3'9"?!"

3'"

”...! It I?

.

m yam

 

 

    

  

ode-3’5.”

‘0‘.’

    

 

< 1:5 J.

 

 

    

  

 

 

   

 

. ”1'10".

-—'r—v-"

 

 

     

   

 
    

 

    

 

 

 

 

    

. _ . .u a D r on. v . uh.
, H. nu.
0. TI. . .. . ..H
an. n! 0.
9.1“..L. .... 1231.. W... q. ..........r...... -.....
G. . . ....
_ m. u“ E. _ ..
— . nun VA: 1. .
v... _. .nU .. I.
VI Si. Ii.
L i . .
. NW NW .
we. 9...» am
E .v. . . . .
.. .. ...... ...”.U.......m.n....n........w.“.H...w.....4§....sn...»«3.501.215352:. 3.“... . .... 1...: ... ..p ...";
.. A ..a....x....,u...fi..: «mfier... ..."? n....22...rn...:w.....fl.£~m.wr. ... u. meMX...

+4;
L I B 33. A R Y
Michigan State
University

    
     

  

..

m __ l‘!
HUM? 8: SONS'
BOOK BINDERY *

 

  

 

       

w LI! RARY BINDERS

on :..--..R' 433535553!

ABSTRACT

AN X-RAY CRYSTALLOGRAPHIC INVESTIGATION OF A

PHASE IN THE YTTERBIUM(III)--OXIDE--CHLORIDE
'SYSTEM

by

John Thomas Richards

A phase which is a probable oxidechloride of ytterbium
(III) was prepared and has been investigated. The green
plate-like cyrstals were found to exhibit hexagonal symmetry
and belong to space group P52m. The observed lattice parame
eters are a = 6.1879 1 0.0016 R and c = 19.764 1 0.015 K. A
partial structural determination locates the 16 metal atom
positions in the unit cell along with the approximate posi-
tions for 19 oxygen atoms. Chemical arguments are used to
construct a complete model for the structure of this phase:
(Yb203)6(YbOC1)4.

This phase was isolated as a side reaction product from
an attempted preparation of an ytterbium chloride phase. The

oxidechloride probably resulted from the presence of traces

of water.

AN X-RAY CRYSTALLOGRAPHIC INVESTIGATION OF A
PHASE IN THE YTTERBIUM<III)--0XIDE--CHLORIDE
SYSTEM

By
John Thomas Richards

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements

for the degree of

MASTER OF SCIENCE

Department of Chemistry

1974

l
(J ACKNOWLEDGEMENTS

The author wishes to thank Professor Harry A. Eick for
the guidance, support and encouragement he provided during
the course of this research.

Lucinda, my wife, is especially to be credited for her
understanding and patience.

My colleagues in the High Temperance Group have been
particularly helpful. Their professionalism and concern
were very helpful.

Financial support of this research by the United States
Atomic Energy Commission under contract number AT (ll-1)-7l6

has been very welcome.

ii

TABLE OF CONTENTS

Page

I. Introduction .............................. 1
II. Experimental .............................. 4
III. X-Ray Structural Analysis ................. 7
IV. Results ................................... 10
V. Structure Determination ................... 14
VI. Discussion ................................ 24
VII. Conclusion ................................ 33
REFERENCES ...................................... 39
APPENDICES ...................................... 41

iii

LIST OF APPENDICES

APPENDIX Page
1. Structure Factor Table ................... 41
II. Observed and Calculated Powder Pattern... 42

iv

LIST OF TABLES

TABLE Page
1. Crystal Data for Ytterbium and Thulium

Chlorides ................................ 2

2. X-ray Diffraction Data Collection .......... 12
3. Patterson Map .............................. 15
4. Properties of Space Group P6/mmm ........... 16
5. Properties of Space Group PBZm ............. 18
6. Refined Atom Positions in Space Group PBZm. l9
7. Patterson Maps from Final Refinement ....... 20
8. Temperature Factors ........................ 21
9. Interatomic Distances ...................... 26
10. Interatomic Angles ......................... 27

LIST OF FIGURES

FIGURE Page
1. Representation of Space Group PBZm. ........ 18
2. Projection of Metal Atom Positions on

(001) Plane ............................. 36
3. Projection of Anion Positions on (001)

Plane ................................... 36
4. Projection of Yb16022Cl4 on (010) Plane... 37

vi

CHAPTER I
INTRODUCTION

Ytterbium exists in two stable oxidation states--
divalent and trivalent. The halides of these two states are
readily prepared1 and their structures have been deter-

d2'4.

mine Recent attention has been focused on the halides

of the other lanthanide series elements to study their less

stable reduced oxidation statess'a.
In the course of investigating the phase diagrams of

the lanthanide series halides, Corbett, et aZ., have detected

many previously unknown phases of mixed oxidation states;

that is, phases composed of combinations of valence states.

As an example, the gadolinium-gadolinium trichloride phase

5 indicates the presence of a phase of composition

diagram
GdC11.58. The crystal structure of this phase has been
solved6 and is interpreted by Corbett to be a mixture of
gadolinium metal atom clusters and gadolinium trichloride.
While phases of mixed oxidation states are reported for
the earlier members of the lanthanide series elements, Fishel

9 the existence of the same type of phases in

and Eick report
the ytterbium.dichloride-ytterbium.trichloride system. They
report the composition as being YbClz.26, detected as the
product of a partial hydrogen reduction of YbClB.

The purpose of this research was to characterize

YbCl2 26 or another ytterbium.chloride phase of intermediate

oxidation state by single crystal structural methods. Of
particular interest to the ytterbium system is the thulium-
thulium(III) chloride phase diagram8 because of the similar-

ities known1

to exist between the chemistry of ytterbium
and thulium. One of the more interesting results of Caro
and Corbett's phase study is the postulated existence of at
least seven phases in the composition range TmClz.o4 to
TmC12.12’ detected by thermal inflections in the cooling
curves at constant compositions. These phases were inter-
preted to result from an ordering of a Tm(II) rich crystal
lattice to include progressively more Tm(III) with the cre-
ation of cation vacancies and the addition of interstitial
anions.

The relevance of the thulium system to this work can be
seen in Table l, which is a listing of crystallographic data

4 4 8 2

for TmCl3 , YbCl3 , TmCl2 and YbCl2 . It is evident that

Table 1: Crystal Data for Ytterbium and Thulium Chlorides

 

 

Symmetry and ,
Compound Structure Type a b c B
TmCl3 Monoclinic A1C13 6.75 3. 11.73 3. 6.39 A llO.6°
YbCl3 Monoclinic AlCl3 6.74 11.66 6.38 110.4
TmCl2 Orthorhombic CaF2 6.55 6.68 6.93
YbCl2 Orthorhombic CeSI 6.69 13.15 6.94

 

there is a very close structural relationship between the
dichlorides and trichlorides of thulium and ytterbium. The

major discrepancy in the lattice parameter b between TmC12

3

and YbCl2 is a result of Fishel's reindexing YbCl2 from the
fluorite to the CeSI structure type, done on the basis of
single crystal structural data. D311 and Klemm10 had ini-
tially erroneously indexed YbCl2 on a fluorite structure
type and obtained the lattice parameters: a - 6.53 A, b =
6.68 A, c = 6.91 A. On the basis of the good agreement of
D311 and Klemm's lattice parameters with those of TmClz, the
close structural relationship with YbCl2 is now more apparent.
Fishel reindexed YbCl2 from fluorite to the CeSI structure
type because the cation to anion radius ratio and the axial
ratios a/c and b/c more closely match the latter structure
type, and suggested TmCl2 might also be better indexed on

the CeSI structure type.

