AN mvxzsmmm .01: mi EURGHUM _ j -; ~ '
‘OXIDEFWORIDE Euaomum sasamoxzag SYSTEM _

Thesis for the Degree of M. S.
MECHIGAN Sm: UNEVERSEE’Y ‘ ,
SANDRA LEGMRD BfiCON ' '

' ' '19::

""w-.~V

 
  
       

  

LI BRA R Y
Michigan Stat:
University

f’n ¢¢¢¢

 

 

In an a
oxidefluoric
“WWW. see
This solutio
at constant -
sequioxide '
not undergo
above the no=

The modq
apparent new
with 91mm ‘

339% red a S

I”Vesti

AN INVESTIGATION OF THE EUROPIUM
OXIDEFLUORIDE - EUROPIUM SESQUIOXIDE SYSTEM

By

Sandra Leonard Bacon

In an attempt to study the vaporization behavior of europium
oxidefluoride, it was learned that the equilibrium product, cubic
europium sesquioxide, forms a solid solution with the oxidefluoride.
This solution was characterized by a continuous decrease in pressure
at constant temperature as the conversion of oxidefluoride to
sesquioxide took place and also by the fact that the sesquioxide did
not undergo its cubic to monoclinic transition at temperature 300°
above the normal transition point.

The mode of vaporization for EuOF was not established due to
apparent reaction of molybdenum, tungsten, thoria and platinum cells
with either the residue or effusate. Condensed effusate always
appeared as a reduced fluoride EUF2.3’

Investigation into the reactions of molybdenum and platinum with
EuF3(s) revealed that some reduction occurs in the solid state during
vaporization at l500° in both cases.

A new compound with the proposed formula Eu3Th2(Eu04)30 was
produced by reaction of Th02 and EuOF under vaporization conditions.
The structure of this compound fits the apatite family with a =
9.590 :_0.002 3 and c = 7.065 t 0.009 3. Confirmation of the formula

by chemical analysis has not been completed.

AN INVESTIGATION OF THE EUROPIUM
OXIDEFLUORIDE - EUROPIUM SESQUIOXIDE SYSTEM

By

Sandra Leonard Bacon

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Chemistry

l97l

 

The ant
Jr. Harry A.
fins investi
797:6”a24re
C. Biefeld i

To the .
SO patient,

The fine

U-3- Atomic

‘._’| (3

ACKNOWLEDGMENTS

 

The author wishes to express her sincere appreciation to
Dr. Harry A. Eick for his guidance and assistance in the course of
this investigation. Also, the aid of the members of the High
Temperature Group, Dr. R. Seivers, D. E. Work, A. V. Hariharan and
C. Biefeld is most gratefully acknowledged.

To the author's husband, Terrance, a special thanks for being
so patient.

The financial support of the National Science Foundation and

U. S. Atomic Energy Commission is gratefully acknowledged.

ii

 

I
L L ‘

lllllll

5- Cha

D. Pha:
l.
2.
3.

E' ThEr
l.
2.
3.

EXPERIME.

Equ;

I.
II.

III.

TABLE OF CONTENTS

INTRODUCTION .......................
THEORETICAL CONSIDERATION ................
A. Vaporization .....................
l. Technique and Theory ...............

2. Limitations of the Knudsen Technique .......

a. Orifice Shape and Cell Geometry .......

b. Orifice Effects ...............

c. Vaporization Coefficient ...........

d. Diffusion ...................

e. Sticking Coefficient .............

f. Interaction with Knudsen Cell .........

B. Characterization of the Gaseous and Condensed Phases .

l. Mass Spectrometry ................
2. X-Ray Fluorescence ................
C. Temperature Measurements and Corrections .......
D. Phase Relationships .................
l. Possible Composition Range for EuOF-Eu203 .....
2. Phase Rule ....................

3 Modes of Vaporization and Their Relation to the
Phase Rule ....................
E. Thermodynamic Calculations ..............
l. General Considerations ..............
2. Third Law Calculations ..............
3. Calculation and Estimation of Thermal Functions. .
a. Free Energy Functions from Vibrational Data. .
b. Estimation of Heat Capacities for Solids . . .
c. Approximation of Standard Entropies of Solids.
EXPERIMENTAL .......................
A. Equipment ......................

l. Target Collection ................

5
6
8
8
9
TO

10

10
ll

l4
l6

l6
l7

l7
l9
19
21
23
23
25
25

26

 

 

$11th

K)“

IV.

2. X-Ray Fluorescence Equipment ............ 26
3. X-Ray Diffraction ................. 26
4. Temperature Measurement .............. 27
5. High Temperature Mass Spectrometry ......... 27
a. Spectrometer ................. 27
b. Mass Spectrometer Cells ............ 27
6. Micrograph ..................... 27
7. Heating Apparatus ................. 27
B. Preparative ....................... 28
l. Europium Oxidefluoride ............... 28
a. Europium Sesquioxide .............. 28
b. EurOpium Trifluoride .............. 28
C. Analysis . . . . .................... 29
l. Powder X-Ray Diffraction Techniques ........ 29
2. Chemical Analysis ................. 30
a. Europium .................... 30
b. Fluorine .................... 30
c. Oxygen ..................... 31
D. Vaporization Techniques ................ 3l
l. Target Collection Method .............. 3l
a. General Procedures Followed for All Experiments. 3l
b. Measurement of Orifice Area ........... 33
c. Specific Conditions for the EuOF Study ..... 33
d. Analysis of Effusates Collected on Targets . . . 33
E. Characterization of the Mode .............. 35
l. Effusate ...................... 35
a. Collection ................... 35
b. Weight Loss ................... .35
c. Mass Spectrometry ............... 35
2. Residue ...................... 36
3. Use of X-Ray Fluorescence to Detect Impurities or
Secondary Reactions ................. 36
4. Reaction of EuF3 with Different Cell Materials . . . 37
RESULTS .......................... 38
A. Results of Preparative Techniques ,,,,,,,,,,, 38

iv

l. Europium Trifluoride ................ 38

2. Europium Oxidefluoride ............... 38
B. Analytical Results ............... . . . 38
l. X-Ray Diffraction . ............. . . . 38
2. Chemical Analysis ................. 39
C. Results of X-Ray Fluorescence Calibration . . . . . . . 40
D. Vaporization Results ................. 40
l. Mode of Vaporization ............ . . . 40
a. Effusate ................... 40
b. Residue .................... 43
c. Attempts to Produce New Phases ......... 44

2. Reaction of Europium Oxidefluoride with Cell
Materials ..................... 44
a. Molybdenum ............... . . . 44
b. Tungsten .................... 45
c. Thoria ..................... 46
d. Platinum ................... 46

3. Reaction of Europium Trifluoride with

Molybdenum and Platinum .............. 47
4. Results of Vaporization Experiments ........ 48
a. Effects of Varying Temperature and Orifice Size 48
b. Results of Constant Temperature Runs ...... 49
E. The Pressure Equation ........ i ......... 5l
F. Thermodynamic Values .................. 5l
l. Europium Oxidefluoride (solid) ........... 52
2. Europium Trifluoride (solid) ............ 53
3. Europium Trifluoride (gas) ............. 53
4 Europium Difluoride (solid and gas) ..... . . . 54
G. Thermodynamic Results ........... . . . . . . 55
l. Results of Various Estimations ........... 55
2. Results of Thermodynamic Reduction of the Data. . . 56
V. DISCUSSION. . . ...................... 57
A. Preparative Work .................... 57
B. Characterization of Mode ................ 57
C Interpretation of Pressure Measurements ........ 59
D Effect of Cell Material on the Reaction ........ 6l
l. Molybdenum ..................... 6l

V

 

EIBLIQGRAPHY
APPENDIX I . .
maisrx II.
IFH’P"
.43.qu III .
! Pm
mam Iv. .

I v
EPPENDIX V .

2. Tungsten ...................... 62
3. Thoria ....................... 63
4. Platinum ...................... 65

E. Unanswered Questions and Their Possible Resolution . . . 65

1. What is the Reaction With Molybdenum? ........ 65
2. What is the Mode of Vaporization of EuOF? ...... 66

3. Does EuF (s) + EuF3(g)? How Does This Reaction
Affect tge Vaporization of EuOF? .......... 66
4. Why the Difference Between SmOF and EuOF? ...... 67
5. Further Comments .................. 68
BIBLIOGRAPHY ........................... 69
APPENDIX I ............................ 73
APPENDIX II ............................ 76
APPENDIX III ........................... 79
APPENDIX IV ............................ 80
APPENDIX V ............................ Bl

vi

 

 

N
3‘ .
IT)

11. Inte
Effu

IIL Rep:

TABLE OF TABLES

I. Weight Loss Data - Percent EurOpium ............ 39
II. Intensity and Appearance Potential Data of the

Effusate of EuOF ...................... 43

III. Reported Thermodynamic Values ............... Sl

vii

The e
attributed
oxidefluor

Zacharia 351

(.1 )

“filo

l";
l. itiuorj .3

Lie.w

I. Introduction

The earliest reported study of the lanthanide oxidefluorides is
attributed to Klemm and Klein], whose investigation revealed that the
oxidefluorides have a cubic fluorite type lattice. Subsequently,

. 2
Zachariasen

reported a rhombohedral structure for all oxidefluorides
except those of Ce, Tm, Yb and Lu. This rhombohedral structure is now
accepted as the correct room temperature phase3 and the cubic phase
found by Klemm and Klein is considered the high temperature modification.
Both stoichiometric LnOF and slightly fluorine rich samples exhibit

the cubic phase with the fluorine rich specimens retaining the cubic
structure more readily upon cooling to room temperature. The transition
temperature for the rhombohedral to cubic change (about 500°) is
relatively cation independent.4’5

Other studies on the lanthanide oxidefluorides have shown regions

of nonstoichiometry and solid solution in the composition range LnOXF
3,6

3-2x
in which O:x_l, and Ln = La, Nd, Sm, Eu, Gd, Er and Y. The regions
where 0.5:xgl appear to be distortions of the basic fluorite structure.
In some cases (La, Y) a tetragonal distortion is evidentz, whereas in
others6 an orthorhombic distortion is observed. In either case, the
distortion may be explained by occupancy of the tetrahedral holes
present in the fluorite lattice by excess fluoride ions. The ability
of fluoride ions to occupy these holes is further evidenced by the

region of solid solution between europium oxidefluoride and europium

trifluoride.3

Iguanas in”

 

 

 

lanthanide
prepare E03
heating the
of starting
that the cor
sesquioxide
630‘ and its
the Eu203 re
monoclinic t1
In their phag
Work and

cxidefluoride

and renorted a
“EOdmium oxid
reactjom my
oxidefluorides

In Light 0
0n EUroplum oxi
itappeared tha1
lieu of Catalang

y. .
apomzatlon wor

Sir
nee europium h-
C

t

the possibmty o

-2 _
A detailed study has not been made on the lanthanide sesquioxide-
lanthanide oxidefluoride systems. Catalano and Bedford3 attempted to
prepare Eu304F by mixing stoichiometric amounts of Eu203 and Eu0F and
heating the mixture to l600°. All attempts resulted in the retention
of starting materials with no evidence of new phases. They concluded
that the composition range between EuOF and Eu203 contained monoclinic
sesquioxide at all temperatures and rhombohedral oxidefluoride below
600° and its cubic form above 600°. They add that in several experiments
the Eu203 retained the low temperature cubic form above its normal
monoclinic transition temperature, but did not include these results
in their phase diagram.
Work and Eick7 have studied the vaporization behavior of samarium

oxidefluoride. They found the mode of vaporization to be:
3SmOF(S) = Sm203(s) + SmF3(g) (1-1)

and reported a AH°298 and AS°298 for this process. In a study of
neodymium oxidefluoride, Shinn4 observed a comparable decomposition
reaction. Discussions on the decomposition of other lanthanide
oxidefluorides cannot be found.

In light of the relatively small amount of information available
on europium oxidefluoride and the previous vaporization study of SmOF,
it appeared that a vaporization study of EuOF would be of interest. In
lieu of Catalano and Bedford's discussion of the phase diagram and the
vaporization work on SmOF, major difficulties were not anticipated, but
since europium has the most stable divalent state of all the lanthanides,

the possibility of EuF2(g) must be considered.

-3-
At the outset of this work, the vaporization of EuOF was considered
a practice system in which to learn the various techniques in hopes of
continuing the study to condensed EuFZ. As it turned out, EuOF became
the whole project. Many interesting features of the system have been
uncovered, but I have been unable to solve the complex EuOF-Eu20 -EuF

3 3
system.

m”?

 

 

This
technique 1

If a 5
Cf molecule

21 ma] be r

II. Theoretical Considerations

A. Vaporization

 

l. Technique and Theory_

 

This vaporization study was carried out by the Knudsen effusion
technique combined with target collection.

If a gaseous species is confined in a closed container, the number
of molecules which strike a unit section of the container wall per unit time,

2, may be represented by
z = n374 molecules/cm2 sec (II-l)

where n is the total number of molecules and U'is the average molecular
velocity.8 If an infinitely thin orifice is drilled into the container,
the number of molecules which would escape through this orifice is

given by (11-2)

N = $02 molecules/volume (11'2)

where So is the area of the orifice.

The effusion of molecules through such an ideal orifice follows
the cosine distribution law.9 A circular target of radius, r, which is
placed coaxially with the orifice at some distance, d, will collect
that part of the molecules whose trajectory lies in the cone defined by
the target and the orifice. By integrating the cosine distribution
law over the volume represented by this cone, the fraction of the total

-4-

-5-
effusate which strikes the target,

2 2

r /(d + r2) (II—3)

is obtained.
Finally, ideal gas behavior may be assumed for the vapor within
the cell as long as conditions of molecular flow are obeyed (pressure

3 atmospheres).8 When equations II—l and 11-3 and from

less than l0-
kinetic theory (8RT/n)1/2 for C'are substituted into the ideal gas law,

equation (II-4) is obtained.
P = [w/sotit2NRT/MJ‘/2[<d2 + r2)/r2] (11-4)

Conversion of all constants to give pressure in units of atmospheres

yields equation (II-5)
P = 3.76 x 10‘4 w/Sot [T/n]‘/2[(d2 + r2)/r2] (II_5)

in which w represents the mass on the target expressed in grams, T, the
temperature in degrees K, M, the molecular weight of the effusing species

and t, the time in seconds during which a target is exposed to the beam.

