's wife was invaluable.
ii




TABLE OF CONTENTS
Pages
I. INTRODUCTION .. 1
Mixtures Containing Fluorocarbons. . . . 1 Perfluoroheptane Isomers .. 2
II. EXPERIMENTAL PROCEDURES . .. . . . . .. 7
Starting Materials ... .7
Sample Preparation 8
Irradiations ........ 11
Analysis ... .12
III. LOW INTENSITY TRAINING REACTOR IRRADIATIONS OF
MIXTURES CONTAINING FLUOROCARBONS . . . . .. 20
CF4+C2F6 Mixtures. .. 20
Mixtures Containing CF4 and Nonfluorocarbon
Components ... .29
IV. COBALT 60 GAMMA IRRADIATIONS OF MIXTURES CONTAINING
FLUOROCARBONS AND OF PURE FLUOROCARBONS . . .. 48
CF4+C2F6 Mixtures . . ... . .. 48
Mixtures Containing CF4 and Nonfluorocarbon
Components. 52
Pure Fluorocarbon Samples . . . . . . 53
V. PERFLUOROHEPTANE ISOMERS. . . .. . . . .. 75
Identification of the Perfluoroheptane Isomers. .75 Estimation of Normal Boiling Points . . .. 78 VI. SUMMARY . . . . . . . . . . . 86
APPENDICES
I. Dosimetry. .. 89
II. Response of Thermal Conductivity Detector Cells .101
iii




LIST OF TABLES
Page
Table 1 Parent Perfluorocarbon Molecules and the
Perfluoroheptane Isomers They Should Yield
as Primary Products Upon Irradiation . . 5
Table 2 Impurities in Starting Materials . . . 9
Table 3 Description of Silica Gel Column No. 79 . .14
Table 4 Description of n-hexadecane Column . ... .15
Table 5 Description of Kel-F No. 90 Grease Column
No. 5 . . .. . ... . ........... 16
Table 6 Analysis of the CF4+C2F6 Mixtures Irradiated
for One Week in the Low Intensity Training
Reactor .. .34
Table 7 Analysis of the Mixtures Containing CF4 and
Nonfluorocarbon Components Irradiated for One
Week in the Low Intensity Training Reactor. .40
Table 8 Analysis of the Mixtures Containing CF4 and
Nonfluorocarbon Components Irradiated for
Four Weeks in the Low Intensity Training
Reactor .. .44
Table 9 Analysis of the CF4+C2F6 Mixtures Irradiated
in the Cobalt 60 Gamma Ray Source . . . 56
Table 10 Analysis of the Mixtures Containing CF4 and
Nonfluorocarbon Components Irradiated in the
Cobalt 60 Gamma Ray Source. . . . . 62
Table 11 Analysis of the Pure Fluorocarbons Irradiated
in the Cobalt 60 Gamma Ray Source . . . 64
Table 12 G Values for the Disappearance of the
Starting Material. ..74
Table 13 G Values and Retention Volumes on the
n-hexadecane Column of All Radiolysis
Products from n-C7F16 to n-CbFls for the
Parent Molecules Listed in Table 1. ... . .76
iv




Table 14 Retention Volumesof the Perfluoroheptane
Isomers on the n-hexadecane Column. . . 79
Table 15 Retention Volumesand Normal Boiling
Points for Fluorocarbon Compounds . . . 81
Table 16 Calculated and Experimental Normal Boiling
Points for Fluorocarbons. . . . . . 83
Table 17 Estimated Normal Boiling Points for Fluorocarbons .. 84
Table 18 Absorbed Dose Rates Measured in Dosimeter
Solution .. 91
Table 19 Ratio of Atomic Number to Atomic Weight for
Various Substances .. 96
Table 20 Area Response of Detector Cell and Correction
Factors for Calculating Weight Fractions
-H2 Carrier Gas . .. .. 103
Table 21 Area Response of Detector Cell and Correction Factors for Calculating Weight Fractions
-He Carrier Gas . 105
V




LIST OF FIGURES
Page
Figure 1 Logarithm of Retention Volume Versus Number
of Carbon Atoms in Normal Saturated Fluorocarbons .. 18
Figure 2 Weight Percent of CF4 and C2F6 in the Products
from the Low Intensity Training Reactor
Irradiations of the CF4+C2Fe Mixtures. ..... .22
Figure 3 Radiolysis Products with Straight-Chain
Structures from the Low Intensity Training
Reactor Irradiations of the CF4+C2F6
Mixtures. 23
Figure 4 Radiolysis Products with Branched Structures
from the Low Intensity Training Reactor
Irradiations of the CF4+C2Fe Mixtures . . 24
Figure 5 G Value of Primary Products Versus Weight
Percent of C2F6 in Original Sample Mixture. 49
Figure 6 G Value of Products Containing Oxygen
Versus Weight Percent of C2F6 in Original
Sample Mixture. 50
Figure 7 Logarithm of Retention Volume on n-hexadecane Column Versus Normal Boiling Point
for Fluorocarbons . .... 82
Figure 8 Absorbed Dose Rates Versus Distance from
Bottom of Sample Space .. 92
Figure 9 Aluminum Dosimeter Tubes . . . . .... 93
Figure 10 Area Response of Thermistor Detector Cell
-H2 Carrier Gas . 106
Figure 11 Area Response of Thermistor Detector Cell
-He Carrier Gas . 107
vi




DEFINITIONS AND SYMYBOLS
Megarad 106 rad Rad 100 erg./gm. Rep 93 erg./gm. Erg 6.242 x 1011 e.V. e.V. Electron-volt Roentgen, absorbed in water 0.975 rad G value Number of molecules produced per 100 e.V. absorbed.
--vr> Used to distinguish reactions brought about by the absorption of ionizing radiation. K Partition coefficient or distribution ratio, ratio of concentration of solute in gas phase to concentration of solute in liquid phase.
vii




ABSTRACT
Mixtures containing fluorocarbons and some selected
pure fluorocarbons were sealed in high purity aluminum tubing and were irradiated by Cobalt 60 gamma rays for absorbed doses of approximately 150 megarads. The mixtures containing fluorocarbons were also irradiated for periods of one week and four weeks in the Low Intensity Training Reactor at Oak Ridge, Tennessee, at approximately 2 x 1013 neutrons per centimeter squared per second.
The mixtures containing fluorocarbons were: CF4+C2F6, CF4 + aluminum powder, CF4+CO,and CF4 + carbon granules. The pure fluorocarbons were: CF4, C2F6, CsFa, n-C4Fo0, i-C4Fjo, n-CsFl2, n-C6F14, 2-CF3Cs5Fi, 3-CFsC5F11 and 2,3-(CF3)2C4Fe. Analysis of the irradiated samples was by gas chromatography and identification of some radiolysis products was by mass spectrometry.
All of the identified products found in this work can be explained by radical formation and recombination. The yields of fluorocarbon products show that C-F as well as C-C bond rupture occurs during the radiolysis of saturated fluorocarbons. Perfluoromethane is the most abundant of all fluorocarbon products and characterizes the radiolysis of saturated fluorocarbons. The abundance of branched fluorocarbon products indicates that F atoms are more difficult to remove from terminal CF3 groups than from viii




other carbon atoms in saturated fluorocarbon molecules.
Perfluoromethane is the most stable to irradiation of all fluorocarbons studied. Carbon granules and CO probably act as radical scavengers during radiolysis of mixtures containing CF4 and this causes extreme degradation of the otherwise stable CF4 to heavier perfluoroalkanes. The presence of aluminum powder during irradiation causes little change in the radiolysis products of CF4. No unsaturated or cyclic products were found in the radiolysis products of the saturated fluorocarbons. Three products that contain oxygen, C2F60eO, C3F80, and (C2Fs)20, were identified and other unknown products are probably oxygen-containing fluorocarbons. The oxygen-containing products are probably formed by radical recombination at the sample tube wall. No corrosion was found in the sample tubes and material balances show that essentially no fluorine is lost during radiolysis of saturated fluorocarbons. Molecular fluorine was not found in any of the radiolysis products. Abstraction reactions, disproportionation reactions and molecular expulsion of F2 probably do not occur in the radiolysis of saturated fluorocarbons.
G values (molecules/lO0 e.V.) calculated for the disappearance of the starting material are approximately 2 to 3 and show that saturated fluorocarbons are more stable to irradiation than saturated hydrocarbons. -G for CF4 is approximately 1.
Retention volumes of eight of the nine perfluoroheptane isomers were determined and the normal boiling points of ix




these compounds were estimated. The perfluoroheptane isomer, 2,2-(CF3)2C5Fi0, was not found in any of the radiolysis products and no other fluorocarbon products with neo structures were found.
x




I. INTRODUCTION
Mixtures Containing Fluorocarbons
Except for the recent work of Mailen(.) and Kevan and Hamlet(a) there have been few comprehensive results published on the radiolysis of saturated fluorocarbons. Other papers'') have been published on the radiolysis of fluorocarbons but the authors were hampered by the use of impure starting materials and relatively poor analytical methods.
Recent published results on the radiolysis of cyclic fluorocarbons are those of Fallgatter and Hanrahan(.) and MacKenzie et al.()
The results of Rexroad and Gordy(f) and of Golden(a) on the radiolysis of polyfluoroethylene (Teflon) are also of interest.
In this work mixtures containing fluorocarbons were
prepared and irradiated in the College of Engineering Cobalt 60. gamma ray source at the University of Florida and in the Low Intensity Training Reactor at Oak Ridge, Tennessee. The mixtures were irradiated for a dose of approximately 100 megarads in the Cobalt 60 gamma ray source and for periods of one week and four weeks in the Low Intensity Training Reactor at approximately 2 x 1013 neutrons per square centimeter per second.
1




2
The fluorocarbon mixtures by approximate mole percent were:
10OCF4 90%C2FI 25#CF4 75%C2F6e 50oCF4 50%C2F6 75%CF4 25%C2F6 90%CF4 10%C2Fe
The mixtures containing fluorocarbons were:
CF4 + aluminum powder CF4 + carbon granules
CF4 + CO
These mixtures will allow surface effects, composition effects, and heterogeneous phase effects in radiolysis of fluorocarbons to be studied.
Perfluoroheptane Isomers
Although the radiolysis mechanisms for fluorocarbon
compounds are not completely understood, there is evidence(-'2'L) that free radical reactions do occur during radiolysis of fluorocarbons.
Therefore, by treating the initial products obtained by irradiating a saturated fluorocarbon compound as 1) fluorocarbon molecules smaller than the parent molecule resulting from recombination of small fluorocarbon radicals, and as 2) fluorocarbon molecules larger than the parent molecule formed by recombination of large fluorocarbon radicals produced during irradiation, it should be possible




3
to synthesize larger molecular weight fluorocarbons from smaller molecular weight parent compounds. Radiolysis products with boiling points greater than the parent perfluorocarbon compound have in fact been observed by Simons and Taylor( ) in the irradiation of C7FI.6e, and Heine( s reports that the major products from irradiating CaFis were higher boiling than the starting material.
Therefore using perfluoronormalpentane as an example
it should undergo the following reactions to give perfluoroheptane isomers.
Rupture to Produce Radicals
n-CsF12 --AAA--- CF3 + n-C4F9
--AAA--- C2F5 + n-C3F7
--WV- F + n-C5Fl1
V--A- F + 2-Cs5F11
--A -+4 F + 3-CsF11
Recombination to Yield Perfluoroheptane Isomers
n-C4F9 + n-CsF7 ~ n-C7F16e n-CsF11 + C2F5 -4- n-C7Fj6
2-C5F11 + C2F5 -a- 3-CF3C6F1s 3-Cs5F11 + C2Fs --- 3-C2FsCs5Fii
It can be seen from the recombination of radicals that three perfluoroheptane isomers should be formed as initial irradiation products from irradiating perfluoronormalpentane. The amount of each isomer produced will depend upon stearic factors, radical diffusion, "cage effects," etc.
It should also be noted that these three perfluoroheptane isomers are the only possible perfluoroheptane




isomers that can be formed as primary recombination products of irradiating perfluoronormalpentane.
Thus, it should be possible to synthesize, by irradiation, all nine of the perfluoroheptane isomers from certain lower molecular weight fluorocarbons.
Table 1 lists the necessary parent molecules and the perfluoroheptane isomers they should produce, which are required to synthesize all nine of the perfluoroheptane isomers.
Reed(10) has resolved the perfluorohexane isomers by gas chrQmatography and it should be possible to resolve all of the perfluoroheptane isomers, so that a chromatogram of the primary irradiation products of perfluoronormalbutane should have two peaks in the perfluoroheptane group. One peak should be the n-C7F16 and the other peak should be the 3-CFsCeF13. Similarly for perfluoronormalpentane there should be three perfluoroheptane peaks: n-C7F16e, 3-CFsC6F3, and 3-C2FsCs5F11.
By comparing these two chromatograms the 3-C2F5CsF11 peak can be identified. In the same manner structures may be assigned to all perfluoroheptane isomers appearing on chromatograms of radiolysis products from the parent molecules given in Table 1. All of these parent molecules will give perfluorohexanes, perfluorooctanes, etc., as primary products, as well as the desired perfluoroheptane products.
Unsaturated products are not expected in the radiolysis products(.) and if they are produced they will probably be




TABLE 1
Parent Perfluorocarbon Molecules and the Perfluoroheptane Isomers
They Should Yield as Primary Products Upon Irradiation.
Perfluoroheptane Isomer
Parent 1 7 7 7 7
Molecule 0 0 a 0 D
I I I q
n-C4Flo
n-CsFla
n-CeFx4
2-CCs1 0 0
H ~ 0i) N- C9 Cu a)C pc) Fri C') CO V) OD C) CO U)C)C IC)C Parent P: ri C Molecule MC
___ __ __ __ __ __ __ C\J n_ _ K___\ CN NM CN n_ Cu _
n- C4F10
i- C4F10 -7
n-C5F12 x x
n-C6F14
2-CF3 C5F11
3-CF3C5F11




6
in very minute quantities. None have been found in any of the work at the University of Florida.
All of the parent molecules listed in Table 1 have been irradiated in the College of Engineering Cobalt 60 gamma ray source for approximately six weeks and for the total dose absorbed it is hoped that the conversion of parent compounds to products will be small enough so the major radiolysis products will be primary products.




II. EXPERIMENTAL PROCEDURES
Starting Materials
Perfluoromethane (CF4) was obtained from the Matheson Company Inc. The gas was chromatographically pure and was used without further purification. The perfluoroethane (C2F6) was obtained as an impure sample from the Minnesota Mining and Manufacturing Company. It was purified on a n-hexadecane column(1-) by gas chromatography.(-) Perfluoronormalbutane (n-C4Flo) was separated by gas chromatography( 01,) from a mixture of low molecular weight material provided by the Minnesota Mining and Manufacturing Company.
The perfluoroisobutane (i-C4Fio) was prepared from
perfluoroisobutene by fluorination with Cobalt trifluoride.(12) The perfluoroisobutene was obtained as a product of the pyrolysis of perfluoropropene from Peninsular Chemical Research Inc. The perfluoroisobutane was purified by gas chromatography.(2)
J. H. Simons prepared the perfluoronormalpentane
(n-CsFl2) by the electrolysis of pyridine in anhydrous HF. It was purified by gas chromatography.(25)
Perfluoronormalhexane (n-C6F14) prepared by Dresdner et al 13) was purified by gas chromatography.( -'__)
7




8
Perfluoro-2-methylpentane (2-CF3CsFI) and perfluoro3-methylpentane (3-CF3CF) were prepared by J. A. Young.(14) These materials were purified by gas chromatography.()
Table 2 lists the impurities detected in the starting materials.
Sample Preparation
Air was removed from the starting compounds by
alternately thawing, freezing and pumping on the compounds in a vacuum system until no residual pressure remained over the condensed compounds. Water was removed by distillation in vacuum through a magnesium perchlorate leg and carbon dioxide was removed by distillation through an ascarite leg. The compounds were then stored, as gas or liquid, in storage bulbs on the vacuum system until needed.
Sample sizes of approximately 1/4 cc. were measured by observing what pressure reading on the manometer of the vacuum system corresponded to a condensed sample size of 1/4 cc.
Sample tubes approximately 30 cm. long were prepared from 1/8 inch o.d. aluminum versatube of 0.025 inch wall thickness from the Wolverine Tube Division of Calumet and Hecla, Inc. After one end of each sample tube was sealed by Heliarc welding under argon, the tubes were tested for leaks at 1000 psig, washed out with benzene, dried under vacuum, and weighed.




9
TABLE 2
Impurities in Starting Materials
Starting Material Mole % Impurity
CF4 None detected CaF6 None detected n-C4FlO 0.18% C02 i-C4F1o less than 1% i-C4F8 n-CsFl2 None detected n-C6F14 None detected 2-CF3CsFl 0.24% C6F12 3-CFsCsFii None detected




10
When canning a sample for irradiation a cleaned sample tube was connected by Flex fittings and vacuum hose to the vacuum system where it was baked out under vacuum with a cool torch flame. The storage bulb containing the starting compound was then opened and the desired amount of sample, as indicated by the system pressure, was bled into the vacuum system from the storage bulb. The entire quantity of sample was condensed into the aluminum sample tube by liquid nitrogen and the tube was pinched shut and Heliarc welded.
After removal from the vacuum system the canned
sample tube was weighed. Several days later the sample tube was weighed again and discarded if it had lost weight.
The CF4+C2F6 mixtures were prepared by distilling the desired quantities of CF4 and C2F6 into large storage bulbs on the vacuum system where the mixtures were allowed to mix for at least 24 hours to insure uniformity. A reference sample of each mixture was sealed under vacuum in a glass reference sample tube to be analyzed when the irradiated samples were analyzed.
When preparing the CF4 + carbon granules and CF4 +
aluminum powder samples, the solid material was packed into the cleaned sample tubes before the tubes were connected to the vacuum system. The tubes were then baked out, filled with CF4, and sealed as before. The height of solid material in the sample tubes was determined by the weight of solid material packed into the tubes.




Because of the high vapor pressure of CO at liquid nitrogen temperature (482 mm. Hg. at -1960C.), the CF4+CO samples were prepared in a different manner from the other samples. After the cleaned sample tube was connected to the vacuum system and baked out under vacuum the desired amount of CF4 was distilled into the sample tube after which the tube was isolated from the rest of the vacuum system by closing a stopcock. Carbon monoxide was then bled into the vacuum system from a pressurized cylinder until a desired reading on the manometer of the vacuum system was observed (approximately 600 mm. Hg.). The stopcock to the sample tube was then opened and the CO was allowed to condense on top of the CF4 in the sample tube until a certain lowering of the manometer reading was observed. The sample tube was pinched shut and sealed as before.
Irradiations
The mixtures containing fluorocarbons were irradiated for approximately six weeks in tube no. 11 of the College of Engineering Cobalt 60 gamma ray source at the University of Florida and for periods of approximately one week and four weeks in the Low Intensity Training Reactor at Oak Ridge, Tennessee.
The pure fluorocarbon samples were not irradiated in the Low Intensity Training Reactor but were irradiated for approximately six weeks in tube no. 11 of the Cobalt 60 gamma ray source.




12
The sample tubes of mixtures containing fluorocarbons were randomly placed in the sample space of the Cobalt 60 gamma ray source and the calculated absorbed doses in these samples are not as accurate as the absorbed doses calculated for the pure fluorocarbon samples. The latter sample tubes were positioned and retained against the walls of the irradiation basket which permitted more accuracy in calculating the absorbed doses for these samples. (See Appendix 1.)
Absorbed doses could not be calculated for the samples irradiated in the Low Intensity Training Reactor, however, the flux to which the samples were exposed was approximately
2 x 1013 neutrons/cm.2 sec. The irradiation temperature in the Low Intensity Training Reactor was approximately 1000C. while-the irradiation temperature in the Cobalt 60 gamma ray source was room temperature.
Analysis
The gas chromatographic analysis of all samples was done with a Perkin Elmer Model 154 Vapor Fractometer. The recording instrument was a one MV full scale range Honeywell strip chart recorder. The peak areas were measured by a Perkin Elmer Model 194 B Printing Integrator which gave 6000 integrator counts per minute for full scale deflection on the recorder.
The response of the thermistor detector cell was calibrated (Appendix 2) so that the peak areas could be converted to weight percentages.




13
A temperature-programmed one meter silica gel column
described in Table 3 was used to resolve compounds containing four carbon atoms or less. A 50-ft. n-hexadecane column described in Table 4 was used to resolve compounds which contained from three to seven carbon atoms, and a two meter Kel-F grease column described in Table 5 was used to resolve compounds which contained more than seven carbon atoms into molecular weight classes.
The aluminum tubes containing the irradiated samples were opened into a vacuum system adjacent to the fractometer where the samples were stored in storage bulbs until analyzed. After the samples had remained in the storage bulbs for at least 24 hours to insure uniformity, they were introduced into the fractometer by a gas sampling value. The average size sample taken for analysis was approximately 0.02 moles.- Since the chromatographic system is able to detect less than 2.0 x 10-7 moles of solute in the carrier gas stream, mole percentages as low as 0.001 can be detected with the analytical system used.
Identification of all saturated fluorocarbon products up through the perfluorohexane isomers was by the appearance time of known standards. The identification of the perfluoroheptane isomers is fully discussed in Chapter V. All isomers of each saturated fluorocarbon compound were found to appear chromatographically between the normal isomer of the compound and the normal isomer of the next higher molecular weight saturated fluorocarbon compound.




14
TABLE 3
*Description of Silica Gel Column No. 79 Length = 1 meter of stainless steel tubing i. d. 0.180 inches packing = 30/60 mesh silica gel from Wilkens Instrument and Research, Inc. Column wrapped with 34 feet of asbestos covered nichrome V strand size 18 resistance wire, 0.41 ohms/ft.
Temperature Program
Time Amperage Setting on Resistance
(Min.) Wire (amps)
0 0
6 1.00
26 1.50 34 1.75 52 2.50 76 3.00
Pi = 28.7 psia,'Po = 14.7 psia. Helium gas flow rate at column exit 42.0 cm.3/min. at 240C.
Compound Appearance Time (Min.)
air 1.7 CF4 4.3 CO2 15.0 C2F6 16.4 C2F60 29.2 CsFe 40.5 C3F8a0 47.7 peak 1 56.3 total C4F1o 58.9 (C2F5s)20 61.6




15
TABLE 4
Description of n-hexadecane Column Length = 50 feet i.d. = 0.197 inches Temperature = 44C. Pi = 34.7 psia Po = 14.7 psia n-C16eHs4 Vol. = 63.7 cm.3 on 30/80 mesh Chromasorb P Helium gas flow rate at column exit = 18.1 cm.3/min. at 24C00.
Appearance Retention Partition Time Volume Vo0 Coefficient,K Compound (Min.) (cm. 3)
air and CF4 18.7
C2F6 19.7 10.2 6.245 C2FO 20.5 18.4 3.462 CsFa 21.6 29.6 2.152 C3F80 21.6 29.6 2.152 (C2F5)20 23.1 45.0 1.416 C02 24.8 62.3 1.022 n-C4Flo 25.0 64.4 0.989 i-C4F1o 26.0 74.6 0.854 n-C5F12. 31.4 129.7 0.491 i-C5F12 33.8 154.2 0.413 n-CeF14 42.1 238.9 0.267 2-CFsC5F11 45.2 270.6 0.235 3-CF3CsF11 47.5 294.1 0.217 2,3-(CF3)2C4F8 52.8 348.2 0.183 n-C7F1re 60.6 427.8 0.149 2-CFsCeFs3 64.7 469.7 0.136 3-CFsCe6F13s 66.4 487.0 0.131 3-C2FsC5F11 70.7 530.9 0.120 2,4-(CFa)aCs5Fo10 74.2 566.7 0.112 2,3-(CFs)2CsFlo 76.8 593.2 0.107 2,2,3-(CF3)3C4F7 83.4 660.6 0.0964 3,3-(CF3)2CsFlo 88.0 707.6 0.0900 n-CeF 8 93.2 760.7 0.0837




16
TABLE 5
Description of Kel-F No. 90 Grease Column No. 5 Length = 2 meters 0.25 inch copper tubing Temperature = 8300. P. = 15.0 psia Po = 14.7 psia
1.20 gm. Kel-F/gm. acia washed 65/100 mesh Celite 545 (Johns-Manville) ml. of solvent per meter of column at RT = 3.0 Helium gas flow rate at column exit = 27.8 cm.3/min. at 240C.
Compound Appearance Time Retention Volume, (Min.) cm.3(Vo)
Air 1.2
n-C4F1o 1.5 8.3 n-CsF12 1.9 19.4 n-C6F14 2.6 38.9 n-C7FIe 4.35 86.1 n-CaFie 7.0 161.2 n-CF20 11.4 283.5 n-CloF22 19.3 503.1 n-C11F24 34.5 925.7 n-Ca12F2 55.0 1495.6




17
Since the linear relationship of the logarithm of the retention time of normal saturated hydrocarbon compounds versus the number of carbon atoms in the compound(1,IZ) seems to apply equally well to the normal saturated fluorocarbon compounds under the experimental conditions used, the appearance times of the normal saturated fluorocarbon compounds above perfluoronormalhexane were estimated by extrapolation of such plots. (See Figure 1.)
Products larger than the perfluoroheptane isomers were grouped into molecular weight classes by the above method. Three products which appeared chromatographically on the silica gel column in a regular pattern following the corresponding perfluoroalkane of the same number of carbon atoms were identified as perfluoroethers. The product perfluorodimethylether (C2Fe60) was trapped out and identified from its mass spectrum.(8) Because of a small sample size the mass spectra of the perfluoromethylethylether (C3F80) peak was inconclusive; it did show, however, that this compound contained oxygen. A known sample of perfluorodiethylether (C2F5)20 was added to the irradiation products and the peak area of the product identified as (C2F5)20 was increased.
CO2 was a product in all irradiations but it could not be quantitatively reported in the CF4+C2Fe mixtures because the C02 peak was obscured by the much larger C2F6 peak.