If the conclusion of Caro and Corbett is correct and
the many phases in the Tm-TmCl3 systems are the result of
successive modifications of the dichloride lattice, then the
close crystal similarity between the thulium and ytterbium
systems indicates the probable existence of several phases
of intermediate stoichiometry between YbCl2 and YbC13. Any
of these expected phases which could be prepared as single
crystals suitable for X-ray structural analysis would sat-
isfactorily meet the stated goal of characterizing this
system.

For the previously mentioned phase Yb012.26' Fishel was
unable to obtain single crystals for analysisz. Thus a
different approach was chosen to prepare a reduced phase, to
obtain the single crystals that apparently could not be found

in the product of Fishel's hydrogen reduction.

CHAPTER II
EXPERIMENTAL

Two reaction procedures were followed for the prepara-
tion of the intermediate phase YbC12.26: l) a mixture of
ytterbium metal and ytterbium trichloride confined in a
sealed ampoule was heated to temperatures between 550 and
l450°C (various runs) and 2) a mixture of ytterbium dichlor-
ide and trichloride confined in a sealed tantalum ampoule
was heated to a temperature of 800°C. These temperatures
were chosen as being less than, equal to and greater than
the melting points of the reactants (the metal, dichloride
and trichloride melt in the range 700 to 865°C).

A consistent heating procedure was used in all prepara-
tions. The reactants were cycled to temperature over a
period of eight hours, heated at temperature for periods of
time which varied from 1.5 to 13 days, then cooled to room
temperature over a period of eight hours. In some prepara-
tions carried out by the first procedure, a very small
amount of iodine was included to serve as a transport agent;
iodine was present in all preparations carried out by the
second procedure.

Ytterbium trichloride was prepared according to the
method of Taylor and Carterll. The overall reaction is as
follows:

Yb203 + 6NH4C1 = 2YbC13 + 3H2qgf'6NHBQD (1).

4

5

Briefly, the method entails dissolving a six-fold excess of
ammonium chloride with the sesquioxide and coprecipitating
the hydrated trichloride and ammonium chloride from evapo-
rating hydrochloric acid. The dried mixture is heated under
a stream of inert gas at 200°C for 1.5 hours, then at 430°C
for eight hours. The product is heated to above 500°C before
cooling.

Ytterbium dichloride was prepared according to the

method of De Rock and Radtkelz.

The overall reaction is as
follows:

2YbCl3 + 2Zn + 2NH4Cl = 2YbCl2 + 2ZnC12 + 2Nngf Hzng).
Zinc metal and zinc chloride are added to YbCl3 to provide a
reducing melt (mole ratios: ZnC12/YbCl3 = 1.2, NH4Cl/YbC13
= 12, Zn/YbCl3 = 2). The reaction temperature is increased
to over 500°C and maintained at that temperature for more
than an hour until a clear melt is obtained. The crude
product is purified by sublimation at 1300°C under high
vacuum.

Routine analyses of the trichloride and dichloride pre-
pared by these methods were effected by an X-ray powder
diffraction technique. An 80 mm Guinier forward focusing
camera which employed Cu Kd radiation was used. Platinum
powder (a =- 3.9237 s 00003 X) was Included in the samples
for powder diffraction analysis. The interplanar d-spacings
of the powder patterns are based on this platinum.standard.

X-ray fluorescence experiments were performed with a
Norelco spectrometer which used a graphite analyzing crystal

and a tugsten anode tube. Three targets were used:

6

platinum, aluminum and quartz. Quantitative analyses were
not attempted, only the qualitative presence of metals was

examined.

Both the ytterbium dichloride and trichloride are

hygrosc0pic materials1

and must be isolated from atmospheric
moisture. Additionally, it was assumed that all intermediate
phases in the dichloride-trichloride system would behave
similarly. All transfer of materials was performed in a dry
box under a dried argon atmosphere and at all other stages

of experimental work precautions were taken to isolate the

experiment from atmospheric moisture.

CHAPTER III

X-RAY STRUCTURAL ANALYSIS

The single crystal diffraction experiment consists of
measuring the intensity of a diffraction maximum and record-
ing its position in space relative to the orientation of the
crystal. The maxima can be viewed as arising from a reflec-
tion of the incident X-ray beam off of planes of atoms in
the crystal and are thus referred to as "reflections."

These reflections are in turn indexed by three integers,
hkl, the Miller indicies. The reflection plane is located
in the unit cell of dimensions a,b,c by the points a/h,
b/k, and c/Z.

One parameter that cannot be measured in the diffraction
experiment is the phase relationship between the incident
and diffracted radiation. This is the so-called "phase

problem"13

in X—ray structural determination: without the
phasing information the structure cannot be derived directly
from the experiment.

From a structural model of the contents of a unit cell,
the diffraction pattern for the crystal can be computed.
These predicted reflections, or structure factors, are de-
pendent on the nature and arrangement of the diffracting
material and include phasing information. A measure of the

precision by which the structural model matches the crystal

structure is provided by the following function:

8
2J1 F01 - 1 Fc ll

0 I (3).

R:
2|

where F0 is the amplitude of the observed reflection and Fc
is the calculated structure factor for that reflection.

Because the system being investigated in this research
contains heavy metal atoms, the method employed to derive a
model structure for the crystal was to compute a Patterson
function13. The Patterson function, or map, is calculated
from|Fo I2 which are independent of the phasing. This map
shows positions and magnitudes of vector peaks in the vector
space u,v,w corresponding to the real space a,b,c. The
peak heights are proportional to the products of the number
of electrons at the ends of the vectors, so when heavy and
light atoms are both present in the unit cell, the relative
orientations of the heavy atoms can be derived from the lo-
cations of the most intense vector peaks.

The structure factors calculated from a derived model
are the Fourier transforms of the electron distribution
within the unit cell, the atoms assumed to be located at
the centers of electron density. Electron density maps or
Fouriers can be calculated, then, from the calculated struc-
ture factors and phasing information. A difference Fourier
is synthesized from calculated phases and the difference
series between F0 and Fe. This difference map may be taken
to indicate where electron density is too large or where

electron density should be increased.

9

A least squares refinement of the atomic positions and
temperature factors simultaneously varies these parameters
to minimize the R function. The temperature factors (or B
values) are measures of the extent to which atomic motion
causes the electron density to be more diffuse than in a
stationary atom. One B value implies that this motion is
being approximated in three dimensions by a sphere, six B
values indicate that the six-parameter general ellipse is
being used. As opposed to the pecision indicated by the R
function, the temperature factors indicate the accuracy of
an atomic position. Negative B values may have no meaning;
very large B values are suspect also. As an example applic-
able to this research, Fishel reports B values between -0.9
and 4.4 for the structure of YbCl2 2.

Calculations were performed using a Control Data 6500
computer. Least squares refinement (LSQXTL), Patterson
function and Fouriers (CRYSTAL) were calculated from.pro-

grams provided by A. Zalkin. Adaptations were made to use

these programs on the CDC 6500.

