 

LlBRAl-W
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

Studies on the Multinary Chalcogenides and
Related Compounds Prepared by the Molten Salt Method

presented by

Kyoung-Shin CHOI

has been accepted towards fulﬁllment
of the requirements for

 

 

 

Ph.D. degree in Chemigtrv
‘ t % [Ii/[M
Major professor
mew—

MSU is an Afﬁrmative Action /Equal Opportunity Institution 0-12771

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 Woes-p.14

 

STUDIES ON THE MULTINARY ANTIMONY CHALCOGENIDES AND
RELATED COMPOUNDS PREPARED BY THE MOLTEN SALT METHOD

B y

Kyoung—Shin CHOl

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Chemistry

2000

 

 

 

ABSTRACT

STUDIES ON THE MULTINARY ANTIMONY CHALCOGENIDES AND
RELATED COMPOUNDS PREPARED BY THE MOLTEN SALT METHOD

By

Kyoung-Shin CHOI

In view of the technological potential of the large class of known natural
sulfosalts, we performed exploratory synthesis for new chalcoantimonate compounds in
order to probe the possibility of novel structure types, characterize their physico-
chemical properties, and obtain a systematic understanding of this class of compounds.
The molten polychalcogenide ﬂux method has been essential for such kind of exploration
and resulted in diverse new chalcoantimonate compounds. These ﬂuxes are formed by the
in situ fusion of AzQ/Sb/Q or AzQ/Sb2Q3/Q (A = K, Rb ; Q = S, Se, Te) to produce
various Ax[Sbsz] units. These discrete species can vary in composition and structure
and can form the basis for generating a large number of ternary and quaternary
compounds.

Reaction of f-elements such as actinide and lanthanide metals with Sb in the
polychalcogenide ﬂuxes resulted in many multinary compounds with novel structure
types(i.e. AzUSng(A=Rb, K; Q=S,Se), RbUSb0,33Te6, KTth28e6, KanZSnZSeg(Ln=Ce,
Sm, Gd, Tb, Yb), EquQ3(Q=Se, Te), and Ba0,5Eu0,SSbTe3). Since the f-elements and Sb
prefer different coordination spheres, these compounds present interesting structural

features with unique distortions. The reactivity of Pb with Bi/Sb in the polyselenide

ﬂuxes were studied because of the potential thermoelectric properties of the resulting
compounds. This investigation resulted in the discovery of a new class of compounds,
A1+be4.2xM7+xSe15(A = K, Rb; M = Sb, Bi), which exhibit interesting thermoelectric
properties with very low thermal conductivities. Several new sulfoantimonates with
alkaline earth metals(SrGSb6Sn, Ba3Sb4,6-6Sw, Ba23ngL6sz4Sm) were obtained by
replacing alkali metals in the ﬂuxes with alkaline earth metals. Lastly a few novel ternary
thorium selenides(AThZSe6(A=K, Rb), BaTh7Se13), which feature interesting patterns of
Se-Se interactions, were prepared and characterized.
In this dissertation, the synthesis, structure, and physicochemical properties of
these new phases will be reported. The synthetic work performed here provides the

ground for further exploration systematic study of these and related systems.

ACKNOWLEDGMENTS

First of all, I would like to thank my advisor, Professor Mercouri G. Kanatzidis
for his invaluable guidance, encouragement, and support. His enthusiasm and passion for
science has been an inspiration to me for the last ﬁve years. He not only taught me
chemistry but also showed me the attitude that true chemists should have towards their

research, for which I have been and will always be deeply grateful.

I would like to thank Professor Thomas J Pinnavaia, Professor Gary J. Blanchard,
and Professor Gregory L. Baker for serving as members of my graduate committee. I
would also like to thank Professor Carl Kannewurf and Professor Ctirad Uher for the
charge transport property measurements, Professor Simon J. L. Billinge for x-ray diffuse
scattering studies, and Professor Marcos Dantus for atomic force microscopy studies.

During my graduate study, I was fortunate enough to have so many talented,

smart, and kind people as my lab-mates in the Kanatzidis group. I appreciate their
support, companionship, and all that I have learned from them. I especially owe a great
deal to Dr. Duck-Young Chung, who provided me with so many useful suggestions and
continuous encouragement throughout my studies. I thank my family and friends for
their love and support.

Finally the Oﬁice of Naval Research and the National Science Foundation are

gratefully acknowledged for ﬁnancial support.

 

 

TABLE OF CONTENTS

12g:
LIST OF TABLES .................................................... x
LIST OF FIGURES ................................................... xvi
LIST OF ABBREVIATIONS xxii

Chapter 1. The Rationale for Exploratory Synthesis of Multinary Antimony
Chalcogenide Compounds Using Molten Alkali Metal Polychalcogenide Fluxes. . . . 1

1. Introduction ................................................... 2
2. Polychalcogenide Flux Method .................................... 4
3. Synthetic Strategy for Multinary Chalcoantimonates ................... 10
4. Thermoelectric Relevance of Sb/Bi Chalcogenides ...................... 12

Chapter 2. Reactions of Uranium Metal with Sb in Polychalcogenide Fluxes;

Discovery of Novel Quaternary Noncentrosymmetric Uranium C halcoantimonates,

AUszQ8 (A = Rb, K; Q = S,Se) and RbUSba33Te6 ............................ 22
1. Introduction ................................................... 23
2. Experimental Section ............................................ 25

2.1 Synthesis ............................................... 25

2.2 Physical Measurements .................................... 26

2.3 X-ray Crystallography ..................................... 29

3. Results and Discussion .......................................... 50
A. RbUZSng and KUZSbSeg ..................................... 50
Structures ............................................... 50
Properties ............................................... 53

B. RbUSbo_33Te6 .............................................. 58
Structure ................................................ 58
Superstructure ........................................... 65
Properties ............................................... 72

4. Conclusions ................................................... 76

 

Chapter 3. Reactions of Thorium Metal with Sb in Polyselenide Fluxes;
A New Quaternary Thorium Selenoantimonate, K Tth 25e6 Featuring One-

Dimensional Double Th Chains Constructed By Dichalcogenide Group ........... 80
1 . Introduction ................................................... 8 1

2. Experimental Section ............................................ 82

2.1 Synthesis ............................................... 82

2.2 Physical Measurements .................................... 83

2.3 X-ray Crystallography ..................................... 85

3. Results and Discussion .......................................... 85

4. Conclusions ................................................... 94

Chapter 4. Reactions of Rare Earth Metals with Sb in Polychalcogenide Fluxes;
Discovery of the Kan28b2889 Family (Ln=Ce, Sm, Gd, Tb, Dy) with a Eight-F old
Superstructure Caused by Three-Dimensional Ordering of the 552 Lane Pair of Sb3 +

Ions. ............................................................... 97
1. Introduction ................................................... 98
2. Experimental Section ............................................ 99

2.1 Synthesis ............................................... 99
2.2 Physical Measurements .................................... 104
2.3 X-ray Crystallography ..................................... 105
3. Results and Discussion .......................................... 110
Substructure ............................................. 110
Superstructure ........................................... 131
Properties ............................................... 136
4. Conclusions ................................................... 144

Chapter 5. Reactions of Europium Metal with Sb in the Polychalcogenide Fluxes;

Novel Europium Chalcoantimonates, EquSe3, 15qu Te 3, and Baa 5Eu0_ 5SbT e 3 with

Distorted T wo—Dimensional Square Se/T e Nets .............................. 147
1. Introduction ................................................... 148

vi

2. Experimental Section ............................................ 150

2.1 Synthesis ............................................... 150

2.2 Physical Measurements .................................... 152

2.3 Electron Diffraction Study (TEM) ........................... 155

2.4 X-ray Crystallography .................................... 155

3. Results and Discussion ....................................... ,. . 157
Structures ................................................ 170

EquSe3 .............................................. 170

EquTe3 .............................................. 175
Ba0.5Euo,SSbTe3 ........................................ 178

Properties ............................................... 183

4. Conclusions ................................................... 189

Chapter 6. Reactions of Lead Metal with Bi/Sb in Polyselenide Fluxes;
K I +be4.2xM7+,S'e 1 5 (A = K, Rb; M = Bi, Sb), A New Class of Solid State Quaternary

Thermoelectric Compounds with Very Low Thermal Conductivity ............... 193
1 . Introduction ................................................... 194
2. Experimental Section ............................................ 195
2.1 Synthesis ............................................... 195

2.2 Physical Measurements ................................... 197

2.3 X-ray Crystallography .................................... 199

3. Results and Discussion .......................................... 212
Structure ................................................ 212

Energy Band-Gap and Thermal Analysis ...................... 218

Charge Transport Properties ................................ 218

Thermal Conductivity ..................................... 223

4. Conclusions ................................................... 227

Chapter 7. Reactions of Sb Metal in Alkaline Earth Metal Polysulﬁde Fluxes;
Discovery of New Sulfoantimonate Compounds with Alkaline Earth Metals,

vii

 

SrGSb6S, 7, Ba3Sb4_66Sm, and Ba2.62Pb,,3ng4Sm .............................. 231
1 . Introduction ................................................... 232
2. Experimental Section ............................................ 234

2.1 Synthesis .............................................. 234

2.2 Physical Measurements ................................... 235

2.3 X-ray Crystallography .................................... 237

3. Results and Discussion .......................................... 239
Structures ............................................... 239
Sr5Sb6Sm ............................................. 239

Ba3Sb4I66Sm and Ba2.52Pb1,3ng4Slo ......................... 252
Relationship to Sulfosalt Minerals ......................... 258
Spectroscopic Characterization and Thermal Analysis ............ 260

Chapter 8. Charge Density Wave Caused by Reducing ThSe3 by One Electron.
Superstructure and Short Range Order in AThZSe6 (A = K, Rb) Studied by X-ray

Diffraction, Electron Diffraction and Diffuse Scattering ........................ 266
1. Introduction ................................................... 267
2. Experimental Section ............................................ 270

2.1 Synthesis ............................................... 270
2.2 Physical Measurements .................................... 271
2.3 X-ray Crystallography ..................................... 272
2.4 Electron Diffraction Study ................................. 275
2.5 Pair Distribution Function Analysis (PDF) .................... 27 5
2.6 Atomic Force Microscopy (AF M) ........................... 27 5
3. Results and Discussion .......................................... 280
Structure ................................................ 280
Properties ............................................... 284
Electron Diffraction Study and Superstructure Model ............. 286

Probing the Existence of Se-Se Single Bonds in Local Structure
by Diffuse Scattering and PDF Analysis ....................... 291

viii

 

4. Conclusions ................................................... 298

Chapter 9. A Novel Ternary Thorium Selenide, BaTh7Se1 3 Featuring Inﬁnite Linear

Se Chains ............................................................ 304
1. Introduction ................................................... 305
2. Experimental Section ............................................ 306

2.1 Synthesis ............................................... 306

2.2 Physical Measurements ................................... 306

2.3 X-ray Crystallography .................................... 308

3. Results and Discussion .......................................... 308
Structure ................................................ 313
Spectroscopic Characterization and Thermal Analysis ............ 320

4. Conclusions ................................................... 320

ix

 

LIST OF TABLES

bet:
Table 1.1 Melting points for some known alkali metal/
polychalcogenide species (A = alkali metal) ...................... 7
Table 2.1 Summary of Crystallographic Data and Structural
Analysis for RbUzsbsg ...................................... 33
Table 2.2 Summary of Crystallographic Data and Structural
Analysis for KUZSbSeg ...................................... 34
Table 2.3 Fractional Atomic Coordinates and Equivalent Isotropic
Displacement Parameter Values for RbUszSg (Ammm)
and KUZSbSeg (Ammm) with Estimated Standard
Deviations in Parentheses .................................... 35

Table 2.4 Fractional Atomic Coordinates and Equivalent Isotropic Displacement
Parameter Values for RbUZSbsg (Cm)
with Estimated Standard Deviations in Parentheses ................ 36

Table 2.5 Fractional Atomic Coordinates and Equivalent Isotropic Displacement
Parameter Values for KUZSbSeg (Cm) with

Estimated Standard Deviations in Parentheses .................... 37
Table 2.6 Selected Distances (A) and Bond Angles (°) for RbUZSng; (Cm) ...... 38
Table 2.7 Selected Distances (A) and Bond Angles (°) for KUZSbSe3(Cm) ...... 39

Table 2.8 Summary of Crystallographic Data and Structural Analysis
for RbUSbo33T€6 ........................................... 41

Table 2.9 Fractional Atomic Coordinates and Equivalent Atomic Displacement
Parameter Values for RbUSb033Te6 (substructure) with Estimated
Standard Deviations in Parentheses ............................ 42

Table 2.10 Selected bond lengths (A) and angles (°) for RbUSbo33Te6
(substructure) ............................................. 43

Table 2.11 Anisotropic Displacement Parameters for RbUSb0_33Te6 ............ 44

Table 2.12 Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (Ueq) Values for RbUSbo_33Te6

(superstructure) with Estimated Standard Deviations in

 

 

Table 2.13

Table 3.1

Table 3.2

Table 3.3

Table 3.4

Table 4.1

Table 4.2

Table 4.3

Table 4.4

Table 4.5

Table 4.6

Table 4.7

Table 4.8

Table 4.9

Parentheses ............................................... 45

Selected bond lengths (A) for RbUSb033Te6 (superstructure) ........ 47
Calculated and Observed X-ray Powder Pattern of KTthZSef, ...... 84
Summary of Crystallographic Data and Structural Analysis

for KTthzSC6 ............................................. 86

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (8“,) Values for KTthZSe6 with

(superstructure) ........................................

xi

Estimated Standard Deviations in Parentheses .................... 87
Selected Distances (A) and Bond Angles (°) for KTthZSeé ......... 88
Calculated and Observed X-ray Powder Pattern of KzGdZSbZSeg ..... 102
Summary of Crystallographic Data and Structural Analysis
for the substructure and the superstructure of K2GdZSbZSe9 ......... 111
Summary of Crystallographic Data and Structural Analysis
for KanZszseg (Ln = Ce, Sm, Tb, Dy) ......................... 112
Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (Ueq) Values for K2Gd28b2$€9
(substructure) ............................................. 1 l4
Anisotropic Displacement Parameters (A2 x 103) for
K2Gd28b28e9 (substructure) .................................. 115
Selected Distances (A) and Bond Angles (°) for K2Gd28b28e9
(substructure) ............................................. 1 l6
Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (ch) Values for KzGdzsbzseg
(superstructure) ........................................ 1 17
Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (Ueq) Values for K2Ce28b28e9
(superstructure) ........................................... l 18
Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (ch) Values for KZSmZSb28e9

. . . . 119

Table 4.10

Table 4.11

Table 4.12

Table 4.13

Table 4.14

Table 4.15

Table 4.16

Table 4.17

Table 5.1

Table 5.2

Table 5.3

Table 5.4

Table 5.5

Table 5.6

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (ch) Values for Kszzsbzseg
(superstructure) ............................................ 120

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (Ueq) Values for KZDyzstSeg
(superstructure) ............................................ 1 2 1

Selected Distances (A) and Bond Angles (°) for KzGdZSbZSeg
(superstructure) ............................................ 1 22

Selected Distances (A) and Bond Angles (°) for K2Ce28sze9
(superstructure) ............................................ l 23

Selected Distances (A) and Bond Angles (°) for K28m28b28e9
(superstructure) ............................................ 1 24

Selected Distances (A) and Bond Angles (°) for Ksz2$b28e9
(superstructure) ............................................ 125

Selected Distances (A) and Bond Angles (°) for KszZSbZSeg
(superstructure) ............................................ 126

Summary of Magnetic Properties of KanzstSeg
(Ln = Ce, Sm, Gd, Tb, Yb) ................................... 142

Summary of Crystallographic Data and Structural Analysis
for EquSe; and EquTe3 .................................... 158

Summary of Crystallographic Data and Structural Analysis
for Bao_5EUolsstC3 ......................................... 159

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (Ueq) Values for EquSe3 with
Estimated Standard Deviations in Parentheses .................... 160

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (ch) Values for EquTe3 with
Estimated Standard Deviations in Parentheses .................... 161

Fractional Atomic Coordinates and Equivalent Atomic

Displacement Parameter (Ueq) Values for Ba05Eu05SbTe3

(substructure) with Estimated Standard Deviations

in Parentheses ............................................. 162

Fractional Atomic Coordinates and Equivalent Atomic

xii

Table 5.7

Table 5.8

Table 5.9

Table 5.10

Table 6.1

Table 6.2

Table 6.3

Table 6.4

Table 6.5

Table 6.6

Table 6.7

Table 6.8

Table 6.9

Table 6.10

Table 6.11

Displacement Parameter (Ueq) Values for Ba0_5Eu0_SSbTe3
(supertructure) with Estimated Standard Deviations

in Parentheses ............................................. 163
Selected bond lengths (A) and angles (°) for EquSe3 .............. 165
Selected bond lengths (A) and angles (°) for EquTe3 .............. 166

Selected bond lengths (A) and angles (°) for Ba0.5Eu0.SSbTe3
(substructure) ............................................. 1 67

Selected bond lengths (A) and angles (°) for Ba0,5Eu0.5SbTe3
(superstructure) ............................................ l 68

Summary of Crystallographic Data and Structural Analysis
for Kl‘25Pb3jBI7258615 and Rb1'45Pb3JSb7‘458615 ................... 202

Summary of Crystallographic Data and Structural Analysis
for K1.45Pb3.]Sb7_458615 and K2.15Pb1.7Sb3.lssC]5 .................... 203

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (Ueq) Values for K1_25Pb3_50Bi7.258e,5
with Estimated Standard Deviations in Parentheses ................ 204

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameters for Rb1_45Pb3,IOSb7_4SSe15 with
Estimated Standard Deviations in Parentheses .................... 205

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameters for Kl_45Pb3‘IOSb7.4SSe15 with
Estimated Standard Deviations in Parentheses .................... 206

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameters for K2,15Pb.,7Sb3.158e15 with

Estimated Standard Deviations in Parentheses .................... 207
Selected Distances (A) for K1_25Pb3_sBi7_258e15 .................... 208
Selected Distances (A) for Rb1_45Pb3_ISb7_4SSe15 ................... 209
Selected Distances (A) for K1,45Pb3.le7,4SSe.5 .................... 210
Selected Distances (A) for K2.15Pb1,7Sb3.158e15 .................... 211

The Energy Band Gap, Melting Point, and Room
Temperature Charge Transport Properties ....................... 219

xiii

 

Table 7.1

Table 7.2

Table 7.3

Table 7.4

Table 7.5

Table 7.6

Table 7.7

Table 7.8

Table 8.1

Table 8.2

Table 8.3

Table 8.4

Table 9.1

Table 9.2

Table 9.3

Summary of Crystallographic Data and Structural
Analysis for SI'6SD6SI7, B32.62Pbl.3gsb4slo, and Ba3Sb4_66Sm ......... 240

Fractional Atomic Coordinates and Equivalent Isotropic
Displacement Parameter Values for Sr6Sb6817 with
Estimated Standard Deviations in Parentheses .................... 241

Fractional Atomic Coordinates and Equivalent Isotropic
Displacement Parameter Values for Ba3Sb4.66Slo with
Estimated Standard Deviations in Parentheses .................... 242

Fractional Atomic Coordinates and Equivalent Isotropic
Displacement Parameter Values for Ba2_62Pb1_338b4Sm

with Estimated Standard Deviations in Parentheses ................ 243
Selected Distances (A) and Bond Angles (°) for Sr6Sb68n ........... 245
Selected Distances (A) and Bond Angles (°) for Ba3Sb4,6(,Slo ....... . 247

Selected Distances (A) and Bond Angles (°) for Ba2_62Pb1,3ng4Slo. . . . 248

Crystallographic Data for Ba3Sb4_66Sm, Ba2_62Pb1_338b4S]0
and the Related Sulfosalt Minerals ............................. 259

Summary of Crystallographic Data and Structural Analysis
for KTh28e6 and RbThZSe6 ................................... 276

Fractional Atomic Coordinates and Equivalent Isotropic
Displacement Parameters (Ueq) Values for KThZSe6 and
RbTh2$e6 with Estimated Standard Deviations in Parentheses ........ 277

Anisotropic Displacement Parameters (A2) for KThZSeé and
RbThZSeé with Estimated Standard Deviations in Parentheses ........ 278

Selected Distances(A) and Bond angles(°) for ATthe6
(A = K, Rb) with Standard Deviations in Parentheses .............. 279

Summary of Crystallographic Data and Structural Analysis
for BaTh7Se13 ............................................. 309

Fractional Atomic Coordinates and Equivalent Atomic
Displacement Parameter (ch) Values for BaTlnSelg with
Estimated Standard Deviations in Parentheses .................... 310

Selected Distances(A) and Bond Angles(°) for BaTh7Se13 .......... 311

xiv

Table 9.4 Anisotropic Displacement Parameters for 8311178613 .............. 312

XV

Figure 2.1

Figure 2.2

Figure 2.3

Figure 2.4

Figure 2.5

Figure 2.6

Figure 2.7
Figure 2.8

Figure 2.9

Figure 2.10

Figure 2.11

Figure 2.12

Figure 2.13

LIST OF FIGURES

Page

The structure of RbUZSbsg (Cm) viewed down the a-axis .......... 51

Coordination environment of (a) the U atoms and the one-dimensional
columns along the a-axis, (b) the Sb, and (c) the Rb ............... 52

Schematic comparison of the ordering and disordering patterns
of Sb3+ and Rb+/K+ in (a) the substructure, (b) the superstructure

of RbUZSbsg, and (c) the superstructure of KUZSbSeg ............. 54
The structure of KUZSbSeg (Cm) viewed down the b-axis .......... 55
Optical absorption spectrum of RbUszSg (—) and

KUZSbSeg (-'-). The band gap value, Eg, is shown in the ﬁgure ...... 56
Inverse molar magnetic susceptibility, xM (per uranium),

of RbUszSg (U) and KUzsbSCg (I) .......................... 56
Raman spectra of (a) RbUZSng and (b) KUZSbSeg ............... 57
The structure of RbUSb0,33Te6 as viewed down the c4axis .......... 59

(a) The one-dimensional zigzag chain found in RbUSb033Te6

and (b) the helical chain found in elemental Te with bond

distances (A) and angles (°). Projections along the c-axis

for each case are shown in the box ............................ 61

(a) Coordination environment of the U(l) and U(2) atoms and

(b) the one-dimensional UTe6 column composed of tri-cappedl

trigonal prisms sharing triangular faces. Projection of the

column along the c-axis is shown in the box ..................... 63

The (a) arrangement of Sb and Rh ions along the [001]
direction, (b) coordination environment of the Sb3+ atom,
and (c) coordination environment of the Rb+ atom ................ 64

The superstructure of RbUSbo33Te6 as viewed down the [111] axis. . 66
The arrangement of Sb and Rb ions along the c-axis in the
superstructure (a) channel type I and (b) channel type II.

(c) The local environments of Rh when Sb is around (left)
and not (right) ............................................ 67

xvi

Figure 2.14

Figure 2.15

Figure 2.16

Figure 2.17

Figure 2.18

Figure 2.19

Figure 3.1

Figure 3.2

Figure 3.3

Figure 3.4

Figure 4.1

Figure 4.2

Figure 4.3

Figure 4.4

Possible Sb-Rb ordering patterns that may appear disordered
as in channel I and II ....................................... 70

(a) The Te-Te modulations in the Te chains in the superstructure.
(b) The dimer, trimer, and pentamer found in the chains ........... 71

Variable temperature thermopower for a single crystal of

RbUSb0_33Te6 ............................................. 74
Variable temperature electrical conductivity for a single

crystal Of RbUSb0_33TC6 ..................................... 74
Inverse molar magnetic susceptibility, xM (per uranium), of

RbUSbo33T66 ............................................. 75
Raman spectrum of RbUSb0_33Te6 ............................. 75

The structure of KTthzSC6 viewed down the a-axis.
The atomic labeling scheme is shown in the inset ................. 89

The double chains of Th atoms running [100] direction.
Each single chain is composed of ThSe6 prisms sharing
opposite triangular faces .................................... 91

The local environments of the Sb3+ ions ........................ 92

Optical absorption spectrum of KTthzSC6.
The band gap value, E,,, is shown in the ﬁgure ................... 93

The overall structure of KzGdZSbZSeg viewed down
the c-axis with labeling ...................................... 127

(a) Coordination environment of the Gd atom.

(b) Polyhedral representation of the Ody-centered bicapped

trigonal prisms. They share comers along the a-axis and

triangular faces along the c-axis to form layers.

(c) Comparison of the layers of KzGdZSbZSeg and ZrSe3 ........... 128

(a) Coordination environment of the Sb atoms. (b) Distortion
in the Sb(1)Se6 and Sb(2)Se6 octahedra. (c) Polyhedral
representation of the SbSe6 octahedral blocks along the c-axis ....... 130

(a) The superstructure of K2Gd28b23e9 viewed down the b’-axis
with labeling. The solid line represents the unit cell for the
superstructure and the dotted line represents the unit cell for
the substructure. (b) corresponding space group-diagrams

xvii

Figure 4.5

Figure 4.6

Figure 4.7

Figure 4.8
Figure 5.1

Figure 5.2

Figure 5.3

Figure 5.4

Figure 5.5

Figure 5.6

Figure 5.7

Figure 5.8

Figure 5.9

Figure 5.10

for Pbam and C 2/m view down the c- and b’-axis respectively ...... 132

Ball and stick representation of the edge-shared SbSe6 octahedra along
the b’-axis ( = c-axis) for (a) the substructure and (b) the
superstructure. Axial Se atoms are omitted for clarity. (c) The

local environment of Sb(la) and Sb(2a) ......................... 133
Optical absorption spectra of KanZSbZSeg

(Ln = Ce, Sm, Gd, Tb, Dy). The band gap value, E2, is shown ...... 137
Inverse molar magnetic susceptibility, l/x M,

(based on Gd metal) of (Ln = Ce, Sm, Gd, Tb, Dy) ............... 139
Raman spectrum of KzGdzsbzseg ............................. 143
The structure of EquSe3 viewed down the a-axis ................. 171

The local environment of the (a) Eu atom and (b) Sb atom.
(c) The Se net in the substructure of EquSe3 .................... 172

Selected area electron diffraction pattern of EquSe3 with

the beam perpendicular to the layers (parallel to the [010]

direction). Schematic of the electron diffraction

pattern corresponding to one unit cell (asub x csub) is shown below. . . . 174

The structure of EquTe3 viewed down the b-axis ................ 176

(a) Comparison of the [Sb2Q4]2' layers in EquTe3 and EquSe3.
(b) The square Te nets in EquTe3 ............................ 177

The structure of BaosEquSbTe3 viewed down the a-axis.
The bonds between Eu and Te are omitted for clarity .............. 179

Comparison of Te nets in the (a) substructure and the
(b) superstructure. Selected Te-Te distances (A) are
shown in the ﬁgure ......................................... 181

Distortion of the Te nets found in the superstructure of SrBiTe3 ..... 182
Optical absorption spectra of (a) EquSe3 and (b) EquTe3(—)

and Bao_5Euo,SSbTe3(---). The band gap value, E8, is shown

in the ﬁgure .............................................. 185

(a) Variable temperature electrical conductivity and
(b) thermopower for a single crystal of EquTe3 ................. 186

xviii

Figure 5.11

Figure 5.12

Figure 6.1

Figure 6.2

Figure 6.3

Figure 6.4

Figure 6.5

Figure 6.6

Figure 6.7

Figure 6.8

Figure 6.9

Figure 7.1

Figure 7.2

Figure 7.3

Inverse molar magnetic susceptibility, xM (per europium),

of EquSe3 ............................................... 187
The Raman spectra of (a) EquSe3 and (b) EquTe3 (——) and

BaoisEuO.sstC3(---) ........................................ 188
The structure 0f K1.25Pb3_5Bl7.258615 viewed down the b-axis ........ 213

The structure of (a) B-KzBIgselg” (b) K2_sBi3,5Se14, and

(c) K1_25Pb3.5Bi-,_258e15 viewed down the b-axis. In each

case, NaCl-, BizTe3- and CdIZ-type fragments found on the

framework are shown as highlighted by the shaded areas ........... 216

Optical absorption spectrum of KLzstgsBinsSels. The
semiconductor energy gap is indicated in the spectrum ............ 221

Variable temperature electrical conductivity and thermopower
for a single crystal of K1,25Pb3,5Bi7,258e15 ....................... 221

Variable temperature electrical conductivity and thermopower
for a single crystal of Rb1,45Pb3JSb7,4SSe15 ...................... 222

Variable temperature electrical conductivity and thermopower
for a single crystal of K1.45Pb3_58b7.458e15 ....................... 222

Variable temperature electrical conductivity and thermopower
for a single crystal Of K2,15Pb1,7Sbg.58e15 ....................... 223

Variable temperature electrical conductivity and thermopower
for an polycrystalline ingot of K1,25Pb3.5Bi7,25$e15 ................. 224

Variable temperature thermal conductivity, Kmeas, for an ingot

of K1_25Pb3,5Bi7,258e15. The inset shows total thermal conductivity,
Km], corrected for radiative loss .............................. 225

The structure of SI'6Sb6SI7 viewed down the b-axis. The inset
shows an asymmetric unit with the atomic labeling scheme ......... 250

(a) Propagation of a single slab along the a-axis The dashed
lines represent weak Sb-S interactions between (Sb3S7)5' units.
(b) Arrangement of the tri-sulﬁde groups ....................... 251

(a) The structure of Ba3Sb4_668m viewed down the a-axis.
(b) Propagation of a single slab along the a-axis .................. 253

xix

 

 

 

Figure 7.4

Figure 7.5

Figure 7.6

Figure 7.7

Figure 7.8

Figure 7.9

Figure 7.10

Figure 8.1

Figure 8.2

Figure 8.3

Figure 8.4

Figure 8.5

Figure 8.6

The local environment of the Sb(5/5’), Ba(1), and Ba(2) atoms in
333813466810 .............................................. 255

The structure 0f B3162Pb1338b4810 VICWCd down the a-axis. The
inset shows an asymmetric unit with the atomic labeling scheme ..... 257

Optical absorption spectrum of Sr6Sb6S17. The band gap
value, Eg, is shown in the ﬁgure ............................... 261

Optical absorption spectra of Ba3Sb4_6(,Sw(—) and
Bag-62Pb1.3ng4Sm('--). The band gap value, Eg, is shown

in the ﬁgure .............................................. 261
Raman spectrum of Sr6Sb6$17 ................................ 262
Raman spectra of Ba3Sb4,66810(—) and Ba2,62Pb1_3ng4Sm(~) ........ 262

DTA diagram Of B32_62Pb1.388b4slo. (—) First CYCIC.
(---) Second cycle .......................................... 263

The overall structure of KTthC6 viewed down the c-axis.
The labeling scheme for the RbThZSe6 is analogous ............... 281

Schematic comparison of the structures of (a) ZrSe3,

(b) KTh2866 and (c) KThzTeG. In KThzTC6, each K+ site is

half occupied. The distribution of chalcogen-chalcogen

distances are indicated ..................................... 282

Electronic absorption spectra of KThZSe6(---) and RbThZSe6(——).
The band gap energies, E2 are indicated ........................ 285

Raman spectra of KThZSe6(---) and RbThZSe6(—) ................ 285

Selected area electron diffraction pattern with the beam parallel

to the [001] zone axis from RbThZSeé showing weak 4 x 4

superlattice. The (hk0) family of reﬂections is shown.

Superlattice peaks around the 310 sublattice reﬂection are

indicated by four small arrows. Also notice that the superlattice

peaks show some streaking along a'-axis due to diffuse scattering. . . 288

AF M image of the surface of a layer of RbThZSes corresponding
to the ab-plane. (A) shows a raw image which has been ﬂattened
with a second order polynomial to account for non-linearity in the
piezoelectric scanner. (B) shows a blown up region of the image
after a spatial Fourier ﬁltration to remove instrumental and

XX

 

 

Figure 8.7

Figure 8.8

Figure 9.1

Figure 9.2

Figure 9.3

Figure 9.4

Figure 9.5

Figure 9.6

environmental noise from the data. The rows run parallel to

the b-axis. The scale bar is 20 A ............................

X-ray powder diffraction pattern from KTthC6. The data are
shown in the form of Q(S(Q)-1) which is the structure function

that is Fourier transformed to obtain the PDF .................

Pair distribution functions in the form of p(r) from KTthC6.
(a) shows the PDF obtained from the data. (b) shows the PDF
calculated from the crystallographic model of the structure.

In both cases an arrow has been placed at a distance of 2.34 A
which is the length expected for a diselenide bond. The lack of
intensity at this position in p(r) for the model reﬂects the fact
that in the average structure this distance does not exist;
however it is clear from the data that this diselenide bond

does exist in the material ..................................

The overall structure of BaTh7Selg view down the c-axis .........

Coordination environment of (a) Th(1), (b) Th(2), (c) Th(3),

and (d) Ba ..............................................

Connectivity of (a) Th(1), (b) Th (2), (c) Th (3), and
(d) Ba along the c-axis. (e) Arrangement of Th(3) and Ba
along the a-axis. The Se-Se distances(A) associated with

the Se chains are shown in the ﬁgure .........................

The [Th48618114- framework view down the c-axis ..............

Optical absorption spectrum of BaTh7Selg. The band gap

value, E“, is shown .......................................

Raman spectrum of BaTh7Se.g .............................

xxi

.. 292

...296

297

..314

..315

..316

.. 318

..321

.. 321

 

ADP
AFM
CCD
DMF
DTA
EDS
HOMO
IR
LUMO
NLO
PDF
SEM
SQUID
TE
TEM
UV/V is
VSEPR

LIST OF ABBREVIATIONS

Atomic Displacement Parameter
Atomic Force Microscopy

Charge Coupled Device
Dimethylforrnamide

Differential Thermal Analysis

Energy Dispersive Spectroscopy
Highest Occupied Molecular Orbital
Infrared Spectroscopy

Lowest Unoccupied Molecular Orbital
Non Linear Optics

Pair Distribution Function

Scanning Electron Microscopy
Superconducting Quantum Interference Device
Thermoelectric

Transmission Electron Microscopy
Ultraviolet/Visible Spectroscopy
Valence Shell Electron Pair Repulsion

xxii

 

 

 

Chapter 1. The Rationale for Exploratory Synthesis of Multinary
Antimony Chalcogenide Compounds Using Molten Alkali Metal

Polychalcogenide Fluxes.

 

I. Introduction

Among many interesting solid state materials one very important class of
compounds is that of Group 15 sulfosalts minerals, which is deﬁned as complex sulﬁdes
with the general formula type of M’xMySz, (M’ = Fe, Mn, Cu, Ag, Zn, Hg, Sn, Pb. etc.;
M= As, Sb, Bi).l Although M’-Q interactions are often covalent, as in
Ag58b84(stephanite)2 or Cu14Sb4S13(tetrahedrite)3, M’ can present alkali or alkaline earth
metals for which ionic bonding is predominant. This class of compounds encompasses a
vast number of naturally occurring materials, which exhibit an asto'undingly rich
structural and compositional diversity. Such solid state compounds are also intriguing for
their potential physical properties (i.e. charge density waves, semiconductivity,
superconductivity, and magnetism) and resulting technological applications. The
plethora of structural and compositional types in the mineral sulfosalts derives from the
ability of the Group 15 elements (As, Sb, Bi) lone ns2 pair electrons to stereochemically
express themselves in various ways. The tendency of this lone pair to localize in a
certain direction diminishes in moving from As to Bi.4 For example, arsenic, the element
directly above antimony in the periodic table, always has its lone pair electrons expressed
and usually adopts trigonal pyramids as its local environment. By contrast, the lone pair
of bismuth, which is directly beneath antimony, is most often stereochemically inactive
and does not have as strong of a tendency to localize in a speciﬁc direction.
Consequently, Bi adopts more symmetric and less distorted local environments such as
an octahedron with small differences in the length of the six Bi-S bonds. The Sb atom
possesses an “in-between” nature and has the ability to express its lone pair electrons in

various degrees. In other words, the lone pair of Sb3+ can express itself in many different

 

ways creating various asymmetric local environments around the Sb3+ ion such as a
trigonal pyramid (C.N.=3), a seesaw (C.N.=4), a square pyramid (C.N.=5), and a

distorted octahedron (C.N.=6), see Scheme 1.1.

 

O, 9 I90

Trigonal pyramid See-saw Square pyramid Distorted octahedron

 

 

 

Scheme 1.1

From a structural chemistry perspective, this feature makes Sb compounds even
more enchanting among Group 15 Chalcogenides because the various interesting
distortions around Sb3+ ions can lead to an enormous number of unique structure types.
Since each of the local coordination environments of Sb is highly asymmetric, the
chances of stabilizing non-centrosymmetric structures with antimony Chalcogenide
building blocks is relatively high. This also encourages the investigation of Sb
compounds because noncentrosymmetric structures are of considerable interest for
potential applications as piezoelectrics, ferroelectrics and nonlinear optics.5

In view of the vast number of known natural sulfosalts and the technological
potential of this class of materials, we were interested in the investigation of the
. corresponding seleno- and telluroantimonate compounds. We would like to probe the
possibility of new structure types, characterize their physico-chemical properties, and
obtain a systematic understanding of this class of compounds. This will also provide us

with an opportunity to look for new solid state materials for possible practical

applications. Except for the characterization of the naturally occuring sulfosalt minerals,
very little investigation has been performed in this area. Most of the synthetic
compounds are ternary phases with alkali or alkaline earth metals (e. g. KSsz 6 KgsbSe4,7
RbSb3Se5,8 CS3Sb5Q9(Q=S,Se),9 c523b48e8,‘° CaZszss,“ Bang68n, 12and snsmsg").
Several ternary compounds with transition metals (e. g. Cu3SbSe414) or main group
metals (e.g. PbSbZSe4,15 Tl4Bi285'6) are also known. However, the number of known
synthetic quaternary compounds is still surprisingly small (i.e. NaZCquS3,l7
AAg28bS4(A=K,Rb),'8 AzAngS4(A=K,Rb),'9 CS3Agzsb383,19 AzAngS, (A = K, Rb,
Cs),20 szAung4Slo,21 AzAquS4(A = Rb, Cs),22 AzAg20Sb4S,9(A=Rb, Cs),23 KHngS3“)

and it dramatically decreases moving from sulﬁdes to telluride compounds (Kanz-

bebe48e12(Ln = La, Ce, Pr, Gd)25 RngSbTe3,26). All these factors combined together

motivated us to further investigate ternary and quaternary chalcoantimonate system. For
this purpose, we have applied alkali metal polychalcogenide ﬂuxes as a synthetic method
to prepare many interesting compounds with various compositions and structures. This
method is well suited for the exploration of new chemistry at intermediate temperatures
(ZOO-600°C) and has been proven to be a powerful synthetic tool for various classes of
Chalcogenide compounds.27 In next chapter, we will discuss the nature and the

advantages of the alkali metal polychalcogenide method in detail.

2. Polychalcogenide Flux Method.

In conventional solid state synthetic methods, elements or simple binary

compounds are stoichiometrically combined based on the formula of the target

compounds and these mixtures are heated at very high temperatures. However, there
exist several limitations in this so-called “heat and beat” method. For example,
homogeneity of the reaction is hard to achieve because the reactions occur via diffusion
between solid particles. Grinding between successive heating periods can be the only
way to obtain homogeneous products. Such a method requires high reaction
temperatures to provide the system with effective diffusion in order to complete the solid
state reactions. Due to this necessity, the products obtained by this method are usually
the most thermodynamically stable ones. Therefore, this classical method is not very
suitable for compounds that decompose at high temperature or whose formation is
kinetically controlled. Finally, it is very hard to obtain well-grown crystals, which makes
structural determination of new phases rather difﬁcult, if not impossible.

‘ The importance of exploratory synthesis in solid state chemistry cannot be
overemphasized considering the dependence of modern technology on the discovery of
novel materials with new or enhanced properties (i.e. thermoelectrics,28 catalysis,29 high
Tc superconductors,” nonlinear optics, electroluminescence,3 ' photovoltaics,32 high
density storage batteries”). Therefore, new solid state synthetic methods that could
decrease the reactions temperatures have been searched and utilized to discover new
materials with more complex compositions and structures over the past decades.
Chemical vapor deposition (CVD),34 hydrothermal and solvotherrnal synthesis,35 eutectic
combination of binary salts,36 and molten ﬂuxesz’7 are good examples.

Molten salts have been used for over 100 year for the purpose of a high
temperature recrystallization for various binary and ternary compounds.38 It was not until

1987 when this method was used at lower temperature to synthesize new compounds.”40

 

For the discovery of new Chalcogenide materials, the use of molten alkali metal
polychalcogenides of the type AzQx (A = alkali metal, Q=S, Se, Te) as solvents is very
appropriate and produces numerous novel and interesting compounds. In many cases,
these salts act not only as solvents but also as reactants, providing species and building
blocks that can be incorporated into to the ﬁnal products. As solvents for intermediate
temperature reactions, AzQx salts are especially well suited because the melting points
range between 200 and 600°C. The AzQx ﬂuxes remain nonvolitile over a wide
temperature range and can be applied for various reaction temperature above their
melting point without conceming solvent loss. The melting points of several alkali metal
polychalcogenide salts are given in Table 1.1.

Various lengths of polychalcogenides chains can be formed by the reaction of
AzQ and Q in situ. Once they are formed, they undergo complex self-redox equilibria of

the type shown in Scheme 1.2.

 

 

 

 

I7 + Z 2

O _ _

(oar (out (03)”
Scheme 1.2

 

Table 1.1. Melting points for some known alkali metal/polychalcogenide species (A =

 

 

 

 

 

 

 

 

 

 

 

 

alkali metal).
AzQ AzQz AZQJ A2Q4 AzQs AzQe
Li28 Li282
900-975 °C 369 °C
Na28 Na2S2 Na283 Na284 Na285
1180 °C 490 °C 228 °C 275 °C 252 °C
K28 K282 K283 K284 K285 K286
840 °C 470 °C 252 °C 145 °C 206 °C 189°C
Rb28 Rb2S2 Rb283 Rb284 Rb285 Rb286
530 °C 420 °C 213 °C 160 °C 225 °C 201 °C
C5282 C5283 C3284 C5285 C5286
460°C 217°C 160°C 210°C 186°C
Na2Se Na2Se2 Na2Se3 Na28e4 Na2Se6
>875 °C 495 °C 313 °C 290 °C 258 °C
K2Se2 K28e3 K2Se4 K2Se5
460 °C 380 °C 205 °C 190 °C
Na2Te Na2Te2 Na2Te6
953 °C 348 °C 436 °C
K2Te K2Te; K2Te4 K2Te5 K2Te6
900 °C 329 °C 266 °C 268 °C 264 °C
Rb2Te Rb2Te;
775 °C 400 °C
CS2Te CS2Te3 CS2Te4 CS2Te5 CS2Te5
820 °C 395 °C 221-237 °C 235 °C 226 °C

 

 

 

 

 

 

 

 

 

The average length of the chain can be varied by changing the A2Q/Q ratio
because they directly affect the ratio of terminal chalcogen which bears a 1- charge and
internal chalcogen which possess no charges. If more Q (or less A2Q) is added to the
reaction mixture, longer polychalcogenide chains will form because it increases the
Qo/Ql' ratio. Conversely, if more A2Q(or less Q) is added to the reaction mixture, shorter

polychalcogenide chains will form, see Scheme 1.3.

 

 

 

 

 

 

 

2K2$e + 4Se Kzse + SSe

i i
4@ AV 2@ +W-

More basic More acidic

 

 

 

 

 

 

Scheme 1.3

The length of the chain plays a critical role to determine the basicity of the ﬂux
because the terminal and internal chalcogen atoms act differently in the oxidation-
reduction reactions with reactive metals. When the metal atoms are added in the ﬂux,
they are drawn into solution by being oxidized by the internal atoms of the (Qx)2' chains.
In doing so, the internal chalcogens themselves are reduced and become terminal
chalcogens by splitting the chain into small fragments. The solvated metal cations are
then coordinated by the terminal chalcogen atoms, which bear a 1- charge and can act as
Lewis bases. Therefore, the ﬂux with longer chains with a higher internal/terminal
chalcogen ratio is more oxidizing than that with shorter chains. Since the basicity of the
ﬂux can be easily controlled by varying the A2Q/Q ratio, it is possible to explore a
chemical system under a wide range of conditions. This is important because some
compounds will form only under certain conditions and this is why the alkali metal
polychalogenide ﬂux method is one of the most powerful synthetic tools for the
exploratory synthesis.

The polychacogenide method is quite suitable to obtain well grown crystals.
When the compounds are formed in the ﬂux, the nucleated species are in equilibrium
with the soluble intermediates, especially if the ﬂux is present in excess, and hence a
solvation/reprecipitation effect occurs. This aids in the growth of single crystals because
the ﬂux can re-dissolve small or poorly formed crystallites and then reprecipitate the
smaller species onto larger, well-formed crystals. Another advantage of the ﬂux methods
is that they allow the reaction system to end up “where it wants” in a kinetic or
thermodynamic sense without forcing it to a certain stoichiometry or structure type.

Provided the temperature and time are appropriate, the reaction systems have all the

 

ingredients and freedom to form a new phase. The beneﬁt of this becomes apparent from
the unusual compositions often found in the new materials that most certainly could not

have been predicted in advance.

3. Synthetic Strategy for Multinary Chalcoantimonates

Based on the discussion above, we have decided to explore the synthesis of
ternary and quaternary chalcoantimonate compounds by the alkali metal
polychalcogenide ﬂuxes. These ﬂuxes are formed by the in situ fusion of A2Q/Sb/Q or
A2Q/Sb2Q3/Q (A = K, Rb ; Q = 8, Se, Te) to produce various Ax[Sbsz] units. These
discrete species can vary in composition and structure and can form the basis for
generating a large number of ternary and quaternary compounds.

To prepare multinary compounds with more complex compositions and structure-
types, we speciﬁcally choose f-elements to react with Sb in the polychalcogenide ﬂuxes.
Our strategy is to incorporate two metals with very different coordination preferences to
induce interesting distortions. Due to the large size of the f-elements, the bonding
between the metals and the chalcogen atoms is primarily ionic in character and the
coordination sphere is large and ﬂexible.41 Several most commonly observed local
environments of actinide or lanthanide metals are shown in Scheme 1.4. This property,
when combined with the property of Sb3+ ions to have stereochemically active lone pairs,
may increase the chances of producing unique structure types in these compounds. Also
by choosing actinide and lanthanide metals as another component in the multinary

compounds, the known structure type of mineral sulfosalts, which are mostly formed with

10

 

transition metals (i.e.Ag, Cu, Zn, Hg) or main group metals (Sn, Pb, Tl) can be

effectively avoided.

 

   

 

 

 

/ \
I \
Bi-capped Tri-capped Mono-capped
trigonal prism trigonal prism square antiprism
Scheme 1.4

We have designed reactions of Sb with various f-elements, which prefer to bear
different oxidation states in a Chalcogenide matrix. For example, U, Gd, and Eu metal
are most stable with +4, +3, and +2 oxidation state, respectively.42 We expected that
changing metal types, which will be oxidized into different valence states in the
polychalcogenide ﬂuxes, will result in different structure types of the ﬁnal products.

Since we were also interested in the changes of the structure-types and properties
of the products depending on the chalcogen types in the ﬂuxes, we investigated the
reactivity of these metals with Sb in each of the polysulﬁde-, polyselenide-, and
polytelluride ﬂuxes. Each chalcogen possesses a different electronegativity and different
tendency to engage in chalcogen-chalcogen interactions. Therefore, changing chalcogen
types in the ﬂuxes often result in the formation of different structure types. The resulting

compounds possess interesting physico-chemical properties due to the presence of f-

11

electrons. Therefore, characterization of these materials can provide us with a better
understanding of the relatively less studied f orbitals.

Another element that we were interested to incorporate into the antimony
Chalcogenide framework was lead. Unlike f-elements lead forms primarily covalent
bonds with chalcogens and the resulting structure types are expected to be somewhat
different from those with actinide or lanthanides. In addition to the structural chemistry,
we were speciﬁcally interested in lead chalcoantimonate compounds because of the
potential thermoelectric properties that these compounds may possess, as will be

discussed in the next chapter.

4. Thermoelectric Relevance of Sb/Bi Chalcogenides

Ever since the solid solutions of Bi2.,,Sb,‘Te3.ySey43 were established as the leading
materials available for thermoelectric applications near room temperature, there has been
continuing efforts to ﬁnd better thermoelectric materials. The challenge lies in achieving
simultaneously high electrical conductivity, high thermoelectric power and low thermal

conductivity. These properties deﬁne the thermoelectric ﬁgure of merit, zT = (SZG/K)T;
where S is the thermopower, 0', the electrical conductivity, K, the thermal conductivity

Owhich is composed of IQ, the phonon thermal conductivity, and K3, the carrier thermal

conductivity. All three of these properties are determined by the details of the electronic
structure and scattering of charge carriers (electrons or holes) and thus are not

independently controllable parameters.

12

 

To date, most investigations were mainly focused on tuning44 the composition of
Bi2Q3 solid solutions, doping45 M2Q3 with other heavy metals and optimizing device
design. However, the synthesis of new structurally or compositionally complex materials
can be another approach to produce better thennoelectricity.

One of the reasonable approaches to ﬁnd better thermoeletric materials would be
to design and prepare compounds that may possess lower thermal conductivity than that
of Bi2.,‘8b,,Te3.ySey because several guide lines are relatively well estabilished to help us
to do so.”47 The phonon contribution to the thermal conductivity can be lowered by
structural complexity, by choosing heavy elements as constituents of the material and by
choosing combinations of elements that normally make moderate to weak chemical
bonds. In these materials a weakly bound atom called a "rattler" is used to lower the
thermal conductivity of the solid without severely affecting the eletronic conduction.

The multinary Bi/ Sb Chalcogenide compounds prepared by alkali metal
polychalcogenide ﬂux method tend to satisfy this description.48 These materials are
composed of heavy elements such as Bi, Sb, Te, and Se and they usually possess large
unit cell with very complicated structures and compositions. Alkali metal ions from the
ﬂux usually incorporated into the structure of the ﬁnal products and stabilized in cavities,
tunnels or galleries of the anionic framework formed by covalent Sb/Bi-chalcogen bonds.
These alkali metal ions are very weakly bound to the framework and always possess the

49 Therefore, they can act as 3

highest thermal displacement parameters in the structure.
rattler and can signiﬁcantly suppress the thermal conductivities of these materials.

Structural and compositional complexity of such materials are also favorable for

high thermopower because they can result in corresponding complexities in the electronic

l3

structure. Such complexities may produce the required asymmetry in 0(E) to obtain large

thermopowers according to the Mott formula below,50 and at the same time possess low

thermal conductivities. The thermopower S is given by :

1: kZT dlno(E)
3 e dE E=Ef

 

S:

where 0(E) is the electrical conductivity determined as a function of band ﬁlling.
The electrical conductivity o= o(E) | 13:13,. where 13f is the Fermi energy. If the carrier
scattering is independent of energy, then 0(E) is proportional to the density of states at E.
In the general case, S is a measure of the difference in 0(E) above and below the Fermi

surface, speciﬁcally through the logarithmic derivative of with respect to E, see the

equation above. So by manipulating the energy dependence of 0(E) one can control
simultaneously 6 and 8. Our previous investigation on the A/M/Q system (A = alkali
metal; M = Sb, Bi; Q = 8, 8e, Te) has produced several ternary compounds with
interesting thermoelectric properties (i.e. BaBiTe3,“ CSBI4Te6,52 K2Bi3813,23 and
KBir5,33810,53 B-K2BigSe13,54 K2Snge13,54 K25Big58e14,54 K25Sb858e1454). These results
proved that compounds with complex composition and structures indeed can decrease the

thermal conductivity without signiﬁcantly decreasing in the electrical conductivity.

Another concept that can help us ﬁnd new materials with low thermal
conductivity is termed "mass ﬂuctuation" scattering of the lattice phonons. 55 This can be

achieved by introducing site occupancy disorder to the system. Such disorder will

generates randomness of the mass, size, and charge of the atoms on a particular lattice

l4

 

position that can strongly scatter lattice phonons carrying heat. 56 To further explore the

effects of structural complexity and mass ﬂuctuation in compounds with large unit cells,
we decided to incorporate Pb metal to the A/Sb/Se (A=alkali metal) system because in
addition to being very heavy, Pb has a well known tendency to disorder with Sb or alkali
metals depending on the local coordination environment.57’58'59 For this purpose, we
extended our investigations to the corresponding Bi system because bismuth
Chalcogenides usually possess smaller band gaps and, therefore, higher electrical
conductivities, which makes these compounds more interesting than the corresponding
Sb compounds from a thermoelectric point of view. Indeed, the reaction of Pb metal with
Bi/Sb in the polyselenide ﬂuxes produced a very interesting family of compounds, which
possess considerable disorder between alkali metal, Pb, and Bi/Sb, and, therefore, very
low thermal conductivities. The related work will be presented and discussed in detail in

Chapter 6.

15

10.

11.

12.

13.

References

Gmelin Handbook of Inorganic and Organomettalic Chemistry, 8th Edition:
TYPIX-Standardized Date and Crystal Chemiccal Characterization of Inorganic
Structure Types, Vol.1, Springer-Verlang Berlin, 1993.

(a) Ribar, B.; Nowacky, W. Acta Crystallogr. 1970, B26, 201. (b) Ribar, B.;
Nowacky, W. Schweizerische Mineralogische und Petrologische Mitteilungen
1969, 49, 379.

(a) Bryndzia, L. T.; Davis, A. M.; Am. Mineral. 1989, 74, 236. (b)Tatsuka, K.;
Morimoto, N. Am Mineral. 1973, 58, 425.

Sheldrick, W. 8.; Wachhold M. Coord. Chem. Rev. 1998, 176, 211-322.

(a) West, A. R. Solid State Chemistry and Its Applications, 1984, John Wiley &
Sons; New York, pp 540—552. (b) Xia, Y.; Chen, C.; Tang, D.; Wu, B. Adv.
Mater. 1995, 7, No.1, 79-81.

Graf, H. A.; Schafer, H. Z. Anorg. Allg. Chem. 1975, 414, 211.

Eisenmann, B.; Zagler, R. Z. Naturforsch. 1989, 44b, 249-256.

Sheldrick, W. 8.; Hausler, H. J. Z anorg. allg. Chem. 1988, 557, 98-104.
Sheldrick, W. S.; Hausler, H. J., Z anorg. allg. Chem. 1988, 561, 149-156.
Sheldrick, W. 8.; Kaub, J. Z anorg. allg. Chem. 1986, 536, 114-118.

Cordier, G.; Schﬁfer, H. Rev. Chim. Miner., 1981, 18, 218-223.

Dorrscheidt, W.; Schafer, H. Z. Naturforsch, 1981, 368, 410-414.

Cordier, G.; Schwidetzky, C.; Schafer, H. Rev. Chim. Miner. , 1982, 19, 179-186.

16

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

Aurivillious, B. Acta Chem. Scand. 1983, A37, 399.

Pﬁtzner, A. Z Kristallogr. 1994, 209, 685.

Skowron, A.; Boswell, F. W.; Corbett, J. M.; Taylor, N. J. J. Solid State Chem.
1994, 112, 251.

Jerome, J. E. ; Wood, P. T.; Pennington, W. T.; Kolis, J. W. Inog. Chem. 1994,
33, 1733.

Schimek, G. L.; Pennington, W. T.; Wood, P. T.; Kolis, J. W. J. Solid State Chem.
1996, 123, 277.

Wood, P. T.; Schimek, G. L.; Kolis, J. W. Chem. Mater. 1996, 8, 721.

Hanko, J. A. Ph. D. dissertation, Michigan State University, 1998.

Hanko, J. A.; Kanatzidis, M. G. J. Chem. Soc. Chem. Commun. 1998, 725.
Hanko, J. A.; Kanatzidis, M. G. J. Alloys Compd. 1998, 280, 71.

Hanko, J. A.; Kanatzidis, M. G. manuscript in preparation.

Imafuku, M.; Nakai, 1.; Nagashima, K. Mat. Res. Bull. 986, 211, 493.

Chen, J. H.; Dorhout, P. K. J. Alloys Compd. 1997, 249, 199.

Li, J .; Chen, Z.;Wang, X.; Proserpio, D. M. J. Alloys Compd. 1997, 262-263, 28.
(a) Kanatzidis, M. G.; Sutorik, A. C. Prog. Inorg. Chem. 1995, 43, 151. (b)
Kanatzidis, M. G. Chem. Mater. 1990, 2, 353.

(a) Rowe, D. M.; Bhandari, C. M. Modern Thermoelectrics, Holt, Rinehart, and
Winston, London, 1983, p.103. (b) Borkowski, K.; Pyzluski, J. Mater. Res. Bull.

1987, 22, 381.

17

 

29.

30.

31.

32.

33.

34.

35.

36.

37.

(a) Chianelli, R. R.; Pecoraro, T. A.; Halbert, T. R.;Pan, W. —H.; Stiefel, E. l. J.
Catal. 1984, 86, 226. (b) Pecoraro, T. A.; Chianelli, R. R. J. Catal. 1981, 67, 430.
(c) Harris, 8.; Chianelli, R. R. J. Catal. 1984, 86, 400.

(a) Malik, S. K.; Vijayaraghavan, R. Rev. Solid State Sci. 1988, 2, 75. (b) Chevrel,
R.; Sergent, M.; Progent, J. J Solid State Chem. 1971, 3, 515. (c) Hebard, A. F .;
Rosseinsky, M. J.; Haddon, R. G; Murphy, D. W.; Glarum, S. H.; Palstra, T. T.
M.; Ramirez, A. P. Nature 1991, 350, 600.

Ballman, A. A.; Byer, R. L., Eimerel, D.; Feigelson, R. 8.; Feldman, B. J .;
Goldberg, L. 8.; Menyuk, N.; Tang, C. L. Applied Optics 1987, 26, 224.

(a) Etman, M.; Katty, A.; Levy-Clement, C.; Lemasson, P. Mater. Res. Bull 1982,

17, 579. (b) Katty, A. Soled, 8.; Wold, A. Mater. Res. Bull 1977, 12, 663.

(a) The Sodium-Sulfur Battery, Sudwirth, J. L.; Tilley, A. R. Eds, Chapman and
Hall, New York, 1985. (b) Fischer, W. Mater. Res. Soc Symp. Proc. 1989, 135,
541. (0) Powers, R. W.; Karas, B. R. J. Electrochem. Soc. 1989, 136, 2787.
Steigerwald, M. in Inorganometallic Chemistry, Fehlner,T. P. Ed. Plenum Press,
New York, 1992, 333.

Rabenau, A. Angew. Chem Int. Ed. Engl. 1985, 24, 1026.

(a) Chen, L. R. Commun. Theor. Phys. 1992, 17(3), 285. (b) Chen L. R. Commun.
Theor. Phys. 1992, 13(2), 147.

(a) Elwell, D.; Scheel, H. J. Crystal Growth from High-Temperature Solution;
Academic Press: London, 1975. (b) Molten Salts Handbook; Janz, G. J ., Eds.;
Academic Press: New York, 1967. (c) Marnantov, G. Molten Salts; Marcel

Dekker: New York, 1969.

18

38.

39.

40.

41.

42.

43.

44

45.

46.

(a) Scheel, H. J. J. Cryst. Growth 1974, 24/25, 669. (b) Sanjines, R.; Berger, H.;
Levy, F. J. Mater. Res. Bull. 1988, 23, 549. (c) Garner, R. W.; White, W. B. J.
Cryst. Growth. 1971, 7, 343.

Sunchine, S. A.; Kang, D.; Ibers, J. A. J. Am. Chem. Soc. 1987, 109, 6202.
Drake, G. W.; Kolis, J. W. Coord. Chem. Rev. 1994, 137, 131.

Freeman, A. J .; Darby Jr., J. B. eds, “The Actinides: Electronic Structure and
Related Properties”, Vol I, II, 1974, Academic Press; New York .

(a) Greenwood, N. N.; Eamshow, E. Chemistry of the elements; Pergamon Press:
New York, 1984; 1434. (b) Guittard, M.; Flahaut, J. Synthesis of Lanthanide and
Actinide Compounds, Kluwer Academic Publisherrs, Netherlands, 1991, 321.

