 

 

 

.. 7
.~ ..
3:!
5...":

V -
Manda“
.. .

.‘n:
.... A...

5......

, ﬂow as.

x .. 3.3. P1...

1.2... 2.

lien-‘5‘

1‘ .

2.
y
.i. . 1.
its. a... .. ... egg
hr; . $51.3... was .
2:53... .2: ..... I ..
.13‘)S\.JIU I‘i‘ Lrubnulil |' V ‘1
.35. s... I: .1 b . .1
7.4.. 3.. .1 :rl . i
12.3.... AI
.22..
. '15..
l I
2: $3.5,
.2. .

5.
ii
.. . I. ha
. .395. 70.5.2.9...- R...’ {,2
c.5339. 5.3. t A a
.r... I .. Inﬂux
n x: 3.... \lnlsazai .
. 3......ilvl...
cur....§:v.a-l¢ .. o t
«5: 2;: at 5 .v
1 {In 3...: I v.3.- . I»... .
. .x.3.$22.3(1.3 2 .ch§..\uanl
x :1 . . E, A
\3..I.....IJ4I a}... «12?. a
.1}... i a}: $0.1 . 131%,
it 3:...2.‘ , ,

Elle" 2‘
Ar. ﬁuumgnwsm
s
i.» ,2
1&xruawmha
“Rb"...
2.3V ._..
a». 2L.
5...:
2...:

. "v r I

. . v

:1 .. I l: :.i.

2.2.1.5231? E35”
9.51....

3.2 2..» k

1...: .. Humiai ,
A Z Il‘l I
if. 1.2..

x? . 2‘. I
32.11 :‘i. ..
.1}! n

 

I“...

300?

J

LIBRARY
gan State
Universi

Mlchi

 

 

 

This is to certify that the
dissertation entitled

Exploratory Synthesis of Complex lnterrnetallic Tetrelides and
Antimonides Using Liquid Aluminum as a Solvent

presented by

Xiuni Wu

has been accepted towards fulﬁllment
of the requirements for the

degree In Chemistry

 

 

wail/ox,

 

Major Professor’ 3 Signature
5] 4 0 / 2 o a

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

 

-.-——— -.- n.—.__._V- .. A

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/CIRC/DateDue.indd-p.1

 

EXPLORATORY SYNTHESIS OF COMPLEX INTERMETALLIC TETRELIDES
AND ANTIMONIDES USING LIQUID ALUMINUM AS A SOLVENT

By

Xiuni Wu

A DISSERTATION
Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Chemistry

2006

ABSTRACT

EXPLORATORY SYNTHESIS OF COMPLEX INTERMETALLIC TETRELIDES
AND ANTIMONIDES USING LIQUID ALUMINUM AS A SOLVENT

By
Xiuni Wu

The metal ﬂux method has been recently suggested as a preparative tool to
synthesize intermetallic compounds. Our group initiated the investigations on the
quaternary systems RE/TM/Al/T r (RE = rare earth metal, TM = transition metal, Tr =
tetrelide) using Al as a ﬂux. In this dissertation we employed the Al ﬂux to study the
systems RE/Au/Al/Ge, TM/Al/T r, RE/Ni/Al and RE/Al/Sb aiming at discovering new
materials with interesting physical properties.

We extended the studies involving first row transition metals to the third row
transition metals such as Au. Our exploratory investigations led to the discovery of two
quaternary families REAuAl4Gez and REAuAl4(AuxGel-x)2, the structures of which are
related and both can be represented as alternating stacking of RE layers and AuAl4Ge2 (or
AuAl4(AuxGe1-x)2) slabs along the c-axis.

By studying the ternary systems TM/Al/T r (TM = V, Co) in liquid aluminum, we
obtained new ternary compounds V2A15Ges, C019A1458im.x (x = 0.13) and CosAlMSiz.
The structure of VzAlsGes features pentagonal columns composed of Al and Ge atoms
with the V atoms forming a long-short alternating chain residing in the center of the
column. Both CoigADSSiHH (x = 0.13) and CosAIMSiz show exceptionally complex
structures and large unit cells. C05A114Si2 shows interesting thermal oxidation resistance

up to 1000°C which is due to the formation of an A1203 layer on the surface of the

crystals.

Th
due to tla
transition.
properties
were SIUdl.
compound
the Fe an
SUSCC’lebllZ

Dur
Phases Yb
thﬁ’l'llloclm~
“'llh the A;

Oxidation S'.

 

The ternary compounds Yb3Ni5A119 and YbHNi3A139, are particularly interesting
due to their unusual physical properties such as mixed valence and metamagnetic
transition. The thermal expansion properties of Yb3Ni5Allg as well as the magnetic
properties with substitutions of Ni by the other transition metals (e. g. Mn, Fe and Cu)
were studied. During the study of substituting Ni by Fe, we discovered a new quaternary
compound YbNi2-xFexAlg. Both magnetic properties and Mossbauer spectra show that
the Fe atoms do not carry a magnetic moment. Temperature dependent magnetic
susceptibility indicates that the Yb atoms are in intermediate oxidation state.

During the investigations of the system Yb/Al/Sb, we isolated two new Zintl
phases Yb3Ale3 and Yb9A13Sb9 from molten Al, among which Yb3Ale3 is the more
thermodynamically stable form. These two compounds exhibit related structure types
with the Ale4 tetrahedral unit as the common feature, and the Yb ions are in divalent

oxidation state.

Fir-
Professor ‘
the past ii
101 from h}:

I u
hlcCuskcr.
discussions
Mohanti an

for me.

[We

 

has been a
infOn'nal an
me many ex

pIOI‘ided ml

ACKNOWLEDGMENTS

First of all, I would like to express my deep and sincere gratitude to my advisor,
Professor Mercouri G. Kanatzidis, for his guidance, encouragement and support during
the past ﬁve years. Professor Kanatzidis is a very knowledgeable chemist and I learned a
lot from him. He has helped me strive to be a successful scientist.

I would also like to acknowledge my committee members, Professors James
McCusker, Thomas Pinnavaia and James Geiger for their valuable suggestions and
discussions. Dr. Reza Loloee helped me with magnetic experiments; Professor S. D.
Mohanti and his students Daniel Bilc and Zsolt Rak did electronic structure calculations
for me.

I want to thank all the Kanatzidis group members for their help and friendship. It
has been a real pleasure working with so many wonderful and talented colleagues in an
informal and friendly environment. Special thanks go to Dr. Susan Latturner, who taught
me many experimental techniques in my ﬁrst year at MSU.

It would not have been possible for me to succeed without the love and support
from my parents, my sister and brother-in-law. They have always supported and
encouraged me to pursue my goals. I would also like to thank my parents-in-law who
provided much needed help to take care of my daughter during the past two years. I have
been very fortunate to have my husband Yiqian with me. His love, his conﬁdence in me
and valuable advice have always been there during the toughest years of my Ph.D. study.

Finally the ﬁnancial support from Department of Energy is gratefully

acknowledged.

iv

LIST OF T
LIST OF E

LIST OF .

Chapter 1.

h)
D)

 

TABLE OF CONTENTS

LIST OF TABLES .............................................................................................................. x

LIST OF FIGURES .......................................................................................................... xv

LIST OF ABBREVIATIONS ......................................................................................... xxii

Chapter 1. Introduction .................................................................................................. 1

1.1 Introduction to Intermetallics ....................................................................... 1

1.2 Crystal Structure and Basic Properties of Intermetallics ............................. 3

1.3 Manipulation of Metal Flux to Synthesize Intermetallics ............................ 4

1.4 Rationale for Al Flux Synthesis ................................................................... 7

1 .5 Experimental techniques ............................................................................ l 3
Chapter 2. REAuAl4Ge2 and REAuAl4(AuxGe1-x)2 ( RE = rare earth element): New

Quaternary Intermetallics Grown from Aluminum Flux ........................... 2]

2. 1 Introduction ................................................................................................ 2]

2.2 Experimental Section ................................................................................. 22

Reagents ......................................................................................... 22

Synthesis ........................................................................................ 23

Isolation .......................................................................................... 24

Elemental Analysis ........................................................................ 24

X-ray Crystallography ................................................................... 24

Magnetic Characterization ............................................................. 32

2.3 Results and Discussion .............................................................................. 32

Synthesis ........................................................................................ 32

Crystal Structure ............................................................................ 34

Magnetic Properties ....................................................................... 42

2.4 Conclusions ................................................................................................ 48

Chapter 3.

'JJ

L!)

'J)
{A

3.4

Chapter 4.

4.1
4,»

'~

4.3

 

Chapter 3.

3.1
3.2

3.3

3.4

Chapter 4.

4.1
4.2

4.3

V2A15Ge5: First Ternary Intermetallic in the V-Al-Ge system Accessible in

Liquid Aluminum ....................................................................................... 51
Introduction ................................................................................................ 5 1
Experimental Section ................................................................................. 52
Reagents ......................................................................................... 52
Synthesis ........................................................................................ 53
Isolation .......................................................................................... 53
Elemental Analysis ........................................................................ 53
X-ray Crystallography ................................................................... 54
Thermal Analysis ........................................................................... 55
Magnetic Characterization ............................................................. 58
Electron Structure Calculation ....................................................... 58
Results and Discussion .............................................................................. 58
Synthesis ........................................................................................ 58
Crystal Structure ............................................................................ 59
Magnetic Properties ....................................................................... 65
Thermal Analysis ........................................................................... 66
Band Structure Calculations .......................................................... 66
Conclusions ................................................................................................ 67

Structurally Complex Cobalt Intermetallics Grown from Liquid

Aluminum: C019A145,Si10.x (x = 0.13) & C05A114Si2 .................................. 72
Introduction ................................................................................................ 72
Experimental Section ................................................................................. 73
Reagents ......................................................................................... 73
Synthesis ........................................................................................ 74
Elemental Analysis ........................................................................ 75
X-ray Crystallography ................................................................... 75
Thermal Analysis ........................................................................... 83
Magnetic Characterization ............................................................. 83
Results and Discussion .............................................................................. 83
Synthesis ........................................................................................ 83

vi

4.4

Jr

Chapter

Part I.

5.1.

5.1.1

 

4.4

Chapter 5.

Part 1.

5.1.1
5.1.2

Part II.

5.2.1
5.2.2

5.2.3

Crystal Structure ............................................................................ 84

Physicochemical Properties ........................................................... 99
Magnetic Properties ..................................................................... 10]
Conclusions .............................................................................................. 1 02

Exploratory Studies on the Ternary System RE/Ni/Al Employing Al as a

Flux .......................................................................................................... 105
Synthesis, Crystal Structure and Magnetic Properties of RE3Ni5A119 (RE =
Sm, Dy, Er, Yb) ....................................................................................... 105
Introduction .............................................................................................. 1 05
Experimental Section ............................................................................... 107
Reagents ....................................................................................... 107
Synthesis ...................................................................................... 107
Isolation ........................................................................................ 108
Elemental Analysis ...................................................................... 108
X-ray Crystallography ................................................................. 110
Physical Properties Characterization ........................................... 119
Results and Discussion ............................................................................ 119
Synthesis ...................................................................................... 119
Crystal Structure .......................................................................... 122
Physical Properties ....................................................................... 125
Conclusions .............................................................................................. 1 36

Flux Synthesis and Characterization of a New Ternary Phase Yb1_1Ni3Alg,9

Introduction .............................................................................................. 140
Experimental Section ............................................................................... 142
Reagents ....................................................................................... 142
Synthesis ...................................................................................... 143
Elemental Analysis ...................................................................... 143
X-ray Crystallography ................................................................. 144
Magnetic Characterization ........................................................... 148
Results and Discussion ............................................................................ 148

vii

Chapter 6

Part1
6.1

6.1

6.1.

 

5.2.4

Chapter 6.

Part I.
6.1.]

6.1.2

Part II.

6.2.]
6.2.2

6.2.3

Crystal Structure .......................................................................... 148
Magnetic Properties ..................................................................... 154

Conclusions .............................................................................................. 1 56

Substitution of Ni in Yb3Ni5A119 with the Other Transition Metals (Cu, Fe,

Mn) and the Discovery of New Intermetallic Phases RENi2-xFexAlg ...... 158
Doping Studies of Yb3Ni5A119 with Cu, Fe and Mn Substitutions .......... 158
Introduction .............................................................................................. 1 5 8
Experimental Section ............................................................................... 162
Reagents ....................................................................................... 162
Synthesis ...................................................................................... 162
Single Crystal X-ray Crystallography .......................................... 163
Magnetic Measurements .............................................................. 163
Results and Discussion ............................................................................ 164
Synthesis and Compositional Variations ..................................... 164
Magnetic Properties ..................................................................... 174
Conclusions .............................................................................................. 1 79

Discovery of New Intermetallic Phases RENi2-xFexAlg (RE = Eu, Yb)

from Liquid Aluminum ............................................................................ 182
Introduction .............................................................................................. 1 82
Experimental Section ............................................................................... 183
Reagents ....................................................................................... 183
Synthesis ...................................................................................... 184
Elemental Analysis ...................................................................... 184
X-ray Crystallography ................................................................. 185
Physical Properties Characterization ........................................... 190
Mossbauer Spectroscopy ....................................... 191
Results and Discussion ............................................................................ 191
Synthesis ...................................................................................... 191

viii

6.2

Chapter 7.

 

\J
b)

Chapter 3

Crystal Structure .......................................................................... 192 .

Physical Properties ....................................................................... 193
Mossbauer Spectroscopy ............................................................. 201
6.2.4 Conclusions .............................................................................................. 203

Chapter 7. Synthesis and Characterization of New Zintl Phases:

Yb9A13Sb9 and Yb3AISb3 ......................................................................... 206

7. 1 Introduction .............................................................................................. 206
7.2 Experimental Section ............................................................................... 208
Reagents ....................................................................................... 208

Synthesis ...................................................................................... 208

Differential Thermal Analysis ..................................................... 209

X-ray Crystallography ................................................................. 209

Charge Transport Measurements ................................................. 212

Magnetic Susceptibility Measurements ....................................... 212

Band Structure Calculations ........................................................ 212

7.3 Results and Discussion ............................................................................ 217
Synthesis and Differential Thermal Analysis .............................. 217

Crystal Structure .......................................................................... 21 7

Charge Transport Measurements ................................................. 220

Magnetic Susceptibility Measurements ....................................... 221

Band Structure Calculations ........................................................ 222

7.4 Conclusions .............................................................................................. 23 1
Chapter 8. Conclusions and Future Work ................................................................. 234

ix

Table 3-1.
Table 2-3.

Table 2-3.

Table 2-4_

Table 2-5.

Table 2-6.

Table 3.7.

Table 2.3

Table 2-9.

Table3-1.

 

Table 2-1.

Table 2-2.

Table 2-3.

Table 2-4.

Table 2-5.

Table 2-6.

Table 2-7.

Table 2-8.

Table 2-9.

Table 3-1.

Table 3-2.

LIST OF TABLES

Crystal data and structure reﬁnements for REAuAl4Gez (RE = Ce, Nd). .26
Crystal data and structure reﬁnements for REAuAl4Ge2 (RE = Gd, Er). ..27

Atomic coordinates and equivalent isotropic displacement parameters (A2
x 103) for REAuAl4Ge2 (RE = Ce, Nd, Gd, Er) ......................................... 28

Anisotropic displacement parameters (A2 x 103) for REAuAl4Ge2 (RE =
Ce, Nd, Gd, Er). ......................................................................................... 29

Crystal data and structure reﬁnements for EuAuA14(AuxGel-x)2 (x = 0.4)
and CeCuAl4(CuxGe1-x)2 (x = 0.62). .......................................................... 30

Atomic coordinates and equivalent isotropic displacement parameters (A2
x 103) for EuAuA14(AuxGe1-x)2 (x = 0.4) and CeCuAl4(CuxGe1-x)2 (x =
0.62). .......................................................................................................... 31

Anisotropic displacement parameters (A2 x 103) for EuAuAl4(AuxGe1-x)2
(x = 0.4) and CeCuAl4(CuxGe1-x)2 (x = 0.62). ........................................... 31

Bond lengths (A) for REAuAl4Ge2 (RE = Ce, Nd, Gd, Er). ...................... 38

Bond lengths (A) for EuAuA14(AuxGe.-x)2 (x = 0.4) and CeCuAl4(CuxGel_
x); (x = 0.62) ............................................................................................... 42

Crystal data and structure reﬁnements for V2A15Ge5. ............................... 56

Atomic coordinates and equivalent isotropic displacement parameters (A2
x 103) for VzAlsGes. .................................................................................. 57

 

Table 3-3

Table 34

Table 4-1

Table 4-3

Table 4.3.

Table 4-4.

 

Table 4-5.

Tabb 4-6.

Table 4.7

Tables‘].:

Table 5.1-:

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 5-1-1.

Table 5-1-2.

Table 5-1-3.

Anisotropic displacement parameters (A2 x 103) for VzAlsGes ................. 57
Bond lengths (A) for vatsoes .................................................................. 65

Selected crystal data and structure reﬁnement details for C019A1458110 x (x
"' -.0 l3) and C05A114Slz. .............................................................................. 76

Atomic coordinates (x 104) and equivalent isotropic displacement

parameters (A2 x 103) for ColgAlassilox (x — 0.13). Estimated standard
deviations are in parentheses. .................................................................... 77

2
Anisotropic displacement parameters (A x 103) for C019A145Si10.x (x =

0.13). .......................................................................................................... 78
Selected bond distances for ColgAl4SSim.x (x = 0.13). .............................. 80
Selected bond distances for Co5A114Si2. .................................................... 80

Atomic coordinates (x 104) and equivalent isotropic displacement

parameters (A2x103) for C05A114Siz. Estimated standard deviations are in
parentheses. ................................................................................................ 81

2 3
Anisotropic displacement parameters (A x 10 ) for C05A114Siz ............... 82

Selected crystal data and structure reﬁnement details for Sm3Ni5A119 and
DY3N15A119. .............................................................................................. I I 1

Selected crystal data and structure reﬁnement details for Er3Ni5A119 and
Yb3Ni5A119. .............................................................................................. l 12

Atomic coordinates (A x 104) and equivalent isotropic displacement
parameters (A2 x 103) for Sm3Ni5Allg and Dy3Ni5A119 ............................ 113

xi

Table 5-1

Table 5-1

Table 5-1

Table 5- 1-

Table 5-3-

Table 5-3-

Table

Jr
J

.
."-
-

t
I

Table 5-

[J

-ﬁ

Table 6-1-}

Table 6-1-:

 

Table 6‘1.3

Table 6~1 4

 

Table 5-1-4.

Table 5-1-5.

Table 5-1-6.

Table 5-1-7.

Table 5-2-1.

Table 5-2-2.

Table 5-2-3.

Table 5-2-4.

Table 6-1-1.

Table 6-1-2.

Table 6-1-3.

Table 6-1-4.

Atomic coordinates (A x 104) and equivalent isotropic displacement
parameters (A2 x 103) for Er3Ni5A119 and Yb3NisA119. ............................ 114

Anisotropic displacement parameters (A2 x 103) for Sm3Ni5A119 and
DY3N15A119. .............................................................................................. I 15

Anisotropic displacement parameters (A2 x 103) for Er3Ni5A119 and
Yb3Ni5A119. .............................................................................................. 116

Bond distances (A) for RE3Ni5Allg (RE = Sm, Dy, Er, Yb). ................... 117
Selected crystal data and structure reﬁnement details for Yb.,.Ni3Alg,9..146

Atomic coordinates (A x 104) and equivalent isotropic displacement

parameters (A2 x 103) for Yb.,lNi3Alg,9. .................................................. 147
Anisotropic displacement parameters (A2 x 103) for Yb._lNi3Alg_9. ........ 147
BOHd lengths (A) for Yb],|Nl3Alg_9. ......................................................... 150

Elemental ratios for the ﬂux synthesis of Yb3Ni5A119, Yb3Ni5-xCuxA119,
Yb3N15-xFexA119 and Yb3N15-anxAllg (unit: 11111101). .............................. I68

Compositional variations of cell volume for Yb3Ni5-xTMxA119 (TM = Cu,
Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of 100 K,
173 K and 298 K. ..................................................................................... 172

Compositional variations of a cell edge lengths for Yb3Ni5-xTMxA119 (TM
= Cu, Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of
100 K, 173Kand 298 K. ......................................................................... 172

Compositional variations of b cell edge lengths for Yb3Ni5-xTMxA119 (TM
= Cu, Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of
100 K,173Kand 298 K. ......................................................................... 173

xii

Table 6-l

Table 6-‘
Table 6-:
Table 6-2

Table 6-2

Table 6-3-
Table 6-3-.

Table 7-1.

Table 7-3 '

Table 7. 3_

TAble 7.4.

 

Table 7.5_

 

 

Table 6-1-5.

Table 6-1-6.

Table 6-2-1.

Table 6-2-2.

Table 6-2-3.

Table 6-2-4.

Table 6-2-5.

Table 7-1.

Table 7-2

Table 7-3.

Table 7-4.

Table 7-5.

Compositional variations of 0 cell edge lengths for Yb3Ni5-xTMxA119 (TM
= Cu, Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of
100K,173Kand298 K. ......................................................................... 173

Magnetic property parameters of Yb3N15-xCuxA119 and Yb3Ni5-xFexA119 (x
= 0, 0.5, 1). ............................................................................................... 176

Selected crystal data and structure reﬁnement details for EuNi2-xFexA13
and YbNi2-xFexAlg .................................................................................... 187

Atomic coordinates (A x 104) and equivalent isotropic displacement
parameters (A2 x 103) for EuNi2-xFexAlg and YbNi2-xFexAlg. .................. 188

Anisotropic displacement parameters (A2 x 103) for EuNiz-xFexAlg and
Ylez-xFCxAlg. ......................................................................................... 189

Bond lengths (A) for 13111~112_,,1=e,,A18 and YbNiz-xFexAlg .......................... 196
57F e Méssbauer spectra parameters for YbNi2-xFexAlg at RT and 80 K. .203

Selected crystal data and structure reﬁnement details for ngAl3Sb9 and
Yb3Ale3. ................................................................................................. 213

Atomic coordinates (A x 104) and equivalent isotropic displacement
parameters (A2 x 103) for Yb9A13Sb9. ...................................................... 214

Anisotropic displacement parameters (A2 x 103) for ngAthg. ............ 215

Atomic coordinates (A x 104) and equivalent isotropic displacement
parameters (A2 x 103) for Yb3AISb3. ....................................................... 216

Anisotropic displacement parameters (A2 x 103) for ngAle3. ............. 216

xiii

Table 74‘

Table 77-—

 

Table 7-6. Selected bond lengths (A) for Yb9A13Sbg. ............................................... 228

Table 7-7. Selected bond lengths (A) for Yb3Ale3. ................................................ 228

xiv

Figure 1-

Figure M

Fig”, 2-8.

 

 

Figure 1-1.

Figure 2-1.

Figure 2-2.

Figure 2-3.

Figure 2-4.

Figure 2-5.

Figure 2-6.

Figure 2-7.

Figure 2-8.

LIST OF FIGURES

Figure 1-1. A typical temperature heating proﬁle used for Al ﬂux
synthesis. .................................................................................................... 14

SEM images of typical crystals of (A) CeAuAl4Ge2.
(B) EuAuAI4(AuxGe1-x)2 (x = 0.4). ............................................................ 33

Cell volume variation of different rare earth metals for REAuAl4Ge2. ..... 35

Structure of CeAuAI4Ge2 viewed down the [010] direction. Large empty
circles: Ce; black dots: A]; lighter shaded circles: Au; darker shaded
circles: Ge. ................................................................................................. 36

Structure of CeAuAl4Ge2. (A) Trigonal pattern of Ce atoms in ab plane.
B) Coordination environment of Ce atom. ................................................ 37

Structure of CeAuAl4Ge2. (A) Coordination environment of Au atom.
B) Coordination environment of Ge atom. ................................................ 38

Structure of EuAuAl4(AuxGe1-x)2 viewed down the [010] direction. Large
empty circles: Eu; black dots: A]; lighter shaded circles: Au; darker
shaded circles: Ge/Au. ............................................................................... 40

Structure of EuAuAl4(AuxGel-x)2. (A) The square net formed by Eu atoms
in ab plane. (B) (C) Local coordination environments of Eu and Au(1)
atoms. ......................................................................................................... 41

(A) Temperature dependent magnetic susceptibility of CeAuAl4Ge2. (B)
Field dependent magnetization at 3K for CeAuAhGez. (C) Temperature
dependent magnetic susceptibility of EuAuA14Ge2. (D) Field dependent
magnetization at 3K for EuAuAl4Gez. (E) Temperature dependent
magnetic susceptibility of DyAuAl4Ge2. (F) Field dependent
magnetization at 3K for DyAuAl4Ge2. ...................................................... 46

XV

Figure I

Figure 3

L”.
'5
”1
('9
7’)

 

 

Figure 2-9.

Figure 3-1.

Figure 3-2.

Figure 3-3.

Figure 3-4.

Figure 3-5.

Figure 3-6.

Figure 3-7.

Figure 4-1.

Figure 4-2.

Figure 4-3.

(A) Temperature dependent magnetic susceptibility of CeAuA14(AuxGe1-
x)2. (B) Field dependent magnetization at 3K for CeAuAl4(AuxGe1-x)2. C)
Temperature dependent magnetic susceptibility of EuAuAl4(AuxGe.-x)2.
(D) Field dependent magnetization at 3K for EuAuAl4(AuxGe;-x)2. ......... 47

SEM image of a typical crystal of V2A15Ge5. ............................................ 55

Structure of V2A15Ge5 in polyhedra of V atoms viewed down the [100]
direction. Large empty circles: Ge; black circles: A1; gray circles: V. ..... 62

Structure of V2A15Ge5 viewed down the a-axis with all Al atoms omitted
for clarity .................................................................................................... 63

Structure of VzAlsGes. (A) Pentagonal column composed of Al and Ge
atoms with V-chaining sitting in the center. (B) — (G) Coordination
environments of Ge and A1 atoms. ............................................................ 64

Temperature dependent magnetic behavior of V2A15Ge5 at 1000 G. ........ 68

Thermal gravimetric analysis of V2A15Ge5 under air. ............................... 68

Density functional theory calculated for V2A15Ge5. .................................. 69

The structure of C019A145Si10.x (x = 0.13) in polyhedral representation
viewed down the b-axis. ............................................................................ 85

Structure of ColgAl45Silo.x (x = 0.13): framework composed of Si(3) and
Si(6)-centered slabs .................................................................................... 86

Structure of C019A1458110-x (x = 0.13): (A) Two Si(3)-centered polyhedra
joined via bridging Co(9) atoms to form a dimer. Si(3)-Si(3) distance:
2.5336 (12) A. (B) The connectivity mode of Si(6)—centered dimers. (C)
Extended organization of Si(6)-centered polyhedra forming a strongly
correlated layer ........................................................................................... 87

xvi

Figure

Figure -

Figure 4

Fleure 4

Figure 4-

Figure 44

Figure 4-1

FlgUre 4-1

Figuleili

 

 

Fi511155.).

 

Figure 4—4.

Figure 4-5.

Figure 4-6.

Figure 4-7.

Figure 4-8.

Figure 4-9.

Figure 4-10.

Figure 4-11.

Figure 4-12.

Figure 5-1-1.

Structure of ColgAl45Silo.x (x = 0.13): the polyhedral view of Co(3), Co(8)
and Co(9) atoms connected by Si-based wire framework. ........................ 88

Structure of ColeAl458im.x (x = 0.13). (A) The layered sub-structure of
Co(3), Co(8) and Co(9) polyhedra viewed down the c-axis. (B) The
connectivity mode of Co(3)-centered clusters. (C) The connectivity mode
of Co(9)-centered polyhedra. (D) The coordination environment of Co(8)
atom ............................................................................................................ 89

Structure of CosAIMSi2 in polyhedral representation down the a-axis. ..... 93

Structure Of C05A114Si2: the layered sub-structure of Co(2) and Co(3)
polyhedra along the b-axis. ........................................................................ 94

Structure of CoSAIMSiZ; (A) Structure of the Co(2)-based dimeric

fragment. (B) Structure of the Co(3)-based dimeric fragment. (C) The
connectivity mode of Co(2) and Co(5) polyhedra along the a-axis ........... 95

Structure of CoSAlMSiZ; the layered sub-structure of Al(l 1) and Al(15)
polyhedra in CoSAIMSi2 along the b—axis ................................................... 96

Structure of CoSAlMSiz; (A) The linkage of Al(15)-based polyhedra along

the a axis. (B) Structure of the Al(11)-based dimer. (C) The fragment
composed of Al(l) and Al(l 1) dimer ......................................................... 97

Coordination environments of (A) Co(3), Co(8), and Co(10) in
C019A1458i10-x (x = 0.13). (B) Co(3), Co(2) and Al(l l) in CoSAlMSiz. (C)
Selected transition metals in MnAl6, a—FeAlSi and o-C04A113. ................ 98

(A) Thermal gravimetric analysis of CoSAlMSi2 in air. (B) Scanning
electron micrograph of C05A114Si2 before annealing. (C) (D) Close up
view of C05A114Si2 before and aﬁer annealing at 1000 °C for 5 hours. ..100

SEM images of typical crystals of Sm3Ni5A119, Er3N15A119 and Yb3Ni5A119
grown from Al ﬂux. ................................................................................. 109

xvii

Figure 3-

Figure 5-.

FlgUre 5_1

Fig—“re 5-1.

Figum 5.1-

 

Figure 5-1-2.

Figure 5-1-3.

Figure 5-1-4.

Figure 5-1-5.

Figure 5-1-6.

Figure 5-1-7.

Figure 5-1-8.

Figure 5-1-9.

(A) Powder X-ray diffraction pattern of Sm3Ni5A119 compared with the
calculated pattern. B) Powder X-ray diffraction pattern of a single crystal
of Sm3Ni5A119 compared with the calculated pattern. ............................. 121

(A) Structure of REzNi4A115 (hypothetical) along the b-axis. (B) Structure
of Y4Ni6Alz3 along the b-axis. (C) Structure of Gd3Ni5A119 along the a-
axis. (D) Structure of YNiAl4 along the a-axis. Large empty circles: rare
earth element; medium gray circles: Ni; black circles: A1 ....................... 123

The structure of Yb3N15A119 in polyhedral view down the a-axis. .......... 126

Coordination environments of Ni atoms (within 3.0 A) and Yb atoms
(within 4.0 A) for Yb3N15A119. ................................................................. 127

(A) Temperature dependent molar magnetic susceptibility and inverse
magnetic susceptibility of Sm3Ni5A119. The susceptibility was measured
with an applied magnetic ﬁeld 1000 Gauss. (B) Field dependent
magnetization for Sm3Ni5A119 at 5 K. Single crystal sample was oriented
with a-axis perpendicular to the external magnetic ﬁeld. ........................ 131

(A) Temperature dependent molar magnetic susceptibility and inverse
magnetic susceptibility of Sm3Ni5A119. The susceptibility was measured
with an applied magnetic ﬁeld 1000 Gauss; (B) Field dependent
magnetization for Sm3Ni5A119 at 5 K. Single crystal sample was oriented
with a-axis parallel to the external magnetic ﬁeld. .................................. 132

A) Temperature dependent of the electric resistivity for a polycrystalline
pellet of Sm3Ni5A119. B) Temperature dependent of the thermoelectric
power for a single crystal of Sm3Ni5A119. ................................................ 133

(A) Temperature dependent molar magnetic susceptibility and inverse
magnetic susceptibility of Er3Ni5A119. The susceptibility was measured
with an applied magnetic ﬁeld 1000 Gauss; (B) Field dependent
magnetization for Er3Ni5A119 at 5 K and 100 K ....................................... 134

xviii

Figure 5-

Figure 5-

Figure 5-.'

Figure 5%

Figure 5-3

FIgUre 5-2

FlgUre 5.3

Flgllre 6-1‘

Flgure 6-] ‘

 

Figure 5-1-10. (A) Temperature dependent molar magnetic susceptibility and inverse

Figure 5-2-1.

Figure 5-2-2.

Figure 5-2-3.

Figure 5-2-4.

Figure 5-2-5.

Figure 5-2-6.

Figure 6-1-1.

Figure 6-1-2.

magnetic susceptibility of Yb3Ni5A119. The susceptibility was measured
with an applied magnetic ﬁeld 1000 Gauss; (B) Field dependent
magnetization for Yb3Ni5A119 at 5 K. ...................................................... 135

Valence of binary (circles) and ternary (triangles) compounds in the Yb-
Ni-Al phase diagram. Closed symbols: trivalent Yb; open symbols: mixed
valent of Yb. Dotted line indicates the border between mixed-valent and
trivalent Yb. ............................................................................................. 142

SEM image of a typical crystal of YbHNi3Alg9 ...................................... 144

Crystal structure onb1_1Ni3Alg,9 viewed down the [010] direction. Large
empty circles: Yb; black circles: Al; gray circles: Ni. ............................. 151

Structure of YbHNi3Algg. (A) Triangular layer composed of Al(l) atoms.
(B) Triangular layer composed of Al(4), Al(5) and Al(6) atoms down the
c-axis. ....................................................................................................... 152

(A) Ordered RE-Al plane in ErNi3A19 viewed down the c-axis. (B)
Partially disordered RE-Al plane in Yb]_1Nl3AIg_9 viewed down the c-axis.
(C) Disordered RE-Al plane in DyNi3A19 viewed down the c-axis. (D) &
(E) Local coordination environments of Yb(1) and Ni atoms in
YbHNi3A139 ............................................................................................. 153

(A) Temperature dependent magnetic susceptibility of polycrystalline
sample of YbHNl3Algg under the ﬁeld of 1000 G. (B) Molar
magnetization of YbHNi3Algg in ﬁelds up to 55000 G measured at 5 K.
inset: derivative (6M/6B) as a function of ﬁeld. ...................................... 155

Cell parameters variations with x for Yb3Ni5-xCuxA119 and Yb3N15-xFexAl.9
(x = 0, 0.5, 1) at 173 K. (A) Cell volume; (B) a cell edge length; (C) b cell
edge length; (D) c cell edge length. ......................................................... 167

(A) Temperature dependent of the cell volume for Yb3Ni5-xTMxA119 (TM
= Mn, Cu, Fe; x = 0, 0.5). (B) Temperature dependent of cell volume for
Yb3Ni5-xTMxAllg (TM = Cu, Fe; x = 0, 1). .............................................. 168

xix

Figure 6-

Figure 6—

Figure 6-

Figure 6‘.

Figure 6-1

Flgm'e 6-:

 

 

Figure 6-2_

Figure 0-2-

Figure 6-1-3.

Figure 6-1-4.

Figure 6-1-5.

Figure 6-1-6.

Figure 6-1-7.

Figure 6-2-1.

Figure 6-2-2.

Figure 6-2-3.

Figure 6-2-4.

Figure 6-2-5.

(A) Temperature dependent of a cell edge for Yb3Ni5-xTMxA119 (TM =
Mn, Cu, Fe; x = 0, 0.5). (B) Temperature dependent of a cell edge for
Yb3Ni5-xTMxAl|9 (TM = Cu, Fe; x = 0, 1). .............................................. 169

(A) Temperature dependent of b cell edge for Yb3N15-xTMxA119 (TM =
Mn, Cu, Fe; x = 0, 0.5). (B) Temperature dependent of b cell edge for
Yb3N15-xTMxA119 (TM = Cu, Fe; X = 0, I). .............................................. 170

(A) Temperature dependent of c cell edge for Yb3Nis-xTMxAl;9 (TM =
Mn, Cu, Fe; x = 0, 0.5). (B) Temperature dependent of c cell edge for
Yb3Ni5-xTMxA119 (TM = Cu, Fe; x = 0, 1). .............................................. 171

Magnetic susceptibility of Yb3Ni5-xCuxA119 (0 _<_ x S 1) compounds as a
function of temperature from 3 K to 300 K. ............................................ 177

Magnetic susceptibility of Yb3Ni5-xFexAllg (0 S x S 1) compounds as a
ﬁanction of temperature from 3 K to 300 K. ............................................ 178

SEM image of a typical crystal of YbNi2-xFexAlg. .................................. 190

Crystal structure of YbNiz-xFexAlg in polyhedran of M(l) and M(2) atoms
viewed down the [001] direction. Large circles: Yb; black dots: Al; darker
shaded circles: M(l); lighter shaded circles: M(2). ................................. 194

Coordination environments of M(l), M(2) and Yb atoms for
YbNi2-xFexAlg. ......................................................................................... 1 95

(A) Temperature dependent magnetic susceptibility of polycrystalline
sample of YbNi2-xFexAlg under the ﬁeld of 1000 G. (B) Molar
magnetization of YbNi2_xFexAlg singles crystals oriented with [001] axis
perpendicular, parallel to the external magnetic ﬁeld and isotropic in ﬁelds
up to 55000 G ........................................................................................... 199

(A) Temperature dependent of the electrical resistivity for a
polycrystalline pellet of YbNi2-xFexA13. (B) Temperature dependent of the
thermoelectric power for a single crystal of YbNi2-,,FexAlg. ................... 200

XX

Figure 6-

Figure 7-

Figure 7

 

Flé’ure 7-15

Figure 6-2-6.

Figure 7-1.

Figure 7-2.

Figure 7-3.

Figure 7-4.

Figure 7-5.

Figure 7-6.

57Fe Mossbauer spectrum of YbNi2-xFexAlg at 80 K and RT ................... 202

Crystal structure of ngAthg viewed down the c-axis. Large empty
circles: Yb; black small circles: Al; gray medium circles: Sb. ................ 224

Structure of Yb9A13Sb9. (A) The fragment of [Alng9]18‘ ribbon. (B) ~ (F)
Coordination environments of Yb atoms. ................................................ 225

Crystal structure of Yb3Ale3 viewed along the b-axis. Large empty
circles: Yb; black small circles: Al; gray medium circles: Sb. ................ 226

Structure of Yb3Ale3. (A) Polymeric chain [Ale3]6' viewed down the a-
axis. (B) (C) (D) Immediate coordination environment of Yb atoms ...... 227

(A) Temperature dependent of the therrnopower for a pressed pellet of
Yb3Ale3 from 300 to 700 K. (B) Temperature dependent of the
therrnopower for a pressed pellet of Yb9A13Sb9 from 300 to 400 K ........ 229

(A) The total DOS of Yb3Ale3. The upper spin-split f-band (fm) is

located near the Fermi level. (B) The total DOS of Lu3Ale3. The ﬁlled f
band is located below —5 eV. (C) The total DOS of La3Ale3 with the
empty f-level located in the conduction band. (D) The total DOS of
CagAle3. A band gap of about 0.4 eV shows the semiconducting
character of the Ca compound. ................................................................ 230

xxi

RE

T.\l

C C D
SEM
EDS
XRD
SQL'ID

BN1

 

RE
TM
CCD
SEM

EDS

SQUID

BM

LIST OF ABBREVIATIONS

Rare Earth

Transition Metal

Charge Couple Device

Scanning Electron Microscope

Energy Dispersive Spectroscopy

X-ray Diffraction

Superconducting Quantum Interference Device

Bohr Magneton

xxii

 

l-l. Int

the imp
appheat
commur
paekagir

S
determin;
Intermeta
aroused 2
Strength 3
high temp
Camdidates
intermetall
ICSISIance.
Chinese a:

reSlOralch

 

“tinned

l
l

 

CHAPTER ONE

Introduction

1-1. Introduction to Intermetallics

Solid state chemistry has been growing rapidly since 19505 as scientists realized
the importance of solid state materials in advanced technology in a higher level. The
applications of these materials span a wide range of ﬁelds from microelectronics,
communication, information, energy conversion and storage, construction, electronic
packaging to domestic and advanced ceramics. '

Solid state chemistry is focused on the design, synthesis and structural
determination of new materials and characterization of their physical properties.
Intermetallics are one of the oldest class of materials in solid state chemistry and have
aroused great interest in materials science and technology due to their mechanical
strength and high temperature performance. For example, the oxidation resistance and
high temperature strength of iron and nickel aluminides make these materials attractive
candidates in industries including steel, chemicals and petroleum.2 Back to BC. 1600,
intermetallics already got extensive use due to their remarkable hardness and wearing
resistance. The bronze mirrors (e.g. composition Cu3ISn3) were used by the ancient
Chinese and Romans; the amalgam (Cu4Hg3, Anggg) found applications as dental
restorative materials in ancient Germany and China.3 Intensive scientiﬁc work on
intermetallics emerged in 19005, initiated by Tammann in Germany and Kurnakov in
Russia.4 Since then a wide variety of intermetallic compounds were discovered; their
crystal structure and properties including mechanical and magnetic properties were

studied. In 19505, more activities aroused investigating intermetallics as high

tempera
its exec
thermal
Conside
material
llldUSIl'l(
tremendl
Ordnanc
back to 1
memory
applicant
Cull-U ;
large farm
Which is ,
(Tl/51311 $1
intermem

0V
undefStanL

example 11

 

Lb Tm, 1

51pm,” d l

l
\

 

Intenn €1,311

dlSQOVeTed

 

temperature structural materials. One of the most extensively studied examples is MoSiz:
its excellent oxidation resistance, high melting point, relatively low density, and high
thermal conductivity found itself an attractive material in advanced engine applications.5
Considerable success had also been achieved on some transition metal aluminides. These
materials, such as TiAl, are of intense interest to the gas turbine and aircraft engine
industries, where their high temperature properties and low density offer prospects for
tremendous weight savings.6 In 1961, Nitinol, which stands for Nickel Titanium Naval
Ordnance Laboratory, discovered that some Ni-Ti alloys after being strained, could revert
back to their original shape at a certain temperature. These materials are called “shape
memory alloys”, and they have been applied to different ﬁelds such as aeronautical
applications and surgical tools. The most effective and widely used alloys are NiTi,
CuZnAl and CuAlNi.7 Another equally important discovery during this period was a
large family of superconducting binary phases with A15 structure type, 8 such as V3Si,
which is a superconductor with critical temperature as high as 17 K.9 Since then, the
crystal structures, electronic structures and superconducting properties of these binary
intermetallics were extensively studied. 10

Over the past twenty years, there have been considerable advances into the
understanding of the structure and behavior of intermetallic compounds. A more recent
example is exhibited by the superconducting quaternary intermetallic LnNi2B2C (Ln =
Lu, Tm, Er, Ho, Y) with critical temperature Tc up to 16.6 K.11 The discovery of
superconducting LnNizBZC has motivated intensive experimental work on these complex
intermetallic systems. In 2001, Mng, a simple binary intermetallic compound

discovered half an century ago,12 was found to be a superconductor, with a surprising

 

high cn
intermet
some kr
of exten
fermion
highly Ct

intermet;

1-2. C ryg

A
that C0012
C0nstitutll
cTYSIal Va
form allo:
Solutions.
They do n
01) a main
POPUlar St

SOIUIIOns,

 

constlmllo;
\valenCe elL‘
lhat many

VllllCh the C

 

 

high critical temperature — 39 K. 13 This discovery stimulated a great interest in
intermetallics, not only the development of new materials, but also the reinvestigation of
some known compounds. Rare earth-containing intermetallics have become the subject
of extensive study due to their special behaviors such as the valence ﬂuctuations, heavy
fermion and Kondo-lattice behavior, which originate from the interaction between the
highly correlated f electrons and delocalized conduction electrons.l4 Such phenomena of

intermetallics will be discussed in detail in the following context.

1-2. Crystal Structure and Basic Properties of Intermetallics

According to the deﬁnition from Schulze, intermetallics are a branch of alloys
that contain two or more metals resulting in crystal structures different from those of the
constitutional elements.15 Under certain conditions, one metal diffuses into another via
crystal vacancies made available by defects, mechanical stress or grain boundaries to
form alloys. Alloys may be homogeneous solid solutions or intermetallics. In solid
solutions, the atoms of one metal are distributed randomly among the atoms of the other.
They do not have speciﬁc chemical formula, and the best way to describe them is based
on a main metal with a certain percentage of other metals added in. For example, a
popular stainless steel, has the composition Fe-18%Cr-8%Ni. Different from solid
solutions, intermetallics have deﬁnite crystal structures and speciﬁc formula. While
constitutional metals are different in atomic size or electronegativity or the number of
valence electrons, intermetallic compounds tend to form instead of solid solutions. Given
that many intermetallics are formed due to the size effect or are electron compounds (in

which the crystal structure types are related to particular valence electron concentrations),

 

they do
surround
ionic bor
contribut

lr
crystal st
can be 0'!
resistanCt
surface p;
makes pr

intermetai

H.313";

Th
group Eler
°C, Fe 15.

the Solid n

 

to Ol'C‘rcor

wElder all

 

 

they do not follow the valency rule. In intermetallics, atoms are preferentially
surrounded by different kinds of atoms; not only metallic bonding, covalent and even
ionic bonding can also be found. It is the strong bonding between different atoms that
contributes to the excellent mechanical properties of intermetallic materials.
Intermetallics are comprised of a huge variety of compounds that differ greatly in
crystal structure, bonding, and properties. However there are some basic properties that
can be observed more or less in intermetallics: high melting point, high hardness, wear
resistance and good oxidation resistance. That is why many have found applications as
surface protecting materials. On the other hand, intermetallics are brittle materials, which
makes processing difﬁcult and it is one of the main challenges for the application of

intermetallics in metallurgy.

1-3. Manipulation of Metal Flux to Synthesize Intermetallics

The melting points of most transition metals, rare earth metals and some main
group elements are typically over 1000 °C. For example, the melting point of Ni is 1455
°C, Fe 1538 °C, Nd 1024 °C, Si 1414 °C. In order to achieve enough diffusion between
the solid metals, synthesis of intermetallics usually requires high temperature techniques
to overcome the energy barrier. Conventionally, high-power equipment such as arc-
welder and radio-frequency reduction ﬁamace has been used for the synthesis of
intermetallic compounds.

In the past two decades, a new method to synthesize intermetallics, which is
called combustion synthesis, has attracted considerable attention due to its unique

characteristics.'6 In this process, the metal powders are mixed and pressed in the form of

cyhndr
the C01
high It
throud

»

huennt

lunuau

innm3:

211010 31

occurs 1

ﬁnned.

duennui
9n§k3c7
bOUUdarlt
St

Single cf}.
propmles
Budgnan

Homer.

   
  

 

and Ihe 10‘
Such (TIC A

lmﬁmﬂMi

 

cylindrical pellets. The pellets are ignited through an electrical coil or a laser beam. If
the combinations of thermochemical and therrnophysical properties are appropriate, a
high temperature reaction front is initiated (1500-3500 °C) which then propagates
through the reactants. ‘7 However, the reactions may not be completed with various
intermediate phases; meanwhile the products formed are generally porous. '8

For the high temperature techniques discussed above, there are two important
limitations. Firstly, thermodynamically stable phases, which are usually simple known
binary compounds, tend to form from these reactions. These phases are too stable to
avoid and they hinder the formation of kinetically stable phases. Secondly, the reaction
occurs too fast to allow single crystals to grow, so polycrystalline products are often
formed. Obviously, the polycrystalline form of the materials limits their structural
determinations. In addition, polycrystalline samples do not have the other chief merits of
single crystals, including uniformity of composition, anisotropy, and the absence of
boundaries between individual grains.

Sometimes extended annealing at high temperatures facilitates the growth of
single crystals. However even then the crystals may not be large enough for desired
properties measurements. Some other methods, such as ﬂoating-Zone, Czochralski and
Bridgman growth methods are often employed to obtain single crystals of intermetallics.
However, these techniques are suitable only when the material is congruently melting;
and the temperature needed to melt the sample may also be inconveniently high. Under

such circumstances other techniques are necessary to grow single crystals of

intermetallics.

at relat
tempera
choosin.
the mel
reaction:
people 1
Crystals
2800 CC
Crystals t
intensive
fr0m sew
FL"
the 6611111
Iempefatu
Other C001
5h0uld 11(1
fOmlation

chemiCal l

 

Periodic 1,.