2. Limitations of the Knudsen Technique.

a. Orifice Shape and Cell Geometry

A machined orifice cannot be made infinitely thin. For any orifice
of finite thickness, L, some deviation from cosine distribution will
occur since the molecules collide with the walls of the orifice and
some may return into the container. This return of molecules causes
the total effusate to be less than the theoretical amount and thus
l0

represents a lower pressure than that actually in the cell. Clausing

made a study of the effects caused by channel orifices and calculated

-5-
for various orifice length to radius ratios, L/p, correction factors which
compensate for the decreased amount of effusate. Pressure equation

(II-5) with a Clausing correction, W0, becomes
P = [3.76 x 10'4 w/Sot] [T/MJVZHd2 + r2)/r2][l/w0] (II-6)

A converging conicalH orifice with a knife edge is a closer
approximation to the ideal situation than is a channel orifice. If such
a conical orifice is combined with a coaxial target of sufficiently
small radius, r, that it collects l% or less of the total effusate, the

1] (cf.

Clausing correction in equation (II-6) may be taken as unity
figure l). The cone defined by the target and orifice (figure la)
contains the same amount of effusate for either the ideal or non-ideal
case since in this volume essentially only those molecules which travel
in a straight line from the surface of the sample to the target are
present. However, to calculate the total amount of effusate which would
effuse if the cell orifice were ideal some correction factor must be

applied. Freeman and Edwards12 have computed the Clausing factors for

converging conical orifices of various lengths, radii and conical angles.

b. Orifice Effects

 

To insure that the equilibrium within the cell is not disturbed
to such an extent by the effusion that only a steady state, and not an
equilibrium pressure is obtained, the diameter of the orifice must be
at least ten times smaller than the mean free path of the gaseous

Species.8

If the radius of the gaseous species is known, the mean
free path may be calculated and the appropriate orifice selected. If
such information is not known, the orifice size may be tested experimentally

by repeating pressure measurements at the same temperature with markedly

.mummsmmm mo coflusnflupmnm .H mosmnm

5:39:33 33:62 5:32:65 32: cotootoo .3 260

3 2 E

    

   

:8 2835‘ :6... contact

b20336 35236
2625. 26250
co 38 \o 266

 

 

392 lllllrlI—

E32

 

 

 

 

differen
pressure!
the orif'

Meas
Since ver
some char
study13 ni

indicated

-8 _
different sized orifices. If at the same temperature identical
pressures are calculated for these different orifices, the sizes of
the orifices are within the allowed limits.

Measurement of the orifice size takes place at room temperature.
Since very high temperatures are reached in the course of an experiment
some change in orifice area due to thermal expansion is expected. A

l3

study made to determine the correction factor for such expansion

indicated that this correction factor would be very close to unity.

c. Vaporization Coefficient

 

Other considerations which must be included in the limitations of
the Knudsen technique do not directly concern the physical set up.
One such consideration is the vaporization coefficient. This coefficient,
ae, may be defined as the ratio between the rate of vaporization into

a vacuum and the rate at which an equilibrium vapor impinges on the

14

sample. The observable effect of a non-unity vaporization coefficient

would be the absence of an equilibrium within the cell, and pressure
measurements would be sensitive to orifice area variation. Rosenblatt15

has shown that for finely divided samples, a is usually unity since

e
the effective vaporization area is large.

d. Diffusion

Processes which are diffusion controlled may hinder establishment
of the correct vapor pressure of a compound. Incongruent vaporization
processes provide an excellent opportunity for diffusion limitations to
occnn'since a new solid product is being formed in the course of an
revaporation and thus the surface composition of the sample is changing

constantly. Ackermann et_al,]4 have calculated that most of the

-9-
effusing molecules vaporize from the outermost atomic layer and therefore,
the changing surface composition might be expected to interfere and to
induce a diffusion controlled process.

Several experiments have been devised to ascertain whether or not
the effects of diffusion are significant. One such test involves the
use of samples of differing particle size in the course of several
experiments. If a systematic trend of increasing measured pressure is
found with more finely divided sample, diffusion may be occuring. A
second approach is to follow the pressure as a function of time at
constant temperature. In this case a drop in pressure with time
indicates diffusion control of the evaporation. Yet another technique
is to vary the temperature in different ways in the course of an
experiment. For example, a series of targets would be collected with
the temperature successively increased, then another series would be
taken as the temperature was decreased through the same temperature
range. If the measured pressures are lower for the series of targets

where the temperature was decreased, diffusion may be present.

e. Sticking Coefficient.

 

Another error in pressure measurement may occur if the effusate
does not adhere to the target. Liquid nitrogen is used to chill the
target so that the energy of the striking molecule will be removed
quickly, but in some instances this cooling may be inadequate. By
placing a chilled disc with a hole in its center in front of the target,
any molecules which pass through the hole in the disc and do not adhere
to the target, but reflect from it should be trapped on the back of the
disc. The amount of effusate collected on the back of the disc can then

be measured and a correction factor calculated.

-10-

f. Interaction with Knudsen Cells.

 

Interaction between the sample and the materials from which the
Knudsen cell is made should be avoided. Such an interaction causes
the thermodynamic activity of the species involved to change. If the
interaction can be accounted for, that is, if a chemical equation
representing the reaction can be written, the effect of the interaction
on the measured pressure may be predicted. For example, if an interaction
is occuring between one of the gaseous species and the cell in an.
incongruent vaporization where two vapor species are present, the
activity of the interacting molecule will be decreased. Since the
equilibrium depends on the activity of both species, the activity of
the non-interacting species will increase. The equilibrium constant
has not changed, but if, as in the typical case, only one of the
gaseous species is used to determine the vapor pressure, this

calculated equilibrium constant will be in error.

B. Characterization of the Gaseous and Condensed Phases

 

Equation II-5 derived to determine the vapor pressure depends
among other things on both the mass of the material collected on the
target and on its molecular weight. Various methods such as total
weight loss, torsion effusion, mass spectrometry, X-ray diffraction,
and chemical analysis all yield information concerning the nature of

the effusate.

l. Mass Spectrometry

 

Grimley16 gives a detailed discussion of the use of mass spectrometry
for characterization of the effusate. All ionic species present in the

mass spectrum can be identified uniquely from the knowledge of the

-1]-
possible combinations of the various isotopes and their percent
abundance. Determination of the neutral precursor for each ionic
species is not simple. From a plot of ion intensity versus ionization
energy, the appearance potential of the ions may be determined by
extrapolation of the linear portion of the curve near the threshold to
zero intensity (cf,figure 2). Appearance potentials of standards
such as mercury or nitrogen must simultaneously be measured to determine
the correction factor necessary to convert the ionization energy
readings of the unknown to appearance potential values. Accurate
values are obtained from this method only when the experimentally
determined curves approximate that illustrated in figure 2.

The appearance potentials can be used to identify the neutral
precursor, usually, by comparison to values obtained from related
systems for which appearance potentials are known. In complex cases
where similar systems have not been studied, additional information
must be obtained by disturbing the equilibrium. By adding (or
removing) certain compounds expected to comprise the effusate, the
concentration of the other components in the effusate change. If
this change in vapor concentration is followed through corresponding
intensity changes of the various ionic species, the neutral precursor

may become evident.

2. X-Ray Fluorescence

 

Once the vapor species has been established one may determine the
total mass collected on a target during the vaporization process.
In this study the amount of europium present on each target was
determined by X-ray fluorescence. A full description of the fluorescence

l7,l8

technique is reported elsewhere, but figure 3 shows the basic

 

)- -i
- -i
i— D —
- -1
i- D ..

.bi- -
3
Q” ‘
Q)
E L "
‘ n
h- c-i
'- "'i
.— D —r
Cl
C]
~ a
)- D _

 

 

[J
()L_J¥§da!3_;1ELLzzl J J J 1 J J .I J J J, I 1 _L
IO 30

Ionization Energy

 

Figure 2. Appearance Potential Curve

43-,

memEogpomam mucmommgo=_u Ammnx on» to orgasmsum < .m mgamwm

LOsomst

 

$3.6 autumn»:

 

Ml 235cm

 

 

 

 

 

 

 

I thl :\W
089:2.“

 

Ib‘l'
one. .8; -X

 

 

-14 -
features of the method. The X-ray tube emits radiation not only of the
wavelengths characteristic of tungsten, but also of a broad continuous
nature. This radiation strikes the target and excites electrons to
higher energy levels and/or is deflected. Electrons which return to

the ground state emit photons in the process of fluorescence. Electrons
located near the nucleus are not affected greatly by the chemical
environment and the energy they emit in going from an excited state to
the ground state will be identical regardless of what compound they

are in. Once the spectrometer has been optimized for a certain element,
these conditions will apply for any compound containing that element.
Detection of the fluorescence is accomplished by use of a single crystal.
Bragg's equation which relates wavelength, X, and hence energy, to

interplanar distances is given as
nX = 2d sine (II-7)

Since a single crystal is used, that is, d is fixed, the energies may
be detected by varying 29. Tables completed by Seivers19 relate the
various electron changes characteristic of an element to observed

26 values for a graphite crystal.

C. Temperature Measurements and Corrections.

 

The Knudsen cell is designed so that temperature may be determined
precisely with confidence. Since a target is situated above the orifice
temperature measurement through the orifice is impossible during the
course of the experiment. Therefore, an identical cavity is made in
the bottom of the cell to replicate conditions of the top LEE figure 4).
This lower orifice is then assumed to be a black body hole from which

any radiation emitted is characteristic of the temperature of the cavity.

-15..

SYMMETRICAL EFFUSION CELL WITH
SAMPLE AND OPTICAL CAVITIES

 

 

 

 

 

 

 

 

6/6 8

Figure 4. Symetric Knudsen Cell

-16-
An optical pyrometer is placed at right angles with a prism which is
located coaxially below the orifice through which the temperature is
to be read. According to Margrave,20 the radiation which is absorbed
by both the prism and the glass of the vacuum line may be accounted

for by application of Wein's and Beer's Laws (II-8)

(l/Tobs - l/T ) = an empirical constant (II-8)

act.

A 3 mm wide band filament lamp is used to determine the Wein's Law
constant by first reading the temperature of the filament and then
rereading it through the prism and glass base of the vacuum line.
Calibration data for each pyrometer have been furnished by the National
Bureau of Standards for the temperature range of 800° to 2000° in
accordance with the l948 International Practical Temperature scale

and are used to convert the pyrometer readings to the actual observed

temperature before the absorption correction factor is applied.

0. Phase Relationships

 

1. Possible Composition Range for EuOF-Eu203

 

No known compounds or regions of nonstoichiometry have been
reported at room temperature between the compositions Eu203 and
Eu0F3, but such observations do not preclude the existence of either
at high temperatures. Indeed, the general formula, Eu01+xF]_2X,
where 0 §_x 5_0.5, would include all possibilities. Previously
unreported compounds which exist at temperatures over l200° would become
evident in the course of a vaporization study, since the equilibrium
pressures will change as new equilibria are established. Knowledge

of the correlation between Gibb's phase rule and vaporization behavior

will help to clarify this point.

-17-

2. Phase Rule

 

The degrees of freedom, F, which exist for a particular system

may be ascertained from the phase rule,
F = C - P + 2 (II-9)

where P is the number of phases present at equilibrium and C represents
the number of components, the smallest number of independently

variable constituents needed to define the composition of all phases.21
Possible degrees of freedom are the variables temperature, pressure,
and composition. With this definition in mind one is able to

investigate the various possibilities which exist in a vaporization

study.

3. Modes of Vaporization and Their Relation to the Phase Rule

There are two modes of vaporization, congruent and incongruent.
A substance that vaporizes without changing its chemical composition
undergoes congruent vaporization. Since there are only two phases, one
condensed and one gaseous, and only one component by definition of
congruency, there will be only one degree of freedom. In other words,
fixing the temperature automatically sets the pressure once equilibrium
is established or vice versa.

In an incongruent vaporization the starting material decomposes to
a new solid product and a gaseous phase of composition different from
that of the solids. For most ternary systems which undergo
incongruent vaporization there will be one gaseous and two solid
phases, or P = 3. If an equation relating all constituents at
equilibrium can be written, C = 2 and F is unity. Experimentally this

fact is evident when a plot of nnP versus l/T yields a straight line.

-l8 _

In a comparable ternary case in which P = 3, and C = 3, two
degrees of freedom exist. Composition along with temperature or
temperature along with pressure must be fixed to yield a unique
pressure or composition, respectively.

Still another possibility in an incongruent vaporization is the
existence of a solid solution which forms at high temperatures. For
a ternary system formation of this solution would decrease P from
three to two. Two degrees of freedom would exist when C equals two.
If composition could not be fixed in the course of a vaporization
experiment, a plot of tnP versus l/T would not necessarily yield a
straight line and, in fact, shouldn't. To obtain any meaningful
thermodynamic data under these conditions one of the variables must
be held constant. From an experimental viewpoint, it is easier to
hold the temperature constant and to investigate the relationship
between pressure and composition than to hold composition constant.