CHAPTER IV
RESULTS

The results of the attempted preparations of an inter-
‘mediate phase were examined visually under a microscope.
Typically, the Yb/YbCl3 reaction products included unreacted
ytterbium metal, a black compound of unknown composition and
colorless to light green crystals of no repetitive morphol-
ogy. The results of the YbClZIYbCl3 preparations were un-
clear, the product being composed of the same green crystals
(YbClz) and white powder (YbC13) that were the reactants.
Seven preparations were run, five of the Yb/Yb013 type and
two of the YbC12/Yb013 type.

When suitable crystals were found, they were mounted on
an X-ray precession camera for preliminary investigation. A
suitable crystal was taken to mean one that was visibly
single and was of an unrecognized type.

One of the colorless crystals was chosen first for ex-
amination and was eventually identified from the unit cell
parameters and space group as YbClz. The orthorhombic unit
cell of this crystal yielded lattice parameters of a - 6.70
1 0.08 A, b = 13.2 t 0.1 A and c - 6.96 1 0.12 As The space

3 values

group was Pbca. Barnighausen, Patow and Beck report
for these parameters to be a - 6.698 A, b - 13.139 A, c -
6.948 A, space group Pbca. This result, together with the
ytterbium metal observed in the product, indicates incom-

plete reaction in these preparations.

10

11

Close examination of all the samples prepared showed
only one to contain crystals of a type different from.YbClZ.
This preparation consisted of metal and trichloride mixed in
YbClz.26 stoichiometric proportions with a small chip of io—
dine (mole ratio: total Yb/Iz:>300) added to the tantalum
ampoule before it was welded shut. The temperature was
raised to 1000°C and held there for two hours, then main-
tained at 800°C for five days. Aside from the unreacted
metal and the recognizable dichloride, there was a small
number of green plate-like crystals. The total yield of
this type of crystal was estimated to be approximately a
milligrams

The majority of these crystals could be seen as twinned
by means of the platelets being stacked on t0p of each other.
Several others were only verified as twinned by precession’
X-ray photography. One good crystal was found for a
structural determination (0.24 x 0.14 x 0.056 mm) and was
mounted and sealed in a glass capillary to seal out at-
mospheriC‘moisture.

The precession photographs of this crystal were found
to be consistent with a hexagonal unit cell with lattice
parameters a - 6.17 i 0.04 A and c - 19.6 i 0.6 A. No gen—
eral or specific extinction conditions were found to limit
the possible reflections. There are five space groups of
hexagonal symmetry which have no general or special extinc-
tion conditions: P622, P6mm, P6m2, P62m.and P6/mmm. Of

these five, only P6/mmm is centrosymmetric.

12

Single crystal diffraction measurements of this crystal
were done on a computer controlled (program FACS I), four
circle, Picker goniometer using Mo Kc radiation and a per-
pendicular graphite monochromator. Room.temperature during
the data collection was 22 i 3°C. Preliminary examination
confirmed that the crystal has hexagonal symmetry and that
it has no general or special extinction conditions. A least
squares lattice parameter fit of twelve data points gave the
following values: a = 6.1879 1 0.0016 A and c = 19.764 a

0.015 A. Table 2 lists pertinent data collection parameters.

Table 2: X-ray Diffraction Data Collection

 

Mo radiation
A = 0.70926 A, no filter, kV - 40, ma = 18
w Scanned
l°/min., no attenuator, background time - 20 sec.
Perpendicular graphite monochromator, d - 3.354 A
sine/A: 0.03691 3‘1 to 1.15494 3‘1
662 reflections measured, 107 absent
Empirical absorption correction applied
v - 655.4 A3
0(ca1c) = 8.3 g/cc

u = 598 emfl

 

In correcting the raw data for background radiation,
the observed intensity was defined to be zero if the raw

intensity was less than two sigma. Sigma, a measure of the

13
uncertainty in the reflection intensity, was calculated from

counting statistics by the method of Ibersla.

An empirical
absorption correction was applied to the data to reflect the
fact that the crystal is not spherical and at different ori-
entations the incident X-ray beam had different path lengths
to travel. Strong reflections were chosen and the crystal
was rotated full-circle around the angle 6 in polar coordi-
nates. Rotating a crystal 3600 through the X-ray beam, as
described, necessarily results in a sinusoidal empirical
absorption curve.

The absorption correction was determined after the
reflection data had been collected. To achieve a sinusoidal
correction the diffractometer had to be realigned. Phi
curves around the 120 and the 360 were collected and averaged
for the empirical absorption correction.

The raw data set contains two equivalent sets of
reflections: the hexagonal symmetry requires that the sets
th and khZ be equivalent because they are related by a
mirror plane. Examination of the absorption corrected data
shows that the reflections which are supposed to be equiva-
lent do not differ from each other by more than three sigma.
These two sets of data were averaged together to give 378

independent reflections.

CHAPTER V

STRUCTURE DETERMINATION

A three dimensional Patterson function was calculated
for the data set after the absorption correction had been
applied to the data and the equivalent reflections averaged.
The centrosymmetric space group P6/mmm was used in the syn-
thesis. Table 3 lists the positions, relative peak heights
and assignments to the vectors observed in the map. Table 4
lists the equivalent positions in P6/mmm that describe the
vector peak positions.

The largest vector peaks were found in six levels of
the Patterson map: fractional w coordinates 0.000, 0.155,
0.187, 0.310, 0.337 and 0.500. These levels were interpreted
as being the six distances of separation of planes of metal
atoms along the long axis of the unit cell. There are only
three types of vector positions in the map: u,0,w; 1/3,2/3,w

15 were used on the

and 0,0,w. Two image seeking functions
map to establish possible real fragments of the heavy atom
positions--(1/3,2/3,z and 2/3,1/3,z) and (0.325,0,z; 0,0.325,z
and 0.675,0.675,z). The image seeking functions indicated
the 1/3,2/3,z position was real and the x,0,z position was
probably real, but for more than one value of x.

The requirement for strong peaks at the level w = 0.187
was found to be satisfied by two eclipsed sets of x,0,z

positions separated by that increment in 2. These positions

14

15

Table 3: Patterson.Map

 

 

P6/mmm Relative
Positions u v w Height Assignment
1a 0.000 0.000 0.000 999 Origin
2c 0.333 0.667 0.000 505 Yb-Yb in same
x,y plane
2e 0.000 0.000 0.098 66 Yb(x,0,z)-
Anion(x,0,z+
0.098)
4h 0.333 0.667 0.098 13 Yb(x,0,z)-
Anion(0,x,z+
0.098)
12n 0.325 0.000 0.155 195 Yb(x,0,z)-
Yb(1/3,2/3,z+
0.155)
2e 0.000 0.000 0.187 387 Yb(x,0,z)-Yb
(no.2)
4h 0.333 0.667 0.187 178 Yb(x,0,z)-Yb
(0.X.2)
12n 0.325 0.000 0.255 16 Yb(l/3,2/3,z)-
Anion(x,0,z+
0.255)
12n 0.350 0.000 0.310* 104 Yb(x,0,z)-Yb
(x,0,z+0.310)
12n 0.325 0.000 0.337* 195 Yb(x,0,z)-Yb
073,2/3,z+0.337)
12n 0.325 0.000 0.405 19 Yb(x,0,z)-Anion
(i,0,z+0.405)
lb 0.000 0.000 0.500 171 Yb(1/3,2/3,z)-
Yb(1/3,2/3,z+
0.500)
2d 0.333 0.667 0.500 96 Yb(l/3,2/3,z)-
Yb(2/3,1/3,z+
0.500)
6k 0.350 0.000 0.500 238 Yb(x,0,z)-Yb