ﬂutes Ila:

 

Molten metal ﬂux synthesis is such a technique that single crystals can be grown
at relatively low temperatures. An excess amount of a metal is used as a high
temperature solvent in which the crystals grow freely in a non-constrained fashion. By
choosing'an appropriate metal, we can run the reaction at a temperature much lower than
the melting point of the constitutional metals. As is well known, low temperature
reactions are likely to produce new kinetically stable compounds. Half a century ago,
people found that this metal ﬂux method was particularly useful to synthesize single
crystals of borides and carbides due to their extremely high melting points (e.g. TiBz
2800 °C, TaCQgg 3983 °C).'9 Rowcliffe and Warren were able to grow 2mm single
crystals of TaCogé by combining Ta and C in iron solutionzo; SiC, one of the most
intensively studied crystals with potential applications in many ﬁelds, has been prepared
from several metal solutions such as Si/Fe, Si/Ag and Si/Ni.2|

For a metal to be suitable as a ﬂux, several key conditions have to be met. First,
the candidate metal should melt at reasonably low temperature; obviously high
temperatures are not convenient with respect to the instrument required. Second, the
other constituents have to be considerably soluble in this metal solvent; and the solvent
should not react with the solute to form stable binary phases, which might prevent the
formation of the desired product. Finally the excess metal can be easily removed by
chemical or physical methods. Based on the criteria mentioned above, inspection of the
periodic table reveals that the group III elements (Al, Ga, In), Sn, Pb, Bi and Zn appear to
be excellent candidates for the high temperature solvents. Indeed, Al, Ga, In, Sn and Zn

ﬂuxes have been actively studied by several research groups including the Kanatzidis

 

group.

anda;

1—4. I

they ;
proper
alumir
the C0;
Candid;
point (6
point ar
range 01
Immedia

t3l€rtlents

cmClblC 0
above 661

Which atta

distinct fr,

 

 

 

group, the Kauzlarich group, the Chan group, the J eitschko group and the Péittgen group,

and a great progress has been made to synthesize new intermetallic compounds.22'23’24'25

1-4. Rationale for Al Flux Synthesis
Aluminides may be one of the most common groups of intermetallic compounds;
they are considered as emerging materials due to their high speciﬁc mechanical

26 An alternative way to synthesize intermetallic aluminides is utilizing

properties.
aluminum as a metal ﬂux, since aluminum tends to serve as a reactive solvent. Based on
the conditions for metal ﬂux discussed above, aluminum appears to be an attractive
candidate for high temperature solvent. First, aluminum has reasonably low melting
point (660 °C) and very high boiling point (2792 °C). The difference between its boiling
point and melting point is almost 2000 °C, so Al can stay in its liquid form over a wide
range of temperatures. Second, inspection of the Al-containing alloy phase diagram
immediately reveals that many transition metals such as Cu, Fe and Ni, and main group
elements such as Si, Ge and Sb are considerably soluble in liquid aluminum. Third, we
can easily ﬁnd appropriate containers for aluminum ﬂux reactions, such as alumina
crucible or graphite tube. Fourth, excess aluminum can be removed either via spin-off
above 660 °C or etching away by simply reacting it with a base solution (e.g. NaOH)
which attacks the desired product at a much slower rate. Finally, what makes aluminum
distinct from the other metal ﬂux candidates is its reducing ability which has been seen
not only from the well-known therrnite reaction in which F6203 is reduced to Fe by Al,

but also from reducing more complicated oxides such as perovskites MTiO; (M = Ca, Sr,

Ba) to intermetallic compounds including M3Au6+xA126Ti27

 

pmpertle
0f lllC‘ We

solvent.

Optimum
point oflrl

A
SI'Slems l
ellVll'Onm
discovere
eliample,
cOmlmsin
Crystal 511'
Ce. Pr, Sn
Nd

 

r Sm‘ (
GdRezAl;

SITUCture; ?
as another

 

Aluminum has been known as an effective solvent for the synthesis of binary and
ternary intermetallics for a few decades. Back to 19705, some Japanese groups started to
utilize aluminum ﬂux to synthesize single crystals of binary boron-rich compounds, such
as LaB6, 28 VB2, 29 Cr3B4 30 and TB; 3 l . Single crystal X-ray diffraction, transport
properties and oxidation resistance were studied on these compounds. Interestingly, most
of the work focused on the early transition metals and Al was serving as a non-reactive
solvent. A few ternary examples were also reported, such as WAlB32 and YbAlB4.33 The

optimum temperature for these reactions is about 1500 °C, much lower than the melting

point of boron which is 2300 °C.
About one decade ago, Jeitschko and his coworkers started to examine the
systems RE/TM/Al (RE = rare earth metal, TM = transition metal) in liquid aluminum

4 . . .
3 Series of new ternary famrlres were

environment and had made a great success.
discovered, which suggests a rich chemistry provided by liquid Al in these systems. For
example, the family RETMzAllo (TM = Mn, Fe, Ru, Os and Re) has a homogeneous
composition, while different reaction conditions led to the discovery of four different
crystal structure types. REanAIm (RE = Y, La-Nd, Sm, Gd-Dy) and RERezAllo (RE =
Ce, Pr, Sm) crystallize in tetragonal CaCrzAllo structure type;35 RETM2A1|0(RE = Y, La-
Nd, Sm, Gd-Lu and TM = Fe, Ru, 05) have the orthorhombic YbFezAllo structure type;36
GdRezAllo and TbRezAllo, crystallize with a stacking variant of the YbFezAllo type
structure?7 RERezAlio (RE = Ho-Lu) form in a new structure type and may be considered

as another variant of YbF e2A110.3 5“ It is noteworthy that Al tends to be a reactive solvent

in these systems and almost all the compounds are Al-rich.

RETM
and ham
line of
intermet.
These 5)
soluble i
and Al-C
never th
Compoun
the cry-51
lﬂCOrpOra
Prepenje,

 

 

Recently our group initiated exploratory studies on the systems RE/Al/Tr and
RE/T M/Al/T r (RE = rare earth metal, TM = ﬁrst row late transition metal, Tr = Si, Ge),
and have achieved a great success in this ﬁeld.38 Although Si and Ge are on the border
line of metals and nonmetals, silicides and gerrnanides are usually regarded as
intermetallics because of their similar characteristics and generally metallic properties.
These systems are of particular interest for several reasons. First, both Si and Ge are
soluble in liquid Al, and they do not form stable binary compounds according to Al-Si
and Al-Ge alloy phase diagram. Second, the quaternary system RE/T M/Al/T r, which has
never been investigated before, imposes challenges to the aluminum ﬂux. If quaternary
compounds do form, they would be of fundamental interest in solid state chemistry from
the crystal structures as well as physical properties point of view. Third, the
incorporation of a tetrelide especially Si into the system, may contribute to the refractory
properties of the material. Silicides are well known for their high hardness, high

”‘40 with MoSiz as the most famous example.”

chemical stability and oxidation resistance
42 Thus they can be used in technological applications such as high temperature wear and
corrosion resistant coatings and thermoelectric energy conversion.43‘44’4S Finally, the rare
earth-containing intermetallics display various interesting physical properties such as
mixed valency, 46 Kondo lattice behavior,47 spin ﬂuctuations 48 and heavy fermion 49,
which are strongly correlated to the electronic structure and position of the 4f electron
level with respect to the Fermi energy.

It was observed that Al can serve as a reactive solvent. It can dissolve Si (or Ge)

along with other metals to produce new quaternary intermerallic compounds such as

5rnzNi(Ni,.sh.,.)Al..s16 (x = 0.18-0.27),38 YNiA14Ge2,38 RE4Fe2+xAl7-xSig (Re = Ce, Pr,

Nd, Sm
(RE = C.
unfamill
design 1:
and relat-
the ﬂux .

I
and seen
metals.
aluminur:
exPlOratu
resulted i

' 1.x
SeneS‘T .,

 

hexagma
Stmcmrg
Ge Instem
exhibiting
Stable bin.
33'5th I
bent;Een S
eXampIe‘ I
Ge) Can be

metal (0 1r

Nd, 5m),50 REgRuleh9819(Aleilz-x) (RE = Pr, Nd, Sm, Gd, Tb, hr),38 and REZMA16814
(RE = Gd, Tb, Dy; M = Au, Pt).5 1 Compounds with such complicated formulas are very
unfamiliar to solid state chemists and obviously it is almost impossible to predict or
design these compounds. However their single crystal form can be obtained from Al ﬂux
and relatively easily characterized. In this sense, exploratory synthesis employing Al as
the ﬂux is a great way to discover novel intermetallic materials.

The achievements of exploratory synthesis in liquid aluminum including the ﬁrst
and second row transition metals stimulated us to investigate the third row transition
metals. Gold is of particular interest not only because it is very reactive in liquid
aluminum, but gold-aluminum alloys have found applications in electronic devices.52 The
exploratory synthesis comprising of a rare earth metal, gold and silicon in Al ﬂux
resulted in a number of new phases such as the Th2(AuxSi1-x)[AuAlz]nSiz homologous
series,38 REAu4A13Si, 53 REzAuAl(r,Si444 and REAU3A17.54 These compounds feature
hexagonal antiﬂuorite-type slabs which can be regarded as fragments of the bulk AuAlz
structure. In the present work, the study of the Au system in liquid Al was extended to
Ge instead of Si. Ge is similar as Si in many aspects: they belong to the same group
exhibiting similar electronic structure; both are soluble in liquid Al and do not form
stable binary phases with Al. Bradley Sieve, a former group member, had studied the
systems RE/TM/Al/Si(Ge) systematically, and he indeed found parallel chemistry
between Si and Ge with both systems tending to form quaternary compounds. For
example, RE2-xTMAl4Tr2(All-yTry)(A11-zTrz)2 (RE = Sm, Dy, Er; TM = Ni, Co; Tr = Si,
Ge) can be formed with both Si and Ge,” when the reaction ratio between the rare earth

metal to transition metal is smaller than 1. The RE-rich family RE2N1(NixSi.-x)Al4Si6

10

 

(RE = I
SmgCot
such as
Nd. Sm;
By stut;
informat
Studies 1‘.
metals 11
the syste
and 1b)
CrEsta] .‘
dlSSEﬂaII
Si
immela
53‘8th 71
Ofmaleri;
of theSe h

thermoc

lc

1” Spite .,
Infomatll

hand. SAUL.

I'
TegjOn 0f

 

(RE = Pr, Nd, Sm, Gd, Dy, Tb) was found as well as their isostructral Ge analogue
Sm2C0(Co,,Al1.,,)A14Ge6.y.55 When the transition metal was changed to the earlier ones
such as Mn and Fe, the results led to the discovery of RE4TM2+XA17-xSig(RE = Ce, Pr,
Nd, Sm; TM = Mn, Fe),50 while isostructural Ge-analogues have not been found so far.
By studying the system RE/Au/Al/Ge systematically, we expect to obtain more
information on the reaction pattern of A1 with tetrelides (Si, Ge); on the other hand, these
studies may help us to compare the chemistry between ﬁrst row and third row transition
metals in these systems. Chapter 2 presents two new quaternary families discovered in
the system RE/Au/Al/Ge: REAuAl4Ge2 (RE = Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Er, Tm
and Yb) and REAuAl4(AuxGe1-x)2 (x = 0.4) (RE = Ce and Eu). The synthesis, single
crystal X-ray diffraction and magnetic measurements will be described in this
dissertation.

So far most of the work described here has focused on rare earth-containing
intermetallics because of the interesting physical phenomena described previously. The
systems TM-Al-Si(Ge) (TM = transition metal) are also of great importance in the ﬁeld
of material science and metallurgy. Extensive research has been done on the applications
of these ternary materials as high temperature structural materials,56 protective coatings,57
thermoelectric power conversion,58 and soft magnetic thin ﬁlms and magnetic sheets.59
In spite of the broad applications of these materials, there is signiﬁcant lack of
information on the crystal structure and phase equilibrium of these phases. On the other
hand, since the discovery of an icosahedral phase in a rapidly quenched Al-Mn alloy,60 a
large number of quasicrystalline approximants have been found to occur in the Al-rich

region of the Al-TM and related ternary systems. or-AlsoMnnSi7 is a well-known

11

protot}

   
      

hllllaAL
unique 1
in half .
int'estig.

IUIOW I CC

eleleOni.
interest“.
and 06g,
electron é
than thos:

dimCult 1

metal.

 

prototype of an approximant crystalline phase presenting Mackay-type clusters. 6'

Mn14A156+,,Ge3.x (x = 0 - 0.6) was recently reported by Wu and Seo, which shows a
unique partially destroyed Mackay icosahedra that retain the icosahedral symmetry only
in half of the individual polyhedra.62 In this work Chapter 3 presents the results of
investigations on the ternary system V/Al/Ge using Al as the ﬂux. To the best of our
knowledge, past explorations of this system using high temperature direct combination
techniques only yielded pseudo-binary phases V3Al(1_x)Gex which belong to the A15

63 We obtained a new ternary phase V2A15Ge5 ﬁom liquid aluminum

structure type.
which crystallizes in a new structure type. Chapter 4 explores the system Co/Al/Si by
using molten A1 as a ﬂux. German has made a tentative study on this system, and found
ﬁve phases though only two of them were structurally characterized.“ In our own study
of this system in liquid Al, we found two complex phases in the Al-rich comer of this
ternary system — ColgAlrssilo.x (x = 0.13) and C05A114Siz. In Chapter 4, the crystal
structures and physicochemical properties of these two phases are described.

Chapters 5 and 6 include studies on the Yb-containing intermetallics. Due to the
electronic transition between Yb2+ and Yb3+ ions, Yb-containing compounds can exhibit
interesting behaviors such as mixed valence,65 Kondo-lattice behavior,66 heavy fermion67
and negative thermal expansion.68 Although Yb (4ﬂ3—4ﬂ4) can be regarded as hole-
electron analogue of Ce (4fl-4f0), the studies of Ce compounds are far more extensive
than those of Yb compounds. One of the main reasons might be that it is relatively more

difﬁcult to synthesize Yb-containing intermetallics due to the high vapor pressure of Yb

metal. Our studies in the system Yb/Ni/Al in liquid Al not only led to the discovery of

12

temary

modi fy

V CICUR
alkaline
or a grit
family,

discover
Alel.‘ 55
the 14-
Euithlnl
The sir.
meagum.
the Yb ,1

Phases;

 

5101chl’0r
ph}SlCa]

antImOnl

 

1~S.E_\.p

A

 

 

 

 

ternary phases such as YbHNi3Algg and Yb3Ni5A119, we also found we might be able to
modify the magnetic states of Yb ions in Yb3Ni5A119 by applying chemical pressure.
Chapter 6 deals with the work when we extended the tetrelides (Si, Ge) to Group
V element —— Sb. There has been increasing interest on the Zintl phases composed of an
alkaline earth metal (Ca, Sr, Ba) or rare earth metal (Eu, Yb), a transition metal (Mn, Zn)
or a group 13 element (A1, Ga, In) and a pnicogen element (P, As, Sb, Bi). In the 9-4—9
family, Agzn4+xPn9 (x = 0 for A: Yb, Pn = Bi, x = 0.5 for A = Ca & Yb, Pn = Sb), the
discovery of mixed valence of Yb in ngZn4Big 69 and extra Zn component in
Agzm58b970 opens the opportunity to reexamine the parental compound Cagszig. In
the 14-1-11 family, Eu14MnSbH shows colossal magnetoresistance effects; 7' while
EuMMnBi“ orders antiferromagnetically and shows a large negative magnetoresistance.72
The single crystals used for structural characterization and physical properties
measurements in most of these systems, were obtained from Sn ﬂux. Here our studies on
the Yb/Al/Sb system using Al as the ﬂux led to the discovery of two new valence-precise
phases: ngAthg and Yb3Ale3. Although these two compounds have the same
stoichiometry, they belong to different structure types and show different chemical and

physical behaviors. Our work indicates that Al is a reactive ﬂux in the synthesis of

antimonides and this study might be extended to the quaternary systems RE/TM/Al/Sb.

1-5. Experimental Techniques
All the reagents including rare earth metal, transition metal, aluminum, tetrelides
and antimony were stored and handled in a nitrogen-ﬁlled dry box. Usually a ten-fold

excess amount of Al was used. Since the rare earth metal and AI will attack the silica

l3

tube at
The mi
(13mm

10“ Tor

L

'l‘cmpcrature

FTgUre 1

for Al 1‘;
higher ”(I
enSure h‘

l“asfﬂ011

0

CA Du:

 

COlll

Aﬁer the

 

 

tube at high temperatures, an inert container such as alumina crucible had to be used.
The mixtures were combined into alumina crucibles which were put into silica tubes
(13mm in diameter). The silica ampoules were then ﬂame-sealed under the vacuum of

10'4 Torr to avoid oxygen contamination.

 

 

 

 

 

 

1000°C 36h
36h

93 T 15h

‘3 500°C
53

D.

E

0.)

{-4

_* Time

Figure 1-1. A typical temperature heating proﬁle used for Al ﬂux synthesis.

The samples were then subjected to heat treatment. A typical heating proﬁle used
for Al ﬂux synthesis is shown in Figure 1-1. The system was brought to a temperature
higher than the melting point of Al metal and kept at this temperature for some time to
ensure homogeneous melting of the constituents in liquid A1. Subsequently the system
was slowly cooled to a temperature lower than the melting point of Al metal (here 500
°C). During the slow-cooling stage, the solution became supersaturated so that crystals
could grow out. Finally the system was quickly brought down to room temperature.
After the reaction, the products were usually embedded in a Al matrix. To get rid of

excess Al, either a physical or chemical method can be utilized. Since Al melts at 660

14

°C, the

Howex t
melting
impract:
convent

tempera

And the
SOlution
tO Stand

IO be [TIC

 

°C, the excess Al can be removed via centrifuge at temperatures higher than 660 °C.
However this method is more applicable to Ga, In and Sn ﬂuxes which have much lower
melting point, since taking the samples out at as high as 700 °C is dangerous and
impracticable. On the contrary, the chemical way to remove Al is much easier and more
convenient: the product was treated with 5M NaOH solution overnight at room

temperature. During this process, Al reacted with NaOH solution by this reaction:

2A1 + 2NaOH + 2H,o —> 2NaAlo2 + 3H,?

And the products were attacked by NaOH solution at a much slower rate. Dilute HCl
solution is another way to get rid of excess Al; however some crystals are too vulnerable
to stand acid solution, therefore for isolation of Al ﬂux reactions, NaOH solution seems

to be the best choice. Finally the products were rinsed and dried by acetone and ether.

15

Referen

 

I A) Sur
b) Cher.-
Rt’t'. 201
Metal]. .
.llicroczt

2 Sauthti
3 Sauthv

4
Gerhar
SONS. In.

53) FIIZL
Berezhn.
LA’I‘ajn ‘

6 ..
\ lSWari
A.2002.

‘1

I a) 12111],

C) “.3375

8A15 is
Pm‘3n. 1.

9
H al’ d E,

‘0 a) N'lor
TeSIaTdL I

RehWald

ll
CaVa‘ l

u

JOHeS‘

 

References:

 

' A) Surdoval, W. A.; Singhal, S. C.; McVay, G. L. P. Electrochem. Soc. 2001, 16, 53.
b) Chen, G.; Dresselhaus, M. S.; Dresselhaus, G.; Fleurial, J .-P.; Caillat, T. Int. Mater.
Rev. 2003, 48, 45. c) Datta, S. K.; Tewari, S. N.; Gatica, J. E.; Shih, W.; Bentsen, L.
Metall. Mater. Trans. A 1999, 30A, 175. d) Cadenhead, R. L.; Frazier, A. B. Int. J.
Microcircuits Electron. Packaging 1995, I8, 51.

2 Sauthoff, G. Intermetallics 2000, 8, 1101.
3 Sauthoff, G. Intermetallics, VCH Verlagsgesellschaft, Weinheim 1995.

4 Gerhard Sauthoff in “Ullmann's Encyclopedia of Industrial Chemistry”, John Wiley &
Sons, Inc., 7‘“, 2005.

5 a) Fitzer, E. Plansee Proc. 1956, 56-79. b) Krockel, O. Silikattechnik 1960, II, 108. c)
Berezhnoi, A. 8.; Repenko, K. N.; German, I. A.; Gul'ko, N. V. Sbornik Nauch. T rudov
Ukrain. Nauch.-Issledovatel. Inst. Ogneuporov 1960, 4, 296. d) Fitzer, F.; Schlichting,
J .; Schmidt, F. K. High Temperatures - High Pressures 1970, 2, 553.

6 Viswanathan, G. B.; Karthikeyan, S.; Hayes, R. W.; Mills, M. .1. Metal]. Mater. Trans.
A, 2002, 32A, 329.

7 a) Izuno, S. J. Electron. Eng. 1983, 20, 89. b) Huang, W. Mater. Design 2002, 23, 11.
c) Wayrnan, C. M. J. Met. 1980, 32, 129.

8 A15 is the Strukturbericht designation for the Cr3Si structure type with the space group
Pm-3n. Other members with this structure: Nb3Al, CF30, Ti3Sb, Ti3Au.

9 Hardy, G. F.; Hulm, J. K. Phys. Rev. 1953, 89, 884.

'0 a) Morin, F. J.; Maita, J. P. Phys. Rev. 1963, 129, 1115. b) Dynes, R. C.; Poate, J. M.;
Testardi, L. R.; Storm, A. R.; Hammond, R. H. IEEE T. Magn. 1977, 13, 640. c)
Rehwald, W.; Rayl, M.; Cohen, R. W.; Cody, G. D. Phys. Rev. B: Solid St. 1972, 6, 363.

H Cava, R. J .; Takagi, H.; Zandbergen, H. W.; Krajewski, J. J .; Peck, W. F.; Siegrist, T.;
Batlogg, B.; Van Dover, R. B.; Felder, R. J. Nature 1994, 367, 252.

‘2 Jones, M.E.; Marsh, R.E. J. Am. Chem. Soc. 1954, 76, 1434.

'3 Nagamatsu, J .; Nakagawa, N; Muranaka, T.; Zenitani, Y.; Akimitsu, J. Nature 2001,
410, 63.

16

 

I
" a) 21.
M.; Sci.
SchonhI
Takaba:
lS

Schul

168)H0
Pro g. .1]

l"Mum:

m .
Ohvc

19 _
Gunn

Komuko
1995.

31:1 Rout:
2Iaillal
b) PlCh’a:

Boston 1

2 Millie,
State C11.

23
Canﬁc

 

 

'4 a) Zlatic, V.; Freericks, J. K. Acta Physica Polonica, B 2001, 32, 3253. b) Getzlaff,
M.; Schmied, B.; Wilhelm, M.; Kubler, U.; Fecher, G. H.; Bansmann, J.; Lu, L.;
Schonhense, G. Phys. Rev. B 1998, 58, 9670. c) Hossain, Z.; Harnashima, S.; Umeo, K.;
Takabatake, T.; Geibel, C.; Steglich, F. Phys. Rev. B 2000, 62, 8950.

'5 Schulze, G. 13. R. Metallphysik Akademie-Verlag: Berlin, 1967.

16 a) Holt, J. B.; Munir, Z. A. J. Mater. Sci. 1986, 21, 251. b) Moore, J. J.; Feng, H. J.
Prog. Mater. Sci. 1995, 39, 275.

'7 Munir, z. A. Ceram. Bull. 1988, 67, 342.

‘8 Oliveira, A. A. M.; Kaviany, M. Int. J. Heat Mass Tran. 1999, 42, 1059.

'9 Gurin, V. N.; Obukhov, A. P.; Mazina, T. I.; Terent'eva, Z. P.; Kozlova, I. R.;
Korsukova, M. M. Izvestiya Akademii Nauk SSSR, Neorganicheskie Materialy 1969, 5,
1995.

20 Rowcliffe, D. J.; Warren, w. J. J. Mater. Sci. 1970, 5, 345.

2' a) Halden, F. A. Silicon Carbide, Proc. Conf., Boston, 1960, Volume Date 1959, 115.
b) Pickar, Paul B., Jr. US. 1967, 5 pp. c) Ellis, R. C., Jr. Silicon Carbide, Proc. Conf
Boston 1960, I959, 420.

22 Millican, J. N.; Macaluso, R. T.; Young, D. P.; Moldovan, M.; Chan, J. Y. J. Solid
State Chem. 2004, 177, 4695.

"-3 Canﬁeld, P. C.; Fisk, z. Philos. Mag. B 1992, 65, 1117.

24 Holm, A. P. P., Park, S.; Condron, C. L.; Olmstead, M.; Kim, H.; Klavins, P.;
Grandjean, F.; Hermann, R. P.; Long, G. J.; Kanatzidis, M. G.; Kauzlarich, S. M.; Kim,
S. Inorg. Chem. 2003, 42, 4660.

25 Kanatzidis, M. G.; P6ttgen, R.; Jeitschko, w. Angew. Chem. Int. Ed. 2005, 44, 6996.

26 a) Kevorkijan, V. Aluminum 2002, 78, 469. b) Satyanarayana, K. G.; Pillai, R. M.; Pai,
B. C. Trans. Indian Inst. Metals 2002, 55, 115. c) Ochiai, S. Mechanical Properties of
Metallic Composite, Marcel Dekker, Inc., New York, 1994.

27 Latturner, s. 13.; Kanatzidis, M. G. Inorg. Chem. 2004,43, 2.

28 Kawai, S.; Takaho, T.; Shigetoshi, M.; Masakazu, A.; Oshima, C. Kinet. Prop.
Electron. Magn. Ceram., Proc. U. S. —Jpn. Semin. Basic Sci. Ceram. 1976, 125.

29 Okada, s.; Tetsuzo, A. Nippon Kagaku kaishi 1984, 3, 416.

17

 

30 Okada, S.; Tetsuzo, A.; Iwami, H.; Yasuo, T. Nippon Kagaku kaishi 1985, 1, 1.
3' Okada, S.; Tetsuzo, A. Kanagawa Daigaku Kogakubu Kenkyu Hokoku 1985, 23, 37.

32 Zhang, Y.; Shigeru, 0.; Tetsuzo, A.; Takeo, Y.; Yasumori, I. Yogyo Kyokaishi 1987,
95, 374.

33 Shigeru, 0.; Kudou, K .; Yu, Y.; Lundstroem, T. JJAP Series 1994, 10, 136.

3“ a) Niemann, 5.; Jeitschko, w. Z. Metallkd. 1994,85, 345. b) Niemann, s.; Jeitschko,
w. J. Solid State Chem. 1995, 114, 337.

35 a) Fehrmann, 13.; Jeitschko, w. Inorg. Chem. 1999, 38, 3344. b) Thiede, v. M. T.;
Jeitschko, W. Zeitschriftfuer Naturforschung, B: Chemical Sciences 1998, 53, 673.

3" a) Thiede, v. M. T.; Ebel, T.; Jeitschko, w. J. Mater. Chem. 1998, 8, 125. b) Reehuis,
M.; Wolff, M. W.; Krimmel, A.; Scheidt, E. W.; Stusser, N.; Loidl, A.; Jeitschko, W. J.
Phys Condens. Matt. 2003, 15, 1773.

37 Fehrrnann, B.; Jeitschko, W. Z. Naturforschung, B: Chemical Sciences 1999, 54, 1277.
38 a) Chen, X. Z.; Sportouch, S.; Sieve, B.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.;
Patschke, R.; Kanatzidis, M. G. Chem. Mater. 1998, 10, 3202. b) Sieve, B.; Chen, X. Z.;
Cowen, J .; Larson, P.; Mahanti, S. D.; Kanatzidis, M. G. Chem. Mater. 1999, 11, 2451.
c) Sieve, B.; Chen, X. Z.; Henning, R.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.;
Schultz, A. J .; Kanatzidis, M. G. J. Am. Chem. Soc. 2001, 123, 7040. d) Latturner, S. E.;
Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G. Chem. Mater. 2002, 14, 1695.

39 Vasudevan, A. K.; Petrovic, J. J. Mater. Sci. Eng. 1992, A155, 1.

4° Williams, J. J.; Akinc, M. Oxid. Met. 2002, 58, 57.

4‘ Lohfeld, s.; Schuetze, M. Mater. Corros. 2005, 56, 93.

42 Kurokawa, K.; Matsuoka, 11.; Nagai, T. T. Mater. Res. Soc. Jpn. 1994, 14A, 255.

43 Liu, X. B.; Yu, L. G.; Wang, H. M. J. Mater. Sci. Lett. 2001, 20, 1489.

4“ Wang, H. M.; Lu, x. D.; Liu, Y. F.; Duan, 6.; Cal, L. x.; Wang, C. M. Mater. Sci.
Foru, 2003, 426-432. 2551.

‘5 Gross, Erwin. Int. J. Powder Metal]. 1995, 31, 239.

46 a) Kishimoto, Y.; Kawasaki, Y.; Ohno, T. Phys. Lett. A, 2003, 31 7, 308. b) Kim, S. J .;
Salvador, J.; Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G. J. Am. Chem. Soc. 2001, 123,

18

 

 

127 .|
Kauzl a

4, 8) H1
Mydos':
Flossair|
K3 Km
48 a ) B"
Magn. 'l
HotTma

w a) Jul
63. b) 1
W.; Rt)
Phys, R,

50 .
Slevc
Papaet‘ri

 

 

59
05th

61’)
Sheehr

61

 

12704. c) Fisher, 1. R.; Bud'ko, S. L.; Song, C.; Canﬁeld, P. C.; Ozawa, T. C.;
Kauzlarich, S. M. Phys. Rev. Lett. 2000, 85, 1120.

47 a) Hossain, Z.; Schmidt, M.; Schnelle, W.; Jeevan, H. S.; Geibel, C.; Ramakrishnan, S.;
Mydosh, J. A.; Grin, Y. Phys. Rev. B, 2005, 71, 060406. b) Singh, Y.; Ramakrishnan, S.;
Hossain, Z.; Geibel, C. Phys. Rev. B, 2002, 66, 014415. c) He, J .; Tsujii, N.; Yoshimura,
K.; Kosuge, K. J. Alloy Compd. 1998, 268, 221.

48 a) Baranov, N. V.; Hilscher, G.; Markin, P. E.; Michor, H.; Yermakov, A. A. J. Magn.
Magn. Mater. 2004, 272-276, 637. b) Taremba, V. 1.; Kaczorowski, D.; Rodewald, U.;
Hofﬁnann, R. D.; Poettgen, R. Chem. Mater. 2004, 16, 466.

49 a) Julian, s. R.; Tautz, F. 5.; McMullan, G. J.; Lonzarich, G. G. Phys. B 1994, 199-200,
63. b) Tanatar, M. A.; Paglione, J .; Nakatsuji, S.; Hawthorn, D. G.; Boaknin, E.; Hill, R.
W.; Ronning, F.; Sutherland, M.; Taillefer, L.; Petrovic, C.; Canﬁeld, P. C.; Fisk, Z.
Phys. Rev. Lett. 2005, 95, 067002.

50 Sieve, B.; Sportouch, S.; Chen, X. Z.; Cowen, J. A.; Brazis, P.; Kannewurf, C. R.;
Papaefthymiou, V.; Kanatzidis, M. G. Chem. Mater. 2001, 13, 273.

5‘ Latturner, s. 13.; Bile, 1).; Mahanti, s. 13.; Kanatzidis, M. G. Inorg. Chem. 2003, 42,
7959.

52 Brevnov, D. A.; Bungay, C. J. Phys. Chem. B 2005, 109, 14529.
5’ Latturner, s. 13.; Kanatzidis, M. G. Chem. Commun. 2003, 18, 2340.

54 Latturner, S. E.; Bilc, D.; Ireland, J. R.; Kannewurf, C. R.; Mahanti, S. D.; Kanatzidis,
M. G. J. Solid State Chem. 2003, I 70, 48.

55 Sieve, B. Ph. D. Dissertation, Michigan State University, 2002.
56 Shah, D. M.; Anton, D. L. Mater. Sci. Eng. A 1992, A153, 402.

57 Vaz, F.; Rebouta, L.; Da Silva, M. F.; Soares, J. C. NA T0 AS] Series, Series 3: High
Technology 1997, 21, 501.

53 a) Ohta, Y.; Miura, 5.; Mishima, Y. Intermetallics 1999, 7, 1203. b) Ono, K.; Kado,
M.; Suzuki, R. O. Steel Res. 1998, 69, 387.

59 Yoshida, S.; Sato, M.; Sugawara, E.; Shimada, Y. J. Appl. Phys. 1999, 85, 4636.
6" Shechtrnan, 1).; Blech, 1.; Gratias, 13.; Cahn, J. w. Phys. Rev. Lett. 1984,53, 1951.

6’ Ma, Y.; Stern, E. A.; Bouldin, C. 13. Phys. Rev. Lett. 1986,57, 1611.

I9

 

62 Wu, L. M.; Seo, D. K. J. Am. Chem. Soc. 2004, 126, 4398.

63 Surikov, V. I.; Borzhitskaya, M. K.; Shtol'ts, A. K.; Zagryazhskii, V. L.; Gel’d, P. V.
Fiz. Met. Metalloved. 1970, 30, 1167.

6“ German, N. v. Vestn. L'VoV. Um‘V. Chim. 1981, 23,61.

65 Mushnikov, N. V.; Goto, T.; Kolomiets, A. V.; Yoshimura, K.; Zhang, W.; Kageyama,
H. J. Phys: Condens. Mat. 2004, 16, 2395.

66 Avila, M. A.; Sera, M.; Takabatake, T. Phys. Rev. B 2004, 70, 100409.

67 Muro, Y.; Haizaki, Y.; Kim, M. S.; Umeo, K.; Tou, H.; Sera, M.; Takabatake, T. Phys.
Rev. B 2004, 69, 020401.

63 Salvador, J.; Guo, F.; Hogan, T.; Kanatzidis, M. G. Nature 2003, 425, 702.

69 Kim, S. J.; Salvador, J .; Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G. J. Am. Chem. Soc.
2001, 123, 12704.

70 Bobev, S.; Thompson, J. D.; Sarrao, J. L.; Olmstead, M. M.; Hope, H.; Kauzlarich, S.
M. Inorg. Chem. 2004, 43, 5044.

7‘ Chan, J. Y.; Kauzlarich, S. M.; Klavins, P.; Shelton, R. N.; Webb, D. J. Chem. Mater.
1997, 9, 3132.

72 Chan, J. Y.; Wang, M. 13.; Rehr, A.; Kauzlarich, s. M.; Webb, D. J. Chem. Mater.
1997,9,2l31.

20

 

2-1. Int

synthes
quality
Physica
RE T.\l
by our
COmpou
pr0pem,
% F e, (j,
comPOU:
Parallel

35141

 

 

CHAPTER TWO

REAuAl4Ge2 and REAuAI4(Au,Ge1-,)2 (RE = rare earth element): New

Quaternary Intermetallics Grown from Aluminum Flux

2-1. Introduction

Recently, Al has been suggested as a high temperature solvent for the exploratory
synthesis of new intermetallic phases.l‘2’ 3 Liquid Al facilitates the growth of large high
quality single crystals of complex intermetallic compounds, and this makes structural and
physical characterization easier and more reliable. Synthetic explorations of the system
RE/TWAUSi(Ge) (RE = rare earth element, TM = ﬁrst row transition metal) conducted
by our former group members led to the discovery of a number of new multinary
compounds, many with novel structures and interesting magnetic and electronic
properties.4 Interestingly, this system is very effective for the late transition metals such
as Fe, Co and Ni. Moreover, both Si- and Ge-containing systems tend to form quaternary
compounds from A1 ﬂux; sometimes isostructural analogues are obtained indicating
parallel chemistry between Si and Ge. For example, the families RE2-xTMA14Tr2(All-
yTry)(All.zTrz)2 (RE = Sm, Dy, Er; TM = Ni, Co; Tr = Si, Ge) can be formed with both Si
and Ge,5 when the reaction ratio of the rare earth metal to transition metal is smaller than
1. The rare earth-rich family RE2N1(NixSi1-x)A148i6 (RE = Pr, Nd, Sm, Gd, Dy and Tb)
was found to be isostructural analogue of szCo(Co,,Al1.,,)A14Ge(,.y,5 although the reason
why rare earth metals show different reaction reactivity is not clear at this point. When
the transition metal was moved to the earlier ones such as Mn and Fe, the results led to
the discovery of the novel series RE4TM2+XAI7-xSi3 (RE = Ce, Pr, Nd, Sm; TM = Mn,

Fe),4 while an isostructural Ge-analogue has not been found so far.

21

interest
numbe:
REAus
antiﬂm
REgAu
\NiAl.
be redl
BaHgll
Ge. but
using A

Sm. Eu.

Eu), 1},

 

 

 

 

The results obtained with the ﬁrst and second row transition metals stimulated
interest in the third row. Gold is particularly active in Al ﬂux and has resulted in a
number of new phases such as the Th2(AuxSi1-x)[AuA12],.Si2 homologous series, 6
REAu4A13Si,7 REzAUAlésl48 and REAu3,A17.9 These compounds feature hexagonal
antiﬂuorite-type slabs which can be regarded as fragments of the bulk AuAlz structure.
REzAuA168i4 is comprised of two different layers — a CaAIZSiz-type layer and a
YNiAlaGez-type layer. The perovskites MTiO3 (M = Ca, Sr, Ba) combined with Au can
be reduced by aluminum to form quaternary compounds M3Au6+xA126Ti with a stuffed
BaHgn structure type.'0 The rich chemistry of the Si systems appears to be parallel to
Ge, but not in an identical fashion. Our systematic studies in the system RE/Au/Al/Ge
using Al as the ﬂux led to two new quaternary families: REAuAIaGez (RE = Ce, Pr, Nd,
Sm, Eu, Gd, Tb, Dy, Er, Tm and Yb) and REAuA14(AuxGe1-x)2 (x = 0.4) (RE = Ce and
Eu). The Cu-analogue of the second family CeCuA14(CuxGe1-x)2 (x = 0.62) was obtained
when we studied the system RE/Cu/Al/Ge. In this chapter, synthesis, structural

characterization and magnetic measurements of these families are described.

242. Experimental Section
Reagents:

The following reagents were used as obtained: rare earth metal (RE = Ce, Pr, Nd,
Sm, Eu, Gd, Tb, Dy, Er, Tm and Yb) (Cerac, 99.9%), Au (shavings from 1 ounce gold

bullion, 99.99%), A1 pellets (Cerac, 99.99%), Ge (Cerac, 99.999%).

22

—.__ _w -__—. _ ._

Syntlr
nitro g
mmol
crucib
vacuur
tempel
d, fOIIt

down 1

n111101 l

 

 

 

Synthesis:

REAuAIaGez (RE = Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Er, Tm and Yb): In a
nitrogen-ﬁlled glove box, 1 mmol RE metal (0.14-0.18 g), 1 mmol Au (0.197 g), 10
mmol Al (0.270 g) and 5 mmol Ge (0.36 g) were combined in an alumina crucible. The
crucible was then placed into a silica tube (13 mm in diameter), which was sealed under
vacuum (~10“1 Torr). The samples were heated to 1000 °C in 12 h, maintained at this
temperature for 30 h, then cooled to 850 °C in 24 h. They were annealed at 850 °C for 3
(1, followed by cooling down to 500 °C in 72 h. Finally the temperature was brought
down to 50 °C in 12 h.

REAuAMAuchH); (x = 0.4) (RE = Ce, Eu): In a nitrogen-ﬁlled glove box, 1
mmol RE metal (Ce 0.140 g, Eu 0.152 g), 1 mmol Au (0.197 g), 10 mmol Al (0.270 g)
and 2 mmol Ge (0.1466 g) were combined in an alumina crucible. The crucible was then
placed into a Silica tube (13 mm in diameter), which was sealed under vacuum (~10‘4
Torr). The samples were heated to 1000 °C in 12 h, maintained at this temperature for 5
h, then cooled to 850 °C in 24 h, ﬁnally slowly cooled down to 50 0C in 72 h.

CeCuAla(Cu,Ge1.x)2 (x = 0.62): In a nitrogen-ﬁlled glove box, 1 mmol Ce metal
(0.140 g), 1 mmol Cu (0.064 g), 10 mmol Al (0.270 g) and 5 mmol Ge (0.36 g) were
combined in an alumina crucible. The crucible was then placed into a silica tube (13 mm
in diameter), which was sealed under vacuum (~104 Torr). The samples were heated to
1000 °C in 12 h, maintained at this temperature for 30 h, then cooled to 860 °C in 24 h.

They were annealed at 850 °C for 3 (I, followed by cooling down to 500 °C in 72 h.

Finally the temperature was brought down to 50 °C in 12 h.

23

150161111

NaOH
rinsed ‘
for RE.
the inil

elem en

Elemen

 

Isolation:

The excess aluminum was removed by soaking the crucibles in aqueous 5M
NaOH solution overnight. The solid product remaining after the isolation procedure was
rinsed with water and acetone. The yields of each were ~80% for REAuAlaGez, ~90%
for REAuA14(AuxGe1_x)2 (x = 0.4) and ~90% for CeCuA14(CuxGe1-x)2 (x = 0.62) based on
the initial amount of RE metal used in the reaction. Single crystals were selected for

elemental analysis, X-ray diffraction and magnetic susceptibility measurements.

Elemental Analysis:

The crystals were picked and placed on a Scanning Electron Microscope (SEM)
sample stub using carbon tape. Chemical composition of the products was determined by
Energy Dispersive Spectroscopy (EDS) performed on a J EOL J SM-35C SEM equipped
with a NORAN EDS detector. Spectra were acquired by applying a 25 kV accelerating
voltage with an accumulation time of 30 s. The atomic ratios in the compounds
CeAuAlaGez, EuAuA14(AuxGe1-x)2 (x = 0.4) and CeCuA14(CuxGe1-x)2 (x = 0.62) were
determined to be 1 : 1.25 : 4.87 : 2.1 (Ce : Au : Al : Ge), 1 : 1.76 : 3.73 : 1.1 (Eu : Au : Al
: Ge) and 1 : 1.86 : 3.84 : 0.86 (Ce : Cu : Al : Ge) respectively, which agreed well with

the results derived from single crystal X-ray analysis.

X—ray Crystallography:
Single crystal X-ray diffraction data were collected for REAuAlaGez (RE = Ce,
Nd, Gd, Er), EuAuA14(AuxGe.-x)2 (x = 0.4) and CeCuA14(CuxGe1-x)2 (x = 0.62) at room

temperature on a Bruker AXS SMART CCD X-ray diffractometer. A data collection

24

(Mo l\'
hemisp
softwar
SADAI
and rct

anisotri

EDS.

 

 

INEL

EXPCI‘ir

“181.31 ,

and PW

(Mo K01 radiation, A = 0.71073 A) was acquired covering either a full sphere or
hemisphere of reciprocal space. Data processing was performed with the SAINTPLUS
software package. ” An empirical absorption correction was applied to the data using the
SADABS program.12 The structures were solved straight forwardly using direct methods
and reﬁned with the SHELXTL package program.” All atomic positions were reﬁned
anisotropically. The resulting stoichiometry agreed well with the elemental analysis from
EDS. Data collection parameters and reﬁnement details for CeAuAlaGez and
NdAuAlaGez can be found in Table 2-1. The reﬁnement details for Gd and Er analog are
listed in Table 2-2. Atomic positions, displacement parameters and anisotropic
displacement parameters for REAuAl4Ge2 (RE = Ce, Nd, Gd, Er) are listed in Tables 2-3
and 2-4. Data collection parameters and reﬁnement details for EuAuA14(AuxGe1_x)2 (x =
0.4) and CeCuAl4(CuxGe1-x)2 (x = 0.62) can be found in Table 2-5; atomic positions,
displacement parameters and anisotropic displacement parameters for these two
compounds are listed in Tables 2-6 and 2-7.

X-ray powder diffraction data were collected at room temperature on a CPS 120
IN EL X-ray diffractometer (Cu K01) equipped with position-sensitive detector.
Experimental powder patterns were compared to the patterns calculated from the single
crystal structure solution (by the CrystalDiffract software) to determine the phase identity

and purity.

25

 

Table 2

Empl

 

Form
Tem;
\Vah
Spac.
Lattit

Volu

Calcr
Abso
F{01_ll
C135:
9 ran
Limi:

Reﬂc
Emu)
Rim

Com}
Reﬁn
V4111;
Good
Final

R ind

Extin

 

11.,

 

 

Table 2-1. Crystal data and structure reﬁnements for REAuAlaGez (RE = Ce, Nd).

 

 

Empirical formula CeAuAlaGez NdAuAlaGez
Formula weight 590.19 594.31
Temperature 293(2) K 293(2) K
Wavelength 0.71073 )1 0.71073 A
Space Group R -3 m (#166) R -3 m (#166)

Lattice constants (A)

a = 4.2384(7) A
c= 31.613(7) A

a = 4.2258(4) A
c = 31.359(5) A

Volume 491.81(16) A3 484.97(10) A3
Z 3 3
Calculated density, (g/cm3) 5.978 6.105
Absorption coefﬁcient, (mm'l) 38.598 40.132
F(000) 759 765
Crystal size, (mm3) 0.63 x 021 x 0,32 0.43 x 0.26 x 0.33
0 range, (°) 193 to 2763 11.22 to 37.15
Limiting indices -5 g h g 5 -7 S h S7
-5 S k S 5 -7 _<_ k .<_ 7
-4031339 5231352
Reﬂections collected 1650 2487
Unique reﬂections 186 347
Rim 0.0454 006%
Completeness to 0 100.0% 93.3 %

Reﬁnement method Full-matrix least-squares on F2

Variables 15 15

Goodness-of-ﬁt on F2 1,111 1.128

Final R indices [I>2cr(l)]a R1 = 0.0214 R1 = 0.0235
sz = 0.0514 WRz = 0.0634

R indices (all data) R1 = 0.0214 R1 = 0.0235
WRz : 0.0514 WRz = 0.0634

Extinction coefﬁcient 0.0121(7) 0.0111(7)

Highest residual peak (e/A3) 1.412 and -2.365 2.920 and -4494

2 2 2 2 1/2
R1 = 2(lFol'chlyleol; WRZ = [2[W(Fo ' Fc l/ [20”le ) I

26

Table ;

 

Volt

Calc
Absr
F100
C135

0 ran
Limi

 

 

Table 2-2. Crystal data and structure reﬁnements for REAuAlaGez (RE = Gd, Er).

 

 

Empirical formula GdAuAlaGez ErAuAlaGez
Formula weight 581.11 671.29
Temperature 293(2) K 293(2) K
Wavelength 0.71073 A 0.71073 A
Space Group R -3 m (#166) R -3 m (#166)

Lattice constants (A)

a = 4.2123(6) A
c = 30.994(6) A

a = 4.2074(4) A
c = 30.717(5) A

Volume 476.26(14) A3 470.91(9) A3
z 3 3

Calculated density, (g/em3) 6.352 6.531
Absorption coefﬁcient, (mm’l) 43.133 46.427
F(000) 777 789

Crystal size, (mm3)

0.27 x 0.19 x 0.28

0.33 x 0.28 x 0.11

0 range, (°) 11.25 to 37.25 11.27 to 37.00
Limiting indices -7 S h S 7 -7 S h S 6
-7 S k S 7 -7 S k S 7
49515.48 -51_<_1551
Reﬂections collected 2472 2448
Unique reﬂections 332 338
Rim 0.0635 0.0793
Completeness to 0 91.0 % 94.4 %

Reﬁnement method

Full-matrix least-squares on F2

Variables 15 15

Goodness-of-ﬁt on F2 1.168 1.108

Final R indices [I>2<5(I)]al R1 = 0.0299 R1 = 0.0283
sz = 0.0696 sz = 0.0772

R indices (all data) R1 = 0.0305 R1 = 0.0297
sz = 0.0708 sz = 0.0780

Extinction coefﬁcient 0.0104(8) 0.0134(10)

Highest residual peak (e/A3) 5.815 and -3.885 2.732 and -4754

2 2 2 2 1/2
R1 = £(lFol-1Fcl)/2|Fo|; wR2 = [2[w(Fo - Fe ]/ [£(wIFaI ) ]

27

Table 2—3. Atomic coordinates and equivalent isotropic displacement parameters (A2 x

103) for REAuAlaGez (RE = Ce, Nd, Gd, Er).

 

 

Atom Wyk. x y 2 U( )‘
Symbol °“
Ce 3]) -0.6667 0.6667 0.1667 8(1)
Au 3a 0 0 0 7(1)
Ge 6c -0.3333 0.3333 0.1081(1) 8(1)
Al(l) 6c -0.3333 0.3333 0.0245(1) 7(1)
Al(2) 6c 0 0 0.0820(1) 9(1)
Nd 31) -0.6667 0.6667 0.1667 7(1)
Au 3a 0 0 0 5(1)
Ge 6c -0.3333 0.3333 0.1091(1) 8(1)
Al(l) 6c 03333 0.3333 0.0250(1) 7(1)
Al(2) 6c 0 0 0.0823(1) 6(1)
Gd 36 -0.6667 0.6667 0.1667 5(1)
Au 3a 0 0 0 7(1)
Ge 6c 03333 0.3333 0.1105(1) 6(1)
Al(l) 6C -0.3333 0.3333 0.0255(1) 6(1)
Al(2) 6c 0 0 0.0832(1) 8(1)
Er 31) -0.6667 0.6667 0.1667 5(1)
Au 3a 0 0 0 6(1)
Ge 6c -0.3333 0.3333 0.1121(1) 4(1)
Al(l) 6C -0.3333 0.3333 0.0260(1) 8(1)
Al(2) 6c 0 0 0.0842(1) 6(1)

 

* UM) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

28

Table 2-4. Anisotropic displacement parameters (A2 x 103) for REAuAl4Ge2 (RE = Ce,
Nd, Gd, Er).

 

 

Atom U11 U22 U33 U23 U13 U12
Ce 9(1) 9(1) 6(1) 0 0 5(1)
Au 7(1) 7(1) 5(1) 0 0 4(1)
Ge 9(1) 9(1) 6(1) 0 0 4(1)
Al(l) 7(1) 7(1) 7(2) 0 0 4(1)
Al(2) 11(1) 11(1) 5(1) 0 0 6(1)
Nd 7(1) 7(1) 7(1) 0 0 4(1)
Au 6(1) 7(1) 3(1) 0 0 3(1)
Ge 9(1) 9(1) 7(1) 0 0 4(1)
Al(l) 6(1) 7(1) 8(1) 0 0 5(1)
Al(2) 5(1) 5(1) 7(1) 0 0 4(1)
Gd 5(1) 6(1) 5(1) 0 0 3(1)
Au 7(1) 7(1) 7(1) 0 0 4(1)
Ge 7(1) 6(1) 7(1) 0 0 4(1)
Al(l) 6(1) 7(1) 7(1) 0 0 4(1)
Al(2) 9(1) 8(1) 8(1) 0 0 5(1)
Er 4(1) 4(1) 7(1) 0 0 2(1)
Au 6(1) 6(1) 6(1) 0 0 4(1)
Ge 5(1) 3(1) 7(1) 0 0 3(1)
Al(l) 6(1) 5(1) 12(1) 0 0 5(1)
Al(2) 6(1) 6(1) 8(1) 0 0 3(1)

 

The anisotropic displacement factor exponent takes the form:
-2n2[h2a*2Ul '+. . .+2hka*b*u‘2]

29

Table
CeCu

Em
F01
Tel
We
Lat
V0
Ca
Ab
F1 1
Cr.
Lil
Re
L11
C 0
Re
Va
00
Fir
R 1

Ex-

R1.