18
lbl 1000
n-C7F,6
Q)
a n-CeF14
o
0
S
n-C5F12
0
100
o
O C
0
*0
10 j 1 1 I I
3 4 5 6 7 8 9
Q)
Number of Carbon Atoms in
Normal Saturated Fluorocarbons Figure 1 Logarithm of Retention Volume Versus )Number of Carbon Atoms in Normal Saturated luorocarbons
.H
4-)
00
3
0ubro abnAosi
~~~~Number of Carbon Atoms in ra auae looabn




19
The unknown peaks which appeared on the chromatograms are identified by retention time or retention volume only. These products are probably fluorocarbon compounds which contain oxygen as will be discussed later. The response of the detector to the unknown peaks was assumed to be the same as the response of the detector to the nearest known peak.
No evidence of unsaturated or cyclic fluorocarbon products was found. Standard reference compounds for unsaturation were perfluoroethylene (C2F4), perfluoropropene (n-CsFe), perfluoroisobutene (i-C4Fs), cis and trans perfluorobutene-2 (cis-C4Fa, trans-C4Fa), perfluoro2-methylpentene-2 and cis and trans perfluoro-2-methylpentene-3. Standard reference cyclic compounds were: perfluorocyclopentane (cyclo-C3F6), perfluorocyclobutane (cyclo-C4Fe), perfluorocyclopentane (cyclo-CsFio), and perfluorocyclohexane (cyclo-C6F12).
Weight percentages reported in the analyses of the irradiated samples are estimated to be correct to better than + 1 percent of the value reported. G values reported for the pure fluorocarbon samples are estimated to be accurate to at least + 5 percent, and G values reported for the mixtures containing fluorocarbons are accurate to at least + 10 percent.




III. LOW INTENSITY TRAINING REACTOR IRRADIATIONS
OF MIXTURES CONTAINING FLUOROCARBONS
CF4 + C2F6 Mixtures
Weight percentages of the radiolysis products of the CF4+C2F6 mixtures are given in Table 6 and the weight percentages of products are plotted versus weight percent of C2F6 in the original mixtures in Figures 2,3, and 4.
All of the identified products of the radiolysis of
the CF4+CaFe mixtures can be explained by radical formation and recombination. The primary radicals formed upon irradiation whether by ionization and subsequent neutralization or by decomposition of excited molecules will be F, CF3, and C2F5 radicals; the concentration of each radical will depend upon the amount of C2Fe in the original mixture.
CF4 > CF3 + F (1) C2F6 --W"--> 2CF3 (2) C2F6 -AM C2F5 + F (3)
Bibby and Carter ) give the C2Fs-F bond strengthi as approximately 5.50 e.V./molecule and the CFs-CF3 bond strength as 6.05 e.V./molecule. Dacey and Hodgins(20) estimate that the strength of the first C-F bond in CF4 is more than 6.68 e.V./molecule. Therefore it seems that reactions (2) and (3) must be somewhat favored over reaction 20




21
(1) and that the concentration of CF3 radicals will always be larger than the concentration of C2F5s radicals during the radiolysis of the CF4+C2Fe mixtures.
Fallgatter and Hanrahan ())have pointed out that implications that the radiation chemistry of fluorocarbons is centered in C-C bonds should be avoided. The radiolysis products found in this work supports their conclusion.
All of the identified products can be described as arising from recombination of radicals. Recombination of the primary radicals will produce CsFe and n-C4Fo10 as primary radiolysis products. The starting materials, CF4 and C2F6, will also be primary products.
CF3 + F CF4 (4) 2CF3 -- C2F6 (5) C2F5 + F C2F6 (6) CF3 + C2F5 C3Fe (7) 2C2F5s n-C4F10 (8)
The primary radiolysis products CsFe and n-C4F1o will also undergo radiolysis and produce secondary radicals which will recombine with primary as well as with other secondary radicals to produce all of the observed secondary products.
C3Fe. -v-- n-CsF7 + F (9)
--- i-C3F-7 + F (10)
-A -- C2Fs + CF3 (11) n-C4Flo --A--. n-C4F9 + F (12)
-- 1-CFsCs3F6 + F (13)
--AvA n-C3F7 + CF3 (14)
-- -4 2C2Fs (15)




22
100
\ /
SCF4/
o 60
o/
0 \ /
0
0o
o 40/ N S4 Original / \ Original C) Amount of--- -Amount of CF4 o C2F6 / C) /e
20 / C2F6
/ N'
20
Q)
0 I I I I
100% CF4 100% C2F6
Weight Percent in Original Sample Mixture Figure 2
Weight Percent of CF4 and C2F6 in the Products
from the Low Intensity Training Reactor Irradiations of the CF4+C2F6 Mixtures




23
8
7
-6
0
5
0
O
o
0 4
-)
P- 3" (D
15
2
n-C4Flo
n-CsFl2 n-CeF14
100% CF4 100% C2Fe
Weight Percent in Original Sample Mixture Figure 3 Radiolysis Products with Straight-Chain
Structures from the Low Intensity Training
Reactor Irradiations of the CF4+C2Fe Mixtures




24
10
9 i-CsF12
8
4D 7
O0
6
6 i-C4Flo
0
0
S
R 5
U
0
4~ (4
.)
Q)
2
2, 3- ( CF3) 2C4Fa
15 .24-(CF3)2 Cs?
-CFsCs?11
100%. CF4 100% C2F6
Weight Percent in Original. Sample Mixture Figure 4 Radiolysis Products with Branched Structures
from the Low Intensity Training Reactor
Irradiations of the CF4+C2Fe Mixtures




25
All of the secondary products can be explained by radical recombination.
R1 + R2 ~ RjR2 (16)
All products larger than CaF1e can be explained by applying the above mechanism to the secondary products.
No unsaturated fluorocarbons were found in the
radiolysis products of the CF4+C2F6 mixtures so disproportionation reactions, abstraction reactions,and molecular expulsion of F2 probably do not occur in the radiolysis of saturated fluorocarbons.
Figure 2 shows that as the weight percent of C2F6
in the original mixture is increased the amount of CF4 in the products is increased by up to five times the amount of CF4 in the original CF4+C2F6 mixture. This indicates quite clearly that reaction (4) is very significant during radiolysis since this is the only method by which CF4 can be made. Figure 2 also shows the great stability of CF4 to irradiation and this is not surprising since the C-F bond strength in CF4 is greater than in any other fluorocarbon. The stability of CF4 to irradiation is also illustrated in the Cobalt 60 irradiations of pure fluorocarbons. (See Table 12.)
As the weight percent of C2F6 in the original mixtures is increased the initial concentration of C2F5 radicals during radiolysis will be increased and the weight percentages of all products greater than C2F6 should be increased also; i.e., reactions (7) and (8) will become more important. This




26
is illustrated in Figures 3 and 4. It should be noted, however, that the weight percentages of C3F8 and n-C4Flo in the radiolysis products reach maximum values and that these primary products are then used up by undergoing radiolysis themselves.
Thus for a one week irradiation period in the Low
Intensity Training Reactor the optimum original composition of a CF4+C2F6 mixture should be approximately seventy weight percent C2F6 in order to make the greatest amount of C3F8 as product. If the period of irradiation is changed the maximum in the C3F8 curve will probably be changed also.
It can be seen from Figure 3 that all of the straightchain fluorocarbon products from the one week Low Intensity Training Reactor irradiation of the CF4+C2F6 mixtures reach a maximum percentage or seem to be reaching a maximum percentage as the weight percent of C2F6 in the original sample mixture is increased, while from Figure 4 none of the fluorocarbon products with branched structures reach a maximum percentage. Therefore for the dose absorbed fluorocarbon radicals with branched structures must be more abundant during radiolysis of the CF4+C2F6 mixtures than fluorocarbon radicals with straight-chain structures. This indicates that F atoms must be harder to remove from terminal CF3 groups than from other carbon atoms in fluorocarbon molecules.
Branched products cannot be made as primary products from the recombination of the initial radicals formed in




27
the CF4+C2F6 mixtures so primary products, secondary products, etc., must be made first and then undergo radiolysis themselves. The large amount of starting material converted to products shown in Figure 2 and the large amount of primary products, secondary products, etc., found in the radiolysis products of the CF4+CaF6 mixtures indicates that this actually happens. For a small dose, however, only the primary products should be formed in the radiolysis of the CF4+C2F6 mixtures and branched fluorocarbon products should not be found in the radiolysis products. This can be seen in the Cobalt 60 irradiations of the CF4 + C2F6 mixtures. (See Chapter IV.)
Thus it appears that all radiolysis products arise from radical recombination, as postulated, and that the ratios of the concentrations of the various radicals during radiolysis change with the initial composition of the CF4+C2F6 mixtures. The instantaneous radical concentrations will also be a function of the total dose absorbed in the mixtures since all products will undergo radiolysis and produce radicals too.
It appears that an equilibrium composition of products in the CF4+C2F6 mixtures has not been reached if indeed an equilibrium condition can be reached in the radiolysis of the CF4+C2F6 mixtures.
The radiolysis products containing oxygen probably arisefrom wall reactions. The handling procedure used (See Chapter II) should ,exclude oxygen as a gas phase




28
contaminant in the samples. The oxygen in the radiolysis products must come from adsorbed oxygen and/or aluminum oxide on the inside walls of the sample tubes.
Fluorine is the most electronegative element known and since fluorine atoms are present during radiolysis it seems quite likely that the oxygen on the sample tube wall will be replaced by fluorine thus producing oxygen as a contaminant in the radiolysis products. Kevan and Hamlet(2) noted that when C2F6 was irradiated in newly made brass cells that C02 was always an irradiation product indicating the presence of oxygen but after the cells were "conditioned" by several irradiations that the radiolysis products usually contained no C02.
Rexroad and Gordy( ) found that when Teflon was
irradiated fluorocarbon radicals were definitely present in the solid and that when the irradiated Teflon was exposed to oxygen the fluorocarbon radicals reacted with the oxygen to form different radicals which contained oxygen.
The random variation in the amount of radiolysis
products which contain oxygen indicates that the amount of oxygen present must vary from one sample tube to the next.
None of the radiolysis products were identified as
cyclic or unsaturated compounds. The unidentified products could be unsaturated or cyclic compounds but since the presence of oxygen in the mixtures is established it seems more likely that these products are perfluorooxygen compounds such as perfluoroethers and/or perfluoroalcohols.




29
Mixtures Containing CF4 and Nonfluorocarbon Components
Weight percentages of the radiolysis products from
the one week Low Intensity Training Reactor irradiations of the mixtures containing CF4 are given in Table 7 and the results of the four week irradiations are given in Table 8. The results of Mailen( ) for pure CF4 are also given in Table 8.
The presence of aluminum powder in the radiolysis of CF4 increases the yield of all products. This is not surprising since the aluminum powder provides a large surface area and surfaces may affect gas-phase radiolysis in which free atoms are intermediates.(16) An unusual amount of oxygen containing material was not found in the products and this may be because the aluminum powder was freshly prepared and probably contained very little aluminum oxide and/or adsorbed oxygen.
The initial compositions of the CF4+CO mixtures were not known exactly, but they were approximately 50 mole percent CF4. The large amount of products and small amount of CF4 left after radiolysis indicates that considerable reaction took place in the CF4+CO mixtures. No CO was left in the radiolysis products.
The CO probably acts as a radical scavenger during radiolysis especially for the very reactive F atoms and this would mean that more fluorocarbon compounds should be found in the radiolysis products of the CF4+CO mixtures than in the radiolysis products of pure CF4. This actually occurs as seen in Table 8.




30
The very large amounts of C2F60, C3F80O, and (C2F5)20 found in the radiolysis products of the CF4+CO mixtures certainly support the identifications of these products. Many unknown products were found in the radiolysis of the CF4+CO mixtures. The retention volumes of most of these products were identical with the retention volumes of unknown products found in the radiolysis of the saturated fluorocarbons and this supports the idea that the unknown products found in the radiolysis of the saturated fluorocarbons may contain oxygen.
The abundance of the fluorocarbon products with branched structures can also be seen in the radiolysis products from the CF4+CO mixtures.
The presence of carbon granules during the radiolysis of CF4 causes extreme degradation of the CF4. For the one week irradiation of CF4 + carbon granules in the Low Intensity Training Reactor only 57.0 percent of the original amount of CF4 was found in the radiolysis products and for the four week irradiation only 32.0 percent of the original amount of CF4 was found in the radiolysis products. However, when pure CF4 is irradiated for four weeks in the Low Intensity Training Reactor approximately 98.3 percent of the original' amount of CF4 is found in the radiolysis products.
Only 90.0 weight percent of the original sample was recovered as gas in the one week irradiation and only 61.0 weight percent of the original sample was recovered as gas in the four week irradiation of the CF4 + carbon granule




31
mixtures. Therefore in both samples some of the tarting material remained in the sample tubes as involatile products.
The relative amounts of the fluorocarbons found in the gaseous products from the CF4 + carbon granule mixtures are almost identical with the product composition found in the radiolysis products of the CF4+C2F6 mixtures. This means that the same processes must be responsible for the radiolysis products in both systems. The radiolysis products can be explained by the formation and recombination of radicals given for the CF4+C2F6 mixtures except that C2F6 is not present as a starting material in the CF4 + carbon granule mixtures.
Kuriakose and Margrave(2) have reacted fluorine with graphite and they found that at temperatures between 315 and 5300C. the graphite gained weight while above 60000. the graphite loses weight with the formation of only gaseous fluorides, mainly CF4. They also suggest that the reaction might involve dissociation of fluorine on the graphite surface.
Since fluorine atoms are present during the radiolysis it certainly seems possible that the fluorine atoms might react with the carbon in the CF4 + carbon granule samples. The loss of fluorine atoms in the radiolysis mechanism would result in an increase in the production of larger fluorocarbon molecules and this would explain the large product yields found in the CF4 + carbon granules mixtures.




32
A material balance on the gaseous products from the one week irradiation of CF4 + carbon granules shows the weight loss of fluorine in the gas to be 0.0406 grams and the weight of carbon in the gaseous products was increased by 0.0005 grams. From this it appears that fluorine is lost to the carbon and that the carbon must act as a radical scavenger during radiolysis.
A material balance on the gaseous products from the four week irradiation of CF4 + carbon granules shows the weight loss of fluorine to be 0.1460 grams and the weight loss of carbon to be 0.0160 grams. The chemical formula of such a product would be CF5s.78 which indicates that more fluorine radicals are lost to the carbon than any other kind of radical. This is logical because fluorine atoms can diffuse faster than any other radical present in the radiolysis mixture.
The material balances are not absolutely correct due to the difficulties in obtaining accurate sample weights but they do show that fluorine and possibly carbon-containing radicals are lost to the carbon granules during the radiolysis of the CF4 + carbon granules mixtures.
The relatively large yield of products that contain oxygen is probably due to a relatively large amount of oxygen originally adsorbed on the carbon granules in the sample mixture. The abundance of the fluorocarbons with branched structures is also evident in the radiolysis products of the CF4 + carbon granule mixtures.




After the gaseous portion of the sample was removed from the CF + carbon granules sample tube which was irradiated for four weeks, half of the tube was connected to a vacuum system and the tube was heated under vacuum with a cool torch flame. A few drops of a clear viscous liquid condensed in the vacuum system. This liquid was soluble in fluorocarbon liquids but a chromatogram of this material dissolved in C7Fl6 showed no fluorocarbon products up through C13F28. This liquid could be a polymer or a mixture of long chain fluorocarbon products but from the abundance of branched structures the liquid material must be highly branched.
The other half of the sample tube was treated with C7F16 solvent and after evaporation of the solvent some clear viscous liquid remained. About 0.050 grams of materialwere recovered from the whole tube and since about 0.1620 grams of sample was lost to the carbon granules it is not known if any of the carbon from the carbon granules was actually transformed into fluorocarbon material. Radioactive carbon tracer studies could answer this question.




34
TABLE 6
Analysis of the CF4+C2F6 Mixtures Irradiated
for One Week in the Low Intensity Training Reactor
Sample No. D-2
Sample Wt. 0.3966
Density 0.5352 gm./cm.3 Original Composition (Wt.%) 86.66 CF4 13.34 C2F6 % Recovery as gas 98.66
Product or Vo on
n-hexadecane column Wt.%
CF4 81.40 C2F6 7.69 CaF60 3.29 C3Fe 1.16 CsF O 0.271 peak 1 3.77
total C4Fio 0.232
n-CsFla2 0.0819 i-CsFl2 0.504 183.8 cm.' 0.551 195.0 cm.3 0.724
218.5 cm.3 0.0455 n-C6F14 0.123
252.2 cm.3 0.0614 2,3-(CFs)2C4F8 0.126 n-C7F16 0.058




55
TABLE 6 (Continued) Sample No. B-4
Sample Wt. 0.4944 gm. Density 0.5352 gm./cm.3 Original Composition (Wt.%) 67.94 CF4 32.06 CaF6 % Recovery as gas 100.28
Product or Vo on
n-hexadecane column Wt. %
CF4 74.73 C2Fe 14.32 C2F60 0.402 CsFe 4.94
C3F0O 0.137 n-C4Flo 1.05 i-C4F10o 1.53
108.3 cm.3 0.0271 n-CsF12 0.236 i-CsF12 1.47
183.8 cm. 0.0065 195.0 cm.3 0.0113 n-CeF14 0.0528 2-CF3C5F11 0.0541 3-CFsCs5F11 0.0358 2,3-(CF3)aC4F8 0.687 2,4-(CF3)aCs5Fo10 0.126 3,3-(CF3)2CsFlo 0.135
CsF18 0.0549




36
TABLE 6 (Continued) Sample No. A-4
Sample Wt. 0.4330 gm. Density 0.4810 gm./cm.3 Original Composition (Wt.%) 39.86 CF4 60.14 CaFe SRecovery as gas 100.60
Product or Vo on
n-hexadecane column Wt.
CF4 58.48 CaFe 15.79 C2F60 0.548 C3Fa 7.76
C3F80 0.266 peak 1 0.138 n-C4Fo 2.01 i-C4Flo 4.29
102.1 cm.3 0.0971 n-CsFl2 0.601 i-CsFl2 4.63
183.8 cm." 0.0225
3
195.0 cm. 0.0464 n-CeF14 0.197 2-CFsCsF11 0.282 3-CF3Cs5F11 0.182 2,3-(CFa)2C4F8 2.34
n-C7Fe 0.0788 2-CF3C6F13 0.0466 3-CF3C6F1s 0.0654




37
TABLE 6 (Continued)
Product or Vo on
n-hexadecane column Wt. %
2,4-(CF3)aCsFo0 0.606
2,3-(CF3)aCsFo10 0.0257 3,3-(CF3)2CsFo 0.517 C8F18 0.726 C9Fao20 0.258 Sample No. C-3
Sample Wt. 0.6208 gm. Density 0.6855 gm./cm.3 Original Composition (Wt. ) 18.53 CF4 81.47 C2F6 % Recovery as gas 100.37
Product or Vo on
n-hexadecane column Wt. %
CF4 50.85 CaFe 12.39
C2F60 0.393 C3F8 7.88 C3F8O0 0.168 n-C4F1o 2.37 i-C4F10o 6.74
n-CsF12 0.859 i-C5FI2 7.52
n-CeFi4 0.273 2-CF3CsF11 0.647 3-CFsC5Fi 0.420 2,3-(CF3)2C4Fe 3.94
n-C7FI6 0.0720




38
TABLE 6 (Continued)
Product or Vo on
n-hexadecane column Wt.%
2-CF3C6F13 0.299 3-CF3C6F13s 0.216 2,4-(CF3)2CsFlo 1.09
2,3-(CF3)2CsFlo 0.0546 3,3-(CF3)2C5F1o 0.942 CaFe 1.51
CFo20 0.843 Clo22 0.441
Sample No. E-1
Sample Wt. 0.6304 gm. Density 0.8507 gm./cm.3 Original Composition (Wt. %) 8.25 CF4 91.75 C2Fe % Recovery as gas 100.29
Product or Vo on
n-hexadecane column Wt. %
CF4 50.44 C2F6 8.55 C2F60 0.558
CsFe 6.50
CsF80 0.232 n-C4Flo 1.81 i-C4F1o 7.92
102.1 cm.3 0.0040 n-C5F12 0.829 i-C5F12 9.52
188.9 cm. 0.0299 n-C6F14 0.343




39
TABLE 6 (Continued) Product or Vo0 on
n-hexadecane column Wt. %
2-CF3Cs5Fii 0.875 3-CF3C5Fll 0.510 2,3-(CF3s)2C4F8 5.17 397.2 cm.3 0.0054 n-C7Fb6 0.0688 2-CF3C6Fi3 0.201 3-CFsC6Fi3 0.389 3-CaFsCs5F11 0.0299 2,4-(CFs)aCsF10 1.49 2,3-(CF3)aCsFlo0 0.142 3,3-(CF3)2C5F10o 1.14 CBFi B 1.35 CoFbo 0.998 CloF2a2 0.814 CaF24 0.0861




40
TABLE 7
Analysis of the Mixtures Containing CF4 and
Nonfluorocarbon Components Irradiated for One
Week in the Low Intensity Training Reactor Sample No. 1-3 CF.+ aluminum powder Sample Wt. 0.3576 gm. CF4 0.5579 gm. al powder % Sample recovery as gas 117.67*
Product or Vo on
n-hexadecane column Wt. %
CF4 92.60
CO2 0.0062
CaFe 4.33.
C2F60 2.41 C3F8 0.404 03F80 0.151
peak 1 0.0533 n-C4F1o 0.0110 i-C4Fo0 0.0161 (C2Fs)20 0.0126 97.1 cm.3 0.0002 115.4 cm.3 0.0013 n-CsF12 0.0002 i-C5F12 0.0009
*Some aluminumpowder lost when sample removed which makes % sample recovery too large.




41
TABLE 7 (Continued) Sample No. G-3 CF4+CO Sample Wt. 0.3101 gm. % Sample recovery as gas 97.52
Product or Vo0 on
n-hexadecane column Wt. %
CF4 32.51 C02 2.99 C2F6 6.38 C2F60 35.01 C3F 1.11 C3FaO0 4.92 peak 1 3.92
total C4F10 0.639 (C2Fs)20 1.46
108.3 cm.3 0.263 118.5 cm.3 0.589 n-C5F12 0.912 i-C5F12 5.25
176.7 cm.s 0.263 195.0 cm.3 0.223 218.5 cm.3 0.383 n-C6F14 1.25
252.2 cm.3 0.0851 284.8 cm.3 0.302
3-CFsCsF1 0.0715 2,3-(CF3)2C4Fs 0.918
382.9 cm.3 0.0362




42
TABLE 7 (Continued)
Product or Vo0 on
n-hexadecane column Wt. %
n-C7F16 0.0831 3-CFsC6FIs 0.158 550.4 cm.3 0.176
2,4-(CF3)2Cs5FIo0 0.0431 609.6 cm.3 trace 650.4 cm.3 trace 3,3-(CFs)2CsFo10 0.0479 CsFx8 0.0376 Sample No. H-3 CF4 + Carbon granules Sample Wt. 0.4036 gm. CF4 0.1194 gm. carbon granules % Sample recovery as gas 90.04
Product or Vo on
n-hexadecane column Wt. %
CF4 63.25 C02 0.554 C2F6 8.73 C2F60 8.57 CsFe 2.79 C3F80 1.22 peak 1 0.178 n-C4Fo10 1.68 i-C4F10 2.56
(C2Fs)20 0.878
97.1 cm.3 0.0105 102.1 cm.3 0.0945 n-CsF12 0.659




43
TABLE 7 (Continued)
Product or Vo on
n-hexadecane column Wt. %
i-CsF12 4.27 183.8 cm.3 0.029 195.0 cm.3 0.086 n-C6F14 0.554 252.2 cm.S 0.0338 2-CFsCsFa. 0.254 3-CFsCs5F11 0.157 2,3-(CF3)aC4F8 2.01 n-C7F16 0.0381 2-CFsC6FI3s 0.0532 3-CFsC6Fls 0.113 3-C2FsCs5F1l 0.0741 2,4-(CF3)aCs5Fo10 0.439 2,3-(CF3)2CsFlo 0.0226 3,5-(CF3)aCsFlo 0.393 C8F18 0.402 C9F2o 0.0917




44
TABLE 8
Analysis of the Mixtures Containing CF4 and
Nonfluorocarbon Components Irradiated for
