P-thuec a .

r5}

uv-wv Irvr,‘av~u-p.w-

 

 

g; . .
'qu '
015‘

:-

xv]

“‘73“ ‘ , E

m ., Y“?~§ +— -
- am‘finfi
‘4 'mzzzuth.

‘ . u
damp" 413.5.
J}

1,1.
u

g . ’ r, 1
":31???" $1:§j.,' . Jam" V ,' ' 9:333
55'; 'P‘i'x’aqrwk- .a‘f é AJ-
0. k 4
1 fl

6

w 7

s.
a

‘31

FM
.’,1.:}‘ -
.. jig?!

- n
“31$
J"

9 32.74, '
".13:

‘1‘ h
5% e m... I.
t Jen -M4lr. u

was“ i

1331:“ | ’1. .. w
.m. "inky,“ a A? . um 3|

. n - ap- Mgr“? 55g
on 3‘}

.1
. i
v.9? nr

.U.‘

m

3
~

“3’“?-

(z s‘
'5‘»

LL31
‘7

. ' 4 “n' i-

{’5‘ I ~ 5‘ align:

"“(flfifip’h . 9591-]. A" w
“in! " .3 '1 “all"

‘£

3‘35
t 3|.“

1,"!
.‘GMVI- a. ‘

.., ; “2mm.”

.3” z-

‘ 5. 4‘
m2“?-
- ‘hfié
3:22:45:
3.;
I
a;

| "J:

2'."

gnu ' " ‘
“'3 3:59?! '

521%

l V
“1 ’él‘hfiu‘dm
t":t.~;.:':z!;r;z..¥.: r“,
3 arm , ;

V a, g“: ' 2F
95;:- ‘Bsrr "-

‘L 22%??? _".'.:
I" ‘1 p

[‘11

‘nv'u‘

I _ ‘ _

'L. .1, .. _ . _.. v

via-u- , r- t- ‘ J > .3

n1.» ,.. » ' . .. .H ’ , 5 n ._ . r p.55; '4

'. “ v ' ' 3‘ ‘ r ’: wa‘u'xm

‘ A u - V r 4 v 7‘ .

. I A. .. v a . -§f f‘, ’ 1' ”:4. l,

‘1”! r1"! , .. ‘ . . "w 1 ":1ng 1‘ “ ”var-M. “'1”, I, v“

' "111;“ i I. 4 , ,. ;<.J:.J.- “41,0 par-3 I w

.1

.-.
't‘éfiwi”
H 'M - ‘ ‘ #Lra.au‘pr I‘Q’
"' " " I ' ‘ .r: ‘ ‘
31$ ' 2‘ 5 ~ ’ '1 '- U Lungxéflj
‘ 3 .:.-':121,;5!.' ';
-‘ _ ...,..4}..a;u._,,,
twat? gm“- .. ’,"
'J.'(.v.m.\. '55}er
N “v m.
a ~ - ' ‘

 

.

... .

,.y .v‘w
mwu " v'

V‘l:“l )‘ " , g - ‘ (

r. 'ggv-"vf,"," , m- .n ‘ .,‘.'.’ ~31"?-

,'u .. W1” _ ‘ V ‘ H ...,u..m

. ‘ fl. .,.

a. __ r 3"a’.-‘

..,.

-.

o-

. ,:-.
II
..

'.Au»4 r
..‘,‘.--b-
;~Ll\.-,‘ m

.

rd .
N‘ ’33.". a.
may. .4

,, r X . m. u.
XI. 7‘ . «I: “
p.49!— v .

.5, .m'fl
(41:131.. ,

l
bu...

. » . .

r 1"”? " IA:
«4 tr ""‘,u'v1'
ll V‘I'O‘II :WL-I’P'fi'
mu .— .1.
a.“ ‘.

(. ¢ v!”

.x.:- :0. 4. , » rm.» Va“

n,""-‘w44— v v a“ c J u . «z», ":u'; ‘ 39‘

i “m." 1‘"qu --:f v}! ".73."?

”"v’p’gr-HV- in..." $631! 0’7"" J.
l I.

won «10v
4.4- w -

’1 K
n g
94239.1“,

33.
‘52).

n . v ‘ 1 y" - u ‘
~ ”win-J" '- .,‘ ~ " cW'r.,. . . . ,I-
.‘-="r'9-.y'»- - ; -.'..:’..J", . ' .”'3"':W.:.. - m." ' . "fl x;
v ‘ n“ ,‘ . V - _ ‘ . v .
“’9‘" . "in!" «my 7. in! ( fi’h‘fi? ‘ mam :Ié'crgg
s r - V H ‘ ’ '9‘",

(’JfiFfiyfi 1' «poy'o'v’q’dflgfi

 

ruesw

MICHIGAN STATE

Hilllilil llllllliHiJ Illnlillllllllllllli

31293 00914 6030

III“

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

I. Syntheses, Structures and Formation Mechanisms of
Metastable Modifications of Metal Dihalides

II. Magnetic and Electrical Properties of Oxidenitrides

presented by

Guo Liu

has been accepted towards fulfillment
of the requirements for

Ph.D. Chemistry

degree in

 

 

:7 filfiflfi/ (jg? Zgigtg

Mljor professor

 

Date ' /(7r/Zfl)

MSU it an Affirmative Action/Equal Opportunity Institution O~ 12771

 

 

 

i LIBRARY
Mlchlgan State
; University

v..—

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

a

 

 

 

 

| .

”—1? fl

__Jl__l;|

2| 1:
II II _

—1 Ti!

MSU Is An Affirmative ActIorVEqual Opportunity Institution
cMmun

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I. SYNTHESES, STRUCTURES AND FORMATION MECHANISMS OF
METASTABLE MODIFICATIONS OF METAL DIHALIDES
and
II. MAGNETIC AND ELECTRICAL PROPERTIES OF OXIDENITRIDES

BY

Guo Liu

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry
1 990

3’
i
V

927-» / 9 ‘

/.
(4/‘

ABSTRACT

I. SYNTHESES. STRUCTURES AND FORMATION MECHANISMS OF
METASTABLE MODIFICATIONS OF METAL DIHALIDES
By
Guo Liu

Among the various low temperature chemical methods utilized for the synthesis of
crystalline solids. those based on topochemical reactions uniquely allow the synthesis of
metastable phases not accessible by conventional methodsbecause the low temperature
preserves the essential features of the precursor structures. Three types of metastable
phases have been produced by combination of a new topochemical reaction. solvolytic
decomposition of mixed-valence halides Lnx,“ and mixed alkaline earth-Ln3+ halides. with
low temperature dehydration. These are: fluorite-type high-temperature modifications of
LnCl, (Ln-Sm. Eu). anti-Fe,P—type 'high-pressure’ modifications of Bax2 (X=Cl. Br), and
Sri,-lV-type 'high-temperature’ forms of Ml2 (M=Sr, Sm, Eu). New vemier- and cluster-
type mixed(-vaient) halide precursors were also prepared. The metastable Srla-Iv
structure was refined by the X-ray Rietveld procedure. Reactions that yield these
metastable phases are analyzed in thermodynamic and kinetic terms and are shown
through layer descriptions to be topotactic. Although low-temperature high-vacuum
dehydrations do not involve major structural changes. those that occur during solvoiytic
decompositions incur significant changes. ln-Iayer atomic rearrangements of the
intermediate lattice lead to formation of the fluorite-type structure; further out-of-layer

displacements yield the anti-FezP-type structure.

a.

 

Guo Liu

ll. MAGNETIC AND ELECTRICAL PROPERTIES OF OXIDENITRIDES

Studies of high Tc superconductors have been centered on oxides. Reports of
anion substitution in these ceramic superconductors are far less extensive than those of
cation substitution. Some of the properties of N" suggest it to be a likely candidate for
O" substitution. Three types of oxidenitrides were then synthesized and their magnetic
susceptibilities and electrical conductivities determined. (1 ) Pyrovskite-related MM0(O,N)3
(Masa, Sr) exhibit complex magnetic behavior at low temperatures and are probably
metallic; (2) CaFe,O4-type BaCe_Lg(O,N); (gs-La. Ce) are insulating and paramagnetic
with Ce exhibiting mixed-valent states; (3) NaCl-type LaO,N,,,‘ are metallic above 6 K for

x=0.45 and 5 K for x2028 and become superconducting below these temperatures.

ACKNOWLEDGMENTS

l gratefully acknowledge the instructions. suggestions, support. and encouragement
of Professor Harry A. Eick throughout my studies at Michigan State University. i would
like to express my appreciations to Dr. Wieslaw Lasocha for his friendly help and
instructions at the very beginning of my research work. and to Professor Shi-Hua Wang
and Stanislaw A. Hodorowicz for their suggestions and discussions. i thank Professor
Mercouri Kanatzidis and Daniel Nocera for their inspirations.

Many thanks to Reza Loloee and Kevin Moeggenborg for their assistance in the
magnetic and electrical measurements.

Support of the Division of Materials Research, Solid State Chemistry Program.
National Science Foundation, Grant DMR-84-00739 for the first part of this work. and
support of the Center for Fundamental Materials Research at Michigan State University
for the second part, are gratefully acknowledged.

Finally, I thank my wife, Ying Yang. for her patience. encouragement, and support.

TABLE OF CONTENTS

LIST OF TABLES ................................................ viii

LIST OF FIGURES ............................................... xi

PART I SYNTHESES, STRUCTURES AND FORMATION

CHAPTER 1
1.1

1.2

CHAPTER 2
2.1

2.2

2.3

CHAPTER 3
3.1

3.2

MECHANISMS OF METASTABLE MODIFICATIONS

OF METAL DIHALIDES ........................... 1
lNTRODUCTiON .................................... 2
Low-Temperature Routes to Solid State Compounds ............ 3
1.1.1 Synthesis Based on Topochemlcal Reactions ............ 3
1.1.2 Synthesis Based on Low Temperature Chemical
Decompositions ................................. 5
Soivolytlc Decompositions and Metastable Phases ............. 8
BACKGROUND lNFORMATION ........................ 10
Structural Chemistry of Lanthanide and Alkaline Earth Halides . . . . 11
2.1.1 Characteristics of Lanthanide and Alkaline Earth Halides . . 11
2.1.2 Mixed and Mixed-Valent Halides .................... 12
Crystal Structures of Related Halides ........ ' .............. 17
2.2.1 The PbClz-Type Structure ......................... 17
2.2.2 The Anti-FezP-Type Structure ...................... 22
22.3 The Fluorite (Cam-Type Structure .................. 25
2.2.4 Fluorite-Related Cluster Type Superstructures .......... 31
The x-ray Rietveld Refinement Procedure .................. 37
2.3.1 introduction ................................... 37
23.2 The Rietveld Refinement XRS—82 System ............. 38
EXPERIMENTAL ................................... 44
Synthesis Procedures ................................. 45
3.1.1 Synthesis of Individual Halides ..................... 47
3.1.2 Synthesis of Precursors .......................... 49
3.1.3 Decomposition of Precursors ....................... 52
instrumentation and Data Processing ...................... 55
32.1 Powder X-ray Diffraction Examinations ................ 55
3.2.2 Automated Powder X-ray Data Collection .............. 55
3.2.3 Data Processing ................................ 55
3.2.4 Rietveld Refinement of the Srlz-lv Structure ............ ° 56

CHAPTER 4 RESULTS ........................................ 60
4.1 Precursors and Structures .............................. 62

42 Decomposition Products and Structures .................... 74

4.2.1 High-Temperature Forms of LnCl2 (Ln = Sm. Eu) ........ 74

42.2 High-Pressure Forms of Bax, (X a Cl, Br. I) ............ 77

4.2.3 High-Temperature Forms of MI, (M a Sr, Sm. Eu) ....... 83

4.3 The Structure of Srlz-iv by X-ray Rietveld Refinement .......... 88

4.4 Thermal Stabilities of Metastable Phases ................... 95

4.5 Soivolytlc Decomposition lntennediates .................... 95
CHAPTER 5 DISCUSSION ...................................... 100
5.1 Soivolytic Decomposition ............................... 101

5.1.1 Energetics .................................... 101

5.1.2 Solvent Effect .................................. 103

52 Low-Temperature Dehydration ........................... 104

5.3 Metastable Structures and Low Temperature Decompositions . . . . 105

5.4 Formation Mechanism of the Metastable Forms .............. 107

5.4.1 Kinetics vs. Thermodynamics ...................... 107

5.4.2 Topochemlcal Transformations ..................... 107

5.4.3 Geometric Correlations . . . . . ...................... 118
CONCLUDING REMARKS ....................................... 121

PART II MAGNETIC ANDELECTRICAL PROPERTIES

CHAPTER 1
1.1

1.2

CHAPTER 2
2.1
22

CHAPTER 3
3.1
3.2

OF OXIDENITRIDES ............................ 122
iNTRODUCTlON ................................... 123
An Overview of Transition Metal Oxidenitn'des ................ 125
1.1.1 Perovskite—Type ................................ 125
1.1.2 NaCl-Type .................................... 126
Goals ............................................. 127
12.1 MMo(O.N), .................................... 127
12.2 BaCeLn(O.N). ................................. 128
12.3 LaO,,NHI ..................................... . 128
MAGNETIC PROPERTIES OF SOLIDS ................... 129
Magnetization and Magnetic Susceptibility .................. 130
Five Basic Types of Magnetic Order ....................... 131 .
22.1 Diamagnetism ................................. 131
22.2 Pararnagnetism ................................ 131
22.3 Ferromagnetism ................................ 134
22.4 Anti-ferromagnetism ............................. 135
22.5 Ferrlmagnetism ................................. 136
ELECTRICAL PROPERTIES OF SOLIDS ................. 138
The Band Theory of Solids ............................. 139
Electrical Conductivity ................................. 141

vi

3.2.1 Temperature Dependence of Metallic Electrical

Conductivity ................................... 142

3.2.2 Semiconductors ................................ 142

3.3 Superconductivity .................................... 144
CHAPTER 4 EXPERIMENTAL ................................... 146
4.1 Synthesis and Structural Characterization of Oxidenitrides ....... 147

4.1.1 List of Chemicals ............................... 147

4.1.2 Synthesis Procedure ............................. 148

4.2 Magnetic and Electrical Measurements ..................... 150

4.3 Data Analysis ....................................... 152

4.3.1 Magnetic Data ................................. 152

4.3.2 Electrical Resistivitles ............................ 154

CHAPTER 5 RESULTS AND DISCUSSION ......................... 156
5.1 Pseudo-Temary BaCeLn(O.N)4 Systems (Ln . La. Ce) ......... 157

5.1.1 Chemical Characterization ......................... 157

5.1.2 Magnetic and Electrical Properties ................... 157

5.1.3 Crystal Structures ............................... 160

5.2 Pseudo-Ternary MMo(O.N)a Oxidenitrides (M a Sr. Ba) ......... 165

52.1 Crystal Structures ............................... 165

52.2 Reaction Conditions ............................. 168

52.3 Magnetic and Electrical Properties .................. 169

5.3 Pseudo-Binary LaO,N,,, Oxidenitrides ...................... 176

5.3.1 Electrical Properties ............................. 176

5.3.2 Magnetic Properties ............................. 180
CONCLUDING REMARKS ........................................ 187
GENERAL REFERENCE BOOKS ................................... 188
REFERENCES ................................................. 189

vii

Table

10.

11.

12.

13.

14.

LIST OF TABLES

Page
Structural modifications of ianthanide halides Lnx2 .................. 13
Structural modifications of alkaline earth halides MX2 ................ 14
Space groups and lattice parameters of mixed-valent Ianthanide halides
(excluding fluorides) ........................................ 15
Space groups and lattice parameters of alkaline earth-related mixed
halides (excluding fluorides) .................................. 16
Atomic positions of the fluorite-type SrCl, in cubic and
hexagonal unit cells ........................................ 30
List of atomic positions for Nd..Cl,, ............................. 32
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for Eu.Cl, ........................................ 63
Miller indices. and observed and calculated interplanar d-spacings and
intensities for Eu,Ci,, ....................................... 64
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for Eu,.Cl,, ....................................... 65
Transformation of 'Ba,LnCl," hexagonal lattice parameters into cubic
lattice parameters ......................................... 67
Miller indices. observed intensities. and observed and calculated interplanar
d-spacings for 6a.,SmmClu .................................. 68
Miller lndlces. and observed (Guinier) and calculated interplanar d-spacings
and intensities for Ba.La,Br, . ._ ................................ 70
Miller indices. and observed (Guinier) and calculated interplanar
d—spacings and intensities for Ba,Nd,Br,., ......................... 72
Lattice parameters and calculated densities of LnCl. (Ln = Sm. Eu) ..... 74

viii

15.
16.
17.
18.
19.

20.
21.

22.
23.
24.

25.
26.
27.
28.
29.

30.
31.
32.

33.

Miller indices. and observed and calculated interplanar d-spacings and

intensifies for fluorite-type SmCl, ............................... 75
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for fluorite-type EuCI, ............................... 76
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for anti-FezP-type BaCl2 ............................. 79
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for anfi-FezP-type BaBrz ............................. 80
Miller indices. and observed and calculated interplanar d-spacings and
intensities for anti-FezP-type Bai, .............................. 81
Structure types and lattice parameters of BaClz .................... 82
Structure types and lattice parameters of BaBr, .................... 82
Miller indices. and observed (Guinier) and calculated interplanar dspacings

and intensifies for Srlz-iv .................................... 84
Miller indices. and observed (Guinier) and calculated interplanar d-spacings

and intensifies for Srla-lV-type Sml2 ............................. 85
Miller indices. and observed (Guinier) and calculated interplanar d-spacings

and intensities for Srlz-iV-type Eul2 ............................. 86
Comparison of selected crystallographic data of Mi2 and Mia-H20 ....... 87
Crystallographic and Rietveld refinement data for Srlz-iv’ ............. 93
Posifional and thermal parameters for Sria-IV ...................... 94
Comparison of selected Srlz-lv and Srla-i bond distances ............. 94
Miller indices. observed (Guinier) intensifies. and observed and calculated
interplanar d—spacings for triclinic NdBr.,(THF)4 ..................... 98
Comparison of atomic coordinates of Srl,-l-i,O and Srlz-IV ............ 112
Structure-ionic radius rafio relafionships for metastable dihalides ........ 116

Crystallographic data of metastable Bax, (X = Cl. Br) prepared from
various precursors ......................................... 117

Geometrical correlafion of the metastable LnCI, (cubic) with their
precursors Ln,.Cl,, (hexagonal) ................................ 120

35.
36.

37.

38.

39.

40.

Geometrical correlation of the metastable BaBr2 with its precursors

Ba.Ln,Br,, (both hexagonal) .................................. 120
Units of quantities related to electrical conducfivity .................. 141
Typical electrical conductivifies for various materials at room
temperature .............................................. 141
Lattice parameters for selected (anm, 2:4) CaFe204-type

oxides and oxidenitrides ..................................... 160
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for PMm CaFe,O.—type BaCe2(O.N)4 .................... 161
Miller indices. and observed and calculated interplanar d-spacings and
intensifies for BaMo(O.N)3 ................................... 167
Esfimated Mobillfies of Oxidenitrides LaO,N,_, ..................... 177

Figure

10.
11.

12.

13.
14.
15.
16.
17.

LIST OF FIGURES

Page
A 9-coordinated tricapped trigonal prism ......................... 19
Projections of the PbClz-type structure down the c axis .............. 20
Cafion arrangement of various structures illustrafing their derivafions from
hexagonaily close-packed layers ............................... 21
Projecfion of the anfi-FezP-type structure down the c axis ............. 23
Two different layers of the anfi-FezP-type stmcture .................. 24
The fluorite-type structure .................................... 26

Anion arrangement in Ln1 .x” showing fiie formation of a Ln.,x,7 cluster . . . 33

Two adjacent cation layers with Ln,x,, clusters .................... 35
Schemafic illustration of layer stackings of various structure types ....... 36
Apparatuses for the solvoiytic decomposition experiment ............. 54
Standard peak for Srl,-lv .................................... 59
Observed (a). calculated (b). and difference (c) X-ray diflracfion patterns

for Srlz-IV ............................................... 90
Comparison of several metal coordination poiyhedra ................ 91
Gaxis projection of the Srlz-iv structure illustrafing the atomic packing . . . 92
FI'IR spectrum of EuCi3(Tl-lF),, (KBr pressed pellet) ................. 97
Projecfion of Srl,-l-l,O (a) and Srlz-IV (b) structures down the b axis ..... 111

Proposed structural changes during soivolytic decomposition in a single
cafion layer of Ln,.Cl,., ...................................... 114

xi

18.

19.

20.
21.

22.

23.

24.

25.
26.
27.

28.

29.

30.
31.
32.

34.

Latfice parameter reiafionships between the M,.X,, superstructure and the
anfi-FezP-type structure (upper left). and between the superstructure and
the fluorite-type structure (lower right) ........................... 119

The low temperature ordering (if any) of neighboring dipoles and
the consequent behavior of spontaneous magnefizafion and/or

suscepfibility ............................................. 137
Schemafic band structures of solids ...... ' ...................... 140
Rafio of the resisfivity of metallic sodium at T to that at 273 K vs. T ..... 143

An illustrafion of the difference between a normal conductor (a) that allows
magnefic flux penetration and a superconductor (b) that expeis magnetic
flux from its interior ........................................ 145
Simple arrangement for four-probe electrical conducfivity measurements . . 151

The magnetic susceptibility of BaCeLa(O.N). plotted against reciprocal

temperature at 5-300 K ..................................... 159
Magnefic behavior of SrMo(O,N)3 at 5300 K ...................... 171
Magnetic behavior of BaMo(O.N), at 10-300 K ..................... 172
Magnefic suscepfibility of SrMo(O,N), vs. reciprocal temperature ........ 173
Conductivity data of SrMo(O.N), at 86-294 K plotted as In 0 against

reciprocal temperature ...................................... 174
Conductivity data of BaMo(O.N)a at 85-295 K plotted as in 0 against

reciprocal temperature ...................................... 175
Electrical resisfivity of LanNm vs. temperature at 5300 K .......... 178
Electrical resisfivity of LaOmNm vs. temperature at 2300 K .......... 179

Magnefic suscepfibiilty of LanNm vs. temperature at 2-300 K
(H, . 200 gauss) .......................................... 182

Magnefic suscepfibiiity of LaOme vs. temperature at 2300 K
(ii0 - 200 gauss) .......................................... 183

Magnefic suscepfibility of LanNo ,2 plotted against reciprocal

temperature to illustrate the Curie behavior at low temperatures
(l-l,-= 200 gauss) .......................................... 184

xii

35.

36.

Magnefic suscepfibility of LaOme plotted against reciprocal
temperature to illustrate the Curie behavior at low temperatures
(H, a 200 gauss) .......................................... 185

Magnetic field and temperature dependences of the magnetic
susceptibility of LaOme above Tc ............................ 186

xiii

PART I

SYNTHESES, STRUCTURES AND FORMATION MECHANISMS OF
METASTABLE MODIFICATIONS or= METAL DIHALIDES

CHAPTER 1 INTRODUCTION

SYNOPSIS
Some low temperature methods important for the synthesis of solid state
compounds are summarized briefly. The new topochemical method developed in this
work, soivolytic decomposifion on mixed-valence halides Lnxm and mixed alkaline earth-

lanthanide halides. is introduced.

3
1.1 Low-Temperature Routes to Solid State Compounds

Convenfional high temperature ceramic mefiiods are most widely used for the
synfiiesis of solid state compounds. However. due to the well-known diffusion problem,
stoichiometries are hard to control and compounds prepared by these methods are
somefimes inhomogeneous. in addifion. high temperature reactions usually preclude the
synthesis of metastable phases. To overcome these limitafions. low temperature chemical
methods have been developed. These new methods have both enabled fiie synthesis of
known solids with higher purity and homogeneity and resulted in the synthesis of new and
metastable phases. Compounds synthesized at low temperatures ae usually finely
divided with large surface areas essenfial for catalysis and desirable for other applicafions.

Some recent experimental results on this subject have been reviewed by Rao and

Gopalakrishnan in their book (listed in the end as General Reference Books).

1.1.1 Synthesis Based on Topochemlcal Reacfions

Among the various chemical methods. fiiose based on topochemical (topotacfic)
reactions uniquely allow the synthesis of metastable phases that cannot be obtained by
convenfionai methods since low temperature synfiiesis preserves fiie essenfial features
of the parent (precursor) structures. The previously known topochemical reacfions can
be mainly classified into three categories:

a. Topochemlcal Redox Reacfions
Examples include fiie well-known intercalafions of alkali metals into layered
transifion metal oxides and chaicogenides. There are numerous papers and review

arficles on this subject. One example (_1_) is shown here.

XC4H9LI + TiS2 > LixTiS2 + x/2 Can.

 

4

The process involves insertion of lithium into the layer spacings of the sulphide;
in the process a redox reaction occurs. This reaction is topotacfic because the structure
of the product. U,TIS,. is directly related to that of the precursor, 118,. The only major

change is expansion of the interlayer distance.

b. Topochemlcal Ion-Exchange Reactions (2)

Some transition metal oxides with layered or tunnel structures can accommodate
mobile ions between the layers or tunnels. The resulfing compounds are usually ionic
conductors. Due to their mobility, fi'Ie guest ions can be exchanged in aqueous solutions
or/and molten salts. Since the reaction temperature is usually far lower than the structure-
breaking temperature. the basic structure is retained except for minor changes required

to accommodate the different-sized incoming ions.

c. Topotacfic Dehydrafions

Tradifionaily the dehydrafion process has been studied by mermal (DTA and TG)
analysis in a flowing nonreacting gas. The products attracted chemists' attenfion; the
structures of the products did not in recent years. however. occasional reports on the
crystal structures of dehydrafion products have surfaced. Low temperature dehydrations
interest solid state scienfists because in some instances they offer a unique procedure for
preparing metastable solid state compounds. Dehydration of MoO,2H,O to yield
metastable M003 (3). and related compounds (4.5) are known topotacfic reacfions.
By this reacfion metastable phases such as metastable bronzes A,MoO,. where A is an
alkali metal (_6_). metastable forms of AiF3 (BoAiF3) (_7_) and MCI2 (M=Ba, Eu) (8.9).

have been prepared.

5
1.1.2 Synthesis Based on Low Temperature Chemical Decomposifions

a. Solid State Precursor Method

This method takes advantage of the low decomposifion temperatures of precursors
that can be isolated from solufions as homogeneous solids with stolchiometric
compositions. The precursors most frequenfiy used are complexes containing organic
ligands that can be decomposed or removed easily. Since it is not always possible to
obtain precursors of all desired compositions. solid solufions including coprecipitated
hydroxides. oxalates and nitrates have also been used. Sol-gel processing is a technique
based on this method. it has been used for the synthesis of many homogeneous ceramic
oxides as Rae and Gopalakrishnan (see GENERAL REFERENCE BOOKS in the end)
have reviewed. Compounds other than oxides can also be prepared by the precursor
method. The preparafion of anhydrous rare earth halides by the ammonium halide
procedure. first introduced by Taylor and Carter (1_0) and later studied carefully by
Meyer and coworkers (1_1.1_2_,1§). also involves solid state precursors such as
(NH4),LnCI,. it is a simple but effecfive method for the syntheSis of anhydrous halides
both in high yields and in high purifies. Other examples include the cubic form of TaN
prepared by thermal decomposifion of a molecular precursor [Np,TaN],-NH,-2C,H,
isolated from a benzene solufion (w. and single phase alloy Cu..,Ni,, prepared by
fiiermolysis of hetero-poiymetailic molecules (m-O)N.Cu..,(Ni(HZO)).Cl, (N-N. N-
diethylnicofinamide) (15).

b. Soluble Deposifion or Precipitation
Isolation of insoluble compounds from solufions through precipitation is one of the

oldest techniques used to prepare solid state compounds. The products are usually either

6
micro crystalline or amorphous materials. Some recent reports indicate that interesfing

materials that usually require high preparation temperatures can be isolated from
molecular precursors in organic solufions. For example, Ni‘l’e containing Ni(0) was
isolated from a toluene solution of Ni(COD),. Et,P and Et,PTe (COD=cyclooctadiene)
(1e). and FeTe and FeTe, from [Cp(Et,P)(CO)Fe]Te,. (1_7). Polycrystals of
semiconducfing MSe, (M-Hg. Cd. Cu). FeSe,. Tl,Se and CulnSe, were isolated by
warming DMF solufions of the polyseienide complexes [M(Se.),j" (MaCd. Hg). [Fe,Se.,]2',
[Cu,Se,J". and [M,Se,,]’ (M-TI, In) with Se-abstracfing reagents such as CN' and n-Bu3P
(jg). Preparafion of solid LnCi2 (Ln=Sm, Eu) from Tl-iF solufions of LnCl,(THF), by

metal reduction. an interesfing low temperature synthesis. also involves precipitation
(19)-

c. Soivolyfic Decomposition

Soivolyfic decomposifion (leaching) is selecfive dissolution of one component from
a solid with the remaining oomponent(s) retained in the solid state. Use of such reactions
as a solid state preparafion method has not received much attention because of solubility
requirement limitations. in 1963 Clifford (2_o._g_1_) reported decomposifion of mixed
valence oxides by an acid. One such reacfion is shown below:

Pr,2022 (s) + 12 H” (aq) --> 4Pr" (aq) + 6H20 + 8PrOz(S).

in this reacfion the lower valent oxide is soluble and removed from the mixed-
vaient oxide latfice; fiie more covalent higher oxide is insoluble and is left in the solid
state. Since elements like praseodymium do not readily form single high valent oxides

by thennai reacfions. acid leaching provides a convenient and unique preparatory

7
procedure. There have been a few reports of acid leaching on Mn,O. and ofiier mixed-

valent oxides @2324). Some delntercaiafion reactions are also solvolytlc

decompositions. For example. acid leaching of the Chevrei phase Cu,Mo,S, results in

Moose (25)-

8
12 Soivolytlc Decompositions and Metastable Phases

The ianthanide trihalides have long been known to be soluble in tetrahydrofuran
(THF) (gs) while the dichlorides LnCI2 (M=Sm. Eu) are insoluble in this solvent (E)-
It is possible to prepare compounds of the lower valent cafion by selecfiveiy dissolving fiie
trivalent halide from a mixed valent compound. Our original goals were to apply soivolyfic
decomposifions to mixed halide systems as a synthefic method. Since for Ho and Gd only
mixed-valent halides have been synthesized, if the soivolytic decomposifion

lio.,Cl11 + xTHF ----------> HoCl3(THF),. + 4HoCl2 (5)

could occur, it would be a unique way of preparing the yet unknown HoCl2. However. our
results on SmCl2 and EuCl, systems indicated that low temperature leaching forms
metastable phases (fl). Therefore. we concentrated on the synthesis. characterizafion
and forrnafion mechanisms of metastable phases.

Our goals can be summarized by the following two-step reacfions:

Add Y Remove Y
W2 (8) ~~~~~ > MX2°IIY > IIIIx2 (s).
at low temp.
(normal) (precursor) (metastable)

 

The first step is the synthesis of appropriate precursors whose structures are
closely related to those of file metastable forms of MX,. The second step is me
decomposifion that yields the metastable forms. When the added component Y is a
trihalide, Lnx,. the precursor is a mixed or mixed-valent halide and the decomposifion

reacfion is solvoiyfic; when Y is water (or an organic solvent). fi'le precursor is a hydrate

9
(or a soivate) and the decomposifion is dehydrafion (desolvafion). Since the structure of

a final metastable product is closely related to that of the precursor. this procedure is
called template synthesis.

Since a thermodynamically less stable (metastable) phase is the expected product
in a successful soivolytic decomposition. solvent (S) must both dissolve the added
component from the mixed halide selecfively. and form a strong and stable solvate
capable of overcoming lattice energies. The main driving force for the decomposition is

formafion of the soluble solvate:

xMXQ-yLnX3(S) + ymS > xMX2(s) + yLnCla-Sm.

 

CHAPTER 2 BACKGROUND INFORMA‘RON

SYNOPSIS
Important structural features of ianthanide and alkaline earth halides and their
mixed or mixed valent halides are summarized. Crystal structures of the PbClz-type. the
anfi-FezP-type and the fluorite-type for Ax2 and the fluorite-related cluster-type M1 4X33
compounds are described with emphasis on layer stackings. A hexagonal unit cell of he
fluorite-type structure is introduced. The Rietveld refinement procedure is briefly reviewed

and the XRS-82 system set of programs is described.

10

2.1

11
Structural Chemistry of Lanthanide and Alkaline Earth Halides

Brown’s book on “Halides of the Lanthanides and Actinides". though somewhat

outdated. is a comprehensive and important source of lnforrnafion. Numerous papers and

review articles dealing wifi'l the synthesis and structures of file ianthanide halides have

appeared since its publicafion. References are given when specific compounds are

mentioned. Some important points are summarized below.

2.1.1 Characterisfice of Lanthanide and Alkaline Earth Halides

a.

ionic radius of Ln3+ decreases across the Ianthanide series steadily but slightly
(La’*:1.172 ---Lu’*:1.001 A);

The 2+ oxidafion state has been achieved mainly with those elements with
conflgurafions of f7 or f“ or close thereto; those that form divalent halides are: Nd.
Sm, Eu, Dy. Tm. Yb. Among them. Sm, Eu and Yb divalency can be achieved by
hydrogen reducfions; others can only be obtained by metal reducfions. Relative
stabilifies are: Eu“ (1’) > Yb" (r‘) > Sm'“ (r);

Ln2+ ions are similar to the alkaline earth ions with respect to ionic radius, acidity.

and ionicity;
ionic radius (28): Sm” Eu2+ SI” Yb”+ Ca2+
(CNa8) A 1.41 1.39 1.40 128 1.26

Halides 0f fiiese elements. especially the fluorides and chlorides. are typical ionic
compounds. The similarity in ionic radii makes comparative studies of Lnx2 and
MX2 (Maaikaline earth) possible, because we can take advantage of the stability
of file alkaline earth ions.

Both Lnx2 and alkaline earth halides exhibit polymorphism (Tables 1 and 2).

Among them. the PbClz-type is a frequently observed form.

12
9. Both Ln” and alkaline earth M2+ form large numbers of mixed and mixed-valent

halides with Ln". Structural informafion is listed in Tables 3 and 4.

2.1.2 Mixed and Mixed-Valent Halides

The mixed (-valent) fluorides have been reviewed by Greis and i-iaschke (2_9).
More recently the crystal chemistry of Ca,YbF, and related structures have been
examined extensively by Bevan and coworkers (30.31.32). in the past decade a
large number of mixed and mixed-valent chlorides and bromides have been prepared
(31,31,35.16_,3_7_.3_8_.39_,fl,4_1,4_2_,4_§). Known mixed halides can

be classified into two series:

a. The vemier-type compounds that are M”-rich and have the general fonnuia Mnxm,
where n-4. 5. 6. 7. 11 are known for chlorides and bromides.

b. The Ln..X,7 cluster-type compounds have the general formula M.Ln,X,., where M
is a divalent alkaline earth or ianthanide ion and Ln is a trivalent ianthanide ion (35.44;).
A common feature of the mixed halides is a fluorite-related superstructure discussed in
2.2.4.

13

 

 

 

Table 1. Structural modificafions of ianthanide halides Lnxz.
Ln\ X F Cl Br i
Nd PbClz (9) 'PbCl. (9) SrBr2 (8)
Sm 0an (8) PbCl2 (9) Pb012 (9) PbCI2 (9. P)
Can (8. H)‘ SrBr2 (8) Eulz (7)
Srlz-lv (7. M)‘
Eu CaF2 (8) PbCl2 (9) PbCI2 (9,P) PbCl2 (9. P)
PbCi2 (9, P) 0an (8, H)‘ SrBr2 (8) Eul2 (7)
Srl2 (7) Srl2 (7)
Srlz-IV (7, M)‘
Dy SrBr2 (8) Srl2 (7) CdCl2 (6)
Tm Can (8) Srl2 (7) Sri2 (7)
CaCi. (6) Cdl. (6)
0t-PbO2 (6)
1102 (6, H)
Yb 0an (8) Srl2 (7) Srl2 (7)
CaCl2 (6) Cdi. (6)
or-PbO2 (6)
1102 (6. H)
Note. ( ): cation coordination number;

P: high-pressure form;

H: high-temperature form;
M: metastable form;

': prepared in this work.