If, as the phase rule indicates, there are only two degrees of freedom,
a linear relationship will exist between tnP and l/T for each composition.
It was stated earlier that application of the phase rule was
helpful in obtaining a phase diagram for a system. In actuality, the
experimental results of the vaporization experiments allow deduction

of the numbers of degrees of freedom. One can work backwards to
decide what phases and components are affecting the equilibrium. For
example, consider a ternary system (C = 2) which yields at constant
temperature a plot of pressure versus time which decreases initially,
then levels out, and finally drops again. For the horizontal region,
dP/dC = 0, F must equal l, therefore P = 3 or two solids and one

gaseous phase are in equilibrium. For the regions where pressure is

-19 -
decreasing there must be more than one degree of freedom; that is,

the composition must be varying. Such a situation is indicative of
solid solution formation. Figure 5 represents a phase diagram which

is consistent with these data.

E. Thermodynamic Calculations

 

l. General Considerations

 

As I have stated in the previous section, a linear relationship
is usually found between tnP and l/T when there is only one degree
of freedom in the system. From such a linear relationship one can
derive values for the vaporization process by use of equations (II-l0)

and (II-ll) on the assumption that the heat capacity change for the

AG° = -RTtnK (II-l0)

-tnKT = (l/T) (AHS/R) - ASS/R (II-ll)

vaporization process is constant over the temperature range of the study.
If the heat capacity change is not constant, its variability must be
considered in the equations.

The AH° and AS° values determined from the slope and intercept of
the linear plot are considered the enthalpy and entropy changes,
respectively, for the vaporization process at the mean temperature of
the study. Through thermochemical cycles these values from the slope
and intercept must be reduced to standard temperature by use of heat
capacity data. Such a data reduction is known as the "second law"
approach and is described in detail by Haschke.22

These Ax°298 values provide information on the relative stability

of the components of the reaction and allow for comparison to other

reactions.

-20 -

 

 

 

 

 

 

A203 AOF
"‘ solution "'
solution solution
A203 + AOF
Composition
A20:5

Figure 5. Hypothetical Phase Diagram

AOF

-2] -

Another approach to data reduction is the "third law method" in
which either derived or estimated free energy functions are used in the
calculations in conjunction with the tnK values experimentally
determined from each data point. The third law technique is generally
more applicable since a linear relationship between tnP and l/T need
not always be established before meaningful thermodynamic values can
be obtained. In a situation where there is formation of a solid
solution so that composition is changing constantly as vaporization
proceeds, it is theoretically possible to establish the tnP to l/T
relationship for each composition. Experimentally this task is
extremely difficult. Each target represents the average pressure of
some average composition during the collection time period. It is
most difficult to repeat experiments so that successive pressure
measurements represent identical average compositions. Extrapolation
procedures can be used to determine the pressure at each composition,
but the extrapolation introduces more uncertainty in the final values.
Only one assumption, that enthalpies of formation vary only slightly
with changing composition, needs to be made to use the third law to
obtain AH°298 values. For a discussion of this method, the reader is
referred to Haschke's dissertation.22

A final approach to data reduction involves calculation of the
activity of one component in a solid solution by determining the ratio
of the pressure of that component in the vapor to its standard pressure

at the same temperature.

2. Third Law Calculations

 

The third law method employs the free energy function, fef, as

a means of obtaining a AH°298 value for each data point. The free energy

-22-

function is represented by equation (II-l2) or (II-l3)
fef = ( G°T - H°298)/T (II-12)

fef=(H° -H

T °298VT ' ( S°T ' 50298) ‘ So298 (11"3)

The change in free energy function, Afef, may be calculated from
the values of fef for each product and reactant by equation (II-l4) where
v represents the stoichiometric coefficients and i and j represent
products and reactants, respectively.

Afef = gvifefi - gvjfefj (II-l4)
The free energy functions of many compounds have been tabulated and for
those not tabulated fef may be readily calculated from equation (II-l3).
The necessary enthalpy, entropy and standard entropy values may also
be approximated (of:II E, 3).

Free energy functions form smooth curves when they are plotted
against temperature and therefore allow interpolation.23 Variation of
composition may also be incorporated into the fef by interpolating
values from a plot of fef versus composition, provided several points
are known or may be estimated to establish the curve.

Once a Afef has been determined for a characterized reaction at
temperature intervals, the calculation of AH°298 values proceeds in a

straightforward manner through equation (II-15),

Afef = (AG°H - AH°298)/T (II-15)

or through use of equation (II-l6) obtained by substitution of

-23...
equation (II-l0) into (II-15).

AH -(Afef + RtnK )T (II-l6)

0298 = T

Alternatively, if AH°298 and Afef are known, tnK can be calculated at

selected temperatures to check hypothetical equilibria.

3. Calculation and Estimation of Thermal Functions

 

In many cases, thermal data have not been determined for lanthanide
compounds, or if they are available they do not cover the temperature
range of interest. However, a statistical thermodynamic approach can
be taken to obtain the appropriate free energy functions if a vibrational
study has been made on the gaseous molecules. For solids, Kopp's

23

Rule may be applied to determine the heat capacity for temperatures

greater than 298 K.

a. Free Energy Functions from Vibrational Data

Thermal functions of a compound can be divided into the translational,
rotational, vibrational and electronic components. By use of the
partition function the amount of energy contributed by each of these
components may be calculated. A detailed discussion of the mathematics
involved in deriving the entropy by use of the partition function can
be found in most Physical Chemistry texts. However, a few of the more
important features of the equation will be pointed out here.24

Since the energy level distribution is not known, the assumption
is made that the molecule is in its ground state, thus for the

electronic contribution, S , is given as Rtngo where 90 is the

25,26

elec

ground state multiplicity. From previous studies the statistical

weight of go is taken as l6 and l4 for EuF2 and EuF3, respectively.

-24 -
The symmetry of the molecule must also be considered in calculating
S°T and the term Rtno accomplishes this. For europium difluoride, o is
taken as two and for europium trifluoride, o = 3.
For the rotational contribution, the moment of inertia must be
known. Again, many texts give equations for determining the moments
of inertia for molecules of any shape. This study relied on the

27 If either internal free or

equations given by Moelwyn-Hughes.
restricted rotation occurs in the molecule, further adaptation must
be made to the rotational contribution.

Vibrational contributions to the entropy depend on the fundamental
vibrational energy levels of the gas. These must be experimentally
determined since comparison with similar molecules does not always
prove valid.

As in the case of entropy, heat capacity and (HT-H0) values may
be calculated if the structure and vibrational levels are known. The
same assumptions that the molecule is in its ground state electronic

configuration and has no internal free or restricted rotation must be

applied to validate equations (II-l7) and (II-l8) which are used for

non-linear molecules.28
3"’° 2 ho /kT hv /kT 2
Cp = 4R + Z R(hvi/kT) [e i /(e i - l) ] (II-l7)
i=l
3n-6
(H° - H° ) = 4RT +- Z [th./(ehvi/kT - 1)] (II-l8)
T O i=l 1
h = Planck's constant
k = Boltzman's constant
v = Vibrational energies
N = Avogadro's number

 

-25-
b. Estimation of Heat Capacities for Solids
To obtain thermal data at various temperatures, the heat capacity
of some compounds must be estimated. For high temperatures, Kopp23
has determined that each atom of a solid contributes 3R to the heat
capacity. The following equations22 can be used to estimate the heat

capacity of a binary or ternary system if the heat capacity of another

binary system with the same heavy element is known.

Cp (AXB )

y [X/UI [Cp(Aqu)] + [(uy - VX)/2] [3R] (II-l9)

Cp (AXB C2)

y [X/u] [Cp(Aqu)] + [(uy - VX)/2 + 2] [3R] (II-20)

The symbol A represents the heaviest element and the two compounds

must have similar ionic character.

c. Approximation of Standard Entropies of Solids
Latimerz9 measured the entropies of many solids and concluded that
each element makes a specific contribution to the entropy for each
oxidation state, regardless of the compound in which the element is

bonded. Westrum30

extended Latimer's study to the rare earth oxides
and developed a set of lattice contributions specific to these elements.
He also estimated the magnetic contribution for each lanthanide. By
cxnnbining Latimer's and Westrum's work, the following equation may be

used to estimate the entropy of lanthanide solids.

30298 = gx1$°.i + ExiMi (II-21)

M.

1 magnetic contribution

x. subscript in the molecular formula

1

 

III. Experimental

A. Equipment

l. Target Collection
13

The apparatus used is basically the unit designed by Kent with
the following two exceptions. l. A copper plate has replaced the
quartz shutter. 2. The collimator is no longer used to determine
the angle that the target subtends. Instead, the area of the target
seen by the X-ray fluorescence beam and consequently the solid

collection angle is limited by a 45° beveled copper insert placed in

the target.

2. X-Ray Fluorescence Equipment

The instrumentation used in quantitatively analyzing the collected
effusate was a Norelco Vacuum fluorescence spectrometer equipped with
a graphite analyzing crystal, a tungsten tube powered by a Norelco

XRG 5000 generator, and a Harshaw NaI(Tl) scintillation detector and

associated electronics.

3. X-Ray Diffraction

An 80 mm radius Haegg-type focussing vacuum X-ray diffraction
camera“ was employed to give crystallographic analyses of powdered
samples. Radiation emitted by a copper X-ray tube powered by a Picker
generator is diffracted from a quartz crystal to obtain the Ka
cxnnponent. This component is directed through the powdered sample and

diffracted by it.
-25 -

-27 -

4. Temperature Measurement

National Bureau of Standards calibrated Leeds and Northrup
disappearing filament-type optical pyrometers (serial numbers l6l9073,
l524388, and l572579) were used to determine the temperature of the
Knudsen cell. For other experiments where knowledge of the temperature
was necessary, either a platinum platinum-l0% rhodium thermocouple
with a Honeywell potentiometer (model 27-20) or a chromel alumel

thermocouple with a Sym-Ply-Trol meter was used.

5. High Temperature Mass Spectrometpy

 

a. Spectrometer

 

Investigation of the effusate was accomplished with the Model

l2-l07, Bendix Time-of-Flight instrument which together with the

associated heating circuits has been described by Pilato.32

 

b. Mass Spectrometer Cells
All studies performed in the mass spectrometer were effected in

molybdenum effusion cells of the design prescribed by Pilato.32

6. Micrograph

 

A Bausch and Lomb Dynazoom micrograph fitted with a Polaroid 4 x 5
Land Camera attachment was used to photograph the cell orifice, which

was illuminated by a hi-intensity external light source.

7. Heating Apparatus

 

Heating of the effusion cells was accomplished by a 20 KVA
Thermonic 200-400 KHz induction generator. Additional heating,
especially that required for cell outgassing and for some preparative
work, was carried out in a high speed vacuum line fitted with a

current concentrator and powered by another similar induction generator.

”Em—'3

-28 -
Laboratory tube furnaces and a Marshall Products Co. platinum-40%

rhodium wound tube furnace were also employed.

8. Preparative

 

l. Europium Oxidefluoride

All preparations of europium oxidefluoride involved the general
method1 of mixing stoichiometric amounts of europium trifluoride and
either the monoclinic or cubic form of europium sesquioxide, and
heating the resulting mixture to 600° in a horizontal tube furnace for
several hours. Platinum boats were used to contain the reactants in
a Vycor tube through which helium previously dried through a P205
tower and a liquid nitrogen chilled trap flowed continuously during
the heating process. All reactants and products were handled in air

and stored in dessicators.

a. Europium Sesquioxide

Europium sesquioxide (99.9% pure, Michigan Chemical Corp.) was
calcined in a muffle furnace at 900° to remoVe absorbed gases. The
body-centered cubic form resulted from this heating. To obtain the
monoclinic form the sesquioxide was heated to above l200°. The
monoclinic phase does not readily convert to the cubic form when it
is cooled. Since the calcining of a sample may decrease surface
area and hence its activity, very finely divided sesquioxide was
prepared by thermal decomposition of europium oxalate at 400° in air.

b. Europium Trifluoride
33

 

Two methods were used for the preparation of europium trifluoride.
The first involved the mixing of europium sesquioxide with a ten-fold

excess of ammonium bifluoride. The NH4HF2 was made by dissolving

 

-29-
NH4F (Baker Analyzed Reagent) in 48% hydrofluoric acid (Allied Chemical)
and boiling off the excess water. A platinum boat which was situated
in a Vycor tube contained the mixture while it was heated to 300° for
ten hours under a flow of helium.

The other method involved the precipitation of europium trifluoride
hydrate and subsequent removal of the waters of hydration. Europium
sesquioxide was dissolved in hydrochloric acid and then filtered to
remove any insoluble residue. Addition of 48% HF reagent to the
filtered solution caused precipitation of the trifluoride hydrate which
was then centrifuged and decanted. After several washings with dilute
HF and subsequently with ethanol, the trifluoride was confined in a
platinum boat and was dried at llO° for several hours. The dried
material mixed with a 6:l mole excess of NH4F and was then placed in

a tube furnace and heated to 600° for 4 hours.

C. Analysis

1. Powder X-Rapriffraction Techniques

 

The general technique of sample preparation and film measurement

3l

for a Guinier camera is given by Stezowski. Powdered platinum was

used as the internal standard. The diffraction data were reduced
with the aid of a least squares program written by Seivers for a
Hewlett-Packard 9lOOB calculator and the program ANIFAC written by

35

Larson, Roof, and Cromer. This latter program which was run on

a CDC 3600 computer calculates from known atomic positions, scattering
factors and estimated thermal parameters, the intensity and sin 26
values for each reflection. An absorption coefficient was estimated.

In most cases published structural data were used for comparison.

 

-30 -

2. Chemical Analysis

 

a. Europium

Europium analysis was accomplished by weight change upon conversion
of the sample to the sesquioxide. Weighed samples were placed in
platinum crucibles which had been heated to constant weight. These
crucibles were in turn placed in a Vycor tube and were heated to 950°
for 3 to 6 hours, depending on the size of the sample. A flow of
oxygen, which was saturated with moisture by bubbling it through

water, was maintained through the tube.

b. Fluorine
The amount of fluoride present in Knudsen cell residues was
determined by difference from a knowledge of the composition and total
mass of the effusate and the composition of the starting material. To
confirm this difference determination the HF produced by the stream of
moist oxygen described above was bubbled through a dilute sodium
hydroxide solution. This solution was then analyzed for fluoride

36 In this procedure the

content by a method devised by Megregian.
bleaching of a zirconium eriochrome cyanine R complex by the fluoride
ion is measured quantitatively with a spectrophotometer.