Four Weeks in the Low Intensity Training Reactor
Sample No. 8 CF4* Sample Wt. 0.1752 gm. % Sample recovery as gas 100.97
Product or Vo0 on
n-hexadecane Column Wt. % CF4 98.31 C2F6 0.95 CaFe60 0.72 C3F8 0.03
*Results from J. C. Mailen(1)
Sample No. G-5 CF4-+C0 Sample Wt. 0.3248 gm. % Sample recovery as gas 96.45
Product or Vo on
n-hexadecane Column Wt. % CF4 23.15 C02 1.91 CaF6 4.41 CaF60 28.49
CsFe 0.692 C3FS0 3.49 peak 1 4.29 total C4F0o 0.270 (c2F5)20 0.851
108.3 cm.s 0.796 118.5 cm.3 1.68 n-C5F12 2.49




45
TABLE 8 (Continued) Product or V0 on n-hexadecane Column Wt. i-CsF12 12.19 176.7 cm.3 0.772 195.0 cm.3 0.851 218.5 cm.3 1.45 n-C6FI4 4.70 252.2 cm.3 1.17 284.8 cm.3 0.469 2,3-(CF3)2C4F8 2.71 382.9 cm.3 0.282 n-C7F16 0.425 3-CF3C6Fl3 0.519 550.4 cm.3 .815 2,4-(CF3)2C5Fio0 0.139 609.6 cm.3 0.0799 650.4 cm.3 0.9487 3,3-(CF3)2C5Fo10 0.214 CasF8 0.644 C9F2o 0.0071




46
TABLE 8 (Continued) Sample No. H-2 CF4 + carbon granules Sample Wt. 0.4190 gm. CF4 0.1239 gm. carbon granules % Sample recovery as gas 61.34
Product or Vo0 on
n-hexadecane Column Wt. %
CF4 52.15 C02 trace C2F6 13.09 C2F60 4.94 C3Fe 4.83 C3F80 1.07
peak 1 0.189 n-C4Flo 1.63 i-C4F1o 2.21
(CaF5)20 0.674 n-C5F12 0.673
i-CsF12 6.71
176.7 cm.3 0.978
195.0 cm.3 0.0936 218.5 cm.3 0.0075
n-CeF14 0.842
252.2 cm.3 0.0483 2-CFsCs5FII 0.796 3-CF3CsF1 0.906 2,3-(CF3)2C4Fe 2.96
382.9 cm.3 0.0723 n-C7F16 0.0755




47
TABLE 8 (Continued) Product or Vo0 on n-hexadecane Column Wt. %
3-CF3C6F,3s 0.365 3-C2F5sCsF11 0o.184 2,4-(CF3)2C5FIo 0.620 2,3-(CF3)2CsFIo 0.189 3,3-(CF3)2CsFlo 0.572 C8Fa 1.80 C6F20 1.18 CloF22 0.156




IV. COBALT 60 GAMMA IRRADIATIONS OF MIXTURES
CONTAINING FLUOROCARBONS AND OF PURE FLUOROCARBONS CF4+C2Fe Mixtures
Weight percentages and G values (molecules/100 e.V.) of the radiolysis products of the CF4+C2F6 mixtures are given in Table 9 and the G values are plotted versus weight percent of C2F6 in the original mixtures in Figures 5 and 6. The G values are average G values and they were calculated from the weight of each product found in the radiolysis products. The initial G value for a given product will probably be somewhat higher than the values given here because for the total doses absorbed some of the initial products are destroyed by undergoing radiolysis themselves.
All of the identified products of the radiolysis of the CF4+C2F6 mixtures can be explained by the processes given for the Low Intensity Training Reactor irradiations on page 21.
The linear relationship between the G values of C3F8 and the weight percent of C2F6 in the original mixture shown in Figure 5 indicates that C3Fe is formed by a first order reaction with respect to the concentration of C2F6 and the curve for C4Fj0 shown in Figure 5 indicates that n-C4F1o is formed by a second order reaction with respect
48




49
0.8
0 117 Megarads 0.7 Q 77.2 Megarads 0 40.5 Megarads
0.6 0
C3Fe
0.5
Q)
0.4
0.3
O
0.2
Total C4F1o0 0
0.1 0
o I I I I,
100% CF4 100% CaF6
Weight Percent in Original Sample Mixture Figure 5 G Value of Primary Products Versus
Weight Percent of C2Fe in Original Sample Mixture




50
0.30
Dose in Mixtures 117 Megarads Dose in Pure CF4 77.2 Megarads Dose in Pure C2F6 40.5 Megarads
0.25
0.20
0
H 0.15
00F60
0.10
C3aF8 0
0.05 Ca0 0
(C2Fs) 20 C4F100
0
100% CF4 100% C2F6
Weight Percent in Original Sample Mixture Figure 6 G Value of Products Containing Oxygen
Versus Weight Percent of C2F6 in Original
Sample Mixture*




51
to the concentration of C2F6 in the original mixture. This certainly supports the hypothesis of free radical reactions occuring during the radiolysis of the CF4+C2F6 mixtures.
The reactions producing fluorocarbon products seem
to be occuring in the bulk of the medium since an absorbed dose variation of from 40.5 to 117 megarads seems to have little effect on the G values of fluorocarbon products.
The G values of products containing oxygen are plotted versus weight percent of C2F6 in the original sample in Figure 6. The shape of the C3F80 curve and the shape of the (C2F5)20 curve indicate that these products are also formed by free radical reactions. The G values of the products containing oxygen seem to be dose dependent and this substantiates wall reactions occuring since diffusion to the wall may control the reaction and the absorbed dose is directly proportional to the length of time of irradiation.
For the total absorbed doses in the Cobalt 60 irradiations of the CF4+C2F6 mixtures there was probably always ample oxygen present in the sample tubes so that the random variation in oxygen-containing products as found in the Low Intensity Training Reactor irradiations was not observed in these samples.
The shape of the C2F60 curve shown in Figure 6 may
reflect the variation in the concentration of CFs radicals at the tube wall during radiolysis as the weight percent of C2F6 is increased in the original sample mixtures.




52
Mixtures Containing CF4 and Nonfluorocarbon Components
Weight percentages and G values of the products from the Cobalt 60 irradiations of the mixtures containing CF4 are given in Table 10. It should be recalled here that the absorbed doses for these samples were calculated by assuming that the solid material had no effect on the gamma rays during radiolysis. This, of course, is a bad assumption so the G values given for these samples are not very reliable.
The presence of aluminum powder during the Cobalt 60 irradiation of CF4 seems to increase the yield of products in keeping with the results of the Low Intensity Training Reactor irradiations.
The presence of CO during the radiolysis of CF4 increases the yield of C02 but completely inhibits the formation of all other products except an unknown product which was found in this sample. CO was also found in the radiolysis products.
Nothing conclusive can be said about the results of
the Cobalt 60 irradiation of CF4 + carbon granules. Fluorocarbon products are being formed in this sample but the extreme degradation of CF4 that was noted in the Low Intensity Training Reactor irradiations of CF4 + carbon granules does not seem to be occuring in this sample. It should be remembered, however, that the absorbed dose calculated for this sample could be very much in error and the destruction of CF4 may be much larger than it appears to be in this sample.




53
Pure Fluorocarbon Samples
Weight percentages and G values of the radiolysis
products from the Cobalt 60 gamma ray irradiations of the pure fluorocarbon samples are given in Table 11.
The free radical reactions predicted for the formation of the perfluoroheptane isomers certainly appear to occur in the radiolysis of the pure parent perfluorocarbon samples. The agreement between the pattern of products predicted by the free radical mechanism and that of the actual products obtained in the radiolysis of the parent perfluorocarbon compounds is almost exact. (See Chapter V.)
G values for the disappearance of the starting material during radiolysis of the pure saturated fluorocarbon samples are given in Table 12. It can be seen from the table that perfluoromethane is the most stable to irradiation of all saturated fluorocarbons studied. CF4 is also the largest product in all saturated fluorocarbon irradiations except in the irradiation of i-C4F10. The i-C4Fo0 contained some i-C4Fa as an impurity and this may account for the low percentage of CF4 in the radiolysis products of the i-C4F1o sample.
The abundance of branched fluorocarbons as products
indicates that F atoms must be harder to remove from terminal CF3 groups than from other carbon atoms during radiolysis of saturated fluorocarbons.
It also appears that larger fluorocarbon molecules aPe
more stable to irradiation than smaller fluorocarbon molecules




excluding CF4. The larger fluorocarbon molecules were present as the liquid phase during irradiations and this may account for the lower G values since radical diffusion in the liquid is slower than in the gas phase. Therefore radical recombination to produce the parent molecule may be more significant in the liquid samples than in the gas samples.
The G values for the disappearance of the starting material for the liquid fluorocarbon samples agree very well with the value of 2 to 3 reported by Simons and Taylor(3) for C7F16.
The radiation chemistry of saturated fluorocarbons is much different in some respects than the radiation chemistry of the analogous hydrocarbons. The great strength of the C-F bond compared to the C-H bond and the weakness of the F-F bond compared to the H-H bond is reason enough to suspect the radiation chemistry of saturated fluorocarbons to be much different from the radiation chemistry of saturated hydrocarbons. Molecular hydrogen is a stable product in the radiolysis of saturated hydrocarbons but molecular fluorine has never been found in the radiolysis products of saturated fluorocarbons. Hydrogen abstraction by hydrogen atoms to form stable molecular hydrogen occurs in the radiolysis of saturated hydrocarbons but fluorine abstraction by fluorine atoms is essentially impossible in fluorocarbons. The absence of unsaturated products in the radiolysis of saturated fluorocarbons indicates that




55
disproportionation reactions which occur in the radiolysis of hydrocarbons do not occur in the radiolysis of saturated fluorocarbons.
For the above reasons saturated fluorocarbons should be more stable to irradiation than saturated hydrocarbons and the G values reported here certainly support this idea.




56
TABLE 9
Analysis of the CF4+C2F6 Mixtures Irradiated
in the Cobalt 60 Gamma Ray Source Sample No. 9 (reanalyzed)* Sample Wt. 0.1852 gm.
Density 0.222 gm./cm. Original Composition (Wt. %) 100.0 CF4 No. days in Cobalt 60 source 33.09 Absorbed dose = 8.92 e.V. = 77.2 megarads
Product or Vo on
n-hexadecane Column Wt. G Value
CF4 99.30 -0.997 C02 0.0931 0.264 C2F6 0.258 0.233 C2F60 0.296 0.240 C3F8 0.0306 0.020 C3F80 0.0226 0.014
* Sample from J. C. Mailen()
Sample No. D-4
Sample Wt. 0.4115 gm.
Density 0.5532 gm./cm.3 Original composition (Wt. %) 86.66 CF4 13.34 C2Fe No. days in Co 60 source 46.89 Absorbed dose 2.52 x 1021 e.V. = 117 megarads % Recovery as gas 104.20
Product or Vo on
n-hexadecane Column Wt. G Value
CF4 87.15 0.549 C2F6 12.10 -0.885 C2F60 0.239 0.153 C3Fe 0.240 0.126 C3F80 0.0663 0.0320 total C4F1o 0.0463 0.0191




57
TABLE 9 (Continued)
Product or Vo on
n-hexadecane Column Wt. % G Value
(C2Fs)20 trace
102.1 cm.3 0.0238
n-CsF12 0.0520 0.0177 i-CsFl2 0.0355 0.0121 n-CeFI4 0.0101 0.0030
2-CFsCsFi1 trace 3-CF3Cs5F11 trace
2,3-(CF3)2C4F8 0.0122 0.0036
Sample B-3
Sample wt. 0.5049 gm.
Density 0.5533 gm./cm.3 Original composition (Wt. %) 67.94 CF4 32.06 C7F6 No. days in Co 60 source 46.89 Absorbed dose 2.51 x 021 e.V. = 117 megarads % Recovery as gas 99.82
Product or Vo on
n-hexadecane Column Wt. G Value
CF4 68.04 0.137 C2Fe6 31.11 -0.833
C2F60 0.124 0.0974 C3Fa 0.414 0.266
03F80 0.119 0.0708 total C4F10 0.0882 0.0449 (C2Fs)20 0.0061 0.0030
102.1 cm.3 0.0198
n-CsF12 0.0298 0.0126 i-CsFl2 0.0345 0.0147 n-C6F14 0.0081 0.0028




58
TABLE 9 (Continued)
Product or Vo on
n-hexadecane Column Wt. % G Value
2-CFsCsF11 trace 3-CFsCsF11 trace
2,3-(CF3)aC4F8 0.0122 0.0043
Sample No. A-5
Sample Wt. 0.4550 gm.
Density 0.4931 gm./cm.3 Original composition (wt. %) 39.86 CF4 60.14 C2Fe No. days in Co 60 source 46.89 Absorbed dose 2.25 x 1021 e.V. = 117 megarads % Recovery as gas 100.09
Product or Vo on
n-hexadecane Column Wt. % G Value
CF4 39.26 -0.833 C2Fe6 59.19 -0.840
C2F60 0.102 0.0812
CsFe 0.693 0.450 CsFs80 0.183 0.109 total C4Flo 0.226 0.116
(C2F5)2O0 0.0275 0.0132
102.1 cm.3 0.0779
n-CsFl2 0.0722 0.0306 i-CsF12 0.0661 0.0280 n-CeF14 0.0305 0.0110
2-CF3C5F11 0.00969 0.00350 3-CFsCs5F11 0.0186 0.00672 2,3-(CFs)2C4Fa 0.0351 0.0127
n-C7F16 0.0163 0.00513




59
TABLE 9 (Continued)
Sample No. C-2
Sample wt. 0.5801
Density 0.6366
Original composition (wt. %) 18.53 CF4 81.47 C2Fe No. days in Co 60 source 46.89 Absorbed dose 2.90 x 102 1 e.V. = 117 megarads SRecovery as gas 100.71
Product or Vo on
n-hexadecane Column Wt. % G Value
CF4 17.64 -1.22 C2F6 80.09 -1.19 C2F60 0.194 0.152 CsFe 0.993 0.638 C3F80 0.244 0.144 total C4F1o 0.485 0.246
(C2Fs)20 0.0777 0.0369
102.1 cm.3 0.115
n-C5F12 0.055 0.0230 i-CsFl2 0.053 0.0222
176.7 cm.3 0.0076
n-C6F14 0.0209 0.00746 2-CFsCsF11 0.0034 0.00121 3-CFsCs5F11 0.0034 0.00121 2,3-(CFs)2C4Fs 0.0188 0.00671 total C7F16 0.0096 0.00298




60
TABLE 9 (Continued)
Sample E-2
Sample wt. 0.6128 gm.
Density 0.8498 gm./cm.3 Original composition (wt. %) 8.25 CF4 91.75 C2Fe No. days in Co 60 source 46.89 Absorbed dose 3.87 x 1021 e.V. = 117 megarads SRecovery as gas 100.94
Product or Vo on
n-hexadecane Column Wt. 5 G Value
CF4 7.65 .652 C2F6 88.70 -2.11
C2F60 0.214 0.133 C3Fe 1.19 0.602 C3FeO 0.371 0.173 total C4F10 0.492 0.197 (C2F5)20 0.0851 0.0320
102.1 cm.3 0.927
n-CsFl2 0.142 0.0471 i-CsF12 0.0865 0.0287
176.7 cm.s 0.0142 195.0 cm.3 0.0072
n-CeF14 0.0542 0.0153 2-CF3CsF11 0.0149 0.00421 3-CFsCsFi 0.0207 0.00585 2,3-(CF3)2C4Fe 0.0269 0.00760 n-C7F1e6 0.0090 0.00222




61
TABLE 9 (Continued)
Sample No. 68 (reanalyzed)* Sample wt. 0.3723 gm.
Density 0.447 gm./cm.
Original composition (wt. %) 100.0 C2F6 No. days in Cobalt 60 source 17.17 Absorbed dose = 9.43 x 102 0 e.V. = 40.5 megarads
Product or Vo on
n-hexadecane Column wt. G Value
CF4 0.279 0.755 CaF6 97.95 -3.542 C2F60 0.0868 0.134 C3F8 0.588 0.745 C3Fs0 0.0703 0.082 total C4F10 0.170 0.170
(Ca2F5)2a0 0.0247 0.0232
102.1 cm.s 0.757
n-CsFla2 0.0199 0.0165 i-C5F12 0.0258 0.0213 n-C6F14 0.0072 0.0051
2-CFsCs5Fi trace 3-CF3C5F11 trace
2,3-(CF3)2C4Fe 0.000144 0.000101
*Sample from J. C. Mailen(




62
TABLE 10
Analysis of the Mixtures Containing CF4
and Nonfluorocarbon Components Irradiated
in the Cobalt 60 Gamma Ray Source Sample No. I-1 CF4 + aluminum powder Sample Wt. 0.2948 gm. CF4 0.5406 gm. aluminum powder No. days in Cobalt 60 source 40.75 Absorbed dose 1.46 x 1021 e.V. = 79.1 megarads % Sample recovery as gas 106.58*
Product Wt. % G Value
CF4 98.73 -1.755 CO2 0.259 0.714 C2F6 0.718 0.631 C2F60 0.225 0.177 CsF8 0.0715 0.0461
*Some aluminum powder lost when sample was removed.
Sample No. G-2 CF4+CO Sample Wt. 0.3209 gm. No days in Cobalt 60 source 40.75 Absorbed dose 1.584 x 1021 e.V. = 79.1 megarads % Sample recovery as gas 98.19
Product Wt. % G Value
CO 10.20 CF4 88.76
C02 0.973 2.70 peak 1 0.068 0.060




63
TABLE 10 (Continued) Sample No. H-1 CF4 + carbon granules Sample Wt. 0.3495 gm. CF4 0.1045 gm. carbon granules No. days in Cobalt 60 source 40.75 Absorbed dose = 1.755 x 10" e.V. = 79.5 megarads % Sample recovery as gas 101.29
Product Wt. % G Value
CF4 99.09 -1.261 C02 0.362 0.997 C2F6 0.552 0.486




64
TABLE 11
Analysis of the Pure Fluorocarbons Irradiated
in the Cobalt 60 Gamma Ray Source Sample No. 2-B n-C4Flo
Sample Wt. 0.4782 gm.
No. days in Co60 source 38.91 Absorbed dose = 4.56 x 1021 e.V. = 152 megarads % Sample recovery as gas 100.23
Product or Vo0 on
n-hexadecane Column Wt. % G Value
CF4 1,.06 0.761
CO2 0.0562 0.0809
C2Fe 0.501 0.230
C2Fe0 0.0342 0.0141 Cs3F 0.814 0.274
CsF80eO 0.0154 0.00416
n-CzFlo 87.59 3.305
102.1 cm.3 0.0178
n-CsFia 0.871 0.192 i-C5F12 1.48 0.324 n-C6F14 0.580 0.109 2-CFaCs5F11 0.133 0.0250 3-CFsCsF11 0.870 0.163 2,3-(CF3)2C4Fe 0.187 0.0350
397.2 cm.3 0.011
n-C7F16 0.368 0.0602 2-CF3Ce6F13 0.0742 0.0121 3-CF3C6F13 1.44 0.236 3-C2F5C5Fii 0.0061 0.0010




65
TABLE 11(Continued)
Product or Vo0 on
n-hexadecane Column Wt. % G Value
2,4-(CFs)aCs5Fo10 0.129 0.0210 2,3-(CF3)a2C5Fo 0.212 0.0346 2,2,3-(CF3)3C4F7 0.005 0.0008 3,3-(CFs)2CsF1o 0.105 0.0172 CaFla8 2.72 0.393
C9F20 0.734 0.0953
Sample No. 6-A i-C4Fio Sample Wt. 0.3466 gm.
No. days in Cobalt 60 source 38.91 Absorbed dose = 3.54 x 1021 e.V. = 164 megarads % Sample recovery as gas 98.90
Product or Vo on
n-hexadecane Column Wt. % G Value
CF4 0.414 0.277
C02 0.968 1.30
C2Fe 0.0988 0.0422
C3Fe 0.673 0.211 i-C4Fo 88.32 -2.894
108.3 cm.3 0.0342
n-C5F12 0.0345 0.00707
i-C5F12 2.24 0.459
n-CeF14 0.0163 0.00284 2-CF3C5F1i 0.131 0.0228 2,3-(CF3)2C4Fa 0.851 0.149 3-CFsCeFia3 1.24 0.188
2,4-(CF3)2C5Fio 0.347 0.0527
2,3-(CFa)2CsFlo 0.0267 0.00406 2,2,3-(CF3)3C4F7 0.0797 0.0121




66
TABLE 11 (Continued)
Product or Vo on
n-hexadecane Column Wt. G Value
CsF18 3.43 0.462 CF20o 0.429 0.0518 CloF22 0.429 0.0470 CIFa4 0.242 0.0242
Sample No. 1-A n-CsFl2
Sample Wt. 0.5637 gm.
No. days in Co60 source 38.91 Absorbed dose = 5.22 x 1021 e.V. = 149 megarads % Sample recovery as gas 97.30
Product or Vo0 on
n-hexadecane Column Wt. G Value
CF4 0.783 0.578 C02 0.0754 0.111 C2F6 0.538 0.253 C3F8 0.536 0.185 n-C4Fo10 0.548 0.149 n-C5F12 87.93 -2.723 i-CsFl2 0.233 0.0525
183.8 cm.3 0.0122 210.3 cm.3 0.0140
n-C6F04 0.461 0.0886 2-CF3C5F11 0.483 0.0928 3-CF3Cs5Fi 0.714 0.137
2,3-(CF3)2C4Fe 0.0143 0.00275
372.7 cm.3 0.0045 398.2 cm.3 0.0103
n-C7F16 0.217 0.0364




67
TABLE 11 (Continued)
Product or Vo on
n-hexadecane Column Wt. % G Value
3-CF3C6F13 0.600 0.0988 3-C2FsCs5F11 0.675 0.113
2,4-(CF3)2C5Fo10 0.0137 0.00229 2,3-(CF3)2C5Fo10 0.0218 0.00371 3,3-(CF3)2CsFlo 0.0375 0.00628
CSF1a8 1.44 0.214 CSF20o 1.95 0.257 CloF22 2.72 0.328
Sample No. 4-A n-C6FI4
Sample Wt. 0.5437 gm.
No. days in Cobalt 60 source 38.91 Absorbed dose = 5.20 x 102 1 e.V. = 153 megarads % Sample recovery as gas 99.06
Product or Vo on
n-hexadecane Column Wt. G Value
CF4 0.724 0.518 CO2 0.0517 0.0740 C2F6 0.548 0.250 CsF8 0.816 0.273 n-C4F1o 0.812 0.215 n-CsFl2 0.830 0.181
i-C5F12 0.0781 0.0171
176.7 cm.3 0.00705
n-CeF14 86.37 -2.539
2-CFsCsF.11 1.03
3-CF3C5F l 0.112 0.0208




68
TABLE ll(Continued)
Product or Vo0 on
n-hexadecane Column Wt. G Value
2,3-(CF3)2C4F8 0.0273 0.00508 397.2 cm.3 0.285 n-C7F16 0.291 0.0471 2-CFsCe6FIs 0.319 0.0518 3-CFsCeFs13 1.29 0.208 3-C2FsCsFll 0.0068 0.0011 2,4-(CF3)2C5Fo10 0.0206 0.00328 2,3-(CF3)aCs5FIo0 0.0185 0.00300 2,2,3-(CF3)sC4F7 0.0096 0.00156 3,3-(CF3)2CsFlo 0.0137 0.00222 C8F18 0.989 0.142 CF2o 2.19 0.282 CloF22 1.57 0.184 CaFa4 1.235 0.131 C2Fa6 0.375 0.0370




69
TABLE 11 (Continued)
Sample No. 5-A 2-CF3CsFi Sample Wt. 0.4555 gm.
No. days in Cobalt 60 source 38.91 Absorbed dose 4.70 x102 1 e.V. = 165 megarads SSample recovery as gas 98.44
Product or Vo0 on
n-hexadecane Column Wt. % G Value
CF4 1.17 0.778
C0o2 0.0553 0.0734
C2F6 0.497 0.210 C3Fs 0.588 0.183 n-C4Flo 0.176 0.0431 i-C4F1o 0.652 0.160
108.3 cm. 0.0196
n-CsFla2 0.578 0.117 i-CsF1a 0.370 0.0750
173.6 cm.3 0.0224 201.2 cm.3 0.0107 218.5 cm.3 0.0125
n-C6F14 0.254 0.0439 2-CF3C5F11i 86.18 -2.388
2,3-(CFs)aC4F8 0.171 0.0296
397.2 cm.3 0.0170
n-C7Fi6 0.0598 0.00599 2-CFsC6F13 0.0624 0.00939 3-CF3C6Fs3 0.569 0.0857
3-C2FsCs5Fii 0.0158 0.00208
2,4'(CFS)2CsFlo 1.48 0.222




70
TABLE 11 (Continued)
Product or Vo0 on
n-hexadecane Column Wt. G Value
2,3-(CF3)2C5Fo10 0.459 0.0691
2,2,3-(CF3)3C4F7 0.0078 0.00117 3,3-(CF3)aCsFlo 0.0428 0.00644
Ce8F18 1.28 0.171
CeF20o 2.08 0.248 CloF22 0.952 0.103 CIF24 1.58 0.157 C12F26 0.660 0.0604
Sample No. 3-D 3-CFsCsF11 Sample Wt. 0.2695 gm.
No. days in Cobalt 60 source 38.91 Absorbed dose 2.79 x 1021 e.V. = 166 megarads % Sample recovery as gas 99.93
Product or Vo on
n-hexadecane Column Wt. % G Value
CF4 1.28 0.844 CO2 0.0567 0.0750 C2F6 0.0883 0.0370 CsFe 0.104 0.0321
n-C4Flo 0.0146 0.00357
i-C4Flo 0.481 0.118
108.3 cm.3 0.0244
n-CsFla 0.231 0.0466
i-C5F12 0.696 0.141
176.7 cm.3 0.0392 221.5 cm.3 0.0451




71
TABLE 11 (Continued)
Product or Vo on
n-hexadecane Column Wt. % G Value
n-C6FI4 0.216 0.0372 3-CFsCsF11 86.494 -2.326
2,3-(CFS)2C4Fe 0.292 0.0503
409.4 cm.s 0.0511
n-C7F16 0.0215 0.00322 3-CFsC6Fs3 0.144 0.0216 3-C2F5F5F11 0.231 0.0346 2,3-(CF3)2Cs5FIo0 0.464 0.0696 3,3-(CF3)2CsFo 2.48 0.373 CeF18 2.45 0.326 C9F20o 1.27 0.151 CoF22 1.57 0.169 C11Fa4 0.929 0.0920 Cla2F26 0.380 0.0346
Sample No. 26 C3F8 (reanalyzed)* Sample Wt. 0.4903 gm.
No. days in Cobalt 60 source 33.10 Absorbed dose 4.97 x 1021 e.V. = 163 megarads
Product or Vo on
n-hexadecane Column Wt. 1 G Value
CF4 1.57 1.06
CO2 0.0158 0.0213 C2F6 1.74 0.747 C2F60 0.0214 0.0050 C3F8 89.56 -3.297 n-C4F1o 1.53 0.382




72
TABLE 11 (Continued)
Product or Vo on
n-hexadecane Column Wt.% G Value
i-C4F10 1.42 0.353 n-CsFla2 0.712 0.147 i-CsF12a 1.17 0.241
n-CeFI4 0.285 0.0496 2-CF3CsFi1 0.640 0.112
3-CF3CsF11 0.0593 0.0104
322.6 cm.s 0.0202
2,3-(CF3)2C4F8 0.507 0.0890
n-C7F16 0.0512 0.00478 2-CF3C6F13 0.0406 0.00622 3-CFsC6Fs3 0.0815 0.0125 2,4-(CF3)aC5Fo0 0.241 0.0368
2,3-(CFa)aCsFlo 0.0226 0.00346 3,3-(CF3)2CsFlo 0.120 0.0184 C8Fi8 0.221 0.0300
* Sample from J. C. Mailen(-)
Sample No. 94 2,3-(CF3)2C4Fe (reanalyzed) Sample Wt. 0.4133 gm.
No. days in Cobalt 60 source 9.91 Absorbed dose 1.31 x 1021 e.V. = 50.7 megarads
Product or Vo on
n-hexadecane Column Wt. % G Value
CF4 0.821 1.78
C02 0.0318 0.138
C2F6 0.0473 0.0653
C3F6 0.381 0.386




73
TABLE 11 (Continued)
Product or Vo on
n-hexadecane Column Wt. % G Value
n-C4Flo 0.0401 0.0521 i-C4Flo 0.0298 0.0238 97.1 cm.3 0.192 108.3 cm.3 0.0185 n-CsFl2 0.0231 0.0153 i-CsFl2 0.224 0.148 176.7 cm.3 0.0095 201.2 cm.s 0.0547 n-C6F14 0.0208 0.0117 2-CFsCs5F11 0.102 0.0577 2,3-(CF3)2C4Fs 92.04 -4.434 n-C7F16e 0.025 0.0123 3-CFsCeF13 0.406 0.199 2,3-(CFs)2CsFlo 0.0649 0.0319 C8F.8 1.80 0.781 CeF20o 1.15 0.447 CloF22 1.40 0.495 C11F24 0.0793 0.0257 C12F26e 0.970 0.290
* Sample from J. C. Mailen(.)




74
TABLE 12
G Values for the Disappearance of the Starting Material
-G Starting Absorbed Dose Sample Material (Megarads)
CF4 0.997 77.2 C2F6 3.542 4o0.5
CsF8 3.297 163 n-C4FIo 3.305 152 i-C4F1o 2.894 164 n-CsFl2 2.723 149 n-C8F14 2.559 153 2-CF3C5F11 2.388 165 3-CFsCs5Fl 2.326 166
2,3-(CF3)2C4F8 4.434 50.7




V. PERFLUOROHEPTANE ISOMERS
Identification of the Perfluoroheptane Isomers
The retention volume (Vo) of n-C7F16 and the retention volume of n-CsF18 on the n-hexadecane column were estimated from Figure 1 to be 427.8 cm.3 and 760.7 cm.3 respectively.
The G values and the retention volumes on the n-hexadecane column of all radiolysis products from n-C7F16 to n-CsF18 are given in Table 13 for the parent molecules listed in Table 1 which were irradiated in the Cobalt 60 gamma ray source to produce the perfluoroheptane isomers.
Table 1 predicts that the radiolysis products of n-C4Fo0 should have two major peaks in the perfluoroheptane group corresponding to the two perfluoroheptane isomers which can be formed as primary radiolysis products assuming that the products are made by radical formation and recombination as stated earlier. This is actually the case as.seen in Table 13.