(a) Jeon, H. -H.; Ha, H. -P.; Hyun, D. -B.; Shim, J. -D. J. Phys. Chem. Solids
1991, 4, 579-585 (b) Testardi, L. R.; Bierly, J. N. Jr.; Donahoe, F. J . J. Phys.
Chem. Solids 1962, 23, 1209 (c) Champness, C. H.; Chiang, P. T.; Parekh, P.
Can. J. Phys. 1965, 43, 653-659 and (d) 1967, 45, 3611-3626

(a) Yim, W. M.; Fitzke, E. V. J. Electrochem. Soc. 1968, 115, 556-560 (b) Yim,
W. M.; Fitzke, E. V.; Rosi, F. D. J. Mater. Sci. 1966, 1, 52-65 (c) Borkowski, K.;
Przyluski, J. Mater. Res. Bull. 1987, 22, 381-387

(a) Chizhevskaya, S. N.; Shelimova, L. E. Inorg. Mater. 1995, 31, 1083-1095 (b)
Horék, J.; Cermak, K.; Koudelka, L. J. Phys. Chem. Solids 1986, 47, 805-809 (c)
Lostak, P.; Horak, J .; Koudelka, L. Phys. Status Solidi 1983, 76, k71-k75 (d)
Zalar, S. M. Adv. Energy Conv. 1962, 2, 105-112

Sales, B. C. Curr. Opin. Solid St. Mater. 1997, 2, 284-289.

19

47.

48.

49.

50.

51.

52.

53.

54.

55.

(a) Nolas G. 8.; Cohn, J. L.; Slack, G. A.; Schujman, S. B. Appl. Phys. Lett. 1998,
73, 178-180. (b) Nolas G. 8.; Slack, G. A.; Morelli, D. T.; Tritt, T. M.; Ehrlich, A.
C. J. Appl. Phys. 1996, 79, 4002-4008. (c) Nolas G. 8.; Weakley, T. J. R.; Cohn,
J. L. Chem. Mater. 1999, 11, 2470-2473.

Chung, D.-Y.; Iordanidis, L.; Choi, K.-S.; Kanatzidis, M. G. Bull. Korean Chem.
Soc, 1998, 19, 1283-1293.

Sales, B. C.; Chakournakos, B. C.; Mandrus, D.; Sharp, J. W. J. Solid State Chem.
1999, 146, 528-532.

Mott, N. R; Jones, H. “The Theory of the Properties of Metals and Alloys”,
Dover Publications, New York, NY.

Chung, D.-Y.; Jobic, 8.; Hogan, T.; Kannewurf, C. R.; Bree, R.; Rouxel, J .;

Kanatzidis, M. G. J. Am. Chem. Soc. 1997, 119, 2505-2515.

Chung, D.-Y.; Hogan, T.; Brazis, P. W.; Rocchi-Lane, M.; Kannewurf, C. R.;
Bastea, M.; Uher, C.; Kanatzidis, M. G. Science 2000, 287, 1024-1027.

(a) Kanatzidis, M. G.; McCarthy, T. J .; Tanzer, T. A.; Chen, L. -H.; Iordanidis, L.;
Hogan, T.; Kannewurf, C. R.; Uher, C.; Chen, B. Chem. Mater. 1996, 8, 1465-
1474.

Chung, D.-Y.; Choi, K-.S; Iordanidis, L.; Schindler, J. L.; Brazis, P.; Kannewurf,
C. R.; Chen, 8.; Hu, 8.; Uher C.; Kanatzidis, M. G. Chem. Mater. 1997, 9, 3060-

3071.

Euken, A.; Kuhn, G., Z anorg. alleg. Chem. 1928, 134, 193.

20

56.

57.

58.

59.

(a) Slack, G. A.,"New Materials and Performance Limits for Thermoelectric
Cooling" in Handbook of Thermoelectrics, ed. by Rowe, D. M. CRC Press, Boca
Raton, 1995. (b) Slack, G. A.; Hussaine, M. A. J. App]. Phys. 1991, 70, 2694.
Takagi, J.; Takéuchi, Y. Acta Cryst. 1972, B28, 649-651.

Take’uchi, Y.; Takagi, J. Proc. Japan Acad. 1974, 50, 76-79.

Skowron, A; Boswell, F. W.; Corbett, J. M.; Taylor, N. J. J. Solid State Chem.

1994, 112, 307-311.

21

Chapter 2. Reactions of Uranium Metal with Sb in Polychalcogenide

Fluxes.

Discovery of Novel Quaternary Noncentrosymmetric Uranium

Chalcoantimonates, AU 2Sng(A=Rb, K; Q=S, Se) and RbUSb0,33Te6

22

1. Introduction

In a chalcogenide matrix, uranium atoms are usually stabilized as U+4 (5f2) with
few exceptions of U3+ and U“, all of which still possess f electrons. ' This makes
uranium compounds of undoubted interest because characterization of these materials
could provide us with a better understanding of the nature of relatively less studied 5f
orbitals. Binary and ternary uranium Chalcogenides have been well studied for their
magnetic properties.2’3’4’5 However, it was not until very recently that quaternary
uranium compounds with more complex structure types were discovered (i.e. CsTiUTe5,
6 CsCuUTe3,6 ngHfsUTe3o_6,7 KCuUSe3,8 Rb4U4P4Se26, 9 K2UP3Se9,'° ). These
compounds show various types of uranium-chalcogen skeletons with different local
environments for the uranium atoms. The low-melting polychalcogenide ﬂux method has
been useful in preparing quaternary compounds because binary or ternary compounds
are, in general, more thermodynamically stable at higher temperature (>900 °C) and their
formation can be avoided by lowering the reaction temperature. We have intended to
combine U metal with 8b metal in the alkali metal polychalcogenide ﬂuxes. The Sb metal
has been known to induce an asymmetric local environment due to the presence of the 552
lone pair.“’12’l3 Therefore, we expected to discover unique structure types by
combining large and ﬂexible U coordinations with highly distorted Sb coordinations.

Reactions of Abesz( A = Rb, K; Q = 8, Se) ﬂux with uranium metal yielded
new compounds RbU2Sng and KU2SbSeg, which present a unique polar
noncentrosymmetric arrangement of atoms. Compounds with noncentrosymmetric

Structures are of considerable interest due to their potential applications as piezoelectrics,

23

ferroelectrics and nonlinear optics.l4 Up to date, none of the known quaternary uranium
Chalcogenides has been reported to adopt a noncentrosymmetric structure.

Changing chalcogen types from Se/S to Te in the ﬂux, we expected to discover
new structure types. The nature of tellurium is quite different from those of Se and S in
engaging in chalcogen-chalcogen interactions due to its more diffuse orbitals. Tellurium
possesses a greater tendency to associate Te-Te bonding interactions with neighboring
tellurium and they often stabilize Te-Te bonds longer than a normal single bond in the
formation of one-dimensional chains or nets by delocalizing electrons. Much of the
interest in these polytelluride compounds stems from distortion of these chains and nets
caused by modulations of Te-Te bonds, which result in interesting superstructures and
changes in the physical properties of the materials.

Indeed, reaction of uranium metal with antimony polytelluride ﬂuxes resulted in
the discovery of a new compound, RbUSbol33Te6, which features a novel way to
polymerize Te atoms one-dimensionally. There exist three inﬁnite zigzag Te chains
wrapped around one-dimensional U arrays, which makes the structure chiral. These
chains are different from the more familiar helical chains found in elemental Te15 or
CuTeX“S or CuTe2Xl7 (X = Cl, Br, I), in their polymerization patterns, formal charges of
the chains, and the Te-Te distances in the chains. In fact, the superstructure reﬁnement
of the compound revealed that the Te-Te distances in the chain modulate as in the case of
extensively studied two-dimensional Te nets (i.e. LnTe2, and LnTe3). Here we report the
synthesis, structural and physicochemical characterization of RbU2Sng, KU2SbSeg, and

RbUSb033T65.

24

2. Experimental Section

2.1 Synthesis.

Reagents. The following reagents were used as obtained: uranium, 99.7%, -60 mesh,
Cerac, Milwaukee, WI; antimony, 99.999%, -200 mesh, Cerac, Milwaukee, WI; selenium
shots, 99.9% Noranda Advanced Materials, Saint-Laurent, Quebec, Canada; sulfur
powder, sublimed, JT Baker Co., Phillipsberg, NJ; tellurium shots, 99.9% Noranda
Advanced Materials, Saint-Laurent, Quebec, Canada; rubidium metal, 99.8%, Johnson
Matthey Co., Ward Hill, MA; potassium metal, analytical reagent, Spectrum Chemical
Mfg. Corp., Gardena, CA. The starting materials K2Se / Rb2S / Rb2Te were prepared by
a stoichiometric reaction of potassium/rubidium metal and selenium/sulfur in liq. NH3.
RbU2Sng. This compound was synthesized from a mixture of 0.0633 g (0.3 mmol)
Rb28, 0.0714 g (0.3 mmol) U, 0.0366g (0.3 mmol) Sb, and 0.0867 g (2.7 mol) 8. The
reagents were thoroughly mixed, sealed in an evacuated quartz ampule, and heated at 650
°C for 5 days (cooling 2 °C/h). Pure reddish black, block-like crystals of RbU28ng were
obtained by isolation in degassed dimethylformamide (DMF) and water (yield > 90%
based on U metal). The crystals are air- and water stable.

KU2SbSeg. This compound was synthesized from a mixture of 0.0471g (0.3 mmol)
K2Se, 0.0714 g (0.3 mmol) U, 0.0366 g (0.3 mmol) Sb, and 0.2132 g (2.7 mmol) Se. The
reagents were thoroughly mixed, sealed in an evacuated quartz ampule, and heated at 650
°C for 5 days (cooling 2 °C/h). Pure golden black, block-like crystals of KU28bSe8 were
Obtained by isolation in degassed dimethylformamide (DMF) and water (yield > 90%

based on U metal). The crystals are air- and water stable.

25

RbUSboxTeg. The compound was synthesized from a mixture of 0.1194g (0.4
mmol) Rb2Te, 0.0952 g (0.4 mmol) U, 0.0244 g (0.2 mmol) Sb, and 0.2552 g (2.0 mmol)

Te. The reagents were thoroughly mixed, sealed in an evacuated quartz ampule, and

heated at 720 0C for 5 days (cooling 2 0C/h). Pure black needle-like crystals of
RbUSbo_33Te6 were obtained by isolation in degassed dimethylformamide (DMF) and
water (yield ~ 80% based on U metal). The crystals are air- and water stable.

The compositions of the materials were analyzed by Scanning Electron
Microscope (SEM)/ Energy Dispersive Spectroscopy (EDS). Homogeneity for all
compounds was conﬁrmed by comparing the powder X-ray diffraction patterns of the

products against ones calculated using X-ray single crystal data.

2.2 Physical Measurements.

Solid-State UV/V is Spectroscopy. Optical diffuse reﬂectance measurements were
performed at room temperature with a Shimadzu UV-3101 PC double-beam, double-
monochromator spectrophotometer operating in the 200-2500 nm region. The instrument
is equipped with an integrating sphere and controlled by a personal computer. BaSO4
was used as a 100% reﬂectance standard for all materials. Samples were prepared by
grinding them to a ﬁne powder and spreading them on a compacted surface of the
powdered standard material and preloaded them into a sample holder. The reﬂectance
versus wavelength data generated were used to estimate a material's band gap by

converting reﬂectance to absorption data as described previously.18

26

 

Infrared Spectroscopy. Optical diffuse reﬂectance measurements were made on the
ﬁnely ground sample at room temperature. The spectrum was recorded in the infrared
region (6000 - 400 cm") with the use of a Nicolet MAGNA-IR 750 Spectrometer
equipped with a Collector Diffuse Reﬂectance of Spectra-Tech. Inc. The reﬂectance
versus wavelength data were used to estimate a material's band gap by converting

reﬂectance to absorption data as described previously. '8

Magnetic Susceptibility. The magnetic susceptibilities for RbU2Sng and KU2SbSe8
were measured over the range 2-300 K using a MPMS Quantum Design SQUID
magnetometer. Sample was ground to a ﬁne powder to minimize possible anisotropic
effects and loaded into PVC containers. The temperature dependent susceptibility studies
were performed at 1000 Gauss applied ﬁled. Corrections for the diamagnetism of the
sample containers were made by measuring the magnetic response of the empty container
under the same conditions of temperature and ﬁeld which were measured for the ﬁlled

container. Diamagnetic contribution of every ion to xMwas corrected according to

Selwood. ‘9

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~20 mg )
of ground crystalline material was sealed in quartz ampules under vacuum. A quartz
ampule of equal mass ﬁlled with A1203 was sealed and placed on the reference side of the
detector. The samples were heated to 800 °C at 5 °C/min, isotherrned for 10 min

f0110wed by cooling at -5 °C/min to 50 °C. Residues of the DTA experiments were

27

examined by X-ray powder diffraction. The stability/reproducibility of the samples was

monitored by running multiple heating/cooling cycles.

Raman Spectroscopy. Raman Spectra were recorded on a Holoprobe Raman
spectrograph equipped with a CCD camera detector using 633nm radiation from a HeNe
laser for excitation. Laser power at the sample was estimated to be about SmW and the
focused laser beam diameter was ca. 10 microns. 5 scans were needed to obtain good

quality spectra. The accumulation time of each scan was 50 sec.

Charge-Transport Measurements. DC electric conductivity and thermopower
measurements were made on single crystals of the compounds. Conductivity
measurements were performed in the usual four-probe geometry with 60- and 25-um gold
wires used for the current and voltage electrodes, respectively. Measurements of the
sample cross-sectional area and voltage probe separation were made with a calibrated
binocular microscope. Conductivity data were obtained with the computer-automated

2° Thermoelectric power measurements were made by using

system described elsewhere.
a slow AC technique”21 with 60 um gold wires serving to support and conduct heat to
the sample, as well as to measure the voltage across the sample resulting from the applied
temperature gradient. In both measurements, the gold electrodes were held in place on the
sample with a conductive gold paste.

Conductivity specimens were mounted on interchangeable sample holders, and

thermopower specimens were mounted on a ﬁxed sample holder/differential heater.

Mounted samples were placed under vacuum (10'3 Torr) and heated to room temperature

28

for 2-4 h to cure the gold contacts. For a variable-temperature run, data (conductivity or
thermopower) were acquired during both sample cooling and warming to check
reversibility. The temperature drift rate during an experiment was kept below 1 K/min.
Typically, three to four separate variable-temperature runs were carried out for each
sample to ensure reproducibility and stability. At a given temperature, reproducibility

was within :t5 %.

2.3 X-ray Crystallography.

RbU2Sng and KU2SbSeg Single crystals of RbU2Sng with dimensions 0.28 x
0.13 x 0.10 mm and of KU2SbSeg with dimensions of 0. 12 x 0.10 x 0.04 mm were
mounted on the tip of a glass ﬁber. Intensity data for RbU2Sng were collected at room
temperature on a Rigaku AF C68 four circle automated diffractometer equipped with a
graphite-crystal monochromator. The unit cell parameters were determined from a least-

squares reﬁnement using the setting angles of 20 carefully centered reﬂections in the

8°<20<300 range. The data were collected with the to scan technique over a full sphere
of reciprocal space, up to 50 deg in 20. Crystal stability was monitored with three
standard reﬂections whose intensities were checked every 150 reﬂections. No signiﬁcant
decay was observed during the data collection period. An empirical absorption correction
was applied based on u! scans.

Intensity data for KU2SbSeg were collected at room temperature on a Siemens
SMART Platform CCD diffractometer using graphite monochromatized Mo Ka

radiation. The data were collected over a full sphere of reciprocal space, up to 50.04 deg

29

in 20. The individual frames were measured with an omega rotation of 0.3 deg and an
acquisition time of 30 sec. The SMART22 software was used for the data acquisition and
SAINT23 for the data extraction and reduction. The absorption correction was performed
using SADABSZ4. The complete data collection parameters and details of the structure
solution and reﬁnement for both compounds are given in Tables 2.1 and 2.2.

Structure solution and reﬁnement for both compounds were performed with the
SHELXTL package of crystallographic programs.25 Systematic absence conditions of the
data sets gave four possible space groups, Cmmm, Cmm2, Amm2, C222. The structures
were solved and reﬁned successfully in the polar Amm2 space group. Due to the absence
of mirror planes perpendicular to the c-axis, these structures are clearly polar and could
not be solved in the centrosymmetric space group Cmmm. All the crystals used for the
structure solution were racemically twinned and the BASF parameters were reﬁned. In
RbU2Sng, the Sb3+ and Rb+ sites were half-occupied and the distance between them was
0.92A indicating that only one of the ions is present at a time with the probability of
50%. KU2SbSeg showed the same disorder pattern between the Sb3+ and K+ ions. These
results suggest that there may exist a supercell which could resolve the disorder.

Intensity data for both compounds were recollected on a Siemens SMART
Platform CCD diffractometer with better grown crystals and a longer acquisition time of
70 sec per frame to probe the existence of a superstructure. It was indeed found that
RbU2Sb83 and KU2SbSe8 possess monoclinic supercells. Interestingly, the supercells are
different for each compound. The vectorial relationships between the supercell

parameters, a’, b’, c’ and the subcell parameters, a,b,c for each compound are as follows:

30

RbU2Sng : a’ = 2b b’ = -2a c’ = -0.5b + 0.50

KU2SbSeg : a’ = 2a b’ = 2b c’ = -a + c

The complete data collection parameters and details of the superstructure solutions and
reﬁnements are given in Tables 2.1 and 2.2.

Both superstructures were solved and reﬁned successfully in the polar Cm space
group. The fact that RbU2Sng and KU2SbSe3 were solved in the same space group is
coincidental as they have different unit cells and different monoclinic b-axes. Again both
of the superstructures are clearly polar (see Figure 1.1 and 1.4) and could not be solved in
the centrosymmetric space group C 2/m.

The coordinates of all atoms, isotropic temperature factors, and their estimated
standard deviations (esd’s) for RbU2Sng and KU2SbSeg for both the substructure and
superstructure are given in Tables 2.3, 2.4, and 2.5. The selected bond distances and

angles for the superstructures are given in Tables 2.6 and 2.7.

RbUSbo—23Te6 Single crystals of RbUSb033Te6 with dimensions 0.18 x 0.05 x
0.03 mm was mounted on the tip of a glass ﬁber. Intensity data were collected at room
temperature on a Siemens SMART Platform CCD diffractometer using graphite
monochromatized Mo Ka radiation. The data were collected over a full sphere of
reciprocal space, up to 58.98 deg in 20. The individual frames were measured with an
omega rotation of 0.3 deg and an acquisition time of 65 sec. The SMART software was
used for the data acquisition and SAINT23 for the data extraction and reduction. The

absorption correction was performed using SADABS.24 The complete data collection

31

parameters and details of the structure solution and reﬁnement are given in Table 2.8.
The coordinates of all atoms, isotropic temperature factors, and their estimated standard
deviations are given in Table 2.9. The selected bond distances and angles are given in
Table 2.10. The anisotropic displacement parameters are listed in Table 2.11.

Structure solution and reﬁnement was performed with the SHELXTL package 25 of
crystallographic programs. The structures were solved and reﬁned successfully in the
chiral P3 space group. The crystal used for the structure solution was racemically
twinned and the BASF parameter was reﬁned. The structure reﬁnement showed several
symptoms that implied the existence of a superstructure such as Sb disorder, short Rb-Sb
distance, and enormously huge anisotropic atomic displacement parameters of Rb.

Intensity data was recollected on a Siemens SMART Platform CCD
diffractometer with better grown crystals and a longer acquisition time of 70 sec per frame
to probe the existence of a superstructure. It was indeed found that RbU033SbTe6
possess R-centered trigonal supercells, cell volume of which is 9 times bigger than that of
the substructure (la’l = V3 |a|, |b'| = ‘13 |b|, lc’l = 3 lcl). The superstructure was reﬁned in the
chiral R3 space group. The R values for the superstructure reﬁnement, however , did not
drop below 0.10 due to a possible twin, poor statistics of the superstructure reﬂections,
and an possible existence of a even bigger superstructure. The complete data collection
parameters and details of the superstructure solutions and reﬁnements are given in Table
2.8. The coordinates of all atoms, isotropic temperature factors, and their estimated
standard deviations for the superstructure are given in Table 2.12. The selected bond

distances are given in Table 2.13.

32

Table 2.1. Summary of Crystallographic Data and Structural Analysis for RbUZSng.

 

 

Substructure Superstructure
Formula RbUZSng RbUZSbS3
Formula weight 939.76 939.76
Crystal habit Reddish black blocks
Space group Amm2 Cm
a, A 5.551(1) 7.9543(9)
b, A 3.978(2) 11.099(1)
c, A 14.057(5) 7.279(1)
B, degree- 90 106.030(2)
z; v, A3 1, 310.4(2) 2, 617.7(1)
Dom, g/em3 5.028 5.053
Temp, K 298(2) 170(2)
MMo K01), A 0.71073 0.71073
“(Mo K01), cm" 333.71 335.38
F(000) 400 800
9 max , deg 24.75 28.26
Total data 1096 2843
Unique data 348 [R....= 0.0657] 1252 [Rint= 0.0503]
No. of variables 26 63
Reﬁnement method Full-matrix least-squares on F 2
Final R indices [I>20] R1“ =0.0580 R1=0.029O
WR2” =0.1408 wR2=0.0697
R indices (all data) R1=0.0580 R1=0.0294
wR2=0.1408 wR2=0.0700
GOF on F2 1.239 1.073
BASF 054(4) 052(1)

 

'R1=2||FoI-|Fcll/Z|Fol ;

b wR2= {2[w(Fo2—F.2)2]/2‘.[w(17.,2)2]}"2

33

Table 2.2. Summary of Crystallographic Data and Structural Analysis for KU2SbSe3.

 

 

Substructure Superstructure
Formula KUszSCg KUszSCg
Formula weight 1268.59 1268.59
Crystal habit Golden black blocks
Space group Amm2 Cm
a, A 5.7811(8) 11.5763(2)
b, A 4.0988(5) 8.2033(1)
c, A 14.184(2) 15.2742(1)
B, degree. 90 112.22(2)
z; v, A3 1, 336.10(8) 4, 1342.75(3)
Dem, g/om3 6.268 6.275
Temp, K 298(2) 170(2)
MMO K01), A 0.71073 0.71073
u(Mo Kd), cm" 479.47 480.06
F(000) 526 2104
9 max , deg 25.02 28.20
Total data 882 5255
Unique data 266 [Riﬂe 0.0554] 3031 [Rm= 0.0447]
No. of variables 26 132
Reﬁnement method F ull-matrix least-squares on F 2
Final R indices [I>26] R1“=0.0355 R1=0.0577
wR2b=0.0904 wR2=0.1839
R indices (all data) R1=0.0355 R1=0.0724
wR2=0.0904 wR2=0.2006
GOF oan 1.157 1.137
BASF 044(4) 076(2)

 

‘R1=2|IFoI-chll/2|Fo

 

9

b wR2= {20608.2—F.2)2]/z[w(F.2)2]}”2

34

Table 2.3. Fractional Atomic Coordinates and Equivalent Isotropic Displacement
Parameter Values for RbU2SbS3 (Ammm) and KUZSbSeg (Ammm) with Estimated

Standard Deviations in Parentheses.

 

 

 

 

 

Atom x y z Ueqf A2 Occ.
RbUZSb33
U 0.5 0 0.0509(1) 0.011(1) 1
Sb 0 0 0.3959(5) 0.015(1) 0.5
Rb 0 0 0.3159(7) 0.015(1) 0.5
S(1) 0.5 0 0.4194(10) 0.013(3) 1
8(2) 0 0 0.0006(8) 0.013(2) 1
8(3) 0.3124(21) 0 0.6710(6) 0.021(2) 1
KUzsbSeg
U 0.5 0 0.0513(1) 0.010(1) 1
Sb 0 0 0.3879(5) 0.015(1) 0.5
K 0 0 0.3174(15) 0.015(1) 0.5
8e(l) 0.5 0 0.4113(3) 0.012(1) 1
Se(2) 0 0 0.0000(3) 0.012(1) 1
Se(3) 0.2964(4) 0 0.6751(3) 0.018(1) 1

 

“ ch is deﬁned as one third of the trace of the orthogonalized Uij tensor.

Table 2.4. Fractional Atomic Coordinates and Equivalent Isotropic Displacement

Parameter Values for RbUZSbsg (Cm) with Estimated Standard Deviations in

 

 

 

Parentheses.
Atom x y z Um,a A2
U(l) 0.3298(1) 2498(1) 0.3126(2) 0.005(1)
Sb(l) 0.0000(1) 0 0.0000(2) 0.008(1)
Rb(1) 0.4596(2) 0 0.8420(2) 0.009(1)
8(1) 0.0097(5) 2404(2) 0.0422(5) 0.008(1)
8(2) 0.2952(5) 0 0.2165(6) 0.009(1)
8(3) 0.1346(4) 0.1637(3) 0.5565(5) 0.010(1)
8(4) 0.1406(4) 0.3523(2) 0.5444(5) 0.010(1)
8(5) 0.8125(5) 0 0.2148(6) 0.009(1)

 

a ch is deﬁned as one third of the trace of the orthogonalized Uij tensor.

36

Table 2.5.. Fractional Atomic Coordinates and Equivalent Isotropic Displacement

Parameter Values for KU2SbSe3 (Cm) with Estimated Standard Deviations in

 

 

Parentheses.
Atom x y z ch,“ A2
U(l) 0.4990(1) 0 0(1) 0.009(1)
U(2) 0.0000(1) 0 0(1) 0.009(1)
U(3) 0.2493(1) 0.2500(1) 0.5000(1) 0.009(1)
8b(1) 0.3337(4) 0.2495(22) 0.1618(3) 0.017(1)
8b(2) 0.5788(3) 0 0.6644(3) 0.012(1)
K(1) 0.3692(13 0.2460(22) 0.2336(9) 0.017(1)
K(2) 0.1128(7) 0 0.7327(6) 0.016(1)
8e(1) 0.0692(3) 0.2500(2) 0.1402(2) 0.013(1)
8e(2) 0.2743(3) 0 0.0498(3) 0.013(1)
36(3) 0.3341(2) 0.2544(2) 0.8745(2) 0.014(1)
8e(4) 0.0400(3) 0.2550(3) 0.8765(2) 0.015(1)
Se(5) 0.7746(3) 0 0.0502(3) 0.013(1)
Se(6) 0.3261(3) 0 0.6407(3) 0.011(1)
Se(7) 0.0251(2) 0.251 1(2) 0.5512(2) 0.014(1)
8e(8) 0.0820(3) 0 0.3779(3) 0.017(1)
8e(9) 0.2832(3) 0 0.3736(2) 0.014(1)
Se(10) 0.8131(3) 0 0.6412(3) 0.014(1)
Se(11) 0.5894(3) 0 0.3754(2) 0.015(1)
Se(12) 0.7930(3) 0 0.3756(2) 0.014(1)

 

" Ueq is deﬁned as one third of the trace of the orthogonalized U),- tensor.

37

Table 2.6. Selected Distances (A) and Bond Angles (°) for RbU2Sng (Cm).

 

Bond Distances

 

 

 

U(l)—S(l) 2.737(4) Rb(1)-S(1) 3.203(3) x 2
U(l)—S(3) 2.752(3) Rb(1)-S(5) 3.322(4)
U(l)-S(1) 2.757(3) Rb(1)-S(2) 3.333(5)
U(1)-S(4) 2.795(3) Rb(1)-S(4) 3.343(4) x 2
U(1)-S(4) 2.818(3) Rb(1)-S(3) 3.367(3) x 2
U(l)—S(3) 2.829(3) Sb(1)-S(2) 2.440(4)
U(l)—S(2) 2.854(1) Sb( 1)-S(5) 2.441(5)
U(l)-S(5) 2.861(1) Sb(l)-S(1) 2.684(2) x 2
S(3)-S(4) 2.09645)

Bond Angles
S(1)-U(1)—S(3) 84.71(10) S(2)-Sb(1)-S(5) 103.6(2)
S(1)-U(1)—S(1) 92.93(12) S(2)-Sb(1)-S(1) 8621(8)
S(1)-U(1)-S(4) 153.36(8) S(5)-Sb(1)-S(1) 8600(9)
S(3)-U(l)-S(4) 8964(9) S(5)-Sb(1)-S(1) 86.00(9)
S(4)—U(1)—S(4) 109.38(12) S(1)—Sb(1)-S(1) 167.4(2)
S(3)-U(l)-S(3) 104.55(1 1) S(1)-Rb(l)-S(1) 128.17(l3)
S(4)-U(1)—S(3) 43.77(1 1) S(1)-Rb(1)-S(5) 69.97(7)
S(1)-U(1)-S(2) 84.07(10) S(5)-Rb(1)-S(2) 76.46(11)
S(3)-U(1)-S(2) 120.35(10) S(1)-Rb(1)-S(4) 135.S4(11)
S(4)-U(1)-S(2) 120.78(1 1) S(5)—Rb(1)-S(4) 9639(9)
S(3)-U(l)-S(2) 78.05(1 l) S(2)—Rb( 1 )-S(4) 150.05(6)
S(1)-U(1)-S(5) 77.37(10) S(1)-Rb(1)-S(4) 8033(9)
S(4)—U(1)-S(5) 75.99(1 1) S(4)-Rb(1)-S(4) 58.71 (1 1)
S(3)-U(l)-S(5) 1 19.14(1 1) S(5)-Rb(l)-S(3) 146.69(6)

 

38

Table 2.7. Selected Distances (A) and Bond Angles (°) for KUzsbSeg(Cm).

 

 

 

 

Bond Distances
U(1)-Se(l) 2.852(3) x 2 Sb(1)-Se(2) 2.588(14)
U(1)-Se(4) 2.917(2) x 2 Sb(1)-Se(5) 2.591(14)
U(1)-Se(2) 2.973(3) Sb(l)-Se(1) 2.868(5)
U(1)-Se(3) 2.984(2) x 2 Sb(1)—Se(1) 2.952(5)
U(l)—Se(5) 2.987(3) Sb(2)-Se(7) 2.594(3) x 2
U(2)-Se(l) 2.852(3) x 2 Sb(2)-Se(6) 2.809(4)
U(2)-86(3) 2.939(2) x 2 Sb(2)-Se(10) 2.866(5)
U(2)-8e(4) 2.969(2) x 2 K(1)-Se(1) 3.147(14)
U(2)-8e(2) 2.972(3) K(1)-Se(1) 3.216(14)
U(2)-Se(5) 2.984(3) K(1)-Se(2) 3.29(2)
U(3)-8e(6) 2.859(3) K(1)-Se(5) 333(2)
U(3)-Se(10) 2.863(3) K(1)-Se(1 1) 333(2)
U(3)-Se(1 1) 2.931(2) K(1)-8e(8) 3 34(2)
U(3)-Se(9) 2.943(2) K(1)-8e(9) 335(2)
U(3)-8e(8) 2.950(3) K(1)-8e(12) 336(2)
U(3)-Se(12) 2.965(2) K(2)-Se(10) 3.213(9)
U(3)-Se(7) 2.980(3) K(2)-Se(6) 3.271(8)
U(3)-Se(7) 2.987(3) K(2)-Se(7) 3.292(7) x 2
Se(3)-Se(4) 2.375(4) K(2)-Se(4) 3.362(6) x 2
8e(8)-8e(9) 2.356(5) K(2)-86(3) 3.378(7) x 2
Se(l l)-Se(12) 2.356(5)

Bond Angles

Se(l)-U(1)-Sc(1) 91 .95(1 1) Se(2)-Sb(1 )-Se(5) 104.7(2)
Se(1)-U(l)-Se(4) 155.68(8) Se(2)-Sb(1 )-Se( 1) 87.2(3)
Se(4)-U(1)—Se(4) 87.13(10) Se(5)-Sb(1)-Se( 1) 87. 1(3)
Se(l)-U(1)-Se(2) 8061(8) Se(2)-Sb(l)-Se(1) 85.5(3)
Se(4)-U(1)-Se(2) 122.54(7) Se(5)-Sb(1)-Se(1) 85.5(3)

39

Table 2.7. (Cont’d).

 

Se(1)-U(l)-Se(3) 156.44(9) Se(l)-Sb(1)-Se(1) 167.9(2)
Se(4)-U(1)-Se(3) 47.44(8) Se(l)—K(1)-Se(1) 130.9(4)
Se(2)-U(1)-Se(3) 7583(8) Se(l)-K(1)-Se(5) 71.1(3)
Se(3)—U(l)-Se(3) 8874(9) Se(l)-K(1)-Se(1 1) 136.7(6)
Se(l)-U(l)-Se(5) 8042(8) Se(2)-K(1)-Se(11) 96.3(4)
Se(4)-U(l)-Se(5) 7530(8) Se(5)-K(1)-Se(1 1) 148.7(5)
Se(2)-U(l)-Se(5) 152.57(12) Se(l)-K(1)—Se(8) 133.6(5)
Se(3)-U(1)-Se(5) 121.78(7) Se(2)-K(1)-Se(8) 150.5(5)
Se(5)—K(1)-Se(8) 95.2(5)
8e(1 1)-K(1)-Se(8) 75.9(3)
Se(8)-K(1)-Se(9) 106.1(4)
Se(2)-K(1)-Se(12) 147.9(5)
Se(5)-K(1)-Seﬂ2) 94.1(4)

 

40

Table 2.8. Summary of Crystallographic Data and Structural Analysis for RbUSbo_33Te6.

 

 

Substructure Superstructure
Formula RbUSb033TC6 RbUSbo,33Te6
Formula weight 1 129.68 1 129.68
Crystal habit black needles
Space group P3 R3
a, A 9.0925(14) 15.741(2)
b, A 9.0925(14) 15.741(2)
c, A 8.129(2) 24.382(4)
Y, deg 120 120
z; v, A3 2, 582.0(2) 18, 5231.9(14)
Dcalc, g/cm3 6.454 6.454
Temp, K 293(2) 293(2)
K(Mo K01). A 0.71073 0.71073
u(Mo K11). cm" 335.30 335.67
F(000) 916 8244
6 max . deg 28.49 24.98
Total data 5473 14489
Unique data 1780 [R.,..=0.0476] 4095 [R....=0.0531]
No. of variables 60 184
Reﬁnement method Full-matrix least-squares on F 2
R indices 11>20] R13 = 0.0682 R1 = 0.1072
wR2b=0.1571 wR2=0.1973
R indices (all data) R1 = 0.0703 R1 = 0.1125
wR2=0.1584 wR2=0.1994
Goodness of ﬁt on F2 1.021 1.151
BASF 014(2) 014(2)

 

3 R1: ZIIFOI — chll/ZIFOI ;

b wR2= {2;[w(1~".2-—I~:.2 )2 ]/2[w(F..2)2]}"2

41

Table 2.9. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter Values for RbUSbo33Te6 (substructure) with Estimated Standard Deviations in

 

 

Parentheses.

Atom x y 2 U eqf A2 Occ.
U(l) 2/3 1/3 0.0014(2) 0.012(1) 1
U(2) 2/3 1/3 0.5014(2) 0.012(1) 1
Te(l) 0.0486(3) 0.4290(3) 0.0015(4) 0.033(1) 1
Te(2) 0.4010(3) 0.3359(4) 0.2490(3) 0.030(1) 1
Te(3) 0.2845(3) 0.0464(4) 0.5011(5) 0.039(1) 1
Te(4) 0.4012(3) 0.3360(4) 0.7539(3) 0.030(1) 1
Rb(l) 1/3 2/3 0.5015(30) 0.117(8) 1
Rb(2) 0 0 0.0027(45) 0.188(15) 1
Sb(l) 1/3 2/3 0.1394(40) 0.019(4) 0.127(14)
Sb(2) 1/3 2/3 0.8648(44) 0.019(4) 0.115(13)
Sb(3) 0 0 0.3532(24) 0.019(4) 020(2)
Sb(4) 0 0 0.6483(25) 0.019(4) 0.20(2)

 

’ ch is deﬁned as one third of the trace of the orthogonalized U”. tensor.

42

Table 2.10. Selected bond lengths (A) and angles (°) for RbUSbo,33Te6 (substructure).

 

 

 

 

Bond Lengths
U(l)-Te(l) 3.130(3) x 3 U(2)-Te(2) 3.179(3) x 3
U(l)-Te(2) 3.153(3) x 3 U(2)-Te(3) 3.133(3) x 3
U(1)-Te(4) 3.152(3) x3 U(2)-Te(4) 3.178(3) x 3
Sb(l)-Te(1) 2.652(14) x 3 Sb(3)-Te(2) 2.493(5) x 3
Sb(l)-Te(2) 3.474(9) x 3 Sb(3)-Te(3) 2.688(9) x 3
Sb(2)-Te(1) 265(2) x 3 Sb(4)-Te(3) 2.685(9) x 3
Sb(2)-Te(4) 3.477(10) x 3 Sb(4)—Te(4) 3.497(6) x 3
Rb(1)-Te(2) 3.935(13) x 3 Rb(2)-Te(1) 3.700(3)
Rb(l)-Te(3) 3.694(3) x 3 Rb(2)-Te(2) 3.94(2)
Rb(l)-Te(4) 3.936(13) x 3 Rb(2)-Te(4) 3.95(2)
Te( 1 )-Te(2) 3.017(5) Te(3)-Te(4) 3.081(5)
Te(2)-Te(3) 3.077(5) Te(4)-Te(1) 3.017(5)
Rb(l )-Sb( 1) 294(4) Rb(2)-Sb(3) 2.85(4)
Rb(l)-Sb(2) 295(4) Rb(2)-Sb(4) 2.88(4)
8b(1)-Sb(2) 223(5) Sb(3)-Sb(4) 2.4%)

Bond Angles
Te(1)-U(1)-Te(2) 57.40(9) Te(l)—Sb(1)-Te(1) 103.4(8)
Tc(2)-U(l)-Te(4) 134.74(8) Te(2)-Sb(l)-Te(2) 113.7(4)
Te(2)—Rb( l )-Te(3) 62.16(14) Te(1)-Sb(1)-Te(2) 160.3(6)
Te(2)-Rb(1)-Te(4) 129.52(5) Te(4)-Te(1)-Te(2) 83.66(10)
Sb(l)-Rb(1)-Te(2) 58.6(3) Te(1)-Te(2)-Te(3) 170.34(11)
Sb(l)-Rb(l)-Te(4) 121.4(3) Te(2)-Te(3)-Te(4) 83.63(10)

Te(3)-Te(4)-Te(1) 170.39(1 1)

 

43

Table 2.11. Anisotropic Displacement Parameters for RbUSbo,33Te6.

 

 

U11 U22 U33 U23 U13 U12
U(l) 0.010(1) 0.010(1) 0.016(1) 0 0 5(1)
U(2) 0.011(1) 0.011(1) 0.015(1) 0 0 6(1)
Te(l) 0.020(1) 0.031(1) 0.047(2) -1(1) -1(1) 11(1)
Te(2) 0.026(1) 0.047(2) 0.025(1) 90) -2(1) 23(1)
Te(3) 0.018(1) 0.029(1) 0.066(2) 1(1) 0(1) 9(1)
Te(4) 0.023(1) 0.045(2) 0.027(1) 9(1) 1(1) 21(1)
Rb(1) 0.027(3) 0.027(3) 0.296(25) 0 0 14(1)
Rb(2) 0.027(3) 0.027(3) 0.510(46) 0 0 13(2)

 

The anisotropic displacement factor exponent takes the form:

-21t2[h2a*2U11+ -------- +2hka*b*U12]

Table 2.12. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for RbUSbo,33Teé (superstructure) with Estimated Standard

Deviations in Parentheses.

 

 

Atom x y 2 U sq,“ A2 Occ.
U(l) 0.3331(1) 0.0001(1) 0.9999(1) 0.011(1) 1
U(2) 0.3324(1) 0.0016(1) 0.1666(1) 0.011(1) 1
Rb(1) 0 0 0.6676(11 0.043(4) 0.498(13)
Rb(l') 0 0 0.6263(15) 0.043(4) 0.502(13)
Rb(2) 0 0 0.3314(11) 0.034(5) 0.502(13)
Rb(2') 0 0 0.3681(16) 0.034(5) 0.498(13)
Rb(3) 0 0 0.9993(8) 0.055(4) 1
Rb(4) 0.3294(9) 0.3299(10) 0.1671(10) 0.033(4) 0.670(1)
Rb(4') 0.3313(18) 0.3380(20) 0.1355(14) 0.033(4) 0.179(12)
Rb(4") 0.3366(18) 0.3364(19) 0.1982(14) 0.033(4) 0.151(12)
8b(1) 0.6667 0.3333 0.1180(7) 0.017(4) 0.502(13)
Sb(2) 0 0 0.2126(6) 0.014(4) 0.498(13)
Sb(3) 0.3274(16) 0.3301(15) 0.6172(8) 0.053(6) 0.179(12)
Sb(4) 0.3357(19) 0.3303(14) 0.0465(11) 0.066(8) 0.151(12)
Te( 1) 0.2700(4) 0.1605(3) 0.6636(3) 0.047(1) 1
Te(2) 0.4234(3) 0.1781(3) 0.5828(2) 0.027(1) 1
Te(3) 0.1732(3) 0.0677(3) 0.1616(2) 0.036(2) 1
Te(4) 0.0868(3) 0.02415(3) 0.0833(2) 0.027(1) 1
Te(5) 0.2684(3) 0.1577(3) 0.3340(2) 0.035(1) 1
Te(6) 0.1536(3) -0.0892(3) 0.9169(2) 0.031(1) 1
Te(7) 0.5587(3) 0.1578(3) 0.1718(2) 0.028(1) 1
Te(8) 0.4221(3) 0.1797(3) 0.0828(2) 0.039(1) 1
Te(9) 0.2708(3) 0.159943) 0.10017(2) 0.036(1) l

 

45

Table 2.12. (Cont’d).

Te(10) 0.4237(3) 0.1771(3) 0.9137(2) 0.029(1) 1
Te(l 1) 0.2753(3) 0.2183(2) 0.1659(2) 0.023(1) l
Te(12) 0.4232(3) 0.1761(3) 0.7529(1) 0.022(1) 1

 

' ch is deﬁned as one third of the trace of the orthogonalized U”. tensor.

46

 

Table 2.13. Selected bond lengths (A) for RbUSbo_33Te6 (superstructure).

 

 

Bond Lengths
U(1)-Te(5) 3.1 19(4) U(2)-Te(1 1) 3.1 1 1(4)
U( l )-Te(9) 3.127(4) U(2)-Te(12) 3.126(4)
U(l)-Te(1) 3.150(5) U(2)-Te(10) 3.137(4)
U(1)-Te(4) 3. 1 70(4) U(2)-Te(2) 3. 147(4)
U( 1)-Te(2) 3.174(4) U(2)-Te(3) 3.159(4)
U(1)-Te(8) 3.176(5) U(2)-Te(4) 3.161(4)
U( l )-Te(6) 3.177(4) U(2)-Te(7) 3.161(4)
U(1)-Te(12) 3.194(4) U(2)-Te(6) 3.170(4)
U(1)-Te(10) 3.204(4) U(2)-Te(8) 3.173(5)
Rb( 1 )-Te( 1) 3 .704(5) Rb(l ')-Te(10) 3.539(12)
Rb(1)-Te(1) 3 .704(5) Rb(l ')-Te(10) 3 .539(1 2)
Rb(1)-Te(] ) 3.704(5) Rb(l ')-Te(10) 3.539(12)
Rb(1)-Te(8) 3.917(14) Rb(l’)-Te(l) 3.813(10)
Rb(1)-Te(8) 3.917(14) Rb(1')-Te(1) 3.813(10)
Rb(1)-Te(8) 3.917(14) Rb(1’)-Te(1) 3.813(10)
Rb(1)-Te(10) 3 .974(1 5) Rb( 1 ')-Te(1 l) 3 .92(3)
Rb(1)-Te(10) 3 .974(1 5) Rb(1')-Te(1 1) 3 .92(3)
Rb(1)-Te(10) 3.974(15) Rb(1')-Te(1 l) 3 .92(3)
Rb(1)-Rb(l ') 1 .01 (4) Rb(1')-Sb(l) 3 .86(4)
Rb(1)—8b(1) 2.86(3)
Rb(2)-Te(5) 3.678(4) Rb(2')-Te(12) 3 .59 1 (14)
Rb(2)-Te(5) 3.678(4) Rb(2')-Te(12) 3.592(14)
Rb(2)-Te(5) 3.678(4) Rb(2')-Te(12) 3 .592(1 4)
Rb(2)-Te(2) 3.907(15) Rb(2)-Te(5) 3.771(10)
Rb(2)-Te(2) 3.907(15) Rb(2')-Te(5) 3.771(10)
Rb(2)-Te(2) 3 .908(1 5) Rb(2)-Te(5) 3.771 Q 0)

 

47

Table 2.13. (Cont’d).

Rb(2)-Te(12) 3.994(15) Rb(2')-Te(1 1) 400(3)
Rb(2)-Te(12) 3.994(15) Rb(2')-Te(1 1) 400(3)
Rb(2)-Te(12) 3.994(15) Rb(2')-Te(l 1) 400(3)
Rb(2)-Rb(2') 090(3) Rb(2)-Te(2) 443(3)
Rb(2)-Sb(2) 290(3) Rb(2)-Sb(2) 379(4)
Rb(3)-Te(9) 3.712(5) Rb(4)-Rb(4") 077(3)
Rb(3)-Te(9) 3.712(4) Rb(4)-Rb(4') 078(3)
Rb(3)-Te(9) 3.712(4) Rb(4)-Sb(3) 285(3)
Rb(3)-Te(6) 3.906(1 1) Rb(4)-Sb(4) 294(4)
Rb(3)-Te(6) 3.906(1 1) Rb(4)-Te(3) 3.598(14)
Rb(3)-Te(6) 3.906(1 1) Rb(4)-Te(7) 3.697(14)
Rb(3)-Te(4) 3.914(1 1) Rb(4)-Te(1 1 3.770(14)
Rb(3)-Te(4) 3.914(1 1) Rb(4)-Te(2) 3.893(16)
Rb(3)-Te(4) 3.914(1 1) Rb(4)-Te(4) 3.921(19)
Rb(4)-Te(8) 3.923(16)
Rb(4)-Te(6) 393(2)
Rb(4)-Te(10 3.965(17)
Rb(4')-Rb(4") 153(4) Rb(4")-Te(10) 356(3)
Rb(4')-Te(12) 357(3) Rb(4")-Te(6) 360(3)
Rb(4')-Te(4) 359(3) Rb(4")-Te(2) 365(2)
Rb(4')-Te(8) 367(2) Rb(4")—Te(7) 366(2)
Rb(4')-Te(1 1) 375(3) Rb(4")—Te(3) 380(3)
Rb(4')-Te(3) 3.76(3) Rb(4")-Te(1 1) 385(3)
Rb(4')-Te(7) 379(3) Rb(4")-Te(1) 401(3)
Rb(4')-Te(5) 403(4) Rb(4")-Te(9) 408(3)
Rb(4')-Te(9) 409(3) Rb(4")-Sb(3) 209(4)
Rb(4')-Sb(3) 362(4) Rb(4")-Sb(4) 370(4)

 

48

Table 2.13. (Cont’d).

Rb(4')-Sb(4) 218(4)
Sb(1)-Te(7) 2.746(8) Sb(2)-Te(3) 2.684(8)
Sb(1)-Te(7) 2.747(8) Sb(2)-Te(3) 2.684(8)
Sb(1)-Te(7) 2.747(8) Sb(2)-Te(3) 2.684(8)
Sb(3)-Te(l) 261(2) Sb(4)-Te(9) 259(2)
Sb(3)-Te(9) 2.658(17) Sb(4)-Te(l) 265(3)
Sb(3)-Te(5) 278(2) Sb(4)-Te(5) 2.733(18)
Sb(3)-Sb(4) 234(3)

Te(1)-Te(2) 3.020(7) Te(7)-Te(8) 3.192(7)
Te(2)-Te(3) 3.159(7) Te(8)-Te(9) 2.990(8)
Te(3)-Te(4) 2.990(7) Te(9)-Te(10) 3.133(6)
Te(4)-Te(5) 3.073(7) Te(10)-Te(] 1) 2.927(6)
Te(5)-Te(6) 3.071(7) Te(l 1)-Te(12) 2.902(60
Te(6)-Te(7) 3.002(7) Te(12)-Te(1) 3.167(7)

 

49

3. Results and Discussions

A. RbUzsbss and KUZSbSes.

Structures. As shown in Figure 2.1, RbUszSg has a two-dimensional character with
layers running perpendicular to the c-axis. The uranium atoms occupy the center of a bi-
capped trigonal prism of S atoms made of two (S2)2‘ units forming the two short parallel
edges of the prism, and four 82' ions at the apex and capping positions. The S-S distance
in the disulﬁde group is normal at 2.096(5)A.26 The formula can therefore be written as
Rb+U4+ZSb3+(Sz')4(S2)2'2, The U4+-centered trigonal prism shares its triangular faces with
neighboring prisms along the a-axis forming one-dimensional columns, which are similar to
those found in ZrSe3,27 see Figure 2.2(a). Then, these columns are connected side by side
sharing S2' atoms on the capping sites to build two-dimensional layers.

The Sb3+ atoms are at the center of a slightly distorted seesaw coordination
environment28 (CN=4), allowing its 582 lone pair to fully express itself, see Figure 2.2(b).
Rb” ions are stabilized in a basically identical environment as the Sb3+ ions. However,
instead of sitting at the center of the seesaw, they move up towards the middle of
interlayer space in order to be coordinated by four more S atoms from the upper layer.
Thus, they are stabilized in 8-coordinate bi-capped trigonal prismatic sites, see Figure
2.2(c).

The difference between the substructure and the superstructure of RbUszSg lies
only in the ordering pattern of the Sb3+ and Rb+ cations. While the substructure shows an
apparent 50/50 statistical disorder between the Rb and Sb ions, the superstructure

resolves this disorder and presents a periodic arrangement of these ions along all the axes.

50

\\\\

@ @y/ ////, ///l
.3.

o .. ‘Ano’i \ff 7151\‘0
0’»: ‘3‘\. 051’ 031/“) (0.}.

2b
@ @ @ «24% S12 73b ”/3

7%

_.
€113

“‘I‘SA‘\\Y.
@

, 3W??- htll'.“ "!..lltlo.

 

Figure 2.1. The structure of RbU2Sng (Cm) viewed down the a-axis.

51

 

 

Figure 2.2. Coordination environment of (a) the U atoms and the one-dimensional
columns along the a-axis, (b) the Sb, and (c) the Rb.

52

The uranium-chalcogen framework in the superstructure remains the same as in the
substructure. Figure 2.3 shows a schematic comparison of the cationic arrangement in the
substructure and the superstructure of RbUZSbS3 .

The structure of KUszSeg (supercell description) is slightly different from that of
RbUszSg in that the Sb3+ and K+ ions are well-ordered only in every other layer and
disordered in the‘remaining layers, see Figure 2.3(c) and Figure 2.4. At this point, it is not
certain if the Sb3+ and K+ ions are really disordered in every other layer or if this
observation is again an artifact caused by the existence of yet a larger supercell that can
remove this disorder. We did indeed observe additional supercell reﬂections which would
double the supercell axis perpendicular to the layer ; a = 8.183(2)A, b = 11.546(3) A, c =
56.42(1)A, (1 = 900°, [3 = 90.0°, y = 900". However, these reﬂections were extremely
weak and we could not collect enough data to precisely probe this additional
superstructure. Increasing the acquisition time per frame to longer than 80 sec did not
improve the quality of the data. Nevertheless, this suggests that the disorder between the

K” and Sb3+ ions is probably a crystallographic artifact.

Properties. These materials are valence-precise and are expected to be semiconductors.
Indeed, RbU2Sng shows an abrupt optical gap at 1.38 eV, see Figure 2.5. The spectrum
also shows intense absorption bands at 0.52 eV, 0.75eV, and 1.02 eV, which can be
associated with the f-f transitions as observed in other compounds with U“ ions.29

The band gap of KUZSbSeg, which appears to be around 0.6 eV, could not be
. deﬁned precisely because the absorption edge is interfered by the f-f transitions, which

exist at the same energy as RbUszSg. The f-f transitions are not usually affected by

53

(a) Substructure (Amm2)
a" at ’x/ 49 Rb

Ziggabos' . Sb

ordered

 

 

Figure 2.3. Schematic comparison of the ordering and disordering patterns of Sb3+ and
Rb+/K+ in (a) the substructure, (b) the superstructure of RbUzsbsg, and (c) the
superstructure of KU2SbSe3.

54

/
ea! a,

     
 
 

  
    
 
 

.//, Sb1 ‘ ”
I. Al . ' .— . Iq‘.‘
C? V . ‘u 2’ . ;, I ‘ , ‘
.‘!.:g“'fv‘2é€im"§ ‘53:!" .46 9.5!! r“
ﬁ’é‘ﬁ [A
0‘“ o .—
@ @
.- : .1
aka“
«922*
e t 4

.952. .m‘;

' - . ‘ — f ’ V ‘ J ;

ewe».

. 22%: "as‘.

c , — -
.d. .d,

a

Figure 2.4. The structure of KUZSbSeg (Cm) viewed down the b-axis.

55

 

  

a/S (Arbitrary Units)

Eg=1.38 eV

 

 

 

IIJJILLIJIIJIIIIIIIIIIIIIJLI

0.6 0.8 l 1.2 1.4 1.6 1.8
Energy (eV)

 

Figure 2.5. Optical absorption spectrum of RbUszSg (—) and KUZSbSeg (---). The

band gap value, E8, is shown in the ﬁgure.

 

 

 

 

a: 400 D
:2 D
.o 350-
0a 0 D
O C] U
o 300- D _ -
Us: 9 250- D D - '
E D I
'g Q) U D D I
3? 200, a 0° ° _ . '
00
g v 150 I I ' I .
§ 100
3
5 50
0 I I I I L I I I I I L I I I I I I I I I I I I I I I I I I
O 50 100 150 200 250 300

Temperature (K)
Figure 2.6. Inverse molar magnetic susceptibility, xM (per uranium), of RbUszSs (El)

and KUszSCs (I).

56

 

479

S-S stretching

\

282

Arbitrary Intensity
236

 

 

lLlLllLllllllllllllllllllllllll_1141_Llll

200 250 300 350 400 450 500 550 600
Raman Shift (cm'l)

 

 

236
252

Se-Se stretching

Arbitrary Intensity
480

 

 

IIIIIIIllIIlllllIlllIllllllIIllllllllll

200 250 300 350 400 450 500 550 600
Raman Shift (cm‘l)

 

Figure 2.7. Raman spectra of (a) RbUzsbsg and (b) KUszSeg.

57

 

external inﬂuences such as changing the ligands from S to Se because the f orbitals are well
shielded by the s and p orbitals.

Magnetic susceptibility measurements show Curie-Weiss behavior between 80K
and 300K with um of 3.03 B.M and 3.21 B.M. and 9 values of 160K and 83K for
RbUZSbsg and KUZSbSeg, respectively, see Figure 2.6. These magnetic moments suggest
sz electron conﬁguration on U4+ . Both crystal ﬁeld splitting and exchange effects act to
reduce the effective momentu"30 below the theoretical value for the free ion (3.58 B.M.)
using the expression of peg = [J(J+l)]”2 (Russell-Saunders coupling applies).31 The
observed um values of these compounds agree well with those reported for US3, USe3,
and UTe3, in which U4+ centers are also stabilized in bicapped trigonal prismatic sites.1
The convex shape of the curves in Figure 6 below 15K for both compounds is probably
due to exchange interactions, which become signiﬁcant at lower temperatures.32

The Raman spectrum of RbUszSg shows shiﬁs at 236 cm], 281 cm'1 and 479
cm’1 and that of KUZSbSeg shows shifts at 236 cm’1 and 252 cm'l, see Figure 2.7. The
shiﬁ at 479 cm’1 for RbUszSg and at 252 cm‘1 for KUZSbSeg are assigned to the
stretching vibration of dichalcogenide groups and these values are in accord with the S-

33 The weak shift appears

S/Se-Se stretching frequencies reported for other compounds.
at 480 cm’1 in Figure 2.7(b) is an overtone. Differential thermal analysis (DTA) showed

that both compounds do not melt below lOOO’C.

B. RbUSbo 331.66.
Structure. The overall structure of RbUSbo33Te6 is shown in Figure 2.8. The

structure is trigonal with one-dimensional character, which is the result of the presence of

58

 

 

Figure 2.8. The structure of RbUSb033T65 as viewed down the c-axis.

59

inﬁnite [UTe6]2' chains along the c-axis. These chains are composed of one-dimensional
rows of U atoms wapped with three inﬁnite zigzag Te chains. These Te chains present a
new way to polymerize Te atoms, which are different from any patterns found in known
Te chains.

An individual Te chain is shown in Figure 2.9(a). The one-dimensional zigzag
chain is composed of alternating bent (Tel and Te3) and linear Te atoms (Te2 and Te4).
The Te-Te distances found in the chain are longer than those found in elemental Te
(2.83A) or in the isolated ditelluride bonds, (Te2)2' of Rb2Te2 (2.78A)32 and of ZrTe
(2.7615034. The helical chain found in elemental Te is shown in Figure 2.9(b) for
comparison. Based on the fundamental VSEPR argument, the oxidation state of the bent

Te atom can be assigned as Te0 and the linear Te atom as Tez', see Scheme 1.

 

bent Te linear Te

Scheme 2.1

Considering that the chains of elemental Te are composed of only bent Teo, the chain
found in RbUSb0_33Te6 can be seen as a reduced version of these chain by ﬂattening every
other bent Te0 atom to a linear Tez' atom. This chain can also be interpreted as a simple

polymerization of Te'1 ions. In either case, the formal charge of the chain is the same as

l/oo(Te(,)6'.

6O

(a) (b)

Te1
3.017(5) \
T92

    

 
    
 
   

’. Te

K 2.8345(8)

103.14(2)
3.081(5)

3.017(5)/

170.3(1) [001]

83.6(1)

 

 

 

 

 

 

 

 

 

Figure 2.9. (a) The one-dimensional zigzag chain found in RbUSbo_33Te6 and (b) the
helical chain found in elemental Te with bond distances (A) and angles (°). Projections

along the c-axis for each case are shown in the box.

61

The coordination environments of U1 and U2 in the column are best described as
distorted tri-capped trigonal prismatic, see Figure 2.10(a). The U1 and U2 trigonal prism
share triangular faces to make one-dimensional columns shown in Figure 2.10(b). The
distortion of the prism is mainly due to the position of the capping Te atoms. Instead of
being exactly in the middle of the rectangular face they are located closer to the apex Te
atoms to form Te-Te bonds, which causes the local environment to bear resemblance to a
pinwheel. Since the distortion occurs in the opposite direction for U1 and U2, U1 can be
considered as a right handed pinwheel and U2 as a left handed one, thus the two would
rotate in the wind in opposing directions. The [UTe6]2' columns can be seen as U4+
centers wrapped with three Te zigzag chains, which are symmetrically related by a 3-fold
axis.

Sb and Rb are stabilized between [UTe6]2‘ columns and serve as bridges between
them. Along the c-axis, Sb and Rb occur in a sequence of -Sb(2)-Sb(1)—Rb(l)-, see Figure
2.11(a). All Sb atoms are stabilized in trigonal pyramidal sites. There also exist three
weaker Sb-Te interactions at the opposite side of the pyramid but their bond distances
are much longer than the bonds forming the trigonal pyramid, see Figure 2.11(b). Rb is
coordinated by 9 Te atoms in tri-capped trigonal prismatic sites, see Figure 2.11(c). The
8b(1)-Sb(2) distance is 2.23(5)A, which is too short to allow both Sb(l) and Sb(2) to be
present at the same time. Also, all Sb atoms initially showed unusually high temperature
factors warranting the occupancy of these sites to be reﬁned. This reﬁnement showed
that all the Sb sites are, on average, 16% occupied; ~16% of the time only Sb( 1) is
present, ~16% of time only Sb(2) is present, and 67% of the time both Sb sites are

empty. Sb(3) and Sb(4) behave the same as Sb(l) and Sb(2) in the neighboring channel.