Table 2-5. Crystal data and structure reﬁnements for EuAuA14(AuxGe1-x)2 (x = 0.4) and
CeCuA14(CuxGe1-x)2 (x = 0.62).

 

 

Empirical formula EuAuA14(AuxGe1-x)2 CeCuA14(CuxGei-x)2
(x = 0.4) (x = 0.62)
Formula weight 726.4 671.29
Temperature 293(2) K 293(2) K
Wavelength 0.71073 A 0.71073 A
Space Group P4/mmm (#123) P4/mmm (#123)
Lattice constants (A) a = 4.3134(8) A a = 4.2207(4) A
c = 8.371(3) A c = 7.9504(11) A
Volume 155.75(7) A3 141 .63(3) A3
Z 1 . 1
Calculated density, (g/cm3) 7.745 9.286
Absorption coefﬁcient, (mm'l) 62.085 42.176
F(000) 305 354
Crystal size, (mm) 0.046 x 0.064 x 0.052 0.033 x 0.058 x 0.071
0 range, (°) 2.43 to 28.25 14.63 to 39.73
Limiting indices -5 S h S 5 -7 S h S 7
-5 S k S 5 -7 S k S 7
4051510 -14Sl_<_14
Reﬂections collected 1745 2341
Unique reﬂections 146 283
Rim 0.0533 0.0401
Completeness to 0 96.7 % 91.3 %
Reﬁnement method F ull-matrix least-squares on F2
Variables 14 14
Goodness-of-ﬁt on F2 1.151 1.498
Final R indices [I>20(I)]a R1 = 0.0276 R1 = 0.0321
sz = 0.0634 sz = 0.0849
R indices (all data) R1 = 0.0289 R1 = 0.0323
sz = 0.0642 sz = 0.0849
Extinction coefﬁcient 0.141(9) 0.12(2)
Highest residual peak (e/A3) 2.923 and -4.569 2.545 and -1.697

 

2 2 2 2
R1 = 2(lFol-IFcI)/2lFol; wR2 = [X[W(Foz- r, 1/ [2(WlFol > 1”

30

Table 2-6. Atomic coordinates and equivalent isotropic displacement parameters (A2 x

103) for EuAuA14(AuxGe1-x)2 (x = 0.4) and CeCuAl.(Cu,Ge.-,)2 (x = 0.62).

 

 

Wyk.Symbol x y z U(eq) Occu.

Au(1) 1a 0.0000 0.0000 0.0000 10(1) 1
Au(2) 2h -0.5000 -0.5000 0.3462(1) 11(1) 0.41 1(4)

Ge 2h -0.5000 -0.5000 0.3462(1) 11(1) 0.590(4)

Eu 1b 0.0000 0.0000 0.5000 13(1) 1

A1 41' 0.0000 -0.5000 0.1739(4) 12(1) 1
Cu(l) 1a 0.0000 0.0000 0.0000 6(1) 1
Cu(2) 2h -0.5000 -0.5000 0.3523(1) 8(1) 0.620(3)

Ge 2h -0.5000 -0.5000 0.3523(1) 8(1) 0.380(3)

Ce 1b 0.0000 0.0000 0.5000 6(1) 1

A1 4i 0.0000 -0.5000 0.1721(2) 9(1) 1

 

* U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

Table 2-7. Anisotropic displacement parameters (A2 x 103) for EuAuA14(AuxGe1-x)2 (x =
0.4) and CeCuA14(CuxGe1-x)2 (x = 0.62).

 

 

Atom U11 U22 U33 U23 U13 U12
Au(1) 12(1) 12(1) 7(1) 0 0 0
Au(2) 14(1) 14(1) 6(1) 0 0 0
Ge 14(1) 14(1) 6(1) 0 0 0
Eu 14(1) 14(1) 10(2) 0 0 0
Al 17(2) 12(2) 7(2) 0 0 0
Cu(l) 7(1) 7(1) 7(1) 0 0 0
Cu(2) 7(1) 7(1) 8(1) 0 0 0
Ge 7(1) 7(1) 8(1) 0 0 0
Ce 5(1) 5(1) 7(1) 0 0 0
Al 13(1) 7(1) 7(1) 0 0 0

 

The anisotropic displacement factor exponent takes the form:
-21t2[hza*2Ul ‘+. . .+2hka*b*u'2]

31

Magnetic Characterization:

Magnetic measurements were conducted on the polycrystalline samples of
CeAuAlaGez, EuAuAlaGez, DyAuAlaGez, CeAuA14(AuxGe1-x)2 and EuAuAl4(AuxGe1-x)2
(x = 0.4) using a Quantum Design MPMS SQUID magnetometer. EDS-analyzed crystals
were ground into powder, and then sealed in kapton tape and placed into the
magnetometer. The data were collected in the temperature range of 3-300 K at 1000 G,
while ﬁeld dependent magnetic measurements, conducted at 3 K, were carried out in

ﬁelds up to i 55000 G. A diamagrretic correction was applied to the data to account for

core diamagnetism.

2-3. Results and Discussion
Synthesis:

The compounds REAuAlaGez were obtained with most rare earth elements and
they tend to crystallize as plates or pyramidal crystals. Figure 2-1 shows the SEM images
of typical single crystals of CeAuAlaGez and EuAuAl4(AuxGe1-x)2 (x = 0.4). The yields
of the reactions were generally > 80% based on the rare earth elements used. Side
products were mainly recrystallized Ge and Au. Reactions with longer annealing times at
1000 °C did not improve the yields. Use of Yb as the RE metal produced the ternary
compound YbAlzGez instead of the target phase. Only when the amount of Yb was
increased (from elemental ratio 1 : 1 : 10 : 5 to 3 : 1 : 10 : 5 for Yb : Au : Al : Ge), could
YbAuAlaGez be obtained as the main phase.

For Ce and Eu, when a much shorter heating proﬁle was used, the phase
REAuA14(AuxGe1-x)2 (x = 0.4) was found as the major product. When Sm was used

under the same heating proﬁle, only SmAuAl4Ge2 could be obtained.

32

 

Figure 2-1. SEM images of typical crystals of (A) CeAuAl4Ge2. (B) EuAuA14(AuxGe1-x)2

(x = 0.4).

33

Crystal Structure of REAuAl4Ge2:

The REAuAlaGez compounds crystallize in a rhombohedral structure with space
group R-3m. These compounds are isostructural to YNiAlaGez,l which suggests that this
structure type is robust and can accommodate a wide variety of non-isoelectronic
transition metals. The difference between the Ni and Au analogs is that RENiAlaGez
forms readily only with late RE metals, whereas almost all RE metals can form the Au
analogs. That the RENiAlaGez compounds form only with late RE metals could be due
to the smaller size of [NiAlaGez] slab (compared to [AuAlaGez] slab) which permits only
the small RE ions to pack in the available space.5 When the Ni atoms are replaced by the
larger Au atoms in this structure, a cell parameter expansion is observed as expected,
with the a-axis being much more expanded than the c-axis. This allows the larger rare
earth elements to ﬁt into the interlayer space. The cell volume for Ce, Nd, Gd and Er
analogs, exhibit the sequential contraction as expected (Figure 2-2).

The structure of CeAuAlaGez can be described as alternating layers of Ce atoms
and [AuAlaGez] slabs, Figure 2-3. The bonds between Ce atoms and the other atoms
were omitted in order to emphasize different layers in the structure. The Ce atoms on the
ab plane are close-hexagonally packed and form regular equilateral triangles with Ce-Ce
distance of 4.2384(7) A, corresponding to the length of the unit cell, a (Figure 2-4A).
Each Ce atom is sandwiched by two puckered layers composed of Ge and Al(2) atoms,
see Figure 2-4B.

The [AuAlaGez] slab consists of distorted A13 cubes, with the lengths of the cube
sides being 2.897(3) A and 3.048 A. Au atoms (as shown in Figure 2-5A), held in the

center of the cube, are surrounded by six Al(l) and two Al(2) atoms above and below to

34

 

achieve an eight-coordinate environment. The Au-Al bond distances are 2.5668(9) A and
2.591(3) A, respectively, comparable to those in gold aluminides such as AuAlz (2.597
A). These cubes are connected to each other via face sharing along the ab-plane.

The Ge atoms form puckered layers with Al(2) atoms which are in chair
geometries when viewed down the c-direction. Each Ge atom is surrounded by three
Al(2) atoms and one Al(l) atom forming an umbrella like geometry (shown in Figure 2-
SB). A similar coordination enviromnent for Ge atoms is found in AEAlzGez (AE = Ca,
Sr) and REAlzGez in which the Ge atom is bonded to three Al atoms and one AE or RE

atom.l4

495

 

490 e

A

00

LII
I

J).
00
O
I

475 L

Cell Volume (A3)

470 .

 

 

 

46 5 I I I L I I I I I I I
Ce Nd Sm Gd Dy Er
Rare Earth

Figure 2-2. Cell volume variation of different rare earth metals in REAuAl4Ge2.

35

 

 

 

 

 

 

Figure 2-3. Structure of CeAuAl4Ge2 viewed down the [010] direction. Large empty
circles: Ce; black dots: A1; lighter shaded circles: Au; darker shaded circles: Ge.

36

O
O 14.23840) A
O O
Q O
O O O

 

Figure 2-4. Structure of CeAuAlaGez. A) Trigonal pattern of Ce atoms in ab plane.

B) Coordination environment of Ce atom.

37

Ge
0/ 1(2)
0

Al(l)

 

Figure 2-5. Structure of CeAuAlaGez. A) Coordination environment of Au atom.

B) Coordination environment of Ge atom.

Table 2-8. Bond lengths (A) for REAuAlaGez (RE = Cc, Nd, Gd, Er).

 

C eAuAI4G82 NdAuAI4G62 GdAuAI4G62 ErAuAl4Ge2

RE-Ge x5 3.0692(7) 3.0342(4) 2.9901(6) 2.9516(5)
Au-A1(1) x6 2.5668(9) 2.5629(6) 2.5575(8) 2.5568(8)
Au-Al(2) x2 2.591(3) 2.5820(19) 2.578(3) 2.585(3)
Ge-Al(1) 2.5824(10) 2.638(2) 2.635(2) 2.645(3)
Ge-Al(2) x3 2.641(3) 2.5806(7) 2.5757(9) 2.5761(9)
Al(l )-A1(1 ) x3 2.897(3) 2.901(2) 2.902(3) 2.906(3)

 

38

 

Crystal Structure of REAuA14(AuxGe,-x)2 (x = 0.4):

The compound REAuA14(AuxGe1-x)2 (x = 0.4) crystallizes in space group
P4/mmm with the KCu483 structure type. The same structure forms with other transition
metals including Ni, Cu and Pd when Ge is replaced by Si.5 As shown in Figure 2-6, the
AuA14(AuxGe1-x)2 layers stack along the c-axis, and the Eu atoms reside in cages formed
by eight Ge/Au atoms and eight Al atoms (Figure 2-7B), similar to those of the Ba atoms
in the BaA14 structure type.15 When viewed on the ab plane, as Shown in Figure 2-7A,
the Eu atoms are arranged in a ﬂat square-net, with an Eu-Eu distance of 4.3134(8)A,
which is equal to the a-cell parameter.

The structure type of the AuA14(AuxGe1_x)2 layer contains a stable unit which
occurs frequently in many intermetallic compounds, such as LaGa(r,Ni1.,,,16 szNiAl4Ge2,4
and szNiGalz.l7 This layer can be described as follows: Au atoms are arranged in a
square network with each Au atom sitting inside a distorted A13 cube, shown in Figure 2-
7C. These cubes share edges with the Al-Al bond distance of 2.912(6) A to form an
inﬁnite slab. The slab is then completed by capping Ge/Au atoms on both sites. Analysis
of the X-ray data shows that the capping atomic sites are in fact occupied by a mixture of
Au and Ge atoms with a ratio of 2 : 3'8 The resulting AuA14(AuxGe1-x)2 layers stack and
link to each other via Ge-Ge bonding between capping Ge atoms along the c-direction.
The distance between Au/Ge atoms across neighboring AuA14(AuxGe1-x)2 layers is

2.575(2) A.

39

 

 

 

 

 

 

 

 

 

 

 

Figure 2-6. Structure of EuAuA14(AuxGe1-x)2 viewed down the [010] direction. Large

empty circles: Eu; black dots: Al; lighter shaded circles: Au; darker shaded circles:

Ge/Au.

40

A)

Eu
0 O O 0
4.3134(8) AI
0 O O O
E O O O O
B) C)

 

Figure 2-7. Structure of EuAuA14(AuxGe1-x)2. (A) The square net formed by Eu atoms in

ab plane. (B) (C) Local coordination environments of Eu and Au(1) atoms.

41

Table 2-9. Bond lengths (A) for EuAuA14(AuxGe1-x)2 (x = 0.4) and CeCuAl4(CuxGe1-x)2
(x = 0.62).

 

 

 

EuAuA14(AuxGe1-x)2 CeCuAl4(CuxGe1-x)2

(x = 0.4) (x = 0.62)
Eu-Ge x10 3.3107(7) Ce-Ge x10 3.2071(4)
Eu-Al x2 3.479(2) Ce-Al x2 3.3536(13)
Au(1)-Al x8 2.6021(17) Cu(l)-Al x8 2.2154(9)
Ge-Al x4 2.5944(18) Ge-Al x4 2.5507(10)
Al-Al 2.912(6) Al-Al 2.9845(3)
Ge—Ge x2 2.575(2) Ge-Ge x2 2.3475(17)
Magnetic Properties:

Crystals were picked from the products and analyzed by EDS. Magnetic
measurements were performed on polycrystalline samples ground from those selected

crystals.

Magnetism of CeAuAl4Ge2, EuAuAl4Ge2 and DyAuAl4Ge2

The magnetic susceptibility as a function of temperature for CeAuAlaGez is
plotted in Figure 2-8A. Up to 300 K, CeAuAlaGez does not obey Curie-Weiss Law at
any temperature range which indicates possible mixed Ce3T/Ce4+ valence in this
compound. The mixed or intermediate valence behavior, which is generally induced by
the hybridization between 4f electrons and conduction electrons, has also been seen in
other Ce-containing compounds such as CeNiA14(Si2-xNix) and CezNiAl(,.,,Ge4.y.5 We
have calculated the magnetic moment of this compound from its susceptibility value xm

measured at T = 280 K according to the equation new = 2.83(Txm/6)'/2(471x10'6) 113. The

42

resulting value ucxp = 0.78 113 is between the theoretical value of 1.1,,” = 0 113 for Ce“ and
um = 2.54 113 for Ce“. However, the temperature dependent magnetic behavior of
CeAuAlaGez is different from well known mixed valence systems such as CeRu3Si2,l9 the
weakly mixed valence system CeSn;20 and the strongly mixed valence system CeRuZZI.
In these typical mixed valence compounds, xm is almost constant at low temperatures,
then increases with increasing temperature and tends to obey the Curie-Weiss Law at
high temperatures. The reason why CeAuAlaGez shows different magnetic behavior
from typical mixed valence compounds might be that, it is a very weakly mixed valence
system that has a much lower ﬂuctuation temperature. Figure 2-8B shows the
magnetization of CeAuAlaGez as a function of ﬁeld. The magnetization increases
linearly up to 10000 G without saturation.

The magnetic susceptibility of EuAuAlaGez obeys the Curie-Weiss Law above 50
K, Figure 2-8C. The neg value, obtained from the data, is 3.34 113, while that of a free-ion
for Eu3+ is 3.40 113. Therefore, we can regard the oxidation state of the Eu ion as 3+,
while Au is in a diamagnetic state. Because of the slightly enhanced stability of half
ﬁlled 4f shells of Eu”, most europium compounds show divalent europium or mixed
valence of Buy/Eu”. To further determine the oxidation state of Eu atoms, l5'Eu
Méssbauer Spectroscopy might be needed. The ﬁeld dependent data (Figure 2-8D)
shows a gradual increase of the magnetization until saturation which occurs at about
30000 G. However magnetization per Eu ion at this point is only 30% of the maximum
value calculated according to the formula umcalc) = gJ 113 (for a free atom the g factor is
deﬁned by the Lande' equation, the total angular momentum J is the sum of the orbital L

and spin S angular momenta).

43

The molar magnetic susceptibility data of DyAuA14G62 vs temperature is plotted
in Figure 2-8E. This material exhibits an antiferromagnetic transition at low temperature
around 11 K and conforms to the Curie-Weiss Law behavior above the transition
temperature. The calculated “eff (9.17 1113), is somewhat lower than the theoretical value
for Dyd+(10.63 113), as is frequently observed, and the difference might be ascribed to the
crystal-ﬁeld effects.22 The measured magnetic moment is entirely attributed to Dy atoms
with the Au atoms being diamagnetic. DyAuAlaGez exhibits ﬁeld-induced metamagnetic
behavior at 3K, as shown in Figure 2-8F. A gradual increase in magnetization is
observed until the ﬁeld reaches 5000 G where a dramatic increment occurs, indicating
spin reorientation. Then the magnetization begins to saturate and does not change much
up to 55000 G. However, the moment per Dy3+ ion at this point is only 60% of its
maximum value (10.60 1.13) which is calculated from upy(ca1c) = g] 1.13. The sharp jump
occurring at 5000 G suggests a possible transition to a more ferromagnetically ordered
state, which can also be supported by the positive sign of the Weiss constant 0. Similar
spin complexity has been observed in other intermetallic compounds including
'DyzAuAlgsiag and B-DyNiGe223. It has been suggested that a trigonal arrangement of
rare earth ions on a plane could create ﬁ'ustration in antiferromagnetic coupling.24 A
ﬁeld higher than 55000 G may be required to achieve complete saturation of the

magnetization.

Magnetism of CeAuAla(AuxGe 1.x): and EuAuAl4(AuxGe1-x)2 (x = 0. 4)
Figure 2-9A gives the thermal dependence of the magnetic susceptibility xm

measured for CeAuAl4(AuxGe1-x)2. The Herr values, calculated by ﬁtting to the Curie-

44

Weiss Law in the temperature range of 50~140 K and 160~300 K, are 1.54 113 and 1.18
113, respectively, which are between the theoretical ones of um = 0 up, for Ce4+ and um =
2.54 1.13 for Ce”. The change in the slope of inverse susceptibility plot may be due to
changes in the Ce valence state over the whole temperature range. This behavior is
different from those compounds with intermediate valence of Ce atoms, which also obey
the Curie-Weiss Law at high temperature while Ce atoms are in 3+ oxidation state.25 The
Au atoms are likely diamagnetic as in the other Al-grown intermetallic compounds
described above.

For EuAuA14(AuxGe1-x)2, the temperature dependent magnetic susceptibility data
at 1000 G Show an antiferromagnetic transition at about 4 K, as shown in the inset of
Figure 2-9 C. When the high temperature (above 20 K) data are ﬁt to the Curie-Weiss
Law, a 11,5 of 6.03 113 is obtained. The calculated effective magnetic moment for Eu2+ 4f7
ions is predicted to be equal to the value of Gd3+ ions, which is 7.94 113. Thus we suggest
that Eu ions in this compound EuAuA14(AuxGe1-x)z are in a divalent 4f7 ground state and
the difference might due to the crystal-ﬁeld effects. This observation is consistent with
the statement that Eu is often found to be divalent in the noble-metal compounds with a
broad 5 band. Examples include EuAg526, EuAu527, EuAuMgzg, Eu(Pd1.,.Au,,)2Si;i_29 and
EuAu4A133i7. The magnetization increases gradually with increasing applied ﬁeld. No

magnetization saturation is observed up to 55000 G (Figure 2-9D).

45

A B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.08 250 0.004 .
0‘07 CeAuAl4Ge2 . . u ' A0003 _ CeAuAl4Ge2 .0 _
"006 " ‘200: 50002 '.
8 ' .a ——> 3x 2 ' ' ‘
% 0.05 ... .1 150 ’5 \0,001 I- .1
=’ I 9 E
E 0.03 ‘ 100 ;,~ 50001 T -
3
5450.02 5 60.002 - . .-
0 01 T 50 20 003 '
. - , D O .-
..~."°~ 0 O o o o o o.
4 I I I I 0 _0004 I I I_ I I I I
0 50 100 150 200 250 300 -15000 —7500 0 7500 15000
Temperature (K) Field (G)
C 1.2 30 D 8
a
l ! DyAuAl,Ge2 . . 25 A 6 .. DyAuAl,Gc2 “u"...
A l‘ . i a 4 - .6
a I I x 75 '
_ 0.8 i . 4 203 e 2 ,l 4
O A \ P
E ' B -. I
\ —> 2. . .
:3 0.6 i! . . "' 15 E O /
E I E V o
v ‘ . \ E '2 '- . ‘l
0.4 .I d 10 a o g
E . I 3 E _4
N O .. 5 g I- . c1
0'2 .. I 5 '6 'g .00.... '-
~O... . . .
0 1 1 l r L. . 0 -8 n I 1 I 1 I n
0 50 100 150 200 250 300 -60000 -30000 0 30000 60000
Temperature (K) Field (G)
E 0.4 200 F 1.5
0.35 EuAuAl4Ge2 . . : 3 l _ EuAuAl4G1:2 ...”~
A 0 3 O - 150 "< 2 °
L3 . . : r g 05 1- .. c1
'2‘ 0 25 . , g \
\ I ——> ‘ 2- 2
a 0 2 1 _ - 100 m m. 0 - -
E . l I: v
3, 0.15 - .' . 3 E
E . . I 3 E ’0.5 '- . '1
x 0.1 - ' o‘ - 50 = o o
. . , v 2 C
0.05 - '5 ‘ '1 '---"' 7
0 .. .... 0 -15 4 r r r r 1 .
0 50 100 150 200 250 300 60000 -30000 0 30000 60000
Temperature (K) Field (G)

Figure 2-8. (A) Temperature dependent magnetic susceptibility of CeAuAlaGez. (B) Field
dependent magnetization at 3K for CeAuAlaGez. (C) Temperature dependent magnetic
susceptibility of EuAuAlaGez. (D) Field dependent magnetization at 3K for EuAuAlaGez.
(E) Temperature dependent magnetic susceptibility of DyAuAlaGez. (F) Field dependent
magnetization at 3K for DyAuAlaGez.

46

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.14 ~ 1400 0.5 "
I
0.12 CeAuA14(AuxGe1-x)2 .. 1200 A . CeAuAl4(AuxGe1-x)2.o' a
O
U ‘ 03x 2
E 0.08 800 3 E ‘
\ " , 0 . ..
0
g 0.06 600 1 “3,
V a 6 ’ 0' ‘
. 3
x3) 04 400 5 5.025 _ .. at
0.02 200 2 .o'
O
0 0 -05 ". . - - n -
50 100 150 200 250 300 -60000 -30000 0 30000 60000
Temperature (K) Field (G)
C
l I I I I I 70 4 ..
c
EuAuA14(AuxGel-x)2 0‘ 60 A 3 .- EuAuA14(AuxGei-x)2 .' q
A 08 . . : LB 2 1- ... .
i=3 . - 50 3x ‘2‘
g 0.6 ' . 40 ’3‘ ‘~. 1 ’ ‘
\ . '———> 9.. Z O _ 4
E 0.4 ' '130 i=1 3? 1
we I.I. g 5 ' " ‘
N u. T 20 ‘5 g -2 ' '. d
0.2 V 2 O
‘ 10 -3 1- ... ‘l
0 1 ..f.“.rc ..‘MO _4 0. 1 - 1 1 -
0 50 100 150 200 250 300 -60000 -30000 0 30000 60000
Temperature (K) Field (G)

Figure 2-9. (A) Temperature dependent magnetic susceptibility of CeAuAl4(AuxGe1-x)2.
(B) Field dependent magnetization at 3K for CeAuA14(AuxGe1-x)2. C) Temperature
dependent magnetic susceptibility of EuAuA14(AuxGe1-x)2. (D) Field dependent
magnetization at 3K for EuAuA14(AuxGe1-x)2.

47

 

 

 

 

2-4. Conclusions

The exploratory reaction chemistry in the system RE/Au/Al/Ge using Al as a ﬂux
led to two types of intermetallic phases REAuAl4Ge2 (RE = Ce, Pr, Nd, Sm, Eu, Gd, Tb,
Dy, Er, Tm and Yb) and REAuAl4(AuxGe1-x)2 (x = 0.4) (RE = Ce and En). These two
families crystallize in different structure types and were characterized by single crystal X-
ray diffraction. However these two structures are related and both can be represented as
consequent stacking of RE layers and AuAhGez (or AuAl4(AuxGe1-x)2) slabs along the c-
axis. Single crystal X-ray diffraction analysis reveals that the Ge atoms in
REAuA14(AuxGe.-x)2 are in a mixed occupancy site with 40% of Au and 60% of Ge,
while for the Cu-analogue, the mixed site is occupied by 62% Cu and 38% Ge.

The discovery of these two families suggests the parallel chemistry between the
third row transition metal and the ﬁrst row transition metal, especially Ni. For example,
the hexagonal phase RETMAl4Ge2 can be formed both with Au and Ni;
REAuAl4(AuxGe1-x)2 and RENiAl4(Si2-xNix) are isostructural phases which have
transition metal and tetrelide on the mixed occupied site.

Magnetic measurements indicate that the magnetic moments are localized on the
RE atoms and Au atoms are in a non-magnetic state. Antiferromagnetic ordering
transitions are observed in DyAuAl4Ge2 and EuAuAl4(AuxGe1-x)2 with Néel temperatures
of 11 K and 4 K respectively, and DyAuAl4Ge2 exhibits metamagnetic behavior at 3K.

The Ce analogs may exhibit valence ﬂuctuations in the temperature range measured.

48

References:

 

' a) Chen, X. Z.; Sportouch, S.; Sieve, B.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.;
Patschke, R.; Kanatzidis, M. G. Chem. Mater. 1998, 10, 3202. b) Sieve, B.; Chen, X. Z.;
Cowen, J .; Larson, P.; Mahanti, S. D.; Kanatzidis, M. G. Chem. Mater. 1999, 11, 2451.

2 a) Thiede, V. M. T.; Fehrmann, B.; Jeitschko, W. Z. Anorg. Allg. Chem. 1999, 625,
1417. b) Fehrmann, B.; Jeitschko, W. J. Alloys Compd. 2000, 298, 153.

3 Kanatzidis, M. G.; Pottgen, R.; J eitschko, W. Angew. Chem. Int. Edit. 2005, 44, 6996.

4 a) Sieve, B.; Sportouch, S.; Chen, X. Z.; Cowen, J. A.; Brazis, P.; Kannewurf, C. R.;
Papaeﬁhymiou, V.; Kanatzidis, M. G. Chem. Mater. 2001, 13, 273. b) Sieve, B.; Chen,
X. Z.; Henning, R.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.; Schultz, A. J.;
Kanatzidis, M. G. J. Am. Chem. Soc. 2001, 123, 7040. c) Sieve, B.; Trikalitis, P. N.;
Kanatzidis, M. G. Z. Anorg. Allg. Chemie 2002, 628, 1568.

5 Sieve, B. Ph. D. Dissertation, Michigan State University, 2002.

6 Latturner, s. 13.; Bilc, D.; Mahanti, s. D.; Kanatzidis, M. G. Chem. Mater. 2002, 14,
1695.

7 Latturner, s. 5.; Kanatzidis, M. G. Chem. Commun. 2003, 18,2340.
3 Latturner, s. 15.; Kanatzidis, M. G. Inorg. Chem. 2003,42, 7959.

9 Latturner, S. E.; Bilc, D.; Ireland, J. R.; Kannewurf, C. R.; Mahanti, S. D.; Kanatzidis,
M. G. J. Solid State Chem. 2003, I 70, 48.

‘0 Latturner, s. 13.; Kanatzidis, M. G. Inorg. Chem. 2004, 43,2.
N Saint, version 4; Simens Analytical X-ray Instruments, Inc., Madison, WI.
12 SADABS, Sheldrick, G. M.; University of Gottingen, Gottingen, Germany.

‘3 Sheldrick, G. M. 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

‘4 Kranenberg, (3.; Johrendt, D.; Mewis, A. Solid State Sci. 2002, 4, 261.
‘5 Andress, K. R.; Alberti, E. z. Metallkunde 1935, 27, 126.

‘6 Grin, Yu. N.; Yarmolyuk, Ya. P.; Rozhdestvenskaya, I. V.; Gladyshevskii, E. 1.;
Kristallograﬁya 1982, 27, 693.

49

 

17Chen, X. Z.; Small, P.; Sportouch, S.; Zhuravleva, M.; Brazis, P.; Kannewurf, C. R.;

Kanatzidis, M. G. Chem. Mater. 2000, 12, 2520.

’8 The Si analogues grown from Al ﬂux such as REAuaAlgsi also has the mixed occupied
site of Au and Si with the ratio 1 : 1. Our assignment in the compound EuAuA14(AuxGe1-
x); supports the assignment that the mixed occupied site in REAmAlgSi is in fact between

Au and Si instead of Au and Al.
'9 Kishimoto, Y.; Kawasaki, Y.; Ohno, T. Phy. Lett. 2003, 31 7, 308.

20 3) Lawrence, J. M.; Riseborough, P. 8.; Parks, R. D. Rep. Prog. Phys. 1981, 44, 1. b)
Lawrence, J. M.; Riseborough, P. S.; Parks, R. D. in: L.M. Falicov, W. Hanke, M.B.

Maple(Eds.), Valence Fluctuations in Solids, North-Holland, Amsterdam, 1981.

2' Tsvyashchenko, A. V.; Fomicheva, L. N.; Sorokin, A. A.; Ryasny, G. K.;
Komissarova, B. A.; Shpinkova, L. G.; Klementiev, K. V.; Kuznetsov, A. V.;

Menushenkov, A. P.; Troﬁmov, V. N.; Primenko, A. E.; Cortes, R. Phys. Rev. B 2002,

65, 174513.
22 Kittel, C. Introduction to Solid State Physics, 7th Ed. John Wiley & Sons, New York,

1996,p.426.
23 Salvador, J. R.; Gour, J. R.; Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G. Inorg. Chem.

2004,43,]403.
2“ Gignoux, D.; Schmitt, D. J. Alloys Compd. 2001, 326, 143.

25 a) Chevalier, B.; Bobet, J. L.; Gaudin, E.; Pasturel, M.; Etourneau, J. J. Solid State
Chem. 2002, 168, 28. b) Tang, J .; Gschneidner, K. A. Phy. Rev. B 1995, 52, 7328.

26 Sampathkumaran, E. V.; Perscheid, B.; Kaindl, G. Solid State Commun. 1984, 51, 701.
27 Van Steenwijk, F. J .; Huiskamp, W. J.; Lefever, H. T.; Thiel, R. C.; Buschow, K. H. J.

Physica B & C 1997, 86-88B, 89.
28 Pottgen, R.; Hoffmann, R.; Renger, J .; Rodewald, U. C.; Moller, M. H. Z. Anorg. Allg.

Chem. 2000, 626, 2257.
2" Abd-Elmeguid, M. M.; Sauer, C.; Koebler, U.; Zinn, w. z. Phys. B: Cond. Matter

1985, 60, 239.

50

 

 

510.1

rapt
llal'e
1131151
are 1
M0,:

re111a]

CHAPTER THREE

V2A15Ge5: First Ternary Intermetallic in the V-Al-Ge system Accessible

in Liquid Aluminum

3-1. Introduction
Recently Al ﬂux has been proven by the Kanatzidis group, J eitschko group and

others to be a powerful method to study the multinary systems including RE/Al/Tr,
RE/l‘M/Al, RE/TM/Al/Tr (RE = rare earth metal, TM = transition metal, Tr = Si or
Ge)"2 Surprisingly rich chemistry was found from these systems in liquid aluminum:
explorations of the new multinary systems led to the discovery of a number of new series
of intermetallics such as RENiAl4Ge2,3 RIS.¢Fe24.,,A17.,,Si3,4 and REAu3A175. In addition,
new phases were obtained by reinvestigation of some old systems. For example, there
are a few ternary rare earth aluminum silicides reported in the literature, such as
RE6AI3Si (RE = Ho, Tm),6 TmzAlSi2,7 and DyAIZSiz.8 Re—inspection of this system using
Al as the ﬂux produced a new family of intermetallics REzAlgsiz (RE = Ho, Er, Tm).9
The single crystal form of these phases obtained from liquid aluminum made the
structural determination easier and physical properties measurements more reliable.
The success on these rare earth-containing systems stimulated us to start
inspecting the systems without rare earth metals, e.g. TM/Al/Tr, TMI/TMz/Al/Tr. There
have been some reports in the literature on ternary transition metal silicides with the

transition metal mainly Mn, Fe, Co, Ni, such as F eAleiIO and N1|6A1819.” Many of them

are pseudo-binary compounds with Al (and Si sitting on the same site, such as

Ni3Alo,68io,4.l3 These systems are of particular interest due to the

M03A10,gSio.2,12
Since 19808, a large number of

remarkable thermal oxidation resistance of silicides.

51

-’

 

 

 

 

 

1111

C61;

5.1m.

quasicrystalline approximants have been discovered in the Al-rich region of the AMT M
and related ternary systems. One well-known example is a-MnAlSi, which is a prototype
of approximants presenting Mackay-type clusters. '4

As to the systems containing Ge, there are even less examples reported in the
literature, MnAlGe,'5 Ni3Alo_gGe0_2“S to name a few. In the system V/Al/Ge, as far as we
know, only a few pseudo-binary phases were reported, such as V3Alo_25Geo,75,i7

V3Alo,3Geo_7l8 and V3,Alo,5Ge0,5.19 All these phases belong to the A15 structure type with

Al and Ge atoms occupying the same site.20 The parent compound V3Ge is a

superconductor with Tc = 6.104 K, and it was found that in the system V3AleeH, the

substitution of Al for Ge expands the lattice and dramatically increases Tc (T6 = 11.17 K

when x = 0.3).21 Our studies on the system V/Al/Ge using Al as the ﬂux led to the

discovery of a new ternary compound VzAlsGes with all the atoms sitting on independent

atomic sites. In this chapter, synthesis, crystal structure, thermal analysis, magnetic

measurements and electronic structure calculations for V2AlsGC5 are reported.

3-2. Experimental Section

Reagents:
The following reagents were used as obtained without further puriﬁcation: V (-325 mesh,

Cerac, 99.7%), A1 pellets (Cerac, 99.99%), Ge (Cerac, 99.999%).

Synthesis:
In a nitrogen-ﬁlled glove box, 2 mmol V metal (0.102 g), 5 mmol Ge (0.36 g) and

10 mmol A] (0.270 g) were combined in an alumina crucible. The crucible was then

52

 

placed into a silica tube (13 mm in diameter), which was sealed under vacuum (~10'4
Torr). The sample was heated to 850 °C in 12 h, maintained at this temperature for 3d,
slowly cooled down to 200 °C at the rate of 15 °C h'l, and ﬁnally brought down to 50 °C

in 5 h. This compound can also be obtained as a pure phase by combining 2mmol V

metal (0.102 g), 5mmol Al (0.135 g) and 5mmol Ge (0.36 g) using the same heating

proﬁle.

 

Isolation:

The excess aluminum was removed by soaking the crucible in aqueous 5M NaOH
solution overnight. The solid product remaining aﬁer the isolation procedure was rinsed
with water and dried with acetone. The yield of the reaction was ~90% based on the

initial amount of V metal used. Single crystals were selected for elemental analysis, X-

ray diffraction, thermal analysis and magnetic susceptibility measurements.

Elemental Analysis:

The crystals were ﬁxed on a Scanning Electron Microscope (SEM) sample plate
using carbon tape. Chemical composition of the products was determined by Energy
Dispersive Spectroscopy (EDS) performed on a JEOL J SM-35C SEM equipped with a

NORAN EDS detector. Data were acquired by applying a 25 kV accelerating voltage

with an accumulation time of 60 s. The atomic ratio averaged from ten crystals was

determined to be 1 : 2.85 : 2.73 ( V : Al : Ge), which agreed well with the results derived

from the single crystal X-ray data analysis.

53

 

 

X-ray Crystallography:
Single crystal X—ray diffraction data of V2A15Ge5 was collected at room
temperature on a Bruker AXS SMART CCD X-ray diffractometer. A data collection
(Mo Ka radiation, 7» = 0.71073 A) was acquired covering a full sphere of reciprocal
space. Data processing was performed with the SAINTPLUS software package.22 An
empirical absorption correction was applied to the data using the SADABS program.”
The structure was solved by direct method and reﬁned with the SHELXTL package
program.24 Systematic absences conditions led to three c-centered space groups: Cmc2,
Cmcm and Ama2. Cmcm is the only centrosymmetric one and has the lowest CFOM
value thus was chosen. Later structure reﬁnement conﬁrmed this choice. All atomic
positions were reﬁned with anisotropic thermal displacement parameters. The resulting
stoichiometry agreed well with the elemental analysis from EDS. Data collection

parameters and reﬁnement details for V2A15Ge5 can be found in Table 3-1. Atomic

positions, displacement parameters and anisotropic displacement parameters for this

compound are listed in Table 3—2 and Table 3-3.

The X-ray powder diffraction data were collected at room temperature on a CPS

120 INEL X-ray diffractometer (Cu Ka) equipped with position-sensitive detector.
Experimental powder patterns were compared to the patterns calculated from the single

crystal structure solution (by the CrystalDiffract program) to determine the phase identity

and purity.

54

 

 

 

Thermal Analysis:

Differential thermal analysis (DTA) was performed on a Shimadzu DTA-SO
differential thermal analyzer with a -A1203 as a standard reference. The sample was
heated to 1000 °C at a rate of 10 oC/min, then cooled to 150 °C at the same rate. The

cycle was then repeated to determine if the compound melted congruently.
1

Thermal gravimetric analysis (TGA) was performed on a Shimadzu TGA-SO

thermal gravimetric analyzer. The sample was heated to 1000 °C at a rate of 10 °C/min

 

under ﬂowing air, held at 1000 °C for 10 min, and then cooled down to room temperature

at the same rate.

 

Figure 3-1. SEM image of a typical crystal of V2A15Ge5.

55

Table 3-1. Crystal data and structure reﬁnements for V2A15Ge5.

 

 

Empirical formula V2A15Ge5

Formula weight 599,73
Temperature 298(2) K
Wavelength 0.71073 A

Space Groun Cmcm (#63)
Lattice constants (A) a = 5407200) A

b = 12.978(2) A ,
c = 11.363(2) A

 

Volume 797.4(3) A3
Z 4
Calculated density, (g/cm3) 4.996
Absorption coefﬁcient, (mm’l) 21,293
F(000) 1084
Crystal size, (mm3) 0.22 x 0.17 x 0.15
0 range, (°) 11.48 to 37.92
Limiting indices -9 g h g 9
-22 S k S 22
-19 S l S 19
Reﬂections collected 6346
Unique reﬂections 1133
Rim 0.0965
Completeness to 0 93.9 %
Reﬁnement method Full-matrix least-squares on F2
Variables 40
Goodness-of-ﬁt on F2 1,064
Final R indices [I>2cr(l)]a R1 = 0.0486
wR2 = 0.0895
R indices (all data) R1 = 0.0323
ng = 0.0996
Extinction coefﬁcient 0.0099(7)

Highest residual peak (e/A3) 2.207 and —1.843

 

R1 = 23(1Fal-1Fel)/>3|Fo|; wR2 = [2[W(F62- Fc2]/ [2(wlFo|2)2]1/2

56

Table 3-2. Atomic coordinates and equivalent isotropic displacement parameters (A2 x

 

 

103) for V2AlsGe5.
Atom Wyk. x y z U(eq).
Symbol

V 8g 0.2526(2) 0.3903(1) 0.7500 3(1)
Ge(l) 4c 0.0000 0.2208(1) 0.7500 6(1)
Ge(2) 4c 0.5000 0.3430(1) 0.5566(1) 6(1)
Ge(3) 8f 0.0000 0.5322(1) 0.6321(1) 8(1)
Al( 1) 4c -0.5000 0.2095(2) 0.7500 1 1(1)
Al(2) 4c 0.0000 0.3331(2) 0.5451(2) 9(1)
Al(3) 8f 0.0000 0.0425(2) 0.6273(2) 9(1)

 

‘ U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

Table 3-3. Anisotropic displacement parameters (A2 x 103) for V2A15Ge5.

 

 

Atom U11 U22 U33 U23 U13 U12
v 4(1) 4(1) 3(1) 0 0 0
Ge(l) 6(1) 4(1) 8(1) 0 0 0
Ge(2) 6(1) 8(1) 4(1) -1(1) 0 0
Ge(3) 7(1) 8(1) 9(1) 3(1) 0 0
Al(l) 8(1) 7(1) 18(1) 0 0 0
Al(2) 11(1) 10(1) 6(1) -3(1) 0 0
Al(3) 9(1) 8(1) 8(1) 3(1) 0 0

 

The anisotro
-27t2[h2a*2U

pic displacement factor exponent takes the form:
1+...+2hl<a*b=1u'2]

57

Magnetic Characterization:

Magnetic measurements were conducted on the polycrystalline samples of
V2A15Ge5 using a Quantum Design MPMS SQUID magnetometer. EDS-analyzed
crystals were ground into powder, which was then sealed in kapton tape and placed into
the magnetometer. The data were collected in the temperature range 3-300 K at 1000 G,
while ﬁeld dependent magnetic measurements, conducted at 3 K, were carried out in
ﬁelds up to i 55000 G. A diamagnetic correction was applied to the data to account for

core diamagnetism. The magnetic contribution from kapton tape was also subtracted for

correction.

Electronic Structure Calculation:

Band structure calculations for V2A15Ge5 were performed using the self-consistent
full-potential linearized augmented plane wave method (LAPW) within density
functional theory (DFT). For the exchange and correlation parts of the potential, the
generalized gradient approximation (GGA) was used. Scalar relativistic corrections were
added and spin-orbital interaction ($01) was incorporated using a second variational

procedure. The calculations were performed using the WIEN97 program.

3-3. Results and Discussion
Synthesis:

V2A15Ge5 was originally found as we were exploring the system V/Co/Al/Ge
using Al as the ﬂux. After excess Al was removed by NaOH solution, VzAlsGes was

obtained in the form of elongated needles, frequently aggregated in bundles. The SEM

58

 

 

image of a typical crystal of V2A15Ge5 is shown in Figure 3-1. Elemental analysis and
single crystal X-ray data revealed that Co did not go into the product and a new ternary
compound V2A15Ge5 formed. This result is surprising since Co is very reactive in liquid
aluminum and several new Co-containing quaternary intermetallics had been discovered
by the Al ﬂux method, such as CeCoAlasig, REzCoAhGez (RE = Sm, Gd, Tb) and
Sm2C0(Co,,Al1.,,)Al.4Ge(,.y.25 Later ﬂux synthesis only incorporating V, Al and Ge was
also successful to produce the title compound as single crystals with the reaction yield
~90%. V2AlsGe5 can also be obtained as a pure phase by combining stoichiometric
amount of V, Al and Ge using the same heating proﬁle; however only microcrystalline
products were obtained. This again shows the effectiveness of Al ﬂux to grow single
crystals of intermetallics.

We have also tried to apply the same conditions to the system V/Al/Si expecting
to obtain Si-analogue of the title compound. Various conditions led to the same product
V3AISi5, which is a variant of the known compound VSiz with A1 and Si atoms sitting on

the same site.26

Crystal Structure of V2A15Ge5:

V2A15Ge5 crystallizes in the orthorhombic space group Cmcm (#63) in what
appears to be a new structure type (Figure 3-2). In this structure a total of seven
crystallographically distinct atomic sites were identiﬁed: V atoms at 8g sites, Al and Ge
atoms occupying three atomic sites: two at 8f and one at 4c. A recognizable building
block in this structure is a distorted pentagonal prismatic column made of stacked

pentagonal prisms deﬁned by Al and Ge atoms with V atoms in the center. The

59

pentagonal prisms have been seen in other compounds such as CeGaMg, in which Ga and
Mg atoms build a three dimensional network with Ce atoms ﬁlling in the pentagonal
channels.27 These pentagonal columns are close-packed hexagonally throughout the
whole structure. To show the structural organization of V2A15Ge5 more clearly, all Al
atoms were removed as shown in Figure 3-3. These pentagons form layered structure
stacking along the c-axis; on the neighboring layer, the pentagons are rotated 180°
pointing to the opposite direction.

In V2A15Ge5 the pentagonal prism is not regular, as described in Figure 3-4A: one
of its pentagonal faces contains one Al(l) atom, two symmetry-related Al(3) atoms and
two symmetry-related Ge(2) atoms, while on the opposite face, correspondingly, there are
one Ge(l) atom, two Ge(3) atoms and two Al(2) atoms. The pentagonal prisms stack
along the a-axis to make a column by sharing the top and bottom faces with two other
prisms. Therefore each V atom is not only surrounded by ﬁve Al atoms and ﬁve Ge
atoms, but also bonded with other two V atoms inside the column forming a vanadium
chain, which is rarely seen in other vanadium-containing intermetallic compounds.28
Another interesting feature of this structure is long—short alternation of the V—V bond
distances in the chain at 2.732(2) A and 2.676(2) A, respectively. Both distances imply
V—V bonding and are comparable to those found in other intermetallic compounds such
as A1V3 and V4A123.29 The difference in bond distance indicates that the V atom is not
located exactly at the center of the prism. In fact, the V atom is closer to the face
containing two Ge atoms and three Al atoms (plane Al(l) — Al(3) — Ge(2), Figure 3-4A).
The local coordination environments of Ge and A1 atoms within 3.0 A are presented in

Figure 3-4B-D and Figure 3-4E-G respectively. Each Al/Ge atom is surrounded by eight

60

atoms, and they exhibit similar coordination geometry. The Ge(l) atom is sitting at the
cross-section of two planes: one plane is deﬁned by atoms Al(1)—V—V—Al(1), the other
plane is perpendicular and deﬁned by atoms Al(2)—Al(3)—Al(3)—Al(2). Ge(l), Al(2) and
V atoms are connected to form a triangular pyramid. The Ge(2) atom shows a similar
geometry as Ge(l), but in this case, those two Al(3) atoms are from different pentagonal
clusters, so there is no bonding between them, as shown in Figure 3-4C. Another
noteworthy point is that the two triangles Al(2).V—Ge(2) connected by Ge(2) atom are
not co-planar. Ge(3) atom also presents a similar coordination environment, but it is
different from those of Ge(l) and Ge(2) atoms, particularly in the triangular pyramid.
One of the A1 atoms in the apex positions is replaced by one Ge(3) atom, with the Ge(3)—
Ge(3) bond distance of 2.6791(15) A. The Al atoms show very similar coordination
geometries to the Ge atoms (shown in Figure 3-4E-G) and will not be discussed here in

detail.

61

 

 

 

Figure 3-2. Structure of V2AlsGe5 in polyhedra of V atoms viewed down the [100]

direction. Large empty circles: Ge; black circles: A1; gray circles: V.

62

l 034
e14

 

0:8}:
018391

O
>
O

{a

 

 

 

671(2) e\/e
o—eO—eo—o
CBC/W G/\©
0330 o o
2588—9

18:
e—li

O
/
e).
O
O

i...
1.
t
SO

1

z._l:

 

Figure 3-3. Structure of V2A15Ge5 viewed down the a-axis with all Al atoms omitted for

clarity.

@8on _< 98 00 mo 35:80:25 c23=€800

A0 - Am .538 2: 5 $53 wcmcﬁno-> 53» £83 00 28 2 mo Bmanoo 5:28 Ecowsgm A< noon—41>? 0.583% .vim Emmi

   

                 
                 

\!
7 \

/‘l

night
/ VI
(-

\
hO/AVO‘ 32
\xeAn
$60 an Izlva
a!
'a.‘ v2
$2 \\\¢l
see a
33

 

2

64

Table 3-4. Bond lengths (A) for V2A15Ge5.

 

 

Bond Distance Bond Distance
V-Ge(1) 2.5890(13) Ge(1)-Al(1) 2.7076(5)
V-Ge(2) x2 2.6447(9) Ge(1)-Al(2) 2.747(2)
V-Ge(3) x2 2.6558(11) Ge(1)-Al(3) 2.701(2)
V-Al(1) 2.701(3) Ge(2)-A1( 1) 2.799(2)
V—Al(2) x2 2.799(2) Ge(2)-Al(2) 2.562(2)
V-Al(3) x2 2.764(2) Ge(2)-Al(2) x2 2.7098(5)
V—V 2.676(2) Ge(2)-Al(3) 2.564(2)
V-V 2.732(2) Ge(2)-Al(3) 2.71 1(2)
Ge(3)-Ge(3) 2.6791 (1 5) Ge(3)-Al(1) 2.662(3)
Al(3)-Al(3) 2.788(4) Ge(3)-Al(2) 2.666(2)
Ge(3)-Al(3) x2 2.7075(5) Ge(3)-Al(2) 2.766(2)

 

Magnetic Properties:

Magnetic measurements were performed on polycrystalline samples ground from
selected single crystals and from direct combination to check the reproducibility of the
result. Originally, the temperature dependent magnetic measurements show
ferromagnetic behavior of the title compound with To around 280 K. However the
magnetization at 3 K is as low as 0.005 in; when the applied ﬁeld is 10000 G. Although
powder X-ray diffraction analysis shows that it is pure phase of VzAlsGes, careful
elemental analysis on the sample detects a very small amount of adventitious iron which
might come from reduction of the crucible cement (a mixture of metal oxides) by Al ﬂux.
To avoid this problem, the reaction was conducted in a one-side—ended crucible as
obtained, so no cement was used. The result conﬁrmed our assumption: as shown in

Figure 3-5, this material was almost temperature independent during the whole

65

temperature range of 3 to 300 K which indicates that VzAlsGes exhibits Pauli
paramagnetic behavior. The very small amount of ferromagnetic fragment at low

temperatures might be from impurities.