Likewise the radiolysis products of n-C5F12 should have three major peaks in the perfluoroheptane group, the radiolysis products of n-C6F14 should have three major peaks in the perfluoroheptane group, and the radiolysis products of 3-CF3C5F11 should have four major peaks in the perfluoroheptane group. This is also the case as seen in Table 13.
75




TABLE 13
G Values and Retention Volumes on the n-hexadecane Column* of All Radiolysis Products from n-C7FI6 to n-CBF1e for the Parent Molecules Listed in Table 1 G Value of Product**
Retention
Volume a cm. ) * 00 CD CD 0O .
*0 .o LC- 0 oE II II II II II w II Parent II II 0
MO OU 0 0 I 0 Molecule o n-C4Flo 0.0602 0.0121 0.0010 0.0210 0.0346 0.0008 0.0172
i-C4F1o 0.188 0.0527 0.oo o 406
n-C5Fl2 0.0364 0.113 0.00229 0.00371 0.00628
n-CeF12 0071 000518 0 0.0011 0.00328 0.0000 0.00156 0.00222
2-CF3CsF11 0.00599 0.00939 0.0857 0.00208 0 0 0.00117 0.00644
3-CF3C5F1 0.00322 03 Z 46 0.0696 0. 373 ]
* Operating conditions of n-hexadecane column given in Table 4.
** Major peaks are outlined




77
Table 1 predicts that the radiolysis products of i-C4F10o should have two major peaks in the perfluoroheptane group and that these two peaks should be different from the two perfluoroheptane product peaks found in the radiolysis products of n-C4F1o. The largest perfluoroheptane product peak found in the radiolysis products of i-C4Flo, contrary to expectation, had the same retention volume as the second major perfluoroheptane peak in the n-C4F1o radiolysis products. The i-C4F1O starting material was not completely pure, however, and this may account for the "unwanted" perfluoroheptane peak. Two major perfluoroheptane peaks in addition to the "unwanted" peak were actually found in the radiolysis products of i-C4F10. These two peaks are probably the desired perfluoroheptane isomers.
Five major perfluoroheptane peaks are predicted in the radiolysis products of 2-CF3C5F11. One of these perfluoroheptane peaks should be 2,2-(CF3)2C5Flo and the retention volume of this peak should be different from the retention volumes of all other peaks in the perfluoroheptane group. Only four major perfluoroheptane peaks are actually found in the radiolysis products of 2-CF3CsF11 and all of these peaks have reten -n volumes which duplicate retention volumes found for other perfluoroheptane peaks. It is concluded that the perfluoroheptane isomer 2,2-(CF3)2CSFo10 is probably not produced as a product in the radiolysis of 2-CF3CsF11. In fact, no fluorocarbon products with neo structures have ever been found in any of the irradiations of fluorocarbons at the University of Florida.(12)




78
By comparing Tables 1 and 13 the retention volumes
on the n-hexadecane column of eight of the nine perfluoroheptane isomers are determined. The retention volumes on the n-hexadecane column of all of the perfluoroheptane isomers except 2,2-(CF3)aCs5Flo0 are given in Table 14. Thus, all of the perfluoroheptane isomers except 2,2-(CPF3)2C5Fo10 have been identified by retention volume on the n-hexadecane column. The partition coefficients for the perfluoroheptam isomers on the n-hexadecane column are listed in Table 4.
The very good agreement between the predicted results
and the experimental results certainly supports the hypothesis of free radical reactions occuring during the radiolysis of saturated fluorocarbons.
Estimation of Normal Boiling Points
It has been shown theoretically by Purnell(6) that for members of a homologous series the logarithm of the retention volume, log Vo, versus the normal boiling point, Tb, should be a straight line. This phenomena has been experimentally verified for many different homologous
(22,23)
series.
Sullivan et al.(-) have also shown that this relationship is approximately true for the hexene and hexane isomers. The data of Desty and Whyman(-4) for hydrocarbons indicate that log Vo0 plotted against Tb for homologous series have one slope while such plots for isomers of individual saturated hydrocarbons have a different slope.




TABLE 14
Retention Volumesof the Perfluoroheptane Isomers on the n-hexadecane Column*
Perfluoroheptane
Isomer and
Retention
(0s 0 0 Volume (cm.3) 00 1
o no o aLO- a v o a o aO
o c 0 r a
00 Q0 ti co 0 O4G Mt- UCJ )CDa H 01 r>,
H 0
*Hprtn cniin ofnh*dcn column gve inTbl .
iL a t [no C0 L1 n 0 c C-- O; M C\J cO(D0 (DOC) U r(\ Pq '-( P~ ,\ Do Fr 0Cc
Hz U--I _I O f ILC) U OL(') OLf Q0 U L-- UOe--. Parent om a co- r
M MII c II l li I l III 1 II 1 1o I Molecule CD) -t 0 11 C\ U
10 10 10 IO "0 0 "0 0 -O IO
n-C4F10
i-C4Flo
n-C5F12
n-C6F14
2-CF3C5F11
3-CF3C5F11
* Operating conditions of n-hexadecane column given in Table 4.
Major peaks are outlined
Predicted isomers are crossed




8o
The experimental retention volumes and normal boiling points for perfluoroalkanes are given in Table 15 and are shown in Figure 7. The slope of log Vo versus Tb calculated for the normal saturated fluorocarbon compounds is 0.00976 while the slope calculated for the perfluorohexane isomers is 0.0907. The slope calculated for the perfluoropentane isomers is 0.0916 which is almost identical with the slope for the perfluorohexane isomers.
Normal boiling points for the normal saturated fluorocarbons were calculated from the normal boiling point of C3Fe and the slope (0.00976) of the log Vo versus Tb plot for the normal fluorocarbon compounds. The results are given in Table 16. The calculated values are almost identical with the measured normal boiling points which indicates that the normal boiling points of normal saturated fluorocarbons can be estimated very accurately using this method. The normal boiling point calculated for n-CsF18 is given in Table 17.
The normal boiling points for the perfluorohexane isomers were calculated from the normal boiling point of n-C6F14 and the slope (0.0907) of the log Vo versus Tb plot for the perfluorohexane isomers. The calculated results shown in Table 16 also agree very well with the experimental normal boiling points.
The normal boiling points of the perfluoroheptane isomers were calculated from the normal boiling point of n-C7F,6 and the slope (0.0907) of the log Vo versus Tb plot




81
TABLE 15
Retention Volumesand Normal Boiling Points
for Fluorocarbon Compounds
Retention Volume, Normal Vo0, on n-hexa- Boiling* Compound decane Column,cm.3 LogloVo Point oC. CsF8 29.6 1.47129 236.70 n-C4Flo 64.4 1.80889 271.2 n-CsFl2 129.7 2.11294 302.4 i-CsF12 145.2 2.18808 303.22 n-CeF14 238.9 2..37822 330.31 2-CF3C5F11 270.6 2.43233 330.90 2,3-(CF3s)2C4F8 348.2 2.54183 332.12 n-C7F16 427.8 2.63124 355.66 2-CF3C6Fs 469.7 2.67182 3-CF3CeFI3 487.0 2.68753 3-C2FsCsFu1 530.9 2.72501 2,4-(CF3)2C5Fo10 566.7 2.75335 2,3-(CF3)2CsF1o 593.2 2.77320 22,2,3-(CF3)sC4F7 660.6 2.81994 3,3-(CF3)2C5F1o 707.6 2.84979 n-C8F18 760.7 2.88121
*Values from T. M. Reed (3)




82
100
mI
I
n-C7F16
2,3-(CF3)2C4F8
a
2-CF3C5F11
S n-CeFI4
0
o I
i-CsFl2
Sin-C5F12 a n- CsFl
100
0
I
r n-C4Flo
0
(D
0
S
0
.H C3F8 p
0
-
0
10 I I I I I I I I 200 300oo 9o00
Normal Boiling Point oC.
Figure 7
Logarithm of Retention Volume on n-hexadecane Column
Versus Normal Boiling-Point for Fluorocarbons




TABLE 16
Calculated and Experimental Normal Boiling
Points for Fluorocarbons
Normal SaturatedFluorocarbons
Experimental Normal* Calculated Normal Compound Boiling Point oC. Boiling Point oC.
C3Fe 236.70 n-C4F1o 271.2 271.2 n-CsFl2 302.4 302.5 n-CeF14 330.31 329.67 n-C7Fe 355.66 355.61
Perfluorohexane Isomers
n-CeF14 330.31 2-CF3C5F11 330.90 3530.91 2,3-(CF3)2C4F 332.12 332.11
*Values from T. M. Reed(3)




84
TABLE 17
Estimated Normal Boiling Points for Fluorocarbons Compound Normal Boiling Point oC.
n-C7F16e 355.66 2-CF3C6F13 356.3 3-CF3C6F13 356.7 3-C2FsCs5F11 357.0 2,4-(CF3)2CsF1o 357.2 2,3-(CFs)2CsFlo 357.7 2,2,3-(CF3s)3C4F7 358.1 3,3-(CF3)a2C5sFo 358.4 n-CaF18 381.2




85
of the perfluorohexane isomers. The slope of the log Vo0 versus Tb plot for the perfluoroheptane isomers was assumed to be the same as the slope for the perfluorohexane isomers. The results are given in Table 17.
Assuming that the experimental normal boiling points are correct, the normal boiling points estimated for the perfluoroheptane isomers and for n-C8F18 are probably accurate to + 0.20C.




VI. SUMMARY
All of the identified products found in this work can be explained by radical formation and recombination. The yields of fluorocarbon products show that C-F as well as C-C bond rupture occurs during the radiolysis of saturated fluorocarbons. Perfluoromethane is the most abundant of all fluorocarbon products and characterizes the radiolysis of saturated fluorocarbons. Fluorocarbons with branched structures are more abundant in the radiolysis products than straight-chain fluorocarbons which may indicate that branched fluorocarbon are more abundant than straightchain fluorocarbon radicals during radiolysis.
This means that F atoms must be harder to remove from terminal CF3 groups than from other carbon atoms during radiolysis of saturated fluorocarbons. Perfluoromethane is the most stable to irradiation of all fluorocarbons studied. Carbon granules and carbon monoxide probably act as radical scavengers during the radiolysis of perfluoromethane mixtures and this causes extreme degradation of the otherwise stable perfluoromethane to heavier perfluoroalkanes.
No unsaturated or cyclic compounds were found in the radiolysis products of the saturated fluorocarbons. Three products that contain oxygen, C2F60, C3FeO, and (C2Fs)20, were identified and other unknown products are 86




87
probably oxygen-containing fluorocarbons. The oxygencontaining products are probably formed by radical recombination at the sample tube wall.
No corrosion was found in the sample tubes and material balances show that essentially no fluorine is lost during radiolysis of saturated fluorocarbons. Molecular fluorine was not found in any of the radiolysis products.
Abstraction reactions and disproportionation reactions which occur in the radiolysis of saturated hydrocarbons probably do not occur in the radiolysis of saturated fluorocarbons. Saturated fluorocarbons are more stable to irradiation than the analogous hydrocarbons.
Retention volumes on the n-hexadecane column of eight of the nine perfluoroheptane isomers were determined and the normal boiling points of these compounds were estimated. The perfluoroheptane isomer, 2,2-(CF3)2CsFo0, was not found in any radiolysis products and no other fluorocarbon products with neo structures were found.




APPENDICES




APPENDIX 1
Dosimetry
The absorbed dose rate distribution in tube II of the College of Engineering Cobalt 60 gamma ray source at the University of Florida was measured by benzene-water dosimetry following Johnson and Martin.(26) The phenol yield in neutral benzene-water solutions is a linear function of the total gamma dose up to a total dose of 70,000-80,000 rep; over the range 100-7,000 rep/min. it is independent of dose rate, benzene concentration (above 0.007M) and product concentration. The phenol concentration is determined from
O
the ultraviolet optical densities measured at 2,900A of a neutral and alkaline sample of the irradiated benzene-water solution. The neutral sample is prepared by diluting the irradiated benzene-water solution with distilled water and the alkaline sample by diluting the irradiated benzenewater solution with 0.06N sodium hydroxide solution.
The optical densities were determined with a Beckman
DK-2 Spectrophotometer at least 10 minutes after the alkaline sample was prepared. Under these conditions the phenol yield is 2.35 molecules/100 e.V.
Distilled water saturated with benzene at room temperature (v 0.010M) was sealed in polyethylene bottles 3.20 cm.
89




90
high and approximately 2.5 cm. in diameter. The bottles had void spaces of 0.29 cm. at the top and 0.33 cm. at the bottom and contained approximately 5 ml. of solution. The bottles were stacked in the center of the sample space and irradiations of from 20 minutes in the high flux zone up to 60 minutes in the low flux zone were done in order to obtain as high a phenol concentration as possible in the dosimeter solution but still remain within the linear function of phenol concentration versus total gamma ray dose.
Two independent runs were made and the results are given in Table 18 and the absorbed dose rates are plotted versus distance from the bottom of the sample space in Figure 8.
The absorbed dose rates determined using the polyethylene bottles are considered average values since the dosimeter solution occupied a large portion of the cross sectional area of the sample zone.
Because the benzene-water dosimeter solution and the radiolysis products are not corrosive to aluminum, (27) aluminum tubing identical to the tubing used for holding the fluorocarbon samples was "coiled" as shown in Figure 9 and the vertical and radial absorbed dose rate distribution in tube 11 was measured.
The results are given in Table 18 and Figure 8.
Because each "coiled" aluminum tube contained less than
2 ml. of dosimeter solution, it was necessary to dilute these samples in order to have enough material to analyze in the spectrophotometer.




Full Text

PAGE 1

EFFECTS OF RADIATION ON MIXTURES CONTAINING FLUOROCARBONS AND THE IDENTIFICATION OF THE PERFLUOROHEPTANE ISOMERS By WILLIAM CREWS ASKEW A DISSERT ATION 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 April 1966

PAGE 2

ACKNOWLEDGMENT. This work was performed under contract to the United States Atomic Energy Commission. The financial support of this agency is greatly appreciated. The author is especially indebted to Dr. T.M. Reed III, chairman of the supervisory committee, whose patience, understanding, and direction made this work possible. The author is also grateful to Dr M. B. Fallgatter and Dr. R. J. Hanrahan for the mass spectra identifications and to Dr. J. A. Wethington, Jr. for help in the spectrophotometric analysis. Appreciation is also extended to the other members of the supervisory comrnittee, Dr. Mack Tyner and Dr. T. O. Moore, for their kind support. The encouragement and patience of the .author's wife was invaluable. ii

PAGE 3

I. II. TABLE OF CONTENTS INTRODUCTION ... . . . .. Pages 1 Mixtures Containing Fluorocarbons. .. 1 Perfluoroheptane Isomers. 2 EXPERIMENTAL PROCEDURES ..... 7 Starting Materials. . . . .7 Sample Preparation . . . 8 Irradiations. . . . .11 Analysis. . . . .. .. .12 III. LOW INTENSITY TRAINING REACTOR IRRADIATIONS OF MIXTURES CONTAINING FLUOROCARBONS ........... 20 CF4+C2F6 Mixtures. . . . .... 20 Mixtures Containing CF4 and Nonfluorocarbon Components. . . . . .29 IV. COBALT 60 GAMMA IRRADIATIONS OF MIXTURES CONTAINING FLUOROCARBONS AND OF PURE FLUOROCARBONS ........ 48 CF4+C2F6 Mixtures ................. 48 Mixtures Containing CF4 and Nonfluorocarbon Components. .. . . .52 Pure Fluorocarbon Samples ............ 53 V. PERFLUOROHEPTANE ISOMERS. ...... 75 Identification of the Perfluoroheptane Isomers .. 75 Estimation of Normal Boiling Points ........ 78 VI. SUMMARy ....................... 86 APPENDICES I. Dosimetry. . . .89 II. Response of Thermal Conductivity Detector Cells ... 101 iii

PAGE 4

Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 LIST OF TABLES. Page Parent Perfluorocarbon Molecules and the Perfluoroheptane Isomers They Should Yield as Primary Products Upon Irradiation. 5 Impurities in Starting Materials. .. 9 Description of Silica Gel Column No. 79 .14 Description of n-hexadecane Column. .15 Description of Kel-F No. 90 Grease Column No.5. .. ............... 16 Analysis of the CF4+C2Fs Mixtures Irradiated for One Week in the Low Intensity Training Reac tor. . . . .. . .34 Analysis of the Mixtures Containing CF4 and Nonfluorocarbon Components Irradiated for One Week in the Low Intensity Training Reactor .. 40 Analysis of the Mixtures Containing CF4 and Nonfluorocarbon Components Irradiated for Four Weeks in the Low Intensity Training Reactor .................... 44 Analysis of the CF4+C2F s Mixtures Irradiated in the Cobalt 60 Gamma Ray Source .. Analysis of the Mixtures Containing CF4 and Nonfluorocarbon Components in the Cobalt 60 Gamma Ray Source .......... 62 Analysis of the Pure Fluorocarbons Irradiated in the Cobalt 60 Gamma Ray Source .... 64 G Values for the Disappearance of the Starting Material ............. G Values and Retention Volumes on the n-hexadecane Column of All Radiolysis Products from n-C7F1S to n-CSF1S for the Parent Molecules Listed in Table 1 ..... iv.

PAGE 5

Table 14 Table 15 Table 16 Table 17 Table 18 Table 19 Table 20 Table 21 Retention Volumesof the Perfluoroheptane Isomers on the n-hexadecane Column. .. 79 Retention Volumes and Normal Boiling Points for Fluorocarbon Compounds Calculated and Experimental Normal Boiling 8 1 Points for Fluorocarbons. .. .... 83 Estimated Normal Boiling Points for Fluoro-carbons. . . . .. 84 Absorbed Dose Rates Measured in Dosimeter Solution . . .. ..... 9 1 Ratio of Atomic Number to Atomic Weight for Various Substances .............. 96 Area Response of Detector Cell and Correction Factors for Calculating Weight Fractions -H2 Carrier Gas. . .. .... 103 Area Response of Detector Cell and Correc tion Factors for Calculating Weight Fractions -He Carrier Gas . .. ... 105 v

PAGE 6

Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 LIST OF FIGURES Logarithm of Retention Volume Versus Number of Carbon Atoms in Normal Saturated FluoroPage carbons. . . . . .18 Weight Percent of CF4 and C2F6 in the Products from the Low Intensity Training Reactor Irradiations of the CF4 +C2F6 Mixtures ..... 22 Radiolysis Products with Straight-Chain Structures from the Low Intensity Training Reactor Irradiations of the CF4 +C2F 6 Mixtures. .. ........... 23 RadiolysisProducts with Branched Structures from the Low Intensity Training Reactor Irradia tions of the CF 4+C2F6 Mixtures .. 24 G Value of Primary Products Versus Weight Bercent of C2F6 in Original Sample Mixture .. 49 G Value of Products Containing Oxygen Versus Weight Percent of C2F6in Original Sample Mixture ................ 50 Logarithm of Retention Volume on n-hexadecane Column Versus Normal Boiling Point for Fluorocarbons. . . . .; 82 Absorbed Dose Rates Versus Distance from Bottom of Sample Space ............ 92 Aluminum Dosimeter Tubes . . 93 Area Response of Thermistor Detector Cell -H2 Carrier Gas. . . . 106 Area Response of Thermistor Detector Cell -He Carrier Gas . . . 107 vi

PAGE 7

DEFINITIONS AND SYlfffiOLS Megarad Rad Rep Erg e.V. Roentgen, absorbed in water G value K vii 106 rad 100 erg./gm. 93 erg./gm. 6.242 x lOll e.V. Electron-volt 0.975 rad Number of molecules produced per 100 e.V. absorbed. used to distinguish reactions brought about by the absorption of ionizing radiation. Partition coefficient or distribution ratio, ratio of concentration of solute in gas phase to concentration of solute in liquid phase.

PAGE 8

ABSTRACT Mixtures containing fluorocarbons and some selected pure fluorocarbons were sealed in high purity aluminum tubing and were irradiated by Cobalt 60 gamma rays for absorbed doses of approximately 150 megarads. The mixtures containing fluorocarbons were also irradiated for periods of one week and four weeks in the Low Intensity Training Reactor at Oak Ridge, Tennessee, at approximately 2 x 1013 neutrons per centimeter squared per second. The mixtures containing fluorocarbons were: CF4+C2F S CF4 + aluminum powder, CF4+CO, and CF4 + carbon granules. The n-C5Fl2, n-CSFl4, 2-CF 3 C 5Fll, 3-CF3C 5Fll and 2,3-(CF3)2C4FS. Analysis 01' the irradiated samples was by gas chromatography and identification 01' some radiolysisproducts was by mass spectrometry. All of the identified products found in this work can be explained by radical formation and recombination. The yields of fluorocarbon products show that C-F as well as C-C bond rupture occurs during the radiolysis of saturated .. fluorocarbons. Perfluoromethane is the most abundant of all fluorocarbon products and characterizes the radiolysis of saturated fluorocarbons. The abundance of branched fluorocarbon products indicates that F atoms are more difficult to remove from terminal CF3 groups than from viii

PAGE 9

other carbon atoms in saturated fluorocarbon molecules. Perfluoromethane is the most stable to irradiation of all rluorocarbons studied. Carbon granules and CO probably act as radical scavengers during radiolysis of mixtures containing CF4 and this causes extreme degradation o f the otherwise stable CF4 to heavier perfluoroalkanes. T h e presence of aluminum powder during irradiation causes little change in the radiolysis products of CF4 No unsaturated or cyclic products were found in the radiolysis products of the saturated fluorocarbons. Three products that contain oxygen, C2FeO, C3FeO, and (C2F5)20, were identified and other unknown products are probably OXYgen-containing f luorocarbons. The oxygen-containin g products are probably formed by radical recombination a t the sample tube wall. No corrosion was found in t h e sam ple tubes and material balances show that essentially no fluorine is lost during radiolysis 01' saturated fluorocarbons. Molecular fluorine was not found in any of the radiolysis products. Abstraction reactions, disproportionation reactions and molecular expulsion of F2 probably. do not occur in the radiolysis of saturated fluorocarbons. G values (molecules/IOO e.V. ) calculated for the disappearance of tne starting material are 2 to 3 and show that saturated fluorocarbons are more stable to irradiation than saturated hydrocarbons. -G for CF4 is approximately 1.' Retention volumes of eight of the nine perfluoroheptane isomers were determined and. the normal boiling pOints of ix I

PAGE 10

these compounds were estimated The perfluoroheptane isomer, 2,2-(CF3)2C5FlO, was not found in any of the radiolysis products and no other fluorocarbon products with neo structures were found. x

PAGE 11

