• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Illustrations
 Introduction
 Apparatus
 Preparation of fluoride salts containing...
 Calculations
 Chemical and radiochemical purity...
 General exchange studies
 Kinetics of the exchange react...
 Discussion of results
 Conclusions
 Bibliography
 Biographical sketch
 Copyright














Title: F¹8 exchange between fluorocarbons and some fluorine-containing compounds.
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00091326/00001
 Material Information
Title: F¹8 exchange between fluorocarbons and some fluorine-containing compounds.
Series Title: F¹8 exchange between fluorocarbons and some fluorine-containing compounds.
Physical Description: Book
Creator: Gens, Theodore A.
 Record Information
Bibliographic ID: UF00091326
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000424014
oclc - 11069281

Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
        Page iii
        Page iv
    Table of Contents
        Page v
    List of Tables
        Page vi
    List of Illustrations
        Page vii
        Page viii
        Page ix
    Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
    Apparatus
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
    Preparation of fluoride salts containing F18
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
    Calculations
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
    Chemical and radiochemical purity and error
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
    General exchange studies
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
    Kinetics of the exchange reactions
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
    Discussion of results
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
    Conclusions
        Page 110
        Page 111
        Page 112
    Bibliography
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
    Biographical sketch
        Page 118
        Page 119
    Copyright
        Copyright
Full Text











F1" EXCHANGE BETWEEN FLUOROCARBONS AND

SOME FLUORINE-CONTAINING COMPOUNDS









By

THEODORE A. GENS


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









UNIVERSITY OF FLORIDA
AUGUST, 1957












ACKNOWLEDGMENTS


The successful performance of these exchange studies was made

possible by the guidance and aid of Dr. John A. Wethington, Jr. In

addition, he frequently applied his skill in laboratory technique to

overcome some of the difficult problems which otherwise might not have

been solved. The burdens introduced by the ORINS Fellowship arrange-

ment were cheerfully accepted by Dr. Wethington.

Dr. A. R. Brosi of the Chemistry Division of the Oak Ridge

National Laboratory (ORNL) watched over the problems of a radiochemical

nature, and his aid in such matters was invaluable. In addition, Dr.

Brosi readily considered and advised on all the problems which arose.

His careful, scientific approach often led to early solution of these

problems.

Dr. E. R. Van Artsdalen, who shared in the supervision of the

early part of this work, was very helpful in planning this research

program and in making arrangements for the design and construction of

the apparatus and instruments.

The personnel of the Fluorine Research Center at Reed Labora-

tory, University of Florida, very generously supplied samples of rare

fluorine-containing compounds. Most of these compounds were pro-

duced by the Electrochemical Process.

The competent aid in the laboratory by D. E. LaValle of the











Analytical Chemistry Division of ORNL made possible the survey with HF

and various inorganic fluorides.

Proton bombardments were planned and supervised in the ORNL

86-inch cyclotron by Dr. J. L. Need of the Applied Nuclear Physics

Group of the Electronuclear Research Division.

MeasuremenU of surface areas were made by the B.E.T. method by

P. K. Melroy of the Special Analytical Services Department of the

Oak Ridge Gaseous Diffusion Plant.

Colorimetric and pyrohydrolysis quantitative analyses of

fluoride ion were performed by the Special Analyses Group of the

Analytical Chemistry Division of ORNL.

Mass spectrometer analyses of fluorocarbons were performed by

J. C. Horton, G. Howell and F. Jones of the Mass Spectrometer Depart-

ment of the Oak Ridge Gaseous Diffusion Plant.

Routine spectrographic analyses were performed by the Spectro-

graphic Group of the Analytical Chemistry Division of ORNL. A

modified Harvey method was employed by M. Murray to analyze for silicon.

In the early search for understanding of the experimental data,

some complicated calculations were coded by E. C. Long for the ORACLE.

Miss B. J. Osborne performed the necessary operations.

Inorganic fluorides were prepared by B. J. Sturm under the

direction of L. G. Overholser of the Chemistry Division of ORNL.

The Instrumentation and Controls Division and the instruments


iii











section of the Chemistry Division and the glass and metal-working shops

at ORNL made essential contributions to this work.

This work was supported by the Oak Ridge Graduate Fellowship

Program of the Oak Ridge Institute of Nuclear Studies. Work done at

the University of Florida was supported by the Office of Naval Research.

The greater part of this work was performed at the Oak Ridge National

Laboratory, which is operated for the U. S. Atomic Energy Commission

by the Union Carbide Nuclear Company.













TABLE OF CONTENTS


Page
LIST OF TABLES ... ....... ....... ......................... vi

LIST OF ILLUSTRATIONS ....................................... vii

Chapter

I. INTRODUCTION ..........................................

II. APPARATUS .............................................. 7

III. PREPARATION OF FLUORIDE SALTS CONTAINING F18 ........... 26

IV. CALCULATIONS ...................................... 34

V. CHEMICAL AND RADIOCHEMICAL PURITY AMD ERROR ............ 43

VI. GENERAL EXCHANGE STUDIES .............................. 54

High Temperature Survey
Low Temperature Survey
Survey of Hydrogen Fluoride and Various Inorganic
Fluorides for Indications of Exchange with SF6 and
Fluorocarbons

VII. KINETICS OF THE EXCHANGE REACTIONS ............,........ 83

VIII. DISCUSSION OF RESULTS ................................. 91

Fluorocarbons
Silicon Tetrafluoride
Other Work
Fluorocarbon Exchange Mechanisms
Suggestions for Future Work

IX. CONCLUSIONS ................................... ...... 110

BIBLIOGRAPHY ........... .......... .................... 113




















LIST OF TABLES


Table Page

1. Six Methods of Preparing Radioactive F8 ................ 29

2. Error ............................................. 50

3. Data from Studies Having Experimental Difficulties or

Negative Results ....................................... 60

4. Exchange Studies between HF and Some Fluorine-Containing

Compounds .............................................. 79

5. Exchange Studies between Some Inorganic Fluorides

and Fluorocarbons at 5000 ............................... 82

6. Kinetics of Exchange at Constant Temperature ............ 83

7. Heats of Reaction from Survey Studies ................... 92













LIST OF ILLUSTRATIONS


Figure Page

1. Exchange System, Sketch ............................... 8

2. Exchange System with Inconel Reactor Tube ............. 9

3. Constant Temperature Exchanger ..................... 13

4. Exchange System with Pyrex Constant Temperature

Exchanger ............................... ....... 14

5. Solenoid-Activated Circulation Pump ................... 15

6. Exchange System Plus Auxiliary Glass System ........... 17

7. All Nickel System for Preparing Inorganic Fluorides,

Sketch ............................... ........ ..... 19

8. All Nickel System for Preparing Inorganic Fluorides ... 20

9. Inorganic Fluoride Exchange Tube ...................... 21

10. Counting and Temperature Control, Sketch ............. 23

11. Instruments ................. ........................ 24

12. Example of Survey Experimental Data ............. 39

13. Gamma Spectrometer Plot of F18 Radiations ............. 47

14. Pressure-Temperature Behavior of C F above CsF ...... 51
4 10
15. Fraction of Exchange Occurring with C3F6 over the

Alkali Fluorides .............................. .... 55

16. Fraction of Exchange Occurring with C3F6 over NaF;

Three Runs ........................................... 57


vii










LIST OF ILLUSTRATIONS Continued


Figure Page

17. Fraction of Exchange Occurring with C4F10 over CsF

and F .................. ... .... .. ....... ..... ........... 58

18. Fraction of Exchange Occurring with C3F6 and C F10

over CsF .. .. .......... .............................. 59

19. Fraction of Exchange Occurring with C3F6 over the

Alkali Fluorides at Lower Temperature ................ 63

20. Fraction of Exchange Occurring with CLF10 over the

Alkali Fluorides at Lower Temperature ................ 64

21. Fraction of Exchange Occurring with (C2F5)20 over

the Alkali Fluorides ................................. 65

22. Fraction of Exchange Occurring with CF4 over RhF

and CsF ....... ................ ....... .. 66

23. Fraction of Exchange Occurring with C3F6 over CsF

by Three Techniques .................................... 67

24. Pressure-Temperature Behavior of CF4 and CF3H over

CsF .............. .... ...................... .. ...... .. 68

25. Pressure-Temperature Behavior of SiF4 Alone and Above

Alkali Fluorides ..................................... 70

26. Results obtained by Passing SIF4 over Alkali Fluorides. 71

27. Pressure-Temperature Behavior of SF6 Alone and Over

the Alkali Fluorides ................................ 73


viii










LIST OF ILLUSTRATIONS Continued


Figure Page

28. Low Temperature Summary Plot .......................... 75

29. Pressure-Temperature Behavior of Argon and Five Gaseous

Fluorine Compounds over the Alkali Fluorides .......... 76

30. Kinetic Studies ................................... 84

31. Kinetic Studies of Exchange Occurring with C3?6

over CsF ............................................. 86

32. Rate Data Taken from Fig. 31 .......................... 88

33. Effect of Duration of CsF Preparation ................. 89

34. Fraction of Exchange Occurring with C3F6 over CsF ..... 94

35. Number of Lattice Layers That Would Be Required to
Account for Values of N Observed in Study AcII ........ 96

36. Second Order Effects with C?6 ........................ 105













CHAPTER I


INTRODUCTION


The existence of the 112 minute fluorine eighteen isotope was

reported by Snell in 1937. Several studies have been made using this

radioactive isotope to follow exchange reactions. Since F18 is a

positron emitter, the 0.511-Mev. annihilation gamma radiation provides

an excellent means of detecting this isotope. Practically all work up

to the present has been in homogeneous, usually gas phase, systems.

Dodgen and Libby made an early study of rates of exchange in the gas

phase between F2 and HF. They found that little exchange occurred

below 2000 C 2 Above this temperature exchange occurred, but this ex-

change was probably catalyzed by metal fluorides on the wall of the

vessel. Nonexchange was explained by the absence of unfilled electron

levels in the-higher shells of fluorine. Such unfilled levels were

thought to allow formation of intermediates of the general formula HX3

with other halogens and the corresponding hydrogen halides. Rapid

exchange has been observed in these systems.

Katz and co-workers have made several studies using F18

Rogers and Katz studied exchange between HF and several interhalogen

compounds .3 They found that the room temperature exchange reactions

between liquid HF and several liquid interhalogen compounds were


- 1 -





-2-


essentially complete in ten minutes while the same reactions in the gas

phase were essentially complete in three minutes. These results made

it appear feasible that an ionic mechanism could be operative in the

liquid phase and that a mechanism postulating an intermediate complex

could explain the gas phase reactions. Essentially zero exchange was

observed between HF and SF6 or CC12P2 and between CIF3 and F2.