14

 

 

Table 2. Structural modifications of alkaline earth halides MX2.
M \ x F Cl Br I
Ba. Pool, (9. P) PbCl, (9) Pool. (9) Pool2 (9)
Can anfi-FezP anti-FezP anti-FezP
(9. M)‘ (9. M)‘ (9. P)‘
Can (8. H)
Sr PbCl2 (9. P) PbCl2 (9. P) PbCl2 (9. P) PbCl2 (9. P)
0an (8) 0an (8) SrBr2 (8) Srlz-l (7)
Srlz-lv (7, M)‘
Ca PbCl2 (9. P) Srl, (7. P) CaCI2 (6)
CaF2 (8) CaCl, (6) or-PbO2 (6. M) Cdl. (6)
or-PbO2 (6. M) TiO, (6. H)

 

Note. ( ): cation coordinafion number;
P: high-pressure form;
l-l: high-temperature form;
M: metastable form;
': prepared in his work.

15

Table 3. Space groups and iatfice parameters of mixed-valent ianthanide halides

 

 

 

 

 

 

 

 

 

 

(excluding fluorides).

II Formula S. G. a(A) b(A) c(A) angle (°) Reference

Vernier-type: Lnnxm,

4 Nd.6r. P 2m 7.741(2) 3016(1) 7.125(1) y=91.80(5) 3_6,g
50.01. P 27o 7.2336(6) 26.299(3) 6.7167(6) y-91.775(5) 35
sin,Bru P 2./m 7.652(2) 3721(2) 7.121(2) 8:90.266) 3.94
50.0., P 2,/m 7214(2) 3517(1) 6.775(2) 3.9034(1) _§

5 Dy,Cl,, P am 7.108(1) 3466(2) 6.632(1) 6.3023(4) a;
l-lo,c1., P 2,/m 7.078(2) 3457(2) 6.603(3) 3-9019(5) 49
Cf.GdCl., P 2,/m 7.130 34.83 6.665 3:90.24 43
Cf.GdBr., P 27m 7.619 36.97 7.040 8:90.18 43
Sm.Br,, I 2/a " 7.649(2) 4444(2) 7.139(2) y=91.30(5) 93.94
Dy.,CI.a l 2/a " 7.099 41 .41 (2) 6.667(1) 1:91 35(3) 33

6 Tm.Cl., 12/a " 7.009(3) 4104(1) 6.557(3) 7.3130(3) 33
men, i2/a " 6.956(11) 4096(6) 6.539(11) 7.912(2) 41,;
march, I 2/a ° 7.004(5) 4100(2) 6.537(5) y=91.2(1) 44,42

7 by,cl,, P n m 6 7.097(1) 4616(1) 6.674(1) 33
Tm,Ci,, P n m 6 7.001(1) 4766(1) 6.571(3) 33

6 1'm.0i,7 ? 6.995(2) 5426(2) 6.560(1) 33

11 sm,.6r,. P 2,/n 7.652(2) 6162(6) 7.130(3) 3.9019(7) gag;

Cluster-type: Ln..x,. with Ln,X,7 clusters
Sm..01,. R3 12.664(2) 24.72(8) this work”
Sm.Nd,CI,. R3 12.694(2) 24.650(6) this work“

14 Sm,Gd,Cl,. n3 12.645(2) 24.694(6) this work'”
Eu..CI,. R3 12.815(4) 24.768(8) 35
Nd,.Cl,, R3 12.967(1) 24.622(3) 44

 

i) Non-conventional setting of space group C 2/c (# 15):
Standard setfing: 5,. g... 9.; transformafion: a .. 9,. b - a, + 9,. g . -_b,.

ii) Directly indexed c values were half those reported here.

16

Table 4. Space groups and lattice parameters of alkaline earth-related mixed halides
(excluding fluorides).

 

n Formula S. G. a(A) b(A) c(A) angle (°) Reference
Vemier-type: M“,,LnX,,,,,

 

 

 

 

 

 

 

Sr.DyCl., P n m a 7210(1) 3516(1) 6.766(1) 33
5 Sr.Nd6r,, P n m 6 7.662(1) 3735(2) 7.140(3) 33
snsuci, P n m a 7.220(2) 3515(1) 6.790(4) g
Sr.Nd0l,. P n m a 7230(5) 3529(2) 6.626(4) g
6 Sr,NdBr,, l2/a 'i 7.642 4462(2) 7.177(6) F91 .1 1 (20) 33
Cluster-type: M.Ln,X3, with MLn.,X,7 clusters
$550.01.. R3 12.654(4) 24.702(8) 31
14 Sr,Nd.CI,. R3 12.908(6) 24.623(10) 33
6a.l.a,6r,, R '3 14.096(2) 26.678(6) this work
6a,Nd,6r,. n '5 14.039(1) 26.477(4) this work
Unknown structures
'Ba,LaCl," 7 17.637(6) 12.495(5) #120 this work
ea1 ,Sm,.01.. P a 3 21.366(2)" this work

 

i) Non-conventional setting of space group C 2/c (# 15): .
Standard setfing: 3,. 3,. 3,; transformafion: 3 = -_t_:,. 3 . 3, + 3,, 3 a :30.

ii) Single crystal parameter.

17
2.2 Crystal Structures of Related Halides

A crystal structure an be described a number of ways. each of which emphasizes
different perspectives. Of course. the most accurate. but the least descripfive
representation of a crystal structure is a set of lattice parameters. a space group and the
atomic coordinates. Descripfions in terms of close packing of atoms and space filling
highlight the geometric relafionshlps and'the extended nature of crystals while those
based on coordinafion polyhedra that share edges. verfices. 0r faces pay more attention
to the local structures and fits chemical nature of central atoms. Even though it is
possible to describe a structure in both terms. for an ionic inorganic solid. close packing
is more convenient; while for a covalent solid. coordination polyhedra are more
meaningful.

Crystal structures of some halides of importance to this work have been described.
For the convenience of later discussions on reacfion mechanisms, selected structures are
summarized briefly. Symmetries of the close packings and correlations with hexagonal
close packing layers are emphasized.

22.1 The PbCL-Type Structure

The normal forms of LnCl2 (Ln-Nd. Sm. Eu) and Bax2 (X=Cl. Br. i). among many
other AB,-type compounds. exhibit this structure. It has been described by Wyckoff (566
GENERAL REFERENCE BOOKS). The EuCI2 structure was refined by BAmighausen

(35).

Space group: anm (# 62); orfiiorhombic.
a . 8.965(2). b . 7.536(1), c . 4.511(1) A; z=4.

All atoms are in me 4c special positions:

16
:l: (u, v, 1/4) 1(1/2-0, v+1/2, 1/4)

with u(Eu) .-. 0.11510(3). v = 0.24952(3). B(Eu) = 126(1); 0(01.) = 0.4279(1), v(Cl,) =
0.1434(1), B(Cl,) .. 132(3); and u(Cl,) = 0.1668(1), v(Cl,) = 0.0238(2), B(Cl,) = 1.56(4).

The cation coordinafion number is hard to define. The coordinafion polyhedron can
be described approximately as irregular 9—coordinated tricapped trigonal prism as shown
in Figure 1. The three capping Eu-Cl bond lengths are: 2.916(1), 2.994(1) and 3.046(1)A;
the remaining six Eu-Cl bond Iengfi'ls are: two 2.925(1). two 3.090(1) and two 3.440(1)A.
it has a mirror plane of symmetry (defined by the capping and file central atoms in Figure
1).

This structure can also be thought of as a considerably distorted close-packing of
the halogen atoms with the metal atoms accommodated in the same plane. The layers
are stacked in an ABABm fashion. Figure 2(a) illustrates a single layer and Figure 2(b)
is the projecfion of file structure down the c axis. showing the double layer feature. The
two layers have idenfical packings and stack such that inversion centers are formed
between filem. .

The arrangement of the metal atoms in this structure can be derived from a single
hexagonally close-packed layer as depicted in Figure 3(a). Exactly half of the cafions are
displaced from the layer by 012 such fiiat the 6-fold axes are destroyed while zigzag metal-
metal chains are formed in each layer and inversion centers are created between any two

adjacent layers.

19

 

 

 

 

 

 

Figure 1. A 9-coordinated tricapped trigonal prism.

 

Figure 2.

 

(a)

(b)

Projecfions of the PbClz-iype structure down the c axis. Small circles
represent cafions and large circles represent anions. (a) A single layer
showing the roughly close-packed anions and the accommodated cafions;
(b) a double layer projecfion with one 9-coordinated polyhedron oufiined
(Open circles: z . 0.25; shaded circles: z . 10.75).

21

.65 .5852 8.9.6 3.8 Em 65 use .65 282%. 8.2.6 coco .3 cam 66.2.6

9.8 e: e... .6. c. .6553... sx..5 as sage 2.. .e as. sea... a .6. use asses 82-252. 2.. .6. 6553»
is. 8.2.6 58 Be 3.3.. 298 8...-dseu...ce 9.. a. 6.36.5... 5...... .225 58 use 2.8. 6.88 6.55.9...
9: Am. .228. 8263-820 2.26596: E9. 22.9.5.6 .85 3.35:... 8.58.56 macros .o EoancmEm 8.80 .n 2:9“.

0
2-----.

O O O O 0 3V

22
2.2.2 The Anti-FeaP-Type Structure

The normal Fe,P structure was described by Wyckoff. Anti-FezP-type Bal, was
prepared under high pressures and the structure was refined by Beck using single crystal
data (1Q).

Space group: p3 2 m (# 189). hexagonal.

a - 9.142(6), c =- 5.173(3) A; z a 3.

Atomic positions:
ea(1): (1b) (0. 0, 12)
83(2): (20) (1/3, 2/3. 0)
I(1): (3f) (0.2563(2). O. O)
l(2): (39) (0.5918(2). o. 112).

In this anti-Fe2P-type structure the metal atoms are 9-coordinated in a way similar
to those in the PbCl,-type structure shown in Figure 1. However, there are two major
differences. First. the cation coordination polyhedra of the anti-FezP-type have 03,,
symmetry while those of the PbClz-type are less symmetric (0,); secondly. there are two
types of cation coordination polyhedra in the anti-FezP-type structure while there is only
one type in the PbCl,-type. The apical and equatorial (capping) Ba-I bond distances of
83(1) differ from those of Ba(2). Ba(1) has six shorter apical (3.491(1) A) and three longer
equatorial (3.734(2) A) bonds, while Ba(2) has six longer apical (3.791(1)) and three
shorter equatorial (3.456(1) A) bonds.

The atoms in the anti-Fe2P-type structure are slightly more densely packed (0.8%
higher packing efficiency for Bal2 (fin than those in the PbCIz-type structure. Therefore,
there is a slight increase in effective coordination numbers in the former structure.

The anti-FezP-type structure exhibits close packing similar to that of the PbClz-type.

23
It is a double layer structure (Figure 4). but unlike the PbClz-type structure the layer

packings are not identical (Figure 5). Therefore a center of symmetry is absent.

The cation positions can also be derived from hexagonal close-packing as shown
in Figure 3(b). Displacing 1/3 of the metal atoms from a hexagonal close-packed layer
by c/2 in such a way that 6-fold inversion axes are retained results in the cation

arrangement of the antl-Fe,P-type structure.

 

Figure 4. Projection of the anti-FezP-type structure down the caxis. Open circles are
at z - $0.5 and semi-shaded circles are at Z - O and 1; small circles
represent cations and large circles represent anions. Two types of 9-
coordinated polyhedra are depicted (----: equatorial bonds).

24

0 PG ---- I O
O O O f5 O I 0 O
O . O ’l . O I, . O .
O 3.9.--- O O (a)

O O

0 0000
O. O. O. 00
00000000

Figure 5. Two different layers of the anti-FezP-type structure. Solid circles represent
cations and open circles anions. (a) z a 0; (b) 2 = 0.5.

25
22.3 The Fluorite (Cam-Type Structure

This is one of the most commonly found structure types for AB,-type compounds.
Among them are the alkaline earth and divalent rare earth fluorides, SrCl,, and the high
temperature forms of MCI, (Ni-Ba. Sm. Eu).

Space group: F m 3 m (It 225). lace-centered cubic.

Unit cell dimension for SrCl,: a - 6.9767 A. z s 4.

Atomic positions:
Sr: (4a) (0, 0, 0)
Cl: (8c) 3: (1/4, 1/4, 1/4).

Since all the atoms occupy special positions. the coordination polyhedra of the
metal atoms are ideal cubic (CN=8. symmetry O..). and those of the anions are ideal
tetrahedra (CN=4, symmetry T,) (Figure 6(a)). The cubes are linked together by sharing
faces. From a cubic coordination polyhedron the ionic radius ratio for the fluorite-type
structure can be evaluated. Assume the cation (A) and anions (B's) are in close contact.
then the body diagonal is 2(rA + r,). The cubic edge must be 2 2's for a compound A82
to adapt this coordination. Therefore 2(rA + r.) 2 13(2r.); that is, an, 2 0.732.

This structure ls commonly described in terms of cation close-packings. When we
view down a three-fold inversion axis (the body diagonal) of a cubic cell. the cations are
arranged in hexagonally close-packed layers normal to the three-fold axis, with the layers
stacked in the ABCABC... sequence (Le face centered cubic stacking). The anions
occupy all the tetrahedral holes to form separate hexagonal close-packed layers parallel
to the cation layers. (Here the term “layer“ is used only for description purposes and does
not imply a layer structure in which inter-layer interactions are weak.)

Many mixed halides have fluorite-related rhombohedral superstructures as will be

described later. To facilitate comparisons between the fluorite cell and the hexagonal

26

 

 

 

 

(a)

all

 

 

 

 

 

(b)

Figure 6. The fluorite-type structure. (a) the cubic unit cell, and the cation (0,.) and
anion (Ta) coordination polyhedra; (b) the choice of the hexagonal unit cell.

27

supercell. we introduce a “hexagonal unit cell“ of the fluorite-type structure. This unit cell

is more effective for describing the fluorite structure layer stackings.

a. Choice of the Hexagonal Unit Cell

The cubic structure can be viewed as a special case of rhombohedral structures
with on - 90°. Therefore, it can be transformed into a hexagonal cell without any loss of
structural information.

The hexagonal cell is chosen such that the 0,, axis coincides with the body
diagonal of the cubic cell, and A“ and 8,, are then the face diagonals as shown in Figure

6(b). Unit cell parameters are transformed according to the following equations:

Ah’ Ao'er.»
a..=-A.+B..
Ch: Afi+B¢+Cc

A. 1 o -1 Ac
°' 8,, 2 -1 1 o B,
on 1 1 1 cc

1 0 -1
The transformation matrix is s. -1 1 0

111

28
Since the determinant of S is |S| :- 3. the volume relationship is

v,, s ave.
From Ac = Be = C6 and the ortho-normallty of these cubic lattice parameters we have
A“ - 3,, - 42A,, and ch = (3A..
It can be shown that the newly chosen unit cell is indeed hexagonal.
Alxa. - (MOM-Ac + a.) = -A.xA. + C.xA. + MB. - chB.
=0+B.+C.-i-A.i=C.
therefore both A.1 and 3,, are orthogonal to C...
M31. = -A.-A. + 0.4. + NB: - Gas.
=-A§+0+0+0=-Ac"’.
Note A.,-Bh - AHBBCOSy - ZAZCOSy. Thus 0081 a -1/2. and y a 120°.

b. Atomic Coordinate Transformation

The inverse matrix of S is

1I3 -1I3 1/3
T: S" = 1!!! 2I3 1I3
-2I3 413 1I3

The coordinate transformation matrix is the transposed inverse matrix of S, or “l”

"3 Va -213
T'= -1I3 213 -1I3
1/3 113 1/3

The space group F m 3 m has 192 general positions. Because the cations occupy
the 4(a) and anions occupy the 8(c) special positions in the cubic cell, we can avoid
unnecessary transformations by defining the following 8 positions as the general position

set (subscript c for cubic) of the cubic unit cell:

29
:l: (xc. Ye, z.) :1; (xc. 1/2+Y.. 024.2.)
s (1/2+x., Ye, 1/2+z.) i (1/2+x., mm, 2,).

ForthecatlonsXc-Y¢=Z€-0andfortheanionsxchcszcs1/4.

By applying the coordinate transformation matrix (subscript h for hexagonal) we

have
X. 113 1/3 -213 Xc
Y. =- -1I3 213 413 V6
2,, 113 113 113 z,
or

x...1/3(x. + Y. -2z.)
i/.,..1/3(-xc +2Y. - z.)

Z,,-:1/C'1l(Xc + Ye + 2:)

The corners of the cubic unit cell have the following coordinates:
(0.0.0; 1.0.0; 0.1.0; 0.0.1; 1.1.1; 0.1.1; 1.0.1; 1.1.0). These
coordinates can be transformed into hexagonal coordinates according to the above
equations:
(0.0.0; 16.-16.16; 16.86.16; 16.-15.16; 0.0.1; 46.16.99; film-96.86; ”bl/33A).

Of the eight hexagonal coordinates. only three. (0.0.0). (16.86.14). and (93.16.99) lie within
the hexagonal unit cell and correspond to the coordinates of Bravais-lattice points.

8 general positions of the hexagonal unit cell can be generated by applying the
same transformations to the previously defined 8 general positions of the cubic unit cell.

When the 8 hexagonal general positions are combined with the three Bravais-Iattice point

30
translations. we finally obtain the 24 general positions of the hexagonal unit cell:

(0. O. 0; ‘6, 96. Va; 85, Va. 96) +

:t (X... Y... Z.) :l: (-1/6+X,., 1/6+Y... 1/3+Z..)
:l: (-1/6+X,., -1/3+Y,.. 1/3+Z..) :l: (1/3+X... 1/6-l-Y... 1/3+Z..).

The atomic coordinates of SrCl2 in both unit cells are listed in Table 5.

In the hexagonal cell it is apparent that the fluorite-type structure is made up of
three cation layers at 2:0. 1/3. 2/3 and six anion layers at Z=1/12 (1. 3. 5. 7. 9. 11).
Projections down the c axis indicate that every cation or anion layer is hexagonally close-

packed. A single layer is shown in Figure 3(c) (Page 21).

Table 5. Atomic positions of the fluorite-type SrCl2 in cubic and hexagonal unit cells.

 

Atom (x.. Y... z.) Multiplicity (x.. Y... z.) Multiplicity

Sr: (00 c) 4 (co 0) 12
Cl: (1/41/4 1/4) a (o c 1/4) 24

 

 

31
2.2.4 Fluorite-Related Cluster Type Superstructures

Mixed and mixed-valent halides of alkaline earth and rare earth halides with
fluorite-related structures are compiled in Tables 3 (Page 15) and 4 (Page 16). The
cluster-type M.Lnsx,., compounds are of special interest to us because many precursors
used in this work have this structure. The structure of the oxygen contaminated Nd..Cl320
was solved by single crystal work (_3§) and later the structure of Nd1 .Claa was solved from
triply-twinned crystals (44). The real symmetry of the triply-twinned crystals is triclinic. but
it can be approximated as rhombohedral.

Space group: 35 (#148). rhombohedral.

Hexagonal lattice parameters for much”: a = 12.937(1), c = 24.822(3) A; 22.

Unit cell contents: 27 Lnx2 and 15 Lnxa; or 3 formula units containing 3 M..X37
(Ln.’*Ln,’*CI3,) clusters.

List of atomic positions: Table 6.

Apparently. clustering of the trivalent cations In the superstructure is due to the
“extra“ halide anions brought in by Ln“. Each Ln" introduces one “extra" X’ when
compared to LnX,. thus there are 15 'extra' X's in a unit cell. Eitel (g). and Bevan and
coworkers (_3_Q,4_7) have described the formation of clusters from the fluorite coordination
polyhedra. Figure 7 illustrates the changes anions undergo when a Ln."’*Ln.;"*x37 cluster
is derived from six face-sharing cubes. The cluster comprises six square anti-prisms of
anions whose centers are occupied by cations. corresponding to 6(Lnx.) + 12X = Ln.x3..
The center of the cuboctahedron formed by the 12X anions is occupied by another X. thus
making it Ln.x,,. In the fluorite structure. however. six cubes of anions only represent
Ln.X,,. 5 anions less than that in a cluster group. Therefore. each unit cell contains 15/5

= 3 clusters.

it

 

 

Table 6. List of atomic positions for Nd..Cl,.,.

Atom Site X Y Z

Nd(1 )' 6(c) 0 0 0.75614(2)
Nd(1 ’) 6(c) 0 0 . 0.72865(1 0)
Nd(2) 1 8(f) 0.09902(2) 0.44864(2) 0.57930(1 )
Nd(3)‘ 18(f) 0.26385(24) 0.21241 (26) . 0.58792(12)
Nd(3’) 18(f) 024689(37) 0.19776(40) 0.58061 (18)
Cl(1) 6(c) 0 0 0.87855(6)
Cl(2) 18(f) 0.314520) 0.436300) 0.62907(3)
Cl(3) 18(f) 0.48795(6) 0.39243(6) 0.54437(3)
Cl(4) 18(f) 0.23341 (7) 0.18049(7) 0.70356(3)
Cl(5) 18(f) 021869(8) 0.33109(9) 0.49998(3)
Cl(6) 18(f) 0.03893(8) 0.18417(9) 0.61225(5)
Cl(7)“ 18(f) 0.0495(16) 0.0583(11) 0.4936(6)

 

Partial occupancles: 0.812(2) for Nd(1) and 0.188(2) for Nd(1'); 0.591 (5) for Nd(3)
and 0.409(5) for Nd(3');

Eitel M did not indicate partial occupancy. This position is very close to the
special 3(b) site. To be consistent with the formula the occupancy factor should

be 0.1667.

Figure 7.

33

(a) Ln,

 

(G) M7

Cluster

Anion arrangement in Ln. 4X.” showing the formation of a Ln.x,, cluster. (a)
6 Lnx. cubes sharing faces in the fluorite structure; (b) rotation of two
opposite shared faces forms two square anti-prisms; (c) addition of four
“equatorial“ anions and a central anion (solid circles) yields a Lri.,X37 cluster
with an anion cuboctahedron.

34

Projections down the 3-fold axis can better illustrate the fluorite-related features of
this superstructure. A single cation layer of Ln.4Cl33 in shown in Figure 3(d) (Page 21).
The hexagonally close packed character is apparent despite minor positional shifts in and
out of the layer plane. Due to clustering trivalent cations are ordered along 3-fold axes
in each layer. Formation of clusters occurs to two adjacent cation layers as shown in
Figure 8. The rhombohedral supercell corresponds to 6-cation layers compared to 3-
cation layers in the fluorite structure.

The layer stackings of the cluster-type structure. together with those of the other
structure types described above. are illustrated in Figure 9. it should be noted that the
cations and anions in Figure 9(a) are not exactly at the designated 2 positions and at the
same planes as shown. Displacements of 0.0205 A from the ideal 2 positions are
calculated. it is interesting to note that the “extra“ anions in this superstructure are the

main differences between the fluorite- and the cluster-type structures.

35

 

Q Ln” at z=0.083

. Ln’+ at z=0.083
O Ln3+ at 2:025
0 Ln2+ at 220.25

Figure 8. Two adjacent cation layers with Ln..X37 clusters.

36

.<ooo.~ .3 as... .<8m.~. .o. . :3 3. as .68 .2 6:668 use. .<-.< 2.. as 8.5-8.5 9:. so: a. .3
35353628 on .p. 26 a. .3 383% e. .ch 2.55.0. o... :5 deacon. 26 ..< use .< .o 6568 2: .se a.
8.6528". N .8233 .3 ”226. .< s .2. =2 n. was. .< c. 22.8 .6 .352 o... 65 2oz a: a. esnseoce
N .oabdsomécm 8. ”Sup 5 «22.208 N .958 .2396; c. 33.2.52. 3. ”Sew c. 8.2.903 N .25
223.0 sx:5 a. .228. 656m 65 c. 98.5 new «condo ”+-+-+- Lo -+.+-+ .--+ :26. 5...? ..... no.8. condo u++++
692.208 N 8823.. 228. 2 Escape 238:2 .82: 2302.» 30.8., .0 353.066 8.8. .o 5.85:... 0.5.89.3 .m 659“.

.3. .0. 3. .3

m +1+l+1+1+1+1 ad H 11+ 11+ 11+ .d. an 1111111111111 a 11 11111111111
o +++++++++++++ o « +++++++++++++ o

H I+I+I+I+I+I+ :‘ o |*|+I+I+I+I* 2‘ H IIIIIIIIIIIII n IIIIIIIIIIIII

‘ IIIIIIIIIIIII

m +1+1+1+1+1+1 ad a 11+ 11+ 11+ ad n 1111111111111 m 1111111111111
v +++++++++++++ a o +++++++++++++ n

a 1+1+1+1+1+1+ ad o 1+1+1+1+1+1+ ad n 1111111111111 r 1111111111111

m +1+1+1+1+1+1 .d. a 11+ 11+ 11+ .d. p 111111 l 111111 a 1111111111111
o +++++++++++++ d. on +++++++++++++ d

H 1+1+1+1+1+1+ :d. o 1+1+1+1+1+1+ ad a 1111111111111 an 11 11111111111

2 1111111111111

m +1+1+J+1+1+1 .d. a 11+ 11+ 11+ ad ”A 1111111111111 ma 11111 1 1111111
«a +++++++++++++ 0 ca +++++++++++++ u

H 1+1+1+1+1+1+ ad a 1+1+1+1+1+1+ ad a 1111111111111 mm 11 11111111111

m +1+1+1+1+1+1 .d. H 11+ 11+ 11+ ad n 1111111111111 he 11111 1 111111 1
v +++++++++++++ a mu +++++++++++++ m

a 1+1+1+1+1+1+ ad o 1+1+1+1+1+1+ ad m 1111111111111 ma 1111111111111

on 1111111111111

m +1+1+1+1+1+1 ad a 11+ 11+ 11+ ad h 1111111111111 as 111111 1111111
a +++++++++++++ d. «« +++++++++++++ d

a 1+1+l+1+1+1+ ad a 1+l+1+1+1+1+ ad a 111 1111111111 nu 1111111111111

m +1+1+1+1+1+1 .d. a 11+ 11+ 11+ ad AH 1111111111111 H 1111111111111

37
2.3 The X-ray Rietveld Refinement Procedure

2.3.1 Introduction

Traditionally. X-ray powder data had been used for many purposes other than
structure solutions because the overlapping reflections limit information content. However.
there does exist a large number of compounds such as metastable phases obtained by
low temperature dehydrations for which it is difficult. if not impossible. to grow single
crystals. Introduction of the whole pattern refinement procedure for neutron powder data
(48) by Rietveld in 1967 and his detailed description of a computer program (5L9) in
1969 marked a significant advance in powder crystallography. In this method it is
assumed that a powder pattern is the sum of a number of Gaussian-shaped Bragg
reflections centered at their respective Bragg angle positions. The intensity counts
actually observed are then used as the least-squares data points. In fact. neutron powder
diffractometers give peaks that are accurately described by a Gaussian function; thus the
information in neutron powder patterns are best used in Rietveld refinements. Since then
numerous compounds have been studied by this technique. Cheetham and Taylor had
reviewed this subject (Q). A more recent review by Albinati and Willis (51) also
included the practice of powder x-ray pattern refinement.

The success of the Rietveld method with powder neutron data has stimulated a
great interest in applying this technique to powder X-ray data because of the easy
availability and relatively low cost of X-ray facilities. The main difficulty in adapting the
method for X-rays is determining a function to describe the peak shape. Generally
applicable and accurate peak shape functions are not available for X-ray patterns.
Furthermore. X—ray data usually contain less information than neutron patterns because

of the angular dependence of X-ray scattering factors.

38
Nonetheless. great progress has been made during the past decade to make it

possible to acquire accurate X-ray intensities from powder data. Various ways of
extracting intensity lnfon'nation from powder patterns have been reported. These include
intensity data from microdensitometer measurements of Guinier-Hagg films
(52.53.355.55). and of Debye-Scherrer films (51)‘ but more accurate X-ray
pattern intensities are obtained by the currently most widely used automated
diffractometers (55.59.59).

The Rietveld method is essentially a refinement method. It requires a good model
structure. Several methods leading to solutions of unknown structures have been
introduced. such as model building (51,52). structural isomorphs @. isostructural
compounds (55.59). new structures (55). and topologically-related structures (55).
Probably the most exciting progress for powder X-ray crystallography is the combination
of the Rietveld refinement technique and the standard Patterson and Fourier techniques
($.55) or the Direct and Fourier methods used for single crystal structure solutions
(_61). Such combinations make ab initio structure solutions possible and have been
used successfully in recent years for some systems. In these methods integrated
intensities of unambiguously indexed peaks in the pattern are used. Partially overlapping
reflections are deconvoluted by employing an appropriate peak shape function @-

2.3.2 The Rletveld Refinement XRS-82 System

This system Is the 1982 version and was edited by Ch. Barlocher. A. Hepp and L
B. McCusker at Institut fur Kristallographie und Petrographie. ET H. Zurich. Switzerland.
It was used in this work for refinement of the metastable structure of Srl, (Srlz-IV) and is

described here.

39
a. Peak Shape Function

As pointed out earlier. the principal problem one encounters In applying the
Rietveld method to X-ray data Is the difficulty in describing the peak shape. As many as
seven peak shape functions have been provided in a recent version of the program
DBW3.ZS. The various profile shape functions have been reviewed by Young and Wiles
(55). In the XRS-82 system. however. no analytic function is used. Instead. a single
non-overlapping observed peak (standard peak) is selected and analyzed numerically to
calculate a peak-profile function ('leamed' peak-shape function). The intensity distribution
function for an individual peak derived by Hepp and Barlocher (19) has the form

0= 1.76.1011 416.011

where H is the half width at half height and A the asymmetry parameter. The G.(r) and
G..(r) are normalized functions representing the symmetric and asymmetric parts of the
'Ieamed' peak-shape function. r = A29/I-l = (26 - 26,)IH (B. is the Bragg angle'). The
lower and upper 29 cutoffs for the standard peak are supplied by the user according to
the observed data. These parameters are evaluated from this peak and then the peak
shape function generated. This procedure gives an accurate description of the observed
peak shape. The dependence of H and A on 29 is then determined from the entire
powder pattern. The XRS-82 system allows more than one peak function.

b. Peak Profile Parameters
In the XRS-82 system. these parameters are refined according to the following
equations:
W0”, - WD,“{(WD1) + (W02)—(29)]

40

PA” =- PA,..1[(PA1) + (PA2)-(26)]
where (W01). (W02). (PAI). (PA2) are variables and WD... and PA... are the estimated
peak width (full width at half maximum or FWHM) and peak asymmetry from the standard
peak. respectively. Note PA :- A. and WD s 2H in the Hepp-Barlocher equation.

c. Least Squares Parameters

The parameters in Rietveld refinement are usually divided into two groups: (1) the
profile parameters that define the lattice dimensions. peak half-widths and asymmetries.
and preferred orientation; (2) the structure parameters that define the contents of the

asymmetric unit cell. In the XRS-82 system. the following parameters are included:

Profile Parameters
Cell parameters: A. 8.0.01. 13.-r.
PAI . PA2: peak asymmetry as defined previously;
WDl, WDZ: peak half width at full maxima as defined previously;
201, zcz: ze zero correction. 29,... - 29.... + (201) + (zc2)(2e);
POF: preferred orientation factor according to I“... a. lo... exp[-(POF)cos2a] where

o is the angle between the scattering vector h k l and the preferred
orientation vector (U V W) (normal of plate-like crystallites).

Structure Parameters
SCL: scale factor (c) defined such that chFch;
UOV: overall isotropic temperature factor defined as the temperature factor
correction T.-exp[-81r’(UOV)(sin9/A)’];

x. Y. Z: fractional coordinates of an atom in the asymmetric unit cell;

41

PP: population parameter. the fractional occupancy of an atomic site.
Independent of site multiplicity;

U: individual isotropic temperature factor. similar to UOV;

U... U... U”. U.,. U.,. U23: individual anisotropic temperature factors defined as
Tmaexp[-(h’B.. + 183.2 + 1’13... + 2hk8,21+ 2hlll.3 + 2mm”. where
8.:23’U.a.'a.'. and a' is a reciprocal axis length.

It is possible to use mixed temperature factors in the XRS-82 system.

d. The Least Squares Refinement

The quantity minimized in Rietveld refinement is:

M = 2 while) - Viewer

where Y.(o) is the observed background-corrected count rate at the ith step of 29. and
Y.(c) the calculated count rate including contributions from all Bragg reflections that
overlap at the ith step. c is the scale factor. and wl the weight. The weight to be applied
during refinement may be specified in one of four different ways as a user option:

1) unit weight (all weights as 1.0);

2) weights are accepted as supplied in the binary profile data file;

3) weights as supplied in the binary profile file are squared;

4) weights are set equal to the reciprocal of o(Y.._) and used as Y... weights.

The program CRYLSP performs the least squares refinement. It is specially
structured to accommodate a whole pattern refinement using the numerical peak-shape
function of a powder pattern as mentioned above. It also allows the user to employ
known interatomic distances and angles as soft restrictions. Two least squares

procedures are available: the Gaussian Algorithm and the Variable Metric Algorithm. Dr.

42
Wieslaw Lasocha stated that his experience indicated that the Gaussian algorithm is more

stable than the latter. The original authors suggest the variable metric algorithm is less

sensitive toward high correlations.

e. Goodness of Fit

The whole pattern plot that Includes the observed. calculated and difference
patterns is a very informative way of describing the goodness of fit. These plots can
reveal discrepancies due to systematic errors such as the systematically lower counts at
low angle region in our observed pattern (Figure 12. Page 90). or due to poor peak shape
functions or impurity peaks. The difference-Fourier map which can be generated by
programs FOURR and PEKPIK is another good way of checking goodness of fit. This
map should be relatively flat since random errors typically produce small fluctuations. In
addition. bond lengths and angles that can be calculated by program BONDLA in the
system can provide more information about the correctness of models and the refined
structures.

In the XRS-82 system. four R-values are used to give quantitative descriptions of
the refinement. They are discussed briefly here.

Structure Factor R,

F‘1 ‘ 215(0) " F110“ / 25(0)
where F.(o) and F.(c) are the observed and scaled calculated structure factors for the ith
reflection. respectively. This R. corresponds to the conventional reliability factor commme
used in single crystal refinements. It measures the agreement between the observed and
calculated structure factors.

Integrated Intensity R or Bragg Intensity R,

43
R. - Elll. - Ll) I 2 I. = 2|(m1'Fi(0)2-F1(CI"’)|/ mirror
where lri| is the lth reflection multiplicity and l the integrated intensity. Note lam-F2.

In a Rietveld refinement both R. and R. (or R.) are biased towards the model

because F(o) and I, are evaluated in proportion to the calculated structure factors.

Profile (or Pattern) R,

Hp ' {$110) ' Yl(c)/c]2 I 2Y1(0)2}"2

Weighted Profile R...
Help " {zwl'lyl(°) ' Yi(c)/c]2 / 2wl°Yl(o)2}"2
These two profile R values indicate how close a calculated profile is to the

observed one. From the view point of profile fitting they are more meaninng than R, and

R.. FL, Is the most significant factor since its numerator is the quantity actually minimized.