I. INTRODUCTION Mixtures Containing Fluorocarbons Except for the recent work of and Kevan and there hav e been few comprehensive results published on the radiolysis of saturated fluorocarbons. Other h a ve been published on the radiolysis of fluorocarbons but the authors ,were hampered by the use of impure starting materials and relatively poor analytical methods. Recent published results on the radiolysis of cyclic fluorocarbons are those of Fallgatter and. Hanrahan(2) and MacKenzie et al. (2) The results of Rexroa d and GOrdY() and of on the radiolysis of polyfluoroethylene (Teflon) are also of int'erest. In this work mixtures containing fluorocarbons were prepared and irradiated in the College of Engineering Cobalt 60. gamma ray source at the University of Florida and in the Low Intensity Training Reactor at Oak Ridge, Tennessee. The mixtures were irradiated for a dose of approximately 100 megarads in the Cobalt 60 gamma ray source and for periods of one week and fbur weeks in the Low Intensity Training Reactor at approximately 2 x 1013 neutrons per square centimeter per second. 1

PAGE 12

were: 2 The fluorocarbon mixtures by approximate mole percent lO%CF4 25%CF4 50%CF4 75%CF4 90%CF4 90%C2Fs 75%C2F6 50%C2F6 25%C2F6 lO%C2F6 The mixtures containing fluorocarbons were: CF4 + aluminum powder CF4 + carbon granules CF4 + CO These mixtures will allow surface effects, compos .ition effects, and heterogeneous phase effects in radiolysis of fluorocarbons to be studied. Perfluoroheptane Isomers Although the radiolysis mechanisms for fluorocarbon compounds are not completely understood,there is that free radibal reactions do otcur during radiolysis of fluorocarbons. Therefore, by treating the initial products obtained by irradiating a saturated fluorocarbon compound as 1) fluorocarbon molecules smaller than the parent molecule resulting from recombination of small fluorocarbon radicals, and as 2) fluorocarbon molecules larger than the parent molecule formed by recombination of large fluorocarbon radicals produced during irradiation, it should be possible

PAGE 13

3 to synthesize larger molecular weight fluorocarbons from sma ller molecular weight parent compounds Radiolysis products with boiling points greater than the parent perfluoro-carbon compound have in fact been observed by Simons and in the irradiation of and reports that the major products from irradiating were h igher boiling than the starting material. Therefore using perfluoronormalpentane as an example it should undergo the following reactions t o give perfl u oro-heptane isomers. Rupture to Produc e Radicals CF3 + n C4FS --"./V'v---C2F S + n-C3 F 7 F + F + F + Recombination to Yield Perfluoroheptane Isomers n-C4Fs + n -C3F 7 n -CSFll + C2F S n C 7 F l S 2 C sFll + C2Fs -3 -CF3 C sFl3 + C2F S I t can be seen from the recombination of radical s that three perfluoroheptane isomers should be formed as initial irradiation products from irradiating perfluoronormalpentane. The amount of each isomer produced wi l l depend up o n steari c fac t ors, radical diffusion, "cage effec ts, etc. I t shoul d also be noted that these three perfluo ro-heptane isomers are the onl y possible perfluoroheptane

PAGE 14

4 isomers that can be formed as primary recombination products of irradiating perfluoronormalpentane. Thus, it should be possible to synthesize, by irradiation, all nine of the perfluoroheptane isomers from certain lower molecular weight fluorocarbons. Table 1 lists the necessary parent molecules and the perfluoroheptane isomers they should produce, which are required to synthesize all nine of the perfluoroheptane isomers. has resolved the perfluorohexane isomers by gas chrDmatography and it should be possible to resolve all of the perfluoroheptane isomers, so that a chromatogram of the primary irradiation products of perfluoronormalbutane should have two peaks in the perfluoroheptane group. One peak should be the n-C7F1S and the other peak should be the 3-CF3C sF13. Similarly for perfluoronormalpentane there should three perfluoroheptane peaks: n-C7F1S, 3-CF3CsF13, By comparing these two chromatograms the 3-C2F5C5Fll peak can be identified In the s ame manner structures may be assigned to all perfluoroheptane isomers appearing on of radiolysis products from the parent molecules given in Table 1. All of these parent molecules will give perfluorooctanes, etc. as primary produbts, as well as the desired perfluoroheptane products. Unsaturated products are not expected in the radiolysis. products(g) and if they. are produced they will probably be

PAGE 15

TABLE 1 Parent Perfluorocarbon Molecules and the Perfluoroheptane Isomers They Should Yield as Primary Products Upon Irradiation. Parent Molecule (I) M IIi (() 0 (I) 0 I C\J Perfluoroheptane M (I) M M IIi IIi If) (() 0 0 If) (I) IIi IIi C\f 0 0 I I j'(\ j'(\ Isomer 0 M IIi If) 0 C\f ..--.. (I) 0 I ;::j-"' C\J t--0 IIi 0 M .j< M IIi 0 IIi If) (I) If) 0 ..--.. 0 C\f (I) C\f ..--.. IIi ..--.. (I) 0 (I) IIi IIi 0 I o j'(\ I "' I j'(\ C\J j'(\ "' "' "' C\J C\J j'(\ \ \ .1). rI' PGI iJDJ 0 aJ C1D 0 ......, J (\jJ "" (\II

PAGE 16

6 in very minute quantities. None have been found in any of the work at the University of Florida. All of the parent molecules listed in Table 1 have been irradiated in the College of Engineering Cobalt 60 gamma ray source for approximately six weeks and for the total dose absorbed it is hoped that the conversion of parent compounds to products will be small enough so the major radiolysis products will be primary products.

PAGE 17

II. EXPERIMENTAL PROCEDURES Starting Materials Perfluoromethane (CF 4 ) was obtained from the Matheson Company Inc. The gas was chromatographically pure and was used without further purification. The perfluoroethane (C2F 6 ) was obtained as an impure sample from the Minnesota Mining and Manufacturing Company. It was purified on a n-hexadecane column (1:Q) by gas tography. Cu) Perfluoronormalbutane (n-C4 FJ.o) was separated by gas chromatOgraphyC1Q,1:.1J from a mixture of low molecular weight material provided by the Minnesota Mining and Manufacturing Company. The perfluoroisobutane (i-C4 FJ.o) was prepared from perfluoroisobutene by fluorination with Cobalt trifluoride.(1:) The perfluoroisobutene was obtained as a 'product of the pyrolysis of perfluoropropene from Peninsular Chemical Research Inc. The perfluoroisobutane was purified by gas chromatography. (1:) J. H. Simons prepared the perfluoronormalpentane (n-C5 FJ.2) by the electrolysis of pyridine in anhydrous HF. It was purified by gas chromatography.(25) Perfluoronormalhexane prepared by Dresdner et was purified by gas 7

PAGE 18

8 Perfluoro-2-methylpentane (2-CF3C5Fll) and perfluoro-3-methylpentane (3-CF3C5Fll) were prepared by J. A. These materials were purified by gas Table 2 lists the impurities detected in the starting materials. Sample Preparation Air was removed from the starting compounds by alternately thawing, freezing and pumping on the compounds in a vacuum system until no residual pressure remained over the condensed compounds. Water was removed by distillation in vacuum through a magnesium perchlorate leg and carbon dioxide was removed by distillation through an ascarite leg. The compounds were then stored, as gas or liquid, in storage bulbs on the vacuum system until needed. Sample sizes of approximately 1/4 cc. were measured by observing what pressure reading on the manometer of the vacuum system corresponded to a condensed sample size of 1/4 cc. Sample tubes approximately 30 cm. long were prepared from 1/8 inch o.d. aluminum versatube of 0.025 inch wall thickness from the Wolverine Tube Division of Calumet and Hecla, Inc. After one end of each sample tube was sealed by Heliarc welding under argon, the tubes were tested for leaks at 1000 psig, washed .out with benzene, dried under vacuum, and weighed.

PAGE 19

9 TABLE 2 Impurities in Starting Materials Starting Material CF4 C2F6 n-C4Fl.o i-C4 Fl.o n-C5Fl.2 n-C6Fl.4 2-CF3C5Fl.l. 3-CF3C5Fl.l. Mole % Impurity None detected None detected 0.18% C02 less than 1% i-C4Fs None detected None detected 0.24% C6Fl.2 None detected

PAGE 20

10 When canning a sample for irradiation a cleaned sample tube was connected by Flex fittings and vacuum hose to the vacuum system where it was baked out under vacuum with a cool torch flame. The storage bulb containing the starting compound was then opened and the desired amount of sample, as indicated by the system pressure, was bled into the vacuum system from the storage bulb. The entire quantity of sample was condensed into the aluminum sample tube by liquid nitrogen and the tube was pinched shut and Heliarc welded. After removal from the vacuum system the canned sample tube was weighed. Several days later the sample tube was weighed again and discarded if it had lost weight. The CF4+C2F e mixtures were prepared by distilling the desired quantities of CF4 and C2Fe large storage bulbs on the vacuum system where the mixtures were allowed to mix for at least 24 hours to insure uniformity. A reference sample of each mixture was sealed under vacuum in a glass reference sample tube to be analyzed when the irradiated samples were analyzed. When preparing the CF4 + carbon granules and CF4 + aluminum powder samples, the solid material was packed into the cleaned sample tubes before the tubes were connected to the vacuum system. The tubes were then baked out, filled with CF4 and sealed as before. The height of solid material in the sample tubes was determined by the weight of solid material packed into the tubes.

PAGE 21

11 Because of the high vapor pressure of CO at liquid nitrogen temperature (48 2 mm. Hg. at -196C.), the CF4+CO samples were prepared in a different manner from the other samples. After the cleaned sample tube was connected to the vacuum system and baked out under vacuum the desired amount of CF4 was distilled into the sample tube after which the tube was isolated from the rest of the vacuum system by closing a stopcock. Carbon monoxide was then bled into the vacuum system from a pressurized cylinder until a desired reading on the manometer of the vacuum system was observed (approximately 600 mm. Hg.). The stopcock to the sample tube was then opened and the CO was allowed to condense on top of the CF4 in the sample tube until a certain lowering of the manometer reading was observed. The sample tube was pinched shut and sealed as before. Irradiations The mixtures containing fluorocarbons were irradiated for approximately six weeks in tube no. 11 of the College of Engineering Cobalt 60 gamma ray source at the University of. Florida and for periods of approximately one week and four weeks in the Low Intensity Training Reactor at Oak Ridge, Tennessee. The pure fluorocarbon samples were not irradiated in the Low Intensity Training Reactor but were irradiated for approximately six weeks in tube no. 11 of the Cobalt 60 gamma ray source.

PAGE 22

12 The sample tubes of mixtures containing fluorocarbons were randomly placed in the sample space of the Cobalt 60 gamma ray source and the calculated absorbed doses in these samples are not, as accurate as the absorbed doses calculated for the pure fluorocarbon samples. The latter sample tubes were positioned and retained against the walls of the irradiation basket which permitted more accuracy in calculat ing the absorbed doses for these samples. (See Appendix L). Absorbed doses could not be calculated for the samples irradiated in the Low Intensity Training Reactor, however, the flux to which the samples were exposed was approximately 2 x neutrons/cm.2 sec. The irradiation temperature in the Low Intensity Training Reactor was approximately 100C. while-the irradiation temperature in the Cobalt 60 gamma ray source was room temperature. Analysis The gas chromatographic analysis of all samples was done with a Perkin Elmer Model 154 Vapor Fractometer. The recording instrument was a one MV full scale range Honeywell strip chart recorder. The peak areas were measured by a Perkin Elmer Model 194 B Printing Integrator which gave 6000 integrator counts per minute for full scale deflection on the recorder. The response .of the thermistor detector cell was calibrated (Appendix 2) so that the peak areas could be converted to weight percentages.

PAGE 23

13 A temperature-programmed one meter silica gei column described in Table 3 was used to resolve compounds containing four carbon atoms or less. A 50-ft. n-hexadecane column described in Table 4 was used to resolve compounds which contained from three to seven carbon atoms, and a two meter Kel-F grease column described in. Table 5 was used to resolve compounds which contained more than seven carbon atoms into molecular weight classes. The aluminum tubes containing the irradiated samples were opened into a vacuum system adjacent to the fracto-meter where the samples were stored in storage bulbs until analyzed. After the samples had remained in the storage bulbs for at least 24 hours to insure uniformity, they were introduced into the fractometer by a gas sampling value. The average size sample taken for. analysis was approximately 0.02 moles.Since the chromatographic system is able to detect less than 2.0 x 10-7 moles of solute in the carrier gas stream, mole percentages as low as 0.001 can be detected with the analytical system used. Identification of all saturated fluorocarbon products up through the perfluorohexane isomers was by the appearance time of known standards. The identification of the perfluoro-heptane isomers is fully discussed in Chapter V. All isomers of each saturated fluorocarbon compound were found to appear chromatographically between the normal isomer of the compound and the normal isomer of the next higher molecular weight saturated fluorocarbon compound.

PAGE 24

14 TABLE 3 of Silica Gel Column No. 79 Length = 1 meter of stainless steel tubing i. d. = 0.180 inches packing = 30/60 mesh silica gel from Wilkens Instrument and Research, Inc. Column 'wrapped with 34 feet of asbestos covered nichrome V strand size 18 resistance wire, 0.41 ohms/ft. Temperature Program Time (Min. ) 0 6 26 34 52 76 Amperage Setting on Resistance Wire (amps) 0 1.00 1.50 1.75 2.50 3.00 Pi = 28.7 psia,' Po = 14.7 psia. Helium gas flow rate at column exit 42.0 cm. 3/min. at 24c. Compound Appearance Time (Min. ) air 1.7 CF4 4.3 CO2 15.0 16.4 C2F6 O 29.2 C3F 6 40.5 C3FSO 47.7 peak 1 56.3 total 58.9 (C2F5)20 61.6

PAGE 25

15 TABLE 4 Description of n-hexadecane Column Length = 50 feet i.d. = 0.197 inches Temperature = 44c. Pi = 34.7 psia Po = 14.7 psia n-C1 6H34 Vol. = 63.7 cm. 3 on 30 / 80 mesh Chromasorb P Helium gas flow rate at column exit = 18.1 cm.3/min. at 24c. Appearance Retention Partition Time Volume Vo Coefficient,K C o mpound (Min. ) ( cm. 3) air and CF4 18.7 C2F6 19.7 10.2 6.245 C2F6 O 20.5 18.4 3.462 C3F S 21.6 29.6 2.152 C3FsO 21.6 29.6 2.152 (C2 FS)20 23.1 45.0 1.416 CO2 24.8 62.3 1.022 n-C 4F1O 25.0 64.4 0.989 i-C4F1O 26.0 74.6 0.854 31.4 0.491 n-C 5 F12 129.7 i-C5F 12 33.8 154.2 0.413 n-C 6 F14 42.1 238.9 0.267 2-CF 3 C sFll 45.2 270.6 0.235 3-CF3C 5Fll 47.5 294.1 0.217 2,3-(CF 3)2C 4 F S .52.'8 348.2 0.183 n-C7F16 60.6 427.8 0.149 2-CF3C 6 F13 64.7 4 69.7 0.136 3-CF3C 6 F13 ,66.4 487.0 0.131 3-C2FsC5 Fll 70.7 530.9 0.120 2,4-(CF3)2C5Fl0 74.2 566.7 0.112 2,3-(CF3)2C5F 1 O 76,8 593.2 0.107 2,2,3-(CF3)3C4 F7 83.4' 660.6 0.0964 3,3-(CF3)2C5F1O 88.0 707.6 0.0900 n-CSF1S 93.2 760.7 0.0837

PAGE 26

16 TABLE 5 Description of Kel-F No. 90 Grease Column No.5 Length = 2 meters 0.25 inch copper tubing Temperature = 83C. P. = 15.0 psia Po = 14.7 psia 1.20 gm. acid washed 65/100 mesh Celite 545 (Johns-Manville) m1. of solvent per meter of column at RT= 3.0 Helium gas flow rate. at column exit = 27.8 cm.3/min. at 24c. Compound Air nC 4FJ.o n-C5FJ.2 n-C 6FJ.4 n-C7FJ.6 n-CsFJ.s n-CSF20 n-CJ.OF22 n-CJ.J.F24 n-:-CJ.2F 26 Appearance (Min. ) 1.2 1.5 1.9 2.6 4.3 7.0 11.4 19.3 34.5 55.0 Time Retention Volume, em. 3 (Vo ) 8.3 19.4 38.9 86.1 161.2 283.5 503.1 925.7 1495.6

PAGE 27

17 Since the linear relationship of the logarithm of the retention time of normal saturated hydrocarbon compounds versus the number of carbon atoms in the compound (ll, .. ..:r) seems to apply equally well to the normal saturated fluoro-carbon compounds under the experimental conditions used, the appearance times of the normal saturated fluorocarbon compounds above perfluoronormalhexane were estimated by extrapolation of such plots. (See Figure lJ Products larger than the perfluoroheptane isomers were grouped into molecular weight classes by the above method. Three products which appeared chromatographically on the silica gel column in a regular pattern following the corresponding perfluoroalkane of the same number of carbon atoms were identified asperfluoroethers. The product perfluorodimethylether (C2F60) was trapped out and identified from its mass Because ofa small sample size the mass spectra of the perfluoromethylethylether (C3FsO) peak was inconclusive; it did show, however, that this com. pound contained oxygen. A known sample of perfluorodiethylether (C2F5)20 was added to the irradiation products and the peak area of the product identified as (C2F5)20 was increased. C02 was a product in all irradiations but it could not be quantitatively reported in the CF4+C2F6 mixtures because the C02 peak was obscured by the much larger C2F6 peak;

PAGE 28

m 8 0 C 8 rl 0 0 m C ro 0 m ro m I C C 0 m 8 rl 0 C 0 rl C m m 10 3 4 5 6 7 8 Number of Carbon Atoms in Normal Saturated Fluorocarbons Figure 1 9 Logarithm'of Retention Volume Versus Number of Carbon Atoms in Normal Saturated Fluorocarbons

PAGE 29

19 The unknown peaks which appeared on the chromatograms are identified by retention time or retention volume only. These products are probably fluorocarbon compounds which contain oxygen as will be discussed later. The response of the detector to the unknown peaks was assumed to be the same as the response of the detector to the nearest known peak. No evidence of unsaturated .or cyclic fluorocarbon products was found. Standard reference compounds for unsaturation were perfluoroethylene (C2F4), perfluoropropene (n-C3F6), perfluoroisobutene (i-C4Fs), cis and trans perfluorobutene-2 (cis-C4Fs, trans-C4Fs), perfluoro-2-methylpentene-2 and cis and trans perfluoro-2-methyl-pentene-3. Standard reference cyclic compounds were: perfluorocyclopentane (cyclo-C3F6), perfluorocyclobutane (dyclo-C4Fs), perfluorocyclopentane (cyclo-CsFl0), and perfluorocyclohexane (cyclo-C6F12). Weight percentages reported in the analyses of the irradiated samples are estimated to be correct to better than + 1 percent of the value G values reported for the pure fluorocarbon samples are estimated to be accurate to at least + 5 percent, and G values reported for the mixtures containing fluorocarbons are accurate to at least + ,10 percent.

PAGE 30

III. LOW INTENSITY TRAINING R EACTOR IRRADIATIONS OF MIXTURES CONTAINING FLUOROCARBON S Weight percentages of the radiolysis prod u cts o f the CF4+C2F6 mixtures are given in Table 6 and the weight percentages of products are plotted versus weight percent of C2F6 in the original mixtures in Figures 2,3, and 4. All of the identified products of the radiolysis of the CF4 +C2F 6 mixtures can be by radical formation and recombination. The primary radicals formed upon irradia-tion whether by ionization and subsequent neutralization or by decomposition of excited molecules will be F, CF3, a nd C2F5 radicals; _the concentration of each radical will depend upon the amount of C2F6 in the -original mixture. CF4 C F 3 + F (1 ) C2F6 --AJ\I'--> 2CF3 (2 ) C2F6 C2F5 + F (3) Bibby and Carter (1..12.) give the C2F5F bond strength as 5.50 e.V./molecule and the CF3-CF 3 bond strength as 6.05 e.V./molecule. Dacey and Hodgins (Q) estimate that the strength of the first C-F bond in CF4 is more than 6.68 e.V./molecule. Therefore it seem s that reactions (2) and (3) must be somewhat favored over r eaction 20

PAGE 31

21 (1) and that the concentration of CF3 radicals will always be larger than the concentration of C2F5 radicals during the radiolysis of the CF4+C2Fe mixtures. Fallgatter and Hanrahan(2)have pointed out that implications that the radiation chemistry of fluoro-carbons is centered in C-C bonds should be avoided. The radiolysis products found in this work supports their conclusion. All of the identified products can be described as arising from recombination of radicals. Recombination of the primary radicals will produce C3Fs and n-C 4 F10 as primary radiolysis products. The starting materials, CF4 and C2Fe also be primary products. CF3 + F --->CF4 (4) 2CF3 C2F e (5) C2F5 + F C2F e ( 6) CF3 + C2F 5 -)00 C3Fs (7 ) 2C2F 5 n-C 4 F10 (8) The primary radiolysis products C3Fs and n-C 4 F10 will also undergo radiolysis and produce secondary radicals which will recombine with primary as well as with other secondary radicals to produce all of the observed secondary products. C3FS n-C3 F7 + F ( 9 ) --I\IV'---ti-C3F7 '+ F (10) C2F5 + CF3 (11) n-C4FS + F (12) l-CF3 C3Fe + F (13) --AM-+ n-C3 F7 + CF3 (14) 2C2F5 ( 15)

PAGE 32

100 rJ) oj-) C) ::::> '"d 0 H p.... oJ '"d ::::> 0 0.. s 0 0 G-i o .p G) C) H G) p.... 80 60 40 o / 22 / Original // Amount of--" C2F6 / / / / / / / / / / / / / / / / / / / / / / / / / Original of CF4 " 100% CF4 100% C2F6 Weight Percent in Original Sample Mixture Figure 2 Weight Percent of CF4 and C2F6 in the Products from the Low Intensity Training Reactor Irradiations of the CF4+C2F6 Mixtures

PAGE 33

[f.l 0 ::i 'd 0 H P-< c: rI 'd c: ::i 0 0. S 0 0
PAGE 34

rJl +.:> C) ::s '0 0 H P-l c:: r! '0 c:: ::s 0 p.. 8 0 0 <+-i 0 +.:> c:: Q) C) H Q) P-l +.:> ..c: QD .r! Q) 10 9 8 7 6 5 4 3 2 1 24 Weight Percent in Original. Sample Mixture Figure 4 Radiolysis Products with Branched Structures from the Low Intensity Training Reactor Irradiations of the CF4+C2Fe Mixtures

PAGE 35

25 All of the secondary products can be explained by radical recombination. (16) All products larger than CSF1S can be explained by applying the above mechanism to the secondary products. No unsaturated fluorocarbons were found in the radiolYpis products of, the CF4 +C2F6 mixtures so disproportionation reactions, abstraction reactions, and molecular expulsion of F2 probably do not occur in the radiolysis of saturated fluorocarbons Figure 2 shows that as the weight percent of C2F6 in the original mixture is increased the amount of CF4 in the products is increased by up to five times the amount of CF4 in the original CF4 +C2F6 mixture. This indicates quite clearly that reaction (4) is very significant during radiolysis since this is the only method by which CF4 can be made. Figure 2 also shows the great stability of CF4 to irradiation and this is not surprising since the C-F bond strength in CF4 is greater than in any other fluorocarbon. The stability of CF4 to irradiation is also illustrated in the Cobalt 60 irradiations of pure fluoro-carbons. (See Table 12j As the weight percent of C2F6 in the original mixtures is increased the initial concentration of C2F5 radicals during radiolysis will be and the weight percentages o f all products greater than C2F6should be increased also; i.e., (7) and (8) will become more important. This