Bernstein and Katz studied the gas phase exchange between

interhalogen compounds and fluorine and found essentially no exchange

below 100 4 Above 100 a measurable rate of exchange was observed,

from which it was possible to propose mechanisms for the exchange re-

action. Small but observable exchange between interhalogen compounds

and several metallic fluorides was reported.

Boggs, Van Artsdalen, and Brosi found no exchange in the gas

phase between HF and fluorinated methanes at 500 over one hour.5

Although no systematic studies have been made of heterogeneous

exchange between HF and metallic fluorides, such exchange has been

observed in several cases.'33

Consideration of other systems in which similar exchange

studies could be performed points out the wide applicability and the

advantages of the fluorocarbon-inorganic fluoride system. Exchange

studies between hydrocarbons and inorganic hydrides, using deuterium

and tritium, are feasible but limited to the few stable inorganic

hydrides. Many exchange studies between deuterium and hydrocarbons on

the surface of various catalysts, such as the cracking catalysts, have





-3-


been made by Taylor and co-workers and others.6,7 These studies, which

were designed primarily to look into surface effects, did not involve

investigation of isotopic distribution within the crystal. Thus, they

differed fundamentally from the studies reported in this manuscript.

However, a series of papers by Wright and Weller describe a more

analogous system.8 Wright and Weller studied isomerization and hydro-

genation of unsaturated hydrocarbons over BaH2 and Cal2. They attri-

buted the catalytic activity to production of dual metal-metal hydride

sites by removal of hydrogen during evacuation at 200 to 300. Ex-

change between hydrogen and deuterium was also found to be catalyzed

by BaH2 and Cal2. Deuterium over Cal2, carefully evacuated at 200,

approached isotopic equilibrium with the hydride within a few hours.

Diffusion of hydride ions between the interior and the surface was

advanced as the explanation of these results. If any reaction mechanism

as rapid as the one which will be discussed with the fluorocarbon-

alkali fluoride system was operative here, its effect was not apparent

at 2000 with the technique used.

Several heterogeneous exchange reactions between completely9

and partially9'10O11 halogenated hydrocarbons and inorganic halides

have been studied. Blau and Willard observed rapid exchange of

chlorine atoms at room temperature between CC14 or partially chlorin-

ated hydrocarbons and AC13 9 Observations made by Kistiakowsky and

Van Wazer0 in the exchange of CH3Br with BaBr2 and AlBr3 may have

resulted from mechanisms similar to those apparently operative in the






-4 -


fluorocarbon-alkali fluoride system. They observed activation energies

of 12 kcal mole-1 with BaBr2 and 4.6 kcal mole-1 with AlBr While

there are innumerable inorganic chlorides, bromides, and iodides which

may be used in exchange studies, fluorocarbons are the only fully

halogenated hydrocarbons in which such great numbers of compounds are

stable. A great many partially chlorinated, brominated, or iodinated

hydrocarbons, while stable, do not have the high vapor pressure of

fluorocarbons and cannot be conveniently studied in the gas phase.

Many such systems have been studied in the liquid phase.2 )l2'1314

In none of these studies have observations resembling those discussed

in this manuscript been made.

Winter found that exchange of oxygen gas with some inorganic

oxides involved subsurface lattice ions, and he was able to correlate

his exchange results with the semiconducting properties of the oxide.15

Kolthoff and O'Brien found that quite rapid exchange occurred between
16
Br2 gas and solid AgBr. The mechanisms involved in such inorganic

systems may resemble those involved in exchange between fluorocarbons

and alkali fluorides.

Fluorocarbons have great thermal stability and resistance to

chemical reaction.17 The dissociation energy of CF4 is reported as
-1 18 -1 a19
130 kcal mole compared with 102 kcal mole for CH." These

properties make fluorocarbons valuable in many unique applications,

but also make it impossible to apply synthesis or degradation re-

actions similar to those normally applied in organic chemistry. To







-5-


obtain significant reaction with fluorocarbons, hot tube reactions,

often under high pressure, are frequently resorted to. In hot tube

reactions the nature of the surface usually affects the reaction. It

is common practice to pack the hot tube with an inert material which

supplies a surface for reaction and possibly produces other unknown

effects.

Many reactions with fluorocarbons which are very feasible

thermodynamically cannot be carried out, even at high temperatures.

There exists great need for catalysts capable of lowering the acti-

vation energy sufficiently to allow smooth low temperature reactions

producing few products in high yield, both for research use and for

successful commercial application of the unique properties of fluoro-

carbons. Most of the science and technology of fluorocarbons has

developed over the last few years. Starting with a small sample of

fluorocarbon supplied by J. H. Simons in 1941, a group at Columbia

University developed the solid and liquid fluorocarbons needed in the

Manhattan project O' At the same time a different method of fluoro-
122
carbon production was being developed at Johns Hopkins University.

Since the field is so new, there is very little published work to

serve as a guide in searching for materials to catalyze fluorocarbon

reactions. A study of the exchange of fluorine between fluorocarbons

and inorganic fluorides using F18 offered the possibility of gathering

much information concerning the effect of various salts in exchange

reactions. It was thought that if salts were found which exchanged






-6-



with fluorocarbons, information thus gathered would be helpful in the

search for catalysts for use in other fluorocarbon reactions.













CHAPTER II


APPARATUS


Photographs are used to show the instruments and the two vacuum

racks containing the apparatus used in this work. In addition, sketches

are included to aid in interpreting the photographs.

Figure 1 is a sketch showing the essential parts of the system

used to study the exchange of fluorocarbons, SIF4, and SF6 with alkali

fluorides. Figure 2 is a photograph of this apparatus. The multiple

unit clam-type furnace has been removed to show the Inconel reactor

tube. Dimensions of the Inconel portion of the reaction tube were:

length, 13.0"; O.D., 0.375"; thickness, 0.025". A Chromel-Alumel

thermocouple was soldered to the middle of the reactor tube with high

melting silver solder. Copper cooling coils through which water was

circulated were soft soldered on the portions of the reactor tube which

extended out of each end of the furnace. These cooling coils prevented

the melting of the soft solder seal on the sleeves directly beneath

the coils. The sleeves connected the Inconel tube to Housekeeper seals,

and thus to the rest of the system, which was made of Pyrex. Inconel

metal was most satisfactory as a reactor tube material because of its

resistance to reaction at high temperature and its unusually poor con-

ductance properties, which made it simpler to protect the Housekeeper

seals. Fans, one of which is visible in Fig. 2, were also used in


-7-









-8-


t
To Manifold,
Pumps, and
System for
Handling Gases


I Furnace I
SReactor
I Furnace I


ti


A Capillary Manometer
B Heavy Lead Shield
C Scintillation Counter
D Solenoid Activated
Circulation Pump


Fig.1. EXCHANGE SYSTEM, SKETCH









-9-


Fig. 2. Exchange System with Inconel Reactor Tube.






- 10 -


cooling the exit and entrance. A split stainless steel sleeve of 1.250"

O.D., 0.250" thickness, and 7.0" length surrounded the Inconel reactor

tube inside the 8.0" long heating coils in the furnace. The large heat

capacity of this steel sleeve prevented rapid temperature fluctuations.

The Chromel-Alumel thermocouple seen in Fig. 2 below the Inconel tube

was inserted into a hole drilled half-way through the length of the

stainless steel sleeve. This thermocouple activated the temperature

controller, as outlined in Fig. 10. This system was built to be used

at temperatures as high as 10000, and at times was actually used at

temperatures of nearly 7000.

The spiral in Fig. 2 was of 4 mm. O.D. Pyrex. It provided the

flexibility needed to open the system at the reactor tube. The stop-

cock just above the spiral was added before the kinetic studies were

made. A cold finger was located beneath the spiral. As the gas entered

the cold finger it was filtered through a type D sintered Pyrex frit.

This frit and a similar frit above the pump on the left side of Fig. 2

isolated the counter from the reactor tube. A well type Nal scintil-

lation counter beneath the lead shield counted the circulating gas.

The solenoid-activated pump just above the counter, and the flowmeter

above the pump, are described below. The capillary mercury manometer,

part of which is visible on the right of Fig. 2, was connected to the

gas system by a capillary Pyrex line which entered above the flowmeter.

Beside this manometer line was the exit to the auxiliary system. At

the bottom of the exchange system was the tap through which radioactive






- 11 -


gas was withdrawn and returned in the counter calibration procedure

which is described under Calculations.

Figure 3 shows a drawing of the Pyrex constant temperature

exchanger used in the kinetic studies. This exchanger was placed in the

position occupied by the Inconel reactor tube in Fig. 2. Figure 4 shows

the exchanger in position. The vapor bath and furnace have been lowered

to show the coil in which the entering gas was preheated. This 4 mm.

0.D. preheater coil was about one yard in length and held 10 cc., nearly

one-tenth of the volume of the system. A Pt, Pt-10% Rh thermocouple,

which was withdrawn for photographing in Fig. 4, was inserted through

the condenser for the experiments, as shown in Fig. 3. The temperature

of the vapor or ice bath was measured from the thermocouple by a Rubicon

slide-wire potentiometer with which the E.M.F. was estimated to the

nearest 0.0001 mv. A 10-inch Pyrex tube with a 50/50 ground glass joint

(the same size tube and joint as in the vapor bath in Fig. 3), wound

with sheet asbestos and Nichrome wire, served as a furnace when it re-

placed the vapor bath. The temperature controller was connected to this

furnace.

The circulation pump and flowmeter are shown in Fig. 5. Similar

pumps are described in the literature.23,24 Other satisfactory solu-

tions to the problem of obtaining circulation in similar systems have

been reported.25,26 This pump circulated the gas with very small hold-

up through a complete cycle in about a minute. A small fan behind the

pump kept the solenoids cool. A ground glass piston filled with soft






- 12 -


Legend for Fig. 3


A. Gas Exit

B. Gas Entrance

C. CsF Salt

D. Porous "D" Pyrex Frit

E. 12/30 Ground Glass Joint

F. Thermocouple

G. Water Cooled Condenser

H. 50/50 Ground Glass Joints

I. Vapor Bath

J. Liquid Used for Vapor Bath

K. Furnace





- 13 -


H






SI


C--
D







Fig. 3. CONSTANT TEMPERATURE EXCHANGER










-14-


Fig. 4. Exchange System with Pyrex Constant Temperature Exchanger.






















'Ii

I s
I ,~
Is


-15-





i.




K-:


Ir ,




A)


Fig. 5. Solenoid-Activated Circulation Pump.