CHAPTER 3 EXPERIMENTAL

SYNOPSIS

This chapter is divided into two sections. In the first section the synthesis
procedure is described as three steps: (1) synthesis of anhydrous individual halides; (2)
synthesis of mixed or mixed-valent halide and hydrate precursors; (3) decomposition of
precursors at low temperatures. Mixed(-valent) halides sz-Lnx3 were prepared by
various methods including partial reduction with hydrogen, thermal decomposition. the
ceramic method. and the molten salt method. In the second section X-ray data
processing including data collection and the x-ray Rietveld refinement procedure are
described.

45
3.1 Synthesis Procedures

List of Important Chemical Reagents
Laps: 99.9%. Research Chemicals. Phoenix. AZ;
No.0,. sup. and W203: 99.99%, Research Chemicals;
Sm203: 99.9%. Alfa lnorganics. lnc., Beverly. MA;
Gd,0,: 99.9%. Michigan Chemical Corp..
BaCI2: 99.9%. Cerac/Pure. Inc. Milwaukee;
BaBr2: synthesized from BaO and NH.Br;
Sr(OH)2~8H20: Baker's Analyzed. J. T. Baker Chemical Co., .Phillipsburg. N. J.;
BaCl2-2H20: Analytic Reagent. Mallinckrodt Chemical Works;
BaBr,-2H,O: Prepurified. Fisher Scientific Company, Fair Lawn. N. J.;
Balg-2H20: Prepurified. Fisher Scientific Company;
Sm chips: distilled. 99.9%. Research Chemicals. Phoenix, AZ;
HgI2: Mallinckrodt chemical works. distilled;
NH4CI: Certified A.C.S. grade. Fisher Scientific Company;
NH.Br: Baker's Analyzed reagent;
NH.I: Analytic Reagent. Mallinckrodt Chemical Works;
THF: Baker's Analyzed reagent. purified;
Pyridine: Analytic Reagent. Mallinckrodt Chemical Works. purified;
H,(g): Matheson prepurified. >99.999%.

46
Sample Handling
Due to the moisture sensitivity of the Ianthanide halides and some alkaline earth

halides. their handling was performed in a glove box continuously purged of both water
by 4A molecular sieves (typical moisture content was 3—30 PPMv) and oxygen by heated
BASF catalyst (2000-3000 PPMv). ‘

Elemental Analysis
Halide ion contents were determined by titration with NaCI standardized AgNOa.
5% K,Cr0. solution served as the indicator; Ba" content was determined gravimetrically

by precipitation as BaSO..

Thermal Test for Metastability

To understand the formation mechanism and then'nai stabilities of the new
structure forms prepared by low-temperature solvolytic decompositions and low-
temperature dehydrations. specimens were heated in a high vacuum (106-10‘ Torr) at
various temperatures and then examined by X-ray powder diffraction for polymorphic
transformations.

47
3.1.1 Synthesis of Individual Halides

Lnx, (X.-Cl, Br, I) One of the principal difficulties encountered in preparing moisture-
sensitive anhydrous halides is preventing the formation of oxidehalide impurities. The
ammonium halide procedure (31.315) is very effective in eliminating oxidehalide impurity
and has been used throughout this work. The general procedure is shown schematically
below.

ana (s) + HX (aq) + NH4X (s) (excess) --—> (NHJnLnxam-tzo (s)

- H20 ~200°c

400°C
Lnxa (s) + NH‘X (s) < (NH.),.LnX3+.. (s)

vacuum

 

NH.X was removed by sublimation to the cool and of the tube. The part of the tube which

contained anhydrous Lnx, was sealed with a hand-held torch.

EuCl, synthesized by hydrogen reduction of EuCl, at 500-600°C. For a ~1 g sample
reduction required ~6 h. Samples were placed In pyrolytic graphite boats and transferred
between the reaction quartz tube and the glove box under argon. Hydrogen gas first
flowed over a Pd catalyst and then through a llquid-nitrogen-chilled trap before being

passed over the sample. Heating was effected in tubular furnaces.

SmCl, prepared by hydrogen reductionat 700-750°C.

YbCl2 The hydrogen reduction product contained a small amount of YbOCl. Pure YbCl2

48
was prepared by induction heating of a 2:1 YbCl3 : Yb mixture in a weld-sealed tantalum

tube at 900-1000°C for 10 h. A 1-2 g sample was prepared each time.

Sml, prepared by the mercury iodide procedure (7_1). A stoichiometric mixture of Sm
and Hg I, was sealed into a quartz tube under a 10*-10° Torr pressure. heated at 300°C

for 1 day. then at 500°C for 2 days.

But, a direct product of the ammonium halide procedure. Unlike the ammonium halide
procedure for other lanthanides where intermediate mixed halides such as (NH4)3LnX.
form. a mixture of Eu,O,. HI. and NH.I was observed to transform to Eulz-HzO and NH.I
below 220°C at which temperature NH.I sublimed out gradually. The reaction is presumed

to proceed as follows:

 

 

<100°C
Eula-tzo (S) > Eulerzo (S) + 1/2 12 (g) + (X-1) H20 (9)
vacuum
>220°C
Eula-H20 (s) > Eula (s) + H20 (9).
vacuum

Iodine released In this reaction was condensed in a liquid nitrogen trap. The preparation
of Eul2 by the ammonium iodide procedure is apparently a simple dehydration of
Eulz-Hzo. Thus Eul2 was also prepared using Eul,~xH,O (Eu.‘.0a dissolved in HI and the

solution evaporated to dryness) as the reactant.

49
3.1.2 Synthesis of Precursors (27.72.13)

Banana“ prepared by the ceramic method. ~1 g of an intimately ground mixture
of BaCl, and LaCl, in a 2:1 molar ratio was sealed into‘a quartz tube under vacuum,
heated at 800°C for 5 days. and then quenched.

Ba.,Ln..Cl.. (Ln=La, Sm) prepared by the molten salt method. ~1 g of a ground
mixture of 40-50% BaCl2 and LnCl, was confined in a pyrolytic graphite boat. heated at
~750°C (molten) for ~0.5-1 h. cooled slowly to 500°C, then quenched. The products
contained excess LnCl... Note that the formula of the precursors was deduced from
crystallographic data as will be discussed in Chapter 4.

Single crystals of Ba. ,SmmCl.‘ were grown from a 1 :1 molar ratio BaClz-SmCIa melt
in a vacuum-sealed quartz tube by melting the mixture (~750°C). cooling it to 500°C at the
rate of 3°C/h. and then to room temperature at 10°C/h. The fused salt was crushed with
an agate mortar and a pestle in the glove box. then extracted by THF for ~1 h. The
crystals were slightly yellowish. They were dispersed in sodium-dried paraffin oil for
microscopic examinations. Appropriate specimens were paper-dried and sealed into
capillaries.

Ba.Ln,Br,, (Ln=l.a, Nd) prepared by the molten salt method. ~1 g of a ground mixture
of 40-50% BaBr, and LnBr, was treated in the same way as those for the chlorides. but

in an argon atmosphere and at 700°C. _ The products contained excess LnBra.

EuCl, (2<x<3) prepared by partial reduction with hydrogen. ~1 g EuCl, was placed
in a graphite boat. heated in flowing H2 at 450-465°C. The sample turned to pale blue

so
slowly. Reduction was stopped after 3~4 h and the product cooled slowly. It was

identified by powder x-ray diffraction as a mixture of two mixed-valence chlorides (55).
the vemler type Eu,Cl.. (major phase). and the cluster type Eu..Cl33 (minor phase) with
traces of EuOCl and an unidentified impurity phase. Even though at these temperatures
EuCI, will be reduced to EuCl, eventually. no trace of EuCl, was detected by X-ray
diffraction. The reaction proceeded slowly enough to prevent complete reduction below
465°C in such a short period of time.

Eu.Cl, prepared by the ceramic method. An intimately ground mixture of EuCI2 and
EuCI, in 3:1 molar ratio was sealed into an outgassed quartz tube under vacuum and
heated at 500°C for 5 d. then quenched. The X-ray diffraction pattern of this blue powder
confirmed the product to be Eu.CI,.

Eu,Cl.. prepared by thermal decomposition or hydrogen reduction. 0.5-1 9 of EuCl3
was heated at 500°C for 5 h in a vacuum of 0.1 to 0.01 Torr; a blue product resulted. The

same product was also obtained by reducing EuCl, with H, at 400°C for 2 d.

Eu..c1,. prepared by thennal decomposition or solid state reaction. Thermal
decomposition of EuCl, (0.5-1 g sample) in vacuum at 440°C for 12 h yielded a black
powder Identified as the title compound by X-ray diffraction. The same compound was
also obtained when a 3:2 mixture of EuCl2 and EuCl, sealed in quartz under vacuum was
heated at 500°C for 5 d and quenched.

Sm..Cl,, prepared by partial reduction with hydrogen. Slow reduction of 1-2 9 SmCl,
at 650°C for 3 h followed by in situ heating to 750°C in flowing H2 produced a melt which

51
with subsequent slow cooling in the oven yielded a black rock-like material. The dark

brown pulverized material consisted of a mixed—valence phase and unreacted SmCl,. This
mixed valence compound is the same as that originally formulated as Sm,Cl7 (75). but
by analogy with Eu.4Cl,, (also originally formulated as Eu,Cl,) is formulated as Sm..CI,.,.

Sm,Gd,Cl,. Reduction of a 1:1 molar ratio mixture of SmCl3 and the irreducible mm,
at 660°C. above the melting point of GdCl, (602°C). for 4-12 h followed by slow cooling
produced a black product. It was dark brown after being pulverized. The product was a
mixture of Srn.Gd,Cl,.. and GdCl,.

Sm.Nd,Cl,, A mixture of SmCla and NdCl, in 9:5 molar ratio was reduced in hydrogen
at 660°C for 7 h. The temperature was increased to 850°C to produce a melt. after which
it was lowered slowly. Weight loss data suggested complete reduction of the Srn’+ ion.
The product had a diffraction pattern similar to that of Sm..Cl,, and was given the title

formula.

Srl,-xt'l,0 (x>2) It was crystallized from an aqueous solution of Sr(OH)2-8H20
neutralized with Hl acid.

Eul,+l,0 sup, was dissolved in excess hydroiodic acid. The aqueous solution was
carefully evaporated to dryness. The obtained yellow powder (Eula-xl-IZO) was confined
in a Pyrex tube connected to the vacuum system and heated slowly at 80-100°C in a 0.1-
0.01 Torr vacuum or at 60°C in a 10°-10° Torr vacuum for 4 h. Only the monohyidrate
was obtained. Eul,-xH,O with x>1 can also be prepared by the hydration procedure
described below.

52
Sml,1xl-I,D prepared by hydration of Sml... 0.5-2 9 Sml, powder confined in a round

bottle was placed in an evacuable glass tube fitted with a stopcock so that samples could
be transferred into the glove box for handling. A sufficient amount of NiCl,6H,O powder
was confined in another tube. The two tubes were connected through glass ground joints.
The system was evacuated. flushed with argon for 3 times. and reevacuated. Hydration
took place by water vapor transport from NiCl,6H,O to Sm la. The degree of hydration
can be monitored both by the color change of the iodide from deep green to reddish
brown and by the weight gain of the sample. For a 0.5 9 sample 5—6 h was sufficient to
ensure complete hydration (x>1). Over hydration should be avoided since samarium

iodide tends to form an aqueous solution and decompose.

3.1.3 Decomposition of Precursors

a. Soivolytlc Decomposition (Leaching)

Solvent Purification

Tetrahydrofuran (THF) was refluxed over sodium metal chips (benzophenone as
indicator) and distilled prior to use. For the Ianthanide (ll) halide experiments. it was also
deoxygenated repeatedly by first freezing with liquid nitrogen. pumping to 10‘5 Torr. and
then warming to room temperature. Pyridine was refluxed with KOH pellets and distilled

prior to use.

Extraction
Soivolytlc decompositions were carried out in a modified Soxhlet extractor fitted
with Teflon stopcocks and joint sleeves. A powdered sample was placed in the removable

53
extraction thimble and closed into the extractor under argon. Then the extractor was

transferred from the glove box to the extraction apparatus. Extraction was effected in a
dry argon atmosphere. The time needed depends on the sample used as will be
discussed individually later. When reaction was complete. the leached product was
evacuated for 1-3 h. blanketed with argon, and transferred into the glove box for further
treatment and analysis.

The solvolytic decomposition and related apparatuses are illustrated schematically
in Figure 10.

b. Dehydration

Samples for dehydration were confined in a glass container which was then placed
in a quartz tube with a stopcock and a ground joint connecting to the vacuum/argon line.
This tube can be transferred into the glove box for moisture sensitive samples. especially
dehydrated. fine powders. Heating was effected in a tubular furnace or an oil bath whose
temperature was controlled carefully.

Hydrates with more than one mole hydration were kept under a dynamic vacuum
at room temperature or heated mildly (below 80°C) overnight. The resulting monohydrates
were then dehydrated in a 10° to 10" Torr vacuum in the temperature range of 110-125°C
for 24 h. Water loss was checked by both weight change and X-ray diffraction.

54

 

      

Vacuum/Argon Line ‘ Quiz-ago" Gauge
' — — _—- 1‘ Heating Wire
v O . . ‘ BASF Catalyst
\ . c‘ . -11
1 .L 1. (R3 I
- 1E“. rfi
' '81 :‘1
fig ')‘. Molecular
I33 1‘1:1 S”
I“ %
hos—4’. W
is
ii
iii
Y I
- I
III TS 'l ‘ ;
j i
I ‘.
Ts/ ‘5 .
500 ml 5009/1 I .
Sulphur
1000 ml
I. Argon purification
ll. Solvent dehydration
lll. Solvent deoxygenatlon
IV. Extraction

TS: Teflon Stopcock

Figure 10. Apparatuses for the solvolytic decomposition experiment.

55
3.2 Instrumentation and Data Processing

3.2.1 Powder X-ray Diffraction Examinations

Due to the moisture sensitivity of most halides we encountered. X-ray diffraction
examinations were carried out in an evacuable Guinier-Hagg camera (diameter 1 14.6mm).
Quart monochromatized Cu K01 radiation (A . 1.54050 A) was used with NBS certified
silicon powder (a, - 5.43062(3) A) as the internal standard. Samples placed on Scotch“
Tape-backed planchets were coated with paraffin oil maintained over sodium chips to

protect them from hydrolysis during transfer to the camera.

3.2.2 Automated Powder x-ray Data Collection

X-ray intensity data suitable for Rietveld refinement were collected with a Philips
APD 3720 diffractometer system equipped with a sample spinner and B-compensating slit.
The extremely moisture-sensitive strontium iodide specimen was gently mixed with 'dry'
Aplezon'“ N grease and spread with a knife onto a Pyrex disk. The diffractometer sample

chamber was flushed continuously with nitrogen gas vaporized from liquid N...
3.2.3 Data Processing

a. Indexing and Lattice Parameter Refinement

Identification of compounds of known structure types and lattice parameters can
be made by comparing the observed patterns with the JCPDS powder files or with
calculated patterns (below). Powder patterns of unknown structures were indexed by
programs IT012 (Z5) and/or TREOR (15). IT 012 requires at least 20 observed

reflections and TREOR can work with fewer than 20 reflections as long as the number is

56
specified by the key word "USE“. In general TREOR is better for high symmetry systems

and IT012 for low symmetry systems. Program APPLEMAN (E) was used to perform

lattice parameter refinement.

b. Systematic Extinctions and Space Groups

Given a crystal system and lattice parameters. program APPLEMAN lists every
possible reflection. Systematic extinctions and subsequently possible space groups were
deduced by comparing the observed pattern to that listed by APPLEMAN.

c. Calculated Intensities

Given a correctly-indexed lattice. the chemical formula. and a proper structural
model. program POWCOM (_7_5) is used to calculate the theoretical reflection intensities
to check the validity of the model structure.

32.4 Rietveld Refinement of the Srlz-IV Structure

a. Structure Solution

62 interplanar d-spaclngs were indexed by the program TREOR to an orthorhombic
unit cell with figures-of—merit (12.50) M(20)-19. F20-28; and M(62)=8. F62=14.
Possible space groups determined from systematic extinctions were ana (if 62) and
Pn2.a (# 33). Both a manual search in the CRYSTAL DATA Deterrninative Tables
effected with lattice parameter ratios and a computer search of the Canadian CRYSTDAT
database failed to identify any likely Isostructural ABz-type compound. Attempts to reduce
the observed lattice parameters to those of a known orthorhombic form of Srlz and Eul,

(91.52.5584) were also unsuccessful. The axial ratios of this new Srl2 structure

57
appeared atypical when compared to those of other ABz-type compounds. suggestive that

this compound may represent a new structure type. It is designated Srl,— IV type according
to Beck’s nomenclature notation (w.

Being interested in the topotactic behavior of low-temperature dehydrations. we
also examined the structure of the precursor. Srlz-I-lzo. When a comparison was made
between the new phase and Srl,+l,O (g). we noticed an unusual similarity in their
lattices. The lattice parameters of SrI,-IV derived from the Guinier data. 5 = 12.365(2).
5 . 4.938(8). and s - 8.408(1) A. were close to those of the well-characterized
monohydrate. 5 - 12.474(2), 5 -.- 4.495(1). and g .. 9.741(2) A. Furthermore. both
structures even could be assigned to the same space group. These results suggested
related atomic arrangements. This was not obvious in their powder patterns, apparently
because of their different unit cell dimensions. However. the observed Srlz-IV X-ray
powder diffraction intensities agreed reasonably well with those calculated with the Sr and
l positional and thermal parameters of Srlziizo. Thus the monohydrate structure devoid

of H20 provided a model for the refinement.

b. Data Reduction

The raw data collected by the diffractometer were stripped of the Cu K011.
component with the APO software (55). All subsequent calculations were effected on
a MICROVAX ll computer with the program XRSBZ (fl). Background and the broad
absorptionband.whlchspanned-15<29<~26°andwascausedbytheprotecting
grease. were removed by using the BG’VALU command and editing the coded data file
manually. Possible systematic 26 errors due to slight misalignment and displacement of
the sample surface from the G-shalt axis were corrected at the profile refinement level.

Observed intensites were multiplied by (sin6)" to correct for the B—compensating slit SLO).

58
Lorent and polarization corrections were applied with the routine STEPCO. Polynomial

scattering factors were used with the real and imaginary dispersion correction terms
(55). The 26-limits of the standard peak were modified until the R-value indicated that
an acceptable description (R<5%) had been achieved. The observed. calculated and the
difference profiles of the standard peak are plotted in Figure 11.

c. Refinement

Refinement began with the scale factor; after 3 cycles R. had decreased to 0.227.
Cell parameters were refined next. followed in alternate cycles by scale. profile and the
26 correction parameters. Unit weights and heavily damped (0.8) structural and profile
factors were employed during the entire refinement process. The refinement converged
quickly when positional parameters were allowed to vary with an overall temperature
factor. Preferred orientation variables and individual isotropic temperature factors were
inhoduced after positional parameter shifts ceased. When refinement under isotropic
conditions was complete (5, - 0.071). temperature factors were made anisotropic. That
for Sr became non-positive. definite; it was converted isotropic and refinement continued
to convergence. The final R, was 0.069. In the final cycles lattice parameters were not
refined.

A difference Fourier after refinement was complete indicated a residual electron
density of - 2 eA" at 0.63. 0.25. 0.09. and a lesser value close to the 1(1) and l(2) sites.
A background level of ~ 1 eA“ was present at numerous sites. Since the analytical data
suggested the presence of up to 1.5 %.oxygen. several attempts were made to reduce
this residual electron density by introducing trace amounts of oxygen. However. when
oxygen atoms were introduced. refinement would not converge. When refinement was

effected with the K1) and 1(2) site occupancy set to unity and that of Sr variable, the Sr

59
occupancy parameter refined to ~ 1.4 with R . 0.085. When the Sr occupancy parameter

was set to unity. those of K1) and l(2) refined to 080(4) and 0.78(5). respectively. with R
- 0.065. Therefore. no further improvement could be made.

It is worth noting that a preferred orientation effect was not observed even though
the sample has a thin-plate structure. The Apiezon'“ grease thus not only served as an
effective protectant during 21 hr of data collection. but also provided a paste in which the
randomly oriented crystallites could not stack. However. a potential problem with the

grease - X-ray beam absorption - could be a source of error.

 

£0-

30"

25"

20-1

COUNTS X10"

 

 

 

 

 

Figure 11. Standard peak for Srla-IV. Top: observed (xx) and calculated (—) profiles
(5000 counts offset); bottom: difference (Obsd. - Calc. ) profile.

CHAPTER 4 RESULTS

SYNOPSIS

Three types of metastable forms of dihalides were prepared at low temperatures.

1. ' Fluorite-type LnClz (Ln=Sm. Eu). which were observed previously only at elevated
temperatures and are not quenchable. have been prepared by solvolytic decompositions

of mixed-valent chlorides.

 

 

 

 

THF 775-785°C
Sm,Ln,Cl,, (s) > SmCl, (3). blue < > SmCl, (s). red
~60°C
high-T form normal form
Eu..Cl. (s)
THF 800°C
Eu,Cl.. (s) } > EuCl, (s). white < > EuCl2(s),white
60°C
Eu...Clas (s) high-T form normal form

2. Metastable high-pressure-like forms of Bax2 (XaCl. Br) with the anti-FezP-type
structure. which could not be prepared under high pressures. have been prepared at low
temperatures and under normal pressure by solvolytlc decompositions of mixed halides

and by dehydrations.

60

61

 

 

 

 

Py 170-500°C

Ba.,Ln.°CI,. (s) > BaCl2 (s) < X BaCI, (s)
~110°C 3.0 Gpa

(LnaLa. Sm) high-P form normal form
THF 500-600°C

Ba.LnsBr,., (s) > BaBr2 (s) < x BaBrz (s)
~60°C 4.0 Gpa

(LnaLa. Nd) high-+P form normal form

I
| 110°C/10° Torr

l
BaBrz-Hzo (s)

125°C 600-1000°C
eal21120 (s) > Balz (s) < Bal, (s)
10‘ Torr 3.0 GPa
high-T form normal form

 

 

3. High temperature forms of MI, (M=Sr. Sm. Eu) which have been observed
previously only at high temperatures and are not quenchable have been prepared by low
temperature high vacuum dehydrations. This new structure (Srlz-IV-type) has been refined

by the X-ray Rietveld procedure with Srl,-I-I,O devoid of H20 as the model structure.

120°C ' > 494°C
Mia-H20 (s) > Ml2 (s) < > Ml2 (s)
10‘ Torr for M = Sm
high-T form normal form

 

 

62
4.1 Precursors and Structures

EtllCI,l (2.0 d x < 3.0)

The structures of the phases prepared had been well characterized as described
in Chapter 2. The various synthesis procedures are summarized briefly by the following
equations. Lattice parameters are summarized in Table 3 (Page 15). Miller indices.
observed and calculated d-spaclngs and intensities of scattered peaks for the three

precursors are listed in Tables 7-9.

 

 

 

 

 

 

”2(9)
> Eu,Cl.. + Eu..Cl,3
450465 °C / 3-4 h
pale blue
“2(9)
> Eu,Cl..
400 °C/2 d
blue
vacuum
> Eu,CI..
500 °Cl5 h
EuCl, blue
white
vacuum
> 50.4033
440 °C/12 it
black
3/2 EuCI,
> EU“C|33
500 °C/5 d
black
3 EuCl,
> Eu.CI,(s)
500 °C/5 d

blue

 

 

 

Table 7. Miller indices. and observed and calculated interplanar d-spacings and
Intensifies for Eu.Cl,. '
h k l d.(A) d..(A) lc l, " h k I d.(A) d.,(A) lc lo "
0 2 0 14.143 5 -- 1 5 .2 2.6673 5 --
0 1 1 6.535 -- -2 8 0 2.5678 2.5605 18 w'
1 0 1 4.921 4.939 31 w“ 2 8 0 2.4895 25 --
0 4 1 4.870 3 --- 2 7 1 2.4681 2.4695 3 vw‘
0 6 0 4.714 2 -- 0 2 2.4605 2.4633 33 vw”
-1 4 1 4.080 4.064 100 m 8 2 2.4350 57
1 4 1 4.000 4.024 98 m -2 2 2 2.4328 2.4370 5 m
0 0 3.615 3.615 21 w 1 10 1 2.4303 5
8 0 3.536 28 -2 2 2.3392 13
3.524 w 2.3266 w. b
-2 2 0 3.529 2 2 2.3088 18
-1 6 1 3.441 3.415 vw 3 1 22685 36
2.2685 w*
0 0 2 3.358 3.376 21 w -2 1 22676 3
-2 4 0 3.260 -2 10 0 2.2620
2 4 0 3.1792 -3 4 1 2.1786 --
-1 1 2 3.0325 3.0470 6 w' -1 12 1 2.1435 19
2.1493 vw
-1 8 1 2.9005 13 3 1 2.1420 9
2.8748 w.b
1 3 2 2.8873 3 3 2.1387 2.1300 3 vw”
0 9 1 2.8466 3 --- 1 12 1 2.1086 2.1155 23 w’
1 8 1 2.8431 9 -3 6 1 2.0678 7
0 10 0 2.8285 8 -1 4 3 2.0523 20
2.0536 w. b
-2 1 2.8072 2.8048 5 w 1 4 3 2.0419 24
2 ’ 2 2.7425 s -2 s 2 2.0399 s
-1 2 2.6965 2.7078 11 w
Note. ‘ Due to poor crystallinity. many relatively weak lines were not observed.

“ v - very; w . weak; m . moderate; s = strong; b a broad. + / - means

slightly stronger / weaker than.

34

 

 

Table 8. Miller indices. and observed and calculated interplanar d-spacings and
Intensities for Eu,Cl...

'1 k ' dc(A) d01A) 'c 'o h k ' dc(A) do(A) 'c lo
0. 2 0 17.565 18.005 5 w 2 12 0 2.2746 5

0 1 1 6.653 6.667 4 w; -3 0 1. 2.2704 2.2726 11 w;
-1 0 1 4.953 15 -3 1 1 2.2656 6 .
1 0 1 4.924 4'937 14 3 3 0 1 2.2619 2'26" 12 w
-1 4 1 4.316 4313 3 w 3 1 1 2.2572 6
1 4 1 4.296 ' 2 012 2 2.2164 2.2184 1 vw
-1 5 1 4.050 4.059 96 vs -3 4 1 2.1963 2.1965 2 vw‘
1 5 1 4.034 4.037 100 vs 213 0 2.1643 21627 1 w
-1 6 1 3.763 3784 3 w -3 5 1 2.1606 ° 5

1 6 1 3.770 ' 2 -1 0 3 2.1588 21568 2 w
2 0 0 3.6069 3.6103 16 m* 3 5 1 2.1533 ° 5

2 1 0 3.5661 4 -1 15 1 2.1192 21
010 0 3.5170 3.5215 27 s' -3 6 1 2.1171 2.1179 4 s
0 0 2 3.3674 3.3664 21 m 1 15 1 2.1169 21

2 5 0 3.2096 3.2129 12 m‘ 3 6 1 2.1103 2.1123 4 w‘
1 1 2 3.0477 3.0507 2 w' -1 5 3 2.0636 2.0641 23 m‘
012 0 2.9306 29323 3 m- 1 5 3 2.0574 2.0576 24 m
2 7 0 2.9301 ' 2 -1 6 3 2.0256 20254 2 w
011 1 2.6914 2.6926 4 w -2 10 2 2.0249 ‘ 6
-110 1 2.6676 28651 9 m 2 10 2 2.0169 2.0166 6 w
110 1 2.6619 ° 9 -1 16 1 2.0092 20087 2 w
-2 6 1 2.6031 2.6067 5 w 1 16 1 2.0072 ' 1

2 6 1 2.7925 2.7952 5 w -1 7 3 1.9635 1.9622 1 w‘
-1 6 2 2.7216 2.7215 6 w’ 1 7 3 1.9778 1
1 6 2 2.7121 2.7135 6 w 215 0 1.9656 1.9666 10 w;
2 9 0 2.6505 2.6567 2 vw -2 11 2 1.9579 1.9592 2 vw
210 0 2.5161 2.5206 35 s -3 10 1 1.9074 1.9065 6 w’
-2 0 2 2.4766 2.4779 14 w‘ 3 10 1 1.9024 1.9037 6 w‘
-2 1 2 2.4705 2.4710 5 w‘ -311 1 1.6511 1.6529 3 w
2 0 2 2.4620 14 . 311 1 1.6465 2
2 1 2 2.4559 2'46” 4 m 0 11 3 1.6446 2 «-
010 2 2.4396 2.4365 57 vs 4 4 0 1.7667 4
211 0 2.3926 2.3939 6 w 0 20 0 1.7565 1.7567 6 w;
-2 5 2 2.3360 2.3366 12 m’ 4 5 0 1.7470 1.7477 5 w‘
2 5 2 2.3237 2.3259 13 m 4 6 0 1.7237 1.7250 3 vw

 

Table 9.

Miler indices. and observed and calculated interplanar d-spacings and
intensifies for £01.01”.

 

 

h k (MA) de(A) I. I. h k | GM) 41(8) '6

1 0 1 10.128 18 --- 4 0 4 2.5320 2
2.5303 vw

0 1 2 8.265 5 --- 2 2 6 2.5309 3

0 0 3 8.256 5 --- 2 3 2 2.4939

1 0 4 5.407 5.425 2 WV 3 2 -2 2.4939 2.4876 4 s'

1 1 3 5.062 3 -- 2 1 -8 2.4910

1 1 -3 5.062 2 --- 1 4 0 2.4218 2.4220 76 m‘

1 2 -1 4.136 5 1 0 10 2.4173

0 2 4 4.132 4.120 5 w 4 1 -3 2.3239 2.3256 3 vw

0 0 6 4.128 40 2 1 10 2.1328 2.1291 29 w

1 2 2 3.973 - 5 1 4 6 2.0889 24

3.970 5 2.0877 rn+

2 1 -2 3.973 1 00 1 4 -6 2.0889 24

3 0 0 3.699 4 --- 4 2 2 2.0679 23
2.0683 w

2 0 5 3.695 4 --- 0 0 12 2.0640 6

2 1 4 3.473 3.472 68 111 4 2 -4 1.9865 1.9880 9 vw+

3 0 3 3.376 5 --- 5 1 -2 1 .9679 2

2 2 -3 2.9867 7 2 3 8 1.9665 1.9665 2 vw

2.9872 w

0 1 8 2.9821 3 3 2 -8 1.9665 2

4 0 1 2.7573 4 4 2 8 1.7364 1.7363 20 1111+

3 1 -4 2.7563 2.7547 8 w 3 3 9 1.6873 1.6860 2 vw

0 3 6 2.7550 2 2 1 -14 1.6301 1.6287 9 vw’

0 4 2 2.7074 3 4 2 ~10 1.6006 1.6003 11 vw’

2 1 7 2.7046 2.7033 7 vw+ 0 7 2 1.5726 8
1.5736 w‘

1 2 -7 2.7046 3 3 5 -2 1.5726 8

3 1 5 2.6144 2.6144 7 vw

 

66
Sm.”Ln,"CI,. (Ln = Nd, Sm, Gd)
All these have the cluster-type structure as discussed previously. Lattice
parameters are listed in Table 3 (Page 15). Their powder diffraction patterns are similar
to that of Eu1‘Cla (Table 9) and are not listed.

BaCl,-LnCl, systems (Ln = La. Sm)

Phase studies on BaCla-LnCl, systems by two separate groups have revealed in
addition to other miscellaneous phases a common 2:1 compound, Ba,LnCi,, that forms
over a wide range of temperatures (89.99). No structural details about these phases
were available. The products which resulted from fusion of the approximate 1:1 BaCiz-
LnCi, mixtures were initially thought to be this 2:1 phase with excess LnCia. Earlier
powder patterns were indexed on a hexagonal unit cell (2;). However, later single crystal
X-ray diffraction studies indicated that the true symmetry for 'Ba,SmCl," is cubic and the
composition is probably Ba.,Sm,oCl.. (below). in fact, the hexagonal lattice parameters
can be transformed into pseudo-cubic parameters by ac . 11312., and ac . 1/6/2-a,,.
Program iT012 also suggested a similar cubic lattice. indexed lattice parameters from
powder patterns, transfonned cubic and observed single crystal data are listed in Table
10. The transformed cubic lattice parameter of 'Ba,LaCl," from a, does not agree well
with that transformed from 6... Therefore the true symmetry of this phase remains in
doubt.

Systematic extinctions from single crystal data of ”BaszCi; correspond to the
unique space group P a 3 (# 205). it was noticed that mis phase has the same space
group and a lattice similar to that of the previously reported mixed-fluoride cfi-phase
M’*,,Ln’*,oF.. (M :- Ca. Sr. Sm, Eu. Yb) (23). Thus the mixed chloride phase is tentatively

formulated as Ba",,Sm°*,°Cl.‘. Density values are (gcm‘): calculated 4.160 with Z a 4;

67
found 4.08(:l:0.07). The theoretical density of 8628mm, (4.098 with Z a 36) is also close

to that observed. thus the experimental value does not permit one to distinguish between
the two formulations. However, the formulation Ba’*,,Sm°*,°Cl.. is more reasonable in
view of the crystal chemistry. The cubic lattice is a superiattice corresponding to 3x3x3
cubic fluorite cells. aJa - 7.122 A for Ba,,Sm,°Ci.. is very close to the cubic unit cell
parameter of BaCl, (ac = 7.311(1) A) (_9_1). its relation with the fluorite type structure is
apparent in the powder X-ray pattern. The “basic reflections” can be indexed as a cubic
cell with a. - 7.1352(6) A. Observed powder diffraction reflections are listed in Table 11.
This formulation corresponds to 10x4 . 40 Srn3+ and 40 “extra" Cl' ions in a unit cell. and
thus suggests 8 BaSm,,Cl37 clusters.

The crystal structure of this phase has not been solved. Direct method procedures
did not give useful phase information and the Patterson map is too. complicated to

interpret.

Table 10. Transformation of 'Ba,LnCl,' hexagonal lattice parameters into cubic lattice
parameters. (units in A)

 

 

 

Lattice 'Ba,LaCl,' 'Ba,SmCl,'
a... or. 17.637(6), 12.495(5) 17.471(6), 12.351(6)
derivedac 21.601, 21.642 21.396, 21.393

 

observed ac ------ 21.366(2)

 

Table 11.