PAGE 36

26 is illustrated in Figures 3 and 4. It should be noted, however, that the weight percentages of C3Fs and in the radiolysis products reach maximum values and that these primary products are then used up by undergoing radiolysis themselves. / ;' Thus for a one week irradiation period in the Low Intensity Training Reactor the optimum original composition of a CF4+C2FS mixture should be approximately seventy weight percent C2Fs in order to make the greatest amount of C3Fs as product. If the period of irradiation is' changed the maximum in the C3Fs curve will probably be changed also. It can be seen from Figure 3 that all of the straightchain fluorocarbon products from the one week Low Intensity Training Reactor irradiation of the CF4 +C2F s mixtures reach a maximum percentage or seem to be reaching a maximum percentage as the weight percent of C2FS in the original sample mixture is increased, while from Figure 4 none of the fluorocarbon products with branched structures reach a maximum percentage. Therefore for the dose absorbed fluorocarbon radicals with branched structures must be more abundant during radiolysis of the CF4+C2FS mixtures than fluorocarbon radicals with straight-chain structures; This indicates that F atoms must be harder to remove from terminal CF3 groups than from other carbon atoms in fluorocarbon molecules. Branched products cannot be made as primary products from the recombination of the initial radicals formed in

PAGE 37

27 the CF4+C2F6 mixtures so primary products, secondary products, etc., must be made first and then undergo radiolysis themselves. The large amount of starting material converted to products shown in Figure 2 and the large amount of primary products, secondary products, etc., found in the radiolysis products of the CF4 +C2F 6 mixtures indicates that this actually happens. For a small dose, however, only the primary products should be formed in the radiolysis of the CF4 +C2F 6 mixtures and branched fluorocarbon products should not be found in the radiolysis products. This can be seen in the Cobalt 60 irradiations of the CF4 + C2F6 mixtures. (See Chapter IV.) Thus it appears that all radiolysis products arise from radical recombination, as postulated, and that the ratios of the concentrations of the various radicals during radiolysis change with the initial composition of the CF4 +C2F 6 mixtures. The instantaneous radical concentrations will also be a function of the total dose absorbed in the mixtures since all products will undergo radiolysis and produce radicals too. It appears that an equilibrium composition of products in the CF4 +C2F 6 mixtures has not been reached if indeed an equilibrium condition can be reached in the radiolysis of the CF4 +C2F 6 mixtures. The radiolysis products containing oxygen probably arisefrom wall reactions. The handling procedure used (See Chapter II) should .exclude oxygen as a gas phase

PAGE 38

28 contaminant in the samples. The oxygen in the radiolysis products must come from adsorbed oxygen and/or aluminum oxide on the inside walls of the sample tubes. Fluorine is the most electronegative element known and since fluorine atoms are present during radiolysis it seems quite likely that the oxygen on the sample tube wall will be replaced by fluorine thus producing oxygen as a contaminant in the products. Kevan and Hamlet(g) noted that when C2Fe was irradiated in newly made brass cells that C02 was always an irradiation product indicating the presence of oxygen but after the cells were "conditioned" by several irradiations that the radiolysis products usually contained no C02. Rexroad and found that when Teflon was irradiated fluorocarbon radicals were definitely present in the solid and that when the irradiated Teflon was exposed to oxygen the fluorocarbon radicals reacted with the oxygen to form different radicals which contained oxygen. The random variation in the amount of radiolysis products which contain oxygen indicates that the amount of oxygen present must vary from one sample tube to the next. None of the radiolysis products were identified as cyclic or unsaturated compounds. The unidentified products could be unsaturated or cyclic compounds but since the presence of oxygen in the mixtures is established it seems more likely that these products are perfluorooxygen compounds such as perfluoroethers and/or perfluoroalcohols.

PAGE 39

29 Mixture s Containing CF4 and Nonfluorocarbon Compone n t s Weigh t percentages of the radiolysis products from the one week Low Intensity Training Reactor irradiations of the mixtures containing CF4 are given in Table 7 and the results of the four week irradiations are given in Table 8. T h e results of for pure CF4 are also given in Table 8. The presence of aluminum powder in the radiolysis of CF4 increases the yield of all products. This is not surprising since the aluminum powder provides a large surface area and surfaces may affect gas-phase radiolysis in which free atoms are intermediates.(16) An unusual amount of oxygen containing material was not found in the products and this may be because the aluminum powder was freshly prepared and probably contained very little aluminum oxide and/or adsorbed oxygen. The initial compositions of the CF4+CO mixtures were not known but they were approximately 50 mole percent CF4 The large amount of products and small amount of CF4 left after radiolysis indicates that considerable reaction took place in the CF4+CO mixtures. No CO was left in t h e radiolysis' products. The CO probably acts as a radical scavenger during radiolysis especially for the very reactive F atoms and' this would mean that more fluorocarbon compounds should be found in the radiolysis products of the CF4+CO mixtures than in the radiolysis products of pure CF4 This actually occurs as seen in Table 8.

PAGE 40

30 The very large amounts of C2F60, C3FsO, and (C2F5)20 found in the radiolysis products of the CF4+CO mixtures certainl y support the identifications of these products. Many unknown products were found in the radiolysis of the CF4+CO mixtures. The retention volumes of most of these products were identical with the retention volumes of unknown products found in the radiolysis of the saturated fluorocarbons and this supports the idea that the unknown products found in the radiolysis of the saturated fluorocarbons may contain oxygen. The abundance of the fluorocarbon products with branched structures can also be seen in the radiolysis products from the CF4+CO mixtures. The presence of c 'arbon granules during the radiolysis of CF4 causes extreme degradation of the CF4 For the one week irradiation of CF4 + carbon granules in the Low Intensity Training Reactor only 57.0 percent of the original amount of CF4 was found in the radiolysis products and for the four week irradiation only 32.0 percent of the original amount of CF4 was found in the radiolysis products. However, when pure CF4 is irradiated for four weeks in the Low Intensity Training Reactor approximately 98.3 percent of the original' amount of CF4 is found in the radiolysis products Only 90.0 weight percent of the original samp l e was recovered as gas in the one week irradiation and only 61.0 weight percent of the original sample was recovered as gas in the four week irradiation of the CF4 + carbon granule

PAGE 41

31 mixtures. Therefore in both samples some of the material remained in the sample tubes as involatile products The relative amounts of the fluorocarbons found in the gaseous products from the CF4 + carbon granule mixtures are almost identical with the product composition found in the radiolysis products of the CF4+C2F6 mixtures. This means that the same processes must be responsible for the radiolysis products in both systems. The radiolysis products can be explained by the formation and recombination of radicals given for the CF4+C2F6 mixtures except that C2F6 is not present as a starting material in the CF4 + carbon granule mixtures. Kuriakose and have reacted fluorine with graphite and they found that at temperatures between 315 and 530C. the graphite gained weight while above GOOC. the graphite loses weight with the formation of only gaseous fluorides, mainly CF4 They also suggest that the reaction might involve dissociation of fluorine on the graphite surface. Since fluorine atoms are present during the radiolysis it certainly seems possible that the fluorine atoms might react with the carbon in the CF4 + carbon granule samples. The loss of fluorine atoms in the radiolysis mechanism would result in an increase in the production of larger fluorocarbon molecules and this would explain the large product yields found in the CF4 + carbon granules mixtures.

PAGE 42

32 A material balance on the gaseous products from the one week irradiation of CF4 + carbon granules shows the weight loss of fluorine in the gas to be 0.0406 grams and the weight of carbon in the gaseous products was increased by 0.0005 grams. From this it appears that fluorine is lost to the carbon and that the carbon must act as a radical scavenger during radiolysis. A material balance on the gaseous products from the four week irradiation of CF4 + carbon granules shows the weight loss of fluorine to be 0.1460 grams and the weight loss of carbon to be 0.0160 grams. The chemical formula of such a product would be CFs.7 8 which indicates that more fluorine radicals are lost to the carbon than any other kind of radical. This is logical because fluorine atoms can diffuse faster than any other radical present in the radiolysis mixture. The material balances are not absolutely correct due to the .difficulties in obtaining accurate sample weights but they do show that fluorine and possibly carbon-containing radicals are lost to the carbon granules during the radiolysis of the CF4 + carbon granules mixtures. The relatively large yield of products that contain oxygen is probably due to a relatively large amount of oxygen originally adsorbed on the carbon granules in the sample mixture. The abundance of the fluorocarbons with branched structures is also evident in the radiolysis products of the CF4 + carbon granule mixtures.

PAGE 43

33 After the gaseous portion of the sample was removed from the CF4 + carbon granules sample tube which was irradiated for four weeks, half tube was connected to a vacuum system and the tube was heated under vacuum with a cool torch flame. A few drops of a clear viscous liquid condensed in the vacuum system. This liquid was soluble in fluorocarbon liquids but a chromatogram of this material dissolved in showed no fluorocarbon products up through This liquid could be a polymer or a mixture of long chain fluorocarbon products but from the abundance of branched structures the liquid material must be highly branched. The other hal f of the sample tube was treated with solvent and after evaporation of the solvent some clear viscous liquid remained.' About 0 050 grams of material were recovered from the whole tube and since about 0 .1620 grams of sample was lost to the carbon granules it is not known if any of the carbon from the carbon granules was actually transformed into fluorocarbon material. Radioactive carbon tracer studies could answer this question.

PAGE 44

34 TABLE 6 Analysis of the CF4 +C2F6 Mixtures Irradiated for One Week in the Low Intensity Training Reactor Sample No. D 2 Sample Wt. 0 .3966 Density 0.5352 gm./cm.3 Original Composition (Wt.%) 86.66 CF4 13.34 C2F6 % Recovery as gas 98.66 Product or Vo on n -hexadecane column Wt. % CF4 81.40 C2F6 7.69 C2F60 3.29 C3.Fs 1.16 peak 1 3 .77 total C4FlO 0.232 n -C5Fl2 0.0819 i-C5Fl20 .504 183.8 cm. 3 195.0 cm. 3 218.5 cm. 3 n C6Fl4 252.2 cm. 3 2,3-(CF3)2C4Fs 0.551 0 .724 0.0455 0.123 0.0614 0.126 0.058

PAGE 45

35 TABLE 6 (Continued) Sample No. B 4 Sample Wt. 0 .4944 gm. Density 0.5352 gm.7cm. 3 Original Composition (Wt.%) 67.94 CF4 32.06 C2FS % Recovery as gas 100.28 Product or Vo on n -hexadecane column Wt. % CF4 74.73 C2Fs 14.32 C2F s O 0.402 C3FS 4 94 C3FSO 0.137 108. 3 cm. 3 0.0271 n C 5Fl2 0.236 i C 5Fl2 1. 47 183. 8 cm. 3 195.0 cm. 3 3,3-(CF3)2C5FlO 0.0065 0.0113 0.0528 0.0541 0.0358 0.687 0.126 0.135 0.0549

PAGE 46

36 TABLE 6 (Continued) Sample No. A-4 Sample Wt. 0 .4330 gm. Density 0.4810 gm./cm.3 Original Composition (wt.%) 39.8 6 CF4 60.14 C2F6 % Recovery as gas 100.60 Product o r Vo on n -hexadecane column Wt. % C2F6 15.79 C2F60 0.548 peak 1 0.138 i 4.29 102.1 cm. 3 0.0971 0.601 63 183. 8 195.0 cm.3 3 cm. 2,3-(CF3)2C4FS 0.0225 0.0464 0.197 0.282 0.182 2.34 0.0788 0.0466 0.0654 .'

PAGE 47

37 TABLE 6 (Continued) Product or Vo on n-hexadecane column Wt. % 0.606 0.0257 0.517 0.726 CSF20 0.258 Sample No. C-3 Sample Wt. 0.6208 gm. Density 0.6855 gm./cm.3 Original Composition (Wt. %) 18.53 CF4 81.47 C2F6 % Recovery as gas 100.37 Product or Vo on n -hexadecane column wt. CF4 50.85 C2F 6 12.39 C2F 6 O 0.393 C3FS 7.88 C3FSO 0.168 2.37 6.74 0.859 7.52 0.273 0.647 0.420 2,3-(CF3)2C4FS 3.94 0.0720

PAGE 48

TABLE 6 (Continued) Product or Vo on n -hexadecane column Wt. % 2-CF3CeFl3 0.299 3 CF3CeFl3 0.216 2,4-(CF3)2C5FlO 1.09 2,3-(CF3)2C5FlO 0.0546 3,3-(CF3)2C5FlO 0.942 CSFlS 1.51 CSF20 0.843 Clo F22 0.441 Sample No. E-1 Sample Wt. 0 .6304 gm. Density 0.8507 gm./cm. 3 Original Composition (Wt. %) 8.25 ,CF4 91.75 C2Fe % R ecovery as gas 100.29 Product or Vo on n -hexadecane column 102.1 cm. 3 Wt. % 50.44 8.55 0.558 6.50 0.232 1.81 7.92 0.0040 0.829 9.52 0.0299 0.343

PAGE 49

TABLE 6 Product or Vo on n -hexadecane column 3-CF3C 5Fll 2,3-(CF3)2C4Fs 397.2 cm.3 2,4-(CF3)2C5 FlO 2,3-(CF3)2C5 FlO 3,3-(CF 3)2C5 FlO 39 (Continued) Wt. % 0.875 0.510 5.17 0.0054 0.0688 0.201 0.389 0.0299 1.49 0.142 1.14 1.35 0.998 0.814 0.0861

PAGE 50

40 TABLE 7 Analysis of the Mixtures Containing CF4 and Nonfluorocarbon Components Irradiated for One Week in the. Low Intensity Training Reactor Sample No. I-3 aluminum powder Sample Wt. 0.3576 gm. CF4 0.5579 gm. al powder % Sample -recovery as gas 117.67* Product or Vo on n-hexadecane column peak 1 3 97.1 cm. 113.4cm.3 Wt. % 92.60 0.0062 4.33. 2.41 0.404 0.151 0.0533 0.0110 0.0161 0.0126 0.0002 0.0013 0.0002 0.0009 Some aluminum powder lost when sample removed which makes % sample recovery too large.

PAGE 51

41 TABLE 7 (Continued) Sample No. G-3 CF4+CO Sample Wt. 0.3101 gm. % Sample recovery as gas 97.52 Product or Vo on n-hexadecane column Wt. % CF4 32.51 C02 2.99 C2FS 6.38 C2FsO 35.01 peak 1 3.92 total C4FlO 0.639 (C2 F5)20 1.46 108.3 cm. 3 118.5 cm. 3 176.7 cm. 3 195.0 cm. 3 218.5 cm. 3 252.2 cm. 3 284.8 cm. 3 2,3-(CF3)2C4Fs 8 3 3 2.9 cm. 0.263 0.589 0.912 5.25 0.263 0.223 0.383 1.23 0.0851 0.302 0.0715 0.918 0.0362

PAGE 52

42 TABLE 7 (Continued) Product or Vo on n -hexadecane column Wt. % n C7Fl6 0 .0831 3-CF3C6 F l 3 0.158 550. 4 cm. 3 0.176 2,4-(CF3)2C5F lO 0.0431 609. 6 cm. 3 650. 4 cm. 3 3;3-(CF3)2C 5 F l O CSFlS trace trace 0.0479 0.0376 Sample No. H 3 CF4 + Carbon granules Sample wt. 0.4036 gm. CF4 0.11 94 gm. carbon granules % Sample recovery as gas 90.04 Product or Vo on n -hexadecane column Wt. % CF4 63.25 CO2 0.554 C2F 6 8 .73 C2F 6 O 8.57 C3Fs 2.79 C3F S O 1.22 peak 1 0.178 n C4F lO 1.68 i-C4F lO 2.56 (C 2 F 5)20 0.878 97.1 cm. 3 0.0105 102.1 cm. 3 0.0945 n-C5Fl2 0.659

PAGE 53

4 3 TABLE 7 (Continued) Product o r Vo on n-hexadecane column 183.8 cm. 3 195.0 cm. 3 252.2 cm. 3 2,3-(CF3 )2C4Fs 3,5-(CF3)2C5Fio Wt. % 4.27 0.9.29 0.086 0.354 0.0338 0.254 0.157 2.01 0.0381 0.0532 0.113 0.0741 0.439 0.0226 0.393 0.402 0.0917 /

PAGE 54

44 TABLE 8 Analysis of the Mixtures Containing CF4 and Nonfluorocarbon Components Irradiated for Four Weeks in the Low Intensity Reactor Sample No.8 CF4* Sample Wt. 0.1752 gm. % Sample recovery as gas 100.97 Product or Vo on n-hexadecane Column Wt. % 98.31 0 95 0 72 0.03 Resul ts from J C. Mailen (.1.) Sample No. G-5 CF4+CO Sample Wt. 0.3248 gm. % Sample recovery as gas 96.45 Product or Vo on n -hexadecane Column CO2 C2FS C2FsO C3Fs C3FsO peak 1 total C4F1O (C2F5)20 108.3 cm. 3 118. 5 cm. 3 n -C5F 12 Wt. % 23 .15 1.91 4.41 28.49 0 692 3.49 4 29 0 270 0 851 0.796 1.68 2 49

PAGE 55

45 TABLE 8 (Continued) Product or Vo on n -hexadecane Column 176.7 cm.3 195.0 cm.3 218.5 cm.3 252.2cm.3 284.8 cm.3 2,3-(CF 3)2C4Fs 382 9 cm.3 550.4 cm.3 2,4-(CF3)2C5FlO 609 6 cm.3 650.4 cm.3 3,3-(CF3)2C5F l O Wt % 12.19 0.772 0 85 1 1.45 4 .70 1.17 0.469 2.71 0.282 0.425 0.519 0.815 0.139 0.0799 0.9487 0.214 0.644 0.0071

PAGE 56

46 TABLE 8 (Continued) Sample No. H-2 CF4 + carbon granules Sample Wt. 0.4190 gm. CF4 0.1239 gm. carbon granules % Sample recovery as gas 61.34 Product or Vo on n-hexadecane Column peak 1 176.7 cm. 3 195.0 cm. 3 218.5 cm. 3 252.2 cm. 3 2,3-(CF3)2C4Fs 382.9 cm. 3 wt. % 52.15 trace 13.09 4.94 4.83 1.07 0.189 1.63. 2.21 0.674 0.673 '6.71 0.978 0.0936 0.0075 0.842 0.0483 0.796 0.906 2.96 0.0723 0.0755

PAGE 57

47 TABLE 8 (Continued) Product or Vo on n-hexadecane Column Wt. % 0 .365 0.184 0.620 0.189 0.572 1.80 CSF20 1.18 0.156

PAGE 58

IV. COBALT 60 GAMMA IRRADIATION S OF MIXTURES CONTAINING FLUOROCARBONS AND OF PUKE FLUOROCARBON S Weight percentages and G values (molecules/100 e.V.) o f the radiolysisproducts of the CF4+C2Fe mixtures are in Ta ble 9 and the G values are plotted versus weight percent of C2Fe in the original mixtures in Figures 5 and 6 The G values are average G v alues and they were calculated from the weight of each product found in the radiolysis products. The initial G value for a given product will probably be some what higher than the values given here because for the total doses absorbed some of the initial products are destroyed-by undergoing radiolysis themselves. All of the identified products of the radiolysis of the CF4 +C2F e mixtures can be explained by the processes given for the Low Intensity Training-Reactor irradiations on page 21. The linear relationship between the G values of C3F S and the weight percent of C2Fe in the original mixture shown in Figure 5 indicates that C3FS is formed by a first order reaction with respect to the concentration of C2Fe and the curve for C4FlO show n in Figure 5 indicates that n-C4 F l O is formed by a second order reaction with respect 48

PAGE 59

0.8 0.7 0.6 0.5 (l) :::J rl co 0.4 :> c:J 0 3 0 2 0 1 o 49 0 117 Megarads G 77.2 Megarads e 40 5 Megarads 0 C3Fs 0 Total C4FJ.o 0 100% CF4 100% C 2Fs Weight Percent in Original Sample Mixture Figure 5 G Value of Primary Products Versus Weight Percent of C2Fs in O r iginal Sampl e Mixture

PAGE 60

50 0.30 Dose in Mixtures 117 Megarads Dose in Pure CF477.2 Megarads Dose in Pure C2FS 40.5 Megarads 0.25 0.20 Q) 0.15 C\j p. o 0.10 0.05 Weight Percent in Original Sample Mixture Figure 6 G Value of Products Containing Oxygen Versus Weight Percent of C2Fs in Original Sample Mixture'

PAGE 61

51 to the concentration of C2Fs in the original mixture. This certainly supports the of free radical reactions occuring during the radiolysis of the CF4+C2Fs mixtures. The reactions producing fluorocarbon products seem to be occuring in the bulk of the medium since an absorbed dose variation of from 40.5 to 117 megarads seems to have little effect on the G values of fluorocarbon products. The G values of products containing oxygen are plotted versus weight percent of C2FS in the original sample in Figure 6. The shape of the C3FsO curve and the shape of the (C2F5)20 curve indicate that these products are also formed by free radical reactions. The G values of the products containing oxygen seem to be dose dependent and this substantiates wall reactions occuring since diffusion to the wall may control the reaction and the absorbed dose is directly proportional to the length of time of irradiation. For the total absorbed doses in the Cobalt 60 irradiations of the CF4+C2F s mixtures there was probably always ample oxygen pr@sent in the sample tubes so that the random variation in oxygen-containing products as found in the Low Intensity Training Reactor irradiations was not observed in these samples. The shape of the C2Fs O curve shown in Figure 6 may reflect the variation in the concentration of CF3 radicals at the tube wall during radiolysis as the weight percent of C2FS is increased in the original sample mixtures.

PAGE 62

52 Mixtures Containing CF4 and Nonfluorocarbon Components Weight percentages and G values of the products from the Cobalt 60 irradiations of the mixtures containing CF4 are given in Table 10. It should be recalled here that the absorbed doses for these samples were calculated by assuming that the solid material had no effect on the gamma rays during radiolysis. This, of course, is a bad assumption so the G values given for these samples are not very reliable. The presence of aluminum powder during the Cobalt 60 irradiation of CF4 seems to increase the yield of products in keeping with the results of the Low Intensity Training Reactor irradiations. The presenGe of CO during the radiolysis of CF4 increases the yield of C02 but completely inhibits the formation of all other products except an unknown product which was found in this sample. CO was also found in the radiolysis products. Nothing conclusive can be said about the results of the Cobalt 60 irradiation of CF4 + carbon granules. Fluorocarbon products are being formed in this sample but the extreme degradation of CF4 that was noted in the Low Intensity Training Reactor irradiations of CF4 + carbon granules does not seem to be occuring in this sample. It should be remembered, however, that the absorbed dose calculated for this sample could be very much error and the destruction of CF4 may be much larger than it appears to be in this sample.

PAGE 63