Standard fluoride solutions were prepared from NaF within the
range of O to 70 ug F750 ml of solution. A 5 ml sample of zirconium
reagent and 5 ml of the eriochrome cyanine R solution were added to
the 50 ml aliquots and then absorbance readings were taken at 527.5 mu
on a Cary l4 spectrophotometer. The sodium hydroxide solutions were

treated in the same manner and concentrations were calculated from

the least squares fit of the absorbance readings of the standard.

 

-3]-

c. Oxygen

No direct tests were made for oxygen content.

0. Vaporization Techniques

 

1. Target Collection Method

 

a. General Procedures Followed for all Experiments

 

Outgassed Knudsen cells of the design shown in figure 4 were
supported within a glass vacuum line (Ef.figure 6). Coaxially above
the cell, ten to twelve copper targets were suspended in a metal
magazine which was cooled with liquid nitrogen. The target-orifice
distance was measured with a cathetometer (Gaertner Scientific Co.).
The cell was heated by induction once the vacuum within the line had

decreased to lo-6

torr. Temperature was increased at a rate such

that the vacuum remained below lO'4 torr. When the same temperature
could be read successively, a state of equilibrium was assumed and a
target was exposed for the time period calculated from data collected

in an initial experiment. This time period was recorded on a laboratory
timer (Lab Chron) and varied between 2 and l20 minutes. The target

was then ejected and the next fell into place. In cases where long
times of collection were required a Latronics ColoRatio Pyrometer and

a Leeds and Northrup temperature controller were used to control the
temperature which was determined by optical pyrometry.

At the conclusion of a few vaporization experiments the target
magazine was replaced with an optical window and the temperature of the
upper orifice was compared with that of the lower orifice. Coil
position was also changed to determine its effect on the temperature

of the upper and lower halves of the cell.

 

-32-

 

 

 

 

 

 

Figure 6.

 

 

Vaporization Apparatus

 

-33 -

b. Measurement of Orifice Area

 

The area of the orifice in the photograph obtained with the
bench micrograph (lOO X) was measured with a polar planimeter
(Keuffel and Esser Co.) and then multiplied by 10'4 to give the real
area. The orifice area was determined both before and after each

experiment.

c. Specific Conditions for the EuOF Stugy

 

The majority of the experiments were effected in molybdenum

cells with orifice areas of 9, 38.2 and lOl x lO'4 cm2. A graphite cell

 

with a thoria cup liner in a molybdenum oven was used once and a
tungsten crucible was tested for possible use in a weight loss
experiment. Later tests were made with platinum containers either

by suspending the platinum tube in an induction field or by lining
molybdenum Knudsen cells with platinum cups. Both the residues and
collected effusates were checked for composition by X-ray fluorescence
and diffraction analysis.

From initial experiments performed with 0.25 9 samples data
points were taken at successively increasing temperatures followed by
decreasing temperatures (usually 25° increments). Later experiments
were carried out at a constant temperature for the duration of the
run. For these constant temperature experiments the sample size
(0.05 g) was such that l0 to l2 targets exposed in succession would

represent complete decomposition of the sample.

d. Analysis of Effusates Collected on Targets

 

The mass of europium in the compound deposited on the target
during a vaporization experiment was determined by X-ray fluorescence.

The fluorescence unit had been optimized for detection of europium by

-34 -
previous investigators in the course of their studies of europium
compounds. Optimization procedures are discussed by Haschke.22

The La energy level was followed by scanning its general region

l
and selecting the 26 value at which the peak maximized. This
procedure was repeated at the beginning of each counting session.

A standard curve of counts versus micrograms of europium was
obtained by counting targets which had deposited on them weighed
amounts of solutions of known concentration of europium. These
solutions were made by dissolving europium sesquioxide in HCl and then
boiling off much of the acid. The solution was then diluted with
distilled water in volumetric flasks. The most reproducible results
for obtaining a value of counts per microgram occur when the solution
is placed on the target in many tiny drops rather than a few large
ones. This procedure gives a more uniform coating and decreases
self-absorption in the spectrometer.

Each group of targets to be used in a vaporization experiment was
handled in the following fashion: Targets were cleaned in concentrated
hydrochloric acid. The collection surface was scraped with fine
sandpapper. The cleaned blank targets were then scanned to obtain a
background correction. One blank target was kept as a standard and
counted several times during each use of the fluorescence unit. Use
of this blank target permitted correlation of background intensity
from day to day. After a vaporization the targets were again counted
along with the standard blank and a target with known mass of europium
deposited on it so that variations in the counts/pg could be taken

into consideration.

I 4
H:- Fug-F.

 

ex;
evi

the

-35 -

E. Characterization of the Mode

 

l. Effusate

a. Collection

 

The effusate of an europium oxidefluoride vaporization was collected
in an inverted quartz cup which was supported over a Knudsen cell.
Collection temperatures were near the extremes of the range llOO-l500°.
The effusate was scraped from the cup and an X—ray diffraction pattern

taken of it.

b. Weight Loss

 

A tungsten Knudsen cell heated to constant weight was loaded with
0.25 g of europium oxidefluoride and heated under vacuum again until
constant weight was achieved. Guinier diffraction pictures were

taken of the residue.

c. Mass Spectrometry

 

Three different experiments were carried out in the mass spectrometer.
For all three, the system was pumped to a pressure of lo"6 torr after
the sample had been loaded. The cells were allowed to outgas below
l000° to ensure no loss of effusate. The temperature was then
increased as rapidly as possible so that the initial rapid rate of
effusion could be detected. The background pressure was maintained
less than 10'6 torr. Background scans were taken at various temperatures
by shuttering the cell effusate from the ionizing beam. In the initial
experiment the temperature at which effusate species were first
evident was established. The ion intensity of the EuF+ peaks was

then monitored as a function of time at constant temperature with a

£2: 50 ev ionization energy until most of the sample had been depleted.

 

-35 -

This same procedure was repeated in the second experiment. Appearance
potential data were obtained for all evident europium-containing
peaks, except Eu+, by measuring the ion intensity while the ionization
energy was decreased from 30 to 93, 8 av. The appearance potentials
of both mercury and nitrogen were taken to calibrate the ionization
energy scale. The temperature range of the various mass spectrometric

investigations was l200-l500°.

2. Residue

The residue of each vaporization experiment was subject to X-ray
diffraction analysis. The diffraction data were examined for the
presence of equilibrium phases and for changes in interplanar d-spacing
or their relative intensities. In some cases chemical analysis was
performed on the residue.

Attempts to prepare phases seen in the diffraction patterns of the
residue were accomplished by sealing various mixtures of Eu203 and
EuOF in quartz tubes under a partial pressure of helium. At first no
container was used, but later the mixture was enclosed in a platinum
capsule before being placed in the quartz ampoule. These samples were
heated to temperatures between 800° and l200° in the Marshall furnace
for varying lengths of time. The ampoules were removed from the
furnace and immediately placed in cold water in hope of quenching a
high temperature phase. The cooled samples were then examined by

X-ray diffraction.

3. Use of X—Ray Fluorescence to Detect Impurities or Secondary
Reactions

Standard targets were made by placing small quantities of the

material of interest on the target with Scotch tape. Various 20

 

-37 .-
regions reported to be characteristic of the radiation emitted by such
elements as molybdenum, thorium, and platinum were scanned and
compared with the background evidenced by blank targets to find the
best region for detection. A powdered sample which was to be
analyzed for one of these elements was placed on a copper target with

Scotch tape and the selected 26 region was scanned again.

4. Reaction of EuF3 with Different Cell Materials

 

-..I-

Europium trifluoride was heated in both molybdenum and platinum
crucibles under a vacuum in the susceptor line to £2: l450° to check
for reaction with the cell. A Vycor cup was placed above the
molybdenum cell so that the effusate could be collected. The platinum
cell was fashioned in such a manner that only the bottom part containing
the EuF3 was heated by the susceptor. The rest of the crucible was
located out of the field so that it remained at a much lower temperature

and the effusate could condense there.

IV. Results

A. Results of Preparative Techniques

 

l. Europium Trifluoride

 

Composition of the trifluoride prepared by dehydration of the
precipitated trifluoride hydrate was confirmed by X-ray diffraction

(5f: IV-B-l ).

2. Eupppium Oxidefluoride

 

All preparative attempts yielded the oxidefluoride, but all
except one sample still contained some cubic europium sesquioxide.
This one exception which was the product of the mixture of EuF3 and
monoclinic Eu203 appeared as pure rhombohedral oxidefluoride in the

X-ray diffraction pattern.

B. Analytical Results

 

l. X-Ray Diffraction

 

Phase purity of a sample was determined by matching the position
and intensity of each diffraction pattern line with the values
produced by the program ANIFAC (EfEIII-C-l). For these calculations

37

the lattice parameters and atomic positions of EuF3, EuOF

(rhombohedral)?38 C—Eu20339 and B-Eu20340 reported by previous
workers were used. Isotropic thermal parameters contained in this
program were used for every compound except B-Eu203, for which the
reported values were substituted. An absorption coefficient of unity

-38-

 

_39 -
was assumed. The only impurity lines found in the rhombohedral
oxidefluoride were those of C-Eu203. The observed and calculated

sin26 values are listed in Appendix I.

2. Chemical Analysis

 

Well annealed cubic sesquioxide X-ray diffraction patterns were
obtained after pyrohydrolysis of the oxidefluoride samples. Weight

loss data are given in Table I.

Table I. Weight Loss Data and Percent Europium

 

EuOF

 

 

 

Sample % Weight Loss % % Eu
Eu 0 + EuOF
2 3

Prep 111 106.9 :_O.8 excess F 80.9 :_O.5

97.5 :_0.8 97.39 81.4 :_0.5
Prep IV 106.2 :_O.8 excess F 81.2 :_O.5

103.4 :_O.8 excess F 81.1 :_O.5
Theoretical - - 81.28

 

Several attempts to collect the hydrogen fluoride produced from
passage of the moist oxygen over commercial NaF samples at 800° proved
unsuccessful. Even after the apparatus had been compacted to minimize
the distance the HF gas had to travel and in which it could contact
the glass apparatus only 78% collection of the fluoride liberated
was obtained in the best results. Besides not collecting all the
fluoride, it was difficult to reproduce the rate of oxygen flow from

run to run. Therefore fluorine analysis was not effected.

-40-

C. Results of X-Ray Fluorescence Calibration

 

A set of copper targets with a varying number of micrograms
of europium was prepared by this researcher and a standard of
2407 :_50 counts/microgram established. This number was within the
standard deviation of the 2442 :_50 counts/pg found by A. V. Hariharan,
a co-worker studying EuClz. Excess acid in my standard solutions
caused some attack on the copper target and on repeated use of the
target loss of europium occurred. Therefore, the standard prepared

by Hariharan was used for the duration of the research.

0. Vaporization Results

 

1. Mode of Vapprization

 

a. Effusate

Collection of the effusate of europium oxidefluoride confined in
a molybdenum crucible resulted in a thin brown film on the collection
cup at both llOO° and 1500°. The diffraction pattern of both
collections exhibited diffuse lines which fit an fcc cubic pattern
(a = 5.78 :_0.01 A). The pattern of the sample collected at 1500°
also contained an extra line which did not fit the cubic indexing.

The results obtained from mass spectrometry are presented in
Table II and Figures 7 and 8.

No peaks corresponding to EuF3 could be found by scanning in the
temperature range 1290° to 1500° and by varying the ionization potential
from 15 ev to 30 ev.

The 19 and 38 mass unit regions were scanned in attempts to
identify F+ or F2+, but the normally high background in these regions
and lack of shutterability of these peaks made a definite assignment

impossible.

 

Intensity

 

 

51TITTTTIITTTTTTTFIT
n
)— d
n
L ..
I. a
on
n
i- a
n
1- DD -
L— EI ..
D
L-
1
n
n
c 4
a
n
I no '1
n

 

 

Ionization Energy,
ev

Figure 7. Appearance Potential Curve for EuF+

 

-42-

 

lTTleIT

1

T

TTTTWT

Intensity

T

I

I

 

fiTTTTTITlTTTIITTIT
fi
q
E]
D ‘i
D "'l
D u

l

l

lALl

l

 

l

l

llLLlJ

LL41

l

l

E]

l

L

[J

JJJlliilJJiJlllJLJL

.4

 

 

{Iii ‘F’T

29

Ionization Energy,
ev

Figure 8. Appearance Potential Curve for EuF2+

-43 _

Table II. Intensity and Appearance Potential Data of the Effusate

 

 

 

of EuOF
Shutterable Relative Intensities Appearance
Species (at 25 ev) Potential
151
Eu+ 33 -
153 -
t
+ 170 .
EuF 100 12.5 :_1 ev
172 e'
+ 189
Equ 24 10.5 :_2 ev
191

 

b. Residue

Residues of a vaporization which had been carried to completion
(i,e,, Eu203) showed X—ray diffraction patterns which were characteristic
of monoclinic europium sesquioxide plus a cubic pattern. The lines of
this cubic pattern were quite faint and only four or five lines could
be read. The indexing of these lines yielded a lattice parameter of
snoiomi.

Residues which had lost :_m 90% fluoride as determined by weight
loss, contained cubic europium sesquioxide and another cubic phase.
This latter cubic phase, when indexed on fcc symmetry exhibited a
lattice parameter of 5.50 :_0.0l A based on the five lines that could

be read.

-44 -

Any residue which had lost / w 90% of the fluoride showed
rhombohedral europium oxidefluoride and cubic sesquioxide. Often the
diffraction patterns of these residues had somewhat broad lines rather
than the sharp lines characteristic of a well-annealed substance.