Miller indices. observed intensities. and observed and calculated interplanar
d-spacings for Ba1 ,SmmCi... Cubic a . 21.406 A (powder); space group

 

 

P 63.

no. h k I MA) duty 1, no. 11 k i dc(A) d,(A)* I.
1 2 0 0 10.703 10.663 wv‘ 35 10 2 0 2.0990 2.0967 vw
2 3 0 0 7.1353 7.1239 vw 36 10 2 1 2.0690 2.0916 vw
3 3 1 1 6.4541 6.4535 W 37 10 2 2 2.0596 2.0601“ w‘
4 3 2 1 5.7210 5.7149 w 38 10 3 1 2.0410 2.0411 wt
5 4 1 0 5.1917 5.1696 vw 39 6 7 1 2.0049 2.0065 wv’
6 3 3 0 5.0454 5.0556 vw 40 9 6 0 1.9790 1.9790 wv
7 5 1 1 4.1196 4.1092“ vs 41 11 1 0 1.9360 1.9391 w
6 5 2 0 3.9750 3.9705 w 42 9 7 2 1.6492 1.6499 917
9 5 2 1 3.9062 3.9055 vw 43 11 4 2 1.6027 1.6036 w
10 4 4 1 3.7263 3.7259 vw 44 12 0 0 1.7636 1.7843” w+
11 6 0 0 3.5677 3.5625" m 45 12 2 0 1.7596 17553 W
12 6 1 1 3.4725 3.4749 vw 46 10 7 0 1.7536 '

13 5 4 0 3.3430 3.3460 vw 47 11 6 1 1.7030 1.7033 vw
14 5 5 0 3.0273 3.0263 w 46 12 4 1 1.6670 1.6661 vw
15 5 5 1 2.9974 2.9962 vw 49 11 7 1 1.6370 1.6371“ w’
16 6 4 0 2.9665 2.9706 vw 50 13 2 0 1.6275 1.6264 w’
17 7 2 0 2.9403 2.9363 w 51 12 6 0 1.5955 1.5957" w
16 7 2 1 2.9130 2.9156 vw 52 10 9 1 1.5667 1.5666 vw
19 7 2 2 2.6353 2.6350 w 53 11 6 0 1.5736 1.5743 w‘
20 7 3 1 2.7666 2.7671 vw 54 13 5 0 1.5369 1.5372 w
21 7 3 2 2.7166 2.7200 w 55 14 2 1 1.5099 1.5105 w
22 7 4 0 2.6551 2.6531 w 56 13 6 1 1.4914 1.4917 01/
23 7 4 1 2.6349 2.6339 w' 57 13 6 2 1.4607 1.4614 vw
24 6 2 0 2.5959 2.5971 vw 56 14 4 2 1.4565 1.4560“ w*
25 6 6 0 2.5227 2.5213" vs 59 14 5 0 1.4399 1.4406 vw
26 7 5 0 2.4884 2.4664 w‘ 60 15 2 1 1.4115 1.4114 wv
27 6 6 2 2.4554 2.4551 vw 61 15 3 0 13993 1.3992 vw
26 6 3 2 2.4394 2.4410 w‘ 62 15 3 3 1.3732 1.3729“ w
29 9 0 0 2.3764 2.3600 w 63 13 10 0 1.3051 13053 1111‘
30 6 5 0 2.2690 2.2694 1111* 64 12 12 0 1.2613 12617” 111‘
31 6 5 2 22197 22204 vw 65 15 9 3 1.2061 1.2063“ w
32 7 7 1 2.1514 2.1501" s’ 66 16 6 0 1.1262 1.1281“ w‘
33 10 1 0 2.1300 2.1313 vw 67 15 9 9 1.0661 1.0890" vw
34 10 1 1 2.1195 2.1199 vw

 

Note.

' powder pattern taken by Guinier camera with ground crystals grown in a melt;

“ these lines define the basic fluorite-type structure with the cubic lattice

parameter a .. 7.1352(6) A.

69
BaBr,-LnBr, systems (Ln = La, Nd)

The mixed bromide systems have received little attention. The vemier-type
ianthanide mixed-valence bromides Nd4Br, (92) SmnBrzm (n:5,6,11) (93,91);
SgLnBr11 and Sr,NdBr,, (_9_5); Cf.GdBr,, (5) have been prepared. Reports of other
alkaline earth-ianthanide bromides could not been found. Our studies revealed a single
mixed bromide phase with excess LnBr3 present. This phase was identified as the cluster-
type Ba,Ln,Br,, which is lsostructural with Eu,.Cl,, (3;). Observed and calculated X-ray
powder diffraction patterns are in good agreement. The refined hexagonal lattice
parameters of rhombohedral Ba,LasBr33 and Ba,NdsBr,3 are listed in Table 4 (Page 16).
Miller indices. calculated and observed interplanar d-spacings and intensities are listed in
Tables 12 and 13.

The structural similarity between BagLnserm and Eu14Cl” can be understood in
terms of their radius ratio values: r(Ba’*)/r(Br’) = 0.724; r(Eu"’*)/r(Ci‘) = 0.691 (Viil
coordinated effective cationic mdii)@. Formation of vemier-type mixed bromides such
as BaaLnBr, and Ba..LnBr,,, would also be predicated by analogy with the europium
chloride system. Other phases were indeed observed at different compositions but no

further effort was made to characterize them in this work.

"1.11.0 (M = Sm, Eu, Sr) When diiodide hydrates were evacuated to 0.1-0.01 Torr
overnight at room temperature or with heating below 80°C, they were converted to

monohydrates with identical othorhombic structures.

70

 

 

Table 12. Miller indices. and observed (Guinier) and calculated interplanar d-spacings
and intensities for Ba.La,Br,3.
h k I dclA) d.(A) I. l. h k I dclA) dolA) 1. I.
1 0 1 11_10211,193 1 vw' 4 0 '4 2.7755 2.7799 2 vw'
1 1 0 7,049 7,074 4 w, 2 2 6 2.7620 2.7637 9 w“
- 2 3 2 2.7412 14
0 0 6 4.446 4.443 21 w‘ 3 2 -2 2.7412 5
2 1 -2 4.361 4.359 47 m 1 1 -9 2.7325 2
- 1 2 6 2.7029 3
2 0 5 4.017 4.022 1 vw 2.7012 vs
2 1 4 3.795 3.600 5 w 2 1 -8 27029 100
3 o a 3,701 3,709 3 vw 1 4 0 2.6643 2.6639 99 vs
1 2 5 3,4903 3.4903 3 w' 1 0 10 2.6063 2.6063 6 vw‘
1 3 1 3,3593 2 _ 2 3 -4 2.5625 2.5656 1 vw
3 1 -1 3,3593 33546 1 w 4 1 3 2.5522 2
1 4 3 2.5522 1
3 1 2 3.2621 2 25530 W.
1 3 -2 3.2621 2 1 4 -3 2.5522 2
2 2 3 3.2765 3 4 1 -3 2.5522 3
3.2763 1111”
- 2 3 5 2.4600 1
2 2 3 3.2765 15 24800 W.
0 2 7 3.2328 3.2363 2 vw 3 2 -5 2.4800 . 3
0 1 6 32169 3.2197 2 vw 0 2 10 2.4446 2.4449 2 vw
4 0 1 3.0325 3.0336 11 w 0 5 1 2.4317 2.4329 4 vw‘
3 1 -4 3.0194 3.0197 16 w’ 3 0 9 2.3960 1
0 1 11 2.3766 2
3 0 6 3.0021 3.0034 2 W 23791 W.
0 3 6 3.0021 6 3 1 8 2.3760 1
0 4 2 2.9754 2.9790 3 vw‘ 3 3 0 23497 23505 2 WI
0 0 9 2.9642 2.9660 3 vw 2 1 10 2.3096 2.3079 24 m“
2 1 7 2.9366 7 14 . 2 4 1 2.2987 1
1 2 -7 2.9366 2'9“ 10 m 0 5 4 22930 1
3 1 5 2.6590 2.6596 16 w’

 

71

 

Table 12. (continued)
1 4 6 2.2854 18 4 4 0 1.7622 1 -~-
2.2841 m'
1 4 ~6 2.2854 18 2 1 ~14 1.7613 1.7612 4 vw’
4 2 2 2.2736 17 3 1 ~13 1.7550 1.7549 2 vw‘
2.2733 w’
2 4 ~2 2.2736 4 4 2 ~10 1.7452 1.7448 10 w’
2 0 11 2.2539 2.2552 1 vw' 5 3 2 1.7294 3
0 4 8 2.2516 1 ~~~ 0 7 2 1.7294 1.7288 6 w1
1 5 ~1 2.1 855 2 3 5 -2 1 .7294 6
2.1864 vw+
5 1 1 2.1855 1 1 5 ~10 1.6940 4
1.6936 vw’
2 3 8 2.1448 1 5 1 10 1.6940 4
2.1449 vw‘
3 2 -8 2.1448 2 6 2 ~2 1.6796 1.6791 2 vw’
5 1 4 2.0831 2.0846 1 vw 1 6 7 1.6729 1
0 6 3 1.9836 1 6 1 -7 1.6729 1.6735 2 vw
6 0 3 1.98361.9827 1 vw 2 4 ~11 1.6717 1
4 1 -9 1.9815 1 4 4 ~6 1.6383 1.6361 3 vw
0 3 12 1.9510 1 ~- 2 5 9 1.6320 1.6317 1 vw’
5 2 3 1.9094 1 1 5 11 1.6266 2
1.9084 vw‘
5 2 ~3 1.9094 1 1 6 ~8 1.6257 1.6249 3 w’
4 2 8 1.8974 1.8957 24 w’ 6 1 8 1.6257 2
2 2 ~12 1.88031.8812 2 vw' 2 1 16 1.5681 1.5672 25 w’
3 3 9 1.8413 1.8409 4 vw 4 5 ~1 1.5605 2
1.5604 vw'
5 2 6 1.7897 2 5 4 1 1.5605 3
2 5 ~6 1.7897 1.7885 1 vw+ 3 5 ~8 1.5455 12
1.5446 1111’
5 2 -6 1 .7897 3 0 7 8 1 .5455 12
0 0 15 1.7785 1 6 3 0 1.5382 1.5379 11 vw
4 3 7 1.7759 1.7750 2 WV

 

72

 

 

Table 13. Miller indices. and observed (Guinier) and calculated interplanar d-spacings
and intensities for Ba.Nd,Br,a.
h k | dc(A)- 4(4) 1. I. h k 1 (WA) 4.18) I. I.
1 0 1 11.049 11.043 2 w‘ 2 1 7 2.9204 13
_ 2.9183 m
0 1 2 8.955 8.971 <1 vw 1 2 -7 2.9204 9
1 1 0 7.019 7.007 5 w 3 1 5 2.8443 16
2.8423 m
0 2 1 5.925 5.928 3 w' 1 3 ~5 2.8443 2
1 0 4 5.814 5.839 <1 vw 3 2 1 2.7739 <1
2.7740 vw
1 1 3 5.495 5.509 <1 vw 2 3 -1 2.7739 <1
2 1 1 4.528 <1 4 0 4 2.7622 2
4.539 vw
1 2 -1 4.528 <1 2 2 6 2.7469 2.7469 9 w'
0 0 6 4.413 4.414 22 m' 2 3 2 2.7293 14
2.7275 m'
2 1 -2 4.341 4.336 49 vs 3 2 ~2 2.7293 5
2 0 5 3.993 3.990 2 vw‘ 1 1 ~9 2.7132 1
2 1 4 3.7748 3.7750 6 w‘ 1 2 8 2.6856 2 6825 3
3 0 3 3.6630 3.6624 3 W 2 1 ~8 2.6656 ' 100 vs
1 2 5 3.4707 3.4710 3 w 1 4 0 2.6531 2.6507 99 vs
1 1 . . . '
3 3345033426 1 w 1 0 10 25871 25865 6 w
3 1 ~1 3.3450 1 2 3 4 2.5704 2.5708 1 vw
3 1 2 32677 2 4 1 3 2.5408 2
1 3 -2 3.2677 1 1 4 3 2.5408 1
3.2598 m‘ 2.5397 w+
2 2 3 32613 3 1 4 ~3 2.5408 2
2 2 ~3 3.2613 14 4 1 ~3 2.5408 3
0 2 7 3.2115 32140 2 w 1 3 7 2.5170 <1
2.5164 w'
0 1 8 3.1934 3.1960 2 vw‘ 3 1 -7 2.5170 <1
4 O 1 3.0197 3. 1 11 ' 2 . 7 1
0 88 m 3 5 2 46 8 2.4672 w*
3 * 1 -4 3.0046 3.0035 16 w’ 3 2 -5 2.4678 3
3 0 6 2.9849 2.9835 2 w 0 2 10 2.4274 2.4291 2 vw'
0 3 6 2.9849 7 ~~~ 0 5 1 2.4214 2.4207 4 w
0 4 2 2.9624 2.9643 3 vw 3 0 9 2.3807 2.3826 2 vw'
0 0 9 2.9419 2.9463 3 vw

 

73

 

Table 13. (continued)
3 1 6 2.3620 1 5 2 6 1.7612 2
0 1 11 2.36112'3610 2 w’ 2 5 -6 1.7812 1.7613 1 w‘
3 3 0 2.3396 2.3393 2 w 5 2 -6 1.7612 3
2 - 1 10 22941 2.2932 24 4 3 7 1.7672 2

. 1.7671 w
2 4 1 2269122832 1 W 0 0 15 1.7651 1
0 5 4 2.2825 1 4 4 0 1.7549 1
1 4 6 2.2736 16 2 1 -14 1.7469 1.7490 4 w‘

2.2711 m
1 4 -6 22736 16 3 1 -13 1.7434 2
4 2 2 22636 22619 16 m 4 2 -10 1.7354 1.7355 10 w
2 4 -2 2.2638 4 0 7 2 1.7221 7
2 3 -7 22449 2.2436 <1 vw' 3 5 -2 1.7221 1.7214 6 w
0 4 6 2.2387 2.2386 1 W 5 3 2 1.7221 3
2 0 11 2.2360 1 1 1 -15 1.7116 1.7120 <1 W
5 0 5 22096 2.2101 <1 w 1 1 15 1.7118 <1
0 0 12 2.2069 <1 5 1 10 1.6646 4
1 5 -1 2.1763 2 1 5 -10 1.6846 4
5 1 1 2.1763 2'1759 1 w 6 2 1 1.6626 1"5826” w
2 3 6 2.1326 1 2 6 -1 1.6826 <1
3 2 ~8 2.1328 21332 2 “r 6 2 -2 1.6725 1.6727 2 w‘
2 4 -5 2.1076 <1 1 6 7 1.6646 1
1 1 -12 2.1049 “062 <1 w’ 6 1 -7 1.6648 1.6641 1 w
5 1 4 2.0737 2.0744 1 vw 2 4 -11 1.6620 1
3 3 6 2.0672 2.0666 <1 vw” 6 2 4 1.6339 1
4 0 10 1396419934 <1 WV 4 4 -6 1.6307 1.6314 3 vw
3 4 .1 1.9931 <1 2 5 9 1.6235 1.6237 1 vw
0 6 3 1.9750 1 1 6 ~8 1.6176 3
6 0 3 1.97501'9755 1 "w 6 1 6 1.6176 1.6169 2 w
4 1 -9 1.9702 1 1 5 11 1.6173 2
0 3 12 1.9376 1 -- 3 3 12 1.6053 1.6063 1 vw‘
5 2 3 1.9011 1 2 1 16 1.5569 1.5562 25 w
1.9023 vw

5 2 -3 1.9011 1 5 4 1 1.5540 1.5567 3 W,
4 2 6 1.6674 1.6636 25 m 4 5 -1 1.5540 2
2 2 -12 1.6660 1 2 6 -7 1.5400 3
3 3 9 1.83121.8320 4 w‘ 3 5 ~8 1.5360 1.5374 12 m
1 5 6 1.82271.8241 1 w‘ 0 7 6 1.5380 11

 

74
4.2 Decomposition Products and Structures

4.2.1 High-Temperature Forms of LnCl, (Ln .- Sm. Eu)

Fluorite-type EuCl, was obtained as a fine white powder when the mixed-valence
europium chlorides were extracted by THF. The fluorite-type SmCl2 that resulted from
extraction of both Sm,.Cl,, and Sm.Gd,Cl,, by THF was a dark blue powder in contrast
to the blood red color of the PbCl,-type SmCl,. Chlorine analysis of this new phase
indimted the composition to be SmCl1 my Extraction of Sm.Nd,Ci,3 also yielded the
fluorite modification, but the product was impure and exhibited a reddish color.

Lattice parameters for both EuCl, and SmCl2 and their calculated densities are
listed in Table 14 together with values from previously observed high temperature phases.
Experimental d values and powder diffraction intensities are in good agreement with the

those calculated for a fluorite-type structure (Tables 15 and 16).

Table 14. Lattice parameters and calculated densities of LnCl, (Ln = Sm. Eu).

 

 

Compound a (A) d, (g cm") T (°C) Reference
SmCl, 7.1496(2) 3.996(2) 775-785 mg

6.9827(5) 4.3167(9) 23(2) this work
EuCl, 7.150(1) 4.073(2) 800 _1_gz

6.961 (1 ) 4.387(2) 23(2) this work

 

75

 

 

Table 15. Miller indices. and observed and calculated interplanar d-spacings and
intensities for fluorite-type SmCl2.
11 n k I dc(A) d,(A) I;= 1o
1 1 1 1 4.019 4.0339 100 vs
2 2 0 0 3.460 3.4976 14 w
3 2 2 0 2.4608 2.4691 95 vs
4 3 1 1 2.0966 2.1052 56 s
5 2 2 2 2.0092 2.0163 6 w
6 4 0 0 1.7401 1.7460 17 w
7 3 3 1 1.5966 1.6023 26 m
6 4 2 0 1.5564 1.5622 9 w
9 4 2 2 1.4207 1.4257 36 m
10 5 1 1 1.3395 15
11 3 3 3 1.3395 "3437 5 w
12 4 4 0 1.2304 1.2343 11 w
13 5 3 1 1.1765 1.1600 20 m
14 4 4 2 1.1600 4
15 6 0 0 1.1600 “634 W
16 ' 6 2 0 1.1005 1.1036 16 w
17 5 3 3 1.0614 1.0642 6 w
16 6 2 2 1.0493 1.0521 3 vw

 

76

 

 

Table 16. Miller indices. and observed and calculated interplanar d~spacings and
intensities for fluorite-type EuCl,.
# n k 1 dc(A) d°(A) 1c 10
1 1 1 1 4.019 4.024 100 vs
2 2 0 0 3.460 3.464 14 w
3 2 2 0 2.4606 2.4610 95 vs
4 3 1 1 2.0986 2.0999 56 s
5 2 2 2 2.0092 2.0117 6 w
6 4 0 0 1.7401 1.7412 17 w
7 3 3 1 1.5968 1.5974 26 m
6 4 2 0 1.5564 1.5576 9 w
9 4 2 2 1.4207 1.4209 36 m
10 5 1 1 1.3395 15
11 3 3 3 1.3395 13396 5 w
12 4 4 0 1.2304 1.2303 11 w
13 5 3 1 1.1765 1.1767 20 m
14 4 4 2 1.1600 4
15 6 0 0 1.1600 “598 W
16 6 2 0 1.1005 1.1001 16 w
17 5 3 3 1.0614 1.0606 6 w
16 6 2 2 1.0493 1.0406 3 vw

 

77
4.2.2 High-Pressure Forms of Bax2 (x - Cl, Br, i)

The mixed chloride precursors Ba, ,anCl,‘ were first leached by THF. No apparent
change was observed even after 3-4 days of continuous extraction. They were then
extracted by pyridine. A new phase was observed after several hours, but solvoiysis
proceeded exceedingly slowly. With a 0.5-0.8 g reactant (Ba,,Sm,°Cl.. and SmCla
mixuire) reaction was complete in 5-6 days. The X-ray powder reflections of the white
powder that remained agreed well with those previously reported for hexagonal anti-FezP-
type BaCl2 (g). The solvoiysis reaction for Ba,LaCi,, incomplete even after 8.5 days.
appeared to have stopped. but the same “hexagonal“ phase was obviously present.

On the other hand. the mixed-cationic bromides were decomposed rapidly by THF.
With 0.5-1.0 g specimens, especially Ba.Nd,Braa, reaction was apparent after 30 min of
extraction and decomposition complete in 2-3 days. The white powder that remained also
exhibited the hexagonal anti-FezP-type structure. Elemental analysis for barium confirmed
that it was Basra: calculated Ba% for Bear” 46.22; found, 45.47. This form was also
prepared by dehydrating BaBrz-l-lzo at 110°C in a high vacuum for 24 h.

Anti-Fe,P modification of Balz, prepared by Beck under high pressure (Q); was
also obtained by high vacuum dehydration of Balz-i-lzo at 110-150°C. The observed
diffraction reflection intensities of the anti-FezP-type Bax2 (X=CI. Br, l) prepared in this
work are in good agreement with those calculated with the positional and thermal
parameters reported for Bai, (Tables 17-19).

Lattice parameter and volume per forrnuia data for the anti-FeaP-type and normal
BaCi2 and BaBr, modifications are presented comparatively in Tables 20 and 21.
respectively. From the tables it is obvious that the anti-FezP modification of both 6aCl2
and BaBr2 is more efficiently packed than the normal modification (the PbCi,~type). They

are considered the “high-pressure" forms similar to that of Bal,. The average volume per

78
formula unit, V/Z, of our anti-FezP-type form is about 1% smaller than that of the PbCl,-

type modification. However. the V/Z data on the anti-FezP form of BaCi2 produced by
dehydrating BaCi2~2H20 (g) at increasing temperatures is slightly larger (at about the
same temperature) than that of the normal form. This less dense material could result
from poor specimen crystallinity as the lattice parameter standard errors are somewhat
large. but more likely results from the matrix environment in which the crystals formed.

The BaCl2 and BaBr2 specimens obtained by soivolytic decompositions are
relatively pure and fairly well crystallized. The most intense reflection of the PbCi2~type
modification was frequently observed as a very weak reflection on the Guinier films. This
impurity content is estimated to be approximately 35% based upon two factors. First. the
NdOBr quantitatively determined impurity level in the BaBr2 sample was 3% (determined
by weighing the mass of the water insoluble residue). Second, the intensities of the
PbCla-type Bax2 and the NdOBr X~ray reflections are comparable, the scattering powers
are similar, and the symmetries are closely related (orthorhombic and tetragond,
respectively).

' it is difficult to eliminate the PbCl2 form. it probably results from the unreacted
barium halide in the mixed-cationic halide precursors. Transformation of the anti-Fe;-
type to the PbCl, type is highly unlikely at the leaching temperatures (GO-115°C) because
thermal tests (discussed later) demonstrated that the transformation was very slow even
at 300°C.

79

Miller indices. and observed and calculated interplanar d~spacings and
intensities for anti-FezP-type BaCl,.

Table 17.

 

 

 

From precursor Ba. ,Sm,°Ci.. '8a,LaCi,'

h k | 11.14) 4(4) 1. 1. (4(4) «(Al I. I.
O 0 1 4.616 4.610 8 vw 4.592 4.619 8 vw
.1 1 0 4.034 4.029 89 3 4.033 4.032 89 s
1 0 1 3.851 3.849 70 m-s 3.837 3.838 70 m
1 1 1 3.0373 3.0356 39 in 3.0301 3.0289 39 rn‘
2 0 1 2.7854 2.7847 100 6 2.7799 2.7774 100 s
2 1 0 ‘ 2.6406 2.6399 10 vw 2.6403 2.6443 10 w
3 0 0 2.3287 2.3289 53 m 2.3285 2.3285 53 s'
0 0 2 2.3080 2.3073 27 w-m 2.2958 27
2 1 1 22920 22902 71 m 22888 2.2883 71 s
3 0 1 2.0791 2.0805 5 vw 2.0767 5
2 2 0 2.0167 2.0185 20 w 2.0166 2.0169 20 m'
1 1 2 2.0032 2.0038 29 w-m 1 .9952 1 .9960 29 m‘
3 1 1 1 .7866 1 .7876 16 w 1 .7850 1.7855 16 w
2 1 2 1 .7377 1 .7388 5 vw 1 .7324 5
3 0 2 1.6393 1.6411 34 vw 1.6348 1.6351 34 w
4 1 0 1 .5245 1 .5248 19 w 1 .5244 1.5246 19 w*
2 2 2 1.5187 1.5190 16 vw 1.5151 1 5135 16 w.
3 2 1 1.5141 1.5140 29 w 1.5131 ' 29

1 0 3 1.5026 5 1 .4950 1 .4949 5 vw’
4 1 1 1 .4479 6 ~~~ 1 .4467 1 .4550 6 vw
1 1 3 1 .4376 3 1 .4309 . 1 .4254 3 vw‘
2 0 3 1.4081 1.4081 11 vw 1.4018 1.4081 11 w
3 3 0 1.3445 1.3452 5 vw 1.3444 5
5 0 1 1.3373 1 .3375 8 vw 1 .3366 1 .3371 8 vw
2 1 3 1.3294 12 ~- 1.3241 12
4 1 2 1 2720 1 2713 20 w 1 2699 1 2695 20 w

 

80

Miller indices. and observed and calculated interplanar d~spacings and
intensities for anti-FezP-type BaBrz.

Table 18.

 

 

 

From precursor Ba.,La,Bras Ba,Nd,,Bras

n k I c.1111 4.14) I. I. 4.14) 4.1» I. I.
1 0 O 7.346 <1 ~- 7.345 <1 ~~~
0 0 1 4.824 4.833 4 vw 4.839 4.839 4 w
1 1 0 4.241 4.243 21 w* 4.241 4.238 21 m
1 0 1 4.032 4.033 1 6 w 4.041 4.039 1 6 w*
2 0 0 3.673 3.679 2 vw 3.672 3.664 2 vw
1 1 1 3.185 3.183 70 vs 3.189 3.187 70 s1
2 0 1 2.9224 2.9206 100 vs 2.9255 2.9208 100 v
2 1 0 2.7766 2.7762 29 m’ 2.7762 2.7738 29 w1
3 0 0 2.4487 2.4484 46 s 2.4484 2.4485 46 s
0 0 2 2.4121 2.4210 31 vw 2.4196 2.4197 31 rn‘
2 1 1 2.4065 2.4062 49 s 2.4081 2.4067 49 m1
3 0 1 2.1835 2.1843 3 vw 2.1847 2.1861 3 vw
2 2 0 2.1206 2.1213 9 w 2.1203 2.1223 9 w
1 1 2 2.0968 2.0980 7 w 2.1016 2.1017 7 w
3 1 0 2.0374 2.0371 5 w 2.0372 2.0377 5 vw‘
2 2 1 1.9414 1.9436 2 vw 1.9421 1.9438 2 vw
3 1 1 1.8769 1 .8770 7 w 1 .8776 1 .8776 7 vw’
4 0 0 1.8365 1.8380 3 vw 1.8363 3 ~-
2 1 2 1.8210 1.8207 16 w 1.8241 1.8239 16 w*
3 0 2 1.7184 1.7186 31 m 1.7210 1.7207 31 m
3 2 0 1.6853 1 .6861 2 vw 1 .6851 1.6863 2 vw‘
4 1 0 1.6031 1.6032 12 w 1.6028 1.6039 12 w
2 2 2 1.5927 1 5911 7 w 1.5947 1.5959 7 vw
3 2 1 1.5910 ' 34 1.5914 1.5912 34 m‘
1 0 3 1 .5709 1 ~~~ 1 .5756 1 ~-
3 1 2 1 .5565 1 .5568 4 w 1.5584 1.5591 4 vw
4 1 1 1.5213 1.5215 6 vw 1.5215 1.5214 6 vw
1 1 3 1 .5036 1 .5037 7 vw 1 .5077 1 .5076 7 vw’
2 0 3 1.4733 1 4713 12 w 1.4769 1.4768 12 w
5 0 0 1 .4692 ' 2 1 .4690 2 ~~~
4 0 2 1 .4612 1 .4628 2 vw 1 .4627 2
3 3 0 1.4138 1.4139 3 vw 1.4136 1.4137 3 vw'
5 0 1 1.4055 1 .4052 8 w 1 .4057 1 .4063 8 vw
2 1 3 1 .3916 1 .3929 9 vw 1 .3947 1 .3948 9 vw“
4 2 0 1 .3883 1 .3894 2 _ vw 1.3881 2
3 2 2 1 .3815 2 ~- 1 .3828 2
4 1 2 1.3351 1 3339 12 w 1.3362 1.3367 12 w’
4 2 1 1.3341 ' 6 1.3343 1.3341 6 vw
5 1 0 1 .3194 1 .3204 2 vw 1.3192 2
5 1 1 1 2727 1 .2717 8 vw 1 .2728 1 2723 8 w

 

Table 19.

Miller indices. and observed and calculated interplanar d~spacings and

intensities for anti-Fe2P~type Baiz. *

 

 

n k l d¢(A) d,(A) 16 l, n k l d¢(A) d°(A) l. 10

1 0 0 7.917 1 -- 4 1 0 1.7277 1.7295 9 vw
0 0 1 5.173 5.197 3 vw 3 2 1 1.7137 17149 38 w.
1 1 0 4.571 4.560 3 vw 2 2 2 1.7127 ' 4

1 0 1 4.3306 4.345 2 vw 3 1 2 1.6739 1.6767 7 vw
2 0 0 3.9566 3 4 1 1 1.6367 1.6376 7 vw
1 1 1 3.4253 3.4249 95 s 1 1 3 1.6134 1.6176 10 wt
2 0 1 3.1437 3.1454 100 s 5 0 0 1.5634 15854 3 w.
2 1 0 2.9924 2.9935 44 m’ 2 0 3 1.5609 ' 13

3 0 0 2.6391 2.6390 42 m‘ 4 0 2 1.5719 4
2 1 1 2.5903 2 5908 39 m. 3 3 0 1.5237 4
0 0 2 2.5665 ° 34 5 0 1 1.5141 1.5162 9 vw
3 0 1 2.3506 2 4 2 0 1.4962 4

2 2 0 2.2655 2.2898 5 vw’ 2 1 3 1.4940 "4977 7 W
1 1 2 22511 1 3 2 2 1.4664 3
3 1 0 2.1956 2.1985 9 w' 4 2 1 1.4373 ”381 5 w.
2 0 2 2.1653 1 4 1 2 1.4367 ' 9

2 2 1 2.0906 2.0937 3 vw' 5 1 0 1.4220 3
3 1 1 2.0213 2.0257 10 vw‘ 5 1 1 1.3711 1.3732 9 vw
4 0 0 1.9793 1.9601 4 vw' 3 1 3 1.3562 3
2 1 2 1.9566 1.9600 26 w 5 0 2 1.3505 4
4 0 1 1.6466 1 3 3 2 1.3126 4
3 0 2 1.6472 1.6503 29 w‘ 4 2 2 1.2951 1.2961 5 vw
3 2 0 1.6163 3 -

 

Sarnpie from dehydration of Bela-H20 at 150°C in a 10‘ Torr vacuum.

82
Table 20. Structure types and lattice parameters of BaCl2. ‘

 

 

 

 

 

Structure
Type T (°C) a (A) b (A) c (A) wz (A’) Reference
PbCi, 25 7.672(1) 9.425(1) 4.7322(3) 67.76 93
Pnam 25 7.881(1) 9.419(2) 4.7296(6) 67.77 this work
412 7.679(2) 9.444(2) 4.734(2) 66.06 g
Anti-Fe2P 25 6.069(1) 4.616(1) 66.76 this work
P62m 12.5 6.103(6) 4.651(5) 66.17 _6_
150 6.113(6) 4.675(5) 66.63 g
200 6.132(7) 4.696(5) 69.63 g
‘ From Ba,,Sm,°Cl.. precursor.
Table 21. Structure types and lattice parameters of Basra. '
Structure
Type rec) a(A) b(A) c(A) V/Z(A’) Reference
PbCi, 6276(6) 9.919(6) 4.956(4) 101.7 (g)
Pnam 25 6266(2) 9.916(3) 4.956(3) 101.6 this work
Anti-FezP 25 6.4626(6) ‘ 4.624(1) 100.2 this work'
P62m

 

From 333L358!” precursor.

83
4.2.3 High-Temperature Forms of Mi, (M a Sr, Sm, Eu)

The iodides obtained by dehydration of Ml,+l,O at 120°C were finely divided
powders and were extremely hygroscopic. The anhydrous Srl2 prepared by dehydration
was pure by X-ray diffraction. The crystallinity of these specimens was good as
evidenced by the relatively narrow lines on the X-ray diffraction patterns. No oxidehalide
was observed. The structure of this new form of strontium iodide (Srl,~iV) was refined by
the Rietveld refinement procedure (§§) and will be described below. The three iodides are
lsostructural and their X~ray diffraction patterns are similar. Observed patterns are
consistent with calculated ones and are listed in Tables 2224. Some crystallographic
data of these metastable phases, together with those of other known modifications and
of the iodide monohydrates, are listed in Table 25 for comparison.

it can be seen from the V/Z data in Table 25 that the new forms of these iodides
prepared by low temperature dehydration are 2.5% (Eul,) to 3.8% (Sri2) less dense than
their normal forms. They therefore can be considered the high temperature forms. Wang
and coworkers conducted thermal analysis studies on the Sml2 system (98). Their
results revealed for Smi2 a reversible polymorphic transformation at 494°C. The high
temperature X-ray diffraction pattern (9_9) was not suitable for indexing, but appeared
to be the same as that of Sri,~lV. They also reported (m) that by dehydrating
Smi,+l,O or Smi,0.5H20 at 170-180°C for 70 min or at 210-240°C for 40 min in a 5x10"
Torr vacuum they obtained a new phase lsostructural with the orthorhombic form of Eula,
6.9. the normal form of strontium iodide, Sriz-l. However, the compound we prepared by
dehydrating Smi,+l,O at 170-180°C is the metastable form lsostructural with Sriz-lv. No
polymorphic transformation of this metastable form was observed even at 250°C in a high
vacuum within 3-4 hours. Thus the structure of the compound reported by Wang et 6!.

(M) is believed to be the same as Srlz-IV instead of being lsostructural with Sriz-i.

Table 22.

Miller indices. and observed (Guinier) and calculated interplanar d~spacings

and intensities for Sriz-IV.

 

 

1- k I 4.14) «1.111) I. I. II It I 4.01) 4.14) I. I.
1 0 1 6.953 6.958 7 w 4 2 0 1.9282 1.9279 17 w‘
2 0 0 6.183 6.209 <1 vw 6 1 0 1.90161.9011 12 w
2 0 1 4.981 4.984 9 w‘ 3 2 ‘2 1.8906 1.8891 19 m‘
0 1 1 4.255 4.256 10 w 4 2 1 1.8794 1.8796 4 vw
1 1 1 4.0237 4.0251 14 w’ 2 1 4 1.8456 1.8450 10 w
1 0 2 3.9801 3.9765 13 w 1 2 3 1.8313 1.8320 1 vw‘
2 1 0 3.8564 3.8506 49 8 2 2 3 1.7739 1.7744 1 vw'
3 0 1 3.7009 3.7058 3 vw 4 0 4 1.7382 17367 12 m
2 1 1 3.5053 3.5105 1 WI 5 1 3 1.7358 ' 2

1 1 2 3.0978 30901 100 vs 3 2 3 1.6891 1.6895 1 vw'
4 0 0 3.0913 ' 35 4 1 4 1.6394 ' 6 «-
3 0 2 2.9431 2.9417 36 s 7 1 1 1.6315 2 ---
4 0 1 2.9014 2.9032 6 w- 7 0 2 1.6285 1.6279 13 w
1 0 3 2.7332 2.7342 1 Wit+ 0 3 1 1.6140 1.6136 <1 vw'
4 1 0 2.6196 2.6188 13 w 0 2 4 1.5999 1.5993 14 m
2 0 3 2.5525 2.5530 2 vw 0 1 5 1.5916 1 ~-
3 1 2 2.5276 2.5246 56 vs 2 3 0 1.5893 15894 3 w
4 1 1 2.5010 2.5023 <1 vw' 4 2 3 1.5885 ' 2

0 2 0 2.4669 2.4653 29 m 1 1 5 1.5786 1.5794 <1 WV
0 1 3 2.4368 2.4380 2 vw 7 1 2 1.5464 1.5464 3 vw’
1 1 3 2.3909 2.3920 3 vw‘ 8 0 1 1.5202 15198 1 w,
1 2 1 2.3249 1 ~- 1 3 2 1.5200 ' 11

3 0 3 2.3176 2.3206 2 vw 8 1 0 1.4750 1.4750 4 w'
2 2 1 2.2106 2.2118 2 WV 4 3 0 1.4519 1.4521 2 w'
5 1 1 2.1382 2.1244 4 Wit 3 3 2 1.4357 1.4359 9 w
0 0 4 2.1019 21002 18 m . 7 1 3 1.4302 2 ~-
1 2 2 2.0968 ' 4 4 2 4 1.4209 1.4204 12 m
4 0 3 2.0763 2.0773 2 vw 6 1 4 1.4102 1.4100 8 w‘
3 2 1 2.0527 2.0548 1 WI 7 2 2 1.3591 ' 13 ~-
5 1 2 1.9567 1.9558 12 w 1 1 6 1.3400 1.3398 7 w

 

6

Sample was annealed in high vacuum at 250°C for 12 hr.