Thermal Analysis:

Differential thermal analysis shows that V2A15Ge5 melts congruently with the
melting point 709 °C and it recrystallizes at 641 °C. Figure 3-6 shows the thermal
gravimetric analysis of single crystals of V2AlsGe5 under ﬂowing air up to 1000 °C. This
material was resistant to air oxidation until 520 °C, and no weight gain or loss was
observed. Above this temperature, however, the weight increased gradually and up to
1000 °C a total of 30% of weight was gained. PXRD shows that this material was ﬁnally
fully oxidized to the mixture of oxides: A1203, V205 and GeOz. The oxidation resistance
properties of germanides are generally not so good as those of silicides, which have been

studied extensively as wear and corrosion resist coating materials.

Band Structure Calculations:

Electronic structure calculations using the linearized augmented plane wave
method within density functional theory show metallic behavior of V2A15Ges, as shown
in Figure 3-7. The Fermi level is at 0 eV with a density of states of about 8
states/eV/2[V2A15Ge5] (Figure 3-7A). The Ge p states are in the energy range -7 to 10
eV, and they hybridize with the Al p and s orbitals indicating the covalent interaction
between Ge and Al atoms (Figure 3-7B, C). The DOS diagram of V d-orbitals, Figure 3-

7D, reveals two bands evidently: one is located between —4 to 0 eV and the other is 0 to 3

66

eV, showing that the d-orbitals of V atoms are partially occupied. Comparison between
the occupied partial DOS diagrams implies that the d-states of V atoms (in the range -4 to
0 eV) hybridize more with Ge p states than Al p and 5 states, suggesting that the covalent

interaction between V and Ge is stronger than that of V and Al atoms.

3—4. Conclusions

By studying the ternary system V-Al-Ge in liquid aluminum, we obtained single
crystals of a new phase V2A15Ge5, which is the ﬁrst real ternary compound in this system.
It crystallizes in a new structure type with the pentagonal column composed of Al and Ge
atoms as the unique building block. The V atoms form a long-short alternating chain
residing in the center of the pentagonal column.

Magnetic measurements show that this material is Pauli paramagnetic, which is
consistent with metallic behavior predicted from band structure calculations. The partial
DOS diagrams reveal covalent bonding between Ge and Al atoms; the d-orbitals of V
atoms are partially occupied and the covalent bonding between V and Ge atoms is
stronger than that between V and Al atoms. VzAlsGes is resistant to air oxidation up to

520 °C; up to 1000 °C this material is fully oxidized to the mixture of metal oxides.

67

A)

 

 

 

 

 

 

 

 

0.003 *
E
=3 0.0025
8
91-:
20.002 .‘0.00.. 00 o o. e ...
B
E
3
E0.0015
X
0.001
0 50 100 150 200 250 300
Temperature (K)
B)
150 _
140 E
A 130}
2". :
£10 120':-
0) l-
3 110 E-
100 :
90 :-- -I.A-l.-.1---la-- I-
0 200 400 600 800 1000
Temperature (°C)

Figure 3-5. A)Temperature dependent magnetic behavior of V2A15Ge5 at 1000G.

B). Thermal gravimetric analysis of V2A15Ge5 under air.

68

40

 

351-

30 t Total DOS ~—

25.
20-

 

15-

States/eV/atom

 

, M11

0.8 .

Partial DOS: Gel p —- ' -
Partial DOS: Gel s

0.6 ..

‘ ' f i i i‘. L-I-Z-ramw

0.4 - i;

States/eV/atom

0.2 - ill

    

I
' “I“. ‘-
1 ﬂ ‘
' ~t‘w ‘~--‘.. ."“~sv,‘u-,—lo--a ,

 

0.8 . Partial DOS: All p—
Partial DOS: All s

0.8 -

0.4 -

0.2 .

States/eV/atom

 

 

Partial DOS: V d — '

 

States/eV/atom

 

o I? I “M

 

 

‘

-1'o -5 o 5 1'0

Energy (eV)

Figure 3-7. Density functional theory calculated for V2A15Ge5.

69

References:

 

l a) Chen, X. Z.; Sportouch, S.; Sieve, B.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.;
Patschke, R.; Kanatzidis, M. G. Chem. Mater. 1998, 10, 3202. b) Sieve, B.; Chen, X. Z.;
Henning, R.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.; Schultz, A. J .; Kanatzidis, M.
G. J. Am. Chem. Soc. 2001, 123, 7040. c) Kanatzidis, M. G.; PGttgen, R.; Jeitschko, W.
Angew. Chem. Int. Edit. 2005, 44, 6996.

2 a) Thiede, v. M. T.; Fehrmann, B.; Jeitschko, w. z. Anorg. Allg. Chem. 1999, 625,
1417. b) Fehrmann, B.; Jeitschko, W. J. Alloys Compd. 2000, 298, 153.

3 Sieve, B.; Chen, X. Z.; Cowen, J. A.; Larson, P.; Mahanti, S. D.; Kanatzidis, M. G.
Chem. Mater. 1999, 11, 2451.

4 Sieve, B.; Sportouch, S.; Chen, X. Z.; Cowan, J. A.; Brazis, P.; Kannewurf, C. R.;
Papaeﬁhymiou, V.; Kanatzidis, M. G. Chem. Mater. 2001, 13, 273.

5 Latturner, S. E.; Bilc, D.; Ireland, J. R.; Kannewurf, C. R.; Mahanti, S. D.; Kanatzidis,
M. G. J. Solid State Chem. 2003, 1 70, 48.

6 Dubenko, I. s.; Evdokimov, A. A.; Ionov, v. M. Kristallografiya 1987,32, 347.
7 Kranenberg, C.; Mewis, A. Z. Anorg. Allg. Chem. 2000, 626, 1448.
8 Schobinger-Papamantellos, P.; Hulliger, F. J. Less-Common Met. 1989, I46, 327.

9 Chen, X. Z.; Sieve, B.; Henning, R.; Schultz, A. J.; Brazis, P.; Kannewurf, C. R.;
Cowen, J. A.; Crosby, R.; Kanatzidis, M. G. Angew. Chem. Int. Edit. 1999, 38, 693.

'0 Geman, N. V.; Zavodnik, V. E.; Yanson, T. I.; Zarechnyuk, O. S. Kristallograﬁya
1989, 34, 738.

” Richter, K.W.; Prots, Yu.; Grin, Yu. N. Z. Anorg. Allg. Chem. 2004, 630, 417.
‘2 Holleck, H.; Benesovsky, F.; Laube, E.; Nowotny, H. Monatsh. Chem. 1962, 93, 1075.

'3 Mohan Rao, P. V.; Satyanarayana Murthy, K.; Suryanarayana, S. V.; Nagender Naidu,
S.V. Phys. Status Solidi, A 1992, 133, 231.

'4 Frank, F. C.; Kasper, J. 5. Acta Cryst. 1996, A52, 125.
‘5 Pearson, w. B. z. Kristallogr. 1985, 171,23.

‘6 Ul-Haq, 1.; Booth, J. G. J. Magn. Magn. Mater. 1986, 62,256.

70

 

'7 Holleck, H.; Benesovsky, F.; Novotny, H. Monatsh. Chem. 1963, 94, 477.
'8 Luo, H. LP.; Vielhaber, E.; Corenzwit, E. z. Phys. 1970, 230, 443.
'9 Kodess, B. N. Phys. Status Solidi, A 1971, 4, 109.

20 A15 is the Strukturbericht designation of the structure type with the space group Pm-
3n. The prototype compound is Cr3Si.

2‘ Smith, T. F. Solid State Comm. 1971, 9, 903.
22 Saint, version 4; Simens Analytical X-ray Instruments, Inc., Madison, WI.
23 SADABS, Sheldrick, G. M.; University of Géttingen, Gdttingen, Germany.

24 Sheldrick, G. M. 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

25 Sieve, B. Ph. D. Dissertation, Michigan State University, 2002.

26 Structural information of V3A1Si5: hexagonal, space group: P6(2)22 #180), a =
4.6173(8) A, c = 6.4445(15) A, v = 118.99(4) A3, 2 = 1, Dc = 4.857 g cm' , u = 9.715
mm"; index range —7<=h<=7, -7<=k<=7, -10<=l<=8; total reﬂections 1153, independent
reﬂections 203 (Rim = 3.48%), and 11 parameters; R1 = 2.59, sz = 6.63, GOF = 1.131.
The 6j site M (occupied by Si in VSiz) was reﬁned to be mixed occupied by Al and Si: Al
17%, Si 83%. What ﬁrst brought to attention was the bond distance of M-M, which
ranges from 2.5145(5) A to 2.6922(8) A, much longer than the bond distance of Si-Si
(2.3~2.5 A), while shorter than that of normal bond distance of Al-Al (2.7~2.9 A).
Moreover, this mixed occupany of Al and Si lowers the R value from 2.68% (fully
occupied by Si) to 2.59%; the reﬁnement result is consistent with the elemental analysis
by EDS (V : Al : Si 3 : 1 : 5), which also conﬁrms this reﬁnement.

27 Krali, R.; Pottgen, R. Chem. Mater. 2003, 15, 2998.
2“ Kuzma, Y. B.; Starodub, P. K.; Izv. Akad. Nauk SSSR, Neorg. Mater. 1973, 9,337.

29 (a) Kodess, B. N. Phys. Status Solidi A 1971, 4A, 109. (b) Ray, A. E.; Smith, J. F. Acta
Cryst. 1960, 13, 876.

71

CHAPTER FOUR

Structurally Complex Cobalt Intermetallics Grown from Liquid

Aluminum: C019Al45Silo-x (X = 0.13) & C05A114Siz

4-1. Introduction

Aluminum matrix alloys are technologically useful materials because of their light

weight, special mechanical strength, high thermal and electrical conductivity, etc.1 For
example, aluminum metal-matrix composites (Al-MMCS) have emerged as a critical
class of material for applications in lightweight automotive structures, forgings for
suspension and drive trains as well as aerospace development products.2 The addition of
transition metals and silicon into an Al matrix contributes to the desirable properties of
these materials such as high temperature wear and corrosion resistant coatings3 and even
thermoelectric energy conversion4. Moreover, a large number of quasicrystal
approximants occur in the Al-rich region of the Al-TM (TM = transition metal) and
related ternary systems.5 One well known example is or-MnAlSi which is a prototype of
approximants presenting Mackay—type clusters.6 Since the quasicrystalline phases are
often found with a composition in the vicinity of crystalline approximant phases, the
detection of new approximants often helps to ﬁnd new quasicrystals and aid in the

understanding of the local order of the latter compounds.

In the ternary phase diagram of Co-Al-Si system, very little information is known

about the structural chemistry of ternary compounds. Nowotny and his coworkers
identiﬁed two ternary phases Co3Al3Si4 and CozAlSi2;7 German has had a tentative study

on this ternary system and found ﬁve ternary phases.8 He conﬁrmed the existence of

72

Co3A13Si4 and CozAlSiz and labeled them as or and [3 phase. The other three phases were
designated as y, 5 and e and their crystal structures were not determined. In 2005 Grin
and his coworkers reported the structure of another ternary compound CoaAl7+xSi2-x,
which shows a covalently bonded Al/Si 3D polyanion presenting ionic interactions with
Co atoms.9 C04A17+,,Sip_.x was assigned by Grin et a]. as x phase. In this chapter we
present two new members in the Al-rich region — ColoAhsSilo.x (x = 0.13) and

Co Al Si grown from aluminum ﬂux.l0 Among them, ColoA145Silo.x (x = 0.13) was
5 14 2

designated by German as y phase however he could not characterize its crystal structure.8
We were able to grow single crystals of these phases by Al ﬂux and determine their
crystal structures by single crystal X-ray diffraction analysis. Both phases exhibit
surprisingly large unit cell parameters and complex structures; CoSAlMSi2 shows
remarkable thermal oxidation resistance in air up to 1000°C. Herein the synthesis, crystal

structure, thermal stability and magnetic properties of the title compounds are reported.

4-2. Experimental Section
Reagents:

The following reagents were used as obtained without further puriﬁcation: Co (-
325 mesh, Cerac, 99.9%), V (-325 mesh, Cerac, 99.7%), Al pellets (Cerac, 99.99%), Si (-

325 mesh amorphous powder, Cerac, 99.999%).

Synthesis:
ColgAhssilh (x=0.13): In a nitrogen-ﬁlled glove box, 1 mmol vanadium (0.051 g),

2 mmol cobalt (0.118 g), 10 mmol Al (0.270 g) and 5 mmol Si (0.14 g) were combined in an

73

alumina crucible. The crucible was put into a silica tube, which was sealed under a
vacuum of 10“11 Torr. The sample was then heated to 1000°C in 15 h, maintained at this
temperature for 5 h, followed by cooling to 850°C in 2 h. It was annealed at 850°C for 3
d, cooled to 200°C in 36 h, and ﬁnally cooled down to 50°C in 2 h.

CoSAIMSizz In a nitrogen-ﬁlled glove box, 3 mmol cobalt (0.177 g), 15 mmol Al
(0.405 g) and 5 mmol Si (0.14 g) were combined in an alumina crucible. The crucible
was put into a silica tube, which was sealed under a vacuum of 104 Torr. The sample was
then heated to 1000°C in 15 h, maintained at this temperature for 5 h, followed by
cooling to 850°C in 2 h. It was annealed at 850°C for 3 d, and then cooled to 200°C in 36
h.

The excess aluminum was removed by soaking the crucible in aqueous 5M NaOH
solution overnight. The resulting crystalline product was rinsed with water and dried
with acetone. The yield for CO|9Al4ssi]0-x (x = 0.13) was ~70% based on the initial
amount of cobalt used. CoSAlMSi2 was obtained as a pure phase with the yield 90%.
Single crystals were selected for elemental analysis, X-ray diffraction, thermal
gravimetric analysis and magnetic measurements. When we tried to produce the phase
ColgAlriSSim.x by arc-melting the stoichiometric amount of the elements (ratio 1.9 : 4.5 :
1.0), the phase C05A1l 4812 formed instead. Arc-melting reactions were run under an Ar
atmosphere on a water-cooled copper plate and the pellets were ﬂipped and remelted

three times to ensure a homogeneous distribution of the reactants.

74

Elemental Analysis:

Selected crystals were ﬁxed on a SEM stub using carbon tape. Chemical
composition of the products was determined by Energy Dispersive Spectroscopy (EDS)
performed on a scanning electron microscope (SEM) JEOL JSM-35C equipped with
NORAN EDS detector. Data were acquired by applying a 25 kV accelerating voltage
with an accumulation time of 30 s. The atomic ratio in the compounds C019A145Si10-x (x
= 0.13) and CoSAlMSi2 were determined to be 18.4 : 41.5 : 10 (Co : A1 : Si) and 5.27 :
13.6 : 2 (Co : Al : Si) respectively, in good agreement with the results derived from the

single crystal X-ray analysis.

X-ray Crystallography:

Single crystal X-ray diffraction data of C019A14_r,Si10.x (x = 0.13) and CosAlMSi2
were collected at room temperature on a Bruker AXS SMART CCD X-ray diffractometer
with graphite monochromatized Mo K01 (A = 0.71073 A) radiation. Data processing was
performed with the SAINTPLUS software package.ll The face-indexing procedure was
used to analytically correct for absorption; and an empirical absorption correction was
applied to the data using the SADABS program.'2 The structure was solved using direct
methods and reﬁned with the SHELXTL package programs.l3 Since Al and Si could not
be distinguished directly from the collected X-ray scattering data, the assignment of Al
and Si positions were made based on the bond distances. All atomic positions were
reﬁned anisotropically. Tables 4-1 shows the crystallographic reﬁnement data for

C019Al45Silo_x (x = 0.13) and CoSAlMSiZ. Tables 4-2, 4-3, 4-4, 4-5, 4-6 and 4-7 list

75

fractional atomic positions, equivalent isotropic thermal displacement parameters and

selected bond distances for ColgA145Silo.x (x = 0.13) and CoSAlMSi2 respectively.

Table 4-1. Selected crystal data and structure reﬁnement details for ColoA145Silo.x (x =

0.13) and CO5A114Slz.

 

 

Empirical formula ColgAhSSilo.x (x = 0.13) C05A114Si2
Formula weight 2614.67 728.55
Crystal system Monoclinic Orthorhombic
Space group C2/c (#15) ana (#62)

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size

Theta range for data collection
Reﬂections collected
Independent reﬂections
Completeness to theta = 37.00°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F2

Final R indices [I>25igma(l)]

R indices (all data)

Largest diff. peak and hole

2 2
R1 = z:(lI'-:ol‘chl)/ZlFoli WRZ : [2[W(Fo ’ 1:c l/ [2(wlFol

a = 19.991(2) A

b = 19.143(2) A

c = 12.8137(2) A

[3 = 123.583(2)°
4085.1(8) A3

4

4.251 Mg/m3

8.773 mrn'l

4952

0.23 x 0.13 x 0.09 mm3
1.62 to 28.10°

21558

4644 [R(int) = 0.0429]
92.9 %

a =13.8948(19)A
b = 23.039(3) A
c = 7.3397(10) A

2349.6(6) A3

8

4.119 Mg/rn3

8.130 mm"

2760

0.22 x 0.18 x 0.26 mm3
1.77 to 28.14°

24268

2789 [R(int) = 0.0328]
94.7 %

Full-matrix least-squares on F2

4644 / 0 / 338
1.030
R1 = 0.0326
sz = 0.0705
R. = 0.0446
sz = 0.0746
-3
1.588 and —0.979 e.A
2 2 1/2

)1

76

2789 / 0 / 200

1.066

R1 = 0.0268

sz = 0.0640

R. = 0.0316

sz = 0.0686

3.007 and -1.111e.A'3

Table 4-2. Atomic coordinates (x 104) and equivalent isotropic displacement parameters

(Azx 103) for ColeA1458ilo.x (x = 0.13). Estimated standard deviations are in parentheses.

 

 

Wyckoff site x y z U(eq)
Co(l) 8f 752(1) 4023(1) 5417(1) 7(1)
Co(2) 8f 1790(1) 5896(1) 8002(1) 5(1)
Co(3) 8f 6144(1) 9017(1) 9009(1) 5(1)
Co(4) 8f 3150(1) 7111(1) 6806(1) 7(1)
Co(5) 8f 3158(1) 4800(1) 6845(1) 8(1)
Co(6) 8f 4223(1) 5238(1) 9509(1) 8(1)
Co(7) 8f 4343(1) 2922(1) 9487(1) 10(1)
Co(8) 4e 5000 5969(1) 7500 7(1)
Co(9) 8f 1374(1) 7814(1) 9263(1) 5(1)
Co(lO) 8f 3327(1) 4063(1) 8855(1) 10(1)
Al(l) 8f 2163(1) 7874(1) 6744(1) 7(1)
Al(2) 8f 4651(1) 7124(1) 8210(1) 8(1)
Al(3) 8f -300(l) 4113(1) 3249(1) 7(1)
Al(4) 8f 2204(1) 3966(1) 6668(1) 6(1)
Al(5) 8f 1290(1) 7129(1) 7360(1) 8(1)
Al(6) 8f 1244(1) 3321(1) 7368(1) 8(1)
Al(7) 8f 2892(1) 6026(1) 10162(1) 9(1)
Al(8) 8f 3833(1) 6653(1) 5789(1) 7(1)
Al(9) 8f 2937(1) 6781(1) 8451(1) 8(1)
Al(10) 8f 3816(1) 5201(1) 5826(1) 8(1)
Al(ll) 8f 4659(1) 4812(1) 8210(1) 7(1)
Al(12) 8f 1190(1) 3510(1) 4091(1) 8(1)
Al(l3) 8f 1209(1) 4736(1) 7299(1) 8(1)
Al(14) 8f 4453(1) 6320(1) 10624(1) 13(1)
Al(15) 8f 1705(1) 4804(1) 4757(1) 15(1)
Al(l6) 8f -473(1) 3295(1) 5210(1) 10(1)
Al(17) 8f 3877(1) 3600(1) 7569(1) 9(1)
Al(18) 8f 6860(1) 7887(1) 9747(1) 16(1)
Al(19) 8f 477(1) 5325(1) 4826(1) 10(1)
Al(20) 4e 0 7681(1) 7500 16(1)

77

Table 4-2. (Continued) Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for ColoAhSSilo.x (x = 0.13). Estimated standard deviations are in

 

 

parentheses.

Wyckoff site x y z U(eq)
Al(21) 8f 2047(1) 6031(1) 6316(1) 12(1)
Al(22) 8f 619(1) 2684(1) 5059(1) 13(1)
Al(23) 8f 3722(1) 5964(1) 7587(1) 7(1)
Si(l) 8f 7485(1) 9093(1) 9607(1) 25(1)
81(2) 4e 5000 8339(1) 7500 6(1)
Si(3) 8f 2516(1) 7828(1) 9157(1) 6(1)
Si(4) 4e 5000 9678(1) 7500 6(1)
Si(5) 8f 2814(1) 5184(1) 8259(1) 8(1)
Si(6) 8f 5249(1) 5928(1) 9655(1) 7(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

Table 4-3. Anisotropic displacement parameters (Azx 103) for ColoA145Silo-,, (x = 0.13).

 

 

ull u22 u33 u23 ul3 ul2
Co(l) 10(1) 6(1) 7(1) 0(1) 6(1) 1(1)
Co(2) 7(1) 6(1) 6(1) 0(1) 4(1) 0(1)
Co(3) 5(1) 5(1) 5(1) 0(1) 2(1) 0(1)
Co(4) 7(1) 8(1) 9(1) 2(1) 5(1) 1(1)
Co(5) 9(1) 8(1) 10(1) -2(1) 7(1) —2(1)
Co(6) 8(1) 8(1) 9(1) 1(1) 5(1) -l(1)
Co(7) 12(1) 7(1) 6(1) 0(1) 2(1) 4(1)
Co(8) 7(1) 7(1) 6(1) 0 4(1) 0
Co(9) 4(1) 5(1) 5(1) 1(1) 3(1) 1(1)
Co(lO) 10(1) 10(1) 10(1) 0(1) 6(1) 1(1)
Al(l) 7(1) 6(1) 9(1) 0(1) 6(1) 1(1)
Al(2) 8(1) 8(1) 7(1) 0(1) 3(1) 0(1)
Al(3) 6(1) 8(1) 5(1) -1(1) 2(1) 0(1)

78

Table 4-3. (Continued) Anisotropic displacement parameters (Azx 103) for ColgAlrsSim.x

 

 

(x=0.13).

ull u22 (J33 (J23 ul3 (112
Al(4) 6(1) 6(1) 5(1) 1(1) 4(1) -l(l)
Al(5) 8(1) 7(1) 6(1) 1(1) 4(1) -l(l)
Al(6) 8(1) 7(1) 8(1) 1(1) 5(1) 0(1)
Al(7) 12(1) 8(1) 7(1) 1(1) 4(1) -2(1)
Al(8) 8(1) 9(1) 6(1) -1(1) 5(1) 0(1)
Al(9) 10(1) 8(1) 8(1) 0(1) 6(1) 1(1)
Al(10) 10(1) 8(1) 8(1) 0(1) 6(1) 0(1)
Al(ll) 8(1) 7(1) 6(1) 1(1) 3(1) 0(1)
Al(12) 10(1) 7(1) 8(1) 1(1) 6(1) 1(1)
Al(13) 11(1) 6(1) 8(1) -l(1) 6(1) -1(1)
Al(14) 10(1) 12(1) 13(1) -4(1) 5(1) 4(1)
Al(15) 22(1) 8(1) 12(1) 3(1) 8(1) 7(1)
Al(16) 11(1) 11(1) 9(1) -2(1) 6(1) 3(1)
Al(17) 9(1) 8(1) 8(1) 0(1) 4(1) 0(1)
Al(18) 25(1) 8(1) 8(1) 1(1) 5(1) 7(1)
Al(19) 11(1) 10(1) 13(1) 1(1) 9(1) -2(1)
Al(20) 8(1) 16(1) 14(1) 0 -1(1) 0
Al(21) 12(1) 15(1) 6(1) 0(1) 4(1) -2(1)
Al(22) 18(1) 11(1) 23(1) -3(1) 18(1) 0(1)
Al(23) 8(1) 7(1) 9(1) 0(1) 6(1) 0(1)
Si(l) 11(1) 34(1) 30(1) 2(1) 12(1) 0(1)
Si(2) 6(1) 8(1) 5(1) 0 3(1) 0
Si(3) 7(1) 5(1) 6(1) 0(1) 4(1) 1(1)
Si(4) 7(1) 6(1) 6(1) 0 4(1) 0
Si(5) 8(1) 7(1) 8(1) -l(1) 5(1) 1(1)
Si(6) 8(1) 7(1) 9(1) 0(1) 6(1) -1(1)

 

The anisotropic displacement factor exponent takes the form:

-21t2[h2a"‘2Ul l+. . .+2hka*b*U12]

Table 4-4. Selected bond distances for ColgAlassilo-x (x = 0.13).

 

 

 

 

 

Bond Length, A Bond Length, A
Co(3)-81(1) 2.3498(16) Al(5)-Co(7) 2.4880(13)
Co(3)-Al(6) 2.5884(13) Al(5)-Al(17) 2.8434(17)
Co(5)-Al(1 1) 2.5013(14) Si(2)-Si(4) 2.562(2)
Co(5)-Co(6) 2.9744(8) Si(3)-Co(9) 2.3587(12)
Al(2)-Al(23) 2.7160(17) Si(3)-Al(7) 2.6517(17)
Alb-Si(6) 2.7666(17) Si(4)-Co(3) 2.3825(10)

Table 4-5. Selected bond distances for CosAlMSiz.

Bond Length, A Bond Length, A
Co(l )-Al(7) 2.4518(12) Si(l)—Al(9) 2.6347(15)
Co(2)-Al(1 1) 2.5148(13) Si(2)-Al(14) 2.5721(16)
Co(3)-Si(2) 2.3686(11) Si(3)-Co(6) 2.6895(19)
Co(4)-Al(14) 2.4538(12) Al(1)-Al(12) 2.9093(16)
Co(5)-81(1) 2.3715(16) Al(3)-Si(2) 2.6162(15)
Co(5)-Co(6) 2.8075(1 1) Al(5)-Co(4) 2.4934(12)

 

80

Table 4-6. Atomic coordinates (x 104) and equivalent isotropic displacement parameters

(Azx 103) for CosAlMSiz. Estimated standard deviations are in parentheses.

 

 

Wyckoff site x y z U(eq)
Co(l) 8d 7853(1) 3393(1) 956(1) 8(1)
Co(2) 8d 5213(1) 6648(1) 1010(1) 8(1)
Co(3) 8d 4393(1) 5755(1) 5767(1) 5(1)
Co(4) 8d 6560(1) 5021(1) -91 8(1) 8(1)
Co(5) 4c 5460(1) 2500 3877(1) 8(1)
Co(6) 4c 7092(1) 2500 6132(1) 12(1)
Si(l) 4c 8753(1) 2500 1136(2) 9(1)
Si(2) 8d 4360(2) 4720(1) 5850(1) 7( 1)
Si(3) 4c 4769(1) 7500 2858(2) 15(1)
Al( 1) 4c 6942(1) 2500 2206(2) 10(1)
Al(2) 8d 6201(1) 5796(1) 1113(2) 6(1)
Al(3) 8d 7672(1) 4218(1) -968(1) 6(1)
Al(4) 8d 4352(1) 5818(1) 2223(2) 8(1)
Al(5) 8d 2984(1) 5256(1) 4086(2) 9(1)
Al(6) 4c 51 10(1) 7500 -778(2) 5(1)
Al(7) 8d 8365(1) 3134(1) 4048(2) 10(1)
Al(8) 8d 4655(1) 5307(1) 8726(2) 11(1)
Al(9) 8d 9607(1) 3415(1) -287(2) 12(1)
Al(10) 8d 523(1) 3104(1) 1052(2) 8(1)
Al(l 1) 8d 8225(1) 3133(1) -2400(2) 14(1)
Al(12) 8d 4395(1) 6352(1) -1920(2) 13(1)
Al(13) 8d 3495(1) 6616(1) 4896(2) 17(1)
Al(14) 8d 8918(1) 4173(1) 2089(2) 11(1)
Al(15) 8d 2900(1) 5708(1) 7494(2) 16( 1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

81

Table 4-7. Anisotropic displacement parameters (Azx 103) for CosAlMSiz.

 

 

ull u22 u33 u23 U13 ul2
Co(l) 10(1) 5(1) 10(1) 0(1) -1(1) 0(1)
Co(2) 8(1) 7(1) 10(1) 2(1) -1(1) -l(l)
Co(3) 5(1) 4(1) 7(1) 0(1) 0(1) 0(1)
Co(4) 7(1) 9(1) 9(1) -2(1) 1(1) -2(1)
Co(5) 8(1) 5(1) 10(1) 0 1(1) 0
Co(6) 11(1) 8(1) 15(1) 0 -2(1) 0
Si(l) 8(1) 6(1) 13(1) 0 2(1) 0
Si(2) 7(1) 5(1) 9(1) 0(1) 0(1) 1(1)
Si(3) 21(1) 9(1) 16(1) 0 -3(1) 0
Al(l) 8(1) 7(1) 15(1) 0 1(1) 0
Al(2) 5(1) 4(1) 8(1) -l(1) 0(1) 1(1)
Al(3) 8(1) 4(1) 6(1) 1(1) 0(1) 1(1)
Al(4) 8(1) 6(1) 10(1) 0(1) 0(1) -2(1)
Al(5) 7(1) 10(1) 9(1) 1(1) 1(1) -1(1)
Al(6) 7(1) 1(1) 7(1) 0 0(1) 0
Al(7) 10(1) 10(1) 10(1) -2(1) -1(1) 2(1)
Al(8) 14(1) 11(1) 9(1) 4(1) -1(1) -1(1)
Al(9) 13(1) 7(1) 15(1) -2(1) 1(1) -l(l)
Al(10) 5(1) 12(1) 9(1) -1(1) 0(1) 0(1)
Al(ll) 11(1) 18(1) 12(1) -4(1) 2(1) -6(1)
Al(12) 20(1) 8(1) 12(1) -2(1) -3(1) 0(1)
Al(13) 15(1) 15(1) 21(1) 0(1) -4(1) 5(1)
Al(14) 6(1) 9(1) 18(1) -5(1) 1(1) -1(1)
Al(15) 9(1) 19(1) 20(1) -9(1) 7(1) -5(1)

The anisotropic displacement factor exponent takes the form:

~2112[h2a*2U l+...+2hka*b*U12]

82

Thermal Analysis:

Thermal gravimetric analysis (TGA) was performed on a Shimadzu TGA-SO
thermal gravimetric analyzer. The sample was heated to 900 °C at a rate of 10 °C/min
under ﬂowing air, held at 900 °C for 10 min, and then cooled down to room temperature

at the same rate.

Magnetic Characterization:

Magnetic susceptibility measurements were conducted on the polycrystalline
samples of C019A14SSilo-x (x = 0.13) and CoSAlMSi2 using a Quantum Design MPMS
SQUID magnetometer. EDS-analyzed crystals were ground into powder, which was
sealed in kapton tape and placed into the magnetometer. The data were collected in the
temperature range 3-300 K at 1000 G, while ﬁeld dependent magnetic measurements,
conducted at 5 K, were carried out in ﬁelds up to i 55000 G. The magnetic contribution

from kapton tape was subtracted for correction.

4-3. Results and Discussion

Synthesis:

C019A1458110-x (x = 0.13) was found in an effort to explore the system V-Co-Al-Si using
Al as the ﬂux. Vanadium did not incorporate into the product and a new ternary phase
C019A145Si 10-x (x = 0.13) formed instead. The yield of the reaction was 70% based on the
initial amount of cobalt element used. The side products were mainly unreacted
vanadium and recrystallized Si. Several attempts to make this compound with only Co
and Si in Al ﬂux or by direct combination with an arc-welder were not successful. These

attempts resulted in C05A114Si2 which is also described here.

83

Crystal Structure of C019A145Si10.x (x = 0.13):

C019A1458110-x (x = 0.13) crystallizes in the monoclinic space group C2/c, and its
large unit cell contains 10 Co, 23 Al and 6 Si independent sites with a total of 296 atoms.
All the atomic sites are ordered except for Si(l) which is only 90% occupied.

The very high structural complexity presents difﬁculties in describing this
structure in a straightforward fashion. A careful examination reveals that no high-
symmetry ﬁrst-neighbor polyhedron can be found around any atomic site. One way to
understand the structure is to inspect the Co(3), Co(8), Co(9), Si(3) and Si(6)-based
polyhedra which are interpenetrating into each other (Figure 4-1). The Si-based
framework, shown in Figure 4-2, is composed of alternating Si(3) and Si(6)-centered
slabs along the a-axis. In order to simplify the description and obtain a better view of the
structure, only the bonds between the center atoms (Si(3) and Si(6) atoms) and the
surrounding atoms are shown. On the local scale two types of structural fragments are
noteworthy. In the ﬁrst fragment, the Si(3) atom is bonded to another Si(3) atom forming
a Si dimer (Figure 4-3A), with the Si(3)-Si(3) bond distance of 2.528(2) A. Although this
Si-Si bond distance is longer than the sum of the metallic radii of silicon, it is comparable
to those in the other intermetallics (2.523 A in Re24Alszilz,M 2.518 A in FeAIZSiIS).
These two Si(3) atoms, together with two Co(9) atoms, form a CozSiz rhombus. This
rhombus is surrounded by Al atoms, some serving to bridge the Si and Co atoms (Al(5),
Al(7), Al(9), Al(12) and Al(17)), while others serving as links to the rest of the structure (

Figure 4-3A).

84

v13

4

(V11

1'44 .
{1‘14 Ix.
' . ' ‘ ‘1 ‘V
"A\1\v§\ 1“ VHW‘A
\A v
A 334‘

9.79»
2.11:"
n: A

.1 ‘7 IVA 'W;

 

85

0.13) in polyhedral representation viewed down the b-axis.

Figure 4-1. The structure of C0|9A14581|0_x (x

.00 SEE >an am ”286

base ME 886 62m 3% ooeeaooézm Ba 55 Co 8838 0263658 “$3 1 5 3.52228 Co 2385 .3 Beam

a/n o\.h|.wd/el>l(l4 .11.. hi ’3
.6 9 Q. 0 ll. .0. 1/70
0V6“ .. 9 9 o o 3 / 9
./. _ we! 48% e/.\
arc/o / e \. o/o ~94.“

 

 

 
 
 
  
  

 
  
 

(A

   

I

Al 10)

87

Al(8)

Al(14)

Al(2)

0.13): (A) Two Si(3)-centered polyhedra joined via bridging Co(9) atoms to form a dimer.

Si(3)-Si(3) distance: 2.528(2) A. (B) The connectivity mode of Si(6)-centered dimers. (C) Extended organization of Si(6)-centered

Figure 4-3. Structure of C019Al4SSi|0-x (x

polyhedra forming a strongly correlated layer.

25 393-5 by 380550 2:on 500 can €00 .

.v—HOBDFEHM

58 no 36$ Beanbag 2: ”GS u as $043.28 ea 6.36am .3. 65mm

5a 55

 

 

88

.ESw A800 mo 55:50:25 532658
2:. Rd 8.6233 8533-850 .«o 258 335558 55. ADV @5520 “550950-950 mo 358 335258 2:. Amy .386
65 :38 8365 88:38 Coo as $8 .58 Co 6588-3... 883 2: 2V .3 _ .o n 5 3.52228 Co 6535 .3. 65w:

k7 2:2

as? 1.13.4
V482");
as. . \

 
  

 

    

 

 

 

 

 

3:2 in 3:2

89

Figure 4-3B depicts the Si(6)-centered polyhedra which conjoin with each other in
similar but not identical fashion to Si(3). In this fragment, two Si(6) and two Co(6)
atoms also form a C028i2 rhombus, however no Si-Si bonding exits with a Si(6)-Si(6)
distance at 3.929 A. The Si(6)-based dimeric polyhedra are linked to each other by
bridging Co(8) atoms; when viewed down the a-axis, these Si(6)-centered polyhedra
form inﬁnite puckered layers, Figure 4-3C.

The overall organization of the structure of C019A1458i10.x (x = 0.13) can be
described as layers deﬁned by Co(3), Co(8) and Co(9) polyhedra and linked with each
other by a Si(3) and Si(6)-based wire framework, Figure 4-4. The slabs composed of
Co(3), Co(8) and Co(9) polyhedra, when viewed down the [001] direction, stack in A-A’
fashion where layer A’ is shifted relative to layer A along the a direction by a/Z (Figure
4-5A). The center of the layer is occupied by Co(9) polyhedra form a chain by vertex-
linking. Both ends of the layer are capped by Co(3) and Co(8) polyhedra alternatively.
Two Co(3) polyhedra are conjoined by bridging Si(2) and Si(4) atoms forming a C028i2
rhombus, Figure 4-SB. As seen in Figure 4—5C, the connectivity mode of Co(9)
polyhedra is very similar to that of Si(6): the C028i2 rhombi composed of two Co(9) and
Si(3) atoms are connected by bridging Al(20) atoms. Each Co(8) atom, is sitting on a site
of icosahedral geometry which contains two Si(6), Al(8), Al(10), Al(ll), Al(23) and

Al(2) atoms (Figure 4-5D).

9O

Crystal Structure of C0 5A], 451'):

CoSAlMSi2 has a different structure type which appears to be new and is also
highly elaborate as in CO|9A145Sl|Q-x (x = 0.13). This compound adopts the orthorhombic
space group ana, which has 6 Co, 3 Si and 15 A1 independent sites with a total of 168
atoms in the unit cell. Although ColgAl458ilo.x (x = 0.13) and CoSAlMSi2 exhibit totally
different structure types, they show notable similarities in the way the various polyhedra
connect to each other. Both structural frameworks derive from the vertex-linkage of
polyhedra with the C028i2 rhombus as the common feature in the structure.

Figure 4-6 depicts this structure in polyhedra of Al(l), Al(ll), Al(15), Co(2),
Co(3) and Co(5) atoms. One layer is based on Co polyhedra and the second is based on
Al polyhedra. To have a clearer view of the layered structure character of C05A114Si2,
only the polyhedra of Co(2) and Co(3) are shown in Figure 4-7. This ﬁgure ShOWS
dimeric units linked to each other by sharing Al(4), Al(9) and Al( 12) atoms along the c
direction. The local structure around these units is shown in Figure 4-8. The manner in
which the Co clusters connect with each other is very similar: two Co(2) polyhedra are
connected by Si(3) and Al(6) atom; while two Co(3) polyhedra are bridged by two Si(2)
atoms. Such linking patterns show resemblance to those of the Co(3) and Co(9) atoms
formed in the structure of ColgAl458i10.x (x = 0.13), which is an indication of the close
structural relation of these two phases. Co(5) atoms, shown in Figure 4-8C, link the
Co(2)-based dimers to form a “ribbon” running along the c-axis, using two Al(12) and
one Al(6) atom as bridges.

The Al-based layer is shown in Figure 4-9. Interestingly, the Al(l 1) and Al(15)-

based polyhedra form a similar layered structure as Co(2) and Co(3). However, a closer

91

view of Al(15)-based polyhedra reveals that they are different from those described
before. As shown in Figure 4-10A, Al(15) polyhedra are linked by Al(5) atoms to form
zig—zag chains along the c axis. The Al(ll) polyhedra are condensed into dimers by
Al(1l)-Al(11) bonding (distance 2.9280(24) A) with Co(6) atoms acting as bridging
atoms, Figure 4-lOB. This unit is surrounded by more Al atoms to form a cluster.
Furthermore, Al(1) polyhedra are connected to this cluster to form a larger trimer-based
cluster.

The local coordination geometries of Co(3), Co(8) and Co(lO) in ColgAI4SSim.x (x
= 0.13), Co(2), Co(3) and Al(l 1) in CoSAlMSi2 are presented in Figure 4-11A & B. As
we can see, the Co atoms are surrounded in a cage composed of A1 and Si atoms. Such
disposition is common in a variety of Al/Ga-rich binaries such as VgGaM,l6 WAllzl7 and
CozAlg18 which also exhibit large and complex structures. Compared to another ternary
intermetallic Co4Al7+,,Si2.x,9 the only signiﬁcant connection with the title compounds is
the trigonal prismatic coordination geometries of Co atoms. Another interesting feature
about the structure of C019A1458i10.x and CosAlMSi2 is that, although there is no obvious
evidence showing that they are approximants themselves, they show structural features
that are reminiscent to some known quasicrystal approximants. For example, Figure 4-
11C shows the ﬁrst-neighbor coordination polyhedra of transition metals in the Mackay-
type approximants MDA15,19 on-FeAlSi20 and o-Co4All3.21 It is evident that for most
polyhedra in the structure of ColgAl4SSilo.x and CoSAlMSiZ, we can ﬁnd corresponding
ones in the approximants, except that they are very distorted from ideal. In this sense,

we might be able to consider these two phases as distorted Mackay-type approximants.

92

.oU ”SEGA—om bﬁw n_< ”83:th xoﬂm .253» 05 :38 5388852 3.60533 8 rm:

_

<moo co oﬁoém .3. 2%:

 

93

 

 

 

94

‘S/

 

 

 

 

Figure 4—7. St ' ~
ructure 0f (30524114312. the layered sub-structure of Co(2) and Co(3) polyhedra along the b-axis

oﬁ CV

.Eoﬁwﬁm 2.586 woman-

68 05 co 23025 Ev

 

.386 05 mac? Smog—on Amvoo Ea Amvou no 258 3338580
.Eoewé 0556 839-58 2: do 238% 2V .3 2%

 

3:? m

E?

52
58

 

$2

55

$2

95

AV AV
\ JIA \V/A
\ \

 

 

 

 

V? VIA VIA VIA
AV A\V AV A\V
I I I I
AV. AV AV- A!
4 “ 4 4
(A VIA VIA V<A

VIA f

VA :19

Figure 4-9. The layered sub-structure of Al(l 1) and Al(15) polyhedra in CosAlMSi2 along the b-axis. '

82%.: 2i O: .556 Ban: :2 2:0 235% Am: .25. a 2: mesa 88:28 Ban-

 

.583 c :2 as. 52 co 8838
G :2 .6 owﬁé 2F 2: .24. 05mm

97

.2250-.. vs. _m_<£-a 32:2 5 232a 8:65.: 3°03 6: 322.8
E 2 :_< E“ 58 .58 Am: .2 _ .o n s: £52228 5 8:8 es. 5:8 .58 33% 358538 89.528“: ._ 3. 05mm

 

98

Physicochemical Properties:

A noteworthy property of CoSAll4Si2 is its inertness to thermal attack. When
crystals of this compound were exposed to air in a thermal gravimetric fumace
environment, there was no obvious ignition point, and up to 900 °C only about 3% weight
gain was observed (Figure 4-12A). To further investigate the oxidation resistance
properties of this material, single crystals of C05A114Si2 were loaded into an alumina
crucible and annealed in air at 1000 °C for 5 hours. The shiny and clean surface of the
crystals was covered by an oxide layer which is evidently seen by comparing the SEM
images of C05A114Si2 before and aﬁer annealing (Figure 4-12B, C & D). Elemental
analysis by energy dispersive spectroscopy showed highly elevated levels of Al on the
surface of the air-annealed crystals and the amount of Co and Si was less than 3%. This
indicates that Al diffused outward and reacted with oxygen on the surface of the material
to form A1203, which was not spallable and blocked further oxygen transportation into
the material.

In aluminides, it is usually found that Al preferentially forms A1203 on the
surfaces of alloys, such as NiAl and TiAl.22 It has also been reported that addition of A1
into ZrSi2 and MoSi2 helps to improve the oxidation resistance properties by forming a
protective A1203 layer.23 However, the presence of A1203 was not detected by the X-ray
diffraction analysis of the material after annealing. Instead all observed Bragg peaks
corresponded to the starting compound which implies that the structural integrity of
C05A114Si2 was not compromised by annealing at 1000 °C. Similar thermal resistant
properties have been found in RE5C04Si14 (RE = Ho, Er, Tm and Yb)24 in which a

protective Si02 layer was formed and this intriguing behavior is due to the Si-rich nature

99

 

A 110 B
105 -

100 -

Weight (°/o)

95'-
90-

IMIHm

 

 

 

85|lnlnlnanl14111

200 400 600 8 00
Temperature (°C)

 

5 ll")

 

Figure 4-12, (A) Thermal gravimetric analysis of C05A114Si2 in air. (B) Scanning electron
micrograph of CosAll4Si2 before annealing. (C) (D) Close up view of CosAll4Si2 before

and after annealing at 1000 °C for 5 hours.

100

of the crystal framework. The difference in the protecting layer between C05A114Si2 and
RE5Co4Sil4 might suggest higher oxygen afﬁnity of aluminum than Co and Si. Although
it has been observed that additions of Si to some aluminide alloys such as Ti3Al result in
a considerable reduction in the oxidation rate,25 this fact does not explain why C05A114Si2
shows better thermal properties than ColgAl4SSi10.x (x = 0.13), since in CosAll4Si2 the Si

level is lower.

Magnetic Properties:

Crystals were picked from the products and analyzed by EDS. Magnetic
measurements were performed on polycrystalline samples ground from those selected
crystals. Magnetic susceptibility measurements showed that both compounds exhibit
ferromagnetic behavior with Tc around 260K, however very low xm values were observed
(~0.015 emu / mol formula at 5K). In addition, although the magnetization determined at
5K does show hysteresis, the low saturation magnetic moment (0.025 B.M. / mol formula
at 10000 0e) indicates that the ferromagnetic behavior is probably caused by magnetic
impurities. These may be nanocrystals of an unidentiﬁed compound occluding the grains
of single crystals, however we could not identify any distinct phase by SEM analysis or

X-ray diffraction analysis. Therefore we conclude that the pure bulk compounds are

probably paramagnetic.

lOl

4-4. Conclusions

The Al flux method is a powerful tool to synthesize new intermetallic compounds.
The compounds C019A1458i10.x (x = 0.13) and C05A114Si2, grown from liquid Al, show
exceptionally complex structures and large unit cells. The extended structures can be
described in terms of distorted polyhedra of Co, Al and Si atoms. The resemblance of the
current structures to the quasicrystal approximants, especially in the ﬁrst-neighbor
coordination environments, might be helpful in discovering new quasicrystals in the
ternary Co-Al-Si system. C05A114Si2 shows interesting thermal oxidation resistance up to

1000°C which is due to the formation of A1203 layer on the surface of the crystals.

102

References:

 

‘ (a) Kevorkijan, v. Aluminum 2002, 78, 469. (b) Satyanarayana, K. (3.; Pillai, R. M.;
Pai, B. C. Trans. Indian Inst. Metals 2002, 55, 115.

2 (a) Suresh, S.; Mortensen, A.; Needleman, A. Fundamentals of Metal-Matrix
Composite, Butterworth-Heinemann, Boston, 1993. (b) Ochiai, S. Mechanical Properties
of Metallic Composite, Marcel Dekker, Inc., New York, 1994.

3 Wang, H. M.; Lu, x. D.; Liu, Y. F.; Duan, G.; Cai, L. x.; Wang, C. M. Mater. Sci.
Forum 2003, 426-432, 2551.

4 Tokita, s.; Amano, T.; Okabayashi, M.; Nishida, 1. A. Mater. Sci. Forum 1993, 12, 197.
5 Tamura, N. Bhil. Mag. A 1997, 76, 337.

6 Frank, F. C.; Kasper, J. 3. Acta Cryst. 1996, A52, 125.

7 Burger, K. 0.; Wittmann, A.; Nowotny, H. Monatsh. Chem. 1962, 93, 9.

8 German, N. v. Bestn. L ’vov. Univ. Chim. 1981,23, 61.

9 Richter, K. W.; Prots, Y.; Grin, Yu. Inorg. Chem. 2005, 44, 4576.

'0 Kanatzidis, M. G.; Pottgen, R.; Jeitschko, w. Angew. Chem. Int. Edit. 2005, 44,6996.
” Saint, Version 4; Simens Analytical X-ray Instruments Inc., Madison, WI.

'2 SADABS, Sheldrick, G. M.; University of Gottingen, Gottingen, Germany.

'3 G.M. Sheldrick, 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

'4 Onogi, T.; Takeuchi, T.; Sato, H.; Mizutani, U. J. Alloys Compd. 2002, 342, 397.

'5 Geman, N. V.; Zavodnik, V. E.; Yanson, T. I.; Zarechnyuk, 0. S. Kristallograﬁya
1989, 34, 738.

‘6 Girgis, K.; Petter, W.; Pupp, G. Acta Crystallogr., Sect. B: Struct. Sci. 1975, B31, 1 l3.
'7 Adam, J.; Rich, J. B. Acta Cryst. 1954, 7, 813.