62

'---
O

I
'0
1
1
1
1

-3 I----
l
l

 

cap Te [001]

 

 

 

t1 .--
I u I
l l
l‘ g
e 1 i :0

 

 

 

 

Figure 2.10. (a) Coordination environment of the U(l) and U(2) atoms and (b) the one-
dimensional UTe6 column composed of tri-capped trigonal prisms sharing triangular

faces. Projection of the column along the c-axis is shown in the box.

63

 

 

Sb2
Te1
Sb1
Te4’ T92” T62
OTeS’
b
( ) 285(1) T94
.— ~~Sb1 [001]
O") O “Orez 8m
3.474(9) {. )

Figure 2.11. The (a) arrangement of Sb and Rb ions along the [001] direction, (b)

coordination environment of the Sb3+ atom, and (c) coordination environment of the Rb+

atom.

64

The sum of all the occupancies, reﬁned without any constraint, adds up to 0.33 equiv.
which is the exact value needed to balance the charge of the compound as
RbIU4+(Sb3+)0.33(Te6)6'. The amount of Sb, as obtained by the x-ray structural analysis,
is consistent with that determined from elemental analysis using the Energy Dispersive
Spectroscopy (EDS) technique.

Even after we adopted a disorder model for the Sb sites, the cationic arrangement
in the channels still shows several problems; the distance between Sb and Rb (2.95 A) is
too short for a cation-cation distance and the Rb atoms show huge U33 anisotropic
atomic displacement parameters. The occupancy for Rb was therefore reﬁned, but did
not result in a decrease of the occupancy nor did it affect the temperature factor. It was
suspected that Rb may displace differently when Sb is present than when it is not, which
could result in positional ﬂuctuation of the Rb atoms along the c-axis. If the presence of
the Sb atoms occurs in a systematic manner, this should result in ordering of Rb along the

channel and may cause a superstructure.

Superstrucure. Indeed, superstructure reﬁnement provides a better understanding
of the cationic arrangement in the channels. There exist two types of channels with
different arrangment of Rb and Sb. In channel type 1, Sb is no longer disordered, see
Figure 2.13(a). There exist two independent Sb atoms, each with occupancy of 50%.
The Rb(1) and Rb(2) positions are resolved into two sites in the superstructure. The
positional disorder of the Rb sites is forced by the partial occupancy of the Sb atoms as
the Rb atom attempts to adjust its positions depending on whether Sb is present or not.

When Sb(l) is absent, Rb can be stabilized in its ideal Rb(1) site. When Sb(l) is present,

65

 

 

 

a) nu .Sb OTe @Rb

Figure 2.12. The superstructure of RbUSbo_33Te6 as viewed down the c-axis.

66

(a) Channel |

@ Rb3 (100%)

(Kip/$13 (50%)

Rb1 (50%)
Rb1’ (50%)

852' (50%)
852 (50%)

54$"? ‘5°°/°’

@ Rb3 (100%)

 

Rb N (N = 1,2,3,4)

(b) Channel II

03 (18%)

3

$54 (15%)

Rb4’ (18%)
Rb4 (87%)
Rb4” 15%
$03

Rb N’/N” (N = 1,2,4)

Figure 2.13. The arrangement of Sb and Rb ions along the c-axis in the superstructure
(a) channel type I and (b) channel type II. (c) The local environments of Rb when Sb is

around (left) and not (right).

the distance between Rb(1) and Sb(l) gets too close (2.66(2)A) and Rb(1) has to move
1.03(2)A up and take the Rb(l') position to allow enough space for Sb(l). The distance
between Rb(l ') and Sb(l) then becomes a reasonable 3.69(2)A, which is reasonable. The
situation for Sb(2), Rb(2), and Rb(2') is the same. Only the position of Rb(3) in the
channel, which is 5.2A apart from both Sb(l) and Sb(2), is not affected by the presence of
Sb ions and is fully occupied in a single position without any disorder. Between Rb(2)
and Rb(1) along the channel, there exists an "Sb vacancy" compared with the Sb array in
the substructure. Therefore, the average Sb occupancy in channel I is 33%
[1/3{50%(Sb1) + 48%(Sb2) + 0%(for the vacancy)}].

In channel II, there exist three independent Sb atoms, all of which are disordered
over two sites as the same manner in the substructure. The three independent Rb atoms
between these disordered Sb are triply disordered again due to the partial occupancy of
the disordered Sb, see Figure 2.13(b). For example for 66% of time when both Sb(3) and
Sb(4) are absent, the Rb(4) position is occupied. For 18% of time when Sb(3) is present,
Rb(4) moves down to the position of Rb(4") to achieve an optimum Rb-Sb distance of
3.88(7) A. For 15% of time when Sb(4) is present, Rb(4) is pushed up to the Rb(4')
position and the distance between Sb(4) and Rb(4') becomes 3.94(7)A. The average
occupancy of Sb in the channel 11 is again ~33% [18%(Sb3) +15%(Sb4)]. The local
environments of Rh when Sb is present and absent are compared in Figure 2.13(c). When
Sb is absent, the Rb ions (Rb(1) through Rb(4)) are 9-coordinate in tri-capped trigonal
prismatic sites with Rb sitting on the same plane with three capping Te atoms. When Sb
is present, Rb sits between the trigonal face formed by three capping Te and one

triangular face of the prism (top or bottom face of the prism depending on the direction

68

from which Sb approaches). This environment is best described as 6-coordinate trigonal
anti-prismatic. Judging from the orientation of the lone pair of Sb and the Rb-Sb
distances, the Sb can be considered to be coordinated to Rb using its lone pair, see Figure
2.13(c).

We cannot exclude a possibility that there may exist a even bigger superstructure
that can lift the disorder of Sb and Rb. Figure 2.14 shows possible well ordered Rb/Sb
arrangements, which would give rise to a super-super structure if they occur in a long
range order. However, they could be averaged out to appear disordered as observed in
channel I and II if no long range periodicity exists.

The superstructure reﬁnement revealed another interesting phenomenon, the
modulation of Te-Te distances in the chain. Instead of having almost equal Te-Te
distances, the inﬁnite chain now separates into dimers, trimers, and pentamers. The Te-
Te distances within these oligomers are shorter (< 3.07A) than the Te-Te distances
between these oligomers (> 3.13A), showing a distinct gap in the Te-Te bonds. The
modulated pattern in the chain is shown in Figure 2.15(a). The three chains around the
uranium center possess the same modulation patterns but these patterns are translated 1/3
of the repeating unit along the c-axis chain by chain. The charge of these oligomers can be
formally assigned as (Te2)2', (Te3)2' and (Te5)6' based on the VSEPR theory as explained
above, see Figure 2.15(b). However, the distances in these oligomers are still longer than a
normal Te-Te bond found in the discrete ditelluride,”33 suggesting that there are
substantial inter-oligomer interactions.

The existence of Te-Te modulations in inﬁnite Te chains of RbUSb0_33Te6 is seems

to be unique as we are not aware of similar 1D Te distortions in the literature. For

69

Channel I

 

 

8 R53 (100%)

%(50%)

R51 (50%)
Rb1’ (50%)

Rb2’ (50%)
R52 (50%)

,8. $52 (50%)

 

 

Averaged

Channel ll

 

 

$53 (18%)
$54 (15%)

%RM'( (18%)
67° )
Rb4”( (15° )

 

 

Averaged

 

 

@ R53

@ R51

@ R52

 

 

50%

 

 

@ R54

@ R54

 

 

67%

models observed in channel I and channel 11.

70

 

 

Q R53

 

 

 

 

545%

@ R54'

ma

@ Rb4’

 

 

18%

Q Rb1’
@ Rb2’
Sb2
50%

h 354

@R54

+ (ks/0

554

@R54'

15% '

 

 

 

Figure 2.14. Possible Sb-Rb ordering patterns that may average out as disordered

 

3.020(7)
3 159 7 E.k/OTel Te 0 T69
' ( )\ T 2 Te T610

9
T53 / 2.990(7) T97 MCQ
Te4 / 3.073(7) T68 T912

Te
\

 
    

f
52)

\ 3.071(7) T510 T52

Te7 3.002(7) T51 1 T53
I
3392(7) T930\‘E)2.990(8) T512 T54
T99 T91 T95
2.927(6) \ 3.133 6
‘ T610 ( ) T92 T66
T511 T53 T97
2.902(6)” Tel?- T94 T58

3.1670? T51 T5 T59
T62 T65
(b) O\,

[Tezlz- ”9312' [Tesle-

0
..g
9.
O

>

Figure 2.15. (a) The Te-Te modulations in the Te chains in the superstructure. (b) The

dimer, trimer, and pentamer found in the chains.

71

example, no modulations have been reported for the Te chains of elemental Te, CuTeZX,
and CuTeX (X=Cl, Br, I). The major difference between these chains and those of
RbUSb033Te6 lies in the formal charge of T6 and the Te-Te distances associated with the
chains. The Te chains found in the elemental Te, CuTeZX, and CuTeX are neutral and
they are composed of normal single Te-Te bonds (~2.8A). Therefore, each neutral T6
with 6 valent electons forms two single covalent Te-Te bonds. These Te atoms satisfy
the Octect rule, which makes the Te chains stable and distortion free. However, the Te
chains found in the substructure of RbUSb033Te6 has negative charges, (Te6)6' and judging
from the signiﬁcantly elongated Te-Te distances, the electrons are not localized in speciﬁc
bonds but delocalized across the chain. In this case, the p band of Te' ions are not
completely ﬁlled but has a hole per Te atom. Therefore, the chain is susceptible to
distort because the total energy of the system can be lowered by modulating the Te-Te
distances. The chain can distort to oligomers, so that the Te-Te distances in the oligomers
are much shortened (near to the normal single bond) and the distance between the
oligomers are signiﬁcantly lengthened. Such distortion can be easily understood as a
version of one dimensional Peierls distortion, which results in a lowering of the ﬁlled
orbital(LUMO) energy and opening a HOMO-LUMO gap.35 The same phenomena has
been often observed in the two-domensional Te nets that possess Te anions, formal

charge of which ranges between -2 and 0.36

Properties. RbUSb0_33Te6 is a narrow gap semiconductor with a band gap smaller than

0.056V. The charge transport properties were measured on single crystal samples. The

electrical conductivity increases with increasing teperature, indicating semiconducting

72

behavior with room temperature conductivity of ~ 30 S/cm, see Figure 2.16. The

thermopower data is shown in Figure 2.17 and gives a room temperature Seebeck
coefﬁcient of ~ 210 uV/K. The positive sign indicates that charge carriers are holes and,
therefore, the material is best described as a p-type semiconductor.

Magnetic susceptibility measurements show Curie-Weiss behavior between 20 K
and 300 K with 11ch = 3.04, B.M. and 0: -6.3 K, see Figure 2.18. The magnetic moment
suggests Sty electron conﬁguration on U4+ . Both crystal ﬁeld splitting and exchange
effects act to reduce the effective moment”29 below the theoretical value for the free ion
(3.58 B.M.).30 The observed ucff value falls in the same range of values for other uranium

"329 Below

(IV) Chalcogenides, where U“ are stabilized in the similar local environment.
20 K, the plot shows deviation from the ideal Curie-Weiss behavior due to pronounced
exchange interactions.

The Raman spectrum of RbUSbo,33Te6 shows shifts at 142 cm", 125 cm", 94 cm'
', and 90 cm", see Figure 2.19. These shifts seem to be related with Te-Te stretching
vibrations because the values are very similar to those observed for elemental Te (144 cm'
', 126 cm], and 95 cm"). The Sb-Te stretching can also appear in this region because Sb
and Te have similar atomic masses and Sb-Te and Te-Te bond distances are similar.
Considering the number of Te-Te bonds is greater than the number of Sb-Te bonds, these

shifts are mainly due to Te-Te stretching. Differential thermal analysis (DTA) showed

that the compound does not melt under 1000°C.

73

 

400
350
300
250

WWH

200

S(uV/K)

150
100

50

 

 

 

0 L I I I I I 1 LJ

1 . n l n n n n n
100 150 200 250 300
Temperature (K)

Figure 2.16. Variable temperature thermopower for a single crystal of RbUSbo,33Te6.

 

 

101
10'1
’5‘ 10'3
2
U)
‘6 10‘5
O
-7 O
10 5

10'9 IIIlIIIIlIJIIILLIIlIIIIIIJIIIIII

0 50 100 150 200 250 300 350
Temperature (K)

Figure 2.17. Variable temperature electrical conductivity for a single crystal of
RbUSbo33TC6.

74

 

300

250

200

150

(mol/emu)

100

Inverse Magnetic Susceptiblity

 

 

0 50 100 150 200 250 300
Temperature (K)

Figure 2.18. Inverse molar magnetic susceptibility, XM (per uranium), of RbUSbo,33Te(,.

 

125

142

 

“’8

Arbitrary Intensity

' LA

 

80 120 160 200 240 280 320 360 400
Raman Shift (cm‘l)

 

 

Figure 2.19. Raman spectrum of RbUSb033Teé.

75

4. Conclusions

By changing chalcogen types in the polychalcogenide ﬂuxes, we were able to
stabilize two different quaternary structure types with uranium metal, AUZSbQ8(A =
Rb, K; Q = S, Se) and RbUSb0,33Te(,. These structure types are all noncentrosymmetric
and proved that incorporating asymmetric Sb building units into the uranium chalcogenide
framework can indeed increase the probability of forming noncentrosymmetric structures.
The AUZSng structure type can potentially be a excellent scaffold for generating NLO
materials. Unfortunately, the presence of f-f transitions of uranium eliminate a
transparent spectral window in the IR region and limit the possibility that these
compounds can be used for nonlinear optics (NLO). However, if the uranium metal could
be properly substituted with other M4+ ions, which do not have f or d electrons such as
Zr4+ and Hf“, the resulting compounds are expected to be optically transparent below
the band-gap.

The discovery of RbUSb0.33Te6 features a novel way to polymerize Te'l anions.
The one dimensional Te chain exhibits interesting modulations of Te-Te distances, which
result in a 9-fold (\/3a x \l3b x 3c) superstructure. The driving force of this distortion
seems to be the further stabilization of the system by lowering the LUMO energy and
opening a HOMO-LUMO gap. This agrees well with the observed semiconducting
behavior of the material. The infrequent presence of Sb and their subtle ordering in the
channel are other components that complicate the structure and signiﬁcantly enlarge the

unit cell.

76

10

11.

12.

l3.

14.

References

Freeman, A. J .; Darby Jr., J. B. eds, “The Actinides: Electronic Structure and
Related Properties”, Vol I, II, 1974, Academic Press; New York .

Noel, H.; Troc, R. J. Solid State Chem. 1979, 27, 123-135.

Noél, H. J. Less Common Met. 1980, 72, 45-49.

Daoudi, A.; Noel, H. Inorg. Chim. Acta. 1987, 140, 93-95.

Daoudi, A.; Noel, H. J. Less Common Met. 1986, 115, 253-259.

Cody, J. A.; Ibers, J. A. Inorg. Chem. 1995, 34, 3165-3172.

Cody, J. A.; Mansuetto, M. F .; Pell, M. A.; Chien, S.; Ibers, J. A. J. Alloys
Compd. 1995, 219, 59-62.

Sutorik, A. C.; Albritton-Thomas, J; Hogan, T.; Kannewurf, C. R.; Kanatzidis, M.
G. Chem. Mater. 1996, 8, 751-761.

Chondroudis, K.; Kanatzidis, M. G. J. Am. Chem. Soc. 1997, 119, 2574-2575.
Chondroudis, K.; Kanatzidis, M. G. C. R. Acad. Sci.,Ser. B 1996, 322, 887-894.
Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G., Inorg. Chem.
1997, 36, 3804-3805.

Choi, K.-S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309.
Hanko, J. A.; Kanatzidis, M. G. J. Chem. Soc. Chem Commun. 1998, 725.

(a) West, A. R. Solid State Chemistry and Its Applications, 1984, John Wiley &
Sons; New York, pp540-552. (b) Xia, Y.; Chen, C.; Tang, D.; Wu, B. Adv. Mater.

1995, 7, No.1, 79-81.

77

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Adenis, C.; Langer, V.; Lindqvist, 0. Acta Cryst. 1989, 45C, 941-942.

(a) Fenner, Von J .; Schulz, H. Acta Cryst. 1979, B35, 307-311. (b) Carkner, P. M.;
Haendler, H. M. J. Solid State Chem. 1976, 18, 183-189. (c) Milius, W. Z. anorg.
allg. Chem. 1990, 586, 175-184.

(a) Milius, W. Z Naturforsch. 1989, 44b, 990-992. (b) Fenner, J. Acta Cryst 1976,
B32, 3084-3086.

McCarthy, T. J.; Ngeyi, S.-P.; Liao, J. -H.; Degroot, D.; Hogan, T.; Kannewurf,
C. R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331.

P. W. Selwood, Magnetochemistry, 2nd ed., 1956,1nterscience Publishers; New
York.

Lyding, J. W.; Marcy, H. 0.; Marks, T. J .; Kannewurf, C. R. IEEE Trans.
Instrum. Meas., 1988, 37, 76-80.

Chaikin, P. 1.; Kawk, J. F. Rev. Sci. Instrum, 1975, 46, 218-220.

SMART: 1994, Siemens Analytical Xray Systems, Inc., Madison, Winsconsin
53719 USA.

SAINT: Version 4, 1994-1996, Siemens Analytical Xray Systems, Inc., Madison,
Winsconsin 53719 USA.

Sheldrick, G. M. University of Gottingen, Germany, to be published.
SHELXTL: Version 5, Sheldrick, G. M. Siemens Analytical Xray Systems, Inc.,
Madison, WI 53719, USA, 1994.

Furuseth, S.; Brattas, L.; Kjekshus, A. Acta Chem. Scand. 1975, A29, 623-631.

Kronert, Von W.; Plidth, K. Z anorg. allg. Chem. 1965, 336. 207-218.

78

28.

29.

30.

31.

32.

33.

34.

35.

36.

A seesaw coordination is deﬁned as one derived from an octahedral coordination in
which two cis-equatorial ligands have been removed.

(a) Gronvold, F .; Drowart, J .; Westrum, E. F. Jr. The Chemical Thermodynamics
of Actinide Elements And Compounds, 1984, International Atomic Energy Agency;
Vienna, pp161-200 and the references therein. (b) Clifton, J. R.; Gruen, D. M.;
Ron, A. J Chem. Phys. 1969, 51, 224-232. (c) Schoenes, J. Phys. Rep. 1980, 63,
301-336.

Dawson, J. K. J. Chem. Soc. 1951, 429-431.

Drago, R. S. Physical Methods for Chemists, 2nd ed., 1992, Saunders College
Publishing; New York, pp480-486.

(a) Suski, W. J. Solid State Chem. 1973, 7, 385-399. (b) Suski, W. Phys. Stat. Sol.
1972, 13, 675-680.

(a) Biittcher, P.; Getzschmann, J .; Keller, R. Z. anorg. allg. Chem. 1993, 619,
476-488.

Furuseth, S.; Brattas, L.; Kjekshus, A. Acta Chem. Scand. 1975, A29, 623-631.
Burdett, J. K. Chemical Bonding in Solids, 1995, Oxford University Press; New
York, pp 48-52.

(a) Stowe, K. J. Solid State Chem. 2000, 149, 155-166. (b) Foran, B.; Lee, S.;
Aronson, M. C. Chem. Mater. 1993, 5, 974-978. (c) Gourdon, O; Hanko, J. A.;
Boucher, F.; Perticek, V.; Whangbo, M.-H.; Kanatzidis, M. G.; Evain, M. Inorg.

Chem. 2000, 39, 1398-1409.

79

Chapter 3. Reactions of Thorium Metal with Sb in Polyselenide

Fluxes.

A New Quaternary Thorium Selenoantimonate, K T thzSeg Featuring

One-Dimensional Double Th Chains Constructed By Dichalcogenide

Groups.

80

1. Introduction

While diverse uranium Chalcogenides have been prepared and studied for their
interesting magnetic properties,l the number of known thorium chalcogenide compounds
is relatively small probably because these materials are diamagnetic with Th4+ centers and
were considered less interesting from a property view point.2 Up to date, only a handful
of binary (i.e. ThSeZ,3 ThSe3,4 ThZSe3,5 T112855,“ Th7Se1 23) and a few ternary thorium
chalcogenide compounds(i.e. AThZSeé,7 KThzTe68 , CSTh2T€6,9 SrThZSeS'O) were
discovered.

From a structural chemistry perspective, however, Th compounds can be very
interesting because they may adopt unique structure types due to the largest ionic radius
of Th4+ metal among few tetra-valent metal ions in a chalcogenide matrix (i.e.
Ti4”(0.605A), Zr4+(0.72A), Hf‘+(0.71 A), U4+(0.89A)).“ In simple binary systems,
these metals with 4+ oxidation state usually form isostructural compounds (i.e. MS; (M
= Ti, Zr, Hf, Th)4’12 without structural variation. However, these metals may produce
compounds with differing structures when the polychalcogenide ﬂuxes are used to
synthesize more complicated ternary or quaternary compounds. The products obtained
by this method can easily vary by the slight difference in the ionic size or
electronegitivity of the reacting metal species because this synthetic method allows the
system to choose and form the most stable compound at any given condition without
forcing a speciﬁc composition or structure upon the system.

Indeed, the reaction of the Th metal with Sb in the polyselenide ﬂuxes provided a

novel quaternary compound KTthzse6 instead of forming an isostructural compound of

81

AUZSbSes (A=Rb, K; Q=S, Se) , which was obtained from the reaction of U metal with
Sb in the same synthetic conditions. KTth2866 presents a new three-dimensional
structure with a unique one dimensional double Th chain construct by di-selenide bonds.

Here, we report the synthesis, structure, and physico-chemical properties of KTthZSeG.

2. Experimental Section

2.1 Synthesis.
Reagents. The following reagents were used as obtained: thorium, -100 mesh, 99.8%,
Cerac, Milwaukee, WI; antimony, 99.999%, -200 mesh, Cerac, Milwaukee, WI; selenium
powder, 99.5+%, 100 mesh, Aldrich, Milwaukee, WI; potassium metal, analytical
reagent, Aldrich, Milwaukee, WI. The starting material Kzse was prepared by a
stoichiometric reaction of potassium/rubidium metal and selenium in liq. NH3.
KTthZSe6. Single crystals were prepared from a mixture of 0.0464 g (0.2
mmol) Th, 0.1922 g (0.4 mmol) SbZSe3, 0.0628 g (0.4 mmol) K2Se, and 0.1579 g (2 mmol)
Se. The reagents were thoroughly mixed, ﬂame-sealed in an evacuated silica tube, and
heated at 540 °C for 5 days (cooling 3 °C/h). Black needle-like crystals of KTthZSe6
were obtained by isolation in degassed dimethylformamide (DMF) and water (yield ~
20% based on Sb metal). The crystals are air- and water stable. Pure material was
obtained by heating Th/KZSe/SbZSeg/Se in a stoichiometric ratio in a quartz tube at 650 °C
for 4 days.

The compositions of the material was analyzed by Scanning Electron Microscope

(SEM)/ Energy Dispersive Spectroscopy (EDS). The homogeneity of the product was

82

conﬁrmed by comparing their powder X-ray diffraction patterns against ones calculated

using X-ray single crystal data, see Table 3.1.

2.2 Physical Measurements.

Solid-State UV/V is Spectroscopy. Optical diffuse reﬂectance measurements were
performed at room temperature with a Shimadzu UV-3101 PC double-beam, double-
monochromator spectrophotometer operating in the 200-2500 nm region. The instrument
is equipped with an integrating sphere and controlled by a personal computer. BaSO4
was used as a 100% reﬂectance standard for all materials. Samples were prepared by
grinding them to a ﬁne powder and spreading them on a compacted surface of the
powdered standard material and preloaded them into a sample holder. The reﬂectance
versus wavelength data were used to estimate a material's band gap by converting

reﬂectance to absorption data as described previously.13

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically, a sample (~20 mg )
of ground crystalline material was sealed in a silica tube under vacuum. A silica tube of
equal mass ﬁlled with A1203 was sealed and placed on the reference side of the detector.
The samples were heated to 800 °C at 5 °C/min, isothermed for 10 min followed by
cooling at -5 °C/min to 50 °C. Residues of the DTA experiments were examined by X-ray
powder diffraction. The stability/reproducibility of the samples was monitored by

running multiple heating/cooling cycles.

83

Table 3.1. Calculated and Observed X-ray Powder Pattern of KTthZSe6.

 

 

h k l dca_lg(A) Dobs(A) I/Imiix(0ble/°)
0 1 1 11.314 11.323 13.59
0 2 0 7.588 7.583 22.99
0 1 2 7.409 7.418 60.59
0 2 1 6.927 6.914 51.87
0 2 2 5.657 5.651 15.74
0 1 4 4.088 4.086 13.82
0 3 3 3.772 3.770 20.50
1 2 1 3.606 3.614 10.31
0 4 2 3.464 3.456 37.71
0 3 4 3.252 3.253 17.78
0 4 3 3.151 3.144 11.44
1 3 -2 3.074 3.074 11.10
1 3 2 3.014 3.016 28.77
0 5 1 2.987 2.978 12.68
1 1 4 2.898 2.900 17.10
0 5 2 3.858 2.849 100.00
1 3 3 2.790 2.787 11.66
1 2 4 2.751 2.752 14.72
0 2 6 2.651 2.654 19.03
0 6 0 2.529 2.522 11.66
1 5 -1 2.455 2.454 12.91
1 5 -2 2.389 2.389 11.78
0 2 7 2.310 2.309 32.05
0 5 5 2.263 2.261 16.42
1 2 6 2.216 2.212 11.44
1 4 5 2.149 2.146 15.40
1 3 6 2.107 2.105 11.44
0 2 8 2.044 2.044 13.02
1 4-6 2.029 2.028 12.23
0 2 9 1.831 1.830 19.59
2 5 2 1.699 1.700 13.48
2 6-2 1.610 1.609 10.31

 

84

2.3 X-ray Crystallography.

A single crystal with dimensions 0.5 x 0.04 x 0.04 mm was mounted on the tip of
a glass ﬁber and intensity data were collected on a Bruker SMART Platform CCD
diffractometer using graphite monochromatized Mo Ka radiation. The data were
collected at room temperature over a full sphere of reciprocal space, up to 25.07° in G.
The individual frames were measured with an omega rotation of 0.3 deg and an acquisition
time of 60 sec. The SMART'4 software was used for the data acquisition and SAINT15
for the data extraction and reduction. The structure was solved by direct methods using
SHELXS-86 package16 of crystallographic programs and full matrix least square
reﬁnement was performed using TEXSAN software package.17 The structure of
KTth28e6 was solved and reﬁned successfully in the P21/c space group, which was
uniquely deﬁned by the systematic absence conditions of the data set. The complete data
collection parameters and details of the structure solution and reﬁnement for each
compound are given in Table 3.2. The coordinates of all atoms, isotropic temperature
factors, and their estimated standard deviations (esd’s) for each compound are given in

Tables3.3. The selected bond distances and bond angles are given in Tables 3.4.

3. Results and Discussion

KTthZSe6 has a three dimensional tunnel framework with tetravalent Th centers,
see Figure 3.1. The K“ ﬁlled tunnels run parallel to the a-axis. The coordination

geometry around the Th is best described as a tri-capped trigonal prism formed by 9

85

Table 3.2. Summary of Crystallographic Data and Structural Analysis for KTth2Se6.

 

 

 

Formula KTth28e6
Formula weight 988.40

Crystal habit Black—silver needle
Space group P21/c

a, A 4.2676(1)

b, A 15.1746(4)

c, A 16.9909(4)

B, deg 92.2170)

Z; V, A3 4, 1099.49(4)
Dcalc, g/6m3 5.971

Temp, K 293(2)

MMo K00. A 0.71073

“(M0 KOL), cm"l 386.17

F(000) 1660

9 max , deg 25.07

Total data 5557

Unique data 2011 [Rim = 0.051]
No. of observations (I>30') 15 30

No. of variables 91

R‘IRw" 0.045/0.050

GOF 1.67

" R= EllFoI — IFcl /2IFoI ; b Rw= {2 w(|F.|—|‘F.|)2/2Mary”

 

86

Table 3.3. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Beq) Values for KTthzSe6 with Estimated Standard Deviations in

 

 

Parentheses.
Atom x y z Beq, a A2
Th 0.3006(2) 0.62954(5) 0.40859(4) 106(2)
Sb(l) 0.7289(3) 0.87293(8) 0.47462(8) 1.10(3)
Sb(2) 0.7522(3) 0.52627(8) 0.20321(8) 102(3)
Se(l) 0.2759(4) 0.8288(1) 0.3727(1) 104(4)
Se(2) 0.7893(4) 0.7069(1) 0.5168(1) 074(4)
Se(3) 0.6816(5) 0.0846(1) 0.4072(1) 156(4)
8e(4) 0.7877(4) 0.6730(1) 0.2870(1) 082(4)
Se(5) 0.3354(4) 0.4688(1) 0.3015(1) l.36(4)
Se(6) 0.1791(4) 0.4863(1) 0.4415(1) 092(4)
K -0.736(l) 0.7878(3) 0.1733(3) 1.9(1)

 

aAnisotropically reﬁned atoms are given in the form of the isotropic equivalent
displacement parameter deﬁned as Beq= (87t2/3)[a2311 + b2322 + 02833 + ab(cosy)B12 +

ac(cosB)Bl3 + bc(cosoc)Bz3].

87

Table 3.4. Selected Distances(A) and Bond Angles(°) for KTthZSe6.

 

Bond Distances

 

 

Th-Se(1) 3.086(2) 8b(1)-8e(1) 3.035(2)
Th-Se(2) 2.968(2) 8b(1)-8e(1) 2.632(2)
Th-Se(2) 3.135(2) 8b(1)-8e(2) 2.629(2)
Th-Se(4) 3.059(2) 8b(1)-8e(3) 3.413(2)
Th-Se(4) 3.024(2) 8b(1)-8e(3) 3.223(2)
Th-Se(5) 3 049(2) 8b(1)-8e(3) 2.791 (2)
Th-Se(6) 3.154(2) Sb(2)-8e(1) 3.262(2)
Th-Se(6) 3.142(2) Sb(2)-8e(3) 3.239(2)
Th-Se(6) 3 052(2) Sb(2)-8e(3) 2.731(2)
Sb(2)-8e(4) 2.644(2)
8e(6)-8e(6) 2.494(4) Sb(2)-Se(5) 3.070(2)
Sb(2)-Se(5) 2.636(2)
Bond Angles
Se(l)-Th-Se(2) 7432(5) 8e(1)-Sb(] )-Se(2) 8985(7)
Se(1)-Th-Se(6) 134.46(5) Se(l)-Sb(1)-Se(3) 175.42(7)
Se(2)-Th-Se(4) 144.05(5) Se(l)-Sb(1)-Se(3) 8919(6)
Se(2)-Th-Se(6) 71 .85(5) Se(2)-Sb(1)-Se(3) 9475(7)
Se(4)-Th-Se(5) 73.72(5) Se(l)—Sb(2)-Se(3) 9223(6)
Se(4)-Th-Se(6) 78.91 (5) Se( 1 )-Sb(2)-Se(4) 170.66(7)
Se(5)-Th-Se(6) 92.81(5) Se(l)-Sb(2)-Se(5) 8606(6)
Th-Se(6)-Th 132.64(6) Se(3)-Sb(2)-Se(4) 9306(6)
Th-Se(6)-Th 8709(5)

 

88

 

 

Figure 3.1. The structure of KTth28e6 viewed down the a-axis. The atomic labeling

scheme is shown in the inset.

89

selenium atoms, see Figure 3.2. Three of the selenium atoms bound to Th belong to di-

selenide groups, Sezz. Therefore, the formula can be described as K+Th4+Sb23+(Sez'
)5(Se2)2'0.5. The Th4+-centered trigonal prism shares its triangular faces with neighboring

prisms along the a-axis forming one-dimensional columns. Such two columns are joined to
form one-dimensional inﬁnite double chains parallel to the [100] direction as shown in
Figure 3.2. The di-selenide ligands play an important role in the construction of the
double chains by bridging single chains. Each Sezz’ is simultaneously connected to four
Th atoms in an arrangement unusual for a di-chalcogenide. This extensive bonding gives
rise to a longer Se-Se distance of 2.494(4)A than normal of 2.34A18 The Th double chains
are separated by bridging [Sb4Se10]n blocks.

The Sb3+ ion occupies two different sites in this structure. Both of them have
three short Sb-Se bonds ranging from 2.629(2)A to 2.791(2)A and three longer bonds
ranging from 3.03 5(2) to 3.413(2)A, indicating that the coordination geometry of Sb3+ is
intermediate between trigonal pyramidal and octahedral, see Figure 3.3. However,
because the longest distance is still shorter than the sum of the Sb-Se van der Waals radii
(3.510131)19 indicating a considerable interaction between Sb and Se, the local symmetry of
Sb3+ is regarded as distorted octahedral. The gross distortion of SbSe6 octahedra is due to
the stereochemically active lone pair of Sb3+ ions. The SbSe6 octahedra share edges with
neighboring octahedra to build [Sb4Sem]n blocks.

The K+ ions are stabilized in the one-dimensional channel parallel to the a-axis.
also coordinated by 9 selenium atoms in tri-capped trigonal prismatic sites with an

average (K-Se) distance of 3.448A.

90

Set

 

 

 

 

 

 

Figure 3.2. The double chains of Th atoms running [100] direction.

Each single chain is composed of ThSe6 prisms sharing opposite triangular faces.

91

 

   

2.644
895 $95
. sag , -
at)
l
Se3 27:“ : 3'239 Se3
{3.262
891

Figure 3.3. The local environments of the Sb“ ions.

92

The compound is valence-precise and is expected to be semiconductors. The
absorption spectrum conﬁrms this by showing the presence of abrupt optical gaps. It
suggests the presence of two band gaps, Eg, at 0.95 and 0.80eV for KTth2$e6., see
Figure 3.4. Differential thermal analysis (DTA) experiments show that KTthZSeé melts
incongruently at 659°C decomposing to ThSe2 and unidentiﬁed K/Sb/Se ternary

compounds (according to the powder X-ray diffraction pattern after DTA).

 

 
   
   

 

 

 

3.5_
3:-
Lu? E
a 2.5-
:5 E
E 2 '-
.4-2 =
.o 15:
i ' " Eg=0.956V
m
3 1-
0'5 ' Eg=0.80eV
0 ......... Llama n .........
05 1 15 2

Energy (eV)

Figure 3.4. Optical absorption spectrum of KTthZSeG. The band gap value, Eg, is

shown in the ﬁgure.

93

4. Conclusions

The alkali metal polychalcogenide ﬂux method has proven to be an extremely
useful low-temperature route to access novel compounds. The reaction of Th in the
antimony polyselenide ﬂuxes resulted in a new quaternary thorium selenoantimonate,
KTthZSe6. The multiple bonded 8e22' groups are key to the formation of this structure-
type. A related study20 shows that this structure type can also be stabilized with Ln3th
centers when alkali metal is replaced with alkaline earth metals (i.e. BaLaBi2Q6 (Q=S,
Se)). The similar sizes between K” and Ba” and Th4+ and La3+ and the retention of
isoelectronic relationship between the K+/La3+ and Ba:!+/La3+ pairs explain the

isostructural nature of these compounds.

94

10.

11.

12.

13.

14.

References

(a) Noel, H.; Troc, R. J. Solid State Chem. 1979, 27, 123-135. (b) Noel, H. J. Less
Common Met. 1980, 72, 45-49. (c) Daoudi, A.; Noél, H. Inorg. Chim. Acta.
1987, 140, 93-95. ((1) Daoudi, A.; Noél, H. J. Less Common Met. 1986, 115,
253-259 and references therein.

Freeman, A. J .; Darby Jr., J. B. eds, “The Actinides: Electronic Structure and
Related Properties”, Vol I, II, 1974, Academic Press; New York.

D’Eye, R. W. M. J. Chem. Soc. 1953, 1670-1672.

Noel, H. J. Inorg. Nucl. Chem. 1980, 42, 1715-1717.

D’Eye, R. W. M.; Sellman, P. G.; Murray, J. R. J. Chem. Soc. 1952, 2555—2562.
Graham, J .; McTaggart, F. K. Austrian Jounal of Chem. 1960, 13, 67-73.

Choi, K.-S ; Patschke, R.; Billinge, S. J. L.; Waner, M. J .; Dantus, M; Kanatzidis,
M. G. J. Am. Chem. Soc. 1998, 120, 10706-10714.

Wu, E. J.; Pell, M. A.; Ibers, J. A., J. Alloys and Compd. 1997, 255, 106-109.
Cody, J. A.; Ibers, J. A.; Inorg. Chem. 1996, 16, 3273-3277.

Narducci, A. A. ; Ibers, J. A.; Inorg. Chem. 1998, 37, 3798-3801.

Shannon, R. D. Acta Cryst. 1976, A32, 751-767.

Furuseth, S.; Brattas, L.; Kjekshus, A. Acta Chem. Scand. 1975, A29, 623-631.
McCarthy, T. J.; Ngeyi, S.-P.; Liao, J. —H.; Degroot, D.; Hogan, T.; Kannewurf,
C. R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331.

SMART: 1994, Siemens Analytical Xray Systems, Inc., Madison, Winsconsin

53719 USA.

95

15.

16.

17.

18.

19.

20.

SAINT: Version 4, 1994-1996, Siemens Analytical Xray Systems, Inc., Madison,
Winsconsin 53719 USA.

Sheldrick, G. M., in Crystallographic Computing 3; Sheldrick, G. M.; Kruger, C.;
Doddard, R., Eds; Oxford University Press: Oxford, England, 1985, 175.
Gilmore, G. J.,Appl. Cryst. 1984, 17, 42.

Kanatzidis, M. G.; Huang, S.-P. Coord. Chem. Rev. 1994, 130, 509.

Pauling, L. “The Nature of the Chemical Bond”, 3rd ed.;Cornell University Press:
Ithaca, NY, 1966; p260.

Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G., Inorg. Chem.

1997, 36, 3804-3805.

96

Chapter 4. Reactions of Rare Earth Metals with Sb in Polyselenide

Fluxes.

Discovery of the K an 2Sb2Se9 Family (Ln = Ce, Sm, Gd, Tb, Dy)
with a E igh t-F old Superstructure Caused by T hree-Dimensional Ordering

of the 552 Lane Pair of Sb3 + Ions

97

1. Introduction

We have applied the chalcoantimonate ﬂux methods to explore the solid state
chemistry of multinary chalcoantimonate compounds. These ﬂuxes are formed by the in
situ fusion of A2Q/Sb/Q or AzQ/Sb2Q3/Q (A = K, Rb, Cs; Q = S, Se) to produce various
Ax[Sbsz] units. These species form the basis for generating a large variety of quaternary
and ternary compounds in a rational and systematic way. Antimony compounds are
expected to have structural diversity due to the tendency of the Sb 53 lone pair to adopt
various stereochemical shapes. Considering that the number of quaternary compounds in
this class is limited (i.e. KHngS3,‘ Kan2-bebe4Se12(Ln = La, Ce, Pr, Gd)?
RngSbTe33) and that the physicochemical characterization of the most mineral
sulfosalts is incomplete, additional study on this type of compounds with various
elements and compositions will provide us with better understanding of solid state Sb
chemistry.

Work with f-elements is of particular interest because the very large f-elements,
along with the lone pair of Sb3+ ions, can induce interesting structures with various kinds
of distortion. For example, reaction of actinide metals with Sb in the polychalcogenide
ﬂuxes resulted in the discovery of AUZSbQ8(A = K, Rb; Q = S, Se)4 , RbUSbo33Te6, and
K'I'thzSe6,5 which show new structure types with unique local environment for the
actinide metals and antimony atoms. After these interesting investigations, we intended
to continue the exploratory synthesis of chalcoantimonate compounds with lanthanide
metals. The lanthanide metals have two major differences from the actinide metals. First,

the most stable oxidation state of the lanthanide metals in a chalcogenide matrix is +3

98

except for few occasions of +2 (i.e. Eu, Yb, Sm). 6’7 Second, the rare earth metals make
more ionic bonds with chalcogens than actinide metals due to their less diffuse and more
localized 4f orbitals.“”7 We expect these features of lanthanide metals affect the structure-
types of the ﬁnal products obtained by the alkali metal polychalcogenide ﬂux methods.
Reaction of lanthanides in the polychalcoantimonate ﬂuxes result in a discovery of
a family of compounds with the composition of KanZszSeg (Ln = Ce, Sm, Gd,8 Tb,
Yb). These quaternary compounds present a new structure type with the lanthanide
metals stabilized in a bi-capped trigonal prismatic sites. The Sb3+ atoms in these
compounds sit on the octahedral sites made by Se/S atoms. At ﬁrst, they appeared to be
positionally disordered, within the same octahedron, over two different sites located 0.26
A away from the center, reminiscent of a double potential well system of equal energy.
However, this model does not represent the true picture in this structure and a closer look
at the x-ray scattering properties revealed the presence of a 2a x 2b x 2c superstructure.
This superstructure lifts the disorder among the Sb atoms and more accurately describes
their distribution. Both the substructure and the superstructure were reﬁned and their
differences are discussed. Here we report the synthesis and structural study of a new
family of lanthanide chalcoantimonate compounds, KanZszseg (Ln = Ce, Sm, Gd, Tb,
Yb). The magnetic properties, thermal stability, and optical, and Raman spectra for all

compounds are reported.

2. Experimental Section

2.1 Synthesis.

99

Reagents. The following reagents were used as obtained: gadolinium, 99.9%, -40
mesh, Cerac, Milwaukee, WI; cerium, 99.9%, metal chips, Research Chemicals, Phoenix,
AZ; sarnarium, 99.9%, metal chips, Research Chemicals, Phoenix, AZ; terbium 99.9%,
metal chips, Research Chemicals, Phoenix, AZ; dysprosium, 99.9%, metal chips,
Research Chemicals, Phoenix, AZ; antimony, 99.999%, -200 mesh, Cerac, Milwaukee,
WI; selenium powder, 99.5+%, 100 mesh, Aldrich, Milwaukee, WI; sulfur powder,
sublimed, JT Baker Co., Philllipsberg, NJ; potassium metal, analytical reagent, Spectrum
Chemical Mfg. Corp., Gardena, CA. The starting material KzSe was prepared by a
stoichiometric reaction of potassium metal and selenium in liq. NH3. SbZSe3 was prepared
by heating a stoichiometric mixture of antimony and selenium at 750 °C for 48 hrs.
KzGdzsbzseg. Amounts of 0.0628 g (0.4 mmol) KZSe, 0.0629 g (0.4 mmol) Gd,

0.0961 g (0.2 mmol) Sb28e3, and 0.1579 g (2 mmol) Se were thoroughly mixed, sealed in an

evacuated Pyrex ampule, and heated at 540 °C for 5 days (cooling 2 oth). Pure black
needles of KzGszszeg were obtained by isolation in degassed dimethylformamide
(DMF) and water (yield > 95% based on Gd metal). The crystals are air- and water
stable.

K2CeszZSe9. Amounts of 0.0628 g (0.4 mmol) KZSe, 0.0629 g (0.4 mmol) Gd,

0.0961g (0.2 mmol) SbQSe3, and 0.1579 g (2 mmol) Se were thoroughly mixed, sealed in an

evacuated Pyrex ampule, and heated at 540 °C for 5 days (cooling 2 °C/h). Pure black
needles of KzGd2$b2$e9 were obtained by isolation in degassed dimethylformamide
(DMF) and water (yield > 95% based on Ce metal). The crystals are air- and water

stable.

100

KzstSbZSeg. Amounts of 0.0628 g (0.4 mmol) KZSe, 0.0629 g (0.4 mmol) Gd,

0.0961g (0.2 mmol) szSe3, and 0.1579 g (2 mmol) Se were thoroughly mixed, sealed in an

evacuated Pyrex ampule, and heated at 540 °C for 5 days (cooling 2 °C/h). Pure black
needles of K2Gd28b28e9 were obtained by isolation in degassed dimethylformamide
(DMF) and water (yield > 95% based on Sm metal). The crystals are air- and water
stable.

KszzstSeg. Amounts of 0.0628 g (0.4 mmol) Kzse, 0.0629 g (0.4 mmol) Gd,

0.0961 g (0.2 mmol) Sb2$e3, and 0.1579 g (2 mmol) Se were thoroughly mixed, sealed in an

evacuated Pyrex ampule, and heated at 540 °C for 5 days (cooling 2 °C/h). Pure black
needles of KzGszszeg were obtained by isolation in degassed dimethylformamide
(DMF) and water (yield > 95% based on Tb metal). The crystals are air- and water
stable.

K2Dy28b28e9. Amounts of 0.0628 g (0.4 mmol) Kzse, 0.0629 g (0.4 mmol) Gd,

0.0961 g (0.2 mmol) SbZSe3, and 0.1579 g (2 mmol) Se were thoroughly mixed, sealed in an

evacuated Pyrex ampule, and heated at 540 °C for 5 days (cooling 2 oC/h). Pure black
needles of KzGdzsszeg were obtained by isolation in degassed dimethylformamide
(DMF) and water (yield ~ 60% based on Dy metal). The crystals are air- and water
stable.

The compositions of the materials were analyzed by Scanning Electron
Microscope (SEM)/ Energy Dispersive Spectroscopy (EDS). Homogeneity for both
compounds was conﬁrmed by comparing the powder X-ray diffraction patterns of the
products against ones calculated using X-ray single crystal data, for example, see Table

4.1 fOI' KzGdZSb28e9.

101

Table 4.1. Calculated and Observed X-ray Powder Pattern of K2Gd23b2869.a

 

 

h k 1 dcalc(A) dobs(A) I/Imax(0b5)(%)
1 1 0 9.630 9.576 13.6
0 2 0 8.831 8.781 100.0
1 2 0 7.001 6.953 80.9
2 0 0 5.744 5.705 35.6
2 1 0 5.462 5.427 29.4
1 3 0 5.239 5.203 6.9
0 4 0 4.415 4.383 32.5
2 3 0 4.111 4.088 14.7
0 2 1 3.808 3.783 9.9
3 l 0 3.742 3.717 21.3
3 2 0 3.513 3.485 49.3
1 3 1 3.287 3.264 17.8
3 3 0 3.210 3.186 19.8
0 4 1 3.051 3.030 54.1
1 4 1 2.949 2.928 52.3
3 4 0 2.893 2.872 10.6
4 1 0 2.835 2.815 43.1
4 2 0 2.731 2.711 70.4
2 4 1 2.694 2.677 17.7
4 3 0 2.581 2.563 20.2
1 7 0 2.464 2.448 9.5
4 4 0 2.408 2.390 16.9
4 1 1 2.353 2.341 22.3
2 7 0 2.310 2.293 13.4
4 2 1 2.293 2.277 13.7
3 5 1 2.211 2.196 14.4
4 3 1 2.202 2.188 20.1
1 8 0 2.168 2.153 57.5
0 0 2 2.110 2.095 14.2
2 8 0 2.061 2.046 24.1
3 6 1 2.042 2.028 14.6
5 2 1 1.967 1.954 8.8
0 8 1 1.956 1.943 6.5

 

102

Table 4.1. (Cont’d).

5

\O-bﬂ-RONMQNNNﬂﬂ-FA-hhomeONw

b)

10

5

NAG—‘00

u—n
O

H
\OMQv-‘OONOO-b-hDJNOO

ﬂ
COO

y—A
O‘iWNW

0

0

WOHHNNNN—‘ﬂN—‘NONNNOON—‘OOO

1

1.926
1.912
1.904
1.857
1.835
1.809
1.766
1.750
1.669
1.634
1.587
1.538
1.512
1.507
1.474
1.452
1.434
1.423
1.394
1.354
1.347
1.257
1.251
1.248
1.205
1.109

1.912
1.899
1.890
1.844
1.823
1.798
1.753
1.737
1.659
1.623
1.578
1.528
1.502
1.498
1.462
1.443
1.424
1.414
1.385
1.347
1.340
1.249
1.243
1.240
1.198
1.101

15.5
8.2
12.0
17.6
5.9
7.4
11.2
16.7
5.0
3.6
4.1
6.8
5.5
4.9
8.1
3.8
8.3
5.4
4.5
5.2
5.5
5.7
5.6
5.6
4.1
3.9

a The hkl indices and the calculated d-spacing are based on the substructure.

103

2.2 Physical Measurements.

Solid-State UV/V is Spectroscopy. Optical diffuse reﬂectance measurements were
performed at room temperature with a Shimadzu UV-3101 PC double-beam, double-
monochromator spectrophotometer operating in the 200-2500 nm region. The instrument
is equipped with an integrating sphere and controlled by a personal computer. 8380,,
was used as a 100% reﬂectance standard for all materials. Samples were prepared by
grinding them to a ﬁne powder and spreading them on a compacted surface of the
powdered standard material and preloaded them into a sample holder. The reﬂectance
versus wavelength data generated were used to estimate a material's band gap by

converting reﬂectance to absorption data as described previously.9

Magnetic Susceptibility. The magnetic response was measured over the range 2-3 00
K using a MPMS Quantum Design SQUID magnetometer. Sample was ground to a ﬁne
powder to minimize possible anisotropic effects and loaded into PVC containers. The
temperature dependent susceptibility studies were performed at 1000 Gauss applied
ﬁled. Corrections for the diamagnetism of the sample containers were made by measuring
the magnetic response of the empty container under the same conditions of temperature
and ﬁeld which were measured for the ﬁlled container. Diamagnetic contribution of every

ion to xMwas corrected according to Selwood.10

104

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~20 mg )
of ground crystalline material was sealed in quartz ampules under vacuum. A quartz

ampule of equal mass ﬁlled with A1203 was sealed and placed on the reference side of the
detector. The samples were heated to 800 °C at 5 °C/min, isothermed for 10 min

followed by cooling at -5 °C/min to 50 °C. Residues of the DTA experiments were
examined by X-ray powder diffraction. The stability/reproducibility of the samples was

monitored by running multiple heating/cooling cycles.

Raman Spectrosc0py. Raman Spectra were recorded on a H010probe Rarnan
spectrograph equipped with a CCD camera detector using 633nm radiation from a HeNe
laser for exitation. Laser power at the sample was estimated to be about SmW and the
focused laser beam diameter was ca. 10 microns. 5 scans were needed to obtain good

quality spectra. The accumulation time of each scan was 50 sec.
2.3 X-ray Crystallography.

Substructure. A Single crystal of KzGdZSb2$e9 with dimensions 0.13 mm x 0.008
mm x 0.008 mm was mounted on the tip of a glass ﬁber. Intensity data were collected at
293K on a Siemens SMART Platform CCD diffractometer using graphite
monochromatized Mo K01 radiation over a hemisphere of reciprocal space. The
individual frames were measured with an omega rotation of 0.3 deg and an acquisition time

of 45 sec per frame for KzGdZszseg. The SMART softwarell was used for the data

105

acquisition and SAINT12 for data extraction and reduction. The absorption correction
was done using SADABS.l3 The complete data collection parameters and details of the
structure solution and reﬁnement are given in Table 4.2.

Structure solution and reﬁnement for both compounds was performed with the
SHELXTL package of crystallographic programs.14 Systematic absence conditions of the
data sets gave two possible space groups, Pbam and Pba2. The structure was solved and
reﬁned successfully with Pbam space group. After successive reﬁnements, the isotropic
temperature factor for Sb(l) in KzGdZSbZSeg was still very high (U,q = 0.04033 A2 with
R1 / wR2 = 0.1267 / 0.2478 for all data) indicating that the Sb atoms may be positionally
disordered in the structure. At a distance of 0.52 A from Sb(l), a high electron density
peak was found which behaved well as a disordered Sb and was assigned as Sb(2). For
ﬁirther reﬁnement, temperature factors, U,,-, for Sb(l) and Sb(2) were constrained to be
equal and the sum of the occupancies of Sb(l) and Sb(2) to be the maximum allowed by
the crystallographic site. After this procedure, temperature factors for Sb atoms and R
indices dropped to reasonable values (Ueq = 0.016 A2 with R1 / wR2 = 0.0337 / 0.0424
for all data). The Sb atoms appear to be disordered over Sb(l) and Sb(2) sites with 50%
occupancy. The structure was successfully reﬁned with reasonable isotropic temperature
factors for the Sb atoms and R values. The temperature factors of the Se(5) atoms were
also high, probably due to the Sb atom disorder. The Se(5)/S(5) atoms are shared by four
Sb atoms in the structure and, therefore, positional disorder of Sb atoms can cause
positional ﬂuctuation of the surrounding Se(5) atoms. Judging from the abnormally high
anisotropic temperature factor, U33, of the Se(5) atoms, the positions of these atoms

seemed to ﬂuctuate mainly along the c-axis. The other Se/S atoms are held not only by Sb

106

atoms but also by the heavy and rigid Gd atoms, and their positions are not affected
signiﬁcantly by the Sb disorder. The coordinates of all atoms, isotropic temperature
factors, and their estimated standard deviations (esd’s) are given in Table 4.4 and the
anisotropic temperature factors are given in Table 4.5. The selected bond distances and

angles for K2Gd2SbZSe9(substructure) are listed in Table 4.6.

Superstructure of KzGdzsbzseg. The positional disorder of the Sb atoms alerted us to
the possibility of a superstructure in these materials. The existence of a superstructure
for KzGszbZSeg was probed by axial photographs on a Rigaku AF C6S diffractometer
with a 4 hr exposure time. The photographs showed weak superlattice peaks around
every crystallographic axis resulting in a 2a x 2b x 2c, 8-fold supercell. Intensity data for
the supercell were collected at 173.1K on a Siemens SMART Platform CCD
diffractometer using the same crystal used for the data collection for the subcell with
longer exposure time, 60 sec per frame. The data were collected over a full sphere of
reciprocal space, up to 28.30° in 0. The collected data were processed in a same way as
described above. The complete data collection parameters and details of the structure
solution and reﬁnement for the superstructure for KzGdZSbZSeg are given in Table 4.2.
The preliminary cell given for the superstructure was orthorhombic with a = 8.4294(2) A,
b = 22.9352(2) A, and c = 35.2682(6) A. As expected from the axial photographs, the
unit cell is doubled along the all three axes resulting in an 8-fold supercell. Each cell
parameter of the supercell is slightly smaller than each subcell parameter multiplied by
two because intensity data for the supercell were collected at lower temperature. The

systematic absence conditions suggested three possible space groups; F mmm, me2, and

107

F222, none of which, however, turned out to be satisfactory. It was suspected that the
crystal system of the supercell might no longer be orthorhombic. If the subtle structural
change, which gives rise to the supercell, deprives the crystal of any of the three diads
(rotation and/or inversion), then the crystal system must be lowered in symmetry even
though the 0t, [3, and y angles remain 90°.

From the substructure, it was possible to deduce the space group of the supercell
by considering the result of doubling all the cell parameters. The screw axes or glide
planes found in the substructure might be lost in the superstructure because their
symmetry operation includes a fractional translation along the axes. The subcell which
has the space group, P2 ,/b2,/a2/m, has four translational symmetry elements; two 21
screw axes along the a-and b-directions; a b-glide plane perpendicular to the a-axis and an
a-glide plane perpendicular to the b-axis. In this case, doubling of all three cell parameters
eliminates all screw axes and glide planes present in the substructure. Only 2-fold axes
parallel to the c-direction and mirror planes perpendicular to the same axis remain. The
loss of both screw axes and glide planes forces the crystal system to be lowered to
monoclinic .

It is obvious that the c-axis, which still has 2-fold axes parallel to it as well as
perpendicular mirror planes, must be the unique b’-axis for the monoclinic setting of the
superstructure. As mentioned previously, systematic absence conditions (hkl : h+k=2n,
h+l=2n, k+l=2n) indicate that the supercell has F-centering. Monoclinic F can be
transformed to monoclinic C by reducing the cell volume by half, see Scheme 4.1.
Consequently, the superstructure has a C-centered monoclinic cell and the unique axis b’

could have a parallel 2-fold axis and/or a perpendicular mirror plane depending on the

108

arrangement of atoms in the supercell. Therefore, the space group should be either C2/m,
Cm, or C2. The vectorial relationships between the supercell parameters a’, b’ c’ and the
subcell parameters, a, b, c are as follows:

a’ = -2a, b’ = 2c, c’= a + b

 

 

 

 

 

  

 

__ F—centered monoclinic lattice

 

 

— — — C-centered monoclinic lattice

 

 

 

Scheme 4.1
The superstructure was solved and reﬁned successfully with space group C 2/m
confu'ming that from all the symmetry elements present in the P bam symmetry of the
subcell, only the 2-fold axes parallel to and the mirror planes perpendicular to the c-axis
(=b’-axis) are valid. No Sb atom disorder was evident in the 8-fold superstructure of

KzGdZszSeg. The apparent Sb atom disorder in the substructure, therefore, is an artifact.

Structures of KanZszseg (Ln = Ce, Sm, Tb, Dy). After the superstructure of

Gd is revealed, the intensity data for KzCeZszSeg, KZSmZSbZSeg, szszseg, and

109

KszZszseg were carefully collected for the superstructure reﬁnement. The data sets
were processed as described above. The superstructures of these compounds were solved
and reﬁned successfully with space group C2/m. They are all isostructural with
KzGdZSbZSeg. The complete data collection parameters and details of the structure
solutions and reﬁnements are given in Table 4.3.

The coordinates of all atoms, isotropic temperature factors, and their estimated
standard deviations (esd’s) for K2Ln28b28e9 are given in Tables 4.7-4.11. The selected

bond distances are listed in Table 4.12-4.16.

3. Results and Discussion

Substructure. Since all the members of the KanZSbZSeg family adopt the same
structure, only KzGdZSbZSeg will be discussed in detail. K2Gd2SbZSe9 has a three-
dimensional tunnel framework with Gd3+ centers, see Figure 4.1. The coordination
geometry around the Gd3+ atom is a bi-capped trigonal prism of Se atoms, see Figure
4.2(a). There are four Se2' ions sitting at two apices and two capping sites of the prism.
The other four Se atoms are from two diselenide groups which form the two short parallel
edges of the trigonal prism. The Se-Se distance is normal at 2.405(1)A. The formula of

the compounds can be described as K+zGd3+2Sb23+(Sez’)5(Se2)2'2, The Gd3+ centered

trigonal prism is sharing its triangular faces with neighboring prisms along the c-axis

‘5 The columns are

forming one-dimensional columns similar to those found in ZrSe3
connected side by side by sharing Se atoms on the capping sites of the trigonal prisms

forming sheets, see Figure 4.2(b). The difference with ZrSe3 lies on the connectivity of

110

Table 4.2. Summary of Crystallographic Data and Structural Analysis for the

substructure and the superstructure of KzGdZszseg.

 

 

Substructure Superstructure
Formula KzGszbzseg KzGdZszSeg
Formula weight 1346.84 1346.84
Crystal habit black needles
Space group P barn C 2/m
a, A 11.4880(3) 22.8783(4)
b, A 17.6612(1) 8.4062(2)
c, A 4.2201(1) 20.970(1)
z; v, A3 2, 856.22(3) 8, 3381.53
Dcalc. g/cm3 5.224 5.291
Temp, K 293(2) 171
MM0 K0414 0.71073 0.71073
p(Mo K01). cm" 303.50 307.39
F(OOO) 1 148 4592
9 max . deg 28.13 28.30
Total data measd 5029 15998
Unique data 1150[R.~m = 0.040] 4306mm = 0.071]

No. of variables
Reﬁnement method

51

162

Full-matrix least-squares on F2

R indices [I>20] R1a = 0.0227 R1 = 0.0797
szb = 0.0403 wR2 = 0.1689

R indices (all data) R1 = 0.0337 R1 = 0.1141
wR2 = 0.0424 wR2 = 0.1782

Goodness of ﬁt on F2 0.976 1.34

 

aR1=2||Fo|—|Fc||/X|Fo| ;

wa2= {314415.246})21/2[w(1r.2)2]}"2

111

Table 4.3. Summary of Crystallographic Data and Structural Analysis for KanZszSeg

(Ln = Ce, Sm, Tb, Dy).