(i,0,z+0.500)

 

*Strongly overlapping peaks

l6

 

 

Table 4: Properties of Space Group P6/mmm16
Point Coordinates of
Positions Summetry Equivalent Positions
12n ‘m x,0,z; 0,x,z; i,i,z; i,o,z;
0,x,z; x,x,z; x,0,z; 0,x,z;
i,i,2; i,0,2; 0,i,2; x,x,z.
6k mm x,0,1/2; 0,x,1/2; i,i,1/2;
i,o,1/2; 0,i,1/2; x,x,1/2.
4h 3m 1/3,2/3,z; 2/3,1/3,z; 1/3,2/3,E;
2/3,1/3.2.
2e 6mm 0,02; 0,0,2.
2d 6m2 1/3,2/3,1/2; 2/3,1/3,1/2.
2c 6m2 1/3,2/3,0; 2/3,1/3,0.
1b 6/mmm 0,0,1/2.
1a 6/mmm 0,0,0.

 

disallowed the space group P622 because the only levels pos-
sible for x,0,z positions in this symmetry are 2 - 0 and

z = 1/2, a separation of z I 0.500. The space group P6m2
‘was rejected for the same reason. Space groups P6mm and
P6/mmm were not considered because the x,0,z positions in
these two symmetries require placing six metal atoms in the
same x,y plane. With six metals in the same plane, the best
metaldmetal separation in x,0,z positions would be 2.06 A,
less than half the Yb-Yb separation found in YbClz, for
example3 .

Of the five possible space groups listed previously,

P62m.is the only one consistent with the Patterson results.

17

Table 5 lists some of the equivalent positions of this
symmetry, and Figure 1 shows a representation of the
symmetry. All the major peaks in the Patterson map are
accounted for by placing metal atoms in the positions shown
in Table 6. The values of x, y and z for the positions are
refined values to be discussed later.

The metal atoms are located in the unit cell in three
sets of special positions. Two sets are x,0,z positions
around 2 = 0.000 and z = 0.500 ("staggered" to each other
with x = 0.638 and x = 0.369) which account for the Patterson
peaks at 0,0,0.l87 and 1/3,2/3,0.l87--each x,0,z set is split
by symmetry into two eclipsed subsets with an average spacing
between them of z = 0.187. Between the staggered sets are
average distances of z = 0.310 and z = 0.500. The third
set of metal atoms occupies the positions l/3,2/3,z at z =
1 0.251. Between this third set and the first two are aver-
age distances of 2 equal to 0.155 and 0.337.

Some metal to anion vectors are found in the Patterson
map at low peak heights. The x,0,z sets of metal atoms are
spaced an average of z I 0.094 above and below the levels
of z = 0.000 and z = 0.500. Thus the vector peaks at
0,0,0.098 and 1/3,2/3,0.098 are interpreted to mean that
anions are located at the zero and 1/2 levels in positions
eclipsed to the surrounding metal atoms. The u,0,w vector
positions at w = 0.255 and 0.405 confirm that the anions at
the zero and one-half levels are in x,0,z positions in real

space.

18

Table 5: Properties of Space Group P62m16

 

 

Point Coordinates of
Positions Symmetry Equivalent Positions
6i m x,0,z; 0,x,z; i,i,z;

x,0,z; 0,x,z; x,x,z.

4h 3 1/3,2/3,z; 1/3,2/3,2;
2/3,1/3,z; 2/3,1/3,z.

 

 

 

16

Figure 1: Representation of Space Group P62m

19

Table 6: Refined Atom Positions in Space Group P62m

 

Atom x y z
Anion 1 0.5957 0.0000 0.0000
Yb 1 0.6383 0.0000 0.0962
Anion 2 0.5892 0.0000 0.1900
Yb 2 0.3333 0.6667 0.2513
Anion 3 0.3716 0.0000 0.2902
Anion 4 0.3333 0.6667 0.3578
Yb 3 0.3609 0.0000 0.4073

 

The arguments outlined above account for the gross
spacing and positioning of the vector peaks in the Patterson
map. However, the peak heights and shapes proved difficult
to completely predict in an a priori manner from the proposed
model. To prove that this model is consistent with the
observed Patterson map, two new Patterson maps were cal-
culated--one for the refined metal atom positions alone and
one for the metal and anion positions listed in Table 6.

The results of these two maps are listed in Table 7. Even
though the final refinement must be regarded as only tenta-
tive (for reasons to be discussed later), the Patterson
syntheses from.calcu1ated structure factors show close
agreement with the vector map derived from the observed
reflection data. Even the three dimensional shapes are

reproduced closely.

20

Table 7: Patterson Maps from Final Refinement

 

P6/mmm Relative Heights
Positions u v w Metals Metals and Anions
la 0.000 0.000 0.000 999 999
2c 0.333 0.667 0.000 485 495
2e 0.000 0.000 0.099 0 69
4h 0.333 0.667 0.099 0 27
12n 0.325 0.000 0.155 196 190
2e 0.000 0.000 0.189 380 377
4h 0.333 0.667 0.189 178 175
12n 0.325 0.000 0.255 0 19
12n 0.350 0.000 0.310 100 100
12n 0.325 0.000 0.337 201 200
12n 0.350 0.000 0.405 0 19
1b 0.000 0.000 0.500 173 173
2d 0.333 0.667 0.500 112 102
6k 0.350 0.000 0.500 240 240

 

*Strongly overlapping peaks

The final least squares refinement of the atomic posi-
tions listed in Table 6 and the corresponding temperature
factors listed in Table 8 contains only those positions
which were statistically stable. After eight cycles of
refinement, the largest shift in a metal position parameter
was 0.00009 and for the anion position parameters 0.0007.
The largest shift in the anisotropic metal B values was 0.05

and for the isotropic anion B values 0.02.

21

Table 8. Temperature Factors

 

 

Atom 3:13 B22 B33 B12 313 B23
Anion 1 -2.44

Yb 1 -l.53 -0 81 0.65 -0.81 -0.02 0.00
Anion 2 -1.74

Yb 2 -1.00 -l.00 1.34 -1.00 0.00 0.00
Anion 3 4.96
Anion 4 0.23

Yb 3 -l.20 -l.27 1.56 -1.27 0.21 0.00

 

Scattering factor tables and a correction for anomalous
dispersion were included in the least squares refinement.

The scattering factors computed by Cromer and W'aberl7 were

used for the ytterbium and chloride ions, those of Tokonami18
were used for the oxide ion. The anomalous dispersion

corrections of Cromer19

were used. The scattering factor
tables allow corrections for an atom interacting with the
incident X-ray beam as a function of the scattering angle.
Anomalous dispersion is an effect by which an atom that is
absorbing radiation strongly will scatter radiation with
different phase changes than the other atoms in the structure.
With only the metal atoms in the refinement and with
isotropic temperature factors, an R value of 0.196 was

attained. Employing anisotropic motion gave an R value of

0.136 for the metal atoms alone. For the final refined

22
parameters listed in Table 6 and Table 8, an R value of
0.101 was achieved.