53 Pure Fluorocarbon Samples Weight percentages and G values of the radiolysis products from the Cobalt 60 gamma ray irradiations of the pure fluorocarbon samples are given in Table 11. The free radical reactions predicted for the formation of the perfluoroheptane isomers certainly appear to occur in the radiolysis of the pure parent perfluorocarbon samples. The agreement between the pattern of products predicted by the free radical mechanism and that of the actual products obtained in the radiolysis of the parent perfluorocarbon compounds is almost exact. (See Chapter v.) G values for the disappearance of the starting material during radiolysis of the pure saturated fluorocarbon samples are given in Table 12. It can be seen from the table that perfluoromethane is the most stable to irradiation of all saturated fluorocarbons studied. CF4 is also the largest product in all saturated fluorocarbon irradiations except in the irradiation of The contained some i-C4Fs as an impurity and this may account for the low percentage bf CF4 in the radiolysis products of the sample. The abundance of branched fluorocarbons as products indicates that F atoms must be harder to remove from terminal CF3 groups than from other carbon atoms during radiolysis of saturated fluorocarbons. It also appears that larger fluorocarbon molecules aPe more to irradiation than smaller fluorocarbon mOleCUle )

PAGE 64

54 excluding CF4 The larger fluorocarbon molecules were present as the liquid phase during irradiations and this may account for the lower G values since radical diffusion in the liquid is slower than in the gas phase. Therefore radical recombination to produce the parent molecule may be more significant in the liquid samples than in the gas samples. The G values for the disappearance of the starting material for the liquid fluorocarbon samples agree very well with the value of 2 to 3 reported by Simons and The radiation chemistry of saturated fluorocarbons is much different in some respects than the radiation chemistry of the analogous hydrocarbons. The great strength of the C-F bond compared to the C-H bond and the weakness of the F-F bond compared to the H-H bond is reason enough to suspect the radiation chemistry of saturated fluorocarbons to be much different from the radiation chemistry of saturated hydrocarbons. Molecular hydrogen is a stable product in the radiolysis of saturated hydrocarbons but molecular fluorine has never been found in the radiolysis products of saturated fluorocarbons. Hydrogen abstraction by hydrogen atoms to form stable molecular hydrogen occurs in the radiolysis of hydrocarbons but fluorine abstraction by fluorine atoms is essentially impossible in fluorocarbons. The absence of unsaturated products in the radiolysis of saturated fluorocarbons indicates that

PAGE 65

55 disproportionation reactions which occur in the radiolysis of hydrocarbons do not occur in the radiolysis of saturated fluorocarbons. For the above reasons saturated fluorocarbons should be more stable to irradiation than saturated hydrocarbons and the G values reported here certainly support this idea.

PAGE 66

56 TABLE 9 Analysis of the CF4+C2 F6 Mixtures Irradiated in the Cobalt 60 Gamma Ray Source Sample No. 9 Sample Wt. 0 .1852 gm. Density 0.222 gm./cm.3 Original Composition (Wt %) 100. 0 CF4 No. days in Cobalt 60 source 33.09 Absorbed dose = 8.92 e .V = 77.2 megarads Product or V o on n -hexadecane Column Sample from J C. Sample No. D 4 Wt. % 99.30 0.0931 0 258 0.296 0 .0306 0.02 26 G Value 0 997 0 264 0 233 0 240 0 020 0.014 Sample wt 0 .4115 gm. Density 0.5532 gm./cm. 3 Original composition (Wt %) 86.66 CF4' 13.34 C2F6 No. days in Co 60 source 46 .89 Absorbed dose 2.52 x 10 2 l e .V. = 117 megarads % Recovery as gas 104.20 Product or Vo on n-hexadecane Column Wt. % G Value CF4 87 .15 0.549 C2F6 12. 10 -0.885 C2F6O 0 239 0 .153 C3FS 0.240 0.1 26 C3FSO 0 .0663 0 0320 total C 4 F lO 0.0463 0.0191

PAGE 67

57 TABLE 9 (Continued) Product or Va on n-hexadecane Column 102.1 cm.3 Sample B-3 Wt. % trace 0.0238 0.0520 0.0355 0 .0101 trace trace 0 .0122 G Value 0.0177 0.0121 0.0030 0.0036 Sample wt. 0.5049 gm. Density 0.5533 gm./cm.3 Original composition (Wt. %) 67.94 CF4 32.06 C7F6 No. days in Co 60 source 46.89 Absorbed dose 2.51 x 1021 e.V. = 117 megarads % Recovery as gas 99.82 Product or Va on n-hexadecane Column wt. % G Value CF4 C2F 6 C2F 6 O C3FS C3FsO total C 4 F10 (C2F5)20 102.1 cm.3 n C 5 F12 i-C5F12 n C 6F14 68.04 31.11 0 .124 0.414 0.119 0.0882 0.0061 0.0198 0.0298 0.0345 0.0081 0.137 -0.833 0.0974 0.266 0.0708 0.0449 0.0030 0 .0126 0.0147 0.0028

PAGE 68

58 TABLE 9 (Continued) Product or Vo on n-hexadecane Column Sample No. A-5 Sample Wt. 0.4550 gm. Density 0.4931 gm./cm.3 Wt. % trace trace 0.0122 G Value 0.0043 Original composition (wt. %) 39.86 CF4 60.14 C2F6 No. days in Co 60 source 46.89 Absorbed dose 2.25 x 102l e.V. = 117 megarads % Recovery as gas 100.09 Product or Vo on n-hexadecane Column C3Fs C3FSO 102.1 cm.3 3-CF3C5Fll Wt. % 39.26 59.19 0.102 0.693 0.183 0.226 0.0275 0.0779 0.0722 0.0661 0.0305 0.00969 0.0186 0.0351 0.0163 G Value -0.833 -0.840 0.0812 0.450 0.109 0.116 0.0132 0.0306 0.0280 0.0110 0.00350 0.00672 0.0127 0.00513

PAGE 69

59 TABLE 9 (Continued) Sample No. C-2 Sample wt. 0.5801 Density 0.6366 Original composition (wt. %) 18.53" CF4 81.47 C2F6 No. days in Co 60 source 46.89 Absorbed dose 2.90 x 10 21 e.V. = 117 megarads % as gas 100.71 Product or Vo on n-hexadecane Column C3FsO 102.1 cm. 3 3-CF3C5 F 11 2,3-(CF3)2C4Fs Wt. % 17.64 80.09 0.194 0.993 0.244 0.485 0.0777 0.115 0.055 0.053 0.0076 0.0209 0.0034 0.0034 0.0188 0.0096 G Value -1.22 -1.19 0.152 0.638 0.144 0.246 0.0369 0.0230 0.0222 0.00746 0.00121 0.00121 0.00671 0.00298

PAGE 70

60 TABLE 9 (Continued) Sample E-2 Sample wt 0.6128 gm. Density 0.8498 gm./cm.3 Original composition (wt. %) 8.25 CF4 91.75 C2Fs No. days in Co 60 source 46.89 Absorbed dose 3.87 x 10 21 e.V. = 117 megarads % Recovery as gas 100.94 Product or Va on n-hexadecane Column C3Fs (C2 F5)20 102.1 cm.3 176.7 cm.3 195.0 cm.3 Wt. % 7.65 88.70 0.214 1.19 0.371 0.492 0.0851 0.927 0.142 0.0865 0.0142 0.0072 0.0542 0.0149 0.0207 0.0269 0.0090 G Value 65 2 -2.11 0.133 0.602 0.173 0.197 0.0320 0.0471 0.0287 0.0153 0.00421 0.00585 0.0076 0 0.00222

PAGE 71

61 TABLE 9 (Continued) Sample No. 68 (reanalyzed)* Sample wt. 0.3723 gm. Density 0.447 gm./cm.3 Original composition (wt. %) 100.0 C2F6 No. days in Cobalt 60 source 17.17 Absorbed dose = 9.43 x 1020 e.V. = 40.5 megarads Product or Vo on n-hexadecane Column (C2Fs)20 102.1 cm.3 Wt. % '0.279 97.95 0.0868 0.588 0.0703 0.170 0.0247 0.757 0.0199 0.0258 0.0072 trace trace 0.000144 *Sample from J. C. G Value 0.755 -3.542 0.134 0.745 0.082 0.170 0.0232 0.0165 0.0213 0.0051 0.000101

PAGE 72

62 TABLE 10 Analysis of the Mixtures Containing CF4 and Nonfluorocarbon Components Irradiated in the Cobalt 60 Gamma Ray Source Sample No. 1-1 CF4 + aluminum powder Sample Wt. 0 2948 gm. CF4 0 5406 gm. aluminum powder No. days in Cobalt 60 source 40.75 Absorbed dose 1.46 x 1021 e.V. = 79.1 megarads % Sample recovery as gas 106.58* Product CF4 C O 2 C2FS C2Fs O C3Fs -*Some aluminum powder Sample No. G-2 CF4+CO Sample Wt. 0.3209 gm. Wt. % G Value 98 73 -1.755 0 259 0.714 0.718 0.631 0 225 0.177 0.071 5 0 0461 lost when sample was removed. No days in Cobalt 60 source 40.75 Absorbed dose 1.584 x 1021 e.V. = 79. 1 megarads % Sample recovery as gas 98.19 Product CO CF-4 CO2 peak 1 Wt. % 10.20 88.76 0.973 0;068 G Value 2.70 0.060

PAGE 73

63 TABLE 10 (Continued) Sample No. H-l CF4 + carbon granules Sample Wt. 0.3495 gm. CF4 .0.1045 gm. carbon granules No. days in Cobalt 60 source 40.75 Absorbed dose = 1.735 x 102J. e.V. = 7 9 5 megarads. % Sample recovery as gas 101.29 Product Wt. % 99.09 0.362 0.552 G Value -1.26 1 0.997 0.486

PAGE 74

64 TABLE 11 Analysis of the Pure Fluorocarbons Irradiated in the Cobalt 60 Gamma Ray Source Sample No. 2-B Sample Wt. 0.4782 gm. No. days in Co60 source 3 8 9 1 Absorbed dose = 4.56 x e.V. = 152 megarads % Sample recovery as gas 100.23 Product or Vo on, n-hexadecane Column CF4 CO2 C2F S C2Fs O C3FS C3FSO 102.1 cm. 3 n-C 5 Fi2 2,3-(CF3)2C4Fs 397.2 cm. 3 Wt. % 1 ,.06 0.0562 0.501 0.0342 0.814 0.0134 87.59 0.0178 '0.871 1.48 0.580 0.133 0.8 70 0.187 0.011 0.368 0.0742 1.44 0.0061 G Value 0.76 1 0.0809 0.230 0.0141 0.274 0.00416 3.305 0.192 0.324 0.109 0.0250 0.163 0.0350 0.06 02 0.0121 0.236 0.0010

PAGE 75

65 TABLE 11(Continued) Product or Vo on n -hexadecane Column 2,4-(C F 3) 2 C 5 F 1 O 2 3-(CF3)2C 5 F 1 O 2,2,3-(CF3)3C4 F7 3 ,3-(CF 3)2C5F1O CSF1 S CSF20 Sample No. 6 A i C 4 F 1 0 Sample Wt. 0 3466 gm. Wt. % 0 129 0.21 2 0 005 0.105 2 72 0 734 No. days in Cobalt 60 source 38 .91 G Value 0 0210 0 0346 0 0008 0 0172 0 393 0 0953 Absorbed dose = 3 54 x 1021 e .V. = 164 megarads % Sample recovery as gas 98.90 Produc't or V o on n -hexadecane Column CF4 C O 2 C2F6 C3Fs i -C4F1O 108.3 cm. 3 n C 5 F 1 2 i C 5 F 12 n C6F 1 4 2 -CF3C5F11 2,3-(C F 3) 2 C 4 F s 3 C F3C6F 1 3 2,4-(C F 3)2C5F1O 2 ,3-(CF3)2C5F 1 O 2, 2,3-(CF3)3C 4 Fi wt 0 414 0 968 0.09 88 0 673 88 32 0.0342 0.0345 2.24 0 0163 0.1 3 1 0.851 1 24 0.347 0.0267 0.0797 G Value 0 277 1 .30 0 0422 0 211 2 894 0.00707 0 459 0 00284 0 0228 0 149 0 188 0 0527 0 00406 0 012 1 /

PAGE 76

TABLE 11 Product or Voon n-hexadecane Column CSF20 Sample No. l-A Sample Wt. 0.5637 gm. 66 (Continued) Wt. % G Value 3.43 0.462 0.429 0.0518 0.429 0.0470 0.242 0.0242 No. days in Co60 source 38.91 Absorbed dose = 5.22 x e.V. = 149 megarads % Sample recovery as gas 97.30 Product or Vo on n-hexadecane Column C3Fs 183.8 cm. 3 210.3 cm. 3 2,3-(CF3)2C4 Fs 372.7 cm. 3 398.2 cm.3 Wt. % 0.783 0.0754 0.538 0.536 0.548 87.93 0.233 0.0122 0.0140 0.461 0.483 0.714 0.0143 0.0045 0.0103 0.217 G Value 0.578 0.111 0.253 0.185 0.149 -2.723 0.0525 0.0886 0.0928 0.137 0.00275 0.0364

PAGE 77

67 TABLE 11 (Continued) Product or Va on n-hexadecane Column Sample No. 4-A Sample Wt. 05437 gm. Wt. % 0.600 0.675 0.0137 0.0218 0.0375 1.44 1.93 2.72 G Value 0.0988 0.113 0.00229 0.00371 0.00628 0.214 0.257 0.328 No. days in Cobalt 60 source 38.91 Absorbed dose = 5.20 x e.V. = 153 megarads % Sample recovery as gas 99.06 Product or Va on n-hexadecane Column wt. % 0 .724 0.548 0.816 0.812 0.830 0.0781 0.00705 86.37 1.03 0.112 G Value 0.518 0.0740 0.250 0.273 0.215 0.181 0.0171 -2.539 0.0208

PAGE 78

68 TABLE 11(Continued) Product or Vo on n-hexadecane Column Wt. % G Value 2,3-(CF3)2C4Fs 0.0273 0.00508 397.2 cm. 3 0.285 n-C7FJ.6 0.291 0.0471 2-CF 3 C 6FJ.3 0.319 0.0518 3-CF3C 6 FJ.3 1.29 0.208 3-C2F5C5FJ.J. 0.0068 0.0011 2,4-(CF3)2C5FJ.O 0.0206 0.00328 2,3-(CF3)2C5FJ.o 0.0185 0.00300 2,2,3-(CF3)3C4 F7 0.0096 0.00156 3,3-(CF3)2C5FJ.o 0.0137 0.00222 CsFJ.s 0.989 0.142 CSF20 2.19 0.282 CJ.oF22 1.57 0.184 CJ.J.F24 1.23 0.131 CJ.2F26 0.375 0.0370

PAGE 79

69 TABLE 11 (Continued) Sample No. 5-A 2-CF3C5Fll Sample Wt. 0.4555 gm. No. days in Cobalt 60 source 38.91 Absorbed'dose 4.70 X1021 e.V. = 165 megarads % Sample recovery as gas 98.44 Product or Vo on n-hexadecane Column 108.3 cm. 3 n-C5 F 12 'i-C5F 12 173.6 cm. 3 201.2 cm. 3 218.5 cm. 3 Wt. % 1.17 0.0553 0.497 0.588 0.176 0.652 0.0196 0.578 0.370 0.0224 0.0107 G Value 0.778 0.0734 0.210 0.183 0.0431 0.160 0.117 0.0750 0.0125 0.254 86.18 0.0439 -2.388 2,3-(CF3)2C4F s 397.2 em' 3 0.171 0.0170 0.0398 0.0624 0.569 0.0138 1.48 0.0296 0.00599 0.00939 0.0857 0.00208 0.222

PAGE 80

70 TABLE 11 (Continued) Product or Vo on n-hexadecane Column 2,2,3-(CF3)3C4 F7 3,3-(CF3 )2C 5FJ.o CSF20 Sample No. 3-D S ample Wt. 0.2695 gm. Wt. % 0.459 0.0078 0.0428 1.28 2.08 0.952 1.58 0.660 No. days in Cobalt 60 source 38 91 Absorbed dose 2.79 x 10 2J. e.V. = 166 megarads % Sample recovery as gas 99.93 Product or Vo on n-hexadecane Column Wt. 2f CF4 1.28 CO2 0.0567 C2F S 0.0883 C3Fs 0.104 0.0146 0.481 108.3 cm. 3 0.0244 0.231 0.696 176.7 em. 3 0.0392 221.5 cm. 3 0.0451 G Value 0.0691 0.00117 0.00644 0.171 0.248 0.103 0.157 0.0604 G Value 0.844 0.0750 0.0370 0.0321 0.00357 0.118 0.0466 0.141

PAGE 81

71 TABLE 11 (Continued) Product or Va on n-hexadecane Column 2,3-(CF3)2C 4 FS 409.4 cm. 3 Wt. % 0.216 86.494 0.292 0.0311 0.0215 0.144 0.231 0.464 2.48 2.45 1 27 1.57 0.929 0.380 Sample No. 26 C3Fs (reanalyzed)* Sample wt 0.4903 gm. G Value 0.0372 -2.326 0 0503 0.00322 0.0216 0.0346 0.0696 0.373 0.326 0.151 0.169 0.0920 0.0346 No. days in Cobalt 60 source 33.10 Absorbed dose 4.97 x e.V. = 163 megarads Product or Va on n-hexadecane Column CF4 C O 2 C2F e C2Fe O C3Fs Wt. % 1.57 0.0158 1.74 0.0214 89.56 1.53 G Value 1.06 0.0213 0.747 0.0050 -3. 297 0.382

PAGE 82

72 TABLE 11 (Continued) Product or Vo on n-hexadecane Column 322.6 cm. 3 2,3-(CF3 )2C 4 Fs Wt. % 1.42 0.712 1.17 0.283 0.640 0.0593 0.0202 0.507 0.0312 0.0406 0.0815 0.241 0.0226 0.120 0.221 Sample from J. C. Sample No. 94 2,3-(CF3)2C4Fs Sample Wt. 0.4133 gm. G Value 0.353 0.147 0.241 0.0496 0.112 0.0104 0.0890 0.00478 0.00622 0.0125 0.0368 0.00346 0.0184 No. days,in Cobalt 60 source 9.91 Absorbed dose 1.31 x e.V. = 50.7 megarads Product or Vo on n-hexadecane Column Wt. % 0.821 0.0318 0.0473 0.381 G Value 1.78 0.138 0.0653 0.386

PAGE 83

73 TABLE 11 (Continued) Product or Vo on n-hexadecane Column 197.1 cm.3 108.3 cm.3 176.7 cm.3 201.2 cm.3 2-CF3CsFll 2,3-(CF 3)2C4Fs 3-CF3C6F13 2,3-(CF 3)2C s Fl0 Wt. % 0.0401 0.0298 0.192 0.0185 0.0' 231 0.224 0.0095 0.0547 0.0208 0.102 92.04 0.025 0.406 0.06 49 1.80 1.15 1.40 0.0793 0.970 Sample from J. C. Mailen (1,.) G V alue 0.0321 0.0238 0.0153 0.148 0.0117 0.0577 .-4.434 0.0123 0.199 0.0319 0.781 0.447 0.495 0.0257 0.290

PAGE 84

TABLE 12 G Values for the Disappearance of the Starting Material -G Starting Absorbed Dose Sample Material (Megarads) CF4 0.997 77.2 C2Fs 3.542 40.5 C3Fs 3.297 163 3.305 152 2.894 .. 164 2.723 149 2.539 153 2.388 165 2.326 166 2,3-(CF3)2C4 FS 4.434 50.7

PAGE 85

v. PERFLUOROHEPTANE ISOMERS Identification of the Perfluoroheptane Isomers The retention volume (V o ) of and the reten tion volume of on the n-hexadecane column were estimated from Figure 1 to be 427.8 cm. 3 and 7 60.7 cm. 3 respectively. The G values and the retention volumes on the n-hexadecane column of all radiolysis products from to are given in Table 13 for the parent molecules listed in Table 1 which were irradiated in the Cobalt 60 gamma ray source to produce the perfluoroheptane isomers. Table 1 predicts that the radiolysis products of should have two major peaks in the perfluoroheptane group corresponding to the two perfluoroheptane isomers which can be formed as primary radiolysis products assuming that the products are made by radical formation and recombination as stated earlier. This is actually the case as. seen in Table 13. Likewise the radiolysis products of should have three major peaks in the perfluoroheptane group, the radiolysis products of should have three major peaks in the perfluoroheptane group, and the radiolysis products of should have four major peaks in the perfluoroheptane group. This is also the case as seen in Table 13. 75

PAGE 86

Retention Volum e cm.S ) Parent Molecule n-94Fl0 i-C4 Fl0 n-CSF12 n-C6 F12 2-CFsCsFll 3-CFsCsFll TABLE 13 G Values and Retention Volumes on the n-hexadecane Column* of All Radiolysls Products from n-C7F16 to n-CSF1S for the Parent Molecules Listed in Table 1 G Value of Product** 0 0\ (\j CD co t-. t-. t--0 r<\ 0 t--0\ co r<\ CD 0\ CD (() (\j CD ..::t Lf\ CD Lf\ CD r-I..::t ..::t Lf\ IIi II II II II t--II 11 II 0 0 0 S' po po 10.0 602 1 0.0121 0.0010 0.0210 0.0346 0.0008 10. i881 10.0527 1 0.00406 10.01211 10.0988 1 0.00229 0.00371 1 0.047111 0.0518 1 0 .0011 0.00328 0.00300 0.00156 0.00599 I 0.009391 1 0 .0857J o .Ob208 1 0.222 1 0.00117 0.00322 10.02161 1 0.03461 I 0.06961 Operating conditions of n -hexadecane column given in Table 4. ** Major peaks are outlined CD t-. t--0 0
PAGE 87

77 Table 1 predicts that the radiolysis products of i-C4 F10 should have two major peaks in the perfluoroheptane group and that these two peaks should be different from the two perfluoroheptane product peaks found in the radiolysis products of n-C4F10. The largest perfluoroheptane product peak found in the radiolysis products of i-C4F10, contrary to expectation, had the same retention volume as the second major perfluoroheptane peak in the n-C4 F10 radiolysis products. The i-C4 F10 starting material was not completely pure, however, and this may account for the "unwanted" perfluoroheptane peak. Two major perfluoroheptane peaks in addi tion to the "unwant'ed" peak were actually found in the radiolysis products of i-C4F10. These two peaks are probably the desired perfluoroheptane isomers. Five major perfluoroheptane peaks are predicted in the radiolysis products of 2-CF3C5F11. One of these perfluoroheptane peaks should be and the retention volume of this peak should be different from the retention volumes of all other peaks in the .perfluoroheptane group. Only four major perfluoroheptane peaks are actually found in the radiolysis products of 2-CF3C5F11 and all of these peaks have reten n volumes which duplicate ret.entionvolumes found for other perfluoroheptane peaks. is concluded that the perfluoroheptane isomer is probably not produced as a product in the radiolysis of 2-CF3 C 5F11. In fact, no fluorocarbon products with neo structures have ever been found in any of the irradiations of fluorocarbons 'at the University of Florida. (12)

PAGE 88

By comparing Tables 1 and 13 the retention volumes on the n-hexadecane column of eight of the nine perfluoro-heptane isomers are determined. The retention volumes on the n-hexadecane column of all of the perfluoroheptane isomers except are given in Table 14. Thus, all of the perfluoroheptane isomers except have been identified by retention volume on the n-hexadecane column. The partition coefficients for the perfluoroheptare isomers on the n-hexadecane column are listed in Table 4. The very good agreement between the predicted results and the experimental results certainly supports the hypothesis of free radical reactions occuring during the radiolysis of saturated fluorocarbons. Estimation of Normal Boiling Points It has been shown theoretically by for members of a homologous series the logarithm of the retention volume, log V o versus the normal boiling point, T b should be a straight line. This phenomena has been experimentally verified for many different homologous series. Sullivan et have also shown that this relationship is approximately true for the hexene and hexane isomers. The data of Desty and for hydrocarbons indicate that log Vo plotted against Tb for homologous series have one slope while such plots for isomers of individual saturated hydrocarbons have a different slope.

PAGE 89

TABLE 14 Retention Volumesof the Perfluoroheptane on the n-hexadecane Column* PerfluoroheptanJ Isomer and Retention t'-Volume (cm.3 ) 0 IIi 0 r-i 0 .j< r-i IIi r-i U IIi III iII (I') III r-i U III .--... U CD (l')t----(1')0 r-i0\ C\Jt----UC\J (l')tD C\JtD r-i r-i iII .--... .--... IIi .--... t----IIi 0\ 1Iit----1Il0 (l')tD (1')1"\ Uo (l')t----COC\J COtD co CD UI"\ IIitD iII 0\ '--'" w IIi 0 r-i..::j-U..::jU..::j-1IlLf\ ULf\ U Lf\ ltD ut---Parent iII (I') (I') IIi '--'" '--'" 1"\ '--'" Molecule t'/I IIi /I IIi II C\J /I I /I I /I II I /I U U U U ..::j-1"\ C\J 1"\ I 0 I cr I 0 I 0 0 '" 0 0 C\J:> 1"\:> 1"\:> C\J:> C\J:> C\J:> n-C4 Fl0 [X 'C><: -i-C4Fl0 D [>( [X n-C5F12 [>( [X [X n-CSF14 [>( 1[>( l><: 2-CF3C5 Fll [>( ex: ex IX 3-CF3C5 Fll l><: [>( C><: IX Operating conditions of n-hexadecane column given in Table 4 Major peaks are outlined Predicted isomers are crossed 0 r-i IIi III U C\J t----.--... (I') 0 IIi cow U C'. r-it----'--'" IIi I /I co /I C\J U '" 0 C><

PAGE 90

80 The experimental retention volumes and normal boiling points for perfluoroalkanes are given in Table 15 and are shown in Figure 7. The slope of log Vo versus Tb calculated for the normal saturated fluorocarbon compounds is 0.009 76 while the slope calculated for the perfluorohexane isomers is 0.0907.. The slope calculated for the perfluoropentane isomers is 0.0916 which is almost identical with the slope for the perfluorohexane isomers. Normal boiling points for the normal saturated fluorocarbons were calculated from the normal boiling point of C3Fe and the slope (0.00976) of the log Vo versus Tb plot for the normal fluorocarbon compounds. The results are given in Table 16. The calculated values are almost identical with the measured normal boiling points which indicates that the normal boiling points of normal saturated fluorocarbons can be estimated very accurately using this method. The normal boiling point calculated for is given in Table 17# The normal boiling points for the perfluorohexane isomers were calculated from the normal boiling point of and the slope (0.0907) of the log Vo versus Tb plot for the perfluorohexane isomers. The calculated results shown in Table 16 also agree very well with the experimental normal boiling pOints. 'I'he normal boiling points of the perfluoroheptane isomers were calculated from the normal boiling point of and the slope (0.0907) of the log Vo versus Tb plot

PAGE 91

TABLE 15 Retention Volumes and Normal Boiling Points for Fluorocarbon Compounds Volume, Vo, on n:-hexaLog10 Vo Compound decane Column,cm. 3 C3F S 29.6 1.47129 n-C 4 Fl.o 64.4 1.80889 n-C 5 F 12 129.7 2.11294 i -C5F 12 145.2 2.18808 n-C 6 F14 238.9 2 37822 2-C F3C5Fll 270.6 2.43233 2,3-(CF3)2C4 Fs 34 8.2 2.54183 n-C7F 1 6 427.8 2.63124 2-CF 3 C 6F1.3 469.7 2.6718 2 3 -CF3C6 F 1 3 487.0 2 68753 3-C2F 5 C 5 Fli 530.9 2.72501 2,4-(CF3)2C5 F 1 O 566.7 2.75335 2,3-(CF3)2C5 F 1 O 593.2 2.77320 2;2,3-(q F 3)3C 4 F7 660.6 2.81994 3,3-(CF3 )2C5F1O 701.6 2.84979 n-CSF1S 760.7 2.88121 *Va1ues from T. M. Normal Boiling;*Point c. 236.70 271.2 302.4 303.22 330.31 330 .90 332.12 355.66

PAGE 92

.... (]) ::r:; (') e u ... s::: e ::> rl 0 0 (]) s::: m u (]) 'tJ m :x: (]) ..c: I s::: s::: 0 (]) e ::> rl 0 :> s::: 0 rl .-P s::: (]) -P (]) p:; 82 100 100 10 200 300 900 Normal Boiling Point cC. Figure 7 Logarithm of Retention Volume on n-hexadecane Column Versus Normal .Boiling 'Point for Fluorocarbons

PAGE 93

TABLE 16 Calculated and Experimental Normal Boiling Points for Fluorocarbons Normal Saturated Fluorocarbons Compound 2-CF3C5 Fll 2,3-(CF3)2C4FS Experimental Normal* Boiling Point DC. 236.70 271.2 302.4 330.31 355.66 Perfluorohexane Isomers 330.31 330.90 332.12 Values from T. M. Reed(.s.Q.) Calculated Normal Boiling Point DC. 271.2 302.5 329.67 355.61 330.91 332.11

PAGE 94