 

 

KzCeZszseq K28m2$b28e9
Formula K2CCszzseg KZSmZSb28e9
Formula weight 1312.58 1333.04
Crystal habit black needles
Space group C 2/m C 2/m
a, A 23.0285(14) 22.796(4)
b, A 8.6380(5) 8.4512(16)
c, A 20.9901(13) 20.905(4)
z; v, A3 8, 3492.0(4) 8, 3375.8(11)
Dcalc. g/cm3 4.993 5.246
Temp, K 171(2) 289(2)
MMo K04), A 0.71073 0.71073
p(Mo K01), cm" 273.84 298.91
F (000) 4496 4560
0 max . deg 22.50 25.00
Total data measd 7059 11541
Unique data 2471 [Rim = 0.0251] 3190 [Rim = 0.0423]
No. of variables 168 168
Reﬁnement method Full-matrix least-squares on F2
R indices [I>29] R18 = 0.0963 R1 = 0.0807
wR2b = 0.2230 wR2 = 0.1872
R indices (all data) R1 = 0.0988 R1 = 0.0868
wR2 = 0.2238 wR2 = 0.1892
Goodness of ﬁt on F 2 1.376 1.406

 

aR1=2||Fo|—|Fc||/2‘.|Fo| ;

bwR2= {mutual-F.2)21/2[w(19.2)2]}"2

112

Table 4.3. (Cont’d). Summary of Crystallographic Data and Structural Analysis for
K2Ln28b28e9 (Ln = C6, Sm, Tb, Dy).

 

 

K2Tb2$b2869 Kzgvzsmsa,
Formula K2TbZSbZSe9 KszzstSeg
Formula weight 1350.18 1357.34
Crystal habit black needles
Space group C 2/m C 2/m
a, A 22.703(3) 22.712(3)
b, A 8.3363(10) 8.3080(9)
c, A 20.875(2) 20.875(2)
z; V, A3 8, 3314.5(7) 8, 3304.9(6)
Deals. g/cm3 5.411 5.456
Temp, K 171(2) 289(2)
MMo Ka). A 0.71073 0.71073
11(Mo K01), cm" 318.91 324.68
F(000) 4608 4624
0 max . deg 24.99 25.00
Total data measd 8615 8375
Unique data 3128 [Rim = 0.0405] 3123 [Rint = 00471]

No. of variables

Reﬁnement method
R indices [I>20]

R indices (all data)

Goodness of ﬁt on F2

162

165

Full-matrix least-squares on F2

R1a = 0.0853
sz" = 0.1778
R1 = 0.0913
wR2 = 0.1804
1.238

R1 = 0.0875
wR2 = 0.1977
R1 = 0.0926
wR2 = 0.1999
1.285

 

3 R1: ZIIFOI — IFCII/EIFOI ;

wa2={gut/(19.2403?]/£[w(r.z)2]}"2

113

Table 4.4. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for KzGd28b28e9 (substructure).

 

 

Atom x y z Ueq,a A2

Gd 0.3838(1) 0.2030(1) 0.5 0.011(1)
Sb(l)b 0.0590(2) 0.1353(1) 0 0.016(1)
Sb(2) 0.0434(2) 0.1076(1) 0 0.016(1)

K 0.2166(2) 0.4435(1) 0.5 0.023(1)
Se(l) 0.3420(1) 0.3208(1) 0 0.013(1)
Se(2) 0.1180(1) 0.2203(1) 0.5 0.016(1)
Se(3) 0.0181(1) 0.3741(1) 0 0.018(1)
Se(4) 0.3160(1) 0.0903(1) 0 0.018(1)
86(5) 0 0 0.5 0.032(1)

 

' U6“ is deﬁned as one third of the trace of the orthogonalized U), tensor.
b Sb(l) and Sb(2) are 50 % occupied.

114

Table 4.5. Anisotropic Displacement Parameters (A2 x 103) for KzGdzsszeg

 

 

(substructure).

U11 U22 U33 U23 U13 U12

Gd 14(1) 10(1) 9(1) 0 0 0(1)
Sb(l)“ 17(1) 15(1) 17(1) 0 0 0(1)
Sb(2) 17(1) 15(1) 17(1) 0 0 0(1)
K 28(1) 21(1) 19(1) 0 0 0(1)
Se(l) 16(1) 12(1) 13(1) 0 0 1(1)
Se(2) 13(1) 16(1) 20(1) 0 0 2(1)
Se(3) 18(1) 21(1) 14(1) 0 0 -5(1)
Se(4) 22(1) 15(1) 16(1) 0 0 -2(1)
Se(5) 26(1) 12(1) 56(1) 0 0 -1(1)

 

a U); for Sb(l) and Sb(2) are constrained to be equal.

115

Table 4.6. Selected Distances(A) and Bond Angles(°) for KzGdZszseg (substructure).

 

 

 

 

Bond Distances
Gd-Se(l) 3.0010(5) x 2 8b(1)-8e(1) 2.611(3)
Gd-Se(2) 3.0123(7) 8b(1)-8e(2) 2.6763(14) x 2
Gd-Se(2) 3.0685(7) 8b(1)-8e(4) 3.058(3)
Gd-Se(3) 2.9478(5) x 2 Sb(l)-Se(5) 3.260(2) x 2
Gd-Se(4) 3.0044(6) x 2 Sb(2)-8e(1) 2.638(2)
8b(1)-Sb(2) 0.522(2) Sb(2)-86(2) 3.024(2) x 2
Se(3)-8e(4) 2.4055(10) Sb(2)-8e(4) 3.146(2)
Sb(2)-Se(5) 2.8830(14) x 2
Bond Angles
Se( 1 )-Gd-Se( 1) 89.3 5(2) Se(l )-Sb(1)-Se(2) 9432(6)
Se(l)-Gd-Se(2) 7682(2) Se(l)-Sb(1)-Se(4) 177.82(9)
Se( 1 )-Gd-Se(2) 8030(2) Se( 1 )-Sb( 1 )-Se(5) 91 . 10(6)
Se(l )-Gd-Se(3) 8430(2) Se(2)-Sb(1)—Se(2) 104.08(8)
Se(1)-Gd-Se(3) 155.13(2) Se(2)-Sb(1)—Se(4) 95 0(5)
Se(l)-Gd-Se(4) 8563(2) Se(2)-Sb(1)-Se(5) 4053(5)
Se(1)-Gd-Se(4) 155.74(2) Se(2)-Sb(1)-Se(5) 87.33(2)
Se(2)-Gd-Se(2) 147.586(13) Se(2)-Sb(l)-Se(5) 166.93(7)
Se(2)-Gd-Se(3) 124.52(2) Se(4)-Sb(1)-Se(5) 9056(5)
Se(2)-Gd-Se(3) 7494(2) Se(5)-Sb(1)-Se(5) 8068(5)
Se(2)-Gd-Se(4) 7893(2) Se(l )-Sb(2)-Se(2) 86. 1 5(6)
Se(2)—Gd-Se(4) 121.98(2) Se(l)-Sb(2)-Se(4) 156.92(8)
Se(3)-Gd-Se(3) 91 .42(2) Se( 1 )-Sb(2)-Se(5) 99.46(6)
Se(3)-Gd-Se(4) 47.66(2) Se(2)-Sb(2)-Se(2) 8848(6)
Se(3)-Gd-Se(4) 109.39(2) Se(2)-Sb(2)-Se(4) 77.42(5)
Se(4)-Gd-Se(4) 8923(2) Se(2)-Sb(2)-Se(5) 88.402(1 1)
Se(2)-Sb(2)—Se(5) 173.3 8(9)
Se(4)-Sb(2)-Se(5) 9620(6)
Se(5)-Sb(2)-Se(5) 9409(6)

 

116

Table 4.7. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for K2Gd28b28e9 (superstructure).

 

 

Atom X y 2 Ueqsa A2
Gd(1) 0.2934(1) 0.2505(2) 0.2030(1) 0.011(1)
Gd(2) 0.0903(1) 0.2483(2) 0.2970(1) 0.011(1)
Sb(la) 0.4608(1) 0.5 0.3648(1) 0.013(1)
Sb(lb) 0.0967(2) 0 0.1318(2) 0.026(1)
Sb(2a) 0.4687(1) 0 0.3930(1) 0.014(1)
Sb(2b) 0.0760(2) 0.5 0.1 106(2) 0.031(1)

K(1) 0.8866(3) 0.2529(9) 0.0564(3) 0.023(1)

K(2) 0.3300(3) 0.2487(9) 0.4436(4) 0.025(1)
Se(la) 0.3292(2) 0.5 0.3187(2) 0.014(1)
Se(lb) 0.0089(2) 0.5 0.1783(2) 0.015(1)
Se(lc) 0.3334(2) 0 0.3227(2) 0.012(1)
Se(ld) 0.0121(2) 0 0.1797(2) 0.012(1)
Se(2a) -0.0513(1) 0.2415(4) 0.2794(2) 0.015(1)
Se(2b) 0.1694(1) 0.2428(4) 0.2206(2) 0.016(1)
86(38) 0.3194(2) 0.5 0.1242(3) 0.020(1)
Se(3b) 0.6112(2) 0 0.4067(2) 0.015(1)
86(3C) 0.3248(2) 0 0.1277(2) 0.014(1)
Se(3d) 0.1955(2) 0 0.3753(2) 0.017(1)
Se(4a) 0.2010(2) 0.5 0.0901(2) 0.018(1)
Se(4b) 0.1965(2) 0.5 0.3729(2) 0.019(1)
Se(4c) 0.2054(2) 0 0.0903(2) 0.017(1)
Se(4d) 0.6142(2) 0.5 0.4128(2) 0.018(1)
Se(5a) 0.5 0.2283(5) 0.5 0.019(1)
Se(Sb) 0 0.2689(6) 0 0.023(1)

 

117

a Ueq is deﬁned as one third of the trace of the orthogonalized U),- tensor.

Table 4.8. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for K2CCszzs€9 (superstructure).

 

 

Atom x y z Ueq,a A2
Ce(1) 0.2923(1) 0.2511(3) 0.2027(1) 0010(1)
Ce(2) 0.0895(1) 0.2477(3) 0.2972(1) 0010(1)
Sb(la)b 0.4618(2) 0.5000 0.3667(4) 0-01 1(2)
Sb(la')b 0.4590(20) 0.5000 0.4010(40) 0.01 1(2)
Sb(lb) 0.0967(2) 0 0.1347(2) 0.01 1(1)
Sb(Za)c 0.4663(3) 0 0.3931(5) 0016(2)
Sb(2a')° 0.4670(30) 0 0.3600(50) 0016(2)
Sb(2b) 0.0722(2) 0.5000 0.1063(2) 0.014(1)

K(1) 0.8860(5) 0.2540(13) 0.0559(5) 0017(2)

K(2) 0.3300(5) 0.2502(13) 0.4437(5) 0018(2)
Se(la) 0.3303(3) 0.5000 0.3220(3) 0-011(1)
Se(lb) 0.0059(3) 0.5000 0.1756(4) 0015(2)
Se(lc) 0.3344(3) 0 0.3281(4) 0014(2)
Se(ld) 0.0089(3) 0 0.1753(3) O009(1)
Se(Za) 0.0517(2) 0.2382(6) 0.2814(2) 0.014(1)
Se(2b) 0.1668(2) 0.2357(5) 0.2185(2) 0012(1)
Se(3a) 0.3157(3) 0.5000 0.1184(3) O-010(1)
Se(3b) 0.6138(3) 0 0.4099(4) 0.012(1)
Se(3c) 0.3245(3) 0 0.1286(4) 0018(2)
Se(3d) 0.1973(3) 0 0.3800(4) 0022(2)
Se(4a) 0.2003(3) 0.5000 0.0901(4) 0013(2)
Se(4b) 0.1964(3) 0.5000 0.3748(3) 0.010(1)
Se(4c) 0.2042(3) 0 0.0863(3) 0.013(2)
Se(4d) 0.6145(4) 0.5000 0.4144(4) 0016(2)
Se(Sa) 0.5000 0.2231(9) 0.5000 0021(2)
Se(Sb) 0 0.2786(8) 0 0.018(2)

 

‘ qu is deﬁned as one third of the trace of the orthogonalized U), tensor.
" 90(2)% Sb(la) and 10(2)% Sb(la').
‘ 90(2)% Sb(2a) and 10(2)% Sb(2a').

118

Table 4.9. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for KZszsbzseg (superstructure).

 

 

Atom x y z Ueq,° A2
Sm(l) 0.2931(1) 0.2508(2) 0.2027(1) 0.011(1)
Sm(2) 0.0904(1) 0.2485(2) 0.2973(1) 0.011(1)
Sb(la)b 0.4623(2) 0.5000 0.3656(4) 0.012(1)
Sb(la')b 0.4632(16) 0.5000 0.3950(30) 0.012(1)
Sb(lb) 0.0972(2) 0 0.1344(2) 0.015(1)
Sb(2a)c 0.4672(2) 0 0.3915(2) 0.018(1)
Sb(2a')° 0.4630(20) 0 0.3500(40) 0.018(1)
Sb(2b) 0.0742(2) 0.5000 0.1079(2) 0.017(1)
K(1) 0.8865(4) 0.2530(10) 0.0562(4) 0.027(2)
K(2) 0.3302(4) 0.2489(10) 0.4435(4) 0.027(2)
Se(la) 0.3293(2) 0.5000 0.3199(2) 0.012(1)
Se(lb) 0.0078(2) 0.5000 0.1779(3) 0.015(1)
8609) 0.3333(2) 0 0.3245(3) 0.015(1)
Se(ld) 0.0105(2) 0 0.1782(2) 0.013(1)
Se(2a) 0.0513(2) 0.2408(4) 0.2799(2) 0.015(1)
Se(2b) 0.1687(2) 0.2396(4) 0.2200(2) 0.015(1)
Se(3a) 0.3180(2) 0.5000 0.1208(2) 0.013(1)
Se(3b) 0.6131(2) 0 0.4086(3) 0.017(1)
8969) 0.3257(2) 0 0.1297(3) 0.020(1)
Se(3d) 0.1966(2) 0 0.3771(3) 0.025(1)
Se(4a) 0.2015(2) 0.5000 0.0909(3) 0.017(1)
Se(4b) 0.1966(2) 0.5000 0.03736(2) 0.013(1)
Se(4c) 0.2045(2) 0 0.0874(3) 0.018(1)
Se(4d) 0.6144(2) 0.5000 0.4134(3) 0.018(1)
Se(Sa) 0.5000 0.2289(6) 0.5000 0.023(1)
Se(Sb) 0 0.2723(6) 0 0.022(1)

 

‘ 93.2(9)% Sb(2a) and 6.8(9)% Sb(2a').

119

..q is deﬁned as one third of the trace of the orthogonalized Ug- tensor.
b 88(2)% Sb(la) and 12(2)% Sb(la').

Table 4.10. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for K2TbZSb28e9 (superstructure).

 

 

Atom x y z Ueq,a A2
Th(l) 0.2937(1) 0.2501(2) 0.2028(1) 0.007(1)
Th(2) 0.0909(1) 0.2490(2) 0.2972(1) 0.006(1)
Sb(la) 0.4648(2) 0.5000 0.3782(3) 0.038(1)
Sb(lb) 0.0963(1) 0 0.1345(2) 0.012(1)
Sb(2a) 0.4659(2) 0 0.3792(3) 0.038(1)
Sb(Zb) 0.0769(1) 0.5000 0.1082(2) 0.01 1(1)
K(1) 0.8869(3) 0.2514(8) 0.0561(3) 0.015(1)
K(2) 0.3308(3) 0.2451(8) 0.4441(3) 0.014(1)
Se(la) 0.3299(2) 0.5000 0.3195(2) 0.009(1)
Se(lb) 0.0102(2) 0.5000 0.1789(2) 0.006(1)
8909) 0.3322(2) 0 0.3211(2) 0.007(1)
Se( 1d) 0.0113(2) 0 0.1807(2) 0.010(1)
8602!) 0.0509(1) 0.2498(4) 0.2794(2) 0.010(1)
Se(Zb) 0.1697(1) 0.2435(4) 0.2204(2) 0.009(1)
Se(3a) 0.3221(2) 0.5000 0.1257(2) 0.013(1)
Se(3b) 0.6131(2) 0 0.4092(2) 0.011(1)
Se(39) 0.3238(2) 0 0.1272(2) 0.010(1)
Se(3d) 0.1957(2) 0 0.3747(2) 0.010(1)
Se(4a) 0.2031(2) 0.5000 0.0921(2) 0.010(1)
Se(4b) 0.1974(2) 0.5000 0.3720(2) 0.01 1(1)
Se(4c) 0.2038(2) 0 0.0888(2) 0.012(1)
Se(4d) 0.6127(2) 0.5000 0.4099(2) 0.012(1)
Se(Sa) 0.5000 0.2495(7) 0.5000 0.023(1)
Se(Sb) 0 0.2677(6) 0 0.014(1)

 

' Ueq is deﬁned as one third of the trace of the orthogonalized Uij tensor.

120

Table 4.11. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for KszZszSeg (superstructure).

 

 

Atom x y z Um,a A2
D70) 0.2939(1) 0.2509(2) 0.2033(1) 0.013(1)
Dy(2) 0.0906(1) 0.2499(2) 0.2967(1) 0.013(1)
Sb(la) 0.461 1(2) 0.5000 0.3654(7) 0.020(1)
Sb(lb) 0.0884(2) 0 0.1225(3) 0.042(1)
Sb(2a) 0.4689(1) 0 0.3914(2) 0.019(1)
Sb(2b) 0.0847(2) 0.5000 0.1204(3) 0.044(1)
K(1) 0.8865(4) 0.2449(9) 0.0557(4) 0.026(2)
K(2) 0.3303(4) 0.2503(9) 0.4440(4) 0.025(2)
Se(la) 0.3304(2) 0.5000 0.3188(2) 0.015(1)
Se(lb) 0.0104(2) 0.5000 0.1807(2) 0.014(1)
8609) 0.3316(2) 0 0.3208(2) 0.014(1)
Se(ld) 0.0119(2) 0 0.1798(2) 0.015(1)
Se(2a) 0.0510(1) 0.2433(4) 0.2792(2) 0.016(1)
Se(Zb) 0.1698(1) 0.2491(4) 0.2207(2) 0.017(1)
Se(3a) 0.3206(2) 0.5000 0.1258(2) 0.019(1)
Se(3b) 0.6106(2) 0 0.4070(2) 0.020(1)
Se(BC) 0.3258(2) 0 0.1287(2) 0.019(1)
Se(3d) 0.1958(2) 0 0.3725(2) 0.019(1)
Se(4a) 0.2028(2) 0.5000 0.0906(2) 0.019(1)
Se(4b) 0.1959(2) 0.5000 0.3733(2) 0.021(1)
Se(4c) 0.2046(2) 0 0.0916(2) 0.020(1)
Se(4d) 0.6142(2) 0.5000 0.4104(2) 0.019(1)
Se(Sa) 0.5000 0.2337(6) 0.5000 0.023(1)
Se(Sb) 0 0.2522(7) 0 0.030(1)

 

8‘ Ueq is deﬁned as one third of the trace of the orthogonalized U), tensor.

121

Table 4.12. Selected Distances(A) and Bond Angles(°) for KzGdzsbzseg

(superstructure).

 

Bond Distances

 

Gd( 1 )-Se(1 a)
Gd(l)-Se(2a)
Gd(1)-Se(lc)
Gd( 1 )-Se(2b)
Gd(1)-Se(3a)
Gd(1)-Se(3c)
Gd( 1 )-Se(4a)
Gd( 1 )-Se(4c)

Sb(l a)-Se(1 a)
Sb(la)-Se(2a)
Sb(] a)-Se(4d)
Sb(l a)-Se(5a)

Sb(lb)-Se(lb)
Sb(lb)-Se(2b)
Sb(lb)-Se(4a)
Sb(lb)-Se(Sb)

K(1)-Se(1b)
K(1)-Se(1d)
K( 1 )-Se(3 a)
K(1)-Se(3c)
K(1)-Se(4a)
K(1)-Se(4c)
K(1)-Se(5b)

Se(3a)«Se(4a)
Se(3b)-Se(4b)

2.965(3)
2.998(2)
3.011(3)
3.055(3)
2.925(3)
2.946(3)
3.003(3)
2.981(3)

2.617(5)
2.609(3) x 2
3.074(5)

3.372(3) x 2

2.599(5)
3.031(4) x 2
3.114(5)
2.790(4) x 2

3.298(7)
3.371(7)
3.359(8)
3.294(7)
3.326(7)
3.371(7)
3.391(6)

2.395(6)
2.413(6)

Gd(2)-Se(lb)
Gd(2)-Se(3d)
Gd(2)-Se(ld)
Gd(2)-Se(2a)
Gd(2)-Se(2b)
Gd(2)-Se(3b)
Gd(2)-Se(4b)
Gd(2)-Se(4d)

Sb(2a)-Se(1c)
Sb(2a)-Se(2a)
Sb(2a)-Se(3b)
Sb(2a)-Se(5a)

Sb(2b)-Se(ld)
Sb(2b)-Se(2b)
Sb(2b)-Se(4c)
Sb(2b)-Se(5b)

K(2)-Se( l a)
K(2)-Se(1c)
K(2)-Se(3b)
K(2)-Se(3d)
K(2)-Se(4b)
K(2)-Se(4d)
K(2)-Se(Sa)

Se(3c)-Se(4c)
Se(3d)-Se(4d)

122

3.014(3)
2.922(3)
2.968(3)
3.056(3)
3.000(3)
2.967(3)
2.949(3)
3.019(3)

2.605(4)
3.068(3) x 2
3.113(5)

2.735(4) x 2

2.619(5)
2.650(4) x 2
3.051(5)

3.313(4) x 2

3.357(8)
3.319(8)
3.380(7)
2.382(6)
3.325(7)
3.312(8)
3.397(6)

2.396(6)
2.382(6)

Table 4.13. Selected Distances(A) and Bond Angles(°) for KzCeZSbZSeo

 

 

(superstructure).

Bond Distances
Ce(1)-Se(3c) 2.991(5) Ce(2)-Se(3d) 3 005(5)
Ce(1)-Se(3a) 3.015(5) Ce(2)-Se(4b) 3.012(5)
Ce(1)-Se(2a) 3 023(5) Ce(2)-Se(2b) 3.023(5)
Ce(1)-Se(4a) 3.034(5) Ce(2)-Se(3b) 3.034(5)
Ce(1)-Se(1 a) 3 043(5) Ce(2)-Se( 1d) 3 060(5)
Ce(1)-Se(4c) 3.059(5) Ce(2)—Se(4d) 3 067(5)
Ce(1)-Se(2b) 3.087(5) Ce(2)-Se(2a) 3 083(5)
Ce(1)-Se(1c) 3.126(6) Ce(2)-Se(1 b) 3 096(5)
Sb(la/l a')-Se(2a) 2.63 1(7)/3 02(4) x 2 Sb(l b)-Se(1d) 2.588(7)
Sb( 1 a/l a')-Se(1 a) 2.634(8)/2.56(3) Sb( 1b)-Se(2b) 2.594(5) x 2
Sb(la/l a')-Se(4d) 3.083(8)/3.25(3) Sb( 1 b)-Se(4c) 3.150(8)
Sb(la/l a')-Se(5a) 3.410(8)/2.96(3) x 2 Sb(lb)-Se(Sb) 3.442(6) x 2
Sb(la)-Sb(la') 0.75(7)
Sb(2a/2a')-Se( l C) 2.557(8)/2.71(5) Sb(2b)-Se(1b) 2.623(8)
Sb(Za/2a')-Se(5A) 2.728(9)/2.28(5) x 2 Sb(2b)-Se(5b) 2.708(6) x 2
Sb(2a/2a')-Se(2A) 3.1 16(9)/2.56(3) x 2 Sb(2b)-Se(2b) 3.142(5) x 2
Sb(2a/2a')-Se(3B) 3.218(9)/2.93(5) Sb(2b)-Se(4a) 3. 147(8)
Sb(2a/2a')-Sb(2A') 0.82(6)
K(1)-Se(1b) 3 .293(1 2) K(2)-Se(1c) 3.293(12)
K(1)-Se(4a) 3.344(11) K(2)-Se(4d) 3.314(12)
K(1)-Se(3c) 3.348(1 1) K(2)-Se(la) 3.346(11)
K(1)-Se(4c) 3 .3 52(1 1) K(2)-Se(3d) 3.364(12)
K(1)-Se(1d) 3 .366(1 1) K(2)-Se(4b) 3.366(1 1)
K(1)-Se(3a) 3.389(1 1) K(2)-Se(3b) 3.380(1 1)
K(1)-Se(5b) 3.420(10) K(2)-Se(5a) 3.421 (10)
Se(3a)-Se(4a) 2.385(9) Se(3c)-Se(4c) 2.402(9)
Se(3b)-Se(4b) 2.388(9) Se(3d)-Se(4d) 2.381(10)

 

123

Table 4.14. Selected Distances(A) and Bond Angles(°) for K2Sm28b23e9

(superstructure).

 

 

Bond Distances

Sm(l)-Se(3c) 2.934(4) Sm(2)-Se(3d) 2.941(4)
Sm(1)-Se(3a) 2.958(3) Sm(2)-Se(4b) 2.953(3)
Sm(1)-Se(1a) 2.986(3) Sm(2)-Se(3b) 2.982(4)
Sm(1)-Se(4a) 2.989(4) Sm(2)-Se(2b) 2.993(3)
Sm(1)-Se(2a) 2.990(3) Sm(2)-Se(1d) 2.996(4)
Sm(l)—Se(4c) 3.012(4) Sm(2)-Se(4d) 3.026(4)
Sm(l)-Se(1c) 3.045(4) Sm(2)-Se(1b) 3.027(4)
Sm(1)-Se(2b) 3.049(3) Sm(2)-Se(2a) 3.048(3)
Sb(la/la')-Se(2a) 2.615(6)/3.02(4) x 2 Sb(lb)-Se(ld) 2.593(5)
Sb(la/ 1a')-Se(1a) 2.633(6)/2.56(3) Sb(lb)-Se(2b) 2.604(4) x 2
Sb(la/ 1a’)-Se(4d) 3.042(6)3.25(3)/ Sb(lb)-Se(4c) 3.095(6)
Sb(la/ 1a')-Se(5a) 3.354(7)/2.96(3) x 2 Sb(lb)-Se(Sb) 3.363(4) x 2
Sb(la)-Sb(la') 060(5)

Sb(2a/2a')-Se(lc) 2.572(5)/2.71(5) Sb(2b)-Se(1b) 2.617(5)
Sb(23/2a')-Se(5a) 2.756(5)/3.38(5) x 2 Sb(2b)-Se(5b) 2.737(4) x 2
Sb(2a/2a')-Se(2a) 3057(4)/2.56(3) x 2 Sb(2b)-Se(2b) 3.077(4) x 2
Sb(2a/2a')-Se(3 b) 3 . 145(6)/2.93(5) Sb(2b)-Se(4a) 3.109(5)
Sb(2a)-Sb(2a') 0.82(6)

K(1)-Se(1b) 3.286(9) K(2)-Se(1c) 3.288(9)
K(1)-Se(3c) 3 .3 1 1(9) K(2)-Se(4d) 3 .305 (9)
K(1)-Se(4c) 3.331(9) K(2)-Se(3d) 3.320(9)
K(1)-Se(4a) 3.333(9) K(2)-Se(4b) 3.326(9)
K(1)-Se(3a) 3.337(8) K(2)-Se(la) 3.336(8)
K(1)-Se(1d) 3.347(9) K(2)-Se(3b) 3.360(8)
K(1)-Se(5b) 3.378(7) K(2)-Se(5a) 3.379(8)
Se(3a)-Se(4a) 2.374(6) Se(3c)-Se(4c) 2.398(6)
Se(3b)-Se(4b) 2.384(6) Se(3d)-Se(4d) 2.373(7)

 

124

Table 4.15. Selected Distances(A) and Bond Angles(°) for KszZSbZSeg

(superstructure).

 

Tb(1)-Se(3c)
Tb(1)-Se(3a)
Tb(1)-Se(4a)
Tb(1)-Se(1a)
Tb(1)-Se(lc)
Tb(1)-Se(2a)
Tb(1)-Se(4c)
Tb(1)—Se(2b)

Sb(l a)-Se(1a)
Sb(la)-Se(2a)
Sb(la)-Se(Sa)
Sb( 1 a)-Se(4d)

Sb(2a)-Se(1c)
Sb(2a)-Se(2a)
Sb(2a)-Se(5a)
Sb(2a)-Se(3b)

K(1)-Se(3c)
K(1)-Se(1b)
K(1)-Se(3a)
K(1)-Se(4c)
K(1)-Se(ld)
K(1)-Se(4a)
K(1)-Se(5b)

Se(3a)-Se(4a)
Se(3b)-Se(4b)

2.912(3)
2.912(3)
2.956(3)
2.961(3)
2.976(3)
2.978(3)
2.981(3)
3.030(3)

2.605(5)
2.814(5) x 2
3.041(6) x 2
3.051(5)

2.584(5)
2.823(5) x 2
3.026(6) x 2
3.047(5)

3.297(7)
3.301(7)
3.315(7)
3.317(7)
3.337(7)
3.341(7)
3.355(6)

2.395(5)
2.425(6)

Bond Distances

Tb(2)-Se(3d)
Tb(2)-Se(4b)
Tb(2)-Se(1d)
Tb(2)-Se(3b)
Tb(2)-Se(4d)
Tb(2)-Se(2b)
Tb(2)-Se(1b)
Tb(2)-Se(2a)

Sb(lb)-Se(ld)
Sb(lb)-Se(2b)
Sb(lb)-Se(4c)
Sb(lb)-Se(Sb)

Sb(2b)-Se(1b)
Sb(2b)-Se(5b)
Sb(2b)-Se(2b)
Sb(2b)-Se(4a)

K(2)-Se(3d)
K(2)-Se(1c)
K(2)-Se(3b)
K(2)-Se(4b)
K(2)-Se(4d)
K(2)-8e(1 a)
K(2)-Se(5a)

Se(3c)-Se(4c)
Se(3d)-Se(4d)

125

2.897(3)
2.925(3)
2.951(3)
2.969(3)
2.969(3)
2.982(3)
2.985(3)
3.033(3)

2.585(5)
2.619(3) x 2
3.065(5)
3.311(4) x 2

2.627(5)
2.758(4) x 2
3.022(3) x 2
3.060(5)

3.294(7)
3.295(7)
3.307(7)
3.315(7)
3.346(7)
3.350(7)
3.352(6)

2.384(5)
2.365(6)

Table 4.16. Selected Distances(A) and Bond Angles(°) for KszszZSeg

(superstructure).

 

Dy( 1 )-Se(3a)
Dy(1)-Se(3c)
DY(1)-Se(la)
Dy( 1 )-Se(4c)
Dy(l)-Se(1c)
Dy(1)-Se(4a)
I0(1)-364281)
D3(1)-Scab)

Sb(la)-Se(la)
Sb(la)-Se(2a)
Sb(la)-Se(4d)
Sb(la)-Se(Sa)

Sb(2a)-Se(1c)
Sb(2a)-Se(5a)
Sb(2a)-Se(2a)
Sb(2a)-Se(3b)

K(1)-Se(3a)
K(l )-Se(1d)
K(1)-Se(4c)
K(1)-Se(3c)
K(1)-Se(4a)
K(1)-Se(1b)
K(1)-Se(5b)

Se(3a)-Se(4a)
Se(3B)-Se(4b)

2.887(3)
2.919(3)
2.933(3)
2.958(3)
2.967(3)
2.967(3)
2.976(3)
3.031(3)

2.572(5)
2.619(4) x 2
3.069(5)

3.296(4) x 2

2.626(5)
2.768(4) x 2
3.015(3) x 2
3.054(5)

3.303(7)
3.307(7)
3.309(8)
3.317(8)
3.345(7)
3.352(8)
3.356(7)

2.358(6)
2.394(6)

Bond Distances

Dy(2)-Se(3d)
Dy(2)-Se(4b)
Dy(2)-Se(3b)
Dy(2)-Se(1b)
Dy(2)-Se(1d)
Dy(2)-Se(4d)
Dy(2)-Se(2b)
Dy(2)—Se(2a)

Sb(lb)-Se(ld)
Sb(lb)-Se(2b)
Sb(lb)-Se(4c)
Sb(lb)-Se(Sb)

Sb(2b)-Se(1b)
Sb(2b)-Se(2b)
Sb(2b)-Se(5b)
Sb(2b)-Se(4a)

K(2)-Se(4b)
K(2)-Se(3d)
K(2)-Se(4d)
K(2)-Se(1c)
K(2)-Se( 1 a)
K(2)-Se(3b)
K(2)-Se(5a)

Se(3c)-Se(4c)
Se(3d)-Se(4d)

126

2.901(3)
2.904(3)
2.947(3)
2.949(3)
2.952(3)
2.974(3)
2.976(3)
3.033(3)

2.590(5)
2.789(5) x 2
3.039(5)

3.062(6) x 2

2.598(5)
2.839(5) x 2
2.999(6) x 2
3.063(5)

3.302(8)
3.305(8)
3.311(8)
3.321(7)
3.338(8)
3.355(8)
3.366(7)

2.419(6)
2.378(6)

 

 

 

Figure 4.1. The overall structure of KzGszszeg viewed down the c-axis with labeling.

127

7\\\\\\/
233*
/

fW

8W
.. /
v

\\>\‘/
22%
V

I

/

Corner Sharing

‘0’ W

K2Gd2$b2$eg

 

ZrSe3

Figure 4.2. (a) Coordination environment of the Gd atom. (b) Polyhedral representation
of the Ody-centered bicapped trigonal prisms. They share corners along the a-axis and
triangular faces along the c-axis to form layers. (c) Comparison of the layers of
KzGdzsbzseg and ZrSe3.

128

single columns to build layers. In ZrSe3, each column shares Se atoms at the apex and
capping sites with the neighboring columns, so that the apices in one become caps in the
next. In contrast to ZrSe3, KzGdZszSeg has columns which share Se atoms only at the
capping sites, see Figure 4.2(c). As a result, each layer of ZrSe3 is built from bi-layers of
Zr atoms while each layer of KzGszszeg consists of only monolayers of Gd atoms.

The antimony atoms in the compounds have a formal charge of +3. Theses atoms
appear to be positionally disordered over two crystallographically different sites, Sb(l)
and Sb(2), with 50% occupancy. The distance between these two sites is 0.522(2) A.
The Sb ions are stabilized in distorted octahedral sites, see Figure 4.3(a). The local
symmetry of the Sb3+ ion can be described intermediate between trigonal pyramidal and
octahedral with three shorter Sb-Se bonds ranging from 2.618 A to 2.883 A and three
longer Sb-Se bonds ranging from 3.024 A to 3.260 A. This type of distorted environment
indicates that the 53 lone pair of Sb is stereochemically expressing itself and it is directed
towards the three longer Sb-Se bonds. Both Sb(l) and Sb(2) sites are located inside of the
same octahedral pocket and have the same type of distortion. For example, Sb(l) is
making three shorter bonds with Se(l), Se(2), Se(2') and three longer bonds with Se(5),
Se(5') Se(4) while Sb(2) is making three shorter bonds with Se(l), Se(5), Se(5') and three
longer bonds with Se(2), Se(2'), Se(4), see Figure 4.3(b). The SbSe6 octahedra share edges
with neighboring SbSe6 octahedra forming an inﬁnite chain along the c-axis, see Figure
4.3(c). Two such single chains also share edges making inﬁnite double chains of Sb atoms,

which then bridge Gd layers together to make the whole framework three-dimensional.

129

(a) 894

(b)

 

 

Figure 4.3. (a) Coordination environment of the Sb atoms. (b) Distortion in the
Sb(l)Se6 and Sb(2)Se6 octahedra. (c) Polyhedral representation of the SbSe6 octahedral

blocks along the c-axis.

130

The K+ ﬁlled tunnels run parallel to the c-axis. The K+ ions are coordinated by 7

selenium atoms in monocapped trigonal prismatic sites with an average (K-Se) distance of

3.36 A.

Superstructure. In the superstructure, Sb atoms are fully ordered occupying Sb(l)
and Sb(2) sites alternatively with full occupancy instead of statistical disorder over the
two sites. The overall superstructure view down the b’-axis is shown in Figure 4.4(a).
The unit cell of the subcell is also shown with a dotted line for comparison. The labeling
scheme for the atoms in the supercell are devised to preserve the numbering of the Sb and
Se atoms in the subcell for the sake of comparison. For example, Se(la), Se(lb), Se(lc),
and Se(ld) in the supercell are derived from the same atom, Se(l), in the subcell. Figure
4.4(b) shows space-group diagrams of Pbam and C2/m view down the c- and b’-axis,
respectively. As explained previously, the a- and b-glide planes and the 21 screw axes
present, in Pbam, no longer exist in C2/m. Instead, an a-glide plane perpendicular to the
b’-axis and a 2, axis parallel to the b’ axis are generated by the positional ordering of the
Sb atoms in the superstructure.

The arrangement of Sb atoms in the Sb double chain along the b’-axis in the
superstructure is shown in Figure 4.5 compared with that in the substructure. Figure
4.5(a) and (b) is a ball and stick representation of edge-sharing Sb octahedra which were
shown previously as a polyhedral representation in Figure 4.3(c). The axial Se atoms
which do not have any structural variations are omitted for clarity. The two rows of Sb
atoms sandwich a row of Se(5a) atoms. Each Sb row is composed of Sb(la) and Sb(2a)

sitting alternatively. This is how the unit cell is doubled along the c(=b’)-axis. Since

131

 

 

 

 

(b) < subcell > < supercell >

L

     

(—

 

 

 

 

 

Figure 4.4. (a) The superstructure of KzGd28b28e9 viewed down the b’-axis with
labeling. The solid line represents the unit cell for the superstructure and the dotted line
represents the unit cell for the substructure. (b) corresponding space group-diagrams for

P bam and C 2/m view down the c- and b’-axis respectively.

132

(a) Substructure

 

subceﬂ

(b) Superstructure

 

O O O 8131 - O Sb2a Se2a
’ o 9 o
O 0 O O Se5a
a ’ e "
O 1 O O J 0 0892a
' supercell 1
D b’(=c)-axis

 

 

Figure 4.5. Ball and stick representation of the edge-shared SbSe6 octahedra along the
b'-axis ( = c-axis) for (a) the substructure and (b) the superstructure. Axial Se atoms are
omitted for clarity. (c) The local environment of Sb(la) and Sb(2a).

133

Sb(la) is making longer bonds with Se(5a), and Sb(2a) is making shorter bonds with
Se(5a), the Sb atom propagation along the b’-axis is not a straight line but zigzag. In the
substructure, the coordination environment of Se(5) seemed to be square planar.
However, the ordering of the Sb atoms in the superstructure reduces the square planar
symmetry and causes the Se(5) atoms to have two short bonds with Sb(2a) and two long
bonds with Sb(la).

The unit cell is doubled along the a- and b- direction because of the way that the
Sb chains are arranged with respect to one another in the space. Every other Sb chain
along the a- and the b-axis is shifted by 1/2 c (=1/4 b’), which is the distance between
Sb(l) and Sb(2) in the Sb chain, so that Sb(l) in one chain and Sb(2) in the next chain can
sit on the same ab-plane causing the repeating unit along the a- and the b- axis to double.

The zigzag arrangement of the Sb atoms, which causes the 8-fold superstructure,
can be interpreted as a result of the ordering of the 58 lone pairs of Sb ions. Figure 4.5(c)
shows two different orientations of the lone pairs of Sb(la) and Sb(2a) ions. The lone
pairs of Sb(la) are directed toward Se(5a) atoms and those of Sb(2a) are directed toward
Se(2a) atoms.

Figure 4.5(b) also shows that the positions of Se(5a) atoms in the superstructure
are modulated along the b'-axis. As a result, there are two different Se(5a)---Se(5a)
distances in the superstructure; one is short and the other is long while all the Se(5)---
Se(5) distances are all equal in the substructure. This explains why the temperature factor
for Se(5) is high in the substructure reﬁnement and why only the U33 component for
Se(5), which represents the positional displacement along the c-axis ( = b'-axis ), has a

high value. Since the superstructure allows this modulation of Se(5) atoms, the

134

temperature factors of Se(5a) and Se(5b), which is derived from Se(5) in the subcell, now
have reasonable values.

The positional variation of Gd, K, and Se atoms (except for Se(5)) in the
superstructure are negligible with respect to those in the substructure. The supercell is,
therefore, entirely attributed to the Sb positional ordering caused by the steric
requirements of the 582 lone pair.

The rest of members of the KanZSbZSeg family share the same structural features
with K2Gd2$bZSe9. The average distance of Ln-Se bonds gets slightly shortened (from
3.03A for KzGdZSbZSeg to 2.95A for KszzstSeg) as the atomic number of the Ln metal
increases due to the lanthanide contraction. KzCeZSb2$e9 and K2$m28b28e9 still show
small amount of Sb disorder (~10%) even in the superstructure reﬁnement, patterns of
which are the same as observed in the substructure. This seems to be caused by
imperfect ordering of Sb chains. The Sb ions must be well ordered within the one
dimensional chains to minimize the steric repulsion between the lone pairs. However,
since these one dimensional chains are well separated from one another, there is a
possibility that these chains may not be well ordered three dimensionally. Ideally, these
chains are arranged in such a way that Sb(1) in one chain and Sb(2) in the neighboring
chains sit on the same ab(a'c')-plane so that Sb(l) and Sb(2) alternate along the a- and b-
directions. If one out of ten Sb chains does not follow this rule and disturb the perfect
alternating of Sb(l) and Sb(2) on the ab plane, this will be averaged out as 10%

disordering of Sb(l) and Sb(2) atoms as observed in KzCeZszSeg and Kzsmzsb2869.

135

Properties. The compounds reported here are valence-precise and are expected to be
semiconductors. The absorption spectra conﬁrm this by showing the presence of abrupt
optical gaps. The absorption spectra of these compounds share similar features with one
another, see Figure 4.6. The band gaps vary from 1.33-1.56 eV depending on the
compounds, which are indicated in the ﬁgures. The absorption spectrum of KZSmZSbZSeg
shows extra absorption bands below the band-gap transition, which are due to the f-f or f-
d transitions.

Magnetic susceptibility measurements of KanZSbZSeg as a function of
temperature are shown in Figure 4.7. All the compounds exhibit antiferromagnetic
transition around 3K, see insets of Figure 4.7. The room, temperature peg values for the
compounds are summarized in Table 4.17. The corresponding theoretical Meir values
calculated by Van Vleck equation for each free Ln3+ ion are also listed for comparison. ‘6
These observed and calculated um values show good agreements, indicating that the rare
earth ions, even in a crystal lattice, behave almost as though they are free ions. This is
because the 4f electrons are effectively shielded from their environment by the completed
58 and 5p subshells. While other compounds show linear dependency of l/xM versus
temperature, 1/xM of KZszsbzseg does not obey the Curie Weiss law over the
temperature range of 4K-300K. This is because the separation between the
ground(J=5/2) and the next excited state(J=7/2) of the Sm2+ (993 cm“) is comparable to
kT at room temperature and so the dependence of MA on T becomes more complicated.”
It is necessary to include more than one J level to calculate the mean susceptibility and

each level has to be properly weighted according to its population. This is also why the

calculated and observed um values for KZszSbZSeg shows the largest discrepancy.

136

 

 

6 .. K2Ce28b2$eg

Ot/S (Arbitrary Units)

 

 

 

 

 

 

   
   

0 -_s - l - I RA; 1 n 1 a
0.5 1 1.5 2 2.5 3
Energy (eV)

8
K28m28b2899

oc/S (Arbitrary Units)
.3)

E9 = 1.39eV

 

 

l.__l_ I. L l _l l l l l l l L l l l l j

0 . . .
0.5 1 1.5 2 2.5 3
Energy (eV)

Figure 4.6. Optical absorption spectra of KanZszseg (Ln = Ce, Sm, Gd, Tb, Dy). The

band gap value, E3, is shown.

137

 

 

: KzGdzsbzseg

b)
' I

01/8 (Arbitrary Units)
N

 

l j l l l l l J l I l l

O I. l n s n n n s n
0.5 l 1.5 2 2.5 3
Energy (eV)

 

 

4 :' K2Tb23b2$69

 

(x/S (Arbitrary Units)
N

 

 

 

1 I.
I 59: 1.44 eV
0 b‘u—l l - n l n n 44 L n n 14 l J
0.5 l 1.5 2 2.5 3
Energy (eV)

Figure 4.6. (Cont’d). Optical absorption spectra of KanZSbZSeg (Ln = Ce, Sm, Gd, Tb,
Dy). The band gap value, Eg, is shown.

138

 

5 " KZDYQszseg

 

a/S (Arbitrary Units)

 

 

 

Energy (eV)

Figure 4.6. (Cont’d). Optical absorption spectra of KanZSbZSeg (Ln = Ce, Sm, Gd, Tb,
Dy). The band gap value, Eg, is shown.

 

 

 

 

 

 

 

400
r? -
"" I
.g KQCGQSDQSGQ I - I
a. .
8 300 - I
m A . I
53 g I
I
U 60.
13$) I 1
8 o 200 " I ’3 g '
50 E I E 40; .
<6 v I o .
h :
2 t I. o s
6 100 s
> 0:---Al-..ml-‘-mluubml
E 0 5 10 15 20 25
Temperature (K)
1 I I I I I. I I I I l I I I I I I I I I I

  

I l I I

0 50 100 150 200 250 300 350
Temperature (K)

 

 

Figure 4.7. Inverse molar magnetic susceptibility, l/x M, of (Ln = Ce, Sm, Gd, Tb, Dy).

139

 

 

 

 

 

 

 

 

>5
:3 2000 - K2$m28b28e9 I I I I
.3 "
o. - I '
8 I I '
U)
Us) E 1500 :' I . I 800.
08$) .. II. E 600:- I .
O.) O - I $3 : I
N v ' . E :If
2 . .- 5 20°."
8 » I 7 0; ........................
5 500':l 0 5 1015 20 25
E ! Temperature (K)
0—1
0 1

0 50 100 150 200 250 300 350
Temperature (K)

 

 

 

 

 

  
 

 

  

 

40 I
a: : K Gd so 3 '
g . 2 2 2 99 .
.5 . .
G» 30 -
8 ' _ '
3A
m =51 . 4
3 0 20 .. I E q
49: .. I ’5 3'» '
Cl 0 E : I
grog A I g 2'.

r o :
8 10 ' .l {e t
B d 0' .........................
> 0 5 10 15 20 25
.5 Temperature (K)
0

0 50 100 150 200 250 300 350
Temperature (K)

Figure 4.7. (Cont’d). Inverse molar magnetic susceptibility, 1/x M, of (Ln = Ce, Sm,

Gd, Tb, Dy).

140

 

 

 

 

 

 

 

30 _

>5
.: r
B 25 L K2Tb28b2899 I
'5 I I
a : -
o ' I
g A 20 :- I
‘3 E : '3. A,
D b b - D i .

O . E E
ED E I- I d) 2!- .
w v - I a E
2 10 - I E 15.5-ff

. I v ,
8 .- X :
u - I 2 E
Q; 5 - O: .........................
:- t o 5 10 15 20 25
'-‘ Temperature (K
I I L I I I I l I I I L

0 InnnlnnnnlnnnninnLnlnl

0 50 100 150 200 250 300 350
Temperature (K)

 

 

   

 

 

 

 

 

I

>,. I
E 30 KZDYZszseg I .
:9. I
13- I
‘8’ I
:1 A I
Cl) 2 20 I 6E
.8 o I 3 E '
o h E 4 =. I

o I 0 :
Eu E I. 70 5 '
CU v E E

I V L.
E 10 I x 2g f
a) 2 E
33 o: .........................
o 0 5 10 15 20 25
E Temperature (K)
I I I I I I l I

O Jannlnnnnlnnnnlnnnnlnnn

0 50 100 150 200 250 300 350
Temperature (K)

 

Figure 4.7. (Cont’d). Inverse molar magnetic susceptibility, 1/x M, of (Ln = Ce, Sm,

Gd, Tb, Dy).

141

Table 4.17. Summary of Magnetic Properties of KanZSbZSeg (Ln = Ce, Sm, Gd, Tb,

 

 

Yb).
Ln3+ No off Ground Observed um Theoretical “eff Antiferromagnetic
electrons state (at 300 K) transition

Ce3+ 1 2135/2 2.57 2.54 <2K

Sm3+ 5 6115/2 1.09 0.84 3K

Gd3+ 7 “87,2 7.87 7.94 3K

Tb3+ 8 71?6 9.58 9.72 3K

Dy3+ 9 6H15/2 10.49 10.63 3K

 

142

The Raman spectra of KzGdzsbzng are shown in Figure 4.8. The shiﬁs at 256
cm"1 and 266 cm'1 are assigned to the stretching vibration of the di-chalcogenide groups
and these values are in accord with the Se-Se stretching frequencies reported for other
compounds. 18’ ‘9’20 The strong shift at 218 cm‘l is attributed to a Sb -Se vibration,
which causes an overtone at 445 cm". The Raman spectra of other compounds show
very similar features with the Se-Se stretching vibrations appearing at the same

frequencies.

Differential thermal analysis (DTA) experiments showed that these compounds

are not thermodynamically stable and decompose gradually above 400 °C.

 

 

 

 

 

co
‘-
N
3‘ 8
I; N
5
c»
H
.5 $53
F
E‘ N
[s
v-
.2 l0
1- 3
<
IIIIIIIIIIIIIJIIIIIIIIIIIII

 

200 300 400 500 600 700
Raman Shift (cm’l)

Figure 4.8. Raman spectrum of KzGdzsbzseg.

143

4. Conclusions

The family of KanzSnZSe9(Ln=Ce, Sm, Gd, Tb, Yb) adopts a new structure type,
which is composed of two-dimensional corrugate layers formed by Ln3+ centers and one-
dimensional Sb chains that bridge these layers. These compounds are a good example of
how the 532 lone pair of the Sb3+ ion can induce a novel and elaborate framework by
distorting its local environment. To reduce steric repulsion between the lone pairs of
adjacent Sb ions in the edge-shared octahedral blocks, the ions adopt a zigzag arrangement,
which results in a 2a x 2b x 2c superstructure. This structure-type seems to be quite
stable and easily formed with Ln3+ centers. Only the reactions with Eu metal which
prefer +2 oxidation state produced compounds with a different structure. These series of
compounds show a great similarity in their physico-chemical properties because the 4f

electrons are effectively shielded by outer 5s and 5p orbitals.

144

10.

11.

12.

References

Imafuku, M.; Nakai, 1.;Nagashima, K. Mat. Res. Bull. 986, 211, 493.

Chen, J. H.; Dorhout, P. K. J. Alloys Compd. 1997, 249, 199.

Li, J.; Chen, Z.; Wang, X.; Proserpio, D. M. J. Alloys Compd. 1997, 262-263, 28.
Choi, K.-S; Kanatzidis, M. G. Chem. Mater. 1999, 11, 2013-2018.

Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G. Inorg. Chem.
1997, 36, 3804-3 805 and the references therein.

Greenwood, N. N.; Eamshow, E. Chemistry of the elements; Pergamon Press: New
York, 1984; 1434.

Guittard, M.; F lahaut, J. Synthesis of Lanthanide and Actinide Compounds,
Kluwer Academic Publisherrs, Netherlands, 1991, 321.

Choi, K. S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309-
319.

McCarthy, T. J .; Ngeyi, S.-P.; Liao, J. —H.; Degroot, D.; Hogan, T.; Kannewurf,
C. R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331.

Selwood, P. W. in “Magnetochemistry” 2nd ed., Interscience Publishers, New
York, 1956.

SMART: Siemens Analytical Xray Systems, Inc., Madison, WI 53719, USA,
1994.

SAINT: Version 4, Siemens Analytical Xray Systems, Inc., Madison, WI 53719,

USA, 1994-1996.

145

13.

14.

15.

16.

17.

18.

19.

20.

Sheldrick, G. M. University of Gottingen, Germany, to be published.
SHELXTL: Version 5, G. M. Sheldrick, Siemens Analytical Xray Systems, Inc.,
Madison, WI 53719, USA, 1994.

Krb'nert, Von W.; Plieth, K. Z anorg. allg. Chem. 1965, 336, 207.

Bottcher, P.; Getzschmann, J .; Keller, R. Z. anorg. allg. Chem. 1993, 619, 476.
Choi, K.-S.; Patschke, R. R.; Billinge, S. J. L.; Waner, M. J.; Dantus, M.;
Kanatzidis, M. G. J. Am. Chem. Soc.l998, 120, 10706.

Nouvel, G.; Zwick, A.; Renucci, M. A.; Lockwood, D. J .; No'e'l, H. J. Phys. C:
Solid State Phys. 1987, 20, 1881-1897.

Drago, R. S. in “Physical Methods for Chemists”, 2nd ed., Saunders College
Publishing, New York, p484, 1992.

Mulay, B. “Theory and Applications of Molecular Paramagnetism”, John Wiley

& Sons, New York, 1976

146

Chapter 5. Reactions of Europium Metal with Sb in Polychalcogenide

Fluxes.

Novel Europium Chalcoantimonates, EquSe 3, EquTe3, and

Bag, 5E uo, 5Sb T e 3 with Distorted T wo—Dimensional Square Se/T e Nets.

147

1. Introduction

We are pursuing exploratory synthesis of ternary or quaternary chalcoantimonate
compounds incorporating lanthanide metals. Our strategy has been to combine large and
ﬂexible coordinations of the lanthanide metals and asymmetric building units of Sb ions,
which are caused by the presence of the 582 lone pair electrons, in order to derive
interesting structure types. However, lanthanide metals usually form isostructural
compounds due to their similar ionic radii and chemical natures, which makes it
relatively hard to achieve structural variations by changing the lanthanide metals. Up to
date, only two types of quaternary Chalcoantimonates, AanZszSeg (Ln = Ce, Sm, Gd,
Tb, Dy)l and K2Ln2-bebe4Se12(Ln = La, Ce, Pr, Gd)2 were reported, which can be
stabilized with most of the trivalent lanthanide metals. To circumvent their formation,
we decided to incorporate divalent lanthanide metals into the chalcoantimonate ﬂuxes.
Only a few lanthanides, however, are stable as +2 ions including Sm, Eu, Tm, and Yb
with Eu being the most stable.3 Therefore, we chose Eu to react with chalcoantimonate
ﬂuxes and indeed discovered three new europium compounds, EquSe3, EquTe3, and
Bao,5Euo,5SbTe3. These compounds commonly possess two dimensional chalcogenide
nets.

The compounds with square chalcogenide nets have been of particular interest
because they have been found to undergo structural distortions that result in interesting
superstructures.“’5‘6 The properties of these compounds can change dramatically
depending on the distortion of the net. This is because the electronic bands near Fermi

level are entirely made up of p orbitals of Se/T e atoms from the square net. If the Se/T e

148

net is ideally square, each chalcogen has four neighboring chalcogens and these four Q-Q
bonds (Q = chalcogen) are of equal length. Usually the Q-Q distance in the ideal net is
signiﬁcantly elongated than normal Q-Q bonds found in discrete dichalcogenide groups
due to the extended coordination number per chacogen atom. In this case, the valence
electrons in the chalcogen atoms delocalize across the net. Since the average charge per
Te atom in the net is usually less than —2, the p bands of Te/Se is not completely ﬁlled,
resulting in metallic properties of the materials. This partially ﬁlled band makes the net
susceptible to distort because the total energy of the system can be lowered by distortion.
The net can be oligomerized into small fragments so that all the chalcogen atoms in the
net can satisfy the octect rule with correct number of neighbors, depending on the formal
charges of each Te/ Se atom. Such distortion can be viewed as a version of Peierls
distortion, in which displacement of atoms causes the atomic coordination to be reduced
and the overal energy of the system to be lowered.7 This process localizes the electron
density into the shortened Q-Q bonds in the oligomers and opens a gap at the Fermi Level
by lowering the energy level of the resulting ﬁlled p band. Consequently, the physical
properties of the material changes from metallic to semiconducting.

Extensive work has been performed in binary transition or lanthanide metal
dichalcogenide or trichalcogenide compounds to reﬁne (or at least model) superstructures
caused by the distortion of the net and to correlate the structure and properties. However,
other than these binary compounds, relatively few structure types with the nets are known
(i.e. KOBBaWAgTez, 8 KCuCeTe4,9 KCquuTe4,l° ALn3Te3(A = K, Rb, Cs; Ln = Ce,

Nd).ll This restricts our opportunity to systematically investigate the relationship

149

between the types of distortion, electronic band structures, and the observed properties.
In this sense, the discovery of new structure types with Se/T e net is quite valuable.
EquSe3, EquTe3, Ba0_5Euo,5SbTe3 all possess superstructures primarily due to
the modulations of the Se/Te net. The structures of EquSe3 and EquTe; are closely
related to each other in that they possess the same structural motifs that build the whole
framework. However, they are not isostructural due to the slight difference in the local
coordination of Sb and the resulting difference in the stacking arrangement of the layers.
The average structures of Ba0,5Euo_5SbTe3 and EquTe3 are identical but they possess
different types of superstructures. This is a rather remarkable observation considering
that Bao_5Euo,5SbTe3 is derived from EquTe3 by isoelectronic substitution of a half
equivalent Eu by Ba. Here, we present the synthesis, structures, and physico-chemical
properties of three new europium Chalcoantimonates, EUSbSC3, EquTe3,
Bao.5Euo,5SbTe3. These compounds are also polychalcogenides as evidenced by the
presence of Q-Q bonds in their structures. For each compound, we will discuss the

presence, type, and, if possible, reﬁnement of the superstructure.

2. Experimental Section

2.1 Synthesis. The following reagents were used as obtained: europium, 99.9%,
250mesh, Alfa Aesar, Ward Hill, MA; antimony, 99.999%, -200 mesh, Cerac,
Milwaukee, WI; selenium shots, 99.9% Noranda Advanced Materials, Saint-Laurent,
Quebec, Canada; tellurium shots, 99.9% Noranda Advanced Materials, Saint-Laurent,

Quebec, Canada; potassium metal, analytical reagent, Spectrum Chemical Mfg. Corp.,

150

Gardena, CA; barium metal, 99%, granules <6mm, Aldrich Chemical Co., Milwaukee,

WI. The starting materials KZSe /K2Te were prepared by a stoichiometric reaction of
potassium metal and selenium/tellurium in liq. NH3. SbQSe3 was prepared by heating a

stoichiometric mixture of antimony and selenitun at 750 °C for 48 hrs.
EquSe3. The compound was prepared from a mixture of 0.0608 g (0.45 mmol) Eu,

0.0961 g (0.2 mmol) Sb28e3, 0.0628 g (0.4 mmol) KzSe and 0.1579g (2.0 mmol) Se. The

reagents were thoroughly mixed, ﬂame-sealed in an evacuated pyrex tube, and heated at

540 °C for 5 days (cooling 2 OC/h). Pure golden black plank-like crystals of EquSe3
were obtained by isolation in degassed dimethylformamide (DMF) and water (yield ~
90% based on Eu metal). The crystals are air- and water stable.

EquTe3. The compound was prepared from a mixture of 0.0304g (0.2 mmol) Eu,
0.0487g (0.4 mmol) Sb, 0.1194 (0.4 mmol) Rb2Te and 0.3062g (2.4 mmol) Te. The
reagents were thoroughly mixed, ﬂame-sealed in an evacuated pyrex tube, and heated at
540 0C for 5 days (cooling 2 °C/h). Pure silver plate-like crystals of EquTe3 were
obtained by isolation in degassed dimethylformamide (DMF) and water (yield ~90%
based on Eu metal). The crystals are air- and water stable.

Bao,5Euo,5SbTe3. The compound was prepared from a mixture of 0.027g (0.2 mmol)
Ba, 0.0304 g (0.2 mmol) Eu, 0.0487 g (0.4 mmol) Sb, and 0.3062 g (2.4 mmol) Te. The
reagents were thoroughly mixed, ﬂame-sealed in an evacuated silica tube, and heated at
850 °C for 5 days (cooling 2 0C/h). Black plate-like crystals of Bao,5Euo_5SbTe3 were
obtained by isolation in degassed dimethylformamide (DMF) and water (yield ~20%

based on Eu metal).