- '-a)


"I






- 16 -


iron operated inside the solenoids. Current to the solenoids was

turned on and off at the desired rate by an ordinary laboratory stirring

motor equipped with a small cam which operated a microswitch. Several

of these pieces of equipment were operated from a small control box be-

hind the lead shield in Fig. 6. Above the pump a Fisher-Porter

variable-area flowmeter was mounted with Apiezon W wax. Between the

flowmeter and pump was a type D sintered Pyrex frit.

Figure 6 covers the same area seen in Fig. 4 and also shows

some of the manifold and gas handling system.. A gas storage bulb was

attached to the manifold. On the right was the mercury diffusion pump

and the vacuum line leading to the Welch pump beneath the vacuum.rack.

In Fig. 7 are sketched the essential parts of the system which

was used to produce salts containing F18 by exchange with HF. This

same system was used for exchange studies between gaseous fluorine

compounds and BF, and between gaseous fluorine compounds and NiF2,

CuF2 PdF2, and CsF. Figure 8 shows a photograph of the same area.

The soda lime trap to prevent HF from entering the pump is also visible

on the right. The fluorothene tube, in which the products from the

LITR bombardment were placed for exchange with liquid HF, is attached

below the filter in Fig. 8. The Booth-Cromer pressure gage27 allowed

measurement of the vapor pressure of the HF. The diffusion pump, which

is on the left side of Fig. 8, was not used when HF vapor was present.

The nickel tube in which the inorganic fluorides were exchanged

with HF and the nut by which it was attached are shown in Fig. 9. A





-17-


R'9


Lb :"


Fig. 6. Exchange System Plus Auxiliary Glass System.


C


r~;--
4'


PI 7


~l~ij~


~J1 ; ;
i4
E4:/IY






- 18 -


Legend for Fig. 7


A. BF Container

B. Threaded Nut for Attachihg Inorganic Fluoride Exchange Tube

C. Thermocouple and Potentiometer

D. Furnace

E. Filter

F. Threaded Nut for Attaching Tube Containing Reactor Products

(Including F18

G. Booth-Cromer Pressure Gage














PUMP -< -0-
PUMP


A E
B





SF
D


Fig.7. ALL NICKEL SYSTEM FOR PREPARING INORGANIC FLUORIDES, SKETCH








































Fig. 8. All Nickel System for Preparing Inorganic Fluorides.






-21 -


L i
Xqqmmqpiim -i i







II










.4
J I i


Fig. 9. Inorganic Fluoride Exchange Tube.






- 22 -


copper washer was used to insure a vacuum tight seal. An inch deep

thermocouple well in the bottom of the tube is not visible.

No reaction of HF was observed to occur on the exposed parts of

this nickel system, except that NiF2 was formed in the exchange tube at

elevated temperatures. The Fulton bellows type valve had inert Teflon

liners.

A sketch of the counting and temperature control apparatus is

shown in Fig. 10. The impulses from the photomultiplier tube and small

preamplifier were fed into the linear count rate meter. The counting

rate was recorded on the Brown chart recorder, along with the temper-

ature, which was measured by the Chromel-Alumel thermocouple on the

Inconel reactor tube. A similar thermocouple activated the temperature

controller which controlled the current to the furnace. The temper-

ature controller was not used during the kinetic studies, nor was tem-

perature recorded on the Brown recorder. For these studies the tem-

perature of the vapor baths was determined more accurately with the

Rubicon potentiometer.

Figure 11 shows some of the instruments used in this work. The

bottom instrument on the left is an Atomic Instruments Model 312 high

voltage supply. It supplied high voltage to the photomultiplier tube.

The ORNL Counter Amplifier is directly above. The dial on this instru-

ment gave the percentage of input impulses that were amplified. At the

counting rates used in this work, this percentage remained at essential-

ly 100. The output from the amplifier went to the ORNL Linear Count















EI Z I
F
I III
E
A NaI Scintillation Counter with Photomultiplier and Preamplifier
B Amplifier and High Voltage Supply
C Linear Counting Meter
D Chart Recorder
E Furnace
F Temperature Control


Fig.10. COUNTING AND TEMPERATURE CONTROL, SKETCH







-24-


--4 A I
.000
m o0


.-- S ~;


Fig. 11. Instruments.






- 25 -


Rate Meter, which is the top instrument on the left. This instrument

was designed and built for this work by Edward Fairstein and co-workers.

The ORNL Specification No. is 136, and the Drawing No. is Q-1511. Vari-

ous multiplier and range settings allowed measurement of counting rates

up to 2 x 10 cpm. The counting rate was read from the face of the

instrument and also recorded on the two point, 10 mv. Brown Chart Re-

corder. Temperature of the reactor tube in the survey experiments was

simultaneously printed on the recorder. The temperature range which

the recorder could follow was from 0 to approximately 10000.

At the lower right of Fig. 11 is the Minneapolis Honeywell

Electroline Temperature Controller, which was capable of seeking and

holding any temperature up to 12000. The pointer at the top of the

controller face reads the actual temperature at the Chromel-Alumel

thermocouple. In this photograph, it is reading room temperature,

about 300

The Harshaw standard NaI scintillation counter crystal was not

visible because of the lead shielding. This crystal was of the well

type and was mounted in a moisture proof can. The crystal was 1-3/4"

in diameter and 2" in depth and had a well of 3/4" in diameter and

1-1/2" in depth. When the crystal was canned, this well was large

enough to hold the Pyrex bulb of 5 cc. volume. An RCA-5819 photo-

multiplier tube was attached to the NaI crystal.

Another instrument not shown was the General Electric ionization

vacuum gage used to measure pressure between 10 and 103 mm. of Hg.
vacuum gage used to measure pressure between 10 and l0~ mm. of Hg.












CHAPTER III


PREPARATION OF FLUORIDE SALTS

CONTAINING F18


Fluorine eighteen can be prepared by using slow neutrons in the

reactions Li (n,t) He ; 0 (t,n) F18. Because this is a two step re-

action in which F8 is produced from tritons which in turn are pro-

duced from slow neutrons, high levels of F18 activity cannot be made by

this method. The most readily available source of slow neutrons was
28
the ORNL Graphite Reactor,28 which produced a slow neutron flux of about
12 -2 -1 29 I)30
1012 neutrons cm2 sec-1 29 The Low Intensity Test Reactor (LITR)3
13 -2 -1
with a slow neutron flux of above 103 neutrons cm-2 sec1 was more

frequently used in this work. Alkali fluoride salts could be prepared

by using LITR bombardments with between 105 and 106 counts per minute

(cpm) of F18 at the beginning of the exchange experiment. Salts with

about 10 times as much F18 activity were prepared by the reaction

F19 (p,pn) F18 in the ORNL 86-inch cyclotron.

The advantage of the higher F18 activity levels produced by the

cyclotron bombardments was counterbalanced by the disadvantage of a

more uncertain cyclotron operating schedule. Therefore, early exchange

work was all based on the LITR and was of necessity carried out at

higher temperature to obtain easily measurable exchange.

The LITR was chosen over the slightly more convenient Graphite


- 26 -






- 27 -


Reactor because of its higher slow neutron flux. In addition experi-

mental results indicated that the distribution of neutron energies was

more favorable in the LITR. An experiment was performed to test this

possibility. Some cobalt-aluminum alloy was included as a slow neutron

monitor31 in the capsules along with the charges in which F8 was to be

produced. After, neutron bombardment, counting rate measurements were

made of both the F18 present and the amount of Co60 produced in the

monitor. This same procedure was followed in both reactors, keeping

all concentrations, bombardment times, and F18 isolation steps (de-

scribed below) as identical as possible. It was found that 43.5 times

as much F8 was produced in the LITR as in the Graphite Reactor, but

that the slow neutron flux was only 10.5 times as great in the LITR.

The fast neutron reaction F19 (n,2n) F18 in the LITR was found experi-

mentally to produce very little F18 and could not account for the

advantage observed for the LITR over the Graphite Reactor. This con-

clusion agrees in general with similar experimental results of Boggs.5

All the work in the high temperature exchange studies used the LITR
18
as a source of F The main disadvantage of this method, besides the

fact that the F18 activity produced was low, was that nuclei such as

Na which are activated by thermal neutrons, were always present in

small concentrations. This presented problems both in purification and

in shielding.

It was found in attempts to prepare PdF2 containing F18 directly

in the LITR by the reaction F19 (n,2n) F18 that large thermal neutron






- 28 -


cross sections of trace impurities and of palladium made this method

impractical. Volatile radioactive trace impurities were found to con-

taminate the fluorocarbon gas stream during experiments. These

impurities could not be removed easily before the experiments. Radi-

ation originating from palladium created problems in shielding and

handling. This method of preparing salts was abandoned.

Table 1 lists some of the methods that have been used to pre-
18
pare F The two nuclear reactions used in this work are included

with references to work in which others have used the same reactions.

This table is not a complete summary. The selections were made from

considerations of variety or practicality. References 32 and 33 give

a more complete survey of the many methods by which F18 has been pro-

duced.

In the last reaction listed in Table 1, it can be seen that

different techniques have been used to mix the necessary nuclei, Li6

and 016 in the charge which was bombarded by slow neutrons. Consider-

ation must be given to the short range of the triton particle.41 It is

preferable to have the oxygen and lithium in the same molecule, but the

formation of gaseous products or water interfered with the use of such

salts in the present work.

To remove the F18 after neutron bombardment, HF was condensed on

the products, mixed thoroughly, and distilled off through the filter

(Figs. 7 and 8). Lithium salts which contain oxygen react with HF to

produce gases or water, or both. Hydrogen fluoride has a higher vapor





- 29 -


TABLE 1


SIX METHODS OF PREPARING


RADIOACTIVE Fl8


Nuclear Reaction Target Radiation Source Comments


F19 (n, 2n) F18 10-40 mg. NH F;5 LITR LRPa
liquid HF;2 Neutrons from the
34
HF, KHF2 cyclotron reaction
Li6 (d, n) Be7

F19 (, n) F18 KHF2 34 Betatron 48 or 84- HRPb
Mev X-rays

F19 (p, pn) F18 10 mg NaF;5 RNL 86-inch HRP, great
AIF 35 cyclotron quantities

016 (, pn) F18 H20;36 Pb037 UCRL 60-inch Erratic
cyclotron yield

F19 (d, t) F18 NaF38 Cavendish Threshold
cyclotron at 6 Mev.