Fourier and difference Fourier series were computed
at several stages in the refinement. With only the metal
atom positions occupied, both the Fourier and the difference
Fourier showed light atoms located directly above and below
the metal atom positions, in agreement with observed
Patterson peaks. Additionally, the difference map showed
that light atoms should be put in the same x,y planes as
the x,0,z metal atoms. The Fourier showed no peaks in these
positions.

0f the light atom'positions indicated by the difference
Fourier, only the four listed in Table 6 could be refined
simultaneously with the metal atoms. This problem will be
addressed later. A Fourier and difference Fourier were
computed from the final refinement. The Fourier only shows
electron density at the seven refined atomic positions. A
structure factor table is included as Appendix I. The
difference map is not flat--1arge peaks are again evident
in the same x,y plane as the x,0,z metal positions. Other-
wise, no peaks are seen in the difference Fourier.

The remaining cyrstals of the same morphology as the
one that has been investigated above were collected for
X-ray fluorescence and powder diffraction experiments. The
powder pattern, although weak because of the small amount of
sample present, was indexed successfully on the hexagonal

unit cell and lattice parameters found for the single

23

crystal. This result is tabulated as Appendix II. Included
with the pattern are both the observed intensities and
relative intensities for the corresponding single crystal
reflections.

The X-ray fluorescence experiment was run on the powder
sample to confirm the presence of ytterbium. For reasons
to be discussed later, an investigation for tantalum was
included in the analysis. The La1 peaks of both metals
were to be observed. At high gain a small but definite
ytterbium peak was seen, but the presence of tantalum could
neither be confirmed nor denied because of interference from
platinum and copper. Of the three materials available as
sample holders, platinum interfered strongly with the
tantalum peak, aluminum contained too much interfering copper
impurity and at the high gains necessary to observe an
ytterbium response, quartz impurities interfered also. Peaks
in the fluorescence spectra of both platinum and copper give
strong responses at the scattering angle of the tantalum

La1 peak.

CHAPTER VI
DISCUSSION

After the metal atom positions of the unit cell had
been determined, it was seen that the phase being investi-
gated contained atoms other than just ytterbium and chlorine.
From a consideration of the closest packing of hard spheres
with radii equal to the crystal radius of the chloride anion
(1.81 X)20, calculations show that between six and seven
layers of three chloride ions each can be packed within the
dimensions of the unit cell. Allowance for negative
deviations from this crystal radius allows for up to eight
or nine layers, or a maximum of 27 negative charges in the
unit cell. Even the assumption that all 16 metal positions
are filled by divalent ytterbium leaves a deficiency of five
negative charges.

This problem may be solved by recognizing oxygen as a

1to

likely contaminant. Ytterbium oxide chloride is known
form in ytterbium chloride systems when a trace of water is
present. From a single crystal structural studyZIthe following
lattice parameters are reported for hexagonal YbOCl: a =

3.726 3., c = 27.830 3.. The atomic positions of YbOCl all

lie on the z-axis of the unit cell, forming linear anion-
metal-anion interactions. The Patterson and Fourier maps in

this study indicate linear aniondmetal-anion interactions,

parallel to the z-axis of the unit cell. An oxide anion has

24

25
a much smaller crystal radius (1.40 A)20 than does the
chloride anion. The double negative charge on the oxide
anion also requires that fewer of these anions be put in the
unit cell. However, a combination of space group symmetry
and closest packing considerations still restricts to three
the number of oxide anions that can be placed in any x,y
plane.

The final refinement listed in Table 6 and Table 8
identifies four anion positions. Anions l, 2 and 4 were
assigned the oxide anion scattering factor table during the
least squares refinement; Anion 3 was assigned the chloride
anion scattering table. This assignment was made in earlier
cycles of refinement; with more than four anion positions
occupied, Anion 3 had metal-anion distances closer to those
of Yb-Cl than Yb-0 (see below). With just four anion posi-
tions occupied, the metal-anion distances of separation have
readjusted to different values, so no definite identification
of the anions should be made from metal-anion separation
distances until after the structure has been completed.

In Table 9 some selected interatomic distances for the
refined atomic positions are listed. These distances repre-
sent the closest approach between layers of atoms and were
obtained from a complete computation of all distances less
than five angstroms. Table 10 lists angles between several
sets of atoms.

From Table 9 it can be seen that the anion to metal

0
distances of separation are no greater than 2.3 A. These

26

Table 9: Interatomic Distances

 

 

Estimated
Atoms Distance Standard Deviation
Anion 1 to Anion 1 3.259 A 0.003 3.
Yb 1 to Yb 1 3.431 0.001
Anion 2 to Anion 2 3.238 0.002
Yb 2 to Yb 2 3.573 0.001
Anion 3 to Anion 3 3.386 0.003
Anion 4 to Anion 4 3.573 0.001
Yb 3 to Yb 3 3.434 0.001
Yb 1 to Anion 1 1.920 0.002
Yb 1 to Anion 2 1.878 0.002
Yb 1 to Yb 2 3.650 0.002
Yb 2 to Anion 2 2.228 0.001
Yb 2 to Anion 3 2.100 0.002
Yb 2 to Anion 4 2.104 0.005
Yb 2 to Yb 3 3.666 0.002
Yb 3 to Anion 3 2.316 0.002
Yb 3 to Anion 4 2.212 0.002
Anion l to Anion 2 3.755 0.003
Anion 2 to Anion 3 2.394 0.003
Anion 2 to Anion 4 3.806 0.005
Anion 3 to Anion 4 2.368 0.003

 

27
distances are shorter than the ytterbium to chloride dis-
tance in YbOCl (2.75 X)21 or in YbC12 (2.76 X)3. They are
comparable to the ytterbium to oxygen distance in YbOCl
(2.2 A)21. It appears then that the distances calculated
from.the atomic positions of the final refinement support

oxide anions in the refinement rather than chloride anions.

Table 10: Interatomic Angles

 

 

Estimated
Angle Degrees Standard Deviation
Yb 1 - Anion 1 - Yb 1 164.21° 0.27°
Anion l - Yb 1 - Anion 2 162.80 0.16
Anion 2 - Yb 2 - Anion 3 67.10 0.07
Anion 2 - Yb 2 - Anion 4 122.95 0.03
Anion 4 - Yb 3 - Anion 3 63.01 0.12
Yb 1 - Anion 2 - Yb 2 125.23 0.03
Yb 2 - Anion 4 - Yb 3 103.09 0.06
Yb 2 - Anion 3 - Yb 3 99.36 0.06

 

Final assignment of anion positions may not be made be-
cause the structure is too incomplete-eanion positions could
not be refined at the z - 1/2 level. Chemically, anions
must occupy this vacant level and are required to balance
the positive charges of the surrounding metal atoms. As
stated previously, the Patterson vectors indicate only anions
eclipsed to the metal atoms at this level. Additionally, the
difference Fourier calculated with only the metal positions

occupied indicates a similar result. In all refinements,

28
however, the least squares program was unable to find a
statistical minimum for the atomic parameters at this level.
Setting the temperature factor to a constant value for anion
positions at the z = 1/2 level did not allow the position
parameters to refine; populating the special positions
1/3,2/3,1/2 and 0,0,1/2 did not help either. Difference
Fouriers calculated with other anion positions occupied did
not even show a small peak at the z = 1/2 level.