151

The compositions of the materials were analyzed by Scanning Electron
Microscope (SEM)/ Energy Dispersive Spectroscopy (EDS). Homogeneity of each bulk
material was conﬁrmed by comparing the powder X-ray diffraction patterns of the

products against ones calculated using X-ray single crystal data.

2.2 Physical Measurements.

Infrared spectroscopy and UV/V is Spectroscopy. Optical diffuse reﬂectance
measurements were made on the ﬁnely ground sample at room temperature. The
spectrum was recorded in the infrared region (6000 - 400 cm!) with the use of a Nicolet
MAGNA-IR 750 Spectrometer equipped with a Collector Diffuse Reﬂectance of Spectra-
Tech. Inc.

Optical diffuse reﬂectance measurements in the UV/V is region were performed at
room temperature with a Shimadzu UV-3101 PC double-beam, double-monochromator
spectrophotometer (200—2500 nm region). The instrument is equipped with an integrating
sphere and controlled by a personal computer. BaSO4 was used as a 100% reﬂectance
standard for all materials. Samples were prepared by grinding them to a ﬁne powder and
spreading them on a compacted surface of the powdered standard material and preloaded
them into a sample holder. The reﬂectance versus wavelength data were used to estimate
a material's band gap by converting reﬂectance to absorption data as described

previously. 12

152

Magnetic Susceptibility. The magnetic response of each compound was measured
over the range 2-300 K using a MPMS Quantum Design SQUID magnetometer. Sample
was ground to a ﬁne powder to minimize possible anisotropic effects and loaded into
PVC containers. The temperature dependent susceptibility studies were performed at
1000 Gauss. Corrections for the diamagnetism of the sample containers were made by
measuring the magnetic response of the empty container under the same conditions of
temperature and ﬁeld which were measured for the ﬁlled container. The diamagnetic

contribution of every ion to XM was corrected according to Selwood.”

Single Crystal UV/V is Spectroscopy. Optical transmission measurements were
made at room temperature on single crystals using a Hitachi U-6000 microscopic FT
spectrophotometer with an Olympus BH-2 metallurgical microscope over a range of 380-

900nm.

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically, a sample (~20 mg )
of ground crystalline material was sealed in a silica tube under vacuum. A silica tube of

equal mass ﬁlled with A1203 was sealed and placed on the reference side of the detector.

The samples were heated to 800 0C at 5 oC/min, isothermed for 10 min followed by

cooling at «5 oC/min to 50 0C. Residues of the DTA experiments were examined by X-
ray powder diffraction. The stability/reproducibility of the samples was monitored by

running multiple heating/cooling cycles.

153

Raman Spectroscopy. Raman Spectra were recorded on a Holoprobe Raman
spectrograph equipped with a CCD camera detector using 633 nm radiation from a HeNe

laser for excitation. Laser power at the sample was estimated to be about lmW and the

focused laser beam diameter was ca. 10 um. 5 scans were needed to obtain good quality

spectra. The accumulation time of each scan was 5 sec.

Charge-Transport Measurements. DC electric conductivity and thermopower

measurements were made on single crystals of EquTe3. Conductivity measurements

were performed in the usual four-probe geometry with 60- and 25-um gold wires used for

the current and voltage electrodes, respectively. Measurements of the sample cross-
sectional area and voltage probe separation were made with a calibrated binocular
microscope. Conductivity data were obtained with the computer-automated system
described elsewhere.'4 Thermoelectric power measurements were made by using a slow

AC technique”’15 with 60 um gold wires serving to support and conduct heat to the

sample, as well as to measure the voltage across the sample resulting from the applied
temperature gradient. In both measurements, the gold electrodes were held in place on
the sample with a conductive gold paste.

Conductivity specimens were mounted on interchangeable sample holders, and
thermopower specimens were mounted on a ﬁxed sample holder/differential heater.
Mounted samples were placed under vacuum (10'3 Torr) and heated to room temperature
for 2-4 h to cure the gold contacts. For a variable-temperature run, data (conductivity or
thermopower) were acquired during both sample cooling and warming to check

reversibility. The temperature drift rate during an experiment was kept below 1 K/min.

154

Typically, three to four separate variable-temperature runs were carried out for each
sample to ensure reproducibility and stability. At a given temperature, reproducibility

was within 15 %.

2.3 Electron Diffraction Study (TEM).

Electron crystallographic studies were carried out on a JEOL lOOCX
Transmission Electron Microscope (TEM) using an electron beam generated by a CeB6
ﬁlament and an acceleration voltage of 120 kV. After the samples were ground to a ﬁne
powder in acetone, the specimens were prepared by dipping a carbon coated copper grid

in the suspension. The samples showed no decomposition under the e-beam .

2.4 X-ray Crystallography.

EquSe; A single crystal with dimensions 0.10 x 0.08 x 0.01 mm was mounted on
the tip of a glass ﬁber and intensity data were collected on a Bruker SMART Platform

CCD diffractometer using graphite monochromatized Mo Koc radiation. The data were
collected at room temperature over a full sphere of reciprocal space, up to 24.98° in 0.

The individual frames were measured with an omega rotation of 0.3 deg and an
acquisition time of 60 sec. The SMART'6 software was used for the data acquisition and
SAINTl7 for the data extraction and reduction. The absorption correction was performed
using SADABS.18 Structure solution and reﬁnement were performed with the SHELXTL
package of crystallographic programs.19 Systematic absence conditions of the data set

gave four possible space groups, Cmmm, Cmm2, Amm2, C222. The structure was solved

155

and reﬁned successfully in the centrosymmetric Cmmm space group. Atom Se(4) was
disordered over two crystallographically equivalent sites. The distance between these
atoms is 0.80(2)A, which is too close to allow both atoms to coexist at the same time.
Therefore, the maximum occupancy allowed for this atom was 50%. Such positional
disorder can be a clue for the presence of a superstructure. Careful examination of the
data frames for weak reﬂections revealed a superstructure with asuper = 8 x asub. We
reintegrated the intensity data according to the new cell and attempted to solve the
superstructure, even though the added reﬂections were very weak and the statistics poor
to give a decent reﬁnement.20 This reﬁnement showed that the 8-fold superstructure
could not lift the disorder of the Se(4) atoms in the net and additionally generated a
complicated disorder in the Sb sites, suggesting the possibility of an even larger
supercell. This was beyond the capability of our x-ray equipment and thus we resorted to
the electron diffraction study(TEM) of this compound.

EquTeg A Single crystal with dimensions 0.13 x 0.10 x 0.01 mm was mounted on
the tip of a glass ﬁber and intensity data were collected as described above. The
individual frames were measured with an omega rotation of 0.3 deg and an acquisition
time of 70 sec. The data were processed as described above. The structure of EquTe3
was solved and reﬁned in the centrosymmetric Pmmn space group. Rotational
photographs showed additional superlattice reﬂections, which were not readily indexable
as simple multiplets of the substructure reﬂections. They seems to lie between 1/6 a* and
1/7a"'. It is possible that these are satellite peaks resulting from an incommensurate
charge density wave state. We attempted to more accurately ﬁnd the modulation vector

by electron diffraction but this was unsuccessful because the three-dimensional

156

morphology of the compound made it difﬁcult to orient specimens along a proper zone-
axis.

Bao,5Euo,5SbTe3 A Single crystal with dimensions 0.16 x 0.08 x 0.04 mm was
mounted on the tip of a glass ﬁber and intensity data were collected as described above.
The individual frames were measured with an omega rotation of 0.3 deg and an
acquisition time of 65 sec. The data were processed as described above. The cell
parameters and observed diffraction symmetry indicate that this compound is

isostructural with EquTe3. However, weak superstructure reﬂections, which makes the

b-axis 6-times longer, were additionally observed. These superlattice reﬂections were

prominent enough to provide a decent intensity data set. The superstructure of

Bao.5Euo,5SbTe3 was solved and reﬁned successfully in the monoclinic P2 space group.
The relationship between the subcell parameters a,b,c and the supercell parameters
a’,b’,c’ are as follows.
a’ = a b’ = c c’ = 6b

The complete data collection parameters and details of the structure solution and
reﬁnement for each compound are given in Tables 5.1 and 5.2. The coordinates of all
atoms, isotropic temperature factors, and their estimated standard deviations (esd’s) for
each compound are given in Tables 5.3-5.6. The selected bond distances for each

compound are given in Tables 5.7-5.10.

3. Results and Discussion

157

Table 5.1.. Summary of Crystallographic Data and Structural Analysis for EquSe3 and

 

 

EquTeg.

Formula EquSe; EquTe;

Formula weight 501.59 656.51

Crystal habit golden black planks silver square plates

Space group Cmmm Pmmn

a, A 4.342(3) 4.6034(9)

b, A 2982(2) 4.4561(9)

c, A 4.294(3) 15.932(3)

z; v, A3 4, 556.0(7) 2, 326.81(11)

Dcalc. g/crn3 6.099 6.671

Temp, K 293(2) 293(2)

K(Mo 1(a)./i 0.71069 0.71069

MMO K00, cm" 355.01 266.21

F(000) 864 540

6 max , deg 24.98 25.00

Total data measd 2272 2653

Unique data 327 [Rim = 0.0910] 374 [Rim = 0.1079]

No. of variables 25 22

Reﬁnement method F ull-matrix least-squares on F 2

R indices [I>29] R1 = 0.0586 R1 = 0.0684
wR2=0.1470 wR2=0.1719

R indices (all data) R1 = 0.0595 R1 = 0.0685
wR2 = 0.1476 wR2 = 0.1720

Goodness of ﬁt on F2 1.149 1.192

 

3 R1: ZIIFOI '— IFCH/ZIFOI ;

b WR2= {Z[W( F02 _FC2)2 ]/Z[W(F02)2 ]}l/2

158

Table 5.2.. Summary of Crystallographic Data and Structural Analysis for

 

 

3210.5511053bTe3-
Substructure Sycrstructure
Formula Bao,5Euo.5SbTe3 Ba0,5Euo,5SbTe3
Formula weight 649.24 649.24
Crystal habit black irregular shaped crystals
Space group Pmmn Pm
a, A 4.6036(9) 4.6036(9)
b, A 4.6834(9) 28.044(6)
c, A 16.186(3) 16.186(3)
B. deg 90.0 9037(3)
z; v, A3 2, 348.97(12) 12, 2089.6(7)
Dcalc. g/cm3 6.191 6.191
Temp, K 293(2) 293(2)
MMo K00. A 0.71073 0.71073
“(M0 K00, cm" 232.83 233.38
F(000) 533 3198
9 max . deg 24.98 25.00
Total data measd 2623 15754
Unique data 395 [Rim = 0.0754] 7184 [Rim = 0.0563]
No. of variables 24 291
Reﬁnement method Full-matrix least-squares on F2
R indices [I>20] R1 = 0.0657 R1 = 0.0878
wR2 = 0.1540 wR2 = 0.2544
R indices (all data) R1 = 0.0660 R1 = 0.1092
wR2 = 0.1541 wR2 = 0.2780
Goodness of ﬁt on F 2 1.363 1.092
BASF N/A 053(6)

 

aR1=z||Fo|-|F.n/2|Fo| ;

b wR2= {2[w(F.;’- -F£ )2 ]/2[w(F02)2 ]}"2

159

Table 5.3. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for EquSe3 with Estimated Standard Deviations in Parentheses.

 

 

Atom x y z Um,a A2 Occ.
Eu 0 0.0827(1) 0.5 0.019(1) 1
Sb 0 0.3058(1) 0 0.082(2) l

Se(l) 0 0.1929(1) 0.5 0.047(1) 1

8e(2) 0 0.3911(1) 0 0.016(1) 1

Se(3) 0.5 0 0.5 0.043(2) 1

36(4) -0.0920(17) 0 0 0.038(2) 0.50

 

a ch is deﬁned as one third of the trace of the orthogonalized Uij tensor.

160

Table 5.4. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for EquTe; with Estimated Standard Deviations in Parentheses.

 

 

Atom 1 x y z Um,”I A2
Eu 0.25 0.25 0.1676(1) 0.025(1)
Sb 0.75 0.75 0.3906(2) 0.031(1)

Te(l) 0.25 0.25 0.3909(2) 0.067(2)

Te(2) 0.75 0.75 0.2142(2) 0.026(1)

Te(3) 0.75 0.25 0.0001(2) 0.032(1)

 

a .q is deﬁned as one third of the trace of the orthogonalized U), tensor.

161

Table 5.5. Fractional Atomic Coordinates and Equivalent Atomic Displacement
Parameter (Ueq) Values for Bao,5Euo,5SbTe3 (substructure) with Estimated Standard

Deviations in Parentheses.

 

 

Atom x y z Ueq,a A2
Eu/Ba 0.25 0.25 0.3296(1) 0.022(1)

Sb 0.75 0.75 0.1087(2) 0.043(1)
Te(l) 0.25 0.25 , 0.1056(2) 0.086(2)
Te(2) 0.75 0.75 0.2813(2) 0.030(1)
Te(3) 0.25 0.75 0.5003(2) 0.042(1)

 

3 eq is deﬁned as one third of the trace of the orthogonalized U), tensor.

162

Table 5.6. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Um) Values for Bao.5Euo_5SbTe3 (supertructure) with Estimated Standard

Deviations in Parentheses.

 

 

Atom x y z Um,a A2 Occ.
Eu(1) -0.0268(5) 0.0829(1) 0.1554(2) 0.013(1) 1
Eu(2) 0.0191(5) 0.2480(1) 0.1627(2) 0.017(1) 1
Eu(3)/Ba(3') 0.9712(6) 0.4154(1) 0.1661(2) 0.018(1) 0.78(7)/0.22(7)
Ba(1)/Eu(1') 0.5251(8) 0 0.8114(2) 0.015(1) 0.78(9)/0.22(9)
Ba(2) 0.5396(7) 0.1681(1) 0.8165(2) 0.020(1) 1
Ba(3) 0.5380(7) 0.3348(1) 0.8270(2) 0.022(1) 1
Ba(4)/Eu(4') 0.5266(10) 0.5000 0.8286(3) 0.028(2) 0.85(9)/0. 15(9)
Sb(l) 0.4572(14) 0 0.3725(3) 0.033(1) 1
Sb(2) 0.4645(10) 0.1646(1) 0.3794(3) 0.037(1) 1
Sb(3) 0.4689(11) 0.3337(2) 0.3856(3) 0.043(1) 1
Sb(4) 0.4572(16) 0.5000 0.3891(4) 0.054(2) 1
Sb(S) 0.0410(7) 0.0799(1) 0.5942(2) 0.012(1) 1
Sb(6) 0.0254(10) 0.2484(2) 0.5989(3) 0.038(1) 1
Sb(7) -0.0158(11) 0.4159(1) 0.6055(3) 0.043(1) 1
Te(l) 0.4681(8) 0 0.2014(2) 0.009(1) 1
Te(2) -0.0060(1 1) 0.0794(1) 0.3774(3) 0.037(1) 1
Te(3) 0.4715(7) 0.1649(1) 0.2071(2) 0.015(1) 1
Te(4) -0.0067(12) 0.2388(1) 0.3842(2) 0.042(1) 1
Te(5) 0.4715(7) 0.3311(1) 0.2135(2) 0.018(1) 1
Te(6) -0.084(l7) 0.4102(2) 0.3901(3) 0.085(2) 1
Te(7) 0.4692(9) 0.5000 0.2164(3) 0.016(1) 1
Te(8) 0.5516(15) 0 0.5877(3) 0.041(2) 1
Te(9) -0.0344(8) 0.0841(1) 0.7655(2) 0.025(1) 1
Te(10) 0.4588(13) 0.1794(2) 0.5944(3) 0.064(2) 1
Te(l 1) 0.0427(7) 0.2527(1) 0.7733(2) 0.024(1) 1
Te(12) 0.4347(11) 0.3423(1) 0.5977(2) 0.035(1) 1
Te(13) -0.0434(9) 0.4177(1) 0.7803(3) 0.027(1) 1
Te(14) 0.4396(15) 0.5000 0.5999(3) 0.037(1) 1
Te(15) 0.0125(10) 0 0.0186(2) 0.014(1) 1
Te(16) 0.4444(8) 0.0809(1) 0.0174(2) 0.023(1) 1
Te(17) 0.0564(9) 0.1694(1) 0.0147(2) 0.025(1) 1

 

163

Table 5.6. (Cont’d).

Te(18) 0.4826(9) 0.2499(1) 0.0090(2) 0.013(1) 1
Te(19) 0.0548(9) 0.3303(1) 0.0062(2) 0.027(1) 1
Te(20) 0.4413(9) 0.4196(1) 0.0044(2) 0.028(1) 1
Te(21) 0.0164(11) 0.5000 0.0040(3) 0.022(1) 1

 

8‘ ch is deﬁned as one third of the trace of the orthogonalized Ug- tensor.

164

Table 5.7. Selected bond lengths (A) and angles (°) for EquSe3.

 

 

 

 

Bond Lengths
Eu-Se(1) 3.289(5) Sb-Se(1) 3.0535(16) x 4
Eu-Se(2) 3.1512(18) x 4 Sb-Se(2) 2.543(5)
Eu-Se(3) 3.285(2) x 2 8e(3)-8e(4) 2.783(5)
Eu-Se(4) 3.293(2) x 2 Se(3)-8e(4) 3.349(6)
Bond Angles
Se(2)-Eu-Se(2) 151.24(12) Se(2)—Sb-Se(1) 89.31(10)
Se(2)—Eu-Se(2) 8706(6) Se(1)-Sb-Se(1) 8935(6)
Se(2)-Eu-Se(3) 129.91(6) Se(4)-Se(4)-Se(3) 129.52(12)
Se(2)-Eu—Se(3) 74.40(6) Se(3)-Se(4)-Se(3) 101.0(2)
Se(3)-Eu-Se(3) 8274(7)
Se(2)-Eu-Se(l) 7562(6)
Se(2)-Eu-Se(4) 7002(1 1)

 

165

Table 5.8. Selected bond lengths (A) and angles (°) for EquTe3.

 

 

 

 

Bond Lengths
Eu-Te(1) 3.557(4) Sb-Te(1) 3.2034(5) x 4
Eu-Te(2) 3.2883(9) x 4 Sb-Te(1) 3.482(4)
Eu-Te(3) 3.476(3) x 2 Sb-Te(2) 2.811(4)
Eu-Te(3) 3.527(3) x 2 Te(3)-Te(3) 3.2034(4)
Bond Angles
Te(2)-Eu-Te(2) 1 53.90( 12) Te(2)—Sb-Te( 1) 90.07(8)
Te(2)-Eu-Te(2) 8885(3) Te(1)-Sb-Te(1) 88.137( 16)
Te(2)-Eu-Te(3) 127.42(6) Te(l)-Sb-Te(1) 179.87(17)
Te(2)-Eu-Te(3) 7487(5) Te(3)-Te(3)-Te(3) 179.9(2)
Te(2)-Eu-Te(3) 73 39(5) Te(3)-Te(3)-Te(3) 91.863(16)
Te(3)-Eu-Te(3) 7973(8)
Te(2)-Eu-Te(l) 7695(6)

 

166

Table 5.9. Selected bond lengths (A) and angles (°) for BansEunsSbTe; (substructure).

 

 

 

 

Bond Lengths
Eu-Te(1) 3.625(4) Sb-Te(1) 3.2840(5) x 4
Eu-Te(2) 3.3752(9) x 4 Sb-Te(1) 3.468(4)
Eu-Te(3) 3.589(3) x 2 Sb-Te(2) 2.793(4)
Ell-Te(3) 3.622(3) x 2 Te(3)-Te(3) 3.2836(4)
Bond Angles
Te(2)-Eu-Te(2) 153.23(1 1) Te(2)-Sb-Te(l ) 9089(9)
Te(2)-Eu—Te(2) 8786(3) Te(1)-Sb-Te(l) 90.971(16)
Te(2)-Eu-Te(3) 127.95(6) Te(1)-Sb-Te(1) 178.22(18)
Te(2)-Eu-Te(3) 7495(5) Te(3)-Te(3)-Te(3) 179.7(2)
Te(2)-Eu-Te(3) 7422(5) Te(3)-Te(3)-Te(3) 90.984(16)
Te(3)-Eu-Te(3) 7978(8)
Te(2)-Eu-Te(1) 7662(6)

 

167

Table 5.10. Selected bond lengths (A) and angles (°) for Ba0,5Eu0.5SbTe3

(superstructure).

 

Eu( 1 )-Te( 1 )
Eu( 1 )-Te(3)
Eu( 1 )-Te(3)
Eu( 1 )-Te(1 )
Eu(1)-Te(16)
Eu( 1 )-Te(2)
Eu(1)-Te(1 5)
Eu(l)-Te(17)
Eu( 1 )-Te(1 6)

Eu(3)-Te(5)
Eu(3)-Te(5)
Eu(3)-Te(7)
Eu(3)-Te(7)
Eu(3)-Te(20)
Eu(3)-Te(6)
Eu(3)-Te(21)
Eu(3)-Te(19)
Eu(3)-Te(20)

Ba(2)-Te(9)

Ba(2)-Te(1 1)
Ba(2)-Te(1 1)
Ba(2)-Te(9)

Ba(2)-Te(l7)
Ba(2)-Te(10)
Ba(2)-Te(l7)
Ba(2)-Te( 1 6)
Ba(2)-Te(18)

Ba(4)-Te(1 3)
Ba(4)-Te(l 3)
Ba(4)-Te(1 3)
Ba(4)-Te( 1 3)
Ba(4LTe(2 1)

3.336(3)
3.351(3)
3.369(3)
3.375(3)
3.551(5)
3.596(6)
3.658(4)
3.671(4)
3.696(5)

3.384(3)
3.389(4)
3.396(3)
3.415(4)
3.520(6)
3.630(7)
3.641(5)
3.671(5)
3.673(6)

3.376(4)
3.371(4)
3.385(4)
3.415(4)
3.513(6)
3.609(6)
3.635(5)
3.637(4)
3.646(5)

3.303(5)
3.303(5)
3.403(4)
3.403(4)
3.441(8)

Eu(2)-Te(3)
Eu(2)-Te(5)
Eu(2)-Te(3)
Eu(2)-Te(5)
Eu(2)-Te(1 8)
Eu(2)-Te( 1 9)
Eu(2)—Te(4)
Eu(2)-Te(1 7)
Eu(2)-Te(1 8)

Ba( 1 )-Te(9)
Ba(1)-Te(9)
Ba( 1 )-Te(9)
Ba( 1 )-Te(9)
Ba(l)-Te(15)
Ba(1)-Te(16)
Ba(l)-Te(16)
Ba(1)-Te(8)
Ba(l)-Te(15)

Ba(3)-Te( l 3)
Ba(3)-Te(l 1)
Ba(3)-Te(13)
Ba(3)-Te(1 1)
Ba(3)—Te(19)
Ba(3)-Te(18)
Ba(3)-Te(1 9)
Ba(3)-Te(20)
Ba(3)-Te(12)

Sb(l)-Te(1)
8b(1)-Te(2)
Sb(1)-Te(2)
8b(1)-Te(2)
8b(1)-Te(2)

168

3.321(4)
3.344(4)
3.387(4)
3.410(4)
3.494(5)
3.580(5)
3.595(6)
3.623(5)
3.731(5)

3.353(4)
3.353(4)
3.405(4)
3.405(4)
3.489(6)
3.584(4)
3.584(4)
3.641(7)
3.688(6)

3.343(5)
3.360(5)
3.371(5)
3.382(4)
3.490(5)
3.574(4)
3.613(6)
3.621(5)
3.718(6)

2.770(6)
3.084(6)
3.084(6)
3.327(7)
3.327(7)

Table 5.10. (Cont’d).

Ba(4)-Te(20) 3.522(5) Sb(l)-Te(8) 3.507(7)
Ba(4)-Te(20) 3.522(5)

Ba(4)-Te(21) 3.674(7)

Ba(4)-Te(14) 3.705(7)

Sb(2)-Te(3) 2.790(6) Sb(3)-Te(5) 2.786(7)
Sb(2)-Te(4) 3.007(6) Sb(3)-Te(6) 3.072(9)
Sb(2)-Te(4) 3.203(7) Sb(3)-Te(6) 3.224(9)
Sb(2)-Te(2) 3.226(6) Sb(3)-Te(4) 3.445(7)
Sb(2)-Te(2) 3.415(6) Sb(3)-Te(12) 3.446(7)
Sb(2)-Te(10) 3.504(7) Sb(3)-Te(4) 3.592(6)
Sb(4)-Te(7) 2.797(8) Sb(5)-Te(9) 2.776(6)
Sb(4)-Te(6) 3.308(9) Sb(5)-Te(8) 2.924(5)
Sb(4)-Te(6) 3.308(9) Sb(5)-Te(2) 3.513(6)
Sb(4)-Te(14) 3.412(8) Sb(5)-Te(8) 3.533(6)
Sb(4)-Te(6) 3.521(9) Sb(5)-Te(10) 3.617(7)
Sb(4)-Te(6) 3.521(9) Sb(5)-Te(10) 3.616(5)
Sb(6)-Te(1 1) 2.827(7) Sb(7)-Te(13) 2.832(8)
Sb(6)-Te(10) 2.953(8) Sb(7)-Te(12) 2.929(6)
Sb(6)-Te(10) 3.063(8) Sb(7)-Te(14) 3.158(7)
Sb(6)-Te(12) 3.380(7) Sb(7)-Te(12) 3.266(7)
Sb(6)-Te(4) 3.486(6) Sb(7)-Te(l4) 3.444(7)
Sb(6)-Te(12) 3.621(6) Sb(7)-Te(6) 3.490(8)
Te(15)-Te(l6) 3.016(4) Te(18)-Te(19) 2.994(4)
Te(15)-Te(16) 3.462(4) Te(18)-Te(19) 3.468(4)
Te(16)-Te(l7) 3.389(4) Te(19)-Te(20) 3.391(4)
Te(17)-Te(18) 2.992(4) Te(19)-Te(20) 3.415(4)
Te(17)-Te(16) 3.384(4) Te(20)-Te(21) 2.984(4)
Te(17)-Te(18) 3.475(4) Te(20)-Te(21) 3.477(5)

 

169

~ Structures.
EquSe; EquSe3 presents a novel two-dimensional structure, which is composed
of (SbSe2)' layers sandwiching a ﬂat (Se2)2' layer, see Figure 5.1. The Eu atoms are
stabilized between these sub-layers in mono-capped square anti-prismatic sites, see
Figure 5.2(a). The local environment of the Sb3+ ions is square pyramidal, see Figure
5.2(b). The SbSe5 square pyramids share edges of their square bases to build a (SbSe2)'
layer. Adjacent (SbSe2)' layers are separated by an apparent van der Waals gap. The
Sb3+ ions in this coordination mode can express their 582 lone pairs below the pyramid’s
square plane and into the van der Waals gap.

The second type of layer, the ﬂat Se net, is chemically more interesting, see
Figure 5.2(c). The closest Se-Se distance in the net is 2.783(5)A, which is signiﬁcantly
longer than Se-Se distance found in isolated diselenide group (i.e. 2.34 A in in ZrSe32‘).
Previous studies on compounds with Se nets show that these nets usually undergo more
severe distortions than the Te nets in corresponding telluride compounds. This is due to
the less diffuse nature of the orbitals of Se, which favor more localized Se-Se
interactions. The most commonly observed distortion pattern of a Se net is the
dimerization. For example, in LaSez, the Se net with equal Se-Se distance of 2.8A is
distorted to generate Sezz' dimers, Se-Se bond of which gets shortened to 2.45 A The
three remaining Se-Se distances range from 3.1 A to 3.4 A.

The Se net in EquSe3 shows symptoms of a modulation in its plane. The Se(4)
atom in the net deviates from its ideal position, which is on the mirror plane, generating a
symmetrically equivalent position opposite the mirror plane. Therefore, the Se(4) atoms

are 50/50 disordered over two sites, distance of which is 0.82(2)A. Consequently, the

170

Se4 Se3

 

.e. .12 r

n e e

w m ma

U 1.. .1. V g
E _ 11.
1.11116 1 ls. 4.. s. I, Tell. 16
...Q .0 s a! 8...! LEE ....ﬂ.§v4@.
\; .1" .D “. 11M? W1 .|' . F. . n.
V s. ____ s .._. s 4.08.“.5. ....Ie.§s
...W 6.. ...Vﬁ‘... 1.. 881111.“...
.mV\ ... _,___ ﬁ_-\&ew ARE“: ._..R§V4&m
I. .mei with“) r 1 ..

“1 sale. Lﬂ.s%w§.wl...q gunk»
ﬁwlww I ..l ",4lb ..‘I .4"? I. II PM”.
...19 0814.“. ii...
. ...l . 1.th WW .0 3|. __ o
m ..lMIQWNﬂalgmx
«W1. a e”..- 54045.1“... ...WQ 4a....
..... ...iu 4...“. 1100.111 .. eh ,3L6m\v..ve
..7 s @6416“. .me,
..uW. 1 . ___ L._.I|oek\»aw.\w.vm§ 5...... hw‘ﬂmm
«V \V... . 0.141%me ,. «.0... at. «ekm...
. l \. I; I I”; n _. v ...i k
w”. .0. “oak pe‘ cool H ““13.

171

Figure 5.1. The structure of EquSe3 viewed down the a—axis.

(b)

892

Se1 @861
Set Sb 591

 

(C) 50/50 disorder 2.783(5)A 3.349(6)A

 

 

 

Figure 5.2. The local environment of the (a) Eu atom and (b) Sb atom. (c) The Se net in
the substructure of EquSe3.

172

Se(4) atoms do not possess four Se-Se bonds of equal length and the Sc net is no longer
ideal. This apparent positional disorder strongly suggest a modulation in the net, which
could result in a superstructure. In fact, the same disordering pattern of the Se net was
observed in the substructure of Rbo_33DySez,67, which possess a 4a x 3c superstructure
(the net lies perpendicular to the b-axis). However, the superstructure of Rbo_33DySe2,67
could not be reﬁned due to the extremely weak supercell reﬂections.

We search for a superstructure in EquSe; with transmission electron microscopy
(TEM). Indeed, electron diffraction studies revealed a very peculiar and complicated
pattern of superlattice reﬂections both along the a*- and c*-axes, see Figure 5.3. The
positions of these reﬂections and the resulting superlattices are more clearly shown in the
cartoon schematic of an electron diffraction pattern for one unit-cell (asub x csub), see
Figure 5.3. Along the a-axis, EquSe3 possesses a 8-fold superstructure, which agrees
with our observations with the x-ray diffraction study. The reﬂections that occur at 1/4
asub‘, which makes a 4-fold superstructure along the a-axis are much stronger than those
occurring at 1/ 8 asub‘. The superstructure along the c-axis is even more complicated and
possibly incommensurate. The distance between the closest reﬂections that are projected
onto the c-axis, implies that there exists a 10-fold superstructure along the c-axis.
However, there exist extremely weak reﬂections, which are only recognizable on the
original negative photographs of the diffraction pattern. These cannot be indexed as
simple multiplets of the substructure reﬂections. This complicated superstructure of
EquSe; explains why the reﬁnement of the 8-fold superstructure along the a-axis could

not lift the disorder of the Se net and generated a very complex Sb disorder.

173

 

 

.o
3
(I)

a:
O

E

,—

2/ 5 c1"sub

c*sub

1/10 C*sub H

 

 

 

 

V

0 1/8 a*sub

H 1/4 a*sub

 

*
= ;asub

Figure 5.3. Selected area electron diffraction pattern of EquSe3 with the beam
perpendicular to the layers (parallel to the [010] direction). Schematic of the
electron diffraction pattern corresponding to one unit cell (asub x csub) is shown
below.

174

EquTe3 The overall structure of EquTegdies on s showthe b-axis, is shown in
Figure 5.4. The structure is closely related to that of EquSe3 in that they have the same
structural motifs that build the whole framework; [SbQ2]' and ﬂat (Q2)2' layers with the
Eu2+ metals stabilized between these layers (Q = Se, Te). The mainses usually
undergomore sevetwo consecutive [SbQ2]’ layers stackresthe Te nets in . This is due
tonature. In EquSe3, the Sb atoms in the upper layer are staggered with the Sb or Se
atoms in the lower layer. Therefore, Sb is truly 5-coord nate and there exist Van der
Waals gap between two [SbSez] layers, see Figure 5.4(a). On the other hand, in EquTe3
the Sb atoms in the upper layer and the Te atoms in the lower layer are eclipsed, so Sb
makes the 6‘hof the ,which favo Te r morfrom the lower layer. This stacking arrangement
makes the Sb atom 6-coordinate and fuse the two [SbTe2]'elayers, eliminating van der
Waals gap, which makes the framework three-dimensional. The local environment of Sb
in EquTe3 is, therefore, distorted octahedral with the 6th bond signiﬁcantly longer
(3.482(4) A) than the other ﬁve(3.125A on average). Therefore, the [szTe4]2' layers are
built from edge-shared SbTe6 octahedra.

The (Te2)2' layer makes a square nets with equal Te-Te distances of 3.2034(4)A,
which are again longer than the normal Te-Te bond distance found in elemental Te
(2.83A) 22 or in the isolated ditelluride bonds, (Te2)2' of Rb2Te; (2.78A) 23 and ZrTe3
(2.76A)24 , see Figure 5.5(b). Unlike the Se net in EquSe3 the Te atoms in the net do not
show any positional disorder. However, the Te atoms do exhibit anolamous behavior in
their anisotropic displacement parameters (U l l=0.012(1) A2, U22=0.052(2) A2,
U33=0.033(2) A2). Unfortunately, due to its incommensurate nature of the

superstructure, no further work could been attempted.

175

 
 
 

      
 
 
 

"6&1'51’8‘1'3'

o ‘\\ 0 ‘:§ 9 \.'s ‘ \ s
— .‘J _ ‘3- —‘\._-:=—| , _
E =1 ‘—:’| V

..p—

/. ‘\ gs“. ~.‘ / K
.1 \ r \'; ,/. \ID

x. 41W 'in '11

4|
— O. i- ’01,, ,,-_— '2:-

ﬁli— “jl "

_d

 
 

 

 

Te1

-l

il'i

i——

‘1
I),
111‘

11
’0

l

    

 

 

 

J
w
1

 

   
   
   

“
\\

l/-.
\

1
.1,—

1
.J
l

\

1.11

“‘/
,/

11;
1’1
.11
11!
l

.4.

 
 
 
 
  
 

  

\‘1 ﬁf'iﬁliﬁu

  

(9.

7’.
_\
,=\

I
O
\1

’/

l
l/
a
l\

\

l
.9

  
         

$.15 ._. —- =Q== =Q= =1.

~[SbTe212' layer

«P.
11_
llui'l
. ._-.’
Ill-W.
g.
_ l
,—
Y3
.= l
' 1
=0

 
 

‘ - ‘V - ‘ h .-
.-- ...e. e» -—= «- = -—=

 
 

 

1
73‘
k

s
1 i

6
”I
6
1;]
’71-
1.3V
'i '1

51
i,

1:1
14:4,.
. 1-1
(all

g.
l
g

7

   
 
  
   

          

 
    

  
    

.r -g 'r ---'/--1 ..»c-.-' \A
.5. - ’05? .‘> I; _ /5I<
ﬁbO—4’0<‘:>Q

Figure 5.4. The structure of EquTeg viewed down the b-axis.

176

 

 

[SbSe2]' layers in EquSea

‘b’ VV
>2/

 

 

 

b

L, \
3.203A
a

Figure 5.5. (21) Comparison of the [Sb2Q4]2’ layers in EquTe3 and EquSeg. (b) The

square Te nets in EquTeg.

 

  

177

Bao,5Euo,sSbTe3 This compound is derived from EquTe; by replacing half the Eu
atoms with Ba atoms. Therefore, BaosEuoijTe; is isostructural with EquTe; with
50/50 disorder between Ba2+ and Eu2+ ions. Due to the larger size of Ba compared with
Eu, the cell parameters of Bao,5Euo,5SbTe3 are slightly increased. Consequently, most of
the bond distances found in Bao.5Euo.5SbTe3 are slightly elongated than the corresponding
distances in EquTe3. For example, the Te-Te distances in the square Te net are
3.2836(5)A, which is 0.08A longer than that in EquTe3.

Unlike EquTe3, BaQSEUQSSbTejg possesses a commensurate 6-fold superstructure
along the b-axis. We could able to successfully reﬁne the superstructure, which revealed
interesting detailed features of the compound. The overall structure of BaojEuoijTe; is
shown in Figure 5.6. Interestingly, The Ba atom is not statistically disordered with the
Eu atom in the superstructure. Instead, they are found segregated in alternate layers.
This arrangement of Ba and Eu allows better close packing of the [szTe4]2' and [Te2]2'
layers. It avoids scattering the larger Ba ions over every layer, which generates
unnecessary space between these sub-layers. The ordering between Ba and Eu, however,
is not perfect and some of the Ba and Eu sites are slightly “contaminated” with the other
kind in different proportions, see Table 5.6. For example, among four different Ba sites,
the Ba(l) and Ba(4) sites contain 22% and 15% of Eu atoms, repectively, whereas the
Ba(2) and Ba(3) sites are fully occupied with Ba atoms. The ordering between fully
occupied Ba/Eu sites and disordered sites along the c’-axis seems to be one of the
components that causes the 6-fold superstructure.

The superstructure also allows different degrees of distortion for each SbTe6

octahedron. In the substructure, there is only one crystallographically independent Sb

178

o::o:: on 0:: o: g» 0:: a: e» or: or. —o‘) O>—0 0

EU2 Eu3

T91 Eu1 T93 . O .
ﬂ . ‘Te . T97 .

’II 0" O» o» .v 0) .
..9' 3.92; 3b,, IbSeGl b4 “ I

  

- _. e- _ ~—- ~—-“ —“
Te15 9 e19 T921

 

Figure 5.6. The superstructure of Bao,5Euo_5SbTe3 viewed down the a-axis. The bonds

between Eu and Te are omitted for clarity.

179

site and, therefore, giving a single local environment. In the superstructure there are 7
crystallographically different Sb sites, which signiﬁcantly deviate from their original sites
showing different distortions, see Figure 5.6.

The Te net in the superstructure clearly shows modulations of Te-Te bonds, see
Figure 5.7. It oligomerizes into trimers, which more accurately presents the positions of
the Te atoms in the net. The distance of Te-Te bonds in the trimers (<3.016(4)A) are
much shorter than the Te-Te distances between these trimers (>3.384(4)A), showing a
distinct gap between these interactions. These trimers are arranged so that they all point
in one direction (i.e. along the a’axis), which makes the superstructure polar in contrast to
the centrosymmetric substructure with the perfect Te net.

The average structure of EquTe3 and BaasEuoijTeg are isostructural with
SrBiTe3, which was ﬁrst reported in the noncentrosymmetric Pmm2, space group.”
Recently, reinvestigation of the structure of SrBiTe; revealed that the compound can be
successfully reﬁned in the centrosymmetric space group Pmmn and it also possesses a 6a
x 2c superstructure due to the distortion of the Te net.26 The distance of Te-Te bonds in
the net varies from 3.121 A to 3.359 A instead of one equal distance of 3.24 A in the
substructure. If a Te-Te bond is drawn between Te atoms which have shorter interatomic
distances than 3.2 A, Te atoms form an inﬁnite stare-cased chain, see Figure 5.8. The
reason why the c-axis of the substructure is doubled in the superstructure is also related
with the Te nets. When the Te net stacks along the c-axis, every other layer of Te net is

shifted by 1/2 a’ along the a’- axis thus doubling the repeating unit along the c-axis.

180

(a) Substructure

V

T93 /

 

 

 

 

 

(b) Superstructure

x O\
a; as (j;
T616 T918 TeC20\U/O 3.46‘8/1 3;
a, 0x0» 3\O/€
(\{O O\ /O 2.994 2.992 \0/3

0

 

 

C!

Figure 5.7. Comparison of Te nets in the (a) substructure and the (b) superstructure.
Selected Te-Te distances (A) are shown in the ﬁgure.

181

re..58

Dis

 

 

oteeth fdthpruurfSBT

182

 

Properties.

The diffuse reﬂectance optical spectrum of EquSe; was taken in the UV-Visible
region. An abrupt optical gap is observed at 0.97eV, indicating that the distortion of the
Se net in the superstructure must be signiﬁcant enough to completely open a gap at the
Fermi level, see Figure 5.9(a). The optical properties of EquTe3 and Bao,5Euo,5SbTe3
were determined by measuring the diffuse reﬂectance spectrum of each in the Mid IR
region (400-60000m’1), see Figure 5.9(b). These spectra share very similar features with
each other and show optical gaps at the same energy of 0.26 eV, indicating that these
materials are narrow gap semiconductors. The observation of an energy gap in these
compounds indicates absence of states at the Fermi level and is consistent with the
observed distortions in their Te nets.

The electrical conductivity of EquTe3 was measured on single crystal samples.
The compound shows a moderate room temperature conductivity value of ~10 S/cm with
a rather peculiar temperature dependence, which does not conform to either metallic or
semiconducting. Below 80K the conductivity decreases to low values suggesting strong
carrier localization. EquTe3 shows quite a large positive Seebeck coefﬁcient value of

335 uV/K at room temperature. The positive sign indicates that charge carriers are holes

(p-type). The Seebeck coefﬁcient increases almost linearly with increasing temperature,
a behavior typically seen in narrow gap semiconductors, see Figure 5.10(b).
Magnetic susceptibility measurements of EquSe3 shows antiferromagnetic

coupling transition at 5.1K. Above this temperature, l/x shows linear dependence on the

temperature following Curie-Weiss law. The Herr calculated from the slope of the line is

183

7.79 B.M. (with Weiss constant 6 = -O.95K.), which is in good agreement with the

thoretical um value of a free ion Eu2+ (7.94 B.M.) with f7 conﬁguration.27 The

magnetization in this material shows linear dependence as a function of ﬁeld in the range
of O<G< 10000.

The Raman spectrum of EquSe3 is shown in Figure 5.12(a). The strong shift at
215cm'l could be due to Sb-Se stretching vibrations, considering that typically highest
energy shifts for Sb-Se stretches usually appear around ~220-230 cm'U’?’8 This strong
Raman shift generates an overtone at 434 cm]. The tiny bumps at 240 cm], 245 cm",
and 255 cm",which are buried under the right side tail of the Sb-Se band, are due to the
various Se-Se stretching vibrational modes in the net. These values are in accord with the
Se-Se stretching frequencies reported for other compounds (i.e.~253 cm'lfor K28e2,24
252cm'1 for KzGdzsbzseg,‘ 266 cm" for KUzsbSegzs). The Raman spectra of EquTe3
and Bao,5Euo,5SbTe3 are also shown in Figure 5.12(b). Based on Te-Te stretching
frequencies reported for the elemental Te (144cm'1, 126cm", and 95cm") or other
polytelluride compounds (i.e. 170 cm'1 for Rb2Te2), most of the bands in the spectra of
these compounds are attributed to the various Te-Te stretches in the net. However, there
is a possibility that some of these bands may also due to the Sb-Te modes given that Sb
and Te have similar atomic masses and Sb-Te and Te-Te bond distances are similar.

Differential thermal analysis (DTA) experiments showed that EquSe3 and
EuSeTe3 decompose at 723°C and 605°C, respectively, while Bao,5Euo.5SbTe3 does not

melt nor decompose under 1000°C.

184

 

    

 

 

 

 

 

 

3 b
,. E (a)
=2 5
D 2:-
5 E
:8 :
:3 1:-
B - Eg~o.97 eV

O.illllllLllllllllllljl

0.8 1.0 1.2 1.4 1.6 1.3
Energy (eV)
- (b)
.9
t:
:3
g / {I
..D
.-
§
- Eg~o.26 eV
" 0.3 0.4 0.5 0.6

Energy (eV)

Figure 5.9. Optical absorption spectra of (a) EquSe3 and (b) EquTe3(—) and
Bao,5Euo_SSbTe3(---). The band gap value, E8, is shown in the ﬁgure.

185

 

0' (S/cm)

 

 

 

O lllllllllIUIIIIIIILIJLLIIIIIIIJII

0 50 100 150 200 250 300 350
Temperature (K)

 

 

 

 

 

350_
E03)
300:
A 250:-
¥ :
> .
3 P
m 200:
150:-
20%)
loo-Jillllllllllllilell+lllllllL

 

O 50 100 150 200 250 300 350
Temperature (K)

Figure 5.10. (a) Variable temperature electrical conductivity and (b) thermopower for a

single crystal of EquTe3.

186

 

 

 

 

  

 

>5 40 I
E I
8 30 I
VJ
a? '
0,5, "
cud I
age 20 .
gig I
I

2 I.
“m’ 10 I
33 I
>
I:
p—q

lllllllllllllLlllllllllllllll

 

O O 50 100 150 200 250 300

Temperature (K)
Figure 5.11. Inverse molar magnetic susceptibility, xM (per europium), of EquSe3.

187

 

 

 

 

 

 

 

 

 

 

 

“,2
N (a)
a l
S
3
E
3?. v
'8 “1' m
< t g
V
IIJJUIJJLIIIIIIIIILLLL
100 200 300 400 500 600
Raman Shift (cm'l)
(b)
3‘
'5
G
8
.5
E
..D
H
<
he .2." ............................... g'u
Ell lllllllllllllllllllllllll

 

 

150 200 250 300 350 400
Raman Shift (cm‘l)

Figure 5.12. The Raman spectra of (a) EquSe3 and (b) EquTe3 (—) and
Bao.sEuo.sSbTes (---)-

188

4. Conclusions

The three new europium compounds, EquSe3, EquTe3, and Euo,5Ba0,5SbTe3
contain strongly modulated Se and Te nets. The modulations of the Se/T e nets give rise
to superstructures in each compound. The structures of EquSe; and EquTe3 are
differentiated due to the different behavior of the lone electron pair of the Sb atoms.

Electron diffraction studies revealed that EquSe3 possesses a very large and
complicated possibly incommensurate superstructure. Interestingly, EquTe3 and
Bao_5Euo.5SbTe3 possess the same average structure but different superstructures.
Considering that the substitution of Ba for Eu is isoelectronic, modulation in the net
seems to depends not only on the electron count but also on the remaining part of the
structure. EquTe3 possesses an incommensurate superstructure, the reﬁnement of which
is beyond our capability. On the other hand, Bao,5Euo.5SbTe3 possess a commensurate 6-
fold superstructure along the b-axis caused by the modulation of the Te net, cationic
ordering between Ba and En, and the distortion of the local environment of Sb. A
collaboration has been initiated to try and solve the extremely weak or incommensurate
superstructure of EquSe3 and EquTe3. The semiconducting behavior of these materials
implies that the modulation of the nets in these compounds are severe enough to

completely open a gap at the Fermi level.

189

References

Choi, K. S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309-
319.

Chen, J. H.; Dorhout, P. K. J. Alloys Compd. 1997, 249, 199.

(a) Greenwood, N. N.; Eamshow, E. Chemistry of the elements; Pergamon Press:
New York, 1984; 1434. (b) Guittard, M.; F lahaut, J. Synthesis of Lanthanide and
Actinide Compounds, Kluwer Academic Publisherrs, Netherlands, 1991, 321.

(a) DiMasi, E.; Aronson, M. C.; Mansﬁled, J. F.; Foran, B.; Lee, S. Phys. Rev. B.
1995, 55(20), 14516. (b) Gweon, G. H.; Denlinger, J. D.; Clack, J. A.; Allen, J.
W.; Olson, C. G.; DiMasi, E.; Aronson, M. C.; Foran, B.; Lee, S. Phys. Rev. Lett.
1998, 81, No.4, 886. (c) DiMasi, E.; Foran, B.; Aronson, M. C.; Lee, S. Chem.
Mater. 1994, 6, 1867.

(a) Foran, B.; Lee, S.; Aronson, M. C. Chem. Mater. 1993, 5, 974-978. (b) Lee,
S.; Foran, B. J. Am. Chem. Soc. 1994, 116, 154-161 and references therein.
Stdwe, K. J. Solid State Chem. 2000, 149, 155-166 and references therein.
Burdett, J. K. Chemical Bonding in Solids, 1995, Oxford University Press; New
York, pp 48-52.

(a) Zhang, Z.; Li, J.; Foran, B.; Lee, S.; Guo, H.-Y.;Hogan, T.;Kannewurf, C. R.;
Kanatzidis, M. G. J. Am. Chem. Soc. 1995, 117, 10513. (b) Gourdon, O; Hanko, J.
A.; Boucher, F.; Perticek, V.; Whangbo, M.-H.; Kanatzidis, M. G.; Evain, M.

Inorg. Chem. 2000, 39, 1398-1409.

190

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

Patschke, R.; Heising, J .; Kanatzidis, M. G.; Brazis, P.; Kannewurf, C. R. Chem.
Mater. 1998, 10, 695.

Patschke, R.; Brazis, P.; Kannewurf, C. R.; Kanatzidis, M. G. J. Mater. Chem.
1999, 9, 2293-2296.

Patschke, R.; Heising, J .; Schindler J .; Kannewurf, C. R.;Kanatzidis, M. G. J.
Solid State Chem. 1998, 135, 111-115.

McCarthy, T. J.; Ngeyi, S.-P.; Liao, J. —H.; Degroot, D.; Hogan, T.; Kannewurf,
C. R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331.

P. W. Selwood, in “Magnetochemistry” 2nd ed., Interscience Publishers, New
York, 1956.

Lyding, J. W.; Marcy, H. 0.; Marks, T. J .; Kannewurf, C. R. IEEE Trans.
Instrum. Meas., 1988, 3 7, 76-80.

Chaikin, P. I.; Kawk, J. F. Rev. Sci. Instrum, 1975, 46, 218-220.

SMART: 1994, Siemens Analytical Xray Systems, Inc., Madison, Winsconsin
53719 USA.

SAINT: Version 4, 1994-1996, Siemens Analytical Xray Systems, Inc., Madison,
Winsconsin 53719 USA.

Sheldrick, G. M. University of Gdttingen, Germany, to be published.
SHELXTL: Version 5, Sheldrick, G. M. Siemens Analytical Xray Systems, Inc.,

Madison, WI 53719, USA, 1994.
The original 8-fold supercell (asupeF-csub, bsupcr= x 8 asub, csupcr= bsub), which was C-

centered orthorhombic, was reduced to a primitive cell (asupeﬁcsub, bsuper= 0.5 x

csub+4 x asub, csupe,= bsub ). This was done in order to reﬁne the structure in the P1

191

21.

22.

23.

24.

25.

26.

27.

28.

space group so to allow every possible ﬁne atomic displacement without

constraining them by symmetry operations.

Unit cell: a = 4.2925(13)A, b = 14.882(5)A, c = 34.019(10)A, CL = 89.953(5)°, B =
89.967(5)°, y = 81.720(5)°; Z = 16; V = 2150.6(11)A3; Dcalc = 6.120 g/cm3; Space
group P1 (no. 1); p(Mo-Ka) = 361.00 cm"; total data, 22919; unique data, 20142
(Rim = 0.07); No. of variables, 526; R1(I>20') = 0.1373, R1(all data) = 0.2476,

GOF = 1.68.

Krtinert, Von W.; Plieth, K.; Z. anorg. allg. Chem. 1965, 336, 207-218.
Adenis, C.; Langer, V.; Lindqvist, 0. Acta Cryst. 1989, 45C, 941-942.

(a) Bbttcher, P.; Getzschmann, J .; Keller, R. Z anorg. allg. Chem. 1993, 619,
476-488.

F uruseth, S.; Brattas, L.; Kjekshus, A. Acta Chem. Scand. 1975, A29, 623-631.

R. Cook and H. Schafer, in Studies in Inorganic Chemistry Vol. 3, edited by R.
Metselaar, H. J. M. Heijligers and J. Schoonman (Elsevier Scientiﬁc Publishing

Company, Amsterdam, 1983), p. 757-760.

Choi, K.-S.; Chung, D.-Y.; Heising, J .; Brazis, P.; Kannewurf, C. R.; Kanatzidis,
M. G. Mat. Res. Soc. Symp. Proc. 1998, 545, 117-122.

Drago, R. S. Physical Methods for Chemists, 2nd ed., 1992, Saunders College
Publishing; New York, pp480-486.

Choi, K.-S.; Kanatzidis, M. G. Chem. Mater. 1999, 11, 2013-2018.

192

Chapter 6. Reactions of Lead Metal with Bi/Sb in Polyselenide Fluxes.
K1+be4.2xM7+,Se15(A=K, Rb; M=Bi, Sb): A New Class of Solid State

Quaternary Thermoelectric Compounds with Very Low Thermal

Conductivity.

193

1. Introduction

Several new concepts that relate structure and thermoelectric(TE) properties are
stimulating renewed interest and research activity in thermoelectric materials.1’2’3'4’5 The
challenge in ﬁnding a superior TE material lies in achieving high electrical conductivity (0),
high thermopower (S), and low thermal conductivity (K) in the same material. These
properties deﬁne the TE ﬁgure of merit ZT = (G'SZ/K)'T. Our approach to obtain
compounds with a high ﬁgure of merit has been to search for new semiconducting
multinary bismuth chalcogenide materials. Speciﬁcally, we target incorporating alkali and
alkaline earth metal ions into inﬁnite anionic three-dimensional metal-chalcogen
frameworks. These ions are stabilized between the layers or in the channels of the anionic
framework making relatively weak ionic bonds with chalcogens and act as “rattlers” in
their local cavities. These rattlers create low frequency vibration modes that scatter
acoustic phonons and thus, reduce the thermal conductivity. Other factors contributing
to low thermal conductivity in materials are the presence of heavy atoms (e. g. Bi, Pb),
large unit cells, and mass ﬂuctuation disorder in the crystal lattice. The compounds
BaBiTe3,6 B-KzBigsel3,7 K2_,Bi8_53e,4,7 CsBi4Te6,8 KzBigSl3,9 and KBi,_33sm9 exhibit
promising TE properties, encouraging us to extend our work to even more complicated
quaternary systems. The structural complexity in these materials can produce
corresponding complexities in the electronic band structure, which according to the Mott
formula10 may give rise to high thermopower coupled with high electrical conductivity.

To further explore the effects of structural complexity and mass ﬂuctuation in

compounds with large unit cells, we began to investigate quaternary systems by

194

incorporating Pb metal to the K/Bi/Se system. In addition to being very heavy, Pb has
similar electronic properties with Bi and has a well known tendency to disorder with Bi
or alkali metals depending on the local coordination environment.1 ”2'13 Site occupancy
disorder in the structure generates randomness of the mass, size, and charge of the atoms
on a particular lattice position that can strongly scatter lattice phonons carrying heat.2
These investigations led to a new class of quaternary solids of the general formula,
K1+be4-2xM7+xSe15 (M = Bi, Sb). The structure type associated with this class is new
and a remarkable characteristic of it is the extensive mixed occupancy of certain
crystallographic sites between K, Pb, and Bi (or Sb) atoms. This property has two
important consequences; ﬁrst, it generates a continuum of isostructural compositions
with varying band-gap, and second, it sets the stage for systems with very low thermal
conductivity. The former is advantageous because it provides a mechanism for controlling
the electrical properties of these materials, whereas the latter is a necessary condition for
any viable TE material. Therefore, certain members of this, apparently large, class of
isostructural compounds may hold high potential for TB applications. We report here the
syntheses, structures, and physico-chemical and thermoelectric properties of the
members, K1251) b3.5Bi7.2586159 Rb1.45Pb3.13b7.4SS€15, K1.45Pb3.1Sb7.453615, and
K2_15Pb1_7Sb8,ISSe1 5. We show that K1_25Pb3_5Bi7_25Se15 in fact is a promising candidate for

TB applications.

2. Experimental Section

2.1. Synthesis.

195

The following reagents were used as obtained: lead metal, 99.999%, 200 mesh,
Cerac, Milwaukee, WI; bismuth selenide, 99.999% -325 mesh, Cerac, Milwaukee, WI ;
antimony -200 mesh, Cerac, Milwaukee, WI; selenium powder, 99.5+%, 100 mesh,
Aldrich Chemical Co., Milwaukee, WI or selenium shots, 99.999%, Noranda Advanced
Materials, Quebec, Canada; rubidium metal, 99.8%, Johnson Matthey Co., Ward Hill,
MA; potassium metal, granules, <6mm, 99% purity, Aldrich Chemical Co., Inc.,
Milwaukee, WI. The starting materials KZSe and RbZSe were prepared by a

stoichiometric reaction of potassium/rubidium metal with selenium in liq. NH3. szSe3

was prepared by heating a stoichiometric mixture of antimony and selenium at 750 °C for

48 hrs.

K1.25Pb3.5Bi7.2550159 Rbl.45Pb3.le7.4sseIS’ and K1.45P b3.isb7.455015- The
compounds were synthesized from a mixture of 2 mmol AZSe (A=K or Rb), 2 mmol Pb, 3
mmol M28e3(M=Bi or Sb), and 10 mmol Se. The reagents were thoroughly mixed, sealed
in an evacuated Pyrex tube, and heated at 540 °C for 5 d (cooling rate of 2 °C/h). Pure
silver colored needles of each compound were obtained by isolation in degassed
dimethylformamide (DMF) and water (yield > 95% based on Pb metal). The crystals are

air and water stable.

K2.15Pb1.73bs.153915- The compound was prepared from a mixture of 4 mmol
KZSe, 1 mmol Pb, 3 mmol SbZSe3, and 10 mmol Se using the same synthetic condition as
described above. The X-ray crystallographic reﬁnement was unable to determine the

exact formulas for the Sb compounds due to sites in the crystal structure with

196

considerable mixed occupancy among K+/Rb+, Pb”, and Sb”, and the associated
difﬁculties in reﬁning occupancies of triply disordered sites. Therefore, all the formulae
we report here were determined by elemental analyses obtained by electron microprobe
EDS analysis. For K..25Pb3.5Bi7,258e15 we were also able to conﬁrm the formula by X-ray
crystallographic analysis as we discuss later. The resulting formula is in good agreement

with those determined by the EDS analysis.

2.2. Physical Measurements.

Infrared Spectroscopy. Optical diffuse reﬂectance measurements were made on the
ﬁnely ground sample at room temperature. The spectrum was recorded in the infrared
region (6000 - 400 cm") with the use of a Nicolet MAGNA-IR 750 Spectrometer
equipped with a Collector Diffuse Reﬂectance of Spectra-Tech. Inc. The reﬂectance
versus wavenumber data were used to estimate a material's band gap by converting

reﬂectance to absorption data as described previously.'4

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically, a sample (~20 mg)
of ground crystalline material was sealed in silica tubes under vacuum. A silica tube of

equal mass ﬁlled with A1203 was sealed and placed on the reference side of the detector.
The samples were heated to 800 °C at 5 °C/min, isothermed for 10 min followed by

cooling at -5 oC/min to 50 °C. Residues of the DTA experiments were examined by X-

197

ray powder diffraction. The stability/reproducibility of the samples was monitored by

running multiple heating/cooling cycles.

Charge-Transport Measurements. DC electric conductivity and thermopower
measurements were made on single crystals of the compounds. Conductivity
measurements were performed in the usual four-probe geometry with 60- and 25-um gold
wires used for the current and voltage electrodes, respectively. Measurements of the
sample cross~sectional area and voltage probe separation were made with a calibrated
binocular microscope. Conductivity data were obtained with the computer-automated

‘5 Thermoelectric power measurements were made by using

system described elsewhere.
a slow AC techniquels’16 with 60 um gold wires serving to support and conduct heat to
the sample, as well as to measure the voltage across the sample resulting from the applied
temperature gradient. In both measurements, the gold electrodes were held in place on the
sample with a conductive gold paste.

Conductivity specimens were mounted on interchangeable sample holders, and
thermopower specimens were mounted on a ﬁxed sample holder/differential heater.
Mounted samples were placed under vacuum (10'3 Torr) and heated to room temperature
for 2-4 h to cure the gold contacts. For a variable-temperature run, data (conductivity or
thermopower) were acquired during both sample cooling and warming to check
reversibility. The temperature drift rate during an experiment was kept below 1 K/min.
Typically, three to four separate variable-temperature runs were carried out for each

sample to ensure reproducibility and stability. At a given temperature, reproducibility

was within :5 %.