Li6 (n,
16 (t
0 (t,


He4;
F18


LiF and Al20 5,34
LiCO39 LiNO 33940
3I03; 3


Slow Neutrons


LRP


a. Low radiochemical purity
b. High radiochemical purity






- 30 -


pressure than water and slow distillation leaves the water behind. How-

ever, to avoid the possibility of entrainment of water it was preferable

not to use the salts which reacted to give water. Experimental procedure

was also simpler if there were no gaseous products to remove. Because of

the small particle size of less than one micron, it was found that Linde

B Alumina worked satisfactorily when mixed with LiF. High purity of

this alumina was also a very desirable property. Analysis of a sample

of this alumina put through a dry run with HF showed some fluoride ion

to be present. It was thought that this result was caused by strongly

adsorbed HF rather than by reaction. Nevertheless, some HF residue was

always discarded in the distillation procedure to insure that no

moisture was distilled off.

At least a 2:1 molar excess of this HF was condensed onto the

fluoride salt in the fluoride exchange tube in Fig. 9 and the tube was

warmed slowly to 700, with the exit valve closed. After allowing about

fifteen minutes for achievement of isotopic equilibrium, the HF was

pumped off through the soda-lime trap. Pumping was continued while the

fluoride salt was heated, usually at 4000. With the alkali fluorides,

increasing difficulty in decomposing the bifluorides was encountered in

progressing through the group from LIF to CsF. With CsF a temperature

between 5500 and 6000 was maintained over at least one-half hour, after

which the CsF was found to have sublimed to the cooler part of the re-

actor tube near the top. Previous work has indicated that even after

this procedure some HF may have been present in the CsF.42 Flushing






- 31 -


with argon at high temperature was required to remove the last of the HF.

Alkali fluoride salts were prepared by cyclotron bombardment for

the low temperature exchange studies and the kinetic studies. Between

0.3 and 0.5 gram of dry alkali fluoride was packed into cylindrical

tubes made of 2S aluminum. These tubes were of 0.250" O.D., 0.0045"

thickness, and 2.75" length. The ends of the tube were flattened,

folded over once, and crimped tightly shut in a vise. The tubes were

flattened and placed in the cyclotron target head behind a 2S aluminum

window of 0.012" thickness. Cooling water was circulated at 40 pounds

per square inch gage pressure at a sufficient rate to allow only a one

degree temperature rise in the cooling water during bombardment. The

protons at the target had an energy of 21 Mev. The beam current was

approximately 80 microamperes. It was possible by this method to pre-

pare one curie of 1 in 15 min. A general description of the cyclotron

is available, as is a discussion of radioisotope production rates44

and a description of target heads 5 quite similar to those used in pro-

ducing F18

The aluminum tubes were opened behind a lead barricade by

quickly ripping them into two pieces with long handled pliers. The

alkali fluorides were washed from the two pieces into a beaker of water.

Alkali fluoride carrier was added, and the solutions were filtered

through Whatman No. 50 filter paper. The water was removed by evapo-

ration on a hot plate. The hot, dry alkali fluorides were placed in

hot weighing bottles to prevent adsorption of atmospheric moisture.






- 32 -


This precaution was not required for LIF and NaF. The weighing bottles

were stored in a desiccator. The weight of a portion of alkali fluoride

removed for an experiment was determined by difference. The portion re-

moved was quickly placed in the exchange system and evacuated.

From this point on the fluoride salts prepared by either method

were handled in the same manner. Evacuation was continued as the salts

were heated. For the survey experiments the salts were heated to approxi-

mately two-thirds of their melting points. The CsF was heated to 4000

for the kinetic studies. These temperatures were maintained until the
-4
pressure as read on the ionization vacuum gage was reduced to 10 mm.

Dry argon was circulated while the alkali fluoride was kept at high

temperature. If any radioactivity was observed in the argon, it was

pumped out and the process was repeated until no more activity was de-

tected. The system was then thoroughly evacuated and brought to the

temperature desired for the start of the exchange study. In the kinetic

studies, the argon was not monitored by circulation through the counter,

but was admitted and pumped off the CsF three times to insure a clean

surface.

Besides the alkali fluorides, CdF2, CrF2, PdF2, MgF2, NiF2 and

CuF2 were prepared by exchange with HF. Palladium difluoride showed

very little tendency to exchange with HF. By use of the HF with high

specific activity described in the survey with HF, it was possible to
8 18
prepare a gram of PdF2 with over 10 cpm of F Many fluorides have

been found to be quite soluble in HF.46 This technique probably can be






33 -



used successfully with a great many inorganic fluorides to incorporate
18
F The technique of direct bombardment by protons also appears to be

applicable to many inorganic fluorides.













CHAPTER IV


CALCULATIONS


The amount of exchange in all experiments described as survey

or kinetic studies was calculated by the following equation:47

_18 18
N = Net fraction exchanged = gas total (1)
gas total

gas
18 1 18

F total = sum of F counting rates, gas and solid phases,

Fgas = weight of fluorine, gas phase,

Ftotal = sum of weights of fluorine, gas and solid phases.

18
When N equaled one, the F8 atoms were randomly distributed

among all fluorine atoms in the two phases.

The following experimental values, which varied, were used to

solve this equation:

Temperature of the Inconel reaction tube, C = Ti,

Pressure of fluorocarbon, mm. of Hg = Pi

Initial counting rate of the alkali fluoride, cpm = (AFCR)o,

Counting rate of the background, cpm = (Bgd)i,

Recorded counting rate of fluorocarbon in the

counting chamber, including (Bgd)i, cpm = (RCR)I,

Counting rate of total gas when counted at the

same geometry as the salt, cpm = (CRTG)i,


- 34 -






- 35 -


Factor to correct (RCR)i-(Bgd)i to (CRTG)i = (CF)i.

All counting rates (RCR)i, (AFCR)o and (CRTG)i were corrected

for decay of F by the exponential law48 R/R = e4t where\ =

0.693/tl/2 and tl/2 = 112 minutes.

Other values, constant within a particular run, needed to solve

Equation 1 were:

Pressure of gas in the exchange system at 2980 K,

mm. of Hg = P298,

Volume of the exchange system, cc. = Vs

Volume of the counting chamber, cc. = Vcc,

Weight of fluorine in one mole of gas, g. = WFgas

Weight of fluorine in the salt, g. = Fsalt'

298 K = T.

The terms in Equation 1 became:

Fgas, g. = (WF gas)(moles of gas)


= (WF gas)(P298Vs/RT), by the ideal gas law.

Total, g. = Fgas + Fsalt
= (WF gas)(P298Vs/RT) + Fsalt

F8gasP cpm. = (CRTG)i = (CF)i (RCR)i (Bgd) .
18 18 18
F total' cpm* = F gas + F alkali fluoride
= (AFCR) corrected for decay, since all F l

was initially in the alkali fluoride.

Substitution yielded:






- 36 -


N= (CF)i IRCR)i (Bgd)i] / (AFCR)0 (2)
(WFgas) (P298Vs/RT)/(WFgas )(P298s/RT) + Fsalt

Radioactive C3F6 was counted as a liquid in an external well

type scintillation counter very similar to the one in the gas line to

give the quantity (CRTG)I. This radioactive C3F6 was placed in the ex-

change system and (CF)i was determined from the observed values of (RCR),

- (Bgd)i. For correct evaluation of N, it was necessary that (CF)i be

determined very similarly to (AFCR) Therefore, (AFCR) was determined

by counting aqueous solutions of the alkali fluoride in the same ex-

ternal counter at the same geometry, at approximately the same counting

rates, and applying counting loss corrections. The empty reactor tube

of Fig. 2, or the empty exchanger of Figs. 3 and 4, was at room tempera-

ture when (CF)i was determined. When the reactor tube or exchanger was

warmed, (CF)I decreased and Pi increased.

Consideration of (CF)i shows that it is a function of the ratio

of the total number of moles of gas present in the system to the number
of moles of gas in the counting chamber. While the reactor tube or

exchanger was at room temperature, this ratio, which will be called

ns/ncc, was equal to the ratio Vs/Vcc, by the ideal gas law. Thus the
following evaluation of (CF)i, as a function of Pi as Ti changed, could

be made:

(CF)298 = k (ns/ncc) = k (V/Vc) (3)
(F18 /F18)
(F8cr dr )EC
where k = 18 18 rEC
(F cr dr GLC1






- 37 -


and (F cr 18dr )EC = the ratio of counting rate to disintegration rate
in the external counter,

crand (Fr GLC = the same ratio in the gas line counter.

(CF)i = k(ns/ncc)Ti = k(ns/k Pi) (4)

where ns = P298 Vs/RT (5)

and k = Vc/RT (6)
since the counter chamber temperature stayed approximately at T, and the
only systems considered were those in which ns stayed constant. Dividing
5 by 6 gave

ns/k = V298 cc7)
Substituting (7) in (4) gave
(CF)i = (k V/Vcc)(298/i (8)
Substituting (3) in (8) gave

(CF)I = (CF)298 (P298/P (9)
Thus, (CF)I could be calculated very simply from the experimentally de-
termined (CF)298. From known values of V and Vs, Equation 3 gave k
from the experimental value of (CF)298. The first calibration gave a
value of 25.27 for (CF)298 and k was calculated to be 1.38. In a
similar calibration in the kinetics chapter, where Vs and (CF)298 were
both different and small adjustments had been made in the count rate
meter, k was calculated to be 1.31. The term k did not enter directly
into calculations, but this agreement showed that the values for k, Vs,
Vc and (CF)298 were all approximately correct.






- 38 -


For the high temperature experiments, the volume of the system,
86.3 cc., was less than in the calibration experiment where a cold
finger had been added. By Equation 3, where Vc = 5,

(CF)298 = 1.38 x 86- = 23.82.
This value of (CF)298 was used for the high temperature studies and the

first few low temperature studies.
It was not necessary to have (CF)298 in Equation 9 if it was
known at some other temperature, since:

(CF)2 = (CF)298(P298/P2),
(CF)1 = (CF)298(P298/P1),
(CF2/CF1) = (P1/P2),

(CF)2 = (CF)1 (PI/P2). (10)
Equation 10 was applied in the kinetic studies.
Usually the background (Bgd)i was substracted from the recorded
counting rate (RCR)i directly as these values were obtained from the
recorder chart paper. Therefore, this new quantity, the net recorded
counting rate, (NRCR)I, and also the value for (CF)I from Equation 9
were introduced into Equation 2. Rearrangement yielded:

C(NRCR), (1)
N = > (11)
Pi(AFCR)

where C = 298 WFgas) P298Vs + RTFsalt]
w(WF as) Vs
Figure 12 shows an example of the experiment data to which
Figure 12 shows an example of the experimental data to which


































10 2It~2 30



Pal83


Fig. 12. Example of Survey Experimental Data.










Equation 11 was applied. The heavy dark line on the left in Fig. 12

shows the temperature of the reaction zone. Times were written on the

left margin with temperature in the adjacent column. For the first

55 minutes a constant temperature of about 480 was maintained. For the

next 40 minutes the temperature was raised at about 5 per minute. The

temperature was then again held constant for 40 minutes at about 2560.