' overlapping with intemai standard Si reflections;

85

 

 

Table 23. Miller indices. and observed (Guinier) and calculated interplanar d~spacings
and intensities for Srl2~lV~type Smiz.
h k 1 4(4) 6.0)) I. I. h k I d.(A) 14(4) 1. I.
1 0 1 6.947 6.991 14 w 5 1 .2 1.948 1.948 5 vw‘
2 0 0 6.145 6.165 4 vw 4 2 0 1.916 1.921' 16 w
2 0 1 4.964 4.973 20 w 6 1 0 1.890 1.891 10 w‘
0 1 1 4.236 4.237 21 wt 3 2 2 1.882 1.880 15 w
1 1 1 4.005 4.011 28 w’ 4 2 1 1.868 6
1 0 2 3.983 1 ~- 2 1 4 1.845 3 ~-
2 1 0 3.832 3.834 18 w 6 1 1 1.844 2
3 0 1 3.684 9 ~- 1 2 3 1.826 2
2 1 1 3.488 4 2 2 3 1 .768 3
1 1 2 3.091 100 4 0 4 1.737 12 .
4 0 0 3.073 3'08“ 33 s 5 1 3 1.730 “732 4 w
3 0 2 2.936 2.930 30 w‘ 3 2 3 1.683 3 -~
4 O 1 2.886 2.893 10 vw 4 1 4 1.637 1.632 10 vw
1 0 3 2.736 3 ~- 7 1 1 1.622 4 ~-
4 1 0 2.603 2.605 20 w 7 0 2 1.620 1.619 15 w
2 0 3 2.553 4 -- 0 3 1 1.604 2 ~-
3 1 2 2.519 2.514 72 m 0 2 4 1.597 1.593 16 w
4 1 1 2.487 2 ~- 0 1 5 1.593 2
0 2 0 2.451 2.453 32 w‘ 1 3 1 1.591 2
0 1 3 2.436 4 4 2 3 1.582 3
1 1 3 2.389 2.386 8 vw‘ 1 1 5 1 .580 2
3 0 3 2.316 2.315 4 vw 7 1 2 1.539 3 -~
1 2 1 2.311 2 1 3 2 1.512 1.511 11 w’
2 1 3 2264 1 ~~~ 8 1 0 1.466 4
2 2 1 2.198 2201 5 vw 4 3 0 1.443 3 ~-
5 1 1 2.126 7 -~ 3 3 2 1.428 1.427 12 vw’
0 0 4 2.105 2.097 20 w’ 4 2 4 1.417 1.414 11 vw
4 0 3 2.072 4 ~- 6 1 4 1 .406 1 .403 7 vw
3 2 1 2.041 4 7 2 2 1352 1.351 16 w’

 

N018.

Observed lines are broad.

overlapping with one intemai standard Si reflections.

86

 

 

Table 24. Miller indices. and observed (Guinier) and calculated interplanar d~spacings
and intensities for Sri2~iV-type Eula.
II It I 41A) 4.14) I. I. II k I «1.14) 4.1/1) I. I.
1 0 1 6.911 6.928 14 w 6 1 0 1.8866 1.8883 10 w
2 0 0 6.136 6.142 4 vw 3 2 '2 1.8746 1.8760 15 m‘
2 0 1 4.947 4.948 20 w 4 2 1 1.8629 1.8646 6 vw
0 1 1 4.218 4.222 21 w‘ 6 1 1 1 .8404 2 ~-
1 1 1 3.989 3.988 28 w’ 2 1 4 1.8342 1.8348 3 w
1 0 2 3.958 1 ~- 1 2 3 1 .8169 2
2 1 0 3.822 3.817 18 m 2 2 3 1.7599 1.7627 3 vw‘
3 0 1 3.674 3.676 9 vw 4 0 4 1.7277 1.7276 12 m
2 1 1 3.476 3.480 4 vw 5 1 3 1 .7236 4 ~-
1 1 2 3.075 3 068 100 vs 3 2 3 1.6759 1.6764 3 vw'
4 0 0 3.068 ' 33 4 1 4 1.6288 1.6298 10 w
3 0 2 2.924 2.925 30 s' 7 1 1 1.6189 16176 4 m‘
4 O 1 2.880 2.882 10 vw 7 0 2 1.6168 ‘ 15
1 0 3 2.718 2.723 3 vw 0 3 1 1 .5984 2 ~-
4 1 O 2.598 2.601 20 m 0 2 4 1.5883 1.5889 16 m
2 0 3 2.532 2.543 4 vw 1 3 1 1.5850 2
3 1 2 2.509 2.508 72 s 0 1 5 1.5824 2 ~-
4 1 1 2.481 2 ~- 4 2 3 1.5761 1.5725 3 vw'
0 2 0 2.4426 2.4414 32 m1 1 1 5 1.5694 2 ~-
0 1 3 2.4212 2.4246 4 vw 7 1 2 1.5349 1.5378 3 vw
1 1 3 2.3754 2.3753 8 w 8 0 1 1.5089 2
3 0 3 2.3036 2 3086 4 w‘ 1 3 2 1.5059 1.5067 11 m‘
1 2 1 2.3030 ' 2 4 0 5 1.4696 2
2 1 3 22522 22560 1 vw 8 1 0 1.4635 1.4654 5 vw
2 2 1 2.1902 2.1932 5 w‘ 4 3 0 1.4383 1.4388 3 vw'
5 1 1 2.1214 2.1244 7 w 3 3 2 1.4227 1.4237 12 w
0 0 4 2.0908 2.0896 20 m 7 1 3 1.4203 3 --
4 0 3 2.0631 2.0657 4 vw 4 2 4 1.4105 1.4118 11 w
3 2 1 2.0342 2.0385 4 vw 6 1 4 1.4006 1.4025 7 w
5 1 2 1.9422 1.9433 5 w 7 2 2 1.3482 1.3498 16 w
4 2 0 1.9109 1.9115 16 w1

 

Note. The sample was annealed in a high vacuum at 310°C for 6 h.

87

 

 

 

 

 

 

 

Table 25. Comparison of selected crystallographic data of Mi2 and Mi,+l,O.
Compound Structure' 6 (A) b (A) c (A) 6 (°) wz (A’) Reference
P n m a 12.474(2) 4.495(1) 9.741(2) 136.6 6_5
Sri2~HZO
12.501(3) 4.473(1) 9.740(2) 1362 this work
Srlz-IV 12.365(2) 4.9336(6) 6.406(1) 126.2 this work
Srl2 Srl2~i 1522(6) 622(3) 790(3) 123.5 22
Srl2~ll 1123(2) 8.889(10) 4.630(6) 115.5 a
Smlz-HZO P n m a 12.472(3) 4.510(1) 9.711(3) 136.6 this work
Srlz-lV 12.290(8) 4.902(3) 6.421(7) 126.6 this work
Sml, Eula-l 7.623(2) 6260(2) 7.903(2) 9792(2) 123.5 this work
Srl,~ll 11.208(6) 6.666(5) 4.579(4) 114.0 a
Eula-i120 P n m a 12.414(2) 4.4667(7) 9.696(2) 135.0 this work
Srlz-lV 12.272(2) 4.6652(6) 6.363(1) 125.4 this work
Eula-l 762(2) 623(2) 7.88(2) 98.0(5) 122.3 22
we Srlz-l 1512(3) 8.18(2) 7.83(2) 121.0 8_3
Srl2~ll 11.184(3) 6.693(2) 4.575(1) 113.6 g

 

’ Note. The structure types are:

(1)
(2)

l3)

(4)
(5)

Sri,~i, normal form, orthorhombic, space group P b c a (#61);

Srlz-li, high-pressure form. PbClz-type structure (orthorhombic. space group
P b n m (#62);
Sria-lli, high-pressure form, probably orthorhombic. Structure unknown.
See reference g5;
Srla-iv, metastable form. orthorhombic. space group P n m a (#62);
Eul,~i, normal form. monoclinic. space group P 2,/c (#14).

88
4.3 The Structure of Srlz-lv by X-ray Rietveld Refinement

Data collection and Rietveld refinement details are summarized in Table 26. The
observed. calculated and difference patterns are presented in Figure 12; positional and
thermal parameters are listed in Table 27. Estimated standard deviations (ESD's) are
calculated according to Scott L191) and are 5-7 times larger than those reported by
the XR682 system in the absence of this option. Selected Sriz-iv and Sr|2~l (norrnai form)
bond distances are compared in Table 28.

The Srlz-iv structure is centrosymmetric. Refinement could not be effected in non-
centrosymmetrlc space group Pn2,a; the z coordinates exhibited large shift/error ratios
even when the x and y coordinates were refined initially. The Srlz-lv lattice parameters
derived from Guinier fllm data are slightly larger than those derived through the Rietveld
reflnement. Generally the lattice parameters refined in the Rietveld procedure are
believed to be more accurate. However, these parameters in our case were not stable
and the standard errors were exceedingly small. presumably due to program errors. Since
the Guinier data are referenced to NBS certified Si, these values are considered more
accurate in this work.

A significance test based on 5 factors with 2623 observations. the number whose
intensity is 20 counts above the sum of background and the standard deviation of the
count rate at that position. suggests that the iodine anisotropic temperature factors are
significant at greater than the 95 % confidence level (192).

An ORTEP drawing of the monocapped trigonally prismatic coordination of Sr is
shown in Figure 13(b). and the atomic packing is presented in Figure 14. The Sr
coordination polyhedron in Srl2~iv is similar to that of Sr in Srlz-HzO (g5) when the water
molecules are ignored (Figure 13(a)). It is also similar to the metal coordination
polyhedron in the normal form of Mi, (Figure 13(c)) (M a: Sr. Eu) (Srlz-l type) w. and

89
in the monoclinic Eui, structure ($1), but with one major difference. In the Srl,~lv

coordination polyhedra (Figure 13(b)) a mirror plane coincident with the lattice mirror plane
passes through atoms l(1)". l(2)". l(1)" and stand relates l(1)' to K1)" and l(2)' to I(2)'; the
cation coordination polyhedra of the Srl,-i and monoclinic Eui, structure types are devoid
of symmetry. This mirror plane. which is perpendicular to the 9 axis, is apparent in Figure
14(a).

The average Sr-l bond lengths of Sri,-lv and Srl2~i are nearly identical. However,
the iodine atoms do not relax to fill completely the coordination sites vacated by the H20.
Thus the metastable form is less dense than is Sriz-i (4.45 vs. 4.59 g cm4 respectively).
The openness of the structure is also reflected in the molar volume, 77.3 crn’lmol, a value
close to that of the monohydrate, 82.24 cm’lmoi. Such packing inefficiency might be
expected because low preparatory and annealing temperatures mitigate against atomic
rearrangement. The capping iodide in Srl,-lv, l(1)". has a rather long bond distance in
comparison to the corresponding Srl,-l distance. The corresponding bond in the hydrate
is opposite to those of coordinated water molecules. its elongation in the hydrate is
apparently due both to the decreased acidity of the strontium atom because of the H20
bonding and the more crowded coordination sphere that results. During the low energy
dehydration all the Srl,~lv atoms remain locked in the 49 special positions just as they
were in the monohydrate; the minor relaxations that occur are insufficient to revert the
lattice to the more efficiently packed Srla-i.

Two types of anion coordination are present in the structure. Atom l(1) is
coordinated in a distorted tetrahedral arrangement to four cations. whereas atom l(2) is
coordinated in a capped trigonal arrangement to only three. These can be visualized

clearly in Figure 14(b).

90

2004

K)
0
A

 

 

COUNTS :1 l0'3

 

 

 

 

 

 

 

8

 

 

 

 

CMTS x l0"

2

 

 

Figure 12. Observed (a). calculated (b), and difference (6) X-ray diffraction patterns

for Srla-lV.

91

 

 

 

 

 

 

 

 

 

 

 

 

Figure 13. Comparison of several metal coordination polyhedra. (a) Sri2+l20 with H
atoms omitted; (b) Srlz-IV; (c) Srl2~i (normal form).

92

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 14. C-axis projection of the Srl,~lv structure illustrating the atomic packing.
(a) mirror planes and layer character are shown clearly; (b) the structure
is projected and then rotated along the 6 axis to reveal the coordination
polyhedra.

93

 

Table 26. Crystallographic and Rietveld refinement data for Sriz-lv.

Pattern 20 (deg) 10-85

Step size 20 (deg) 0.020

Count time (sec/step) 20

Standard peak: hki (20) 210, (2310“)

R (standard peak) 0.036

Space group (number), Z £11352), 4

Systematic extinctions _i1k0: D. = 21.1.”
O|_(_I: 5+1 = 211+1

g (A) 12.328(2)

.12 (A) 4.920(1)

g (A) 6.364(1)

Volume (A’) 509.3

Density (D,, 9cm") 4.452

Number of observations 3682

Number of reflections 196

No. of refined structural parameters 19

No. of refined profile parameters 9 (incl. 2, 9, g)

B, = Zia-Fajita 0.069

B " 2“(1:01" 691/2132 0087

6. ~ (summed/211.1212)“ 0.188

B... =i£mlm(9.l-xi(£)/9.l’/Ev_vm(2)’l"’ 3184

Maximum shift/error

0.09

 

94

 

 

Table 27. Positional and therrnai parameters for Srl2~lv.

Atom 5 x 9.11" 9.22” flee" lire”
sr 0.170(1) 0.25 0.130(5) 11(4) ‘

l(1) 0.392(1) 0.25 0.377(3) [23(4)]" 20(9) 27(11) 27(11) 14(14)

l(2) 0.380(1) 0.25 0.883(3) [32(5)]°

37(11) 40(13) 17(12) 13(15)

 

x103 (A’).

from isotropic refinement; form: exp-{81:2Q(sin8/A)’}. . .
° Anisotropic factor form: exp[~(h_"’6,,+..+2h_kl3.2+..)]; l‘ill = (212% g, ).

 

 

Table 28. Comparison of selected Sri2~lv and Srl2~l bond distances.
Srl2~lV Bond distance (A) Srl2~l Bond distance (A)
Sr~l(1)' 334(3) Sr-l(2)" 3.369(6)
Sr~l(1)'I 334(3) Sr-l(1) 3268(10)
Sr-l(2)' 331(2) Sr-l(2)" 3.339(9)
Sr-l(2)' 331(2) Sr~l(1)' 3256(9)
Sr~l(1)" 343(3) Sr-l(2)‘ 3.415(9)
Sr-l(2)"' 331(3) Sr~l(2) 3.417(6)
Sr-l(1)" 344(2) Sr(1)' 3.390(1 1)
Average

Sr~7xi 3.35 Sr—7xl 3.351

l(1 )-4xSr 3.38 l(2)-4xSr 3.385
l(2)-3xSr 3.31 l(1)-3xSr 3.305

 

95
4.4 Thermal Stabilities of Metastable Phases

The fluorite modifications of EuCI2 and SmCl, are definitely stable toward phase
transformations at ambient temperature. After 9 h at 300°C only a very small fraction of
the fluorite forms of both chlorides had transformed. Even after 11 h at 360°C or 4 h at
430°C only a portion of the EuCl2 specimen had transformed to the normal PbClz-type
structure. At 440°C the fluorite form of SmCl2 (blue) changed quantitatively into the
thermodynamically stable low-temperature (red) form within hours.

Similarly transformation of the 'high pressure' forms of Bax, (X=CI. Br) occurs
only at temperatures far higher than the solvolytic decomposition temperatures. The anti
FeaP-form of BaCI, at 350°C and that of BaBr, at 300°C evidenced no obvious change in
6 h in a vacuum of 10‘ to 10‘7 Torr. At 400°C about 90% of the former transformed into
the PbCla-type (the thennodynamically stable form) within 5 h. and about 50% of the latter
transformed in 7 h. All these results conflnn that the new forms are metastable phases

and their formation is kinetically controlled.

4.5 Soivolytlc Decomposition Intermediates

We observed formation of a crystalline inbn'nedate in our early studies (12) when
some Ba.Ln,Br” (Ln-La. Nd) precursors were examined after 30 min of extraction by
THF. Later experiments revealed similar behavior in other systems. it is clear now that
the intermediates are actually Ianthanide trihalide THF complexes.

The triclinic intermediate we reported Q2) is NdBr,(THF)‘. This composition was
determined by the weight loss of the complex when it was heated gradually at 130°C for
many hours in a high vacuum and the product was identified as the normal form of NdBr,.
Its powder X-ray diffraction pattern is identical with that of a sample prepared by
contacting anhydrous NdBr, with THF, and then evacuating to dryness. With 1-3 9 of

96
precursor containing LnBr,. the complex was apparently observable in the X-ray diffraction

patterns after 3 h of extraction. Miller indices. observed intensities. observed and
calculated interplanar d-spacings for NdBr,(THF)4 are listed in Table 29.

The EuCl,(THF),, (1<n<2) complex that formed after 30 min THF extraction of
EuCIm mixed-valent chlorides containing free EuCla had the same X-ray diffraction pattern
as that formed between pure EuCl, and THF. This compound was not observed by X-ray
diffraction after 6 h. When a pure precursor Eu4Cl, was used the complex was not
observable even after 30 min extraction. However, solid state FT IR spectra of these
specimens indicated a trace of the THF complex even after 5 h extraction. Fifteen hours
later it was not visible by IR. A typical IR spectrum of EuCl:,(Tl-ll'-')n is shown in Figure 15.

The mixed-valent chlorides LnClm (Ln-Sm, Eu) were decomposed quickly by
THF; mixed Ba.Ln,Br,3 was also decomposed quickly, but less rapidly than the LnClm.
Extraction of BaCl,-LnCl, mixed halides proceeded much more slowly. The color change
from blue or black to white when EuClM was leached became apparent almost instantly.
When the specimens were examined by x-ray diffraction. no trace of the mixed-valent
precursor EuClM was observed even after 30 min extraction while the mixed halide
precursor Ba,,Sm,°Cl.. disappeared only in a course of 5-6 days. In all the systems

examined no trace of any other new phase was observed.

97

09V

3.3 3865 .9: 6:505 .6 53586 EL .2 2:9...

“alEUC LODEDCI>03

 

 

 

 

 

 

 

 

 

 

 

000 OONd DOW.“ OOON OOVN 000mm Comm” 00C“ 000'
I4. + i- .- (i ivl + l+|.i.ililT|-:-- (Til. :
.3 ..
a a
L.- a
'
‘
t.
u
#1 c U
a C
W
o!
.4
a.
U
w
45'
a C
#1 g
B
a. 2
w? "mm ”mm mm LQ< on C m.“ EIP\ “HUJUN.NHUJNM vb

98

 

 

Table 29. Miller indices. observed (Guinier) intensities, and observed and calculated
interplanar d-spaclngs for triclinic NdBr,(THF).; a - 6.352(1), b =
15.915(3). c = 9.429(1) A; 01 a 100.02(2), B = 105.26(2), and y . 87.54(2)°.

11 11 k 1 dc(A) d°(A) 1° 11 h k 1 dc(A) d°(A) 1°

1 0 1 0 15.672 15.696 wt 27 0 3 1 4.2149

2 0 0 1 6.9660 6.9626 s 26 -1 -2 2 4.1666 4.1707 vw

3 0 -1 1 6.4190 6.4063 s 29 0 1 2 4.1296 4.1306 vw

4 1 0 0 6.0576 6.0436 vs 30 -1 1 2 4.0962

5 0 2 0 7.6362 7.6352 vs 31 -2 0 1 4.0942 4.0941 m'

6 0 1 1 7.2714 7.2727 m 32 1 -3 1 3.9426

7 -1 1 0 7.1750 7.1675 vw‘ 33 -1 3 1 3.9421 3.9414 m‘

6 1 1 0 7.1573 34 92 1 0 3.9049

9 -1 0 1 6.9595 6.9579 w’ 35 2 1 0 3.9992 3.9046 vw

10 -1 -1 1 6.6670 6.6694 117* 36 -2 1 1 3.6915 3.6930 vw

11 0 -2 1 6.4656 6.4666 w 37 0 .4 1 3.6357 3.6369 vw’

12 -1 1 1 6.0777 6.0732 m‘ 36 -2 -2 1 3.7442 3.7441 vw

13 -1 2 0 5.6263 39 -1 -3 2 3.6976 3.6976 vw

14 1 2 0 5.6092 5.6123 s 40 -1 -4 1 3.6166

15 -1 -2 1 5.5706 5.5704 vw‘ 41 -1 2 2 3.6139 3.6151 w'

16 0 3 0 52241 42 -2 2 0 3.5675 3.5665 w

17 1 -1 1 5.2230 52230 w‘ 43 1 -1 2 3.5631 3.5619 w

16 1 1 1 4.9054 44 -2 2 1 35234

19 -1 2 1 4.9006 4.9020 w’ 45 1 4 0 3.5194 3.5212 w

20 0 -3 1 4.6669 46 1 -2 2 3.4074 3.4056 w‘

21 1 -2 1 4.6414 4.6446 w 47 -2 -3 1 3.3449 3.3476 vw

22 0 -1 2 4.5164 4.5166 w 46 -2 -2 2 3.3435

23 0 0 2 4.4630 4.4651 vw‘ 49 -2 3 1 3.1126 3.1121 vw’

24 -1 -1 2 4.4664 4.4662 vw'- 50 -2 -3 2 3.0646 3.0643 vw

25 -1 -31 4.4621 51 1 2 2 3.0761

26 1 2 1 42174 42175 w’ 52 -1 0 3 3.0746 3.0750 vw

 

99

 

 

Table 29. (continued)

11 h k 1 d.(A) d,(A) I. 1: 11 k 1 dc(A) d,(A) 10
53 -2 2 2 3.0369 72 02 3 2.6476

54 0 -1 3 3.0314 3.0332 vw’ 73 -2 -3 3 2.6196 2.6197 vw’
55 -1 -5 1 3.0062 3.0074 vw’ 74 1 -1 3 2.6193

56 0 0 3 2.9667 2.9663 w 75 1 0 3 2.5911 2.5912 vw+
57 0 -2 3 2.9640 2.9651 vw 76 1 5 1 2.5900

56 -2 -4 1 2.9415 2.9409 w 77 2 1 2 2.5696

59 0 1 3 2.6466 2.6463 w 76 -3 2 1 2.5797

60 0 5 1 2.6146 79 1 -2 3 2.5759 2.5766 vw
61 -2 4 0 2.6132 2.6129 w 60 -1 -6 1 2.5556 2.5556 vw‘
62 0 -3 3 2.6063 2.6060 vw 61 3 2 0 2.5364 2.5392 vw
63 2 4 0 2.6046 62 -2 -5 2 2.4936

64 -1 -5 2 2.7667 2.7676 vw 63 -2 2 3 2.4930 2.4929 vw
65 -2 -4 2 2.7654 64 2 -3 2 2.4639

66 -2 -2 3 2.7477 2.7464 vw‘ 65 1 6 0 2.4626 2.4630 w
67 2 3 1 2.7464 66 2 4 1 2.4725

66 -2 3 2 2.7364 2.7362 vw 67 1 -3 3 2.4707 2.4712 w’
69 -1 2 3 2.7079 2.7065 vw 66 -3 -3 2 2.4407 2.4405 vw
70 -3 -2 1 2.6609 2.6610 w' 69 -3 2 2 2.4196 2.4203 w
71-3 1 0 2.6487 2.6492 vw 90 -1 -5 3 2.4129 2.4119 w

 

CHAPTER 5 DISCUSSION

SYNOPSIS
Synthesis conditions are discussed. Reactions that yield metastable phases are
analyzed in thermodynamic and kinetic terms and are shown through layer descriptions
to be topologically controlled with minimal structural reorganizations. The low-temperature
high-vacuum dehydrations do not involve major structural changes while changes that
occur during soivolytic decompositions are significant. ln-layer atomic rearrangement of
the intermediate lattice leads to formation of the fluorite-type structure; further out-of-layer

displacement yields the anti-FezP-type sthcture.

100

1 01
5.1 Soivolytlc Decomposition

5.1.1 Energetics

As has been pointed out in Chapter 2, for a soivolytic decomposition reaction that
produces a thermodynamically less stable form to occur, the leaching solvent must form
a strong complex to compensate for the loss of lattice energy. On the other hand. the
stability of the metastable form relative to that of the mixed precursor is also important.
The inability to prepare NdCl2 by solvolytic decomposition of NdClm could be attributed
to instability of the assumed fluorite form of NdCl2. This fluorite form would be the
expected direct product of a leaching experiment because of the similarity of structural
properties of neodymium compounds and those of the samarium and europium analogues
(ionic radius (26): Nd", 1.43; Sm“, 1.41; Eu”, 1.39 A). The normal form of NdCl2 (Peel,-
type), even though preparable by neodymium or lithium reduction, is apparently much less
stable than that of LnCl, (Ln :- Sm, Eu. Yb). The fluorite form of NdCla is thus expected
to be highly unstable at ambient conditions. The low solubility of the NdClaxTHF solvate
in THF, another energetically unfavorable factor, could also contribute to the difficulty of
hIdCl2 formation through leaching.

Contamination of mlxed(-valent) halide precursors by oxidehalide impurities may
also hinder soivolytic decompositions. Many of the mixed halide precursors have cluster-
type structures. Given the moisture sensitivity of M",Ln"*5x,, phases and the formation
of oxidehaMe phase. Nd’*.Nd’*.Cl,20, whose structure is almost indistinguishable from
that of the Nd..X,. phases, the oxidehalide impurities in the mixed(-valent) halides must
be considered. Oxidehalides are expected to have higher lattice energies than those of
the mixed halides because Mao bonds are stronger than M-Cl bonds. In the oxidehalide

structure oxygen atoms occupy the cluster centers Q3) where the structure undergoes

102
significant changes upon soivolytic decomposition. Therefore the M.Ln,X,,O structure is

probably too difflcult to be decomposed by THF. The fact that the 'Sm,Gd,CI,;
specimens prepared by the ceramic method could be only partially extracted by THF
probably resulted from contamination by oxygen. That is, the specimen may have
contained Srn.Gd.Cl,,O or Sm.Gd,CluO. Oxygen could come from the quartz container.
For the various samarium mixed chlorides prepared by hydrogen reduction. the
formulation Sm,M.Cl,, (M . Gd and Nd) was based upon X-ray and mass balance data
since GdCI, cannot be reduced to a discrete Gd" ion even when elemental Gd is the
reducing agent. NdCl,, on the other hand, can be reduced to NdCI2 with strong reducing
agents Qflm but not with hydrogen (105). Oxidehalide contamination was
probably a problem, especially in the SmCl,-NdCla system in which a substantial amount
of the 'M,.X,; phase was unbreakable even though fluorite-type SmCl2 did form.
Based on these considerations, efforts should be devoted to the elimination of
oxidehalldes during the preparation of mixed or mixed-valent halides. Oxygen- and water-
free conditions are essential. Lower temperatures and shorter heating times also tend to
minimize oxidehalide contamination possibilities. For the europium mixed-valent chlorides
the various methods used. partial reduction with hydrogen. thermal decomposition. and
even solid state reactions between EuCl, and EuCl,. did produce fairly pure compounds
that could be decomposed completely by THF. For the samarium mixed chlorides.
hydrogen reduction of the SmCl,-GdCla 1:1 mixture gave the best result. The barium
mixed bromides were not obviously subject to oxygen contamination. But the barium
chloride precursors. especially the BaClz-LaCl, system that required reaction temperatures
above 800°C, may have contained some oxidechloride since complete extraction was not
achieved. BaCl, is known to attack quartz at high temperatures (3). 8a,,Sm,°Cl.. is a

better precursor than “BaeLaCI7'.

1 03
5.1.2 Solvent Effect

The choice of an appropriate solvent is important to the success of a soivolytic
decomposition reaction. However, this choice is usually severely limited by solubility
requirements. THF is a good solvent for the Ianthanide chlorides and barium bromide.
The ianthanide trichlorides are soluble in THF and are expected to form more stable
complexes with THF than with pyridine. The reason that the barium mixed chlorides can

be leached by pyridine but not by THF is unknown.

1 04
5.2 Low-Temperature Dehydration

To monitor the degree of dehydration higher hydrates were always converted first
to monohydrates. This step proceeded easily at temperatures lower than 80°C under a
dynamic vacuum. Removal of the last mole of water from monohydrates usually requires
temperatures higher than 80°C. Previous thennal studies had shown that the sump
to Srl2 conversion occurs at 280°C or even higher temperatures under normal pressure
(£6). Under the high vacuum conditions of our experiments the dehydration
temperature decreases as low as 110°C. At this temperature loss of water is much slower
than at 280°C, but dehydration was usually complete within 24 h. The process may be

summarized as follows:

 

vacuum 103-10" Torr
Mlz-tzo (s) --------> MIZHZO (S) > Ml2 (s).
overnight 110-125°C / 24 hr
(1<x<4) (MaSr,Sm,Eu)

Such a low temperature or a high vacuum is not always necessary. For example,
Eula-H20 can be dehydrated at 220-230°C in a 10" Torr vacuum in 24 h; Smlz-Hzo can
be dehydrated at 170-180°C in 10" Torr within about 1 h (98). However, it is important
to perform the experiment at a lower temperature to insure that the reaction proceeds
slowly. Slow removal of water molecules with minimal disturbance of the crystal lattice
is the key to the synthesis of a metastable modification. Fast dehydration at high
temperatures can lead to a mixture or to formation of a different modification from that
obtained by slow dehydration (8). In addition, low temperature dehydration in a high
vacuum will prevent the following hydrolysis reactions that may produce oxidehalides such
as EuOCI and SrpCl. (g).

sz-x H20(s) heat

 

> MOX, M4OX. (5) etc. + HX (g)
or MX2(s) + H20 (9)

105
5.3 Metastable Structures and Low Temperature Decompositions

Fluorite-type modifications of SmCl, and EuCl2 were observed previously at
elevated temperatures by X-ray diffraction (107.108). evidenced transformation
temperatures of 76814 (M) and 745d:5°C (up). respectively, and were not
quenchable. The anti-FezP-form of Bax2 (X=Cl, Br) prepared by low-temperature
soivolytic decompositions could not be synthesized at high temperatures under high
pressures. Nor is the unquenchable high temperature form of Smlz. also observed by high
temperature diffraction above the polymorphic transformation temperature (9_9_). X-ray
diffraction pattern of a Eul, specimen heated in a sealed quartz tube to near melting
(520°C) and quenched in liquid nitrogen gave no trace of the metastable form even though
the metastable form can be synthesized by low temperature dehydration. In view of the
need to have stable specimens at ambient temperature to study physical properties, it is
significant that the low temperature soivolytic decompositions and dehydrations can both
provide convenient mild-condition synthesis routes for some known metastable
modifications. and provide possibilities to synthesize compounds not easily accessible by
other means.

It was mentioned briefly in Chapter 1 that dehydrating BaCl2-2H20 could yield the
metastable 'high-pressure' (anti-FezP) form of BaCL. Since in this reaction the metastable
form was observed together with the high—temperature form (8). dehydration is not an
approprlate synthesis method for anti-Fe,P-type BaClz. On the other hand. dehydration
is a unique low temperature approach (91) to the high temperature form of BaCl2. Our
room-temperature high-vacuum dehydration experiment with BaCL-ZHp yielded only the
high-temperature form (CaF2-type). in conclusion. dehydration of BaCl2-2H20 favors the
formation of the fluorite-type high-temperature modification of Back. while soivolytic

decompositions on mixed halide precursors uniquely allow synthesis of the high-pressure

106

form.

A soivolytic decomposition reaction requires that the two components in the
precursor have distinctly different solubilities in the solvent. For this reason soivolytic
decomposition becomes difficult or impractical for bromides and iodides since they are
soluble. For these systems low temperature dehydration or more generally, low
temperature desolvation provides an alternative way of achieving the same goal. Slow
removal of solvent molecules from a solvate lattice during desolvation resembles the slow
removal of the soluble species from the lattice of a mixed precursor in a soivolytic
decomposition. In fact. fluorite-type EuCl2 was observed during dehydration of EuCl, H20
(9) and the pure “high-pressure“ form of BaBr2 could be prepared by dehydrating
BaBrz-Hzo. Therefore, the two methods are complementary. The combination is

applicable to many more systems than is either individual method.

107
5.4 Formation Mechanism of the Metastable Forms

5.4.1 Kinetics vs. Thermodynamics

These metastable structures are stable indefinitely at room temperature in the
absence of moisture and oxygen. As the thermal results demonstrate. polymorphic
transformations into the normal forms take place only at temperatures above 300°C. 0n
the other hand. formation of metastable forms ls relatively fast at the reaction
temperatures (BS-125°C) (see section 4.5). These data indicate fairly low activation
energies for the precursor to metastable MX2 conversions; but high activation energies for
the metastable to normal form conversions. Therefore. the metastable forms are trapped

as kinetically-controlled products in an energy well.

5.4.2 Topochemlcal Transformations

Thermodynamic considerations enable us to understand why the metastable
modifications are formed. but do not tell us how they are formed and why the precursor
to metastable MX, conversions have low energy barriers. We must examine our systems
at the microscopic level to answer these questions.

At first glance one might speculate that metastable dihalide formation during
soivolytic decompositions is a simple dissolution-reprecipitation process. That is. the
divalent and trivalent cations with associated anions dissolve together and then the
dihalide salts precipitate from solution as metastable solids. Such a mechanism was
postulated for the formation of y-MnO, when Mn,0. was leached by a mineral acid such
as l-lCl(aq) (_2_§). Mn,0. dissolves into Mn2+ and Mn" ions with the latter
disproportionating into Mn" and y-Mno, (6). However. such a dissolution-reprecipitation
mechanism for the halides contradicts other experimental results. When SmCI3(Tl-IF),, is

108
reduced in THF, the precipitate, SmCI, (s). is the normal (red) (19) rather than the

metastable (blue) form. Thus there is a major difference between the mixed halides and
Mn,0.. Dihalide salts are only slightly soluble in the solvent, while in Mn,0. both
components (MnO and Mn,0,) are soluble. Based on thermodynamic considerations. the

following equilibria should exist between the solids and the solution:

MX2.nLnX, (s)

(1) Solvent (L)

 

f
LnXa-Lx
+
(2) (3)
iiilx2 (s) a M2*(L) + 2X'(L) =2 sz (s)
metastable normal

Were dissolution (1) to occur, the dlhallde MX, would have to dissolve rapidly since
formation of the metastable MX2 is fast. This is unlikely. if not impossible. because of the
very low MX2 solubility (thus a very small solvatlon energy), and the need to destroy the
entire precursor lattice which would require a high activation energy. Were dissolution of
MXz rapid and precipitation reaction (2) faster than (3), one would expect a gradual
conversion of the metastable form (kinetic product) into the thermodynamically more stable
(normal) form. This has not been observed. In addition, if dissolution-reprecipitation is
involved, the precipitate should have poor crystallinity because a low temperature rapid
precipitation precludes annealing; and the unit cell dimension of the metastable form
should be independent of the dimension of the mixed halide precursor. In fact. the

109
metastable modifications are well crystallized as evidenced by the sharp X-ray reflections.

It took only about 30 min for ~1 g Eu.Cl, (s) to be decomposed completely into metastable
EuCl, (s) and EuCl3(THF),, by THF; no trace of the normal form of EuCl2 (s) was observed
as'a result of 24-48 h continuous extraction. In the Bax2 (X . Cl, Br) systems, even
though trace amounts of normal forms were observed in the leached product. the fact that
lattice parameters of the metastable forms are related to the type of mixed-halide
precursor used (Page 117) argues strongly against a dissolution-reprecipitation
mechanism.