198

Thermal Conductivity Measurement. Thermal conductivity was determined using
a longitudinal steady-state method over the temperature range 4—300 K. The samples
were attached (using either a low melting point solder or silver-loaded epoxy) to the cold
tip of the cryostat, while the other end of the samples were provided with a small strain
gauge resistor (thin ﬁlm) which serves as a heater. The temperature differences across the

samples were measured using a differential chromel-constantan thermocouple.

2.3 X-ray Crystallography.

A single crystal of each compound was mounted on the tip of a glass ﬁber.
Intensity data were collected on a Siemens SMART Platform CCD diffractometer using
graphite monochromatized Mo Ka radiation over a fullsphere of reciprocal space. The
individual frames were measured with an omega rotation of 0.3 deg and an acquisition time
of 45 sec per frame. The SMART software17 was used for the data acquisition and
SAINT18 for data extraction and reduction. The absorption correction was done using
SADABS.19 Structure solution and reﬁnement for all compounds were performed with
the SHELXTL package20 of crystallographic programs. All compounds are isostructural,
and were solved and reﬁned successfully in the P21/m space group. The complete data
collection parameters, details of the structure solution and the reﬁnement for each
compound are given in Tables 6.1 and 6.2. The coordinates of all atoms, isotropic
temperature factors, and their estimated standard deviations for each compound are given
in Tables 6.3, 6.4, 6.5, and 6.6. The selected bond distances for each compound are given

in Tables 6.7, 6.8, 6.9, and 6.10.

199

K1,25Pb3,50Bi-,~,258e15. Since Pb2+ and Bi3+ have similar atomic weight and
scattering power, resolution between them is impossible by X-ray diffraction. Nominally
the 8-coordinate sites were assigned as Pb(l) and Pb(2) and the 6-coordinate sites as
Bi(l) through Bi(9) because in the Sb analogs Pb mainly occupies the 8-coordinate sites
and Sb occupies the 6-coordinate sites. Aﬁer successive reﬁnements, the atomic
displacement parameters (ADPs) for the Pb(l) and Pb(2) remained high, indicating that
these sites should be disordered with the lighter K atoms. The occupancy reﬁnement
shows that the Pb(l) and Pb(2) sites contain 10.9(6)% of K(l ’) and 14.3(6)% of K(2’),
respectively. The octahedral Bi sites had reasonable ADPs and did not display any mixed
occupancy with K atoms. Once the amount of K+ ions in the compound was determined,
the ratio of Pb” and Bi3+ could be calculated under the assumption that the compound is
valence precise (i.e. no mixed valency). The resulting ﬁnal formula K125Pb350Bi7258e15 is
in good agreement with that proposed by the results of elemental analysis (SEM/EDS).
Rbl.45Pb3JSb7,4SSe15. Unlike K1,25Pb3_5Bi7,ZSSe15 for which we practically dealt
with the disorder between only two species (K and indistinguishable Pb/Bi), here there
are three distinguishable elements, Rb, Sb, and Pb. If a site is occupied by three different
atoms, mathematically there are more than one way to assign partial occupancy for each
element so that the resulting calculated scattering electron density for the site is the same
as the observed one. In addition, when we reﬁned occupancies for more than three triply
disordered sites, the reﬁnement became unstable and diverged. Therefore, the Rb/Pb/Sb
ratio was determined from the elemental analysis results provided by EDS. For the sites
with abnormally high ADPs, we introduced occupancy disorder with only one other

element. For example, the Pb(2) site with a high ADP (Ueq = 0.026 A2) can be disordered

200

with lighter Rb and/or Sb but this site was arbitrarily disordered with only Rb, which
resulted in 82.1(5)% Pb / 17.9(5)% Rb with a lowered ADP (Ucq = 0.021 A2). Among
the Sb sites, Sb(3), Sb(4), and Sb(6) had lower ADPs indicating mixed occupation with
heavier Pb atoms. The reﬁnement resulted in 29.4(8)%, 27.3(7)%, and 19.9(7)% of Pb in
Sb(3), Sb(4), Sb(6) sites, respectively. We cannot exclude the possibility that the other
Sb sites with reasonable ADPs may also be disordered with heavier Pb and lighter Rb at
the same time in such ratios that the ADPs remain reasonable.

K1,..5Pb3JSb7458e, 5. The 8-coordinate Pb(l) and Pb(2) sites once again had high
ADPs and mixed occupation with 9.2(4)% of K(l ’) and 29.4(5)% of K(2’). Since the
ADPs for the Sb atoms were normal, a Pb contribution was not considered for these sites.
However, the EDS results requires more Pb content than the total Pb amount suggested
by the reﬁnement for the 8-coordinate sites, implying that the Sb sites may have varying
contributions from both Pb and K atoms.

szstlJSstsSeu. Both Pb(l) and Pb(2) sites showed high ADPs. For the
Pb(l) site introducing mixed occupancy did not drop the ADPs and the amount of K
atom reﬁned for this site was negligible. This indicated that the high ADPs for the Pb(l)
site was mainly due to the positional ﬂuctuation so called “rattling effect” of the Pb
atoms in this large pocket. The Pb(2) site was reﬁned to 50.5(5)% Pb / 49.5(5)% K. The
difference in the total Pb content in the 8-coordinate sites for both K1,45Pb3_ISb7_458e15 and
K2_15Pb1,7Sb3,15Se1 5 are not signiﬁcant. Therefore, the difference should lie in the Pb
amount disordered with Sb on the octahedral sites. The composition K2.15Pb1.7Sb3,lsSe15,

indicated by EDS/SEM, contains only half of the Pb present in Kl.45Pb3.le7,45Se15.

201

Table 6.1. Summary of Crystallographic Data and Structural Analysis for
K1.25Pb3.sBi7.253615 and Rbl.45Pb3.le7.4SSClS-

 

Formula
Formula weight
Crystal habit
Space group

a, A

b, A

c, A

[3, degree.

2; v, A3

Dcalc, g/cm3

Temp, K
K(Mo K01), A

u(Mo K01), cm'l
F(000)

9 max 3 deg
Total data
Unique data
No. of variables

Reﬁnement method
Final R indices [I>2o]

R indices (all data)

GOF on F2

“R1=2llFol-IFcll/ZIF0

 

K1.25Pb3.sBi7.253615
3473.54

silver needles
P21/m

17.4481(8)
4.1964(2)
21.6945(10)
98.8500(10)

2, 1569.54(l3)
7.350

293(2)

0.71069

767.31

2845

28.30

12085

4208 [Rim-r— 0.0821]
166

Rb1.45Pb3.ISb7.453€15
2857.66

silver needles
P21/m

17.3160(7)
4.1406(2)

21 .6401(8)
99.139000)
2,1531.87(11)
6.195

171(2)

0.71069

435.29

2396

28.84

10198

4089 [Rim-‘- 0.0427]
174

Full-matrix least-squares on F2

Rlb =0.0537
wR2°=0.1126
R1=0.0910
wR2=0.1211
0.921

R1=0.0432
wR2=0.0931
R1=0.0603
wR2=0.9864
0.994

b wR2= {23./(11:349.2 )21/2[w(F..2)2]}"2

202

Table 6.2. Summary of Crystallographic Data and Structural Analysis for
K1.45Pb3.13b7.453615 and K2.15Pb1.7Sb8.1sSels-

 

Formula
Formula weight
Crystal habit
Space group

a, A

b, A

c, A

B, degree.

Z; V, A3

Dcalc, g/cm3

Temp, K
K(Mo K01), A

p(Mo K01), cm’l
F(000)

0 max , deg
Total data
Unique data
No. of variables

Reﬁnement method
Final R indices [I>2o]

R indices (all data)

GOF on F2

”R1=leFol-IFcll/ZIF0

 

9

K1.45Pb3.1Sb7.453615
2790.42

silver needles
P21/m

17. 1204(6)
4.1568(2)

21 .6362(8)
98.706(1)

2, 152202(11)
6.089

171(2)

0.71069

417.05

2343

28.35

10547

3992 [11.m = 0.0546]
166

K2.15Pb1.7Sb8.153615
2612.95

silver needles
P21/m

17.164(4) .
4.1494(9)

21 .684(5)
98.664(3)

2, 1527.0(6)

5.683

293(2)

0.71069

345.88

2212

28.70

12322

3992 [Rim= 0.0433]
165

Full-matrix least-squares on F2

R1=0.0533
wR2=0.1095
R1=0.0850
wR2=0.1212
1.029

R1=0.0502
wR2=0.1308
R1=0.0605
wR2=0.1364
1.043

b wR2= {Emmi—Ff )2 ]/2[w(F02)2 1}"2

203

Table 6.3. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for K1,25Pb3.5oBi7,258e15 with Estimated Standard Deviations in

 

 

Parentheses.

Atoms x y Z Ueq,a A2
Pb(1)/K(1’)b 0.3300(1) 1/4 0.7735(1) 0.020(1)
Pb(2)/K(2’)° 0.7769(1) 1/4 0.6832(1) 0.019(1)
Bi(l) 0.4546(1) 1/4 0.1669(1) 0.011(1)
Bi(2) 0.7563(1) 1/4 0.8799(1) 0.011(1)
Bi(3) 0.0340(1) 1/4 0.0928(1) 0.014(1)
Bi(4) 0.1848(1) 1/4 0.9413(1) 0.013(1)
Bi(5) 0.6063(1) 1/4 0.0266(1) 0.013(1)
Bi(6) 0.0535(1) 1/4 0.7146(1) 0.012(1)
Bi(7) 0.3253(1) 1/4 0.4814(1) 0.010(1)
Bi(8) 0.0681(1) 1/4 0.4323(1) 0.011(1)
Bi(9) 0.8064(1) 1/4 0.3938(1) 0.010(1)
K(1) 0.5069(4) 1/4 0.6328(3) 0.018(2)
Se(l) 0.4619(2) 1/4 0.9036(1) 0.015(1)
8e(2) 0.3207(2) 1/4 0.0454(1) 0.010(1)
Se(3) 0.8937(2) 1/4 0.9855(1) 0.011(1)
Se(4) 0.5705(2) 1/4 0.2659(1) 0.011(1)
Se(5) 0.6255(2) 1/4 0.7802(1) 0.011(1)
Se(6) 0.1761(2) 1/4 0.1839(1) 0.014(1)
Se(7) 0.430(2) 1/4 0.8476(1) 0.015(1)
8e(8) 0.7373(2) 1/4 0.1190(1) 0.013(1)
Se(9) 0.3350(2) 1/4 0.3565(1) 0.014(1)
Se(10) 0.0684(2) 1/4 0.3059(1) 0.010(1)
Se(ll) 0.8231(2) 1/4 0.2711(1) 0.011(1)
Se(12) 0.5711(2) 1/4 0.4922(1) 0.010( 1)
Se(13) 0.8149(2) 1/4 0.5511(1) 0.011(1)
Se(14) 0.575(2) 1/4 0.5788(1) 0.009(1)
Se(15) 0.3016(2) 1/4 0.6274(1) 0.011(1)

 

“ Ueq is deﬁned as one third of the trace of the orthogonalized Uij tensor.
b 89.1(6)% Pb(l) and 10.9(6)% K(l').
° 85.7(6)% Pb(2) and 14.3(6)% K(2').

204

Table 6.4. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameters for Rb1,45Pb3_me7_458e15 with Estimated Standard Deviations in Parentheses.

 

 

Atoms x y z Ueq, A2
Pb(l) 0.3284(1) 1/4 0.7800(1) 0.019(1)
Pb(2)/K(2)a 0.7864(1) 1/4 0.6842(1) 0.021(1)
Sb(l) 0.4551(1) 1/4 0.1641(1) 0.018(1)
Sb(2) 0.7564(1) 1/4 0.8809(1) 0.014(1)
Sb(3)/Pb(3’)b 0.0383(14) 1/4 0.0953(14) 0.017(2)
Sb(4)/Pb(4’)° 0.1829(13) 1/4 0.9438(13) 0.016(2)
Sb(5) 0.6117(1) 1/4 0.0299(1) 0.019(1)
Sb(6) /Pb(6’)d 0.556(12) 1/4 0.7160(9) 0.014(2)
Sb(7) 0.3196(1) 1/4 0.4790(1) 0.013(1)
Sb(8) 0.0653(1) 1/4 0.4304(1) 0.017(1)
Sb(9) 0.8042(1) 1/4 0.3900(1) 0.013(1)
Rb(1) 0.5052(1) 1/4 0.6344(1) 0.015(1)
Se(l) 0.4595(1) 1/4 0.9036(1) 0.019(1)
Se(2) 0.3187(1) 1/4 0.0408(1) 0.016(1)
Se(3) 0.8914(1) 1/4 0.9876(1) 0.018(1)
Se(4) 0.5684(1) 1/4 0.2590(1) 0.011(1)
Se(5) 0.6235(1) 1/4 0.7845(1) 0.021(1)
Se(6) 0.1786(1) 1/4 0.1817(1) 0.034(1)
Se(7) 0.0348(2) 1/4 0.8475(1) 0.051(1)
86(8) 0.7352(1) 1/4 0.1187(1) 0.016(1)
Se(9) 0.3297(1) 1/4 0.3592(1) 0.013(1)
Se(10) 0.0664(1) 1/4 0.3095(1) 0.014(1)
Se(] 1) 0.8212(1) 1/4 0.2724(1) 0.014(1)
Se(12) 0.5799(1) 1/4 0.4946(1) 0.014(1)
Se(13) 0.8198(1) 1/4 0.5535(1) 0.013(1)
Se(l4) 0.0580(1) 1/4 0.5820(1) 0.013(1)
Se(15) 0.2996(1) 1/4 0.6304(1) 0.014(1)

 

“ 82.1(5)% Pb and 17.9(5)% K. b 70.6(8)% Sb and 29.4(8)% Pb. ° 72.7(7)% Sb and
27.3(7)% Pb. d 80.1(7)% Sb and 19.9(7)% Pb.

205

Table 6.5. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameters for K1.45Pb3.10Sb7,4sSe15 with Estimated Standard Deviations in Parentheses.

 

 

Atoms x y z Ueq, A2
Pb(1)/K(1’)a 0.3277(1) 1/4 0.7770(1) 0.021(1)
Pb(2)/K(2)b 0.7795(1) 1/4 0.6852(1) 0.026(1)
Sb(1) 0.4547(1) 1/4 0.1675(1) 0.022(1)
Sb(2) 0.7561(1) 1/4 0.8821(1) 0.022(1)
Sb(3) 0.0338(1) 1/4 0.0935(1) 0.015(1)
Sb(4) 0.1811(1) 1/4 0.9427(1) 0.017(1)
Sb(5) 0.6101(1) 1/4 0.0312(1) 0.024(1)
Sb(6) 0.0530(1) 1/4 0.7138(1) 0.018(1)
Sb(7) 0.3220(1) 1/4 0.4778(1) 0.024(1)
Sb(8) 0.0669(1) 1/4 0.4300(1) 0.022(1)
Sb(9) 0.8051(1) 1/4 0.3900(1) 0.016(1)
K(1) 0.5039(3) 1/4 0.6341(2) 0.020(1)
Se(1) 0.4603(1) 1/4 0.9004(1) 0.024(1)
Se(2) 0.3208(1) 1/4 0.0400(1) 0.019(1)
Se(3) 0.8904(1) 1/4 0.9866(1) 0.020(1)
Se(4) 0.5674(1) 1/4 0.2630(1) 0.014(1)
Se(5) 0.6255(1) 1/4 0.7816(1) 0.025(1)
Se(6) 0.1766(1) 1/4 0.1797(1) 0.035(1)
Se(7) 0.0361(2) 1/4 0.8483(1) 0.050(1)
Se(8) 0.7353(1) 1/4 0.1188(1) 0.020(1)
Se(9) 0.3380(1) 1/4 0.3597(1) 0.018(1)
Se(10) 0.0696(1) 1/4 0.3096(1) 0.017(1)
Se(1 1) 0.8232(1) 1/4 0.2724(1) 0.018(1)
Se(12) 0.5754(1) 1/4 0.4939(1) 0.031(1)
Se(13) 0.8159(1) 1/4 0.5532(1) 0.023(1)
Se(14) 0.0566(1) 1/4 0.5820(1) 0.017(1)
Se(15) 0.3001(1) 1/4 0.6294(1) 0.017(1)

 

° 90.8(4)% Pb and 9.2(4)% K. " 70.6(5)% Sb and 29.4(5)% Pb.

206

Table 6.6. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameters for K2_15Pb1,7SngSSe.5 with Estimated Standard Deviations in Parentheses.

 

 

Atoms x y z Ueq, A2
Pb(1) 0.3281(1) 1/4 0.7767(1) 0.041(1)
Pb(2)/K(2’)“ 0.7758(1) 1/4 0.6839(1) 0.050(1)
Sb(1) 0.4545(1) 1/4 0.1674(1) 0.027(1)
Sb(2) 0.7568(1) 1/4 0.8826(1) 0.033(1)
Sb(3) 0.0342(1) 1/4 0.0934(1) 0.026(1)
Sb(4) 0.1815(1) 1/4 0.09430(1) 0.027(1)
Sb(5) 0.6108(1) 1/4 0.0314(1) 0.030(1)
Sb(6) 0.0528(1) 1/4 0.7127(1) 0.026(1)
Sb(7) 0.3219(1) 1/4 0.4787(1) 0.028(1)
Sb(8) 0.0676(1) 1/4 0.4303(1) 0.028(1)
Sb(9) 0.8049(1) 1/4 0.3898(1) ' 0.022(1)
K(1) 0.5036(2) 1/4 0.6338(2) 0.031(1)
Se(1) 0.4609(1) 1/4 0.9011(1) 0.031(1)
Se(2) 0.3204(1) 1/4 0.0393(1) 0.025(1)
Se(3) 0.8902(1) 1/4 0.9864(1) 0.026(1)
Se(4) 0.5663(1) 1/4 0.2626(1) 0.020(1)
Se(5) 0.6259(1) 1/4 0.7822(1) 0.033(1)
Se(6) 0.1747(1) 1/4 0.1778(1) 0.042(1)
Se(7) 0.0355(2) 1/4 0.8481(1) 0.062(1)
Se(8) 0.7351(1) 1/4 0.1182(1) 0.025(1)
Se(9) 0.3388(1) 1/4 0.3609(1) 0.025(1)
Se(10) 0.0678(1) 1/4 0.3101(1) 0.022(1)
Se(1 1) 0.8237(1) 1/4 0.2731(1) 0.022(1)
Se(12) 0.5765(1) 1/4 0.4935(1) 0.032(1)
Se(13) 0.8168(1) 1/4 0.5530(1) 0.029(1)
Se(14) 0.0574(1) 1/4 0.5826(1) 0.022(1)
Se(15) 0.3002(1) 1/4 0.6292(1) 0.022(1)

 

° 50.5(5)% Pb and 49.5(5)% K.

207

Table 6.7. Selected Distances(A) for K1.25Pb3,5Bi7_25Se15.

 

Pb(1)-Se(4) x 2 2.934(2) Pb(2)-Se(9) x 2 2.905(2)
Pb(1)-Se(15) 3.133(3) Pb(2)-Se(13) 3.038(3)
Pb(1)-Se(1) 3.355(3) Pb(2)-Se(10) x 2 3.399(3)
Pb(1)-Se(11) x 2 3.418(3) Pb(2)-Se(6) x 2 3.559(3)
Pb(1)-Se(8) x 2 3.473(3) Pb(2)-Se(5) 3.622(3)
Bi(1)-Se(4) 2.715(3) Bi(6)-Se(7) 2.920(3)
Bi(l)-Se(5) x 2 2.857(2) Bi(6)-Se(14) 2.958(3)
Bi(1)-Se(1) x 2 3.091(2) Bi(6)-Se(10) x 2 2.972(2)
Bi(1)-Se(2) 3.241(3) Bi(6)-Se(11) x 2 2.988(2)
Bi(2)-Se(6) x 2 2.866(2) Bi(7)-Se(9) 2.741(3)
Bi(2)-Se(5) 2.892(3) Bi(7)-Se(12) x 2 2.772(2)
Bi(2)-Se(3) 3.051(3) Bi(7)-Se(13) x 2 3.218(2)
Bi(2)-Se(2) x 2 3.081(2) Bi(7)-Se(15) 3.257(3)
Bi(3)-Se(7) x 2 2.900(2) Bi(8)-Se(10) 2.742(3)
Bi(3)-Se(6) 2.923(3) Bi(8)-Se(13) x 2 2.910(2)
Bi(3)-Se(3) x 2 3.090(2) Bi(8)-Se(14) x 2 3.017(2)
Bi(3)-Se(3) 3.108(3) Bi(8)-Se(14) 3.211(3)
Bi(4)-Se(8) x 2 2.917(2) Bi(9)-Se(11) 2.721(3)
Bi(4)-Se(7) 2.950(3) Bi(9)-Se(15) x 2 2.811(2)
Bi(4)-Se(2) 3.013(3) Bi(9)-Se(14) x 2 3.156(2)
Bi(4)-Se(3) x 2 3.077(2) Bi(9)-Se(13) 3.394(3)
Bi(5)-Se(8) 2.799(3) K(1)-Se(9) x 2 3.445(6)
Bi(5)-Se(1) x 2 2.944(2) K(1)-Se(4) x 2 3.458(6)
Bi(5)-Se(2) x 2 3.014(2) K(1)-Se(12) x 2 3.531(6)
Bi(5)-Se(1) 3.376(3) K(1)-Se(12) 3.407(7)

K(1)-Se(5) 3.533(7)

K(1)-Se(15) 3.566(8)

 

208

Table 6.8. Selected Distances(A) for Rb1.45Pb3,1Sb7,45Se15.

 

Pb(1)-Se(4) x 2
Pb(1)-Se(15)
Pb(1)-Se(1)
Pb(1)-Se(8) x 2
Pb(1)-Se(1 1) x 2

Sb(1)-Se(4)
Sb(1)-Se(5) x 2
Sb(1)-Se(1) x 2
Sb(1)-Se(2)

Sb(2)-Se(6) x 2
Sb(2)-Se(5)
Sb(2)-Se(3)
Sb(2)-Se(2) x 2

Sb(3)-Se(7) x 2
Sb(3)-Se(6)
Sb(3)-Se(3) x 2
Sb(3)-Se(3)

Sb(4)-Se(2)
Sb(4)-Se(8) x 2
Sb(4)-Se(3) x 2
Sb(4)-Se(7)

Sb(5)-Se(8)
Sb(5)-Se(1) x 2
Sb(5)-Se(2) x 2
Sb(5)-Se(1)

2.9449(11)
3.197(2)
3.220(2)
3.3261(14)
3.3696(13)

2.606(2)
2.8025(14)
3.0508(14)
3.271(2)

2.807(2)
2.849(2)
3.014(2)
3.0891(14)

281(2)
282(2)
311(2)
3.16(3)

289(2)
296(2)
296(2)
304(2)

2.639(2)
2.9060(14)
2.9434(14)
3.483(2)

Pb(2)-Se(9) x 2
Pb(2)-Se(13)
Pb(2)-Se(10) x 2
Pb(2)-Se(6) x 2
Pb(2)-Se(5)

Sb(6)-Se(14)
Sb(6)-Se(7)

Sb(6)-Se(10) x 2
Sb(6)-Se(11) x 2

Sb(7)-Se(9)
Sb(7)-Se(12) x 2
Sb(7)-Se(13) x 2
Sb(7)-Se(1 5)

Sb(8)-Se(10)
Sb(8)-Se(13) x 2
Sb(8)-Se(14) x 2
Sb(8)-Se(14)

Sb(9)-Se(1 1)
Sb(9)-Se(15) x 2
Sb(9)-Se(14) x 2
Sb(9)-Se(13)

Rb(1)-Se(4) x 2
Rb(1)-Se(5)
Rb(1)-Se(9) x 2
Rb(1)-Se(12) x 2
Rb(1)-Se(12)
Rb(1)-Se(15)

209

2.9341(12)
2.976(2)
3.2696(12)
3.538(2)
3.822(2)

291(2)
292(2)
295(2)
295(2)

2.626(2)
2.7060(11)
3.1735(13)
3.350(2)

2.619(2)
2.8533(12)
2.9561(13)
3.303(2)

2.608(2)
2.7307(11)
3.1449(13)
3.505(2)

3.487(2)
3.555(2)
3.515(2)
3.599(2)
3.474(2)
3.547(2)

Table 6.9. Selected Distances(A) for K1,45Pb3_ISb7,45Se15.

 

Pb(1)-Se(4) x 2
Pb(1)-Se(15)
Pb(1)-Se(1)
Pb(1)-Se(8) x 2
Pb(1)-Se(11) x 2

Sb(1)-Se(4)
Sb(1)-Se(5) x 2
Sb(1)-Se(1) x 2
Sb(1)-Se(2)

Sb(2)-Se(6) x 2
Sb(2)-Se(5)
Sb(2)-Se(3)
Sb(2)-Se(2) x 2

Sb(3)-Se(7) x 2
Sb(3)-Se(6)
Sb(3)-Se(3) x 2
Sb(3)-Se(3)

Sb(4)-Se(2)
Sb(4)-Se(8) x 2
Sb(4)-Se(3) x 2
Sb(4)-Se(7)

Sb(5)-Se(8)
Sb(5)-Se(1) x 2
Sb(5)-Se(2) x 2
Sb(5)-Se(1)

2.960(2)
3.157(2)
3.234(2)
3.363(2)
3.362(2)

2.604(2)
2.806(2)
3.041(2)
3.309(2)

2.811(2)
2.874(2)
2.971(2)
3.093(2)

2.791(2)
2.841(3)
3.113(2)
3.107(2)

2.937(2)
2.951(2)
2.954(2)
2.967(3)

2.637(2)
2.916(2)
2.940(2)
3.519(3)

Pb(2)-Se(9) x 2
Pb(2)-Se(13)
Pb(2)-Se(10) x 2
Pb(2)-Se(5)
Pb(2)-Se(6) x 2

Sb(6)-Se(14)
Sb(6)-Se(10) x 2
Sb(6)-Se(11) x 2
Sb(6)-Se(7)

Sb(7)-Se(9)
Sb(7)-Se(12) x 2
Sb(7)-Se(13) x 2
Sb(7)-Se(15)

Sb(8)—Se(10)
Sb(8)-Se(13) x 2
Sb(8)-Se(14) x 2
Sb(8)-Se(14)

Sb(9)-Se(1 1)
Sb(9)-Se(15) x 2
Sb(9)-Se(14) x 2
Sb(9)-Se(13)

K(1)-Se(5)
K(1)-Se(4) x 2
K(1)-Se(9) x 2
K(1)-Se(12) x 2
K(1)-Se(12)
K(1)-Se(15)

210

2.953(2)
3.012(2)
3.303(2)
3.602(2)
3.574(2)

2.861(2)
2.943(2)
2.950(2)
2.968(3)

2.613(2)
2.730(2)
3.140(2)
3.355(2)

2.613(2)
2.874(2)
2.948(2)
3.321(2)

2.609(2)
2.740(2)
3.140(2)
3.508(2)

3.535(5)
3.404(4)
3.399(4)
3.564(4)
3.439(5)
3.475(5)

Table 6.10. Selected Distances(A) for K2,]5Pb1,7Sb3,158€15.

 

Pb(1)-Se(4) x 2
Pb(1)-Se(15)
Pb(1)-Se(1)
Pb(1)-Se(8) x 2
Pb(1)-Se(11) x 2

Sb(1)-Se(4)
Sb(1)-Se(5) x 2
Sb(1)-Se(1) x 2
Sb(1)-Se(2)

Sb(2)-Se(6) x 2
Sb(2)-Se(5)
Sb(2)-Se(3)
Sb(2)-Se(2) x 2

Sb(3)-Se(7) x 2
Sb(3)-Se(6)
Sb(3)-Se(3) x 2
Sb(3)-Se(3)

Sb(4)-Se(2)
Sb(4)-Se(8) x 2
Sb(4)-Se(3) x 2
Sb(4)-Se(7)

Sb(5)-Se(8)
Sb(5)-Se(1) x 2
Sb(5)-Se(2) x 2
Sb(5)-Se(1)

2.9639(14)

3.163(2)
3.257(2)
3.378(2)
3.379(2)

2.599(2)
2.803(2)
3.045(2)
3.329(2)

2.804(2)
2.883(2)
2.958(2)
3.101(2)

2.793(2)
2.799(2)
3.108(2)
3.125(2)

2.923(2)
2.949(2)
2.955(2)
2.994(3)

2.625(2)
2.917(2)
2.931(2)
3.524(2)

Pb(2)-Se(9) x 2
Pb(2)-Se(13)
Pb(2)-Se(10) x 2
Pb(2)-Se(5)
Pb(2)-Se(6) x 2

Sb(6)-Se(14)
Sb(6)-Se(10) x 2
Sb(6)-Se(11) x 2
Sb(6)-Se(7)

Sb(7)-Se(9)
Sb(7)-Se(12) x 2
Sb(7)-Se(13) x 2
Sb(7)-Se(15)

Sb(8)-Se(10)
Sb(8)-Se(13) x 2
Sb(8)-Se(14) x 2
Sb(8)-Se(14)

Sb(9)-Se(1 1)
Sb(9)-Se(15) x 2
Sb(9)-Se(14) x 2
Sb(9)-Se(13)

K(1)-Se(5)
K(1)-Se(4) x 2
K(1)-Se(9) x 2
K(1)-Se(12) x 2
K(1)-Se(12)
K(1)-Se(15)

211

2.923(2)
3.025(3)
3.379(2)
3.581(3)
3.640(2)

2.833(2)
2.920(2)
2.949(2)
2.996(3)

2.613(2)
2.718(2)
3.155(2)
3.341(2)

2.605(2)
2.856(2)
2.968(2)
3.335(2)

2.600(2)

2.7394(13)

3.134(2)
3.514(2)

3.570(5)
3.406(4)
3.398(4)
3.560(4)
3.459(5)
3.478(4)

3. Results and Discussion

Structure.

The structural motif associated with the AI+XM4-2xM’7+xSe15 (A = K, Rb; M =
Pb, Sn; M’ = Bi, Sb) family is Shown in Figure 6.1. This is an anisotropic three-
dimensional monoclinic structure, which propagates along the b-axis with the very short
repeating length of ~4.2A. In fact, this is a persistent feature of all ternary or quaternary
bismuth and antimony chalcogenide compounds and it is responsible for their highly
anisotropic needle-like crystal growth habits. All atoms are situated on crystallographic
mirror planes lying perpendicular to the b-axis.

The three-dimensional “M4-2xBi7+xSe15” framework is assembled from two
different structural modules, both of which are inﬁnite along the b-axis. The ﬁrst is a
stepped layer having the thickness and structural features of a BizTe3-type layer. These
stepped layers are parallel to the ab-plane, see Figure 6.1. The second structural element,

which is also found in B—K2B138613 and K2,5Bi8,SSe14,7 is an inﬁnite rod, which can be

regarded as an excised block from the NaCl lattice. In cross section, the rod is rectangular
and 3-Bi octahedra by 2-Bi octahedra wide. These so-called 3x2 rods, indicated in Figure
6.1, serve as pillars between the stepped layers to form the three-dimensional structure.
The tunnels, which accommodate the alkali metal cations, then form between the pillars.
The connection points between the rods and the layers are metal atom sites of 8
coordination. These are the sites most prone to mixed occupancy with alkali metals or

with different metals in the framework. For example in B-KzBlgSC” and K251318583”, the

corresponding sites exhibit mixed occupancy between Bi and K. In Al+be4-2xM7+xSe15,

212

    

2 o- .se'e :0
o v :~Q5~‘ =72- ‘. ”I
8 12 1:11 e..: :6 ,2} 112M «272”
.ﬁ. 1.4“? 1.: ~. ' _ ~ 3" S..." 3'3 01.: ‘5685727 :6 3'.
O 9‘ 3.? 28?: ‘ {/l .E E = = Q " . I‘ . V 5'.
‘11.. 1.2-11w, <1: ==1:< ’ 1W 21111111116“
.041 é’djé-'=-%-; -. 9: . . t” Fifi": "0 5753‘
In"! ill “6 1., ‘1‘- E-s: =3: :0 6.\"”""0 "'51"

'4. --.'- 1.1-1111 2 .Ir ‘3“: < 5.3-e: '
[Mrs 1‘45? ‘2 ’,- :4; as: 3.41 2 o 11 ll" ‘ \
"118 i 1%.: Zv.‘ 7' 93 ‘ .. \
2 .11.. "I.“ o, ' g} {8 '\\\\§'\8\\‘8\‘0
‘ #1» Ir," 27:: 4i._ \ \‘ \ \ §\\

‘77 \ //- 1 .-
‘h‘ 0; /— ”.42-"w v
III/Ix 24,812,391 9.0"

    
   

 
  
  
 

. I \-‘ \‘
"”|“lll'l'illln’ili///.°
or .2: 09‘ _ m . ‘ m.
...JllﬂlllnlyLLIuIy) , '°'(.’"1'-‘P"-!111\\1«,.
@ ..l'vlrllllﬂtntllltlh'ill ..\ ‘3 . .. / ., ”3'68.
‘8 . I] .“‘- ‘\ .§ _\ ‘-\ . I ///I'/’ /’l/

. 1.111111111111111"

 

    
 

    

‘ V, ./ '0,
\\\-'\\\0\ 9 O a. 4' 4m
\\\.;\\\. 10% ,

,
// C
I A‘\
" l

v \ \\ ‘ V?— :9}? .39,” 3‘47], (7,
' \\“o‘ \\\‘i 8‘: $1 \,, Q, 42")”,2. .,.. M.,,
7‘6; a

o‘ o . \\,. \. .. _
'81:! ‘2 ‘6' 8 ~ °
9'. §.§;;‘

@K on oSb 086

Figure 6.1. The structure of K1,25Pb3_5Bi7.258e15 viewed down the b-axis.

213

these sites Show triple occupancy between A/Pb/M. The remaining metal atom sites in
the framework are distorted octahedra, the degree of distortion being subject to the
identity of the metal. The distortions of Sb octahedra found in Rb1,45Pb3,le7,4SSe15,
K1,45Pb3.1Sb7.458e15, and K2,15Pb1,7Sbg_15Se,5 are more pronounced than those of Bi
octahedra found in K125Pb35Bi7258e15. This is due to the greater stereochemical activity
of the SS2 lone pair electrons of Sb3+ ions. In Rb1.45Pb3,le7,4SSe15, the Sb(3), Sb(4), and
Sb(6) sites, which contain a considerable amount of Pb atoms, Show much milder
distortions with relatively equal Sb-Se distances compared with other “clean” Sb
octahedral sites. This is due to the tendency of Pb atoms to form more regular octahedral
sites. Atom Se(7), which binds to Sb(3), Sb(4), and Sb(6), shows a very high ADP in all
the Sb compounds. Releasing the full occupancy constraint for Se(7) did not change the
occupancy, suggesting the presence of positional ﬂuctuation. Thus Se(7) responds to the
local distortions associated with the mixed occupancy at its immediate neighbors in order
to adjust the coordination environment of Sb or Pb. That means Se(7) displaces
differently depending on whether Sb or Pb occupies the disordered Sb(3), Sb(4), and
Sb(6) sites.

The 8-coordinate sites, which are mostly occupied by Pb atoms, connect NaCl-
and BizTe3- type fragment together. The Pb(1) and Pb(2) sites sites are best described as
bi-capped trigonal prismatic. The ADPs for the Pb sites remain high even after
introducing mixed occupancy with alkali metal ions indicating that either random
positional ﬂuctuation or “rattling” of the ions is occurring in these relatively spacious

pockets. This condition can help the system achieve very low thermal conductivity.

214

The structure of Al+be4-2xM7+xSe15 is closely related to those of B-KzBigselg, and

K2,SBig_SSe14. The local environment of alkali metal ions and the size of the NaCl-type

block in K1,25Pb3_50Bi7,258e15 are also shared by B-KzBigseD and K2.5Bi3,5Se14, see Figure

6.2. Only the size of the BizTe3-type unit in each compound is different. The
relationship between these three compounds can easily be seen if the formulae are broken
down into two parts, the anionic ﬁamework and the alkali metal cations in the tunnels.

For example, in B-K2B138613, one K+ ions is stabilized in the tunnel while the other is

disordered with Bi atoms in the anionic framework as described above. Therefore, the
formula can be written as K+[KBi38e13]' or K+[M9$e13]' (M = K + Bi in the anionic
framework). In the same way, K2,5Bi3,58e14 can be described as K+[MIOSe14]' and

Kl+be4-2xBi7+xSe,5 as K+[M1 1Se,5]' (M = K + Bi + Pb in the anionic framework).

[MSe] [MSel
K+[M98e13]' ———p K+[M1oSe14]' -—-> K"'[M11Se15]'

Therefore, K25Bi358e14 and K1+be4-2xBi7+xSe15 are derived by successively adding neutral

“MSe” units to K2B138e13 as follows.

Therefore, the three compounds are members of the homologous series
K[M9+,.Se1 31"] in both a compositional and structural sense. This structural homology is
ﬂexible enough to preserve the basic framework through successive addition of “MSe”
equivalents by adjusting the width of the BizTe3-type blocks. It would be interesting to

investigate if higher order homologs predicted by the K[M9+nSe13+n] series exist.

215

(a) B-KzBi3Sel3 (b) K2.SBi8.55614

Cdl2 type 3121.93 type NaCl type i ‘ BizTea type
. 9 o 9 ‘ .1)

9‘ ‘2 . . 1' Q . . . a e . .0

. 11312115113 .12. 2

' .J'Q.. . e “38‘ 0‘ ' 0..

.. , ° :3 ° .% 0 .
#:3513592? ‘2." 11*
30%?228932158'5‘: '- ‘9

9 G Fl#ﬁ3§ﬁ~f "(.0 3 at
2211.6: :9 12» . by
3‘ o .0 1“ ,dg‘oa. :‘s ‘8 8..
3‘13‘8521401 2"
"" ’41124ﬁ1‘2m’ "If"

 

Figure 6.2. The structure of (a) B~K2BisSei3, (b) K258113586”, and (c)

K1_25Pb3_5Bi7_258e15 viewed down the b-axis. In each case, NaCl-, BizTe3- and CdIz-type

fragments found on the framework are shown as highlighted by the Shaded areas.

216

(C) K1+be4-2xBi7+x3615

Bi2T93 type

 

 

 

 

 

 

 

 

 

91 3°
Bi 312 Bi i4 3‘5
© 394 $95 $66 397 3°
6 pb 86 S1 1
Se - Bi6 Bi
i7 '

Se12 3" 3° 4 8°15
@
K1

@

 

a

Figure 6.2. (Cont’d). The structure of (a) B—K2B13Sel3, (b) K2.5Bi3,5Se14, and (c)

K1.25Pb3_5Bi7,25Se15 viewed down the b-axis. In each case, NaCl-, BizTe3- and CdIz-type

fragments found on the framework are Shown as highlighted by the shaded areas.

217

The modular construction of this type of compounds is a major underlying
characteristic and makes understanding of their structure and their interrelationships easier
to achieve. It is also expected to be useful in the prediction of new compounds based on

different size and shapes of the modules. The parent structure of course of all modules is

that of NaCl.

Energy Band-Gap and Thermal Analysis.

The A1+be4-2xM7+xSe15 compounds are valance precise and are narrow gap
semiconductors. The band gap was easily detectable spectroscopically in the infrared
region, as a well deﬁned absorption. A typical absorption spectrum revealing the energy-
gap of K1_25Pb3,5Bi7,2sSe15 is shown in Figure 6.3. The band gaps vary from 0.35-0.60 eV
depending on the compound and they are summarized in Table 6.11. The size of the band
gap is critical to the electrical characteristic of a potential thermoelectric material. To be
viable for near room temperature applications a material needs to have E8 of < 0.55 eV,
although a correct band gap in and of itself is not a sufﬁcient property.21 Certainly, the
band gaps observed for most members of this family lie in the desirable range.

DTA experiments indicate that they are thermodynamically stable compounds

melting congruently in the narrow region between 570-690 °C, see Table 6.11 for details.

Charge Transport Properties.
The charge transport properties of K1_25Pb3,50Bi7.25Se15 were measured on single
crystal samples. The room temperature conductivity of K1_25Pb3.50Bi7,258e15 was ~ 260

S/cm, and showed a metallic-like trend increasing between 300 K and 50 K, however it

218

Table 6.11. The Energy Band Gap, Melting Point, and Room Temperature Charge

 

 

Transport Properties.
Compound Eg (RT) melting point on Sm
(eV) (°C) (S/crri (uV/K)
K1,25Pb3,5oBi7,258615 0.53 685 260 -150
Rb,_..5Pb3,ISb7..5Sel5 0.36 578 1.7 x 10'2 1331
K1...5Pb3,ISb7,..SSe15 0.45 576 2.6 x 10‘3 1029
K2_15Pb1,7Sb3,158€15 0.60 576 1.2 X 10.3 1071

 

 

219

reaches a maximum at 50 K and subsequently decreases, see Figure 6.4.

K.,25Pb3_50Bi7_sze.5 shows quite large negative Seebeck coefﬁcients, -150 uV/K at room

temperature. The negative sign indicates that charge carriers are electrons (n-type). The
Seebeck coefﬁcient decreases almost linearly with decreasing temperature trending to zero
at 0 K. This temperature dependence suggests much higher Seebeck values at >300K and
implies that the maximum ZT for this material lies above room temperature. The charge
transport properties are highly dependent on the degree of doping. Varying degrees of
doping can be achieved in several ways, including adding doping impurities deliberately or
changing the synthetic conditions. For example ingot samples of this compound are
prepared differently and exhibit different values of electrical conductivity and
thermopower (see below). Systematic doping studies with this compound will be
reported later.

The charge transport properties of Sb compounds were measured on single crystal
samples. They exhibit semiconducting behavior with room temperature conductivities of
1.7 x 102, 2.6 x 103, and 1.2 x 10'3 S/cm, for RblstbquMsSels, Kl,45Pb,.,Sb7,4SSe,,,
and K2.15Pb1,7Sb3,1sSe15, respectively, see Figure 6. These values are too low for TB
applications. Unlike K1_25Pb3.50Bi7.258e15, the Sb compounds are doped p-type and
possess enormous room temperature thermopower of 1331, 1029, and 1071 uV/K for
R11145103313137.”se 15, K1.4510331311745565, and K2_15Pb1_7Sb3_158615, respectively.
K1_45Pb3,.Sb7_458e15with a higher Pb content Shows higher electrical conductivity than

K2_15Pb1.7Sbg,lsSe1 5 with comparable thermopower (Figures 6.5, 6.6, and 6.7).

220

 

Ot/S (Arbitrary Units)

 

 

l l .L l L l l l l l l l l A l

0.3 0.4 0.5 0.6 0.7

 

 

Energy (eV)
Figure 6.3. Optical absorption spectrum of K1,25Pb3_5Bi7_25Se15. The semiconductor

energy gap is indicated in the spectrum.

 

500 _ , 0
450 :

400_

0' (S/cm)
()I/ATl)S

350 :

 

300 C

 

      

 

I
1
l l l l I l l l I l l l l l l l 1 A l l l I 1 l I

250 . . .
0 50 100 150 200 250 300
Temperature (K)

 

Figure 6.4. Variable temperature electrical conductivity and thermopower for a single

crystal OfK1,25Pb3.5Bl7,258615.

221

 

* 2000

10-2 1500
w .

 

A U)
E ..o A
t
a)? _ °o° -1000 <
v 0° %
b 10'3'.‘ O 1 v
:<——.° .
; .° 1500

W

 

 

l
‘41111111111111111111111O

10-4
180 200 220 240 260 280 300
Temperature (K)

 

Figure 6.5. Variable temperature electrical conductivity and thermopower for a single

crystal Of Rb],45Pb3,1Sb7_45SC]5.

 

10-2

0' (S/cm)
ES

 

 

..1...‘0
280 300

260

l I I I l I I l I I

_6, . . . .
180 200 220 240
Temperature (K)

10

Figure 6.6. Variable temperature electrical conductivity and thermopower for a single

crystal 0f K1,45Pb3.5Sb7,458€15.

222

 

 

 

 

10-2 < 2000
10-3 1500
”a“ 52
3°, 10-4 41000 E
‘6 35
10-5 500
-6 “11--.: ..1...1...1..240

180 200 220 240 260 280 300
Temperature (K)

Figure 6.7. Variable temperature electrical conductivity and thermopower for a single

crystal Of K2_15Pb1_7Sb8_15S€15.

Thermal Conductivity. Unfortunately, it was not possible to measure the thermal
conductivity of small single crystals. For this measurement, large ingots of the material
are required, and therefore the thermal transport properties of KL25Pb350Bi725Se15 were
measured on elongated ingots. The ingot was prepared by melting a stoichiometric
mixture of KzSe, Pb, Bi, and Se metals in a carbon-coated silica tube. The doping levels of
such ingots can vary and usually are different from those of the single crystals discussed
above. For example, the ingot prepared exhibited a much higher electrical conductivity of
~800 S/cm and lower thermopower of -25 uV/K at room temperature, see Figure 6.8.
Therefore the “as made” ingot samples are heavily doped with much larger number of
carriers than the Single crystal samples. Heating the stoichiometrically combined elements
seems to produce defects in the ingot sample(perhaps anti-site) that generate more

carriers (i.e. electrons and holes), whereas the ﬂux method, used to prepare the single

223

crystals, provides relatively Se-rich media, which as well as slow crystal growth
conditions keep the deping level of the single crystals relatively low. The doping effects
associated with this material and their dependence on preparation condition will be

investigated further in the future.

 

 

 

’0
-5
-10 m
I.- A
"E
"-__ -15 <
'. .I ——-> 5
O I. '20
o. I.
.0 ' -25
. O
L11L1....1....1....1...I1...._30
0 50 100 150 2 250 300

Temperature (K)

Figure 6.8. Variable temperature electrical conductivity and thermopower for an

polycrystalline ingot of K._25Pb3_5Bi7_258e15.

The measured thermal conductivity, Kmeas of an ingot of K1_25Pb3_50Bi7_25Se.5 is
shown in Figure 6.9. The room temperature Kmms value is ~1.50 W/m-K. The measured
thermal conductivity can be written as Kmcas = Klan + Kc, + Kmd, where Klan is the lattice

contribution, Kc. electronic contribution, and Krad the apparent contribution due to

224

 

 

 

 

 

 

 

 

1.6
1 J
l- . .
I . 0
A 1.2} .. .
. o
E . . o ‘ .
é ; ..00'. ,_. 1.4,
. E i .I"
V 08- M E I..'--
' - \ 1.2:
>4E ,‘ {E 0.6;-
1 + E
0-4? 3’ 0.2g-
50 100 150 200 250 300
Temperature (K)
O..x.l....l.x.xL.x.jl.x..l...L
O 50 100 150 200 250 300
Temperature (K)

Figure 6.9. Variable temperature thermal conductivity, Kmeas, for an ingot of

K1.25Pb3,5Bi7,258e15. The inset shows total thermal conductivity, Km), corrected for

radiative loss.

225

radiative loss, which originates from non-ideal sample shape and from experimental
artifacts. The latter term increases linearly with T3 becoming signiﬁcant at high

temperatures (>180 K) and has to be properly subtracted from Kmeas in order to determine
the sample’s total thermal conductivity (K1621 = Klan + Kc] ). The Ktotal, after the radiative
loss term is corrected, 22 is shown in the inset of Figure 8. The room temperature Krad is

estimated as 0.24 W/m'K. Thus the total thermal conductivity is ~1.26 W/m-K, which is

16 % lower than that of optimized BizTe3 alloy (Km. ~1.5 W/m°K).23 Given that the

electronic contribution, which is calculated from the electrical conductivity data using the

Wiedeman—Franz law,24 is 0.56 W /m-K, the lattice contribution, roan is estimated to be

only ~0.70 W/m°K at room temperature.
The total conductivity of the single crystal samples is expected to be even lower

because the single crystals have fewer carriers and, therefore, the contribution from the Kc)
part is expected to be smaller. The K61 of the single crystals is calculated to be 0.18 W

/m-K at room temperature based on the Wiedeman—Franz law and their observed electrical

conductivity (see Figure 4). Assuming that the Klan of the single crystals is the same as

that of the ingot (0.70 W/m-K) because it is only affected by the structural features, the

Km] of the single crystals is estimated at ~0.9 W/m°K at room temperature. This thermal

conductivity value is among the lowest reported for potential thermoelectric materials
under investigation, and is well below the value for BizTe3 alloys.

The structural and compositional complexity, the rattling effect of alkali metal and
Pb2+ ions in cavities, and the alloy-type occupancy disorder among the K, Pb, and Bi in

K125Pb350Bi725Se15 all combine to result in a very low lattice thermal conductivity indeed.

226

At this point the ﬁgure of merit (ZT) for the single crystals of K1,25Pb3_50Bi7,sze15 is
estimated to be ~0.2 at room temperature. The ZT of pure BizTe3 is 0.5 and the
optimized BizTe3 alloy is approximately ~0.9. Despite this, a ZT ~0.2 is high enough to
suggest that K1.25Pb3.50Bi7_258e15 should be considered a viable candidate for thermoelectric

research that warrants serious experimental efforts at ZT optimization.

4. Conclusions

A new family of materials A1+be4-2xM7+xSe15 (A = K, Rb; M = Bi, Sb) has been
identiﬁed and characterized. The adopted structure type is novel but related to those of

B-K2Bi38e13 and K2.sBi8_5Se,4. Among the A1+be4-2xM7+xSe15 family, the

K1_25Pb3_50Bi7_sze.5 member is a potential new thermoelectric material because it
possesses a favorable combination of enhanced electrical and thermal transport
properties. The complexity of the structure with its selected Sites exhibiting mixed
occupancy of K/Pb/Bi atoms creates a system in which heat ﬂow is signiﬁcantly
frustrated. Consequently, this system possesses one of the lowest thermal conductivities
reported for any crystalline compound. This and previous work‘s'9 demonstrate that it is
not only possible but actually easy to achieve lower thermal conductivity in quaternary
chalcogenide compounds with complex composition and structures, compared to simpler
high—symmetry binary compounds. The fact that there exist Sb analogs of
K,_25Pb3,5Bi7.258e15 should make systematic Bi-Sb solid solution studies possible, without

disrupting the structure, and enhance the probability of TE property optimization.

227

References

(a) Tritt, T. M.; Kanatzidis, M. G.; Lyon, H. B.; Mahan, G. D. Ed., Mat. Res.
Soc. Symp. Proc. 1997. (b) Tritt, T. M.; Kanatzidis, M. G.; Lyon, H. B.; Mahan,
G. D. Ed., Mat. Res. Soc. Symp. Proc. 1998.

(a) Slack, G. A. “New Materials and Performance Limits for Thermoelectric
Cooling” in CRC Handbook of Thermoelectrics; Rowe, D. M., Ed.; CRC Press:
Boca Raton, 1995, pp. 407-440. (b) Slack, G. A. Solid State Physics; Ehrenreich,
H.; Seitz, F .; Tumbull, D., Ed.; Academic: New York, 1997, 34, 1

(a) Sales, B. C., Mater. Res. Bull.. 1998, 23, 15-21. (b) Sales, B. C.; Mandrus, D;
Williams, R. K., Science 1996, 272, 1325-1328. (c) Sales, B. C.; Mandrus, D;
Chakournakos, B. C.; Keppens, V.; Thompson, J. R, Phys. Rev. B. 1997, 56,
15081-15089.

Tritt, T. M., Science 1996, 272, 1276-1277.

Chung, D.-Y.; Iordanidis, L.; Choi, K.-S.; Kanatzidis, M. G. Bull. Korean Chem.
Soc, 1998, 19, 1283-1293.

Chung, D.-Y.; Jobic, S.; Hogan, T.; Kannewurf, C. R.; Bree, R.; Rouxel, J.;
Kanatzidis, M. G. J. Am. Chem. Soc, 1997, 119, 2505-2515.

Chung, D.-Y.; Choi, K-.S; Iordanidis, L.; Schindler, J. L.; Brazis, F.; Kannewurf,
C. R.; Chen, 8.; Hu, 8.; Uher C.; Kanatzidis, M. G. Chem. Mater., 1997, 9, 3060-

3071.

228

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Chung, D.-Y.; Hogan, T.; Brazis, P. W.; Rocchi-Lane, M.; Kannewurf, C. R.;
Bastea, M.; Uher,C.; Kanatzidis, M. G. Science, 2000, 28 7, 1024-1027.

(a) Kanatzidis, M. G.; McCarthy, T. J.; Tanzer, T. A.; Chen, L. -H.; Iordanidis,
L.; Hogan, T.; Kannewurf, C. R.; Uher, C.; Chen, B. Chem. Mater. 1996, 8, 1465-
1474.

CRC Handbook of T hermoelectrics; Rowe, D. M. Ed., CRC Press, Inc.: Boca
Raton, FL, 1995 and references therein.

Takagi, J .; Takéuchi, Y. Acta Cryst. 1972, 828, 649-651.

Take'uchi, Y.; Takagi, J. Proc. Japan Acad., 1974, 50, 76-79.

Skowron, A; Boswell, F. W.; Corbett, J. M.; Taylor, N. J. J. Solid State Chem.
1994, 112, 307-311.

McCarthy, T. J .; Ngeyi, S.-P.; Liao, J. -H.; Degroot, D.; Hogan, T.; Kannewurf,
C. R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331-340.

Lyding, J. W.; Marcy, H. 0.; Marks, T. J .; Kannewurf, C. R. IEEE Trans.
Instrum. Meas., 1988, 37, 76-80.

Chaikin, P. I.; Kawk, J. F. Rev. Sci. Instrum, 1975, 46, 218-220.

SMART: Siemens Analytical X-ray Systems, Inc., Madison, WI 53719, USA,
1994.

SAINT: Version 4, Siemens Analytical X-ray Systems, Inc., Madison, WI 53719,
USA, 1994-1996.

SADABS : Area-Detector Absorption Correction; Siemens Analytical X-ray
Systems, Inc., Madison, WI 53719, USA, 1996 : G. M. Sheldrick, University of

thtingen, Germany.

229

20.

21.

22.

23.

24.

SHELXTL: Version 5, Siemens Analytical Xray Systems, Inc., Madison, WI
53719, USA, 1994. : G. M. Sheldrick, University of thtingen, Germany.
Mahan, G. J. Appl. Phys. 1989, 65, 1578-1583.

The Km, contribution was estimated as follows. F irst, Kel was calculated from the

electrical conductivity based on the Wiedeman-Franz law over the temperature

range 4-300 K and was subtracted from chas. The remaining part, which is the
sum of Klan and Kmd, was plotted against T. The Klan, which follows a l/(A+BT)

dependence (Rosenberg, H. M. The Solid State 2"d Ed., Clarendon Press Oxford,
197 8, p.95), was modeled based on the fact that radiation losses are negligible

below 70K. Thus the Krad was obtained from Km“, — (Klan + Kel). When Km, is

plotted against T3 a straight line is obtained which is consistent with the expected
T dependence for radiative losses.

Encyclopedia of Materials Science and Engineering; Thermoelectric
Semiconductors; MIT Press: Cambridge, MA; Pergamon Press: Oxford, 1986;
p4968.

Kittel, C. Introduction to Solid State Physics, 6th ed.; John Wiley & Sons, Inc.:

New York, 1986; p150.

230

Chapter 7. Reactions of Sb metal in Alkaline Earth Metal Polysulfide

Fluxes.

Discovery of New Sulfoantimonate Compounds with Alkaline Earth

Metals, sr6Sb6SI7r Ba3Sb4.66slor and Ba2.62PbI.38Sb4S10

231

1. Introduction

For the last several years, we have explored new phases in the broad class of
ternary and quaternary chalcoantimonate compoundsl’z'3’4 We discovered unique
structure types stabilized mainly due to the ability of Sb3+ ions to adopt various local
environments and due to numerous ways available to combine thio- and selenoantimonate
building units and metal centers. The molten ﬂux method has been the tool of choice for
this type of exploratory synthesis. The ﬂux forms by the in situ fusion of AzQ (A =
alkali metals; Q = chalcogen), Sb/Sb2Q3, and chalcogens. Alkali metals from the ﬂux
usually incorporate into the structure and are stabilized between the layers or in the
cavities of the anionic framework. Recently, we extended this approach to alkaline earth
metal chalcogenide ﬂuxes. Alkaline earth metals are smaller and possess higher positive
charge than alkali metals. Therefore, replacing alkali metal ions with alkaline earth metals
in the ﬂux should lead to different structure types. Up to date, only a few alkaline earth
chalcoantimonates have been reported (e.g. Ba4Sb4Se1 1,5 BaSb284,6 Sr3Sb489,7 ABX3(A =
Sr, Ba; B = Sb, Bi; X = Se,Te)8), which suggests that ﬁirther work in this area could
generate interesting new chemistry.

Our investigation using (Sr,Ba)beSy ﬂuxes resulted in the discovery of three new
sulfoantimonate compounds, Sr6Sb6Sl7, Ba3Sb4_66Sm, and Ba2_62Pb1.33Sb4S10. The novelty
in Sr6Sb6Sn lies in the presence of discrete tri-sulﬁde (S3)2' groups, which align in rows to
generate a new non-centrosymmteric structure type. Ba3 $64,668“) and Ba2,6sz,_33Sb4Slo
are similar in structure but not isomorphous. They are closely related to the known lead

arsenosulﬁde minerals, the rathite group, where rathite-I((Pb,Tl)3As4(As,Ag)S,0),9

232

rathite-Ia(Pb3_5As4,5S10),1° and rathite-III(Pb3AssSlo),l° and dufrenoysite(Pb4As4Sm)'1’12
belong. These compounds Share basic structural features with one another but the
difference in composition results in slight modiﬁcation in atomic positions, which makes
each compound to adopt a different space group. For example, rathite-I was reported
with the centrosymmetric space group, P21/a, while dufrenoysite and rathite-Ia were
reported with the noncentrosymmetric Space group, P21 and exhibit strong
piezoelectricity.10’ll However, the published structure reﬁnements for some of these
mineral sulfosalts are incomplete due to the lack of ﬁne structural tools in the early 60’s,
when most of these studies were performed. Among these compounds, Pb4As4Sm is the
only one, whose formula preserves charge neutrality. The others probably require an
introduction of mixed or partial occupancies to certain sites in order to obtain realistic
formulas, which can balance charge.

Ba3Sb4,668.0 and Ba2,62Pb1,33Sb4Sm are analogs of these arsenosulﬁdes with
substitution of Pb by Ba ions partly or entirely. Interestingly, szszSs, which would be
an antimony analog of Pb4AS4810 without any Ba content, was reported to adopt a
different structure type.13 The structures of Bang4,6GSlo and Ba2_62Pb1,33Sb4Sw are
closely related to each other but they exhibit ﬁne alterations in the atomic positions in
order to compensate their different divalent (Ba2+, Pb”) to tri-valent ion (Sb3+) ratios as
in the case of the rathite group compounds. This modiﬁcation again affects the polarity
of these compound and only Ba2,62Pb1,338b4Sm presents a polar non-centrosymmetric
structure. Therefore, a complete structural study on Ba3Sb4l66810 and Ba2.62Pb1_3ng4S10
will help us to understand the general structural features of the rathite group of minerals

such as common trends for disordering or partial occupancy for speciﬁc sites. Here we

233

report the synthesis, structures and physicochemical characterization of Sr68b6Sl7,

Ba3313466510, and 832.62Pb1388b4810-

2. Experimental Section

2.1 Synthesis.

The following reagents were used as obtained: antimony, 99.999%, -200 mesh,
Cerac, Milwaukee, WI; sulfur powder, sublimed, JT Baker Co., Philllipsberg, NJ; lead
metal, 99.999%, 200 mesh, Cerac, Milwaukee, WI; strontium metal, 99%, granules
<6mm, Aldrich Chemical Co., Milwaukee, WI; barium metal, 99%, granules <6mm,
Aldrich Chemical Co., Milwaukee, WI. The BaS was prepared by a stoichiometric
reaction of barium metal and sulfur in liquid NH3.
Sr6Sb6S17. The compound was prepared from a mixture of 0. 1402 g (1.6 mmol) Sr,
0.048 g (0.4 mmol) Sb, and 0.1539g (4.8 mmol) S. The reagents were thoroughly mixed,
ﬂame-sealed in an evacuated silica tube, and heated at 800 °C for 5 days (cooling 2 °C/h).
Pure red irregular Shaped crystals of Sr6Sb6Sl7 were obtained by isolation in degassed
dimethylformamide (DMF) and water (yield ~ 50% based on Sb metal). The crystals are
air- and water stable.
Ba3Sb4,66Sm. The compound was prepared from a mixture of 0.1355 g (0.8 mmol) BaS,
0.144 g (1.2 mmol) Sb, and 0.1283 g (4.0 mmol) S. The reagents were thoroughly mixed,
ﬂame-sealed in an evacuated silica tube, and heated at 720 °C for 5 days (cooling 2 °C/h).