'8 Bostroem, M.; Rosner, H.; Prots, Y.; Burkhardt, U.; Grin, Yu. Z. Anorg. Allg. Chem.
2005, 631, 534.

103

 

'9 Stenberg, L.; Sjovall, R.; Lidin, s. .1. Solid State Chem. 1996, 124, 65.

2" Corby, R. N.; Black, P. J. Acta Crystallogr. 1977, 33B, 3468.

2' Grin, J.; Burkhardt, U.; Ellner, M.; Peters, K. J. Alloy Compd. 1994, 206, 243.
22 Grabke, H. J. Mater. Sci. Forum 1997, 251-254, 149.

23 Perez, P.; Lopez, M. F.; Jimenez, J. A.; Adeva, P. Intermetallics 2000, 8, 1393.

24 Salvador, J. R.; Malliakas, C.; Gour, J. R.; Kanatzidis, M. G. Chem. Mater. 2005, 17,
1636.

25 Jazayeri-G, A.; Buckley, R. A.; Davies, H. A. Mater. Sci. Tech. 2002, 18, 1485.

104

CHAPTER FIVE

Exploratory Studies on the Ternary System RE/Ni/Al
Employing Al as a Flux

PART 1. Synthesis, Crystal Structure and Magnetic Properties of RE3N15A119
(RE = Sm, Dy, Er, Yb)

5-1-1. Introduction
The use of liquid A] as a high temperature solvent for exploratory synthesis has

' These phases exhibit

led to the discovery of many novel multinary intermetallics.
complex structures and interesting physical properties. Examples include RE2TM3A19
(TM = Rh, Ir, Pd), 2 YNiAere2, 3 RE4Fe2+xA17-xSig (RE = Ce, Pr, Nd, Sm), 4
REgRu12Al498i9(Alei12-x) (RE = Pr, Nd, Sm, Gd, Tb, Er),5 and RE2TMA16Si4 (RE = Gd,
Tb, Dy; TM = Au, Pt).6 The rich chemistry possible in the liquid metal environment can
be exempliﬁed by the family RETM2A110 (TM = Mn, Re, Fe, Ru, US). It has a
homogeneous composition while four different crystal structure types are present which
can be considered as stacking variants of each other. REMn2Al 10 (RE = Y, La-Nd, Sm,
Gd-Dy) and RERe2A110 (RE = Ce, Pr, Sm) crystallize in the tetragonal CaCr2A110
structure type;7 RETM2A110 (RE = Y, La-Nd, Sm, Gd-Lu; TM = Fe, Ru, 0s) have the
orthorhombic YbFe2A110 structure type;8 and the middle rare earth (Gd, Tb) and late rare
earth (Ho-Lu) analogues of RERe2A110 crystallize in a different stacking variant of the
YbFe2A110 type structure. 7“ ’9

The ternary system RE/Ni/Al is probably one of the most extensively studied

intermetallic systems. There are quite a few ternary families in this system reported in

105

the literature, examples include RE4Ni6A123 (RE = Ce, Sm),lo RENiAl2 (RE = Gd, Dy,
Ho, Er),” RENiAh (RE = La, Ce, Sm, Tb, Dy, Ho, Er, Tm),”’ ‘2 RE3Ni6Al2 (RE = Y,
Sm, Gd, Dy, Ho, En,” RENi2A15 (RE = Ce, Pr),” RENi4A1 (RE = La, Ce, Pr, Nd, Sm,
Gd, Dy, Ho, Er),15 RENiAl (RE = Ce, Gd, Tb, Ho),'6 RENi3A19 (RE = Gd, Dy, Er),l7
RENi2Al3 (RE = Pr, Nd)'8 and RE3Ni5Allo (RE = Gd, Yb).l9 RENiAl crystallizes in the
hexagonal ZrNiAl structure type, and for most rare earth metals, low temperature
ferromagnetic ordering has been observed. 20 Interestingly, it was found that the
incorporation of hydrogen into the crystal lattice of RENiAl drastically reduced the
magnetic ordering temperatures due to modiﬁcation of the density of electronic states at
the Fermi level.21 YbNiAl was shown to be the ﬁrst Yb-based example of a heavy-
Ferrnion antiferromagnet with a Neel temperature TN of 3 K.22 Ferromagnetic ordering at
low temperature was also observed for the RENiaAl family, and the location of the
magnetic moments was identiﬁed to be within the hexagonal plane.23 In the case of
CeNiaAl, mixed-valence behavior was observed by X-ray photoemission spectroscopy
(XPS) and magnetic susceptibility measurements.24

Most of these RE/Ni/Al compounds were obtained by direct combination
reactions using an arc-welder, and this stimulated us to study the RE/Ni/Al system using
the less common Al ﬂux method. Based on our experience, Al tends to serve as a
reactive solvent, so ternary compounds are expected from this RE/Ni/Al system. In
addition, the metal ﬂux method facilitates the growth of high quality large single crystals,
which not only make the structure determination relatively easier, they also allow
anisotropic measurements of physical properties. Our investigations of the RE/Ni/Al

system employing A1 flux led to the discovery of compounds RE3Ni5A119 (RE = Sm, Dy,

106

Er, Yb) which crystallize in the orthorhombic Gd3Ni5Allg structure type.19a

Among these
compounds Yb3Ni5A119 was reported by Bauer and coworkers in 2004 using Al ﬂux
method and it exhibits intermediate valence behavior.19b We used different reaction
conditions to form Yb3Ni5A119 from that of Bauer et al; additionally, our studies showed
different magnetic behavior of this compound. In this section, we will describe the

crystal growth, structure characterization and physical properties of RE3Ni5A119 (RE =

Sm, Dy, Er, Yb).

5-1-2. Experimental Section
Reagents:

The following reagents were used as obtained: rare earth metals (RE = Sm, Dy,
Er, Yb) (Cerac, 99.9%), Ni (99%, 325 mesh, Sargent, Buffalo Grove, IL), Al (Cerac,

99.5%, -20 mesh).

Synthesis:

In a nitrogen—ﬁlled glove box, 1 mmol RE metal (0.15 — 0.18 g), 2 mmol Ni
(0.118 g) and 10 mmol Al (0.270 g) were combined in alumina crucibles. The crucibles
were then placed into silica tubes (13 mm in diameter), which were sealed under vacuum
(~104 Torr). The samples were heated to 1000 °C in 15 h, maintained at this temperature
for 5 h, then cooled to 850 °C in 2 h. They were annealed at 850 °C for 3 (1, followed by
cooling down to 500 °C in 36 h. Finally the temperature was brought down to 50 °C in

10 h. The yield of the reaction for the Yb-analogue was ~50% with the side products

107

mainly YbA13 and YbHNi3A13925 ; while for the Sm-analogue the yield was 90% and

Sm4Ni6Al23 was the side product.26

Isolation:

The excess aluminum was removed by soaking the crucible in aqueous 5M NaOH
solution overnight. The crystalline product remaining after the isolation procedure was
rinsed with water and acetone. The crystals of RE3Ni5Ali9 crystallized in the form of
needles. Single crystals were selected for elemental analysis, X-ray diffraction, and

physical properties measurements.

Scanning Electron Microscopy and Elemental Analysis:

The crystals were picked and placed on a Scanning Electron Microscope (SEM)
sample plate using carbon tape. Chemical compositions of the products were determined
by Energy Dispersive Spectroscopy (EDS) performed on a JEOL ISM-35C SEM
equipped with a NORAN EDS detector. Data were acquired by applying a 25 kV
accelerating voltage with an accumulation time of 60 s. The atomic ratio for Yb3N15A1|9
was determined to be 3 : 5.55 : 19.61 for Yb : Ni : Al, which agreed well with the results
derived from single crystal X-ray analysis. The SEM images of typical crystals of
Sm3Ni5A119, Er3Ni5Al.9 and Yb3Ni5A119 are shown in Figure 5-1-1. The crystals were

formed with rod or needle morphology, sometimes bundled together.

108

 

Figure 5-1-1. SEM images of typical crystals of Sm3Ni5A119, Er3Ni5A119 and Yb3Ni5A119
grown from A1 ﬂux.

109

X-ray Crystallography:

Single crystal X-ray diffraction data were collected for RE3Ni5A119 (RE = Sm, Dy,
Er and Yb) at room temperature on a Bruker AXS SMART CCD X-ray diffractometer.
The data collections (Mo Ka radiation, 7b = 0.71073 A) were acquired covering a full
sphere of reciprocal space. Data acquisition and cell reduction were done with the
SMART27 software package and data processing was performed with the SAINTPLUS
program. 28 An empirical absorption correction was applied to the data using the
SADABS program.29 The structures were solved using direct methods and reﬁned with
the SHELXTL package program.3O Systematic absence conditions led to three possible
space groups: Cmc21, Cmcm and Ama2. The IE2 -1 I value was 0.876 suggesting a
centrosymmetric space group. So Cmcm was chosen and later structural reﬁnement
conﬁrmed this choice. All atomic positions were reﬁned anisotropically. Data collection
parameters and reﬁnement details for Sm3Ni5A119 and Dy3N15A119 can be found in Table
5-1-1. The reﬁnement details for Br and Yb analogs are listed in Table 5-1-2. Atomic
positions, displacement parameters and anisotropic displacement parameters for
RE3Ni5A119 (RE = Sm, Dy, Er and Yb) are listed in Tables 5-1-3 and 5-1-4.

X-ray powder diffraction data were collected at room temperature on a CPS 120
IN EL X-ray diffractometer (Cu K01) equipped with position-sensitive detector.
Experimental powder patterns were compared with the patterns calculated from the single

crystal structure solution (by the CrystalDiffract software program“) to determine the

phase purity.

110

Table 5-1-1. Selected crystal data and structure reﬁnement details for Sm3Ni5A119 and

 

 

Dy3Nl5A119.

Empirical formula Sm3Ni5A119 Dy3Ni5A119
Formula weight 1257.22 1293.67
Crystal system Orthorhombic Orthorhombic
Space group Cmcm (#63) Cmcm (#63)

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient
F (000)

Crystal size

Theta range for data collection
Limiting indices

Reﬂections collected
Independent reﬂections
Completeness to theta = 37.00°
Reﬁnement method
Data / restraints / parameters

2
Goodness-of-ﬁt on F

Final R indices [I>2sigma(l)]
R indices (all data)

Largest diff. peak and hole

a = 4.1013(9) A

b = 16.050(4) A

c = 27.105(6) A
1784.2(7) A3

4

4.680 Mg/m3

15.783 mm"

2292

0.24 x 0.032 x 0.05 mm3

1.50 to 28.05°
-55h55
aoskszo
-35 51534

9577
1228 [R(int) = 0.0380]
97.4 %

a = 4.0747(4) A

b = 15.963206) A

c = 27.029(3) A
1758.1(3) A3

4

4.887 Mg/rn3

18.747 mm"

2340

0.22 x 0.18 x 0.26 min3

1.51 to 27.96°
65hss
-20 s k s 19
-3551535

8481
1169 [R(int) = 0.0313]
94.7 %

Full-matrix least-squares on F 2

1228 I 0 I 86
1.328

Rl = 0.0256

sz = 0.0577

R1 = 0.0296

sz = 0.0615

1.720 and —2.209 e.A'3

221/2

R1 = sharing/211:3; WR2 = [Z[W(Foz- 1331/ [2(wIFOI ) 1

111

1169 I 0 I 86
1.440

R1 = 0.0270

sz = 0.0675

R1 = 0.0280

sz = 0.0713

2.114 and -2.371 e.A'3

Table 5-1-2. Selected crystal data and structure reﬁnement details for Er3Ni5A119 and

 

 

Yb3Ni5A119.

Empirical formula Er3Ni5A119 Yb3Ni5A119
Formula weight 1307.95 1325.29
Crystal system Orthorhombic Orthorhombic
Space group Cmcm (#63) Cmcm (#63)

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient
F(000)

Crystal size

Theta range for data collection
Limiting indices

Reﬂections collected
Independent reﬂections
Completeness to theta = 37.00°
Reﬁnement method

a = 4.0523(8) A

b = 15.878(3) A

c = 27.927(5) A
1732.6(6) A3

4

5.014 Mg/m3

20.617 mm"

2364

0.20 x 0.12 x 0.15 mm3

1.50 to 28.36°
-5£h£5
205k520
-35 515 35

8740
1231 [R(int) = 0.0764]
97.5 %

a = 4.0573(2) A

b = 15.8745(9) A

c = 26.936505) A
1734.92(16) A3

4

5.074 Mg/rn3

22.249 mm'1

2388

0.23 x 0.16 x 0.09 mm3

1.51 to 28.23°
-55h55
205k520
-34 515 35

9907
1228 [R(int) = 0.0332]
98.2 %

F ull-matrix least-squares on F 2

Data / restraints / parameters 1231 / 0 / 86 1228 / 0 / 86
Goodness-of-ﬁt on F2 1.739 1.209
Final R indices [I>28igma(1)] R1 = 0.0499 R1 = 0.0217
wR2 = 0.1172 wR2 = 0.0428
R indices (all data) R1 = 0.0533 R1 = 0.0243
wR2 = 0.01371 wR2 = 0.0434
Largest diff. peak and hole 5.790 and —2.757 e.A'3 1.335 and -2.178 e.A”3

 

2 2
R1 = 2(lFoi'chD/Z11:015 WRZ = [2[W(Fo ' Fc ]/ [2(w1Foi

112

221/2

)1

Table 5-1-3. Atomic coordinates (A x 104) and equivalent isotropic displacement

parameters (A2 x 103) for Sm3Ni5A119 and Dy3Ni5A119.

 

 

Atom Wyk. x y z U(eq)‘
Sm(1) 4c 0 3865(1) 2500 7(1)
Sm(2) 8f 60% 6685(1) 3649(1) 6(1)
Ni(1) 8f 0 6597(1) 4581(1) 6(1)
Ni(2) 8f -5000 5531(1) 5835(1) 8(1)
Ni(3) 4c -5000 5450(1) 2500 7(1)
Al(l) 8f 0 5949(1) 5370(1) 9(1)
Al(2) 8f 0 5780(1) 2982(1) 9(1)
Al(3) 8f 0 7917(1) 4118(1) 8(1)
Al(4) 8f -5000 7345(1) 4717(1) 8(1)
Al(5) 8f 0 3702(1) 3654(1) 9(1)
Al(6) 8f -5000 2758(1) 3067(1) 10(1)
Al(7) 8f -5000 4321(1) 5302(1) 9(1)
Al(8) 4c -5000 6929(2) 2500 9(1)
Al(9) 8f 0 5364(1) 4001(1) 9(1)
Al(10) 8f 60% 4649(1) 3265(1) 10(1)
Dy(1) 4c 0 3883(1) 2500 7(1)
Dy(2) 8f ~5000 6659(1) 3648(1) 6(1)
Ni(1) 8f 0 6595(1) 4575(1) 6(1)
Ni(2) 8f -5000 5537(1) 5845(1) 8(1)
Ni(3) 4c -5000 5454(1) 2500 7(1)
Al(l) 8f 0 5941(2) 5368(1) 9(1)
Al(2) 8f 0 5799(2) 2983(1) 9(1)
Al(3) 8f 0 7909(2) 41 12(1) 8(1)
Al(4) 8f -5000 7343(2) 4716(1) 8(1)
Al(5) 8f 0 3702(2) 3651(1) 9(1)
Al(6) 8f -5000 2766(2) 3064(1) 9(1)
Al(7) 8f -5000 4325(2) 5310(1) 9(1)
Al(8) 4c -5000 6939(2) 2500 9(1)
Al(9) 8f 0 5369(2) 3993(1) 9(1)
Al(10) 8f -5000 4650(2) 3259(1) 10(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

113

Table 5-1-4. Atomic coordinates (A x 104) and equivalent isotropic displacement

parameters (A2 x 103) for Er3Ni5Allo and Yb3Ni5Altg.

 

 

Atom Wyk. x y z U(eq).
Er(l) 4c 0 3892(1) 2500 4(1)
Er(2) 8f -5000 6663(1) 3647(2) 4(1)
Ni(1) 8f 0 6586(1) 4573(2) 4(1)
Ni(2) 8f -5000 5540(1) 5853(2) 5(1)
Ni(3) 4c -5000 5454(2) 2500 4(1)
Al(l) 8f 0 5930(3) 5373(2) 6(1)
Al(2) 8f 0 5801(3) 2984(2) 5(1)
Al(3) 8f 0 7907(3) 41 10(2) 4(1)
Al(4) 8f -5000 7348(3) 4713(2) 5(1)
Al(5) 8f 0 3698(3) 3645(2) 6(1)
Al(6) 8f —5000 2770(2) 3063(2) 6(1)
Al(7) 8f —5000 4334(2) 5316(2) 5(1)
Al(8) 4c -5000 6941 (4) 2500 6(1)
Al(9) 8f 0 5364(3) 3986(2) 6(1)
Al(10) 8f -5000 4649(3) 3253(2) 6( 1)
Yb(l) 4c 0 3912(1) 2500 5(1)
Yb(2) 8f -5000 6664(1) 3646(1) 4(1)
Ni(1) 8f 0 6591(1) 4581(1) 4(1)
Ni(2) 8f -5000 5534(1) 5850(1) 5(1)
Ni(3) 4c -5000 5473(1) 2500 5(1)
Al(l) 8f 0 5937(1) 5377(1) 6(1)
Al(2) 8f 0 5826(1) 2983(1) 6(1)
Al(3) 8f 0 7904(1) 4105(1) 6(1)
Al(4) 8f -5000 7347(1) 4710(1) 6(1)
Al(5) 8f 0 3700(1) 3642(1) 7(1)
Al(6) 8f -5000 2785(1) 3062(1) 7(1)
Al(7) 8f ~5000 4327(1) 5317(1) 6(1)
Al(8) 4c -5000 6968(2) 2500 7(1)
Al(9) 8f 0 5381(1) 3983(1) 7(1)
_ Al(10) 8f -5000 4662(1) 3253(1) 6(1)

 

114

U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

Table 5-1-5. Anisotropic displacement parameters (A2 x 103) for Sm3Ni5A119 and

 

 

Dy3N15A119.
Atom U11 U22 U33 U23 U13 U12
Sm(l) 8(1) 7(1) 6(1) 0 0 0
Sara) 6(1) 6(1) 5(1) 0(1) 0 0
Ni(1) 7(1) 6(1) 6(1) 0(1) 0 0
Ni(2) 8(1) 7(1) 9(1) -1(1) 0 o
Ni(3) 8(1) 7(1) 6(1) 0 0 0
Al(l) 12(1) 10(1) 5(1) 3(1) 0 0
Al(2) 7(1) 14(1) 6(1) -2(1) 0 0
Al(3) 9(1) 6(1) 9(1) 1(1) 0 0
Al(4) 7(1) 8(1) 8(1) -2(1) 0 0
Al(5) 11(1) 8(1) 9(1) 0(1) 0 0
Al(6) 5(1) 5(1) 7(1) 0 0 4(1)
Al(7) 10(1) 12(1) 7(1) 2(1) 0 o
Al(8) 12(1) 8(1) 8(1) 0 0 0
Al(9) 9(1) 8(1) 11(1) 1(1) 0 0
Al(10) 10(1) 11(1) 7(1) 0(1) 0 0
Dy(1) 9(1) 5(1) 6(1) 0 0 0
Dy(2) 8(1) 5(1) 6(1) 0(1) 0 0
Ni(1) 8(1) 5(1) 6(1) 0(1) 0 0
Ni(2) 10(1) 5(1) 9(1) 0(1) 0 0
Ni(3) 10(1) 5(1) 6(1) 0 0 0
Al(l) 14(1) 8(1) 5(1) 2(1) 0 0
Al(2) 8(1) 13(1) 5(1) -1(1) 0 0
Al(3) 9(1) 4(1) 9(1) 1(1) 0 0
Al(4) 8(1) 7(1) 9(1) -3(1) 0 0
Al(5) 12(1) 6(1) 9(1) 1(1) 0 0
Al(6) 11(1) 9(1) 7(1) 2(1) 0 0
Al(7) 10(1) 6(1) 12(1) 0(1) 0 0
Al(8) 12(1) 6(1) 1 1(2) 0 0 0
Al(9) 10(1) 6(1) 12(1) 2(1) 0 o
Al(10) 12(1) 9(1) 9(1) 1(1) 0 0

 

The anisotropic displacement factor exponent takes the form:
-21t2[hza*2Ull+...+2hka*b*U12]

115

Table 5-1-6. Anisotropic displacement parameters (A2 x 103) for Er3Ni5A119 and

 

 

Yb3N15A119.
Atom U11 U22 U33 U23 U13 U12
Er(1) 6(1) 6(1) 0(1) 0 0 0
Er(2) 6(1) 6(1) 0(1) 0(1) 0 o
Ni(1) 5(1) 6(1) 0(1) 0(1) 0 0
Ni(2) 7(1) 6(1) 1(1) -1(1) 0 0
Ni(3) 8(1) 5(1) 0(1) 0 0 0
Al(l) 10(2) 9(2) 0(2) 4(2) 0 0
Al(2) 6(2) 10(2) 0(2) -1(2) 0 0
Al(3) 6(2) 6(2) 0(2) 1(2) 0 0
Al(4) 6(2) 8(2) 0(2) -2(2) 0 0
Al(5) 7(3) 10(3) 0(3) 0 0 0
Al(6) 8(2) 7(2) 4(2) 2(2) 0 0
Al(7) 7(2) 7(2) 2(2) -2(2) 0 0
Al(8) 7(3) 10(3) 0(3) 0 0 0
Al(9) 8(2) 7(2) 4(2) 2(2) 0 o
Al(10) 9(2) 7(2) 0(2) 1(2) 0 0
Yb(l) 6(1) 5(1) 3(1) 0 0 0
Yb(2) 5(1) 5(1) 4(1) 0(1) 0 0
Ni(1) 4(1) 4(1) 4(1) 0(1) 0 0
Ni(2) 6(1) 5(1) 5(1) 0(1) 0 0
Ni(3) 6(1) 5(1) 4(1) 0 0 0
Al(l) 10(1) 8(1) 1(1) 4(1) 0 0
Al(2) 5(1) 10(1) 4(1) -1(1) 0 0
Al(3) 7(1) 6(1) 6(1) 0(1) 0 0
Al(4) 6(1) 6(1) 6(1) -1(1) 0 0
Al(5) 8(1) 7(1) 6(1) -1(1) 0 0
Al(6) 9(1) 7(1) 5(1) 0(1) 0 0
Al(7) 6(1) 4(1) 7(1) 0(1) 0 0
Al(8) 7(1) 7(1) 6(1) 0 0 0
Al(9) 7(1) 7(1) 8(1) 1(1) 0 0
Al(10) 8(1) 5(1) 7(1) 1(1) 0 0

 

The anisotropic displacement factor exponent takes the form:
-21t2[h2a*2U '+. . .+2hka*b*U'2]

116

Table 5-1-7. Bond distances (A) for RE3Ni5A119 (RE = Sm, Dy, Er, Yb).

 

 

Bond Distance Bond Distance
Sm(1)-Al(5) x2 3.139(2) RE(2)-Al(4) 3.099(2)
Dy(1)-Al(5) 3.125(3) 3.086(3)
Er(1)-Al(5) 3.099(4) 3.068(4)
Yb(1)-Al(5) 3.095(2) 3.063(2)
RE(1)-Al(6) x4 3.1187(16) RE(2)-Ni(1) 2.379(2)
3.1066(18) 3.2296(9)
3.095(3) 3.2137(16)
3.1002(15) 3.2353(7)
RE(1)-Al(8) 3.107(3) Ni(1)-Al(1) 2.379(2)
3.103(4) 2.386(3)
3.097(6) 2.393(5)
3.086(3) 2.382(2)
RE(1)-A1(10) x4 3.1762(17) Ni(1)-Al(7) x2 2.5433(14)
3.139(2) 2.5308(16)
3.108(3) 2.516(3)
3.1048(16) 2.5128(13)
RE(1)-Ni(3) 3.2679(12) Ni(2)-Al(5) x2 2.7638(16)
3.2320(12) 2.7349(19)
3.203(2) 2.719(3)
3.2023(10) 2.7322(15)
RE(2)-Al(2) 3.0742(12) Ni(2)-Al(9) x2 2.5429(14)
3.044(2) 2.5360(16)
3.028(3) 2.520(3)
3.0126(15) 2.5356(13)

 

117

Table 51.7. (Continued) bond distances (A) for RE3Ni5A119 (RE = Sm, Dy, Er, Yb).

 

 

Bond Distance Bond Distance
RE(2)-Al(3) 3.1486(17) Ni(3)-Al(2) x4 2.4885(13)
3.1157(19) 2.4819(15)
3.092(3) 2.471(3)
3.0844(15) 2.4747(12)
Ni(3)-Al(10) x2 2.439(2) Al(3)-Al(4) x2 2.772(2)
2.419(3) 2.763(3)
2.396(4) 2.743(4)
2.402(2) 2.7477(19)
Al(4)-Al(4) x2 2.609(3) Al(3)-Al(5) x2 2.715(2)
2.600(3) 2.702(2)
2.594(5) 2.693(4)
2.606(3) 2.6964(19)
Al(7)-Al(7) 2.725(4) Al(4)-Al(7) 2.675(3)
2.731(5) 2.664(3)
2.714(8) 2.673(6)
2.735(4) 2.657(3)
Al(7)-Al(9) x2 2.834(2) Al(5)-Al(10) x2 2.762(2)
2.817(3) 2.751(2)
2.805(4) 2.739(4)
2.809(2) 2.7473(19)
Al(6)-Al(8) x2 2.888(2)
2.866(2)
2.852(4)

2.8433(19)

 

118

Physical Properties Characterization:

Magnetic measurements were conducted on a single crystal of Sm3Ni5A119 and
polycrystalline samples of Er3Ni5A119 and Yb3Ni5A119. Field-cooled and zero-ﬁeld
cooled dc magnetization measurements were performed for the above samples using a
Quantum Design MPMS SQUID magnetometer. For polycrystalline samples, EDS-
analyzed crystals were ground into powder, sealed in a kapton tape and placed into the
magnetometer by using a straw. Thermal dependent data were collected in the
temperature range of 5 ~ 400 K at 1000 G, while ﬁeld dependent magnetic
measurements, conducted at 5 K, were carried out in ﬁelds up to i 55000 G. The
magnetic contributions from the core diamagnetism and kapton tape were subtracted for
correction.

Electrical resistivity was measured over the temperature range 5 K ~ 300 K using
a four-probe dc technique with contacts made using silver paste on a pressed-pellet
sample of Sm3Ni5A119. Single crystals of Sm3Ni5A119 were selected, ground into powder
and pressed into a pellet of 3 mm length, 2 mm width and 0.2 mm thickness. The pellet
sample was annealed at 350 °C for 3h. Thermopower data were collected on single
crystals of Sm3Ni5A119 from 300 K to 400 K with a MMR Technologies, Inc. Seebeck

measurement system.

5-1-3. Results and Discussion
Synthesis:
Gd3Ni5A119 — the ﬁrst member belonging to the RE3N15A119 family, was

synthesized by arc-melting stoichiometric amounts of the elements followed by annealing

119

at 800 °C for two weekslga Our systematic explorations on the system Yb/Ni/Al using
Al as a ﬂux led to the discovery of the second member of this family — Yb3N15A119.
This phase is very stable and it was formed with a variety of Yb : Ni ratios ﬁ'om 1:1 to
1:2 to 2:1. The reaction with equal amount of Yb and Ni produced Yb3N15A119 in 70%
yield, and the side products were mainly YbAl3. Bauer has shown that Yb3Ni5A119 could
be synthesized by loading equal amount of Yb and Ni in liquid A1, then heating the
samples up to 1100 °C and slowly cooling down to 600 °C at which point excess Al was
removed by centrifuging.19b However no yield or phase purity was given in the paper.
Our experimental studies show that larger single crystals of Yb3Ni5A119 could be obtained
when the ratio of Yb : Ni was 1 : 2 or 2 : 1, however under these conditions another
ternary compound formed — YbHNi3Algg, which is a new phase and will be described in
Part II of this chapter.

The synthesis of the Sm-analogue is noteworthy. Reactions with equal amount of
Sm and Ni (1 : 1) led to the formation of a pure phase of SmaNi6Al23, which belongs to
the family Rﬁzn+mT4n+mAllsn+4m with m = 2, n = 1.26 This family will be described in
more detail in the crystal structure section. When extra amount of Ni was added to have
a ratio of Sm : Ni as 1 : 2, powder X-ray diffraction indicated that the major phase was
Sm3Ni5A119, shown in Figure 5-1-2A. Most experimental Bragg peaks match fairly well
with the ones calculated from single crystal X—ray data of Sm3Ni5A119 except for a weak
peak at 22° in 20. This peak belongs to SmaNi6A123, the other peaks of which overlap
with the ones of Sm3Ni5A119. Even the reaction with stoichiometric amount of Sm and Ni
(1 :1.67) could not avoid this problem. Sm3Ni5A119 and Srr14Ni6Al23 could not be

distinguished by their crystal habits because both of them crystallize in needles.

120

E:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3500 5000
*2
g 3000 ' - 4000 ET
3 2500 l- g
3* - 3000 52.
3’5 2000 - 3;
D. . experimental - 2000 t:
5 1500 - , . , 2,
‘ - l ' i » ‘ - 1000 °
11111 11111111111111»
t: t:
a . _ 0 g
.5 500 - . a
calculated (Sm3N15All9) 9-
0 “1.1....1..,.1....1....1....1.44.3_1000
10 20 30 40 50 60 70 80
2 theta
B)
4
1 10 5000
g 4000 5
o "' F9
8 8000 r- g
'C‘. E”...
8. - 3000 Q
8 6000 - 3:
3 experimental _ 2000 c
<’ 4000 - . ‘o’
E I l . 1000 :3,-
' 1111111111 E
,3 2000 III Will1 ' 1 0 g;
5: calculated (Sm Ni A1 )
0 ....l.anl.Arnla...144.41.3..5.19.A-. _1000
10 20 30 40 50 60 70 80
2 theta

Figure 5-1-2. A) Powder X-ray diffraction pattern of Sm3Ni5A119 compared with the
calculated pattern. B) Powder X-ray diffraction pattern of a single crystal of Sm3Ni5A119

compared with the calculated pattern.

121

Attempts to make the other rare earth analogs of RE3Ni5A119 were not successful so far;

instead stable binaries such as REAl3 and Ni86A114 formed.

Crystal Structure:

All four compounds — Sm3Ni5A119, Dy3Ni5A119, Er3Ni5A119 and Yb3Ni5A119
crystallize in the Gd3Ni5A119 structure typewa According to Gladyshevskii, Gd3Ni5Allg
belongs to the series RE2n+mT4n+mA115n+4m where n is the number of the RE2T4A115 slab
and m is the number of the RENiA14 slab. AThe monoclinic structure of RE2T4A115, shown
in Figure 5-1-3A, is hypothetical since no crystallographic data has been observed. The
RENiAl4 slab crystallizes in the YNiAl4 structure type and is depicted in Figure 5-1-3D.32
Both the RE2T4A115 slabs and the YNiAl4 slabs can be considered to form from the
stacking of RE-centered pentagonal prisms and T-centered trigonal prisms. Gd3Ni5Allg is
the simplest member in this family with m = 1, n = 1. So the structure of Gd3Ni5A119 can
be considered as intergrowth of two kinds of slabs, one deriving from the orthorhombic
YNiAl4 lattice and the other one corresponding to the translation unit of a hypothetical
RE2T4A115 lattice. Another member of this family was identiﬁed as monoclinic
Y4Ni6Alz3 with n = 1, m = 2, shown in Figure 5-1-3B.33

RE3Ni5Allg crystallizes in an orthorhombic space group Cmcm with two
crystallographically independent RE sites, three Ni sites and ten Al sites. All atomic sites
are fully occupied and no disorder is observed in the structure. It is evident from Tables
5-1-1 and 5-1-2 that from Sm to Dy to Er, the unit cell parameters decrease as expected

(known as “Lanthanide Contraction”). However an anomaly is found in the Yb analogue

122

 

 

 

 

 

 

 

 

Wﬁ‘lﬁﬂlﬁ

 

Gd3Ni5All9

Figure 5-1-3. A) Structure of REzNi4A115 (hypothetical) along the b-axis. B) Structure of
Y4Ni6A123 along the b-axis. C) Structure of Gd3Ni5A119 along the a-axis. D) Structure of
YNiAl4 along the a-axis. Large empty circles: rare earth element; medium gray circles:

Ni; black circles: A1.

123

 

i. e. the cell volume of the Yb-analogue is larger than that of the Er analogue, suggesting
that the Yb ions are in a divalent or at least an intermediate valence state which will be
discussed in detail in the physical properties section.

For the structure description of RE3Ni5A119 we will use the Yb analogue as an
example. The whole structure can be viewed in polyhedra of three Ni sites with Yb
atoms sitting in the channels down the a-axis (Figure 5-1-4). The channel is composed of
two Ni(1) polyhedra, four Ni(2) polyhedra and two Ni(3) polyhedra. These channels
form columns along the b-direction by sharing Ni polyhedra. Each unit cell contains two
such columns, rotated by 180° with respect to each other. Inside the channel, the
polyhedra of Ni atoms are interconnected by bridging Al(6) atoms with Al(6)-Al(6) bond
distance of 2.939(3) A.

The local coordination environments of three Ni sites and two Yb sites are shown
in Figure 5-1-5. All three Ni sites are sitting in matrixes composed of Al atoms with
coordination numbers of 8, 9 and 7, respectively. The Ni(1) atom is bonded to eight Al
atoms with Ni-Al bond distance ranging from 2.3 82(2) to 2.548(2) A. The Ni(2) atom is
sitting in a tri-capped trigonal prism composed of nine Al atoms with Ni-Al bond
distance in a similar range as the Ni(1) atom. This tri-capped trigonal prism geometry is
often seen in other transition metal almninides, such as CozAls.34 The Ni(3) atom is
coordinated by seven Al atoms with one trianglular face composed of one Al(8) and two
Al(2) atoms, the other face being a parallelogram containing two Al(2) and two Al(10)
atoms.

The two Yb sites have similar coordination geometry: both of them are in a penta-

capped pentagonal prism composed of thirteen Al atoms and two Ni atoms. The

124

geometry of Yb(2) atom, however is slightly more distorted than the Yb(l) atom (Figure
5-1-5D & E). The distance between Yb(l) and Al atoms ranges from 3.085(3) to
3.306(2) A, while for Yb(2) atoms, the range is slightly broader which is from 3.0126(15)
to 3.3516(20) A. Obviously in this compound we cannot assess a size difference for the

two Yb ions by comparing Yb-Al bond distances.

Physical Properties:

Initially magnetic measurements of Sm3Ni5Al.9 were conducted on
polycrystalline samples ground from single crystals. Ferromagnetic behavior was found
with magnetic susceptibility xm = 0.35 pg at 5 K. This behavior was also supported by
the fact that single crystals were attracted to a magnet. To conﬁrm this observation, 3
large crystal of Sm3Ni5A119 about 3 mg was selected for the magnetic measurements.
Powder X-ray diffraction analysis on this crystal showed that every Bragg peak matched
with the calculated one for Sm3Ni5A119 (Figure 5-1-28). The magnetic measurements
obtained from this single crystal indicated paramagnetic instead of ferromagnetic
behavior. Thus the observed ferromagnetic behavior of the ground sample was most
likely caused by impurities. To avoid Ni particles attaching on the surface of the crystals,
the crystals were sonicated in acetone for ten minutes. After the crystals were dry, we
found that fewer of them were attracted to the magnet. Attempts to further clean the

crystals with aqua-regia failed since acid attacked the sample.

125

 

Figure 5-1-4. The structure of Yb3Ni5A119 in polyhedral view down the a-axis.
Large circles: Yb; black circles: Al; gray circles: Ni.

126

.Q oé 55:5 mega § 93 Q o.m £536 853 _Z we 3:08:835 cosmE—zooo .m- Tm 8:3”—

 

127

The temperature dependent magnetic susceptibility of Sm3Ni5Allg measured on a
single crystal sample with the a-axis perpendicular and parallel to the external magnetic
ﬁeld are shown in Figures 5-1-6A and 5-1-7A, respectively. With the applied ﬁeld
perpendicular to the a-axis, Sm3Ni5Allg ﬁrst undergoes a ferromagnetic transition at ~150
K and then an antiferromagnetic transition at the low temperature about 5 K.
Interestingly we did not observe the curvy feature in the inverse magnetic susceptibility
data due to the close spacing of the 6H5/2 and 6H7/2 multiplet levels.35 Instead, above 300
K, this material follows the Curie-Weiss Law with an effective magnetic moment of 2.36
pg. This number is much larger than the theoretical value for Sm3+ ions (um = 1.46 pa
per formula), and the reason might be due to the effect of the low-lying excited levels.

The transition from ferromagnetic to antiferromagnetic state is supported by the
ﬁeld dependent magnetization data, Figure 5-1-6B. The moments are rapidly aligned
with the applied ﬁeld indicating a ferromagnetic state until 2000 G; then the
magnetization increases more slowly in a linear function until 55000 G which is not
enough to saturate the spins.

When the ﬁeld is applied parallel to the a-axis of Sm3Ni5Allg, no
antiferromagnetic transition is found at low temperature; besides, the hump at 150 K
indicates that the ferromagnetic interaction is much weaker (Figure 5-1-7A). However
the higher temperature data is very similar to those obtained with the ﬁeld perpendicular
to the a-axis: a Curie-Weiss behavior is found between 320 K to 400 K with an effective
magnetic moment of 2.45 us. This orientation dependent behavior implies that the

magnetic spins are conﬁned to the bc-plane. When the external ﬁeld is parallel to the a-

128

axis, i.e. perpendicular to the magnetically ordered spins, the spins are forced out of the
plane by the external ﬁeld which suppresses the ferro- or antiferromagnetic ordering.

Figure 5-1-8A shows the temperature dependence of the electrical resistivity p(T)
of Sm3Ni5A119, which indicates metallic behavior of this material. The resistivity value at
room temperature is about 550 pﬂ-cm, and it decreases with decreasing temperature until
it reaches a plateau of 400 pﬂ'cm below 20 K. The thermoelectric power, shown in
Figure 5-1-8B, is about 3 pV/K at room temperature. The small magnitude of
therrnopower is also indicative of the metallic system of Sm3Ni5Al19; and the positive
values suggest a p-type material.

The temperature dependent magnetic susceptibility for Er3Ni5A119 is plotted in
Figure 5-l-9A. This material shows Curie-Weiss behavior from 50 K to 400 K without
obvious magnetic ordering at low temperatures. The par value, obtained from ﬁtting the
data to the Curie-Weiss law, is 16.2 pg (per formula), in excellent agreement with the
theoretical value for Er3+ (16.6 pg). This indicates that Ni atoms do not contribute to the
magnetic moment. The very low 0 value (-1.85 K) implies weak antiferromagnetic
interaction between the Er atoms. Figure 5-1-9B shows the isothermal magnetization
behavior of Er3Ni5Allg at different temperatures (5 K and 100 K). In both cases, the
magnetization increases gradually with applied external ﬁeld which is characteristic of
the paramagnetic material; and no sign of saturation is observed up to 55000 G.

The temperature dependent magnetic susceptibility for Yb3Ni5A119 is shown in
Figure 5-1-10A. From the temperature range 5~300 K, this compound shows
paramagnetic behavior; above 80 K, the magnetic susceptibility follows the Curie-Weiss

law with an effective magnetic moment peg = 6.77 pg. This value is smaller than the

129

theoretical value for Yb3+ ion (7.86 pg per formula), which could be an indication of
mixed valent behavior of the Yb ions. This observation is also supported by the
comparison of the unit cell parameters between Er3Ni5A119 and Yb3Ni5A119. If the Yb
ions were in a 3+ oxidation state as the Er ions, we would expect to see a unit cell
parameter contraction from the Er analogue to the Yb analogue; however the cell volume
of the Yb analogue is larger than that of the Er analogue, implying either a divalent or an
intermediate valence state of the Yb ions. The unusual negative large Weiss constant 0 (-
551.2 K) indicates strong antiferromagnetic interaction between the Yb ions; however, no
magnetic ordering was observed down to 5 K. Bauer and his coworkers also claimed to
observe intermediate valence behavior of Yb ions in Yb3Ni5A119:l9b the calculated
effective magnetic moment was found to be 7.36 pg, lower than the theoretical value of
Yb3+ (7.86 pg); a broad hump in the temperature dependent magnetic susceptibility data
was present at ~100 K, which is typical of intermediate valence system. 36 Our
experimental results did not show the broad hump at low temperatures; and our effective
magnetic moments (Herr = 6.77 pg, 0 = -551.2 K) are slightly different from the results
reported by Bauer ((Hcff = 7.36 pg, 0 = -731 K). These differences might be due to
experimental error or differences in the samples because of different reaction conditions
employed in the synthesis; however they all support the argument that Yb atoms in

Yb3Ni5Allg are in intermediate valence states.

130

 

 

 

 

 

0.05 r 300
“err: 2-36 Ila ,"
E 0.04 v. 1 250
3 0 = 203.9 K v j
E o i .' 200
‘9‘ .'o. ' a
v— 0.03 P .. ’ :
o : f .
E t . ,v , 1 150
D 0.02 " . 'v E
E 3 . “a“, 1 100
x5 0.01 : -
O
0 50 100 150 200 250 300 350 400
Temperature (K)
B)
1;; 0.08
'5‘ Q
g 0.06 - . .
c9. . '
.— 0.04 - , .
E . -
\ 0.02 - ..o"
2 I
. 0
a l
"a -0.02 - J...
E -0.04 - , . '
'5 o
a -0.06 - . .
(264108 .lllr.ll...l.r.l...l.

 

 

 

6104-4104-2104
Field (G)

0 2104 410‘ 6104

/I

X

Figure 5-1-6. (A) Temperature dependent molar magnetic susceptibility and inverse

magnetic susceptibility of Sm3Ni5A119. The susceptibility was measured with an applied

magnetic ﬁeld 1000 Gauss; (B) Field dependent magnetization for Sm3Ni5Al.9 at 5 K.

The single crystal sample was oriented with the a-axis perpendicular to the external

magnetic ﬁeld.

131

A)

 

 

 

 

0.03 ‘ 300
1; 0.025 11m: 2-45 MB .I'é 250
'5‘ ' j
E 0.02 -_ 0—209.l K ’7 J 200
- : ' . ,v '
"c3 : ° . ,I
E 0.015 _- . '. 1150
B I o . 'v —->:
E 0.01 :- 3!. $100
v v .
E . V' o .
x0.005 v" -..-50
0 ‘0
0 50 100 150 200 250 300 350 400
Temperature (K)
B)
E 0.06 .
:2 0
E .'
e 0.04 " .0
E 002 ° .
— I
\ ' ..-
2. 0 .
E
g -0.02 - .
.3 .-'
a -004 - ,-
ﬁt ,-
CU
241063.1..41“.1...r...r.

-6

Figure 5-1-7. (A) Temperature dependent molar magnetic susceptibility and inverse
magnetic susceptibility of Sm3Ni5Allg. The susceptibility was measured with an applied
magnetic ﬁeld 1000 Gauss; (B) Field dependent magnetization for Sm3Ni5A119 at 5 K.

The single crystal sample was oriented with the a-axis parallel to the external magnetic

ﬁeld.

 

 

 

Field (G)

132

1044104-2104 0 2104 4104 6104

A)

 

 

 

 

 

 

 

 

600
550 - . 0
’g 500 k . '
{3. . °
Q 450 - . °
400 '
350 .1..l..l..l..lr.l..
0 50 100 150 200 250 300 350
Temperature (K)
B)
8
g 7 r . '
> o
3 .
(I) 6 _, °
‘5 o
3
8" .- o
a 5 °
0.) o
e 4 -
3 ' , I . I n l L i A l
300 320 340 360 380 400 420
Temperature (K)

Figure 5-1-8. A) Temperature dependence of the electric resistivity for a polycrystalline
pellet of Sm3Ni5Allg. B) Temperature dependence of the thermoelectric power for a

single crystal of Sm3Ni5A119.

133

3

 

 

 

 

 

5 .14

’8
'5‘
E
“-4
'3
E
:3
E
3

E
X

0 50 100 150 200 250 300 350 400
Temperature(K)

B)
r; 20
3 C . .
E : . - '
‘3 15- .
O ' o
E ; .
2' 10'- .
e; » -
r: I '
.g 5- '
g o v V V V
’6 : 'v V ' ' '
é Or-‘VM
2 21....1....1....1....r....r....

 

 

 

0 110“ 2104 3104 410‘ 510‘ 610“
Field (G)

Figure 5-1-9. (A) Temperature dependent molar magnetic susceptibility and inverse
magnetic susceptibility of a polycrystalline sample of Er3Ni5A119. The susceptibility was
measured with an applied magnetic ﬁeld 1000 Gauss; (B) Field dependent magnetization
for Er3Ni5Allg at 5 K and 100 K.

134

A)

 

 

0.045 160
A 0.04 J 140
.2 1
E 0.035 . 1 120
c9. 0.03 v 1
"o" 1 100 v:
a 0.025 . x
\ “eff: 6-77 113 .' 30 3
:3
E 0.02 1
20.015 -
x 1 40
0.01 . 1
....... . . . . . . ..1
0005 . L l L l l_. . I . . l . . 1 . . 20

 

O 50 100 150 200 250 300

 

Temperature(K)
B)
E 2i
E 1.5:- ..
. .0
‘9‘ 1; .0
'6 E .0
E 0.5;-
: .
g -05 - .
E -1 - ,'.
o ..0
é ~15 :g.
2 _2:. .1...r...r...r...r.

 

 

 

-61044104-2104 0 2104 4104 6104
Field(G)

Figure 5-1-10. (A) Temperature dependent molar magnetic susceptibility and inverse
magnetic susceptibility of polycrystalline sample of Yb3Ni5A119. The susceptibility was
measured with an applied magnetic ﬁeld 1000 Gauss; (B) Field dependent magnetization
for Yb3Ni5A119 at 5 K.

135

5-1-4. Conclusions

We have presented in this part the compounds RE3Ni5A119 (RE = Sm, Dy, Br and
Yb) synthesized by the molten Al method. This series of compounds crystallize in the
Cmcm space group with the Gd3Ni5Al|9 structure type. The discovery of these
compounds proves the power of the metal ﬂux method to grow high quality large single
crystals of intermetallics.

For all the compounds the magnetic moments are from the RE ions and the Ni
atoms have ﬁlled d-orbitals. The Sm-analogue shows interesting magnetic properties: it
has two different magnetic orderings occurring at 5 K (antiferro-) and 150 K (ferro-) with
the applied magnetic ﬁeld perpendicular to the a-axis. When the applied ﬁeld is parallel
to the a-axis, these transitions are much less obvious and we argue that this effect is
because the spins are conﬁned to the bc-plane. In the case of Er and Yb analogues,
negative Weiss constant 0 values imply antiferromagnetic interactions although no
magnetic ordering is observed in both cases. Magnetic susceptibility measurements
indicate that the Yb ions of Yb3Ni5Al|9 are in intermediate oxidation state of Yb2+ and

Yb3+ which is in agreement with a previous report by Bauer.

136

References:

 

' Kanatzidis, M. (3.; Pottgen, R.; Jeitschko, w. Angew. Chem. Int. Edit. 2005, 44, 6996.

2 Niermann, J.; Fehrmann, 3.; Wolff, M. W.; Jeitschko, w. J. Solid State Chem. 2004,
177,2600.

3 Sieve, B.; Chen, X. Z.; Cowen, J .; Larson, P.; Mahanti, S. D.; Kanatzidis, M. G. Chem.
Mater. 1999, 11, 2451.

4 Sieve, B.; Sportouch, S.; Chen, X. Z.; Cowen, J. A.; Brazis, P.; Kannewurf, C. R.;
Papaeﬁhymiou, V.; Kanatzidis, M. G. Chem. Mater. 2001, 13, 273.

5 Sieve, B.; Chen, X. Z.; Henning, R.; Brazis, P.; Kannewurf, C. R.; Cowen, J. A.;
Schultz, A. J .; Kanatzidis, M. G. J. Am. Chem. Soc. 2001, 123, 7040.

6 Latturner, S. E.; Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G. Inorg. Chem. 2003, 42,
7959.

7a) Fehrmann, B.; Jeitschko, w. Inorg. Chem. 1999, 38, 3344. b) Thiede, v. M. T.;
Jeitschko, w. Z. Naturforsch. B 1998, 53, 673.

8 a) Thiede, v. M. T.; Ebel, T.; Jeitschko, w. J. Mater. Chem. 1998, 8, 125. b) Reehuis,
M.; Wolff, M. W.; Krimmel, A.; Scheidt, E. W.; Stusser, N.; Loidl, A.; Jeitschko, W. J.
Phys Condens. Matt. 2003, 15, 1773.

9 Fehrmann, B.; Jeitschko, w. Z. Naturforsch. B 1999,54, 1277.

'0 Gout, D.; Benbow, E.; Gourdon, 0.; Miller, G. J. J. Solid State Chem. 2003, 174, 471.
“ Rykhal', R. M.; Zarechnyuk, o. s.; Pyshchik, A. v. Dopov. Akad. NaukA 1973, 6,568.
12 a) Mizushima, A.; Isikawa, Y.; Maeda, A.; Oyabe, K.; Mori, K.; Sato, K.; Kamigaki,
K. J. Phys. Soc. Jpn. 1991, 60, 753. b) Fomasini, M. L.; Raggio, R.; Borzone, G. Z. Krist.
— New Cryst. St. 2004, 219, 75.