The character of the residue seemed more dependent on the
percent of fluoride lost, than on temperature of decomposition,
although at higher temperatures the sample was more apt to lose all
the fluoride because of the increased vapor pressure.

It should also be noted that the color of the residue changed
from an off-white through purple to black as the fluoride content

was depleted.

c. Attempts to Produce New Phases

 

The attempts to prepare the two new cubic phases seen in the
residue by mixing various amounts of Eu203 and EuOF proved unsuccessful.
In addition, no substantial evidence for solid solution could be
obtained even when samples were heated in sealed platinum capsules
for times which varied from hours to several days, and were then
quenched. Rhombohedral EuOF and cubic sesquioxide were always present.
X-Ray diffraction patterns of some samples showed rather broad lines

on a random basis.

2. Reaction of Europium Oxidefluoride with Cell Materials

 

(cf Appendix IV)

a. Molybdenum

 

Molybdenum Knudsen cells showed no evidence of attack at the
end of a vaporization experiment in that the residue fell out readily
and all surfaces were shiny and smooth. Weight loss data was not

taken with a molybdenum cell, but later vaporization experiments were

 

-45-

weighed before and after decomposition. Several of these showed a
weight loss of greater than 100% of that expected by analogy to
equation (I-l).

The X-ray fluorescence spectrometer was exceedingly sensitive to
a small amount of pure powdered molybdenum at its K0] and K62
(26 = 12.13 and 12.21 respectively, graphite monochrometer) emission
region. When small amounts of the vaporization residue were scanned
in this region, results varied. Residues which had not changed much
in color (usually heated below 1400°) showed no evidence of Mo. 0n
the other hand, the deeper the color of the residue, the greater the
intensity of the fluorescence at peaks assignable to molybdenum.

One residue which showed B-Eu203 and the cubic (5.40 :_0.0l A)
pattern and gave positive results when tested for molybdenum was heated
in air. The black residue changed to bright yellow after heating, but
evidenced no change in the diffraction pattern. The small weight 1055
recorded upon heating could not be considered conclusive since the
original sample was so small (25 mg).

One sample of effusate which was collected at 1500° from a
molybdenum cell also showed the presence of molybdenum but the X-ray

fluorescence peak was not very intense.

b. Tungsten

When an 85 mg sample of eur0pium oxidefluoride was completely
decomposed in a tungsten cell, a weight loss of 99.4 :_0.8% was recorded
(gf,IV-D-2a). However, the mass of the tungsten cell decreased by
6 mg. The X-ray diffraction pattern of the dark blue residue showed
monoclinic Eu203 and a cubic pattern similar to the one seen when a

molybdenum cell was used. The X-ray diffraction lines of the cubic

-46-
phase in the tungsten cell were more intense than those observed in
any of the molybdenum cells, and enough lines were present to confirm
that the phase should be indexed on body centered cubic symmetry with
a lattice parameter 10.760 :_0.001 A.

c.T_h9_ri_a
Thoria reacted more completely with europium oxidefluoride than
either molybdenum or tungsten. After vaporization of EuOF at m 1400° l
the thoria cup had changed from white to yellow, and brown crystals [‘1
were growing from the sides and bottom of it. When the residue was 1
shaken loose from the cup, brown crystals could be seen growing out of
a yellow solid. No X-ray diffraction lines of either of these phases
corresponded with those of Eu203, EuOF or ThOZ. The brown residue
gave a definite hexagonal pattern (a = 9.590 :_0.002 A and c =
7.065 :_0.009 A). The pattern given by the yellow substance could not
be indexed and may represent more than one phase.
The presence of thorium could not be detected readily by X-ray
fluorescence. It did not produce sharp intense peaks when a powdered
sample was introduced in the X-ray fluorescence spectrometer; the
most intense peak was an La1 emission (26 = 16.4,graphite monochrometer).
When the brown sample was examined for thorium in this 26 region, no
peak could be discerned; examination of the yellow sample produced a

small peak.

d. Platinum
When europium oxidefluoride was heated to l400° in a long platinum
tube suspended in the susceptor line, no interaction with the solid

was apparent. The residue remained off-white in color and showed

-47 -
X-ray diffraction lines which correspond to rhombohedral EuOF and
cubic Eu203. The tube was suspended in such a manner that the effusate
could collect on the upper walls; however, only a thin gold-colored
layer appeared on the lid. This gold-colored layer was scraped off
and its X-ray diffraction pattern corresponded to that of two phases,
one cubic (a = 5.780 :_0.005 A) and one unidentified. X-Ray fluorescence
analysis of the residue for Pt gave negative results, but a piece of
the upper Pt wall showed intense europium peaks when it was rubbed [4f
clean and placed in the spectrometer.

Another test run was made by placing a platinum cup in a molybdenum .-

 

Knudsen cell gave very different results. No change was found in the
platinum cup after the Mo and Pt had been outgassed together at 1500°.
The EuOF was then placed in the apparatus and decomposition was
effected at l475°. When the cell was removed from the susceptor line,
the residue was dark in color and the now brittle platinum cup had
slot like holes in two places. A small piece of the platinum cup
which was placed in the X-ray fluorescence spectrometer yielded
positive results for molybdenum, but not for europium.

Fluorescence analysis of the effusate and residue evidenced trace
amounts of molybdenum, but their X-ray diffraction patterns showed
only a cubic pattern (a = 5.790 :_0.001 A) and a C-Eu203-rhombohedral
EuOF mixture, respectively.

3. Reaction of Europium Trifluoride with Molybdenum and Platinum

(pf,Appendix IV)

The condensed effusate of a sample of EuF3 heated in a molybdenum

cell at 1500° appeared as two distinctly colored phases. One which

possessed a pinkish beige color gave a fluorescence peak at 26 = 12.2

~48-
(graphite monochrometer) characteristic of molybdenum and an X-ray
diffraction pattern that could be indexed as cubic (a = 5.791 i
0.001 A) plus another set of unidentified lines. The second, a
dark brown phase, showed only the lines of the unidentified diffraction
pattern described above for the beige phase. This latter phase also
gave positive results for molybdenum.

The orange-colored residue showed evidence of melting. No further
tests were made on it.

When a platinum tube was used to hold the trifluoride, a clear
to white effusate was collected on the upper walls and an orange-gold
residue was left in the bottom. Neither effusate nor residue gave
any evidence of platinum by fluorescence spectrometry.

X-Ray diffraction analysis of the effusate gave lines which
could be matched completely with the reported trifluoride lines,
but the intensities of these lines did not correspond to those
reported. The residue also showed EuF3 and a second set of cubic

lines (a = 5.78 10.01 X).

4. Results of Vaporization Experiments

 

a. Effects of Varying Temperature and Orifice Size

 

The data collected from initial runs were calculated by use of
equation (11-5) on the assumption that Equ was the gaseous species.

The pressure of fluorine atoms was then calculated by equation (IV-1)

1/2

P (IV-l)

= P [M /M ]
F EuF2 F EuF2

The equilibrium constant, K, was derived and £nK was plotted

against 1/T. Plots of data collected in the initial runs made by

 

-49 -

increasing the temperature and then decreasing it failed to yield a
straight line. The points fell onto two lines, one line for ascending
temperature and another for descending temperatures. A few points
from a successive experiment might fall on one of these lines or they
might establish a completely new line. It was noticed that in all
runs, the last few data points taken seemed to drop below the previous
ones of that run regardless of whether temperature was being increased
or decreased. Also, when because of a mechanism failure the
experiment had to be interupted, the apparatus disassembled and the
run started over, the first target collected upon reheating had
greater amounts of europium than the last one collected at the same
temperature.

The size of the orifice seemed to have no affect on the

vaporization results.

b. Results of Constant Temperature Experiments

The temperature could be held within a 5° range over a period of
4 to 5 hours with or without the use of the two color pyrometer and
the controller. The controller would hold the temperature steady
for a long period, then deviate wildly and settle again to the
initial temperature, but the deviation was usually enough to cause
the data collected from the target to be rejected.

Figure 9 shows the results of a typical constant temperature
experiment. For some of these experiments the total amount of effusate
lost was determined by having some target exposed for the entire
period of heating. This procedure allowed the average composition

(If the solid material to be calculated for each target exposed.

-50-

 

.5 ITTTTTTTTINTITITTTTTTTIT

0 1
. 1652 4

—i

T
1

PD 1
"' "l
- -l
.E ~° a
5 * i
- E]
ii: 3
- a
E]
“:2“ i

T
J

l
L

E]
.- U .4
.. D .-
- El .1
r- -1

 

 

oLLLJLLJLIILLILJJJJLIIJLI

Time , minutes 275

 

Figure 9. Results of a Constant Temperature Experiment

-5]_
Appendix II gives the results of these runs calculated on the basis
of EuF2 and F as the gaseous species.

One exception to the drop in vapor pressure with time was
found when EuOF was mixed with an excess of monoclinic Eu203. The
pressure dropped off initially but then increased, dropped again and

appeared to be increasing on the last target exposed.

E. The Pressure Equation

 

A pressure equation has not been established for reaction (IV—2)

3EuOF(S) + Eu203(s) + EuF2(g) + F(g) (IV—2)

primarily because the pressure does not remain constant with temperature;
thus one equation can not represent the entire process. Secondly, in
light of the presence of molybdenum in some of the residues, it

appears that the pressure measurements may not represent the above

reaction. These comments will be discussed later (Efgv-B).

F. Thermodynamic Values

 

In Table III various relevant thermodynamic values are tabulated.

Table III. Reported Thermodynamic Values

 

 

Sample 4H°f 298 in kcal/gfw fef 1500 Reference
Eu203 cubic -386.5 - 41
EuF3(s) -39l.O - 42

-4ll.O - 43

-52-
Table III. (Continued)

 

 

AH°f 298 in kcal/gfw fef 1500 Reference
EuF2(s) -282.0 - 44
-294.0 - 33
F(g) 18.86 -42.201 45
F2(g) o -55.l67 45

 

Estimates of other thermodynamic functions were established in

the following manner:

1. Eurppjum Oxidefluoride (solid)

 

Free energy functions values were obtained for the oxidefluoride

by estimating the heat capacity and the S°298 value (gfl_Il-E-3b).

40

The heat capacity data given by Holley et_al, for the cubic form

of eurOpium sesquioxide were used to calculate equation (IV-3).

3

Cp (EuOF) = 19. l l l x 10- T -

35 i- '14 + ( '72 i- '02)
(1.86 :_l.36) x 10 /T

A plot of Cp (Eu203) versus temperature was linear to 1400 K, the
highest temperature at which data were reported. Due to its linearity,
extrapolation to the temperature range of this study (1500-1850 K)
seemed valid.

29

By use of Latimer's estimates for the entropy contributions of

oxygen and fluorine and Westrum's30 lattice and magnetic contributions

-53-
for Eu, (Ef:II-E-3c), 5°298(Eu0F) was calculated to be 22.1 :_1 en

where the error reported is estimated.

The AH°f 298(EuOF) was estimated by comparison with Work's7

reported value for SmOF (-274.6 :_2.5 kcal/gfw) and reaction (IV-4)

C‘M203 + MF3(S) + 3MOF(S) (IV'4)

The AH°r for M = Sm was calculated to be 126 :_14 kcal/gfw.41’43

41,43

This AHr Was assumed for the case M = EU to give AH f 298 =

-224 :_18 kcal/gfw with the error estimate calculated from
2 1/2
(4° errori ) ’

2. Europium Trifluoride (solid)

 

Heat capacity data have been measured for many of the lanthanide

47

trifluorides, but unfortunately not for europium. By use of

relationship (IV-5)
EU(S) + MF3(S) = EUF3(S) + M(S) (IV‘S)

with the assumption that nfef for the reaction is zero, the fefs
for EuF3 were interpolated. Free energy functions for M = Pr, Ce,

46 and five values were

Nd, Gd, and Dy were taken from Hultgren
obtained for fef (EuF3). These values were then averaged to give the
numbers listed in Appendix III.

3. Europium Trifluoride (gas)
48

 

A recent spectroscopic study of the structure of EuF3(g)
allowed a statistical approach to be used to obtain various thermodynamic
functions. Vibrational data were not taken on the gas, but rather

on various condensed matricies of the gas and of an inert gas.

 

-54..

Margrave §t_al,50 have pointed out that the data obtained with a
nitrogen matrix are a more true representation of the real gas than
those taken with any other type of matrix gas. Therefore, the
reported values48 of o], 62, v3 and 64 obtained in a N2 matrix were
used for calculating Cp° and H°T-H°298 values as discussed in
section (II-E-3).

Moments of inertia needed for calculating S°T require both a
value for the bond angle and a bond length. A value of 117° is

assigned48

49

to the angle F-Eu-F, but no bond length is reported. Other

workers have reported bond lengths for La, Y and Nd trifluorides on

50 of the various

the assumption of a linear model. By adding the radii
+3 metals to the radius of F- (1.36 A) a factor (sum of radii -
reported bond length) was established to be applied to the sum of

+3 and F" radii. This method yielded the bond length of 2.15 A for

Eu
EuF3. Moments of inertia were then calculated to be Ix = I =

y
2.22 x 10'38 38.

and I2 = 4.24 x 10' Values obtained subsequently for

S°T are included in Appendix III.

4. Europium Difluoride (solid and_gas)

Thermodynamic data have not been published on EuF2(s). Several
estimates do appear in the literature, but these values are quite
varied. Since the enthalpy of formation of EuF2(s) was needed, the
enthalpies of formation of the group II fluorides42 were used as a
guide. A value of AH°f 298 EuF2(s) (290 :_4 kcal/gfw) was chosen by
comparison since the enthalpies for the group II fluorides cluster

around this number.

0 . . . . 52
The AH sub 298 was determined by use of b0111ng p01nt data,

Cp(1 + g) (-10 eu) and Cp(s + g) (-6 eu), to be 102 kcal/gfw. From

-55-
these values AH°f EuF2(g) is found to be -188 :_10 kcal/gfw, where
the reported error is an estimate.
Other thermodynamic values for Equ gas were obtained by a
statistical treatment from the vibrational information reported by

Margrave, et_al,50

All parameters needed to calculate Cp°,
H°T-H°298 and S°T were taken directly from Margrave's paper. These

values are reported in Appendix III.