Pure red irregular shaped crystals of Bal,sSb2_33S5 were obtained by isolation in degassed

234

dimethylformamide (DMF) and water (yield ~60% based on Sb metal). The crystals are
air- and water stable.
Bauszl 33Sb4810. The compound was prepared from a mixture ofO. 102g (0.6 mmol)
BaS, 0.124 g (0.6 mmol) Pb, 0.146 g (1.2 mmol) Sb, and 0.106 g (3.3 mmol) S. The
reagents were thoroughly mixed, ﬂame-sealed in an evacuated silica tube, and heated at
800 0C for 5 days (cooling 2 °C/h). Pure silvery red needle-like crystals of
Ba._31Pb0.698b2S5 were obtained by isolation in degassed dimethylformamide (DMF) and
water (yield >90% based on Sb metal).

The compositions of the materials were analyzed by Scanning Electron
Microscope (SEM)/ Energy Dispersive Spectroscopy (EDS). The homogeneity of each
product was conﬁrmed by comparing their powder X-ray diffraction patterns against

ones calculated using X-ray single crystal data.

2.2 Physical Measurements.

Solid-State UV/V is Spectroscopy. Optical diffuse reﬂectance measurements were
performed at room temperature with a Shimadzu UV-3101 PC double-beam, double-
monochromator spectrophotometer operating in the 200-2500 nm region. The instrument
is equipped with an integrating sphere and controlled by a personal computer. BaSO4
was used as a 100% reﬂectance standard for all materials. Samples were prepared by
grinding them to a ﬁne powder and spreading them on a compacted surface of the

powdered standard material and preloaded them into a sample holder. The reﬂectance

235

versus wavelength data were used to estimate a material's band gap by converting

reﬂectance to absorption data as described previously.14

Single Crystal UV/V is Spectroscopy. Optical transmission measurements were
made at room temperature on single crystals using a Hitachi U-6000 microscopic FT
spectrophotometer with an Olympus BH-2 metallurgical microscope over a range of 380-

900nm.

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically, a sample (~20 mg )
of ground crystalline material was sealed in a silica tube under vacuum. A silica tube of

equal mass ﬁlled with A1203 was sealed and placed on the reference side of the detector.
The samples were heated to 800 °C at 5 °C/min, isothermed for 10 min followed by

cooling at -5 °C/min to 50 °C. Residues of the DTA experiments were examined by X-
ray powder diffraction. The stability/reproducibility of the samples was monitored by

running multiple heating/cooling cycles.

Raman Spectroscopy. Raman Spectra were recorded on a Holoprobe Raman
spectrograph equipped with a CCD camera detector using 633nm radiation from a HeNe
laser for excitation. Laser power at the sample was estimated to be about ImW and the
focused laser beam diameter was ca. 10 microns. 5 scans were needed to obtain good

quality spectra. The accumulation time of each scan was 5 sec.

236

2.3 X-ray Crystallography.

Sr6Sb6S1-7. A single crystal with dimensions 0.25 x 0.03 x 0.07 mm was mounted on
the tip of a glass ﬁber and intensity data were collected on a Bruker SMART Platform
CCD diffractometer using graphite monochromatized Mo K01 radiation. The data were
collected at room temperature over a ﬁrll sphere of reciprocal space, up to 28.71 ° in 0.
The individual frames were measured with an omega rotation of 0.3 deg and an acquisition
time of 45 sec. The SMART15 software was used for the data acquisition and SAINTl6
for the data extraction and reduction. The absorption correction was performed using
SADABS.17 Structure solution and reﬁnement were performed with the SHELXTL
package of crystallographic programs.18 The structure of sr6Sbésl7 was solved and
reﬁned successfully in the chiral space group, P2 ,2 12,, which was uniquely deﬁned by the
systematic absence conditions of the data set.

Ba3Sb4,6-,Sm. A single crystal with dimensions 0.27 x 0.17 x 0.06 mm was mounted on
the tip of a glass ﬁber and intensity data were collected on a Bruker SMART Platform
CCD diffractometer using graphite monochromatized Mo K01 radiation. The data were
collected at room temperature over a ﬁrll sphere of reciprocal space, up to 28.25 ° in 0.
The individual frames were measured with an omega rotation of 0.3 deg and an acquisition
time of 40 sec. The data were processed as described above. The structure of A
Ba3Sb4_67Slo was solved and reﬁned in the centrosymmetric P21/c space group. After
successive reﬁnement, one of the Sb atoms, Sb(5), showed an abnormally high atomic
displacement parameter(ADP), implying a partial occupancy for the site. The occupancy

of Sb(5) was reﬁned to be 66.7(3) % without any constraint, and this result gave a

237

nominal composition of Ba3Sb4,67Sm, which satisﬁes charge neutrality. However, the
ADP for Sb(5) remained high, and at this stage we suspected that the Sb atom might be
described as positionally disordered over more than one site. Unlike other 3-coordinate
Sb atoms in the structure, Sb(5) occupies a large 6-coordinate site and may ﬁnd more than
one proper position to sit in this large pocket. A notable electron density peak was
found near the Sb(5) site (0.6A apart) and assigned as Sb(5’). Introduction of the
positional disorder dropped the ADPS to normal values and the ﬁnal occupancies for
Sb(5) and Sb(S’) are 59.5(3)% and 7.2(3)%, respectively.

Ba2,62Pb138Sb4Sm. A single crystal with dimensions 0.31 x 0.10 x 0.04 mm was
mounted on the tip of a glass ﬁber and intensity data were collected on a Bruker SMART
Platform CCD diffractometer using graphite monochromatized Mo K01 radiation. The
data were collected at room temperature over a full sphere of reciprocal space, up to
2900" in 0. The individual frames were measured with an omega rotation of 0.3 deg and
an acquisition time of 60 sec. The data were processed as explained above. The structure
of Ba2_62Pb1.3ng4Sm was solved and reﬁned successfully in the chiral P2, space group.
After successive reﬁnement, all the Pb sites showed high ADPS and mixed occupancy
with Ba was introduced to these sites. The 8(1) and 8(5) sites also showed unusually
high ADPS, which could not be dropped by reﬁning the occupancies. Near both sites,
prominent electron density peaks were found and assigned as S(l ’) and S(5’),
respectively. The sum of the occupancies of 8(1) and S(l ’) was constrained to be the
maximum allowed in these crystallographic sites and so was that of the occupancies of
8(5) and S(S’). The introduction of positional disorder decreased ADPS to the reasonable

values.

238

An attempt to solve the structure in the centrosymmetric space group, P21/m was
made but it failed because the structure obviously does not possess mirror planes
perpendicular to the b-axis, see description below. Solving the structure in the
centrosymmetric Space group, P21/c was also considered because its analog, Ba3Sb4,67810
adopts the same space group. However, the systematic absence conditions showed that
the c-glide plane does not exist in this structure (22% of the hOI reﬂections with l = 2n+1,
which must be systematically absent if c-glide planes exist, are observed with the
intensities greater than 30 ). Nevertheless, we still attemted to solve and reﬁne the
structure in P2l/c. This gave a signiﬁcantly higher values of R1 = 14.05 and wR2 = 29.86
for all data.

The complete data collection parameters and details of the structure solution and
reﬁnement for each compound are given in Table 7.1. The coordinates of all atoms,
isotropic temperature factors, and their estimated standard deviations (esd’s) for each
compound are given in Tables 7.2, 7.3 and 7.4. The selected bond distances for each

compound are given in Tables 7.5, 7.6, and 7.7.

3. Results and Discussion

Structures.
Sr6Sb6Sl-7 The overall structure viewed down the b-axis is shown in Figure 7.1. This

compound is composed of Sr2+ ions stabilized between [Sb3S7]5' units and tri-sulﬁde
groups, (S3)2'. Therefore, the formula can be written as (Sr2+)6[(Sb3S7)5']2(S32) There

exist six independent Sb atoms, all of which are stabilized in trigonal pyramidal sites.

239

Table 7.1. Summary of Crystallographic Data and Structural Analysis for Sr6Sb68n,
332.62Pb1383b4310, and Ba3Sb466SIO-

 

Formula

Formula weight

81158136817
1801.24

BaZ(BalaSb0.66)Sb4SIO
1301.19

Baz(Pb1.38,Bao.62)Sb4Slo
1453.88

Crystal habit Red, irregular shape Red, irregular shape Silver needles
Space group P212121 P21/c P21
a, A 8.2871(9) 8.955(2) 8.8402(2)
b, A 15.352(2) 8.225(2) 8.2038(2)
c, A 22.873(3) 26.756(5) 26.7623(6)
B. degree- N.A. 100.29(3) 99.488(1)
2; v, A3 4; 2909.9(5) 4, 1939.0(7) 4, 1914.34(8)
Deane, g/cm3 4.112 4.457 5.045
Temp, K 298(2) 170(1) 298(2)
K(Mo K0114 0.71073 0.71073 0.71073
11(Mo K01). cm" 175.90 134.47 240.60
F(OOO) 3224 2265 2496
6...... deg 28.71 28.25 29.00
Total data 23682 15455 17144
Unique data 6985 4618 9089

[R.,,.= 0.0541] [R.,..= 0.0332] [R,,.= 0.0501]
No. of variables 263 169 339
Final R indices R1“ = 0.0321 R1 = 0.0262 R1 = 0.0579
[I>201 wR2b = 0.0645 wR2 = 0.0602 wR2 = 0.1509
R indices R1 = 0.0446 R1 = 0.0294 R1 = 0.0740
(all data) wR2 = 0.0677 wR2 = 0.0613 wR2 = 0.1586
GOF on P"- 0.978 1.106 0.017
BASF 0.058(8) N.A. 0.501(12)

 

‘ R1=21|F0|-|Fc||/2|F0|

240

b wR2={231M193“1'762)21/2111’(1'702)21}“2

Table 7.2. Fractional Atomic Coordinates and Equivalent Isotropic Displacement

Parameter Values for Sr6SbéSl7 with Estimated Standard Deviations in Parentheses.

 

 

Atom x y z Ueq,a A2
Sb(1) 0.0002(1) 0.3071(1) 0.2496(1) 0.015(1)
Sb(2) 0.4760(1) 0.2911(1) 0.2318(1) 0.018(1)
Sb(3) 0.5113(1) 0.1814(1) 0.0852(1) 0.017(1)
Sb(4) 0.0087(1) 0.6747(1) 0.4399(1) 0.019(1)
Sb(5) 0.5330(1) 0.4224(1) 0.0917(1) 0.019(1)
Sb(6) 0.0164(1) 0.4281(1) 0.1077(1) 0.018(1)
Sr(1) 0.2444(1) 0.1655(1) 0.3764(1) 0.012(1)
31(2) 0.2489(1) 0.6665(1) 0.1247(1) 0.013(1)
Sr(3) 0.2501(1) 0.0498(1) 0.2126(1) 0.017(1)
Sr(4) 0.2462(1) 0.5547(1) 0.2890(1) 0.020(1)
Sr(5) 0.2498(1) 0.0596(1) 0.5449(1) 0.013(1)
Sr(6) 0.2433(1) 0.4429(1) 0.4566(1) 0.014(1)
S(l) 0.2916(3) 0.1467(2) 0.6834(1) 0.019(1)
S(2) 0.2502(3) 0.3448(2) 0.3067(1) 0.016(1)
S(3) 0.0011(3) 0.1540(1) 0.2793(1) 0.013(1)
8(4) 0.5001(3) 0.1503(1) 0.2810(1) 0.014(1)
S(5) 0.2861(3) 0.2316(2) 0.1567(1) 0.015(1)
S(6) 0.5016(3) 0.0294(1) 0.1192(1) 0.014(1)
8(7) 0.7285(3) 0.2288(2) 0.1489(1) 0.017(1)
S(8) 0.2763(3) 0.3723(2) 0.0048(1) 0.018(1)
S(9) 0.0036(3) 0.0384(1) 0.1169(1) 0.014(1)
S(10) 0.7218(3) 0.3761(2) 0.0101(1) 0.016(1)
S(l 1) 0.4930(3) 0.0680(1) 0.4460(1) 0.012(1)
S(12) 0.2676(3) 0.0272(2) 0.8363(1) 0.016(1)
S(13) 0.0081(3) 0.5623(1) 0.0511(1) 0.014(1)
S(14) 0.2381(3) 0.4765(2) 0.1698(1) 0.021(1)
S(15) 0.4991(3) 0.2959(1) 0.4285(1) 0.014(1)
S(16) 0.1398(2) 0.2488(1) 0.5019(1) 0.012(1)
S(17) 0.0048(3) 0.2938(1) 0.4309(1) 0.013(1)

 

eq is deﬁned as one third of the trace of the orthogonalized Ug- tensor.

241

Table 7.3.. Fractional Atomic Coordinates and Equivalent Isotropic Displacement

Parameter Values for Ba3Sb4,66S10 with Estimated Standard Deviations in Parentheses.

 

 

Atom x y z Ueq, A2
Ba(l) 0.0910(1) 0.6183(1) 0.4317(1) 0.008(1)
Ba(2) 0.5276(1) 0.2686(1) 0.2992(1) 0.007(1)
Ba(3) 0.0233(1) 0.2673(1) 0.2974(1) 0.007(1)
Sb(1) 0.7044(1) 0.3346(1) 0.1606(1) 0.007(1)
Sb(2) 0.2420(1) 0.3132(1) 0.1515(1) 0.009(1)
Sb(3) 0.1421(1) 0.6501(1) 0.0590(1) 0.009(1)
Sb(4) 0.3262(1) 0.1549(1) 0.4554(1) 0.015(1)
Sb(5)“ 0.4256(1) 0.0357(1) 0.0566(1) 0.019(1)
Sb(S') 0.4183(8) , 0.052(1) 0.0770(5) 0.019(1)
S(l) 0.0813(2) ' 0.6560(2) 0.3164(1) 0.009(1)
S(2) 0.4928(2) 0.1607(2) 0.1801(1) 0.010(1)
8(3) 0.7403(1) 0.4980(2) 0.2366(1) 0.009( 1)
S(4) 0.2471(1) 0.4718(2) 0.2265(1) 0.008(1)
S(5) 0.3754(1) 0.5049(2) 0.1023(1) 0.009(1)
S(6) 0.0424(2) 0.0061(2) 0.4005(1) 0.01 1(1)
S(7) 0.1777(2) 0.6192(2) 0.6162(1) 0.010(1)
3(8) 0.3144(2) 0.3493(2) 0.3870(1) 0.010(1)
S(9) 0.1736(2) 0.1850(2) 0.0058(1) 0.011(1)
S(10) 0.5833(2) 0.2066(2) 0.0050(1) 0.015(1)

 

aSb5 and Sb5' are positionally disordered with occupancy of 59.5(3)% and 7.2(3)% ,

respectively.

242

Table 7.4. Fractional Atomic Coordinates and Equivalent Isotropic Displacement

Parameter Values for Ba2,62Pb1,338b4Sm with Estimated Standard Deviations in

 

 

Parentheses.

Atom x y z Ueq, A2
Pb(1)/Ba(1)a 4124(1) 3005(2) 8241(1) 34(1)
Pb(2)/Ba(2)b 947(2) 8216(2) 1868(1) 33(1)
Pb(3)/Ba(3)c 5994(1) 4078(2) 6814(1) 39(1)
Pb(4)/Ba(4)d 815(1) 4091(2) 6805(1) 39(1)

Ba(5) 5220(2) 1 100(2) 451(1) 28(1)

Ba(6) 228(2) 1022(2) 475(1) 31(1)

Ba(7) 5224(2) 1097(2) 5472(1) 25(1)

Ba(8) 256(2) 1092(2) 5470(1) 25(1)

Sb(1) 2836(2) 5513(3) 960(1) 36(1)

Sb(2) 1840(2) 523(3) 9030(1) 36(1)

Sb(3) 8305(2) 2302(3) 1999(1) 35(1)

Sb(4) 3707(2) 2297(3) 2013(1) 36(1)

Sb(5) 1427(2) 4973(3) 2912(1) 43(1)

Sb(6) 6541(2) 4893(2) 3068(1) 23(1)

Sb(7) 6939(3) 1630(3) 3985(1) 53(1)

Sb(8) 2042(2) 1739(2) 4097(1) 20(1)

S( 1 )e 4950(40) 7260(40) 709(12) 53(5)

S( l ') 5690(40) 7260(30) 686(1 1) 53(5)

S(2) 413(9) 7146(9) 663(3) 40(2)
8(3) 2574(7) 3887(1 1) 192(3) 36(2)
S(4) 7620(7) 3910(1 1) 202(3) 37(2)
8(5) 298(13) 3629(15) 1474(5) 42(2)
S(5')f 1240(20) 3650(20) 1484(8) 42(2)
3(6) 5766(12) 3786(10) 1494(3) 64(3)
S(7) 8073(8) 19(1 1) 1373(3) 35(2)
S(8) 3355(7) 55(10) 1371(3) 31(2)
S(9) 6230(12) 633(12) 2543(3) 63(3)
SLIOL 1307(1 1) 672(12) 2538(4) 62(3)

 

243

Table 7 .4. (Cont’d).

S(l 1) 3372(6) 6519(8) 2464(2) 24(1)
S(12) 91 1(6) 1559(8) 7542(2) 24(1)
S(13) 1845(7) 6931(8) 3640(3) 25(1)
S(14) 3144(7) 2273(9) 6372(3) 30(2)
S(15) 4751(5) 3517(7) 3532(2) 19(1)
S(16) 8797(5) 3510(6) 3530(2) 17(1)
S(17) 7369(6) 3136(7) 4751(2) 19(1)
S(18) 2425(7) 3380(7) 4856(2) 22(1)
S(19) 4286(6) 4(7) 4308(2) 18(1)
S(20) 38@) 0(6) 4310(2) 13(1)

 

“ 72(1)% Pb(1) / 28(1)% Ba(l)
b 36(l)% Pb(2) / 63(1)% Ba(2)
° 83(1)% Pb(3)/ 17(1)% Ba(3)
d 87(l)% Pb(4)/ 13(1)% Ba(4)
c49(3)% S(1)/ 51(3)% S(l’)
f63(1)% 8(5) / 37(1)% S(S’)

244

Table 7.5. Selected Distances (A) and Bond Angles (°) for Sr68b6S17.

 

 

Bond Distances
Sb(1)-8(1) 2.417(2) Sb(4)-S(8) 2.429(2)
Sb(1)-S(3) 2.445 8(16) Sb(4)-S(9) 2.4639(16)
Sb(1)-S(2) 2.517(2) Sb(4)-S(10) 2.615(2)
Sb(1)-S(5) 3.387(2) Sb(4)-S(7) 3.092(3)
Sb(1)-S(7) 3 .43 6(3) Sb(4)-8(5) 3.407(2)
Sb(2)-S(4) 2.4460(16) Sb(5)-S(11) 2.4066(16)
Sb(2)-S(5) 2.502(2) Sb(5)-S(10) 2.538(3)
Sb(2)-S(2) 2.668(2) Sb(5)-S(12) 2.662(2)
Sb(2)-S(7) 2.981(2) Sb(5)-S(8) 3.012(3)
Sb(2)-8(1) 3.394(3) Sb(5)-S(14) 3.138(3)
Sb(3)-S(7) 2.428(2) Sb(6)-S(13) 2.4345(16)
Sb(3)-S(6) 2.4597(17) Sb(6)-S(14) 2.438(3)
Sb(3)-S(5) 2.598(2) Sb(6)—S(12) 2.523(2)
Sb(3)-S(8) 3.120(3) Sb(6)-S(8) 3.304(2)
Sb(3)-S(10) 3.359(3) Sb(6)-S(10) 3.404(2)
Sr(1)-S(1 1) 3.003(2) Sr(4)-S(14) 2.981(3)
Sr(1)-S(3) 3.006(2) Sr(4)—S(6) 2.987(2)
Sr(1)-S(4) 3.050(2) Sr(4)-S(3) 2.994(2)
Sr(1)-S(17) 3.063(2) Sr(4)-S(9) 2.998(2)
Sr(1)-S(12) 3.099(3) Sr(4)-S(4) 3.023(2)
Sr(1)-S(13) 3.104(2) Sr(4)-S(7) 3.034(3)
Sr(1)-S(1 5) 3.143(2) Sr(4)-S(2) 3.247(3)
Sr(1)-S(2) 3.183(3) Sr(5)-S(6) 3.016(2)
Sr(1)-S(16) 3 260(2) Sr(5)-S(9) 3.024(2)
Sr(2)-S(4) 3.007(2) Sr(5)-S(1 1) 3.032(2)
Sr(2)-S(3) 3.025(2) Sr(5)-S(13) 3.064(2)
Sr(2)-S(l3) 3.062(2) Sr(5)-S(10) 3.093(3)
Sr(2)—S(1 1) 3.079(2) Sr(5)-S(15) 3.100(2)
Sr(2)-S(14) 3.094(3) Sr(5)-S(17) 3.136(2)
Sr(2)-S(15) 3.130(2) Sr(5)-S(16) 3.200(2)
Sr(2)-S(17) 3.141(2) Sr(5)-S(1) 3.456(3)
Sr(2)-S( 1) 3.185(3) Sr(6)-S(13) 2.987(2)
Sr(2)-S(16) 3.230(2) Sr(6)-SQ) 3.026(2)

 

245

Table 7.5. (Cont’d).

Sr(3)-S(9) 3.001(2) Sr(6)-S(6) 3.039(2)
Sr(3)-S(6) 3.002(2) Sr(6)-S(8) 3.048(3)
Sr(3)-S(4) 3.019(2) Sr(6)—S(1 1) 3.048(2)
Sr(3)-S(3) 3.023(2) Sr(6)-S(17) 3.080(2)
Sr(3)-S(12) 3.069(3) Sr(6)-S(15) 3.162(2)
Sr(3)-S(5) 3.084(3) Sr(6)-S(16) 3.270(2)
Sr(3)-S(1) 3.1 10(3) Sr(6)-S(2) 3.746(2)
S(15)-S(16) 2.090(3) S(16)-S(17) 2.089(3)

 

246

Table 7.6. Selected Distances (A) and Bond Angles (°) for Ba3Sb4,6(,Slo.

 

 

Bond Distances
Ba( 1 )-S( 1) 3.0861(14) Sb(1)-8(1) 2.4078(14)
Ba(1)-S(9) 3.1815(16) Sb(1)-8(3) 2.4115(14)
Ba(1)-S(10 3.1832(17) Sb(1)-S(2) 2.5027(14)
Ba(1)-S(7) 3.1857(15) Sb(1)-S(6) 3.3349(15)
Ba(1)-S(9) 3.1902(15) Sb(1)-3(5) 3.3809(16)
Ba(1)-S(6) 3.3059(15) Sb(2)-S(4) 2.3861(13)
Ba(1)-S(8) 3.3442(15) Sb(2)-8(5) 2.4920(14)
Ba(2)-S(2) 3.2687(16) Sb(2)-S(2) 2.5671 (14)
Ba(2)-S(2) 3.2826(14) Sb(2)-8(6) 3.1 107(16)
Ba(2)-S(3) 3.2869(14) Sb(2)-8(1) 3.4167(15)
Ba(2)-S(7) 3.2877(17) Sb(3)-S(7) 2.4223(14)
Ba(2)-S(4) 3.3 161 (14) Sb(3)-8(6) 2.43 86( 1 4)
Ba(2)-S(3) 3.3366(15) Sb(3)-8(5) 2.5039(15)
Ba(2)-S(4) 3.3375(16) Sb(3)-S(9) 3.3294(17)
Ba(2)-S(8) 3.3487(16) Sb(3)-S(10) 3.4459(17)
Ba(2)-8(5) 3.4044(15) Sb(4)-S(8) 2.4179(14)
Ba(3)-S( 1) 3.1572(15) Sb(4)-S(9) 2.4652(15)
Ba(3)-S( 1) 3.2644(14) Sb(4)-S(10) 2.6954(16)
Ba(3)-S(8) 3.2821(17) Sb(4)-8(6) 2.9633(16)
Ba(3)-S(3) 3.3011(14) Sb(5)-S(10) 2.5645(19)
Ba(3)-S(7) 3 .3096(1 6) Sb(5)-S(10) 2.579(2)
Ba(3)-S(3) 3.3461(15) Sb(5)-S(9) 2.7122(17)
Ba(3)-S(4) 3.4103(14) Sb(5)-S(8) 2.9645(17)
Ba(3)-S(4) 3.4349(15) Sb(5)-S(7) 3.2175(19)
Ba(3)-S(6) 3 .4776(1 5) Sb(5)-S(2) 3.409(2)
Sb(5’)-S(9) 2.852(8)
Sb(5’)-S(2) 2.864(13)
Sb(5’)-S(10) 2.920(1 1)
Sb(5’)-S(7) 2.921(9)
Sb(5’)-S(8) 2.936(7)
Sb(5’)-S(10) 3.054(12)
Sb(5)-8M5 ’) 0.575(12)

 

247

Table 7.7. Selected Distances (A) and Bond Angles (°) for Ba2,62Pb1,3SSb4S10.

 

 

Bond Distances

Pb(1)/Ba(1)-S(8) 2.848(7) Sb(1)-S(3) 2.429(8)
Pb(1)/Ba(1)-S(1) 2.86(3) Sb(1)-S(1) 253(3)
Pb(1)/Ba(1)-S(7) 2.869(8) Sb(1)-S(2) 2.540(8)
Pb(1)/Ba(1)—S(1') 291(3) Sb(1)-S(5') 263(2)
Pb(1)/Ba(1)-S(9) 2.989(9) Sb(1)-8(6) 3 .089(1 0)
Pb(1)/Ba(1)-S(12) 3.350(6) Sb(1)-S(1') 3.09(3)
Pb(1)/Ba(1)-S(11) 3.365(6) Sb(1)-S(5) 3.214(12)
Pb(2)/Ba(2)-S(10) 2.681(9) Sb(2)-S(4) 2.427(9)
Pb(2)/Ba(2)-S( 12) 2.81 1 (6) Sb(2)-S( 1 ') 262(3)
Pb(2)/Ba(2)—S(1 1) 2.820(6) Sb(2)-S(2) 2.637(8)
Pb(2)/Ba(2)-S(7) 3.048(7) Sb(2)-8(5) 2.638(1 1)
Pb(2)/Ba(2)-S(8) 3.084(7) Sb(2)-S(6) 3.074(10)
Pb(2)/Ba(2)-S(2) 3.298(8) Sb(2)-S( 1) 315(3)
Pb(3)/Ba(3)-S(1 1) 2.846(6) Sb(2)-S(5') 3.233(17)
Pb(3)/Ba(3)-S(14) 2.993(7) Sb(3)-S(7) 2.498(9)
Pb(3)/Ba(3)-S(13) 2.999(7) Sb(3)-S(5) 2.661 (1 2)
Pb(3)/Ba(3)-S(10) 3.008(10) Sb(3)-8(6) 2.710(10)
Pb(3)/Ba(3)-S(19) 3.067(6) Sb(3)-S(9) 2.871(9)
Pb(3)/Ba(3)-S(9) 3.093(9) Sb(3)-S(10) 3.106(10)
Pb(4)/Ba(4)-S( l 2) 2.855(6) Sb(3)-S(5’) 3 .32(2)
Pb(4)/Ba(4)-S(14) 2.931(7) Sb(4)-S(8) 2.502(8)
Pb(4)/Ba(4)-S(13) 3.028(7) Sb(4)-S(5') 2.637(17)
Pb(4)/Ba(4)-S(20) 3.048(5) Sb(4)-8(6) 2.750(10)
Pb(4)/Ba(4)-S(10) 3.063(10) Sb(4)-S(9) 2.795(10)
Pb(4)/Ba(4)—S(9) 3.156(10) Sb(4)-S(10) 3 .03 8( 1 O)
Ba(5)—S(1') 3.16(3) Sb(4)-S(5) 3.305(11)
Ba(5)-S( 1) 3 23(3) Sb(5)-S(13) 2.504(7)
Ba(5)-S(1') 3 23(3) Sb(5)-S(12) 2.572(6)
Ba(5)-S(1) 324(3) Sb(5)-S(1 1) 2.586(6)
Ba(5)-S(3) 3 262(8) Sb(5)-S(16) 3 294(5)
Ba(5)-S(4) 3 276(8) Sb(5)-S(15) 3.347(6)
Ba(5)-S(8) 3 294(7) Sb(6)-S(15) 2.443(5)
Ba(5)-S(4) 3 337(8) Sb(6)-S(16) 2.446(5)

 

248

Table 7.7. (Cont’d).

 

Ba(5)-S(3)
Ba(5)-S(7)
Ba(5)-S(6)
Ba(6)-S(2)
Ba(6)-S(2)
Ba(6)-S(4)
Ba(6)-8(3)
Ba(6)-S(3)
Ba(6)-S(4)
Ba(6)-S(7)
Ba(6)-8(5)
Ba(6)-S(8)
Ba(6)-S(5')
Ba(7)-S(19)
Ba(7)-S(18)
Ba(7)-S(19)
Ba(7)-S(13)
Ba(7)-S(18)
Ba(7)-S(17)
Ba(7)-S(17)
Ba(7)-S(15)
Ba(7)-S(14)
Ba(8)-S(20)
Ba(8)-S(l 8)
Ba(8)-S(20)
Ba(8)-S(18)
Ba(8)-S(17)
Ba(8)-S(13)
Ba(8)-S(14)
Ba(8)-S(17)
Ba(8)-S(16)

3.342(8)
3.344(8)
3.527(8)
3.145(8)
3.219(8)
3.303(8)
3.305(8)
3.307(8)
3.322(8)
3.404(7)

3.415(13)

3.439(7)
346(2)

3.217(6)
3.265(6)
3.275(6)
3.286(7)
3.317(6)
3.326(6)
3.365(6)
3.401(6)
3.403(7)
3.206(6)
3.263(6)
3.278(6)
3.306(6)
3.324(6)
3.325(7)
3.356(8)
3.377(6)
3.408(6)

Sb(6)-S(14)
Sb(6)-S(1 1)
Sb(6)-S(12)
Sb(7)-S(17)
Sb(7)-S(15)
Sb(7)-S(16)
Sb(7)-S(l9)
Sb(7)-S(20)
Sb(8)-S(18)
Sb(8)-S(20)
Sb(8)-S(19)
Sb(8)-S(16)
Sb(8)-S(15)
S(1)-S(l')

S(5)-S(5')

249

2.449(7)
3.278(6)
3.289(6)
2.370(6)
2.615(6)
2.686(5)
2.949(6)
3.043(5)
2.414(6)
2.415(5)
2.431(5)
3.348(5)
3.369(5)
067(3)

0.833(19)

 

 

 

 

 

 

 

 

 

Figure 7.1. The structure of Sr6Sb6S17 viewed down the b-axis. The inset shows an

asymmetric unit with the atomic labeling scheme.

250

 

A

(a) T S b)

S

 

 

 

1 ‘C- 90 ° Rotation

S15 S17

S16
3% V939 3%
A

 

 

 

 

 

 

 

 

[100]

 

Figure 7.2. (a) Propagation of a single slab along the a-axis The dashed lines represent
weak Sb-S interactions between (Sb3S7)5' units. (b) Arrangement of the tri-sulﬁde

groups.

251

Sb(1), Sb(2), and Sb(3) pyramids are joined together by sharing corners to form (Sb3S7)5'
units and so are Sb(4), Sb(5), and Sb(6), see the inset of Figure 7.1. This unit has a ﬁnite
length and does not form an inﬁnite chain. The closest distance between the Sb atom in

one unit and sulfur atoms in the neighboring units ranges from 2.981 A to 3.436 A, which

are much longer than the average Sb-S bond distance of 2.500 A in the (Sb3Sy)5' unit.
Nevertheless, if we take this weak Sb-S interaction into consideration, these units
associate to form one-dimensional ribbon-like slabs parallel to the a-axis, see Figure 72(a).
Two such slabs are paired by Van der Waals interactions to form a sandwich-like blocks
parallel to the a-axis. Between these blocks Sr2+ ions are stabilized either in a tri- or in a
mono-capped trigonal prismatic site.

A novel feature of this compound is the presence of discrete tri-sulﬁde groups,
(S3)2' , which are embedded between these Sr2+ ions. The tri-sulﬁde group is bent with
the S-S-S bond angle of 11375" and the S-S distances of 2.09 A. These values are similar
to those observed in BaS3, C5283, Rb2S3, and K2S3. 1930’” The arrangement of tri-sulﬁde
groups are shown in Figure 72(b). These arrow-shaped anions are well aligned in one
direction forming a polar row. The polarity, however, is cancelled by a neighboring row,
which runs in the opposite direction making the compound non-polar, albeit it remains
non-centrosymmetric.

Ba3Sb4,66Sw and Bazanung4Sm Interestingly, these two compounds
represent a cetrosymmetric and a polar version of the same structural type. The
Ba3Sb4.66S10 is composed of Ba2+ ions stabilized between two-dimensional layers
perpendicular to the c-axis. The overall structure of Ba3Sb4_66S10 viewed down the a-axis

is shown in Figure 7.3(a). Each layer is built from one-dimensional slab parallel to the a-

252

 

 

 

 

" $8 "9" " [100]

SM W}! '
o 4 S S {“ .
0
Sb 3” $02,. .
o ‘6
$20 .8335 ﬂag. . ..

/
. Sb1 .9”)
$1 a . . ‘. C . .
, ’45

Figure 7.3. (a) The structure of Ba3Sb4,66810 viewed down the a-axis. (b) Propagation of

a single slab along the a-axis.

253

axis stitched together by Ba(l) and Sb(5/5’) ions. A single slab is highlighted with a
shaded box. A single slab along the a-axis is shown in Figure 7.3(b) with the atomic
labeling scheme. There exist ﬁve independent Sb atoms in the structure. Sb(1), Sb(2), and
Sb(3) are stabilized in 3-coordinate trigonal pyramidal sites. These pyramids share
comers with one another to from (Sb3S7)5’ units, as seen in Sr6Sb6Sn. Sb(4) forms its
own pyramid, which is weakly attached to the (Sb3S7)5' unit since Sb(4) has a moderate
interaction with the 4th nearest sulfur atom, S(6) from the (Sb3S7)5' unit (2.96A). These
(Sb481 .)'°’ fragments then are connected together by high coordinate Sb(5/5’) and Ba(l)
atoms in order to form a one-dimensional slab along the a-axis. The local geometry of the
6-coordinate Sb(5/5’) and 7-coordinate Ba(l) are shown in Figure 7.4. The Sb(5/5’) and
Ba(l) atoms are stabilized in basically the same local environment. The difference is that
seventh nearest sulfur atom from Sb(5/5’) is too far away to be considered
bonding(~3.9A). The Sb(5/5’) atom is shifted relative to Ba(l) in order to more tightly
bind to six sulfur atoms. For the smaller Sb atom, even a 6-coordinate site is too spacious
causing the Sb to be positionally disordered over two sites, Sb(5) and Sb(5’). The
Sb(5/5’) and Ba(l) atoms play an important role in forming the structure because they
serve as stitching points for the (Sb481 1)”). units to build a single slab, see Figure 7.3(a).
Between these layers, Ba(2) and Ba(3) atoms are stabilized in 9-coordinate tri-capped
trigonal prismatic sites, see Figure 7.4(b) and (c).

The compound can be described as (Ba2)(Ba1,Sboi67)(Sb4S10) to break down cations
into three different groups depending on their local environments; The ﬁrst group is the
cations stabilized in the 9-coordinate sites, the second group is in the (6+1)-coordinate

sites, and the third group is in the 3-coordinate sites.

254

 

Figure 7.4. The local environment of the Sb(5/5’), Ba(1), and Ba(2) atoms in
383513466310

255

Ba2_62Pb1.38Sb4S10 is structurally related to Ba3Sb4,66S10, see Figure 7.5. The
formula can be described as Ba2(Ba0.62,Pbl,33)Sb4Sm by the rules above, which gives a
better idea of how these two compounds are related to each other. From the formula, the
9-coordinate Ba sites and 3-coodinate Sb sites remain intact. It is the (6+1) coordinate
sites that can accommodate mixed occupancy and presumably generate a variety of
compositions. The Pb atoms are segregated in the (6+1)-coordinate sites and all are
disordered with Ba atoms in different proportions, see Table 7.4. The coordination
environments of the Pb atoms are similar to those of Sb(5) in Ba3Sb4,66S10., These sites
are now fully occupied by di-valent Pb2+/Ba2+ while the Sb(5/5’) site in Ba3Sb4_66Sm is
only two thirds occupied by the tri-valent Sb3+ so as to balance the charge. The cationic
arrangement pattern of the compound is easily understood by considering the size of
cations. The larger Ba2+ ions prefer the larger 9-coordinate sites ﬁrst and then the
remaining ones are placed along with Pb atoms in the next larger (6+1)-coordinate sites.
The disorder between Pb and Ba seems to be responsible for the generally high ADPS of
the sulfur atoms around these sites. Unlike the Ba2+ ion, the Pb2+ ion possesses its 652
lone pair, which tends to distort the local environment. Therefore, the lone pair may
force the Slllﬁll‘ atoms to adjust their positions slightly differently from the positions
they take when Ba2+ is present and even causes a positional disorder for S(1) and 8(5).

While Ba2,62Pb1,338b4S10 shares the same basic structural features with
Ba3Sb4.66S10, the atomic positions of the former deviate enough from those of the latter to
destroy the c-glide plane perpendicular to the b-axis. As a result Ba2_62Pb1.33Sb4Sm lacks a
center of inversion and adopts the polar space group, P21. In order to visualize the

difference in these two compounds, a center of inversion of the Ba3Sb4.66$10 structure and

256

 

 

 

 

 
  

.20
~ Sb8 1
36 S15 9. 9

'5' 8137
S16 S18
35 ‘- 7 Ba8 .0517
O O C
86 3'34 39 Pb3 $14 887

    
  
 
 
    
  

 

 

 

Figure 7.5. The structure of Ba2.62Pb1,33Sb4Sm viewed down the a-axis. The inset shows

an asymmetric unit with the atomic labeling scheme.

257

the potential center of inversion of the Ba2.62Pb1.3ng4S10 structure are x—marked in each
unit cell, see Figures 7.2 and 7.5. A careful inspection of these ﬁgures reveals that most
of the atoms in Ba2.62Pb1_338b4Sm and their “potential centrosymmetric pairs” are not well

related by the “center of inversion”.

Relationship to Sulfosalt Minerals.

These two compounds are closely related to the mineral lead sulfoarsenides,
(Pb,Tl)3As4(As,Ag)S.0, Pb3.5As4_5S10, Pb3A55S10, and Pb4As4S10, which are classiﬁed as
the rathite group. These rathite group compounds have the same basic framework as
those of Ba3Sb4,66810 and Baziészmng4Sm regardless of their compositional differences.
However, the atomic positions of each compound are altered enough to make each
compound adopt different space groups. The cell parameters and the space groups for
the minerals and the compounds reported here are shown in Table 7.8. All compounds
possess 21 screw axes parallel to the b-axis but the b-axes of dufrenoysite and rathite-Ia
are differently chosen from the rest of the compounds, so the screw axes in these
compounds lie in different directions. It is also worthwhile to notice that Rathite-I,
Ba3Sb4_66S10 and Ba2,62Pb1.3ng4S10 have B angles of ~100 ° while the others have B angles
of ~90 °. It is not due to any signiﬁcant structural distortions but arises because one of
the cell axes is chosen differently in such a way that the resulting cell volume is \/2 times
larger than those of the compounds with the B of ~90°. Therefore, strictly speaking these
compounds are not really isostructural with one another. Only dufrenoysite and rathite-

Ia are isostructural and so are Ba3Sb4,66S]0 and rahite-I with the a- and c-axis switched.

258

Another main difference among these compounds lies in the di-valent (Pb2+) to tri-
valent (A5”) ion ratio. As mentioned previously Pb4As4S10 is the only compound that
satisﬁes the charge neutrality with a divalent to trivalent metal ratio equal to 1.

Therefore, other compounds, which possess the ratio smaller than 1, are suspected to
have partial occupancy for certain As sites as the Sb(5/5’) site of Ba3Sb4,6(,Slo. It is
mostly the (6+1) sites that have a disorder between As and Pb while the 9-coordinate
sites are fully occupied with Pb atoms and 3-coordinate site are ﬁlled with As ions. It
seems that this structure type can tolerate considerable compositional variety without
disrupting the basic framework. This is possible by allowing mixed occupancy between
tri-valent and di-valent ions or partial occupancy of the tri-valent ion in the highly ﬂexible

(6+1) sites.

Table 7.8. Crystallographic Data for Ba3Sb4.66S10, Ba2.62Pb1.33Sb4S10 and the Related

Sulfosalt Minerals.

 

 

 

Formula Mineral a (A) b (A) c (A) 5 (°) S.G.
Pb4As4Slo dufrenoysite 7.90 25.74 8.37 90.21 P21
Pb35As45S10 rathite-Ia 7.91 25.80 8.43 90 P21
(Pb,Tl)3As4(As,Ag)Sm rathite-I 25.16 7.94 8.47 100.28 P21/a
Pb3A85810 rathite-III 24.52 7.91 8.43 90 P21
Ba3Sb4_66810 8.955 8.225 26.756 100.29 P21/c
Ba2,62Pb1.33$b4S10 8.8402 8.2038 26.7623 99.488 P21

 

259

Spectroscopic Characterization and Thermal Analysis.

The materials described here are valence-precise and are expected to be
semiconductors. The optical absorption spectrum of Sr6Sb6Sn is shown in Figure 7.6.
The compound shows an abrupt optical gap at 2.12 eV. Ba3Sb4,66S10 and
Ba2_62Pb1.3ng4S10 also exhibit clean optical band gaps at 2.14 eV and 1.64 eV,
respectively, showing that introducing heavy Pb atom to the structure decreases the band
gap by 0.5 eV, see Figures 7.7. The polar structure of Ba2_62Pb1_38Sb4S10 may make it an
interesting material for NLO Optical investigations. The spectrum of Ba2,62Pb1.33Sb4Sw
below the band gap transition shows very little absorption which bodes well for optical
NLO experiments.

The Raman spectrum of Sr6Sb6S17 is shown in Figure 7.8. The shift at 431 and
466 cm'1 is assigned to the SS stretching vibration of the tri-sulﬁde groups and these
values are in accord with the S-S stretching frequencies reported for other compounds
such as 453 cm"1 for B-Na282,22 473 cm'1 KzLaZSb289,3 and 479 cm'1 for RbUszsg.4 The
shifts at lower energies are due to the various Sb-S vibration modes. The Raman spectra
of Ba3Sb4.66Slo and Ba2,62Pb1_338b4Sm are also shown in Figure 7.9. They share similar
features with that of Sr6Sb6S17, only this time the S-S stretching modes are absent. The
highest energy shifts appear at 360 cm'1 for Ba3Sb4I66Sm and 348 cm'1 for
Bag-62Pb133Sb4Sm, both of which are generated by Sb-S vibrations. The reason for the
lower energy shift at 348 cm'I as well as the other peaks is attributed to the participation

of the Pb atoms bonded to the thio-antimonate framework.

260

 

 

 

 

2'01
a; 1.5;-
D .
E 10'-
b h
.5.
g 0.51
m I
B . Eg = 2.12 eV

0 - .
1.6 1.8 2.0 2.2 2.4 2.6
Energy (eV)

Figure 7.6. Optical absorption spectrum of Sr68b6S17. The band gap value, Eg, is shown

in the ﬁgure.

 

Ot/S (Arbitrary Units)

 

 

 

c...
"" .1..21.

L

 

1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
Energy (eV)

Figure 7.7. Optical absorption spectra of Ba3Sb4.66S10(—) and Ba2.62Pb1.3ng4S10(---).

The band gap value, E3, is shown in the ﬁgure.

261

 

   

 

 

 

 

 

N
\O
N
b
E _, S-S stretching
O ('0
a 9 /
E as.
e Q
<
\O
\0
v
100 ‘ 200‘ ‘300‘ ‘ 400‘ ‘ 500‘ T600 ”700 ‘ 800 ‘ 900
Raman Shift (cm-1)

Figure 7.8. Raman spectrum of Sresbasn.

 

 

 

 

 

 

 

b
'53

s:

a

t:

U-I-t

.2

'3 ,

166 666” 366‘ 466‘ ‘ 566‘ 606‘ "766‘ 800 5‘ 900

Raman Shift (cm-1)

Figure 7.9. Raman spectra of Ba3Sb4,66Slo(—) and Ba2.62Pb1.38$b4Sm(---).

262

Differential thermal analysis (DTA) showed that these compounds melt
congruently at 729°C, 770 °C, and 749 °C for Sr6Sb6$l7, Ba3Sb4.6éSlo, and
Ba2.62Pbl.338b4Slo, respectively. For each compound, multiple heating/cooling cycles were
monitored as shown in Figure 7.10 for Ba2.62Pb1_338b4S10, which exhibits well deﬁned
melting and crystallization points. The powder XRD patterns, which were taken before
and after DTA measurement, are identical indicating that these materials do melt

congruently.

 

1 00
Crystallization

_ 677 °C
50- \
Melting Point

 

 

   

 

 

l 4

0 ‘ 2010‘ i ‘400‘ ‘ ‘600 ' A8005 ‘ 1000

 

Temperature ( 0C)

Figure 7.10. DTA diagram of Baz,62Pb1.388b4Sm. (—) First cycle. (---) Second cycle.

263

10.

11.

12.

13.

14.

References

Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G., Inorg. Chem.
1997, 36, 3804-3805.

Chung, D.-Y.; Choi, K.-S.; Iordanidis, L.; Shindler, J. L.; Brazis, P. W.;
Kannewurf, C. R. ; Chen, B.; Hu, 8.; Uher, C.; Kanatzidis, M. G. Chem. Mater.
1997, 9, 3060-3071.

Choi, K.-S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309-
319.

Choi, K.-S.; Kanatzidis, M. G. Chem. Mater. 1999, 11, 2013-2018.

Cordier, G; Cook, R.; Schéifer, H. Angew. Chem. Int. Ed. Engl. 1980, 92, 310.
Cordier, G; Schéifer, H. Z. Naturforsch. 1979, 34b, 1053-1056.

Cordier, G; Schwidetzky, C.; Schafer, H. Rev. Chim. Mine'ral 1982, 19, 179-186.
Cook, R.; Schafer, H. Studies Inorg. Chem. 1983, 3, 757-760.

Marumo, F.; Nowachi, W. Z. Kristallogr. 1965, 122, 433-456.

Le Bihan, M.-Th. Bull. Soc. Franc. Minér. Crist. 1962, LXXXV, 15-47.
Marumo, F.; Nowachi, W. Z. Kristallogr. 1967, 124, 409-419.

Ribar, B.; Nicca, Ch.; Nowacki, W. Z Kristallogr. 1969, 130, 15-40.

Smith, P. P. K.; Hyde, B. G. Acta Cryst. 1983, C39, 1498-1502.

McCarthy, T. J .; Ngeyi, S.-P.; Liao, J. —H.; Degroot, D.; Hogan, T.; Kannewurf,

C. R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331.

264

15.

16.

17.

18.

19.

20.

21.

22.

SMART: 1994, Siemens Analytical Xray Systems, Inc., Madison, Winsconsin
53719 USA.

SAINT: Version 4, 1994-1996, Siemens Analytical Xray Systems, Inc., Madison,
Winsconsin 53719 USA.

Sheldrick, G. M. University of Gottingen, Germany, to be published.
SHELXTL: Version 5, Sheldrick, G. M. Siemens Analytical Xray Systems, Inc.,
Madison, WI 53719, USA, 1994.

(a) Yamaoka, S.; Lemley, J. T.; Jenks, J. M.; Steinﬁnk, H. Inorg. Chem. 1975, 14,
129-131. (b) Miller, W. S.; King, A. T. Z Kristallogr. Kristallgeom. Kristallphys.
Kristallchem. 1977, 144, 439-446.

Bdttcher, P. Z. anorg. allg. Chem. 1980, 461, 13-21.

Bottcher, P. Z. anorg. allg. Chem. 1977, 432, 167-172.

Bottcher, P.; Getzschmann, J.; Keller, R. Z. anorg. allg. Chem. 1993, 619, 476-

488.

265

Chapter 8. Charge Density Wave Caused by Reducing ThSe3 by One
Electron. Superstructure and Short Range Order in AThZSe6 (A = K,

Rb) Studied by X-ray Diffraction, Electron Diffraction and Diffuse

Scattering.

266

1. Introduction

The reactions of lanthanides in AZQx (A=alkali metal, Q=S, Se, Te) ﬂuxes have
led to new phases which repeat several structural motifs seen in known binary and
ternary Chalcogenides.l Recently, we have extended the molten salt synthetic method to
thio-2 and seleno-antimonate ﬂuxes which are little investigated.3‘4 We were attracted to
the possibility of combining lanthanides and actinides with chalcoantimonate fragments
because of the potentially interesting structural distortions which could develop around
the Sb3+ centers (due to the lone pair of electrons) and the rather large f-elements.1’5’6 In
most cases, depending on the ﬂux basicity, quaternary phases were obtained at
temperatures < 650°C, while binary or ternary phases were found to be
thermodynamically stable at higher temperatures. For example, the KTth28663 forms
under such conditions, but when we attempted to prepare A/Th/Sb/ Se quaternary phases
at temperatures > 700°C, phase separation occured and instead the ternary compounds,
AThZSe6 (A=K, Rb) were isolated. These compounds have a two-dimensional structure
with layers that are equivalent to those seen in ZrSe37 with alkali metal ions between the
layers. Since ThSe; also adopts the ZrSe3 structure types, we recognized that AThZSeé
represent a well crystallized version of intercalated ThSe3 with 0.5 equivalents of an alkali
metal. This is signiﬁcant because, normally, intercalated samples have much poorer
crystallinity than the pristine parent ones. The lack of ability to perform single crystal
structure investigations in these samples leaves many of the ﬁne structural details

unresolved. Often these details are important in understanding the properties of

267

intercalated layered materials such as the nature of intralayer guest sites for alkali ions, the
effect of charge density waves on electrical conductivity, ion mobility, color, etc.

Two kinds of structural modiﬁcations of the ZrSe3-type structure are expected for
the AThZSe6 compounds. The ﬁrst structural modiﬁcation is that of the layer itself.
This layer, in which the formal charges balance as Th°+Se2'(Se22'), undergoes reduction by
the alkali metal. Either the diselenide group, Sezz', or the Th4+ center could act as a
possible reduction sites. We wanted to investigate where and how the layers could
accommodate the extra electrons from the alkali metal and how the presence of these
electrons inﬂuences the ZrSe3-type structure. In fact, this system has a lot of similarity
to the Li intercalated ZrSe3 which has been extensively investigated for possible battery
applications. With LierSe3 (0<xS3),9 there is still uncertainty about the intercalation
mechanism, the local environment of the Li ion, and the structural change of the host
material during the intercalation process. In this respect, our study on AxThSe3 (x=0.5;
A=K, Rb) could shed new light on the intercalation process of Li in ZrSe3-type
compounds by providing structural insight as to how the host structure responds when
electrons are inserted into it. The second type of structural modiﬁcation is attributed to
the stacking arrangement of the layers. In ZrSe3, the stacking arrangement of the layers
produces trigonal prismatic sites between the layers. These sites, however, are not large
enough to accommodate the I(+/Rb+ ions. Therefore, the layers must stack in such a way
that they can provide enough space for the newly introduced alkali metals.

One of these compounds, KThZSe6, was reported10 while this manuscript was in
preparation. Only the space group of this compound was mentioned, without complete

characterization of the structure or physical properties, and it was concluded that

268

KThZSe6 is isostructural with KThzTeélO and CsThzTe6ll 11based on the unit cell
parameters and the systematic absences determined by oscillation and Weissenberg
photographs. However, according to our results, KThZSeG is crystallographically
different from that of KThzTe6. The known difference between selenium and tellurium to
engage in chalcogen-chalcogen interactions does not allow ATthe6 (A=K, Rb) to have the
exact same stacking arrangement of the anionic layers and cations as KThzTeG, even
though their compositions and basic frameworks [Th2Q6] (Q = Se, Te) are the same. In
fact, it is instructive to compare these two structures and understand why the selenide
adopts a slightly different structure.

Here we report a complete structural study of AThZSe6 (A=K, Rb) and discuss it
with respect to the structures of ZrSe3 and KThzTe6. In addition to a single crystal x-ray
crystallographic analysis which gives the average structure, electron diffraction studies on
AThZSe6 have revealed a subtle, but signiﬁcant, structural modiﬁcation which originates
from an electronic instability caused by the extra electrons added on the ThSe3
framework. With the aid of Raman and optical spectroscopy as well as magnetic
susceptibility measurements, we suggest that this structure distortion is due to the
reductive cleavage of one quarter of the total diselenide groups in the lattice and
represents a charge density wave (CDW) on the ideal structure of ThSe3.12 In addition,
we also carried out an atomic pair distribution function (PDF) analysis of the X-ray
diffuse scattering generated by KThZSe6 in order to probe the local structure around the
diselenide group. This veriﬁed the structure distortion and the presence of short Se-Se

bonds, which are “invisible” by the single crystal x-ray crystallographic analysis.

269

2. Experimental Section

2.1 Synthesis.

The following reagents were used as obtained: Thorium, 100 mesh, Cerac,
Milwaukee, WI; selenium powder, 100 mesh, Aldrich, Milwaukee, WI; potassium metal,
analytical reagent, Mallinckrodt Inc., Paris, KY; rubidium metal, analytical reagent,
Johnson Matthey, Ward Hill, MA. The KZSe/RbZSe starting materials were prepared by

a stoichiometric reaction of potassium/rubidium metal and selenium in liq. NH3.

ATbZSeG (A =K, Rb) These compounds were synthesized from a mixture of 0.1256 g
(0.80 mmol) of KZSe (or 0.200 g (0.80 mrnol) of RbZSe for RbThZSeé), 0.1856 g (0.80
mmol) of Th, 0.3158 g (0.40 mmol) of Sb and 0.4158 g (4.0 mmol) of Se. This reagents

were thoroughly mixed, sealed in a carbon coated fused silica tube (9 mm diameter) and

heated at 720 °C 5 days (cooling 2 oC/h). Pure golden-black planks of A ThZSe6 were
obtained by isolation in degassed dimethylformamide (DMF) (>90% yield based on Th).
The compositions of the materials were analyzed by Scanning Electron Microscope
(SEM)/ Energy Dispersive Spectroscopy (EDS). No Sb impurities were detected in the
compounds. Homogeneity was conﬁrmed by comparing the powder X-ray diffraction
pattern of the product against one calculated using X-ray single crystal data. It is
interesting to note that even though the Sb was not incorporated into the compound, it is
needed to grow large crystals of this material in high yield by forming a suitable AbeSey

(A = K, Rb) ﬂux. Other reaction mixtures without Sb did not give as pure products or as

270

large crystals as the ones obtained from the AbeSey ﬂuxes. The reactions at lower

temperature ( < 700°C) gave a lower yield of the compounds.

2.2 Physical Measurements.

Semiquantitative microprobe analyses. The analyses were performed using a JEOL
J SM-35C Scanning Electron Microscope (SEM) equipped with a Tracor Northern Energy
Dispersive Spectroscopy (EDS) detector. Data were acquired on several crystals using an

accelerating voltage of 20 kV and 40 sec. accumulation time.

Solid-State UV/V is Spectroscopy. Optical diffuse reﬂectance measurements were
performed at room temperature with a Shimadzu UV-3101 PC double-beam, double-
monochromator spectrophotometer operating in the 200-2500 nm region. The instrument
is equipped with an integrating sphere and controlled by a personal computer. BaSO4
was used as a 100% reﬂectance standard for all materials. Samples were prepared by
grinding them to a ﬁne powder, spreading them on a compacted surface of the powdered
standard material and preloaded them into a sample holder. The reﬂectance versus
wavelength data generated were used to estimate the material's band gap by converting the

reﬂectance data to absorption data as described elsewhere. '3

Raman Spectroscopy. Raman Spectra of the crystal specimens were recorded on a

Holoprobe Raman spectrograph equipped with a 633 nm HeNe laser and a CCD camera

271

detector. The instrument was coupled to an Olympus BX60 microscope. The spot size

of the laser beam was 10 microns when a 50x objective lens was used.

Magnetic Susceptibility. The magnetic response of the compound was measured
over the range 2-3 00 K using a MPMS Quantum Design SQUID magnetometer. Samples
were ground to a ﬁne powder to minimize possible anisotropic effects and loaded into
PVC containers. The temperature dependent susceptibility measurements were
performed at 1000 Gauss. Corrections for the diamagnetism of the sample containers
were made by measuring the magnetic response of the empty container at the same
magnetic ﬁeld that was used for the ﬁlled container. The core diamagnetic contribution of

every ion to 70,1 was subtracted according to Selwood.l4

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically, a sample (~20 mg )
of ground crystalline material was sealed in a quartz ampoule under vacuum. A quartz

ampoule of equal mass ﬁlled with A1203 was sealed and placed on the reference side of
the detector. The samples were heated to the desired temperature at 5 °C/min,

isothermed for 10 min. followed by cooling at -5 oC/min to 50 °C. The stability and
reproducibility of the samples were monitored by running multiple heating/cooling cycles.

Residues of the DTA experiments were examined by X-ray powder diffraction.

2.3 X-ray Crystallography.

272

Single crystals of KTh2$e6 with dimensions 0.41 x 0.04 x 0.06 mm and of
RbTthe6 with dimensions of 0.29 x 0.13 x 0.03 mm were mounted on the tip of a glass
ﬁber. Intensity data for KThZSe6 were collected at room temperature on a Rigaku
AFC6S four circle automated diffractometer equipped with a graphite-crystal

monochromator. The unit cell parameters were determined from a least-squares

reﬁnement using the setting angles of 20 carefully centered reﬂections in the 8032033 0°
range. The data were collected with the (1) scan technique over a full sphere of reciprocal
space, up to 50 deg in 20. Crystal stability was monitored with three standard reﬂections
whose intensities were checked every 150 reﬂections. No signiﬁcant decay was observed
during the data collection period. An empirical absorption correction was applied based
on 111 scans. Since the cell parameters given by the four-circle automated diffractometer
had high standard deviations, Guinier x-ray powder diffraction patterns (calibrated with a
Si standard) were used to obtain more precise cell parameters for KThZSe6. LATCON in

Xtal32 software15 was used to reﬁne the cell parameters with 38 reﬂections selected in

the 130520544o range and the resultant values were used for the calculation of bonds and
angles.

The crystallographic data for RbThZSeG were collected at 173.1K on a Siemens
SMART Platform CCD diffractometer using graphite monochromatized Mo K01
radiation. The data were collected over a full sphere of reciprocal space, up to 56 deg in
20. The individual frames were measured with an omega rotation of 0.3 deg and an
acquisition time of 25 sec. The SMART16 software was used for the data acquisition and

SAINTl7 for the data extraction and reduction. The absorption correction was performed

273

using SADABS“. The complete data collection parameters and details of the structure
solution and reﬁnement for both compounds are given in Table 8.1.