On the other hand, if the structural changes upon soivolytic decompositions are
taken into account, one finds that experimental results can be explained satisfactorily. An
important feature of these metastable phases is that they have symmetries higher than
those of the corresponding normal forms. The highest symmetry element, the 3-fold
inversion axis in the fluorite-related precursors. was retained in both the high-temperature
forms of LnCI2 (Ln - Sm, Eu) and the high-pressure forms of Bax, (X . Cl, Br). It is
apparent that the structures of the metastable forms and those of the precursors are
cloSely related. These relationships can be explained in topotacticchemistry terms. The

various reactions are discussed separately below.

1 10
a. Srlz-l-lzo to Srlz-IV Transformation

The three iodide monohydrates Ml,+l,O (M . Sr. Sm, Eu) are lsostructural; and
so are the metastable iodides Mlz. To demonstrate the structural changes involved in the
monohydrate to anhydrous metastable form transformation. projections of Srlz-HZO and
Srlz-IV lattices down the shortest (b) axis are shown in Figure 16 and atomic positions of
SrI,-Iv and Srl,-l-I,O are compared in Table 30.

The Srl,-H,O to Srl,-IV transformation is a relatively straightfonlvard topochemical
conversion because only water molecules are removed from the precursor lattice. As can
be seen in Figure 16. slow removal of coordinated water molecules leaves the frame
structure intact except for minor atomic shifts. This can also be seen in Table 30. In fact,
it is the topochemical nature of the reaction that allowed us to use Srl,-l-l,O as a model
structure for refinement of the SrI,-lv structure.

A significant feature of this topotactic reaction Is retention of the mirror planes
perpendicular to the b—axis. one of which contains atoms Sr', l(1)“. l(2)" and l(1)” in Figure
13 (Page 91), despite the sizable b axis expansion and c axis shortening which resulted
from collapse of the parent lattice upon dehydration. Since atomic movements are
restricted to the planes, the Srlz-lv atoms remain locked in the 4c special positions in the
space group as they were in the monohydrate. Since all symmetry transformations are
retained. the space group remains unchanged.

Similar features were observed in the topotactic dehydration of the layered oxide
hydrates MO,-l-l,O (3,4). Both Srl,+l,O and Srlz-lv also can be viewed as layer-like
structures (Figure 14(a). Page 92). The major difference is that the interlayer distance (b
axis) of the iodide increases while that of the oxide decreases upon dehydration. a result
apparently attributable to the mobility of atoms in the dominantly ionic iodides.

111

.of "89.0 9.8 cm 822.0 code .35 3. ”8.2.0 code 093
83558 a as 3865 685:2 6.3 e as See masses 3. 2-6m Be 3 0.5.5 o... 5 58%...

.3

E

.9 2.5m

 

112

Table 30. Comparison of atomic coordinates of Srlz-HZO and Srlz-IV.

 

Compound Srlz-Hzo (85) Srlz-IV

 

 

 

 

 

Space group ana ' ana
Z 4 4
x 0.1967(1) 0.170(1)
Sr y 0.25 0.25
0.1208(2) 0.130(5)
x 0.10358(7) 0.108(1)
l(1) y 0.75 0.75
-0.1 1399(8) 0123(3)
x 0.10784(7) 0.120(1)
l(2) y 0.75 0.75

0.35664(8) 0.383(3)

N

 

113 .
b. Ln,.Cl,a to Fluorite-Type LnCl, (Ln = Sm. Eu) Transformation

The fluorite-type structure of 1.an and the fluorite-related structure of Ln2*,Ln°*5X”
have been described in Chapter 2 where we emphasized layered hexagonal close packing
of the cations. It was shown that introduction of the “extra“ anions that accompany
trivalent cations leads to clustering. In a soivolytic decomposition, the major chemical
change is the selective removal of these trivalent cations and accompanying anions from
the mixed valent halide lattice. Thus the clusters are the decomposition centers.

When the trivalent cations and associated anions are removed from the layers so
gently at low temperatures (Figure 17(a)) that the rest of the lattice is perturbed only
slightly, we can assume that the intermediate (Lnx2) lattice (Figure 17(b)) is comprised of
Ln2+ and Cl layers with a large number of vacancies. This intermediate should be highly
unstable, and should collapse almost instantly. If the ions simply rearrange in the layers
to eliminate the vacancies. each resulting layer is then hexagonally close packed (Figure
17(c)). If the inter-layer stacking sequence (ABCABC ...... ) is retained, with minimal
adjustments the product should be the observed fluoriteotype structure. This structural
change can be understood better by comparing the layer stackings in Figures 9(a) and
9(b) (Page 36).

Even though direct observations of this microscopic process have not been
possible. we believe this is the most likely mechanism for two reasons. Firstly, such a
change is nearly a one step process. The intermediate can neither be trapped under the
experimental conditions nor be observed easily. This is consistent with our experimental
results that no trace of any new phase is seen by X-ray diffraction. Secondly, such an in-
layer ion rearrangement involves minimal structural change and should have a low

activation energy.

114

oooooooocoooooo
oooooooooooooooo
oceooooccooooooo
oooooooooooooooo
oocooooooooooooo
oooooooocoooooo (a)
oooooooooooooooo
00000000000000
00000000000000.
0000000000000

0 000 0 000 O
000 0 000, O OO
O O 000 0 000
0 000 O 000 O
O 00 O 000 O 00 (b)
O 000 0 000
O 000 O 000 O
OO 0 000 O O
C) 000 O 000
000 O 000

0 Li 0 O O O O O O O
O O O O 0 O O O O
O O O O O O O O O O
O O O O O O O O O
D O O O O O O O O O
O O O O O O O O O

(C)

Figure 17. Proposed structural changes during soivolytic decomposition in a single
cation layer of Ln,.Cl,:,. (solid circles: Ln“; open circles: Ln”) (a) Before
extraction; (b) proposed intermediate; (c) after reorganized.

115
c. Ba,Ln,Br” to Anti-FezP-Type BaBr, (Ln a La, Nd) Transformation

The title precursors have the cluster-type Ln“Cl33 structure. The structure of
Ba1 ,l.n,°Cl“ is unknown, but has been shown to be fluorite-related. The soivolytic
decomposition product has the hexagonal anti-FeaP-type structure. but not the fluorite-type
structure even though the fluorite-type high-temperature form of BaCl2 can be retained at
ambient temperature as described previously. No trace of fluorite-type BaCl2 or BaBr2 was
observed in either experiment. These results suggest another mechanism.

The major difference between Bax2 and Lnx2 is their cationic radii (_28). Ba2+ (1.56
A) is ~10°/e larger than both Sm“ (1.41 A) and Eu“ (1.39 A). From Table 31 it appears
that the M”’/Ln3+ radius ratio is the major factor in determining product structure.

Thus when LnX3 species are removed from the mixed halides BaX2-LnXa the lattice
collapses in a way different from that in the mixed valent chlorides LnClz-LnCl3 presumably
due to the largerosized Ba”. We demonstrated in Chapter 2 that the cation positions in
the anti-FezP-type structure can be derived through displacement of hexagonal close-
packed layers. The observed formation of this phase upon soivolytic decomposition
suggests that the in-layer rearrangements of the cations to form hexagonal close packing
layers is not an energetically favorable process for Bax? instead. an out-of-layer
displacement occurs. Cation displacement leads to the mixing of the cations and anions
in the same layers and eventually to the formation of anti-FezP-type structure as can be
seen in Figure 9(c) (Page 36).

Metastable phases formed in this way are more densely packed than the normal
forms, in contrast to the significantly less closely packed fluorite forms. This result
suggests that the barium halides in the mixed halides are probably under intemai
pressure. The intemai pressure may be created by the size incompatibility mentioned

above, and is probably the cause of the atomic displacements.

size
06)

le

T:

I

Jilin

901ml?!

.9? I

l

1 16
It is noteworthy that the 'high-pressure' forms prepared from precursors of different

sized ianthanide cations in Bax2-Lnx, differ slightly. As is shown in Table 32, the c axis
deviation appears significant and the precursor with the larger cation produces the anti-
FeaP-form with the smaller unit cell volume. The deviations are small but are apparently
not due to experimental errors since they are reproducible for both the chlorides and the
bromides, and are apparent in the observed X-ray diffraction patterns (Table 17 (Page 79)
and Table 18 (Page 80)). The larger La3+ cation tends to create higher ”internal

pressures' than the smaller Nd" and Sm” cations.

Table 31. Structure-ionic radius ratio relationships for metastable dihalides.

 

 

 

 

Precursor Radius ratio MX2 structure
system XIM” M’VLn” type
SmCl2-SmCl, 1.16 1.16 Fluorite
EuClz-EuCl, 1 .20 1 .14 Fluorite
BaBrz-NdBr, 1 .17 1 .28 Anti-FezP

 

BaCIz-SmCl, 1.07 126 Anti-FezP

 

117

.930: o .6. 9.80 an U259.— _m_.= 6.5 05 E2. $5 mu 06m» 05 $05.m 9.35 6:80» 35 5 8:350 8:85 as...

.8526 B 8583 used 228 23 in a. as: B usage .53 em; as .2: es s 5282.. 9: .

 

 

 

 

233852 58. 83... m. .8 888.4 8333 65.....sz

355:5 .984 a 332. 93.5 38 $93. 833 6563.8
232862 58. 38.5 e58 £33. £88.» 65.3.8

.5... 283 5.33852 :52 53.5 a. EN $83... $883 2.0533.

.as .2: 5.33852 58. 836 38 338.4 $6.83 h.0535

5.82 222.85 232852 58. 835 38 game... 883 .6348
flcoEEoo «<0 r<v> A30 23 39823

ceased genes 52 3.365 E ._o i xv Nxem assess: .6 see “523.8330 .mo 23»

1 18
5.4.3 Geometric Correlations

The preceding discussions on topochemical relations are based on symmetries of
the precursors and the metastable products. Further justifications can be developed from

geometric considerations.

3. Ln,4Cl,. to LnCl, (Ln . Sm. Eu) Transformation

It has been shown that the fluorite-type LnCl, forms from the cluster-type Ln,.Cl,,.
Correspondingly. the cubic lattice parameter 2c of LnCl, can be related to the Ln,.CI”
hexagonal supercell lattice parameters 3,,(super) and c,,(super). Figure 18 illustrates the
relation of a,(super) to ac in a cation layer. From the Figure we find 1/2-aJah(super) a
26.5(mm)/34.0(mm), or ac - 0.551 a..(super). From our previous descriptions we know
that c..(super) is parallel to the cubic body diagonal (=- vaag and the supercell is 6
layered while the cubic cell is 3 layered. Thus we have (6/3)‘13-ac . c..(super), that is, ac
== (~13/6)c,,(super) :- 0.2887c,,(super). The cubic lattice parameters derived from both
a,.(super) and 0,,(super) are listed in Table 33. Derived parameters are close to observed

values.

b. Ba.Ln,Br,, to BaBr, (Ln . La. Nd) Transformation
Similar derivations can be made. In Figure 18 the relationship between the two
sets of lattice parameters are ah(BaX,)/a,,(super) - 23.5(mm)!36.0(mm). or ah(Bax,) ~
0 -653a.(super). and 0,,(Baxz) - c,.(super)/6. Derived parameters are listed in Table 34.
Here the metal atoms must undergo significant in-layer reorganization and out-of-layer
displacement to form the anti-FezP-type structure (comparing Figures 9(a) and 9(6). Page
36). As a result of the displacement the an axis (in-layer) shortens substantially while the

Cr. axis elongates.

119

 

Figure 18. Lattice parameter relationships between the M,.X,, superstructure aid the
anti-FezP-type structure (upper left), and between the superstructure and
the fluorite-type structure (lower right).

120

Table 33. Geometrical correlation of the metastable LnCl, (cubic) with their
precursors Ln“Cim (hexagonal) (units in A).

 

 

 

 

 

 

Precursor Metastable LnClz. ac
Ln“Clm Derived Observed
a..(super) 12.864(2) 7.09 ' 6.9827(5)
Sm,.Cl,=.
0,,(super) 24.72(8) 7.186 6.9827(5)
a..(super) 12.815(4) 7.06 6.961 (1 )
Eu,.Cl,a
c..(super) 24.768(8) 7.150 6.961 (1 )
Note. Derived ac - 0.551a,,(super). or ac - 0.2887-ch(super).

Table 34. Geometrical correlation of the metastable BaBr2 with its precursors
6a.l.n,sr,. (both hexagonal) (units in A).

 

 

 

 

 

Precursor Metastable BaBr2
LnuClaa Derived Observed
ah 14.039(1) 9.17 8.4814(8)
Ba,Nd,Br,,
0,. 26.477(4) 4.413 4.8393(7)
afl 14.098(2) 9.21 8.4826(6)
839L358!”
ch 26.678(6) 4.446 4.824(1)

a¥

Nate. Derived: a,,(BaX,) ~ 0.653-ah(super); c,,(BaX2) a c,,(super)/6.

CONCLUDING REMARKS

Low temperature dehydrations and soivolytic decompositions of mixed(-valent)
halides have enabled syntheses of metastable phases either not observed previously or
observed previously. but not quenchable. Layer structure analysis suggested that the
reactions are topologically controlled. Nearly all the low temperature reaction products
have symmetries equal to or higher than those of their precursors. The leaching product
of Ba,Ln,Brm is the anti-FezP-type BaBr, but not the PbClz-type BaBrz. The latter has a
similar layer structure (Page 36) and similar coordination polyhedra but is less symmetric.
Symmetry factors apparently are playing significant roles in these low temperature
reactions. In the soivolytic decomposition the layers appear to have higher reactivities and
the rotational axes orthogonal to the layers are the characteristic symmetry elements
retained during reactions. The author believes that the topological analysis developed in
this work will help future design for and possibly allow prediction of product structures
based on known precursor structures. A combination of the soivolytic decomposition and

desolvation reactions can probably be applied to more systems.

121

PART II

MAGNETIC AND ELECTRICAL PROPERTIES OF OXIDENIT RIDES

122

CHAPTER 1 INTRODUCTION

SYNOPSIS
The possibility of achieving mixed-valent states and making new compounds with
potentially interesting magnetic and electrical properties by effectively substituting N$ for
O" is discussed. The chemistry of transition metal oxidenitrides is reviewed and three

systems are proposed for studies.

Since the discovery of high T, ceramic oxide superconductors in 1986 (_111)

and subsequent characterization of the T6 = 90 K ’123' phase, YBa,Cu,O,,, mg) in

1 987, Cu-containing oxides have become the most intensely examined compounds for
sunerconductivlty. Numerous compounds of the ’123' phase (see. for example.
mdlaflmmug, some of which exhibit superconductivity.
and new types of superconductors such as the Te - 125 K Tl,CazBa,Cu,O,w and related
compounds (119129), and other types of cuprates (see, for example,

1 21 122 123 125%) have been synthesized and characterized.

123

124

Reports of anion substitution in these ceramic superconductors are far less
extensive than those of cation substitution. Such a result is not surprising because not
only is it more difficult to synthesize an anion- than a cation-substituted compound, but
also the number of possible substitutions is much more limited. Low-temperature
fluoridation has been found to passlfy ’123' compounds to hydrolysis (126) with no loss
of superconductivity. However, substitution of fluoride and chloride anions for oxide
anions in ’123’-related compounds suggests that at least for these anions Tc is decreased.
The Nd,CuO.-type compounds appear to behave differently to anion substitution. When
La,CuO. is treated with F,(g) at 200°C a ’new' KzNiFj-type structure forms (1_27). This
new phase has Tc . 40-50 K. somewhat higher than values reported for Oz-treated
La,CuO..

On the basis of size and electronegativity considerations. N’ appears to be a likely
candidate for O" substitution. The ionic radius (_2_§) of four-coordinated N", 1.32A, is only
slightly larger than that of four-coordinated O". 1.24A; and the electronegativity of N
(3.04) and O (3.44) expressed on the Pauling scale. are relatively close. Some metal
nitrides have long been known as low Tc superconductors (_1_2§). in view of the

possibility of achieving a mixed-valence state whose existence in many compounds is an
important factor for high Te superconductivity and uncommon stoichiometries, and of
making new compounds with potentially interesting magnetic and electrical properties by
effectively substituting N“ for O". we initiated an invesfigation of selected oxidenitrides,

mainly ianthanide-containing oxidenitrides.

125
1.1 An Overview of Transition Metal Oxidenitrides

Previously known transition metal oxidenitrides may be classified into two
categories: (1) pseudo-ternary oxidenitrides including perovskite-related AB(O.N), where
A is an alkali, alkaline earth. Y or Ianthanide cation; 8 is W, Mo. V, Nb, Ta or TI

1291_3§,1_31,132_,1_3_3); and Scheelite-type AB(O.N)4 including LnWOaN
(Ln :- Nd, Sm, Gd. Dy) @330. and MOsO,N (M . K. Cs) (135.136) whose
structure is closely related to that of BaSO4. (2) pseudcbinary NaCI-type oxidenitrides
LnO,N,,, where Ln is a Ianthanide cation and so far Ln .. La-Er except for Dy
(137,138,139) have been known. The transition metals in these oxidenitrides

typically have a lower than maximum oxidation state.

1.1.1 Perovskite-Type

MTaO,N (M .. Ca. Sr. Ba) and BaNbO,N (119) were prepared by one step heating
of intimately ground mixtures of BaCO, and N120, in NH, (9) at ~1000°C. On the other
hand, for synthesis of compounds such as BaerJaOzN and SrTaO,N the ternary oxides
were first prepared and then these oxides were nitridized in NH, m. Structures of
some pseudo-ternary oxidenitrides were examined by neutron diffraction. BaMOzN (M =
‘l’a. Nb) compounds were found to remain cubic to the lowest temperature studied (4 K)
with N and O atoms distributed statistically over the anion sites (133). while the structure
of SrTaO,N Q3) was found to be tetragonal rather than the cubic observed by X-ray
diffraction (£9), again with O and N atoms distributed statistically in two crystallographic

sites. Mixed-cation Ba.;Sr,TaO,N was found to be cubic (139).
Another series of perovskite-type oxidenitrides, LnWO;N3,, (0.6<x<0.8). was
prepared by heating the mixed oxide Ln2w20. in NH,(g) at 700-900°C (Lag). LnWOo..N2_.

Was also found to be tetragonal by neutron diffraction (131); the N and O atoms are

015i

31)

inc
ior
ital

0*!

126
distributed statistically. The W atoms in LnWOu 1,Nz,4 exhibited mixed VNl oxidation states

and the compound was semiconducting with powder conductivity values of 10-15 a" cm“.
Thermal power measurements indicated electrons to be the charge carriers. Some of

these oxidenitrides are potentially interesting dielectric materials.

1.1.2 NaCI-Type

Structuracomposition relationships of the NaCI—type oxidenitrides. LnO,N,,,, have
been examined for most ianthanides (137.1%.139). In most systems the cubic lattice
parameter decreases almost linearly as oxygen content increases. Magnetic and transport
properties of some of these compounds, primarily NdO,N,., and the Eu-related systems
Eu,.,Ln,OHN, (Ln . Nd. Eu. Gd) (140,141,312), have been studied extensively.
The NdO,N,,, phases are ferromagnetic and metallic. Conductivity increases with
increasing oxide content and resistivities vary from 10" n-crn for _x . 0.09 to 10* (tom
for x a 0.22. The Eu,Lm,,O,N,,, compounds are also ferromagnetic. However, their
transport properties are apparently different from those of the NdO,N,., systems. The
oxygen-rich Eu,Ln;,.O,N,,, oxidenitrides are semiconductors, while the nitrogen-rich
specimens are metallic. Europium is found in both the +2 and +3 oxidation states.
Magnetic properties are closely related to the predominantly localized! electrons.

1 27
1.2 Goals

Many aspects of the solid state oxidenitride chemistry deserve further studies. Our
primary goals in this work include the synthesis. structural characterization. magnetic and
electrical property studies of new and interesting known oxidenitrides. Described below
are three systems. The MMo(O.N), system presents the importance of NH, reductive
substh for the preparation of mixed-valent or lower valent oxidenitrides; the
BaCeLn(O,N), system utilizes N” as a means'of introducing mixed vaiences; the LaO,N,,,
system demonstrates the need for the examination and understanding of the magnetic and
electrical properties of oxidenitrides.

1.2.1 MMo(O.N),

it is worth noting that in the reported pseudo-ternary oxidenitrides described above
the transition metals are either at the highest oxidation states or at mixed-valences only
slightly lower than the highest oxidation states. These transition metals did not seem to
be easily reducible by N". Such electronic configurations are not desirable for important
magnefic and electrical properties including superconductivity. Since N’ is a fairly strong
reducing reagent especially at high temperatures. we expected reductive substitution to
be possible under favorable conditions.

it has long been known that MoO, can be readily reduced to Moo, by both H, and
NH, (m) below 470°C, and MMOO, (Mo°‘. d°; M 3 Ba, Sr, Ca) can be reduced by
hydrogen to perovskite-type MMoO, (Mo‘*, d“) m. These perovskites are metallic
conductors as determined by electrical (145) and magnetic measurements (_1_4_6).
lf reductive substitution of MMoO, with NH, occurs. oxidenitrides with Mo oxidation states
higher than that in MMO, are expected.

128
1.2.2 BaCeLn(O.N),

Because cerium is known to exhibit multiple valency its nitride and oxidenitride
systems seemed of special interest. Room temperature magnetic susceptibility data
suggest the Ce“ content in CeN may be as high as 89% (£7); Ce“, as expected,
has also been reported in both Li,CeN, and Ce,N,O (148). These observations
suggest that some cerium ions tend to be in their highest oxidation state even in the
presence of reducing N’ ions. On the other hand. cerium forms calcium ferrite-type
BaCezo, that presumably contains Ce“ ions (14;) and is probably stabilized by lattice
energy. Effectively substituting N" for O" in BaCezo4 by fusing CeN with BaCeO, thus
might both lead to a stable compound and also introduce Ce‘+ into the lattice to give a

mixed valence compound.

1.2.3 LaO,,NHI

Since La3+ has neither unpaired nor valence electrons. the LaO,N,_, oxidenitrides
are expected to differ from the other pseudo-binary oxidenitrides mentioned above. Both
LaN and LaO have the NaCI-type structure (1_5_0,m.15_2). I The nature of the
LaN conductivity was still the subject of a recent study (_15_3); it is probably a small
band overlap semimetal. it is interesting to note that LaN undergoes a superconducting
transition at ~1 K Qfl) even though the lanthanum ion is probably trivalent (_15_0).
LaO prepared under high pressures is also reported to be metallic with trivalent La ions
(150,152). Thus Leo is better formulated as La(lll)(O,§), where 3 represents a free
electron. Since La(O.g) has a relatively high solubility in LaN (137), substitution of
nitrogen by oxygen in LaN should enhance its conductivity by increasing the carrier

electron density.

CHAPTER 2 MAGNETIC PROPERTIES OF SOLIDS

SYNOPSIS
Some fundamental magnetic properties of solids are reviewed briefly. The origin
and behavior of each of the five basic types of magnetic order. diamagnetism,
paramagnetism. ferromagnetism, antl-ferromagnetlsm, and ferrimagnetlsm, are described

for the convenience of later calculations and discussions.

129

130
2.1 Magnetization and Magnetic Susceptibility

When an external magnetic field H is applied to a substance. the magnetic field

within the body differs from that surrounding the body. The magnetic induction 8 is

definedas

B=H+4nM
which can be rearranged as

%=1+4rt%

where M is the magnetization (magnetic moment per unit volume).

I . A]; is the volume magnetic susceptibility (emu cm");

x, a % is the gram magnetic susceptibility (emug") where d is the density; and

I" = 16' u. . x L2“: is the molar magnetic susceptibility (emurnol“) with M. the

molecular weight.

131
2.2 Five Basic Types of Magnetic Order

22.1 Diamagnetism

This is a non-cooperative phenomenon associated with closed electron shells.
When placed in an inhomogeneous magnetic field. a diamagnetic substance will tend to
move towards the weakest region of this field. its origin can be understood qualitatively
by considering Lenz's law of electricity. Shielding currents are induced in atoms of filled
electron shells by an applied magnetic field. The shielding magnetic field thus induced
will repel the external magnetic field. Some features are summarized as follows:
(1). Susceptibilities are always negative.
(2). it is a property of all matter since closed shells exist in all substances.
(3). It persists only when an extemai magnetic field is applied.
(4). Susceptibilities are field-independent and nearly temperature-independent.
(5). Susceptibilities are very small (10*5 -10‘ emu-moi").

Theoretical calculation of the diamagnetic susceptibility is possible. Values for
many ions can be found in Selwood (15_5) and other books (see. for example,

156.157).

2.2.2 Paramagnetism

This is also a non-cooperative phenomenon. Atoms with unpaired electrons
possess magnetic moments. A substance is paramagnetic if magnetic moments of
various atoms in this substance are uncorrelated (randomly oriented) in the absence of
an extemai magnetic field. The dipoles tend to become aligned in an applied magnetic
field. Thus a paramagnetic substance will move to the strongest region of the field.

Paramagnetic behavior is shown schematically in Figure 19(a) (Page 137). Some

132

common characteristics are summarized below:

(1).
(2).
(3)-
(4)-

Susceptlbilities are always positive.

It is a property of substances with unpaired electrons.

Susceptibilities are field-independent.

Susceptibilities are significantly higher (for localized electrons. 10"-10*5 emu moi")

than diamagnefic susceptibilities.

Curie and Curie-Weiss Laws

Paramagnetic susceptibilities of compounds with localized and unpaired electrons

are strongly temperature-dependent. The temperature dependence follows the Curie or

Curie-Weiss Law.
Curie Law:
0
xm ' ‘7".
Curie-Weiss Law:
x = i
"' T- 0

where C is the Curie constant and 8 the Weiss constant. Deviation from the Curie law is

caused by the interaction (exchange coupling) between atomic moments in a magnetic

field.

b.

Spin-Only

In compounds of the first transition (3d) series. the orbital contribution to the

magnetic moment is almost completely quenched by the ligand fields. Only the spins

contribute to the effective magnetic moment it... (in Bohr magnetons):

133

“‘8' ¢S(§+1)=Jn(n+2)

where S is the total spin quantum number and n the number of unpaired electrons.

c. Spin-Orbital Coupling
For a free. paramagnetic ion such as a Ianthanide ion in which spin-orbital
coupling causes a splitting of the ground state that is larger than kT, the effective

magnetic moment may be calculated from the formulas

Ila: UYJIJ+ 1)

1+ J(J+1)+sgs+1)-L(L+1)
2J(Jo+ 1)

 

a:

where g is the electron gyromagnetic ratio, and J. S, and L refer to the ground-state

spectroscopic term L38"

d. Temperature-Independent Paramagnetism (TIP)

This arises from the admixing of higher energy states that fie above the ground
state by an amount much greater than kT into the ground state. It has been discussed
by Jolly (1_§§). Correction for TIP is necessary when it is significant.

6. Pauli Paramagnetism

Metals have unpaired itinerant electrons. Most metallic materials, except those few
that are ferromagnetic, are either diamagnetic 0r feebly paramagnetic with nearly
temperature-independent magnetic susceptibilities. This behavior can be understood on

the basis of a free electron gas model. in this model it is assumed that the valence

134
electrons are delocalized. These mobile electrons occupy the conducfion band and are

thus called conduction electrons. Most conduction electrons do not respond to the applied
magnetic field because most states are fully occupied. Only those electrons near the
Fermi level E, within the energy range of ~kT contribute to the susceptibility. The volume
susceptibility is given by the Pauli-Peierls equation I

m: I
3m“

 

X . M(¥)m.pg.(1-

h2

where p, is the Bohr magneton ( . -9.27x10"‘" erg gauss"). n the electron density, at, the

mi

3m‘2

electron rest mass. and m' the electron effective mass. The Landau term -

 

applies the diamagnetic correction. m' can be greater or smaller than m, (157). Many
transition metals have values of m'lm, »1. Hence from the above equation we would
expect them to have comparatively large paramagnetic susceptibilities (observed 10‘-10‘I

emu-mor‘). A few metals such as Bi, Hg, and Cu have m'lm, «1.

2.2.3 Ferromagnetism

This is a cooperative phenomenon and exists only in the solid state. In a
ferromagnetic solid the spins tend to be aligned parallel below the ferromagnetic Curie
temperature T,. As a result. a spontaneous magnetization (M,) exists in such materials;
that is. even in the absence of a magnetic field there is a magnetic moment. The
magnitude of the spontaneous magnetization drops to zero at T, because thermal
energies randomize the spins. Above T,, the paramagnetic susceptibility obeys the Curie-

1 35
Weiss law:

0
r- r,

 

Xm'

In practice, when a magnetic field is applied to a ferromagnetic specimen, the
magnetization may vary from zero to a maximum value‘called saturation magnetization.
According to Weiss’ postulation, there exist many domains in a bulk material. Although
each domain is spontaneously magnetized, the direction of magnetization may vary from
one domain to another. The overall magnetic moment of the specimen is the vector sum
of the magnetic moments of individual domains. Thus the average magnetization and
magnetic susceptibility of a ferromagnetic material depend on the applied magnetic field.
Saturation occurs when the maximum domain alignment is achieved at a certain magnetic
field. The magnetic susceptibility of a ferromagnet also depends on the previous magnetic
history. Hysteresis, a characteristic behavior of ferromagnetic materials, is not discussed

here. The ferromagnetic behavior is illustrated in Figure 19(0) (Page 137).

2.2.4 Anti-ferromagnetism

Anti-ferromagnetism, like ferromagnetism, is a cooperative magnetism of long-
range order among identical, spontaneous moments, and exists only in the solid state.
Ideally, magnetic ions occupy crystallographicaliy equivalent sites. A simple anti-
ferromagnet can be envisaged as composed of two identical sub-lattices of opposite (anti-
parallel) spins. The exchange coupling is thus negative. There is overall no spontaneous
magnetization.

The temperature dependence of the magnetic susceptibility is complicated. Above
the Neel temperature, T... where the thennal energy of a moment exceeds the exchange

energy. the substance is paramagnetic and follows the Curie-Weiss law

136

c
r- r,

 

In."

Below T" in the anti-ferromagnetic range, the susceptibility decreases with decreasing
temperature. In an ideal anti-ferromagnet, x is zero at 0 K.

The anti-ferromagnetic behavoir is illustrated in Figure 19(b).

2.2.5 Ferrimagnetlsm

A ferrimagnetic material may be defined as one which below a certain temperature
possesses a spontaneous magnetization that arises from a non-parallel arrangement of
the strongly coupled atomic dipoles or from two or more magnetic spades that are
chemically different. These different spades occupy different crystallographic sites. The
non-parallel arrangement can be anti-parallel like that in an anti-ferromagnet. However,
the net magnetization is not zero because of the difference in magnetic moments of
different magnetic species.

The temperature dependence of the magnetic susceptibility is similar to that of a
ferromagnet in a ferrimagnetic material, however, the spontaneous magnetization M,
usually decreases more rapidly than that in a ferromagnet with increasing temperature up
to the Neel temperature T9,. In the paramagnetic range there is appreciable deviation
from the Curie-Weiss behavior, particularly close to Tm. The ferrimagnetic behavior is
shown in Figure 19(d).

In addition to the above five basic types of magnetic order, Hurd fig) recently
reviewed nine other types that are derived from these basic types. These relatively new

magnetic phenomena are beyond this work and therefore not discussed.

Figure 19.

137

 

 

‘I/hI/
3\--'\/
o 4r

 

 

 

 

 

 

II I I I
I I I (b)

g- I I l

o T: T4

‘i ‘ j l
l I 1 (d)

M, I I
’I’TI/X.)
o r. FA

The low temperature ordering (if any) of neighboring dipoles and the
consequent behavior of spontaneous magnetization and/or susceptibility,
for (a) paramagnetlsm; (b) anti-ferromagnetism; (c) ferromagnetism; (d)

ferrimagnetlsm.

CHAPTER 3 ELECTRICAL PROPERTIES OF SOLIDS

SYNOPSIS
The classification of electrical materials into metals, semimetals, semiconductors,
and insulators according to the Band Theory of solids is reviewed briefly. Typical
properties of superconductors and the thermal behavior of electrical conductivity of metals

and intrinsic semiconductors are described.

138

139
3.1 The Band Theory of Solids

The electronic structures of solid state compounds are usually described by the
band theory of Bloch and Wilson developed from molecular orbital theory. In this theory
it is assumed that an extremely large number of atomic orbitals of the same symmetry
overlap to form an energy band with nearly continuous energy levels. The space between
two different bands is called the band gap E9. The highest filled band (HOMO) is called
the valence band, and the lowest empty (LUMO) band the conduction band. The highest
occupied energy level In a band is called the Fermi level E,. In a metal. EF lies within the
partially filled band. In an insulator or a semiconductor. EF lies between the valence band
and the conduction band.

According to this theory, a metal has a strong overlap of the conduction and the
valence band. In other words. the highest occupied band is only partially filled with
electrons in a metal. An insulator has either completely full or completely empty bands
with a large band gap (E, > 3.0 eV); semiconductors can be classified as intrinsic (pure)
or extrinsic (doped) depending on the configurations of the energy levels. Extrinsic
semiconductors can be further divided into P- (positive or hole) and N- (negative or
electron) types according to the nature of carriers. An intrinsic semiconductor is N type
and has a completely filled valence band and a completely empty conduction band at 0 K,
but has a fairly small band gap (3.0 W > E, > 0). The small thermal population of the
conduction band by electrons from the valence band is responsible for the
semiconductivity at higher temperatures. The conductivity of extrinsic semiconductors
arises either from population of the conduction band by the small number of dopant
electrons (electron donor dopant. N-type) or from population of the empty energy levels
provided by the electron-accepting dopant (P-type). The band structures of various
materials are shown schematically in Figure 20.

140

It is worth noting that there are several classes of compounds that cannot be
described easily as either metals or semiconductors. They are called semimetals because

the conduction and valence band overlap is smaller than that in metals.

Conduction
5:5 “in" j '1' EzECIg . ”,I I I

 

 

, c ( Iszso“""’ i=5.
l “T I GEO 111')
I up
E=E -- _.
El E=E W // E=EV E:EA- .
Valence bond
lol ibl icl ldl

 

 

m
H
m

 

 

lei ifi l9l

Figure 20. Schematic band structures of solids. (a) insulator; (b) intrinsic
semiconductor; (c) and (d) extrinsic semiconductors; donor and acceptor
levels in n- and p-type semiconductors are shown respectively. (9)
compensated semiconductor; (f) metal; (9) semimetal.

141

3.2 Electrical Conductivity

A The conductivity (0) of a material is proportional to the carrier density (n) and the

carrier mobility (u):

c=ne°tt

where e, = 1.602x10"° Coulomb is the carrier (electron) charge. p = Us is defined as the

resistivity. Units of the above quantities in both the Si and CGS systems are listed in

Table 35.

Typical conductivity values for metals, semiconductors, insulators, and

superconductors are listed in Table 36.

 

 

 

 

Table 35. Units of quantities related to electrical conductivity.

Quantity n u c p

SI rna mz-V‘ s" n" m“ n-m

CGS cm“ crn’-\r1 s“ 0" cm" ncrn

Table 36. Typical electrical conductivities for various materials at room temperature.

 

Metals Semiconductors Insulators superconductors

 

c (fl-cm)"

10‘-10‘ 10‘-10’ 10"'-10‘ 10”

 

142
3.2.1 Temperature Dependence of Metallic Electrical Conductivity

The resistivity of a typical metal decreases with decreasing temperature because
of weakening lattice vibrations Upon cooling. Above a certain temperature, the resistivity
is linearly proportional to the temperature. pm as T; at very low temperatures. p(T) ~ T°
or T“. In practice, the resistivity reaches a plateau value controlled by impurities and
lattice imperfections at a low temperature. Matthiessen's Rule is often observed: the

contribution of Impurities and imperfections to resistivity is a constant (1mm).

1 1 1

s +

°(T) a” OMITI
A typical resistivity vs. T curve for sodium metal (159) is shown in Figure 21 (Page

 

 

 

143).