G. Thermodynamic Results

 

1. Results of Various Estimations

 

In the course of interpreting the pressure data taken in this
study, the possibility of several reactions was encountered. In an
effort to decide which was thermodynamically most possible, the
enthalpies of various reactions were investigated. Most crucial to

interpretation were reaction (IV-6) to (IV-8).

EUF3(g) + EUF2(g) + F(g) (IV-6)
EUF3(S) * EuF2(g) + F(g) (IV-7)
EuF2(g) + F2(g) + EuF3(g) + F(g) (IV-8)

. .. o _ 43
For equation (IV—6) combination of AH f 298(EuF3(g)) - -31O kcal

and AH°f 298(EUF2(g)) values of a) -l6O kcal/gfw43 or b) -188 kcal/gfw
(g:.IV-F-4) yield AH°298 values of 141 :_14 kcal/gfw or 169 :_14 kcal/gfw,
respectively. By combining these data with the free energy functions
established in the statistical treatment (Appendix III) an equilibrium
constant of 7.97 x 10'8 was calculated for reaction (IV-6) at the

highest temperature studied (1838 K). (Note that AH°f EuF2 is taken as

-188 kcal/gfw in this as well as in subsequent calculations.)

 

-55 -

The AH° for equation (IV-7) is found to be 242 :_14 kcal/gfw.

298
From this value and the free energy functions a AGOlB38 of
104 :_20 kcal/gfw is obtained.

Equation (IV-8) has a pressure independent equilibrium constant

and also contains the well established equilibrium constant for

reaction (IV-9).

F + 2F (IV-9)

2

The value of K1600 for reaction (IV-8) was found to be 7.47 x 1010

from the enthalpies and fefs discussed above.

2. Results of Thermodynamic Reduction of the Data

 

Since an equation could not be established definitely for the
reaction taking place in this study the data could not be reduced.
When the reaction is established methods discussed in II-E-2

may be applied for obtaining AH°298 values.

V. Discussion

A. Preparative Work

 

It is most interesting that a 1:1 mixture of monoclinic Eu203
and EuF3 produce phase pure EuOF while the same mixture of cubic
Eu203 and EuF3 always showed excess Eu203 in the final product. The
full significance of this behavior is not understood, but could be
due to the formation of solid solution between the cubic oxide and
EuF3. This possibility will be discussed later.

It should also be pointed out that the preparations which resulted
in excess sesquioxide were used for the bulk of the vaporization
experiments on the assumption that since Eu203 is one of the equilibrium
products, its presence would not interfere and, if anything, would
hasten attainment of equilibrium. Vapor pressure measurements bore
out the validity of this assumption, since both the pure EuOF sample

and those which contained excess C-form oxide gave consistent results.

B. Characterization of Mode

 

Mass spectrometry results seem rather conclusive that Equ is

one of the vapor species. Appearance potentials of the ionic species

53

generated by electron bombardment from EuF3 are reported to be

13.5 10.7 ev, Eurz”; 19.5 1 0.7 ev, Eur“; and 27.0 1 0.7 ev Eu+.
Although the appearance potential of Equ+ from EuF3 found in this
study (10.5 ev) is fairly close to the reported value that for EuF+

-57 -

-53 -
is markedly different. No break in intensity could be found around
19 ev as the ionization energy was decreased, an indication that
EuF3 was not providing any of the EuF+ species present. Other
evidence which supports the fact that EuF2 is the only neutral species

is the work done on the Group II fluorides.“-56

The appearance
potentials of these parent ions and the parent minus one fluorine
ion are very similar in value (x 12-13 ev). Also, in the study of
MgF2,56 the only species for which the MFZ+ intensity is adequate
for interpretation, the intensity ratio of the ionic species, Mg+,
MgF+, MgF2+, is very similar to that seen in this study.

The X-ray diffraction data of the effusate are not so conclusive
as the mass spectrometry. The cubic pattern which appears regardless
of temperature or cell material does not coincide with the lattice
parameter (a = 5.840 A) reported for EUF2.OO‘57 Catalano and
Bedford have completed a moderately extensive study on the phases
which exist between EuF2 and EuF3. They report a composition region
from EUF2.29'EUF2.44 which has an unsolved structure, but its strongest
lines can be indexed on an fcc cell with a = 5.78 A. The color of
the phases which possess this composition vary from dark brown to
brown. Both the lattice parameter and color correspond to the
effusate collected from EuOF. It is also noteworthy that this
same range of stoichiometry was duplicated time after time even though
the temperature of the collection surface was not controlled and the
temperature of the reaction was varied. The most reasonable explanation

for this consistency is that EuF2 and F are evolved and some interaction

occurs as the two hit the collection surface. The composition range

 

-59 -
Eu 2.29-EuF2.44 may be a particularly stable region and therefore
easily reproduced, and beyond the composition EuF2.44 only EuF3 is

an allowed second phase.

A mixture of EuF3(g) and EuF2(g) seems to be ruled out on the
basis of mass spectrometer results. If EuF3 were one of the species,
it should be more abundant at lower temperatures, yet no peak could
be found even at 1290°. However, the possibility of an EuF2(g) and
MoFx(g) mixture can not be ruled out. The effusate does evidence
molybdenum when examined on the X-ray fluorescence spectrometer. Mass
spectrometry was not useful in supporting this conclusion since the
unshutterable background in the region Mo+ to MoF4+ is very intense,
and molybdenum has a large number of isotopes.

The analysis of residue, on the other hand, proved to be more
straightforward. The oxidefluoride definitely decomposes to the cubic
form of the sesquioxide with no intermediate phases. The two other

cubic phases will be discussed in the following sections.

C. Interpretation of Pressure Measurements

 

The exponential type drop in pressure with time at constant
temperature may be explained by either a diffusion controlled process
(gf,II-A-2a) or a reaction which has more than one degree of
freedom (pf:II-D-3). A diffusion problem alone does not seem to be
the case since cooling a sample and reheating it to the same
temperature gives a pressure increase rather than that pressure
recorded before the cooling-reheating cycle occurred. Diffusion is
caused by a change in surface composition and cooling and reheating

should not change this composition. The pressure should be the same.

 

-50 -
All evidence points toward solid solution formation between EuOF
and Eu203, a case in which F = 2 (C = 2 and P = 2). As the solution
is formed the vapor pressure is lowered, and as the sample under
study becomes depleted in fluoride, the pressure falls off even more.
The tendency of pressure to increase on cooling and subsequent reheating
indicates that the solid solution is temperature dependent. Apparently
the solution is not formed completely as the sample is heated, so the
higher vapor pressure of the pure oxidefluoride must be present
initially. Another indication of the temperature dependence of this
solution is the fact that the cubic form of the sesquioxide and the
rhombohedral form of EuOF are present by the time the residue of a
vaporization has reached room temperature. Only two of the many
samples that were studied gave any X-ray diffraction evidence of the
solid solution, and those were the ones where the rhombohedral EuOF
had not completely grown in, but an fcc pattern (a = 5.50 :_0.01 A)
was observed instead. Since Shinn4 reports that cubic EuOF has a
lattice parameter of 5.535 A, this new phase is probably the solid
solution. The a = 5.50 A parameter is intermediate between that of
cubic EuOF and that of Eu203 indexed on fcc (a = 5.433 A) rather than
on bcc (a = 10.866 A)39 symmetry.
It is very reasonable that a solution would form between C-Eu203

and EuOF at high temperatures. Above 600° EuOF has a fluorite

4

structure and Eu203 itself has a structure which is based on fluorite

with only anion vacancies. A number of examples of fluorite structures

which form solid solutions have been reported in the 1iterature;3’6’59’60

many systems closely related to that under study have been proven to

have regions of solid solution. The most pertinent is the EuF -EuOF

3

 

-6]-
system studied by Catalano and Bedford.3 They point out the availability
of interstitial holes in the EuOF fluorite lattice, and they reason that
a pair of fluoride ions enters these vacancies as the solution is
formed. Initially as only a slight excess of F' is introduced the cubic
structure is distorted tetrahedrally and then as the fluoride concentration
is increased a solid solution which is stable at room temperature forms.

Other examples of fluorite type solid solutions are the EuFZ-EuF357

59 _Eu 0 60

system, the Th0 -ThOF system and the Th02 2 3 system in addition

2
to other MF —MOF6 systems. All of these observations substantiate the

3
ability of Eu203 and the fluorite type EuOF to form solutions.

Some comment on the stability of the solution formed between
Eu203 and EuOF can also be made from the results obtained in this
study. It is definitely temperature dependent. Apparently the
stability of the rhombohedral form of EuOF at room temperature is the
cause of this behavior. If one pictures the formation of solution
to result from the ability of the fluoride ions to migrate through the
Eu203 lattice at high temperatures and situate themselves in the
vacancies of the bcc structure as temperature is lowered, cubic EuOF
could result in the midst of Eu203. Brauer and Roether5 report that
M01.0F1.O (M = La, Nd, Sm and Gd) in the cubic form cannot be

quenched. Thus, since the solution may contain only EuO1 OFl O

surrounded by Eu203, it may not be quenchable.
D. Effect of Cell Material on the Reaction (cf Appendix IV)

1. Molybdenum

 

Molybdenum was originally chosen for use as cell material since

the study of SmOF had been completed in molybdenum with no apparent

 

-62-
interaction. No weight loss data were taken initially since the first
few X—ray diffraction patterns showed no new phases in either the
residue or effusate. The fact that EuF2 appeared as a gaseous species
was thought to indicate that EuF2 was the more stable gaseous species
and not that cell material was interferring. It was not until one of
the later vaporization experiments that the fcc phase (a = 5.40 :_
0.01 A) was seen. This cell is even smaller than those of EuOF or
the sesquioxide and it was hard to explain why a solution of these
two would cause a decrease in cell size. Molybdenum, which is a
smaller atom than europium, could be replacing some of the europium in

the cubic lattice and thus shrinking the cell size.

2. Tungsten

The reaction between EuOF and tungsten seems to confirm the
idea of interaction cited above. With tungsten, this new cubic phase
was more intense in the X-ray diffraction pattern, indicative that
tungsten was more reactive to EuOF than molybdenum. Enough lines
appeared in the pattern to define the cell as body centered cubic,
and therefore to double the cell parameter to a = 10.760 :_0.001 A.
This observation seemed to confirm that the bcc Eu203 was being
affected as was suspected in the case of molybdenum. A literature
search produced no reports of a cubic europium tungstate or molybdate,
however.

The mechanism of the reaction of EuOF with W or M0 is not yet
understood, but it is apparent that molybdenum appears in both the
solid and effusate. One explanation of this phenomenon is that
gaseous EuF3 escapes from the solid solution and interacts with the

cell walls forming some molybdenum fluoride and europium difluoride.

-53-

The molybdenum fluoride vapor can then either escape through the
orifice or react with the solid solution, displacing more europium
from it.

The results of a test in which EuF3(s) was evaporated in a
molybdenum cell gives some support to this premise; a substantial
amount of molybdenum was found in the effusate along with a new phase,
the X-ray diffraction pattern of which does not match any known
europium—fluorine phases. This observation is confusing since the

vaporization study of EuF3 was carried out in molybdenum or graphite.53

50

A more recent article has reported that reaction (V-l) occurs in

graphite cells, and my research gives similar results with molybdenum.

EuF3(S) + EuF2(g) + 1/2 F2(g) (V-l)

Catalano and Bedford3’57 report reaction with molybdenum in both

the case of EuF3 and fluorine rich samples of EuO]_xF1+2x and also remark
that the trifluoride is an unstable, ill-defined substance above

1400°. However, they do not elaborate in this later report as to what

is happening to the EuF3.

When molybdenum and tungsten proved to be the improper choice

of cell material, platinum and thoria were tried.

3.99m

The small brown crystals which formed from the reaction of EuOF
with thoria yielded a hexagonal diffraction pattern. Attempts to find
one crystal suitable for a single crystal structural analysis proved
difficult. On the basis of the X-ray diffraction pattern and a few
single crystal layer pictures that were obtained, the structure

appears to belong to the apatite family. This structure was realized

-54 -
when it was noted that the same pattern (with lattice parameter shifts)
appeared in the diffraction patterns of the green powder scraped from
the walls of a quartz tube in which EuOF had been heated, and the pink
powder from a quantz and a thoria cup used in a study of SmOBr by

D. E. Work. A literature search produced the apatite structures
41.67(5'04)30'62
a = 9.72, c = 7.20, all of whose

5m4(5104)3,°' a = 9.497, c = 6.949 3, Ce a = 9.659,

c = 7.121 and La5(SlO4)3O].5,63
intensity data matched those seen in the films of the EuOF and SmOBr
reactions.

By using the program ANIFAC the observed intensities and
interplanar spacings could be recreated for both the europium and
samarium silicates for the empirical formula Ln4.67(SiO4)3O with the
atomic position reported for Ca5(PO4)3F64 (gngppendix V). The
apatite structure found in the thoria cup proved more difficult to
regenerate. The intensities of the Eu-Th compound are very similar
to those observed in the Eu-Si structure, but use of the formula
Eu4067(Th04)30 in the program ANIFAC gives intensities which do not
match the observed pattern at all. Further investigation revealed
that some apatite compounds have the general formula A3B2(C04)3X.64
The compound Eu3Th2(EuO4)3O fits this general formula and its generated
intensity data is much closer to that observed in the X-ray diffraction
pattern. The exact positioning of the europium and thorium atoms
could bring the intensities into the proper perspective, but such
positioning can not be undertaken without the availability of single
crystals. If such crystals can be obtained the confirmation of the
proposed structure would be of value due to the unusual tetrahedral

arrangement around the europium.