The counting rate record starts on the right of the chart. A full

scale value of 10,000 cpm (10 K in Fig. 12) was sufficient for 62

minutes; 20,000, 67 minutes; 100,000, 98 minutes; 200,000, the re-

mainder. A straight line was drawn for the slightly changing back-

ground. Data for two crude constant temperature studies and one survey

run were obtained from this chart. The survey run is discussed in

connection with Fig. 19 as Run C14a. The constant temperature results

are tabulated-as RbF, C3F6 in Table 6. For the kinetic studies, no

temperature data were recorded as in Fig. 12, and the speed of the re-

corder was increased so that the distance covered in 10 minutes in

Fig. 12 was covered in one minute.

As an example of actual calculations as they were made from

chart data, the (NRCR)i of 7050 at 810 and 59 minutes from Fig. 12 is

substituted along with other experimental data in Equation 11 to give

the following value for N:

N (7983)(7050) = 8.686 x 10-4
(287)(2.259 x 108 )

The experimental values used in calculating C were as follows:


- 40 -






- 41 -


C = (25.27) E114)(283)(91.3) + (1.86 x 107)(.0184 = 7983
(114)(91.3)

The particular value calculated above appears in Fig. 19 at a value of

2.825 for 1000/T, oK.

By use of Equation 11, only the exchange visible to the counter

could be evaluated. Exchange which is invisible to the counter also

occurs. Once an F18 atom enters the gas phase, the probability that it

will re-exchange with a fluorine atom in the alkali fluoride is as great

as is the probability that any other fluorine atom in the gas phase will

exchange with a fluorine atom in the alkali fluoride. The results

obtained by Equation 11 had to be correlated with the mechanistic

picture which developed during this work. The following paragraphs

discuss this correlation and its effect.

As will be seen in the discussion of results, all the evidence

indicated that both a fast and slow exchange reaction occurred. The

fast reaction soon led to a pseudo-equilibrium between the gas phase

and a portion of the alkali fluoride. The calculated magnitude of N

was not an actual measure of the amount of exchange, since exchange con-

tinued after random distribution by the fast reaction without changing
18
the F8 concentration in the gas. Thus, N should be considered as the

fraction of the alkali fluoride which was in equilibrium with the gas

phase when the fast reaction is being considered. No corrections for

exchange invisible to the counter needed to be made on the results of

Equation 11 as long as the fast reaction was being considered, since






- 42 -


equilibrium concentration of F18 was being measured. Results of the

survey studies will be shown to represent almost entirely the fast ex-

change equilibrium.

Therefore, N, which could only be defined as net fraction

exchanged early in this work, was redefined after analysis of all

observations as fraction of alkali fluoride in exchange equilibrium

with the gas phase. Use of N was continued with its original meaning

because the necessity of the second definition did not become apparent

until after the kinetic studies were made.

The same treatment of the data was applied with SIF4 and SF6

because of its convenience and because of the desirability of comparing

these results with those of the fluorocarbons.

In the kinetic studies, the determination of the rate of the

slow reaction required a different treatment of the exchange results

calculated by Equation 11. To calculate the rate of exchange, a cor-

rection had to be added to the slow increase in counting rate observed

in the gas to account for the exchange which removed F18 from the gas

phase. This same problem is always faced in kinetic studies involving

49
isotopic exchange, and the standard treatment of McKay was used. At

first consideration, it appeared that McKay's treatment, which was worked

out for systems in which the tracer was distributed homogeneously

throughout the different phases, could not be applied to gas-solid ex-

change experiments. The reason that it could be used in this work is

explained in connection with proposed mechanisms in the discussion of

results.












CHAPTER V


CHEMICAL AND RADIOCHEMICAL PURITY AND ERROR


Chemical Purity of Gases and Salts


All fluorocarbons were treated as follows to remove air,

moisture and other impurities. The gases were distilled from storage

bulbs into a liquid nitrogen-cooled trap. This trap was open to the

mechanical and diffusion pumps to remove any traces of air. After the

fluorocarbons passed twice through a tube filled with P205, a Regnault

molecular weight determination5 was made. This procedure established

that all molecular weights were within 1% of the theoretical value.

Perfluoropropene was purchased from Peninsular Chem-Research,

Inc. The 6nly suspected impurity was C3F H. Mass spectrographic

analysis of this CF6 and also of C3F prepared by a similar process

in the laboratory showed an estimated 0.3 mole percent of C F H. No

other impurities were detected. It seemed unlikely that this con-

centration of C F H could have any serious effect on fluorine exchange,

at least at the lower temperatures used. No attempt was made to remove

this impurity.

Perfluorobutane was prepared in the laboratory by the Electro-

chemical Process51 starting with C4H9COOH. Acidic products were re-

moved by passage through cold aqueous base. Preliminary single plate

distillation was performed in an open system, and material boiling near


- 43 -






- 44 -


0 was removed. After removing air and drying with P205 this fraction

was carefully distilled through a column of approximately 30.plates.

The material used in these studies boiled between -2.250 and -1.60.

Infrared analysis of C4F10 prepared by a similar procedure indicated

a mixture of the isomers perfluoroisobutane and perfluoro-n-butane.

Perfluorodiethylether, CF4, CF3H, and C2F6 were all prepared

by a procedure similar to that described for C4F10, except that only a

single plate distillation was required to isolate a pure sample of the

low boiling CF4. Oxygen difluoride was removed from these gases by

passage through a buffered KI solution. Infrared analyses indicated no

impurities in these materials. No impurities were detected in a mass-

spectrographic analysis of the C2F6.

The compounds SIF4 of 99.5% purity and SF6 of 99.0% purity were

used as purchased from the Matheson Company, except that air and other

low boiling materials were removed by the same process described above,

and SF6 was dried by passage through a tube packed with P205. Since

SiF4 reacts on contact with moisture, the drying step was not required.

A single plate vacuum distillation was performed in the vacuum system

with both SF6 and Si4, and the first and last portions of the dis-

tillates were discarded.

Baker Analyzed LIF and Fisher Certified Reagent NaF were used

as purchased. Spectrographic analysis indicated that the LIF contained

as much as 1% calcium. Baker and Adamson ACS Reagent Grade KF was first

dried and then sublimed before use. Rubidium fluoride was obtained





- 45 -


from General Chemical Company. The reported impurities were about 1% K

and traces of the other alkali metals. Cesium fluoride was obtained

from the A. D. Mackay Company. The reported impurities included 0.12%

Rb and traces of the other alkali metals. Spectrographic analysis of

this CsF also showed as much as 0.1% Al and traces of Si. A sample of
42
the very pure CsF used by Bredig, Bronstein, and Smith was obtained

for the last kinetic study. This CsF also showed traces of Al by

spectrographic analysis. Similarity of experimental results in the

last kinetic study to all previous results indicated that impurities

were not the cause of the unique observations in this work. Upon

heating at 4000 under vacuum, RbF and CsF changed color from pure

white to grey. This discoloration appeared to be as intense when the

very pure CsF was used as when the CsF purchased from the A. D. Mackay

Company was used. Support for a defect-caused mechanism was found in

this discoloration. Analysis of samples of all the alkali fluorides by
52
the spectrophotometric method5 both before and after exchange with

fluorocarbons showed no change in composition from the theoretical

values. The standard pyrohydrolysis method53 of analysis for fluoride

was found to be unsatisfactory for all the more volatile fluorides

(KF, RbF, and CsF).

Anhydrous nickel fluoride was prepared by hydrofluorination of

Special Reagent Grade Baker and Adamson NiC12 6H20 at 400 to 5000.

Anhydrous cupric fluoride was prepared by hydrofluorination at 250 to

300 of the Baker and Adamson ACS Reagent Grade, partially hydrated






- 46 -


product. Hydrofluorination produced salts which were spectroscopically

pure in any materials which formed volatile fluorides. Anhydrous chro-

mous fluoride was prepared by the thermal decomposition of (NH4)3CrF6

in a stream of hydrogen at 700. The (NH4)3CrF6 was prepared by the

reaction of Fisher Scientific CrF3 3-1/2 H20 with NH4F at 1000. A

mixture of PdF2 and PdF3 was prepared by direct fluorination of spectro-

scopically pure PdC12 at 4000. An attempt was made to convert the mixture

completely into the trifluoride by dissolving in excess BrF3 and dis-

tilling off the Br2 and BrF3. The resulting addition compound

PdF3 BrF3 was found to decompose at 200 under vacuum to yield PdF3,

but X-ray analysis showed that the decomposition was incomplete. Heating

the impure PdF3 to 4000 in the presence of HF, followed by an additional

hour of heating at 4000 under vacuum, yielded a product which gave the

theoretical fluoride analysis for PdF2 by the pyrohydrolysis method.53


Radiochemical Purity

In early attempts to prepare alkali fluorides by using LITR bom-
18
bardments as a source of F it was found that other radiations besides

those of F were present in the alkali fluorides. The filter shown in

Figs. 7 and 8 was added to the apparatus. This filter was packed with

a matting made of fine nickel wire which effectively stopped radioactive

materials from contaminating the product. Investigations of purity

were made with the ORNL 60-channel gamma spectrometer.54 Figure 13

shows a typical plot of counting rate against channel number for pure













_I >1



-lo
Sd


*O.


0


*
**





N-
.0


*.0


20 30
CHANNEL NUMBER


Fig.43. GAMMA SPECTROMETER PLOT OF F18 RADIATIONS


-
ii..^^^


0


0

B-


0 00
o00


40


,


I


- 47 -






- 48 -


F8 radiations. It was deduced from such a plot and the known amount
24
of Na24 radiation in the reactor irradiated sample that there was less

than one ppm contamination of the product by material from the original

sample after installation of the filter. Background variations did not

allow any closer evaluation of contamination. Many plots similar to

Fig. 13 were made after installation of the filter, and no evidence of

radiochemical contamination appeared.

In Fig. 13 the very small peak at 1.0 Mev was caused by the

occurrence of two gamma rays of 0.51 Mev being registered by the

spectrometer simultaneously as a single radiation. At first, the peak

at 0.7 Mev was not understood. To insure that this peak was not the

result of radiochemical impurity, further studies were made. It was

found to have a 112 minute half life and therefore was almost certainly

associated with F Since the 180 backscattered Compton radiation

from a 0.511-Mev gamma ray would have an energy of about 0.20 Mev and

would be in coincidence with the primary forward gamma of the annihi-

lation pair, the 0.7-Mev peak could result from this scattering process.

In the case of LITR-produced F18 the activity was always associated

with approximately a gram of salt which could act as a back-scatterer.