3.2.2 Semiconductors
For intrinsic semiconductors. the number of conduction electrons is the same as

the number of holes. Thus the conductivity is

d = 0,001" + up)
where
)—— .5.
n, 3 ”O. ”V m - )
2kT
here N, and N, are the effective densities of states in the conduction and valence bands.
respectively. and u’s are the mobilities. The final expression of the conductivity is

E E
d . Zeu,(b+1)(21t‘/m,,m,%')”'oxp(—2—kl— = gum—fil—

where b . unlup. Since the pre-exponential factor (0,) is insensitive to temperature, a plot

143
of in 0 vs. 1/T should yield a straight line with slope . -E,/2k, from which the band gap

can be estimated.
For extrinsic semiconductors. the temperature dependence of conductivity is
complicated because the electron and hole densities, n, and n,, are no longer equal.

Since it is irrelevant to this work, this subject is not discussed here.

 

(”Pm

 

 

 

 

TIKI

Figure 21. Ratio of the resistivity of metallic sodium at T to that at 273 K vs. T. The
essentially linear variation at T > 100 K is typical of simple metals.

1 44
3.3 Superconductivity

Superconductivity is the loss of electrical resistance below a critical temperature
T, called the superconducting transition temperature. Typical resistivities in the
superconducting state are ~10” 0cm in comparison to the ---10’7 O-crn resistivities of the
very good metallic conductors Cu and Au at 100 K. A superconductor must be a good
metallic conductor above T,. Some important features of superconductors are

summarized here.

The Meissner-Ochsenfeld Effect
In the superconducting state materials exhibit perfect diarnagnetism. The magnetic

induction 8 inside the specimens is zero.

B=H+4RM=0

Thus the volume magnetic susceptibility in the perfect diamagnetic state is

1

x8—8-—

H 41:
The result of this diarnagnetism can be observed by magnetic field exclusion which
causes the Ievitatlon phenomenon. Since the Meissner-Ochsenfeld effect is a
fundamental property of the superconducting state, it is often used for detecting
superconducting transitions.
The interactions of the normal conducting and superconducting states with a
magnetic field are illustrated schematically in Figure 22.

145

       

“Perfect“
Conductor

   

 

 

 

 

 

 

 

 

4)

I Superconductor

T>T,. T<T,. T<T.-
H>0 0<H<H,. H=O

   

 

 

 

 

 

 

Figure 22. An illustration of the difference between a normal conductor (a) that allows
magnetic flux penetration and a superconductor (b) that expels magnetic
flux from its interior.

CHAPTER 4 EXPERIMENTAL

SYNOPSIS
Synthesis, magnetic and electrical measurement procedures for various
oxidenitrides are described. The oxidenitrides were prepared either by nitridation of
oxides with NH, (9) or by direct reaction of oxides with nitrides. The four—probe method
for the conductivity measurement was used for both pressed pellets and sintered rods of
poly-crystalline specimens. Formulas used to elucidate the magnetic and electrical data

are listed.

146

147

4.1 Synthesis and Structural Characterization of Oxidenitrides

The nitrides and some oxidenitrides are air-sensitive and were handled in the glove

box when necessary. Nitrogen content was determined by Galbraith Laboratories, Inc.

Powder x-ray diffraction and pattern treatment procedure are as described in Part I of this

dissertation.

4.1.1 Ust of Chemicals

La. Ce chips: Research Chemicals, Phoenix, AZ.

CeO,:
La,O,:
ThO,:
MoO,:
BaCO,:

Sr(NO,),:
Ba(NO,),:

NH: (9)3

99.9%, Research Chemicals, Phoenix, AZ.

99.9%, Research Chemicals, Phoenix, AZ.

anhydrous. purified. Fisher Scientific Company, Fair Lawn, N. J.
99.5%. Reagent, Merck 8 CO. Inc., Rahway. N. J.

“Baker Analyzed“, J. 1'. Baker Chemical Co., Phillipsburg, N. J.
Certified A.C.S., Fisher Scientific Company, Fair Lawn, N. J.
Reagent A.C.S., Matheson Coleman &Bell, Norwood. Ohio.
Matheson Gas Products, E. Rutherford, N. J.

1 48
4.1.2 Synthesis Procedure

MMoO, (M = Sr, Ba) Precipitated from 1:1 solutions of M(NO,), and (NH,),MoO, (MoO,
dissolved in Nl-l,-l-l,O) at pH . 9-10, then annealed at 800°C for 2-5h.

mm, (M -.- Ce, Th) Perovskite-type BaCeO, and BaThO, were synthesized by heating
intimately ground 1:1 molar mixtures of 8300, and M02 (M = Ce, Th) at 950°C for 24 h,

then quenching them.

LnN (Ln a La, Co) 5.6 g of metal chips in corundum crucibles were heated in a slowly
flowing ammonia gas stream at 700°C for 12 h. Pulverized CeN was dark brown and
contained ~10% CeO, by mass; LaN was black and contained ~5% La,O,. Both samples

are moisture and oxygen sensitive.

BaCeLn(O,N), (Ln = La, Co) A 08-10 g intimately ground and nominally stoichiometric
mixture of BaCeO, and LnN was sealed into an outgassed quartztube under vacuum or
an argon atmosphere. heated at 800°C for 24 h. then quenched by removal from the
furnace.

For electrical conductivity measurements ~1.5 g specimens of each product were
inserted into 6.35 mm id. Ta cylindrical crucibles whose ends were crimped closed. The
containers were then heated inductively to 1200-1300 °C for ~2 h in a 10‘ Torr vacuum

to sinter the material. Specimens thus prepared were ~5 mm in diameter and 10 mm

long.

LaO,N,., Two oxidenitride compositions were prepared by inductively heating
intimately-ground mixtures of the LaN and La,O, for 2-3 h in vacuo to 1500-1600°C. The

 

149
~15% (l) and ~26% (II) (by weight) La,O, mixtures were confined in closed 6.35 mm id

Ta tubes. The products. sintered rods ~4 mm o.d. and 6-8 mm long, were used as
synthesized for resistance measurements. Crushed samples were used for magnetic
measurements and X-ray characterization. The oxidenitrides are air and moisture

sensitive.

MMo(O,N), (M = Sr, Be) A sample of 0.64.2 g MMoO4 was weighed carefully and
contained in a clean platinum or silver crucible, which was then placed in an alumina boat
for stability. Reduction was carried out in a flowing NH, stream. Reaction conditions vary
depending on the system. For SrMoO4 heating at 750°C for 48 h appeared to be
appropriate; for BaMoO, various temperatures from 650 to 900°C and reaction time from
2 to 6 days have been tried. Reaction products were weighed carefully to check for mass

loss.

150
4.2 Magnetic and Electrical Measurements

Powder samples were contained in plastic capsules and their magnetic moments
were measured on a computer-controlled Quantum Design SQUID magnetometer at
various magnetic field strengths. Three scans were made and then averaged for each
data point. In most of our experiments magnetic moments were measured as a function
of both temperature (2 or 5-300 K) and magnetic field strength (200 gauss to up to 5
teslas) by changing the magnetic field at each temperature.

A. C. resistances of pressed pellets of polycrystalline MMo(O,N), (M = Sr, Ba)
were measured on a Hewlett-Packard 4192A LF Impedance Analyzer by the 4-point
arrangement (the van der Pauw method). Specimens were round pellets with a 2.38 mm
diameter and a uniform thickness (1.24 mm for BaMo(O,N), and 1.01 mm for SrMo(O,N),).
Pressure sensitive contacts were employed. Applied voltage was 1.00 V at 1 kHz.
Temperatures ranged from 295 K to ~85 K (liquid nitrogen temperature). The probe
arrangement is shown in Figure 23(a).

D. C. resistances of sintered LaO,N,,, specimens were measured from 2-300 K by
the four-probe method on the Quantum Design SQUID magnetometer. Insulated copper
wire probes were spot-welded onto samples in a nitrogen-filled glove bag. To eliminate
the thermal e.m.f. contribution to the voltage readings measurements were performed with
positive and reversed polarities for 20 seconds each and then averaged. The
measurement was also computer controlled with a constant D. C. current supply of 10 mA.
Voltage drops ranged from 1.3 to 140 mV in the normal conducting state. The probe
arrangement is shown in Figure 23(b). ’

Conductivity of the sintered BaCeLn(O,N), specimens were checked with a digital

ohmmeter.

151

 

Figure 23. Simple arrangement for four-probe electrical conductivity measurements.
(a) Four equally spaced probes on a pressed round pellet of uniform
thickness; (b) four probes on a cylindrical sample of uniform diameter;
(0) The arrangement for van der Pauw measurements. Note that in (a)
and (c) current is driven between two probes and the voltage is measured
between other two probes.

1 52
4.3 Data Analysis

4.3.1 Magnetic Data
3. Calculations

Directly obtained values of the magnetic measUrements were longitudinal magnetic
moments in emu units. Molar magnetic susceptibilities were obtained according to the
following formula:

x" . W
n-H,

where n is the number of moles of the material and H, is the applied magnetic field
(gauss). The xm's were plotted against 1/H when data in different magnetic fields were
available and straight lines were extrapolated to zero reciprocal field (1/i-l —> 0) to
eliminate contributions of ferromagnetic impurities. If necessary. extrapolated
susceptibilities x..." were finally corrected for diarnagnetism according to the following

equation:

«In at &
Xllr ' Xm " Xcr
Diamagnetic susceptibilities were estimated by the following procedure developed
by Pascal:

I: ' 21 ni'XKA) '* 2:, "I'XKQ

where n' is the number of atoms of type iin a molecular formula, x, the ionic (or atomic)
gram diamagnetic susceptibility, and x, is the susceptibility associated with n] structural
features such as double or triple C-C bonds. In this work 111 . 0. Diamagnetic
susceptibilities were taken from Selwood fig). It is worth noting that the correction also

153
included the diamagnetism of the capsule containers used in this experiment.

Because the diamagnetic contribution is always negative, the corrected molar
susceptibility is always greater than the uncorrected value.
For a paramagnetic substance that obeys the Curie or Curie-Weiss Law, the

effective magnetic moment )1... (in Bohr magnetons) can be calculated with the formula:

 

1m= amt/armr- el

8 = 0 when the Curie Law is obeyed.

(Numerous computer programs for the above calculations have been written in
FORTRAN-77 and are available in directory [LIU] in the CEMVAX system. If sufficiently

good data are available, computer extrapolation of x,“ vs. 1/H can be used.)

b. Plotting
x...” vs. 1/T.

If a straight line is observed the data follow the Curie Law. The slope of the
straight line is the Curie constant C; the line should pass though the origin.

If a straight line which intercepts the positive 1,, axis at UT = 0 is obtained,
temperature-independent paramagnetism may be present, or the diamagnetic correction
may have been over-estimated.

If a straight line which intercepts the negative x", axis at 1/T = 0 is obtained. the
diamagnetic correction probably has been underestimated.
1m,” vs. T

If the x...“ vs. 1/T plot is not linear, then plot 1/x,,,°°" vs. T. A straight line indicates
that the data follow the Curie-Weiss Law. The intercept of the line is the Weiss constant

8; the slope is still the Curie constant.

154
xmm" vs. T

Sometimes magnetic susceptibilities do not have a simple temperature
dependence. For example. an anti-ferromagnetic transition will have a complex magnetic

feature. In this case. a straightforward plot of 1 vs. T is probably most informative.

4.3.2 Electrical Resistivitles
For the sintered LaO,N,,, specimens, D.C. resistivities (n-cm) can be calculated

with the formula:

p=R

~|>

gnl’:
I

where R Is the resistance (9) obtained by the method shown in Figure 23(b) (Page 151).
A is the cross sectional area of the sample rod, I the distance (cm) between the two
voltage probes. and r- d/2 is the radius (cm) of the rod. Resistivity errors are estimated
to be less than 5% and are mainly from inaccuracies in the size measurement of the
irregular-shaped specimens.

For powder samples of MMo(O,N), (M = Ba, Sr), the conductivity 6 (0" cm") can
be calculated according to the following van der Pauw equation L1§_9_) that is derived for
arbitrarily positioned contacts with an irregularly shaped sample of uniform thickness d

(cm) (Figure 23(c), Page 151).

°= '“il 2 Iii—”WI
"d ”Accofnccoc Recall

where Rs are the resistances (a) between the designated probes, and F is a slowly

 

varying function of its argument. In our experiment, the samples were Around disks and

the four probes were equally spaced, thus

155

and the van der Pauw equation is simplified as

ln2
03——

0’
1rd}?

132'

CHAPTER 5 RESULTS AND DISCUSSION

SYNOPSIS
Crystal structures derived from X-ray powder diffraction data, chemical analysis
results, and magnetic and electrical properties of various oxidenitrides are presented and
discussed. Perovskite-related MMo(O,N), (M a Ba, Sr) exhibit complex magnetic behavior
at low temperatures and are probably metallic; CaFe,(O,N),-type BaCeLn(O.N)4 (Ln = La,
Co) are insulating and paramagnetic with Ce exhibiting mixed-valent states; NaCI-type
LaO,N,,, are metallic above 6 K for x = 0.45 and above 5 K for x = 0.28 and become

superconducting below these temperatures.

156

157
5.1 Pseudo-Temary BaCeLn(O.N). Systems (Ln . La, Ce)

5.1.1 Chemical Characterization

Reddish-colored BaCeLn(O,N), (Ln = La, Ce) powders were obtained both by
quenching and upon slow cooling. They gave similar ~X-ray powder diffraction patterns,
decomposed quickly upon contact with water, and evolved ammonia gas that was easily
detected both by its smell and by moist gHygrign paper. They also decomposed slowly
in air with a weight gain and a color change first to light yellow and eventually to brown.

Duplicate nitrogen analyses were performed on two identical but separately
packaged (La, Ag) BaLaCe(O.N), specimens. Found for A: 1.60, 1.55%; for _B_: 1.62,
1.49%. (The second member of each pair is lower than the first as would happen if
hydrolysis occurred between analyses and may indicate a slightly low analysis result)
Combining all four results yields %N . 1.56 :l: 0.05. If all anion sites are occupied, the
formula is BaCeLanmN, 54121- The nitrogen content is slightly higher than this; as is
indicated below, the specimen contained ~10% BaLa,O,. .

I Duplicate nitrogen analyses were also performed on comparably packaged
BaCe,(O,N), specimens. Found for specimen A: 1.23, 1.56%; for g: 2.35, 2.41%. In lieu
of the data on the Ln - La compound, the disparity between the two A results, and the
magnetic data, the A set is rejected. The remaining analytical data suggest the formula,
BaCe,O,,,(,,N,,,,,,.

5.1.2 Magnetic and Electrical Properties

Both oxidenitrides are strongly paramagnetic between 10-300 K. in the high
temperature region they obey the Curie law as is illustrated in the Figure 24 plot of
x(BaCeLa(O,N),) _vs T‘. For BaCeLa(O,N)4 the presence of Ce3+ is thus clear. Observed

158

magnetic moments are lower than those calculated on the basis of an 1‘ localized
configuration for Ce". indicative that the cerium ions exhibit mixed valence states. The
observed magnetic moments of 1.83 and 1.70 it, at 300 K for V2 BaCe,(O,N), and
BaCeLa(O,N),, respectively, correspond to about 1.44 Ce3+ in BaCe,(O,N), and 0.68 Ce3+
in BaCeLa(O,N),. Approximately 70% of the cerium ions in each compound are in the
trivalent state.

Magnetic data suggest the formulas BaCeLaO,,,N,,, and BaCe,O,,,,N,,,, again
on ; assumption of complete anion site occupancy. They support the mixed anion
composition, partial reduction of Ce“ in both CeN and BaCeO, by N°‘ when the latter
reacts with CeN at 800°C, and a greater Ce" ion content in the oxidenitride than in the
reactant CeN. For BaCeLa(O,N), the magnetic data also underestimate the N-content
because of the BaLa,O, impurity.

Conductivity tests with sintered specimens indicated insulating properties of these
oxidenitrides. The resistances of both sintered specimens at room temperature exceeded
2 megohms. Unidentified impurity reflections were observed in the X-ray powder
diffraction patterns of the sintered specimens. These resistivities are so large that even
in the presence of a small impurity level the products must be insulators. Although
precise resitivities could not be obtained, these data suggest localized trivalent and

tetravalent cations.

159

1.8

 

1.6 - ,’
1.4 'i l
1.2- /

1.0- I

0.8‘ ,’

0.6- I

0.4 -

Susceptibility (emu/moi) 10E 2
\

0.2‘

 

 

 

0 40 80 120 150 200

1 000/T

Figure 24. The magnetic susceptibility of BaCeLa(O,N), plotted against reciprocal
temperature at 5300 K.

160
5.1.3 Crystal Structures

The BaCe,(O,N), product contained a trace unidentified impurity-two very weak
reflections in the X-ray powder diffraction photograph. The powder X-ray diffraction
pattern can be indexed on orthorhombic symmetry with figures-of-merit, F(20) :- 26 and
F(31) .. 23. Lattice parameters of both BaLn,(O,N), compounds are listed In Table 37.
Systematic extinctions consistent with space groups anm (at 62, centrosymmetric) or
an21 (if 33, noncentrosymmetric) suggest the CaFe,O,-type structure (3&9).
Intensities calculated using the atomic parameters for CaSczo, m and isotropic
temperature factors of 1.5, 1.0, and 0.9, respectively, for 0", Ba”. and Cek“ agree well
with observed values. Miller indices and observed and calculated lnterplanar d-spacings

and intensities for BaCe,(O,N), are presented in Table 38.

Table 37. Lattice parameters for selected (ngm, 2:4) CaFe,O.-type oxides and

 

 

oxidenitrides.
Formula §(A) p(A) 9(A) V(A’) Reference
BaCe,O, 12.564 10.641 3.676 4922 (1_49)

BaCe,(O.N), 12.565(1) 10.644(1) 3.6593(4) 469.4 this work

BaCeLa(O.N), 12.573(3) 10.637(3) 3.657(1) 469.1 this work

saLa,o, 12.652(2) 10.666(2) 3.7077(6) 501.3 this work
12.662 10.675 3.705 500.6 (lg)

 

161

 

 

Table 38. Miller indices, and observed and calculated interplanar d~spacings and
intensities for £91m CaFe,O,—type BaCe,(O,N),.
n k .I still 9M) 1: I. I1 6 .I 9.14) still I: I.
0 2 0 5.322 5.331 vw 3 3 41 1.9142 - 2
2 2 0 4.061 4.065 w 9 0 0 2 1629716263 w 20
1 0 1 3.513 3.513 w 11 1 5 1 1320713210 vw1
1 1 1 3.336 3.336 w 9 6 3 0 1803518031 w 14
4 0 0 3.141 3.142 m 45 6 1 1 1791617911 w 17
2 3 0 3.069 3.066 vs 100 4 4 1 1.775417734 m 20
2 1 1 3.031 3.029 s 67 0 6 0 1.7740 ' 12
0 2 1 3.015 32 2 5 1 1.7659 14
4 1 0 3.013 3'0” m 6 4 5 0 1.7623 ”651 w 4
1 2 1 2.932 2.932 w 7 6 2 1 1720017202 w 6
2 2 1 2.716 2.711 w 2 3 5 1 1646713646 vw 2
3 1 1 2.6677 2.6677 m 26 2 2 2 1.6661 1.6667 vw 2
1 3 1 2.4965 2.4963 w 9 6 4 0 1645716459 vw 3
2 4 0 2.4503 4 5 4 1 1.6346 - 4
3 2 1 2.44712'4483 w 2 7 1 1 1593415931 3
2 3 1 2.3606 - 3 1 6 1 1.56361_5806 w 2
4 1 1 2.3259 2.3257 w 6 4 0 2 1.5610 14
3 3 1 2.1764 11 2 3 2 1.5743 33
4 2 1 2.1453 2'17“ m 31 6 0 0 1.5706 1573‘ m 2
0 4 1 2152121510 w 21 4 1 2 1563915636 vw 3
1 4 1 2.1212 2.1206 w 12 6 1 0 1553815536 vw 5
5 0 1 2.0716 2.0713 vw 6 4 6 0 1.544715442 w 12
2 4 1 2.0360 2.0344 vw 2 7 2 1 1.5424 ' 1
6 2 0 1946719466 w 12 6 2 0 1506415012 WI 1
5 2 1 1930513301 w 5 6 4 1 1.5009 5

 

162
The product of the BaCeO,-LaN reaction is a mixture of two CaFe,O,-type

structures: BaCeLa(O,N), and BaLa,O, (L49). The data compiled in Table 37 illustrate the
close relationship between the lattice parameters of BaCeLn(O,N), (Ln = La, Ce) and
BaLn,O,. The unit cell volume of BaCe,(O,N), is slightly less than that of BaCe,O,. This
result would not be expected from oxidation number and ionic radii considerations. The
Ce“, La°” and Ce“ CN Vi ionic radii are 1.01, 1.03, and 0.67 A, respectively; 02' and Ni
CN IV ionic radii are 1.36 and 1.46 A (26). If the compounds are considered ionic the
number of N“ and Ce“ ions must be equal. The volume of a N3" ion is 2.026 A3 larger
than that of the 0‘ ion while a Ce‘+ ion is 1.557 A3 smaller than a Ce3+ ion. We would
thus expect a unit cell volume increase of 0.470 A’lN atom. The volume decrease
therefore must reflect either more efficient lattice packing (which is unlikely), a smaller size
for the nitrogen atom than its ’ionic’ radius indicates, or the presence of a greater quantity
of Ce“+ than is required for charge balance. Magnetic and analytical data suggest the
nitrogen atom is smaller than predicted by its ionic radius.

The oxygen needed to substitute for loss of nitrogen came from the small amount
of La,O, or CeO, present in the reactants. In reactions which involved CeN this
contaminant disappeared during reaction. in those which involved LaN the product
contained a BaLa,O, impurity, an indication of insufficient Lap, to allow the reaction to
go to completion. The reacfion can typically be illustrated by the following equation

800°C
6BaCeO,(_§) + 4LaN(s) + La,O,(s) ----------> 68aCeLaO,,N,,(s) + V2N,(g)

Gas evolution indeed occurred; quartz tubes which contained the products were

pressurized when they were opened in the glove box. When research was initiated the

163
reaction was envisioned a 1:1 BaCeO,:LnN combination and oxide was not considered

necessary. The results indicate that the (BaCe,O,N) 3:1 O:N product is less stable than
less nitrogen-rich compounds; additional oxide is therefore necessary for reaction.

Reduction of Ce“ by N” also occurred in the CaO-Ceoz-CeN system when we
attempted to prepare CaCe,(O.N), at 950°C. The quartz tubes again contained a gas
under pressure and a Ce,O,-like phase (presumably Ce,(O,N),) resulted; CaCe,(O.N), did
not form. The Ca'“ ionic radius is presumed too small to stabilize the compound.

Efforts to synthesize BaThCe(O,N), using the difficultly reducible Th“ to substitute
for Ce‘+ in BaCeO, were unsuccessful even though perovskite-type BaThO, could be
prepared (162). The 0.94 A Th‘+ ionic radius is very close to that of Ce“+ (0.87 A) and
reasonably close to that of Ca3+ (1.01 A) (2_8), suggestive that a mixed valence Ce-
compound might form. However, a BaThO, - CeN reaction did not occur even at 950°C.
3 temperature higher than that for the BaCeO,-CeN reaction (800°C) the only product
was again a Ce,O,-like phase which resulted from reaction between CeN and the Geo,
impurity. This observation suggests mixed valence to be an important factor for the
formation of salvl,(o,N),-type oxidenitrides -- for cerium to exhibit both 3+ and 4+
oxidation states the related central cation must exhibit some degree of trivalency.

It is noteworthy that the cerium oxidenitrides characterized so far have structures
and lattice parameters closely related to those of their oxide counterparts. The X-ray
diffraction pattern previously reported for the (Ce"’-containing) oxidenitride, Ce,N,O (148).
said to be lsostructural with Th,N,O, is almost identical to that of Ce,O, prepared by H,
reduction of CeO, at 1200°C LLQI- This similarity can be demonstrated better by
comparison of their hexagonal lattice parameters : 'Ce,N,O": a .. 3.660 A, g .. 6.057 A ys_.
0e20,: fl . 3.891 A, c - 6.063 A. Given the essentially identical X-ray scattering powers

of N’ and O”, the small size difference between them, and the counter-balancing size

1 64
difference between Ce“ and 06”, it is difficult to distinguish Ce,N,O (03“). 0620, (063*)

and Ce,NO, (063* and Ce“) and other mixed valence states by x-ray diffraction.
Similarly, the possibility that the reported Li,CeN2 (14_8) with Ce‘+ could be a mixed valent
Li,Ce(O,N), cannot be dismissed. Since anion ordering has been observed in the cerium
oxidenitride CeO,N,_,, L138), the true symmetries of these oxidenitrides may differ from
predictions based upon X-ray diffraction data. Further studies, A9,. neutron diffraction.

appear necessary to characterize these compounds definitively.

165
5.2 Pseudo-Ternary MMo(O,N), Oxidenitrides (M =- Sr, Ba)

5.2.1 Crystal Structures

SrMo(O,N), This product was obtained as dark blue to deep purple fine powders. its
color is in sharp contrast to the deep red of SrMoO,. The X-ray diffraction pattern of the
oxidenitride is essentially indistinguishable from that of SrMoO,. Both are perovskite-types
with a - 3.9765(3) A for SrMo(O,N),, and a . 3.9751(3) A (L42) for SrMoO, (3.9761(5) A
from our result). Repeated experiments indicated that the weight loss upon NH, reduction
corresponds to an empirical formula very close to SrMoO,N. Found N (%): 3.09 :1: 0.03.
Derived chemical formula is SrMoO,,.N,,,. Thus Mo appears to be in a mixed-valent

state; its average oxidation number is +4.51.

BaMo(O,N), It was obtained as a black powder, also distinctly different from the dark red
of BaMoO,. Samples prepared between 700-950°C were mixtures of at least three
phases. A typical sample contained in addition to BaMo(O,N), 5-10% perovskite-type
BaMoO,, and ~5% of another unidentified impurity which shows broad reflections at d =-
2.500 and 2.465 A. Found N (%): 3.66 :h 0.15. Derived chemical formula is
BaMoOwNM, The real composition of the major phase. typically about 90% in the
mixture, while uncertain, obviously has a high nitrogen content. This result suggests for
molybdenum a higher oxidation state than that in SrMoO,.

The X-ray powder pattern of the BaMo(O,N), phase has been indexed successfully
by the program TREOR on a rhombohedral lattice with the refined hexagonal parameters
a = 5.9701(6) and c = 21.510(4) A. Possible space groups deduced from systematic

' 166

extinctions include 332, 33m, and 33m. While the structure of this new

rhombohedral phase remains a mystery, a possible model structure BaRuO,. better

described as a nine-layer close packing of BaO, layers (164). It has the space

group 33m. Recently Ba,lr,,Mn,,O27 has been shown by single crystal studies

(1_6_5_) to be lsostructural with BaRuO,. The model structure can be derived by a
rhombohedral distortion of the cubic perovskite structure. Calculated diffraction reflection
intensities based on this model generally matched those observed. Miller indices,
observed and calculated interplanar d spacings and intensities are listed in Table 39.
However, the structure of BaMo(O,N), could not be refined to a satisfactory degree by the
X-ray Rietveld procedure. Refinements have been tested with both programs XRS-82 and
DBW32S (mentioned in Part l). Refinement did converge; final R, was 14%. However,
the isotropic temperature factor of Mo(1) at (0.0, 0.0, 0.5) was unusually high (U~0.42
compared to ~0.01 for Mo(2). Alternatively, when it was allowed to refine with a fixed
overall temperature factor, the occupancy factor of the Mo(1) site refined to ~0.1. Such
a low occupancy suggests a chemical formula close to Ba,Mo,(O,N),,,,, which appears

to contradict mass loss data obtained upon NH, reduction of BaMoO,.

167

 

 

 

Table 39. Miller indices, and observed and calculated interplanar d-spacings and
intensities for BaMo(O,N),.

'1 k | dc(A) dslA)" I.” I.” h k | dc(A) do(A) i. '0
0 0 3 7.1701 7.1737 7 <1 2 0 11 1.5596 1.5614 0 1
1 0 1 5.0396 5.0396 4 <1 3 0 6 1.5533 --- 20
0 1 2 4.6598 --- 6 0 2 2 0 1.4925 1.4929 23 23
1 0 4 3.7293 3.7293 3 1 0 1 14 1.4728 1.4741 <1 2
0 0 6 3.5851 --- 5 <1 2 1 10 1.4464 1.4464 29 34
0 1 5 3.3070 3.2995 67 98 0 0 15 1.4340 1.4343 2 5
1 1 0 2.9850 2.9802 100 100 3 0 9 1.3979 1.3991 <1 2
1 1 3 2.7558 5 0 2 2 6 1.3779 --- 2 <1
1 0 7 2.6415 ---- 3 <1 3 1 5 1.3604 1.3607 12 22
2 0 2 2.5135 2.5150 0 5 2 0 14 1.3208 1.3215 <1
0 2 4 2.3299 2.3302 10 7 1 0 16 1.3011 --- 2 O
1 1 6 2.2940 2.2931 11 1 1 15 1.2926 1.2924 10 23
2 0 5 2.2159 2.2136 46 57 2 2 9 1.2660 1.2667 0 3
1 0 10 1.9860 1.9854 40 35 4 O 4 1.2568 1.2570 1 2
0 2 7 1.9782 --- 3 0 0 4 5 1.2379 1.2382 7 10
1 1 9 1.8657 1.8662 3 1 2 14 1.2078 1.2082 <1 2
2 0 8 1.8635 --- 2 0 1 3 10 1.1932 1.1932 15 16
1 2 5 1.7792 1.7787 24 56 2 3 5 1.1435 1.1434 7 12
3 0 0 1.72341.7234 22 27 4 1 0 1.12821.1281 12 12
0 2 10 1.6535 1.6533 25 27 4 O 10 1.1079 1.1065 7 6
1 2 8 1.5808 1.5816 <1 1 2 1 16 1.1076 --- 2 0
Note. 1) d, Guinier data;

2) I, calculated from the BaRuO, model;

3) l, diffractometer data.

1 68
5.2.2 Reaction Conditions

Two independent reactions can be described as follows:

6MMOO,(S) + 8NH,(g) --------> 6MMOO,N(S) + 12H,O(g) + N,(g) (1)

3MMOO,(s) + 2NH,(g) -------- > 3MMOO,(S) + 3 H,O(g) + N,(g) (2)

(1) is a reductive substitution that yields Mo’”; (2) is a simple reduction similar to the
hydrogen reduction that produces Mo“ perovskites. Apparently both reactions occurred
for BaMoO, since BaMoO, was observed. It partial reductive substitution occurs, a mixed-
valent oxidenitride with Mo‘“ will be obtained. It is not clear whether SrMo(O,N), is a
mixed-valent compound or simply a mixture of SrMoO, and SrMO,N. But BaMo(O,N),
must either be a mixed-valent compound or a reductively substituted oxidenitride
BaMoO,.N with Mo". For the NH, reactions with BaMoO, temperature appears to a very
important factor in detenninlng the relative reaction rates. Numerous experiments at
various temperatures performed to synthesize the pure oxidenitride met with little success.
Too high a reaction temperature (2750°C) increases the yield of BaMoO,. When the
temperature was 700-900°C, not only was BaMoO, formed but another unknown reaction
also occurred. When it was lowered to 650°C, the reaction proceeded slowly with the
formation of essentially a single phase; that is, the rhombohedral oxidenitride. However,
the reaction was incomplete even in 6 days; it seemed to have stopped after the first 2~3
days of reaction, probably because of a diffusion problem. Subsequent regrinding and
heating or raising the temperature unavoidably lead to formation of impurity phases.
Efforts to synthesize mixed metal oxidenitrides Ba.Sr,.,,Mo(O,N), produced mixtures of the
cubic and rhombohedral phases. The relative composition varies with the reaction

169
temperature and the Ba/Sr ratio. The cubic phase predominates even with Ba/Sr up to

3. Preparation of a single cubic phase at a small Ba/Sr ratio appeared possible. however,
the weight loss indicated that unlike SrMoO, in which nearly one mole of oxygen can be
substituted by nitrogen simple reduction predominates for Ba,Sr,,,MoO,.

5.2.3 Magnetic and Electrical Properties

Both SrMo(O,N), and BaMo(O,N), are feebly paramagnetic and magnetic
susceptibilities are field dependent. The extrapolated and diamagnetically corrected
susceptibilities are plotted against the temperature in Figures 25 and 26. Room
temperature susceptibilities of SrMo(O,N), and BaMo(O,N), are 123x104 and 5.80x10‘
emu-mor‘, respectively, and increase only about 5% down to 90 K. These values are
significantly lower than those expected for localized d‘ or d’ electronic configurations and
are also lower than susceptibilities of the metallic oxides (1L6), SrMoO, (2.01x10‘) and
BaMoO, (2.15x104 emu-mor‘). The difference in magnetic behavior between the
oxidenitrides and oxides becomes more apparent in the low temperature range.
Susceptibilities of the oxides are constant to within 3% from room temperature to that of
liquid helium; the oxidenitrides exhibit interesting features below 90 K. First, there is a
peak around 50 K on curves of both SrMo(O,N), and BaMo(O,N),. It was suspected to
be from liquified molecular oxygen (mp 54.35. bp 90.18 K, from reference 1_6_6_).
However, repeated experiments with samples carefully purged prior to measurements
indicated that the peak results from the specimens. It probably represents an anti-
ferromagnetic transition. BaMo(O,N), has another weaker peak around 30 K. Since
among the three phases present in the BaMo(O,N), mixture BaMoO, has a temperature-
independent paramagnetism, the different magnetic behavior observed is probably
principally that of the oxidenitride. Secondly, the magnetic susceptibility of SrMo(O,N),

170
increases rapidly with decreasing temperature below the transition point and appears to

obey the Curie law as shown in Figure 27. The susceptibility of BaMo(O,N), at the lower
temperature region is only slightly higher than that at higher temperatures.

The feeble and nearly temperature-independent paramagnetic susceptibilities of
BaMo(O,N), and SrMo(O,N), above 90 K suggest metallic behavoir. Typical powder
resistivity values are: for SrMo(O,N),, 1.50x10‘2 n-cm at 294 K and 1.77x10’2 item at
86 K; for BaMo(O,N),, 0.416 Ocm at 295 K and 0.652 0cm at 85 K. The resistivity of
both samples increases with decreasing temperature. This is a typical feature of a
semiconductor. However, the non-linear relationships between In a and VT shown in
Figures 28 and 29 do not support the semiconducting model. It is believed that observed
resistances are predominately contact resistances, especially for the BaMo(O,N),
specimen because it contained a small amount of insulating BaMoO,. Since sintering
tends to decompose the oxidenitrides. we have not been able to get more accurate data.
When all the experiments are considered, the fairly low resistivities of SrMo(O,N), are

considered an indication of metallic conductivity.

4.2

s» s» s»
O 48 (I)

N
0)

on

Susceptibility x 104 (emu moi—1)
i S

'0

Figure 25.

171

 

 

 

 

I I I I T
50 100 150 200 250

Temperature (K)

Magnetic behavior of SrMo(O,N), at 5300 K.

T
300

 

350

240
220

x~<200

.0“
o

9’
o

5‘
c:

120

100

9"
c:

Susceptibility x 105 (emu mol
'0

.4‘
c:

Figure 26.

172

 

 

 

 

 

I I I I I
50 100 150 200 250

Temperature (K)

Magnetic behavior of BaMo(O,N), at 10-300 K.