 

-55 -

4. Platinum

Platinum did not prove to be a satisfactory cell material either.
Not only does it react with molybdenum so that a Knudsen cell would
have to be machined completely from it, but it also showed evidence
of gas phase attack since the upper walls contained europium.

It is interesting that europium was abstracted into the Pt.
Catalano and Bedford65 report that such a reaction occurs only with
the difluoride. This observation would indicate that Equ was the
gaseous species. Indeed, this thought is consistent with the results
obtained in molybdenum, but EuF3 is not reduced by platinum according

57 My own results of the

to further work by Catalano and Bedford.
sublimation of EuF3 in Pt do not clarify this point (g£.Appendix 4).
The effusate appears to remain untouched (EuF3) while the residue

has been reduced (EUF2.5’EUF2.9)

E. Unanswered Questions and Their Possible Resolution

 

1. What Is the Reaction with Molybdenum?

 

As section V-D-2 infers, it is believed that either condensed EuOF
or gaseous EuF3 is reacting with the molybdenum, thus mobilizing the
molybdenum so that it ends up in both the effusate and the residue.
Such reaction requires the formation of a molybdenum fluoride or
some mixed Eu-Mo fluoride. If there were thermodynamic data on either
of the types of fluorides, the possibility of several reactions could
be investigated. Unfortunately, few molybdenum fluorides have been
studied thermodynamically. Since the question can not be answered
theoretically perhaps other tests may shed some light on the problem.

By suspending a strip of Mo over EuOF in the same manner as described

-55 -

65 in their study of Eu-Pt alloys, one COUld

by Catalano and Bedford
determine if the reaction was strictly a gas phase or not. Further
tests with EuF3 and Mo might also be helpful.

It is obvious that the solution to this question is important
since it would allow interpretation and data reduction of the pressure

measurements made in this study.

2. What Is The Mode of Vaporization of EuOF?

The only way to answer this question is to find some refractory
substance that will not react with EuOF at temperatures up to 1600°
and possibly higher. Since the cell material interfered, the
temperature range for a vaporization study could be considerably
higher than that observed in this study. Suggestions for cell
material are limited to a fluoride passified nickel for temperatures
below l400°, and some type of platinum alloy. A platinum-rhodium
or platinum-iridium alloy would give the needed temperature range

and possibly passify the platinum so that it would not interact.

+ 7 . .
3. Does EuF3(S) EuF3(g). How Does This Reaction Affect the

 

Vaporization of EuOF?

 

From the existing thermodynamic estimates (gfilv-E G-l) EuF3 is
the more stable gaseous species at the temperatures of this study. It
would, therefore, be expected to be one of the equilibrium products of
the decomposition of EuOF. There is definite reason to question the
AHS(EuF3) value reported by Margrave53 since more recent works report
the reduction of EuF3 (gftv-E-l) by Mo and graphite, the materials
used in the sublimation study. If indeed this value of AH; is in
error, the estimate of AHf EuF3(g) would be altered possibly to the

extent where EuF2(g) would be the more stable gaseous species at high

-57 -

57 report a pressure build up for

temperatures. Catalano and Bedford
the compositions EuF2 5-EuF2 99 in sealed capsules at 1600°, but no
such build up for the lower fluorides. Their observation indicates

that EuF is more stable with respect to vaporization than EuF3(S)

2(5)
at 1600°, although it still must be questioned whether the cell
interfered (a Pt-10% Rh alloy was used). The stability of EuF2
over EuF3 would refer to a AGOT of sublimation which would be more
negative for Equ. If EuF3(S) + EuF3(g) at this temperature and
EuF2(S) + EuF2(g) then the entropies of the two reactions must be
reasonably similar leaving the enthalpies as the thermal quantities

. . . +
which differ. On the other hand, if EuF3(S) + EuF2(g)F(g) then the
additional entropy gain over EuF2(S) + EuF2(g) could be enough to
cause the reaction to go at lower temperatures than EuF2(s) + EuF2(g).

Further study on both Equ and EuF3 are required before any of these

suppositions can be justified.

4. Why the Difference Between SmOF and EuOF?

 

There are many other inferences which can be made from this study
which unfortunately lack much substantiation. One such inference is
that the enthalpy of formation of EuOF is more positive than that of
SmOF. Measureable vapor pressures were recorded for EuOF at temperatures
much lower than those for SmOF.7 However, the interference of
molybdenum may have been enough to increase the activity of Eu to give
the greater pressure at lower temperatures, since SmOF showed no
evidence of reaction with molybdenum. The estimated enthalpy of
formation of EuOF (gf,IV-F-l) is substantially lower than that of SmOF
since it is based on the known difference between the enthalpies of

formation of Eu203 and Sm203.

-68 .-
Another basic question which can not be answered with existing
data is why Sm and Eu compounds behave so differently. The unusual
behavior of europium compounds in molybdenum has been noted by another

66 when he studied the vaporization of EuOBr and SmOBr in

worker
molybdenum. The europium oxidebromide reacted with the Mo to the
extent that europium molybdate lines were identified in the X-ray
diffraction pattern. The SmOBr gave no evidence of attack with weight

loss data equaling 100% of theoretical.

5. Further Comments

 

Continuation of the study of solid solution formation between
EuOF and the sesquioxide would be of value. It is possible that at
high temperatures (in the range of 1200°-l600°) EuOF may not be a
compound but rather a 1:1 solution of EuF3 dissolved in Eu203. From
vapor pressure measurements made on EuOF with no cell interference
the activity of the species in solution may be calculated and further
data reduction allowed (gngI-E-l). If EuF3 is the species in solution,
the activity data should extrapolate to the pressure of EuF3(g) over
pure EuF3(s). Further information about the activity of the
trifluoride could be gained also by studying the vaporization behavior
of the fluoride rich side of EuOF.

It is unfortunate that the interpretation of my study depends so

heavily on estimates and the observations of only two other groups,
3,57,64 1 48,50,53

that of Catalano and Bedford and that of Margrave gt_
The contradictory results can not be solved unless the thermodynamics
of Equ and EuF3 are investigated more deeply in cognizance of the

various cell materials which interfere.

BIBLIOGRAPHY

10.
ll.
12.

13.

14.

15.

16.

#wm

BIBLIOGRAPHY

Klemm and H. A. Klein, Z. anorg. allg. Chem., 248, 167 (1941).

 

Zachariasen, Acta Cryst., 4, 231 (1951).

 

Bedford and E. Catalano, J. Solid State Chem., 2, 585 (1970).

 

WCDIZ

V

W

R.

D. . Shinn, Ph.D. Thesis, Michigan State University, East Lansing,
MI (1 68).
G. Brauer and U. Roether, Proceedings of the 7th R. E. Research
Conf., Volume 1, Colorado, 253 (1968).

D. J. M. Bevan, R. S. Cameron, A. W. Mann, G. Brauer and U. Roether,
Inorg. Nucl. Chem. Letters, 4, 241 (1968).

 

 

D. Work and H. A. Eick, J. Phys. Chem., 74, 3130 (1970).

J. L. Margrave, "Physicochemical Measurements at High Temperatures",
J. Bockris, J. L. White, and J. D. MacKenzie, eds., Butterworths
Scientific Publications, London (1959) p 225.

K. D. Carlson, "The Characterization of High Temperature Vapors",
J. L. Margrave, ed., Wiley, New York, N.Y. (1967) p 115.

P. Clausing, Physica, 9, 65 (1929).

J. W. Ward, AEC Report, LA-3509, Los Alamos, N.M. (1966).

R. D. Freeman and J. G. Edwards, "The Characterization of High
Temperature Vapors", J. L. Margrave, ed., Wiley, New York, N.Y.
(1967) p 508.

R. A. Kent, Ph.D. Thesis, Michigan State University, East Lansing,
MI (1963).

R. J. Ackermann, R. J. Thorn, and G. H. Winslow, "The Characterization
of Hi h Temperature Vapors", J. L. Margrave, ed., Wiley, New York,
N.Y. Il967) p 427.

 

G. M. Rosenblatt, P. Lee, and M. B. Dowell, J. Chem. Phys., 45,
3454 (1966).

R. Grimley, "The Characterization of High Temperature Vapors",
J. L. Margrave, ed., Wiley, New York, N.Y. (1967) p 195.

-69..

17.

18.

19.

20.

21.

22.

23.

24.

25.
26.

27.

28.

29.

30.

31.

32.

33.

34.
35.

-70-

G. H. Stout and L. H. Jensen, "X-Ray Structure Determination - A
Practical Guide", MacMillian, New York, N.Y. (1968).

W. Parrish, "Advances in X—Ray Diffraction", Centurex Publishing
Co., Eindhover (1962).

R. L. Seiver, AEC Report COO-716-067 (1970).

J. L. Margrave, "Physicochemical Measurements at High Temperatures",
J. Bockris, J. L. White, and J. D. MacKenzie, eds., Butterworths
Scientific Publications, London (1959) p 6.

A. Findlay, "The Phase Rule and Its Applications", A. N. Campbell,
N. O. Smith, eds., Dover Publications (1951).

J. Haschke, Ph.D. Thesis, Michigan State University, East Lansing,
MI (1969).

G. N. Lewis and M. Randall, "Thermodynamics", Revised by K. S. Pitzer
and L. Brewer, McGraw Hill Book Co., Inc., New York, N.Y. (1961).

K. K. Kelley and E. G. King, Bureau of Mines Bulletin 592,
U. S. Government Printing Office, Washington, D.C. (1961).

L. L. Ames, P. N. Walsh, and D. White, J. Phys. Chem., ll, 2707 (1967).

 

S. Smoes, P. Coppens, C. Bergman, and J. Drowart, Trans. Farad. Soc.,
65, 682 (1969).

 

E. A. Hoelwyn-Hughes, "Physical Chemistry", 2nd ed, Pergamon Press,
New York, N.Y. (1961).

K. K. Kelly, Bureau of Mines Bulletin 584, U. S. Government Printing
Office, Washington, D.C. (1960).

W. M. Latimer, "Oxidation Potentials", 2nd ed., Prentice Hall,
Englewood Cliffs, N.J. (1952), Appendix III.

E. F. Westrum, Jr., "Advances in Chemistry Series", No. 71,
American Chemical Society, Washington, D.C. (1967).

J. J. Stezowski, Ph.D. Thesis, Michigan State University, East
Lansing, MI (1968).

P. A. Pilato, Ph.D. Thesis, Michigan State University, East Lansing,
MI (1968).

D. Brown, "Halides of the Lanthanides and Actinides”, Wiley, New
York, N.Y. 1968.

R. H. Powell and O. Menis, Anal. Chem., 30, 1546 (1958).

 

A. C. Larson, R. B. Roof, D. T. Cromer, U. S. Atomic Energy
Commission Report LR 3335, Los Alamos, N.M. (1965).

 

36.
37.

38.

39.

40.

41.

42.

43.
44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

-7]-
S. Megregian, Anal. Chem., 26, 1161 (1954).

 

R. W. G. Wyckoff, "Crystal Structures", Vol. II, Interscience
Publishers, New York, N.Y. (1964).

N. C. Baenziger, J. R. Holden, G. E. Knudsen, and A. I. Popov,
J. Amer. Chem. Soc., 76, 4734 (1954).

 

S. Geller, P. Romo, and J. P. Remeika, Zeit. fur Kristallographie,
124, 1361 (1967).

 

D. T. Cromer, J. Phys. Chem., El, 753 (1957).

 

C. E. Holley, Jr., E. J. Huber, Jr., F. 8. Baker, "Progress in the
Science and Technology of the Rare Earths", L. Eyring ed. Vol. 3,
343, Pergamon Press, New York, N.Y. (1968).

L. Brewer, G. R. Somayajulu, and E. Brackett, Chem. Rev., 63,
111 (1963).

 

K. F. Zmbov and J. L. Margrave, J. Inorg. Nucl. Chem., 29, 59 (1967).

C. E. Wicks and F. E. Block, Bureau of Mines Bulletin 605, U. S.
Government Printing Office, Washington, D.C. (1963).

"Janaf Thermochemical Tables", 0. R. Stull, ed., Dow Chemical Co.,
Midland, MI (1965).

R. Hultgren, R. L. Orr, P. D. Anderson and K. K. Kelley, "Supplement
of Selected Values of Thermodynamic Properties of Metals and
Alloys", private communication, R. Hultgren.

J. V. Charlee, A. K. Chaudhuri, and J. L. Margrave, High Temp. Sci.,
2, 1 (1970).

 

R. H. Hauge, J. W. Hastie and J. L. Margrave, J. Less Common Metals,
23, 359 (1971).

 

P. A. Akishin, N. G. Rambidi and V. P. Spiridonov, "The Characteriation
of Hi h Temperature Vapors, J. L. Margrave, ed., Wiley, New York,
N.Y. I1967).

J. W. Hastie, R. H. Hauge and J. L. Margrave, High Temp. Sci., 3,
56 (1971).

 

F. A. Cotton and G. W. Wilkinson, "Advanced Inorganic Chemistry”,
2nd ed., Interscience Publishers, Wiley, New York, N.Y. (1966).

L. Brewer, "The Chemistry and Metallurgy of Miscellaneous Materials",
L. L. Quill, ed., McGraw Hill, New York, N.Y. (1950) paper 8.

K. F. Zmbov and J. L. Margrave, "Advances in Chemistry Series", #72,
American Chemical Society, Washington, D.C. (1968).

54.
55.

56.
57.

58.

59.

60.

61.

62.
63.
64.

65.

66.

-72-
D. L. Hildenbrand, J. Chem. Phys., 483 3657 (1968).