Structure solutions and reﬁnements were performed, for both compounds, using
the SHELXTL19 package of crystallographic programs. Systematic absence conditions of

the data sets gave three possible acentn'c space groups, I222 (# 23), I212121 (# 24), Imm2

(# 34) and one centric space group, Immm (No. 71). The MISSYM algorithm20 as

1 revealed a center of inversion in the structure,

implemented in the PLATON program2
suggesting that the correct space group is Immm. The space group reported for K ThZSe6
earlier was Cmcm10 because it was assumed that KThZSe6 was isostructural with
KThzTe6, which crystallizes in this space group. The only evidence that led these
authors to this conclusion was two systematic absence conditions hkl, h+k=2n+l; h01,
l=2n+1 obtained from oscillation and Weissenberg photographs. These two conditions,
however, are not sufﬁcient enough evidences to conclude the correct space group because
they are also sub-conditions of I-centering; hkl, h+k+l = 2n+1. As a matter of fact, we
found that 912 reﬂections out of the 1812 total reﬂections collected for KThZSeé violate
the systematic absence conditions of Cmcm such as hkl, h+k = 2n+1; 0k], k = 2n+l ; hOI,
h, l = 2n+l, etc. Nevertheless, we still tried to reﬁne the structure using the atomic
coordinates reported for the Cmcm space group. This gave a signiﬁcantly higher values of
R1 = 0.0876 and wR2 = 0.1860 for all data. In addition, the isotropic temperature factor
for the K+ ions, which are equally disordered over two crystallographically equivalent
sites in the Cmcm space group, was very large (U eq = 0.398). The structure was solved

and reﬁned successfully using the 1mm space group with R indices shown in Table 8.1.

The coordinates of all atoms, isotropic displacement parameters, and their estimated

274

standard deviations are given in Table 8.2. The anisotropic displacement parameters for

both compounds are listed in Table 8.3. The selected bond distances and angles are listed

in Table 8 .4.

2.4 Electron Diffraction Study (T EM).

Electron crystallographic studies were carried out on a JEOL lOOCX
Transmission Electron Microscope (TEM) using an electron beam generated by a C636
ﬁlament and an acceleration voltage of 120 kV. After the samples were ground to a ﬁne
powder in acetone, the specimens were prepared by dipping a carbon coated copper grid

in the suspension. The samples showed no decomposition under the e-beam .

2.5 Pair Distribution Function Analysis (PDF).

The procedures used to perform atomic pair distribution function analysis of x-
ray diffraction data have been published elsewhere.”23 Powder diffraction data were
collected out to 20mm of 140 deg on an in-house powder diffractometer in Bragg-Brentano

geometry using Mo K01 radiation.

2.6 Atomic force microscopy (AF M).

Contact AFM images were acquired using an Park Scientiﬁc Instruments
Autoprobe CP scanning probe microscope, with a ﬁve um high resolution scanner.
Microlever cantilevers (Park Scientiﬁc Instruments) having a spring constant of
approximately 0.5 N/m and an integrated silicon nitride probe with a radius of curvature

of approximately 50 nm, according to the manufacturer, were used. The experiment was

275

Table 8.1. Summary of Crystallographic Data and Structural Analysis for KTh2Se6 and

 

RbThZSe6.

Formula KThZSe6 RbTthe6

Formula weight 976.94 1023.31

Crystal habit golden-black needle golden-black plate

Space group Immm Im

a, A 4.1899(5) 4.2031(1)

b, A 5.6337(5) 5.6347(1)

c, A 21.860(4) 22.4714(1)

2; v, A3 2 ; 516.01(13) 2 ; 532.19(2)

Dcaic, glcm3 6.288 6.386

Temp, K 298 173

MW K90, A 0.71073 0.71073

MMO K00 cm" 502.74 528.98

F(000) 806 842

20 max , deg 50,0 56.5

Total data 1 812 2822

Unique data 296 [Rim = 0.0399] 420 [Rim = 0.0960]

No. of variables 19 19

Reﬁnement method Full matrix least-squares on F 2

Final R indices [I>29] a111:0.0212 R1=0.0390
bWR2=0.0528 wR2=0.0875

R indices (all data) R1=0.0239 R1=0.0405
wR2=0.0564 wR2=0.0880

goodness of ﬁt on F2 1.120 1.112

 

“R1=leFol-IFcll/2lFol

b wR2= {31./(17.2 —F.2 )2 ]/ 2[w(1~:.2)2 1}"2

276

Table 8.2. Fractional Atomic Coordinates and Equivalent Isotropic Displacement

Parameters (Ueq) Values for KTh28e6 and RbTthe6 with Estimated Standard Deviations

in Parentheses.

 

 

 

 

 

Atom x y z Ueq,a A2
KTthe6

Th 1/2 0 0.3225(1) 0.010(1)

Se(1) 0 0.2420(2) 0.3976(1) 0.020(1)

Se(2) O O 0.2267(l) 0.010(1)

K 0 0 0 0.030(1)
RbThZSe6

Th 1/2 0 0.3204(1) 0.006(1)

Se(1) 0 0.2421(2) 0.3936(1) 0.015(1)

Se(2) 0 0 0.2273(1) 0.005(1)

Rb 0 0 0 0.016(1)

 

aUeq is deﬁned as one third of the trace of the orthogonalized Ur) tensor.

277

Table 8.3. Anisotropic Displacement Parameters (A2) for KThZSeé and RbTh2Se6 with
Estimated Standard Deviations in Parentheses.

 

 

 

 

 

 

KTh28e6
Th Se(1) Se(2) K
U11 0.010(1) 0.012(1) 0.012(1) 0.024(3)
U22 0.006(1) 0.032(1) 0.005(1) 0.042(3)
U33 0.015(1) 0.016(1) 0.012(1) 0.022(3)
Ulz 0.0000 -0.008(1) 0.0000 0.0000
U13 0.0000 0.0000 0.0000 0.0000
U23 0.0000 0.0000 0.0000 0.0000
RbTh28e6
Th Se(1) Se(2) Rb
U11 0.005(1) 0.006(1) 0.006(1) 0.012(1)
U22 0.003(1) 0.028(1) 0.002(1) 0.026( 1)
U33 0.010(1) 0.011(1) 0.006(1) 0.010(1)
U12 0.0000 -0.007(l) 0.0000 0.0000
U13 0.0000 0.0000 - 0.0000 0.0000
U23 0.0000 0.0000 0.0000 0.0000

 

278

Table 8.4. Selected Distances(A) and Bond angles(°) for AThZSeé (A = K, Rb) with
Standard Deviations in Parentheses.

 

 

 

KTthe6 RbThzs€6
Bond Distances
Th-Se(l) 2.9913(9) x 4 2.9970(9) x 4
Th-Se(2) 2.9615(11) x 2 2.9658(10) x 2
Th-Se(2) 3.0153(6) x 2 3.0147(5) x 2
A-Se(1) 3.3922(9) x 8 3.4989(9) x 8
Se(1)-Se(1) 2.727(2) 2.728(3)
Bond Angles
Se(l)-Th-Se(1) 8891(3) x 2 8905(3) x 2
Se(1)-Th-Se(1) 5423(4) x 2 54.14(5) x 2
Se(1)—Th-Se(l) 113.36(4) x 2 113.43(4) x 2
Se(2)-Th-Se(2) 7539(2) x 4 75.46(2) x 4
Se(2)-Th-Se(2) 9004(4) 8905(3)
Se(2)-Th-Se(2) 13 820(6) 138.30(5)
Se(l)-Th-Se(2) 76.71(3) x 4 76.70(3) x 4
Se(1)-Th-Se(2) 128.45(2) x 4 128.36(2) x 4
Se(1)-Th-Se(2) 152.09(2) x 4 152.14(3) x 4
Se(1)-Th-Se(2) 8385(3) x 4 83.71(2) x 4

 

279

conducted in air using either pristine (as made) or a freshly cleaved surfaces of single
crystals of RbTthe6.

The AF M data were recorded in constant height mode under controlled ambient
conditions (35% humidity and 21°C). In this mode of imaging the deﬂection of the
cantilever is measured as the sample is raster scanned beneath the tip. The force used to
acquire these images was 10 nN. The samples were scanned at a rate of 32 lines per
second. Each image was 100-150 A and 512 x 512 pixels. Distances were calibrated using

the 5.1 A lattice spacing obtained from freshly cleaved muscovite mica.24

3. Results and Discussion
Structure.

ATh28e6 (A = K, Rb) has a two-dimensional structure composed of inﬁnite
[Tthe6]' layers that lie perpendicular to the [001] direction, see Figure 8.1. The Th
atoms in this layer are coordinated by 8 Se atoms in a distorted bicapped trigonal
prismatic geometry. Each trigonal prism has two edges much shorter than the rest. These
short edges are formed by the diselenide groups, Sezz'. The Th centered trigonal prisms
make one dimensional chains parallel to the [100] direction by sharing triangular faces
with each other. The layers are formed when each chain shares monoselenide ions, Se2', at
the apex and capping sites with neighboring chains.

Each [ThZSeJ layer itself can be described as ZrSe37 type except for the stacking
arrangement of the layers. In ZrSe3, all Se atoms of the diselenide groups in one layer are
staggered with the corresponding Se atoms in the next layer to minimize the steric

repulsion, see Figure 8.2(a). The only reasonable place for the alkali metal to reside

280

 

 

 

 

 

<26
@ K@ @ @

$91
6— f-

\

 

   
   
   
 

   
 
 
 

  

 

e?\—o‘

 

 

 

 

 

$92 Th
ﬂ‘ SW:
\
.g—‘\_/. .0 ‘\\cl In!" ‘3

  

88 a
L...

Figure 8.1. The overall Structure of KTthe6 viewed down the c-axis. The labeling

 

scheme for the RbTthe6 is analogous.

281

(a) ZrSe3

(b) KTthes

 

2.727(2)A 2.907(2)A

(c) KThgTee

  

 

 

3.057(3)A 3.085(3)A

Figure 8.2. Schematic comparison of the structures of (a) ZrSe3, (b) KTthe6 and (c)

KThzTe6. In KThZTe6, each K+ site is half occupied.10 The distribution of chalcogen-

chalcogen distances are indicated.

282

between the layers are trigonal prismatic sites, the size of which is not big enough to
accommodate the K+/Rb+ ions. Therefore, in AThZSeé (A = K, Rb), all the [TtheJ
layers are stacked in an eclipsed fashion by shifting every other layer 1/4 along the b-axis
to generate larger square prismatic sites. Since the crystallographically observed Se-Se
distance in the diselenide bond is 2.727(2) A for KTh2Se6 (2.728(3) A for RbThZSe6) and
the shortest distance between the diselenide groups is 2.907(2) A, (2.907(3) for
RbThZSe6), the resultant stacking arrangement generates two square prismatic sites with
different sizes, see Figure 8.2(b). The K‘A/Rb+ ions are stabilized in the larger square
prismatic site and are coordinated by 8 Se atoms, with K---Se distances of 3.3922(9) A
for KThZSe6 and 3.4989(9) A for RbThZSe6.

In KThzTe6, the Te-Te distance within the ditelluride fragment (3.057(3) A) is
very similar to the Te-Te distance between ditelluride fragments (3.085(3) A).
Considering that tellurium is more basic and softer than selenium, the electrons on the Te
atoms can easily delocalize to neighboring Te atoms, giving rise to, essentially, inﬁnite
chains along the [010] direction. In this case, the eclipsed stacking of the ditelluride
fragment of the [ThzTe6]' layers is no longer preferred to make suitable pockets for the
alkali metal (Figure 8.2(c)). Therefore, KThzTe6 adopts a slightly different stacking
arrangement of layers by staggering the ditelluride units of one layer to those of the next
layer. All square prismatic sites between the layers are still present but are of equal size
which is suitable to accommodate the K+ ions, thus resulting in statistical disorder of the
cations over all such sites.

The assignment of formal oxidation states in these phases appears to be Th4+Se2'

(Sezz's'), judging from the results of the single crystal X-ray diffraction analysis. We ﬁnd

283

it rather unusual that the observed Se-Se distance in the diselenide bond of AThZSe6 is

signiﬁcantly longer at 2.727 A than the nonna Se-Se bond distance of ~2.35 A, suggesting

that the oxidation state of Se(1) in the diselenide group of [Th28e6]‘ is between -1 and -2,
namely, -125. In other words, it would seem that the 0.5 equivalent of electrons from
the alkali metal is located on all the Se(1) atoms in the diselenide groups, and this
elongates the Se-Se bond by 0.4 A. This corresponds to 0.5 electron per Se-Se bond
which would give essentially partially-broken bonds. This situation could cause

paramagnetism in the compounds.

Properties.

The optical properties of the compounds were determined by measuring the solid
state UV/V is diffuse reﬂectance spectra, which show the presence of two abrupt optical
transitions at 0.90 eV and 2.16 eV for KThZSeé and 0.90 eV and 2.09 eV for RbThZSeé,
suggesting that the materials are semiconductors and the extra electrons introduced into
the ZrSe3-type layer are localized, see Figure 8.3. The origin of the two transitions is not
clear, but they are also present (at slightly different energies) in ThSe3 itself.25 One
possibility is that they are dichroic due to the anisotropic nature of the electronic band
structure given the lamellar nature of these materials. In fact visual inspection of the
crystals does suggest the presense of dichroism.

DTA measurements show that KThZSe6 and RbTh2$e6 do not melt below 1000°C.

The Raman spectra display shifts at ~124 cm", ~160 cm", ~199 cm‘l, and ~234
cm'1 for both compounds (Figure 8.4). Since the Se-Se stretching should appear at a

higher energy than any Th-Se stretching, the absorption at ~234 cm'1 for ATthe6 can be

284

 

 

 

 

 

 

5.0:—
:6: E
a 4.0~
: i
i
E‘ 3.6:
E! E
.e .
< 2.0;"
m E
B 10:
0.0; JLL1_L1J_llll|llllIJILllllll
0.5 1.0 1.5 2.0 2.5 3.0 3.5

Energy (eV)

Figure 8.3. Electronic absorption spectra of KTh2S66(---) and RbThZSe6(—). The band

gap energies, Eg are indicated .

 

Arbitrary Intensity

 

 

 

IljlllllljllllilllJllllll

 

100 200 300 400 500 600
Raman Shift (cm-1)

Figure 8.4. Raman spectra of KThZSe6(---) and RbTh2Se6(—).

285

tentatively assigned to the diselenide group. The wavenumber at which the stretching
vibrations of a normal Se-Se bond occur varies depending upon both the local geometry of
the selenium atom and the metal with which it makes bonds. For example, that of USe326
appears at ~285 cm",27’28 that of Kzsez at ~253 cm'l, 29 and that of crystalline Se occurs
near 238 cm'l.28 This frequency variation occurs despite the fact that the Se-Se distance
in all three compounds ranges narrowly between 2.33-2.40 A. Based on the
crystallographically observed Se-Se distance of 2.73A, which is lengthened by 0.35 A, we
expect the Se-Se stretching vibration to appear at much lower energy. The Raman shift at
~234 cm'1 for AThZSe6, therefore, is inconsistent with the presence of a 2.73A Se-Se
bond, and more indicative of normal diselenide groups (i.e. with a bond distance close to
2.35 A).

The magnetic susceptibility of RbThZSe6 was measured from 2 to 300K at 1000G,
and the material appears to be diamagnetic. This is also discrepant with the model from
the x-ray diffraction study, which suggests paramagnetic behavior if an unpaired electron
is assumed on every Sezz' group.

These apparent inconsistencies between the spectroscopic data and the structural
model suggest that the local structure associated with the Se-Se unit is more complicated
than depicted from the x-ray crystallographic results. Therefore, we proceeded to
examine the structure of ATthe6 in greater detail using other methods such as electron

diffraction, atomic force microscopy and X-ray diffuse scattering analysis.

Electron Diffraction Study and Superstructure Model.

286

In spite of the good structural reﬁnement of the single crystal x-ray diffraction
data, we noticed that one of the anisotropic temperature factors, U22 for Se(1) belonging
to the diselenide group, was relatively high compared with that of Se(2) and Th atoms for
both the K- and Rb- compounds (See Table 8.3). This factor is three times larger than
U11 and U33 for Se(1). Notice that the K/Rb atoms also have relatively high U22 values,
which is not surprising because the alkali atoms are surrounded by Se(1) atoms and their
temperature displacement factors are directly affected by the behavior of Se(1). This fact
gave us a clue for the potential presence of a superstructure, along the b-axis, due to
positional long-range ordering of Se(1) atoms in the diselenide groups brought about by
electronic coupling and resulting in a charge density wave. The fractional oxidation states
for Se(1) derived from the structure determination above, in fact, makes this system a
potential candidate for CDW as has been observed in other systems where similar
structure/oxidation state relationships exist, e.g. LnSe3, LnTe3 (Ln=rare earth),
K0.3333510667138182-3O

The possibility for superstructure was examined by transmission electron
microscopy (TEM). Electron beams interact more strongly with matter than x-rays, and
thus electron diffraction is more sensitive to the presence of a weak structure modulation.
Indeed, electron diffraction studies revealed a weak superstructure along both the a- and
the b-axis. Both compounds, KThZSeG and RbThZSe6, showed very similar electron
diffraction patterns. Figure 8.5 shows the electron diffraction pattern for RbThzse6 with
the beam perpendicular to the ab-plane.31 The reﬂections associated with the new
superstructure present in AThZSe6 are very weak and occur along both the a°- and b'-

direction with a modulation vector qsuper = 0.25a°,ub + 0.25b‘sub essentially resulting in a

287

 

diffuse scattering

Figure 8.5. Selected area electron diffraction pattern with the beam parallel to the
[001] zone axis from RbThZSe6 showing weak 4 x 4 superlattice. The (hkO) family
of reﬂections is shown. Superlattice peaks around the 310 sublattice reﬂection are
indicated by four small arrows. Also notice that the superlattice peaks show some

streaking along a‘-axis due to diffuse scattering.

288

4a x 4h supercell. Here, a'sub and b‘sub are the lengths of the reciprocal sublattice based on
the AThZSe6 structure. Four weak supercell peaks around the (310) subcell reﬂection are
indicated by small arrows in this pattern to make them more recognizable. Notice that
these supercell peaks are streaking along the a°- direction due to some diffuse scattering
while their shapes are sharp along the b'- direction. The streaking indicates that the
superstructure in real space along the a-direction is more poorly deﬁned than along the b-
direction.

The substructure, observed by the x-ray diffraction analysis above, represents the
average positions of the Se(1) atoms. The weak superstructure seems to be caused by
small displacements from the ideal local arrangement of Se(1) atoms causing the
anisotropic temperature factors (in the substructure) for Se(1) to be larger. Since only
U22 for Se(1) has an unusually high value, we expect that the positional parameters for

Se(1) will vary mostly along the b-axis.

diselenide groups 362‘
' I 1 r_'l—I
0'0 0'0 0'. o 0 0'.
"9 9 -B {9 g)
° .‘:'.‘jﬂ.‘:’.‘:ﬂ.‘: .
c 0 0‘0 0‘. o I o 0‘.

I ’1

no Se-Se bond

 

 

l‘
bsuper=4b

Scheme 8.1. Separation of Sezz' and Sez' groups along the b-axis and corresponding

superstructure.

289

Based on these results, we can consider a plausible superstructure model with a
quadrupled cell along the b-axis. If all the electrons from the alkali atoms are transferred
to the diselenide groups in AThZSeb, it would correspond to 1e per two 8622' groups, or
2e per four Sezz' groups. Of course, two electrons can cleave, reductively, one diselenide
bond into two single selenide ions, Sez'. Therefore, the remaining three diselenide groups
are expected to be normal Sezz' units instead of Se22'5', with normal bond distances and
Raman frequencies. This bond cleavage, which can happen only in one out of four SeZZ'
groups, probably causes the CDW and the quadrupling of the a- and b-crystallographic
axes, see tentative model in Scheme 8.1. This model requires that all electrons are paired
and thus agrees with both the diamagnetism and semiconducting nature of these
compounds. It is also consistent with the Raman spectroscopic data which support the
presence of single Se-Se bonds.

The quadrupling of the cell along the b-axis is due to the ordering of the three Se22'
and two Sez' species along this direction. The streaking of the spots along the at-direction
suggests ordering along the a-direction is not as well deﬁned as along the b-direction. This
can be easily understood if we consider that along the a-axis direction there is no
substantial interaction between the parallel Sezz' groups, since they are spaced 42A apart.
Therefore, any particular ordering of the Sezz' groups and pairs of Sez' atoms, along the a-
axis, will not have as strong of a driving force as it does along the b-axis. If it occurs at all,
it might be irregular and of short range. Scheme 8.2 shows a possible 4a x 4b

superstructure model with only the Se(1) atom network on the ab-plane.

290

Z

0—0 0—0
H 0—0 00
o—o

asuper

 

 

 

 

bsuper

[j Sublattice Superlattice

Scheme 8.2. A possible in-plane ordering of 8e22' and Se2' groups consistent with the

observed superstructure.

Observations with Atomic Force Microscopy.

In order to investigate whether the 4a x 4b superstructure could be observed
directly on the surface of a RbThZSe6 layer, we used atomic force microscopy (AF M).
The AF M image of the surface of RbTtheb, shown in Figure 8.6A, clearly resolves the
periodic rows of Se atoms in the crystals. The spacing between the rows of Se atoms
were measured to be 4.1 i 0.2 A which corresponds well to the lattice constant in the a-

direction of 4.203 A. While the rows of Se atoms are clearly visible, we were unable to

291

4 °‘V v '2" rh' .‘ M ' ’ . ‘ ' .51 ll f'. 'rl'z'” ‘ .
mag - .° ' A .. 7° ‘ gutwmtlhlzt 1

- - 45M ”:75”: r - H H , ‘1 - V .. auﬂ‘lfﬂv"
‘tﬁ‘f:,. y': .— ff.
' may:

WV

 

Figure 8.6. AFM image of the surface of a layer of RbTh2$e5 corresponding to the ab-
plane. (A) shows a raw image which has been ﬂattened with a second order polynomial
to account for non-linearity in the piezoelectric scanner. (B) shows a blown up region of
the image aﬁer a spatial Fourier ﬁltration to remove instrumental and environmental

noise from the data. The rows run parallel to the b-axis. The scale bar is 20 A.

292

resolve the 2.8 — 3 A spacing of atoms within the rows. Figure 8.6B is an enhanced region
of the original image, which shows a recognizable undulation along the rows of Se atoms
(white regions), but no well resolved atomic structure. It is reasonable to expect the Sezz’
groups to lie slightly higher on the surface than the regions of Se?” ions, because of the
more narrow Se—Th-Se angles associated with the diselenide groups. In this respect the
higher, white, regions in the photo of Figure 8.6B should be due to the Se22' groups. The
average spacing of these higher lying regions seems to be ~12-15 A not ~24A which
would be expected from the quadrupling of the lattice spacing. This discrepancy may be
attributed to a certain degree of surface air oxidation of Sez' ions to 8622' groups, which
will increase the density of the latter on the exposed surface. The lack of resolution along
the b-axis could be due to a number of factors including (a) poor deﬁnition of the
superstructure along this direction, which is consistent with the observed weak
superstructure in the electron diffraction patterns, (b) surface modiﬁcation upon exposure
to the ambient or (c) limitations associated with the AF M technique. While the AF M
results were consistent with the X-ray and electron diffraction data, they could not

unequivocally address the issue of Se-Se bonding along the superstructure direction.

Probing the Existence of Se-Se Single Bonds in Local Structure by Diffuse
Scattering and PDF Analysis.

In order to determine directly the presence or absence of diselenide bonds at the
normal bond distance of 2.34 A it is necessary to use a probe of the local structure. To
do this we carried out an atomic pair distribution function (PDF) analysis of x-ray

diffraction datazz’ ’32. In this technique, powder diffraction data are corrected for

293

experimental artifacts, normalized by the photon ﬂux and the number of scatterers, and
divided by the sample-average atomic form-factor to recover S(Q), the total scattering
function (i.e. diffraction pattern), see eq. (1). This is then Fourier transformed to obtain

the atomic pair distribution function (PDF), p(r), see eq. (2).

5(0) = N(f....(Q))2[l + J 4m2p(r)i%_QI.dr eq. (1)
_°° r

41tr2p(r) = 41tr2p0 + 2%? Q[S(Q)]sin erQ eq. (2)
0

Where: Q=47tsin0/A, S(Q)= normalized scattering function, f= average atomic scattering

fK(Q) + 2fn.(Q) + 6fs.(Q)
ZK + 2ZT,, + 6Z3,

 

factor, fm(Q) = , Z=atomic number, N= number of scattering

atoms, p(r)= atomic pair density function .

The PDF represents all the interatomic vectors in the structure. The approach is
the same as has been used to study glasses and amorphous materials.” 23 The PDF is a
measure of the microscopic density in the solid. Recently, we have been extending its
application to disordered crystalline materials”, where we can carry out a ﬁill proﬁle-
ﬁtting regression modeling analysis, similar to Rietveld reﬁnement, except that it yields
the local rather than the long-range ordered structure.33 In the present case, we are
interested simply in determining whether short Se-Se bonds exist in KThZSe6 and we have
not carried out a complete full proﬁle-modeling of the PDF data. This will be reported in

the future.

294

The corrected data in the form of I(Q) = Q(S(Q)-1) are shown in Figure 8.7. The
data contain Bragg peaks and diffuse scattering. Both of these contributions to the
scattering are Fourier transformed to obtain the PDF. This is important since the diffuse
scattering contains the short-range order information which is not considered in a
conventional Rietveld or single crystal crystallographic reﬁnement. 34

Figure 8.8 shows PDFs in the form of p(r) from the data (a) and calculated from
the crystal structural model (b). Overall, the PDF from the data and the model are
qualitatively similar, as expected; however, there are signiﬁcant differences as we describe.
Peaks occur at positions, r, in p(r) when two atoms are separated by this distance in the
solid. Thus, the ﬁrst peak to appear is the nearest-neighbor distance. In the
crystallographic model determined from the single crystal reﬁnement (see above), this is
at 2.73 A and is a Se-Se distance. This feature is not resolved from the strong Th-Se peak
centered around 2.98 A in the crystallographic model. This peak is the sharpest and
tallest feature in the model-PDF calculated from the crystallographically determined
structure. The peak maximum occurs at 0.125 A3 and is cut off in the ﬁgure. In the
model this peak is almost 2x higher than the next strongest peak at r=5 A. Also, in the
crystallographically determined model, there is no intensity in p(r) at the position of the
diselenide bond, 2.34 A. In contrast, the PDF derived from the observed data shows that
the peak at 2.98 A is signiﬁcantly broadened, to the extent that it appears weaker than the
second strong peak at r=5 A. Furthermore, a well developed shoulder is evident at 2.34
A, indicated in Figure 8.8 with an arrow. These factors taken together are strong evidence
for the existence of signiﬁcant numbers of short diselenide bonds in the material and

support the existence of the superstructure observed with the electron diffraction data

295

 

 

 

 

 

 

 

 

 

IIII'IIIIITTr
0" ..1
)- ..
:‘w- —
| _ .

Em- -

VJ

l—l __

0' ‘
O r
0'2 ....11.1.1...

5 10

Figure 8.7. X-ray powder diffraction pattern from KThZSeé. The data are shown in the
form of Q(S(Q)-1)'which is the structure function that is Fourier transformed to obtain
the PDF.

296

 

 

E l # I I T r j V f
. (a) Data ,
o
q— .
o
E'Se-Se
A
2,..— l -
g .
m
°.r -
o

 

 

q
T.
.11.
d
.1-

 

10014)

0.02 0.04 0.08 0.080

 

 

i

 

 

 

r(&)

Figure 8.8. Pair distribution functions in the form of p(r) from KTthCa. (a) shows the

PDF obtained from the data. (b) shows the PDF calculated from the crystallographic
model of the structure. In both cases an arrow has been placed at a distance of 2.34 A
which is the length expected for a diselenide bond. The lack of intensity at this position
in p(r) for the model reﬂects the fact that in the average structure this distance does not

exist; however it is clear from the data that this diselenide bond does exist in the material.

297

(see above). A more complete modeling analysis is underway to fully determine the local

arrangement of atoms in this material.35

4. Conclusions

AThZSe6 (A=K, Rb) has a two-dimensional structure related to the ZrSe3-type
and possesses a charge density wave. The stacking arrangement of these layers is slightly
modiﬁed from that of the ZrSe3 structure to accommodate the added alkali metal ions
between the layers. This structure is also different from that of the telluride analog, due
to the different nature of the chalcogens. TEM and PDF studies reveal that the short Se-
Se distances of 2.7 A, determined by single crystal X-ray diffraction, are an artifact of the
averaging effect of the single crystal structure and that there is a CDW modulation due to
an extra electron from the alkali metal which cleaves one out of four diselenide bonds.

The resultant structural distortion establishes a 4a x 4b superstructure due to the ordering
of Sef’ groups and Sez' ions. The PDF studies, which take into account the diffuse X-ray
scattering in the total diffraction pattern, directly and unequivocally expose the presence
of the Sezz' groups. These conclusions are supported by magnetic susceptibility
measurements, optical and Raman spectroscopic data. Finally, the well deﬁned
crystallographic sites of the K°r in the structure of ATh2Se6 may serve as a model for the
analogous sites for Li ions in LierSe3. The difference would be that in the latter only
tetrahedral or octahedral sites would be likely, formed by the negatively charged selenium
atoms from the cleaved diselenide bonds. The enhanced Li+-Se2‘ interactions could

stabilize the alkali metal in the interlayer space, preventing it from facile deintercalation.

298

Finally, the observation of CDW distortion in AThZSeG suggests that similar phenomena
may also exist in the related KThzTe6 and warrants a closer look into the structure of this

compound.

299

10.

References

(a) Sutorik, A. C.; Albritton-Thomas, J .; Kannewurf, C. R.; Kanatzidis, M. G. .1.
Am. Chem. Soc. 1994, 116, 7706-7713. (b) Kanatzidis, M. G.; Sutorik, A. C.
Progr. Inorg. Chem. 1995, 43, 151-265. (c) Sutorik, A. C.; Albritton-Thomas, J .;
Hogan, T.; Kannewurf, C. R.; Kanatzidis, M. G. Chem. Mater. , 1996, 8, 751-761.
(d) Sutorik, A. C.; Kanatzidis, M. G. Chem. Mater. 1996, 9, 387-398.

(a) Hanko, J. A. Ph. D. Dissertation, Michigan State University, 1998. (b) Hanko,
J. A.; Kanatzidis, M. G. J. Chem. Soc. Chem. Commun. 1998, 725-726. (c)
Hanko, J. A.; Kanatzidis, M. G. J. Alloys Comp. 1998, 280, 71.

Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G. Inorg. Chem.
1997, 36, 3804-3 805 and the references therein.

Chung, D.-Y.; Choi, K.-S.; Iordanidis, L.; Schindler, J. L.; Brazis, P.; Kannewurf,
C. R. ; Chen, B.; Hu, 8.; Uher, C.; Kanatzidis, M. G. Chem. Mater. 1997, 9, 3060-
3071.

Choi, K.-S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309.
Choi, K.-S ; Kanatzidis, M. G. Chem. Mater. 1999, 11, 2013-2018.

Kriinert, Von W.; Plieth, K.; Z. anorg. allg. Chem. 1965, 336, 207-218.

Noel, H. J Inorg. Nucl. Chem. 1980, 42, 1715-1717.

Sourisseau, C.; Gwet, S. P.; Gard, P. J. Solid State Chem. 1988, 72, 257-271 and
references therein.

Wu, E. J.; Pell, M. A.; Ibers, J. A. J. Alloys and Camp. 1997, 255, 106-109.

300

11.

12.

13.

14.

15.

l6.

17.

Cody, J. A.; Ibers, J. A. Inorg. Chem. 1996, 16, 3273-3277.

Found primarily in one- and two-dimensional metallic compounds, the CDW state
is characterized by a periodic density wave appearing in the conduction electron
density and a concomitant periodic superstructure of a longitudinal lattice-
distortion wave. The properties of CDW systems have been studied extensively
during the past 25 years, mostly in di- and trichalcogenides such as TaSz and
NbSe3. A large fraction of this research has been devoted to studies of the
nonlinear electronic transport exhibited by CDW systems and the nature of the
superstructrure. For comprehensive reviews of the electronic properties of CDW
systems see (a) Monceau, P. Electronic Properties of Quasi-One-Dimensional
Materials; Reidel: Dordrecht 1985; Pt. II, p. 139; (b) Griiner, G. Rev. Mod. Phys.
1998, 60, 1129. and references therein.

McCarthy, T. J.; Ngeyi, S.-P.; Liao J .-H; Degroot, D.; Hogan, T.; Kannewurf, C.
R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331-340.

Selwood, P. W. Magnetochemistry, 2nd ed.; Interscience Publishers: New York,
l956;p78.

Hall, S. R.; Flack, H. D.; Stewart, J. M. 1992, Editors. Xtal3.2 Reference Manual.
University of Western Australia, Australia, Geneva, Switzerland, and Maryland,
USA.

SMART: 1994, Siemens Analytical Xray Systems, Inc., Madison, Wisconsin
53719 USA.

SAINT: Version 4, 1994-1996, Siemens Analytical Xray Systems, Inc., Madison,

Wisconsin 53719 USA.

301

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

Sheldrick, G. M. University of Gottingen, Germany, to be published.
SHELXTL: Version 5, 1994, G. M. Sheldrick, Siemens Analytical Xray Systems,
Inc., Madison, Wisconsin 53719 USA.

Le Page, Y. J. Appl. Crystallogr. 1987, 20, 264-269.

Spek, A. L. Acta Crystallogr. Sect. A 1990, 46, C34.

B. E. Warren, X-ray Diﬁfaction ; Dover: New York, 1990.

(a) Egami, T. Mater. Trans. JIM 1990, 31, 163. (b) Billinge S. J. L.; Egami, T.
Phys. Rev. B 1993, 47, 14386.

Sharp, T. G.; Oden, P. 1.; Buseck, P. R. Surf Sci. Lett. 1993, 284, L405-L410.
Choi, K.-S.; Kanatzidis, M. G. unpublished work.

Ben Salem, A.; Meerschaut, A.; Rouxel, J. C. R. Acad. Sci, Paris [I 1984, 299,
617-619.

Nouvel, G.; Zwick, A.; Renucci, M. A.; Lockwood, D. J .; Noél, H. J. Phys. C:
Solid State Phys. 1987, 20, 1881-1897.

Nouvel, G.; Zwick, A.; Renucci, M. A. J. Less-Common Met. 1986, 121, 253-259.
Bottcher, P.; Getzschmann, J .; Keller, R. Z. anorg. allg. Chem. 1993, 619, 476-
488

(3) Lee, S.; Foran, B. J. J. Amer. Chem. Soc., 1996, 118, 9139-9147. (b) Zhang,
X.; Li, J.; Foran, B. J.; Lee, S.; Guo H.; Hogan, T.; Kannewurf, C. R.; Kanatzidis
M. G. J. Amer. Chem. Soc., 1995, 117, 10513-10520. (c) Lee, S.; Foran, B. J. J.
Amer. Chem. Soc., 1994, 116, 154-161. (d) Prodan, A.; Hla, S. W.; Marinkovic,
V.; Bohm, H.; Boswell, F. W.; Bennett, J. C., Phys. Rev. B-Cond. Matter. 1998,

57, 6235-6238.

302

 

31.

32.

33.

34.

35.

(a) Since we are looking down the [001] direction, the hk0 reﬂections which
satisfy the h+k=2n+l condition such as 100, 010, 210 should be systematically
absent due to the extinction rule for an I-centered cell, hkl, h+k+l=2n+1. The
weak reﬂections present at these positions (where the systematic absences occur)
are not violations, but result from the interaction of the Ewald sphere with
diffraction spots in the ﬁrst order [hkl] layer. This is due to the very thin
dimensions of the observed specimens (< 100A) along the crystallographic c-axis
which broaden the reciprocal lattice spots signiﬁcantly along the c°-axis. (b)
Reimer, L. Transmission Electron Microscopy, 2nd ed.; Springer Series in Optical
Sciences: Springer Verlag, 1989; pp 276-284.

An overview of the use of this and other scattering techniques for studying
disordered crystals is given in Billinge, S. J. L. Curr. Opinion in Solid State and
Mater. Sci, 1996, l, 477.

Billinge, S. J. L. Local Structure ﬁ'om Diﬁfaction, eds Billinge, S. J. L. and Thorpe,
M. F.; Plenum: New York, 1998, in press.

The data and PDF shown represent a preliminary analysis where corrections for
multiple scattering and Compton scattering have not been explicitly taken into
account and the sample form factor is approximated by smoothing the data. Our
experience is that these PDFs are qualitatively, though not quantitatively, correct
and that this preliminary analysis will not affect the qualitative observations we
are making here. A more complete analysis is in progress.

Choi, K.-S.; Billinge, S. J. L.; Kanatzidis, M. G. work in progress.

303

Chapter 9. A Novel Ternary Thorium Selenide, BaTh7Se18 Featuring

Infinite Linear Se Chains.

304

1. Introduction

After handful of binary thorium compounds (i.e. ThSez,l ThSe3,2 ThZSe3,3
Tthe5,2’4 Th7Selz') were reported more than two decades ago, not many investigations
have been done on the synthesis of new thorium chalcogenide compounds. It was very
recently that more complex ternary or quaternary thorium chalcogenides started to be
discovered (i.e. KTth28e6,5 AThZSeéf KThzTe67 , CsThzTe6,8 CuThzTeé, 9 SrThZSesg).
The compounds, AThZSe6 (A = K, Rb), are composed of alkali metal ions stabilized
between two-dimensional layers that are equivalent to those seen in ZrSe3. Since ThSe3
also adopts the ZrSe3 structure type, we recognized that AThZSe6 represents a well-
crystallized version of intercalated ThSe3 with 0.5 equiv of an alkali metal. The diselenide
bonds present in ThSe3 act as an reduction center, so the electron from the alkali metal is
used to cleave one out of four Se-Se bonds in a ordered fashion resulting in a 4a x 4h
superstructure. After this interesting observation, we intended to make BaTh2$e6
compound, which would be the intercalated version of ThSe3 with divalent Ba” ions. If
this phase could be stabilized, the diselenide groups in the layer would be reduced now by
two electrons from the Ba metal causing a different type of modulation in Se-Se distances
and, therefore, a different type of superstructure. Our attempt to synthesize BaThZSeé,
however, resulted in a discovery of an unexpected new compound, BaTh7Se18 instead,
discovery of which is serendipitous to us for several reasons. This compound is very
peculiar in composition and its presence would have not been predicted in advance. The
novelty of the compound is in its inﬁnite linear Se chains which extend along two

mutually perpendicular crystallographic axes. Linear chalcogenide chains are often

305

observed in polytelluride compoundsg'g’m but they are rarely found in polyselenide
compounds. This compound is also unique in that it introduces new type of coordination
environments for Th ions, which are quite different from most commonly observed mono-
, di-, and tri-capped trigonal prisms. Here we report the synthesis, structure, properties

of the new ternary thorium selenide, BaTh7Se18.

2. Experimental Section

2.1 Synthesis.

The following reagents were used as obtained: thorium, -100 mesh, 99.8%, Cerac,
Milwaukee, WI; selenium powder, 100 mesh, 99.5+%, Aldrich, Milwaukee, WI; barium
granules, <6mm, 99%, Aldrich, Milwaukee, WI.

BaTh7Se13 was synthesized from a mixture of 0.055 g (0.4 mmol) Ba, 0.186 g (0.8

mmol) Th, 0.205 g (2.6 mmol) Se. The reagents were thoroughly mixed, sealed in an

evacuated silica ampule, and heated at 800 0C for 5 days (cooling 2 °C/h). Pure copper
colored block-like crystals of BaTh7Se13 were obtained by isolation in degassed water
(yield > 90 % based on Th metal). The crystals are air- and water stable.

The compositions of the materials were analyzed by Scanning Electron
Microscope (SEM)/ Energy Dispersive Spectroscopy (EDS). Homogeneity for both
compounds was conﬁrmed by comparing the powder X-ray diffraction patterns of the

products against ones calculated using X-ray single crystal data.

2.2 Physical Measurements.

306

Solid-State UV/V is Spectroscopy. Optical diffuse reﬂectance measurements were
performed at room temperature with a Shimadzu UV-3101 PC double-beam, double-
monochromator spectrophotometer operating in the 200-2500 nm region. The instrument
is equipped with an integrating sphere and controlled by a personal computer. BaSO4
was used as a 100% reﬂectance standard for all materials. Samples were prepared by
grinding them to a ﬁne powder and spreading them on a compacted surface of the
powdered standard material and preloaded them into a sample holder. The reﬂectance
versus wavelength data generated were used to estimate a material's band gap by

converting reﬂectance to absorption data as described previously.ll

Differential Thermal Analysis (DTA). DTA experiments were performed on a
computer-controlled Shimadzu DTA-50 thermal analyzer. Typically a sample (~20 mg )
of ground crystalline material was sealed in quartz ampules under vacuum. A quartz

ampule of equal mass ﬁlled with A1203 was sealed and placed on the reference side of the
detector. The samples were heated to 800 °C at 5 OC/min, isothermed for 10 min

followed by cooling at -5 °C/min to 50 0C. Residues of the DTA experiments were
examined by X-ray powder diffraction. The stability/reproducibility of the samples was

monitored by running multiple heating/cooling cycles.

Raman Spectroscopy. Raman Spectra were recorded on a Holoprobe Raman
spectrograph equipped with a CCD camera detector using 633nm radiation from a HeNe

laser for exitation. Laser power at the sample was estimated to be about SmW and the

307

focused laser beam diameter was ca. 10 microns. 5 scans were needed to obtain good

quality spectra. The accumulation time of each scan was 50 sec.

2.3 X-ray Crystallography.

A single crystal of BaTh7Selg with dimensions 0.11 mm x 0.03 mm x 0.02 mm was
mounted on the tip of a glass ﬁber. Intensity data were collected at 293K on a Bruker
SMART Platform CCD diffractometer using graphite monochromatized Mo K01 radiation
over a fullsphere of reciprocal space. The individual frames were measured with an omega
rotation of 0.3 deg and an acquisition time of 65 sec per frame. The SMART software12
was used for the data acquisition and SAINTl3 for data extraction and reduction. The
absorption correction was done using SADABS.14 The complete data collection
parameters and details of the structure solution and reﬁnement are given in Table 9.1.

Structure solution and reﬁnement for the compound were performed with the
SHELXTL package of crystallographic programs.15 Systematic absence conditions of the
data set gave four possible space groups, Cmmm, Cmm2, Amm2, C222. The structure
was solved and reﬁned successfully in Cmmm. The coordinates of all atoms, isotropic
temperature factors, and their estimated standard deviations (esd’s) are given in Table 9.2.

The selected bond distances and angles are given in Table 9.3.

3. Results and Discussion

308

 

Table 9.1. Summary of Crystallographic Data and Structural Analysis for BaTh7Se13.

 

Formula
Formula weight
Crystal habit
Space group

a, A

b, A

c, A

Z; V, A3

Dcalc: g/Ctn3

temp, K
MMo K01), A

“(Mo K01), cm'l
F(000)

0 max , deg
Total data
Unique data
No. of variables

Reﬁnement method
Final R indices [I>20]

R indices (all data)

goodness of ﬁt on F2

a R]: ZIIFOI — IFCH/ZIFOI ;

BaTh7Se13

3182.9

Golden black plates
Cmmm

11.073(3)
21.613(5)
5.6644(14)

2, 1355.6(6)

7.798

293(2)

0.71073

638.27

2596

28.55

6655

970 [Rim = 0.0539]
47

Full-matrix least-squares on F2
R1a = 0.0409
WR2b = 0.1146

R1 = 0.0423

WR2 = 0.1153
1.186

b Wl{2:{2[W(1702 —Fc2 )2 ]/2[W(F02)2 ]}1/2

309

Table 9.2. Fractional Atomic Coordinates and Equivalent Atomic Displacement

Parameter (Ueq) Values for BaTh7Se13 with Estimated Standard Deviations in

 

 

Parentheses.

Atom x y z Ueq,a A2
Th(l) 0 0.2199(1) 0 0.014(1)
Th(2) 0.2577(1) 0.1263(1) 1/2 0.007(1)
Th(3) 0 0 0 0.013(1)
Se(1) 0 0.3547(1) 0.2510(5) 0.017(1)
Se(2) 1/4 1/4 0.2494(4) 0.015(1)
Se(3) 0 0.1639(2) 1/2 0.014(1)
Se(4) 0.1865(2) 0.0998(1) 0 0.017(1)
Se(5) 0.1221(3) 0 1/2 0.011(1)
Se(6) 0.3687(4) 0 1/2 0.022(1)

Ba 1/2 0 0 0.057(1)

 

a Ueq is deﬁned as one third of the trace of the orthogonalized U0. tensor.

310

 

Table 9.3. Selected Distances(A) and Bond Angles(°) for BaTh7Se13.

 

Bond Distances

 

 

 

Th(1)-Se(1) x 2 3.243(2) Th(3)-Se(4) x 4 2.986(2)
Th(1)-Se(2) x 4 3.1753(13) Th(3)-Se(5) x 4 3.1382(14)
Th(1)-Se(3) x 2 3.0793(15)
Th(1)-Se(4) x 2 3.316(3) Ba-Se(1) x 4 3.447(2)
Ba-Se(6) x 4 3.1835( 19)
Th(2)-Se(1) x 2 3.0589(15)
Th(2)-Se(2) x 2 3.0282(15) Se(1)-Se(1) 2.820(5)
Th(2)-Se(3) 2.9668(13) Se(1)-Se(1) 2.844(5)
Th(2)-Se(4) x 2 2.9951(10) Se(2)-Se(2) 2.825(5)
Th(2)-Se(5) 3.1 156(17) Se(2)-Se(2) 2.83 9(5)
Th(2)-Se(6) 2.9940(18) Se(5)-Se(5) 2.703(6)
Se(5)-Se(6) 2.731(5)
Se(6)-Se(6) 2.908(8)
Bond Angles
Se(1)-Th(l )-Se(4) 134.68(4) Se(4)-Th(3)-Se(4) 87.51(10)
Se(2)-Th(1)-Se(2) 156.32(4) Se(4)-Th(3)-Se(4) 180.0
Se(2)-Th(1)-Se(2) 5282(8) Se(4)-Th(3)-Se(5) 107.33(4)
Se(2)-Tb( l )-Se(2) 121.34(8) Se(5)-Th(3)-Se(5) 5103(10)
Se(3)-Th(1)-Se(3) 133 .78( 1 2) Se(5)-Th(3)-Se(5) 128.97(10)
Se(3)-Th(1)-Se(2 1 1932(4) Se(4)-Th(3)-Se(5) 7267(4)
Se(3)-Th(1)-Se(2) 7082(4) Se(4)-Th(3)-Se(4) 9249(1)
Se(1 )-Th(2)-Se( 1) 5490(9) Se(6)-Ba-Se(6) 125.66(13)
Se(1)-Th(2)-Se(5) 122.70(6) Se(6)-Ba-Se(6) 54.34(13)
Se(2)-Th(2)-Se(5) 139.47(5) Se(6)-Ba-Se(1) 6847(4)
Se(3 )-Th(2)-Se(5) 77.10(8) Se(6)-Ba-Se( l) 1 1 153(4)
Se(4)-Th(2)-Se(1) 79.66(6) Se(6)-Ba-Se(1) 4873(8)
Se(4)-Th(2)-Se(1) 133 .82(7) Th(2)-Se(6)-Ba 100.809(10)
Se(6)-Th(2)-Se(2) l44.73(5) Se(5)-Se(6)-Ba 1 1 717(6)

 

311

Table 9.4. Anisotropic Displacement Parameters for BaTh7Se13.

 

 

U11 U22 U33 U23 U13 U12
Th(1) 0.016(1) 0.013(1) 0.014(1) 0 0 0
Th(2) 0.006(1) 0.008(1) 0.007(1) 0 0 0.000(1)
Th(3) 0.009(1) 0.009(1) 0.019(1) 0 0 0
Se(1) 0.004(1) 0.013(1) 0.033(1) -0.006(1) 0 0
Se(2) 0.013(1) 0.002(1) 0.029(1) 0 0 0.000(1)
Se(3) 0.005(1) 0.021(1) 0.017(1) 0 0 0
Se(4) 0.028(1) 0.018(1) 0.006(1) 0 0 0.014(1)
Se(5) 0.012(1) 0.005(1) 0.015(1) 0 0 0
Se(6) 0.036(2) 0.003(1) 0.026(2) 0 0 0
Ba 0.123(5) 0.031(2) 0.016(2) 0 0 0

 

The anisotropic displacement factor exponent takes the form:
-27t2[h2a*2U11+ -------- +2hka*b*U12]

312

Structure.

The overall structure of BaTh7Se13 is shown in Figure 9.1. BaTh7Selg has a
complicated three-dimensional structure with three independent Th4+ centers. The local
enviromnents of these Th ions are unconventional and have not been observed for other
thorium chalcogenides, Figure 9.2. Th(l) is 10-coordinate and there is no proper name to
describe this type of environment. Five of the ten Se atoms form a puckered pentagonal
face at one end and four more form a planar square face at the other end. The remaining
one is used to cap the pentagonal face. The coordination of Th(2) is best described as a
distorted mono-capped square anti-prism with Se(5) capping one square face from a
severely tilted angle. Th(3) is 8-coordinate and is stabilized at the center of a cage formed
by two interpenetrated selenium rectangles with different size. The coordination
environment of Ba is very similar to that of Th(3). Only the shape of the rectangles and
the way they interpenetrate are slightly different, see Figure 92(d). Ba ions show high
atomic displacement parameters which are not affected by occupancy reﬁnement,
suggesting that these ions sit in large cages undergoing a rattling motion.

Each Th4+ center forms an one-dimensional Th chain by sharing corners or edges
of its polyhedra with neighbors as shown in Figure 9.3. All these chains run parallel to
the c-axis. The arrangement of Ba along the c-axis is also shown in Figure 9.3(d), which
are quite similar to that of Th(3). Both Th(3) and Ba chains are built by sharing edges of
horizontal rectangles around these ions.

These different kinds of one-dimensional chains then are interconnected to
fabricate a three-dimensional framework. The simplest way to look at this complicated

structure is to decompose it into Th(2)-selenide skeleton and Th(1), Th(3), and Ba

313

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure. 9.1. The overall structure of BaTh7Se13 view down the c-axis.

314

 

$95 895 891 861

 

Se4 Sei Se1

Figure. 9.2. Coordination environment of (a) Th(1), (b) Th(2), (c) Th(3), and ((1) Ba.

315

(a) 2.844(5) 2.820(5)

         
 
 
     
 
 

2.825t5)
————> [001]

2.839(5)

2.820(5)
2.844(5)
i l

0;;
Ali"

2.
O

  
     

   

    
  
  
   

   
  
      
 

   
 
    

       
 
     
     
  
 
        

 

   
  
   

 

    
    

      
 

  
    

    

 

/ [001]
(9) Set “

Se4 Se4 .0391

o o 0

$95 Se6 .e6 . '1 . 0
395 ..m- '. ”.3“er
$95 $95 I
g 595 o A 0 9 Q 2.908(8)
Se4 Se4 s . 2.703(6) 2.731(5) 0
61Set

.—————> [100]

Figure. 9.3. Connectivity of (a) Th(1), (b) Th (2), (c) Th (3), and ((1) Ba along the c-axis.
(e) Arrangement of Th(3) and Ba along the a-axis. The Se-Se distances(A) associated

with the Se chains are shown in the ﬁgure.

316

cations. Th(2) is coordinated by all six independent Se atoms and the resulting Th(2)
polyhedra are three-dimensionally connected to construct a basic skeleton for a whole
structure, see Figure 9.4. This simpliﬁed [Th4Se18]'4' framework clearly shows a channel
and cages where Th(1), Th(3), and Ba can reside.

A unique characteristic of this structure is the presence of inﬁnite linear Se chains.
There exist three independent Se chains, two of which run along the c-axis and the third
propagates along the a-axis. The ﬁrst chain is composed of Se(1) with two different Se-Se
distances, 2820(5)A and 2.844(5)A, alternating. The second chain is composed of Se(2)
with two Se-Se distances of 2825(5)A and 2839(5)A alternating. These Se(1) and Se(2)
chains are shared by Th(l) and Th(2) chains and they run parallel to the c-axis, see
Figures 9.3(a) and (b). The third chain, which is parallel to the a-axis, is composed of
Se(5) and Se(6) in an order of -Se(5)-Se(5)-Se(6)-Se(6)-, see Figure 9.3(e). Such chains lie
side by side perpendicular to the c-axis and are separated by the Th(3) and Ba ions ,
which sit between these chains. The arrangement of Th(3) and Ba coordinated by Se(5)-
Se(6) chains is shown in Figure 9.3(e). The Se-Se distances found in this chain is
2703(6)A, 2.731(5)A, and 2908(8)A for Se(5)-Se(5), Se(5)-Se(6), Se(6)-Se(6),
respectively. The shortest Se(5)-Se(5) bond forms the edge of the horizontal rectangle
around Th(3) and the longest Se(6)-Se(6) bond makes the edge of the horizontal rectangle
around the Ba. Between Th(3) and Ba compartment, exist empty rectangles with no
cation inside. The Se(5)-Se(6) bond forms the edges of these empty rectangles.

The Se-Se distances observed in these chains are signiﬁcantly longer than the
normal Se-Se distances found in di-selenide groups (i.e., 2.34 A for ZrSe3).16 The

assignment of formal oxidation states of Se ions involving these chains can be expected

317

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure. 9.4. The [Th48e13]14' framework view down the c-axis.

318

between -1 (as Se' in the diselenide group) and -2 (as normal Sez' ion). To preserve charge
neutrality of the compound, the formal oxidation state of the Se in the Se chain should be
assigned as -1.5, so that the compound can be described as Ba2+(Th4+)7(Se2')6(Se"'5)12.
These signiﬁcantly elongated chalcogen-chalcogen interactions are not commonly
observed in polyselenide compounds while it is often observed in polytelluride
compounds. The more diffuse Te ions with less electronegativity have the greater
tendency to form Te---Te bonding interactions that range between a normal single bond
and no bond, resulting in the formation of Te chains (i.e. LizTe6,l° Rb2Te5”), rings (i.e.
RbTe6 18, Cs3Te2219) and nets (i.e. NdTe320, LaTeZZ') with almost equal Te-Te distances.
When Se chains or nets are formed, they prefer to distort forming fragments such as di-
selenide or tri-selenide with normal Se-Se distances. The distances between these
fragments become much longer after the distortion.”23
This known tendency of selenide ions prompted us to carefully look for a
possible superstructure that might reveal a modulation of Se-Se distances. In addition,
large U33 anisotropic displacement parameters (ADPS) of Se(1) and Se(2) as well as large
U1 1 ADPs of Se(6) and Ba(l) can also be a indication of a superstructure, see Table 9.4.
Indeed, when we recollected intensities data from much bigger crystals with longer
exposure time per frame, we did observe additional supercell reﬂections which would
double the c-axis. However, these reﬂections were extremely weak and we could not
collect enough data to precisely examine what causes this superstructure. We attempted

to probe the superstructure by electron diffraction expecting an even more complicated

superstructure along the c-axis or other directions but it failed because the three-

319

dimensional morphology of the compound made it difﬁcult to orient specimens along a

proper zone-axis.

Spectroscopic Chracterization and Thermal Analysis.
The compound is a semiconductor and its absorption spectrum shows the
presence of an abrupt optical gap at 0.63 eV, see Figure 9.5. Differential thermal analysis

(DTA) experiments showed that BaTh7Se13 does not melt under 1000 °C. The Raman

spectrum shows a broader shift at 208 cm], which is assigned to the stretching vibration
of the Se-Se bonds in the Se chains, see Figure 9.6. This value is relatively low in
frequency compared with those for other Se-Se/S-S stretching vibration reported for other
-1 for

compounds with discrete di-selenide groups with normal Se-Se distances (i.e. ~253 cm

K2862, 24 "”266 CIT].l fOl' KzGdzsbzsegzs).

4. Conclusions

The unexpected discovery of BaTh7Se18 revealed a new structure type, which
stabilizes inﬁnite Se chains in a complicated three-dimensional framework. The striking
feature of the Se chains is that they are composed of signiﬁcantly elongated Se-Se
distances without severe modulations. The ﬁnely fabricated three-dimensional structure
of the compound seems to restrict the freedom of distortion in the Sc chains and forces
the telluride-like behavior on the selenide ions in the chain. The formal charges of the

selenide ions associated with the chains are not certain at this point. Complete structural

320

 

    

 

 

 

2.0:-
g :
S 1.5:"
2:: I
g 1.0:-
m 0 5
B . Eg = 0.63 CV

0.4 0.6 0.8 1.0 1.2 1.4

Energy (eV)
Figure. 9.5. Optical absorption spectrum of BaTh7Se13. The band gap value, E8, is

shown.

 

188

 

Arbitrary Inensity

 

 

LlllllllllllJlll

100 150 200 250 300 350 400
Raman Shift (cm-1)

Figure. 9.6. Raman spectrum of BaTh7Se13.

321

reﬁnement of the superstructure is necessary to determine the structural details of the

inﬁnite Se chains.

322

10.

ll.

12.

13.

References

D’Eye, R. W. M. J. Chem. Soc. 1953, 1670-1672.

Noél, H. J. Inorg. Nucl. Chem. 1980, 42, 1715-1717.

D’Eye, R. W. M.; Sellman, P. G.; Murray, J. R. J. Chem. Soc. 1952, 2555-2562.
Graham, J.; McTaggart, F. K. Aust. J. Chem. 1960, 13, 67-73.

Choi, K.-S.; Iordanidis, L.; Chondroudis, K.; Kanatzidis, M. G., Inorg. Chem.
1997, 36, 3804-3805.

Choi, K.-S ; Patschke, R.; Billinge, S. J. L.; Waner, M. J .; Dantus, M; Kanatzidis,
M. G. J. Am. Chem. Soc. 1998, 120, 10706-10714.

Wu, E. J.; Pell, M. A.; Ibers, J. A., J. Alloys and Compd. 1997, 255, 106-109.
Cody, J. A.; Ibers, J. A.; Inorg. Chem. 1996, 16, 3273-3277.

Narducci, A. A. ; Ibers, J. A.; Inorg. Chem. 1998, 37, 3798-3801.

B6ttcher, P.; Keller, R. J. Less Common Met, 1985, 109, 311 and references
therein.

McCarthy, T. J .; Ngeyi, S.-P.; Liao J .-H; Degroot, D.; Hogan, T.; Kannewurf, C.
R.; Kanatzidis, M. G. Chem. Mater. 1993, 5, 331-340.

SMART: Siemens Analytical X-ray Systems, Inc., Madison, WI 53719, USA,
1994.

SAINT: Version 4, Siemens Analytical X-ray Systems, Inc., Madison, WI 53719,

USA, 1994-1996.

323

14.

15.

l6.

17.

18.

19.

20.

21.

22.

23.

24.

25.

G. M. Sheldrick, University of Géttingen, Germany, to be published.
SHELXTL: Version 5, G. M. Sheldrick, Siemens Analytical Xray Systems, Inc.,
Madison, WI 53719, USA, 1994.

Kr6nert, Von W.; Plieth, K.; Z anorg. allg. Chem. 1965, 336, 207—218.

Btittcher , F.; Krestschmann, U. .1. Less Common Met. 1983, 95, 81.

Sheldrick, W. S.; Schaaf, B. Z Naturforsch. 1994, 49, 993.

Sheldrick, W. S.; Wachhold, M. Angew, Chem. 1995, 107, 490; Angew. Chem. Int.
Ed Engl. 1995, 34, 450.

Norling, B. K.; Steinﬁnk, H. Inorg. Chem. 1996, 5, 1488.

Wang, R.; Steinﬁnk H.; Bradley, W. F. Inorg. Chem. 1996, 5, 142

(a) Be’nazeth, P. S.; Carré, D.; Laruelle, P. Acta Crystallogr. 1982, B38, 33. (b)
Bénazeth, P. S.; Carre’, D.; Laruelle, P. Acta Crystallogr. 1982, B38, 37.

van der Lee, A.; Hoistad, L. M.; Evain, M.; Foran, B. J .; Lee, S. Chem. Mater.
1997, 9,218.

Béttcher , P.; Getzschmann, J .; Keller, R., Z anorg. allg. Chem. 1993, 619, 476.
Choi, K. S.; Hanko, J. A.; Kanatzidis, M. G. J. Solid State Chem. 1999, 147, 309-

319.

324

 

 

 

11111111))1111111))11111311111