I
300

350

173

 

4.20

3.80 —-

s» s»
O 48
O O
l i

I"

at

o
l

00
o
J

Susceptibility x 104 (emu moi”)
6 a
I l

 

 

I I

 

'o
o

I ' I
0.0 40.0 80.0 120.0

1000/T

I l
160.0 200.0

Figure 27. Magnetic susceptibility of SrMo(O,N), vs. reciprocal temperature.

174

 

4.25

4.20 —

4.15 '-

Inc

4.10 a

4.05 '-

 

 

 

3.0 5.0 7.0 9.0 11.0 13.0

1000/T

Figure 28. Conductivity data of SrMo(O,N), at 86-294 K plotted as in 0 against
reciprocal temperature.

175

 

0.9

0.8 —

0.7 -

lncr

0.6 -

 

 

 

3.0 5.0 7.0 9.0 11.0 13.0
1000/T

Figure 29. Conductivity data of BaMo(O,N), at 85295 K plotted as In 0 against
reciprocal temperature.

176
5.3 Pseudo-Binary LaO,N,_, Oxidenitrides

Both compositions (Page 149) yielded cubic single-phase products; specimen I was
dark blue with a - 5.2540(5) A while II was deep red with a - 5.2273(6) A. Similar colors
were reported by Brown and Clark (1_3_Z). Upon exposure to air the products decomposed
gradually with evolution of NH,(g). The X-ray powder diffraction lines were fairly broad,
indicative of inhomogeneous solid solutions over a certain range of compositions.
Approximate compositions deduced from a Vegard-law plot of LaO,N,_, - lattice parameter
data (_1_31) indicated the formulas LaO,,,N,,, (I) and LaO,,,N,,, (II), respectively. Since
the lattice parameter remained invariant at Q as 5.223 A when 5 2 0.46 in LaO,,N,,,, the
oxygen content in II, whose lattice parameter is slightly less than this invariant value, may

be actually higher than 0.45.

5.3.1 Electrical Properties

The resistivity-temperature data of l and II are shown in Figures 30 and 31,
respectively. Both samples exhibit characteristic metallic conductivity similar to that of
Sodium (Figure 21, Page 143) in the temperature regions 6-300 K. In this normal
conducting state the resistivities of both specimens decrease linearly with decreasing
temperature to the superconducting transition point where a sharp transition occurs at T,
= 5.6-6 K for LaOmNm and at T, . 4—5 K for LaO,_,,N,,,. Comparison of these values
to the LaN T,, 1.3 K @4) suggests that the superconducting transition temperature
increases with increasing oxygen content.

The LaO,,,N,,, specimen is more conducting than LaO,,,N,72 at the normal
conducting state. At room temperature the resistivity of the former is nearly two orders
of magnitude less than that of the latter, while LaO,,,N,,, has a greater thermal coefficient

than LaO,,,N,,,. The greater conductivity value with higher oxygen content suggests that

177

an increase in the oxygen content increases the number of electrons in the conduction
band and thus confirms the forrnulatlon La"O,N,,,§,. Similar behavoir has been observed
for NdO,N,., (l3_9).

Even though powder resistivities are usually greater than single crystal resistivities,
our data on sintered samples are considered quantitative because the temperature and
composition dependences and even the zero resistance behavior at the superconducting
state are well resolved. Therefore contact resistances are negligible. Based on the
deduced compositions, the lattice parameters and the free electron formulation, electron
mobilities can be estimated as discussed Chapter 3. Mobilities calculated for the two
samples at three different temperatures are listed in Table 40. These data suggest that
LaO,,,N,,, is nearly a free carrier conductor. The reason for the substantially lower

mobilities of the less oxygen-rich LaO,,,,N,,, is not clear.

Table 40. Estimated Mobilities of Oxidenitrides LaO,N,,,.

 

 

Compound Electron dens. Mobility (cm’volt" sec")
electronscm“ 10 K 150 K 300 K

LaO,,,N,,, 7.72x10" 3.04 0.376 0.166

LaO,,,N,,, 126x10” 23.3 13.4 6.39

 

The temperature dependence curves of the resistivities of both samples are linear
but describable as two straight lines of different slopes above the T,’s. The cross point
of the two straight lines is ~50 K for LaO,,,N,,, and ~85 K for LaO,,,N,,,. Below these
temperatures the thermal coefficients are smaller than those above it. This transition is
probably related to a structural change. such as an anion ordering as was observed in

CeO,,N,,,I by electron diffraction (138) or a cubic to tetragonal transformation as was

178
observed in NdOmN”, (139). Structural studies, especially using neutron or electron

diffraction as a function of temperature, are necessary to reveal the changes.

 

 

 

 

5.0
/.
-1 /
/
/

A c
€4.0" /

Q /

l 4 I

I

E .x‘

03.03 .1
V I
n

O 4 //

"‘ 2

x204 /

I.

a? . .r'

.2 J

“H I

.31 0—1 I

a) l

O: , ,/

a...”
f
0.0")" I

I I I I I
0 50 100 150 200 250 300 350
Temperature (K)

Figure 30. Electrical resistivity of LaO,,,N,,, vs. temperature at 5-300 K.

179

 

 

 

 

 

 

 

 

6.0
A50 -1 '
E .
U
E 4.0 4 °
.C
O ..
v
“’0 .3 0 ‘i '3 O
"I ‘ ”‘0‘...- - -' " no
X / l _2 O '—
320 a l x
.->- I l 3‘
2.: -| ,‘ -1 0 :g
.3 l l E
G) 1.0 “I I 8
C: I :::;.L I r O O D:
a 0 10 20
I K
0 O “'1'“ T F r l T r i r i I i I r
0 50 100 150 200 250 300 350

Temperature (K)

Figure 31. Electrical resistivity of LaO,_,,N,,, vs. temperature at 2300 K.

1 80
5.3.2 Magnetic Properties

The molar susceptibility-temperature curves for LaOmNm and LaOmNm are
depicted in Figures 32 and 33, respectively. The flux exclusion phenomenon is observed
at the superconducting state with transition temperatures consistent with those observed
in the electrical measurements.

The temperature dependence of the susceptibility in the normal conducting state
appears complicated. it remains almost constant above ~40 K for LaOmNm, and above I
~80 K for LaOmNm as can be seen from Figures 32 and 33; between these
temperatures and the superconducting transition temperatures Curie behavior is observed
as shown in Figures 34 and 35. However, the susceptibility is apparently field-dependent.
indicating the presence of ferromagnetic impurities in the sample. Such a dependence
is illustrated for It in Figure 36 together with the diamagnetically-corrected and zero
reciprocal field-extrapolated susceptibilities, which should be the true paramagnetic
susceptibilities of the samples. The Curie type susceptibilities in the low temperature
regions probably result from a very small amount of paramagnetic impurity. The nearly
constant and small extrapolated susceptibilities in the higher temperature regions are in
agreement with a Pauli-type electron gas model of metallic materials. However. the
susceptibility values of 7 - 10 x 1045 emu-moi" are significantly higher than expected.
Using the electron densities given in Table 40, we can estimate the magnetic susceptibility
from the Pauli-Peierls equation discussed in 2.2.2e (Page 134). Even neglecting the
diamagnetic term (-rn,’/3m°’) in the computation, with a volume of 21.50 cm°-mor‘, the
molar Pauli susceptibility tor LaOmNm would be about 1.1 x 10“, at least seven times
lower than the observed value. The observed high susceptibilities are probably an
indication of a narrow 5d 9. band structure (jig).

Metal nitride superconductors with the NaCI-type structure have long been known.

181
Among them, NbN has the highest Te (~18 K) (128). Phillips applied the bond length

anomalies arising from anharmonic stabilization of harmonic lattice instabilities m and
three golden coordinates. the suitable averaged valence-electron numbers N, the orbital
radii differences AR, and the metallic electronegativity differences AX (lg). to analyze
the superconducting transition temperatures. The above described oxidenitrides LaO,,NHI
are a group of new materials. it appears that the increase of superconducting transition
temperature of LaO,N,., is due to the increased electron densities. Further studies on
related oxidenitrides and measurements of the electronic and magnetic properties of
La(lll)Og at low temperatures are necessary to facilitate our understanding of the
superconductivity of oxidenitrides and to search for ways of increasing the transition

temperatures.

182

 

0.80

.0
Ci
0
1

 

 

 

 

 

0.40 — . r1-0

1““ 0- ~ n

r O

. -o.o "

‘I’
if ‘ X

0.20 ~ , ——1 o

I
" T— T r I I 15-20
0 10 20 30
K

 

 

Susceptibility x 103 (emu moi-1)

 

.0
o
o

l I

I I I l
o 50 100 150 200 250 300 350
Temperature (K)

Figure 32. Magnetic susceptibility of LanNm vs. temperature at 2-300 K (H, = 200
gauss).

183

 

 

 

 

 

 

 

 

6.0
i
f 0.5
I“ f .
I «1‘
65.6 't ,...._. ~0.0 O
E :, [I _ a;
3 I. ’
, --a.5
S 1. v” -
V l
V 4.9“ L r I I I i ‘1.0
O i‘ O 4 8 12 A
t K
X \
—- \
glad \
a. “~~~
m K~‘
U “."‘--.--_
U) -----—.
3
U)
2.0 1 r r r i i
O 50 100 150 200 250 300 350

Temperature (K)

Figure 33. Magnetic susceptibility of Laomlilm vs. temperature at 2-300 K (H, = 200
gauss).

184

 

8.0

.0’ .\‘
O O
1 l

.0"
O
1

Susceptibility x 104 (emu moi-1)

 

 

 

.4"
o

0 100 200
lOOO/T (K)

Figure 34. Magnetic susceptibility of Laoo 1”No.72 plotted against reciprocal temperature
to illustrate the Curie behavior at low temperatures (Ho = 200 gauss).

185

6.0

 

A
O
l

Susceptibility x 104 (emu moi")

 

 

 

I\)
Q

T i

so 100 150
1000/T (K)

C

Figure 35. Magnetic susceptibility of LaOmNm plotted against reciprocal temperature
to illustrate the Curie behavior at low temperatures (Ho =- 200 gauss).

186

 

.A.

N (0)

9’
o
l

 

 

T" (b)

1i
0

' we)

P
o
J

Susceptibility x 104 (emu moi-1)
#- .
O
1 l
—i fr

il

' . = 4(a)

T T T T
100 150 200 250 300 350

Temperature (K)

 

.0
o

 

 

O
O"
0

Figure 86. Magnetic field and temperature dependences of the magnetic susceptibility
of Laome above T,. Applied magnetic field is: (a) 5000 gauss; (b) 1

teslas; (c) 2 teslas; (d) extrapolated to H —> oo.

CONCLUDING REMARKS

Oxidenitrides are potentially interesting magnetic. electrical, or dielectric materials.
The preliminary results presented in this work indicate some possible directions of this
research area. LaO,N,_, are especially interesting because of their superconducting
transitions at low temperatures. Oxidenitrides of non-ianthanide elements without f
valence electrons such as Y0,N,,, and ScO‘N”, if they exist. are expected to have
interesting electrical and magnetic properties similar to those of Lacy” if the “free
electrons“ are responsible for these properties. Further studies are necessary to
understand the superconducting mechanism and to increase the superconducting
transition temperature Tc. Transition elements such as Fie, Mn, Fe. and Cs exhibit
multiple valences and form perovskite-type BaMO, oxides. These oxides are likely
candidates for nitrogen substitutions. Neutron diffraction studies are desirable for

structural studies of these oxidenitrides.

187

GENERAL REFERENCE BOOKS

C. N. Fl. Rao and J. Gopalakrishnan. “New Directions in Solid State Chemistry.
Structure, synthesis, properties, reactivity and materials design“. Cambridge
University Press. 1986.

A. K. Cheetham and P. Day, “Solid State Chemistry. Techniques”. Clarendon
Press. 1987.

D. Brown, “Halides of the Transition Elements. Halides of Lanthanides and
Actinides". John Wiley & Sons Ltd., 1968.

M. F. C. Ladd and R. A. Palmer. ”Structure Determination by X-Flay
Crystallography', Plenum Press, Second Edition, 1985.

G. H. Stout and L H. Jensen, 'X-Fiay Structure Determination. A Practical Guide".
1972.

B. D. Cullity, “Elements of X-Ray Diffraction“, Addison-Wesley Publishing
Company, Second Edition, 1978.

H. D. Megaw, “Crystal Structures: A Working Approach", W. B. Saunders
Company, 1973.

R. W. G. Wyckoff, "Crystal Structures”, Vol. 1, lnterscience Publishers. Second
Edition, 1963.

J. S. Blakemore. “Solid State Physics“, Cambridge University Press, Second
Edition, 1985.

188

IN |-*

lib I00

I01

I03

IV

IQ

g

189

REFERENCES

M. S. Whittingham. Prog. Solid State Chem. 12, 41-99 (1978).

W. A. England. J. B. Goodenough, and P. J. iMseman, J. Solid State Chem. 49,
289 (1983).

J. R. Giinter, J. Solid State Chem. 5, 354-359 (1972).

L. Ganapathl, A. Ramanan. J. Gopalakrishnan. and C. N. R. Rao, J. Chem. Soc.,
Chem. Commun., 6263 (1986).

T. Kudo. A. lshikawa. H. Okamoto, and K. Miyauchi, J. Solid sate Chem. 77, 412-
415 (1988).

A. K. Ganguli. L. Ganapathi, J. Gopalakrishnan, and C. N. Fl. Rao. J. Solid State
Chem. 74, 228-231 (1988).

A. Le Bail, C. Jacoboni, M. Leblanc. R. De Papa, H. Huroy. and J. L. Fourquet,
J. Solid State Chem. 77. 96-101 (1988).

A. Haase and G. Brauer, Z. Analy. AIIg. Chem. 441. 181-195 (1978).

A. Haase and G. Brauer, Z. Anorg. AIIg. Chem. 450. 36-44 (1979).

M. D. Taylor and C. P. Carter, J. Inorg. Nucl. Chem. 24, 387-390 (1962).

G. Meyer and P. Ax, Mater. Res. Bull. 17, 1447-1455 (1982).

G. Meyer, T. Staffei, S. Dbtsch, T. Schleid, Inorg. Chem. 24, 3505-3507 (1985).
G. Meyer. S. Dbtsch and T. Staffel. J. Less-Common Metals 127, 155-160 (1987).
M. M. Banaszak Holl. M. Kersting, B. D. Pendley and P. T. Wolozanskl, Inorg.
Chem. 29, 1518-1526(1990); M. M. Banaszak Holl, P. T. Wolczanski. G. D. Van
Duyne, and A. T. Harrison, 1990 ACS meeting, Boston, Inorg Chem Division,
abstract no. 143.

G. Davies. 8. C. Giessen and H. L Shao, 1990 ACS meeting. Boston, lnorg Chem
Division. abstract no. 351.

In

It}?

I?

ISISIBISIER

I“ lm l‘é’

I‘tilc‘fi

I32

190

J. G. Brennan. T. Slegrist, S. M. Stuczynski, and M. L. Steigerwald, J. Am. Chem.
Soc. 111. 9240-9241 (1989).

M. L. Steigerwald, Chem. Mater. 1, 52-57 (1989).

K.-W. Kim, S. Dhingra and M. G. Kanatzidis, 1990 ACS Meeting, Boston, lnorg.
Chem. Division. abstract no. 142.

K. Rossmanith. Monatsh, Chem. 110, 109-114 (1979).

A. F. Clifford and K. J. Hughes. In Rare Earth Research Ill; L Eyring, Ed., Gordon
and Breach. 1965.

A. F. Clifford, in Rare Earth Research ll. K. S. Vorres. Ed.; Gordan and Breach,
New York, 1963:1110. 45-50.

C. Brauer and B. Pfeiffer, J. Less-Common Met. 5, 171-176 (1963).

R. Giovanoli, W. Feitknecht, R. Maurer. and H. Haeni, Chimia 30, 307-309 (1976).
M. M. Thackeray and A. De Kock, J. Solid State Chem. 74, 414-418 (1988).

R. Chevrei. M. Sergent and J. Prigent. Mater. Res. Bull. 9, 1487 (1974).

K. Rossmanith and C. Auer-Welsbach, Monatsh. Chem. 96, 602-605 (1965).

a. Uu and H. A. Eick, Inorg. Chem. 27, 2161-2163 (1988).

R. D. Shannon, Acta Cryst. A32, 751-767 (1976).

O. Greis and J. M. Haschke, Rare Earth Fluorides, inHandaook on the Physics
and Chemistry of Rare Earths, Vol. 5, Chapter 45, pp 387-460. Eds. K. A
Gschneldner, Jr. and L Eyring, North-Holland. 1982.

S. E. Ness, D. J. M. Bevan. H. J. Rossell, Eur. J. Solid State Inorg. Chem. 25.
509-516 (1988).

D. J. M. Bevan. M. J. McCall, S. E. Ness, and M. R. Taylor. Eur. J. Solid State
Inorg. Chem. 25. 517-526 (1988).

D. J. M. Bevan. S. E. Ness, and M. R. Taylor, Eur. J. Solid State Inorg. Chem. 25,
527-534 (1988).

C. Rinck, Ph. D. Dissertation, University of Karlsruhe, 1982, and references therein.

D. M. Laptev. V. *F. Goryushkin. l. S. Astakhova and G. G. Polyakova. Russ. J.
Inorg. Chem. (Engl. Transl.) 24, 730-733 (1979).

H. BAmlghausen, J. M. Friedt and J. P. Sanchez, Int. Rare Earth Conf., ETH
Ziirlch, Switzerland March 3-8, 1985; private communication. J. F. Sanchez, M.

l8

IE}

Ialfilfilal‘to’lg

I: IS IS It

I‘AISIS'HBISIE’IISISIS

191

Malki. J. M. Friedt. and H. Barnighausen. Hyperfine Interactions 28. 811-814
(1986).

S. A. Hodorowicz, M. Olelak-Chodan. and H. A. Eick. J. Solid State Chem. 71.
205-213 (1987).

M. Olejak-Chodan, S. A. Hodorowicz. and H. A. Eick, Cryst. Res. Tech. 22, 285-
289 (1987).

U. LOchner, Dissertation. Universitét Karlsruhe, 1980.

H. Barnighausen, Proc. Rare Earth Res. Conf., 12th. 1, 404-413 (1976).

U. LOchner, H. Bamighausen, and J. D. Corbett, inorg. Chem. 16. 2137 (1977).
H. Luke and H. A. Eick. Proc. Rare Earth Res. Conf., 12th, 1, 424 (1976).

H. Lake and H. A. Eick. lnorg. Chem. 21, 965 (1982).

G. J. McCarthy, J. J. Rhyne and H. B. Silber, The Rare Earths in Mad. Sci. and
Tech, Vol. 2, Plenum Press. New York. 1978.

M. Eitel. Ph.D. Dissertation, University of Karlsruhe, 1985.

H. Barnighausen, Rev. Chim. Min. 10, 77-92 (1973).

H. P. Beck. J. Solid State Chem. 47, 328-332 (1983).

D. J. M. Bevan. J. Strahie and 0. Greis, J. Solid State Chem. 44,75-81 (1982).
H. M. Rietveld, Acm Cryst. 22, 151-152 (1967).

H. M. Rietveld. J. Appl. Cryst. 2. 65-71(1969).

A. K. Cheetham and J. c. Taylor, J. Solid State Chem. 21, 253275 (1977).

A. Albinati and e. T. M. Willis. J. Appl. Cryst. 15, 361-374 (1982).

G. Malmros and J. 0. Thomas. J. Appl. Cryst. 10, 7 (1977). >
J. 0. Thomas. N. O. Fresson and Y. Anderson. J. Appl. Cryst. 13, 605-607 (1980).
K. E. Johansson; T. Palm and P. E. Werner. J. Phys. E. 13. 1289-1291 (1980).
P.-E. Werner. Chem. Scripts A26. 57-64 (1986).

J. 0. Thomas. Chem. Scripts A26. 7-10 (1986).

P. Thompson and 1. G. Wood. J. Appl. Cryst. 16. 458-472 (1983).

Is lat-6 is".

1321813

|8l8|8i

IS

Ia‘ IS

I?!

I;I lo‘o‘lRa‘

ldlald

192

A. Clearfield. L a. McCuster. and P. a. Rudolf. Inorg. Chem. 33. 4679-4662
(1964).

P. Rudolf and A. Clearfieid. A018 Cryst. 341. 418-425 (1985).

W. Lasocha and H. A. EICK. J. Solid State Chem. 75. 175-182 (1988); and 77. 90-
95 (1988).

Ch. Banocher. Proc. of the 6th int. Conf. on Zeolites. Reno. U.S.A. P. 823. Eds.
D. Olson and A. Bisio. Butterworths. London. 1984.

J. Paul Attfield. Acta Cryst. 344. 563-568 (1988).
A. G. Casado. A. Mendiola and I. Raslnes. Inorg. Chem. 21. 2902-2903 (1982).

o. E. Smith. J. E. Maskasky. and L. a. Spaulding. J. Appl. Cryst. 15. 488-492
(1982).

G. Uu and H. A. Eick. J. Less-Common Metals 156. 237-245 (1989).
J. E. Berg and P. E. Werner. Z. Kristallogr. 145. 310-320 (1977).

Y. Laligant. A. Le Bail. G. Ferey. M. Hervieu. B. Raveau. A. Wilkinson and A. K.
Cheetham. Eur. J. Solid Sme lnorg. Chem. 25. 237-247 (1988).

P. Lightfoot. A. K. Cheetham and A. w. Sleight. Inorg. Chem. 26. 3544-3547
(1987).

a. A. Young and o. 6. Wiles. J. Appl. Cryst. 15. 430-438 (1982).
A. Hepp and Ch. Barlocher. Aust. J. Phys. 41. 229-236 (1966).

S. H. Wang. a. King. s. H. Lin and T. .Brown. J. Solid State Chem. 69. 224-229
(1967).

G. Uu and H. A. Eick. J. Less-Common Metals149. 47-53 (1989).
G. Uu and H. A. Eick. to be published.

l. S. Astakhova. v. F. Goryushkln. and D. M. Laptev. Russ. J. Inorg. Chem. (Engl.
Transl.) 24. 1096-1097 (1979).

J. W. V1888. J. Appl. Cryst. 2. 89-95 (1962).
P. E. Werner. L Eriksson and M. Westdahl. J. Appl. Cryst. 18. 367-370 (1985).

D. E. Appieman. D. S. Handwerker and H. T. Evans. Program X-ray. Geological
Survey. U. S. Department of Interior. Washington. DC. 1966.

lat

ISISIBISISIB

ISIS-la

IS

IS IS

IS!

l8l8|fl8|$

193
D. K. Smith. M. C. Nichols and M. E. Zolensky. “A FORTRAN IV Program for
Calculating X-Ray Powder Diffraction Patterns. Version 10.‘ The Pennsylvania
State University. University Park. PA.. 1983.
P. M. de Wolff. J. Appl. Cryst. 1. 108-113 (1968).
G. S. Smith and R. L Snyder. J. Appl. Cryst. 12. 60-65 (1979).
H. Barnighausen and E. Th. Rietschel. z. Anory. Allg. Chem. 354. 23-26 (1967).
E. Th. Rietschel and H. BAmighausen. Z. Anom. Allg. Chem. 368. 62-72 (1969).
H. Barnighausen and N. Schultz. Acta Cryst. 825. 1104-1110 (1969).
H. P. Beck. Z. Anorg. Allg. Chem. 459. 81-86 (1979).
H. D. Lutz. W. Buchmeier and B. Engeien. Acta Cryst. B43. 71-75 (1987).
J. Ladell. A. Zagofsky. and J. S. Pearlman. J. Appl. Cryst. 8. 499-506 (1975).

Ch. Barlocher and A. Hepp. "The X-Ray Rietveld System.“ lnstitut f. Kristailogr. u.
Petrographie. ETH. Zfirich. Switzerland. 1982.

J. A. Ibers and W. C. Hamilton. Ed. “International Tables for x-Ray
Crystallography.“ Vol. IV. Kynoch Press. Birmingham. England. 1974.

G. Vogel and A. Schneider. lnorg. Nucl. Chem. Lett. 8. 513-521 (1972).

R. Blachnik. G. Aiberts and E. Enninga. Z. Anorg. Allg. Chem. 522. 207-216
(1985).

JCPDS Powder diffraction File No. 24-95. lntemational Centre for Diffraction Data.
1601 Park Lane. Swarthmore. PA 19081.

R. Bachman. Diplomaibelt. Universiiat Karlsruhe, 1980.

J. M. Haschke. lnorg. Chem. 15. 298-303 (1976).

H. Barnighausen and J. M. Haschke. inorg. Chem. 17, 18-21 (1978).
W. Martin. Diplomarbelt. Universitat Karlsruhe, 1976.

JCPDS Powder diffraction File No. 24-94. International Centre for Diffraction Data.
1601 Park Lane. Swarthmore. PA 19081.

JCPDS Powder diffraction File No. 24-89. Intemationai Centre for Diffraction Data.
1601 Park Lane. Swarthmore. PA 19081. '

S-H. Wang. S-B. Jiang. Z-X. Tang. H-Z. Yang. X-M. Li. Huaxue Tongbao. 654-656
(1981).

|§|8

_A
o
.5

_S
o
N

|§|§|8|

ls

N

IS

l6

0

d
d

_L
'0

l3

h

E

a}

194
S-H. Wang. private communication; 1989.
X-M. Li and S-H. Wang. in New Front. Rare Earth Sci. Appl.. Proc. lntemah‘onai
Conf. on Rare Earth Development and Applications. Vol. 1. pp.338-341. Scientific
Press; 1985.
H. G. Scott. J. Appl. Cryst. 16. 159-163 (1963).
W. C. Hamilton. Acta Cryst. 18. 502-510 (1965).
L. F. Druding and J. D. Corbett. J. Am. Chem. Soc. 83. 2462-2467 (1961).
G. Meyer and T. Schleid, Z. Anorg. Allg. Chem. 528. 55-60 (1985).

L. R. Morss. in Smmm Potentials in Aqueous Solution; A. J. Bard. R. Parsons
and J. Jordan eds; Dekker: New York. 1985; pp 587-629.

H. D. Lutz. H. J. Kluppel. F.-J. Marner. and H.-J. Kesterke. Z. Naturforsch. 283.
8-11 (1973). -

D. M. Laptev. I. S. Astakhova. N. M. Kulagin and N. F. Bomko. Russ. J. Inorg.
Chem. (Engl. Transl.) 21. 643-644 (1976).

D. M. Laptev and l. S. Astakhova. Russ. J. Inorg. Chem. (Engl. Transl.) 26. 914-
915 (1981).

D. M. Laptev. V. F. Goryushkln. N. M. Kulagin. and E. S. Vorontsov. Russ. J.
Inorg. Chem. (Engl. Transl.) 21. 1438-1441 (1976).

D. M. Laptev. N. M. Kulagin and I. S. Astakhova. Russ. J. lnorg. Chem. (Engl.
Transl.) 20. 1108-1109 (1975).

J. a. Bednorz and K. A. Miiller. 2. Phys. 864. 189 (1986).

M. K. Wu. J. R. Ashbum. C. J. Tomg. P. H. Hor. R. L. Meng. L Gao. Z. J. Huang.
Y. Q. Wang. and C. w. Chu. Phys. Rev. Lett. 58. 908 (1987).

J. M. Tarascon. W. R. McKinnon. L H. Greene. G. W. Hull. and E. M. Vogel.
Phys. Rev. B 36. 226-234 (1987).

Y. LePage. T. Siegrist. S. A. Sunshine. L. F.Schneemeyer. D. W. Murphy. S. M.
Zahurak. J. V. Waszczak. W. R. McKinnon. J. M. Tarascon. G. W. Hull. and L. H.
Greene. Phys. Rev. B 36. 3617-3621 (1987).

s. A. Hodorowicz. w. Lasocha. A. Lasocha. and H. A. Eick. J. Solid State Chem.
77. 145-146 (1968).

W. Kang. H. J. Schulz. D. Jerome, S. S. Parkin. J. M. Bassat. and Ph. Odier.
Phys. Rev. B 37. 5132-5138 (1988).

la‘:

_5
.5
‘D

_A
N
03

—L
lie
\I

ISIS

E

195

J. M. Tarascon. L. H. Greene. P. Barboux. W. R. McKinnon. G. W. Hull. T. P.
Orlando. K. A. Delin. S. Foner. and E. J. McNiff. Jr.. Phys. Rev. B 36. 8393-8400
(1987).

D. M. de Leeuw. C. A. H. A. Mutsaers. H. A. M. Van Hal. H. Verwei]. A. H. Carim.
and A. C. A. Smoorenburg. Physica C158. 126-132 (1988).

S. S. P. Parkin. V. Y. Lee. E. M. Engler. A. I. Nazzal. T. C. Huang. G. Gorman. R.
Savoy. and R. Beyers. Phys. Rev. Lett. 60. 2539 (1988).

S. S. P. Parkin. V. Y. Lee. A. I. Nazzal. R. Savoy. T. C. Huang. G. Gorman. and
R. Beyers. Phys. Rev. B 38. 6531-6537 (1988).

F. A. M. Van Woerden and D. M. Deleeuw. Physica C 165. 221 (1990).

M. A. Subramanian. J. Gopalkrishnan. and A. W. Sleight. J. Solid State Chem. 84.
413 (1990).

I. Felner. U. Yaron. Y. Yeshurun. E. R. Yacoby. and Y. Wolfus. Phys. Rev. B 40.
1 1366 (1989).

Y. K. Tao. M. Bonvalot. Y. Y. Sun. R. L. Meng. P. H. Hor. and C. W. Chu. Physica
C165. 13 (1990).

E. A. Early. N. Y. Ayoub. J. Beille. J. T. Markert. and M. 8. Maple. Physica C 160.
320 (1989).

A. Tressaud. B. Chevalier. B. Lepine. K. Amine. J. M. Dance. L Lozano. J. Darriet.
J. Etoumeau. R. Toumier. A. Sulpice. P. Lejay. J. L Diot. and P. Maestro. Solid
State lonics 32l33. 1109 (1989).

B. Chevalier, A. Tressaud. B. Lepine. K. Amine. J. M. Dance. L Lozano. E.
Hickey. and J. Etoumeau. Physica C 167. 97 (1990).

J. 0. Phillips. Phys. Rev. Lett. 26. 543-546 (1971).

R. Marchand. F. Pors. and Y. Laurent. Rev. Int. Hautes Temper. Refract. Fr. 23.
11-15 (1986).

F. Pors. P. Bacher. R. Marchand. Y. Laurent. and G. Roult. Rev. Int. Hautes
Temper. Refract. Fr. 24. 239-246 (1987-1988).

P. Bacher. P. Antione. R. Marchand. P. L'Haridon. Y. Laurent. and G. Roult. J.
Solid State Chem. 77. 67-71 (1988).

P. Antoine. R. Marchand. Y. Laurent, C. Michel. and B. Raveau. Mater. Res. Bull.
23. 953-957 (1988).

F. Pors. R. Marchand. Y. Laurent. P. Bacher. and G. Roult. Mater. Res. Bull. 23.
1447-1450 (1988).

I9 E

I§I§I§

1..

S

d
d

N

1.1.

I5

IE IE IS It; IE

Eli:

IE

IS

196

P. Antoine. R. Marchand. and Y. Laurent. Rev. Int. Hautes Temper. Refract. Fr.
24. 43-46 (1987).

Y. Laurent. R. Pastuszak. P. L'Haridon. and R. Marchand. Acta Cryst. B 38. 914-
916 (1982).

R. Pastuszak. P. L'Haridon. R. Marchand. and Y. Laurent. Acta Cryst. B 38. 1427-
1430 (1982).

R. C. Brown and N. J. Clark. J. Inorg. Nucl. Chem. 36. 2287-2290 (1974).

R. C. Brown and N. J. Clark. J. Inorg. Nucl. Chem. 36. 1777-1782 (1974).

C. Mourgout. B. Chevalier. J. Etoumeau. J. Portier. P. Hagenmuller and R.
Georges. Rev. Int. Hautes Temper. Refract, Fr. 14. 89-96 (1977). and references
therein.

B. Chevalier. J. Etoumeau. and P. Hagenmuller. Mater. Res. Bull. 12. 473-479
(1977)

J. Etoumeau. B. Chevalier. P. Hagenmuller. and R. Georges. J. Phys. Colloq. c
5. 193-204 (1980).

B. Chevalier. J. Etoumeau and P. Hagenmuller in The Rare Earths in Mod. Sci.
and Tech.. Eds. G. J. McCathy. J. J. Rhyne. and H. B. Silber. 1980. pp 403-408.

F. A. Cotton and G. Wilkinson. “Advanced inorganic Chemistry.“ 4th ed. John
Wiley & Sons. inc.. 1980. p 847.

R. Scholder and L. H. Brixner. Z. Naturforsch.. 10 D. 178-180 (1955).

L. H. Brixner. J. Inorg. Nucl. Chem. 14. 225-230 (1960).

G. H. Bouchard. Jr.. and M. J. Sienko. Inorg. Chem. 7. 441-443 (1968).
U. Von Essen and W. Klemm. 2. Mary. Ally. Chem. 317. 25-34 (1962).

M. G. Barker and l. 0. Alexander. J. Chem. Soc. Dalton Trans. (20). 2166-2170
(1974). '

N. M. Maister and I. M. Lopato. lnorg. Mater. (Eng. Transl.) 9(1). 57-60 (1973).
G. L Oicese. J. Phys. F. 9. 569-78 (1979).

J. M. Leger. N. Yacoubi. and J. Loriers in The Rare Earths in Mad. Sci. and Tech..
Eds. G. J. McCathy. J. J. Rhyne. and H. B. Silber. 1980. pp 203-208.

J. M. Leger. N. Yacoubi. and J. Loriers. J. Solid State Chem. 36. 261-270 (1981).

I8

—5
0'1
(’1

Iii

—L
lei
‘1

93

_5
(II
‘0

IS

197

M. R. Norman. H. J. F. Jansen. D. D. Keolling. and A. J. Freeman. Solid State
Commun. 52. 739-741 (1984).

J. J. Veyssie. D. Brochier. A. Nemoz. and J. Bianc. Phys. Left. 14. 261-262
(1965).

P. W. Selwood. 'Magnetochemistry.‘ 2nd ed.. lnterscience Publishers. lnc.. New
York. N.Y. 1956. p 78. '

W. L. Jolly. 'The Synthesis and Characterization of Inorganic Compounds.”
Prentice-Hall. lnc.. Englewood Cliffs. N.J.. 1970.

A. H. Morrish. 'The Physical Principles of Magnetism.“ John Wiley & Sons. inc..
1965.

C. M. Hurd. Contemp. Phys. 23. 469-493 (1982).

A. K. Cheetham. and P. Day. ”Solid State Chemistry. Techniques.“ Clarendon
Press. Oxford. 1987. p. 289-290.

R. W. G. Wyckoff. "Crystal Structures." Vol. 3. 2nd Ed. lnterscience. 1965. pp 87-
91.

H. Mfilier-Buschbaum and H. G. Schnering. Z. Anorg. Allg. Chem. 336. 295-305
(1965).

JCPDS Powder Diffraction File No. 34102. international Centre for Diffraction
Data. 1601 Park Lane. Swarthmore. PA 19081.

JCPDS Powder Diffraction File No. 23-1048. international Centre for Diffraction
Data. 1601 Park Lane. Swarthmore. PA 19081.

P. C. Donohue. L Katz. and R. Ward. Inorg. Chem. 4. 306-310 (1965).

Ch. Lang and Hk. Miiiier-Buschbaum. J. Less-Common Metals 157. 301-306
(1990).

N. N. Greenwood. and A. Earnshaw. “Chemistry of the Elements.“ 1st ed.
Pergamon Press. 1984. (reprinted with corrections in 1986). p. 706.

P. Wllars and J. C. Phillips. Phys. Rev. B 37. 2345-2348 (1988).

ICHIGRN STRT

E UNIV. LIBRARIES
III III IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
6030

3129300914