 

J. W. Green, G. 0. Blue, T. C. Ehlert, and J. L. Margrave, J. Phys.
Chem., 41, 2245 (1964).

J. Berkowitz and J. R. Marquart, J. Chem. Phys., 37, 1853 (1962).

 

E. Catalano, R. G. Bedford, V. G. Silveira, and H. H. Wickman,
J. Phys. Chem., g0, 1613 (1969).

A. F. Wells, "Structural Inorganic Chemistry", 3rd ed., Clarendon
Press, Oxford (1962).

J. Lucas and J. P. Rannou, C. R. Acad. Sci. Paris, Ser C, 266,
1056 (1968).

 

 

Von K. A. Gingerick and G. Brauer, Zeit. anorg. allg. Chem , 324,
48 (1963).

G. J. McCarthy, W. B. White, and R. Roy, J. Inorg. Nucl. Chem., 29,
253 (1967).

 

J. Felsche, Naturwissenschaften, 56, 325 (1969).

 

J. Felsche, Ibid., 56, 212 (1969).

R. W. G. Wyckoff, "Crystal Structures", Vol. III, Interscience
Publishers, New York, N.Y. (1965).

R. G. Bedford and E. Catalano, Proceedings 8th R. E. Research Conf.,
Reno, Nevada (1970).

D. E. Work, Michigan State University, private communication (1971).

 

APPENDICES

 

Appendix I
X-Ray Diffraction Analysis

 

EuF3 orthorhombic X = 1.54050 A
CALCULATED OBSERVED
Intensity sinzo sin26 pkg_
19.9 .0427 .04237 011
56.0 .0442 .04389 101
59.4 .0482 .04785 020
100.0 .0563 .05592 111
76.8 .0662 .06585 210
3.6 .0848 .08454 201
34.2 .0924 .09225 121
2.6 .0969 211
1.6 .1023 220
16.1 .1228 .12257 002
31.8 .1330 .13285 221
3.9 .1363 102
41.6 .1484 .14825 112
46.2 .1525 .15245 301
67.0 .1526 .15245 131
62.4 .1625 .16253 230
37.6 .1646 .16462 311
28.3 .1710 .17112 022
31.6 .1845 .18447 122
42.2 .1890 .18901 212
25.1 .1927
35.3 .2006 .20076 321
19.1 .2165 .21700 400
18.8 .2369 .23687 141
3.7 .2446 302
13.0 .2447 '24490 132
15.2 .2566 312
11.5 .2592 411
24.0 .2609 331
17.9 .2646 .26481 420
7.6 .2775 241
23.4 .2853 232
17.6 .2884 013
7.9 .2927 322
6.2 .3019 113
12.9 .3155 042
2.5 .3249 430
9.4 .3290 142

-73-

-74-

EuF3 orthorhombic (Continued)

 

 

 

CALCULATED OBSERVED
Intensity sin26 sin26 hk£_
15.4 .3304 .33043 203
4.5 .3318 051
6.4 .3380 123
9.4 .3393 402
25.6 .3451 341
14.0 .3453 '34559 151
2.8 .3530 332
12.1 .3552 250
4.0 .3556 431
5.8 .3689 501
36.5 .3786 .37862 223
EuOF rhombohedral A = 1.54050 Cu K 1 a = 6
a=3
CALCULATED OBSERVED
pkg_ sin26 Relative Intensity sin26
111 .0143 6.6 .0139
222 .0571 29.4 .0568
110 .0588 100.0 .0586
211 .0778 35.8 .0776
221 .0921 4.4
333 .1284 2.2 .1281
322 .1301 4.6 .1299
332 .1539 11.4 .1537
lOT .1573 12.7 .1575
433 .2110 2.7 .2112
321 .2144 7.1 .2143
200 .2161 4.1 .2160
220 .2352 2.6 .2351
443 .2443 1.5
432 .2858 2.2
422 .3113 3.4
544 .3204 1.4
44g_ .3683 1.3
211 .3735 3.7 .3732
543 .3856 1.1
310 .3925 2.0 .3913

-75 -

Eu203 cubic A = 1.54050 a = 10.866 A
CALCULATED OBSERVED
pk§_ sin26 Intensity sin26
112 .0301 5.6 .0300
222 .0602 100.0 .0602
123 .0702 1.4 .0701
004 .0804 33.1 .0802
114 .0904 4.4 .0903
042 .1004 0.7
233 .1104 5.1 .1103
134 .1305 4.2 .1304
125 .1506 0.8 .1506
044 .1607 12.2 .1606
334 .1705 0.5 .1705
116 .1908 0.6 .1906
145 .2108 0.9 .2107
226 .2208 9.6 .2207
163 .2309 1.8 .2309
444 .2408 2.4 .2406
345 .2508 0.8 .2508
336 .2709 0.5
127 .2711 0.5 '2709
246 .2810 0.5
237 .3112 0.9
008 .3216 1.7 .3200
455 .3311 0.5
147 .3313 0.5 .3310
118 .3315 0.6
356 .3512 0.7
266 .3814 2.9

 

Appendix II

Vapor Pressure Measurements

 

III-B G.F. = 206.273 a0 = .01012 cm2

Time ugEu T°K PEuF PF Composition

2
. -5 -6
30.00 mln 11.172 1598 1.0348x10 3.2727x10 Eu0].065F.87
30.00 6.120 1598 5.6687x10'6 1.7928x10'6 Euo1 10F 80
29.92 5.402 1603 5.0236006 1.5887x10‘6 Euo1 12F 77
30.20 4.999 1598 4.5997x10'6 1.4547x10'6 Euo1 13F 73
30.00 4.418 1600 4.0949x10'6 1.2951x10’6 Euo1 15F 70
30.00 4.111 1603 3.8128006 1.2108006 [:00] 17F 67
30.05 3.909 1598 3.6147x10'6 1.1432x10'6 1:001 181‘ 64
30.02 3.457 1597 3.1989006 1.0117x10“6 Eu0] 19F 61
30.00 3.336 1597 3.0890x10‘6 9.7692x10’7 Euo1 21F 59
Euo1.22F.56
111-0 G.F. = 199.962 a0 = 38.5x10‘
10.00 4.284 . 1651 3.0833x10‘5 97513006 15001 10F 78
10.05 3.713 1651 2.6592x10"5 8.4100x10‘6 Euo1 14F 73
15.00 4.611 1652 2.2132x10'5 6.7148x10'6 Euo1 16F 67
25.00 6.333 1651 1.8231x10'5 57658006 15001 20F 61
25.00 5.359 1655 1.5446x10"5 4.8850x10‘6 £00] 25F 51
-6 -6

30.00 4.111 1653 9.8692x10 3.1212x10 Eu0].29F.42

 

 

-75-

 

 

-77-

III-D (Continued)

 

 

 

 

 

 

 

Time ugEu T°K PEqu PF Composition
40.00 4.984 1652 8.9708x10'6 2.8371x10'6 E”°1.33F.34
50.00 4.591 1654 6.6148x10'6 2.0920x10'6 Eu0].38F.24
60.00 4.452 1656 5.3486x10'6 1.6915x10‘6 Eu0].43F.15

EUO1.48F.05
Vap V-B G.F. = 203.12 a0 = 38.25x10'4
pure EuOF

Time ugEu T°K PEuF2 PF

5.00 6.888 1783 1.0533x10'4 3.3313x10'5
5.00 5.350 1787 8.1916x10’5 2.5907x10'5
8.00 6.620 1785 6.3310x10'5 2.0022x10'5

10 00 5.584 1787 4.2749x10'5 1.3520x10"5
15.00 6.040 1785 3.0806x10'5 9.7428x10‘6
15.01 4.917 1784 2.5054x10'5 7.9238x10‘6
20.01 3.482 1784 1.3249x10‘5 4.1903x10‘6
20.00 1.071 1785 4.095900“6 1.2955x10’6
Vap V-A G.F. = 202.471 a0 = 38.25x10'4

EuOF with excess b-Eu203

5.99 1.1863 1696 1.4722x10'5 4.6561x10’6
8.02 2.042 1719 1.9053x10'5 6.0259x10'6
8.00 2.369 1727 2.2211 7.0247
12.00 2.439 1726 2.1491 6.7965
12 00 3.020 1725 1.8866 5.9667

Vap V-A (Continued)

Time

 

11.99
15.00
20.00
20.00

IV-B

Time

 

50.00
50.02
50.00
50.00

IV-C

25.00
25.00
25.00
25.00

 

 

“9

3.283

3.876

4.435

5.003
ugEu Temp
5.418 1524
4.651 1523
4.637 1522
3.638 1513
4.635 1585
4.221 1580
3.762 1582
3.362 1583

NNNU)

(Job-D

T°K

1724
1725
1724
1727

G.F.

P
EuF2

 

.1278x10-
.6328
.6749
.0929

G.F.

.4576x10-
.9800
.4236
.9553

-78—

 

 

EuF2 _1:
2.0519 6.4892
1.9372 6.1268
1.6620 5.2564
1.8765 5.9349
= 220.897 a0 = 1.019x10'4
PF Composition
6 -7
9.8919x10 Eu0].00F.99
8.4845 EuOIOOIF.98
8.4595 E“°1.02F.97
6.6191 Euo]_02F 96
Euo1.03F.95
= 221.343 a0 = 1.019x10'4
6 -6
1.7260x10 Eu0].03F.94
1.5750 EuOLMF.93
1.3990 Eu01.04F.92
1.2509 E”01.05F.91
Euo1.05F.90

Temp

1500
1600
1700
1800
1900

°K

13.841
13.849
13.856
13.862

Appendix III

Thermodynamic Va1ues

 

H°T-H°298 in ca1 S°T in eu
17410 96.899
18892 87.856
20374 98.754
21858 96.611
16475 97.430
17715 98.323
19100 99.163
20486 99.954

Free Energy Functions

EuF

3(5)

-50.809
-52.573
-54.337
-56.101
-57.865

-79-

fef eu

-85.293
-86.048
-86.930
-87.458

-86.543
-87.251
-87.928
-88.573

 

EuOF(S)

—38.052
-39.068
-40.084
-41.100
-42.116

oo oz
02 moogp
oo oz

oz

-80-

oz

0:

mo moocz

mocmomogoozd.
zom1x

sh mmoco m 1 30__w>

oopom__ou poz so oz N u o .mm.m n o - czoom
< oo.o_ u o ooo +
ooooo__oo ooz pooh oz ooo + momoo-o - oozm
oo.~aoo-om.mooo oo oo
Amm.m n o .ooov czoom oz momom-o + zoom - zoom
omozo o + om.Nooo - o_oo oo oo momao + zoom - ooozz
omozo w u czocm
mmmca m+

ow Noam 4 mm Noam u wmwom oz moom + Awk.m u ow oom omcooo

A< ow.op u my ooo
oo mono-om Nooo oz momoo-ov zoozm 1 ozoooo

Fo.o + mn.m n o m N

oom - czogm zoom oz oz A o om1o + dozmvomzom 4 wows:

cowpooommwo moomomoco:_d oowpooowmwo

xmmux homux >om-x

mpomommm moowmmm

m

m_o_ooooz __oo zoo; zoo ooo zoom 4o moooooooz

>H xwoooooz

N

meow + ooh

oozm + 3

ooo oz ow ooom + oo

A z ozv zoom + oo

maom + oz

loom + 02

——mo

Appendix V

X-Ray Diffraction Ana1ysis of Eu-Th Compound

 

 

Eu2Th3(Eu04)3O a = 9.590 :_0.002 c = 7065 :_0.009
CALCULATED FOR
OBSERVED CALCULATED Eu6(Si04)30
sinze Re1ative Intensitx_hk2 sigf§_ Intensitx Intensity
100 .0086 3.9 4.9
.02591 5 110 .0258 2.7 4.7
.03451 10 200 .0344 2.3 24.5
.03780 10 111 .0377 1.2 26.6
.04766 5 002 .0475 19.6 10.0
.05626 20 102 .0561 8.6 40.5
.06037 20 120 .0602 18.0 38.9
.07220 100 121 .0721 100.0 100.0
.07346 70 112 .0733 57.1 82.5
.07751 40 300 .0774 35.0 52.1
.8198 4 202 .0819 2.1 3.1
.10305 2 220 .1032 0.0 1.6
.11174 5 130 .1118 0.2 1.8
.12474 2 302 .1249 0.4 2 1
.13274 10 113 .1328 0.7 7.5
.13754 4 400 .1376 1.3 2.6
.15079 15 222 .1507 6.7 12.7
.15936 10 132 .1593 1.5 7.1

-8] -

-32-

CALCULATED FOR

 

OBSERVED CALCULATED Eu6(SiO4)30
sinze Re1ative Intensitx_hk2 §1fl39_ Intensity Intensitx
.16342 5 320 .1634 1.0 2.5
.16726 20 213 .1672 7.0 6.3
.17534 7 231 .1753 3.0 4.0
.18071 10 140 .1806 1.2 4.1
.18515 15 033 .1844 0.1 6.8
.19027 5 004 .1902 2.3 3.0
.22433 2 204 .2246 0.1 1.4
.22823 2 142 .2281 0.4 0.7
.24050 4 240 .2408 1.7 2.4
.24411 4 331 .2441 1.7 2.4
.25044 5 124 .2504 1.8 4.1
.26254 7 052 .2625 3.4 3.

150 .2666 1.0 1.5
.26713 8 304 .2676 3.8 4.9
.27033 8 323 .2704 2.5 3.8
.27888 5 151 .2785 3.2 4.1
332 .2797 1.4 3.1
404 .3278 0.6 1.2
.34719 2 251 .3473 1.4 2.1
324 .3536 0.6 1.4
.35667 2 602 .3571 1.1 1.8
125 .3574 2.9 2.4
342 .3657 1.1 2.0
160 .3698 0.9 1.0

OBSERVED

sin2 Re1ative Intensity hk

.36968 2 414
153
161
.38208 7 252

-83 -

CALCULATED FOR

CALCULATED Eu6(Si04)30
§1n3__ Intensitx Intensitx
.3708 1.3 4.4
.3736 1.5 1.8
.3817 0.2 1.4

.3829 0.9 3.8

'- -

MICHIGAN STATE UNIVERSITY LI

3 1193 03082 5529

11111“