When measurements were made with high specific activity F18 produced

on the cyclotron, the 0.7-Mev peak was absent. It was possible to pro-

duce a 0.7-Mev peak by placing the high specific activity source be-

tween a scatterer and the scintillation spectrometer detector. It

was concluded that the fluoride salts containing F8 prepared with this





- 49 -


system were of high radiochemical purity.

Less danger of radiochemical contamination existed in the pro-

duction of F18 with the cyclotron, since the impurities have cross

sections comparable with the cross section of F19 under proton bombard-

ment.55 Thus, radiochemical contamination was approximately proportional

to the concentration of the bombarded nuclei, and by using moderately
18
pure salts, the specific activity of F8 for a few hours after bombard-

ment was such that other radiations could not be detected. This

technique also produced radioactive isotopes from the cations, but the

half lives were such that their radiations caused no difficulty in the

case of the alkali fluorides. For example, with a sample of LiF in

which the total F18 activity was over 110 cpm after bombardment, a

counting rate of about 105 cpm from the 53 day Be isotope was detected

after the F18 had decayed. Several half-life studies on the radioactive

gas, after exchange with these alkali fluorides, indicated that only

F radiations were present. Half-life determinations were also made on

the gases after exchange with salts prepared by the technique based on

the LITR, and only F18 radiations were observed.


Error

The following table presents the most important sources of

error which entered into the terms in Equation 11. These estimates

were made on the typical example of calculations, which accompanied

Fig. 12. The first error in Table 2 has the greatest effect on N.

This error has a small effect on the precision within a particular run.






- 50 -


TABLE 2

ERROR


Estimated
Term Limits Source
of Error

(AFCR)o t 2% tl/2, over 4 hr.

(NRCR)i + 1% Statistical fluctuation of count
rate meter

P298' Pi 1% Isochore plots obtained with
capillary Hg manometer

t 1% Temperature uncertainty

Vs + 2% Estimation for loss of
calibrating water

Fsalt 1 Weighing
C t 1% P298 Vs' Fsalt

N t 3% Accumulated error



Most systematic errors were kept small by the technique of

comparing counting rates of salts and gases on the same external counter.

Other potential sources of error, not considered significant in any of

the data presented, are discussed below.

The higher boiling materials C4F10 and (C2F5)20 adsorbed suf-

ficently on the counting chamber walls to cause experimental difficulty.

However, a very slight increase in the counting chamber temperature,

with increasing Ti, completely removed this effect with both materials.

Figure 14 demonstrates that much adsorption occurred on the system walls









280F--


E/
E /
275 -





270 I I
100 200 300
Fig. 4. PRESSURE-TEMPERATURE BEHAVIOR OF C ABOVE
Fig. t4. PRESSURE-TEMPERATURE BEHAVIOR OF C4FO ABOVE CsF





- 52 -


and on the salt with C4Fi at low reactor chamber temperatures. There-

fore, no data is presented for CF10, and (C2F5)20 below a reactor

chamber temperature of 270 although these materials exchanged readily

at 00, the lowest temperature used.

Heating rate was approximately 50 min.-1. Deliberate variation

of this rate indicated that results were not appreciably affected.

Some background variation almost always occurred during low

temperature runs. Figure 12 presents a typical example. This variation

was not significant.

There was some possibility that all HF was not removed,

especially with CsF, in the high temperature survey work. Agreement

in results from the same systems between low and high temperature

surveys indicates that there was no error from this source.

The rate of gas flow was constant at about one cycle per minute.

Determination of the rate of the slow exchange process was not hindered

by this rate of flow. The fast exchange process was probably delayed

in its approach to equilibrium. No error was introduced in the kinetic

studies, since only equilibrium data for the fast reaction were re-

ported. A small delay must have been involved in following N during

survey runs, but this small delay was common to all runs.

In the kinetic studies, the temperature of the constant tem-

perature bath was determined with a precision of + 0.1 by a Pt,

10% Pt-Rh thermocouple. Calibration by ice baths and steam baths

showed this thermocouple to be within t 0.1. of the true temperature






53 -



at these two points. The most favorable conditions were chosen and

the best techniques developed in the survey studies were used to keep

all errors at a minimum.












CHAPTER VI


GENERAL EXCHANGE STUDIES


High Temperature Survey


At first these studies were performed on various fluoride

salts in order to find fluorides capable of exchanging with fluoro-

carbons at moderate temperatures. In exchange studies with C3F6 up

to 5000 and 340 over CrF2 and CuF2, respectively, these fluorides

produced few counts in the gas phase beyond what might be attributed

to statistical variation in background. Sodium fluoride produced

easily detectable counts with C3F6 at 4000. Further experiments in

both the high and low temperature surveys were confined to the alkali

fluorides.

Both C3F6 and C4F10 were passed over CsF, KF, NaF and LiF in

the apparatus shown in Figs. 1 and 2. Temperature was increased at

the rate of 50 per minute. With C F10, the counting rate was found to

be sufficient for accurate calculations only with CsF and KF and at

high temperatures.

The results were always found to be almost linear up to large

values of N when the logarithm of N was plotted against reciprocal

temperature. Figure 15 shows such plots for C3F6 on each of the four

fluorides studied. Curvature started between N values of 0.2 and 0.4


- 54 -




- 55 -


1.0 I.1 1.2 1.3 1.4 1.5 1.6 4.7
1000/ T, K
Fig.45. FRACTION OF EXCHANGE OCCURRING WITH
CGF OVER THE ALKALI FLUORIDES


0.5


0.05



0.05






- 56 -


with CsF and increased as the run was continued. Some curvature was

observed with NsF at about the same values of N as with CsF. In Fig.

16, three runs with CF6 and NaF were compared. The fact that the

lowest points fell below the line on all three runs is significant for

reasons which become apparent when results are discussed in connection

with Fig. 36.

Figure 17 shows the behavior of C4F10. Low temperature survey

plots included on the same graph showed general reproducibility. The

CsF curve was to the right of the KF curve, as it was in Fig. 15. In

Fig. 17 the CsF plot continued to be linear past the temperature where

curvature started with C3F6. Values of N were lower at comparative

temperatures with C F10 than with C3F6, as Fig. 18 illustrates. Curva-

ture appeared to accompany large values of N and was not affected by

the temperature.

Wide variations in concentration of either phase did not pro-

duce any noticeable change in results. This is shown by Fig. 16, where

approximately the same weight of NaF was used in all three runs, but

CF6 pressures at room temperature varied from about 200 to 600 mm.

Runs which yielded fragmentary information, because of low

counting rate or counter chamber contamination, were summarized in

Table 3.


Low Temperature Survey

Four fluorocarbon gases, CFg6, C F0, (C2F5)20, and CF4 were




- 57 -


0.1


















0.01--










4.0 1.1 1.2 1.3 4.4 1.5
4000 / 7; K
Fig.46. FRACTION OF EXCHANGE OCCURRING WITH C3F
OVER NaF; THREE RUNS









- 58 -


,I :


I I I I I I


CsF


0.1













0.04













0.004


CsF


I I I I I I I I


4.4 4.2 1.3 1.4 1.5 4.6 1.7 1.8
1000 / T0K
Fig.47. FRACTION OF EXCHANGE OCCURRING
WITH C4F oOVER CsF AND KF


CsF





- 59 -


- Ism~


0-- \


I I I


C4 F


I I II I I I I


1.0 4.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
1000 / T, K
Fig.18. FRACTION OF EXCHANGE OCCURRING WITH
WITH C3F6 AND C4F40 OVER CsF


0.1













0.01






- 60 -


TABLE 3

DATA FROM STUDIES HAVING EXPERIMENTAL
DIFFICULTIES OR NEGATIVE RESULTS


Run No. and Temperature at
Run No. and
Material Which Exchange Results
Materials
Was First Observed
14 No exchange up to 3250. Activity
CuF 7 ?6 in CuF^ was such that less than
C2- 6 10 exchange could not be detected.


Cr19 300 Less than 8% exchange at 4250
CrF2-C?6


cKF(c 2 About 0.05% exchange at 600.


C4a Less than 5% exchange at 6250. Back-
KF-(C )0 ground would not return to normal.
(2 520No evidence of decomposition.

C8a 0.4% exchange at 4500. No decom-
CF-(C) position. Background would not re-
-(25) 2 turn to normal.


iC5a 325 Background would not return to normal.
LIF-SF6


C6a 3100 0.5% exchange at 4470. Background
LiF-SF6 would not return to normal.


C7c 350o Less than 0.2% exchange at 4500
K.-SFG6 Background would not return to normal.


ClOc 4000 Background would not return to normal.
C-SF6


C12d
RF -SF6


No exchange up to 3500.






- 61 -


TABLE 3 Continued


un o. Temperature at
Materials Which Exchange Results
a Was First Observed


NF3bSF No exchange up to 2800.


C4b 600 Less than 5% exchange at 6500.
KF-CF10 Decomposition started at 600.


C13a 0.04% exchange at 3000. Background
NaF-C F10 would not return to normal.


NaF-Cb Less than 0.2% exchange at 300.
4 10

C16d 2900 Decomposition started at 2900
CsF-CF3 H


CsF8a 250-3000 Less than 0.03% exchange at 4000.
CsF-C2 6


S-C2g No exchange up to 500.


C12b Room Physical adsorption interfered with
RbF-(CF3 )3N temperature calculations.


44-111 5000 Pressure dropped throughout the
KF-C1CF=CF 2 experiment.


39-107 4750 Pressure increase occurred before
LiF-CF SF5 counts were detected.


4o-lo6
LiF-CF N=CF2
3 =C2


4000


Pressure increased slowly throughout
the experiment.






- 62 -


passed over the five alkali fluorides in the same apparatus used in the

high temperature survey. Figures 19 through 22 show the results, which

were very similar to those of the high temperature survey. In Fig. 19

the same relationship between different fluorides observed in Fig. 15

occurred again. Two runs with RbF showed good reproducibility. The

LiF prepared with the cyclotron contained some volatile material, possi-

bly a complex fluoride. By preparing the LiF very carefully, results

obtained with C4F10 and (C2F5)20 are regarded as dependable.

In Fig. 20, the general conclusion can be drawn that CsF and

RbF exchange better with C F10 than does KF or LIF. With (C2F5)20,

in Fig. 21, CsF is indicated as being superior to RbF, which in turn is

superior to LiF.

Few points were obtained with CF4, since high temperatures were

avoided and detectable exchange did not occur until at about 3000 with

CsF and at nearly 4000 with RbF. Figure 22 shows similar slopes with

these two salts, but CsF appears to exchange better than RbF. The

slopes of Fig. 22 are greater than with the other fluorocarbons.