‘3 Yartys', v. A.; Pavlenko, v. v. Koordinats. Khim. 1992, 18,424.

'4 Yarmolyuk, Y. P.; Rykhal', R. M.; Aksel'rud, R. D.; Zarechnyuk, O. S. Dopov. Akad.
Nauk A 1981, 43, 86. b) Sikawa, Y.; Mizushima, A.; Sakurai, J .; Mori, K.; Munoz, A.;

Givord, F .; Boucherle, J. X.; Voiron, J .; Oliveira, I. S.; Flouquet, J. J. Phys. Soc. Jpn.
1994, 63, 2349.

137

 

'5 a) Takeshita, T.; Malik, s. K.; Wallace, w. E. J. Solid State Chem. 1978, 23, 271. b)
Achard, J. C.; Givord, F.; Percheron-Guegan, A.; Soubeyroux, J. L.; Tasset, F. J. Phys.
Paris 1979, 40, 218.

‘6 Dwight, A. E.; Mueller, M. 11.; Conner, R. A. jr.; Downey, J. W.; Knott, H. T. Metal].
Soc. Aime 1968, 242, 2075. b) Oesterreicher, H. J. Less-Common Met. 1973, 30, 225. c)
Maletta, H.; Sechovsky, V. J. Alloys Compd. 1994, 207, 254.

'7 Gladyshevskii, R. E.; Cenzual, K.; Flack, H. D.; Parthe, E. Acta Crystallogr. B 1993,
49, 468.

'3 Rykhal', R. M.; Zarechnyuk, o. s.; Kuten', Y. I. Dopov. Akad. NaukA 1978,40, 1136.
b) Rykhal', R. M.; Zarechnyuk, O. S.; Yanson, T. I. Dopov. Akad. Nauk A 1979, 41,
1057.

'9 a) Gladyshevskii, R. E.; Cenzual, K.; Parthe, E. J. Solid State Chem. 1992, 100, 9. b)
Bauer, E. D.; Bobev, S.; Thompson, J. D.; Hundley, M. F.; Sarrao, J. L.; Lobos, A.;
Aligia, A. A. J. Phys. .' Condens. Matter 2004, 16, 4025.

20 Tuan, N. C.; Sechovsky, V.; Divis, M.; Svoboda, F.; Nakotte, H.; de Boer, F. R.; Kim-
Ngan, N. H. J. Appl. Phys. 1993, 73, 5677.

2' Kolomiets, A. V.; Havela, L.; Yartys, V. A.; Andreev, A. V. Zh. Fiz. Doslidzhen 1999,
3, 458.

22 Steglich, F.; Geibel, C.; Gloos, K.; Olesch, G.; Schank, C.; Wassilew, C.; Loidl, A.;
Krimmel, A.; Stewart, G. R. J. Low Temp. Phys. 1994, 95, 3.

23 a) Tolinski, T.; Schafer, W.; Kockelmann, W.; Kowalczyk, A.; Hoser, A. Phys. Rev. B
2003, 68, 144403. b) Tolinski, T.; Schafer, W.; Kowalczyk, A.; Andrzejewski, B.; Hoser,
A.; Szlaferek, A. J. Alloys Compd. 2004, 385, 28.

24 Tolinski, T.; Kowalczyk, A.; Chelkowska, G.; Pugaczowa-Michalska, M.;
Andrzejewski, B.; Ivanov, V.; Szewczyk, A.; Gutowska, M. Phys. Rev. B 2004, 70,
064413.

25 YbN13A1|9 crystallizes in the ErNi3A119 structure type; the synthesis, crystal structure
and physical properties of this compound will be described in Part II of this chapter.

26 Fomasini, M. L.; Raggio, R.; Borzone, G. Z. Krist. - New Cryst. St. 2004, 219, 77.
27 SMART, version 5; Siemens Analytical X-ray Systems, Inc.: Madison, WI, 1998.

28 Saint, Version 4; Simens Analytical X-ray Instruments Inc., Madison, WI.

138

 

29 SADABS, Sheldrick, G. M.; University of Gottingen, G6ttingen, Germany.

30 GM. Sheldrick, 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

3‘ CrystalDiffract is © 1995-1996, Dr. David C. Palmer.

32 Rykhal, R. M.; Zarechnyuk, o. s. Yarmolyuk, Y. P. Sov. Phys. Crystallogr. Sect. C
1992,17,453.

33 Gladyshevskii, R. E.; Cenzual, K.; Parthé, E. Acta Crystallogr. Sect. C1992, 48, 232.
3“ Newkirk, J. 3.; Black, P. J.; Damjanovic, A. Acta Crystallogr. 1961, 14, 532.

35 Bourdreaux, E. A.; Mulay, L. N. In Theory and Applications of Molecular
Paramagnetism; John Wiley and Sons: New York, 1976.

36 Sales, B. C.; Wohllebent, D. K. Phys. Rev. Lett. 1975, 35, 1240.

139

CHAPTER FIVE

Exploratory Studies on the Ternary System RE/Ni/Al
Employing A] as a Flux

PART II. Flux Synthesis and Characterization of a New Ternary Phase
th.tNi3Als.9

5—2-1. Introduction

In the last two decades, there has been increasing interest in Yb-containing
intermetallic compounds due to a variety of interesting properties they display such as
heavy fermion,l intermediate valence2 and Kondo lattice behavior.3 These characteristics
are associated with the fact that for this element 4f'3 and 4f‘4 electronic states are close in
energy and they hybridize very easily with the conduction electrons. In spite of the
interesting properties, studies on Yb-based compounds are very limited, possibly because
of the difﬁculty in synthesizing Yb-containing compounds due to the high vapor pressure
of this element.4 To overcome this problem, we suggested using metal (A1, Ga, In) as a
ﬂux to synthesize Yb-based compounds. 5 This method allows the reaction to be
conducted below 1000 °C, much lower than the traditional method (induction furnace,
arc-welder). Besides this high temperature solution facilitates the growth of single
crystals; thus allowing the structure determination to be much easier and more reliable.

One of the systems which people have been interested in is the ternary family Yb-
TM-X (TM = transition metal, X = A1, Ga, In). Steglich and coworkers have investigated
tentatively the Yb-Ni-Al ternary phase diagram which is shown in Figure 5-2-1.6 A

slightly curved line starting near YbNi and passing close to szNizAl and YbNiAl

140

separates the region with mixed valent Yb (on the Yb and Al-rich side) from the region
with trivalent Yb (on the Ni-rich side). The valence of Yb for szNizAl is between +3
and +2; studies of szNizAl indicate that it is a heavy fermion compound without any
indication of magnetic ordering down to 2 K.6 YbNiAl orders antiferromagnetically and
shows metamagnetic behavior in a magnetic ﬁeld.7 YbNiAlz shows a magnetically
ordered ground state with stable trivalent Yb ions.6 Although Steglich and his coworkers
claimed that they found a simple composition dependence of the Yb-valence in the
ternary Yb-Ni-Al system as depicted in Figure 5-4-1, some compounds such as
YD3N15A119 do not ﬁt into this diagram.8 Yb3Ni5A119 is a member of the RE3N15A119
family which was described earlier in this dissertation and also reported by Bauer and
coworkers. According to the phase diagram by Steglich, the Yb ions in this compound
are trivalent (Yb 11.1%, Ni 18.5%, A1 70.4%), however magnetic susceptibility
measurements indicated intermediate valence behavior of the Yb ions.

Previous studies employing the Al ﬂux method reveal that Al tends to serve as a
reactive solvent which usually incorporates into the product.9 Therefore Al ﬂux is an
appropriate tool to synthesize ternary compounds in the Yb-Ni-Al system. Our
exploratory studies on this system led to the discovery of a new ternary compound
YbHN13A139, which is a disordered variant of the ErNi3A19 structure type.10 Herein we

report synthesis, crystal structure and magnetic properties of Yb._1Ni3A13,9.

141

  
 

 

\ 4e
o\° LO ’\'
"~ 60 Mixed valent /
.8 at.
39,0 221 ,, 60 '3}
,”‘ 111 .\
20 ° , 2°
/ Trlvulent .
o e 1 . LU“
Al 100 30 60 ’00 20 0 Ni
Al-(ot °/e)

Figure 5-2-1. Valence of binary (circles) and ternary (triangles) compounds in the Yb-Ni-
A1 phase diagram. Closed symbols: trivalent Yb; open symbols: mixed valent of Yb.

Dotted line indicates the border between mixed-valent and trivalent Yb.

5-2-2. Experimental Section

Reagents:

The following reagents were used as obtained: Yb (Cerac, 99.9%), Ni (99%, 325

mesh, Sargent, Buffalo Grove, IL), A1 (Cerac, 99.5%, -20 mesh).

Synthesis:

In a nitrogen-ﬁlled glove box, the reaction mixture containing 1 mmol Yb metal
(0.173 g), 2 mmol Ni (0.118 g), 10 mmol Al (0.270 g) was combined into an alumina
crucible. The crucible was then placed into a silica tube (13 mm in diameter), which was
sealed under vacuum (~10‘4 Torr). The sample was subjected to the following treatment:

it was heated to 1000 °C in 15 h, maintained at this temperature for 5 h, and then cooled

142

to 850 °C in 2 h. It was annealed at 850 °C for 3 (1, followed by cooling down to 500 °C
in 36 h. Finally the temperature was brought down to 50 °C in 10 h. YbHNi3Alge
crystallized as a major phase in 50% yield with Yb3N15A1|9 and YbAl3 as minor phases.
The excess aluminum was removed by soaking the crucible in aqueous 5M NaOH
solution overnight. The remaining crystalline products after the isolation procedure were
rinsed with water and acetone. Single crystals were selected for elemental analysis, X-

ray diffraction and magnetic susceptibility measurements.

Scanning Electron Microscopy and Elemental Analysis:

The crystals were picked and afﬁxed on a Scanning Electron Microscope (SEM)
sample plate using carbon tape. Chemical compositions of the products were determined
by Energy Dispersive Spectroscopy (EDS) performed on a JEOL JSM-35C SEM
equipped with a NORAN EDS detector. Data were acquired by applying a 25 kV
accelerating voltage with an accumulation time of 30 5. Several crystals were analyzed
and averaged with the resulting approximate atomic ratio of l : 3.25 : 8.92 for Yb : Ni :
Al, which agreed fairly well with the results derived from the single crystal X-ray
analysis. The crystals of YbHNi3A139 are metallic with shiny and smooth surfaces, as

shown in Figure 5-2-2.

143

 

Figure 5-2-2. SEM image of a typical crystal of YbHN13A189.

X—ray Crystallography:

X-ray diffraction data were collected at 173 K on a Bruker AXS SMART CCD X-
ray diffractometer. A single crystal 0fo]_|N13A13_9 with the size 0.24 x 0.16 x 0.16 mm3
was cut from a larger crystal and mounted on glass ﬁbers. A data collection (Mo K01
radiation, A = 0.71073 A) was acquired covering a full sphere of reciprocal space. The
SMARTll software was used for data acquisition and cell reduction, and the integration
was performed with the SAINTPLUS software package.12 An empirical absorption
correction was applied to the data using the SADABS program.[3 The structures were
solved using direct methods and reﬁned with the SHELXTL package program.14

Five space groups were given by XPREP program as possible candidates: R-3,
R3, R3m, R32, R-3m. Initially two centrosymmetric space groups R-3 and R-3m were

examined which also have the lowest CFOM values. However no satisfactory solutions

144

could be obtained. So the chiral space group R32 was chosen which was consistent with
that of ErNi3A19 reported by Gladyshevskii.10 In the structure of Yb 1, 1N13A139 a total of
nine atomic sites were identiﬁed. The 6c Yb(l) site, 1 8f Ni(1) site, 18fAl(l), 9d Al(3),
and three 6c Al sites — Al(4), Al(5) and Al(6) were assigned unambiguously and found
fully occupied. Disorder was observed between the Yb(2) and Al(2) atoms: the distance
between the Yb(2) and Al(2) atoms was found be to very close at 1.522(2) A;
furthermore, both sites were found partially occupied with Yb(2) 10% and Al(2) 90%.
So the Yb(2) and Al(2) sites were constrained to make the occupancy sum of Yb(2) and
Al(2) equal to 1. Similar disorder was observed in the isostructural compounds YNi3A19
and DyNi3A19, while in both cases an additional Al(7) site was found which also
exhibited partially disordered arrangements with RE(l) atom in a triangular mesh.lo
Interestingly the Er and Gd analogues have ordered arrangements of rare earth and
aluminum triangles: only one RE site, one Ni site and six Al sites were identiﬁed and all
were reﬁned fully occupied.

In the structure of Yb],.Ni3A13_9, all atomic sites were reﬁned anisotropically.
Data collection parameters and reﬁnement details for YbHN13A139 can be found in Table
5-2-1. Atomic positions, displacement parameters and anisotropic displacement
parameters are listed in Tables 5-2-2 and 5-2-3.

X-ray powder diffraction data were collected at room temperature on a CPS 120
IN EL X-ray diffractometer (Cu Ka) equipped with a position-sensitive detector.
Experimental powder patterns were compared to the patterns calculated from the single
crystal structure solution (by the CrystalDiffract15 software program) to determine the

phase identity and purity.

145

Table 5-2-1. Selected crystal data and structure refinement details for Yb.,1Ni3Alg_9.

 

 

Empirical formula YbHNi3A139
Formula weight 591.99

Crystal system Rhombohedral
Space group R32 (#155)

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient
F(000)

Crystal size

Theta range for data collection
Limiting indices

Reﬂections collected
Independent reﬂections
Completeness to theta = 28.12
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F 2

Final R indices [I>2$igma(l)]
R indices (all data)

Largest diff. peak and hole

a = 7.2314(4) A

c = 27.140(3) A

1229.1 1(17) A3

6

4.799 Mg/m3

18.975 mm"

1626

0.24 x 0.16 x 0.16 mm3

3.34 to 28.12°
95h59
osks9
8651535

3654

635 [R(int) = 0.0522]

95.5 %

F ull-matrix least-squares on F2
635 / O / 42

1.211

R. = 0.0260, wR2 = 0.0606

R1 = 0.0267, wR2 = 0.0616
2.299 and —1.840 e.A'3

221/2

R1 = 2(lrol-lrclyzlrol; wR2 = [Z[W(Foz' Fc21/ [2(wlFoI ) 1

146

Table 5-2-2. Atomic coordinates (A x 104) and equivalent isotropic displacement
parameters (A2 x 103) for YbHNi3A139.

 

 

 

 

 

Atom Wyk. x y z U(eq)t Occu
Symbol
Yb(l) 6c 0 O 1668(1) 10(1) 1
Ni 18f 3334(2) 48(2) 861(1) 7(1) 1
Al(l) 18f 3329(4) 3348(4) 1003(1) 9(1) 1
Al(3) 9d 3340(3) 0 0 9(2) 1
Al(4) 6c 0 0 522(1) 8(1) 1
Al(5) 6c 0 0 2815(1) 8(1) 1
Al(6) 6c 0 0 3865(1) 9(1) 1
Yb(2) 3b 0 0 5000 3(1) 0.097
A142) 9e 2 1 05(3) 0 5000 3(1) 0.903
U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.
Table 5-2-3. Anisotropic displacement parameters (A2 x 103) for Yb1,1Ni3Alg_9.
Atom Ull U22 U33 U23 Ul3 U12
Yb(l) 11(1) 11(1) 8(1) 0 0 5(1)
Ni 8(1) 8(1) 5(1) 0(1) 0(1) 4(1)
Al(l) 7(1) 9(1) 13(2) 4(1) 1(1) 4(1)
Al(3) 10(1) 11(2) 8(2) -1(1) 0(1) 5(1)
Al(4) 9(1) 9(1) 7(1) 0 0 4(1)
Al(5) 8(1) 8(1) 8(1) 0 0 4(1)
Al(6) 9(1) 9(1) 8(1) 0 0 5(1)
Yb(2) 3(1) 3(1) 2(1) 0 0 2(1)
Al(2) 3(1) 3(1) 2(1) 0 0 2(1)

 

The anisotropic displacement factor exponent takes the form:
-27t2[h2a"‘2Ul 1+. . .+2hka*b*Ul2]

147

Magnetic Characterization:

Magnetic measurements were conducted on polycrystalline samples of
YbHNigAlgg. Field-cooled and zero-ﬁeld cooled dc magnetization measurements were
performed for the above samples using a Quantum Design MPMS SQUID magnetometer.
EDS-analyzed crystals were ground into powder, which was sealed in kapton tape and
placed into the magnetometer. The data were collected in the temperature range 3-300 K
at 1000 G, while ﬁeld dependent magnetic measurements, conducted at 5 K, were carried
out in ﬁelds up to i 55000 G. A diamagnetic correction was applied to the data to

account for core diamagnetism.

5-2-3. Results and Discussion
Crystal Structure:
The crystal structure Ofo|_|N13A13_9 viewed down the b-axis is shown in Figure
5-2-3. This structure is related to the ErNi3A19 structure type with a partially disordered
triangular arrangement of Yb and Al atoms.IO As described by Gladyshevskii, this
structure can be viewed as three different kinds of triangular layers stacking along the c-
axis: A] layer, Ni layer and Yb-Al layer. Al(l) and Al(3) atoms form similar triangular
arrangements on the ab plane with Al-Al distance longer than 4 A, as shown in Figure 5-
2-4A. Al(4), Al(5) and Al(6) atoms also form triangular patterns, with Al(4)-Al(4) bond
distance of 2.831(6) A and Al(5)-Al(6) distance of 2.848 (6) A. The Ni layer, is inserted
between the Al(l) layer and the Al layer composed of Al(4), Al(5) and Al(6) atoms.
What makes this compound interesting and different from the structure of

ErNi3A19 is the triangular mesh composed of Yb and A1 atoms. As shown in Figure 5-2-

148

5A, in the Er2A13 layer the Er and Al atoms form ordered triangular mesh with regular
bond distances between Er and Al atoms of 2.977(2) A. However for the Y- and Dy-
analogues, two additional atomic sites, i.e. one rare earth metal and one Al site, were
detected.'0 Figure 5-2-5C depicts the disordered arrangement of RE and Al atom
triangles, for the Dy compound the sites RE(l) and RE(2) were found to have the
occupation of 80% and 40%, respectively; while for the Y-analogue the corresponding
occupation factors were 97% and 6%. While in the present compound YbHNi3Alg.9, the
Yb(l) site is fully occupied and a partly disordered triangular arrangement of Yb-Al
atoms is observed, producing unreasonably short distance between Yb(2) and Al(2)
atoms: 1.522(2) A. The occupancy factors of Yb(2) and Al(2) are 10% and 90%,
respectively.

The compounds ErNi3A19, DyNi3A19, YNi3A19 and Yb1_1N13A13_9 crystallize in the
same structure type, however with a different type of disorder in the RE-Al triangle layer.
It was suggested by Gladyshevskii that this disorder originates from the non-

10 For a structure with

homogeneous stacking of the RE2A13 layer along the c-axis.
maximum disorder of RE and Al atoms triangles, both RE sites would have a occupation
of 67% and Al sites 33%. This results in a structure with a space group P-6m2 and
smaller unit cell: a = a’/\/3, c = c’/3 (a’ & c’ are the cell parameters of ErNi3A19). This
argument has been conﬁrmed by RE0_67NizGa6-thx (Tt = Si/Ge) in which a complete
disorder was observed in the RE-Ga plane. The random stacking of RE-Ga planes along

the c-axis causes the elongated diffuse streaks along this direction on the MI and 0k] zone

photographs. '6

149

The local coordination environments of Yb(l) and Ni atoms are shown in Figure
5-2-5D & E. The Yb atoms are surrounded by 17 atoms which form a bicapped hepta-
layer geometry. These ﬁve monoatomic layers are in the order of Ni-Al(1)-Al(2)-Al(l)-
Ni atoms with Al(4) and Al(5) as capping atoms. The bond distance between Yb(l) and
Al(2) is 2.9566(16) A, slightly shorter than the other Yb-Al distances. The Ni atoms sit
in a distorted cubic geometry composed of eight Al atoms with Ni-Al distance ranging

from 2.3369(15) A to 2.6033(17) A.

Table 5.2.4. Bond lengths (A) for Yb.,.Ni3Alg_9.

 

 

Bond Distance Bond Distance
Yb(l)-Al(l) 3.012(3) Ni(1)—Al(5) 2.5679(17)
Yb(1)-A1(2) 2.9566(16) Ni(1)-Al(6) 2.6033(17)
Yb(1)-Ni(1) 3.23 86(13) Al(1)-Al(2) 2.766(3)
Yb(2)-Al(1) 2.999(3) Al(1)-Al(4) 2.746(4)
Yb(2)-Al(2) 1.522(2) Al(1)-Al(5) 2.754(3)
Yb(2)-Al(6) 3.082(3) Al(6)-Al(1) 2.719(3)
Yb(2)-Ni(1) 3.2800(13) Al(3)-Al(4) 2.800(2)
Ni(1)-Al(1) 2.419(2) Al(3)-Al(5) 2.7883(18)
Ni(1)-Al(1) 2.451 (2) Al(3)-Al(6) 2. 8065(1 8)
Ni(1)-Al(1) 2.454(2) Al(4)-Al(4) 2.831(6)
Ni(1)-Al(2) 2.3734(17) Al(5)-Al(6) 2.848(6)
Ni(1)-Al(3) 2.3369(15) Al(2)-Al(2) 2.636(4)
Ni(1)-Al(4) 2.5649(17)

 

150

A10)

Al(4). Al(5) and A1(6)

‘Al(3)

“MONONQWM. ..‘N.N.M®
VON MQNOWOW ..NONOMO
9...”?! 000......
WWW... WWW.” ch
«4 a . .D»... .WWm.ao ..wmuuw »

z

 

 

ewed down the [010] direction. Large

cles: Yb; black circles: Al; gray circles: Ni.

Figure 5-2-3. Crystal structure of YbHNi3Algg vi

empty cir

151

A)
O O O
X
0 O 0
4.148 A 4.189 A
.4——> -4—> . o
y I O O
O O
B)
" Al(6)" 0’
Al(5).’ ‘0
Al(4)., I
0’ f .1
O”. ."
0'. ." y
0’. ." .’
X

Figure 5-2-4. Structure of YbHNi3Algg. A) Triangular layer composed of Al(l) atoms. B)
Triangular layer composed of Al(4), Al(5) and Al(6) atoms down the c-axis.

152

$2.22.; a was. 2 85 E.; 80
35:59:25 585308 803 Am 0% AD .38...» 05 :38 @0303 823739 E 283 _<-m.m 35285 G .3820 05 :32. 3303

323229» E 28E _<-m_.m 38986 383.8.— Am 28.9 2.: :38 3303 £<£Zcm E 283 37mm 3830 A< .m-m-m 035

iii!
‘1'!
a
..._<._z_._§
O
,. V.
O O
O
0.
ox O
O
O WW0.
\\ \
O

 

153

Magnetic Properties:

The temperature dependent molar magnetic susceptibility xm for YbHNi3Algg is
shown in Figure 5-2-6A. Above 50 K, the magnetic susceptibility can be ﬁt to a Curie-
Weiss law yielding an effective magnetic moment ,ueff = 4.14 ,uB/mol Yb, slightly lower
than the theoretical value for Yb3+ (peg = 4.54 ﬁg). This implies that the Yb ions in
YbHNi3A139 are in an intermediate oxidation state of Yb2+ and Yb“. This compound is
another example that does not ﬁt into the diagram suggested by Steglich.6 The negative
Weiss constant (9 = -18.41 K) indicates the presence of weak antiferromagnetic
interactions between the Yb ions.

The isothermal magnetization of Yb1.1Ni3A13_9 conducted at 5 K up to 55000 G is
plotted in Figure 5-2-6B. Clear metamagnetic behavior without hysteresis effects is
observed. This metamagnetic transition is more obvious when plotting the derivative
(8W6B)T vs B shown in the inset of Figure 5-2-6B. The sharp peak at 1000 G (T = 5 K)
corresponds to the ﬁeld-induced change in the magnetization. Similar metamagnetic
transition is also observed in szRh3A19 and Yb21r3A19; however this transition
disappears as TN is reached.2 The magnetization value M for Ybl, 1Ni3A139 at 55000 G

reaches a value of 0.9 ,uB/Yb, which is much lower than the ordered value which is 4.5

Ala/Yb.

154

A)

 

   
   
   

 

0.6 ‘ ‘ 140
05 u =4_14 BM. 1 12°
9eff
A =-1 .4
§ 04 8 'K 100
'8 .
E 80 C
3 0.3 ‘ 3x
E 60
3
0.2
XE 40
0.1 20
0 O
0 50 100 150 200 250 300
Temperature (K)

 

0.5 '

Moment (B. M. / mol Yb)
o

 

 

 

-1 ,' .
-60000 -3 0000 0 30000 60000
Field (G)

Figure 5-2-6. A) Temperature dependent magnetic susceptibility of a polycrystalline
sample of YbLlNl3Algg under the ﬁeld of 1000 G. B) Molar magnetization of
Yb...Ni3Alg.9 in ﬁelds up to 55000 G measured at S K. inset: derivative (6M/6B) as a

function of ﬁeld.

5-2-4. Conclusions

155

Single crystals of a new ternary intermetallic compound YbHNigAlgg have been
synthesized by the A1 ﬂux method. This compound has a crystal structure related to that
of ErNi3Alg and the whole structure can be viewed as a stack of monoatomic layers along
the c-axis. As in the isostructural compounds DyNi3A19 and YNi3Alg, a partially
disordered triangle arrangement of Yb and Al atoms was observed for YbHNigAlgg on
the Yb-Al plane. In this disordered model every Yb(2) atom is surrounded by three Al(2)
atoms with very close distance of 1.522(2) A; and the occupancy factors of Yb(2) and
Al(2) were found 10%/90%.

Temperature dependent magnetic measurements reveal that the Yb ions in this
compound are in an intermediate oxidation state of Yb2+ and Yb“. As in the other
intermetallic compounds, transition metal Ni atoms do not contribute to the magnetic
moment. The negative Weiss constant 6 (-18.41 K) implies weak antiferromagnetic
interactions. A clear ﬁeld-induced metamagnetic transition occurs at 1000 G and a ﬁeld

higher than 55000 G is required to saturate the magnetic moments.

156

References:

 

I Kaczorowski, D.; Andraka, B.; Pietri, R.; Cichorek, T.; Zaremba, V. 1. Phys. Rev. B
2000, 61, 15255.

2 Trovarelli, 0.; Geibel, C.; Buschinger, B.; Borth, R.; Mederle, 8.; Grosche, M.; Spam,
G.; Steglich, F.; Brosch, O. Donnevert, L. Phys. Rev. B 1999, 60, l 136.

3 Singh, Y.; Ramakrishnan, 8. Phys. Rev. B 2003, 68, 054419.
4 Fisk, 2.; Maple, M. B. J. Alloys Compd. 1992, 183,303.
5 Kanatzidis, M. G.; Pottgen, R.; Jeitschko, W. Angew. Chem. Int. Ed. 2005, 44, 6996.

6 Geibel, C.; Klinger, U.; Buschinger, M.; Weiden, M.; Olesch, G.; Thomas, F .; Steglich,
F. Phys. B 1996, 223 & 224, 370.

7 a) Diehl, J.; Davideit, H.; Klimm, S.; Tegel, U.; Geibel, C.; Steglich, F .; Horn, S. Phys.
B 1995, 206 & 207, 344. b) Schank, C.; Olesch, G.; Kohler, J.; Tegel, U.; Klinger, U.;
Diehl, J.; Klimm, 8.; Spam, G.; Horn, 8.; Geibel, C.; Steglich, F. J. Magn. Magn. Mater.
1995, 140-144, 1237.

8 Bauer, E. D.; Bobev, S.; Thompson, J. D.; Hundley, M. F.; Sarrao, J. L.; Lobos, A.;
Aligia, A. A. J. Phys. .' Condens. Matter 2004, I6, 4025.

9 Sieve, B. Ph. D. Dissertation, Michigan State University, 2002.

‘0 Gladyshevskii, R. E.; Cenzual, K.; Flack, H. D.; Parthé, E. Acta. Com. 1993, B49, 468.
” SMART, Version 5; Siemens Analytical X-ray Systems, Inc.: Madison, WI, 1998.

'2 Saint, Version 4; Simens Analytical X-ray Instruments Inc., Madison, WI.

'3 SADABS, Sheldrick, G. M.; University of Gottingen, Gottingen, Germany.

'4 G.M. Sheldrick, 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

‘5 CrystalDiffract is © 1995-1996, Dr. David C. Palmer.

'6 Zhuravleva, M. A.; Chen, X. Z.; Wang, X.; Schultz, A. J.; Ireland, J.; Kannewurf, C.
K.; Kanatzidis, M. G. Chem. Mater. 2002, 14, 3066.

157

CHAPTER SIX

PART I. Doping Studies of Yb3Ni5A119 with Cu, Fe and Mn Substitutions

6-1-1. Introduction
In the last two decades, the strongly correlated f—electron systems, especially Ce

' These compounds often

and Yb-based ones, have attracted considerable attention.
exhibit interesting behavior such as Kondo lattice,2 heavy fermion3 and mixed valence.4
These properties are a result of the competition between the magnetic ordering
(Ruderman-Kittel-Kasuya-Yosida interaction) and the Kondo effect.5 The Kondo effect
results from the exchange interaction between the conduction electrons and the local
magnetic moments leading to scattering events in which the electron spin is ﬂipped. In a
Kondo system, the resistivity drops to a minimum at low temperature and then proceeds
to rise at lower temperatures. These two effects (the RKKY interaction and the Kondo
effect) are characterized by the hybridization strength J between the 4f electrons and
conduction electrons. For the lower J values, i. e. the RKKY interaction dominates, the
compound can exhibit magnetic ordering; on the other hand, for the higher J values, the
Kondo effect takes over, and the ground state is nonmagnetic. This phenomenon is
clearly described in the well-known ‘Doniach’s magnetic phase diagram’.5

The J value (hybridization strength) of a compound can be tuned by applying
external pressure or chemical pressure. For Ce-based compounds, applying external
pressure (or positive chemical pressure) can cause them to switch from magnetic ordering

to intermediate valence behavior (Kondo effect). Generally, with decreasing cell volume

the 4f electrons and conduction band hybridization J value increases, favoring the Kondo

158

effect. This theory is clearly exempliﬁed by Ceszsiz and CeRhZSiz; in both cases the
pressure induces a transition from antiferromagnetism to a nonmagnetic state.6 More
interestingly, these compounds become superconductors at pressures close to the critical
pressure where magnetic ordering vanishes. The chemical pressure effect on CeNi has
been extensively studied by substituting Ni with Cu or Pt. 7 CeNi crystallizes in
orthorhombic CrB-type structure with Ce ions in intermediate valence states. Since the
size of Cu or Pt is larger than Ni, nonmagnetic to magnetic ordering transition is expected
as this “negative pressure” is applied. Indeed, in CeNiHPtx compounds, a continuous
evolution from the intermediate valence behavior in CeNi to a ferromagnetically ordered
Ce3+ state in CePt was found, passing through a Kondo lattice behavior and heavy
fermion state. 8 These properties can be well understood by modiﬁcations of 4f-
conduction band hybridization driven by cell volume changes.

The Yb-based systems, usually present mirror-like behavior to Ce—compounds due
to their hole-electron symmetry: P (Ce3+) vs fl3 (Yb3+). Therefore, under similar pressure
conditions, the opposite effect is expected for Yb-based compounds, which means that
external pressure might drive a non-magnetic Yb-system to a magnetically ordered state.
This phenomenon has been observed and reported for some Yb-containing intermetallic
compounds. For example, szNizAl is a nonmagnetic heavy fermion compound
crystallizing in orthorhombic MoszNi structure type.9 When the applied pressure is
more than 8 GPa and the temperature is below 2 K, a crossover from a nonmagnetic state
to magnetically ordered state is observed; however this crossover is not connected with a
change of the Yb valence state. YbCuzsiz is another well-known nonmagnetic compound

with Yb in an intermediate valence state.lo When the external pressure is higher than 8

159

GPa, this compound is also converted from a nonmagnetic to magnetic ordered state with
localized Yb moments. In this case, the transition is accompanied by a valence change

towards the Yb“ state.ll

Chemical substitution may result in a similar effect, such as
Yb(Cu1-xNix)ZSi2, in which the chemical pressure obtained by gradual substitution of Cu
by Ni changes the nonmagnetic ground state of YbCuZSiz to the magnetically ordered one
of YbNi28i2 ( TN = 2.1 K).12 Different from the examples described above, Bauer and
coworkers found that the substitution of Cu by A1 in YbCus caused a crossover from non-
magnetic 4fl4 state to the magnetic 4fl3 state in YbCu3A12.13 Since the substitution of Cu
by A1 results in an expansion of the cell volume, chemical pressure might be excluded as
a driving force of this change of ground state properties; instead it was suggested that the
physical origin was the modiﬁcation of the electronic structure of the compound which
was associated with different number of electrons provided by Al and Cu.

In Part I of Chapter Five, we described a ternary intermetallic compound
Yb3Ni5A119, in which the Yb ions show intermediate valence behavior.14 It might be
interesting to investigate the magnetic behavior of this Kondo compound by applying
chemical pressure, thus we can better understand the correlations between compositions
and magnetic states. In this section, we will describe the effects of substitution of Ni by
Cu, Fe or Mn on the magnetic properties of Yb3Ni5A119. By applying this chemical
pressure, two effects are expected. On one hand, the metallic radii of Mn, Cu, Fe and Ni
are 1.37 A, 1.28 A, 1.26 A and 1.25 A, respectively. By substituting Ni with Mn, Fe or
Cu, the unit cell volume is expected to expand slightly in all the cases, which corresponds
to negative pressure. For the Yb-based systems, it is customary to decrease J values by

applying external pressure. On the other hand, the number of electrons provided by the

160

substitutional element (Mn, Fe or Cu) is different from that of Ni, which might change
the electronic structure of the parent compound. These two effects can be competing and
lead to interesting properties induced by valence ﬂuctuations in the Yb ions.

Here it is noteworthy to mention another important reason that we initiated this
study. YbGaGe was reported to show zero thermal expansion (ZTE) between 10 and 300
K.15 Magnetic measurements indicated that Yb ions were in an intermediate valence state
and it was proposed that the ZTE behavior was derived from internal electronic charge
transfer of the Yb ions. Later it was found that only when graphite crucibles were used as
the reaction container, could the results be repeated. Although Fisher and his coworkers
found that lightly C- or B- doped YbGaGe samples tended toward zero volume

'6 we suspect that Fe — which is the second most abundant impurity element

expansion,
in the graphite crucible, might inﬂuence the unusual thermal expansion property. It is
known that with substitutions of transition metals, the thermal expansion coefﬁcient

7 Neutron diffraction analysis

could be changed. CeNi..,,Cux is such an example.1
revealed that the relative changes of the cell parameters for CeNiogCum between 300 K
and 1.5 K were smaller than those observed for CeNi. However there has been no
discussion about the mechanism behind this observation and no further investigations
were reported on the thermal expansion of this system. So we investigated the

temperature dependent lattice variation of Yb3Ni5-xTM,,A119 (TM = Cu, Fe, Mn) as well

as their magnetic properties.

161

6-1-2. Experimental Section
Reagents:

The following reagents were used as obtained: Yb (Cerac, 99.9%), Ni (99%, 325
mesh, Sargent, Buffalo Grove, IL), Cu (99.5%, -325 mesh, Cerac, Milwaukee, WI), Fe
(99.99%, ﬁne powder, Aldrich Chemical, Milwaukee, WI), Mn (99.95%, ﬁne powder,

Cerac, Milwaukee, WI), Al (Cerac, 99.5%, -20 mesh).

Synthesis:

The elemental starting materials were stored and manipulated in a nitrogen-ﬁlled
single dry box. All the compounds were synthesized by the Al ﬂux method. The ratios
used for Yb : Ni : TM : Al (TM = Cu, Fe, Mn) were summarized in Table 6-1-1. The
constitutional metals were combined into alumina crucibles, which were then placed into
silica tubes (13 mm in diameter) and sealed under vacuum (~104 Torr). The samples
were heated to 1000 °C in 15 h, maintained at this temperature for 5 h, then cooled to 850
°C in 2 h. They were annealed at 850 °C for 3 d, followed by cooling down to 500 °C in
36 h. Finally the temperature was brought down to 50 °C in 10 h. The yield of the
reactions was ~70% with YbAl3 as the side product. When x in Yb3Ni5-xFexA119 was 1,
another side product YbNi2-xFexAlg appeared which will be described in Part II of this
chapter.

Excess aluminum was removed by soaking the crucible in aqueous 5M NaOH
solution overnight. The crystalline product remaining after the isolation procedure was
rinsed with water and acetone. Single crystals were selected for elemental analysis,

single crystal X-ray diffraction analysis and magnetic susceptibility measurements.

162

Table 6-1-1. Elemental ratios for the ﬂux synthesis of Yb3Ni5Al19, Yb3Ni5-xCuxAllg,
Yb3Ni5-xFexAl|9 and Yb3Ni5-anxAl.9 (unit: mmol).

 

 

 

 

 

 

 

 

 

 

 

 

Compound Yb Ni Cu/F e Al
Yb3Ni5A119 1 1 0 10
Yb3Ni5-xCuxAllg (x = 0.5) 1 0.9 0.1 10
Yb3Ni5-xCuxAl|9 (x = 1) 1 0.8 0.2 10
Yb3Ni5-xFexAl|9 (x = 0.5) 1 0.9 0.1 10
Yb3Ni5-xFexA119 (x = 1) 1 0.8 0.2 10
Yb3Ni5-anxA119(x = 0.5) 1 0.9 0.1 10

 

Single Crystal X-ray Crystallography:

Single crystal X-ray diffraction data were collected for Yb3Ni5-xTMxAl.9 (TM =
Cu, Fe, Mn) at different temperatures (100 K, 173 K and 298 K) on a Bruker AXS
SMART CCD X-ray diffractometer. A data collection (Mo Ka radiation, A. = 0.71073 A)
was acquired covering a hemisphere of reciprocal space. The data acquisition and cell
reduction were done with the SMART'8 software package and data processing was
performed with the SAINTPLUS software package. '9 An empirical absorption correction
was applied to the data using the SADABS program.20 The structure of Yb3Ni5Al.9 was

used as the starting model and then reﬁned with the SHELXTL package program.2|

Magnetic Measurements:
Magnetic measurements were conducted on the polycrystalline samples of

Yb3Ni5-xTMxA119 (TM = Cu and Fe). Field-cooled and zero-ﬁeld cooled dc

163

 

magnetization measurements were performed on the samples using a Quantum Design
MPMS SQUID magnetometer. EDS—analyzed crystals were ground into powder, sealed
in a kapton tape and placed into the magnetometer. The data were collected in the
temperature range 3-300 K at 1000 G, while ﬁeld dependent magnetic measurements,
conducted at 5 K, were carried out in ﬁelds up to i 55000 G. Diamagnetic corrections

were applied to the data to account for core diamagnetism and kapton tape.

6-1-3. Results and Discussion
Synthesis and Compositional Variations:

As was discussed in Part I of Chapter Five, large single crystals of Yb3Ni5Allg
could be obtained by using a molar ratio of 1 : 2 for Yb : Ni. So the ﬁrst series of
substitutional reactions were conducted based on the following ratios: Yb : Ni : Cu : Al (1
: 2-x : x :10), x = 0.2, 0.4, 0.6 and 0.8. However powder X-ray diffraction pattern
showed no presence of Yb3Ni5-xCuxA119. Instead elemental analysis and single crystal X-
ray analysis indicated that the products were YbNi4Al8 and Yb(NiCu)4Al8 which
crystallize in the tetragonal Tthn structure type.22 It is obvious that the disruption
coming from the large amount of Cu is enough to destabilize the original structure and
form different compounds.

When the reactions were conducted with equal amount of Yb and transition
metals (Table 6-1-1), we obtained the desired phases Yb3Ni5-xTMxA119. Here the values
of x were calculated from the ratio of the initial amount of transition metals. Figure 6-1-1
shows the single crystal X-ray data of Yb3Ni5-xTMxAl.9 (TM = Cu, Fe) (x = 0, 0.5, 1)

determined at 173 K. In both cases the cell volume expanded with increasing x as

164

expected since the metallic radii of these transition metals are in the order of rcu > r1:e >
rm.” Therefore the orthorhombic structure of Yb3Ni5A119 was retained and Ni had been
successfully substituted by small amounts of Cu or Fe. Elemental analysis showed very
low level of substitutional element (Cu or Fe): about 1% when x = 0.5 and 3% with x = 1,
The radius of Cu atom is larger than that of Fe, thus the cell parameters of Yb3Ni5_
xCuxAllg are larger than the corresponding ones of Yb3Ni5-xFexAllg. Although the
relative volume expansion was very similar, a linear Vegard law was observed for the
Yb3Ni5-xCuxA119 compounds, Figure 6-1—1A. In the case of Yb3Ni5-xFexAl.9, the a cell
parameter contracted slightly and then increased until x was 1. Since the b and c edge
lengths of Yb3Ni5-xFexA119 expanded normally, the overall cell volume increased by
0.2%. The orthorhombic structure of Yb3Ni5-xFexA119 can be stabilized up to x = 1.
When x was 1, elemental analysis and XRD showed Yb3Ni5-xFexA119 was formed along
with side products YbAl3 and YbNi2-xFexAlg. YbNi2-xFexAlg is a new phase also
crystallizing in needle morphology; and its crystal structure and physical properties will
be described in Part II of this chapter. With more Fe added into the reaction, the
percentage of YbNi2-xFexAlg was higher and ﬁnally was the main phase in the product.
The temperature dependent cell parameters of Yb3Ni5-xTMxA119 (TM = Mn, Cu,
Fe) obtained on single crystals are shown in Figures 6-1-2, 6-1-3, 6-1-4 and 6-1-5. The
data were summarized in Tables 6-1-2, 6-1-3, 6-1-4 and 6-1-5. When the temperature
increased from 100 K to 298 K, the relative change in cell volume for Yb3Ni5A119 was
dV/V = 0.57 %. For the Fe and Mn-substituted samples, the corresponding changes were
very similar to that of the parent compound. While the cell volume of Yb3Ni5-xCuxAl.9

changed much more slowly when the temperature increased from 100 K to 298 K, which

165

were dV/V = 0.32 % (x = 0.5) and dV/V = 0.49 (x = 1). The slow change of the cell
volume for Yb3Ni5-xCuxAl.9 (x = 0.5) came mainly from the a-axis. When the
temperature increased from 173 K to 298 K, the a axis contracted so the overall change
from 100 K to 298 K was only 0.03%. As we discussed before, all the cell parameters of
Yb3Ni5-xTMxA119 are expected to be larger than that of Yb3Ni5A119; however this
happened only at 173 K (Figure 6-1-2). At higher temperature (298 K), except Yb3Ni5-
xCu,,A119 (x = l), the cell volume of all the Yb3Ni5-xTMxA119 compounds was equal to or
smaller than that of Yb3Ni5A119. Although experimental error cannot be excluded as one
possible reason, the electronic valence transition between Yb” and Yb” ions might also

be responsible for this abnormal behavior.

166

A) B)

 

 

    
  
 
  

      

 

 

 

 

 

 

 

 

    

 

 

 

 

 

 

1741 [ 1741 4.062 4.062
1740) 4 1740 4.061 4 4 4.061
p 1739L Yb3N‘5~xC”xAll9 4 1739 4.06 4 Yb,Ni,_,Cu,Al,9 44.06
°$ l p
E 1738 . 4 1738 A 4.059 (4.059
3 s
g 1737 4 1737 a: 4.058 4 4.058
= .’ ﬂ 4- '
9 1736 4' . 4 1736 4.057 4 . 4 4.057 '
U ‘ ‘-,-'a’ Yb3N15_chxA]|9 ‘‘‘‘‘ 5’ YbJNlS-XFCXAllg
1735 4 1735 4.056 4 4 4.056
1734 l 1 4 4 4 4 1734 4.055 . 4 1 1 . 4.055
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
X X
C) I3)
'5'89 6 ‘5'89 26.97 A 26.97
15.89l Yb3N1,,,Cu,Al,9 4 15.89
26.97 4 4 26.97
15.89 4 4 15.89 .
26.96 4 Yb3N15_,Cu,Al,9 4 26.96
15.88 4 4 15.88 D
°< ,4 26.96 ' 4 26.96
v 15.88 4 «0 41588 <
‘3 .- 6"
26.95 4 26.95
15.88 .9 , 15.88
’ Yb3N15_,Fe,Al,9 l .
l5.88 4 15.88 26.95 “1’ . 4 26.95
15.88. 7 15.88 26.94 4 26.94
15.87 if 4 1 . . ' 15.87 I
26.94 1 1 1 4 ' 26.94
0 0.2 0.4 ’06 0.8 1 0 0.2 0 4 0.6 0 8 1
X

Figure 6-1-1. Cell parameter variations with x for Yb3Ni5-xCuxA119 and Yb3Ni5-xFexAl|9
(x = 0, 0.5, 1) at 173 K. A) cell volume; B) a cell edge length; C) b cell edge length; D) c
cell edge length.

167

A)

 

    

 

 

 

 

 

 

 

 

1748 1748
+Yb3Ni5All9
1746 - —-v—1t1>31~115.1‘M11x14.119 (x=0.5) 7 1746

A 1744 4 +Yb3Nf54xcuxAl19(x=O'5) - 1744

0:: + Yb3N1MFexAllg (x = 0.5)

V 1742 P _ 1742

O)

6

£2 1740 — 4 1740

g

.4 1738 ~ ~ 1738

T)

U 1736 4 4 1736
1734 4 4 1734
1732 1 l 1 A 1 1732

80 120 160 200 240 280 320
Temperature(K)
B)
1748 1748
+Yb3NisAlm
1746 ~ +Yb3NihCu‘All9 (X: l) - 1746
+Yb Ni Fe Al (x=l)

p 1744 4 3 5-. x 19 4 1744

°$
1742 4 4 1742

Q)

E

2 1740 4 1 1740

g

:2 1738 — ~ 1738

8
1736 — - 1736
1734 4 4 1734
1732 1 i 1 1 I 1732

80 120 160 200 240 280 320

Temperature(K)

Figure 6-1-2. A) Temperature dependence of the cell volume for Yb3Ni5.xTMxAl.9 (TM =
Mn, Cu, Fe; x = 0, 0.5). B) Temperature dependence of cell volume for Yb3Ni5-xTMxAllg
(TM = Cu, Fe; x = 0, 1).

168

A)

 

     
   
 

 

 

 

 

 

 

 

4.064 4 +YblesAlw 4 4.064
+Yb3N1MCuxA119 (x = 0.5)
4.062 _ +Yb3Nis_xFexAllg (x=0.5) _ 4.062
+ YbSNithxAlw (x = 0.5)
4.06 4 4 4.06
E
m 4.058 4 4 4.058
4.056 4 4 4.056
4.054 4 4 4.054
4.052 4 1 1 1 1 4.052
80 120 160 200 240 280 320
Temperature(K)
B)
4.066 4.066
+Yb3NiquuxAll9 (x = 1)
4.064 4 +YbsN‘s-4FexA’19‘x: 1) 4 4.064
—e— Yb Ni Al
3 5 19
4.062 4 1 4.062
g 406 4 4 4.06
it!
4.058 4 4 4.058
4.056 4 4 4.056
4.054 1 1 1 1 1 4.054
80 120 160 200 240 280 320

Temperature(K)

Figure 6-1-3. A) Temperature dependence of the a cell edge for Yb3Ni5-xTMxAllg (TM =
Mn, Cu, Fe; x = 0, 0.5). B) Temperature dependence of the a cell edge for Yb3Ni5-
xTMxAlm (TM = Cu, Fe; x = 0, 1).

169

A)

 

 

 
   

 

 

 

 

    

 

 

15.91 15.91
+Yb3NisAll9
+Yb3Ni5-anxA119(x = 0.5)
15.9 4 +Yb3Nis-xCuxAll9(x= 0.5) 4 15.9
—*—-Yb Ni FeAl (x=0.5
3 5411 x 19 ,
15.89 4 4 15.89
E
"D 15.88 4 4 15.88
15.87 4 4 15.87
15.86 1 1 1 1 1 15.86
80 120 160 200 240 280 320
Temperature(K)
B)
15.92 15.92
+Yb3Ni5All9
15.91 4 +Yb3Nis_xCuxAlm(x= 1) 4 15.91
+YbNi FeAl (x=1)
3 541: x 19
15.9 4 4 15.9
42‘
V 15.89 4 4 15.89
..D
15.88 4 4 15.88
15.87 4 4 15.87
15.86 1 1 1 1 1 15.86
80 120 160 200 240 280 320

Temperature(K)

Figure 64144. A) Temperature dependent of b cell edge for Yb3Ni5-xTMxA119 (TM = Mn,
Cu, Fe; x = O, 0.5). B) Temperature dependent of b cell edge for Yb3Ni5-xTMxA119 (TM =

Cu, Fe; x = 0, 1).