84 TABLE 17 Estimated Normal Boiling Points for Fluorocarbons Compound Normal Boiling Point cc. n-C7F 1S 355.66 2-CF 3 C s F13 356.3 3-C F 3 CSF 13 356.7 3-C2 F 5 C 5 Fll 357.0 2,4-(CF3)2C5F1O 357.2 2,3-(CF3 )2C 5 F 1 O 357.7 2,2,3-(CF3 )3C4F7 358.1 3,3-(CF3)2C5 F 1 O 358.4 n-CSF1S 381.2

PAGE 95

85 of the perfluorohexane isomers. The slope of the log Va versus Tb plot for the perfluoroheptane isomers was assumed to be the same as the slope for the perfluorohexane isomers. The results are given in Table 17. Assuming that the experimental normal boiling points are correct, the normal boiling points estimated for the perfluoroheptane isomers and for n-CSF1S are probably accurate to + O.2DC.

PAGE 96

VI. SUMMARY All of the identified products found in this work can be explained by radical formation and recombination. The yields of fluorocarbon products show that C-F as well as C-C bond rupture occurs during the radiolysis of saturated fluorocarbons. Perfluoromethane is the most abundant of all fluorocarbon products and characterizes the radiolysis of saturated fluorocarbons. Fluorocarbons with branched structures are more abundant in the radiolysis products than straight-chain fluorocarbons which may indicate that branched fluorocarbon are more abundant than straightchain fluorocarbon radicals during radiolysis. This means that F atoms must be harder to remove from terminal CF3 groups than from other carbon atoms during radiolysis of saturated fluorocarbons. Perfluoromethane is the most stable to irradiation of all fluorocarbons studied. Carbon granules and carbon monoxide probably act as radical scavengers during the radiolysis of perfluoromethane mixtures and this causes extreme degradation of the otherwise stab+e perfluoromethane to heavier perfluoroalkanes. N o unsaturated or cyclic compounds were found in the radiolysis products of the saturated fluorocarbons. Three products that contain oxygen, C2FsO, C3FSO, and (C2F s)20 were identified and other unknown products are 86

PAGE 97

probably oxygen-containing fluorocarbons. The oxygen) containing products are probably formed by radical recombina-tion at the sample tube wall. No corrosion was found in the sample tubes and material balances show that essentially no fluorine is lost during radiolysis of saturated fluorocarbons. Molecular fluorine was not found in any of the radiolysis products. Abstraction reactions and disproportionation reactions which occur in the radiolysis of saturated hydrocarbons probably do not occur in the radiolysis of saturated fluorocarbons. Saturated fluorocarbons are more stable to irradiation than the analogous hydrocarbons. Retention volumes on the n-hexadecane column of eight of the nine perfluoroheptane isomers were determined and the normal boiling points of these compounds were estimated. The perfluoroheptane isomer, 2,2-(CF3)2Cs FlO, was not found in any radiolysis products and no other fluorocarbon products with neo structures were found.

PAGE 98

APPENDIC.ES

PAGE 99

APPENDIX 1 Dosimetry The absorbed dose rate distribution in tube 11 of the College of Engineering Cobalt 60 gamma ray source at the University of Florida was measured by benzene-water dosimetry following Johnson and Martin. (26) The phenol yield in neutral benzene-water solutions is a linear function of the total gamma dose up to a total dose of 70,000-80,000 rep; over the range 100-7,000 rep/min. it is independent of dose rate, benzene concentration (above 0.007M) and product concentration. The phenol concentration is determined from o the ultraviolet optical densities measured at 2,900A of a neutral and alkaline sample of the irradiated benzene-water solution. The neutral sample is prepared by diluting the irradiated benzene-water solution with distilled water and the alkaline sample by diluting the irradiated benzene-water solution with 0.06N sodium hydroxide solution. The optical densities were determined with a Beckman DK-2 Spectrophotometer at least 10 minutes after the alkaline sample was prepared. Under these conditions the phenol yield is 2.35 molecules/100 e.V. Distilled water saturated with benzene at room tempera-ture was sealed in polyethylene bottles 3.20 cm. 89

PAGE 100

90 high and approximately 2.5 cm. in diameter. The bottles had void spaces of 0.29 cm. at the top and 0.33 cm. at the bottom and contained approximately 5 ml. of solution. The bottles were stacked in the center of the sample space and irradiations of from 20 minutes in the high flux zone up to 60 minutes in the low flux zone were done in order to obtain as high a phenol concentration as possible in the dosimeter solution but still remain within the linear func-tion of phenol concentration versus total gamma ray dose. Two independent runs were made and the results are given in Table 18 and the a 'bsorbed dose rates are plotted versus distance from the bottom of the sample space in Figure 8. The absorbed dose rates determined using the poly-ethylene bottles are considered average values since the dosimeter solution occupied a large portion of the cross sectional area of the sample zone. Because the benzene-water dosimeter solution and the radiolysis products are not corrosive to aluminum, (27) aluminum tubing identical to the tubing used for holding the fluorocarbon samples was "coiled" as shown in Figure 9 and the vertical and radial absorbed dose rate distribution in tube 11 was measured. The results are given in Table 18 and Figure 8. Because each "coiled" aluminum tube contained less than 2 ml. of dosimeter solution, it was necessary to dilute these samples in order to have enough material to analyze in the

PAGE 101

91 TABLE 18 Absorbed Dose Rates Measured in Dosimeter Solution Date Measured 9-23-65 Polyethylene Bottle Numbered from B otto m of Sampl e Space 1 2 3 4 5 6 7 11 Coi ledll Alum inum Tube Numbered from Botto m of SamI21e SI2ace 1 2 3 4 5 6 Absorbed Dose Rates in Dosimeter Solution ( rads!min. ) Run 1 Run 2 2907 2866 3322 3339 2917 2685 1604 1 784 523 487 319 114 52 5 Absorbed Dose Rates in Dosimeter Solution radsLmin. 2 Inside Aluminum Outside Aluminum Tubes Tubes 2509 3833 3038 3586 2255 2851 2 443 2 2 82 1832 661 1359

PAGE 102

..--.. 0 S () '-"" (l) () ro 0.. CI) (l) rl 0.. s ro CI) tH 0 S 0 .p .p 0 p:) s 0 H tH (l) () e ro .p rfl o r! A 24 22 20 18 16 14 12 10 8 6 4 2 0 92 0 1000 () Polyethylene Bottles () Outside Aluminum Tubes Inside Aluminum Tubes o o 2000 3000 '4000 Absorbed Dose Rate in Dosimeter SO,lu tion (rads 0 /min 0 ) Figure 8 Absorbed Dose Rates Versus Distance from Bottom of Sample Space

PAGE 103

93 "Coiled" Shape of Aluminum Dosimeter Tubes Top View Outside Aluminum Tubes Top View Inside Aluminum Tubes ", ..... ---.......... ,;' ........... ( )----'A--'-./ 000 o 0 o 0 o 3.47 cm. o 0 _0 __ 1",,---0 0 0 1 o 0 o 0 2.69 cm. 0 0 o 0,_0 __ --'__ Figure 9 Aluminum Dosimeter Tubes

PAGE 104

94 The absorbed dose rates measured in the II coiled" aluminum are not very accurate due to the small dosimeter sample but they do show that there is considerable radial variation in absorbed dose rates through the sample space. Since the absorbed dose rates in the outside aluminum tubes were generally about 10 percent higher than the absorbed dose rates measured in the polyethylene bottles and the absorbed dose rates in the inside aluminum tubes were generally about 10 percent lower than the absorbed dose rates measured in the polyethylene bottles, the average absorbed dose rates in the polyethylene bottles will be assumed to be equal to the average absorbed dose rates in aluminum tubing. Since the absorption of Cobalt 60 gamma rays by aqueous, biological, most organic substances, and other substances of low atomic number is predominantlyby the Compton process, absorbed dose in the fluorocarbon samples may be calculated from the absorbed dose in the dosimeter samples by where, (Z/A) FC (Z/A) D DFC = absorbed dose in the fluorocarbon sample DD ="absorbed.dose in the dosimeter sample (1)

PAGE 105

95 (Z/A) = ratio of atomic number to the atomic weight for each material If the materials are not pure elements, mean values of Z/A must be calculated from, where, Z!A = W.(Z/A). l l (2) W. = weight fraction of ith element in the material. l For a pure chemical compound Z!A is simply the sum of the atomic numbers of the atoms present divided by the molecular weight of the compound. Values of Z/A for compounds of interest are listed in Table 19. For Equation 1 to be valid it is necessary that both materlals be exposed to the same beam of incident radia-tion, that both materials be in electronic equilibrium, and that the gamma rays are not appreciably attenuated by passing through the absorbing media. All these requirements are met in this work since the dosimeter solution and the fluorocarbon samples were irradiated in identical aluminum tubing and the volume of each material was quite small. The first requirement is not fully met, however, because in some irradiations the sample tubes were randomly placed in the sample space. In these samples the average absorbed dose rate distribution was used in calculating absorbed doses .and the accuracy of the absorbed doses are probably + 10 percent. More care

PAGE 106

96 TABLE 19 Ratio of Atomic Number to Atomic Weight for Various Substances Substance m aluminum 0.4818 H2O 0.5556 CF4 0.4773 C2FS 0.478 3 C3FS 0.4787 C 4 F 1O 0.4790 C5F12 0.4792 CSF14 0.4793

PAGE 107

97 was taken with the positioning of samples in subsequent irradiations and the absorbed doses in these samples are more accurate. In these samples the average absorbed dose rates in Figure 8 were increased 10 percent since the position of these sample tubes was the same as the position of t h e outside aluminum tubes used in the dosimetry measurements and the absorbed dose rates in the outside aluminum tubes were generally about 10 percent higher than the averag e absorbed dose rates. The accuracy of the absorbed doses in these samp les is probably + 5 percent. The average absorbed dose rate over the entire sample tube of samples that contained only one phase was found by graphically integrating the absorbed dose rate versus distance curve shown in Figure 8 over the length of the sample tube and dividing this integral by the length of the sample tube Because the intensity of radiation and the absorbed dose rate of a radioactive source decreases due to radio-active decay, the average absorbed d9se rate measured with the dosimeter solution was time corrected to the sample irradiation period by, where, D t = absorbed dose rate at start of sample irradiation Do = absorbed dose rate measured by dosimeter solution

PAGE 108

98 t = time in months from dosimeter irradiation to start of sample irradiation decay constant, 0.01096 month-1 for Cobalt 60 gamma rays The total absorbed dose was then calculated by, ( 4 ) where, D = total absorbed dose D t = absorbed dose rate at start of sample irradiation {\ = decay constant = 2.537 X 10-7 min.-1 t = minutes sample was exposed to radiation The correction from dosimeter solution to fluorocarbon sample (Equation 1) was also made. In the samples which contained both liqu i d and vapor phases, the liquid phase was assumed to occupy the lower portion of the sample tube and the vapor phase to occupy the upper portion of the sample tube.. The volume and weight of each phase was calculated from the liquid and vapor by, (5 ) ( 6)

PAGE 109

99 where, by, where, V L = volume of the liquid VG = volume of the vapor V T = total volume of the sample tube P L = density of the liquid P G = density of the vapor WT :=: total weight of sample The height of liquid in the sample tube was calculated HL = height of the liquid phase from the bottom of the sample space r = inside radius of sample tube The average absorbed dose rate in each phase w a s calculated as previously described and the average absorbed dose rate for the entire sample was found by, (8) where, DAVE = average absorbed dose rate for entire sample

PAGE 110

100 D LAVE = average absorbed dose rate for liquid phase D average absorbed dose rate for vapor phase GAVE = = weight of liquid WG = weight of vapor W T = total sample weight The total absorbed dose in the sample was calculated as before. In the samples which contained aluminum powder and carbon granules the solid phases were assumed to have no effect on the absorbed doses for the samples and the absorbed doses were calculated based only on the CF4 present in the sample tubes.

PAGE 111

APPENDIX 2 Response of Thermal Conductivity Detector Cells The area response of the thermistor detector used in the chromatographic analysis of the irradiated samples was found to be neither proportional to the moles of solute in the carrier gas nor to the weight of solute in the carrier gas. The response was found to be linear with amount of material analyzed, however; e.g., injecting twice the amount of CF4 would result in twice the area response from the detector. The response of the detector was also a function of its temperature and of the carrier gas flow rate. The response was greatest at low detector temperatures and it increased with decreasing carrier gas flow rates. The type of carrier gas used also affected the response of the detector. Since both hydrogen and helium gas were used in analyzing the irradiated samples, the response of the detector cell was calibrated with hydrogen and with helium carrier gas. Even though the actual response of the detector depended upon the c.onditiQn imposed, fluorocarbon mixtures analyzed with different conditions of detector temperature and carrier gas flow rates always gave the same area per-centages for the 'components of the mixture. 101.

PAGE 112

102 The chromatographic .area response of the detector to measured quantities of and other compounds was measured and the results are given in Tables 20 and 21 and the area response is plotted versus molecular weight in Figures 10 and 11. Included in the fluorocarbon com-pounds used for calibrating the response of the detector were an unsaturated compound, cyclic compounds and various saturated compounds. Perfluorotributylamine, perfluorodiethylether, air, carbon monoxide, and argon were the other compounds used in calibrating the response of the detector cell. The experimental response given in integrator counts from the Perkin Elmer 194B printing integrator per gram of compound injected are given in Tables 20 and 21 for the compounds used in the calibration. In FigureslO and 11 the experimental response was plotted versus the molecular weight of the compound analyzed and smooth curves were drawn through the experimental points. The detector cell response for fluorocarbons and fluorine-containing compounds was taken from the smooth curves. The actual response values taken from the curves are given in Tables 20 and 21 where they are normalized to 1.000 for CF4 It is clearly seen from the tables and figures that \ the response of the detector cell to solutes in the carrier gas decreases with increasing molecular weight.

PAGE 113

TABLE 20 Area Response of Detector Cell and Correction Factors for Calculating Weight Fractions -H2 Carrier Gas Molecular Area/gm.* Area/gm.* Normalized Correction ComI2ound Weight EXI2erimental From Fig.l0 CF4=1.000 Factor CO 28 1.000 0.990 2.292 0.436 air 29 0.922 0 .936 2.167 0.461 Argon 40 0.703 0.708 1.639 0.610 CO2 44 0.648 1.269 0.788 CF4 88 0.432 0.432 1.000 1.000 C 2 F 4 100 0.406 0.940 1.064 C2F 6 138 0.342 0.349 0.808 1 .238 J-l 0 C3F 6 150 0.342 0.335 0.775 1.290 \>l C2F 6 0 154 0.330 0.764 1.310 C3Fs 188 0.299 0.295 0.683 1.464 CYCC 4 F S 200 0.282 0.283 0.655 1.527 C3F s O 204 0 .280 0 .648 1.543 C 4 F 1 O 238 0.256 0 .265 0.613 1.631 C 4 F1OO 254 0.260 0.602 1.661 C5F12 288 0.250 0.579 1.727 C 6F14 338 0.242 0.560 1.786 C7F16 388 0.244 0.238 0 .551 1.815 (CF3)2Cy-C6Flo 400 0.232 0.237 0.549 1.821 CSF1S 438 0.237 0.549 1.821 CSF20 488 0.247 0.232 0.537 1 .862

PAGE 114

TABLE 20 (Continued) Molecular Area/g m. Area/gm. Normalized Correction Com:Qound Weight EX:Qerimental From Fig. CF4=1.000 Factor CloF22 538 0.228 0.528 1.894 CllF24 588 .0.225 0.521 1.919 Ca.2F26 638 0.222 0.514 1.946 C12F27N 671 0.217 0.220 0.509 1.965 Values are in integrator counts per gram of solute x 10-8

PAGE 115

TABLE 21 Area Response of Detector Cell and Correction F actors for Calculating Weight Fractions -He Carrier G a s Helium Carrier Gas at 17.9 cc./min. at 23C. and One Atmosphere with Detector at 44C Compound air argon C 02 CF4 C2F6 C3F6 C2F6 0 C3Fs C3FsO C4F10 (C2F5)20 C5F12 C 6 F14 C7F 16 CSF1S CSF20 CloF22 CllF24 'C12F 26 C1 2F27N Molecular Weight 29 40 44 88 138 150 154 188 204 238 254 288 338 388 438 488 538 588 638 671 Area/gm.* Experimental 14.09 10.81 6.86 5.61 5.75 4.95 4.25 4.16 3.93 3.57 3.02 Area/ gm. -x From Fig.ll 14.09 10.80 10.05 6.86 5.60 5.36 5.30 4 '.83 4.66 4.35 4.22 4.00 3.76 3.60 '3.48 3.38 3.27 3.17 3.08 3.02 Normalized C F4=1.000 2.054 1.574 1.465 1.0010 0.816 0".781 0.773 0.704 0.679 0.634 0.615 0.583 0.548 0.525 0.507 0.493 0.477 0.462 0.449 0.440 t f 1 t X 10-s Values are in integrator coun s per gram 0 so u e Correction Factor 0.487 0.635 0.683 1.000 1.225 1.280 1.294 1.472 1.577 1.626 1.715 1.824 1.906 1.971 2.030 2.098 2.164 2.227 2.273 J-l o Vl

PAGE 116

1.2 S Cd H 1.0 0 H 0) p.. rIl .p 0.8 s:: ::sen 01 00 rl H K 0 .pO) Cd.p 0.6 H ::s Q()rl 0) 0 .pC!) s:: HIi-l 0 0.-4 s:: 'M S Cd H 0 0.2 """-Cd 0) H <:J:! CO o 100 200 300 400 Molecular Weight Figure 10 500 600 700 Area Response of Thermistor Detector Cell -H2 Carrier Gas J--l o (j)

PAGE 117

s co H 0 H <: 0 -P
PAGE 118

To report a chromatographic analysis in weight fractions the area fractions of the compounds in the analysis must be "weighted" by correction factors which are the reciprocals of the normalized' response values of the compounds in the analysis. Correction factors for various compounds ,are given in Tables 20 and 21. The weight fraction of each component in the analysis is obtained by dividing its weighted area fraction by the sum of all the weighted area fractions in the analysis. EXAMPLE CALCULATION Hydrogen Carrier Gas Area Weighted Area Component Fraction Fraction CF4 0.500 0.500xl.000=0.500 C2F e 0.300 0.300xl.238=0.371 CsFs 0.200 0.200xl.464=0.293 1.164 Weight Fraction 0.430 0.319 0.252

PAGE 119

1. 2. 3. 4. 5. 6. 8. 9 10. 11. 12. 13. BIBLIOGRAPHY J. C. Mailen, Ph.D. Dissertation, University of Florida (1964). L. Kevan and P. Hamlet, J. Phys. Chem., 42, 2255 (1965). J. H. Simons and E. H. Taylor, J. Phys. Chem., 63, 636 (1959). R. E. Florin, L. A. Wall, and D. W. Brown, J. Res Nat. Bur. Stds., 64 A 269 (1960). R. F. Heine, J. Phys. Chem., 2116 (1962). M. B. Fallgatter and R J. Hanrahan, J. Phys. Chem., 69, 2059 (1965). D. R. MacKenzie, F. W. Bloch, and R. H. Wismall, Jr., J. Phys. Chem., 6 9 2526 (1965). H. N. Rexroad and vi. Gordy, J. Chem. Phys., 30, 399 (1959). J. H. Golden, J. Polymer Sci., XLV, 534 (1960). T. M. Reed III, J. Chromatog 9, 419 (1962). T. M. Reed III, J. T. Walter, R. R. Cecil, and R. D. Dresdner, Ind. Eng. Chem. 51, 271 (1959). T. M. Reed III, J. C Mailen and W. C Aske w Final Report to the United States Atomic Energy Commission on Contract No. AT-(40-1)-2846, September (1965). R. D. Dresdner, T. M. Reed III, T. E. Taylor, and 'j J. A. Young, J. Org. Chem., 25, 1464 (1960). 14. J. A. Young, T. M Reed III, K. C. Ramey, and G. A. Crowder, First Annu a l Report to the Natioval Science Foundation on Grant NSFG 14591, September (1961). 109

PAGE 120

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 110 J. C. Mailen, T. M. Reed III, and J. A. Young, Anal. Chern., 36, 1883 (1964). H. Purnell, Gas Chromatography, John Wiley and Sons, Inc., New York (1962). R. L. Pecsok, Principles and Practice of Gas Chromatograbhy, John Wiley and Sons, Inc., New York ( 1 959) M. B. Fallgatter and R. J. Hanrahan, Personal Communication. M. M. Bibby and G. Carter, Trans. Faraday Soc., 59, 2455 (1963). R. Dacey and J. W. Hodgins, Can. J. Research, 28, 173 (1950). A. K. Kuriakose and J. L. Margrave, J. Phys. Chern., 2772 (1962). H. M. Tenney, Anal., Chern. 30, 2 (1958). L. J. Sullivan, J. R Lotz, andC. B. Willingham, Anal Chern., 28, 495 (1956) D. H. Desty and B. H. F. Whyman, Anal. Chern., 29,320 (1957). T. M. Reed III, Anal. Chern., 30, 221 (1958). T. R. Johnson and J. J. Martin, Nucleonics, 20, 83 (1962). F. W. Bloch, F. J. Haasbroek, and D. R. MacKenzie, Nucleonics, 22, 77 (1964) J. W. T. Spinks and R. J. Woods, An Introduction to Radiation Chemistry, John Wiley and Sons, Inc., New York (1964). A. J. Swallow, Radiation Chemistr Pergamon Press, New York 1 960 T. M. Reed III, Fluorine Chemistry Vol. V, Ed. by J. H. Simons, Academic Press, New York (1964).

PAGE 121

BIOGRAPHICAL SKETCH William Crews Askew, son of William King Askew and Juanita DeLoach Askew, was born January 5, 1940, at Athens, Alabama. In June, 1958, he was graduated from Auburn High School, in Auburn, Alabama In June, 1958, he enrolled in Auburn University where he received the degree of Bachelor of Science in Chemical Engineering in June, 1962. He also received an appointment as Second Lieutenant inthe United States Army Reserve. Following summer employment with Shell Oil Company in New Orleans, Louisiana, he enrolled in the Graduate School of Auburn University in September, 1962. He worked as a teaching assistant in the School of Chemistry until June,. 1963, and received the degree of Master of Science in August, 1963. In September, 1963, he enrolled in the Graduate School of the University of Florida. Until the present time he has worked as a graduate research assistant in the Department of Chemical Engineering and has worked toward the degree of Doctor of Philosophy. William Crews Askew is married to the former Casinel Lavonne Tisdale and is the father of one son. He is a member of the American Institute of Chemical Engineers, Phi Lambda Upsilon, Tau Beta Pi, Phi Kappa Phi, and Lambda Chi Alpha. 111

PAGE 122

This dissertation was prepared under the direction of the chairman of the candidate's supervisory committee and has been approved by all members of that committee. It was submitted to the Dean of the College of Engineering and to the Graduate Council, and was approved as partial fulfillment of the requirements for the degree of Doctor of Philosophy. April 23, 1966 7 Dean, College of Engineering Dean, Graduate School Supervisory Committee: Chairman /