There are several possible explanations for this in-
ability of the entire structure to refine. Some of these
include a bad crystal or data set, a bad absorption correc-
tion, the wrong space group or even the wrong assumptions
about the chemistry of the reaction that formed the crystal.

Although very small, the crystal gave good sharp re-
flections with the 001 reflections found to have a mosaic
spread of 0.5° during a 26 scan. This spread is indicative
that there was not a large amount of disorder in the crystal.
Some bad reflections may be in the data set, though. In
the course of refinement it was noted that the calculated
structure factor for a strong 006 reflection was twice the
observed value. When this reflection was deleted from.the
data set, the R value dropped considerably and a few addi-
tional anion positions could be located. The 006 reflection
was the only strong reflection to stand out in this manner,
and it was the only one deleted from the data set. The
complete list of observed and calculated structure factors

is given in Appendix I.

29

Table 2 lists a very large linear absorption coefficient
(598 cm-1) for the crystal being investigated. As described
earlier, the empirical absorption measured small angular
variations in the intensity of a reflection. For crystals
that absorb X-rays strongly, the empirical correction measures
small variations in a large absorption effect, allowing the
correction to be imprecise.

As mentioned previously, P62m is the only space group
of the five groups considered which could accommodate the
‘metal orientation that was consistent with the Patterson map.
Yet there is a very special type of crystal twinning that
could be operating in this system--Buerger22 warns that in
apparently hexagonal cells which have no general or special
extinction conditions, the possibility of trilling should be
considered. Trilling is a twinning mechanism by which three
orthorhombic cells of symmetry 2mm are intergrown in a manner
that yields an additional mirror plane for a combined sym-
metry as high as 6/mmm. Intergrowth such as this masks the
true extinction conditions.

The observation of pseudo-extinctions or periodic ar-
rangements of weak reflections might support this or other
twinning mechanisms. A pseudo-extinction is especially
noticeable along the 001 axis where the intensity for a re-
flection with an even numbered Z is consistently greater
than for an odd numbered 1. In defense of the solution, how-
ever, it should be pointed out that the least square refine-
ment and calculated structure factors match this odd-even

behavior quite closely.

30

The assumption that ytterbium was the only heavy atom
in the crystal was also investigated. The crystal was iso-
lated from a reaction carried out in a sealed tantalum
ampoule, so tantalum is a possible heavy metal contaminant.
Additionally, the possibility of tantalum.being present in
the crystal was investigated because of certain similarities
between published structures in a tantalum.cxide system23’24
and results of this study. Specific points of similarity
are some tantalum oxygen distances of less than 2.0 A and
negative temperature factors for some of the metal atoms
and anions. In the tantalum contamination model of the
structure being investigated in this work, the two x,0,z
sets of metal positions can be interpreted as groups of
composition Ta6015, with nine oxygen atoms above, below and
bridging the x,0,z metals. Six oxygen atoms occupy positions
previously mentioned as being large peaks in the difference
Fourier calculated from metal occupancy only: anion positions
almost in the same plane as the metal atoms, in x,0,z
positions staggered to the metal positions.

This tantalum contamination theory is rejected. One
reason is that the oxide anion position in the same planes
as the x,0,z metal positions cannot be refined, the least
squares program being unable to find a position for it.
Additionally, this oxide anion position gives a closest
oxygen-oxygen approach of less than the sums of the two

crystal radii.

31

The peak in the difference map that gave rise to the
tantalum oxide impurity speculation is not attributed to a
real anion position but rather to a reflection of the aniso-
tropic motion of the nearby metal atom. The short ytterbium-
oxygen anion distances listed in Table 9 and the negative
temperature factors in the final refinement are attributed
to a lack of electrons in the refinement--i.e., an incomplete
structure. Small changes in the position coordinates of the
atoms in a complete refinement could add 0.2 A to the short
distances. The negative temperature factors are a result
of the least squares program trying to match the observed
structure factors. If more electrons could be included in
the refinement, the temperature factor would not have to be
varied so much to calculate good structure factors.

The results of this crystal study can be tied to other
work with lanthanide series elements. Caro has reported a

study25

of the crystal structures of rare-earth oxides and
oxide salts and is able to describe these structures as 0M4
tetrahedra sharing edges. Specifically, the structure of an
oxide chloride is described in this theory as sheets of OM4
tetrahedra, each sharing four edges.

Caro's theory is consistent with only part of the
structure found in this study. Tetrahedral orientation for
the metal atoms cannot exist for the x,0,z positions at z =
0.096 and z = 0.407. Each of these sets of metal atoms is
eclipsed by symmetry, so oxygens at z = 0 and z = 1/2 do not

conform to Caro's theory. Tetrahedral orientation does

32
exist between x,0,z and the l/3,2/3,z metal positions which
are staggered to each other. Anions located in planes be-
tween these metal positions must necessarily be located in
a tetrahedral M4 net although the last three angles listed
in Table 10 show large deviations from the 1090 angles of a
regular tetrahedron.

2

Fishel attempted to prepare crystals of Yb304Cl anal-

ogous to the phase Y304C126. From his work a new phase of
unknown composition resulted. A comparison of the X-ray
powder diffraction pattern listed in Appendix II to that
reported by Fishel shows no relationship between these two

phases.

CHAPTER VII

CONCLUSION

The intent of this research was to characterize a
mixed oxidation state in the ytterbium(II)—ytterbium(lII)
chloride system. Based on ionic size considerations and
observed metal-anion distances from the partial structure
refinement that was possible, the phase probably contains
oxygen atoms.

Sixteen ytterbium atoms are contained in the unit cell
in two types of special positions: two sixfold x,0,z
positions and a fourfold l/3,2/3,z position. Moreover,
anions occupy only these two types of positions as indicated
by the Patterson map, Fouriers and difference Fouriers, and
the partial refinement of three metal and four anion
positions. As a necessary consequence of space group P62m,
these sets of ions are located in a layered pattern along
the long c—axis of the unit cell, in good agreement with the
plate-like morphology of the phase.

The ytterbium positions and the partial set of refined
anion positions may be interpreted in terms of three possible
models for the complete crystal structure. All three models
assign both x,0,z sets of metal atoms to Yb203 structural
units. The first model consists of the l/3,2/3,z metal
atoms as part of linear units of YbClz, the second puts
Yb0C1 units in these positions, and the third has YbO units

in the fourfold position.
33

34
These three models require a linear group around the
fourfold position. Of the three possibilities, Yb0 units

le

must be linear, units of YbOCl are reporte to be linear,

and units of YbCl2 are reported3 to have a bent configuration.

1 contaminant in

Because the oxide chloride is a known
ytterbium chloride systems, the model chosen for the complete
structure is the one having YbOCl units in the fourfold
positions. The final structure by this model would consist
of sesquioxide and oxide chloride units to build the
Yb16022C14 unit cell, or (Yb203)6(Yb0C1)4. This formula was
used to calculate a density and linear absorption coefficient
for the crystal, as listed in Table 2.