170

A)

 

 

 

 

 

    

27 27
+Yb3NiSAll9
2699 — +Yb3Ni$_anxAl'9(x= 0.5) 1 7 746-99
26.98 b +Yb3NihCuxAll9 (x= 0.5) ‘ 26.98
+ YbSNiMFexAI'9 (x = 0.5)
26.97 4 4 26.97
°$ 26.96 4 4 26.96
0
26.95 4 4 26.95
26.94 4 4 26.94
26.93 4 4 26.93
26.92 4 1 1 1 1 26.92
80 120 160 200 240 280 320
Temperature(K)
B)
27.02 27.02
—e—Yb Ni Al
3 .5 19
27 4 +Yb3N15-xcuxAll9 (x= 1) 4 27
+Yb Ni Fe Al (x = l)
3 54x 1: 19
26.98 4 4 26.98
°$
O
26.96 4 1 26.96
26.94 4 4 26.94
265%!

 

 

 

1 1 1 1 J 2692
80 120 160 200 240 280 320
Temperature(K)

Figure 641-5. A) Temperature dependence of the c cell edge for Yb3Ni5-xTMxA119 (TM =
Mn, Cu, Fe; x = 0, 0.5). B) Temperature dependence of the c cell edge for Yb3Ni5-
xTM,,A119 (TM = Cu, Fe; x = 0, 1).

171

Table 641-2. Compositional variations of the cell volume for Yb3Ni5-xTMxA119 (TM = Cu,

Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of 100 K, 173 K and

298 K.

 

 

 

 

 

 

 

 

Yb3Ni5-xTMxA119 3 V 3 V 3 V %
(A , 100 K) (A , 173 K) (A , 298 K) (100 K-298K)

x = 0 1734.0(3) 1734.9(2) 1743.9(7) 0.57
x = 0.5 (TM 4 Cu) 1734.9(4) 1737.6(4) 1740.4(10) 0.32
x = 1 (TM = Cu) 1738.1(4) 1740.2(4) 1746.7(4) 0.49
x = 0.5 (TM = Fe) 1733.4(2) 1735.8(2) 1744.2(4) 0.62
x = 1 (TM = Fe) 1733.6(2) 1738.3(4) 1743.6(8) 0.58
x = 0.5 (TM = Mn) 1732.9(2) 1737.2(4) 1742.6(4) 0.56

 

 

 

 

 

 

Table 641-3. Compositional variations of the a cell edge lengths for Yb3Ni5-xTMxAl|9

(TM = Cu, Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of 100 K,

173 K and 298 K.

 

 

 

 

 

 

 

 

Yb3Ni5"TM"Al'9 (A, 100 K) (A, 173 K) (A, 298 K) (100 1229819
x = 0 4.0578(4) 4.0573(2) 4.0625(9) 0.12
x = 0.5 (TM = Cu) 4.0541(6) 4.0587(6) 4.0553(13) 0.03
x = 1 (TM = Cu) 4.0596(5) 4.0610(5) 4.0657(5) 0.15
x = 0.5 (TM 4 Fe) 4.0550(3) 4.0566(3) 4.0631(6) 0.20
x = 1 (TM 4 Fe) 4.0551(3) 4.0601(5) 4.0634(11) 0.20
x = 0.5 (TM = Mn) 4.0555(3) 4.0597(5) 4.0618(5) 0.16

 

 

 

 

 

 

172

Table 641-4. Compositional variations of the b cell edge lengths for Yb3Ni5-xTMxA119

(TM = Cu, Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of 100 K,

173 K and 298 K.

 

 

 

 

 

 

 

 

 

 

 

 

W3Ni5'XTM‘Al'9 (A, 1100 K) (A, 123 K) (A, 238 K) (10013298K)
x = 0 15.868606) 15.8745(9) 15.906(4) 0.24
x = 0.5 (TM = Cu) 15.885(2) 15.886(2) 15.902(5) 0.11
x = 1 (TM = Cu) 15.882(2) 15.889(2) 15.910(2) 0.18
x = 0.5 (TM 4 Fe) 15.871600) 15.8802(10) 15.902(2) 0.19
x = 1 (TM = Fe) 15.8707(11) 15.8823(18) 15.891(4) 0.13
x = 0.5 (TM = Mn) 15.869103) 15.882(2) 15.898308) 0.18

 

Table 641-5. Compositional variations of the 6 cell edge lengths for ngNi5.xTMxAl;9

(TM = Cu, Fe, Mn; x = 0, 0.5, 1) obtained on single crystals at temperatures of 100 K,

173 K and 298 K.

 

Yb3Nis-xTMxA119

C

C

C

%

 

 

 

 

 

 

 

(A, 100 K) (A, 173 K) (A, 298 K) 000 K4298K)
x = 0 26.929(3) 26.93650 5) 26.989(6) 0.22
x = 0.5 (TM = Cu) 26.939(4) 26.948(4) 26.98800) 0.18
x = 1 (TM 4 Cu) 26.957(3) 26.970(3) 27.003(4) 0.17
x = 0.5 (TM = Fe) 26.933307) 26.945807) 26.995(4) 0.23
x = 1 (TM = Fe) 26.937408) 26.957(3) 27.002(7) 0.24
x = 0.5 (TM = Mn) 7.6926(2) 26.942(2) 269850) 0.22

 

 

 

 

 

173

 

 

Magnetic Properties:

The temperature dependent magnetic susceptibility data for Yb3N15-xTMxA119
(TM = Cu, Fe; x = 0, 0.5, 1) are displayed in Figures 641-6 & 6-1-7. As the pristine
Yb3Ni5A119 compound, all the derivatives show paramagnetic behavior and follow Curie-
Weiss Law at temperatures above 50 K. As the content of TM 4—4 x increases from 0 to 1,
the magnitude of the susceptibility increases. The Curie temperature 0 and calculated
effective magnetic moment pen for these samples are summarized in Table 641-6. Herein
we recall from the ﬁrst part of this chapter, the parent compound Yb3Ni5A119 exhibits
intermediate valence behavior and strong antiferromagnetic interactions. As shown in
Table 64146, the Curie temperatures 0 for Yb3Ni5-xCuxAllg and Yb3Nis-xFexAl|9 are lower
than that of Yb3Ni5A119. This effect indicates a weakening of the Kondo interaction,
since in the scope of single-ion Kondo theories 0 is related to TK by TK = m l0 I , where
the coefﬁcient m is usually of the order of 1.24 Although the Kondo effect is weakening,
we have not observed any magnetic ordering in either of the cases until x = 1.

The last column in Table 64146 is the effective magnetic moment peg determined
by ﬁtting the high temperature data to the Curie-Weiss law. For the Fe-based samples,
the new value does not change much. This is possibly because that the Yb valence can
change as a function of composition only in materials that undergo a relatively large
change in unit cell volume.25 In Yb3Ni5-xFexA119 the cell volume expands only 0.19%
with x changing from O to 1. In the case of Yb3Ni5-xCuxA1.9, there is no apparent trend
with x with regard to the pen values. The substitution of Ni by Cu results in the cell
volume expansion as conﬁrmed by single crystal X-ray diffraction. Therefore this

system is driven by negative chemical pressure toward the Yb2+ form while still being in

174

YbZJr/Yb3+ intermediate state. This result is consistent with the fact that the ionic radius
of the nonmagnetic Yb ion (4f'4-state) is larger than that of the magnetic one (4143 4state).
With more Cu added into the structure, the effective magnetic moment pen for Yb3Ni5-
xCuxAl 19 (x = 1) is 7.90 113, which appears to be very close to that of the free Yb3+ ion
(theoretical value 11ch = 7.86 113 per formula). Hence the magnetic 4fl3 conﬁguration is
stabilized by the substitution of Ni with Cu. Here the chemical pressure cannot be the
reason for this magnetic properties variation, since the substitution causes a volume
expansion. The substitution of Ni by Cu should lead to an increment of the number of
conduction electrons, and hence to a rise of the Fermi level. Therefore the hybridization
between the 4f electrons and the conduction band should decrease causing a quenching of
the Kondo effect (indicated by the value of 0). This phenomenon has been observed for
the other systems as well, such as YbCu5. 26 The substitution of Cu by Al or Ga shifts the
Yb ions from non-magnetic divalent 4f"4 state to localized trivalent 4fl3 state. Since the
substitution of Cu by Al/Ga also causes a volume expansion, chemical pressure can be
excluded as a driving force of the valence state changes. The number of conduction
electrons provided by A1 or Ga is more than that of Cu leading to the rise of the Fermi
level. For YbCu_:,.,,Alx (1.5 S x S 2), the 4f band is far below the Fermi level and the s~f J
mixing is negligible. This example and our experimental observations indicate that in
order to explain the changes in the hybridization between 4f electrons and conduction

band, both volume and Fermi level modiﬁcation have to be taken into account.

175

Table 641-6. Magnetic property parameters of Yb3Nis-xCuxA119 and Yb3Ni5-xFexAllg (x =
0, 0.5, 1).

 

 

 

 

 

 

Compound x c 0 (K) 113M113)

Yb3Ni5A119 0 5.73 4551.2 6.77
Yb3Ni5-xCuxA119 0.5 4.59 4276.9 6.06
Yb3Ni5-xCuxA119 1 7.81 4399.0 7.90
Yb3N15-xFexAllg 0.5 5.38 4222.5 6.56
Yb3Ni5-xFexA119 1 6.20 4271.4 7.04

 

 

 

 

 

 

 

Note: pent (theoretical value Yb3Ni5A119) = 7.86 113 (Yb3+)

176

0.1

 

 

 

 

”c?
'5 0'08 I Yb Ni Al
E 3 5 19
o - _
1; 0.06 V Yb3N15_xCuxAl19 (x -- 0.5)
E 0 Yb3N154xcuxA119 (x = 1)
g 0.04
3
E
X 0.02 00
000000
:.:::::: i i i 2 i 2 2
0 llllll4Jlllllllllll‘lllLlUl

0 50 100 150 200 250 300
Temperature (K)

Figure 641-6. Magnetic susceptibility of Yb3Ni5-xCuxAllg (0 S x S. 1) compounds as a
function of temperature from 3 K to 300 K.

177

...q4—“#_. ‘_ ‘- \ -1.

 

0.07 -

 

 

 

 

 

E 0'06 Yb Ni Al
5 3 5 l9
g 0.05 Yb3Ni5 Fe A119 (x = 0.5)
“-4 -X X
g 0.04 Yb3NiMFexAl19 (x = 1)
a 0.03 "
3
E 0.02

>4 ,....

0.01 Ill-III. 3 ‘ . . . . .

0 111111111Innnnlnnnnlnnnnlnnnn
0 50 100 150 200 250 300

Temperature (K)

Figure 641-7. Magnetic susceptibility of Yb3Ni5-xFexAl(9 (0 S x _<_ 1) compounds as a
function of temperature from 3 K to 300 K.

 

178

641-4. Conclusions

The transition metal Ni in Yb3Ni5Allg can be substituted to a limited extent by
other transition metals such as Mn, Fe or Cu, which are slightly larger and can cause a
variation in the number of conduction electrons. Lower cell volume values have been
observed for Yb3Ni5-xTMXA119 (TM = Mn, Fe) than that of Yb3Ni5A119 at 100 K and 298
K; and the electronic valence transition of Yb ions might be one reason causing this
abnormal behavior. Substitution of Ni by Mn or Fe does not change signiﬁcantly the
thermal expansion property of Yb3Ni5-xTMxA119 from 100 K to 298 K; however the
thermal expansion coefﬁcient of Yb3Nis-xCuxAllg is much lower than that of the parent
compound Yb3N15A119. Magnetic susceptibility measurements indicate that effective
magnetic moment can be changed by a modest amount due to the substitution of Ni by
Cu, and this may be explained by the effect of cell volume change and fermi level

modiﬁcation.

179

 

References:

 

1 a) Chandran, L.; Krishna-Murthy, H. R.; Ramakrishnan, T. v. J. Phys. Condens. Mat.
1992, 4, 7067. b) Zlatic, V.; Costi, T. A.; Hewson, A. C.; Coles, B. R. Phys. Rev. B 1993,
48, 16152. c) Jaccard, D.; Link, P.; Vargoz, E.; Alami-Yadri, K. Physica B 1997, 230-
232, 297. d) Mushnikov, N. V.; Goto, T.; Kolomiets, A. V.; Yoshimura, K.; Zhang, W.;
Kageyama, H. J. Phys. Condens. Mat. 2004, I6, 2395.

2 Singh, Y.; Ramakrishnan, 8. Phys. Rev. B 2003, 68, 054419.

3 Kaczorowski, D.; Andraka, B.; Pietri, R.; Cichorek, T.; Zaremba, V. 1. Phys. Rev. B
2000, 61, 15255.

4 Trovarelli, 0.; Geibel, C.; Buschinger, B.; Borth, R.; Mederle, S.; Grosche, M.; Spam,
G.; Steglich, F.; Brosch, O. Donnevert, L. Phys. Rev. B 1999, 60, 1136.

5a) Doniach, S. Phy. B 1977, 71, 231. b) Thompson, J. D. and Lawrence, J. L. In
Handbook on the Physics and Chemistry of Rare Earths; Gschneidner, Jr., K. A., Eyring,
L., Lander, G. H., Choppin, G. R., Eds.; North-Holland: Amsterdam, 1994; Vol. 19, p
383.

6 a) Link, P.; Jaccard, D. Physica B 1996, 223342248, 303. b) Grosche, F. M.; Julian, s.
R.; Mathur, N. 1).; Carter, F. V.; Lonzarich, G. G. Physica B 1997, 23 74238, 197.

7 a) Marcano, N.; Paccard, D.; Espeso, J. I.; Allemand, J .; Moreau, J. M.; Kurbakov, A.;
Sekine, C.; Paulsen, C.; Lhotel, E.; Gomez-Sal, J. C. J. Magn. Magn. Mater. 2004, 2 72-
276, 468. b) Garcia, S. J.; Gomez-Sal, J. C.; Rodriguez, F. J .; Espeso, J. I.; Monconduit,
L.; Allemand, J .; Paccard, D. Physica B 1997, 2304232, 117.

8 Blanco, J. A.; Podesta, M. de; Espeso, J. 1.; Gomez-Sal, J. C.; Lester, C.; McEwen, K.
A.; Patrikios, N.; Rodriguez, Fernandez, J. Phys. Rev. B 1994, 49, 15126.

9 Winkelmann, H.; Abd-Elmeguid, M. M.; Micklitz, H.; Sanchez, J. P.; Geibel, C.;
Steglich, F. Phys. Rev. Lett. 1998, 81, 4947.

'0 Sales, B. (2.; Viswanathan, R. J. Low Temp. Phys. 1976, 23, 449.

H Winkelmann, H.; Abd-Elmeguid, M. M.; Micklitz, H.; Sanchez, J. P.; Vulliet, P.;
Alami-Yadri, K. J accard, D. Phys. Rev. B 1999, 60, 3324.

'2 Andreica, D.; Amato, A.; Gygax, F.; Pinkpank, M.; Schenck, A. Physica B 2000, 289-
290, 24.

180

 

'3 a) Bauer, E.; Hauser, R.; Keller, L.; Fischer, P.; Trovarelli, O.; Sereni, J. G.; Rieger, J.
J .; Stewart, G. R. Phys. Rev. B 1997, 56, 711. b) Bauer, E. J. Magn. Magn. Mater. 1999,
1964197, 873.

" Bauer, E. D.; Bobev, S.; Thompson, J. D.; Hundley, M. F .; Sarrao, J. L.; Lobos, A.;
Aligia, A. A. J. Phys. : Condens. Matter 2004, 16, 4025.

‘5 Salvador, J. R.; Guo, F.; Hogan, T.; Kanatzidis, M. G. Nature 2003,4125, 702.

'6 Drymiotis, F. R.; Lee, Y.; Lawes, G.; Lashley, J. C.; Kimura, T.; Shapiro, S. M.;
Migliori, A.; Correa, V.; Fisher, R. A. Phys. Rev. B 2005, 71, 174304.

'7 a) Gignoux, D.; Givord, F .; Lemaire, R. J. Less-Common Met. 1983, 94, 165. b)
Marcano, N.; Paccard, D.; Espeso, J. I.; Allemand, J .; Moreau, J. M.; Kurbakov, A.;
Sekine, C.; Paulsen, C.; Lhotel, E.; Gémez Sal, J. C. J. Magn. Magn. Mater. 2004, 272-
276, 468.

'8 SMART, version 5; Siemens Analytical X-ray Systems, Inc.: Madison, WI, 1998.
'9 Saint, Version 4; Simens Analytical X-ray Instruments Inc., Madison, WI.
20 SADABS, Sheldrick, G. M.; University of Géittingen, G6ttingen, Germany.

2‘ G.M. Sheldrick, 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI. '

22 Florio, J. V.; Rundle, R. E.; Snow, A. I. Acta Crystallogr. 1952, 5, 449.
23 The metallic radii of Mn, Fe, Ni and Cu are 1.37 A, 1.26 A, 1.25 A and 1.28 A,
respectively. In Wells, A. F. Structural inorganic chemistry, 5‘h edn, Clarendon Press,

Oxford (1984).

2‘ Hewson, A. The Kondo Problem to Heavy Fermions, Cambridge Studies in Magnetism,
Vol. 2 (Cambridge University Press, 1993).

25 Bomick, R. M.; Stacy, A. M. Chem. Mater. 1994, 6, 333.
3" a) Bauer, E.; Tuan, L.; Hauser, R.; Gratz, E.; Holubar, T.; Hilscher, G.; Yoshimura, K.

J. Magn. Magn. Mater. 1995, 1404144, 1247. b) Bauer, E. J. Magn. Magn. Mater. 1999,
1964197, 873. c) He, J.; Ling, G.; Ye, Z. J. Alloys Comd. 2001, 325, 54.

181

CHAPTER SIX

PART II. Discovery of the New Intermetallic Phases RENi2-,FexAlg (RE
= Eu, Yb) from Liquid Aluminum

642-1. Introduction
Yb-containing intermetallics continue to attract considerable attention in the study

3In

of interesting behaviors such as Kondo lattice,l heavy fermion2 and mixed valence.
such systems, an array of 4f moments is embedded in a metallic environment, and the
nature of the ground state (GS) is determined by the competition between Kondo effect
and RKKY (Ruderman-Kittel-Kasuya-Yosida) interactions.4 The Kondo effect results
from the exchange interaction between the conduction electrons and the local magnetic
moments leading to scattering events in which the electron spin is ﬂipped. The Kondo
effect favors a nonmagnetic GS with a characteristic energy given by the Kondo
temperature TK. The RKKY interaction favors a magnetically ordered GS characterized
by TRKKy. In general, the Kondo effect is expected for TRKKY > TK, heavy fermion
behavior for TRKKY E TK, and mixed valence behavior for TK > Tmy.

In Part I of Chapter Five, we described crystal growth, structure and physical
properties of ternary intermetallics RE3Ni5A119 (RE = Sm, Dy, Er, Yb) accessible in
liquid Al. In this family Yb3NisAllg is a Kondo compound with Yb ions showing
intermediate valence behavior. 5 We also found that magnetic properties of this
compound could be tuned by minor substitutions of Ni by Cu/F e. The structural integrity

of Yb3Ni5-xFexA119 was retained until x was 1. With more Fe added into the reaction,

another new phase appeared in the product and dominated with increasing x — YbNiz-

182

 

 

._‘.._ ___u».‘A_._. - ..— ‘im ~—

xFexAlg. This reaction can be optimized by combining stoichiometric amount of Ni and
Fe; later we discovered the Eu analogue under the same experimental conditions.
RENi2-xFexAlg (RE = Eu, Yb) belong to the ternary family RETM2X3 (TM = C0,
Fe; X = A1, Ga) crystallizing in the CaCozAlg structure type.6 Most compounds reported
belonging to this family are ternary with the transition metal mainly Fe, Co or Ni.
Among them CeFezAlg is a rare example of the valence ﬂuctuation compound containing '

element F e.7 Méssbauer spectra Show that the Fe atoms in CCFCzAlg do not carry local

 

magnetic moment. 8 PFCOzAlg orders antiferromagnetically at 5 K, with a clear
metamagnetic transition occurring at a critical ﬁeld of 9000 G.9 Quite a few gallides
have also been prepared: the series REFezGag (RE = Ce, Pr, Nd, Sm),10 RECOzGag (RE =
Ce, Pr, Eu, Yb)lo‘” and RERuzGag (RE = La, Ce, Pr, Nd).12 The indium-containing
isostructural compound Euhalng was prepared and reported by Péttgen and co-
workers.'3 The compounds RENi2-xFexAlg (RE = Eu, Yb) we discovered by using Al
ﬂux method are the ﬁrst quaternary examples reported in the literature in this family. In
this part of the dissertation, the synthesis, crystal structure, physical properties and 57Fe
Mbssbauer spectroscopy measurements of RENi2-xFexAlg (RE = Eu, Yb) will be

described.

642-2. Experimental Section
Reagents:

The following reagents were used as obtained: rare earth metals (RE = Eu, Yb)
(Cerac, 99.9%), Ni (99%, 325 mesh, Sargent, Buffalo Grove, IL), Fe (99.99%, ﬁne

powder, Aldrich Chemical, Milwaukee, WI), Al (Cerac, 99.5%, 420 mesh).

183

Synthesis:

Rare earth metal, Ni and Fe were combined with a large excess of Al in a
nitrogen-ﬁlled glove box. Alumina crucibles containing the reaction mixture of 1 mmol
RE metal (Eu 0.152 g, Yb 0.173 g), 1 mmol Ni (0.059 g), 1 mmol Fe (0.056 g) and 10
mmol Al (0.270 g) were placed into silica tubes (13 mm in diameter), which were then
sealed under vacuum (~104 Torr). The samples were heated to 1000 °C in 15 h,
maintained at this temperature for 5 h, and then cooled to 850 °C in 2 h. They were

annealed at 850 °C for 3 (1, followed by cooling down to 500 °C at a rate of 10 °C/h.

Finally the temperature was brought down to 50 °C in 10 h.

The excess aluminum was removed by soaking the crucible in aqueous 5M NaOH
solution overnight. The crystalline product remaining aﬂer the isolation procedure was
rinsed with water and acetone. RENi2-xFexAlg was the major phase in the product with a

yield of 70 %. For RE = Yb, the ternary phase Yb3Ni5A1|9 was also observed.

Scanning Electron Microscopy and Elemental Analysis:

The crystals were picked and placed on a Scanning Electron Microscope (SEM)
sample plate using carbon tape. Chemical compositions of the products were determined
by Energy Dispersive Spectroscopy (EDS) performed on a JEOL JSM-35C SEM
equipped with a NORAN EDS detector. The SEM image of a typical crystal of YbNiz-
xFexAlg is Shown in Figure 6-2-1. Data were acquired by applying a 25 kV accelerating
voltage with an accumulation time of 40 8. Several crystals were analyzed with the
resulting elemental composition corresponding to the ratio 1:1:128, which agreed well

with the results derived from the single crystal X-ray analysis.

184

X—ray Crystallography:

Single crystal X-ray diffraction data were collected for RENi2-xFexAlg (RE = Eu,
Yb) at room temperature on a Bruker AXS SMART CCD X-ray diffractometer. A data
collection (Mo K01 radiation, A = 0.71073 A) was acquired covering a full sphere of
reciprocal space using exposure time 20 s/frame. The data acquisition and cell reduction
were done with the SMART14 software package and data processing was performed with
the SAINTPLUS program.‘5 An empirical absorption correction was applied to the data
using the SADABS program.16 The structures were solved using direct methods and
reﬁned with the SHELXTL package program.l7 Systematic absence conditions led to
two possible space groups: Pbam and Pba2. The mean value of IE2-1 Iwas 0.969
indicating that the structure was likely centrosymmetric. Pbam is the centrosymmetric
one and has a much lower CFOM value. So Pbam was chosen and proved to be correct
after ﬁnal reﬁnement of the structure.

In the structure of YbNiz-xFexAlg, a total of twelve atomic sites including one Yb,
two transition metal Sites (Ni/Fe) and nine Al sites were identiﬁed. Originally the X-ray
data was collected at 173 K covering a hemisphere of reciprocal space. Since the atomic
numbers of Ni and Fe are very close to each other, different assignments on these two
transition metal sites M(l) and M(2) were examined. Two satisfactory solutions were
obtained: 1) M(l) Site was occupied by Ni and M(2) was occupied by Fe with both sites
fully occupied; 2) Both M sites were occupied by a mixture of Ni/Fe: for M( 1), Ni and Fe
was found to have an occupancy of 84.6 and 15.4%, respectively; while for M(2), the
corresponding values were 33.5 and 66.5%. To further conﬁrm the assignment, we

collected single crystal X-ray data covering a full sphere of reciprocal space. We found

185

 

 

that if the M(l) site was assigned to Ni and the M(2) assigned to Fe, the occupancy
factors were 0.94 and 1.06 respectively, R. and wR2 were 2.39% and 5.72%. If we
exchanged these two sites, the occupancy factors were 1.04 and 0.96; R1 and wR2 values
were 2.40% and 5.77%. As a result of that, M(l) and M(2) sites were reﬁned as a mixed
occupancy of Ni and Fe: 42%/58% on M( 1) and 62%/38% on M(2) site for Ni/Fe. This
reﬁnement gave the lowest R values with R; 1.86 % and wR2 3.95 %. And the resulting
stoichiometry is in fair agreement with the elemental analysis from EDS.

Data collection parameters and reﬁnement details for BUN12-XFCxAl3 and YbNiz-
xFexAlg can be found in Table 642-1. Atomic positions, displacement parameters and
anisotropic displacement parameters for EuNi2-xFexAlg and YbNi2-xFexAlg are listed in
Tables 642-2 and 642-3.

X-ray powder diffraction data were collected at room temperature on a CPS 120
INEL X-ray diffractometer (Cu K01) equipped with position-sensitive detector.
Experimental powder patterns were compared to the patterns calculated from the single
crystal structure solution (by the CrystalDiffract program”) to determine the phase

identity and purity.

186

Table 642-1. Selected crystal data and structure reﬁnement details for EuNiz-,,Fe,,Alg and

 

 

YbNiz-xFCxAlg.

Empirical formula EuNi2-,,FexAlg YbNi2-xFexAlg
Formula weight 482.36 503.44
Crystal system Orthorhombic Orthorhombic
Space group Pbam (#55) Pbam (#55)

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient
F(000)

Crystal size

Theta range for data collection
Limiting indices

Reﬂections collected
Independent reﬂections
Completeness to theta = 37.00°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F2

Final R indices [I>28igma(l)]
R indices (all data)

Largest diff. peak and hole

a = 125300) A

b = 14.503(3) A

c 4 4.0326(8) A
732.8(2) A3

4

4.372 Mg/m3

13.806 mm"

884

0.17 x 0.16 x 0.25 mm3

2.15 to 27.70°
46Sh$16
485k518
45 515 5

7162
941 [R(int) = 0.0282]
95.4 %

a = 12.4311(9) A

b = 14.430900) A

c = 3.9749(3) A
713.06(9) A3

4

4.663 Mg/m3

18.404 mm"

912

0.22 x 0.30 x 0.24 mm3

2.82 to 27.79°
416ShS 16
4195ks 18
455155

6852
929 [R(int) = 0.0275]
95.6 %

Full-matrix least-squares on F2

941 / 0 / 70
1.162

R, = 0.0186

wR2 = 0.0395

R, = 0.0200

wR2 = 0.0399

0.593 and 40.963 e.A'3

929 / 0 / 72
1.356

R, = 0.0178

wR2 = 0.0397

R, = 0.0186

wR2 = 0.0400

1.064 and —1.407 e.A'3

221/2

R1 = 2(|F0|4|FC|)/Z|Fol; wR2 = [Z[W(F02' Fczl/ [2(wlFol ) 1

187

Table 642-2. Atomic coordinates (A x 104) and equivalent isotropic displacement

parameters (A2 x 103) for EuNi2-,,FexAlg and YbNi2-,,FexAlg.

 

Wyk.

 

Atom Symbol x y z Occu. U(eq).
Eu 4h 6592(1) 1811(1) 5000 1.0 8(1)
Ni(1) 4h 5342(1) 5913(1) 5000 0.43 7(1)
Fe(l) 4h 5342(1) 5913(1) 5000 0.57 7(1)
Ni(2) 4h 8505(1) 4015(1) 5000 0.35 6(1)
Fe(2) 4h 8505(1) 4015(1) 5000 0.65 6(1)
Al( 1) 2a 5000 5000 0 1.0 9(1)
Al(2) 4g 8399(1) 1197(1) 0 1.0 7(1)
Al(3) 4g 7642(1) 3275(1) 0 1.0 8(1) .
Al(4) 4h 4051(1) 2545(1) 5000 1.0 8(1)
Al(5) 4g 4737(1) 1330(1) 0 1.0 7(1)
Al(6) 4g 6682(1) 72(1) 0 1.0 8(1)
Al(7) 4g 5481(1) 3203(1) 0 1.0 8(1)
Al(8) 2d 0 5000 5000 1.0 8(1)
Al(9) 4h 6639(1) 4536(1) 5000 1.0 13(1)
Yb 4h 6579(1) 1809(1) 5000 1.0 9(1)
Ni(1) 4h 5341(1) 5924(1) 5000 0.35 6(1)
Fe(l) 4h 5341(1) 5924(1) 5000 0.65 6(1)
Ni(2) 4h 8488(1) 4023(1) 5000 0.49 5(1)
Fe(2) 4h 8488(1) 4023(1) 5000 0.51 5(1)
Al( 1) 2a 5000 5000 0 1.0 8( 1)
Al(2) 4g 8374(1) 1211(1) 0 1.0 8(1)
Al(3) 4g 7633(1) 3253(1) 0 1.0 8(1)
Al(4) 4h 4027(1) 2546(1) 5000 1.0 7(1)
Al(5) 4g 4752(1) 1331(1) 0 1.0 6(1)
Al(6) 4g 6678(1) 88(1) 0 1.0 8(1)
Al(7) 4g 5481(1) 3188(1) 0 1.0 7(1)
Al(8) 2d 0 5000 5000 1.0 6(1)
Al(9) 4h 6627(1) 4545(1) 5000 1.0 12(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

188

Table 642-3. Anisotropic displacement parameters (A2 x 103) for EuNi2-,,FexAlg and

 

 

YbNiz_xFexAlg.

Atom U11 U22 U33 U23 U13 U12

Eu 8(1) 11(1) 7(1) 0 0 41(1)
M(l) 8(1) 6(1) 7(1) 0 0 40(1)
M(2) 6(1) 6(1) 6(1) 0 0 0(1)
Al(l) 12(1) 6(1) 8(1) 0 0 42(1)
Al(2) 5(1) 8(1) 8(1) 0 0 0(1)
Al(3) 7(1) 10(1) 8(1) 0 0 42(1)
Al(4) 9(1) 6(1) 10(1) 0 0 0(1)
Al(5) 7(1) 8(1) 7(1) 0 0 41(1)
Al(6) 9(1) 7(1) 9(1) 0 0 42(1)
Al(7) 7(1) 6(1) 10(1) 0 0 0(1)
Al(8) 6(1) 6(1) 10(1) 0 0 42(1)
Al(9) 9(1) 17(1) 13(1) 0 0 4(1)
Yb 8(1) 12(1) 6(1) 0 0 41(1)
M(l) 7(1) 5(1) 5(1) 0 0 0(1)
M(2) 6(1) 5(1) 4(1) 0 0 0(1)
Al(l) 11(1) 6(1) 6(1) 0 0 41(1)
Al(2) 6(1) 9(1) 8(1) 0 0 41(1)
Al(3) 7(1) 10(1) 7(1) 0 0 42(1)
Al(4) 10(1) 6(1) 7(1) 0 0 0(1)
Al(5) 6(1) 7(1) 6(1) 0 0 41(1)
Al(6) 10(1) 7(1) 6(1) 0 0 41(1)
Al(7) 7(1) 7(1) 8(1) 0 0 41(1)
Al(8) 5(1) 8(1) 6(1) 0 0 42(1)
Al(9) 8(1) 16(1) 11(1) 0 0 4(1)

 

-27t2[h2a*2Ull+...+2hka*b*U12]

189

The anisotropic displacement factor exponent takes the form:

 

 

 

Figure 6-2-1. SEM image of a typical crystal of YbNi2-xFe,,Alg.

Physical Properties Characterization:

Magnetic measurements were conducted on single crystal and polycrystalline
samples of YbNi2-,,Fe,,Alg. Field-cooled and zero—ﬁeld cooled dc magnetization
measurements were performed for the above samples using a Quantum Design MPMS
SQUID magnetometer. EDS-analyzed crystals were ground into powder, which was
sealed in kapton tape and placed into the magnetometer. The data were collected in the
temperature range 34300 K at 1000 G, while ﬁeld dependent magnetic measurements,
conducted at 5 K, were carried out in ﬁelds up to i 55000 G. A diamagnetic correction
was applied to the data to account for core diamagnetism.

Electrical resistivity of YbNiz-,,FexAlg was measured over the temperature range 5
K ~ 300 K using a four-probe dc technique with contacts made using silver paste on a
pressed-pellet sample. Single crystals of YbNi2-,,Fe,,Alg were selected, ground into
powder and pressed into a pellet of 3 mm length, 2 mm width and 0.2 mm thickness. The

pellet sample was annealed at 350 °C for 3h. Thennopower data were collected on the

190

pellet from 300 K to 400 K with a MMR Technologies, Inc. Seebeck measurement

system.

Mo'ssbauer Spectroscopy:

The Méssbauer spectra were taken on a polycrystalline sample of YbNi2-,,Fe,,Alg
with a constant—acceleration spectrometer, equipped with a 57C0 source in a Rh matrix.
The spectrometer was calibrated with metallic iron, and the isomer shift values were
reported relative to a-Fe. A closed-loop refrigerator system was used for the low
temperature measurements. An Oxford Instrument Variox 316 cryostat was used for

measurements at liquid—helium temperature.

642-3. Results and Discussion
Synthesis:

YbNi2-xFexAlg was discovered during an attempt to substitute Ni in Yb3Ni5Al,9
with Fe with the molar ratio 0.8 : 0.2 for Ni : Fe. PXRD showed that besides Yb3Ni5A119,
another phase appeared in the product and elemental analysis indicated almost equal
amount of Ni and Fe. Single crystal X-ray diffraction analysis conﬁrmed the existence of
YbNi2-,,Fe,,A13 which crystallizes in a different structure from Yb3Ni5A119. To optimize
this reaction, equal amount of Ni and Fe metals were combined to form YbNi2-,,Fe,,Alg
with a yield as high as 70%; however a large percentage of Yb3Ni5Al,9 remained as the
Side product. Interestingly, when we tried to grow larger single crystals of YbNi2-,,FexAlg
by scaling up the reaction, i.e. the molar ratio of Yb : Ni : Fe : A1 as 3 : 3 : 3 : 30, the

yield was greatly improved (90 %). To make the other rare earth analogs of this phase,

191

we tried the reactions with equal amount of Ni and Fe metals. So far only EuNiz-,,Fe,,Alg
was obtained which will also be described in this section. For the other rare earth metals

only RE3Ni5Al,9 was found in the product.

Crystal Structure:

RENi2-,,Fe,,Alg (RE = Eu, Yb) crystallize in a known structure type of CaC02A13
with the space group Pbam.6 This is a rather stable structural arrangement as a number of
isotypical ternary phases have been reported which we already described in the
introduction section. The RENi2-,,Fe,,A13 family seems to be the ﬁrst quaternary analogue
of the CaCozAlg structure type. Herein we will brieﬂy describe the structure and local
coordination geometries with YbNi2-xFexAlg as an example. In this structure, there are
one crystallographically independent Yb site, two M sites (mixed occupied by Ni and Fe)
and nine Al sites. Figure 642-2 depicts the structure of YbNi2-,,Fe,,Alg in polyhedral view
down the c-axis. The M(1)- and M(2)-centered polyhedra, composed of Al atoms, form a
three-dimensional framework with Yb ions sitting in the channels. Each channel is
composed of two M(1)-polyhedra and three M(2)-polyhedra which connect with each
other by sharing Al corners. As is shown in Figure 64243, the coordination environments
of M(l) and M(2) atoms are similar and best described as tri-capped trigonal prisms
composed of Al atoms. On the ab plane, the M(l) polyhedra are condensed into dimers
by sharing faces containing two Al(l) and two Al(9) atoms. The bond distance between
M( 1) atoms is 2.8051(12) A. Along the c-direction, these M(l)-centered polyhedra share
Al trianglular faces containing Al(l), Al(2) and Al(7) atoms. The distances between

these Al atoms range from 2.664(2) A to 2.997(2) A. The coordination geometry of the

192

 

 

M(2) atom resembles that of M(l); each M(2) atom is surrounded by nine Al atoms
which form a tri-capped trigonal prism. Along the c-axis, these trigonal prisms share the
Al-based trigonal faces but without forming M(2)-dimers on the ab-plane.

The Yb atoms reside in distorted penta-capped pentagonal prisms composed of Al
atoms. This geometry is similar as those of the Yb atoms in Yb3Ni5A1,9 which are also in
distorted fashion.5 The distance from Yb to Al(9) atoms is 3.917(2) A, much longer than
the distances from Yb to the other Al neighbors which are between 3.1009(11) A to

3.2717(3) A.

Physical Properties:

Because of the explicit needle morphology and the large crystal size of YbNiz-
xFe,,Alg, we were able to study its magnetic properties both isotropically (polycrystalline)
and anisotropically (single crystal). The temperature dependent molar magnetic
susceptibility data conducted on a polycrystalline sample is shown in Figure 6-2-4A.
Above 50 K, the inverse magnetic susceptibility data can be ﬁtted to the Curie-Weiss
Law, with calculated Heir = 2.19 113, which is between the theoretical number for Yb2+ (0
113) and Yb3+ (4.54 113). This suggests that the Yb ions are of mixed valency nature in
this compound. It is noteworthy that no broad hump was observed in the susceptibility
data which is typical of mixed-valent systems.'9 In some other mixed-valent Yb-
containing compounds, this broad hump is not observed either, such as YbGaGe20 and

“2113665.”

193

‘11—‘—

 

 

 

Figure 6-2-2. Crystal structure of YbNi2-,,Fe,,A13 in polyhedra of M(l) and M(2) atoms
viewed down the [001] direction. Large circles: Yb; black dots: Al; darker shaded
circles: M(l); lighter shaded circles: M(2).

194

 

Al(6) Al(7)

Figure 642-3. Coordination environments of M(l), M(2) and Yb atoms.

195

Table 642-4. Bond lengths (A) for EuNi2-,,Fe,,Alg and YbNi2-,,FexAlg.

 

 

Bond Distance Bond Distance
Eu-Al(2) x2 3.1599(11) M(2)-Al(8) 2.3559(7)
Yb-Al(2) x2 3.1164(10) 2.3544(6)
RE-Al(3) x2 3.2102(11) M(2)-Al(9) 2.4560(16)
3.1695(11) 2.4380(15)
RE-A1(4) x2 3.2187(14) Al(1)-A1(2) x2 2.6523(14)
3.1892(14) 2.6768(14)
RE-Al(5) x2 3.1555(11) Al(1)-Al(7) x2 2.6752(14)
3.1009(11) 2.6870(13)
RE-A1(6) x2 3.2308(12) Al(1)-Al(9) x4 2.9551(11)
3.1890(11) 2.9162(10)
RE-Al(7) x2 3.1751(11) Al(2)-Al(4) x2 2.8397(14)
3.1326(10) 2.8028(13)
RE- A1(g) 3.2982(5) Al(2)-Al(6) 2.7004(19)
3.2717(3) 2.664(2)
M( 1 )-Al( 1) x2 2.4497(5) Al(2)-Al(7) 2.7494(19)
2.4353(4) 2.7644(19)
M(1)- 111(2) x2 2.5925(10) Al(3)-Al(4) x2 2.9316(14)
2.5874(9) 2.8834(14)
M(1)-Al(4) 2.363 2( 1 5) Al(3)-Al(5) 2.687(2)
2.3470(14) 2.708(2)
101(1)- Al(7) x2 2.6029(10) Al(3)-Al(6) 2.740(2)
2.5804(9) 2.7887(19)
M(1)-Al(9) x2 2.5740(16) Al(3)-Al(7) 2.710(2)
2.5576(15) 2.682(2)

 

196

Table 6-2-4. (Continued) bond lengths (A) for EuNi2-,,Fe,,Alg and YbNi2-,,FexAlg.

 

 

Bond Distance Bond Distance
M(1)-M(1) 2.7825(13) Al(3)-Al(9) x2 2.9980(15)
2.8051(12) 2.997(2)
M(2)-Al(3) x2 2.5270(9) Al(4)-Al(5) x2 2.812503)
2.5177(9) 2.8048(13)
M(2)-Al(4) 2.3638(15) Al(4)-Al(7) x2 2.8612(14)
2.3653(14) 2.8470(14)
M(2)-Al(5) 2.5885(10) Al(5)-Al(6) 2.7004(19)
2.5892(10) 2.717(2)
M(2)-Al(6) 2.5432(10) Al(5)-Al(7) 2.8717(19)
2.5256(9) 2.8349(19)

 

197

The negative value of Weiss constant (0 = 492.2 K) implies strong antiferromagnetic
coupling and/or a Kondo effect. However no magnetic ordering was observed down to 3
K. Below 20 K, the inverse magnetic susceptibility deviated from a linear behavior,
probably due to thermal depopulation of the crystal ﬁeld effects.

The magnetization behavior of YbNiz-,,Fe,,Alg with the ﬁeld perpendicular or
parallel to the c-axis is Shown in Figure 6-2-4B. In both cases there was no sign of
magnetic saturation up to 55000 G; and substantial magnetic anisotropy was observed.
When the ﬁeld was perpendicular to the c—axis, the magnetization values were higher
than those when the ﬁeld was parallel to the c-axis. This orientation dependent behavior
implies that the magnetic spins are conﬁned to the ab plane. In the case of applied ﬁeld
perpendicular to the c-axis, the magnetization reached 0.16 1.13 at 55000 G, which is only
about 4% of the saturation value.

Figure 64245A shows the temperature dependent electrical resistivity p(T) of
YbNi2-,,Fe,,Alg, which indicates metallic behavior of this material. The resistivity value at
room temperature was ~185 uQ-cm, and it decreased with decreasing temperature
tending towards a constant value at low temperatures. A broad shoulder was observed

9 The thermoelectric

around 160 K, characteristic of intermediate valence systems.l
power, shown in Figure 642458, was about 41 uV/K at room temperature. The small

magnitude of the therrnopower is also indicative of the metallic system of YbNi2_,,Fe,,Alg;

and the negative values suggest that it is an n-type material.

198

 

  

 

 

 

 

 

 

 

0.025 700
- 600
0.02
:5 -1 500
——>
i: 0.015 . 400 1:
a a"
3E 0.01 ‘ 300
x 4 200
0.005 100
“.0 . Q . . . . . .
O l I l l L O
0 50 100 150 200 250 300 350
Temperature (K)
B)
0.20
:3 V H i c axis
E 0.16 - O Hiso , ' -
; AH || caxis , '
c ' q
a 0.12 1- '
8 v
'13 0.08 4 v 4
N
'4': v
g v
35’ 0.04 - v' . . . -
2 " e ' . . . :. a A
o .1233. 9 ‘ i ‘ 1 1
0 210‘ 410‘ 610‘
Field (G)

Figure 64244. A) Temperature dependent magnetic susceptibility of a polycrystalline
sample of YbNi2-,,Fe,,Alg under the ﬁeld of 1000 G. B) Molar magnetization of YbNiz-
xFe,,A13 single crystals oriented with [001] axes perpendicular, parallel to the external

magnetic ﬁeld and isotropic in ﬁelds up to 55000 G.

199

A)
188

 

186 1' o
184 4 .-
182 "' o.

180 " o.

Resistivity (119-cm)

178 - .°

I
176 La"

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

 

 

 

13)

 

Thermopower S (11V/K)

 

 

 

-7 I 14 LI I I I I I I I I I I I I I

300 320 340 360 380 400
Temperature (K)

Figure 64245. A) Temperature dependence of the electric resistivity on a polycrystalline
pellet of YbNi2-xFe,,Alg. B) Temperature dependence of the thermoelectric power on a
single crystal of YbNi2-,,Fe,,Alg.

200

Mo'ssbauer Spectroscopy:

To further probe the electronic state of Fe in RENi2-,,Fe,,A13, 57Fe Méssbauer
spectroscopy was taken at temperatures of 80 K and RT and the spectrum was ﬁtted with
Lorentzian form, shown in Figure 642-6. The Méssbauer spectrum parameters are listed
in Table 642-5.

Each spectrum consists of two doublets indicting two Fe atomic sites (M1 and
M2) in the structure. The doublet with large quadrupole splitting is assigned to site Fe(l)
while the doublet with the smaller quadrupole value is attributed to site Fe(2), because
Fe(l) is also bonded to another Fe(l) atom besides nine Al atoms and the principal
electric ﬁeld gradient at Fe(l) is larger than that of Fe(2). For the Fe(l) site, the
quadrupole splitting values AEq at RT and 80 K are the same — 063(2) mm/s; For the
Fe(2) site, the AEq values are 0.37(2) mm/s and 0.41(2) mm/s at RT and 80 K,
respectively. The similarities of the quadrupole splitting values AEq at different
temperatures suggest the absence of any transition in this temperature range. The isomer
shift value of Fe(l) is 0.25(2) mm/s at 80 K, very close to the values of metallic Fe;
however the isomer shift value of Fe(2) is much higher — 0.42(2) mm/s which has more
ferric character. Similar nonmagnetic character of the Fe atoms was observed in other
Fe-containing intermetallic compounds, such as REFezAllo,22 REzFe3Si5,23 RE4Fe2+xAl7-

xSi324 and RE4FeGa12_,,Ge,,.25

201

100

 

 

 

 

 

 

 

99-
80K
98-
’3
°\ .
v
c: 974
.9.
.12 J
g 96 . , i
c: 44 42 2
1‘3
5" 100
0)
.2
3 99
o _
m
98—
974 RT
%‘
44 -2 0 2

Velocity (mm/s)

Figure 642-6. 57Fe Mbssbauer spectrum of YbNi2-,,FexAlg at 80 K and RT.

202

Table 642-5. 57Fe Méssbauer spectra parameters for YbNi2-,,Fe,,Alg at RT and 80 K.

 

 

 

 

 

Site M(l) Site M(2)
IS 6 AEq Area IS e AEq 0
Temperature (K) (mrrl/S) (mm/S) (%) (mnli/s) (mm/s) Area (/o)
300 015(2) 063(2) 67(2) 028(2) 037(2) 33(2)
80 025(2) 063(2) 66(2) 042(2) 041(2) 34(2)

 

 

 

 

 

 

 

642-4. Conclusions

New quaternary intermetallic aluminides RENi2-,,Fe,,Alg (RE = Eu and Yb) were
synthesized by using Al as a high temperature solvent. The RENi2-,,Fe,,Alg species are
the ﬁrst quaternary analogues of the CaCozAlg structure type. Mixed occupancy is
present on both transition metal sites (M1 and M2) by Ni and Fe. Both magnetic
properties and Méssbauer spectra show that the Fe atoms do not carry magnetic
moments. Temperature dependent magnetic susceptibility indicates that the Yb atoms are
in an intermediate oxidation state with effective magnetic moment ucfr = 2.19 113. The
magnetization behavior studies point out that the ab plane is the easy plane on which the
magnetic moments are conﬁned. Both resistivity and therrnopower measurements exhibit
metallic character of YbNi2_,,Fe,,Alg. Furthermore ”Fe M(‘Sssbauer spectra investigations

conﬁrm the absence of magnetic ordering in this material.

203

 

References:

 

' Singh, Y.; Ramakrishnan, 8. Phys. Rev. B 2003, 68, 054419.

2 Kaczorowski, D.; Andraka, B.; Pietri, R.; Cichorek, T.; Zaremba, V. 1. Phys. Rev. B
2000, 61, 15255.

3 Trovarelli, O.; Geibel, C.; Buschinger, B.; Borth, R.; Mederle, S.; Grosche, M.; Spam,
G.; Steglich, F.; Brosch, O. Donnevert, L. Phys. Rev. B 1999, 60, 1136.

‘1 Doniach, s. Physica B&C 1977, 913, 231.

5 Bauer, E. D.; Bobev, S.; Thompson, J. D.; Hundley, M. F.; Sarrao, J. L.; Lobos, A.;
Aligia, A. A. J. Phys. : Condens. Matter 2004, I6, 4025.

‘1 Czech, E.; Cordier, G.; Schaefer, H. J. Less-Common Met. 1983, 95, 205.

7 Koterlin, M. D.; Morokhivskii, B. S.; Lapunova, R. V.; Sichevich, O. M. Sov. Phys.
Solild State 1989, 31, 1826.

8 Tamura, I.; Mizushima, T.; Isikawa, Y.; Sakurai, J. J. Magn. Magn. Mater. 2000, 220,
31.

9 Tougait, O.; Kaczorowski, D.; Noel, H. J. Solid State Chem. 2005, 1 78, 3639.

‘0 Sichevits, O. M.; Lapunova, R. V.; Grin, Yu. N.; Yarmolyuk, Ya. P. Izv. Akad. Nauk
SSSR, Metally 1985, 6, 117.

‘1 Gladyshevskii, R. E.; Yarmolyuk, Ya. P.; Grin, Yu. N. Kristallografiya 1983,28, 1090.
”- Schliiter, M.; Jeitschko, w. Inorg. Chem. 2001, 40, 6362.

'3 P6ttgen, R. Kumann, D. z. Anorg. Allg. Chem. 2001, 62 7, 55.

'4 SMART, version 5; Siemens Analytical X-ray Systems, Inc.: Madison, WI, 1998.

'5 Saint, Version 4; Simens Analytical X-ray Instruments Inc., Madison, WI.

'6 SADABS, Sheldrick, G. M.; University of Géttingen, thtingen, Germany.

'7 G.M. Sheldrick, 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

‘3 CrystalDiffract is © 1995-1996, Dr. David C. Palmer.

204

 

‘9 Sales, B. C.; Wohlleben, D. K. Phys. Rev. Lett. 1975,35, 1240.
2" Salvador, J. R.; Guo, F.; Hogan, T.; Kanatzidis, M. G. Nature 2003, 425, 702.
2' Singh, Y.; Ramakrishnan, S. Phys. Rev. B 2003, 68, 054419.

22 Tiede, v. M. T.; Bbel, T.; Jeitschko, w. J. J. Mater. Chem. 1998, 8, 125.

23 Moodenbaugh, A. R.; Cox, D. E.; Vining, C. B.; Segre, C. U. Phys. Rev. 1984, 829,
109.

24 Sieve, B.; Sportouch, S.; Chen, X. Z.; Cowen, J. A.; Brazis, P.; Kannewurf, C. R.;
Papaefthymiou, V.; Kanatzidis, M. G. Chem. Mater. 2001, 13, 273.

 

25 Zhuravleva, M. A.; Wang, X.; Schultz, A. J.; Bakas, T.; Kanatzidis, M. G. Inorg.
Chem. 2002, 41 , 6056.

205

 

CHAPTER SEVEN

Synthesis and Characterization of New Zintl Phases:

Yb3Ale3 and ngAl3Sb9

741. Introduction

In the past decade, there has been increasing interest in Zintl phases composed of
an alkaline-earth metal (Ca, Sr, Ba) or divalent rare earth metal (Eu, Yb), a transition
metal (Mn, Zn, Cd) or a group 13 element (Al, Ga, In) and a pnicogen element (P, As,
Sb, Bi).l In these systems, the classic octet rule (8-N) is satisﬁed and closed shell
conﬁguration is achieved. These Zintl phases present a striking variety of structural
features and interesting electronic and magnetic properties. The 1441-11 family
R14(TM)Pn11 ( R = Ca, Sr, Ba, Eu, Yb; TM = Mn, Zn, Cd; Pn = P, As, Sb, Bi) is such an
example that offers a wide range of interesting properties. Eu14MnSb11 is an intermetallic
collosal magnetoresistive material, presenting a metal-insulator transition associated with
a ferromagnetic phase transition at 92 K.2 EulaMnBin orders antiferromagnetically and
shows a large negative magnetoresistance possibly associated with strong ferromagnetic

3 Yb14MnSb1, exhibits a ferromagnetic transition at 52 K with well-

ﬂuctuations.
separated Mn3+ ions acting as local moments.4 Yb14A1Sb11 is a nonmagnetic nonmetallic
Zintl compound; when substituting trivalent aluminum ions with divalent Zn, the
isostructural compound Yb142nSb11 is formed with Yb ions in intermediate valence to
balance the charge.5 The 1441-11 family is not the only example that is reluctant to
deviate from balanced charge. ngZnaBig crystallizes in the orthorhombic CaglnaBio

structure type, and magnetic susceptibility measurements indicated intermediate

oxidation state of Yb2+/Yb3+.6 So this compound was rationalized as (Yb3+)(Yb2+)g(Zn2'

206

)4(Bi2')2(Bi")7. This discovery raised questions as to the nature of the parent compound
CaolnaBio, which contains the same [ZnaBig]‘9' ribbons.7 However, there are only nine
divalent cations (Ca2+) in this compound and obviously the electron count falls one
electron short. Later Kauzlarich and co-workers reported Zintl compounds ngZn4+be9
and Caozna58b9, the structure of which can be regarded as an interstitially stabilized
variant of the Ca9ZmBi9 structure type.8 The charges of these two compounds can be
balanced by accommodating additional Zn atoms, with both ytterbium and calcium ions
in +2 oxidation state, which is supported by magnetic measurements. This discovery
opened up the opportunity to reexamine the nature of previously reported fully
stoichiometric compounds AEgTM4Pn9 (AE = Ca, Sr; TM = Zn, Cd, Mn; Pn = Sb, Bi). It
might be interesting to study this system by substituting divalent Zn with trivalent Al
atoms in an attempt to tune the electronic properties.

Herein we describe the synthesis, structural characterization and physical
properties of new Zintl phase Yb9A13Sb9 grown from A1 ﬂux. It also crystallizes in the
CaoznaBio structure type; the Zintl concept in this compound still holds with one of the
Al sites half occupied. During an attempt to explore the quaternary system Yb/Al/Sb/Si
in liquid A1, another new phase Yb3Ale3 was formed. It crystallizes in the Ca3Ale3
structure type and is the ﬁrst rare earth analogue of the family AE3MPn3 (AB = Ca, Sr,
Ba; M = A1, Ga; Pn = Sb, Bi). Notably, Yb3Ale3 has the same stoichiometry as
YboAl3Sb9, while they present different structure types and properties. In this chapter,
synthesis, crystal structure, physical properties and electronic structure calculations of

Yb3Ale3 and Yb9A13Sb9 Will be described.

207

 

 

 

742. Experimental Section
Reagents:
The following reagents were used as obtained: Yb (Cerac, 99.9%), Sb (Cerac,

chips, 99.999%), Al (Cerac, 99.5%, 420 mesh).

Synthesis:

All manipulations were performed in a nitrogen-ﬁlled dry box. Yb, Sb and Al
metals were loaded in alumina crucibles, which were then put into the silica tubes (13mm
in diameter). These silica ampoules were sealed under vacuum (~104 Torr).

Initially Yb9A13Sb9 was obtained by combining 1 mmol Yb metal (0.173 g) and 1
mmol Sb (0.122 g) with a large excess amount of Al (0.810 g). The sample was heated to
1000 °C in 15 h, maintained at this temperature for 36 h, then slowly cooled to 250 °C in
3 (1. Finally the temperature was brought down to 50 °C in 5 h. The excess aluminum
was removed by soaking the crucible in aqueous 5M NaOH solution overnight. The solid
product remaining after the isolation procedure was rinsed with water and acetone. The
yield of Yb9A13Sb9 was about 60% with the side products mainly Yb14Ale111° and
Ale".

After the structure of YboAl3Sb9 was determined by single crystal X-ray
diffraction, we tried to produce a pure phase of Yb9A13Sb9 by stoichiometric reactions.
The mixtures of the elements in ratio of 3:1 :3 for Yb:Al:Sb were heated to 750 °C in 2 d,
stayed at this temperature for 24 h and then slowly cooled to room temperature.
Yb9A13Sb9 was obtained as the major phase (80% yield) with trace amount of impurity

phases Yb1 1Sb1010 and Yb14Ale115.

208

Yb3Ale3 was found during an attempt to explore the system Yb/Al/Sb/Si using
Al as a ﬂux. 1 mmol Yb metal (0.173 g), 3 mmol Si (0.084 g), 1 mmol Sb (0.122 g) and
20 mmol Al (0.270 g) were combined in an alumina crucible. The sample was heated to
1000 °C in 48 h, maintained at this temperature for 36 h, then cooled to 500 °C in 36 h.
Yb3Ale3 can also be formed by direct combination reactions. The mixture with molar
ratio l.5:0.5:l.5 for Yb:Al:Sb was heated to 1000 °C in 15 h, maintained at this
temperature for 4 h, then cooled to 850 °C in 2 h. It was annealed at 850 °C for 8 (1,
followed by Slowly cooling down to 50 °C at a rate of 8 °C/h. Yb3Ale3 was found as

the major phase with traces of Yb5Sb3 as the side product.11

Differential Thermal Analysis:

Differential thermal analysis (DTA) was performed on Yb3Ale3 and Yb9A13Sb9
with a Shimadzu DTA-50 thermal analyzer. The ground samples (~20 mg) were sealed
in carbon-coated Silica ampoule under vacuum. A silica ampoule containing equal
amount of alumina was sealed and placed on the reference side of the detector. The
samples were heated to 1000 °C at 10 °C/min and held for 5 min, followed by cooling at
—1 0 °C/min to 50 °C. The residue of the DTA experiment was examined with powder X-

ray diffraction.

X-ray Crystallography:
Single crystal X-ray diffraction data were collected for ngAthg and Yb3Ale3
at room temperature on a Bruker AXS SMART CCD X-ray diffractometer. Single

crystals grown from A1 ﬂux were selected and mounted on glass ﬁbers. The data

209

collections (Mo K01 radiation, A = 0.71073 A) were acquired covering a full sphere of
reciprocal space. The data acquisition and cell reduction were done with the SMART12
software package and data processing was performed with the SAINTPLUS software
package. '3 An empirical absorption correction was applied to the data using the
SADABS program.14 The structures were solved using direct methods and reﬁned with
the SHELXTL package program. ‘5

Inspection of systematic absence conditions of ngAthg revealed two possible
space groups: centrosymmetric space group Pbam (#55) and non-centrosymmetric Pba2
(#32). These two candidates presented similar ﬁgure of merit. The centrosymmetric
space group Pbam was chosen and later structure reﬁnement conﬁrmed this choice.
Direct methods gave a structure solution with ﬁve Yb sites, ﬁve Sb sites and two Al sites.
Among them there were four atomic Sites — Al(2), Yb(4), Yb(5) and Sb(5) showing
unreasonable high thermal parameters. Examination of Al(2) site revealed that it was
only 50% occupied (Al is the lightest element in this compound). After this cycle of
reﬁnement, the R, factor dropped to 12.54% and the thermal parameters of Al(2) changed
back to normal. All the other three atomic sites Yb(4), Yb(5) and Sb(5) were found fully
occupied; and an analytical absorption correction was applied but could not solve the
problem of high thermal parameters. The bond distances between Al(2) and Yb(4) or
Yb(5) atom are 2.879(11) A, 3.044(11) A and 3.106(17) A, respectively, which are
Shorter than the other Yb-Al bond distances (3.262(2) A ~ 3.503(17) A); while the bond
distance from Al(2) to Sb(5) is 3.027(17) A, much longer than the other Al-Sb bond

distances (2.728(9) A ~ 2.768(9) A). We believe that it is the disordered Al(2) site that

210

raises the thermal shifts of Yb(4), Yb(5) and Sb(5) atoms resulting in high thermal
parameters, position uncertainty and high standard deviations of bond distances.

The structure of Yb3Ale3 was solved in a straightforward fashion. Inspection of
the systematic absence conditions brought two possible space groups: Pna21 (#33) and
ana (#62). ana is a centrosymmetric space group; plus its ﬁgure of merit is much
lower (2.59), so ana was chosen and later structure reﬁnement proved that this was the
right one. Moreover, review of the literature revealed that Ca3Ale3 has similar cell
parameters and same space group: ana, a = 12.835(5) A, b = 4.489(2) A, c = 14.282(5)
A.16 This implies that Yb3Ale3 is an isostructural compound of Ca3Ale3_ All the seven
atomic sites were proved to be fully occupied and reﬁned anisotropically. The R, factor
ﬁnally dropped to 2.44% with highest residual peak 1.98 e'/A'3, almost the same as the
deepest hole 441.87 e'/A'3.

Data collection parameters and reﬁnement details for Yb9A13Sb9 and Yb3Ale3
can be found in Table 741. Atomic positions, displacement parameters and anisotropic
displacement parameters for ngAthg and Yb3A1Sb3 are listed in Tables 7-2, 7-3, 744
and 745.

X-ray powder diffraction data were collected at room temperature with a CPS 120
INEL X-ray difﬁactometer (Cu Ka) equipped with a position-sensitive detector.
Experimental powder patterns were compared to the patterns calculated from the single
crystal structure solution (by the CrystalDiffract program 17) to determine the phase

identity and purity.

211

Charge transport measurements:

Electrical conductivity and therrnopower measurements were made on Yb9A13Sb9
and Yb3Ale3. The samples were ground and made into pellets. Conductivity
measurements were performed in the usual four-probe geometry. Thermopower data
were from 300 K to 400 K with a MMR Technologies, Inc. Seebeck measurement

system.

Magnetic Susceptibility Measurements:

Magnetic susceptibility measurements were performed with a MPMS Quantum
Design SQUID magnetometer. The data were collected in the temperature range 34300 K
at 1000 G, while ﬁeld dependent magnetic measurements, conducted at 5 K, were carried
out in ﬁelds up to i 55000 G. A diamagnetic correction was applied to the data to

account for core diamagnetism.

Band Structure Calculations:

To understand the bonding ian3Ale3, we carried out ab initio band structure
calculations using the density functional theory (DFT) formalism. The calculations were
performed using the self-consistent full-potential linearized augmented plane wave
method (LAPW), with the generalized gradient approximation (GGA) for the exchange

and correlation potential. Spin-orbit interactions were included in the calculations.

212

 

Table 7-1. Selected crystal data and structure reﬁnement details for Yb9A13Sb9 and

 

 

Yb3A1Sb3.

Empirical formula Yb9A138b9 Yb3Ale3
Formula weight 2734.05 911.35
Temperature 293(2) K 293(2) K
Crystal system Orthorhombic Orthorhombic
Space group Pbam (# 55) ana (#62)

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient
F (000)

Crystal size

Theta range for data collection

Limiting indices

Reﬂections collected

Independent reﬂections

Completeness to theta = 37.000

Reﬁnement method

Data / restraints / parameters

Goodness-of-ﬁt on F 2

Final R indices [I>28igma(l)]

R indices (all data)

Largest diff. peak and hole

a = 12.433(2) A
b = 21.916(4) A
c = 4.5876(9) A
1250.1(4) A3

2

7.264 Mg/m3

42.905 mm'l

2256

0.16 x 0.12 x 0.10 mm3

1.86 to 27.94°
455h515
-28$k£28
465135

12689
1612 [R(int) = 0.0765]
95.7 %

a = 12.838(5) A
b = 4.502807) A
c = 14.220(5) A
822.0(5) A3
4
7.364 Mg/m3
43.497 mm'l
1504
0.43 x 0.26 x 0.33 mm3

2.14to 28.21°
416Sh516
455k55
41751518

8651
1099 [R(int) = 0.0512]
97.3 %

Full-matrix least-squares on F2

1612 / 0 / 73
1.282
R, = 0.0648
wR2 = 0.1010
R, = 0.0838
wR2 = 0.1053

5.987 and — 7.848 6A3

2 2 2 2 172
R1 = Z(lFol-IFeIVZIFol; wR2 = [2[W(Fo - Fe ]/ [2(WlFol ) l

213

1099 / 0 / 44
1.092

R, = 0.0244

wR2 = 0.0517

R, = 0.0339

wR2 = 0.0541

1.978 and —1.868 e.A'3

Table 742. Atomic coordinates (A x 104) and equivalent isotropic displacement

parameters (A2 x 103) for Yb9A13Sb9.

 

Wyk.

 

Atom Symbol x y z Occu. U(eq).
Yb(l) 2b 0 0 5000 1.0 14(1)
Yb(2) 4g -1347(1) 2366(1) 0 1.0 11(1)
Yb(3) 4g 1247(1) 1325(1) O 1.0 9(1)
Yb(4) 4g 4844(1) 4153(1) 0 1.0 33(1)
Yb(5) 4h 42910(1) 958(1) 5000 1.0 25(1)
Sb(l) 4h 4451(1) 1498(1) 5000 1.0 10(1)
Sb(2) 4h 42031(1) 3303(1) 45000 1.0 8(1)
Sb(3) 4g -3932(1) 1920(1) 0 1.0 9(1)
Sb(4) 4g 41886(2) 38(1) 0 10 10(1)
Sb(5) 2d 0 5000 45000 1.0 53(2)
Al( 1) 4h 1097(7) 23 80(4) 45000 1.0 9(2)
Al(2) 4h -2298(l 3) 4545(8) 45000 0.5 9(6)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.

214

 

 

Table 743. Anisotropic displacement parameters (A2 x 103) for Yb9A13Sb9.

 

Atom
Yb( 1)
Yb(2)
Yb(3)
Yb(4)
Yb(5)
Sb( 1)
Sb(2)
Sb(3)
Sb(4)
Sb(5)
Al( 1)
Al(2)

U11
12(1)
11(1)
10(1)
11(1)
26(1)

8(1)
10(1)
7(1)

11(1)
123(5)
10(4)
0(9)

U22
14(1)
6( 1)
8(1)
8(1)
28(1)
11(1)
3(1)
6(1)

5(1)
28(2)
4(4)
8(10)

U33
15(1)
15(1)
10(1)
78(3)
21(1)
11(1)
10(1)
14(1)
14(1)

9(2)
12(4)
19(11)

C
I»)
DJ

OOOOOOOOOOOO

(3..

OOOOOOOOOOOO

2(1)
2(1)
2(1)
-3(1)
16(1)
0(1)
0(1)
42(1)
0(1)
41(3)

1(3)
-3(6)

 

42n2[h2a*2U1l+...+2hka*b*U12]

215

The anisotropic displacement factor exponent takes the form:

 

 

 

Table 744. Atomic coordinates (A x 104) and equivalent isotropic displacement

parameters (A2 x 103) for Yb3AISb3.

 

 

 

 

 

Atom Symbol x y 2 U(eq).
Yb(l) 4c 9404(1) 2500 6114(1) 13(1)
Yb(2) 4c 2272(1) 2500 7212(1) 13(1)
Yb(3) 4c 6500(1) 2500 4941(1) 14(1)
Sb(l) 4c 7562(1) -2500 6190(1) 11(1)
Sb(2) 4c 8864(1) 2500 3905(1) 11(1)
Sb(3) 4c 9599(1) 2500 8507(1) 11(1)
Al 4c 9334(3) 2500 2031(2) 13(1)
Um) is deﬁned as one third of the trace of the orthogonalized Uij tensor.
Table 745. Anisotropic displacement parameters (A2 x 103) for Yb3Ale3.
Atom U11 U22 U33 U23 U13 U12
Yb(l) 10(1) 13(1) 16(1) 0 0(1) 0
Yb(2) 12(1) 11(1) 16(1) 0 42(1) 0
Yb(3) 14(1) 13(1) 15(1) 0 41(1) 0
Sb(l) 10(1) 8(1) 15(1) 0 1(1) 0
Sb(2) 11(1) 11(1) 10(1) 0 0(1) 0
Sb(3) 9(1) 9(1) 14(1) 0 1(1) 0
Al 12(1) 12(1) 14(1) 0 1(1) 0

 

The anisotropic displacement factor exponent takes the form:

42n2[h

2211

a*U

+. . .+2hka*b*U'2]

216

 

743. Results and Discussion
Synthesis and Differential Thermal Analysis:

Initially both YboAl3Sb9 and Yb3Ale3 were prepared in molten Al. The isolated
crystals were shiny and black with needle morphology. The stoichiometries of the two
compounds are identical, however they adopt different structure types which will be
described in detail in the crystal structure section.

Both compounds can be made by stoichiometric reactions, and small amount of
impurity phases such as YbnSbm and Yb5Sb3 were also detected by powder X-ray
diffraction analysis. YboAl3Sb9 was synthesized at 750 °C, but reactions at high
temperatures (850 °C) formed Yb3Ale3, which suggests that Yb3A1Sb3 is probably a
more thermodynamically stable phase. This argument is also supported by DTA
experiments. Powder X-ray diffraction pattern after DTA showed that Yb9A13Sb9
converted to ngAle; along with some Yb5Sb3. The more thermodynamically stable

phase Yb3Ale3 melts at 820 °C and recrystallizes at 801 °C.

Crystal Structure of Y boAl 3Sb9:

The crystal structure of ngAlngo is shown in Figure 741. The main feature of
this structure, as described elsewhere,” is the ribbon-like chains (Figure 742A) composed
of four Alea tetrahedral units running along the c-axis. The four Ale4 units are
connected with each other by Sharing Sb corners. Sb(2), Sb(3) and Sb(4) atoms are
shared by Al(1)- and Al(2)-centered tetrahedra to form a ribbon, see Figure 7-2A. The
Sb(S) atom is bridging two double-chains in a linear fashion with Al(2)-Sb(5)—Al(2) angle

of 180°. The ﬁve different Yb cations occupy the empty spaces between the chains to

217

 

 

 

counter-balance the charge. Their coordination environments are shown in Figure 7-
ZB~F, all featuring a distorted octahedral geometry composed of Sb atoms. The bond
distances between Yb and Sb atoms fall in the range of 3.090(2) ~ 3.423(2) A. For the
Yb(3), Yb(4) and Yb(5) atoms, they are also capped by four, two and one Al atom,
respectively. The disorder of Al(2) results in unusual close contact of Yb(4)-Al(2) which
is 2.879(11) A, much shorter than the other Yb-Al bond distances (3.044(11) ~ 3.503( 17)
A).

An interesting feature of the structure of YboAl3Sb9 is on the Al(2) site. There are
two independent Al Sites: Al(l) is fully occupied, while Al(2) Site is only 50% occupied.
The possibility that this site could be occupied by a lighter element was excluded since
EDS analysis did not detect any other element.

According to the Zintl concept, the formal charge of a four-bonded Zn atom can
be assigned as an' and those one-bonded Sb and two-bonded Sb atoms can be assigned
as sz' and Sb". So for the .413ngan family (AB = Ca, Sr; TM = Zn, Cd; Pn = Sb,
Bi), the anion lattice can be described as [(TMZ')4(1b-Pnz')2(2b-Pnl’)7]'9’, which requires
19 electrons to balance the charge. However the divalent alkaline metal cations can only
provide 18 electrons, which challenges the correctness of this structure. The publication
of the charge-balanced Zintl phase ngZmBig ﬁrst raised this concern.6 In this formula
the Yb ions are mixed-valent of Yb2+/Yb3+ instead of divalent, and this argument is
supported by magnetic susceptibility measurements. So YbolmBio can be rationalized as
[(Yb3+)(Yb2+)3][(Zn2')4(lb-Biz')2(2b-Bil')7]. Later two other members ngZn45Sb9 and
CaoZn45Sb9 were discovered.8 Additional Zn atoms were found in these two structures

which connect two [Zme9]'9' ribbons to form a [Zna.5Sb9]'8’ anionic sub-lattice.

218

Therefore the ytterbium or calcium cations in ngZn4,5Sb9 and Ca92n4_5Sb9 are in divalent
oxidation state. The present compound Yb9A13Sb9 demonstrates its Zintl nature by
requiring less Al components since divalent Zn ions are substituted by trivalent Al atoms.
According to the simple formal oxidation rules, this Yb9A13Sb9 composition gives a

complete charge balance: (Yb2+)9[(Al3+)3(Sb3')9].

Crystal Structure of Yb3Ale3:

AS mentioned before, Yb3Ale3 crystallizes in the Ca3Ale3 structure type.16
Ca3GaAS3, Ca3InP3 and Sr3InP3 are isostructural compounds that belong to this family.'8
As shown in Figure 743, this structure is composed of Ale4 tetrahedra which form an
inﬁnite chain along the b-axis by sharing Sb corners; these [Ale3]6' anionic species are
separated by Yb ions, which occupy the empty spaces and provide charge balance.
Interestingly, Ba3Ale3 and Ba3GaSb3 have the same stoichiometry as Yb3Ale3;
however in these two compounds two distorted MSb4 (M = A1, Ga) tetrahedra are
connected by a common Sb-edge to form an isolated dimeric anion [MZSb6]'2'. Another
compound with the same stoichiometry was reported by Cordier 4— Sr3GaSb3, in which
distorted GaSb4 tetrahedra are connected by common Sb corners, to give strands with
groups of four tetrahedra as the repeating unit.18 It is not clear at this point whether size
determines the structure type; since Yb and Ca are similar in size, although Sr is a larger
cation than those two.

Figure 744A depicts how the Ale4 tetrahedra form a polymeric chain by sharing
Sb atoms. The MPna (M = A1, Ga; Pn = As, Sb, Bi) tetrahedron is a very common unit

and it has been observed in many Zintl phases, such as A5M2Pn6,l9 A11MPn920 and

219

A14MPn,,.20"’ 2' In Yb3Ale3, the Al-centered tetrahedra are built of three
crystallographically inequivalent Sb atoms: Sb(l), Sb(2) and Sb(3). The bond distances
are quite similar: 2.713(4) A, 2.732(4) A and 2.744(2) A for Al-Sb(l), Al-Sb(2) and Al-
Sb(3) respectively. The Ale4 tetrahedron unit is in a distorted fashion with Sb-Al-Sb
angles 112.46(8)° and 108.96(9)°, which are deviated ﬁom the ideal tetrahedral angle of
1095".

The three independent Yb sites occupy the space between the Ale4 tetrahedra;
and their local coordination environments are shown in Figure 744B,C&D. Both Yb(l)
and Yb(3) atoms are sitting in distorted octahedral geometries composed of six Sb atoms.
The Yb-Sb bond distances are in the range of 3.1173(9) A ~ 3.4113(15) A, which are
comparable to the sum of the covalent radii of Yb and Sb atoms (3.20 A). The Yb(2)
atoms are also in distorted octahedral geometries; in addition, the Yb(2) atom is also
bonded to two capping Al atoms with Yb-Al distance 3.237(3) A.

According to the Zintl concept, the formal charge of a four-bonded Al atom can
be assigned as Al" and those one-bonded and two-bonded Sb atoms can be assigned as
sz' and Sb". So the formula Yb3Ale3 can be rationalized as (Yb2+)3[(Al")(1b-Sb2'
)2(2b-Sbl')]. Alternatively one can also describe the formula as (Yb2+)3[(Al3+)(Sb3')3]

using formal oxidation numbers.

Charge Transport Measurements:
To assess the thermoelectric properties of these two Zintl phases, we conducted
preliminary studies of the electrical conductivity and thermoelectric power on Yb3A1Sb3

and Yb9A13Sb9. Yb3Ale3 and Yb9A13Sb9 exhibit similar charge transport properties.

220

 

 

The electrical conductivity for Yb3Ale3 and Yb9A13Sb9 at room temperature was 230
S/cm and 272 S/cm, respectively, comparable to the other Zintl phases such as
BaaIn3Sb16' (135 S/cm) and Yb5m2Sb6 000 S/cm).22 The thermoelectric power of
polycrystalline pellets of Yb3Ale3 and ngAl3Sb9 was measured as a function of
temperature (Figure 745). So far the therrnopower of Yb9A13Sb9 could only be measured
up to 400 K due to the fact that pressed pellets often lose contact at higher temperatures.
In both cases of ngAle; and Yb9A13Sb9, the positive values of thennopower indicate
holes as the charge carrier. The value of the therrnopower for Yb3Ale3 was about 10
uV/K at room temperature and rose to ~21 uV/K at 700 K. For Yb9A13Sb9, the
therrnopower varied from +29 to +45 uV/K between 300 and 400 K, and showed a linear

increase with rising temperature.

Magnetic Susceptibility Measurements:

Single crystals of Yb3Ale3 and ngAl3Sb9 grown from Al ﬂux were picked and
ground into powders for the magnetic susceptibility measurements. In both cases, the
high temperature (above 60 K) magnetic susceptibility data was nearly temperature
independent, as expected for divalent Yb ions (pm = 0). This observation is consistent
with electron counting and band structure calculations. The low temperature magnetic

susceptibility might come from impurities such as szO3.

221

Band Structure Calculations:

In order to compensate for the inability of Density Functional Theory (DFT) to
accurately model the highly correlated f-electrons, the calculations were performed on
four different compounds Yb3Ale3, Lu3Ale3, La3Ale3 and Ca3A1Sb3, three of which
have different f-level conﬁgurations. The Ca system does not have an f-level in the
energy range of interest and it is isostructural with the three rare earth systems. The
density of states (DOS), as computed for the above mentioned compounds are shown in
Figure 746. There are several similar features in these diagrams: a narrow band located at
410 eV, corresponding to the S levels of the Sb ions, a smaller band at 46 and 45 eV, due
to the Al s level and ﬁnally the wide valence and conduction bands which indicates the
strong hybridization between the Al s, p and Sb p states.

The Sharp, narrow bands corresponding to the f-level of Yb, Lu and La are visible
in Figure 7-6A, B and C respectively. In the cases of Yb and Lu, the f-levels are very
narrow (0.1 eV) and are energetically split in two groups because of the spin-orbit
coupling. The crystal ﬁeld splitting can also be seen on the f-level of the Lu ion. In
contrast the empty f-level of La is much broader (0.5 eV), due to strong hybridization
with the conduction band.

The DOS at the Fermi level of LU3Ale3 and La3A1Sb3 is nonzero, indicating
metallic character of these compounds. In the case of the Yb system, the bands near the
Fermi level are composed of f states of Yb, p and S orbitals of Al and p states of Sb. The
nonzero DOS at the Fermi level suggests that this is also a metallic system. However, the
number of electrons occupying the conduction band may be very small. Further a small

downward shift in the position of the f-level will make Yb3Ale3 a semiconductor. In

222

 

this case the Yb ions are in a 2+ oxidation state, consistent with our charge balanced
formulism and the diamagnetic nature of the compound. The DOS plot in Figure 746D
shows a semiconducting gap of about 0.4 eV for Ca3Ale3. Since DFT tends to
underestimate the band gap value, the actual semiconducting gap may be larger than 0.4

eV.

223

 

 

 

Figure 7-1. Crystal structure of Yb9A13Sb9 viewed down the c-axis. Large empty circles:

Yb; black small circles: Al; gray medium circles: Sb.

224

 

ease .3 co aeoeeoeseo eoeeeecoou a z a donor 3.74547; co accuse oﬁ 2 32.34;. co 2368a .3 new;

 

5,5 I
:5 3 0“
A E 38$th

N are

E .

Com 3%

G a
$5
5 . 2 m
0
a?

333.. Am .m 2:.

 

E 2 .m 2

225

 

 

 

 

 

 

Oo '-
Oom 06

"S-

 

 

 

Figure 743. Crystal structure of Yb3Ale3 viewed along the b-axis. Large empty circles:

Yb; black small circles: Al; gray medium circles: Sb.

226

 

A)

 

 

 

 

Sb(2)
Sb(l) l
Sb(2) Sb(3) D) Sb(2)
O ' ?
Sb(3)
*7me Sb(” 5%:
Sb(2) swig? Sb(l)
.
Sb(3) Sb(3)

Sb(3)

Figure 744. Structure of Yb3Ale3. A) Polymeric chain [Ale3]6‘ viewed down the a-axis.

B) C) D) Immediate coordination enviromnent of Yb atoms.

227

Table 7-6. Selected bond lengths (A) for Yb9A13Sb9.

 

 

 

 

 

Bond Distance Bond Distance
Yb(1)-Yb(3) x4 4.0117(11) Yb(4)-Al(2) 3.044(11)
Yb(1)-Sb(1) x2 3.330(2) Yb(4)-Sb(2) 3.3029(17)
Yb(1)-Sb(4) x4 3.2813(14) Yb(4)-Sb(3) 3.345(2)
Yb(2)-Yb(3) 3.9506(19) Yb(4)-Sb(4) 3.423(2)
Yb(2)-Yb(3) 4.1452(19) Yb(4)-Sb(5) 3.1318(10)
Yb(2)-Yb(4) x2 3.967(2) . Yb(4)-Yb(4) 4.265(3)
Yb(2)-Yb(5) x2 4.3084(19) Yb(4)-Yb(5) x2 4.317(2)
Yb(2)-Sb( 1) x2 3.1818(17) Yb(5)-Al(2) 3.106(17)
Yb(2)-Sb(2) x2 3.1937(16) Yb(5)-Sb(1) 3.278(3)
Yb(2)-Sb(3) 3.359(2) Yb(5)-Sb(3) x2 3.3653(18)
Yb(2)-Sb(3) 3.386(2) Yb(5)-Sb(4) x2 3.3086(18)
Yb(3)-Yb(4) x2 3.765(2) Yb(5)-Sb(5) 3.3404(18)
Yb(3)-A1( 1) x2 3.262(6) Sb(1)-Al(l) 2.728(9)
Yb(3)-Al(2) x2 3.489(12) Sb(2)-Al(l) 2.767(9)
Yb(3)-Sb(l) x2 3.1404(16) Sb(2)-Al(2) 2.743(17)
Yb(3)-Sb(2) x2 3.2427(17) Sb(3)-Al(1) x2 2.760 (5)
Yb(3)-Sb(4) 3.090(2) Sb(4)-Al(2) x2 2.730(9)
Yb(4)-Al(2) 2.879(11) Sb(5)-Al(2) 3.026(16)
Table 747. Selected bond lengths (A) for Yb3Ale3.

Bond Distance Bond Distance
Yb(1)-Yb( 1) x2 4177203) Yb(2)-Sb(3) 3.158204)
Yb(1)-Yb(3) 4.0839(15) Yb(3)-Sb(l) x2 3.1747(10)
Yb(1)-Sb(1) x2 3.2670(11) Yb(3)-Sb(2) 3.3733(14)
Yb(1)-Sb(2) x2 3.1645(11) Yb(3)-Sb(3) x2 3.3500(10)
Yb(1)-Sb(3) 3.4113(15) Yb(3)-Sb(3) 3.2910(12)
Yb(2)-Al(1) x2 3.237(3) Sb(1)-Al(1) 2.713(4)
Yb(2)-Sb(1) x2 3.2205(10) Sb(2)-Al(1) 2.732(4)
Yb(2)-Sb(2) x2 3.1173(9) Sb(3)-Al(1) x2 2.744(2)

 

228

A)

M

bh

 

 

 

 

 

 

 

 

I
I
20 4 I ' .
A I
g is 4
3 I
w
2 16 4 ' '
a I
o
E
E 14 -
l- I
I
12 ~
I
lor l 1 I I I 1 l
300 350 400 450 500 550 600 650 700
Temperature (K)
B)
50
45 4
I
g .
>
3 404 I
m I
3‘ I
g I
8- 35
E I '
15 c
304 '
I
25 L 1 L I L l L l A l

300 320 340 360 380 400 420

Temperature (K)

Figure 745. A) Temperature dependence of the therrnopower for a pressed pellet of
Yb3Ale3 from 300 to 700 K. B) Temperature dependence of the therrnopower for a
pressed pellet of chAl3Sb9 from 300 to 400 K.

229

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100 . . ' ; . ~ 100
80 ' A) 80 . B)
8 60 ~ 60 .
a I
40 i 1 I»; : III», 1 40 i i ' ' I M. q
.51 1I 1 ‘ f." m; ‘ ,- 3 M '(F .
~ , 1’ ‘ -’ ‘I I , 1' , A .
0 1.6:! i I) I ‘1 ' i": . 0 {PJ- : 1' I" if") 1“)"; A
15 -10 -5 O 5 10 15 15 —10 -5 0 5 10 15
100 ' f f ' ' 100 . . .
C) I 1 , D) J
80 “ l i ‘ 8O . ‘
I l . i I I
60 ' ' I . 60 I I
3 III . . (1,91; 1’1
D 40 . j ' . 4O I ”I! I .1" 3.1‘1'11'111'115'
20 . 1 1,311“, If" ‘IIIMIIV’YI « 20. Ill '1 II .1 1
J11} M, I ‘I .I i I" l"! Ii 1 1, 1
O z . . l ‘i .‘ I; l a L 0 A i 1.1 1' - A
15 -10 -5 O 5 10 15 15 -10 -5 0 5 10 15
Energy (eV) Energy (eV)

Figure 7-6. (A) The total DOS of Yb3Ale3. The upper spin-split f-band (fm) is located

near the Fermi level. (B) The total DOS of Lu3Ale3. The ﬁlled f band is located below
-5 eV. (C) The total DOS of L33AISb3 with the empty f-level located in the conduction
band. (D) The total DOS of Ca3AISb3. A band gap of about 0.4 eV shows the

semiconducting character of the Ca compound.

230

 

7-4. Conclusions:

We have synthesized and structurally characterized the new ternary zintl phases
Yb3AISb3 and Yb9A13Sb9. Both phases can be obtained by the Al ﬂux method and direct
combination reactions. The fact that Yb3Ale3 was obtained at higher temperature
implies that Yb3Ale3 is a more thermodynamically stable phase, and DTA experiments
conﬁrmed this conclusion. Although these two phases present the same stoichiometry,
they exhibit different structure types. Yb3Ale3 is composed of Ale4 tetrahedra forming
inﬁnite chains along the b-axis; while the main feature of ngAthg is a ribbon-like
chain composed of four Ale4 tetrahedra which connect with each other by sharing Sb
comers. The discovery of Yb9A13Sb9 demonstrated the zintl nature of the 9-4-9 family:
this structure managed to maintain charge-balance by reducing ‘A of the Al atoms. Thus
the charge of the anionic chain is —18; and the Yb cations are in divalent oxidation state

which is supported by magnetic susceptibility measurements.

231

References:

 

I a) Corbett, J. D. Angew. Chem. Int. Ed. 2000, 39, 670. b) Chan, J. Y.; Olmstead, M. M.;
Kauzlarich, S. M.; Webb, D. J. Chem. Mater. 1998, 10, 3583. c) Fisher, I. R.; Budko, S.
L.; Song, C.; Canﬁeld, P. C.; Ozawa, T. C.; Kauzlarich, S. M. Phys. Rev. Lett. 2000, 85,
1120. (1) Kim, S. J .; Hu, S. Q.; Uher, C.; Kanatzidis, M. G. Chem. Mater. 1999, 11, 3154.
e) Holm, A. P.; Park, S. M.; Condron, C. L.; Olmstead, M. M.; Kim, H.; Klavins, P.;
Grandjean, F.; Hermann, R. P.; Long, G. J.; Kanatzidis, M. G.; Kauzlarich, S. M.; Kim,
S. J. Inorg. Chem. 2003, 42, 4660. 0 Park, S. M.; Kim, S. J.; Kanatzidis, M. G. Inorg.
Chem. 2005, 44, 4979.

2 Chan, J. Y.; Kauzlarich, S. M.; Klavins, P.; Shelton, R. N.; Webb, D. J. Chem. Mater.
1997, 9, 3132.

3 Chan, J. Y.; Kauzlarich, s. M.; Klavins, P.; Shelton, R. N.; Webb, D. J. Phys. Rev. B
1998,57,8103.

4 Fisher, 1. R.; Wiener, T. A.; Budko, S. L.; Canﬁeld, P. C.; Chan, J. Y. Kauzlarich, S. M.
Phys. Rev. B 1999, 59, 13829.

5 Fisher, 1. R.; Budko, s. L.; Song, C.; Canﬁeld, P. C.; Ozawa, T. C.; Kauzlarich, s. M.
Phys. Rev. Lett. 2000, 85, 1120.

6 Kim, S. J .; Salvador, J .; Bilc. D.; Mahanti, S. D.; Kanatzidis, M. G. J. Am. Chem. Soc.
2001, 123, 12704.

7 Brechtel, E.; Cordier, G.; Schafer, H. Z. Naturforsch. B 1979, 34, 1229.

8 Bobev, S.; Thompson, J. D.; Sarrao, J. L.; Olmstead, M. M.; Hope, H.; Kauzlarich, S.
M. Inorg. Chem. 2004, 43, 5044.

9 Owen, E. A.; Preston, G. D. P. Phys. Soc. London 1924, 36, 341.

'0 Clark, H. L.;Simpson, H. D.;Steinﬁnk, H. Inorg. Chem. 1970, 9, 1962.

” Brunton, G. D.; Steinﬁnk, H. Inorg. Chem. 1971, 10, 2301.

'2 SMART, version 5; Siemens Analytical X-ray Systems, Inc.: Madison, WI, 1998.
'3 Saint, Version 4; Simens Analytical X-ray Instruments Inc., Madison, WI.

'4 SADABS, Sheldrick, G. M.; University of Gottingen, Gottingen, Germany.

232

 

 

'5 GM. Sheldrick, 1995, SHELXTL. Structure Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments, Inc. Madison, WI.

'6 Cordier, D.;Sohaefer, H.;Stelter, M. Z. Naturforsch. B 1984,39, 727.

'7 CrystalDiffract is © 1995-1996, Dr. David C. Palmer.

‘8 a) Cordier, G.; Schafer, H.; Stelter, M. Z. Naturforsch. B 1987, 42, 1268. b) Cordier,
G.; Schafer, H.; Stelter, M. Z. Naturforsch. B 1985, 40, 1100.

‘9 a) Cordier, (3.; Schafer, H.; Stelter, M. z. Naturforsch. B 1984,39, 727. b) Cordier, (3.;
Schafer, H.; Stelter, M. Z. Naturforsch. B 1985, 40, 975. c) Cordier, G.; Schafer, H.;
Stelter, M. Z. Naturforsch. B 1985, 40, 1100. d) Kim, S.; Ireland, J. R.; Kannewurf, C. R.;
Kanatzidis, M. G. J. Solid State Chem. 2000, 155, 55.

20 a) Cordier, G.; Schafer, H.; Stelter, M. Z. Naturforsch. B 1985, 40, 868. b) Bobev, S.;
Fritsch, V.; Thompson, J. D.; Sarro, J. L.; Eck, B.; Dronskowski, R.; Kauzlarich, S. M. J.
Solid State Chem. 2005, 1 78, 1071.

21 a) Cordier, G.; Schafer, H. Z. Anorg. Allg. Chem. 1984, 519, 183. b) Brock, S. L.;
Weston, L. J .; Olmsted, M. M.; Kauzlarich, S. M. J. Solid State Chem. 1993, 107, 513.

22 Kim, S. J .; Ireland, J. R.; Kannewurf, C. R.; Kanatzidis, M. G. J. Solid State Chem.
2000, 155, 55.

233

 

CHAPTER EIGHT

Conclusions and Future work

It has been demonstrated that the Al ﬂux method is an effective approach to
synthesize single crystals of new intermetallic phases. To better understand the reaction
pattern of A1 with a transition metal and silicon or germanium, we systematically
investigated the system RE/Au/Al/Ge in liquid Al. In this system, Al is serving as a
reactive solvent and it tends to form quaternary compounds with Ge. In addition, parallel
chemistry was found between the ﬁrst row and the third row transition metals, especially
between Au and Ni. The hexagonal phase RETMAhGez can be formed both with Au and
Ni; REAuA14(AuxGe1-x)2 and RliNiA14(Si2.,.Ni,.)l are isostructural phases which have a
transition metal and tetrelide on the mixed-occupied site.

Our results in the system RE/Au/Al/Ge also gave a clear comparison between Si
and Ge chemistry. Au reacts with RE and Si in liquid A1 to form a number of new phases
such as the Th2(AuxSi.-x)[AuA12]nSiz homologous series,2 REAu4A13Si,3 REzAuA16Sia4
and REAu3Al7.S The common feature of these compounds is the hexagonal antiﬂuorite-
type slabs regarded as fragments of the bulk AuAlz structure, which is absent ﬁom our
Ge-containing compounds REAuAl4Ge2 and REAuA14(AuxGet-x)2 (x = 0.4). To further
understand both the similarities and differences between Si and Ge systems, reactions
with the second row transition metals especially Ru, Rh and Ag should be looked at in
the future.

Our investigations of the ternary systems V/Al/Ge and Co/Al/Si led to the
discovery of three new phases: VzAlsGes, C019A145Si.0-,, (x = 0.13) and CosAlMSiz.

V2A15Ge5 was the ﬁrst known ternary member in the V/Al/Ge system. Its structure

234

 

 

features unusual pentagonal columnar building blocks with a one dimensional V-V chain
running along the central axis of the pentagonal columns. Magnetic susceptibility
measurements indicate Pauli paramagnetism for this compound, which is consistent with
the metallic behavior predicted by electronic structure calculations. C019A145Si 10., (x =
0.13) and C05A114Siz are two new silicides exhibiting highly complicated structures and
large unit cells. C05A1.4Siz shows remarkable oxidation resistance with only less than
3% weight gain up to 900 °C. These ternary systems can be extended to the quaternary
ones TMl/TMz/Al/SKGe). Although some alloys formed in these systems have been
used as advanced materials, there is a signiﬁcant lack of structural information about
these phases. Considering the fact that the metal ﬂux method is very powerful to grow
single crystals of intermetallics, explorations on the systems TM.fTM2/Al/Si(Ge) might
be helpful in understanding the composition, structure and properties of these alloys.
Given the fact that there are more than twenty transition metals available, a rich
chemistry is expected from the systems TMI/TMz/Al/SKGe). The systems TM/Al/Si(Ge)
and TMrfTMz/Al/SKGe) are of particular interest not only from a viewpoint of
discovering new materials but also desirable properties, especially oxidation resistance
properties.

Of all the rare earth-containing intermetallic compounds, Yb-based ones are
particularly interesting because they often exhibit unique properties associated with
mixed valency of Yb2*/Yb3*. ngNisAlto and YbNi3Alo have been characterized and
both show intermediate valence behaviors. Other rare earth analogues of Yb3Ni5A119
have also been synthesized, among which the Sm3Ni5Allo analogue shows interesting

magnetic properties. It has two different magnetic ordering events occurring at 5 K

235

(antiferro-) and 150 K (ferro-) with the applied magnetic ﬁeld perpendicular to the a-axis.
When the ﬁeld is parallel to the a-axis, these transitions are far less signiﬁcant and it
could be due to the fact that the spins are conﬁned to the bc-plane.

The discovery of YbGaGe, which showed zero thermal expansion (ZTE) between
10 K and 300 K, prompted us to carry out more investigations on the Yb-containing
compounds.6 Since YbGaGe and Yb3Ni5Allo show similar intermediate valence
behavior, Yb3Ni5A119 could be a good candidate to study the unusual zero thermal
expansion property. We successfully substituted Ni in Yb3Ni5A119 with Fe and other
transition metals and observed larger cell voltune of the substituted compounds.
Substitution of Ni by Fe or Mn did not signiﬁcantly change the thermal expansion
property of Yb3Ni5-xTMxA119 from 100 K to 298 K; while the thermal expansion
coefﬁcient of Yb3N15-xCuxAl|9 was much lower than that of the parent compound
Yb3N15AlIo, however, no clear trend was observed when the x values increased. On the
other hand we found that substitutions of Ni with other transition metals, especially Cu,
can modify the magnetic properties of the parent compound ngNisAllo. The substitution
of Ni by Cu causes cell volume expansion so this system is driven by negative chemical
pressure towards the Yb2+ regime while still in the YbZVYb3+ intermediate state. With
more Cu added into the structure, the effective magnetic moment 11ch for Yb3N15-xCuxAl.9
(x = 1) appears to be very close to that of the free Yb3+ ion (theoretical value pea = 7.86
113 per formula), which could be due to an increment of the number of conduction
electrons, and hence a rise of the Fermi level. Therefore the hybridization between the 4f
electrons and the conduction band is decreased, causing a quenching of the Kondo effect.

In order to explain the changes in the hybridization between 4f electrons and the

236

 

 

 

conduction band, both volume and Fermi level modiﬁcation have to be taken into
account.

The use of chemical pressure to tune Yb hybridization can be also applied to the
other Yb-containing compounds, such as the charge-balanced Zintl phase YboAl3Sbo in
which the Yb ions are in a divalent oxidation state. The partial substitution of Al ions by
larger Zn ions might cause changes of the 4f-conduction electron hybridization, which
can be tracked by magnetic susceptibility measurements.

When we tried to substitute Ni with Fe in Yb3Ni5A119, new quaternary
intermetallics RENi2-xFexAlg (RE = Eu and Yb) were isolated from liquid aluminum.
The RENi2-xFexAlg species are the ﬁrst quaternary analogues with the C3C02A13 structure
type.7 Both magnetic properties and Mossbauer spectra of YbNi2-xFexAlg show that the
Fe atoms do not carry magnetic moments. Temperature dependent magnetic
susceptibility measurements indicate that the Yb atoms are in an intermediate oxidation
state with effective magnetic moment peg = 2.19 113. The magnetization behavior studies
indicate that the ab plane is the easy plane on which the magnetic moments are conﬁned.
The discovery of these new phases might give us a direction to explore the systems
RE/TMl/TMz/Al using Al as a ﬂux. It has been discovered that in the presence of a
divalent rare earth metal or alkaline earth metal, Au reacts with an early transition metal
to form cubic phases M3Au6+xA126T (M = Ca, Sr, Eu, Yb; T = early transition metals from
groups 4-7).8 The crystal structure is a stuffed variant of the BaHgn type and the
transition metal site can be a host for a variety of early transition metals. Therefore it
would be interesting to examine the systems containing a rare earth metal, an early

transition metal, a late ﬁrst row transition metal in liquid aluminum. We would be

237

especially interested in Ce— and Yb-containing intermetallic phases with an eye on the

behaviors associated with 4f-conduction electron interactions.

238

References:

 

‘ Sieve, B. Ph. D. Dissertation, Michigan State University, 2002.

2 Latturner, S. E.; Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G. Chem. Mater. 2002, 14,
1695.

3 Latturner, s. E.; Kanatzidis, M. G. Chem. Commun. 2003, 18,2340.
4 Latturner, s. E.; Kanatzidis, M. G. Inorg. Chem. 2003, 42, 7959.

5 Latturner, s. E.; Bilc, D.; Ireland, J. R.; Kannewurf, C. R.; Mahanti, s. D. ; Kanatzidis,
M. G. J. Solid State Chem. 2003, 170,48.

6 Salvador, J. R.; Guo, F.; Hogan, T.; Kanatzidis, M. G. Nature 2003, 425, 702.

7 Czech, E.; Cordier, G.; Schaefer, H. J. Less-Common Met. 1983, 95, 205.

3 Latturner, s. E.; Kanatzidis, M. G. Inorg. Chem. 2004, 43, 2.

239

 

AR

IIIIIIIIIIIIIIIIII