Figure 23 compares high and low temperature CsF-C3F6 runs

(AC I and AC II are discussed in the kinetics studies). This figure,

plus a further comparison between high and low temperature runs, dis-

cussed below in connection with Fig. 36, indicates that similar pro-

cesses take place and that the different methods of preparing fluorides

yield similar results.

The curve for CF H in Fig. 24 shows an inflection just under




- 63 -


4x10 NaF
Cl





04
+-




ix(52 x
+-

+ CsF -
+\- + C0b -'



x -x


RbF
C14a
RbF
C42c



1.4 1.3 1.5 4.7 1.9 2.1 2.3 2.5 2.7 2.9 3.1 3.3
1000 / 7; K
Fig.49. FRACTION OF EXCHANGE OCCURRING WITH C3F6 OVER
THE ALKALI FLUORIDES AT LOWER TEMPERATURE







- 64 -


4 x10-2


4 x4 0-3
x1 .0
1.0


I I I I I I I I


RbF
*


CsF


LiF


I I I I I I I I


4.1 1.2 4.3 1.4
4000 /


1.5 1.6
T7 K


1.7 1.8 4.9


Fig. 20. FRACTION OF EXCHANGE OCCURRING WITH C4 F40
OVER THE ALKALI FLUORIDES AT LOWER TEMPERATURE





- 65 -


I I I I I I I I I


CsF
C9a


LiF
C6c


4 x 4 -2










4 x 10


1.0 1.1 1.2 1.3 1.4 1.5 1.6 4.7 1.8 1.9 2.0
1000/ T oK
Fig. 24. FRACTION OF EXCHANGE OCCURRING WITH (C2F5)2
OVER THE ALKALI FLUORIDES


RbF
C4a
I:I I I I I*




- 66 -


RbF
C47a




CHQ \


1 x40-5
1.3


Fig. 22.


4.4


FRACTION
WITH CF4


1.5
4000 / TOK


1.6 1.


OF EXCHANGE OCCURRING
OVER RbF AND CsF


4 x 0-4


CsF
C46a


I




- 67 -


36-405


0.1
















o.o \





0
0.0 01
I I I I
4.0 1.2 1.4 1.62.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6
1000 / 7 K
Fig. 23. FRACTION OF EXCHANGE OCCURRING WITH C F6
OVER CsF BY THREE TECHNIQUES






- 68 -


SI I l I l I I I I




OL! 0
oo *

-00





0

0
0 0

o i
O LLJ




\ 0 D


0-




Ld
a-




W
cn

0 0
0 L F-
I
Soo





w

cJ
od

" t O 'O (D o O 0 D Io 0
bH ^~






- 69 -


3000. As listed in Table 3, radioactivity was first observed in the

gas phase at about this temperature with the CF3H-CsF system. Tri-

fluoromethane is known to have thermal and chemical stability approach-

ing that of CF 456 The CF4-CsF system also produced detectable radio-

activity in the gas phase at 3000. It therefore appears possible that

a similar process was involved in both cases. With CF3H, decomposition

must have occurred to yield fewer molecules in the gaseous state, while

CF4 was not transformed into new materials.

When SIF4 was passed over the alkali fluorides, decreasing slopes

of the isochores were observed. Figure 25 shows that with LiF this slope

did not differ appreciably from that of SiFt alone, but the slope de-

creased at higher temperatures with KF, RbF, and CsF. One run with KF

was followed beyond the inflection point, and the pressure continued to

decrease. The quantity N could no longer be easily determined after a

significant quantity of SIF4 was removed from the gas phase. Therefore,

in Fig. 26, the values of N go only up to the temperatures where devi-

ations in the pressure-temperature curves were observed. In Fig. 25,

it can be seen that various starting pressures were used. The two runs

with KF had nearly equimolar quantities of SiF4 but Run C4c had nearly

twenty times as much KF as did Run C7b. This variation of concentration

appears to have separated the curves from these runs in Fig. 26. The

run with LIF had molar concentrations of SiF4 and salt similar to those

of Run C4c with KF, and the results of these two runs in Fig. 26 are

quite similar. A fluorocarbon run was included for comparison in



















500


-I-
E
E KFC4c




300 '
*- e---**-- .CsF

100 200 300 400 500
T,
Fig.25. PRESSURE-TEMPERATURE BEHAVIOR OF SiF4 ALONE AND ABOVE
ALKALI FLUORIDES










0,3






0.2






0.1


100 200 300 400 5C

Fig. 26. RESULTS OBTAINED BY PASSING SiF4 OVER ALKALI FLUORIDES






- 72 -


Fig. 26. Logarithms of the results with SiF4 plotted against reciprocal

temperature, as shown in Fig. 28, were not linear. No traces of Li2SiF6

could be found by spectroscopic analysis for silicon in the LiF after

the experiment, while KF, RbF, and CsF all showed large amounts of

silicon.

Exchange studies with SF6 were also carried out in the same

manner. Table 3 shows that radioactivity was detected at temperatures

above 3000 with different alkali fluorides. In this respect, SF6

resembled CF4 and CF3H. Comparison of temperatures at which counts

were detected with SF6 to the isochores in Fig. 27 shows that a pres-

sure decreasing reaction accompanied the radioactivity. Pumping

failed to remove up to as much as three-fourths of the radioactivity

present at the end of these experiments. The pressure reducing re-

action might be 2SF6 ) SF10 + F2. The compound S2F10 has a

normal boiling point of 290 and would be adsorbed by the cool counter

chamber surface.

The surface areas of representative samples of alkali fluorides

used in the low temperature survey, as measured by the BET method using

krypton gas, were found to be:
2 -1
CsF = 0.11 m. g.
2 -1
RbF = 0.08 m. g.
2 -1
KF = 0.08 m. g.
2 -1
NaF = 0.22 m. g.
2 -1
Li = 1.1 m. g.









I I I

600 -
SF6 alone

-CsF





!= RbF ,

+KF


x LiF


500 I I I
200 300 400 500
7T, C
Fig. 27. PRESSURE-TEMPERATURE BEHAVIOR OF SF6 ALONE AND OVER
THE ALKALI FLUORIDES






- 74 -


Figure 28 is a collection of examples of all the low temperature

exchange studies.

Figure 29 shows examples of isochores from five gases. None of

these gases which produced curves varying significantly from that of

argon were used in further exchange studies, with the exception of SiF4.


Survey of Hydrogen Fluoride and Various Inorganic Fluorides

for Indications of Exchange with SF6 and Fluorocarbons


Nonexchange of F18 between HF and fluorinated methanes has been

reported.5 This work used HF of low specific activity; consequently

values of N below about 0.10 could not be detected. Since hydrogen may

be considered the first member of the alkali group, this work was re-

peated with HF of high specific activity.
18
The cyclotron was used to prepare EF containing F The

aluminum tubes containing KF were broken open and dropped into the

fluorothene tube shown in Fig. 8. This tube was attached to the vacuum

line and evacuated. One to two cc of HF was condensed from the HF

storage vessel into the fluorothene tube by cooling with liquid nitrogen.

The fluorothene tube was warmed to room temperature to allow the HF to

exchange with the KF and pick up F 8. Potassium fluoride was chosen be-

cause it goes into solution with a rapid exothermic reaction, thus

causing random distribution of the F18
18
A molar excess of HF was used, so that most of the F8 was re-

moved with the HF when it was distilled away. This HF usually had




- 75 -


A,ft


-t ,* .


*


3.--. -..0-


"--0.


*

Ix


CsF
RbF
KF
LiF
NaF


C3F6
- C4Fo0
- -- (CzF5)20
------CF4
------SiF4
4


4000/ T, K


4 2 3
Fig. 28. LOW TEMPERATURE SUMMARY PLOT


1-3










10-4


,* f.l













600


QE 500








400-
300

Fig.-29.


400 500 600
T oC
PRESSURE-TEMPERATURE BEHAVIOR OFARGON AND FIVE GASEOUS
FLUORINE COMPOUNDS OVER THE ALKALI FLUORIDES






- 77 -


sufficient F18 activity to read one roentgen per hour at contact with a

laboratory monitor on the outside of a 20 cc nickel storage vessel.

A portion of this HF was admitted to the manifold at a pressure

of 200 to 400 mm and condensed into the nickel exchange tube shown in

Fig. 9. A sample of the gas to be studied, also at a pressure of 200

to 400 mm, was admitted to the manifold and then condensed into the ex-

change tube with the HF. The exchange tube was quickly warmed to 500 +

100 with a small furnace. This temperature was maintained for one hour.

The furnace was removed and the reactor tube was cooled with

liquid nitrogen. The exchange tube was opened to the manifold and the

pressure measured. A small residual pressure was often noted, and with

CF6 considerable pressure was observed. Therefore the two U tubes on

the exit side of the manifold were cooled with liquid nitrogen and the

reaction products were distilled into these U tubes while the noncon-

densible gases were pumped off.

Any F18 removed in this manner introduced error into the de-

termination. Since the residual pressure was small above the products

cooled with liquid nitrogen, this error was small, except for the

experiment with C3F6. The reaction products were condensed into a

fluorothene tube packed with moist NaOH pellets. This fluorothene tube

was warmed to room temperature to cause reaction of the BF with the

NaOH pellets. Unreacted gas was distilled off through the filter, con-

densed into a small nickel counting vessel, and its counting rate was

determined. The NaOH pellets and the newly formed NaF were washed from






- 78 -


the fluorothene tube and diluted to 100 ml in a volumetric flask.

Aliquots were removed'and counted. Precautions very similar to those

described under the survey calculations were taken during counting to

reduce error.

No evidence of HF was observed in any of the gases counted when

they were bubbled slowly through an indicator solution. The gas from

Run VII had an unmistakable odor of H2S.