Figures 2-4 show three projections of this structural
model. Figure 2 is a projection of the refined metal atom
positions on the (001) plane, Figure 3 is a projection of the
refined anion positions on the (001) plane. Numbering of
positions on both figures corresponds to the numbering system
of the final refinement (Table 6). Figure 4 is a projection
of the Yb16022C14 model on the (010) plane. The darker
circles of Figure 4 represent atomic positions in the (010)
plane, the fully dashed circles represent anion positions
that are not refineable. Figure 3 shows the near-eclipsing
of x,0,z anion positions, Figure 4 shows the layered packing
of ions in the unit cell.

The unit cell for the above model may be considered to
be derived from the hexagonal symmetry of YbOCl with that of
the cubic symmetry exhibited in ytterbium sesquioxide.

35

 

 

 

Figure 2: Projection of Metal Atom Positions on (001) Plane

 

 

 

Figure 3: Projection of Anion Positions on (001) Plane.

z = 0.000
2 = 0.096
2 = (0.12)
z = 0.190
2 = 0.251
2 = 0.290
2 = 0.358
2 = 0.407
2 = 0.593
2 = 0.593
2 = 0.642
2 = 0.710
2 = 0.749
2 = 0.810
2 = (0.88)
z = 0.904
2 = 1.000
Figure 4:

 

36

 

 

’\
I

"

P

 

.0
T’O

~"

\o

m

Oxygen

. Ytterbium
‘5 (Chlorine)

Oxygen

< Ytterbium
Oxygen

C Oxygen

Ytterbium

(Oxygen)

Ytterbium

< Oxygen

Oxygen
C Ytterbium

Oxygen

' (Chlorine)
Ytterbium

 

Oxygen

 

Projection of Yb16022C14 on

(010) Plane.

37

Wyckoff27

indicates that the sesquioxide unit cell contains
16 sesquioxide units; the metal atoms are in special posi-
tions at four levels on the z-axis, with 24 atoms in an
x,0,z closest packing type of position in the Ia3 space
group and eight in special positions at z = i 1/4. The
oxygen atoms are in the 48 general positions. Substitution
of chloride ions for some of the oxygen atoms in the cubic
sesquioxide lattice probably distorts the symmetry to
hexagonal because the large chloride ion increases the anion
to cation radius ratio. The structure remains similar, the
symmetry changes.

The structure is not refineable further than the metal
positions and four anion positions, most likely the result
of a faulty data set or a bad absorption correction. Pos-
sibly a few strong reflections were measured at an intensity
less than the actual value. This condition would be similar to
that mentioned earlier in which the 006 reflection had to be
deleted from the data set. Reasons for the reflections being
recorded at too low intensity values are not known but
probably reflect an absorption or extinction problem. As
discussed earlier, for a small crystal that absorbs X-rays
strongly, a good absorption correction may be difficult to
obtain.

The possibility of twinning cannot be completely re-
jected until the crystal structure is solved: the proba-

bility of twinning is lessened by the sharp diffraction peaks

38
observed in the data set. A twinning geometry would have
to be of a precise nature that did not introduce disorder
into the crystal and broaden the diffraction peaks.

The final solution will require at least the collection
of another data set--h0pefully from a larger crystal which
is not now available. Since the full structure is not known,
a direct reaction to obtain more of this phase may not be
possible. A.repetition of the one reaction which led to
this phase may reproduce the conditions which led to its
accidental formation.

The original intent of this research was to increase an
understanding of the structural aspects of phases of inter-
mediate oxidation states in the ytterbium(II)-ytterbium(III)
chloride system. The new phase that was isolated was con-
taminated by oxygen and only a partial structure could be
obtained. This result may not help characterize the chlor-
ide system, but it does help characterize the reactions used
to obtain the phases in the chloride system. The phase that
was isolated as a side product of these reactions and its
presence can now be rapidly identified from.the X-ray powder

diffraction pattern.

REFERENCES

REFERENCES

l) D. Brown, "Halides of the Lanthanides and Actinides,"
John Wiley and Sons Ltd., London, 1968.

2) N. A. Fishel, Ph. D. Thesis, Michigan State University,
East Lansing, MI, 1970.

3) H. Barnighausen, H. Patow and H. P. Beck, Z. Anorg. AZZg.
Chem., 403, 45 (1974).

4) D. H. Tem leton and G. F. Carter, J. Phys. Chem., 58,
940 (1954 . .

5) J. D. Corbett, Rev. Chim. Miner., 19, 239 (1973).

6) D. A. Lokken and J. D. Corbett, Inorg. Chem., 12, 556
(1973).

7) B. c. McCollum, M. J. Camp and J. 0. Corbett, ibid.,
12, 778 (1973).

8) P. E. Caro and J. D. Corbett, J. Less-Common Metals, 18,
1 (1969).

9) N. A. Fishel and H. A. Eick, J. Inorg. Nucl. Chem., 33,
1198 (1971).

10) ‘w. 0311 and w. Klemm, z. Anorg. AZZg. Chem., ggl, 239
(1939).

11) M. D. Taylor and C. P. Carter, J. Inorg. Nucl. Chem.,
24, 387 (1962).

12) C. W. De Rock and D. D. Radtke, ibid., 32, 3687 (1970).

13) G. H. Stout and L. H. Jensen, "X-ray Structure Determin-
ation," The Macmillan Company, New York, NY, 1968.

14) P. W. R. Corfield, R. J. Doedens and J. A. Ibers, Inorg.
Chem., 9, 197 (1967).

15) M. J. Buerger, "Vector Space," John Wiley and Sons, Inc.,
New York, NY, 1959, Chapter 10.

39

16)

17)

18)
19)
20)

21)
22)

23)

24)
25)
26)
27)

40

N. F. M. Henry and K. Lonsdale, Eds., "International
Tables for X-Ray Crystallography," Vol. 1, International
Union of Crystallography, Kynoch Press, Birmingham,
1952.

D. T. Cromer and J. T. Waber, Acta Cryst., 88 104
(1965).

M. Tokonami, ibid., 18. 486 (1965).

D. T. Cromer, ibid., l8, 16 (1965).

L. Pauling, "The Nature of the Chemical Bond," 3rd
Edition, Cornell University Press, Ithaca, NY, 1960,
p. 514.

G. Brandt and R. Diehl, Mater. Res. Bull., 8, 411 (1974).

M. J. Buerger, "Crystal-Structure Analysis," John Wiley
and Sons, Inc., New York, NY, 1960, Chapter 5.

N. C. Stephenson and R. S. Roth, Acta Cryst., B27,
1010 (1971).

N. C. Stephenson and R. S. Roth, ibid., 881, 1018 (1971).
P. E. Caro, J. Less-Common Metals, 18, 367 (1968).
S. Natansohn, J. Inorg. Nucl. Chem., 88, 3123 (1968).

R. W. G. Wyckoff, "Crystal Structures," Vol. 2, John
Wiley and Sons, New York, NY, 1964.

APPENDICES

41

Structure Factor Table

APPENDIX I

( L Fflfl< FCAL K L FDRS FCAL

L rnus VCAL

roux 'CAL K

 

7H"517705‘H‘h75975“??01QR“6?R56H5RQQRQQQ1QO§qq77flnRQHQU n7.31505““9°nhqqnn5759799579g?“5n0055,000
70,475hQ55576597Q7Sn?03020550625766??566 2747nQBR?03°Q7 53.7553‘076?“2537SR703666900700785.89%]
75W9$70“HIPWHH“qsa742?fi05?39§7§7569697575457569Q70527§93906“35“1830?08999.3055855597095989O6075
1
«In0‘6076940404106790206900749P3476085705935818Q759124§44 110036135036 3204803060455071350 093600
‘9‘"6"‘"QQ?‘?Q7HQ“QQQQ‘Q“ S5“fi???‘“7135766quah13HQQ?755 ,5 5355“ 6960955 1n977‘7‘36n91 “72
$70010Q5lqg“?75070Q1??705 g0“gA?R5q96g5525n?70&985‘65767 25 65960 0‘069‘9 573‘sng0907g7 Q35
1? 5 P7555 ?5 5 2 ? ‘7 ??5 5 ’5 55 3? 23 5552287 5 S 2 2 ‘ ? 5 35 575 5 55. 5?
6’HU"520575“$57“905?TQQfi7H90.05?.3Lq578905?15q50523556739MHH05PJQSfi7R905?3“5670521$S67N9m5?H052300
5555??? 5555555 55555555
5. H
O
77????715‘1‘15511111‘1‘11‘11h4b‘44““h“bhbbhbhgqggq§qggsqql1‘115113131111aaajbbbfih“““bkkkhu555600
nachu1?17«A8609708€1719916711QQRIQQ175440982668hfiQl??91flfi?lQ§9067792A091306497§oI677Q?§60§n§a010
7Q5hgr’1Q?QH“Q“?TQQQQQQ90"97RQ“QR““TASRQQA36QO5O‘SR1Q8732485hq564nqqq;7qn058g070$686566“QI097RH7
57H77fi?hUH50UHH $559 5SN6629399“571h§996015359h?“770?flfl7756?0503556926662 “6179996457QQSEQH/0?5H9
‘5 5?5 5 55 5 55 3? 3“ 75713 21? 55 5? 5 555 511 125 “‘53“ 75?““ 2
7
5hPFQEH5RQHQfin“aflPQB‘HQ351?Q650fi‘6nhhfl577““305371176&65n71664q“090595650067QQ75 963570fi3555fifl5‘7
Ak'qnflknfi77?7flg QQHRQQQP5QJQQ7OHQK 0096?QO0557OQH0§511?Q?QRhl?795fl96?h?a 7RP‘7‘ 0q7R35735flnR4qqz
5H07051finq14)qq 7&507q071?51004??‘ "7034459“Rq?hfla7?nnaQ5Tnnpnn507fl?6117 717640 OQQ110675°441§73
5 5. .5? 5?.555 55 55 55 5 “1.51“ «(Ila/11:. ,1? 55 5? 5 5 5 5.5? ??5 35‘35th ???h“ 3
‘bthHQU5?505?3“6678905?3“q67n°0520523QS67RQ05233557a905236567390573Q50523“567RH0523“§678905?3“5
5555555??fil? 5555555555??? 5555555555 555555 555555

0
O
P7?)7??7?7)555‘113111‘1133331331‘3“Q““Q““5““Q““““‘““Q“5555SSSSSSSSS§556565565550???2???2?722,??2

0507h?71153305 5R51ORAQ15QO57QQO1GQIQ

 

 

 

 

 

Gullnzwuhlaulnwfiah1 R01641afi? 84103§?043766HI7Q5071885425RAPI
5'55hnnu55NU5hthfi77hh555’nnn5q “56“7‘O1QHR‘5R07507Q701 05h5.‘45““7&59hfinQRQ5QIOHH589758102454““HIQ
anzlla 215 026171??§ QAMPQ???IIQ10an?7a1?760la II1PQ|61680R66760H414lR Lo?aauonnwl??€?§l?oonuna7
5 1 ? 5 5 ? ? a 5 2 6 5 q 5 5 5‘15hq 12““ 1‘? 5 1755555 55 5 55
5
“WWI"“000’55’5nhORh 050“]?6nhahs sfiohbq0‘77070709997505050 oqqqana76205OQQSPDQhQQ5h55?]qfinnn1a?no
71“, 5 05105 R q” H 10775n5qfl 5Q7nn “75 h 1 ““556 7 5Q Q5qon>qq9161 7gb17n51nfifi5n5ARhQ?Rn957h
WI?) 5 ‘1055 h 5‘ Q ??51“1“?Q “7770 “78 q 1 5?9“8 h a“ 6&06775350qa hqn3q7qgn611H?’ho7kn5597q
5 5 ? 5 5 5 5 1 7 1 5 1 5 2 q 5 b 5 5 1 :5QB5AQ ??3“b 71? 1? )5 ll 5
5/505 545h,“9052HM55’RQO505?‘5555RQ052148570523kgh7nq0523H05?3“§57RQ052‘“gfi7aqo525hn521QQ6789052
7)? .18? 55555555 5555 5555555555222)? 555
I
O
1‘36kkéhhhbh“hhhhkh“5““““qsgqqsgggqqsgqsqg565566666666566. L l ?222??2227222
OhUH5“55W”55’Mq5)0H5536h2h56“““OQQOQhQ??97)Hfiq“OjneQQ6°“?76“53270°QQ6789476554n67058725738505h78
n,\5QIIHH)7U"57“"GQ7WHHH?176n55fin na7075¢77nQ551K?Q‘?97HQR17an1q5OQQ7P7QQH‘5gn5h?75a7?Q473Q‘in?
n7hk5655hU)ththQ?601‘Rq 557 6 5 h???° 1 R59 5 15n?6595 ??5QI 515973 5“5?h15??5??95?7§fl87“9719“
) 2 B H55 7 ? h 5 ? 5 5 5 5 5 5 5 2 5 5 q 7 1 7 1 h 1 2 7 ‘
5”“061555URMNHIQHQRfiHQQ55)Q“0)‘IRRPO 1? ”:70‘20.00095§9qa7“555750105010086005gio701?6q575501h6079“?
“55h05n5hflhbfih“HRHHH555HHOQO55915Q ?00775005 ? 7596923h5qg72 7 q A 10° 6‘55 0 Qflhh5hAQOQP06Rqoq
“5)WWQ5fln5Q‘MQ”H5Q92574955285‘Io55 0285“5R?B 1 $2656?8?A?5?5 1 k n 3?? P76, “ 55°51?486754“52’6
2 P 5 55M5‘ ? 25“ ? 2 5 5 5 5 5 2 8 7 6 2 3 h 3 2 ? 3
”)5 9. 5:3905t16§fi7fl905258505?34§6739052345678905710505?35567890523hq67690583“05?3“§67“905236567H90
5555555555?P???? >I????2 555555555522??? p
O
I
O00000000000000000000000555555555555555555555555552222222222282222222?22222]33333333333333333333

42

 

 

APPENDIX II: Observed and Calculated Powder Pattern
Intensity
d-value Relative Single
th* Calculated Observed Observed Crystal
100 5.3589 R 5.3535 3 vw 25
111 3.0567 3.0450 m 54
112 2.9526 2.9377 ‘m 41
113 2.8005 2.7965 VW 16
114 2.6223 2.6212 vw 47
202 2.5861 2.5781 vw 17
115 2.4364 2.4309 w 47
211 2.0149 2.0099 vw 22
300 1.7863 1.7811 m 88
306 1.5703 1.5672 vw 69
221 1.5423 1.5357 vw 50

 

*Indexing based on a = 6.1879, c = 19.7637 3