Table 4 demonstrates the stepwise development of the calcu-

lations of the exchange experiments with HF. From the pressure of HF

(column 2) and of the gaseous fluorine-containing compound (column 3),

the fraction of total fluorine atoms (column 7) in the fluorocarbon or

SF6 was calculated from the ideal gas law as follows:
nRT
v
and P = kn,

since V = volume of manifold,

and T = room temperature;

(PFC) (Fs/mole FC)
Column 7 = (PF)(Fs/mole FC) + (P H)(Fs/mole HF)


(P ) (Fs/mole FC)
(PFC)(Fs/mole FC) +(PF)

where PFC = pressure of gaseous fluorine-containing compound,

Fs/mole FC = fluorine atoms per mole of gaseous fluorine-con-

taming compound,

PHF = pressure of HF
HF






- 79 -


TABLE 4

EXCHANGE STUDIES BETWEEN HF AND
SOME FLUORINE-CONTAINING COMPOUNDS


SP Pressure Counting Rates
Run, Page Pressure of Residual
and
Compound of HF, mm. Fluorine- Pressure NaF Gaseous
Containing Compound
Compound
(1) (2) (3) (4) (5) (6)


I,B15 ,C36
IIa,B16,SF6
IIIa,B26,CF4
IIIb,B30,CF4
IV,C3,CF3H
V, C5, CF3H
VI,C7, CF3H
VII, C9, SF6
VIII,C10,SF6
IX, C12,C4F10
X, C14, C4F10


380
379
370
387
230
247
237
281
245
270
261


3.437xl05
2.064x107
3.950x106
4.811xl06
9.952x107
4.237xl07
6.105xl07
1.368xo08
1.609xlo8
8.038x107
7.665xl06


2.212xl05
1.485x105
1.317xl05
4.255xl04
7.691xl05
1.002x105
7.236x105
8.310xlo5
2.114xlo6
2.006x1o6
2.356x105


143
367
368
300
350
280
276
310
297
302
302






- 80 -


TABLE 4 Continued


Counting Rates
Fraction F
in Gaseous Sum of At 100% o Percent
Compound (5) and (6) Exchange T C Exchange Remarks

(7) (8) (9) (10) (11) (12)


5.649xl05
2.079x107
4.082x106
4.854x106
1.003x108
4.247xlo7
6.173x107
1.376xo08
1.630xl08
8.239xl07
7.901xl06


5.316x105
1.790xl07
3.270x106
4.o48x1o6
3.982x107
3.083xlO7
4.451x07
1.163x108
1.356xlo8
7.407xo07
7.079x106


500
500
500
400
400
500
250
500
250
500
250


41.6
0.83
4.0
1.1
1.9
0.33
1.6
0.72
1.6
2.7
3.3


PONa
PON
VLRb
VLR
VLR
VLR
VIR
VLR, H2S
VLR, no H2S
VLR
VLR


a. Pumped off noncondensibles
b. Very little residual pressure


0.941
0.861
0.801
0.834
0.397
0.726
0.721
0.845
0.832
0.899
0.896






- 81 -


Fs/mole HF = 1 = fluorine atoms per mole of HF.

The fraction of total fluorine atoms in the fluorocarbon or SF6

was multiplied by the total F18 counting rate (column 8), which was the

sum of the counting rates of the HF (column 5) and of the gaseous

fluorine-containing compound (column 6), to give the counting rate

possible at random distribution, or 100% exchange. The product of

observed counting rate of the gaseous fluorine-containing compound in

column 6 multiplied by 100 and divided by the counting rate correspond-

ing to 100% exchange in column 9, yielded percent exchange (column 11).

Column 10 gives the temperature which was held for one hour in the re-

action tube. The remark PON in column 12 stands for pumped off non-

condensibles. The amount of F18 pumped off was significant only where

PON was noted. VLR means that the residual pressure over the reaction

products at liquid nitrogen temperature was so small that no significant
18
amount of F8 was pumped off with the noncondensible materials.

After Run X of Table 4, the nickel reactor tube was removed.

The tube was coated with NIF2 with a sufficient counting rate to

possibly account for the small percentages of exchange listed in

Table 4. Studies were then performed by heating C4F10 over NiF2 and

CF4 over CuF2, PdF2 and CsF for one hour. The results are summarized

in Table 5. Calculations were of the type discussed above with HF.

The four inorganic fluorides were prepared by exchange with HF. Sur-

face areas of these fluorides were measured by the BET method using Kr.

As has been discussed above, CsF is very difficult to prepare free of






- 82 -


HF by this method without loss by sublimation. Because of such loss,

not enough CsF remained to allow surface area measurement. The number

0.01 m2 is a reasonable maximum estimate.


TABLE 5

EXCHANGE STUDIES BETWEEN SOME INORGANIC FLUORIDES
AND FLUORMOfARBONS AT 5000



Run, Page Materials Surface Area % Exchange


XI, C16 C4F10, NiF2 4.7 m2 19.5
XII, C19 CF4, CuF2 2.3 m2 11.8
XIII, C22 CF4, PdF2 1.2 m 29.1
XIV, C24 CF4, CsF <0.01 m2 2.8












CHAPTER VII


KINETICS OF THE EXCHANGE REACTIONS


The results of the rate studies performed in combination with

the survey studies are listed in Table 6. Figure 30 shows the graphs

from which these data were obtained.


TABLE 6

KINETICS OF EXCHANGE AT CONSTANT TEMPERATURE



Materials T, oC Rate, N, min-


NaF, C3F6 22 t 2 5.14 x 106
NaF, C3F6 260 4 6.18 x 10-5
RbF, C3F 6 46 2 6.53 x 10-6
RbF, C36 256 3 8.49 x 10-5
CsF, CFg6 203 t 3 8.25 x 10-5
RbF, (C2F5)20 300 t 5 3.49 x 10-6
CsF, CF4 368 t 3 3.16 x 10-5


In the two systems, NaF, C3F6 and RbF, C3F6, lower and higher constant-

temperature data were taken from the beginning and end, respectively,

of the same survey experiment. Considerable increase in rate with

temperature was evident in these four studies. Table 6 shows quite

similar rates with C3F6 above NaF, RbF, and CsF when the temperatures

were approximately equal. Both (C2F5)20 and CF4 appeared to produce


- 83 -











200|


405 300 5 --


1 00 Z u--

~A G

S5,5 4. -






NoF- 3G 3


RbF- C F x 103 256 30
XX S 6 X-

0 10 20 30
A t, min
Fig. 30. KINETIC STUDIES






- 85 -


slower rates than did C3F6 on the same salts at comparable temperatures.

Figure 30 indicates that after a few minutes at constant temperature the

rates became constant.

Careful rate studies were made in the exchange apparatus shown

in Figs. 3 and 4. The procedure was modified slightly from that used in

survey studies. To prevent physical adsorption difficulties, the C3F6

or CF4 was admitted to all of the system except the exchanger. When

the gas was admitted to the exchanger, by opening the stopcocks which

isolated this part of the system, the pressure decreased in the counter

chamber and no adsorption of radioactive fluorocarbon gas occurred.

Simultaneously, a timer was turned on and times were recorded to the

nearest second.

After collecting data for one-half hour or longer, the gas was

condensed in the cold-finger with liquid nitrogen. A vapor bath con-

taining a higher boiling liquid was installed, and the entire procedure

was repeated using the same gas and salt. For Study AC I, the pro-

cedure was repeated for a third time at a still higher temperature.

Figure 31 shows the results for the Run AC Ia at 98.90, AC Ib at 130.50

and AC Ic at 182.50 and for Run AC IIa at 0.00 and AC IIb at 55.4.

AC Ic was continued for 90 minutes with no deviation from linearity

becoming apparent. In AC I, 5.518 x 10-3 mole of CsF was used; in
-2
AC II, 1.007 x 102 mole. Equal amounts of C3F6 were used in the two

studies.

The rates obtained from the slopes in Fig. 31 are shown in













0.001






- 0.005








0.040


40 20 30 40 50 60 70


t, min
Fig.34. KINETIC STUDIES OF EXCHANGE OCCURRING WITH
C3Fg OVER CsF






- 87 -


Arrhenius Plots in Fig. 32. The inflection in AC I at a temperature of

about 4250 K corresponds to the expected Tammann temperature57 at about

one-half the melting point. Activation energies calculated from the

slopes of Fig. 32 are 4.0 kcal mole-1 for AC II and 0.8 kcal mole- for

AC I below the Tammann temperature. Above the Tammann temperature, the

activation energy in AC I was estimated as 17 kcal mole-.

In Fig. 31, extrapolations of the linear plots back to zero time

did not intercept the ordinate at zero. Logarithms of the values of N

at the ordinate intercepts are plotted against 1000/T0K in Fig. 23.

This combination plot shows that the survey results were caused almost

entirely by the fast reaction.

One study, AC III, was attempted with CF4 over CsF at 990 and

2090. At 990 an immediate exchange occurred which gave an N value of

2.1 x 10-5. No change in this value was observed over one-half hour.

At 2090 no change in N could be detected. The counting rate was then

too close to the natural background to follow the exchange further.
2 -1
The measured BET surface area on this salt was 0.11 m. g.

The study AC IV shown in Fig. 33 was undertaken to determine

whether different heating periods in the CsF preparation were respon-

sible for the differences between AC I and AC II shown in Figs. 31 and

32. Two samples of CsF were compared. One was prepared at 4000 for

1/2 hr., the other at 46 for 1-1/2 hr. Each sample of CsF was used

in the same exchange procedure with C3F6 as was used for other kinetic

studies, except that at 20 min. the ice bath was quickly replaced by a




- 88 -


1x 10-4





















E x


1 x 10"5


-ACI









ACH








- I
- -
\


2.0 2.2


2.4 2.6 2.8 3.0
4000 / To


3.2
K


3.4 3.6 3.8


Fig. 32. RATE DATA TAKEN FROM FIG. 31





































40
t, min


Fig. 33. EFFECT OF DURATION OF CsF PREPARATION


100






0
X



40






- 90 -


boiling-water bath, which was subsequently quickly replaced at 40 min.

by a boiling-nitrobenzene bath. The CF6 was not removed between instal-

lation of different constant temperature baths, and therefore the CsF

changed temperature gradually with C3F6 flowing constantly. The sample

of CsF heated for 1-1/2 hr. during preparation produced larger values

of N at the two higher temperatures. This CsF had a BET surface area
2 -1
of 0.11 m. g. while the CsF heated for only 1/2 hr. during prepa-
2 -1
ration had 0.13 m. g. Nearly equal concentrations of CsF and C3F6

were used in both parts of the study.













CHAPTER VIII


DISCUSSION OF RESULTS


Fluorocarbons


The order of exchange ability was found to be CsF> RbF > KF,

NaF, LiF regardless of which fluorocarbon was being considered. Fair

agreement between high temperature and low temperature surveys as

shown in Figs. 17, 23 and 36 shows that the different techniques used

in preparing the alkali fluorides did not noticeably affect the results.

Reproducibility of results with CF6 and alkali fluorides was not

affected by concentrations, which varied widely. Results could not,

have been induced by F18 radiations, since the specific activity

varied by a factor of 10 with no apparent change in results.

The fluorocarbon survey results, expressed as the logarithm of

N plotted versus reciprocal temperature, were always linear. This

fact suggested that either the Arrhenius AE or Van't Hoff AE was

being measured. The latter is more probable, since reproducible

linearity is unlikely in an Arrhenius plot at a constant heating rate.

Table 7 lists slopes from the survey studies and AE values calculated

from these slopes. These slopes appear to result essentially from ex-

change by a fast reaction mechanism up to the highest temperatures

used in the low temperature survey. Figures 23 and 36 demonstrate


- 91 -




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs