\ "
q. -

 

 

 

 

 

 

   

 

w

 

 

’<"|ru,‘- -x\tq‘-1 ”nu...”

 

   

 

 

 

 

 

 

 

'v '
_ .A ~.‘ ‘

, f .. . '
. ‘ . '..\’ :

 

 

 

 

 

 

 

 

 

 

 

“11:15:13

3003

 

' 'LTB RARY
Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

EXPLORATORY SYNTHESIS OF QUATERNARY RARE EARTH
TRANSITION METAL ALUMINUM TETRELIDES:
UTILIZING MOLTEN AL AS A SOLVENT

presented by

Bradley Joseph Sieve

has been accepted towards fulfillment
of the requirements for

Ph. . . Ch 1
D degree m em stry

 

 

 

Major professor

Date ‘0' lo 2002

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJCIRC/DatoDuopss-p. 15

EXPLORATORY SYNTHESIS OF QUATERNARY RARE EARTH TRANSITION
METAL ALUNIINUM TETRELIDES: UTILIZING MOLTEN AL AS A SOLVENT

Volume I

By

Bradley J. Sieve

A DISSERTATION
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
DOCTOR OF PHILOSOPHY

Department of Chemistry

2002

IFT—i 'L 2‘12.“

ABSTRACT

EXPLORATORY SYNTHESIS OF QUATERNARY RARE EARTH TRANSITION
METAL ALUMINUM TETRELIDES: UTILIZING MOLTEN AL AS A SOLVENT

By

Bradley J. Sieve

Metallic fluxes have recently emerged as an important synthetic tool in solid state
chemistry for both crystal growth studies and exploratory studies of new chemical
systems. Our research group is currently employing Al metal fluxes to study quaternary
intermetallics in the RE/M/Al/T t systems (RE = rare earth, M = transition metal, Tt = Si
or Ge) in the hopes of discovering new structural relations and interesting physical
properties.

We have discovered that by utilizing Al fluxes a wide variety of products with
novel structures and compositions including REzNi(NixSi,_x)Al4Si6, REzNiA14Ge2,
RENiAl4Ge2, REHMZAIHS‘iy, REZ,,M2A14Tt2(All,yTty)(All_thZ)2, RE4Fe2+xAl7,xSi8,
REFe4A198i6 and REgRu12A1498i9(Al,Si12,x) can be produced. Along with the new structure
types, several isostructural analogues of known compounds such as RENiAl4(Si2_,Nix)
and RENiiAl,,,‘Ge,Ly have been synthesized for the first time.

In addition to exhibiting interesting structures, the compounds have displayed
several unusual physical properties such as valence fluctuations, metamagnetic phase
transitions, and orientation dependent magnetism. Of particular interest, however, is the

apparent filling of the transition metal (1 bands in the systems. This unexpected behavior

 

appears to be caused by electron transfer from the more electropositive RE and Al atoms
to the less electropositive transition metals.

This dissertation presents the synthesis, structural relationships, characterization
and properties of the newly discovered compounds. The synthetic work performed here

provides the groundwork for further exploration of these and related systems.

 

ACKNOWLEDGEMENTS

First and foremost I would like to thank my advisor, Professor Mercouri G.
Kanatzidis, for his support, guidance, and patience during the last five years. Without
these things I would not be at this point today. I have grown as both a scientist and a
person in the time I have been at Michigan State because of him.

I also would like to thank all the people I have had the pleasure to work with over
the years both within the Kanatzidis lab and outside of it. I especially want to thank Dr.
Chen who took the time to teach me many things when I was starting both in terms of
instrumentation as well as the use of metal fluxes. I want to thank Professor Kannewurf
for charge transport measurements, Dr. Trikalitis for the TEM work in the thesis, Dr.
Sportouch for her work examining the band structures, Dr. Loloee, Professor Mahanti
and the late Professor Cowen for their endless help with aspects of magnetism both large
and small and Dr. Schultz and Dr. Henning at Argonne National Lab for their work
studying compounds presented in this thesis using neutron diffraction. Also I would like
to thank Dr. Pinnavaia, Dr. Weliky and Dr. Geiger for serving as members of my
guidance committee. As you can see I have had the opportunity to work with many
outstanding people over the years for which I am very grateful. Funding for this work is
also gratefully acknowledged from Department of Energy.

Finally I would like to thank my family who has always encouraged me to strive
to be the best and my friends both old and new who have made my tenure here at

Michigan State an enjoyable one.

 

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
LIST OF ABBREVIATONS

Chapter 1

Chapter 2:

Chapter 3:

Introduction

1. Introduction

2. Flux Synthesis

3. Rationale for Al Flux Synthesis
4. A1 Matrix Alloys

5. Subject Matter of the Thesis

Synthesis and Characterization 0fRE2Al3Si2 (RE = Tb, Dy, H0, Er,
Tm) in the Y2A13Si2 Structure type
1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Method ll
Method 2
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Charge Transport Analysis
Magnetic Characterization
3. Results and Discussion
Synthesis and Thermal Behavior
Structural Description
Charge Transport Properties
Magnetic Properties
4. Conclusions

Synthesis and Characterization of the Phases RENiA14(Si2,xNix)
(RE = La, Ce, Pr, Nd, Eu), CeCoAl4Si2, RECuAl4(Si2,xCux)
(RE = Ce, La and Sm) and LanAl4(Si2_de,).
1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography

Page

xvii
xxiv

oouhwr—u—

14
15
15
15
15
15
15
16
l6
16
17
17
21
21
21
26
29
3O

39
4O
4O
4O
40
41
41
41

Charge Transport Analysis
Magnetic Characterization
3. Results and Discussion
Synthesis and Thermal Behavior
Structural Description
Charge Transport Properties
Magnetic Properties
4. Conclusions

Chapter 4. Routes to the Multinary Aluminum Tetrelides
REzNi(Ni,Si,_x)Al4Si6 (RE = Pr, Nd, Sm, Gd, Dy, Tb) and
Sm2C0( C 0,Al ,_X)Al4Ge6_).

1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Method 1
Method 2
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Magnetic Characterization
3. Results and Discussion
Synthesis and Thermal Behavior
Structural Description
Magnetic Properties
4. Conclusions

Chapter 5. Synthesis and Characterization of the Compounds REer'Al40e2
(RE = Gd, Tb, Dy, Er) and REzCoAl4Ge2 (RE = Sm, Gd, Th)
1. Introduction
2. Experimental Section

Synthesis
Reagents
Synthetic Method
Method 1
Method 2
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Magnetic Characterization
3. Results and Discussion
Synthesis
Structural Description

vi

42
55
55
55
56
57
62
62

66
67
67
67
67
67
67
68
68
68
69
84
84
84
91
91

95
96
96
96
96
96
97
97
97
98
98
107
107
107

 

Chapter 6.

Chapter 7.

Chapter 8.

Magnetic Properties
4. Conclusions

Synthesis, Structural and Property Characterization of the Phases,

REzNiAlé_xGe4_y (x~0.24, y~1.33) (RE: Ce, La, Pr, Nd, Sm)
1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Electron Diffraction Studies (TEM)
Magnetic Characterization
3. Results and Discussion
Synthesis
Structural Description
Magnetic Properties
4. Conclusions

Synthesis and Characterization of RENiAl4Ge2 (RE = Sm, Gd, Tb,
Dy, Ho, Er, Tm, Lu, Y) Containing a Trigonal Lattice of RE ions.
1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Magnetic Characterization
3. Results and Discussion
Synthesis
Structural Description
Magnetic Properties
4. Conclusions

Synthesis and Characterization of the Ni Rich Phases
RE,_,MZAI5.,.Siy (RE: Nd, Sm, Tb, Tm, Yb, Y; M: Ni and Pd) and
RE2_,MzAl4Tt2(Al,_).Tt),)(Al,_thz)2(RE= Sm, Er, Dy; M: Ni, Co;
Tt= Si, Ge)

1. Introduction

2. Experimental Section

vii

112
112

116
117
117
117
117
117
117
118
125
125
126
126
126
133
134

138
139
139
139
139
139
139
140
141
150
150
150
155
163

169
170

Chapter 9.

Chapter 10.

Synthesis
Reagents
Synthetic Method
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Charge Transport Analysis
Magnetic Characterization
3. Results and Discussion
Synthesis
Structural Description
Charge Transport Properties
Magnetic Properties
4. Conclusions

Synthesis and Characterization of the Quaternary Aluminum
Silicides RE4Fe2+,Al7_,Si8 (RE = Ce, Pr, Nd, Sm) and
RE4Mn2,,Al7_,Si8 (RE = Ce, Pr, Nd, Gd).
1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Method 1
Method 2
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Electronic Structure Calculations
Charge Transport Analysis
Magnetic Characterization
Mossbauer Spectroscopy
3. Results and Discussion
Synthesis
Structural Description
Electronic Structure
Charge Transport Properties
Magnetic Properties
Mossbauer Spectroscopy
4. Conclusions

Synthesis, Structure and Physical Characterization of REF e4AlgS i,5

(RE=Tb, Er, Gd, Dy, Ho)
1. Introduction

viii

170
170
170
171
171
171
172
172
189
189
191
194
203
207

210
211
211
211
212
212
212
213
213
213
214
228
230
230
231
231
232
236
243
246
248
254

259

2. Experimental Section
Synthesis
Reagents
Synthetic Method
Method 1
Method 2
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Magnetic Characterization
Mossbauer Spectroscopy
3. Results and Discussion
Synthesis
Structural Description
Magnetic Properties
Mossbauer Spectroscopy
4. Conclusions

Chapter 11. Cubic Quaternary Aluminum Silicides RE gRu ”Al 495i9(Al,Si ,2“)
(x~4) (RE = Pr, Nd, Sm, Gd, Tb, or Er) from Molten Aluminum.
Empty (Al,Si),2 Cuboctahedral Clusters and the Assignment of the

Al/Si Distribution with Neutron Difiraction
1. Introduction
2. Experimental Section
Synthesis
Reagents
Synthetic Method
Method 1
Method 2
Method 3
Physical Measurements
EDS Analysis
Single Crystal X-ray Crystallography
Charge Transport Analysis
Magnetic Characterization
3. Results and Discussion
Synthesis
Structural Description
Charge Transport Properties
Magnetic Properties
4. Conclusions

Chapter 12. Conclusions and Future Work

260
260
260
260
260
261
261
261
261
273
273
274
274
275
276
282
285

288
289
289
289
289
289
290
290
291
291
291
303
303
304
304
306
314
314
323

329

Table 2-1.

Table 2-2.

Table 2-3.

Table 2-4.
Table 2-5.

Table 3-1.

Table 3-2.

Table 3-3.

Table 3-4.

Table 3-5.

Table 3-6.

Table 38.

Table 3-9.

Table 4-1.

LIST OF TABLES

Crystal data and structure refinement for RE2A13Si2
(RE = Tb, Dy).

Atomic coordinates ( x10“) for RE;.Al;,Si2 (RE = Tb, Dy).

Anisotropic displacement parameters (A2 x 103) for RE2A13Si2
(RE = Tb, Dy).

Selected Bond Distances (A) in Tb2A13Si2
Magnetic properties of RE2A13Si2 (RE: Tb, Dy, Ho, Er, Tm)

Crystal data and structure refinement for RENiAl4(Si2_,Ni,)
(RE = La, Ce).

Crystal data and structure refinement for RENiAl4(Si2,xNix)
(RE = Pr, Nd).

Crystal data and structure refinement for EuNiAl4Si2 and
CeCoAl4Si2.

Crystal data and structure refinement for RECuAl4(Si2_,,Cux)
(RE = Ce, La).

Crystal data and structure refinement for SmCu1+xAl4Si2,, and
Lan1+xAl4Si2_,.

Atomic coordinates ( x10“) and occupancies for RENiAl4(Si2,,Ni,)

(RE: La, Pr, Nd, Eu. Ce), CeCoAl,,Si2, RECuAl4(Si2-,Cu,)
(RE: Ce, La, Sm) and LanAl4(Si2-,Pd,).

Anisotropic displacement parameters (A2 x 103) for RENiA]4(Si2_
,Ni,) (RE: La, Pr, Nd, Eu. Ce), CeCoAl4Si2, RECuAl,(Si2,,Cu,)

(RE: Ce, La, Sm) and LanA14(Si2,,de).
Selected bond distances [A] for LaNiAl4(Si2,,Ni,).

Crystal data and structure refinement for REzNi(Ni,Sil,,,)Al,,Si6
(RE = Sm,Pr).

Page

18

19

19
20

37

43

45

46

47

48

51

54

70

 

Table 4-2.

Table 4-3.

Table 4-4.

Table 4-5.

Table 4-6.

Table 4-7.

Table 4-8.

Table 4-9.

Table 4-10.

Table 5- 1.

Table 5-2.

Table 5-3.

Table 5-4.

Table 5-5.

Table 5-6.

Crystal data and structure refinement for REzNi(Ni,,Si,,x)Al,,Si6
(RE = Nd, Gd).

Crystal data and structure refinement for REzNi(Ni,Si,_,)Al4Si6
(RE = Tb, Dy).

Atomic coordinates ( x10“) and occupancies for
REzNi(NixSi,_x)Al4Si6(RE= Pr, Sm, Nd, Gd, Tb, Dy).

Anisotropic displacement parameters (A2 x 103) for
REZNi(Ni,,Si,_,,)Al4Si6 (RE: Pr, Sm, Nd, Gd, Tb, Dy).

Selected bond distances (A) for REgNi(NixSi1-x)Al4Si6
(RE: Pr, Sm, Nd, Gd, Tb, Dy)

Crystal data and structure refinement for szCo(Co,Al l,,,)A14Ge6_y.

Atomic coordinates ( x104) and occupancies for
Sm2C0(CoxAl,_,)Al4Ge6,y.

Anisotropic displacement parameters (A2 x 103) for
Sm2C0(Co,All_,)Al4Gegy.

Selected bond distances (A) for szCo(CoxAll,x)A14Ge6,y,

Crystal data and structure refinement for REzNiAl4Ge2
(RE = Tb, Gd).

Crystal data and structure refinement for REzNiAhGe2
(RE = Dy, Er).

Atomic coordinates ( x10“) for REZNiAhGe2
(RE: Tb, Gd, Dy, Er).

Anisotropic displacement parameters (A2 x 103) for REzNiA14Ge2
(RE: Tb, Gd, Dy, Er).

Crystal data and structure refinement for RE2C0A14Ge2
(RE = Sm, Gd).

Atomic coordinates ( x104) for RE2C0A14Ge2 (RE: Sm, Gd).

xi

71

72

73

76

79

82

82

83

98

99

100

101

102

103

Table 5-7.

Table 5-8.

Table 6- 1.

Table 6-2.

Table 6-3.

Table 6-4.

Table 6-5.

Table 6-6.

Table 7-1.

Table 7-2.

Table 7-3.

Table 7-4.

Table 7-5.

Table 7-6.

Table 7-7.

Table 7-6.

Anisotropic displacement parameters (A2 x 103) for REzCoAIJGe2

(RE: Sm, Gd).
Selected bond lengths [A] for szNiAhGez.

Crystal data and structure refinement for RENiAl£,_,,Ge,,.y
(RE = Ce, La).

Crystal data and structure refinement for RENiAlngeH
(RE = Pr, Nd).

Crystal data and structure refinement for SmNiAlMGeH.

Atomic coordinates ( x10“) and occupancies for REZNiAhMGe,y
(RE: Ce, La, Pr, Nd, Sm).

Anisotropic displacement parameters (A2 x 103) for
Riazrlnstlgpe,y (RE: Ce, La, Pr, Nd, Sm).

Selected bond lengths [A] for CezNiAlngeH.

Crystal data and structure refinement for RENiAl,,Ge2
(RE = Sm, Gd).

Crystal data and structure refinement for RENiAhGe2
(RE = Tb, Dy).

Crystal data and structure refinement for RENiAl4Ge2
(RE = Ho, Er).

Crystal data and structure refinement for RENiA14Gez(
RE = Tm, Lu).

Crystal data and structure refinement for YNiAl4Ge2.

Atomic coordinates ( x104) for RENiAl4Ge2
(RE: Sm, Gd, Dy, Ho, Tm, Lu, Y).

Anisotropic displacement parameters (A2 x 103) for RENiAl4Ge2
(RE: Sm, Gd, Dy, Ho, Tm, Lu, Y).

Selected bond distances for SmNiAl4Ge2

xii

104

105

118

119

120

121

122

123

141

142

143

144

145

146

147

148

Table 7-7.

Table 8-1.

Table 8-2.

Table 8-3.

. Table 8-4.
Table 8-5.

Table 8-6.

Table 8.7

Table 8-8.

Table 8-9.

Table 8—10.

Table 8-1 1.

Table 8.12

Table 8.13

Table 8- 14

Table 8-15.

Table 9-1.

Temperature dependent magnetic behavior of RENiA14Ge2
(RE = Sm, Gd, Tb, Tm)

Crystal data and structure refinement for RE,_,,M2A15_ySiy
(RE = Nd, Sm).

Crystal data and structure refinement for RE,_,,M2A15_ySiy
(RE = Tb, Tm).

Crystal data and structure refinement for REHMZAlgySiy
(RE = Yb, Y).

Crystal data and structure refinement for Y1_,Pd2A15_ySiy.
Atomic coordinates ( x104) for RE,_,M2A15_ySiy.

Anisotropic displacement parameters (A2 x 103) for
RE,,,M,A1,_,Si,.

Selected Bond Distances (A) for N d,_,Ni2A15_ySi/y.

Crystal data and structure refinement for
REZ-XM2A14Tt2(Al l -yT[y)(Al l-thJZ'

Crystal data and structure refinement for
RE2,,M2A14Tt2(All_yTty)(All_thl)2.

Atomic coordinates ( x104) for RE2_,M2A14Tt2(Al,,,Gey)(Al1,2Gez)2.

Anisotropic displacement parameters (A2 x 103) for
RE,-,M,A1,Tt,(A1,.,Ge,)(Al,_,Ge,),_.

Atomic coordinates ( x10“) for Dy2_,Ni2Al4Si2(Al1,ySiy)(Al,_,Siz)2.

Anisotropic displacement parameters (A2 x 103) for
Dy2,,Ni2Al4Si2(Al1_ySi,)(Al,_zSiz)2.

Selected bond distances (A) for Sm2_,CozAl4Ge2(A11_yGey)(A11_
zGe,)2 and Dy2,,N i2A14Ge2(Al,_yGey)(Al,_zGez)2.

Selected bond distances (A) for Dy,.,Ni._A1,Si,(Al,,,Si,)(Aluse),

Crystal data and structure refinement for RE4Fe._,+,,Al7_,,Si8
(RE = Sm, Ce).

xiii

164

173

174

175

176

177

179

180

181

182

183

185

186

186

187

188

215

Table 9-2.

Table 9-3.

Table 9-4.

Table 9-5.

Table 9-6.

Table 9-7.

Table 9-8.

Table 9-9.

Table 9-10.
Table 10-1.
Table 10-2.

Table 10-3.

Table 10-4.
Table 10-5.

Table 10-6.

Crystal data and structure refinement for RE4Fe2,,,A17_xSi8
(RE = Pr, Nd).

Crystal data and structure refinement for RE,,Mn2+,,Al7_,,Si8

(RE = Ce, Pr).

Crystal data and structure refinement for RE4Mn2,,,Al7,,,Si8
(RE = Nd, Gd).

Atomic coordinates ( x10“) and occupancies for RE4MMA1MSi8
(RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

Anisotropic displacement parameters (A2 x 103) for
RE,,M2,,,A17_,,Si8 (RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

Selected Bond Lengths (A) for Sm4Fe2,,A17_,Si8.
Exponents and parameters used in the extended Hijckel
calculations. The contraction coefficients used in the double-

C expansion are c1 and c2.

Calculated electron densities for the Y,,Fe2Al7Si8 compound
as a function of the electron count.

Mbssbauer Parameters, collected at 20 K, of RE,,Fe2,,,Al7.,Si8
(RE: Ce, Pr).

Crystal data and structure refinement for REFe4A198i6
(RE = Tb, Er).

Crystal data and structure refinement for REFe,,A19Si6
(RE = Gd, Dy).

Crystal data and structure refinement for HoFe4A198i6.

Atomic coordinates ( x 10“) for REFe,,Al,,Si6
(RE=Tb, Er, Gd, Dy, HO).

Anisotropic displacement parameters (A2 x 103)“ for REFe,,Al9Si6
(RE=Tb, Er, Gd, Dy, Ho).

Selected Bond Distances ( A) for TbFe4A19Si6.

xiv

216

217

218

219

222

227

255

263

264

265

266

269

272

'1!

Table 10-7.

Table 11-1.

Table 1 1-2.

Table 11-3.

Table 1 1-4.

Table 1 1-5.

Table 1 1-6.

Table 1 1-7.

Mossbauer Parameters, Collected at 18 K, of REFe,,Al.,Si6
(RE = Tb, Er).

Crystal data and Structure refinement for REgRu12A149Si9(Al,Si12-x)
(RE = Sm,Pr).

Crystal data and structure refinement for REBRu12A149Si9(A1,Sim)
(RE = Nd, Gd).

Crystal data and structure refinement for REsRu12A149Si9(A1,Sim)
(RE = Br, Tb).

Atomic coordinates ( x10“) for REgRu12A1498i9(Alein-x)
(RE: Sm, Pr, Nd, Gd, Er, Tb).

Anisotropic displacement parameters (A2 x 103) for
RE,Ru,,A1,,Si,(A1,Si,,,,) (RE: Sm, Pr, Nd, Gd, Er, Tb).

Selected bond distances for SmBRule149Sig(Al,Si 12_,,)

Temperature dependent magnetic behavior of
REgRu12A1498i9(Al,Si,2_,) (RE: Pr, Nd, Sm, Gd, Tb, or Er)

XV

286

293

294

Figure 2-1.

Figure 2-2.

Figure 2-3.

Figure 2-4.

Figure 2-5.

Figure 2-6.

Figure 2-7.

Figure 2-8.

Figure 2-9.

Figure 2-10

Figure 2-1 1.

Figure 3-1.

Figure 3-2.

LIST OF FIGURES
Images in this dissertation are presented in color

Structure of REQA13Si2 viewed down the b axis.

AIIZSi2 layer which runs parallel the b axis in RE2A13Si2.
Coordination environments in RE2A13Si2.

Electrical resistivity and thermopower measurements for single
crystals of Y2A13Si2. Two separate single crystals were measured

for resistivity analysis.

Electrical resistivity and thermopower measurements of single

crystals of Dy2A13Si2. Two separate single crystals were measured

in each analysis.

Magnetic behavior of Tb2A13Si2 (top) susceptibility as a function
of temperature (middle) inverse susceptibility as a function of
temperature and (bottom) magnetization as a function of field.

Magnetic behavior of Dy2A13Si2 (top) susceptibility as a function
of temperature (middle) inverse susceptibility as a function of
temperature and (bottom) magnetization as a function of field.

Magnetic behavior of HozAl3Si2 (top) susceptibility as a function
of temperature (middle) inverse susceptibility as a function of
temperature and (bottom) magnetization as a function of field.

Magnetic behavior of Er2A13Si2 (top) susceptibility as a function
of temperature (middle) inverse susceptibility as a function of
temperature and (bottom) magnetization as a function of field.
Magnetic behavior of TmzAl3Si2 (top) susceptibility as a function
of temperature (middle) inverse susceptibility as a function of

temperature and (bottom) magnetization as a function of field.

Single crystal magnetic response of H02A13Si2 with the field
aligned both parallel and perpendicular the b axis.

Structure of REMA14(Si2_,M,) viewed down the b axis.

The MA14(Si2_,M,) layer viewed down the c axis.

xvi

Page

23

24

25

27

28

31

32

33

34

35

36
58

59

Figure 3-3.

, Figure 3-4.

Figure 3-5.

Figure 4-1

Figure 4-2.

Figure 4-3.

Figure 4-4.

Figure 4-5.

Figure 5-1.

Figure 5-2.

Figure 5-3.

Figure 5-4.

Coordination environments of each atomic position in
REMA14(Si2,,M,,).

Conductivity (top) and thermopower (bottom) behavior of
CeNiAl4(Si2,,Ni,).

Magnetic behavior of CeNiAl4(Si2_xNix)

The structure of RE2M(Mth1_x)Al4Tt6. Note the MA]8 cube
layers at the top and bottom of the cell. In the middle there is
a layer formed by Tt(4) atoms and the disorder position. The
two different layers are connected through bonds between
Tt(2) and the disorder position. No bonds are drawn around
RE atoms.

The MAl8 cube layer in ab plane of RE2M(Mth1-x)A]4Tt6.
Note that each MAl8 cube only connects four (not six) other
MAI8 cubes by sharing the four paralleled edges.

(a) The double "Tt"-net and (b) Half of the net which is obtained
by cutting the one in (a) at the middle in ab plane.

The local coordination environments for different atoms in
szNi(NixSi1-x)Al4Si6.

Magnetic behavior of GdzNi(Ni,Si1,,)Al,,Si6 (top) susceptibility
as a function of temperature (middle) inverse susceptibility as a
function of temperature and (bottom) magnetization as a function
of field.

Structure of REzMAl4Ge2 viewed down the b axis.

Views of the MAl4Ge2 layer viewed down (a) the b axis and
(b) the c axis.

Local coordination environments of individual atomic positions
in REzMAl4Ge2.

Magnetic behavior of szNiAl,,Ge2 (top) susceptibility as a
function of temperature (middle) inverse susceptibility as a
function of temperature and (bottom) magnetization as a function
of field.

xvii

60

61

64

86

87

88

89

92

108

109

110

113

Figure 6-1.

Figure 6-2.

Figure 6-3.

Figure 6-4.

Figure 6-5.

Figure 6-6.

Figure 7-1.

Figure 7-2.

Figure 7-3.

Figure 7-4.

Figure 7-5.

Structure of REzNiAl(,,,,Ge,,_y (x~0.24, y~1.34) viewed down the

b axis.

Views of the NiAl,,Ge2 layer viewed down (a) the b axis and

(b) the c axis.

Views of the heavily vacant A12_,,Ge2_y (x~0.24, y~1.34) layer
viewed down (a) the b axis and (b) the c axis. The disordered

Al atoms are shown.

Coordination environments in REZNiAlMGeer

(a) Selected area diffraction pattern of CezNiAlngegy, viewed
down the c axis (hk0 zone), showing the 3ax3b supercell and
(b) one-dimensional peak intensity profile from the boxed area
indicated in the pattern clearly showing the weak ( 5 60) and

(Z 60) commensurate supercell reflections.

Magnetic behavior of CezNiAhHGe,Ly (top) susceptibility as a
function of temperature (middle) inverse susceptibility as a

function of temperature and (bottom) magnetization as a function

of field.

The structure of RENiAl4Ge2 viewed down the a-axis.

The highlighted layer in RENiAl,,Ge2 showing Ni stuffed As-type

layer in the center and the umbrella-like arrangement of the Ge

atoms at the outer edges.

Immediate coordination environments of the (a) Ge atoms
exhibiting an umbrella-like and (b) Ni atoms exhibiting bonding
to the six Al atoms in the As-type layer along with 2 more A]

atoms situated above and below the layer.

Magnetic behavior of SmNiAl,,Ge2 (top) susceptibility as a
function of temperature (middle) inverse susceptibility as a

function of temperature and (bottom) magnetization as a function

of field.

Magnetic behavior of TbNiAl4Ge2 (top) susceptibility as a
function of temperature (middle) inverse susceptibility as a

function of temperature and (bottom) magnetization as a function

of field.

xviii

128

129

130

131

134

135

151

152

153

155

157

Figure 7-6.

Figure 7-7.

Figure 7-8.

Figure 8-1.

Figure 8-2.

Figure 8-3.

Figure 8-4.

Figure 8-5.

Figure 8-6.
Figure 8-7.
Figure 8-8.

Figure 8-9.

Magnetic behavior of TmNiAl,,Ge2 (top) susceptibility as a

function of temperature (middle) inverse susceptibility as a

function of temperature and (bottom) magnetization as a function

of field. 159

Magnetic behavior of GdNiAl,,Ge2 (top) susceptibility as a

function of temperature (middle) inverse susceptibility as a

function of temperature and (bottom) magnetization as a function

of field. 161

Real portion of the magnetic susceptibility of GdNiAl4Ge2

single crystals measured in an AC field (top). Field dependence
response (bottom) of single crystal of GdNiAl4Ge2 orientated

parallel and perpendicular the c axis. 165

SEM images of (a) Y,_,,Ni2A15_ySiy (b) Dy2_,Ni2Al4Si2(Al,_ySiy)
(AlHSiz)2 and (c) Tb1_,,Ni2Als,ySiy exhibiting typical hexagonal rod-
like crystal morphology. 190

Crystal structure of RE,_,M2A15-ySiy. 195

Building units of RE1,,,M2A15,ySiy including (a) stuffed As-type
NiAl layer (b) full disorder layer with all disorder atoms shown

and (c) disorder layer displaying the ordering of atoms to produce
the a\/3 x a‘i3 cell. 196

RE,,,,A1 layer of RE1,,,M2A15,ySiy (a) layer with all disorder atoms
shown and (b) model layer displaying the ordering of atoms to
produce the a\/3 x a\/3 cell. 197

X-ray zone photos of Tbl_,Ni2Als,ySi in ab plane (top) and ac
plane (bottom) showing the aV3 x ai3 supercell in the ab plane

and diffuse scattering along the c* axis. 198
Coordination environments of RE,,,,M2A15,ySiy 199
Crystal structure of RE2.,,M2A14Tt2(Al,_yTty)(Al,_,TtZ)2 200

Coordination environments in RFQ,,M2A14Tt2(Al,,yTty)(AlHT t1)2 201

Charge Transport Properties of two selected single crystals of
Y,_,,NiA15_ySiy 202

xix

Figure 8-10.

Figure 8-11.

Figure 9-1.
Figure 9-2.

Figure 9-3.

Figure 9-4.

Figure 9-5

Figure 9-6.

Figure 9-7.

Figure 9-8.

Figure 9-9.

Magnetic behavior of Ybl_,,NiA15,ySiy single crystals. Data
shown with squares corresponds to crystal alignment with the
field parallel the c axis. Circles correspond to the field aligned

perpendicular the c axis 205
Magnetic behavior of Dy2.,,Ni2Al,,Si2(Al,,ySiy)(A1,.zSiz)2 single

crystals. Data shown with squares corresponds to crystal alignment
with the field parallel the c axis. Circles correspond to the field

aligned perpendicular the c axis 206
SEM image of crystals of Sm,,Fe2,,,Al7_,,Si8 exhibiting typical
needle-like morphology 234
Structure of RE,,Fe;,_,,,Al7,,Si8 (RE: Ce, Nd, Pr, Sm) and
RE,,Mn2,,,Al7,,,Si8 (RE: Ce, Pr, Nd, Gd) viewed down the c axis 237
Coordination environments in RE,,Fe2,,,Al7,,,Si8 and

RE,,Mn2,,,Al7,,,Si8 of the individual atpms out to 3 A, except for

the RE atom which is shown to 3.5 A 238
Structure of the repeating layer along the c axis. Only the layer

of the Fe analog's is shown for simplicity 239
Total DOS for Y,,Fe2Al7Si8 and contributions to the DOS

of Fe 3d orbitals and Si 3p orbitals. The Fermi level is drawn
by a dashed line for 162 valence electrons. (taken from
reference 1) 240

Resistivity measurements of Ce4Fe2,,Al7_,Si8, Pr4Fe3,,,Al7_,Si8
and Sm4F +,,Al7,,,Si8 244

Thermopower Measurements of (a) RE4Fe2,,Al7,,,Si8
(RE = Pr and Sm) and (b) Ce,,Fe2,,,Al7_,,Si8 showing small
absolute values indicative of metallic compounds 24S

Inverse magnetic susceptibility data for (a) Ce4Fe2,,Al7_,Si3
and (b) Pr,,Fe2,,Al7_,,Si8 standardized to one RE per formula
unit 249

Inverse magnetic susceptibility data for (a) Nd,,Fe2,,,Al,,,,Si8
and (b) Sm4Fe2,,,Al7_,,Si8 standardized to one RE per formula
unit 250

XX

Figure 9-10.

Figure 9-11.

Figure 9-12.

Figure 10-1.

Figure 10-2.

Figure 103.

Figure 10-4.

Figure 10-5.

Figure 10-6.

Figure 10-7.

Figure 10-8.

Magnetic behavior of Ce,,Mn2,,,Al7_,,Si8 (top) susceptibility

as a function of temperature (middle) inverse susceptibility as a
function of temperature and (bottom) magnetization as a function
of field standardized to one RE per formula unit

Magnetic behavior of Nd4Mn2,,Al7,,Si3 (top) susceptibility as

a function of temperature (middle) inverse susceptibility as a
function of temperature and (bottom) magnetization as a function
of field standardized to one RE per formula unit

Typical Mossbauer spectra of Pr,,Fe2,,,Al7,,Si8 and Ce,,Fe2,,,Al7_,,Si8
taken at 20 K.

Structure of REFe,,A198i6 (RE = Tb, Er) viewed down the c axis
to Show the tetragonal symmetry of the structure

The 2:0 layers in (a) NdRh4Alm showing a highly disordered
Al atom arrangement and in (b) REFe,,Al.,Si6 showing an ordered
Al and Si layer

Structure of REFe,,A19Si6 (RE = Tb, Er) shown being built in
stages as each elemental type is added to the structure. First
showing the (a) corrugated Al layers (b) then addition of the
Fe atoms and the (c) Si atoms to the structure and finally the
(d) RE atoms are added. The unit cell outline is represented
by the thick black lines

Local coordination environments of the individual non rare
earth atomic positions in REFe,,Al,,Si6 out to 3 A. Bonds are
drawn for interatomic distances less than 2.4 A for Fe-Si and
Si-Si, 2.7 A for Fe-Al and Si-Al, and 2.9 A for Al-Al

Local coordination environment of rare earth atomic position
in REFe4A19$i6 drawn out to 3.5 A

Temperature dependent magnetic susceptibility for (a) TbFe4A193i6
and (b)ErFe4A198i6. Insets Show the inverse susceptibility for
each compound

Field dependent magnetic susceptibility for (a) TbFe4Al.,-,Si6
and (b)ErFe4A198i6.

Typical Mbssbauer spectra of ErFe,,A19Si6 taken at 18 K.

Velocity corresponds to the isomer shift (18) which is
referenced to (Jr-Fe.-

xxi

 

251

252

255

277

278

279

280

281

283

284

286

 

Figure 11-1.

Figure 11-2.

Figure 11-3.

Figure 11-4.

Figure 11-5.

Figure 11-6.

Figure 11-7.

Figure 1 1-8.

Figure 11-9.

Figure 11-10.

Figure 11-11.

Figure 11-12.

SEM image of crystals of PrgRu12A149Sig(Al,Si12-x) exhibiting
typical morphology.

Structure of REgRU 12Al498i9(Al,Si12,,) (one units cell shown)

Structural units of REgRu12A149Si9(Al,Si12,,) (A) AlSi6 unit at
body center position, (B) Ru2A14Si unit at center of cell faces
(two units of Ru2A14Si removed for clarity), (C) SiAl8 units

at center of cell edges, and (D) the M12 unit at the cell comers.

View of the AlSi6 octahedron at the cell center sharing Si atoms
with the Ru2A14Si clusters at the face centers.

(A) The SiAl8 cube and the Ru2A14Si clusters bonding through
Al-Al and Al-Ru bonds (B) and a SiAl8 cube linked to a M12
cluster in a through Al-Al bonds.

Coordination environments of the various atoms in the unit cell.

(A) Thermopower values for Sngu12A1498i9(Al,Sim). Values
Shown have been corrected for the contribution of the Au
electrodes used to make contacts to the samples.(B) Conductivity
values of SngunAl498i9(Al,Si12.x). Values shown have been
corrected for the contribution of the Au electrodes used to make
contacts to the samples.

(a) Magnetic susceptibility vs. temperature of
Sngu12A1498i9(AlmSix) (b) inverse magnetic susceptibility
and (0) response as a function of field.

(a) Magnetic susceptibility vs. temperature of
PrgRu12A1498i9(Aln,xSix) (b) inverse magnetic susceptibility
and (0) response as a function of field.

(a) Magnetic susceptibility vs. temperature of
ngRu12A1498i9(AlmSix) (b) inverse magnetic susceptibility
and (c) response as a function of field.

(a) Magnetic susceptibility vs. temperature of
Gngu12A1498i9(Al12,,Six) (b) inverse magnetic susceptibility
and (c) response as a function of field.

(a) Magnetic susceptibility vs. temperature of

ErgRunAl498i9(AlmSix) (b) inverse magnetic susceptibility
and (0) response as a function of field.

xxii

305

309

310

311

312

313

315

317

318

320

321

322

Figure 11-13. (a) Magnetic susceptibility vs. temperature of
TbgRu12A1498i9(Al 12_,,Si,,) (b) inverse magnetic susceptibility
and (c) response as a function of field 324

xxiii

CCD
EDS
SEM
SQUID

TEM

LIST OF ABBREVIATIONS

Charge Couple Device

Energy Dispersive Spectroscopy

Scanning Electron Microscope
Superconducting Quantum Interference Device

Transmission Electron Microsc0pe

xxiv

 

Chapter 1
Introduction

1. Introduction.

The effects of solid state chemistry are everywhere around us in the technology
and materials we use each day. Solid state research is responsible for the discovery of
new materials daily replacing traditional ones as greater demands are made for
applications. These new materials or the adaptation of known phases often decides the
fate of emerging technologies such as energy conversion, superconductivity, information
storage and structural alloys.I Once one realizes society's dependence on these materials
the importance of solid state chemistry and ongoing research cannot be questioned.
One interesting class of compounds is the intermetallic tetrelides (i.e. Si and Ge).
These materials are studied for many types of applications utilizing inherent properties
Such as chemical stability,2 high melting points,3 and superconductivity.4 Several types of
applications based on these properties have been developed such as advanced structural
materials,’ thermoelectric energy conversion6 and high temperature coatings.3 The
Promise of such materials has thus prompted a resurgence in the studies of intermetallic
tetrelide compounds not simply a review of previously known binary and ternary
Compounds, but also the synthesis and characterization of new higher order phases which
Can be used in ever more demanding applications.
Traditional synthesis of intermetallic tetrelide phases involves direct combination

of reactants at elevated temperatures, normally well over 1500° C. This requires the use

irvfi

 

of high-energy synthetic methods such as induction furnaces, arc-melting and tube

furnaces and often prohibits any amount of predictability in the reactions. The extreme
synthetic conditions are required to facilitate the transport of reactants through diffusion
barriers present in solid state systems. Often, however, high temperatures alone are not
enough to overcome these barriers and the samples need to be ground to powders several
times during the synthesis to expose fresh surfaces on which reactions can occur.

Along with the costs associated with required equipment several other drawbacks
have plagued researchers over the years. First the reactions tend to favor thermodynamic
phases which are often simple binary compounds hindering one's ability to study more
complex phases that are not thermodynamically favored or are formed through kinetic
pathways. These thermodynamic phases due to their high lattice stabilities can often act
as synthetic traps for systems preventing the synthesis of other more interesting phases.
Another major hindrance in the method is that normally powder products are received.
Both the rapid cooling of reactants from high temperatures along with the repeated
powdering of the samples, as mentioned above, do not create a favorable environment for
Crystal growth. Single crystals can sometimes form through extended annealing of the
Powders though even then the growth of crystals large enough for analysis is not always
Seen. This microcrystalline product can than limit the proper characterization of the new

material both structurally and physically, particularly when knowledge about anisotropic
effects is desired, and may even prevent the proper identification of .the reaction products
in extreme cases.

To avoid the problems associated with high temperature synthesis several

a] tEtl‘native synthetic methods have been explored such as electrochemical synthesis,7

 

chemical vapor deposition,8 and molten fluxes.9 These methods do not require extensive

temperatures but each have hindrances of their own

2. Flux Synthesis.

After reviewing the alternative methods it was found that the use of low
temperature fluxes presented an attractive alternative to traditional synthetic methods for
the synthesis of intermetallic tetrelides. First the molten state of the flux removes the
solid state diffusion barriers and allows reactants to diffuse over long distances while still
maintaining relatively low temperatures. This approach also allows for crystal growth
during initial synthesis due again to the solvent nature of the reaction") This allows for

even mixtures of products to be easily separated and identified where before powders
often limited the useful information that could be obtained. This growth also means that
relatively pure synthesis is not required to yield useful and informative products.
Conceptually this represents a significant deviation from traditional methods in not
rEEquiring a target composition or compound when planning reactions instead allowing the
SYstem to form the most stable phases for the conditions in easily selectable crystals for
analysis.

Several types of low temperature flux systems (<1500° C) have been studied in
the past including the use of molten salts,“ elemental metals such as Pb12 and Bi,13 and
low melting intermetallic phases.” The selection of solvent (i.e. flux) depends on several

factors including solubility, reactivity, means of product isolation, and the availability of

r e<Illired equipment for the particular system.7 In choosing a solvent though several of the

factors may not be known requiring several rounds of trial and error before an appropriate
solvent is found.

Traditionally flux research has concentrated on one of two end goals. The first
being the recrystallization of known phases for property or structural analysis while the
second is the use of fluxes in a true exploratory manner. The recrystallization approach is
quite old in the literature though recently many groups such as Canfieldlo and Takei '5
have made great strides using various fluxes growing large single crystals of known
phases to study properties. Historically the second approach using molten fluxes for
exploratory work has been limited.16 Though as the benefits become more apparent

several groups such as Jeitschko,‘7 zur Loye,18 and Kanatzidis8 have had great success

forming many new and interesting phases utilizing exploratory flux synthesis methods.

3. Rationale for Al Flux Synthesis.

For the synthesis of intermetallic tetrelide compounds the option of molten metal
fluxes presents a promising approach particularly with the use of low melting metals of
the B family (Al, Ga, and In). These metals have low melting points, large solubility
lil‘nits for tetrelides at moderate temperatures and do not form binary compounds with
tetrelides,19 and are easily isolable from the final product through either chemical or

Physical means.
Using this rationale we selected A] as a possible solvent for synthesis of complex
inte-l‘metallics compounds containing Si and Ge. Many different reasons exist which
W0llld make Al an excellent solvent in exploratory synthesis. The melting point of Al is

Well below traditional synthetic temperatures, 660° C, hopefully allowing pathways to

both kinetic and thermodynamic phases. Also upon inspecting the binary phase diagrams

one can note that Al will dissolve large amounts of both Si and Ge at relatively (<1000°
C) lbw temperatures.16 It is also seen that these systems do not form binary phases that
could possibly scavenge the tetralide from the system removing it from a reactive state.
Once the tetrelide is dissolved into the system other elements can be introduced, even if
only present in small amounts, to hopefully react and form new phases. Examining other
binary Al phase diagrams it is seen that most elements dissolve into molten Alto some
degree below 1000° C. This allows for the study of a wide range of elements using Al as
the solvent. Finally after the reaction is completed the products can be isolated from the
excess Al in several ways depending on the stability of the product. These methods
include dissolution of Al in acids and bases or the removal of the flux through physical
methods such as filtration at high temperatures.

Another interesting aspect of Al as a solvent is the ability to use oxide starting
materials as reactants (i.e. metal oxides or complex silicate compounds). The use of oxide
Starting materials is possible due the powerful driving force of the formation of A1203,
Which drives the reactions in a manner reminiscent of the thermite reaction.20 The Al
SOlvent reduces the oxide, likely to elemental components though this is not certain, and

traps the oxygen in the form of A1203. Once this oxide as formed it appears to be stable

even in the A1 flux environment completely removing the oxygen from the system. This
in turn provides the system with finely dispersed metal atoms or fragments to the react
with other components of the reaction solution. An approach based on the use of oxide
starting materials is attractive for many reasons including cost and ease of sample

preparation. Costs of reactants are reduced if oxide starting materials are used versus pure

elemental materials, especially in the case of rare earth elements. Second when using
oxide starting materials the manipulation of the sample can be completed in an oxygen
containing atmosphere which is significantly easier than working in a oxygen vacant
environment both in terms of equipment and general ease to the scientist.

The idea of using molten Al fluxes is not purely based in theory though. In the
literature molten Al metal has proven to be an excellent reactive medium in the synthesis
of many phases including boride compounds21 at high temperatures (>1400° C) along
with the formation of many ternary compounds at temperatures below 1000° C both with
and without the presence of a tetrelide.14 For these reasons it was concluded that the same
approach could be used in the discovery of many new quaternary and higher aluminum
tetrelide compounds in the hopes of forming complex phases exhibiting interesting
structures and physiochemical properties.

To this end we decided to primarily study quaternary systems containing elements
from differing parts of the periodic table. Using a variety of different element types it was
hOped that simple substitution of known structure types would not be predominate and
instead promote the creation of new structural arrangements. To this end it was decided

that the bulk of the exploratory reactions studied would then contain a rare earth element,
tI'ansition metal, aluminum and either silicon or germanium.

Along with promoting new structural arrangements the use of differing types of
e1fiftients can lead to interesting physical properties such as magnetic coupling and charge
transport behavior. For example the use of rare earth ions and their non-interacting 4f

e1€=Ctrons creates the opportunity to study many different magnetic ions in a relatively

uni form environment. Because the rare earth ions all exhibit similar sizes and prefer 3+

 

oxidiation states studies of structural series can provide important information about the
effects of varying the strength and type of magnetic interactions (due to the different
numbers of unpaired spins) in the system without changing important structural
parameters or charges on other atoms in the system. The magnetic systems can become
even more complex if a second magnetic ion is added to the system such as a transition
metal ion. With this addition several different behaviors can occur including simple
addition of the moments and varies forms of coupling between the two magnetic sub-
lattices.22 Also interesting charge transport properties are found in some Al compounds
such as RuAl2 which exhibits behavior typical of semiconducting materials. This is due to

the formation of a psuedo-gap due to hybrization of electron bands in the material.23

4. Al Matrix Alloys.

Another important benefit of studying Al based systems is the relationship of
Compounds grown in metal solution to those found in metal matrix composites.24 These
Phases form when small percentages of impurities, either accidental or intentional, react
to form complex phases within the sea of the parent matrix material affecting the overall
material's properties. This reaction can occur at any point in the lifetime of the composite
including preparation, processing, or over time through use. These newly formed
Complex phases, though only representing a small percentage of the overall amounts can

greatly influence the overall material's properties. Often these small impurities are the
Very source of the desired properties or may be the cause of material failure?"5 Detailed
knowledge of these phases and their behavior, both chemical and mechanical, can

thel‘efore be critical in understanding known materials as well as designing new materials

for future use. One example of this behavior is the formation of the phase CuzMggAISSi6
in A] matrix materials. This phase has been proven to significantly increase the total
matrix's strength without significantly increasing the material's overall weight.26

The problem in studying such materials is that the size of the precipitants which
form inside the actual composite are normally on the micron size scale. From such small
samples only limited amounts of information can be extracted such as crystallographic
system and approximate elemental ratios, but normally information such as complete
crystallographic structures and physical properties are difficult to extract. Understanding
of these properties is often critical in comprehending the material's effects on the overall
system.27 If these phases can be grown as larger crystals in a metal flux reaction, which is
similar in principle to processing in the original extended matrix, they can likely yield
important information needed to understand the behavior of the total metal matrix
composite. Such approaches are not evident in the literature though groups have
attempted studies of these systems using other means such as scaling up the original

fOrrnation reactions or through complex microanalysis techniques.28

5. Subject Matter of the Thesis.

Now that the concepts and benefits of molten Al as a solvent have been discussed
the balance of the thesis will discuss the synthesis, structure and physical properties of
SeVeral tetrelide intermetallic systems synthesized using Al as a solvent. Chapter 2

Pre$ents the ternary compounds RE2A13Si2 (RE: Tb, Dy, Ho, Er, Tm) of the YzAl-jSi2
Stl‘u<:ture type.29 Though the structure type is not new this work discovered many new

analogs of the system and explores their physical properties for the first time. Chapters 3

through 6 present a large series of related structures each containing the MAl,,Tt2 (M: Co,

Ni, Cu, Pd; Tt=Si or Ge) structural layer. This layer appears to be prevalent in later
transition metal systems (Co, N i, Cu) where the rare earth to transition metal ratio is
greater than or equal to 1. Each of the structures presented however present various
binding modes between the layers or the addition of secondary units leading to a wide
variety of new structure types.

Chapter 7 explores the RENiAl,,Ge2 (RE: Sm, Gd, Tb, Dy, Ho, Er, Tm, Lu, Y)
system which crystallize in the hexagonal space group R-3m and exhibits a trigonal layer
of RE ions. This lattice raises interesting questions concerning the magnetic interaction of
the ions and the possible existence of magnetic frustration which is explored in the
Chapter. Chapter 8 studies systems that are transition metal rich for the first time
Presenting the related structure types of RE,_,M2A15-,Siy (RE: Nd, Sm, Tb, Tm, Yb, Y;
M: Ni and Pd) and REZ_,,M;,,A1,,Tt2(Al,_yTty)(Al,_th,)2 (RE: Sm, Er, Dy; M: Ni, Co; Tt=
Si. Ge).

After chapter 8 the thesis moves away from Co, Ni, and Cu systems and looks at
the reactivity and structures of systems containing the earlier transition metals Fe, Mn,
and Ru. Chapter 9, the RE,,M2,,,A17,,,Si8 (M: Fe, Mn; RE=Ce, Pr, Nd, Sm, Gd) structure
tYpe, shows that while Fe and Mn form isostructural compounds to each other the
SL1‘llcztural components are strikingly different from those seen in earlier chapters. Of
particular interest is the apparent reduced electronic state which Fe and Mn ions exhibit.

Chapter 10 explores the transition metal rich Section of the RE/Fe/Al/Si systems with the

REFqugSig (RE = Tb, Er, Gd, Dy, Ho) structure type. Here again we find that the

S trUCtures differ from the later transition metal rich phases highlighting the apparent

structural transition between Fe and Co as one moves across the first row transition
metals. The final chapter presents the REgRu12A1498i9(Alei12_x) (RE = Pr, Nd, Sm, Gd,
Tb, or Er) structure type. The Ru work was completed to compare and contrast the
behavior of a first row transition metal, Fe, to that of the second row counterpart. While
the structural chemistry has changed significantly moving down in the periodic table the

Ru atoms interestingly still appear to exhibit a similar diamagnetic behavior to that which

was seen earlier in the Fe and Mn compounds.

10

References

 

‘ (a) Disalvo, F.J. Solid State Commun. 1997 102, 79-85
2 (a) Shah, D.M.; Berczik, D.; Anton, D.L.; and Hecht, R. Mater. Sci. Eng. A 1992 155,

45-57. (b) Meschter, P. J.; Schwartz, D. J. Metals, 1989 41, 52. (c) Inui, H.; Moriwaki,
M.; Ito, K.; Yamaguchi, M. Philosophical Mag. 1998 77, 375.

3Meier, G.H. High-Temperature Ordered Intermetallic Alloys 1] Stoloff, N .S.; Koch, C.;
Liu, OT; and Izumi, O., Eds.; Materials Research Society Symposium Proceedings 81;
Materials Research Society: Pittsburgh, 1987 p 443.

4 Shah, D.M.; Berczik, D.; Anton, D.L.; Hecht, R. Mater. Sci. Eng, A 1992 155, 45-57

5 Surcsh, S.; Mortensen, A.; Needleman, A. Fundamentals of Metal-Matrix Composites,

Butterworth-Heinemann , Boston , 1993

6 CRC Handbook of Thermoelectrics; Rowe, D.M., Ed.; CRC Press: Boca Rogue, F1 1995

and references therein.

7 Shapoval, V.I.; Malyshev, V.V.; Novoselova, LA; and Kushkhov, K.B. Russ. J. Appl.

Chem, 1994 67, 828-833.
8 Madar, R.; Thomas, N.; and Bernard, c. Mater. Sci. Eng, B 1993 17, 118-125, and

Te f erences therein.

9 (a) Okada s, Kudou K, Miyamoto M, Hikichi Y Nippon Kagaku Kaishi 1991 12, 1612-

1 6 l ‘7 (b) Okada S, Suda T, Kamezaki A, Hamano K, Kudou K, Takagi K, Lundstrom T

Mater. Sci. Eng., A 1996 209, 33-37
10
E1 Well, D.; Scheel HJ. Crystal Growth From High-Temperature Solutions, Academic

Press, New York, 1975

11

 

” Kanatzidis, M. Curr. Opin. Solid St. Mater. 1997 2, 139-149

‘2 Okada S, Kudou K, Miyamoto M, Hikichi Y, Lundstrom Nippon Kagaku Kaishi 1993
5, 681-684

13Cant‘teld, P.C.; Fisk, z. Philos. Mag. B 1992 65, 1117-1123

'4 Canfield, P.; Gammel P.L.; Bishop D.J. Physics Today 1998 51, 40-46

5 Yan, Z.; Takei, H. J. Cryst. Growth 1997 171, 131-135

‘6 Luzhnaya, N.P. J. Cryst. Growth 1968 97-107

‘7 (a) Nieman, S.; Jeitschko, W., J. Solid State Chem. 1995 114, 337-341. (b) Nieman, S.;
Jeitschko, W., Z. Kristallogr. 1995 210, 338-341. (c) Nieman, S.; Jeitschko, W., J. Solid
State Chem. 1995 116, 131-135. ((1) Nieman, S.; Jeitschko, W., J. Alloys Comp. 1995
221 , 235-239. (e) Prots', Yurij; Pottgen, R.; Niepmann, D.; Wolff, M.; Jeitschko, W. J.
Solid State Chem. 1999 142, 400—408

'8 (a) Stitzer, K.E.; Smith, M.D.; zur Loye, H.C. J. Alloys Comp. 2002 338, 104-111 (b)
Stitzer,K.E.; Smith, M.D.; zur Loye, H.C. Chemical Communications 2001 17, 1680-

168 1
19 Binary Alloy Phase Diagrams, T. B. Massalski, ASM International; Publisher William

W. Scott Jr, 1990.
m Nebergall, W.; Schmidt, F.; Holtzclaw, H. Jr. General Chemistry; D.C. Heath and

C0l'l'rpany; Lexington, MA 1972 p. 845
2! Okada, 3.; Yu, Y; Lundstrom, T.; Kudou, K.; and Tanaka T. Jpn. J. Appl. Phys. 1996

35 . 4718-4723. (b) Okada, 3.; ; Kudou, K.; Yu, Y; Lundstrom, T. Jpn. J. Appl. Phys.

1 994 33, 2662-2666
22
Myers, H.P. Introductory Solid State Physics Taylor & Francis, Bristol, PA 1997 p. 356

12

 

23 Mandrus, D.; Keppens, V.; Sales, B.C.; Sarrao, J.L. Phys. Rev. B 1998 58, 3712
2‘ (a) Suresh, S.; Mortensen, A.; Needleman, A. Fundamentals of Metal-Matrix
Composites, Butterworth-Heinemann , Boston 1993 (b) Ochiai, S. Mechanical Properties
of Metallic Composites Marcel Dekker, Inc., New York, NY 1994
25 (a) Chawla, N.; Shen, Y. Adv. Eng. Mater. 2001 3, 357 (b) Tjong, S.C.; Ma, M.A.
Mater. Sci. Eng., R. 2000 29, 29-113
26 (a) Amberg, L.; Aurivillius, B. Acta Chem. Scand, Ser. A, 1980 34A, 1-5
27 Radmilovic, V.; Kilaas, R.; Dahmen, U.; Shiflet, G.J. Acta Mater. 1999 47, 3987-3997
28 (a) Phragmen, G. J. Inst. Met. 1950 77, 521 (b) Zheng, J.G.; Vincent, R.; Steeds, J.W.
Philos. Mag. A 1999 79, 2725-2733
29 Yanson, T.I.; Manyakov, M.B.; Bodak, O.I.; Gladyshevskii, R.E.; Cemy, R.; Yvon, K.

ACta Crystallogr. Sect. C 1994 50, 1377

13

Chapter 2

Synthesis and Characterization of

RE2A13Si2 (RE: Tb, Dy, Ho, Er, Tm)

1. Introduction.

We began the work with molten aluminum by first exploring the ternary RE/Al/Si
systems before moving into quaternary reactions. The idea of studying these systems is
appealing for two reasons first to explore the chemistry of the ternary phases and second
to prove that molten Al could be used as a solvent for exploratory synthesis. In early
reactions several structural analogs of the compound Y2A13Si2 were discovered with
differing RE ions. This structure type was fairly uncommon within the literature with
0111 y reporting the Y analog as a crystallographic solution.1 This series of compounds are
Particularly interesting as proof that complex phases can be synthesized in molten Al in

addition to several interesting physical properties the compounds exhibit. These
Properties include a metamagnetic transitions under the application of a magnetic fields
and a distinct anisotropy in the magnetic response present in the Ho and Tm analogs.
Here we present the synthesis, structure and property characterization of the newly
f0I‘l'ned analogs, RE2A13Si2 (RE = Tb, Dy, Ho, Er, Tm) formed within molten Al. Early
Work in this system has been previously published including the crystallographic solution

of the Ho, Er, Tm analogs and initial magnetic analysis on the systems by our research

«g1.()l_,|p.2

14

L

 

2. Experimental Section.

Synthesis.

Reagents.
The following reagents were used as obtained: Tb, Dy, Ho, Er, Tm, Y 99.9%, -40

mesh, Cerac, Milwaukee, WI; Si, 99.96%, -325 mesh, Cerac, Milwaukee, WI; A1, 99.5

%, -20 mesh, Milwaukee, WI.

Synthetic Method.
Method 1. Synthesis of RE3A13812 was conducted through mixing 1RE: lsi and

placing into an alumina crucible with excess Al. This mixture was sealed under vacuum
and heated at 1000 °C for 5 days. The sample was then slowly cooled to 300° C in a

Period of 96 hours to promote crystal growth. Isolation of the sample was completed by

the procedure described below

Method 2. To conduct a single crystal neutron analysis to differentiate A1 and Si
Positions a larger crystal was needed than that formed from method 1. To form larger
crystals the reaction scale was increased 10 fold and the heating profile was slightly
adj usted. After mixing the reactants and sealing under vacuum the reaction was heated to
950 °C for 72 hours then cooled to 50 °C in 12 hours. This method produced crystals

SeVeral mm in length, large enough for single crystal neutron analysis.

In both methods, 5M NaOH (aq.) solution was used to remove excess Al flux

from the product. Purity of the final product was confirmed through comparison of the

15

experimental X-ray diffraction powder patterns, taken of the bulk product, to theoretical

patterns calculated from the refined single crystal data.

Physical Measurements.

EDS Analysis.

Quantitative rnicroprobe analysis of the two compounds was performed with a
JEOL ISM-6400 Scanning Electron Microscope (SEM) equipped with Noran Energy
Dispersive Spectroscopy (EDS) detector. Data were acquired using an accelerating
voltage of 25 kV and 100 sec accumulation time. Standards where recorded under the

same experimental conditions to yield correction factors. After calibration the Tb

Compound gave an elemental ratio of 2 RE: 2.8 A1: 1.78 Si.

Single Crystal X-ray Crystallography.
RE2A13Si2 crystallize in the Y2A13Si2 structure type previously published

by Gladyshevskii.2 As part of this work two analogs, RE2A13Si2 (RE=Tb, Dy), of the

S'er‘ies were crystallographically solved and refined. This work coincided with the

di Scovery of three other analogs by Dr. Chen in our lab namely RE2A13Si2 (RE=Ho, Er,

Tm). Single crystal X-ray diffraction data of Tb2A13Si2 was collected at room temperature

uSing a Siemans Platform CCD diffractometer using Mo K010» = 0.71073 A) radiation.
The SMART software3 was used for the data acquisition and the program SAINT‘ was
uSed for the data extraction and reduction. An empirical absorption correction using
S‘4\IDABSS was applied to the data and the structure was solved in the SHELXL package

of PI'Ograms6 using the Sm analog as a starting point. Data collection parameters, atomic

16

 

positions, anisotropic thermal parameters information is provided in Tables 2-1 to 2-3.
Selected bond distances for Tb2A13Si2 are listed in Table 2-4.
The second analog, Dy2A13Si2, was collected at room temperature using a Rigaku
4 circle diffractometer with Mo K01 (A = 0.71073 A) radiation. An empirical absorption
correction based on ‘1’ scans was applied to the data. The structure was solved with direct
methods using SHELXS 867 and refined with SHELXS 5.03 package of programs. Data

collection parameters, atomic positions, anisotropic thermal parameters information is

provided in Table 2-1 to 2-3.

After the single crystal solutions were found and atomic assignments of Al and Si
were made, based on bond distances, a neutron structural refinement of H02A13Si2 was
Conducted to verify the assignments. This refinement confirmed the assignments of all

the atoms supporting the premise that bond distances are indeed a viable means of

making such assignments.

Charge Transport Analysis.

DC electrical conductivity and thermopower measurements were completed on
Selected single crystals of Y2A13Si2 and Dy2A13Si2. Conductivity measurements were
Performed with the conventional four-probe technique.8 Thermopower measurements

Were made using a slow AC technique as described elsewhere.9

Magnetic Characterization.

Magnetic susceptibility data for RE2A138i2 (RE = Tb, Dy, Ho, Er, Tm, Y) were

r"He-asured, using polycrystalline powders, as a function of both temperature and field

17

Table 2-1. Crystal data and structure refinement for RE2A13Si2 (RE = Tb, Dy).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

2, Calculated density
Absorption coefficient (11)
RM)
Crystal size
9 range

Limiting indices

. Ref, collected / unique
Rtinr)

. Refinement method

Data / restraints / parameters

GOOdness-of-fit on F2
Final R indices [I>ZO'(I)]

R iIldices (all data)

EXtinction coefficient

Largest diff. peak and hole 1.703 and -1.763 e.A‘3
R 1 =2||1=,| - |FC|IIX|F0|, wR2=[Z(w|F02 - r-‘,2|)2/2‘.(w1=,"-)2]"z

szAl3Si2
454.96

298

0.71073 A

C2/m (#12)

a = 10.194(6) A

b = 4.040(2) A

c = 6.606(4) A

[3 = 101.083(9)
267.0(3) A3

1, 5.659 Mg/m’
27.046 mm"

394

0.12 x 0.15 x 0.24 mm
3.14 to 28.61 °
-l3<=h<=13
-5<=k<=5

-8<=l<=8

1384 / 378

0.0341

DyzAl3Si2
462.12

298

0.71073 A

C2/m (#12)

a = 10.1740(13) A
b = 4.0370(9) A

c = 6.5960(8) A

B = 101.01(1)
265.92(8) A

1, 5.771 Mg/m3
28.090 mrn'l

311

0.5 x 0.3 x0.2 mm
4.08 to 30.00 °
-l4<=h<=l4
0<=k<=5
-9<=l<=9

837 I437

0.0837

Full-matrix least-squares on F2

378 / 0 / 24
1.227

R1 = 0.0219
wR2 = 0.0598
R1 = 0.0220
wR2 = 0.0599
0.057(3)

18

437 / 0 / 24

2.426

R1 = 0.0428

wR2 = 0.1041

R1 = 0.0438

wR2 = 0.1042
0.147(11)

4.967 and -5.885 e.A‘3

Table 2-2. Atomic coordinates ( x104) for RE2A13Si2 (RE = Tb, Dy).

 

 

Wyckoff x y 2
position
Tb 4m 6199(1) 0 3228(1)
Dy 6918(1) 0 3229(1)
Si 4m 9072(3) 0 3600(4)
9073(4) 0 3600(7)
Al(1) 2a 0 0 0
0 0 0
Al(2) 4m 3059(3) 0 1339(5)
3054(5) 0 1347(8)

Table 2-3. Anisotropic displacement parameters (A2 x 103) for RE2A13,Si2 (RE = Tb, Dy).

 

 

U11 U22 U33 U23 U13 U12

Tb 4(1) 7(1) 6(1) 0 2(1) 0
Dy 2(1) 7(1) 5(1) 0 2(1) 0
Si 6(1) 6(1) 6(1) 0 2(1) 0
1(2) 6(2) 9(2) 0 0(1) 0
Al(1) 7(1) 9(1) 9(1) 0 1(1) 0
17(3) 15(4) 2(3) 0 0(2) 0
Al(2) 15(2) 10(2) 6(2) 0 0(1) 0
8(2) 15(3) 2(2) 0 1(2) 0

\

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +
2hka*b*U12]

19

Table 2-4. Selected Bond Distances (A) in szAl3Si2

 

Tb(1)-Si(1) 2.891(3)
Tb(1)-Si(1) 2.962(2)
Tb(1)-Si(1) 3.008(2)
Tb(l)-Al(1) 3.0199(11)
Tb(1)-Al(2) 3.185(3)
Tb(1)-Al(2) 3.205(4)
Tb(1)-Al(2) 3.248(4)
Tb(l)-Tb(1) 3.696(2)
Si(1)-Si(1) 2.378(6)
Si(1)-Al(2) 2.605(3)
Si(1)-Al(1) 2.723(3)
Al(2)- Al(2) 2.781(5)

20

using a MPMS Quantum Design SQUID magnetometer. An initial study of field
dependence was conducted to find a suitable field for the variable temperature studies.
Temperature dependent magnetic measurements on RE2A13512 (RE = Tb, Dy, Y) were
conducted at 200 G while the field dependent measurements were conducted at 5 K. The
field dependent measurements of the other analogs RE = Er, Tm, Ho were conducted at
2.5 K while the temperature dependent measurements were conducted at 200G, 100G,
and 1000 G respectively. Diamagnetic correction was applied to the data to account for

both core diamagnetism and diamagnetic behavior due to the sample container.

3. Results and Discussion.

Synthesis and Thermal Behavior.

Exploratory synthesis utilizing molten Al has proven to be an excellent means to
Study the reactivity of RE metals with Al and Si. Liquid A1 creates an environment that
increases the reactivity of the elements and promotes crystal growth. This allows for large
Well faceted needle-like crystals (>1 cm) of RE2A13$i2 to form. These crystals often grow
together in bundles but can be easily separated by mechanical means without harm being
done to the individual crystals. They are stable in air at ambient temperature but appear to

form an oxide layer on the surface and lose their mechanical stability when exposed to

elevated temperatures (~1000° C) for more than 12 hours.

StI‘uctural Description.
Rl-3,2A13Si2 (RE = Tb, Dy, Ho, Er, Tm,Y) crystallize in the Y2A13Si2 structure

tyPe.‘ The structure of RE2A13Si2 is shown in Figure 2-1 viewed down the b axis. To

21

describe the structure it is useful to regard it as a three-dimensional anionic (A13Siz)6'
framework filled with rare-earth cations. In the framework, parallel Al-Al zigzag chains
of Al(2), running along the b axis, bridged by Si-Si dimers to form AlZSizlayers
perpendicular to ac plane as shown in Figure 2-2. The distances in Tb2A13,Si2 of Al(2)-
Al(2), Al(2)-Si, and Si-Si are 2.781(5), 2.605(3), and 2.378(6) A, respectively. Within the
A1231, layer, chair-like hexagonal rings, formed by Si and Al(2) atoms can be seen
sharing edges along the b axis. These layers are then linked along the c axis by Al(l)
atoms with a Si-Al distance of 2.721 A to form the total three-dimensional structure of
the framework.

The structure exhibits parallel tunnels along b direction with two rows of rare-
earth atoms occupying each tunnel. Within each tunnel, the M-M distances are 4.040 A in
eFitch row and 3.770 A between the two rows for the Tb analog. The shortest RE-RE
diStance of 3.696(2) A is actually found between those RE atoms in adjacent tunnels and
I‘ight above and below the Si-Si dimers, see Figure 2-1

The different atomic positions in the structure are shown in Figure 2-3. The A1
atoms exhibit 2 types of bonding environments a linear environment for Al(l) bonding
and a distorted square planar arrangement for Al(2). Al(l) is bonded to 2 Si atoms in the
A128i2 layers (Figure 2-3a) while Al(2), contained within the Alzsi2 layer, exhibits self
bonding to 2 other Al(2) atoms, forming the Al zig-zag chains, and also bonds to 2 Si
atoms to complete the hexagonal rings. The Si position exhibits a distorted trigonal
pyramidal arrangement bonding to two Al(2) atoms, one Al(l) atom, and one Si atom,
See Figure 2-3c. The Si atoms also display interactions with 5 RE atoms which are within

a I‘adius of 3 A of the central Si atoms. Finally the RE ions sit in the tunnels of the

22

 

z
OREOAInSi >

X

Figme 2-1. Structure of RE2A13Si2 viewed down the b axis.

23

Si Al(2)

Figure 2-2. AlZSi2 layer which runs parallel the b axis in RE2A13Si2.

24

Si Al(2)
Al(2)

A|( 1) Al(2)
Si

Si

 

3
Al(2)
Al(2)

Fi ' '
glue 2-3. Coordination environments in REzAl3812

25

framework exhibiting bonding interactions with 5 Si atoms, 5 Al(2) atoms and 2 Al(l)

atoms .

Charge Transport Properties.

To explore the properties of this system two of the analog's charge transport
properties were measured. These analogs, Y2A13Si2 and Dy2A13Si2, exhibit metallic
conductivity with resistivity values of approximately 20 uQ-cm for the Y compound and
25 119-cm for DyzAl3Si2 at room temperature, see Figures 2-4 and 2-5. Both the low
resistivity values and the temperature dependence of increasing resistivity with increasing
temperatures are in agreement with metallic behavior.

Thermopower values, on the order of a few uV/K, are also consistent with
metallic behavior, shown in Figures 2-4 and 2-5. Y2A13Si2 exhibits positive thermOpower
Values, i.e. p-type behavior, with a veryflat response as a function of temperature until 50
K. Below .50 K an increase in thermopower occurs with cooling temperatures. The two
Crystals of DyzAl3Si2 both Show positive thermopower values while one Show a similar
temperature response as that which was seen in the Y analog. The other crystal shows a
general decrease in thermopower with decreasing temperatures till temperatures are lower
than 50 K where the same behavior occurs as in the first crystal. The small thermopower
Values exhibited by these crystals is in accordance with the expected metallic behavior of

the compounds.

26

N
I 11

 

-* '0
L11 0
I 1 I I I I I I l

A
CL
I Y I I

Resistivity
(,uQ-cm)

 

  
 
  
 
  

 

 

#1 J 1 1 1 1 l 1 1 1 1 l 1 1 1 1 I 1 1 1 1 l 1 1 1 1 l 1 4 1 1
50 100 150 200 250 300 350
Temperature (K)

 

0"
T—r111T
.. .

Thermopower
(,uV/K)

 

 

1

 

FiglIre 2-4. Electrical resistivity and thermopower measurements for single crystals of

Y2

1111111111111111L1111
100 150 200 250 300

Temperature (K)

A13812. Two separate single crystals were measured for resistivity analysis.

27

#
ICD

 

Resistivity

(pQ-cm)
N N w 00
”cannot . HCID‘ . '01

._s
01

 

10”

 

 

 

1 1 1 1 1 1 1 1 l 1 1 1 1 l 1 1 1T1 1 1 1 1 l 1 1 1 1 l 1a_1 1
0 50 100 150 200 250 300 350
Temperature (K)

 

OD #- 01 O)
IIIIIIIIIIIIIIUUF
0..
O

Thermopower
(pV/K)
N

o .4
TIT!

441

 

 

1 L4 11111 #141 l lll¥1A111 l

r

l' 1 1 1 1

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

 

I
A

Flglll'e 2-5. Electrical resistivity and thermopower measurements of single crystals of

Dy2Al3Si2. Two separate single crystals were measured in each analysis.

28

Magnetic Properties.

Each analog of RE2A13Si2 was studied in terms of magnetic behavior to
understand the charges of the rare earth ions as well as study any possible. magnetic
interactions which may be present in the systems. All show antiferromagnetic transitions
at low temperatures with the exception of Er2A13Siz, in which no transition is seen, and
the Y2A13Si2 which is Pauli paramagnetic, see table 2-5. The transition temperatures range
from about 15 K in szAl3Si2 to 9 K in TmzAl3Siz. A trend in Tm, for these systems is
seen between transition temperatures and RE size with T111111 dropping as the size of the
RE ion decreases. At high temperatures well above Tm all analogs follow Curie-Weiss
behavior. From this behavior the 11,," values for each analog are found to be very close to
those expected for RE ions in a 3+ oxidation state, see Table 2-5.

One note of interest is that the Ho and Tm analogs appear to display broad
antiferromagnetic transitions. This behavior was found to be due to differing orientations
of crystals in the powder sample and corresponds to an averaging of orientations in the
sample. The high temperature transition was found to be due alignment of the filed along
the b axis of the crystal while the lower temperature transition is due to the field
Perpendicular to the b axis. This can be seen in Figure 2-11 where a sharp Single
t1'Elnsition is seen with the field along the b axis. The magnetic response with the crystal
Perpendicular still shows both transition temperatures though the high temperature
tranSition is less evident and presumably only present due to the a slight misorientation in
the Samme.

When the field dependent behavior is examined all analogs Show a general

“crease in magnetization out to the highest fields measured W1th no Signs of magnetic

29

saturation present. The Ho analog shows a marked difference from the other analogs, due
to orientation effects, and the fact that a hysteresis in magnetization is seen between with
increasing versus decreasing fields at about 40,000 G. The Tm analog however does not
show similar field dependent effects even though the behavior is similar in the

temperature dependent susceptibility measurements.

4. Conclusions.

Utilizing Al flux crystals of RE2A13Si2 form up to several mm in length with well
facetted edges. These large crystals allowed for easy analysis of physical properties
including charge transport and magnetic behavior. The charge transport behavior exhibits
typical p-type metal with values of 15 and 25 119-cm at room temperature for Y2A13,Si2
and Dy2A13Si2, respectively. Magnetically the systems all display behavior corresponding
that of RE ions in 3+ oxidation states while the other ions are magnetically silent. No
magnetic saturation is seen upon application of a magnetic field up to the maximum
applied field of 55,000 G. The H02A13812 and TmzAlfii2 analogs do however Show a
slightly more complex magnetic behavior with differing T111111 values, in the temperature
dependent measurements, effected by the orientation of the crystal in relation to the

aPplied field.

30

(emu / mole RE ion)

1/x

(mole RE ion / emu)

nefizafion

Map
(B.M. mole RE ion)

 

 

 

 

 

 

 

 

 

 

 

 

oafi.<
I. O
0.15— l o
v 0
I
O
0.1- o .
0.05-
0 1 1 1 1
o 20 40 60 80 100
Temperature (K)
30
O
25- ’
O
O
O
20- .
O
I
15- .
l
O
101 o.
O
.0
C ..l l l 1 l 1
ti 5'0 100 150 200 2&0 300 350
Temperature (K)
L “U
.-
1.5- J.-
00
1- 00
O.
I.
0.5- a“
0- f
05 0".
o . .1 fl
.-
-1. ..o’
.0
4.54 ’0‘
.0.
-2 1 1 1 1 1
-610‘ -410‘ -210‘ 0 210‘ 410‘ 610‘
Field (G)

Figure 2-6. Magnetic behavior of Tb2A13Si2 (top) susceptibility as a function of

temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

31

 

 

 

 

 

 

 

 

 

 

 

 

0.4
A I".
c .
,9 0.35 —o'-.,,.33 .
LLI .00...
II . e
a) .
sf 6 03 . .
s t
\ I
e . -
0.25 -
3 o
0.2,.1111141L1l1111dn111la111
0 10 20 30 40 50
Temperature (K)
20
I o
. O
’5 ~ '
E 15 e . °
0 r
\ i . .
E 8 ~ 0
\ “" 10 e o
.— m 1 .
m 0
O
s 1 .-
E 5 '
o P1144111111111141111111.11111111
0 50 100 150 200 250 300 350
Temperature (K)
4: .
I O
3E
1.? : 0 .
.9 2 L . '
St“ E . '
gm 1 :- . 0
N2 o
'3 O 0 :- '
C) E >- .
C r o
g\ -1 :— . o
255 -2: . . °
v 1 o
-3 L o
_ O
_4*91111111111111111A1111
-610‘ -410‘ -210‘ 0 210‘ 410‘ 610‘

Field (G)

Figure 2-7. Magnetic behavior of Dy2A13Si2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

32

 

 

 

 

 

 

 

 

 

 

 

 

1.0
t.
L
€0.80;
0 b
.. _ .
UJ . o
1110.60 - ',
d) r 0
EB p
R E 1. O
\0.40: ° . .
a I ' o
00.20 _
0.0» 11L14L11111J11L11L111
O 10 20 30 40 50
Temperature (K)
20
’ I
: o
A O
3 r
E 15 ~ '
G) ' '
\ " o
E C : 0
id .9 1o __ .0
,. LU . .0
t: .0
2 I .°
0 _ o
g 5.0 /
00 1 11111111111111L111111111 1111
O 100 150 200 250 300 350
Temperature (K)
15 _ .3
; 1:.
A 10 L :0 /
C I O
c-9 i 0
gm 5; o
to“: i
QE 0 _ /
m0 1
CE I
O) .
«3‘ -5 _ o
22 i o
. - .
m ,
v '10 r :
: V
.151 A114+L4 1411111111
-610‘ -210‘ 0 210‘ 410‘ 6104
Field (G)

Figure 2-8. Magnetic behavior of H02A13Si2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

33

A
O

 

p
(n
O
1
o

 

 

 

 

 

 

 

 

 

 

 

’EE
.9 l ',
111 , 0,
0: 0.60 —
82 ' o
x o t .
E 0.40 - .
3 C ' .
E b . . 0
33, 0.20 -
0.0.41111111111111111111...,
0 10 20 30 4o 50
Temperature (K)
30 .
E O
’5 25 T o
' O
E ’ .
‘1’ 20 '— o
“m ' o
' o
E c : .
3 Q 15 .— o
1'- UJ _ .
(I : .
.9 1° :' o
O C .0
g 50 L .9.
0.0“11141111111114111111LLLJ11,,,
0 50 1 00 1 50 200 250 300 350
Temperature (K)
2.0 _
i 1
1.5 r o
A ’ .
c: : .
83 1.0 :- ... 0
_ . .
‘50: 0.50 :- ....
N a: -
8'6 00 f /
83E : 0'.
m\. '0.50 ':' ....
22. -1.0 - o .
m I o
v I .
-1.5 - I
15
-2.0 . J 4 l 1 L k 1 1 g: 1 1 1 1 1 1 g
'110‘ . 5000 110‘

0
Field (G)

Figure 2-9. Magnetic behavior of Er2A13Si2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field. °

34

Xm
(emu / mole RE ion)

3
E
Q)
\~
C
PJE.S2
-
FLU
m
2
C)
E

Magnetization
(BM. / mole RE ion)

—5
O

F>
a)
c:

9’
.h
C

9’
A)
c:

25

20

15E-

10

1

110

21)

‘LO

.0
8

 

 

llllllLJLJLlllll

 

 

20 30 40 50

Temperature (K)

 

'YY‘Y'I'Y

I'Ifij'I'i

 

l l l I l J L L l I

 

 

00

150

200 250 300 350

Temperature (K)

 

IrY'V'TV

l‘1'

IT‘VTI’TY‘VIII

 

1".

1 J

 

L l l A l A J_L 1 L

 

-5000

5000 1 10‘

0
Field (G)

Figure 2-10 Magnetic behavior of TmzAl3Si2 (top) susceptibility as a function of

temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

35

 

 

 

 

0.14 E
O
0.13- o
”E? ' 0
.2 0.12- 0 parallel
a 011 . °
8% . - 0 O . .
x E 0.1. ° ° ‘ . 0
\ 0 . .
5 009 - . 0
g - ‘ perpendicular .
v 0
0.08- ’ .
0
0'07 1 I I I IT1 I I I I I I I I1 rrrifi IIIIIIII
0 5 10 15

Temperature (K)

Figure 2-11. Single crystal magnetic response of H02A13Si2 with the field aligned both

parallel and perpendicular the b axis.

36

Table 2-5. Magnetic properties of RE2A13Si2 (RE: Tb, Dy, Ho, Er, Tm)

 

 

RE fleaCB-MJ “theo (BM) 9 (K) me(K)
Tb 9.92 9.72 -38.6 15

Dy 12.09 10.6 -433 14

Ho 11.47 10.6 -94 . 10

Er 10.35 9.57 -l4.6 NA

Tm 7.58 7.63 6.3 9

 

37

References

 

‘ Yanson, T.I.; Manyakov, M.B.; Bodak, 0.1.; Gladyshevskii, R.E.; Cemy, R.; Yvon, K.
Acta Crystallogr. Sect. C 1994 50, 1377

2 Chen, X.Z.; Sieve, B.; Henning, R.; Schultz, A.; Brazis, P.; Kannewurf, C.R.; Cowen.
J.; Cosby, R.; Kanatzidis, M.G. Angew. Chemie Int. Ed. 1999 38, 693

3 SMART. Data Collection Software for the SMART System. Siemans Analytical X—ray
Instruments Inc., 1995

‘ SAINT. Data Processing Software for SMART System. Siemans Analytical X-ray
Instruments Inc., 1995

5 Sheldrick, G.M. University of Gdttingen, Germany. Manuscript to be Published

6 Sheldrick, G.M. SHELXL Structural Determination Programs, Version 5.0; Siemans
Analytical X-Ray Instruments Inc.: Madison, WI, 1995

7 (a) G.M. Sheldrick, In Crystallographic Computing 3 ; Sheldrick, G.M., Kruger, C.,
Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985 pp. 175-189.(b)
SHELXTL: Version 5, 1994 G.M. Sheldrick, Siemens Analytical X-ray Systems, Inc.,
Madison Wisconsin, USA.
8 Lyding, J.W.; Marcy, H.O.; Marks, T.J.; KanneWurf, C.R. IEEE Trans. Meas. 1988 37,

76~80.

9 Marcy, H.O.; Marks, T.I.; Kannewurf, C.R. IEEE Trans Instrum. Meas. 1990 39, 756-
760.

38

Chapter 3.
Synthesis and Characterization of the Phases RENiAl4(Si2_xNix)
(RE = La, Ce, Pr, Nd, Eu), CeCoAl4Si2, RECuAl4(Si2,xCux) (RE

= Ce, La and Sm) and LanAl4(Si2,dex)

1. Introduction.

Al solvents have proven useful in exploratory synthesis for ternary compounds
such as RE2A13Si2 (see chapter 2) but the whether the methods could be used for the
formation of quaternary phases still remained. To explore possible uses in quaternary
systems reactions containing Ni were conducted in attempts to form true quaternary
compounds. It was found that the inclusion of Ni in the reaction mixtures allowedfor the
formation of quaternary compounds containing RE, Ni, Al, and Si.

While working with the systems with early rare earth metals (RE: La-Nd, Eu) in
the RE/Ni/Al/Si systems a new series of compounds were found. These phases
RENiAl4(Si2_,Ni,) adopt the KCu,,S3 structure type1 which appears to be very stable
forming with a variety of differing transition metals including Ni, Cu, Co, and Pd as well
as Several rare-earth metals. This existence of isostructural compounds with varying
transition metals as well as rare earth metal ions may provide an interesting opportunity
to ti3Xplore the effects of this variation on the systems properties. This chapter presents the

Synthesis, and crystallographic structure of the title compounds along with preliminary

Charge transport and magnetic characterization of the CeNiAl4(Si2_,Ni,) analog.

39

2. Experimental Section.
Synthesis.
Reagents.

The following reagents were used as obtained: Ce 99.9%, metal chips, Research
Chemicals, Phoenix, AZ; La, Pr, Nd 99.9%, -40 mesh, Cerac, Milwaukee, WI; Eu 99%, -
325 mesh, Alfa Aesar, Ward Hill, MA; Ni, 99%, 325 mesh, Sargent, Buffalo Grove, 11;
Co 99.8%, -325 mesh, Cerac, Milwaukee, WI; Cu 99.5%, -325 mesh, Cerac, Milwaukee,
WI; Pd 99.95%, -325 mesh, Cerac, Milwaukee, WI; Si, 99.96%,-325 mesh, Cerac,

Milwaukee, WI; A1, 99.5 %, -20 mesh, Milwaukee, WI.

Synthetic Method.

RENiAl4(Si2,,Ni,) (RE = La, Ce, Pr, Nd, Eu), CeCoAl,,Si2, RECuAl4(Si2_xCux) (RE
= Ce, La and Sm) and LanAl,(Si2_,Pd,) was prepared by mixing 1 mmol of elemental
RE, 1 mmol of the transitional metal, 0.26 mg (10 mmol) A1, 0.056 mg (2 mmol) Si in a
N2 atmosphere and placing inside an A1203 crucible. The alumina tube was then sealed in
an evacuated (1.0x10'4 Torr) 13 mm o.d. by 11 mm id. fused silica tube and heated at
1000° C for 8 hours. It was then cooled to 860° C and maintained for 48 hours.
Subsequently the reaction was cooled to 260° C in 36 hours and then to 50° C in 6 hours.
After heating, the final products were isolated from the solidified Al matrix with 5 M
N aOH solution. After the matrix was removed the product was washed and dried with
acetone and ether yielding stable silver plates and black powder both of the desired

phase, in 56% yield based on the Ni metal used. Purity of the final product was confirmed

40

through Comparison of the experimental X-ray diffraction powder patterns, taken of the

bulk product, to theoretical patterns calculated from the refined single crystal data.

Physical Measurements.
EDS Analysis.

Quantitative rnicroprobe analysis was performed with a JEOL J SM-6400
Scanning Electron Microscope (SEM) equipped with Noran Energy Dispersive
Spectroscopy (EDS) detector. Data were acquired using an accelerating voltage of 25 kV
and 100 sec accumulation time. Standards where recorded under the same experimental
conditions to yield correction factors. After calibration samples of NdN i,_5Al4Si 1.5 yielded
a consistent elemental ratio of 1 Nd: 1.8 Ni: 4.0 A1: 1.9 Si was found, well within

experimental errors from the actual formula.

Single Crystal X-ray Crystallography.

Data were collected on RENiAl4(Si2_xNix) (RE = La and Ce), CeCoAl4Si2 and
CeCuh,,,Al4Si2_x using a Rigaku AFC6S four circle automated diffractometer. Room
temperature data were collected using MoKoc (1:0.71073 A) radiation and an empirical
absorption correction was made based on 111 scans. The structures were solved utilizing
direct methods and refined with the SHELXL package of programs.2 For these
compounds, as with all Al/Si compounds, the Al and Si positions could not be
distinguished directly from the collected x-ray scattering data, due to Al and Si's similar
X-ray scattering .cross sections. The identification of Al and Si atoms in the structure was

made through analysis of bond distances involving the particular site as will be discussed

41

later. The crystallographic and refinement data for each analog are listed in Tables 3-1 to
3-8. Selected bond distances for LaNiAl4(Si2,,Ni,) are shown in Table 3-9.

Single-crystal X—ray analysis was completed on the other analogs RENiAl4(Si2_
xNix) (RE = Pr, Nd), RECuAl4(Si2,,Cux) (RE = La, Sm) and LanAl4(Si2,,de) using a
Siemens SMART Platform CCD diffractometer with MoKa (8:071073/3.) radiation.
After data collection, cell refinement and data processing was performed using the
program SAINT .3 After data processing an empirical absorption correction was applied
with the use of the SADABS program. The final structure was solved by direct methods
and refined with the SI-{ELXL3 package of programs using the LaNiAl4(Si2,xNix) solution
as a basis. The crystallographic and refinement data for the structural analogs are listed in

Tables 3-1 to 3-8.

Charge Transport Properties.

DC electrical conductivity and thermopower measurements were completed on
selected single crystals of CeNiAl4(Si2,,Ni,). Conductivity measurements were performed
using the conventional four-probe technique.4 Thermopower measurements were made
using a slow AC technique as described elsewhere.5 Attempts to measure the other Ce

analogs have been unsuccessful to date due to problems with crystal size and shape.

42

Table 3-1. Crystal data and structure refinement for RENiAl4(Si2,,Ni,) (RE = La, Ce).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11.)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
Rim
Completeness to 0um

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=£IIF.I - IF.Il/EIF.I. sz=12(wIF.2 - F.2I>2/2(wF.2)21“2

LaNil.46Al4Sil.54

C3Ni116A14Si184

375.77 369.20
293(2) K 298(2) K
0.71073 A 0.71073 A
P4/mmm (#123) P4/mmm (#123)
a=4.l788(9) A a=4.1279(13)A
c = 7.959(1) A c = 7.882(4) A
138.99(4) A3 134.30(8) A3
1, 4.489 Mg/m’ 1, 4.565 Mg/m3
13.714 mm" 12.811 mm“
173 166
0.16 x 0.2 x 0.08 mm 0.5 x 0.4 x 0.05 mm
15.01 to 29.96° 2.58 to 28.18°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=h<=5
-11<=l<=11 -lO<=l<=10
448/ 127 1291 / 131
0.0311 0.0899
80.9 % 98.5 %
Full-matrix least-squares on F2
127/1/15 131/1/15
3.855 1.226
R1 = 0.0788 R1 = 0.0565
wR2 = 0.1844 wR2 = 0.1478
R1 = 0.0807 R1 = 0.0565
wR2 = 0.1845 wR2 = 0.1478
005(6) 004(2)

8.417 and -5032 WA3

43

3.296 and 2.432 e-/A3

Table 3-2. Crystal data and structure refinement for RENiAl,(Si2,xNi,) (RE = Pr, Nd).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume
Z, Calculated density

Absorption coefficient (11)

F(000)
Crystal size
0 range

Limiting indices

Ref. collected / unique
Ram)
Completeness to 0m

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=Z||Fo| - (Rn/mm, wR2=[2(w|F,2 — air/2612921!”

Pr Nil.38Al4Si 1.62

NdNi1_5Al,,SiL5

374.71 382.33
293(2) K 293(2) K
0.71073 A 0.71073 A
P4/mmm (#123) P4/mmm (#123)
a = 4.1336(6) A a = 4.109(3) A
c = 7.9614(16) A c = 7.924(7) A
136.03(4) A3 133.80(16) A3
1, 4.574 Mg/m3 1, 4.745 Mg/m3
13.236 mm" 14.055 mm"
167 168
0.03 x 0.03 x 0.06 mm 0.03 x 0.06 x 0.09 mm
2.56 to 28.22° 4.96 to 28.3l°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-10<=l<=10 -10<=l<=10
1326/ 132 1279/ 127
0.0546 0.0779
97.1% 94.8%
Full-matrix least-squares on F2
132/1/15 127/1/15
1.111 1.260
R1 = 0.0293 R1 = 0.0393
wR2 = 0.0699 wR2 = 0.0889
R1 = 0.0292 R1 = 0.0393
wR2 = 0.0699 wR2 = 0.0889
0.086(14) 030(4)

1.156 and -2191 e-IA3

1.769 and -2.948 e-/A3

Table 3-3. Crystal data and structure refinement for EuNiAl,,Si2 and CeCoAl,Si2.

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient ()1)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R601)
Completeness to 0max

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient
Largest diff. peak and hole

R1=2||F°| - |Fc||/Z|Fo|, wR2=[z(w|F,2 ~ F,2|)2/Z(WF.2)21”2

3.411 and -3.800 e-IA3

45

EuNiAi,8i, CeCoAl4Si2
375.03 363.45
293(2) K 293(2) K
0.71073 A 0.71073 A
P4/mmm (#123) P4/mmm (#123)
a = 4.1484(6) A a = 4.1484(6) A
c = 8.049(2) A c = 8.049(2) A
138.51(4) A3 140.74(4) A3
1, 4.596 Mg/m3 1, 4.288 Mg/m3
14.235 mrn'l 17.096 mrn‘l
179 176
0.03 x 0.06 x 0.09 mm 0.16 x 0.16 x 0.08 mm
4.90 to 30.00 2.53 to 29.97°
0<=h<=5 -5<=h<=5
-4<=k<=4 -5<=k<=5
-ll<=l<=ll -11<=l<=ll
162/159 1594/ 156
0.0136 0.0419
99.4 % 100 %
Full-matrix least-squares on F2

159/1/14 156/0/13

. 1.300 1.260
R1: 0.0417 R1: 0.0219
wR2 = 0.1018 wR2 = 0.0480
R1=0.0419 R1 =0.0221
wR2 = 0.1019 wR2 = 0.0481
017(2) 037(2)

1.773 and -1.276 e—/A3

Table 34. Crystal data and structure refinement for RECuAl4(Si2_xCux) (RE = Ce, La).

Formula
Formula weight
Temperature

. Wavelength
Space group

Unit cell dimensions

Volume

2, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int) =
Completeness to 0mm,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient
Largest diff. peak and hole

R1=2||Fo| - |Fc||/2|Fo|, wR2=[Z(WIFoz - Fc2|)2/2(wF,,2)2]"2

CCCU1.86A14Si1.14

LaCUL36Al4Si1J4

398.26 397.06
293(2) K 293(2) K
0.71073 A 0.71073 A
P4/mmm (#123) P4/mmm (#123)
a = 4.2052(6) A a = 4.2116(9) A
c = 7.912(2) A c = 7.934(3) A
139.91(4) A3 140.72(6) A3
1, 4.727 Mg/m3 1, 4.685 Mg/m3
15.623 rrtrrt'l 11.943 mm"
179 152
0.04 x 0.12 x 0.20 mm 0.2 x 0.2 x 0.05 mm
2.57 to 29.78° 2.57 to 28.28°
-5<=h<=5 -5<=h<=5
-4<=k<=4 -5<=k<=5
-11<=l<=11 ~10<=l<=10
932/ 156 1370/ 137
0.0344 0.0310
98.7 % 99.3 %
Full-matrix least-squares on F2
154/1/15 137/1/15
1.272 1.177
R1 = 0.0169 R1 = 0.0174
wR2 = 0.0435 wR2 = 0.0426
R1 = 0.0673 R1 = 0.0183
wR2 = 0.2062 wR2 = 0.0433
0.038(7) 0.016(4)

2.702 and -3209 e-/A3

46

0.798 and -0944 e-IA3

Table 3-5. Crystal data and structure refinement for SmCu1,,,Al,,Si2_x and Lan,,xAl4Si2_,.

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density

Absorption coefficient (11.) ‘

F (000)
Crystal size
0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]‘

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=2||F,| - |Fc||/Z|Fo|, wR2=[E(w|F°2 - 1=,2|)2/‘2:(wF,2)2]”2

smcul66Al45ir34

LaP dl.30A14Si1.70

404.22 409.41
293(2) K 293(2) K
0.71073 A 0.71073 A
P4/mmm (#123) P4/mmm (#123)
a = 4.146(3) A a = 4.2659(17) A
c = 7.978(7) A c = 8.026(5) A
137.14(18) A3 146.05(12) A3
1, 4.894 Mg/m3 1, 4.922 Mg/m3
15.397 mm'l 11.157 mm“1
171 183
0.03 x 0.06 x 0.18 mm 0.06 x 0.09 x 0.18 mm
2.55 to 28.16° 2.54 to 27.90°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-10<=l<=10 -10<=l<=10
1338/ 134 1376/ 137
0.0688 0.0240
97.8 % 99.3%
Full-matrix least-squares on F2
134/1/15 137/1/15
1.263 1.171
R1 = 0.0389 R1 = 0.0266
wR2 = 0.1077 wR2 = 0.0709
R1 = 0.0389 R1 = 0.0266
wR2 = 0.1077 wR2 = 0.0709
0.058(17) 0.015(7)

1.731 and -3313 e-IA3

47

3.963 and -1.348 e-IA3

Table 3-6. Atomic coordinates ( x10“) and occupancies for RENiAl4(Si2,xNi,) (RE: La,
Pr, Nd, Eu. Ce), CeCoAl4Si2, RECuA14(Si2_,Cu,) (RE: Ce, La, Sm) and LanAl4(Siz_
.Pd.)-

 

 

Wyckoff x y 2 Occupancy
Position

La la 0 0 0 1
Ce 0 0 0 1
Pr 0 0 0 1
Nd 0 0 0 1
Eu 0 0 0 1
Ce 0 0 0 1
Ce 0 0 0 1
La 0 0 0 1
Sm O 0 0 1
La 0- 0 0 1
Ni lb 0 0 5000 1

0 0 5000 1

0 0 5000 1

0 0 5000 1

0 0 5000 1
Co 0 0 5000 1
Cu 0 0 5000 1

0 0 5000 1

0 0 5000 1
Pd 0 0 5000 l

 

48

Table 3-7. (continued) atomic coordinates ( x10“) and occupancies for RENiAl,(Si2_,Ni,)
(RE: La, Pr, Nd, Eu. Ce), CeCoAl,,Si2, RECuAl,(Si2.,Cux) (RE: Ce, La, Sm) and
LanAl4(Si2_,Pd,).

 

 

Wyckoff x y 2 Occupancy
Position

Ni(1) 2g 5000 5000 1478(8) 0.23
5000 5000 1479(5) 0.16
5000 5000 1460(3) 0.18
5000 5000 1460(3) 0.25

N A*

N A*

Cu( 1) 5000 5000 1477(1) 0.42
-5000 -5000 1480(1) 0.43
5000 5000 1457(4) 0.33

Pd(1) 5000 5000 1473(2) 0.15

Si( 1) 5000 5000 1478(8) 0.77
5000 5000 1479(5) 0.84
5000 5000 1460(3) 0.81
5000 5000 1460(3) 0.75
5000 5000 1465(4) 1
5000 5000 1440(2) 1
5000 5000 1477(1) 0.57
-5000 -5000 1480(1) 0.57
5000 5000 1457(4) 0.77

5000 5000 1473(2) 0.84

 

* These analogs do not exhibit a refinable amount of transition metal on this site.

49

Table 3-7. (continued) atomic coordinates ( x10“) and occupancies for RENiA14(Si2,xNi,)
(RE: La, Pr, Nd, Eu. Ce), CeCoAl4Si2, RECuAl4(Si2_,Cu,) (RE: Ce, La, Sm) and
LanAl4(Si2,,P¢).

 

 

Wyckoff x y 2 Occupancy
Position
Al 41 5000 0 3352(7) 1
5000 0 3332(4) 1
5000 0 3335(2) l
5000 0 3325(3) 1
5000 0 3366(3) 1
5000 0 3392(2) l
5000 0 3272(2) 1
5000 0 3279(2) l
5000 0 3327(3) 1
5000 0 3293(2) l

 

50

Table 3-8. Anisotropic displacement parameters (A2 x 103) for RENiAl,(Si2.,Ni,) (RE:
La, Pr, Nd, Eu. Ce), CeCoAl,Si2, RECuA1,(Si,.,Cu,) (RE: Ce, La, Sm) and LanAl4(Si2,

dex)‘

 

U11 U22 U33 U23 U13 U12

 

La 4(1) 4(1) 13(2) 0 0 0
Ce 8(1) 8(1) 10(1) 0 0 0
Pr 4(1) 4(1) 6(1) 0 0 0
Nd 5(1) 5(1) 3(1) 0 0 0
Eu 8(1) 8(1) 12(1) 0 0 0
Ce 7(1) 7(1) 10(1) 0 0 0
Ce 5(1) 5(1) 6(1) 0 0 0
La 7(1) 7(1) 8(1) 0 0 0
Sm 8(1) 8(1) 10(1) 0 0 0
La 8(1) 8(1) 11(1) 0 0 0
Ni 3(2) 3(2) 10(2) 0 0 0

3(1) 8(1) 8(2) 0 0 0

5(1) 5(1) 5(1) 0 0 0

4(1) 4(1) 2(1) 0 0 0

8(1) 8(1) 6(1) 0 0 0
Co 7(1) 7(1) 4(1) 0 0 0
Cu 7(1) 7(1) 6(1) 0 0 0

8(1) 8(1) 8(1) 0 0 0

7(1) 7(1) 8(1) 0 0 0
Pd , 8(1) 8(1) 10(1) 0 0 0

 

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +
2hka*b*U12]

51

Table 3-8. (continued) anisotropic displacement parameters (A2 x 103) for RENiAl4(Si2,
,Nix) (RE: La, Pr, Nd, Eu. Ce), CeCoAl,Si2, RECuAl4(Si2,xCux) (RE: Ce, La, Sm) and
LanAl,(Si2_,Pd,).

 

U11 U22 U33 U23 U13 U12

 

Ni(1) 8(2) 8(2) 14(3) 0
7(2) 7(2) 13(2) 0
6(1) 6(1) 7(1) 0
6(1) 6(1) 6(2) 0

COCO
COCO

NA*

NA*

Cu(l) 7(1) 7(1) 5(1)
8(1) 8(1) 6(1)
10(1) 10(1) 8(2)

Pd(l) 9(1) 9(1) 6(1)

COCO
CCOC
COCO

Si(l) 8(2) 8(2) 14(3)
7(2) 7(2) 13(2)
6(1) 6(1) 7(1)
6(1) 6(1) 6(2)
9(1) 9(1) 6(1)
9(1) 9(1) 3(1)
7(1) 7(1) 5(1)
8(1) 8(1) 6(1)
10(1) 10(1) 8(2)
9(1) 9(1) 6(1)

CCCCCCCCCC
OCCCCOCCCC
OOCCCCCCCC

 

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +
2hka*b*U12]
* These analogs do not exhibit a refinable amount of transition metal on this site.

52

Table 3-8. (continued) anisotropic displacement parameters (A2 x 103) for RENiAl4(Si2_
,Nix) (RE: La, Pr, Nd, Eu. Ce), CeCoAl,Si2, RECuAl4(Si2_XCux) (RE: Ce, La, Sm) and
LanAl,(Si2_,Pd,).

 

U11 U22 U33 U23 U13 U12

 

Al 7(3) 16(3) 10(2) 0

10(2) 16(2) 9(2) 0

6(1) 12(1) 6(1) 0

5(1) 13(1) 3(1) 0
9(1) 10(1) 6(1) 0
9(1) 10(1) 3(1) 0
8(1) 11(1) 5(1) 0
9(1) 13(1) 7(1) 0
6(1) 13(2) 7(1) 0
8(1) 14(1) 7(1) 0

CCOCCOCOCC
CCOCCOCOCC

 

The anisotropic displacement factor exponent takes the form: —2n2(h2a*2U11 + +
2hka*b*U12]

53

Table 39. Selected bond distances [A] for LaNiAl,(Si,,,Ni,).

 

BOnd Distance (A)
La-Si(1)/Ni(1) 3.180(2)
Ni-Al 2.467(3)
Si(l)/Ni(1)-Si(l)/Ni(l) 2.353(13)
Si(l)/Ni(1)-Al 2.567(5)
Al-Al 2.624(11)
Al-Al 2.9549(4)

54

Magnetic Characterization.

Magnetic susceptibility for CeNiAl4(Siz,xNi,) was measured as a function of both
temperature and field using a MPMS Quantum Design SQUID magnetometer to explore
the magnetic behavior of the system. An initial Study of field dependence was conducted
to find a field suitable for the temperature dependence Studies. Temperature dependence
measurements on polycrystalline samples were then conducted under increasing
temperature within a 200 G field. Field dependent measurements were conducted at 2K in
fields between :t 1000 G. A diamagnetic correction was applied to the data to account for
core diamagnetism though no correction was made for the sample container however as
the correction factor was well over an order of magnitude smaller than the sample Signal

itself.

3. Results and Discussion.
Synthesis and Thermal Behavior.

Exploratory work led to the formation of an extensive class of compounds of the
KCu4S3 structure type. These phases form predominately in the reaction of early rare
earth metals, transition metals, and Si the presence of excess Al crystallizing as large
plate-like crystals up to several mm per edge. The crystals are often twinned stacking on
top of each other though can be easily separated by mechanical means to yield single
crystals for analysis. The compounds are stable in water and bases but dissolve in dilute

acids (10% HCl) in the period of a day. Thermally the compounds are indefinitely stable

55

at ambient temperatures but upon exposure to air at 1000° C for longer than 24 hours

surface attack is seen.

Structural Description.

Before the structure can be discussed in detail the positions of Al and Si need to
be assigned. In this compound due to the limited types and number of unique atoms
atomic assignments can be made based on bond distances. The Al positions in
LaNiAl4(Si2_xNix) involves bonding to Ni and the mixed Si/Ni position at distances which
corresponding to A1 in the position at 2.467(3) A and 2.567(5) A respectively. The Si
position, which is occupied largely by Si, exhibits distances one would typically expect
for a Si atom namely RE-Si distances of 3.180(2) A, Si(1)/Ni(1)-Si(l)/Ni(1) 2.353(13) A,
and Si(1)/Ni(1)—Al with a distance of 2.567(5) A.

The structure of REMA14(Si2,,Mx) (RE: early rare earth metal, M: transition
metal) crystallizes in the space group P4/mmm (#123) with the KCu,,S3 structure type
shown in Figure 3-1. The structure is best described as stacking of MAl4(Si2_xM,) layers,
see Figure 2, along the z axis. These layers are then joined through bonding between the
capping Si/M mixed occupancy positions to form the total 3D structure. The RE ions then
occupy cavities between the joined MA1,,Si2 layers. The MA1,,Si2 type unit, or the Ge
analog, is a very stable structural unit, related to the antiflourite structure type, and is seen
recently in many intermetallic compounds such as LaGagNiH,6 szNi(Si1_,,,Ni,,)Al.,Si67
and szNiA14Ge2.8

The disordered M/Si position, mostly occupied by Si, exhibits a 5 coordinate

square pyramidal arrangement with 4 Al-Si bonds directed into the MAl4(Si2_,M,) layers

56

while the fifth bond serves to link the adjacent layer, see Figure 3a. The A1 position in the
structure displays a five coordinate environment which can be described as a capped
tetrahedral arrangement bonding to 2 Ni atoms, 2 Si atoms and one capping Al atom, see
Figure 3b. The Al atoms lie on a plane and define a perfect square net with Al-Al
distance of 2.955(1)A in the plane for LaNiAl4(Si2,,Nix), slightly longer than that
normally considered a strong bonding interaction, however weak interactions are likely
present. The pure Ni position exhibits a square prismatic (nearly cubic) environment
between two of the Al planar layers bonding to 8 Al atoms, Figure 3c. The rare earth ion
in the structure exhibits bonding (defined as <3.5 A) to 4 Al atoms and 5 Ge atoms in a 9
coordinate environment situated, as mentioned before, between the joint MA14(Si2_,M,)

layers, see Figure 3d.

Charge Transport Properties.

CeNi,,,Al,,Siz_, displays metallic behavior with a measured conductivity value of
about 20,000 S/cm at room temperature. The data also displays trends expected for a
metallic material with a general decrease in the values as temperatures are increased.
When thermopower measurements are conducted normal metallic behavior is again seen
with increases thermopower values with increasing temperatures. The material appears to

be a p-type metal with very small thermopower values of only about 1.5 uV/K at room

temperature.

57

 

ORE.M 0 Al. Si

of REMA1,(Si,_,M,) viewed down the b axis.

Figure 3-1. Structure

58

Position

r
e
d
r
o
.B
D

 

.M OAIo Si

 

Figure 3-2. The MAl4(Si2,,M,) layer viewed down the c axis.

59

 

©RE.MOAIO Si

Figure 3-3. Coordination environments of each atomic position in REMA14(Siz,,M,)

60

._L
O
O)

 

Conductivity (S/cm)
5;.

 

 

111L111 1

11111111411111 11111111111
50 100 150 200 250 300 350
Temperature (K)

.5
o-h
O

 

 

1.8
1.6

IflIII

1.4}
1.25—
15-
0.8”—
0.6 E.

Therrnopower (uV/K)

0.4 :-

.0

.0
02 LllLJllJLllllllllelllll
0

50 100 150 200 250 300
Temperature (K)

 

 

 

Figure 3-4. Conductivity (top) and thermopower (bottom) behavior of CeNiAl4(Si2,,Ni,).

61

Magnetic Properties.

CeNiA14(Siz_xNi,) shows a very complex magnetic behavior without exhibiting
Curie-Weiss at any temperature. At low temperatures a ferromagnetic transition is seen
occurring at about 20 K while at high temperatures a second weak transition is seen at
about 200 K significantly decreasing the magnetization with increasing temperatures.
This behavior is expressed by the dramatic change in the slope of the inverse
susceptibility plot and may be due to valence fluctuations in Ce over the measured
temperature range.9 This behavior is very similar to the behavior seen in the compound
CezNiAlg,Ge4,y (x=0.24 y=1.34).9 LaNiAl4(Siz,,Nix). The formal valency of the Ce atom
in CezNiAlngegyis likely 3+ while the Ni ions are probably in a diamagnetic state as has
been seen in other compounds described in this thesis. When the field dependence was
studied at 5 K a gradual strengthening of the moment is seen with increasing fields with
the absence of hysteresis, see Figure 3-5. No signs of magnetic saturation occurred
however before 55000 G. Studies conducted on LaNiAl4(Siz,,Ni,) showed that the phase
is Pauli-paramagnetic though a small amount of an unknown impurity was present in the

samples displaying molar susceptibilities on the order of 10‘5 emu per mole.

4. Conclusions.

A new series of quaternary aluminum tetrelides of the general formula
REMA14Si2, crystallizing in the KCu,,S3 structure type, can be accessed readily from
molten aluminum solutions. These results underscore the great potential of molten Al as a

reaction medium for accessing novel Al-containing multinary tetrelides. The structure is

62

of interest due to the existence of the "MAl4Siz" subunits in the structure that is seen in
many compounds presented in this thesis. This structural series represents the direct
bonding version between the layers. The series can also be formed with many different
transition metals replacing the Ni ions including Cu, Co, and Pd. These results could
signify that a large diversity in composition and structure based on the NiAl4Tt2 (Tt= Si
or Ge) unit may indeed be form with the other transition metals as well. The
CeNil+,,Al,,Si2_x analog is metallic and shows complex magnetic behavior likely caused by

mixed valency in the Ce atoms over the measured temperature range.

63

10

 

X
(emu / mole) '

.1

h
"1:3...
I ..
- .0. Q
C
r C
.

0LllllAllillllllLLAJ_Ll4lJ_ll+l;

0 50 100 150 200 250 300
Temperature (K)

 

 

 

100

 

T

80 '

60

1/x
(mole / emu)

I
.

40L

20-5,‘
.
OVJIIILLLILIARIIIAAlAlAAAIIAAJAlLALL

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

 

 

 

 

0.8
0.6
0.4 'r
0.2

0 I—
-0.2
-0.4
-0.6

_0.8"...lln...1...113L1.1..1.l....
-1.510‘ -110‘ -5000 0 5000 110‘ 1510‘

Magnetic Field (G)

I'V'I

(B.M. / mole)
\

Magnetization

it

Y‘ITII '1‘ II

 

 

 

Figure 3-5. Magnetic behavior of CeNiAl,(Si2_xNi,)

64

References

 

‘ Ruedorff, W.; Schwarz, H.G.; Walter M. ZAACA 1952 269, 141-152

2 Sheldrick, G.M. 1995, SHELXL. Structural Determination Programs, Version 5.0.

Siemens Analytical X-ray Instruments Inc., Madison, WI.

3 SAINT, version 4, Siemens Analytical X-ray Instruments Inc., Madison WI.

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

5 Marcy, H.O.; Marks, T.J.; Kannewurf, C.R. IEEE Trans Instrum. Meas. 1990 39, 756-
760.

° Grin, Yu. N .; Yarmolyuk, Ya.P.; Rozhdestvenskaya I.V.; Gladyshevskii, E.I. Sov. Phys.

Crystallogr., 1982 27, 418-419.

7 Chen, X.Z.; Sportouch, 8.; Sieve, B.; Brazis, P.; Kannewurf, C.R.; Cowen, J. A.;

Patschke, R; Kanatzidis, M.G. Chem. Mater, 1998 10, 3202-3211.

3 Sieve, B.; Trikalitis, P.N.; Kanatzidis, M.G. Z. Anorg. Allg. Chemie 2002 628, 1568-

1574

‘0 Kwei, G.H.; Lawrence, J .M.; Canfield, P.C. Phys. Rev. B, 1994 49, 14708-1410.

65

Chapter 4.
Routes to the Multinary Aluminum Tetrelides
RE2N1(Ni,,Sil_,,)Al4Si6 (RE = Pr, Nd, Sm, Gd, Dy, Tb) and

Sm2C0(C0,,A11_,,)Al,,Ge6_y

1. Introduction.

After the discovery of several quaternary compounds with early rare earth metals
in the RI-3NiA1,,Si2 systems (see chapter 3) it was decided to continue to explore the
chemistry of similar systems including later rare earth metals. This exploration of the
systems yielded a new structural series REzNi(Ni,,Sil,,,)Al,,Si6 (RE: Pr, Nd, Sm, Gd, Tb,
Dy) along with a later discovery of a Ge analog, Sm2C0(Co,,All,,,)Al,,Ge6,y . The three
dimensional structure of these compounds is best described as the merging of two
structural layers along the c axis including a (N i,Co)AL,Tt2 (Tt= Si or Ge) layer (seen in
chapter 3) and an unusual tetrelide rich layer. This tetrelide layer has been seen in the
literature before in the compound Ce2(Ge0_9Gao_l)7l though appears to be a very rare
structural unit in solid state chemistry.

In this chapter we present the synthesis and Structural characterization of the
RE,Ni(Ni,,Si1,,,)Al,,Si6 (RE: Pr, Sm, Gd, Tb, Dy) system along with SmZCo(Co,Al,_
,)A14Gegy along with preliminary magnetic behavior of the Gd analog, GdzNi(Ni,Sil_
x)Al48i6, is also presented. Early structural work in the series as well as a preliminary

magnetics study of szNi(Ni,Sil,,)Al,Si6 has previously been published by this group.2

66

2. Experimental Section.
Synthesis.
Reagents.

Sm203, 99%, powder, Rhone-Poulenc, Princeton NJ; N i0, 99%, powder, J .T.
Baker Chemical Co., Phillipsburg, NJ, Pr, Nd, Gd, Dy, Tb 99.9%, -40 mesh, Cerac,
Milwaukee, WI; Sm, 99.9%, metal chips, Research Chemicals, Phoenix, AZ; Ni, 99 %,
powder, Aldrich, Milwaukee, WI ; Co 99.8%, -325 mesh, Cerac, Milwaukee, WI; Si,
99.96%, -325 mesh, Cerac, Milwaukee, WI; Al, 20 mesh, Fisher, Fair Lawn, NJ; Ge

99.999%, -325, Milwaukee, WI.

Synthetic Method.

Method 1. The initial synthesis of szNi(Ni,Sil_,)Al4Sié3 was done by Dr. Chen
through reacting $111203, NiO, Si, and Al in different molar ratios with excess of A1 metal
inside an alumina container which was flame sealed in a quartz tube under high vacuum.
The sample was then heated to 800 °C, kept at that temperature for 8 days, and finally

cooled to 300 °C (-5.2°/h).

Method 2. Later all the analogs were formed using the following procedure. In a

Nz-filled dry box, elemental reactants RE:Ni:Al:Si or szCozAlzGe were mixed in a

2:1:20:7 ratio and transferred into an alumina tube which was sealed in a quartz tube as

67

described above. The sample was first heated at 1000 °C for 15 hours, then cooled to 860
°C in 10 hours, kept at 860 °C for 4 days, and finally cooled to 360 °C (-10.4°/h).

In both methods, 5M NaOH (aq.) solution was used to remove excess Al flux
from the product. Purity of the final product was confirmed through comparison of the
experimental X-ray diffraction powder patterns, taken of the bulk product, to theoretical
patterns calculated from the refined single crystal data. Pure phases were obtained for all

analogs using method 2.

Physical Measurements.
EDS Analysis.

Semiquantitative microprobe analysis of the compound was performed with a
JEOL ISM-35C Scanning Electron Microscope (SEM) equipped with a Tracor Northern
Energy Dispersive Spectroscopy (EDS) detector. Data were acquired using an
accelerating voltage of 20 kV and 100 sec accumulation time. The results for
szNi(Ni,,Si,,,,)Al,,Si6 were quite consistent with an average formula of Sm2,oNiA13,5Si3,7.
It was found from analysis of standards at the same time that the composition of Al and
Si was lower than that obtained from X-ray refinement and therefore a correction factors
of 1.14 for Al and 1.82 for Si was applied to the results to yield the final corrected values.

These final corrected values are then very close to the crystallographic refined values.
Single Crystal X-ray Crystallography.

Single crystal X-ray diffraction data of REzNi(Ni,Si,,,)Al,Si6 (RE = Pr, Nd, Sm,

Gd, Dy, Tb) and szCo(Co,Al,,,,)Al,,Ge6_y were collected using a Siemens SMART CCD

68

diffractometer with Mo Koc (A = 0.71073 A) radiation. Absorption correction was applied
to the data using the SADABS program while cell refinement and data processing were
performed with the program SAINT .4 The structure of szNi(Ni,,Si,,,,)Al,,Si6 was first
solved by direct methods within the T EXSAN 5 crystallographic software package. The
structure was refined with the SHELXL° package of programs. The other analogs were
solved in SHELXL bydirect methods and refined using the Sm,Ni(Ni,Si,,,)Al,Si6
solution as starting values. Collection parameters, atomic positions and anisotropic
thermal parameters information is provided in Tables 4-1 through 4-6. Selected bond

distances for szNi(Ni,Sil,,)Al4Si6 analog are shown in Figure 4-7.

Magnetic Characterization.

Magnetic susceptibility for GdzNi(Ni,,Si,,,,)Al,Si6 were measured as a function of
both temperature and field using a MPMS Quantum Design SQUID magnetometer to
explore the magnetic behavior of the system. An initial study of field dependence was
conducted to find a field suitable for the temperature dependence studies. Temperature
dependence measurements on polycrystalline samples were then conducted under
increasing temperature within a 500 G field. Field dependent measurements were

conducted at 5K in fields between i 55000 G.

69

Table 4-1. Crystal data and Structure refinement for REzNi(Ni,Si,,,,)Al,,Si6 (RE = Pr,Sm).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11.)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int)

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Extinction Coefficient
Final R indices [I>20(I)]

R indices (all data)

Extinction Coefficient
Largest diff. peak and hole

R1=£IIF0| - chll/EIFOI, WRZ=IZ(W|F.12 - HIV/ENVY?”

PerilmAhSiW
647.90

293(2) K
0.71073 A
P4/nmm (#129)
a = 5.939(2) A
c = 15.169(8) A

szNimALSim
672.23

298 K

0.71073 A
P4/nmm (#129)
a = 5.8060(3) A
c = 14.845(1) A

535.1(4) A3 500.43(5) A3
2, 4.02 Mg/m3 2, 4.461 Mg/m3
11.7575 mrn'l ' 14.928 mm’1
592 612
0.18x0.18x0.07mm 0.13x0.07x0.13mm
2.69 to 27.73 ° 2.74 to 28.42 °
-7<=h<=7 -7<=h<=5
-7<=k<=7 -7<=k<=7
-19<=l<=l9 -19<=l<=16
4344 / 417 3070 I419
0.0635 0.047
Full-matrix least-squares on F2
417 / 0 / 30 419/
1.107 1.101
0.0011(9) 0.013(1)
R1 = 0.0651 R1 = 0.0252
wR2 = 0.1609 wR2 = 0.0634
R1 = 0.0659 R1 = 0.0324
wR2 = 0.1615 wR2 = 0.0663
0.0011(9) 0.013(1)

4.776 and -2751 e-/A3

70

1.554 and -2.038 e-/A3

Table 4-2. Crystal data and structure refinement for RE.2Ni(Ni,,Si,_,,)Al4Si,5 (RE = Nd, Gd).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique

Rn.

Refinement method

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

Final R indices [I>20(I)]

R indices (all data)

Extinction Coefficient
Largest diff. peak and hole

R1=Z||F°| - |1=,||/2|F,|, wR2=[2(w|F,2 - RAY/Emmi)?“

NdzNilmALSiW
621.66

298.0

0.71073 A

P4/nmm (#129)

a = 5.843(3) A

c = 1492(1) A
509.6(5) A3

2, 4.051 Mg/m’

13.07 mm"

597

0.10 x 0.14 x 0.30 mm
1.36 to 28.19°

Gd,Ni,,,Al,Si,_,,
688.76

298.0

0.71073 A

P4/nmm (#129)

a = 5.8093(1) A

c = 14.9868(2) A
505.774(14) A3

2, 4.523 Mg/m3
16.530 mm'1

622

0.03 x 0.06 x 0.12 mm
2.72 to 28.130

-7<=h<=7 -7<=h<=7
-7<=k<=7 -7<=k<=7
-18<=l<=19 -19<=l<=19
4879/416 5772/415
0.0447 0.0356
Full-matrix least-squares on F2
416/0/31 415/1/31
1.284 2.833
R1=0.0245 R1 = 0.0227
wR2=0.0667 wR2 = 0.0506
R1=0.0254 R1 = 0.0243
wR2=0.0670 wR2 = 0.0507
0.0089(7) 0.0120(7)

0.882 and -l.162 e—/A3

71

1.023 and -0933 e—/A3

Table 4-3. Crystal data and structure refinement for RE,Ni(Ni,,Si,,,)Al,,Si,5 (RE = Tb, Dy).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique

R(int)

Refinement method

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

Final R indices [I>20(I)]

R indices (all data)

Extinction Coefficient

Largest diff. peak and hole

szNil.34Al4Si6.66

D y 2Ni 1.27Al4Si6.73

965.92 696.50
293(2) K 293(2) K
0.71073 A 0.71073 A
P4/nmm (#129) P4/nmm (#129)
a = 5.7349(8) A a = 5.8214(1) A
= 14.774(3) A c = 14.8772(3) A
485.9(1) A3 504.17(2) A3
1.4.757 Mg/m3 2, 4.588 Mg/m3
17.886 mm" 17.635 mm"
722 620
0.12x0.12x0.18mm 0.03x0.03x0.06mm
2.76 to 28.01 deg. 1.37 to 29.95 deg.
-7<=h<=7 -8<=h<=8
-7<=k<=7 -8<=k<=8
-19<=l<=18 ~20<=l<=20
4479 / 397 5748 / 489
0. 0950 0.0655
Full-matrix least-squares on F2
397/0/31 487/1/30
1.125 1.313
R1 = 0.0309 R1 = 0.0471
wR2 = 0.0776 wR2 = 0.1288
R1 = 0.0344 R1 = 0.0524
wR2 = 0.0789 wR2 = 0.1377
0.015(1) 0.006(1)

1.640 and -1339 e—/A3

R1=2||Fo| - IF.l|/2IF.I. wR2=[Z(w|F02 - F,2|)2/z(wF,2)21‘”

72

4.481 and -3.123 c-/A3

Table 4-4. Atomic coordinates ( x10“) and occupancies for REzNi(Ni,Si1,,)Al4Si6
(RE: Pr, Sm, Nd, Gd, Tb, Dy).

 

 

Wyckoff x y 2 Occupancy
position
Pr 4f 7500 2500 2615(1) 1
Sm 7500 2500 2643(1) 1
Nd 7500 2500 2617(1) 1
Gd 7500 2500 2630(1) 1
Tb 7500 2500 2642(1) 1
Dy 7500 2500 2621(1) 1
Ni 2a 7500 2500 0 1
7500 2500 0 1
7500 2500 0 ' 1
7500 2500 0 1
7500 2500 0 1
7500 2500 0 1
Ni(1) 2c 7500 -2500 3587(6) 006(3)
7500 -2500 3587(2) 027(1)
7500 -2500 3570(3) 006(2)
7500 -2500 3538(2) 036(1)
7500 -2500 3458(1) 034(2)
7500 -2500 3554(4) 0.27(2)

 

73

Table 44. (continued) atomic coordinates ( x10“) and occupancies for

R1=.,Ni(Ni,,Si,_,)Al,Si6 (RE: Pr, Sm, Nd, Gd, Tb, Dy).

 

 

Wyckoff x y 2 Occupancy
position

Si(l) 2c 7500 -2500 3587(6) 094(3)
7500 -2500 3587(2) 073(1)
7500 -2500 3570(3) 094(2)
7500 -2500 3538(2) 063(1)
7500 ~2500 3664(14) 065(2)
7500 -2500 3554(4) 0.73(2)

Si(2) 2c 7500 7500 1953(6) 1
7500 7500 1964(2) 1
7500 7500 1917(3) 1
7500 7500 1952(2) 1
7500 7500 1968(2) 1
7500 7500 1953(3) 1

Si(3) 2c 2500 2500 2008(6) 1
2500 2500 2059(2) 1
2500 2500 2023(3) l
2500 2500 2055(2) l
2500 2500 2076(2) l
2500 2500 ‘ 2051(3) 1

 

74

Table 4-4. (continued) atomic coordinates ( x10“) and occupancies for

RE,Ni(Ni,Si1_,)Al,Si6 (RE: Pr, Sm, Nd, Gd, Tb, Dy).

 

 

Wyckoff x y 2 Occupancy
position
Si(4) 8j 4599(7) 406(7) 4225(3) 1
4611(3) 389(3) 4218(1) 1
4603(3) 397(3) 4212(2) 1
4640(3) 360(3) 4205(1) 1
4646(4) 354(4) 4207(2) 1
4599(5) 401(5) 4208(2) 1
Al 8j 4956(5) 44(5) 867(3) 1
4959(2) 41(2) 891(1) 1
4950(3) 50(3) 877(1) 1
4957(2) 43(2) 887(1) 1
4959(1) 41(1) 893(2) 1
4948(3) 52(3) 886(2) 1

 

75

Table 4-5. Anisotropic displacement parameters (A2 x 103) for

RE,Ni(Ni,Si,_,)A1,Si, (RE: Pr, Sm, Nd, Gd, Tb, Dy).

 

U11

U22

U33

U23

U13

U12

 

Pr

Sm
Nd
Gd
Tb

Ni

Ni(1)

5(1)
5(1)
8(1)
3(1)
8(1)
9(1)

6(1)
5(1)
7(1)
2(1)
7(1)
3(1)

11(3)
8(3)

13(1)
12(1)
16(1)
14(2)

5(1)
5(1)
8(1)
3(1)
8(1)
9(1)

6(1)
5(1)
7(1)
2(1)
7(1)
3(1)

11(3)
8(3)

13(1)
12(1)
16(1)
14(2)

12(1)
12(1)
8(1)
5(1)
3(1)
9(1)

10(2)
12(2)
8(1)
4(1)
0(1)
2(1)

21(4)
23(4)
26(2)
34(2)
2(8)

30(3)

CCOCCC CCOCCC

CCOCCC

CCOCCC CCOCCC

CCOCCC

CCOCCC CCOCCC

CCOCCC

 

The anisotropic displacement factor exponent takes the form: -2n2[h2a*2U11 + +

2hka*b*U12]

76

Table 4-5. (continued) anisotropic displacement parameters (A2 x 103) for
RE,Ni(Ni,Si,,,)Al,Si, (RE: Pr, Sm, Nd, Gd, Tb, Dy).

 

U11 U22 U33 U23 U13 U12

 

Si(l)

Si(2)

Si(3)

11(3)
8(3)

13(1)
12(1)
16(1)
14(2)

7(3)
7(2)
8(1)
4(1)
10(1)
4(1)

7(3)
6(2)
9(1)
6(1)
9(1)
5(1)

11(3)
8(3)

13(1)
12(1)
16(1)
14(2)

7(3)
7(2)
8(1)
4(1)
10(1)
4(1)

7(3)
6(2)
9(1)
6(1)
9(1)
5(1)

21(4)
23(4)
26(2)
34(2)
2(8)

30(3)

15(4)
12(4)
9(2)
5(1)
1(2)
0(2)

16(4)
15(4)
7(2)
4(2)
3(2)
1(2)

CCOCCC CCOCCC

CCOCCC

CCOCCC CCOCCC

CCOCCC

CCOCCC

CCOCCC

CCOCCC

 

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +

2hka*b*U12]

77

Table 4-5. (continued) anisotropic displacement parameters (A2 x 103) for

RE,Ni(Ni,Si,_,)Al,Si, (RE: Pr, Sm, Nd, Gd, Tb, Dy).

 

U11 U22 U33 U23 U13 U12

 

Si(4) 16(1) 16(1) 24(2) 1(1) -l(l) 0(2)
16(1) 16(1) 26(2) 1(1) -l(1) 0(2)
23(1) 23(1) 23(1) 2(1) 2(1) -6(1)
32(1) 32(1) 15(1) 0(1) 0(1) -16(1)
38(1) 38(1) 10(1) -l(1) 1(1) -18(1)
24(1) 24(1) 16(2) 1(1) -1(1) -8(2)

Al 6(2) 6(2) 13(2) 0(1) 0(1) 1(2)
6(1) 6(1) 13(2) 1(1) -1(1) 1(2)
9(1) 9(1) 7(1) 0(1) 0(1) -l(l)
3(1) 3(1) 5(1) 1(1) -1(1) 0(1)
8(1) 8(1) 1(1) 0(1) 0(1) 0(1)
6(1) 6(1) 1(1) 0(1) 0(1) 0(1)

 

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +

2hka*b*U12]

78

—=::::;_ -._. -. _ ..

Table 4-6. Selected bond distances (A) for RE2N1(NixSi1_x)Al4Si6 (RE: Pr, Sm, Nd, Gd,

 

 

Tb, Dy).
Bond Bond distances (A) Bond Bond distances (A)
Pr-Si(1)/N 1(1) 3.312(4) Ni-Al 2.481(2)
Sm 3.223(1) 2.442(9)
Nd 3.052(2) 2.445(2)
Gd 3.0300(8) 2.4471(9)
Tb 3.1 1(1) 2.423(3)
Dy 3.225(3) 2.445(2)
RE-Si(2) 3.156(3) Si(l)/Ni(l)-Si(2) 2.535(12)
* 3.073(1) 2403(4)
3.102(2) 2.468(6)
3.207(1) 2.376(4)
2.987(1) 249(4)
3.0760(2) 2.383(6)
RE-Si(3) 3.108(3) Si(1)/Ni(1)-Si(4) 2.631(7)
3.029(9) 2.551(2)
3.053(2) 2.578(4) ,
3.077(1) 2.554(3)
3.036(1) 2.45(1)
3.032(2) 2.579(5)
RE- Si(4) 3.235(4) Si(2)-Al 2.658(7)
3.128(1) 2.625(2)
3.170(2) 2.618(3)
3.143(2) 2.629(3) -
3.087(2) 2.597(3)
3.149(3) 2.632(3)
RE-Al 3.378(4) Si(3)-A1 2.692(7)
3.314(1) 2.662(3)
3.320(2) 2.651(4)
3.323(1) 2.671(3)
3.280(2) 2.645(3)
3.301(2) 2.658(5)

 

79

Table 4-6. (continued) selected bond distances (A) for REzNi(NixSil-x)Al4Si6 (RE: Pr,

Sm, Nd, Gd, Tb, Dy).

 

 

Bond Bond distances (A) Bond Bond distances (A)
Si(4)-Si(4) 2.460(9) Al-Al 2.916(6)
2.407(4) 2.856(2)
2.442(5) 2.863(3)
2.545(4) 2.855(2)
2.415(5) 2.819(2)
2.444(6) 2.850(3)
Si(4)-Si(4) 2.481(8) Al-Al 2.916(6)
2.451(3) 2.950(2)
2.458(4) 2.980(3)
2.486(3) 2.955(2)
2.463(4) 2.915(2)
2.447(7) 2.971(3)
Al-Al 2.643(9)
2.647(3)
2.617(4)
2.661(3)
2.652(5)
2.639(6)

 

80

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11.)
F(000)

Crystal Size

0 range

Limiting indices

Ref. collected / unique
R(int)

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>20(I)]

R indices (all data)

Extinction Coefficient

Largest diff. peak and hole

Table 4-7. Crystal data and structure refinement for szCo(Co,Al,.x)Al4Gegy.

szCo,.ozAlmGe,53
895.97

293(2) K

0.71073 A

P4/nmm (#129)

a = 5.8835(5) A

c =15.0394(19)A
520.60(9) A3

2, 5.716 Mg/m3
32.564 mm‘1

852

0.03 x 0.18 x0.18
2.71 to 27.95°
-7<=h<=7 ,
-7<=k<=7
-l9<=l<=19

4415 / 399

0.0904

Full-matrix least-squares on F2
399/0/30

1.073

R1 = 0.0401

wR2 = 0.1126

R1 = 0.0474

wR2 = 0.1145
0.0085(11)

5.765 and -5.208 e/A3

' R1=2||Fo| - ch||IZ|FO|, wR2=[§.".(w|F02 - F62|)"’/ZZ(wF02)2]1.’2

81

Table 4-8. Atomic coordinates ( x104) and occupancies for szCo(CoxA11_,)Al4Ge6_y.

 

 

Wyckoff x y 2 Occupancy
position
Sm 4f 7500 2500 2579(1) 1
Co 2a 7500 2500 0 1
Co(1) 2c 7500 -2500 3578(4) 0.02(2)
Al(l) 2c 7500 -2500 3578(4) 098(2)
Ge(2) 2c 7500 7500 1953 088(1)
Ge(3) 2c 2500 2500 2081(2) l
Ge(4) 8j 4625(2) 375(2) 4183(1) 1
All 8j 4945(4) 55(4) 856(2) 1

 

Table 4-9. Anisotropic displacement parameters (A2 x 103) for Sm2Co(Co,All,x)Al,Ge6,y.

 

 

U11 U22 U33 U23 U13 U12

Sm 9(1) 9(1) 5(1) 0 0 0
Co 7(1) 7(1) 2(2) 0 0 0
Co(l) 8(3) 8(3) 2(3) 0 0 O
Al(l) 8(3) 8(3) 2(3) 0 O 0
Ge(2) 10(1) 10(1) 5(1) I 0 0 0
Ge(3) 10(1) 10(1) 5(1) 0 0 O
Ge(4) 20(1) 20(1) 12(1) -1(1) 1(1) -2(1)
Al 9(1) 9(1) 2(2) 0(1) 0(1) 1(1)

 

The anisotropic displacement factor exponent takes the form: -2n2[h2a*2U11 + +

2hka*b*U12]

82

Table 410. Selected bond distances (A) for Sm,Co(Co,Al,,,)Al,Ge,,,

 

Bond Bond distances (A)
Sm-Ge(3) 3.0355(7)
Sm-Ge(2) 3.0889(4)
Sm—Ge(4) 3.2005(13)
Sm—Al(2) 3.303(3)
Co-Al(1) 2.4464( 17)
Al(2)-Ge(2) 2.444(6)
Al(2)-Ge(4) 2.560(3)
Al(2)-Sm 3.303(3)
Ge(2).-Co(2) 2.444(6)
Ge(2)-Al(l) 2.691(3)
Ge(3)-Al(1) 2.745(4)
Ge(4)-Ge(4) 2.500(3)
Al(1)-Al(1) 2.575(7)
Al(1)-Al( 1) 2.876(4)
Al(1)-Sm 3.324(3)

 

83

3. Results and Discussion.
Synthesis and Thermal Behavior.

To obtain pure REZM(M,,Si,_,,)Al,,Tt6 (RE: Pr, Sm, Nd, Gd, Tb, Dy; M: Ni, Co;
and Tt= Si, Ge) it is necessary to react stiochemeteric ratios of RE, M and Tt in excess Al
for several days. The use of stiochemeteric or slight excess of Tt reactants is important to
avoid the synthesis of RENiAl4(Si2,,Ni,) type compounds, particularly in the earlier RE
metals Pr and Nd where both structure types can form readily, This reaction produces
large single plate-like crystals of the compounds of up to several mm in size. These
compounds are stabile in air, water, and N aOH but are destroyed immediately upon
contact of dilute HCl. This decomposition in acid releases silane gasses that explodes on
contact with air. For this reason these systems should never be isolated from excess Al
using acids. The compounds Show no signs of decomposition at normal temperatures
over several months though at exposure elevated oxygen appears to attack the material

over a period of 48 hours.

Structural Description.

For clarity in describing the structural arrangement the discussion will highlight
the Si analogs, differences in the Ge analog will be noted and discussed at the end of the
section. The three-dimensional structure of RE2N1(NixSi1-x)Al4Si6 is shown in Figure 4-
1. The compound crystallizes in a new structure type within the P4/nmm (No. 129) space
group. 81(1) and Ni(1) are disordered on the same site with 77% and 23% occupancies,

respectively, based on the SmgNi(NixSi1-x)Alasi6 crystal selected. Why the observed

narrow range of Ni/Si ratios in szNi(NixSi1-x)Al4Si6 is preferred is an interesting issue

84

which may have to do with the stability of the entire structure. The other RE analogs
however Show a wide variations in x, ranging from only a few percent in the Pr and Nd
cases to over 30 % Ni occupancy in the analogs with smaller RE ions. The
szCo(Co,,A1,_,,)Al,,Ge6_y analog does not Show M/Tt disorder instead a mixed Al/Co
occupancy in the position, largely occupied by Al (98% Al). This site is the only one
exhibiting disorder in the analogs and is noted as such in the structural figures.

The 3 dimensional structure contains two different types of alternating layers
which are linked along the c-axis though bonding of the Tt(2) to disorder site. The first
layer, a MAl,,Tt2 layer,7 is five atomic layers thick with a stacking sequence
Tt/Al/M/Al/Tt. The M atoms occupy the centers of distorted Al cubes, see Figure 3(b).
Each of these NiAlg cubes share its four parallel edges with four other NiAlg cubes to
make an infinite layer, see Figures 4-2 and 4-3. This infinite layer is seen in many of the
compounds present in this work and appears to be a very stable structural unit in
compounds synthesized from excess molten A] (see chapters 3,5 and 6). This type of
NiAlg cubic coordination is exhibited by several compounds reported in the literature
including NiAl which has a CSCl-type structure.8 In szNi(NixSi1-x)Al4Si6 the cube
dimensions are 2.647 x 2.856 x 2.950 A, while those in MM are 2.88 x 2.88 x 2.88 A.
Therefore, there is Al-Al bonding present within the layers in REzNi(NixSi1-x)Al4Si6.

The second type of layer in RE2M(MxSil-x)Al4Tt6 in a largely tetrelide
layer which also includes the disorder position. In the Si analogs the layer is formed by
the disordered position and Si(4) atoms (Figure 4-2 and 4—4). This Si-based layer is two
atomic layers thick and consists of 4-member planar rings formed by Si(4) atoms [Si(4)-

Si(4) bond 2.450 A], 5-member rings formed by four Si(4) and one Si( l)/Ni(1) atoms,

85

Unit:. 3.. ”-
3.9."

   
 
  
   
    

. 0 RE
09.0.. 00‘ o M
.' 0 Al

Figure 4-1. The structure of RE2M(Mth1_x)Al4Tt6. Note the MAI8 cube layers at the
top and bottom of the cell. In the middle there is a layer formed by Tt(4) atoms and the
disorder position. The two different layers are connected through bonds between Tt(2)

and the disorder position. N 0 bonds are drawn around RE atoms.

86

O M
0 Al
O Tt

 

Figure 4—2. The MAI,3 cube layer in ab plane of RE2M(Mth1_x)A14Tt6. Note that each
MAlg cube only connects four (not six) other MAIg cubes by sharing the four paralleled

edges.

87

 

iii

 

Figure 4-3. (a) The double "Tt"—net and (b) Half of the net which is obtained by

cutting the one in (a) at the middle in ab plane.

88

Si(1)/Ni(1)

Si(2)

  
 

Si(1)/Ni(1) Si(4)

Figure 4-4. The local coordination environments for different atoms in szNi(NixSi1_

x)Al4Si5.

89

and 6-member rings (boat—like) formed by four Si(4) and two Si(1)/Ni(l) atoms [Si(4)-
Si(l)/Ni(1) bond 2.564 A]. The square and hexagonal rings can be seen clearly in Figure
4-3(b) while S-member rings are better viewed from Figure 4-2 and Figure 4-4(a). Si
hexagonal nets are known in several Silicides, for example, CaSi2 and B-USiz both
contain hexagonal Si rings.

In the Ge analog, the introduction of Ge for Si atoms in the tetrelide layer appears
to cause vacancies on the Tt(2) site with the occupancy refining to only 88%. Along with
the change in occupancies a replacement of Al for the disorder Tt atom occurs in the
disorder position yielding Al/Co disorder in the structure instead of the Tt/Co disorder
seen in the Si analogs. This substitution of A1 onto the disorder position and the vacancies
introduced onto the Ge sites may be important features of the structure allowing the Ge
analog to form, a unique feature to this system when the REzNi ratio is greater than 1.
The CoA1,,Ge2 layer exhibits no changes other than simple substitution of elements for the
Ge compound.

The different coordination environments for the three different pure Si/Ge atoms
in RE2M(Mth1-x)A14Tt6 are shown in Figure 5. Tt(2) exhibits a square pyramidal
arrangement bonding to 4 Al atoms and the disorder position. Tt(3) coordination similar
to Tt(2) however the apex atom has been removed though still bonding to 4 Al atoms, see
Figure 4-5, while Tt(4) exhibits a distorted tetrahedral site bonding to 3 other Tt atoms
and the disorder position. The disorder position itself exhibits a square pyramidal
arrangement with 4 Tt(4) atoms, each within the tetrelide layer, as the base and a Tt(2)
atom, in the MAl,,Tt2 layer, at the apex. The transition metal (M) atom exhibits a distorted

cubic environment bonding to 8 Al atoms. The Al atoms bond to 2 M atoms, 2 Tt atoms,

90

and 3 Al atoms however 2 other A1 atoms are at 2.96 A likely exhibiting weak interaction
with the Al atom. Finally the RE ion exhibits bonding to 2 atoms in the disorder position,

10 Tt atoms and 4 Al atoms, see Figure 4—4.

Magnetic Properties.

The temperature dependent magnetic susceptibility for GdzNi(NixSi1_x)A14816
exhibits a antiferromagnetic interaction at low temperatures with a Tmax near 10 K. A fit
to the high temperature data, corrected for TIP, gives a lief, value of 8.37 B.M. for the
effective moment which is close to the theoretical value for Gd“ of 7.94 B.M. There
appears to be no magnetic moment on the nickel atom as has also been observed in many
RENiTt2 compounds (RE = rare earth metal, Tt = Si, Ge) which were studied by neutron
diffraction9 as well as many of the phases present in this work. From the linear fit the
Weiss constant of 24 K is found indicating the weak ferromagnetic interaction are present
even at high temperatures. When the field dependence of GdzNi(NixSil-x)Al4Si5 was
measured a nearly linear increase with increasing fields is seen up to the maximum field
measured of 55,000 G. No signs of magnetic saturation is seen at maximum fields as

expected since only 6 B.M. have been accounted for.

4. Conclusions.

A series of quaternary aluminum tetrelides, RE,Ni(Ni,Si,_,,)Al,,Si6 (RE = Pr, Nd,
Sm, Gd, Dy, Tb) and Sm2C0(Co,A11,,)Al,Gegy, has been synthesized from molten
aluminum. The compounds exhibit a new structure type comprised of alternating layers

of MAl4Tt2 and an unusual tetrelide unit stacking along the c axis. The compound is of

91

special interest due to several structural features exhibited including the presence of
MA1,,Tt2 layers (also seen in several other chapters of this work, see chapters 3 through
6), the formation of an unusual tetrelide layer, existence of Ge/Si analog structures, and
the presence of disorder in the tetrelide position. Magnetically the GdzNi(NixSi1-x)A14$i6
analog exhibits an antiferromagnetic transition at low temperatures and Curie-Weiss
behavior at high temperatures. Fitting the data at high temperatures one can assign a
charge of 3+ to the oxidation state of the Gd ion while the M ions are magnetically silent

as has been the trend in many of the compounds recently synthesized from molten A1.

92

 

m

X
(emu / mole RE ion)

1/x

(mole RE ion / emu)

nefizafion

Map
(B.M. mole RE ion)

0.8 _

(17
(16
(15
031
013
(12
OJ

35

30

25

20

15

1O

-2

-4

-6

-610‘

-..—-_- _2 “.24. .1". _ _

 

I'IT

E

)-
b

IIYTII’IIIIIIII

IIII'

TII

 

1 1 J L 41 J L l l l 1 1 .l 1 l l l l A

 

 

20 30
Temperature (K)

50

 

50 1

llllLL

 

l L A A n l n
150 200
Temperature (K)

lLlLLLLJLL

250 300

1 AL

00 350

 

ITIfI’YT“

TjT‘I

 

-4 10‘

 

1 l 4 l l.

410‘ 610‘

ALLJ

210‘

-2 10‘ 0
Field (G)

A Figure 4-5. Magnetic behavior of GdzNi(Ni,Si ,_,)A14Si6 (top) susceptibility as a function

of temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

93

 

——— i-“ A -75vmf‘. I 7 .

References

 

1 Yarmolyuk, Y.P.; Pecharskii, V.K.; Gryniv, I.A.; Bodak, O.I.; Zavodnik, V.E. Sov.
Phy. Crystallogr. 1989, 34(2), 174

2 Chen, X.Z.; Sportouch, S.; Sieve, B.; Brazis, P.; Kannewurf, C.R.; Cowen, J. A.;
Patschke, R; Kanatzidis, M.G. Chem. Mater, 1998 10, 3202-3211.

3 Chen, X.Z.; Sportouch, S.; Sieve, B.; Brazis, P.; Kannewurf, C.R.; Cowen, J. A.;
Patschke, R; Kanatzidis, M.G. Chem. Mater, 1998 10, 3202-3211.

4 SAINT, version 4; Siemens Analytical X-ray Instruments Inc., Madison, WI.

5 "TEXSAN -- TEXRAY Structure Analysis Package", Molecular Structure Corporation,
The Woodlands, TX, 1992.

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

7‘ This same building unit has been seen as a Ga and In analogs in compounds such as
LaGaéNim, HoCoGa, and DyCoIn, though the Al analog appears to have been absent
from the literature until the publication of several compounds presented in this thesis
including szNi(Si1_x,Nix)Al4Si6, szNiAl4GeL and CezNiAlngegy.

8 Dutchak Ya.I. and Chekh V.G., Russian Journal of Physical Chemistry, Translated
From Zhurnal Fizicheskoi Khimii 1981 55(9), 1326-1328.

9 Gil, A.; Szytula, A.; Tomkowicz, Z.; Wojciechowski, K. J. Magn. Magn. Mater. 1994

129, 271-278 and references 7 and 8 therein.

94

Chapter 5.
Synthesis and Characterization of the Compounds
REzNiAl4Ge2 (RE = Gd, Tb, Dy, Er) and RE2C0A14Ge2 (RE =

Sm, Gd, Th)

1. Introduction.

During attempts to grow large crystals of TbNiA14Ge2 (see chapter 7), utilizing
molten metal fluxes, a new compound szNiA14Ge2 was synthesized. After this discovery
other rare earth and transition metal systems were explored and a new structural series
was established RE2N1A14Ge2 (RE = Gd, Tb, Dy, Er) and RE2C0A14Ge2 (RE = Sm, Gd).
These compounds exhibit a new structural arrangement which is best described as
stacking of (Ni,Co)Al,,Ge2 layers, previously seen as a Si analog in other compounds such
as szNi(Ni,,Si,,,,)Al,,Si6 and RENiAl4(Si2_,Ni,)(chapterS 3 and 4). The structure is very
closely related to that of RENiAl4(Si2_,Ni,) (see chapter 3) however no direct bonding
between the (Ni,Co)Al,,Ge2 layers is seen.

These phases are of particular interest due to both the existence of the Ge form of
the MAl4Tt2 (Tt= Si or Ge) unit along with the compounds ability to be formed through
either flux methods or more traditional arc-melting approaches. This ability to form
through both synthetic methods appears to be highly unusual this and related phases
which has been synthesized in molten Al. Our research group highlighted the structure

and magnetic properties of the szNiA14Ge2 analog in the first paper released highlighting

95

these systems.1 Here we present the synthesis and characterization of the broader series of

compounds R1~:,MAI,Ge2 (RE = Gd, Tb, Dy, Er) and R13,CoAl,Ge2 (RE = Sm, Gd).

2. Experimental Section.

Synthesis.

Reagents.

The following reagents were used as received: Sm, 99.9%, metal chips, Research
Chemicals, Phoenix, AZ; Gd, Tb, Dy, Er, Tm, Lu, Y 99.9%, -40 mesh, Cerac,
Milwaukee, WI; Ni, 99%, 325 mesh, Sargent, Buffalo Grove, 11; Co 99.8%, -325 mesh,
Cerac, Milwaukee, WI; A1, 99.5 %, -20 mesh, Milwaukee, WI; Ge 99.999%, -325,

Milwaukee, WI.

Synthetic Methods.

Method 1. In a N2 filled glove box RE, M, Al, and Ge powders were mixed in a
2: 1:10:2 ratio for RE = Gd, Tb, Dy, Er while M = Ni and RE = Sm and Gd while M = C0.
The mixture was then placed into an alumina crucible and sealed within a fused silica
tube under vacuum (<1x10“‘ Torr). The tubes were heated to 1000° C over a period of 24
hours and kept there for 48 hours. They were then slowly cooled to 500° C over 48 hours
followed by cooling to 50° C in 12 hours. The alumina tubes were then removed from the
silica and submerged in 5M NaOH for 24 hours to remove the excess Al matrix. This
procedure yielded two types of products, silvery plate-like crystals and black
microcrystalline powder both of which were shown to be the same phase, by comparison

of experimental powder patterns to patterns generated from the solved crystal structure.

96

The szNiAl,,Ge2 reaction shows yields greater than 85%, based on Tb, with greater than

95% purity.

Method 2. In a N2 filled glove box RE, Ni, Al, and Ge powders were mixed in a
2:1: 10:2 ratio for RE 2 Tb and Er. This mixture was pressed and loaded into an arc
wielder and melted, under an Ar atmosphere, for approximately 30 seconds until a good
melt was observed. The samples were then flipped and remelted several times to ensure
homogeneity in the overall sample. After cooling, crystals were observed on the surface
of the ingot imbedded in the excess Al. The Al was removed by submersion in 5M N aOH
for 24 hours. After the isolation was complete an X-ray powder diffraction pattern was
recorded which matched well with the calculated pattern. In the szNiAl,,Ge2 reaction
yields were on the order of 80% based on the amount of Tb used. Attempts to synthesize
the Tb analog without excess Al to date do not produce pure product but instead mixtures

of phases including TbNiAl,,Ge23 and TbZNiA14Ge2.

Physical Measurements.
EDS Analysis.

Quantitative microprobe analysis of RE,MA1,,Ge2 was performed with a JEOL
JSM-6400 Scanning Electron Microscope (SEM) equipped with Noran Energy
Dispersive Spectroscopy (EDS) detector. Data were acquired using an accelerating
voltage of 25 kV and 100 sec accumulation time. Standards where recorded under the

same experimental conditions to yield correction factors. After calibration szNiAl,,Ge2

97

gas:

1"" I I
dial

”7‘10

hair“)
I

v
n

«L:

gave an elemental ratio of 1.99 Tb: 1 Ni: 4.01 Al: 1.95 Ge, in good agreement with the

actual formula.

' Single Crystal X-ray Crystallography.

Single crystal X-ray diffraction data for RE2N1A14Ge2 (RE 2 Gd, Tb, Dy, Br) and
GdZCoAl4Ge2 were collected at 298 K on a Siemens Platform CCD diffractometer using
Mo K0t (1:0.71069 A) radiation. The SMART software2 was used for the data
acquisition and the program SAINT3 was used for the data extraction and reduction. An
empirical absorption correction using SADABS4 was applied to the data. These structures
were solved and refined with the SHELXL package of programs.’ The crystallographic
and refinement details are listed in Table 5-1 through 5—16.

Single crystal X-ray diffraction data of szCoAl4Ge2were collected at room
temperature using a Rigaku 4 circle diffractometer with Mo Ka (A = 0.71073 A)
radiation. An empirical absorption correction based on ‘1’ scans was applied to the data.
The structure was solved by direct methods and refined with the SHELXL package of
programs. Data collection parameters, atomic positions, anisotropic thermal parameters

information is provided in Tables 5-12 and 5-13.

Magnetic Characterization.

Magnetic susceptibility for single crystals of szN iA1,,Ge2 was measured as a
function of both temperature and field using a MPMS Quantum Design SQUID
magnetometer both parallel and perpendicular to the c-axis. An initial study of field

dependence was conducted to find a suitable field for the variable temperature studies.

98

Table 5-1. Crystal data and structure refinement for REzNiA14Ge2 (RE = Tb, Gd).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)I

R indices (all data)

Extinction coefficient
Largest diff. peak and hole

R1=X||Fo| - |Fc||/Z|Fo|, wR2=[2(w|F02 - Fc2|)2/2(wF02)2]“2

Tb,NiAl,Ge2 Gd,NiA1,,Ge2
626.69 626.29
293(2) K 293(2) K
0.71073 A 0.71073 A
I4/mmm (#139) I4/mmm (#139)
a=4.l346(2)A a=4.1557(1)A
c = 19.3437(7) A c = 19.4446(5) A
330.68(3) A3 335.805(14) A3
2, 6.29 Mg/m3 2, 6.19 Mg/m3
33.00 m" 31.14 mm"
548 544
0.12 x 0.15 x 0.24 mm 0.20 x 0.08 x 0.02 mm
2.11 to 28.24 ° 2.09 to 28.22°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-25<=l<=20 -25<=l<=25
1413/ 158 1701 / 162
0.0683 0.0651
98.8 % 99.4 %
Full-matrix least-squares on F2
158/0/14 162/0/14
1.093 1.150
R1 = 0.0252 R1 = 0.0393
wR2 = 0.0698 wR2 = 0.0866
R1 = 0.0266 R1 = 0.0407
wR2 = 0.0703 wR2 = 0.0875
0.0160(15) 0.041(4)

1.819 and -2.182 e.A'3

99

2.189 and -5.162 e.A'3

Table 5-2. Crystal data and structure refinement for REzNiAl4Ge2 (RE = Dy, Er).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (1t)
F(000)

Crystal size

0 range

Limiting indices

Ref.‘collected / unique
R(int)
Completeness to 9m

Refinement method

Data I restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=EIIF.I - Ian/21m, wR2=[Z(WIF.2 - F.2|)2/2(wF.2)21'°

DyzNiAl4Ge2
636.89

293(2) K
0.71073 A
I4/mmm (#139)
a = 4.127(1) A
c = l9.331(7) A
329.2(2) A3

2, 6.478 Mg/m3
34.65 mm“1
552

0.04 x 0.03 x 0.01 mm
2.11 to 24.98
-4<=h<=4
-4<=k<=3
-22<=l<=22
930/ 115
0.0506

100.0%

EeriAl,,Ge2
646.41

293(2) K
0.71073 A
I4/mmm (#139)
a = 4.1067(4) A
c = 19.231(3) A
324.33(6) A3

2, 6.618 Mg/m3
38.01 mm“1

560

0.06 x 0.06 x 0.03 mm
4.24 to 28.36°
-5<=h<=5
-5<=k<=5
-22<=l<=24
1495/ 154
0.0602

97.5 %

Full-matrix least—squares on F2

115/0/14
1.224

R1 = 0.0267

wR2 = 0.0686

R1 = 0.0284

wR2 = 0.0697
0.0086(12)

2.900 and -1.171 e.A'3

100

154 / 0 / 14
1.248

R1 = 0.0312

wR2 = 0.0786

R1 = 0.0320

wR2 = 0.0791
0.021(2)

1.921 and -2.848 e.A'3

Table 5-3. Atomic coordinates ( x104) for RE,NiAl,,Ge2 (RE: Tb, Gd, Dy, Er).

 

 

Wyckoff x y 2
position
Tb 4e 0 0 1864(1)
Gd 0 0 1864(1)
Dy 0 0 1864(1)
Er O 0 1868(1)
Ge 4e 0 0 3383(1)
0 0 3390(1)
0 0 3376(1)
0 0 3364(1)
Ni 2a 0 0 0
0 0 0
0 0 0
0 0 0
A1 8g -5000 0 672(2)
-5000 0 669(2)
-5000 0 678(2)
-5000 0 679(2)

 

101

Table 5-4. Anisotropic displacement parameters (A2 x 103) for RE2N1A14Ge2 (RE: Tb,

 

 

Gd, Dy, Er).
U11 U22 U33 U23 U13 U12
Tb 3(1) 3(1) 6(1) 0 0 0
Gd 2(1) 2(1) 0(1) 0 0 0
Dy 16(1) 16(1) 7(1) 0 0 0
Er 9(1) 9(1) 4(1) 0 0 0
Ge 3(1) 3(1) 6(1) 0 0 0
2(1) 2(1) 0(1) 0 0 0
16(1) 16(1) 10(1) 0 0 0
13(1) 13(1) 8(1) 0 0 0
Ni 3(1) 3(1) 3(1) 0 0 0
2(1) 2(1) 0(2) 0 0 0
17(1) 17(1) 8(2) 0 0 0
10(1) 10(1) 3(2) 0 0 0
Al 3(1) 7(2) 8(2) 0 0 0
2(2) 4(2) 0(2) 0 0 0
16(2) 20(2) 7(2) 0 0 0
10(2) 12(2) 4(2) 0 0 0

 

102

1;.
.‘lvfilvli
1‘ r ..
p

:1.
113”

1. ”1

.-K‘_.1

.”i
‘4‘

Table 5-5. Crystal data and structure refinement for RE2C0A14Ge2 (RE = Sm, Gd).

Empirical formula
Formula weight
Temperature
Wavelength

Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (ll)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected/ unique
R(int)
Completeness to 9m...

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient
Largest diff. peak and hole

a R1=Z||F°| - |Fc||/2|F°|, wR2=[2§(w|F02 - Fc2|)2/§‘..(wF02)2]”2

SmZCoAl4Ge2 GdzCoAl,,Ge2
612.67 626.57
293(2) K 293(2) K
0.71073 A 0.71073 A
I4/mmm (#139) I4/mmm (#139)
a=4.1610(6) A a=4.1263(8) A
c = 19.670(4) A = 19.483(5) A
340.56(10) A3 331.72(13) A3
2, 5.976 Mg/m3 2, 6.272 Mg/m3
28.47 m“ 31.52 mm"
534 542
0.18x0.11 x0.06 mm 0.28x0.10x0.06mm
2.07 to 30.07° 2.09 to 28.42°
0<=h<=5 -5<=h<=5
0<=k<=5 -5<=k<=5
0<=l<=14 -25<=l<=25
228/131 1667/ 161
0.1244 0.0535
69.3 % 97.6 %
Full-matrix least-squares on F2
131/0/14 161/0/14
1.333 1.234
R1 = 0.0516 R1 = 0.0347
wR2 = 0.1387 wR2 = 0.0901
R1 = 0.0593 R1 = 0.0350
wR2 = 0.2048 wR2 = 0.0905
0.009(4) 0.017(2)

2.751 and -3.569 e.A'3

103

3.572 and -1.877 e.A'3

Table 5-6. Atomic coordinates ( x104) for R15‘32C0A14Ge2 (RE: Sm, Gd).

 

 

Wyckoff x y 2
position
Sm 4e 0 0 1838(1)
Gd 4e 0 0 1844(1)
Ge 4e 0 0 3386(4)
4e 0 0 3377(1)
Co 2a 0 0
2a 0 0
Al 8g -5000 0 649(6)
8g ' -5000 0 656(2)

 

104

Table 5-7. Anisotropic displacement parameters (A2 x 103) for RE2C0A14Ge2 (RE: Sm,
Gd).

 

U11 U22 U33 U23 U13 U12

 

Sm 3(1) 3(1) 0(4) 0 0 0
Gd 9(1) 9(1) 2(1) 0 O 0
Ge 4(1) 4(1) 4(5) 0 0

9(1) 9(1) 3(1) 0 0 O

Co 4(2) 4(2) 0(10) 0 0 0
10(1) 10(1) 2(2) 0 0 0
Al 2(4) 8(4) 16(11) 0

C

11(2) 13(2) 4(2) 0

 

The anisotropic displacement factor exponent takes the form: -21r2[h2a*2U11 + +

2hka*b*U12]

105

Table. 5-8. Selected bond lengths [A] for szNiAl,Ge2.

 

Bond Distance (A)
Tb-Ge 2.937(2)
Tb-Al 3.097(3)
Tb-Tb 3.821(1)
Ge-Al 2.760(2)
Ge-Tb 2.9623(4)

N i-Al 2.442(2)
Al-Al 2.599(6)
Al-Al 2.9236(1)

 

106

These measurements on all samples were then conducted under increasing temperature
using a 500 G applied field. The field dependent measurements, conducted at 5K, were
carried out between 0 to 55000 G. A diamagnetic correction was applied to the data to
account for core diamagnetism. No correction was made for the sample container because
the measured moment was well over an order of magnitude smaller than the sample

signal itself.

3. Results and Discussion.
Synthesis.

Of particular interest is the ability of RE,NiAl,,Ge2 (RE = Gd, Tb, Dy, Br) and
RE,CoAl,,Ge2 (RE = Sm, Gd) to form through either arc-melting or through flux
techniques, which has not seem in most of the systems presented in this work. The

‘ traditional synthetic method produced a polycrystalline ingot containing small crystals
while the flux method produces large single crystals. The traditional method produced
smaller crystals but it was significantly faster taking only a couple hours from start to

finish, while flux procedures take about a week to complete.

Structural Description.
REzMAhGe2 (M: Ni, Co) crystallizes in the Space group I4/mmm (#139) with the
structure shown in Figure 5-1. It is best described as MAl,,Ge2 (M: Ni, Co) slabs alternating
With a bilayer of rare earth atoms. The MAl,,Ge2 unit is a very stable structural unit, related to

the antiflourite structure type, and is seen in many intermetallic compounds such as

107

LaGaGNiM,6 szNi(Si1.,,,Ni,,)Al4Si(,,2 and LaNi 1+,,Al,,Si2_,,.7 The slab is five atomic layers thick
with the stacking sequence Ge/Al/M/Al/Ge, see Figure 5-2. The transition metal atoms
occupy the center of slightly distorted Al cubes to form MAI8 units. The N i-Al distances in
szNiAl4Ge2 are themselves consistent at 2.442(2)A, however, Al-Al distances in the cube
differ with distances of 2.599(6)A for interaction along the c axis and 2.9236(1)A for
interactions orientated in the ab plane. Each MAl8 unit then merges with four other sections,
Sharing edges, to form an infinite layer with one half of the cubes filled with a M center. The
Ge atoms then cap above and below the A18 cubes not centered by M atoms.

The Ge atoms themselves are in a four coordinate square pyramidal GeAl4 arrangement
that requires all Ge-Al bonds to be pointed towards the center of the MAl4Ge21ayer, Figure 5-
3a. The Al atoms in the layer exhibit a five coordinate environment which can be described
as a capped tetrahedral arrangement namely they are bonded to 2 M atoms, 2 Ge atoms and
one capping Al atom, see Figure 5-3b. The A1 atoms lie on a plane and define a perfect
square net. The Al-Al distance in the plane 2.9236(1) A, slightly longer than that normally
considered a strong bonding interaction, however weak interactions are likely present. The
nickel atoms exhibit a square prismatic (nearly cubic) environment of 8 Al atoms, Figure 5-
3c. In szNiAl4Ge2 the Ni-RE distance in this compound is 3.606 A. The rare earth ion in the
structure exhibits bonding (defined as <3.5 A) to 4 Al atoms and 5 Ge atoms in a 9

coordinate environment.

108

 

viewed down

of REZMALGe2

Figure 5-1. Structure

109

b)

 

Figure 5-2. Views of the MAl,,Ge2 layer viewed down (a) the b axis and (b) the c axis.

110

©Ge©

  

Figure 5-3. Local coordination environments of individual atomic positions in

RE,MA1,Ge,.

111

Magnetic Properties.

Magnetic susceptibility data for szNiAl4Ge2, measured on a single crystal, show
an antiferromagnetic transition with a maximum temperature at about 20 K for both
orientations of the crystal measured orientations of the crystal in relation to the applied
field, see Figure 5-4. Above this temperature both orientations exhibit Curie-Weiss
behavior as shown in the middle plot of Figure 5-4. With the crystallographic c-axis
oriented along the applied field a Weiss constant, 0, of -92.78 K is found. The 11,“,

v calculated from the Curie-Weiss behavior is 9.96 113 per Tb atom. With the field oriented
perpendicular to the c-axis, i.e. in the ab-plane, the compound exhibits a pm of 9.77 118.
and a 0 of -21.6 K. The 11,,“ values are very close to the theoretical magnetic moment for
Tb3+ of 9.72 113,8 indicating that Ni is diamagnetic as is the case in other transition metal
aluminum compounds}3 Well below the transition temperature, at 5 K, the magnetization
curve shows a linear increase with increasing field up to the maximum measured field of
55,000 G with no indications of magnetic saturation, see Figure 5-4. The differences seen
between the two orientations of the crystal are typical many systems containing heavy
rare earth metal it has been attributed to the interaction of Columb crystal field and a non-
spherical 4f charge distribution.9 This seems to indicate that the spin prefer to align along

the c axis or possible not on a crystallographic axis.
4. Conclusions.

Because of its ability to dissolve large amounts of Ge and aid in crystal growth,

liquid Al is an excellent reactive solvent for the formation of complex REZMAhGez. The

112

phases form large Single crystals at temperatures well below those needed for traditional
synthesis. Of particular interest is the observation of isostructural NiAl4Ge2 building
block seen in many of the aluminum intermetallics recently discovered in this work. This
underscores its exceptional stability, in a variety of environments and synthetic
conditions and probably all other RE ions in these analogs, suggesting it should be
viewed as a basic building block. Finally from magnetics data collected for TbZNiA14Ge2
it can be seen that the Tb ions exhibits a 3+ oxidation state while the Ni does not

contribute to the magnetism of the compound suggesting it posses a filled d band.

113

m

x
(emu / mole RE ion)

m

1/x

(mole RE ion/emu)

Magnetization per mole Tb ion

(BM. / mole RE ion)

0.2

 

 

 

 

 

 

 

 

 

 

 

 

0
0
0.15 ,
. n o
0.1 . Perpendicular C axns
° 0
O
O O
O
0.05 ’ ° '
Parallel C axis ° ° 3 g 3
l
04—11141L111111LL111 J 1L1
0 50 100 150 200 250 300
Temperature (K)
35 _
E .
30 [- °
; . -
25 T Parallel C axis '
E . -
20 ~
t . .
i ' . .
15 r . ° Perpendicular C axns
l.
4!.
10 f 9
_ 0
50,1 [ILJIILLmelAlllllnllemt
50 100 150 200 250 300 350
Temperature (K)
1.6 a
I
1.4 o.
I
. O
1'2 PerpendicularCaxis . g
1 o 0
.0
0.8 . . . . o
. O
06 . O . . O
' . o . . o '
0.4 0 7 . o ' Parallel C axis
I . .
a . .
0.2 a .-
30'
o r111in..111.r4..1111+L411111
0 110‘ 210‘ 310‘ 410‘ 510‘ 610‘

Field (Gauss)

Figure 54. Magnetic behavior of szNiA14Ge2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)
magnetization as a function of field.

114

References

 

‘ Sieve, B.; Trikalitis, P.N.; Kanatzidis, M.G. Z. Anorg. Allg. Chemie 2002 628, 1568-
1574

2 SMART. Data Collection Software for the SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

3 SAINT. Data Processing Software for SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

4 Sheldrick, G.M. University of Gdttingen, Germany. Manuscript to be Published

5 Sheldrick, G.M. SHELXL. Structural Determination Programs, Version 5.0; Siemans
Analytical X-Ray Instruments Inc.: Madison, WI, 1995 8
° Grin, Yu. N.; Yarmolyuk, Ya.P.; Rozhdestvenskaya I.V.; Gladyshevskii, E.I. Sov. Phys.
Crystallogr., 1982, 27, 418-419

7 Sieve, B.; Kanatzidis, M.G. work in progress

8 Bourdreaux, E.A.; Mulay, L.N. in Theory and Applications of Molecular
Paramagnetism, John Wiley and Sons, New York 1976.

9 Zhuravleva, M. Chen, X.C. Wang, X.; Schultz, A.; Ireland, J .; Kannewurf, C.R.

Kanatzidis, M.G. Chem. Mater. submitted

115

Chapter 6
Synthesis, Structural and Property Characterization of the
Phases, REzNiAl6_,,Ge4,y (x~0.24, y~1.33)

(RE: Ce, La, Pr, Nd, Sm)

1. Introduction.

During the last decade metal flux techniques have emerged as an excellent
synthetic method for the synthesis and physical studies of intermetallic compounds.
Many new phases have been synthesized exhibiting new structure type along with
variations on known ones providing important information about the relationships
between atomic arrangements and properties of the overall materials. Attempts to form
analogs in the REZNiAl4Ge2 structure type with early rare earth metals failed instead
forming a new series of compounds RE2N1A16,,Ge4_y. This structure is comprised of
stacking NiAl,,Ge2 layers and Al/Ge PbO type layers Stacking along the c axis.

In this chapter the systems of REzNiAlMGeH (RE = Ce, La, Pr, Nd, Sm) are
presented as synthesized from molten Al. They represent a Structural variation of the
known structure type CezNiGa10 that has been previously studied by another research
group.1 The structural arrangement is presented and structural question highlighting about
atomic disorder and vacancies are discussed. Along with the structural description
preliminary physical properties of the Ce analog are presented. Selections of this chapter

inCluding the structure and properties of CezNiAlnge,y have been published recently.2

116

2. Experimental Section.
Synthesis.
Reagents.

Ce and Sm 99.9%, metal chips, Research Chemicals, Phoenix, AZ;- La, Pr, Nd
99.9%, -40 mesh, Cerac, Milwaukee, WI; Ni, 99%, 325 mesh, Sargent, Buffalo Grove, 11;
A1, 99.5 %, -20 mesh, Milwaukee, WI; Ge, 99.999%, 3-6 mm pieces, Cerac, Milwaukee,

WI.

Synthetic Method.

In a N2 filled glove box RE (La, Ce, Pr, Nd, or Sm), Ni, Al, and Ge powders were
mixed in a l:1:30:1 ratio. The mixture was then placed into an alumina crucible and
sealed within a fused silica tube under vacuum (<1x10“‘ Torr). The tubes were heated to
850° C over a period of 20 hours and kept there for 96 hours then slowly cooled to 500°
C over 72 hours and finally to 50° C in 12 hours. After heating the alumina tubes were
removed from the silica tubes and submerged in 5M NaOH for 24 hours. This dissolved
the excess Al and yielded both plate-like crystal and black microcrystalline power. Both
product types were identified as a Single phase by comparison of experimental powder
patterns to those generated from the solved crystal structure. This reaction shows yields

of around 80% based on Ge.

Physical Measurements

EDS Analysis.

117

Quantitative microprobe analysis was performed with a JEOL J SM-6400
Scanning Electron Microscope (SEM) equipped with N oran Energy Dispersive
Spectroscopy (EDS) detector. Data were acquired using an accelerating voltage of 25 kV
and 100 sec accumulation time. Standards where recorded under the same experimental
conditions to yield correction factors. The elemental analysis for CeZNiAIMGe,y (x~0.24,
y~1.34) yielded a formula of CezNiLmAlmGem. This ratio is very close to the final
crystallographically refined elemental ratios except for Al which exhibited an error of

10% over the actual formula of the compound.

Single Crystal X-ray Crystallography.

Single crystal X-ray diffraction data for REzNiAlngeH (RE = La, Ce, Pr, Nd,
Sm) were collected at 298 K on a Siemens Platform CCD diffractometer using Mo Kor
(A=0.71069 A) radiation. The SMART software3 was used for the data acquisition and
the program SAINT4 was used for the data extraction and reduction. An empirical
absorption correction using SADABS5 was applied to REzNiAlMGe,y data (RE 2 La, Pr,
Nd, Sm) while a face indexed absorption correction was performed on the data for
CezNiAlMGeH utilizing the SHELXL software. 6 These structures were solved and
refined with the SHELXL package of programs. The crystallographic and refinement
details are listed in Table 6-1 through 6-3. The fractional atomic positions, displacement .
parameters (U values) and selected bond distances are listed in Tables 6-4 and 6-5. Bond
distance for CezNiAIMGeH are listed in Table 6-6. During refinement it was found that
the occupancy of Al(2) in REzNiAlngeH (RE = La, Pr, Nd, Sm) refines to unreasonable

values (>25%) likely due to difficulties with the absorption correction used in these

118

Table 6-1. Crystal data and structure refinement for REzNiAlngeH (RE = Ce, La).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11.)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Largest diff. peak and hole

Rl=>3llFol - Ian/2113.1. wR2=[2(WIF.2 - F.2|)°/£(wF,2)21‘”

CezNiAlmGem
686.275

298 K

0.71073 A
I4/mmm (#139)
a = 4.1951(9) A
c = 26.524(7) A
466.8(2) A3

2, 4.883 Mg/m3
20.38 mm"
606.98

0.03 x 0.16 x 0.20 mm
1.54 to 28.65 °

La,NiAl,Ge,,,
690.086

298 K

0.71073 A
I4/mmm (#139)
a = 4.2320(6) A
c = 26.828(5) A
480.49(14) A3
2, 4.770 Mg/m3
19.30 mm"
610.24

0.05 x 0.07 x 0.19 mm
1.52 to 28.45 °

-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-25<=l<=20 -34<=l<=31
2307 / 221 2390/ 305
0.0324 0.0492
98.2% 97.0 %
Full-matrix least-squares on F2
221/0/24 305/0/23
1.205 1.082
R1 = 0.0382 R1 = 0.0497
wR2 = 0.1065 wR2 = 0.1312
R1 = 0.0382 R1 = 0.0497
wR2 = 0.1065 wR2 = 0.1312

1.714 and —2.999 e.A'3

119

1.190 and -1934 e.A'3

Table 6-2. Crystal data and structure refinement for REzNiAlngegy (RE = Pr, Nd).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>20(I)]

R indices (all data)

Largest diff. peak and hole
R1=2||Fo| - [Fell/EIFol, wR2=[Z(w|F02 - Fc2|)2/ZZ(wF02)2]“2

PeriAléGez.78
692.086

293 K

0.71073 A
I4/mmm (#139)
a = 4.1559(13) A

, c = 26.230(12) A

453.0(3) A3

2, 5.074 Mg/m3

16.93 mm"

507.92

0.07 x 0.09 x 0.13 m
11.29 to 23.21°

Nd,NiAl,Ge,_,,
695.41

298 K

0.71073 A

I4/mmm (#139)

a = 4.1483(7) A

c = 26.239(7) A
451.53(16) A3

2, 5.115 Mg/m3

17.22 mm“1

507.36 '

0.12 x 0.12 x 0.15 mm
1.55 to 28.60°

-4<=h<=4 -5<=h<=5
—4<=k<=4 ~5<=k<=5
-28<=l<=25 -33<=l<=34
757/112 2270/218
0.0463 0.0615
81.8 % 98.2 %
Full-matrix least-squares on F2
112/0/23 218/0/24
1.279 1.111
R1 = 0.0557 R1 = 0.0508
wR2 = 0.1272 WR2 = 0.1362
R1 = 0.0561 R1 = 0.0508
wR2 = 0.1277 WR2 = 0.1362

1.739 and -3723 e.A'3

120

3.279 and -2.118 e.A-3

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal Size

0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0max

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Table 6-3. Crystal data and structure refinement for szNiAl6_xGe4_y.

szNiAlgGel26
701.53

293(2) K
0.71073 A
I4/mmm (#139)
a = 4.1224(6) A
c = 25.862(5) A
439.51(13) A3
2, 5.727 Mg/m3
26.76 mm‘1
668.64

0.03 x 0.08 x 0.12
1.57 to 28.54 °
-5<=h<=5
-5<=k<=5
-33<=l<=34
2191 /213
0.1103

98.2 %
Full-matrix least-squares on F2
213 / 0 / 24
1.333

R1 = 0.0512
wR2 = 0.1441
R1 = 0.0529
wR2 = 0.1445
0.0042(14)

Largest diff. peak and hole 3.783 and -2.388 e.A'3
R1=E||Fo| - |1=,||/2|F,|, wR2=[>:(w|F,2 - Fc2|)2/Z(wF02)2]“2

121

Table 6—4. Atomic coordinates ( x10“) and occupancies for REZNiAlngegy (RE: Ce, La,
Pr, Nd, Sm).

 

 

Wyckoff x y 2 Occupancy
position
Ce 4e 0 0 1481(1) 1
La 0 0 1474(1) 1
Pr 0 0 1498(1) 1
Nd 0 0 1491(1) 1
Sm 0 0 1525(1) 1
Ni 2a 0 0 0 1
0 0 0 1
0 0 0 1
0 0 0 1
0 0 0 1
Ge( 1) 4e 5000 5000 11 19(1) 1
5000 5000 1 100(1) 1
5000 5000 1 141(1) 1
5000 5000 1138(1) 1
5000 5000 1 171(1) 1
Al(l) 8g 5000 0 491(1) 1
5000 0 486(2) 1
5000 0 494(2) 1
5000 0 495(2) 1
5000 0 509(2) 1
Ge(2) 4d 5000 0 2500 0.32
5000 0 2500 0.33
5000 0 2500 0.39
5000 0 2500 0.37
5000 0 2500 0.63
Al(2) l6n 1210(4) 0 2903(5) 0.22
4e 5000 5000 2038(6) l .0
l6n 1120(8) 0 2897(12) 0.25
l6n 1220(50) 0 2884(6) 0.25
l6n 1020(60) 0 2847(9) 0.25

 

122

Table 6-5. Anisotropic displacement parameters (A2 x 103) for REzNiAl6,,,Ge,,_y (RE: Ce,
La, Pr, Nd, Sm).

 

 

U11 U22 U33 U23 U13 012
Ce 10(1) 10(1) 10(1) 0 0 0
Pr 9(1) 9(1) 18(2) 0 0 0
Nd 16(1) 16(1) 10(1) 0 0 0
Sm 7(1) 7(1) 7(1) 0 0 0
Ni 8(1) 8(1) 13(2) 0 0 0

7(2) 7(2) 17(3) 0 0 0
14(1) 14(1) 12(2) 0 0 0
4(2) 4(2) 6(2) 0 O 0
Ge(l) 17(1) 17(1) 23(1) 0 0 0
11(2) 11(2) 22(2) 0 0 0
19(1) 19(1) 18(1) 0 0 0
6(1) 6(1) 13(2) 0 0 0
Al(l) 9(2) 10(2) 10(2) 0 0 0
6(3) 7(3) 20(4) 0 0 0
15(2) 15(2) 9(2) 0 0 0
6(3) 14(3) 2(3) 0 0 0
Ge(2) 53(5) 53(5) 28(5) 0 0 0
39(8) 39(8) 19(9) 0 0 0
58(5) 58(5) 30(6) 0 0 0
45(4) 45(4) 11(3) 0 0 0
Al(2) 40(11) 14(7) 24(6) 0 -19(6) 0
80(40) 40(20) 63(14) 0 16(14) 0
52(14) 27(10) 22(9) 0 -18(7) 0
180(70)100(40)15(8) 0 -39(19)0

 

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +
2hka*b*U12]

123

Table 6-6. Selected bond lengths [A] for Ce,NiAl,_,Ge,,.

 

Bond Distance (A)
Ce-Al(2) 3.10(1)
Al(l)-Ce 3.361(3)
Ce-Ge(1) 3.118(1)
Ni-Al( 1) 2.469(2)
Ge(1)-Al(l) 2.64(1)
Ge(l)-Al(1) 2.678(3)
Ge(l)-Ce 3.118(1)
Al(1)-Al(1) 2.604(7)
Al(1)-Al(1) 2.9664(6)
Ge(2)-Al(2) 192(2)
Ge(2)-Al(2) 2.408(6)
Al(2)-Al(2) 0.72(2)
Al(2)-Al(2) 102(3)
Al(2)-Ge(2) 282(2)

124

compounds. To rectify this the occupancy of Al(2) in these systems has been set to 25%

in the refinements.

Electron Diffraction Studies (TEM).

Electron crystallographic studies were carried out on a JEOL 100CX transmission
electron microscope (TEM) equipped 'with a CeB6 filament and an accelerating voltage of
120 kV. The samples were gently ground into a fine powder and the specimens were
prepared by dipping a carbon coated copper grid into the microcrystalline powder. The

samples showed no decomposition under the electron beam.

Magnetic Characterization.

Magnetic susceptibility for CezNiAlMGe,y (x~0.24, y~1.34) was measured as a
function of both temperature and field using a MPMS Quantum Design SQUID
magnetometer with powder samples were used in measuring CezNiAlngeH. An initial
study of field dependence was conducted to find a suitable field for the variable
temperature studies. These measurements were then conducted under increasing
temperature using a 500 G applied field. Field dependent measurements, conducted at
5K, were carried out between 1‘ 55000 G. A diamagnetic correction was applied to the
data to account for core diamagnetism though no correction was made for the sample
container however as the measured moment was well over an order of magnitude smaller

than the sample signal itself.

125

3. Results and Discussion.
Synthesis.
The reaction of the early rare earth metals (La, Ce, Pr, Nd, Sm), Ni and Si in
excess Al metal produces large single crystals of REzNiAlMGe,y up to several mm in
' size. These plate-like crystals exhibit shiny surfaces. Chemically the crystals are stabile in

strong bases, dilute acids, air and water.

Structural Description.

The compounds REzNiAlngeH (RE = La, Pr, Nd, Sm) are of the CezNiGalo structure
type7 with substantial structural deviations, see Figure 6-1. The structure consists of
alternating RE layers (A), layers of N iAl,,Ge2 (B) and A12_,,Ge2,y layers (C) (a defect PbO-type
layer). These layers alternate in a ABACABA fashion along the c axis of the unit cell.

The NiAl,,Ge2 unit is a very stable structural unit, related to the antiflourite structure
type, and is seen in many intermetallic compounds such as LaGa,5Ni,,,,,8 szNi(Si1-
x,Ni,,)Al.iSi6.2 and LaNi,,,,Al4Si2_,.9 The slab is five atomic layers thick with the stacking
sequence Ge/Al/Ni/Al/Ge, see Figure 6-2 The transition metal atoms occupy the center of
slightly distorted Al cubes to form NiAl8 units. The Ni-Al distances are 2.469(2)A while the
Al-Al distances in the cube are 2.9664(6) A for bonds in the ab plane and a distance of
2.604(7) for AL-Al bonds aligned along the c axis in CezNiAlngeH. Each NiAl8 unit then
merge with four other units, sharing edges to form an infinite layer, with one half of the total
formed cubes filled with a Ni center. Ge atoms then cap above and below the A18 cubes not

centered by Ni atoms.

126

The difference from the CezNiGa10 structure type is exhibited in the second layer, the
PbO layer. The parent compound exhibits ideal and complete Ga layers of PbO-type while in
REzNiAlngegy these layers are now AlGe with vacancies on the Ge position and a
displacement of Al atoms off the 4e special position (site symmetry 4mm), see Figure 6-3.
The shifting of the Al atoms off the 4e position, in all analogs except LazNiAlnge,y (see
below), moving it to a l6n position creates four closely lying symmetry related atoms, Al-Al
distance of 0.72 A in the Ce analog. For this quartet, three bond distances to the neighboring
Ge atom (2.817 A, 2.408 A, and 2.408 A) are reasonable whereas the fourth is too short to be
realistic at 192(2) A. The occupancy of this Al position in the Ce analog refines to 22% of
the total full position suggesting that it is present approximately 1/4m of the time. This would
be expected for a simple displacement of a Single Al atom from a 4e to l6n position. At the
same time the occupancy of the Ge(2) atoms in the layer refined to only 32 %. The vacancies
in the Ge layer then alleviates the need for the short 192(2) A distances, which are only
apparent and the result of averaging the structure to a I4/mmm unit cell.

It was found that the La analog is of particular interest in terms of the disorder in the
PbO-Type layer. The Al atoms in the A12,,Ge2,y layer, which in the other analogs are shifted
off the special position, appear to maintain the 4e position. The Ge(2)-Al(2) bond distance in
this analog are 2.462 A, which is Short but not unreasonable for an Al-Ge bond. The La
analog, the largest RE atom which also exhibits the largest unit cell with an a axis of
4.2320(6) A, appears to be large enough to allow reasonable bond distances in the A12,,Ge2,y
layer without shifting of the Al. The Ge occupancy, however, still refines to approximately
one third total occupancy indicating that this vacancy is not purely due to steric reasons

based in the spatial relationships to the neighboring Al atoms. Another point of interest is that

127

the Sm analog shows differences in the displayed vacancies on the Ge(2) position refining to
2/3 occupied when the other analogs refine to 1/3 occupancy. This is possibly due to the
smaller size of the Sm atoms compared to the other analogs, though more conclusive work
either defining the superstructure of the material or the synthesis of the smaller rare earth
analogs need to be completed before this hypothesis can be tested.

The Ge(l) atoms, in the NiAl,,Ge2 layer, exhibit a five coordinate square pyramidal
GeAl5 arrangement with 4 bonds pointing towards the center of the NiAl4Ge21ayer, Figure 6-
43, while the apex bond connects the NiAl4Ge2 unit to the A12_,Ge2,y layer. Ge(2)'s
coordination environment, contained within the AIMGez,y layer, is hard to define without
more information about the supercell nature of the structure but likely bonds to 4 Al(2) atoms
on the outside of the A12_,Ge2,y layer, Figure 6-4b. Al(l) exhibits a five coordinate
environment which can be described as a capped tetrahedral arrangement bonding to 2 Ni
atoms, 2 Ge atoms and one capping Al atom, see Figure 6-4c. The Al atoms lie on a plane
and define a perfect square net with an Al-Al distance in the plane 2.9664(6) A slightly
longer than normal, however weak interactions are likely present. While the Al in the
disorder PbO type layer, Figure 6—4d, likely bonds to 4 Ge atoms 3 within the layer and one
within the NiAl,,Ge2 layer through the connecting bond. The nickel atoms exhibit a distorted
cubic environment of 8 Al atoms, Figure 6-4e. The rare earth ion exhibits a 16 coordinate
environment considering interactions out to 3.5 A bonding to 4 Al(l), 4 Al(2), 4 Ge(l), and 4

Ge(2) assuming only one of each 4 Al(2) is present, Figure 6-4f.

128

 

(x~0.24, y~1.34) viewed down the b axis.

of R13,NiAl,,Ge,,y

Figure 6-1. Structure

129

 

Figure 6-2. Views of the NiAl,,Ge2 layer viewed down (a) the b axis and (b) the c axis.

130

 

Figure 6-3. Views of the heavily vacant A12_,,Ge2,y (x~0.24, y~1.34) layer viewed down

(a) the b axis and (b) the c axis. The disordered Al atoms are shown.

131

Al(2) “(2)

  

Al(2 ) ‘Ge( (2) .
Al(2)

 

Ge(1)

 
   
    

Al(2)

 

Ge(2) Ge(2

Al(t) Al(1) Al(1) Al(1) 1‘) Al(1..)Al(1)AI(1)

   

Ge(2) Ge(2) Ge(2)

Figure 6-4. Coordination environments in REZNiAlngeH

132

The observed vacancies and disorder raises the question of possible long range
ordering in the Al/Ge layer and the existence of a larger structural supercell. To examine this,
electron diffraction experiments were conducted on samples of CezNiAlngegy. Supercell
peaks are indeed seen in selected area electron diffraction patterns indicating long range
ordering in the structure, see Figure 6-5. The observed diffraction pattern be can indexed to a
commensurate 3 x 3 supercell with lattice constants of a=12.58 A and b=12.58 A. We have
no information on ordering along the c axis as electron diffraction examining this axis was
very difficult due to the thin plate morphology of the crystals. A satisfactory X—ray
refinement based on this larger cell could not be reached, however, due to the very low

inherent reflection intensity exhibited by the supercell reflections.

Magnetic Properties.

Samples of CezNiAlMGe,y (x~0.24, y~1.34) exhibit complex magnetic properties
with no Curie-Weiss behavior at any temperature. At low temperatures a ferromagnetic
transition can be observed at 4K, while at higher temperatures the compound shows a
weak transition, at about 200K (indicated by an inflection point), which Significantly
decreases the magnetization with rising temperatures. This behavior is expressed by the
dramatic change in the slope of the inverse susceptibility plot and may be due to valence
fluctuations in Ce over the measured temperature range.10 The Ce atom formal valency in
this compound however is likely 3+ while the Ni ions are likely in a diamagnetic state as
has been seen in other compounds grown from Al fluxes. When the field dependence was

studied at 5 K a gradual saturation of the moment is seen with increasing fields with the

133

absence of hysteresis, see Figure 6-6. No signs of magnetic saturation occur however

before 55000 G.

4. Conclusions.

The ability to dissolve large amounts of reactive Ge has proven molten Al to be
an excellent solvent for the synthesis of the phases REQNiAlngeH (RE = La, Ce, Pr, Nd,
Sm). Using flux methods large single crystals up to several mm in dimensions in a plate-
like morphology can be grown of the title compounds while with traditional arc-melting
techniques only polycrystalline ingots are obtained. Of particular interest in the Structure
is the presence of once again the “NiAl4Ge2” unit in the structure. The reoccurring
existence of this unit in several systems underscores its exceptional stability as a
structural unit. The existence of the same structural unit in several compounds with
differing transition metals as well as rare-earth metals may provide an interesting
opportunity for properties studies using structural series. The second layer of the
structure, the PbO-type layer, appears to be the cause of a larger structural unit cell
possibly containing several of the currently modeled unit cells. One explanation of this
may be ordering of the vacancies in the layer along with positional ordering of the Al(2)

atoms (discussed above).

134

 

 

(-660)
(-360

Intensity (arb. units)

(-560)
(-460)

 

 

 

d* (reciprical angstroms)

Figure 6-5. (a) Selected area diffraction pattern of CezNiAlngeH, viewed down the c
axis (hk0 zone), showing the 3ax3b supercell and (b) one-dimensional peak intensity

profile from the boxed area indicated in the pattern clearly showing the weak (-560) and

(-460) commensurate supercell reflections.

135

21

m

X
(emu /mole Ce ion)

CT
V

m

1/x
(mole Ce ion / emu)

10

0.1

250

200

150

100

50

 

V Vd‘VIT' . '1 III

 

 

A 1 l A l A l A I 1 1 1 J A A L L A L A j

 

150 200 250 300 350
Temperature (K)

 

IlTrY

 

 

 

0 50 100 150 200 250 300 350

Temperature (K)
0.8

 

O
v

0.6 [- 0....
0.4 L 0'

0.2

(3M)
0
,..,.. ,.

0.2 :

-O.4 :— 0'

Magnetization per mole Ce

'0.6 :— ‘0'.

 

 

l L l l l 1 J l 1 l 1 J l 1 ‘L 1 J
-210‘ 0 210‘ 410 610‘
Field (Gauss)

Figure 6-6. Magnetic behavior of CezNiAIMGe,Ly (top) susceptibility as a function of

 

-0.8 .
-6 10‘

temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

136

References

 

' Yarmolyuk, Ya.P.; Grin, Yu. N .; Rozhdestvenskaya I.V.; Usov, O.A.; Kuz’min, A.M.;
Bruskov, V.A.; Gladyshevskii, E.I. Sov. Phys. Crystallogr., 198227, 599-600

2 Sieve, B.; Trikalitis, P.N.; Kanatzidis, M.G. Z. Anorg. Allg. Chemie 2002 628, 1568-
1574

5 SMART. Data Collection Software for the SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

‘ SAINT. Data Processing Software for SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

5 Sheldrick, G.M. University of Gdttingen, Germany. Manuscript to be Published

5 Sheldrick, G.M. SHELXL. Structural Determination Programs, Version 5.0; Siemans
Analytical X-Ray Instruments Inc.: Madison WI, 1995

7 Yarmolyuk, Ya.P.; Grin, Yu. N .; Rozhdestvenskaya I.V.; Usov, O.A.; Kuz’min, A.M.;
Bruskov, V.A.; Gladyshevskii, E.I. Sov. Phys. Crystallogr., 1982 27, 599-600

8 Grin, Yu. N .; Yarmolyuk, Ya.P.; Rozhdestvenskaya I.V.; Gladyshevskii, E.I. Sov. Phys.
Crystallogr., 1982 27, 418-419

9 Sieve, B.; Kanatzidis, M.G. work in progress

1° (a) Kwei, G.H.; Lawrence, J.M.; Canfield, P.C. Phys. Rev. B, 1994 49, 14708-1410 (b)
Thermopower measurements were then conducted to probe the possible Ce valence
fluctuations. The thermopower values ranged from i5 uV/K, in the temperature range of
50-300 K suggest metallic behavior but provided no significant information about

possible of valence fluctuations in the Ce atoms.

137

Chapter 7.
Synthesis and Characterization of RENiA14Ge2 (RE = Sm, Gd,
Tb, Dy, Ho, Er, Tm, Lu, Y) Containing a Trigonal Lattice of

RE ions

1. Introducion.

Continuing to study the chemistry of Ni containing compounds reactions were
conducted substituting Ge for Si. The initial reactions using Sm lead to the discovery of
the first quaternary aluminum germanide for our group, the trigonal phase SmNiAl4Ge2.
Since the discovery it has been found that several analogs can be made containing a
variety of late rare earth metals creating a sizable isostructural series. Initial results on
this series including the Sm structure and magnetism have recently been published in the
literature.‘

Here we continue in exploring the system and report the synthesis, novel structure
and magnetic properties of the expanded series RENiAl4Ge2 (RE= Sm, Gd, Tb, Dy, Ho,
Er, Tm, Lu, Y) which all crystallized from molten Al exhibiting the novel structure type.
Along with the structures an extensive Study of the magnetic behavior of the phases is
presented including a investigation of a possible existence of spin frustration originating

from the trigonal lattice of rare earth ions in the structure.

138

2. Experimental Section.
Synthesis.
Reagents.

The following reagents were used as obtained: Sm, 99.9%, metal chips, Research
Chemicals, Phoenix, AZ; Gd, Tb, Dy, Ho, Er, Tm, Lu, Y 99.9%, -40 mesh, Cerac,
Milwaukee, WI; Ni, 99%, 325 mesh, Sargent, Buffalo Grove, 11; Ge, 99.999%, 3-6 mm

pieces, Cerac, Milwaukee, WI.

Synthetic Method.

RENiAl,,Ge2 (RE = Sm, Gd, Tb, Dy, Ho, Er, Tm, Lu, Y) was prepared by the
reaction of 0.1 mole of RE metal, 0.059 g (0.1 mole) of Ni metal, .0405 g (1.5
mole) of A1 metal, and .0363 g ( 0.5 mole) of Ge mixed in an alumina tube. This tube
was then sealed in an evacuated (1.0 *10-4 Torr) 13 mm o.d. x 11 mm i.d. quartz tube
and heated at 800° for 4 days, followed by cooling at a rate of -6.25° cm to 500° C and
then -50° cm to 50° C. The products were isolated in ~5 M NaOH and washed and dried
with acetone and ether. This reaction gave black powder and small silver colored plates
in 26% yield, based on Sm. Purity was measured by comparison of experimentalpowder
patterns, both powder and plates, to theoretical calculated patterns. Purity was shown to
be ~95% in the SmNiAl,,Ge2 reaction with the impurity peaks being due to residual

elemental Ge.

Physical Measurements.

EDS Analysis.

139

Quantitative micr0probe analysis of RENiAl,Ge2was performed with a JEOL
ISM-6400 Scanning Electron Microscope (SEM) equipped with Noran Energy
Dispersive Spectroscopy (EDS) detector. Data were acquired using an accelerating
voltage of 25 kV and 100 sec accumulation time. Standards where recorded under the
I same experimental conditions to obtain correction factors. After calibration the Tb
compound gave an elemental ratio of 1 Tb: 1 Ni: 4.5 A1: 2.3 Ge, within experimental

errors from the actual formula.

Single Crystal X-ray Crystallography.

Single crystal X-ray diffraction data of SmNiA14Ge2 were collected at room
temperature using a Rigaku 4 circle diffractometer with Mo K01 (A = 0.71073 A)
radiation. An empirical absorption correction based on ‘1’ scans was applied to the data.
The structure was solved with direct methods using SHELXS 862 and refined with
SHELXS 5.03 package of programs. Data collection parameters, atomic positions,
anisotropic thermal parameters information is provided in Tables 11-1 through 11-3.
Selected bond distances for the Sm analog are shown in Table 11-6.

Single crystal X-ray diffraction data of RENiAl4Ge2 (RE: GD, Tb, Dy, Ho, Er,
Tm, Lu, Y) were collected at room temperature using a Siemans Platform CCD
diffractometer using Mo KOL (A = 0.71073 A) radiation. The SMART software3 was used
for the data acquisition and the program SAINT4 was used for the data extraction and
reduction. And empirical absorption correction using SADABS5 was applied to the data

and the structure was solved in the SHELXL package of programs6 using the Sm analog

140

as a starting point. Data collection parameters, atomic positions, anisotropic thermal

parameters information is provided in Tables 11-4, 11-5, and 11-7 through 11—23.

Magnetic Characterization.

The magnetic susceptibilities for RENiAl4Ge2 (RE = Sm, Tb, Tm, Gd) were
measured over the range 2-300 K using a MPMS Quantum Design SQUID
magnetometer. Samples were ground to a fine powder to minimize possible anisotropic
effects and loaded into PVC containers. The temperature dependent susceptibility studies
were performed at 200 Gauss. Corrections for the diamagnetism of the sample containers
were made by measuring the magnetic response of the empty container under the same
conditions of temperature and field which were measured for the filled container.
Diamagnetic contribution of every ion to xM was corrected according to Selwood.7

AC single crystal analysis was conducted on GdNiAl,,Ge2 between 8 and 10 K in with an

field of 5 G alternating at 1000, 100 and 10 Hz.

141

Table 7-1. Crystal data and structure refinement for RENiA14Ge2(RE = Sm, Gd).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

6 range

Limiting indices

Refl. collected / unique
R(int)
Completeness to Bmax

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient
Largest diff. peak and hole

111=2||Fo| - IFCIIIZIFOI. wR2=[£(wIF.f - FCZIY/mwrffl‘”

SmNiAl,Ge2 GdNiAl,Ge2
462.19 469.08
293(2) K 293(2) K
0.71073 A 0.71073 A
R-3m (#166) R-3m (#166)
a=4.112l(6) A a=4.089(2) A
c = 31.109(6) A c = 30.89(2) A
455.55(13) A3 447.2(5) A3
3, 5.04 Mg/m3 3, 5.22 Mg/m3
22.82 m" 24.54 mm"
618 624
0.12 x 0.08 x 0.16 mm 0.08 x 0.05 x 0.02 mm
l.96 to 29.96 deg. 1.98 to 28.10 deg.
-5<=h<=5 -5<=h<=5
-5<=k<=5 -4<=k<=5
-42<=l<=42 -3 8<=l<=40
1744/207 1219/ 166
0.0262 0.1417
99.5 % 93.3 %
Full-matrix least-squares on F2
205/0/15 166/0/15
0.888 1.187
R1 =0.0121 R1 =0.0440
wR2 = 0.0378 wR2 = 0.1142
R1 = 0.0183 R1: 0.0440
wR2 = 0.1260 wR2 = 0.1142
0.0030(4) 0.0008(7)

0.984 and -l.056 6A3

142

2.329 and -3329 e.A'3

Table 72 Crystal data and structure refinement for RENiAl,,Ge2 (RE = Tb, Dy).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

2, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

9 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20‘(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole
R1=Z||Fol - IFCII/EIFOI. wR2=[Z(w|F,2 - Fc2|)2/Z(wF02)2]”2

TbNiAl4Ge2
470.76

293(2) K
0.71073 A
R-3m (#166)

a = 4.0991(6) A
c = 30.881(6) A

DyNiAl4Ge2
474.33

293(2) K
0.71073 A
R-3m (#166)

a = 4.0951(6) A
c = 309100) A

449.36(13) A3 448.91(13) A3

3, 5.22 Mg/m3 3, 5.25 Mg/m3

25.13 mrn‘l 25.83 mrn'l

627 630

0.09 x 0.08 x 0.04 mm 0.06 x 0.05 x 0.02 mm

1.98 to 28.62 deg. 1.98 to 28.94 deg.

~5<=h<=5 -5<=h<=5

-5<=k<=5 -5<=k<=5

-40<=l<=39 ~40<=l<=40

1470/ 180 1535/ 186

0.1128 0.1102

96.8 % 96.4 %
Full-matrix least-squares on F2

180/0/15 186/0/15

1.182 1.162

R1 = 0.0320 R1 = 0.0445

wR2 = 0.0768 wR2 = 0.1258

R1 = 0.0324 R1 = 0.0447

wR2 = 0.0773 wR2 = 0.1259

0.0020(6) 0.0037(12)

1.611 and -2.067 e.A'3 4.038 and 4.207 e.A'3

143

Table 7-3. Crystal data and structure refinement for RENiAl,,Ge2 (RE = Ho, Er).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size (mm)

0 range

Limiting indices

Reflections collected / unique
R(int)

Completeness to 0 = 28.27
Refinement method

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

Final R indices [I>20’(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=z||F,| - |Fc||/Z|Fo|, wR2=[Z(w|F02

HoNiAl4Ge2
476.76

293(2) K
0.71073 A

R-3m (#166)

a = 4.0704(11) A
c = 30.717(12) A
440.7(2) A3

3, 5.40 Mg/m3
27.05 mrn‘l

633

0.10 x 0.07 x 0.03 mm
1.99 to 28.27 deg.
—5<=h<=5
-5<=k<=5
-39<=l<=36
1399/ 178
0.0859

99.4 %

ErNiA14Ge2
479.76

293(2) K

0.71073 A

R-3m (#166)

a = 4.079(2) A

c = 3069(2) A
442.2(5) A3

3, 5.40 Mg/m3
27.78 mm1

636

0.20 x 0.20 x 0.15 mm
1.99 to 28.25 deg.
-5<=h<=5
-5<=k<=5
-39<=l<=39
1436/ 173
0.1831

96.6 %

Full-matrix least-squares on F2

178/0/15
1.206

R1 = 0.0274

wR2 = 0.0790

R1 = 0.0274

wR2 = 0.0790
0.0033(7)

2.267 and -2.640 6A3

_ Fc2l)2/Z(WF°2)2]1/2

144

173/0/15

1.213

R1 = 0.0542

wR2 = 0.1471

R1 = 0.0542

wR2 = 0.1471
0.014(3)

4.930 and -2.116 6A3

Table 7-4. Crystal data and structure refinement for RENiAl4Ge2 (RE = Tm, Lu).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected/ unique
R(int)
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole
R1=2||Fo| - IFCIIIZIFOI, wR2=[ZZ(w|F,,2 - Fc2|)2/22(wF02)2]”2

TmNiAl4Ge2
480.76

293(2) K

0.71073 A

R-3m (#166)

a = 4.0703(18) A

c = 3065(2) A
439.8(4) A3

3, 5.46 Mg/m3
28.75 mm"

639

0.6 x 0.3 x 0.01 mm
1.99 to 28.49 deg.

LuNiAl4Ge2
486.80

293(2) K

0.71073 A

R-3m (#166)

a = 4.0650(8) A

c = 30.652(9)/1.
438.64(17) A3

3, 5.52 Mg/m3

30.53 mm"

645

0.09 x 0.08 x 0.01 mm
1.99 to 23.21 deg.

-5<=h<=5 -4<=h<=4
-5<=k<=5 —4<=k<=4
~40<=l<=39 ~34<=1<=34
1466/ 177 1045 / 111
0.1414 0.0543
97.3 % 100.0 %
Full-matrix least-squares on F2
177/0/15 111/0/15
1.126 0.996
R1 = 0.0420 R1 = 0.0286
wR2 = 0.1026 wR2 = 0.0870
R1 = 0.0429 R1 = 0.0286
wR2 = 0.1026 wR2 = 0.0870 ‘
0.0080(14) 0.0061(12)

1.920 and -5339 e. A3

145

1.244 and -2032 e.A'3

EmpiIical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Reflections collected / unique
Completeness to‘ 0max

R(int)

Refinement method

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

Final R indices [I>20'(I)]
R indices (all data)

Extinction coefficient
Largest diff. peak and hole

Table 7-5. Crystal data and structure refinement for YNiAl4Ge2.

YNiAl4Ge2

400.74

293(2) K

0.71073 A

R-3m (#166)

a = 4.0959(11) A

c = 30.958(11) A
449.8(2) A3

3, 4.44 Mg/m3
23.04 mm"

549

0.2 x 0.2 x 0.16 mm
1.97 to 28.15 deg.
-5<=h<=5
-5<=k<=5
-40<=l<=38

1456

176

0.1110

Full-matrix least-squares on F2
176 / 0/ 15

1.270

R1 = 0.0464

wR2 = 0.1285

R1 = 0.0470

wR2 = 0.1289
0.007(2)

2.576 and -1.406 e.A'3

R1=2||Fo| - [Fell/ZIFOI, wR2=[>:(w|1=,2 - Ff|)2/Z(wF.,2)21”2

146

Table 7-6. Atomic coordinates ( x104) for RENiA14Ge2 (RE: Sm, Gd, Dy, Ho, Tm, Lu,

 

 

Y).
Wyckoff x y z
Position

Sm 3a 0 0 0
Gd 0 0 0
Dy 0 0 0
Ho 0 0 O
Tm 0 0 0
Lu 0 0 0
Y 0 0 0

Ge 6c -3333 3333 575(1)

-3333 3333 570(1)

—3333 3333 562(1)

-3333 3333 559(1)

-3333 3333 553(1)

-3333 3333 548(1)

-3333 3333 567(1)

Ni 3b -3333 3333 -1667

-3333 3333 -l667

-3333 3333 -1667

-3333 3333 -1667

-3333 3333 -l667

-3333 3333 -1667

-3333 3333 ~1667

Al(l) 6c -3333 3333 1444(1)

-3333 3333 . 1439(2)

-3333 3333 1442(2)

-3333 3333 1438(1)

-3333 3333 1439(1)

-3333 3333 1435(2)

-3333 3333 1441(2)

Al(2) 6c 3333 -3333 889(1)

3333 -3333 892(2)

3333 -3333 888(2)

3333 -3333 885(1)

3333 -3333 884(1)

3333 -3333 884(2)

3333 —3333 886(2)

 

147

Table 7-7. Anisotropic displacement parameters (A2 x 103) for RENiAl,Ge2 (RE: Sm,

 

 

Gd, Dy, Ho, Tm, Lu, Y).
U11 U22 U33 U23 U13 U12
Sm 7(1) 7(1) 5(1) 0 0 4(1)
Gd 9(1) 9(1) 0(1) 0 0 4(1)
Dy 7(1) 7(1) 10(1) 0 0 4(1)
Ho 3(1) 3(1) 2(1) 0 0 1(1)
Tm 6(1) 6(1) 5(1) 0 0 3(1)
Lu 19(1) 19(1) 15(1) 0 0 10(1)
Y 2(1) 2(1) 5(1) 0 0 1(1)
Ge 7(1) 7(1) 4(1) 0 0 4(1)
9(1) 9(1) 0(1) 0 0 4(1)
6(1) 6(1) 10(1) 0 0 3(1)
2(1) 2(1) 2(1) 0 0 1(1)
5(1) 5(1) 4(1) 0 0 2(1)
17(1) 17(1) 14(1) 0 0 9(1)
7(1) 7(1) 11(1) 0 0 4(1)
Ni 6(1) 6(1) 2(1) 0 0 3(1)
11(1) 11(1) 0(2) 0 0 5(1)
5(1) 5(1) 10(1) 0 0 3(1)
1(1) 1(1) 1(1) 0 0 0(1)
4(1) 4(1) 2(1) 0 0 2(1)
16(2) 16(2) 10(2) 0 0 8(1)
6(1) 6(1) 10(1) 0 0 3(1)
Al(l) 10(1) 10(1) 6(1) 0 0 5(1)
10(2) 10(2) 4(3) 0 0 5(1)
8(2) 8(2) 13(2) 0 0 4(1)
2(1) 2(1) 6(1) 0 0 1(1)
6(1) 6(1) 6(2) 0 0 3(1)
15(2) 15(2) 23(4) 0 0 8(1)
8(2) 8(2) 12(2) 0 0 4(1)
Al(2) 9(1) 9(1) 8(1) 0 0 5(1)
12(2) 12(2) 1(2) 0 0 6(1)
9(1) 9(1) 12(2) 0 0 4(1)
3(1) 3(1) 3(1) 0 0 1(1)
8(1) 8(1) 4(2) 0 0 4(1)
20(3) 20(3) 9(3) 0 0 10( 1)
9(2) 9(2) 12(2) 0 0 5(1)

 

The anisotropic displacement factor exponent takes the form: -21t2[h2a*2U11 + +
2hka*b*U12]

148

Table 7-6. Selected bond distances for SmNiAl,,Ge2

 

Bond Distance (A)
Sm-Ge 2.9733(5)
Ge-Al(2) 2.5664(8)
Ge-A1(1) 2.702(2)
Ni-Al(2) 2.420(2)
Ni-Al(l) 2.4731(6)
Al(1)-A1(1) 2.749(2)
Al(1)-Al(2) 2.936(1)

149

3. Results and Discussion.
Synthesis.

Germanium is even more soluble in molten aluminum than Si. Based on this we
set out to use aluminum flux to prepare the Ge analog of szNi(NixSi1.x)Al4Sig.8 It was
found that indeed the aluminum flux approach can be extended to germanium

compounds, however, isostructural analogs to corresponding Si compounds are rarely
observed. The new compound, RENiAl4Ge2, forms as well shaped hexagonal plates when

the RE:Ni:Ge react in molten A1 in the ratio 1:1:n (n=2-5). When the RE:Ni:Ge ratio is
2: 1: n, a new tetragonal phase REzNiAl4Ge2 is obtained,8 which strongly implying that in
this system alone the synthetic chemistry is relatively rich.

This structural arrangement appears stable with the later RE metals where the
earlier RE ion seem to favor tetragonal structures such as the previously mentioned
REZNiAhGe2 and REZNiAlngegy (x~ 0.24, y~ 1.34) exhibiting the tetragonal based
N iA14Ge2 unit. Sm however appears to have the ability to form in each of the separate

structure types likely due to the fact that it's size falls in the center of the RE se1ies.

Structural Description.

RENiA14Ge2 adopts a well-ordered rhombohedral structure with NiAl4Ge2 layers
separated by Sm atoms, see Figure 7-1. The layers possess trigonal symmetry, as seen in
chapter 8 and unlike the tetragonal units seen in chapters 3 through 6 and are made of a
continuous network of A1 and Ge atoms. Each layer is approximately 8 A thick and
contains seven atomic layers of Ge and Al in the stacking sequence Ge-Al-Al-Ni-Al-Al-

Ge. This generates Al-Al bonding and a puckered aluminum layer in the middle of the

150

NiAl4Ge2 layer with an As-type structure. This structure type requires that both surfaces

of each NiAlr4Ge2 layer be terminated with Ge atoms, see Figure 7-2. A closer look at the
structure reveals umbrella like Ge atoms which requires all Ge-Al bonds to be directed on
the same side of the Ge atom and towards the center of the NiAl4Ge2 layer, see Figure 7-
2. Presumably, the opposite side of the Ge atoms involves lone pairs of electrons
interacting with the Sm atoms. Similar Ge symmetry is also seen in CaAlzGeZ.9

The puckered Al-hexagons (cyclohexane chair conformation) in the As-type layer
are large enough (Al-Al bond lengths of 2.749 A) to accommodate the relatively small Ni
atoms. Each Ni atom is thus surrounded not only by the six Al atoms in the As-type layer
but also by two Al atoms above and below this layer to achieve an eight- coordinate
environment with Ni-Al bonds ranging from 2.40 to 2.47 A, see Figure .7 -2.

The RE atoms occur in planes perpendicular to the c axis close packed so that they form a
perfect triangular net. The triangles are equilateral with RE-RE edges of 4.12 A in
szNiAl4Ge2. The spacing between the Sm layers is 10 A. Therefore, the well separated
planar triangular arrays of paramagnetic Sm atoms have the potential of creating spin

frustration, which could lead to unusual magnetic phenomena.lo

151

8.88

r.“
I.
.w

 

$«.80 O
. '§!

0

 
  

. o .8.

®

 

@

.I

  

‘ka'lM
”1”.
Ali;

6 68.88.

    

O

 
 

Figure 7-1. The structure of RENiA14Ge2 viewed down the a-axis.

152

    

 
   
   

\O/I RO/
041.13% 54"?

 

 

iii \' 7.

    
     
    
      

3.819118
.8 a):

     

   

Figure 7-2. The highlighted layer in RENiA14Ge2 showing Ni stuffed As-type layer in the

center and the umbrella—like arrangement of the Ge atoms at the outer edges.

153

 

Figure 7-3. Immediate coordination environments of the (a) Ge atoms exhibiting an
umbrella-like and (b) Ni atoms exhibiting bonding to the six Al atoms in the As-type

layer along with 2 more Al atoms situated above and below the layer.

154

Magnetic Properties.
The most intriguing property of RENiAl4Ge2 is its magnetic behavior. The ideal

triangular arrangement of RE atoms in parallel planes separated by a 10 A spacing,
suggests that nearest neighbor magnetic RE-RE interactions within the plane will be
substantially stronger than those between planes. This could set the stage for an
interesting case of geometrical spin-frustration.“ In ideal two-dimensional triangular
spin lattices the magnetic ground state is degenerate and unstable and as a consequence

unusual magnetic phenomena can arise.9
SmNiAl4Ge2 was first studied as a function of magnetic field and temperature.

The compound exhibits a broad maximum in the zero-field cooled susceptibility, see
Figure 7-4, and the onset of irreversibility between zero-field cooled and field-cooled
magnetization which commences already at ~300 K. A well-defined hysteresis loop is
observed, see Figure 7-4, in which the magnetization begins to fully saturates at 10000
Gauss. Both these properties could be characteristic of disordered or spin-glass systems.
The compound, however, cannot be classified unequivocally as a spin glass because it
fails several important tests. First, the hysteresis loop does not shift in

position as a function of applied field during cooling of the sample and the AC
measurements show absence of frequency dependence of the susceptibility peak ( more
below for the other analogs). From the saturation value a magnetic moment of 0.28 BM
per formula unit can be calculated which is much lower than what would be expected

from Sm3+ ions (0.80 BM) and is even lower than the expected value for Sm2+ ions (1.65

155

 

 

 

 

2.5
I...
28
pg. .. FC
A o
2 g .0 .
(8315- °
E ' ° °
x \ ZFC . .
3 O o
E l~ o .
(D O 0
V o 8 '
O
0.5.. 3 .
O . O
0 ‘ ‘r i 1 f + 1
O 50 100 150 200 250 300 350

Temperature (K)

 

 

 

 

4
3.5.
3. .
g 2.5-
E on ' 2
<< 3 2- .
F 5 1.54'ZFC - '
é ' : '
1-1. p . g :
0.5-$333838” ' z .
0 FC
0 5'0 100 750 200 250 300 350

Temperature (K)

 

 

 

 

0.3
0.2‘ . . . o 0
0000'...
c A
.9 a) 0.1- l
o-o 6 .0
.‘e" E .s
a» 0‘ -.
82 o. -"
2 9i ' ' ‘ . l
-0.2- o 0 ° . . ...
03 1 1 : ‘r :
-1.510‘ -1 10‘ -5000 o 5000 1 10‘ 1.510‘
Field (G)

Figure 7-4. Magnetic behavior of SmNiAl,,Ge2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)
magnetization as a function of field.

156

BM). Although the magnetic moment strongly suggests the presence of Sm3+ ions rather
than Sm“ ions, the reason for the low 116,, value is not clear. In this context, we point out
that Sm magnetism, in general is very complicated and often the experimental values of
magnetic moments in its corresponding compounds deviate substantially from the
calculated ones.12

Magnetism of the Tb and Tm analogs show Curie-Weiss behavior at high
temperatures. The Tb analog exhibits and an antiferromagnetic transition at low
temperatures with a Tn of about 15 K while the Tm analog shows no such transition at
any measured temperature. When the high temperature data is fit 11,“ values of 9.65 and
7.16 B.M. are found for Tb and Tm, respectively. These values are very close to those
expected for the RE ions in a 3+ oxidation state, 9.72 B.M. for Tb and 7.63 B.M. for Tm,
further supporting a assumption that the Ni ions do not exhibit any overall magnetism in
these systems. The analogs show very different Weiss constants (8w) however with a
values of 18.80 K for TbNiAl,,Ge2 and -14.06 K for the Tm analog. This data seems to
suggest that at high temperatures ferromagnetic interactions are present in the Tb analog
while antiferromagnetic interactions are dominant in the case of TmNiAl4Ge2. When field
effects are studied in both systems a gradual increase in magnetization is seen with
increasing fields. This increase is rather dramatic in both cases until approximately
20,000 G for TbNiAl,,Ge2 and 10,000 G for the Tm analog where the slope flattens
significantly. The reason for this are not known as the expected magnetic saturation

values are still significantly higher then the magnetization reported at 55,000 G.

157

1/x

(mole / emu)

Magnetization

(BM. / mole)

(emu / mole)

1.4

1.2

0.8 f-

(16

(14

(12

30

25

2O

15

10

 

'IY'IIV'

 

 

1 4 1

 

20

100

Temperature (K)

 

Y f

Y'TTITII'

 

L l l A l l A J l l l A AL J A 1 L

 

LAL

 

150 200 250 300 350

Temperature (K)

 

VIT'

'rjj'l',

 

LJ.

1

 

r l 1 r 1 1 1 1

 

-410‘

o 210‘ 410‘ A 610‘
Field (G)

Figure 7-5. Magnetic behavior of TbNiAl,,Ge2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

158

1.5

 

 

 

 

 

 

 

 

 

 

 

 

 

D
1? 13
CE) .
xs\ :
g .:
3 0.5»;
"In....
0 1 1 ' '1' 0 9 o 910 1 h
0 50 100 150 200 250 300 350
Temperature (K)
50 .
40'— .
’5 ' o
E 30- .
em ~ .
X\ *
h on C '
'6 20— °
8 -°
10—
Obrrrlr all. 14111LL111411111111LL1
0 so 100 150 200 250 300 350
Temperature (K) ‘
4
I
.513 2f .-
wo 0
AS of /
‘8‘- * .-
382. — .-°
29 -2~ ,0.
'4” J
x:;r 14L1J11111411111111L
610‘ -410‘ -210‘ o 210‘ 410‘ 610‘
Field(G)

Figure 7-6. Magnetic behavior of TmNiAl,,Ge2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field.

159

The last analog magnetically studied, GdNiA14Ge2, exhibits an antiferromagnetic
transition at 12K and Curie-Weiss behavior at temperatures above the transition
temperature. The slope of the inverse xm plot corresponds to a system exhibiting a ueff of
7.97 B.M. which is the value for the Gd ions in a 3+ oxidation state once again indicating
no magnetic contribution from the Ni atoms as in the other analogs. The Weiss constant
from the fit of -2.47 indicates that the antiferromagnetic interactions are weakly present
even at high temperatures. The field dependent measurement deviate dramatically from
the other analogs in the fact that no signs of magnetic saturation are seen up to 55,000 G
but instead a steady increase in the magnetic response is seen at all fields measured.

The possible existence of spin glass behavior in these systems is of great interest
due to the geometric arrangement of RE ions. Spin glasses materials exhibit a variety of
characteristic responses in both DC and AC magnetic measurements that can be studied
to verify such behavior. For a geometrically frustrated magnetic three characteristics are
seen in the DC analysis. First the material must be antiferromagnetic at low temperatures
normally exhibiting divergence of the zero field cooled and field cooled xm values below
this transition. Second no long-range ordering can is present until well below I ®w1 and
finally the inverse susceptibility must be linear to temperatures well below I GW1 .‘3
Examining the experimental data, excluding the Tm analog since no transition is seen and
the Sm due to inherent complexities in its behavior,22 one does not see the characteristic
behaviors except for the presence of an antiferromagnetic trasition. For TbNiAl,,Ge2 the
I Gwi is ~19 K with the transition temperature is roughly 15 K while for the Gd analog

the transition temperature ~12 K is actually higher then the I Owl value of roughly

160

0.5

 

0.4 E 0
0.3 f. .

0.2 - °

x
(emu / mole)

0.1

 

 

YTYTIV‘

 

o l l I L l A L L L l L l A A J A A J 1 l I L A I

0 10 20 30 40 50
Temperature (K)

 

1/x
(mole / emu)
'33

 

 

11L]lllllLllllLllllllllllllllll

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

 

Magnetization
( BM. / mole)
O
I

I
N
l
.

.4L.

 

 

1'1 1 l l PLLL L 1 I ll; 1 l J 1

-610‘ -410‘ -210‘ 0 210‘ 410‘ 610‘
Field(G)

 

Figure 7-7. Magnetic behavior of GdNiAl,,Ge2 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)
magnetization as a function of field.

161

3 K. Also in both compounds the data no divergence is seen in the plots of field cooled
and zero field cooled data.

Along with the characteristic behavior in DC measurements spin glass materials
exhibit distinct behavior in AC measurements. The behavior includes a shifting of the
antiferromagnetic transition temperature with the application of varying AC fields as well
as the existence of a peak in the imaginary portion of the magnetic response (xm").”
Figure 7-8 shows the real portion of the magnetic response for GdNiAl,,Ge2 single
crystals, aligned with the applied field parallel the c axis, with various AC frequencies
applied. The response with the field perpendicular to the c axis exhibits similar. behavior
and is not shown. The data shows no shifting in the xm' peak with the application of AC
fields. Also upon inspecting the imaginary portion of the data (not shown) one does not
see any peaks in the data. It should be noted however that the xm" data is very noisy due
to the small sample sizes and the small values normally received in such analysis. Both
the AC and DC data than suggests that magnetic frustration is not present in these
systems even though the geometry is present to promote this type of behavior.

Interest remained however in the plataus seen in the field dependent data for the
analOgs. To examine this further field dependence study was conducted on the single
crystals of GdNiAl4Ge2. The data shows that GdNiAl,,Ge2 passes through several
metamagnetic states, see Figure 7-8, with increasing fields at temperatures lower than
Tm. This is likely why several of the analogs appear to magnetically saturate at moments
_ much lower than the total moment available when they are actually only exhibiting field
stabilized states. When samples are measured with the field perpendicular to the c axis

several field stabilized magnetic states are seen with reversibility between increasing and

162

decreasing fields. The response with the field aligned parallel however is slightly more
interesting in that it exhibits a non-reversibility as the sample passes through one of the
magnetic states. When increasing fields are applied the moment overlaps the response of
the perpendicular orientation until roughly 3235000 G. At this point a significant increase
in the moment is seen by a jump in the compounds magnetic response. When decreasing
passes past this point however the moment does not return to the lower state until about
i15000 G, see Figure 7-8. This hysteresis corresponds to the energy required to pass in
and out of the metamagnetic state.

From the magnetic data one can conclude that the RE ions are in a 3+ oxidation
state while the Ni ions do not exhibit a magnetic moment. This has been seen in several
systems presented in chapters of this thesis as well as in the literature.” Also spin glass
behavior is not evident for the compounds though the Gd analog does exhibit interesting
field dependent behavior that may explain the appearance of saturation in the other

analogs field dependence measurements.

4. Conclusions. .

Again the use of molten Al in exploratory synthesis has lead to the discovery of a
new structural series, RENiAl,,Ge2 (RE: Sm, Gd, Tb, Dy, Ho, Er, Tm, Lu, Y). The
structure of the series can be thought of as stacking of "NiA14Ge2" units, the hexagonal
form version, and a flat trigonal lattice of RE ions. The structure type appears to be very
stable forming both with various late rare earth metals though it does not appear to be
formed containing the early rare earth metals. This is likely due to a needed small size to

allow the rare earth ions to sit in a planar arrangement between the NiAl,Ge2 units.

163

The magnetism of these systems is of extreme interest due to the trigonal lattice of
magnetically active centers formed by the rare earth ions. This arrangement is ideal for
the formation of a frustrated magnetic system due to competing magnetic coupling
interactions from the odd number of neighboring rare earth ions. The magnetic
measurements however do not confirm the presence of a frustrated magnetic system for.
the compounds even though several analogs do exhibit low temperature
antiferromagnetic transitions. This absence of frustration even in the presence of the
trigonal lattice could be due to a variety of reasons and more detailed studies would be
needed to insolate the exact cause of the absence. Though the compounds to date are not
interesting as magnetic frustrated systems they do exhibit several interesting behaviors

with the application of high fields including metamagnetic transitions.

164

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

13
12- .
A ' '.:'o ' I 'V
2 ',:iIII8""5I.: '::" .
O 11_ vle' vv "
E . .
\ .'
:53 10d 3'
3 ,'
.E 9_ .l
x I 0 1000 Hz
l ' 100 Hz
8- ,' - 10 Hz
U
1!
7 I I I I l T I I I r r I I I I I I I I
8 8.5 9 9.5 10
Temperature (K)
10
5-
833
'5 O
S E
'S\ 0-1
0 o
82.
2‘5
-51 0
° Perpendicular
+Parallel
‘10 4I I I [4I I r {T f I l I I I I 4I I I I “I I I
-610 -410 -210 0 210 410 610‘

Field (G)

Figure 7-8. Real portion of the magnetic susceptibility of GdNiAl,,Ge2 single crystals
measured in an AC field (top). Field dependence response (bottom) of single crystal of
GdNiA14Ge2 orientated parallel and perpendicular the c axis.

165

Table 7-7. Temperature dependent magnetic behavior of RENiA14Ge2 (RE = Sm, Gd, Tb,

 

 

Tm)

RE “eff “111:0 9 (K) Tn (K)
Sm * 0.84 * 50

Gd 7.97 7.94 -2.47 l2/9**
Tb 9.65 9.72 18.80 3

Tm 7.16 7.63 -l4.06 NA

* The Sm analog does not exhibit Curie—Weiss behavior at any temperature
** T, appears to be around 12 K for powder samples and 9 K in single crystal

measurements.

166

References

 

‘ Sieve, B.; Chen, X.Z.; Cowen, J. A.; Larson, P.; Mahanti, S. D.; Kanatzidis, M.G.,
Chem Mater, 1999 11, 2451-2455.

2 (a) G.M. Sheldrick, In Crystallographic Computing 3 ; Sheldrick, G.M., Kruger, C.,
Doddard, R., Eds.; Oxford University Press: Oxford, England, 1985 l75-189.(b)
SHELXTL: Version 5, 1994, G.M. Sheldrick, Siemens Analytical X-ray Systems, Inc.,
Madison Wisconsin, USA.

3 SMART. Data Collection Software for the SMART System. Siemans AnalyticalX-ray
Instruments Inc., 1995

4 SAINT. Data Processing Software for SMART System. Siemans Analytical X-ray
Instruments Inc., 1995

5 Sheldrick, G.M. University of Gottingen, Germany. Manuscript to be Published

6 Sheldrick, G.M. SHELXL. Structural Determination Programs, Version 5.0; Siemans
Analytical X-Ray Instruments Inc.: Madison, WI, 1995

' 7 P. W. Selwood, Magnetochernistry, 2nd ed., 1956, Interscience Publishers; New York.
8 Sieve, B.; Trikalitis, P.N.; Kanatzidis, MG. Z. Anorg. Allg. Chemie 2002 628, 1568-
1574

9 Gladyshevskii, E. 1.; Kripyakevich, P. 1.; Bodak, O. I. Ukr. Phys. J. 1967 12, 447-452.
‘0 Schiffer, P.; Ramirez, A. P., Comments Cond. Mat. Phys., 1996 18, 21-50, references
therein.

” In an ideal geometrically frustrated Heisenberg spin system without disorder, the
ground states are not necessarily separated by energy baniers, leading to a continuum of

equivalent ground states. The ground state character of geometrically frustrated magnetic

I67

 

systems is an unsolved problem in condensed matter physics and has been an active area
of theoretical research. Both the nature of the ground state in different systems (e.g.,
lattice type, spin size and dimensionality) and the process of relaxation between
metastable states under the influence of quantum and thermal perturbations have been the
subject of controversy. While much theoretical effort has been directed toward these
issues, the problems are difficult to resolve either analytically or computationally (due to
the high degeneracy of ground states), and there is a strong need for complementary
experimental work to focus the theoretical effort.

‘2 (a) Carlin, Richard L. in Magnetochemistry, Springer-Verlaq, Berlin, 1986. (b)
Theory Appl. Mol. Paramagn., Edited by Edward Boudreaux, Wiley, New York, 1976
257-270

’3 Schiffer, P.; Ramirez, A.P. Comments Cond. Mat. Phys. 1996 18, 21-50

1‘ Li, D.X.; Shiokawa, Y.; Nozue, T.; Kamimura,T.; Sumiyama,K. J. Magn. Magn.
Mater. 2002 241, 17—24

‘5 Gil, A.; Szytula, A.; Tomkowicz, Z.; Wojciechowski, K. J. Magn. Magn. Mater. 1994

129, 271-278 and references 7 and 8 therein.

168

EXPLORATORY SYNTHESIS OF QUATERNARY RARE EARTH TRANSITION
METAL ALUMINUM TETRELIDES: UTILIZING MOLTEN AL AS A SOLVENT

Volume H

By

Bradley J. Sieve

A DISSERTATION
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of
DOCTOR OF PHILOSOPHY

Department of Chemistry

2002

Chapter 8.
Synthesis and Characterization of the N 1 Rich Phases RE],
xM2A15,_ySiy (RE: Nd, Sm, Tb, Tm, Yb, Y; M: Ni and Pd) and
RE2_,M2A14Tt2(Al1,,Tty)(All.,th)2(RE= Sm, Er, Dy; M= Ni, Co;

Tt= Si, Ge)

1. Introduction.

Recently we have reported the synthesis and properties of many quaternary
transition metal aluminium silicide or germinide compounds formed in liquid aluminum
including SrnzNi(Si1.,,,Ni,,)Al4Sigl RENiAl.,Ge2 (RE: Sm, Tb, Y),2 RE,,Fe2.,,,Al7_,Si8 (RE:
Ce, Pr, Nd, Sm),3 szNiAl4Ge2, and CeZNiAlg,Ge,_y (x~0.24, y~1.34).4 These new
compounds exhibit a wide variety of structural arrangements and several of them share
similar building units, such as the "NiAl,Tt2" layer. These compounds are also of interest
due to the apparent lack of magnetic moments associated with the transition metals.

Continuing our investigations we decided to explore the chemistry of the systems
under transition metal-rich conditions. To this point in the thesis the work has
concentrated on systems were the reactions exhibited a rare earth to transition metal ratio
of greater than or equal to one. In this chapter we examine what happens when this ratio
is smaller than one. We describe here two transition metal rich systems RE,_,,M2A15,ySiy
(RE: Nd, Sm, Tb, Tm, Yb, Y; M: Ni and Pd) and RlE,,.,,M2Al,,Tt2(Al,,yTty)(Al,,,Tt,)2
(RE: Sm, Er, Dy; M: Ni, Co; Tt= Si, Ge). We report the synthesis, structure and

preliminary charge transport and magnetic properties of these two structural series.

169

2. Experimental Section.
Synthesis.
Reagents.

Sm, 99.9%, metal chips, Research Chemicals, Phoenix, AZ; Nd, Tb, Dy, Er, Tm,
Yb, Y 99.9%, -40 mesh, Cerac, Milwaukee, WI; Ni, 99%, 325 mesh, Sargent, Buffalo
Grove, 11; Pd 99.95%, -325 mesh, Cerac, Milwaukee, WI; Si, 99.96%, -325 mesh, Cerac,

Milwaukee, WI; Ge, 99.999%, 3-6 mm pieces, Cerac, Milwaukee, WI.

Synthetic Method.

RE,_,,M2A15,ySiy was formed by mixing RE:M:Al:Si in a 1:2:20:2 molar ratio in a
Nz-filled dry box and placing them inside an alumina tube which was sealed within a
fused silica tube under vacuum. The mixture was heated at 1000 °C for 5 hours and
cooled to 850 °C over 2 hours. This temperature was maintained for 72 hours and then
cooled to 50 °C at a rate of -22° C per hour. The RI7,,,,,M2A1,,Tt2(All,yTty)(Al,,,Tt.L)2 phase
was formed through mixing reactants in a 3:4:20:6 RE:M:Alth ratio which was placed
inside an A1203 crucible and sealed under vacuum. This mixture was then heated as
described above for RE,,,M2A15,ySiy.

The compounds after heating were isolated from the excess Al flux by placing
them in 5M NaOH solution for 24 hours. Once freed from the matrix the products were
washed and dried revealing silver shiny crystals with a very distinct hexagonal rod shape
(desired phase) and silver plates (impurities). Both reactions yielded mixtures of phases
though the majority of the product was the target compound (>80 %). Overall yield of

products appears to be only about 50 percent based on the Ni amounts used in the

170

reaction. The balance of the reactants are unaccounted for though likely formed base
sensitive products which were destroyed during the isolation phase. The impurity plates
present were found to include RE/Al/Si phases, rare earth rich phases, Tm/Al/Si phases
and elemental Si plates depending on the specific system. The desired products can be
easily separated from the bulk however by selection of single crystals due to the unique
morphology of the title compounds (see synthesis section below). It is difficult however,
to separate the two title compounds from each other if both present as they exhibit nearly

identical morphologies.

Physical Measurements.
EDS Analysis.

Quantitative microprobe analysis was performed with a JEOL JSM-6400
Scanning Electron Microscope (SEM) equipped with Noran Energy Dispersive
Spectroscopy (EDS) detector. Data were acquired using an accelerating voltage of 25 kV
and 100 sec accumulation time: Standards were recorded under the same experimental
conditions to yield correction factors. The elemental analysis for RE,,,M2A15,,Si, yielded
a formula of 0.6 RE: 1.9 M: 3.6 A1: 1.3 Si while analysis of RE,,,M2AI,Tt2(Al,_yTty)(All,
zTt,,)2 yielded ratios of 1 RE: 1.5 M: 4 A1: 1.8 Si both within experimental error of the

final crystallographically refined formulas of the compounds.
Single Crystal X-ray Crystallography

Single crystal X-ray diffraction data for RE,,,,M2A15,ySiy and RE2.,M2A14Tt2(Al,,

yTty)(Al,_th,)2 were collected at 298 K on a Siemens Platform CCD diffractometer using

171

Mo K01 (A=0.71069 A) radiation. The SMART software5 was used for the data
acquisition and the program SAINT‘5 was used for the data extraction and reduction. An
empirical absorption correction using SADABS7 was applied and the structures were
solved and refined with the SHELXL software. 8 The crystallographic and refinement
details are listed in Table 8-1 through 8-4, 8-8 and 8-9. The fractional atomic positions,
anisotropic displacement parameters (U values) of the non-disordered positions are listed
in Tables 8-5, 8-6, 8-10 through 8-13. Bond distance for Nd,,,,Ni2A15_ySiy are listed in
Table 8-7 while distances for Sm2_,CozAl,,Ge2(Al,,yGey)(All_,Ge,)2, Dy2_,NizAl4Ge2(Al,_
yGey)(Al,_,Ge,)2 and Sm2_,,Ni2Al,,Si2(Al,,ySiy)(Al,,,Si,)2 are listed in 8-14. The reported
occupancy values for the positions Al(3) in RE,_,,M2Al,.ySiy and Al(2) in REZ,

,M2A14Tt2(Al,.yTty)(Al,_,Tt,)2 correspond to refined total occupancies after the correction

has been made for the fact that only 1/3 of the position is present at any time.

Charge Transport Measurements.

DC electrical conductivity and thermopower measurements were performed on
selected single crystals of Y,,,Ni2Als,ySiy. Conductivity measurements were conducted
with the conventional four-probe technique.9 Thermopower measurements were made

using a slow AC technique as described elsewhere.10
Magnetic Characterization.

Magnetic susceptibility for Yb,,,,NizA15,ySiy and Dy2_,,Ni2Al,,Si2(Al,_),Siy)(Al,_,Si,)2

single crystals were measured as a function of both temperature and field using a MPMS

172

Table 8-1. Crystal data and structure refinement for REHMZAI Si (RE = Nd, Sm).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (1.1)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R

Completeness to 0,,W

int

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

Nd,_,,Ni,.4.1,.,s1y
325.82

298(2) K
0.71073 A
P63/mmc (#194)
a = 4.1412(12) A
c = 15.735(7) A

5-y y

8111,,,,1~Ii,.4.1,,,81y
328.98

298(2) K
0.71073 A
P63/mmc (#194)
a = 4.1153(6) A
c = 15.718(3) A

233.69(14) A3 230540) A3
2, 4.63 Mg/m3 2, 4.74 Mg/m3
8.047 mm'1 21.421 mrn'1
168 356
0.2 x 0.05 110.05 mm 0.3 x 0.1 x 0.1 mm
2.59 to 28.21° 2.59 to 29.94°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-20<=l<=20 -22<=l<=22
2109/ 145 2404/ 162
0.0313 0.0570
100.0 % 100.0 %
Full-matrix least-squares on F2
145/0/15 162/0/15
1.900 1.273
R1=0.0538 R1=0.0418
wR2 = 0.1232 wR2 = 0.1139
R1 = 0.0556 R1 = 0.0506
wR2 = 0.1249 wR2 = 0.1179
0.012(5) 0.054(9)

2.839 and -3597 e/A3

R1=Z||F°| - IF.ll/ZIF.I. wR2=[E(w|F°2 - F,2|)2/E(wF.2)2]”2

173

1.972 and -1333 e/A3

Table 8-2. Crystal data and structure refinement for RE1,,,M2Als,ySiy (RE = Tb, Tm).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected/ unique
R18
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole
R1=2||F0| - chll/EIFOL wR2=[2(w|F,,2 - Fc2|)2/E(WF,,2)2]”2

Tb,,,.,,Ni,Al,_,Siy Tm(,_54Ni2A1,,ySiy -
350.80 343.49
298(2) K 293(2) K
0.71073 A 0.71073 A
P63/mmc (#194) P63/mmc (#194)
a = 4.1490(8) A a = 4.0697(6) A
c = 15.678(4) A = 15.591(3) A
233.73(9) A3 223.62(6) A3
2, 4.98 Mg/m3 2, 5.10 Mg/m3
19.261 mm'l 32.951 mrn‘l
240 g 428
0.1 x 0.75 x 0.75 mm 0.3 x 0.2 x0.2 mm
2.60 to 28.24° 2.61 to 29.94°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -4<=k<=4
-20<=l<=20 -21<=1<=21
2217/ 145 1376/ 158
0.0382 0.0648
100.0 % 100.0 %
Full-matrix least-squares on F2
145/0/15 158/0/15
1.767 1.368
R1 = 0.0318 R1: 0.0291
wR2 = 0.0768 wR2 = 0.0708
R1 = 0.0338 R1: 0.0411
wR2 = 0.0771 wR2 = 0.0766
0.014(4) 0.037(6)

1.760 and -2495 e/A3

174

3.112 and -1.283 e/A3

Table 8-3. Crystal data and structure refinement for RE1_,,M,_A15,ySiy (RE = Yb, Y).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
Rint
Completeness to 0max

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>2o(I)] '

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

Yb0_52Ni2Al,_ySiy
342.24

298(2) K

0.71073 A
P63/mmc (#194)

a = 4.091(5) A

c = 1559(3) A
226.0(5) A3

2, 5.03 Mg/m3
17.953 mm"

353

0.4 x 0.1x 0.1mm
2.61 to 28.27°

Y0_,,Ni2Al,,ySiy
301.18

298(2) K

0.71073 A

P63/mmc (#194)

a = 4.1012(10) A

c = 15.652(5) A
227.99(11) A3
2.4.39 Mg/m3
13.934 mm"

322

0.1 x 0.05 x 0.05 mm

2.60 to 28.15°

-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-20<=l<=20 -20<=l<=20
2033/ 144 1983/ 142
0.0688 0.0422
100.0 % 100.0 %
Full-matrix least-squares on F2
144/0/15 142/0/15
1.190 1.084
R1 = 0.0781 R1 = 0.0369
WR2 = 0.2074 WR2 = 0.1095
R1 = 0.0784 R1 = 0.0369
WR2 = 0.2105 WR2 = 0.1095
005(2) 0.064(13)

4.312 and -5125 e/A3

R1=ZIIFol - IFcll/ZIFJ, WR2=[Z(w|F02 - F,2|)2/2(w1=,2)2]m

175

2.816 and -1.367 e/A3

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (1.1)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
Rint
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

Table 84. Crystal data and structure refinement for Y1,,Pd2A15_ySiy.

Y0_35Pd2A15,ySiy
283.40

298(2) K

0.71073 A
P6(3)/mmc (#194)
a = 4.2798(6) A

c = 16.255(3) A
257.85(7) A3

2, 5.471 Mg/m3
18.749 mm'1

382

0.4 x 0.1 0.1 mm
2.51 to 28.22°
-5<=h<=5
-5<=k<=5
-21<=1<=21
2828/ 156
0.0416

100.0 %
Full-matrix least-squares on F2
156 / 0/ 15 I
0.864

R1 = 0.0425

wR2 = 0.1083

R1 = 0.0425

wR2 = 0.1083
0.061(10)

1.375 and -1.655 e/A’3

R1=2||1=,| - |F,||/2|F,|, wR2=[2(w|F02 - F,2|)2/z(wF.2)2]‘”

176

Table 8-5. Atomic coordinates ( x104) for RE1_,M2A1

 

 

Wyckoff x y 2 Occupancy
posrtion

Nd 2c 3333 6667 2500 0.52
Sm 3333 6667 2500 0.51
Tb 3333 6667 2500 0.62
Tm 3333 6667 2500 0.54
Yb 3333 6667 2500 0.52
Y 3333 6667 2500 0.55
Y 3333 6667 2500 0.35

Ni 4f 3333 6667 6101(1) 1

3333 6667 6103(1) 1

3333 6667 6103(1) 1

3333 6667 6116(1) 1

3333 6667 6118(2) 1

3333 6667 6112(1) 1

Pd 3333 6667 6099(1) 1

Al(l) 4e 0 0 1315(4) 1

0 0 1332(4) ' l

0 0 1329(2) 1

0 0 1364(2) 1

0 0 1350(2) l

0 O 1345(2) 1

0 0 1369(4) 1

 

177

Table 8-5. (continued) atomic coordinates ( x104) for RE1.,M2A15,ySiy.

 

 

Wyckoff x y 2 Occupancy
position
Al(2) 4f 3333 6667 469(3) 1
3333 6667 469(3) 1
3333 6667 471(2) 1
3333 6667 476(2) 1
3333 6667 470(4) 1
3333 6667 475(2) 1
3333 6667 474(3) 1
Al(3)* 6h 5373(18) 750(40) 2500 l
5360(20 710(40) 2500 1
5415( 10) 830(20) 2500 1
533 8( 1 2) 680(20) 2500 1
5392(12) 780(20) 2500 1
5344(] 1) 690(20) 2500 l
5370(18) 740(40) 2500 1

 

* occupancy values reported are after correcting for only 1 in 3 atoms present

178

Table 8-6. Anisotropic displacement parameters (A2 x 103) for RE1_,M2A1

Sm, Tb, Tm, Yb, Y; M: Ni, Pd).

Si

5-y y

(RE: Nd,

 

 

U11 U22 U33 U23 U13 U12
Ni 9(1) 9(1) 6(1) 0 0 5(1)
5(1) 5(1) 9(1) 0 0 2(1)
9(1) 9(1) 11(1) 0 0 4(1)
4(1) 4(1) 10(1) 0 0 2(1)
8(1) 8(1) 16(2) 0 0 4(1)
4(1) 4(1) 13(1) 0 0 2(1)
Al(l) 11(2) 11(2) 16(2) 0 0 6(1)
3(1) 3(1) 14(3) 0 0 1(1)
9(1) 9(1) 15(2) 0 0 4(1)
2(1) 2(1) 8(1) 0 0 1(1)
7(2) 7(2) 13(2) 0 _ 0 3(1)
3(1) 3(1) 14(2) 0 0 1(1)
Al(2) 10(2) 10(2) 3(2) 0 0 5(1)
4(2) 4(2) 8(2) 0 0 2(1)
10(1) 10(1) 6(1) 0 0 5(1)
4(1) 4(1) 7(1) 0 0 2(1)
8(2) 8(2) 9(2) 0 0 4(1)
4(1) 4(1) 9(1) 0 0 2(1)

 

The anisotr0pic displacement factor exponent takes the form: -21t2[h2a*2U11 + +

2hka*b*U12]

179

Table 8-7 Selected Bond Distances (A) for NdHNiZAl s1

5-y y'

 

 

Bond Distances (A)
Nd-Al(1) 1.463(13)
Nd-Al(1) 2.966(10)
Nd-Al(2) 3.032(4)
'Nd-Ni 3.2495(17)
Ni-Al(1) 2.388(6)
Ni-Al(2) 2.4144(11)
Ni-Al(3) 2.590(2)
Al(l)-Al(1) l.61(2)
Al(l)-Al(1) 2.53(2)
Al(1)-Al(2) 2.799(4)
Al(2)—Al(3) 2.736(4)
Al(3)-Al(3) 2.810(5)

 

180

Table 8-8. Crystal data and structure refinement for RE2_,M2A14Tt2(Al,,yTty)(Al,_th,)2.

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R18

Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>20'(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole
R1=Z||Fo| - |Fc||/Z|Fo|, wR2=[}:.(w|F,,2 - F,,2|)2/Z(wF02)2]”2

SmmCozAlmGe,35
764.96

293(2) K

0.71073 A
P63/mmc (#194)

a = 4.154(14) A

c = 2997(12) A

ErL53Ni2A1537Ge,63
781.81

293(2) K

0.71073 A
P63/mmc (#194)

a = 4.1417(5) A

c = 29.346(5) A

448(3) A3 435.95(11) A3
2, 5.56 Mg/m3 2, 5.56 Mg/m3
24.795 mrn'l 30.168 mm'1
663 564
0.3 x 0.05 x 0.05 mm 0.2 x 0.1 x 0.1 mm
1.36 to 28.26° 2.78 to 28.25°
-5<=h<=5 -5<=h<=5
-5<=k<=5 -5<=k<=5
-38<=l<=38 -39<=l<=37
4126 / 273 4019 / 267
0.0626 0.0664
99.6 % 99.6 %
Full-matrix least-squares on F2
273/2/24 267/2/24
1.512 2.078
R1 = 0.0408 R1 = 0.0509
wR2 = 0.0851 wR2 = 0.1196
R1 = 0.0432 R1 = 0.0539
wR2 = 0.0858 wR2 = 0.1203
0.0017(4) 0.0166(19)

3.375 and -2359 e/A3

181

2.455 and -5212 e/A3

Table 8-9. Crystal data and structure refinement for RE2,,M2A14Tt2(Al1_yTty)(All_,th)2.

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R

Completeness to 0m,

int

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>20'(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=2||Fol - chll/EIFOI, wR2=[‘£(w|F,2 - 11,21)2/2(wr:,=’-)2]V2

I Dy 1.53Ni2A1546Ge354

768.94

293(2) K

0.71073 A
P6(3)/mmc (#194)
a = 4.1581(7) A

e“ = 29.436(7) A
440.7505) A3

2, 5.794 Mg/m3
26.516 mm"

656

0.4 x 0.2 x 0.2 mm
1.38 to 28.45°
-5<=h<=5
-5<=k<=5
-37<=1<=37

3767 / 266

0.1038

97.8 %

DylglNizAl7Si2
607.80

293(2) K
0.71073 A

'P6(3)/mmc (#194)

a = 4.1134(6) A
c = 29.218(6) A
428.13(12) A3
2.4.715 Mg/m3
14.014 mm"
498

0.2 x 0.1x 0.1mm
2.79 to 28.45°
-5<=h<=5
-5<=k<=5
-38<=l<=37
3712 / 258
0.0965

96.3 %

Full-matrix least-squares on F2

266 / 2 / 24
2.309

R1 = 0.0695

wR2 = 0.1560

R1 = 0.0697

wR2 = 0.1561
0.011(2)

3.938 and -4.664 e/A3

182

258 / 0 / 21
1.151

R1 = 0.0467

wR2 = 0.1096

R1 = 0.0494

wR2 = 0.1112
0.030(4)

2.325 and -2074 e/A3

Table 8-10. Atomic coordinates ( x104) for RE,,,M2A14Ge2(All_yGey)(Al,,zGe,)2.

 

 

Wyckoff x y 2 Occupancy
position
Sm(l) 2a 0 0 5000 1
Er(l) 0 0 5000 1
Dy(l) 0 0 5000 1
Sm(2) 2d -3333 -6667 2500 0.52
Er(2) -3333 -6667 2500 0.53
Dy(2) 3333 6667 2500 0.53
Co 4f 3333 -3333 1745(1) 1
Ni 3333 -3333 1730(1) 1
3333 -3333 1729(1) 1
Al(l) 4f -3333 3333 3537(2) 1
-3333 3333 3540(3) 1
-3333 3333 3540(3) 1
Al(3) 4f —6667 6667 4073(2) 1
-6667 6667 4094(3) 1
-6667 6667 4088(3) 1

 

183

Table 8-10. (continued) atomic coordinates ( x104) for RE2_,M2Al4Ge2(Al1_yGey)(Al1_
zGe,)2.

 

 

Wyckoff x y 2 Occupancy
position
Ge(1) 4f -3333 3333 4409(1) 1
-3333 3333 4428(1) l
-3333 3333 4423(1) . 1
Al(2)* 6h 4634(10) 9270(20) 2500 0.21
4640(20) 9290(40) 2500 0.43
4580(20) 9170(50) 2500 0.56
Ge(2)* 6h 4634(10) 9270(20) 2500 0.79
4640(20) 9290(40) 2500 0.57
4580(20) 9170(50) 2500 0.44
Al(4) 4f 0 0 3136(1) 0.47
3108(3) 0.47
0 0 31 16(4) 0.45
Ge(4) 4f 0 O 3136(1) 0.53
0 3108(3) 0.53
0 31 16(4) 0.55

 

* occupancy values reported are after correcting for only 1 in 3 atoms present

184

Table 8-11. Anisotropic displacement parameters (A2 x 103) for

RB,,,M,A1,oe,(A1,,,Ge,)(A1,,,oe,),.

 

U11 U22 U33 U23 U13 U12

 

Sm(l) 2(1) 2(1) 4(1) 0 0 1(1)
Er(l) 6(1) 6(1) 13(1) 0 0 3(1)
Dy(l) 13(1) 13(1) 9(1) 0 0 6(1)
Co 3(1) 3(1) 3(1) 0 0 2(1)
Ni 6(1) 6(1) 15(2) 0 0 3(1)
12(1) 12(1) 9(2) 0 0 6(1)
Ge(1) 2(1) 2(1) 2(1) 0 0 1(1)
5(1) 5(1) 14(2) 0 0 3(1)
12(1) 12(1) 10(2) 0 0 6(1)
Al(l) 4(2) 4(2) 3(3) 0 0 2(1)
3(2) 3(2) 13(4) 0 0 1(1)
8(3) 8(3) 7(4) 0 0 4(1)
Al(3) 6(2) 6(2) 3(2) 0 0 3(1)
7(3) 7(3) 16(4) 0 3(1)
16(3) 16(3) 6(4) 0 8(2)

 

The anisotropic displacement factor exponent takes the form: -21tz[hza*2U11 + +

2hka*b*U12]

185

Table 8-12. Atomic coordinates ( x104) for Dy2_,Ni2Al4Siz(Al,,ySiy)(Al,,zSi,)2.

 

 

Wyckoff x y 2 Occupancy

pos1tron
Dy(l) 0 0 5000 1
Dy(2) -3333 -6667 2500 0.51
Ni 3333 -3333 1744(1) 1
Al(l) -3333 3333 3541(2) 1
Al(3) -6667 6667 4106(2) 1
Si(l) -3333 3333 4415(1) 1
Al(2) 4631(18) 9260(40) 2500 1
Al(4) 0 0 3110(2) 1

 

Table 8-13. (continued) anisotropic displacement parameters (A2 x 103) for

Dy,,,Ni,A1,s12(A1,,,Si,)(A1,,,81,),.

 

 

U11 U22 U33 U23 U13 U12
Dy(l) 5(1) 5(1) 17(1) 0 0 3(1)
Ni 6(1) 6(1) 21(1) 0 0 3(1)
81(1) 8(1) 8(1) 10(2) 0 0 4(1)
Al(2) 8(1) 8(1) 14(2) 0 0 4(1)
Al(3) 8(1) 8(1) 14(2) 0 0 4(1)

 

The anisotropic displacement factor exponent takes the form: -2n2[h2a*2U11 + +

2hka*b*U 12]

186

Table 8-14. Selected bond distances (A) for Sm2,,,CozAl,,Ge2(Al,,yGey)(All,,Ge,)2 and
DYz-xNi2A14Gez(Al1-yGey)(Al1-20692.

 

 

Bond Bond distances (A) Bond Bond distances (A)
Sml-Ge(1) 2.981(8) Al(3)-Al(1) 2.885(9)
Dy -Ge( 1) 2.941(2) 2.893(7)
Co-Al(4)/Ge(4) 2.424(8) RE(2)-Al(2)/Ge(2) 1.462(9)
Ni 2.443(7) 1.50(2) ‘
M-Al(2)/Ge(2) 2.447(9) RE(2)-Al(2)/Ge(2) 297(1)
240(1) 2.950)
Ge(1)-Al(3) 2.602(8) Al(2)/Ge(2)-Al(2)/Ge(2) 1.62(l)
2.595(4) 156(3)
Ge(1)-Al( l) 2.6 1(1) Al(2)/Ge(2)-M 2.447(9)
2.59(1) 2.443(7)
Al( 1 )-M 2.543(8) Al(2)/Ge(2)- Al(2)/Ge(2) 2.53(2)
2.527(3) 2.60(3)
Al(1)-Al(4)/Ge(4) 2.682(8) Al(2)/Ge(2)—Al(4)/Ge(4) 2.832(7)
2.705(7) 2.776(7)
Al(1)-Al(3) 2.885(9) Al(4)/Ge(4)—M 2.424(8)
2.893(7) 2.444(2)
Al(3)—M ' 2.45(1) Al(4)/Ge(4)—Al( 1) 2.682(8)
240(1) 2.705(7)
Al(3)-Ge(1) 2.602(8)
2.595(4)

 

First distance of each pair corresponds to the distances for Sm2.,C02Al,Ge2(Al,,yGey)(All,
,Ge,)2 while the second corresponds to Dy2_xNizAl4Ge2(Al,,yGey)(Al,,zGe,_)2.

187

Table 8-15. Selected bond distances (A) for Dy2_xNizAl4Si2(Al1,ySiy)(All,,Siz)2,

 

 

Bond Distances (A)
Dy -Si(1) 2.926(2)
Ni-Al(4) 2.4129(5)
Ni-Al(2) 2.395(5)
Si(1)-Al(3) 2.541(2)
Si(1)-Al(l) 2.553(6)
Al(1)-Ni 2.516(1)
Al(1)-Al(4) 2.689(2)
Al(1)-Al(3) 2.892(4)
Al(3)-Ni 2.483(5)
Al(3)-Si( 1) 2.541(2)
Al(3)-Al(1) 2.891(4)
Dy(2)-Al(2) 1.45(1)
Dy(2)-Al(2) 2.948(9)
Al(2)-Al(2) 1.60(2)
Al(2)-Ni 2.395(5)
Al(2)-Al(2) 251(2)
Al(2)-Al(4) 2.734(1)
Al(4)-Ni 2.4129(5)
Al(4)-Al(1) 2.689(2)

 

188

Quantum Design SQUID magnetometer. An initial study of field dependence was
conducted to find a suitable field for the variable temperature studies. The measurements
were then conducted under increasing temperature using a 500 G applied field. Field
dependent measurements, conducted at 5K, were carried out between i 55000 G for Yb,_
xN i2A15_ySiy while only i1000 G was measured for Dy2,,Ni2Al4Si2(Al,_ySiy)(All,,Si,)2. A
diamagnetic correction was applied to the data to account for core diamagnetism but no
correction was made for the sample container because the measured moment was well

over an order of magnitude smaller than the sample signal itself.

3. Results and Discussion.
Synthesis.

Reactions of RE, M and Tt in excess Al produces crystals of RE,,,M2A15,ySi, (RE:
Nd, Sm, Tb, Tm, Yb, Y; M: Ni and Pd) and RE2_,M2A14Tt2(Al,,yTty)(Al,,,Tt,)2(RE= Sm,
Er, Dy; M: Ni, Co; Tt= Si, Ge) up to several mm in length, see Figure 1. These phases
form as rod-like crystals often tapered at one or both ends. In several systems both
structure types co-exist in the product which may cause difficulties in identifying and
isolating each title phase separately. This separation can be achieved however through
elemental analysis or crystallographic screening of individual crystals. This coexistence
of the phases in products can easily be understood once the structural and compositional
relationships are examined (see structural description below). The crystals appear to be
stable in air, water, acids and bases however surface attack is apparent after crystals were

exposed to air for 24 hours at 1000° C.

189

 

Figure 8-1. SEM images of (a) Y,,,Ni,Al,,,s1y (b) Dy,,,Ni,1A.1,Si,(Al,_,Si,)(A1,_,Si,)2 and (c)

Tbl_,,NizA15(ySiy exhibiting typical hexagonal rod-like crystal morphology.

190

Structural Description.
RE ,,,MZAI5,,Siy

RE1_,M2A15,ySiy, exhibits a three-dimensional structure that can be thought of as
merging of 2 individual layers, a MAl,,_ySiy layer and a disordered RE,,,Al layer. These
layers stack in an alternating fashion along the c axis, see Figure 8-2. The MAlgySiy layer
can be thought of as the merging of two (A1,Si) layers, As-type, stuffed with Ni atoms,
Figure 8-3a. The layers merge through Ni-Al bonds with a bond distance of 2.590(2) A
and Al-Al bonds with a distance of 2.736(4) A in NdHNizAlgySiy. Ga/Ge structural
analogsll exhibit a Ga/Ge disorder in this layer which is likely also present in these Al/Si
analogs. However due to the similar scattering power of A1 and Si the existence of
disorder on the site could not be explored.

The second layer, REHAl, is a monatomic thick layer shown Figure 8-4. Here the
RE and Al interatomic vectors are only about 1.5 A, Figure 8-4a. Such distances are not
realistic and prompted an examination of the occupancies for the crystallographic
positions involved. It was found that indeed vacancies do exist on the sites with
approximately 2/3 of the RE ions present and 1/3 of the total Al atoms present. It should
be noted that the Pd analog of the structure shows significantly less RE atoms present
than the Ni analogs with only about 35% RE occupancy.

Similiar disorder has been seen in several structurally related phases including
RENi3A19,12 RE,,Pt.,A12,,l3 RED-mNizGasfldGex,” and Gd1_33PI3(Al,SI)8 and
Odo-67Pt2(Al,Si)5.” Parthé13 pr0posed an explanation stating that one can realistically
model the layer by removing 2/3 of the Al atoms and 1/3 of the RE atoms as shown in

Figure 8-4b. This model only exhibits reasonable Al-Al and RE-Al distances of 253(2) A

191

 

and 2.96(1) A, respectively. This new structure is best described as a supercell of the
original structure with cell dimensions of a‘l3 x a\l3. Initial support of this model can be
found experimentally by the fact that the refined occupancies do indeed refine to the
predicted values of 1/3 Al and 2/3 of the RE ions.

The existence of a larger structural cell is also shown by long exposure zone
photos taken of single crystals, Figure 85 From these photos two important observations
can be made. First the ab zone photo shows that indeed an a\/3 x a\l3 cell is present as
predicted due to an ordering of the REHAl layers. Second when the ac and be zone
photos are examined diffuse streaking is seen instead of normal Bragg reflections along
the c* axis indicating that the repeating unit along this axis is likely irregular in nature.
These observations lead one to conclude that the REHAl layers themselves are ordered,
but they do not repeat periodically along the c axis (i.e. stacking disorder). Several
attempts have been made to refine the larger structural cell using X-ray diffraction but all
have failed to date due to the inherent weak intensity and diffuse nature of the diffraction
spots caused by the stacking disorder in the extended structure.

The local atomic environments of the structure are shown in Figure 8-6 (the
triplets of Al(3) atoms contained in the REHAl have been excluded for clarity). Al(l) sits
inside a 6 membered ring of 3 Al(2) atoms and 3 Ni atoms, Figure 8-6a. While Al(2)
exhibits a 10 coordinate environment bonding to a 6 membered ring, containing 3 Al
atoms and 3 Ni atoms, and a NiAl3 umbrella units, see Figure 8-6b. In addition to these
interaction 3 RE ions are positioned above the ion at bonding distances of 2.979 A. The
Ni atom of the structure is contained within the As-type (A1,Si) layer and also exhibits

bonding to a neighboring As-type layer merging the layers through a Ni-Al bond. The RE

192

ion exhibits a 14 coordinate environment which is best described as sandwiched between
2 6-membered (Al(2),N i) rings with the addition of 2 Al(l) atoms bonding through the

center of the rings.

RE2_,A/12Al4th(Al,_,Tty)(Al,_,Tt,)2

RE2_,,M2A1,,Tt2(All,yTty)(Al,_,Tt,)2 crystallizes in the space group P63/mrnc with the
structure, shown in Figure 8—7, exhibiting two structural layers stacking along the c axis.
The first layer is a pure trigonal RE layer while the second layer is a RE,_,M2Al,Tt2(All,
yTty)(All,th,)2 layer, closely related to the REl,,,M2A15,ySiy structure (discussed above).
These layers stack in an alternating fashion along the c axis to form the total structure.

The RE1_,,M2A1,,Tt2(All,yTt,)(Al,,,Tt,)2 layer conceptually contains 3 merged
sublayers including an Ath As-type layer (layer A), see Figure. 8-8, a stuffed As-type
layer(layer B) and a RE,_,(All,yTty) layer isostructural to the RE,_,A1 layer (layer C) the
latter 2 are also seen in the structure of RE,,,,M2A15_ySiy (see above). These sublayers stack
displaying an A,B,C,B,A ordering to form the total RE1,,,M2Al4Tt2(Al1,,Tty)(Al,,,Tt,)2
layer. For comparison the total structure of RE,_,,M2A15,,ySiy can be described by the layers
stacking in an B,C,B,B,C,B ordering.

This group of compounds forms with both Al/Si and Al/Ge analogs so the Alth
distribution in the structure could be explored unlike the RE1_,M2A15_,Siy systems. Indeed
disorder was found between A1 atoms and tetrelide atoms on two separate positions in
this structure. The first instance of disorder, similar to what is expected for RE,,,M2A15_
Siy,12 was seen in layer B where the Al(4)/Ge(4) positions exhibit mixed Al/Ge

Y

occupancy. The second case of disorder is in layer C, the RE,,,,(A1,Tt) layer, with the non-

193

 

rare earth positions exhibiting Al/Ge disorder. This type of disorder is not seen however
in the Ga/Ge examples of the RE,,,,M2A15,ySiy system. 11 First evidence supporting this
assignment was seen when the non-rare earth position in Layer C was refined as pure Al
in that it led to greater than 175% occupancy,15 indicating more electron density was
needed on this position. Also several separate atoms acquired non-positive definite
temperature factors when pure Al in the position was refined. When Ge was added to the
site the total occupancy refined to normal values (shown in Table 8-10) and the
temperature factors became reasonable. Simultaneously the R values of the refinement
dropped in all analogs upon the addition of Ge to the site.

RE(2), Ni, Al(l) and Al(4) in R13,,,,,M2A14Tt2(Al,_yTty)(All_,Tt,)2 all exhibit similar
coordination environment as were seen in RE,,,M2A15_ySiy, see Figure 8.8 (here the
triplets formed by Al(2) have been removed for clarity). RE(l) atoms in the pure RE
layer separating the RE,.,,M2Al,,Tt2(Al,_),Tty)(All_,Tt,)2 layers exhibit an octahedral
arrangement bonding to 6 Tt atoms. The pure tetrelide position exhibits an umbrella like
arrangement bonding to 4 Al atoms along with 3 other interactions with RE ions at
distances of 2.934 A, see Figure 8-8b. This type of arrangement for a tetrelide atom has
also been seen in the hexagonal phase RENiA14Ge2 (see chapter 7) as well as in the
literature for compounds such as CaAlzGez.16 The final position, Al(3), exhibits a capped

trigonal prism arrangement bonding to 3 tetrelide atoms, 3 Al(l), and a Ni atom.
Charge transport properties.

Charge transport properties of Yl_,,NiA15,ySiy single crystals, Figure 8-9, are

typical of metal species in both the measured values and the temperature dependence.

194

 

 

Figure 8-2. Crystal structure of RE,_,M2Als_ySiy.

195

 

Figure 8-3. As-type building units of RE1_,M2A15,ySiy including (a) stuffed arsenic-type
layer seen in both compounds and (b) the Ath unstuffed version seen in RE,_

.M.AI.T8(A1..Tt.)(Al...To..

196

 

 

A , .2
/O\A \\
, \ /O
A o ,\
z \

2AlgySiy (a) layer with all disorder atoms shown and

Figure 8-4. REHAI layer of REMM

197

(b) model layer displaying the ordering of atoms to produce the a\/3 x a\l3 cell.

 

1" , . '1'»...

Figure 8-5. X-ray zone photos of Tbl_,Ni2Al,.,Si, collected for 10 minutes rotating along
the c axis (top) and rotating along the b axis (bottom) showing the a‘l3 x a‘lS supercell in

the ab plane and diffuse scattering along the c* axis.

198

a) RE

Al(1)

   

Figure 8—6. Local coordination environments of the atoms in REHMZAlMSiy.

199

Ath layer

RBI-XMZAW‘Z MA1(A11-thz) layer
(All-yTty)(A11-2th)2
RE 1-x(A11-yTty) layer

layer
MAl(A11-thz) layer

 

Ath layer

RE layer

 

 

Figure 8-7. Crystal structure of RE,_,M2AI,Tt2(Al,,yTt,)(Al,_,Tt,)2.

200

 

Figure 8-8. Coordination environments in RE2,,M2A1,,T12(A11_yTt,)(Al,,,Tt,)2.

201

 

 

 

 

 

 

 

300

A 250}

E _

o :

5: 200:

3» E

31501

.5 . .

325 100}

(I) .

‘D :

CE 50:
0:11L1111111LL311illllrLinlilki
0 50 100 150 200 250 300

Temperature (K)

8.

:2 7E3

\ :

E 63’.

a 52”:

E 45— .

Q. : Q

0 3; .

E E 0

2:) 27 7 /

'— 1? WM
0: 1 141L111 1 11

 

 

 

1 1 1 1 1 1 1 1 1 1 r 1 1 l 1 l 1 1
0 50 100 150 200 250 300
Temperature (K)

Figure 8-9. Charge Transport Properties of two selected single crystals of YHNiAlgySiy.

202

The exhibited thermopower values are very small, ~1 11V/K at room temperature while
the gradual decrease in thermopower values as temperatures approaches zero is also
consistent with this expectation. The compound exhibits resistivity values of about 150

[152-cm at I'OOII'I temperature.

Magnetic Properties.

The magnetic behavior of single crystal samples of Ybl,,NiAl,,ySiy, with the field
aligned parallel and perpendicular to the c axis, is shown in Figure 8-10. No distinct
transitions are apparent at low temperatures though a slight effect of cooling in a field is
seen with the field aligned along the c axis. At higher temperatures a transition appears to
take place at about 200 K for both orientations, seen in the l/xm plot, with magnetization
decreasing at a much greater rate past this temperature. Fitting the high temperature data
to Curie-Weiss behavior yields a 11eff of 4.43 B.M., in good agreement with the calculated
value of 4.50 B.M. for Yb in a 3+ oxidation state with the field aligned along the c axis, a
value of only 2.51 B.M. is observed. This type of magnetic behavior with broad high
temperature transitions has been previously studied in systems where the Yb ion
undergoes a valence fluctuation over themeasured temperature range.17 The field
dependence shows a weak linear response in the perpendicular arrangement, but when
oriented parallel the system shows a much more dramatic increase. In both cases however
only a small percentage of the total moment (< 0.5 B.M.) is accounted for at the
maximum field measured.

Dy2,xNi2Al4Si2(Al,,ySiy)(Al 1,281,), exhibits a low temperature antiferromagnetic

cusp at approximately 8 K, Figure 8-11. This transition is predominately seen with the

203

field aligned in the ab plane though subtle indications of it can also be seen with the field
parallel to the c axis. High temperature data, in the parallel orientation, obeys Curie-
Weiss behavior with a corresponding 11eff of about 14 BM. This value is slightly higher
than the calculated value of (10.6 B.M.) though is not unreasonable considering the small
sample size used (<1 mg) in the single crystal measurements. When the corresponding
measurements were conducted with larger powder samples (>3 mg) a 11eff of 10.54 B.M.
was determined indicating that the Dy in the samples is actually in the 3+ oxidation state
and the Ni ions do not contribute to the magnetism as has been the case in related
structures)?"15 With the field oriented perpendicular to the c axis an increase of l/xm with
increasing temperatures is seen though no linear response section that can be fit to Curie-
Weiss behavior. The field dependence for both orientations shows a linear increase with
increasing fields and no signs of magnetic saturation up to the maximum field measured
of 1000 G.

Of interest in this compound is the presence of a triangular lattice of pure RE ions
and the possibility of spin frustration in the system due to this geometric arrangement.l8
Measurements conducted to probe spin frustration however did not exhibit characteristics
of a frustrated system such as a shifting of the antiferromagnetic peak with the
application of an AC field, the existence of a peak in the imaginary section of the AC
measurements and divergence of zero field cooled and field cooled data at moderate

temperatures. '9

204

046

 

 

 

0.14
E 0.12
E 0.1
0)
,53 0.08
E
\ 0.06
3
E 0.04
3 II... I I I
.0000 I
0.02 ....::::'I
0 rerlrrAILILLL11L1414L3L’

 

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

 

 

 

 

 

 

 

 

mm
, r
A _ O
E 150 - ’
2 i 8
E E »
§ 2 100 e '
1— L” ’ l I
a: I
1 C I
.9 8 ' I I
g 50 00000 ' I . I
O Irnrlrrrrlrrnrlrrrrl11111111111111
0 50 100 150 200 250 300 350
Temperature (K)
05
A . ....I
c: > I
.9 .'
5m “ '
1.3:: _ l .
.§£§ ' .elocloO".‘..
*‘O 0 l- . 0'.
m p . . . .
CE .0 O 0 .
£2 -'
m. r I.-
v .-
L...
_O'5 ’- A L 1 1 A L l l A I 1 l L 1 L l L 1 L l L 1 L
4310‘ .410‘ -210‘ 0 210‘ 410‘ 610‘

Field (G)

Figure 8-10. Magnetic behavior of Ybl,,,NiAls,ySiy single crystals. Data shown with
squares corresponds to crystal alignment with the field parallel to the c axis. Circles

correspond to the field aligned perpendicular to the c axis.

205

 

 

_.
(I N

.‘v'vv—vvtvv1v'
I

xm
(emu / mole RE ion)

1

I...
(emu / mole RE ion)

0
(3
vv‘.

 

 

O

LAalkkAlA‘AIAA.l.AA1LLLJ AAAAA

. o 2 4 a a 10 12 14

Temperature (K)
'7‘.
\I.
'O'II I I
I
Ll[ALA—PALILAAJLAL!1!AL!L!A!1’AALL

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

 

 

 

 

Xm
(emu / mole RE ion)

 

 

AllllllJllllllALlllllll111

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

 

 

 

0.4 . i
03 I
E? . . I
c.9 0'2 I
QUJ I . . 0
«(am 0.1 I . . . o
N m ' . I
=- 0 I
o o . .
C:E . 0". I
8)\ '0.1 . . . I .
22. 43.2 . '
E1 I
I
-0.3 .
_0.4 .L 1 L 1 1 1 1 L A l A 1 1 1 l 1 1 1
~1000 -500 0 500 1000
Field (G)

Figure 8-11. Magnetic behavior of Dy2,,N izAl4Siz(Al l_ySiy)(Al ”$12), single crystals. Data
shown with squares corresponds to crystal alignment with the field parallel to the c axis.

Circles correspond to the field aligned perpendicular to the c axis.

206

4. Conclusions.

The reaction of rare earth metal transition metal and a tetrelide (i.e. Si or Ge) in
excess molten Al has lead to the discovery of two new transition metal rich structural
series RE,,,,M2A15_ySiy (RE: Nd, Sm, Tb, Tm, Yb, Y; M: Ni and Pd) and REZ,
xM2A1,,Tt2(Al,.yTty)(Al1.,Tt,)2(RE= Sm, Er, Dy; M: Ni, Co; Tt= Si, Ge). These structures
are very stable appearing in a wide variety of systems when the concentration of
transition metal is greater than that of the rare earth ions in the reaction. These results
highlight the ability of A1 fluxes to facilitate the synthesis of both transition metal rich
phases along with the previously seen rare earth rich phases.

The structures of the two phases are closely related displaying similar building
units including a disordered RE/Al layer, similar to layers seen in several other Al and Ga
intermetallics. This layer shows ordering on scales larger than the currently refined cell
dimensions in the ac plane though attempts to refine this larger cell so far have failed due
to the weak nature of the supercell reflections. This ordering is apparent, however, when
long exposure zone photos of Tb1_,,NizAl,,ySiy are examined. Along with this ordering in
the ab plane, the photos show diffuse scattering along the c axis likely caused by stacking
faults in the crystal. Electrical conductivity and thermopower measurements show
metallic behavior for Y,_,,NiA15_ySiy similar to many newly synthesized phases from Al
flux. The magnetism of both systems, lack of spin glass behavior, does exhibit interesting
orientation effects and possible valence fluctuations in the Yb system. The Dyz,
xNi2A1,,Si2(All_ySiy)(Al,,,Si,)2 magnetism conforms to Curie-Weiss behavior at high
temperatures with the Dy atoms in a 3+ oxidation state and the Ni ions magnetically

silent.

207

References

 

‘ Chen, X.Z.; Sportouch, S.; Sieve, B.; Brazis, P.; Kannewurf, C.R.; Cowen, J. A.;
Patschke, R; Kanatzidis, M.G. Chem. Mater, 1998 10, 3202-3211.

2 Sieve, B.; Chen, X.Z.; Cowen, J. A.; Larson, P.; Mahanti, S. D.; Kanatzidis, M.G.,
Chem Mater, 1999 11, 2451-2455.

3 Sieve, B.; Sportouch, S.; Chen, X.Z.; Cowan, J .A.; Brazis, P.; Kannewurf, C.R.;
Papaefthymiou, V.; Kanatzidis, M.G. Chem. Mater., 2001 13,273-283

4 Sieve, B.; Trikalitis, P.N.; Kanatzidis, M.G. Z. Anorg. Allg. Chemie 2002 628, 1568-
1574

5 SMART. Data Collection Software for the SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

6 SAINT. Data Processing Software for SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

7 Sheldrick, G.M. University of Gottingen, Germany. Manuscript to be Published

8 Sheldrick, G.M. SHELXL. Structural Determination Programs, Version 5.0; Siemans
Analytical X-Ray Instruments Inc.: Madison, WI, 1995

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

'0 Marcy, H.O.; Marks, T.J.; Kannewurf, C.R. IEEE Trans Instrum. Meas. 1990 39, 756-
760.

” Zhuravleva, M. Chen, X.C. Wang, X.; Schultz, A.; Ireland, J .; Kannewurf, C.R.

Kanatzidis, M.G. Chem. Mater. in press

208

 

‘2 Gladishevskii, R.I.; Cenuzal, K.; Flack, H.D.; Parthe E. Acta C ryst. 1993 (B49), 468-
474

‘3 Thiede, V.M.; Fehrmann, B.; Jeitschko, W. Z. Anorg. Allg. Chem. 1999 (625) 1417-
1425

‘4 Lattumer, S.E.; Kanatzidis, M.G. Inorg. Chem. in press

‘5 To refine the addition of disorder on the Al(2) site the overall occupancy of the site had
to be fixed to 1/3 occupancy matching what was seen in the Ga/Ge analogs.

‘6 Gladyshevskii, E.I.; Kripyakevich, P.I.; Bodak, O.I.; Ukr. Phys. J. 1967 12, 447—452

'7 Vanna, C.M. Rev. Mod. Phys. 1976 (48), 219

‘8 Schiffer, P.; Ramirez, A. P., Comments Cond. Mat. Phys., 1996 18, 21-50, references

therein.

209

Chapter 9.
Synthesis and Characterization of the Quaternary Aluminum _
Silicides RE4Fe2,,,Al7,,,Si8 (RE = Ce, Pr, Nd, Sm) and

RE,Mn,,,Al,_,Si8 (RE = Ce, Pr, Nd, Gd)

1. Introduction.

Work to this point of the thesis has centered around the transition elements of Cu,
Ni, Co, and their elemental families. The results have lead to the discovery of a wide
variety of structural arrangements as well as several compounds exhibiting interesting
physical properties. One surprise from the work was to learn that Ni atoms appear to
achieve a diamagnetic electronic configuration. The electron density seems to be stable
due to the presence of more electropositive RE and Al atoms in the structure. Because the
neighboring transition metals Mn, Fe and Co have similar electronegativities to Ni, we
decided to comparatively study the corresponding RE/Mn/Al/Si, RE/Fe/Al/Si,
RE/Co/Al/Sil systems. We are particularly interested in the electron donor/acceptor
behavior of these transition metals, as well as the effects they have on both composition
and structure of the resulting quaternary intermetallics vis-a-vis the corresponding Ni
systems. We find significantly different chemistry than that exhibited by Ni leading to a
wide variety of new structures and properties.

Reports of quaternary Fe compounds in the literature are relatively scarce. These

studies have mostly concentrated on substitution studies of known temaries, forming

210

pseudo-temaries of known structure types as in szFe 1HCoxAl3 in the ThZZnl7 structure
type,2 or examinations Of quaternary Fe quasi-crystalline materials such those formed in
the alloys of Al-Cu-Co-Fe.3 Recent studies of quaternary Mn are also scarce
concentrating on psuedo-temary systems, i.e. FeanmAlmB,‘ and Y(Ni,,,Mn,)2B2C,5
and microstrucural work with alloys such as those found in Al/Mn/Cr/Si systems.‘5
Recently our group published the synthesis and characterization of a novel series
of quaternary rare earth iron aluminum Silicides.7 This chapter expands on that work and
reports results on both the RE/Fe/Al/Si and the RE/Mn/Al/Si systems, namely the
synthesis, structure, electronic band structure calculations, electrical properties, and
magnetic properties of RE,,(Mn,Fe)2,,,Al7,,,Si8 in addition to results of M'ossbauer .

spectroscopy conducted on (Ce,Pr)4Fe2,,Al7,,Sig.

2. Experimental Section.
Synthesis.
Reagents.

The following reagents were used as obtained: Ce and Sm, 99.9%, metal chips,
Research Chemicals, Phoenix, AZ; Pr, Nd, Gd 99.9%, -40 mesh, Cerac, Milwaukee, WI;
CeOz, PrzO3 and N d203, Sylvania, Towanday, PA; szO3, Rhone-Poulenc, Princeton NJ;
Fe, 99.99%, fine powder, Aldrich Chemical, Milwaukee, WI; Mn, 99.95%, fine powder,
Cerac, Milwaukee, WI; Si, 99.96%, -325 mesh, Cerac, Milwaukee, WI; Al, 20 mesh,

Fisher, Fair Lawn, NJ.

211

Synthetic Methods.

Method 1. RE,,Fe2,,,Al7,,Si8 (RE = Ce, Pr , Nd and Sm) and RE,,Mn2,,,Al,,,Si8 (RE
2 Ce, Pr, Nd, Gd) were prepared by mixing 0.0368 g (0.6 mmol) Fe metal or 0.0330 (0.6
mol) Mn metal, 1.2 mmol elemental rare earth, 0.2669 g (9.0 mmol) Al, and 0.0674 g
(2.4 mmol) of Si in a nitrogen atmosphere and placing the mixture in an alumina tube.
The alumina tube was then sealed in an evacuated (1.0x10‘4 Torr) 13 mm o.d. by 11 mm
i.d. fused silica tube and heated at 850° C for 4 days. The tube was then cooled to 500° C

in 3 days (-4.86° C/h), and finally to 50° C in 12 hours (-37.5° C/h).

Method 2. RE,,Fe2,,,Al7,,,Si8 (RE = Ce, Pr , Nd and Sm) were later synthesized
substituting rare earth oxides for the elemental rare earth as a starting material. Using the
oxide starting materials title compounds were prepared by mixing 0.0368 g (0.6 mmol)
Fe metal, 1.2 mmol rare earth in oxide form, 0.2669 g (9.0 mmol) Al, and 0.0674 g (2.4
mmol) of Si mixed in air and placed in an alumina tube. The alumina tube was then
sealed in an evacuated (1.0x10'4 Torr) 13 mm o.d. by 11 mm i.d. fused silica tube and
heated at 850° C for 4 days. The tube was cooled to 500° C in 3 days (-4.86° C/h), and
finally to 50° C in 12 hours (-37.5° C/h). The synthesis of the Mn analogs were not
attempted using oxide starting materials due to the quality of products received from the
elemental reactants.

After heating, the final products from both methods were isolated from the
solidified Al matrix with 5 M NaOH solution. After the A] matrix was removed the
product was washed and dried using acetone and ether. The products were then

submerged in a 50% HCl solution (by volume) for 30 minutes to remove any acid

212

sensitive byproducts. The final isolation yielded stable silver needles and black powder,
in 40% yield based on the transition metal used. Purity of the final product was confirmed
through comparison of the experimental X-ray diffraction powder patterns, taken of the

bulk product, to theoretical patterns calculated from the refined single crystal data.

Physical Measurements.
EDS analysis.

Quantitative microprobe analysis of the compound was performed with a JEOL
ISM-35C Scanning Electron Microscope (SEM) equipped with Noran Vantage Energy
Dispersive Spectroscopy (EDS) detector. Data were acquired by using an accelerating
voltage of 25 kV and 100 sec accumulation time. Crystals selected from the different
synthetic methods were each measured showing no significant differences in their final
elemental ratios. The EDS analysis however consistently yielded low values for Al and Si
as seen in the past with other compounds. To correct for the error a standard was
measured under the analysis conditions yielding an appropriate correction factor, which

was then used to adjust the measured ratios of Al and Si.

Single Crystal X-ray Crystallography.

Data were collected on the first two Fe analogs, Sm,,Fe2+,,Al7_,,Si8 and Ce4Fe2,,,Al,_
,Sis, using a Rigaku AFC6S four circle automated diffractometer. Room temperature data
were collected using MoKoc (A20.71073A) radiation and an empirical absorption
correction was made based on 11! scans. The structures were solved utilizing direct

methods and refined with the SHELXL package of programs.8 For these compounds, as

213

with all Al/Si compounds, the Al and Si positions could not be distinguished directly
from the collected X-ray data, due to Al and Si's similar scattering cross sections. The
identification of Al and Si atoms in the structure was made through analysis of bond”
distances involving the particular site as will be discussed later. The crystallographic and
refinement data for Sm4Fe2,,,Al7_,,Si8 and Ce4Fe2,,,Al7,,Si8 are listed in'Tables 9-1 to 9-5.
Selected bond distances for Sm,,Fe2+,Al7,,,Si8 are shown in Table 9-6.

Single—crystal X-ray analysis was completed on the other Fe homologues (Nd and
Pr) and the Mn phases, RE.,Mn2.,,,Al7,,,Si8 (RE = Ce, Pr, Nd, Gd), using a Siemens
SMART Platform CCD diffractometer with MoKoc (A=0.71073A) radiation. After data
collection, cell refinement and data processing were performed using the program
SAINT .9 After processing an empirical absorption correction was applied with the use of
the SADABS program.” The final structure was solved by direct methods and refined
with the SHELXL” package of programs using the Sm,,Fe2,,,Al7_,Si8 solution as a basis.
Thecrystallographic and refinement data for the structural analogs are listed in Tables 9-2

to 9-6.

Electronic Structure Calculations.

The crystal structure of RE,,Fe2,,,Al7.,,Si8 possesses a disordered site
between A1 and Fe atoms corresponding to the 2d Wyckoff site. Indeed, Al atoms occupy
over 84% of this site. Consequently, because a mixed site cannot be described in the
extended Hiickel tight biding theory, one simpler way to carry out an electronic structure
calculation in disordered compounds is to consider that the 2d Wyckoff site is fully

occupied by Al. In addition all the rare earth elements involved in these series of

214

Table 9—1. Crystal data and structure refinement for RE,,Fe2,,Al7,,Si8 (RE = Sm, Ce).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20’(I)]

R indices (all data)

Extinction coefficient
Largest diff. peak and hole

R1=2||Fo| - |Fc||/X|F°|, wR2=[}:(w|I~‘,2 - Fc2|)2/Z(wF02)2]”2

Sm,Pe,,,Al,_,,8i8 Ce,re,_,,Al,_,,s18
1131.34 1086.75
293(2) K 293(2) K
0.71073 A 0.71073 A
Cmmm (#65) Cmmm (#65)
a = 10.860(2) A a = 10.936(2) A
b = 16.206(3) A b = 16.404(3) A
c = 4.0911(8) A c = 4.1532(3) A
720.0(3) A3 745.1(3) A3
2, 5.218 Mg/m3 2.4.594 Mg/m3
19.22 m" 14.91 mm-1
1010.16 974.78
0.22 x 0.1 x 0.1 mm 0.06 x 0.03 x 0.03 mm
2.26 to 28.32° 2.24 to 28.25°
-10<=h<=14 -14<=h<=14
-20<=k<=21 -21<=k<=20
-5<=l<=5 -5<=1<=5
2261 / 543 3706 / 554
0.0408 0.0264
97.8 % 96.7 %
Full-matrix least-squares on F2
543/1/43 554/1/43
1.065 0.575
R1 = 0.0266 R1 = 0.0197
wR2 = 0.0628 wR2 = 0.0508
R1 = 0.0295 R1 = 0.0224
wR2 = 0.0639 wR2 = 0.0542
0.0014(2) 0.00182(17)

2.444 and -1.524 e.A‘3

215

1.642 and -1.1 19 e.A°3

Table 9-2. Crystal data and structure refinement for RE,,Fe2+,,Al7_,,Si8 (RE = Pr, Nd).

Formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Pr4Fe2‘04Alg96Si8
1090.07

293(2) K

0.71073 A
Cmmm (#65)

a = 10.893(2) A

b = 16.305(3) A

c = 4.1290(8) A
733.3(3) A3

2, 4.936 Mg/m3
16.030 mm"

983

0.07 x 0.04 x 0.04 mm
4.19 to 31.28 deg.
-15<=h<=15
-22<=k<=23

-5<=l<=5

Reflections collected/ unique4393 / 677

R(int)
Completeness to 0m,m >

Refinement method

0.0395
93.8 %

1~Id,,1=e,,,Al,,,Si8
1102.24

293(2) K

0.71073 A

Cmmm (#65)

a = 10.898(4) A

b = 1631(1) A

c = 4.131(1) A
734.5(6) A3

2, 4.984 Mg/m3
16.84 mm'1

990

0.1x 0.04 x 0.04 mm
2.25 to 28.22 deg.
-l4<=h<=14
-21<=k<=21
-5<=l<=5

3612 / 556

0.0324

99.1 %

Full-matrix least-squares on F2

Data / restraints / parameters 677/ 1 / 43

Goodness-of-fit on F2

Final R indices [I>20'(I)]

R indices (all data)

Largest diff. peak and hole

1.320

R1 = 0.0286
wR2 = 0.0748
R1 = 0.0287
wR2 = 0.0749

2.241 and —l.623 e.A‘3
R1=2||F,1 - |F,||/>:|l=,|, wR2=[2(w|1=,2 - 1=.,2|)2/>3(wF.2)2]"2

216

556/ 1 I 41

1.106

R1 = 0.0192

wR2 = 0.0420

R1 = 0.0250

wR2 = 0.0437

1.400 and —l.055 e.A‘3

Table 93 Crystal data and structure refinement for RE,,Mn2.,,,Al7,,,Si8 (RE = Ce, Pr).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected/ unique
R(int)
Completeness to 0max

Refinement method

Data / restraints / parameters

Goodness-of—fit on F2

Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=2IIF.I - IFCIIIZIFOI, w82=12(wIF.3 - F.2|)2/2(w1=,2)21m

Ce4Mn2016Alfi84Si3
1088.41

293(2) K
0.71073 A
Cmmm (#65)

a = 109430) A
b = 1652(1) A

Pr,,Mn2.08Al,5,92Si8
1089.34

293(2) K
0.71073 A
Cmmm (#65)

a = 10.940(2) A
b = 16.509(3) A

c = 4.143(3) A c = 4.1396(8) A
748.9(8) A3 747.7(2) A3
2, 4.826 Mg/m3 2, 4.838 Mg/m3
14.68 mrn'l 15.50 mm"1
973 979
0.12x0.04x0.04mm 0.18x0.04x0.04mm
2.23 to 28.30° 2.23 to 28.43°
-14<=h<=14 -l4<=h<=l4
-21<=k<=21 -21<=k<=21
-5<=l<=5 -5<=l<=5
3792 / 570 3624 / 554
0.0400 0.0503
98.4 % 94.2 %
Full-matrix least-squares on F2
570/1/43 554/1/43
1.096 1.099
R1 = 0.0217 R1 = 0.0266
wR2 = 0.0494 wR2 = 0.0663
R1 = 0.0260 R1 = 0.0299
wR2 = 0.0499 wR2 = 0.0678
0.0017704) 0.0019(2)

1.923 and -l.448 e.A‘3

217

1.507 and -l.387 e.A'3

Table 94. Crystal data and structure refinement for RE,,Mn2,,,Al7,xSi8 (RE = Nd, Gd).

Empirical formula
Formula weight
Temperature
Wavelength
Space group

Unit cell dimensions

Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected/ unique
R(int)
Completeness to 0mm

. Refinement method

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole
R1=Z||FO| - |Fc||/2|Fo|, wR2=[Z(w|F02 - F.2|)2/2(wF.,2)2]"Z

Nd4Mnm6AlmSig
1107.69

293(2) K
0.71073 A
Cmmm (#65)

a = 10.9211(3) A
b = 16.4552(1) A

Gc1,Mn,,,/s.1,,,si8
1162.81

293(2) K
0.71073 A
Cmmm (#65)

a = 10.942(4) A
b = 16.471(7) A

c=4.1040(1)A c=4.111(2)A
737.53(3) A3 740.9(5) A3
2, 4.988 Mg/m3 2, 5.212 Mg/m3
16.72 m" 20.61 mm'1
992.24 1026.88
0.08 x 0.03 x 0.02 mm , 0.1 x 0.03 x 0.03 mm
2.24 to 28.27 deg. 2.23 to 28.34 deg.
~14<=h<=14 -14<=h<=13
-21<=k<=21 -21<=k<=21
-5<=l<=5 -5<=1<=5
3754 / 557 3520 / 544
0.0456 0.0347
98.9 % 94.6 %
Full-matrix least-squares on F2
557/1/43 544/1/43
1.053 1.143
R1 = 0.0221 R1 = 0.0250
wR2 = 0.0497 wR2 = 0.0680
R1 = 0.0292 R1 = 0.0277
wR2 = 0.0510 wR2 = 0.0690
0.001 12(12) 0.00092(14)

1.513 and -l.l93 e.A‘3

218

1.853 and -l.205 e.A'3

Table 9—5. Atomic coordinates ( x10“) and occupancies for RE,.,M2,,,A17_,,Si8 (RE=Ce, Pr,

Nd, Sm, Gd; M: Fe, Mn).

 

 

Wyckoff x y z Occupancy
posntion
Sm 8q 2678(1) 1294(1) 5000 1
Ce 2687(1) 1293(1) 5000 1
Pr 2681(1) 1293(1) 5000 1
Nd 2683(1) 1295(1) 5000 1
Ce 2656(1) 1285(1) 5000 l
Pr 2653(1) 1285(1) 5000 1
Nd 2647(1) 1284(1) 5000 1
Gd 2646(1) 1283(1) 5000 1
Fe(1) 4i 0 1348(1) 0 1
0 1336(1) O 1
O 1339(1) O 1
0 1342(1) 0 1
Mn(1) 4i 0 1351(1) 0 1
0 1353(1) O 1
0 1358(1) O 1
0 1358(1) 0 1
Si(l) 8p 1419(2) 2452(1) 0 1
1410(1) 2444(1) 0 1
1413(2) 2446(1) 0 1
1416(1) 2446(1) 0 1
1413(2) 2452(1) 0 1
1415(2) 2449(1) 0 1
1415(2) 2458(1) 0 1
1416(2) 2458(1) 0 1

 

219

Table 9-5. (continued) atomic coordinates ( x10“) and occupancies for RE,,M2.,,,A17,,,Si8

(RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

 

 

Wyckoff x y 2 Occupancy
posmon
Si(2) 4j 0 730(1) 5000 l
0 731(1) 5000 l
0 729(1) 5000 1
0 729(1) 5000 1
0 732( 1) 5000 1
0 730(1) 5000 l
0 732(1) 5000 1
0 735(2) 5000 1
Si(3) 4g 3486(2) 0 0 1
3494(2) 0 0 1
3492(2) O 0 1
3487(2) O 0 1
3497(2) 0 0 1
3489(2) 0 0 1
3501(2) 0 0 1
3504(3) 0 O 1
Al(l) 41 5000 1235(2) 0 1
5000 1244(7) 0 1
5000 1240(2) 0 l
5000 1241(1) 0 1
5000 1238(1) 0 1
5000 1241(1) 0 l
5000 1229(2) 0 1
5000 1225(2) 0 1

 

220

Table 9-5. (continued) atomic coordinates ( x10“) and occupancies for RE,,M2,,,A17_,Si8

(RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

 

 

Wyckoff x y 2 Occupancy
mosition

Al(2) 4g 1218(3) 0 0 1
1213(2) O 0 1
1212(3) 0 0 1
121 1(2) 0 0 1
1200(2) 0 0 1
1202(2) 0 0 1
1207(2) 0 0 1
1205(3) 0 0 1

Al(3) 4j 0 2353(2) 5000 1
0 2338(1) 5000 1
0 2341(2) 5000 1
0 2344(1) 5000 1
0 2332(1) 5000 1
0 2333(2) 5000 1
0 2338(2) 5000 1
0 2343(2) 5000 1

M(1) 2d 5000 0 5000 0.16 Fe + 0.84 A1
5000 0 5000 0.03 Fe + 0.97 A1
5000 0 5000 0.04 Fe + 0.96 A1
5000 0 5000 0.006 Fe + 0.993 A1
5000 0 5000 0.158 Mn + 0.842 A1
5000 0 5000 0.08 Mn + 0.92 A1
5000 0 5000 0.26 Mn + 0.74 A1
5000 0 5000 0.37 Mn + 0.63 Al

 

221

Table 9-6. Anisotropic displacement parameters (A2 x 103) for RE,,M2,,,A17_,,S18 (RE=Ce,
Pr, Nd, Sm, Gd; M: Fe, Mn).

 

 

U11 U22 U33 U23 U13 U12
Sm 5(1) 4(1) 6(1) 0 0 0(1)
Ce 4(1) 5(1) 4(1) 0 0 0(1)
Pr 8(1) 4(1) 4(1) 0 0 0(1)
Nd 4(1) 3(1) 5(1) 0 0 0(1)
Ce 6(1) 4(1) 5(1) 0 0 -1(1)
Pr 7(1) 6(1) 6(1) 0 0 0(1)
Nd 8(1) 3(1) 7(1) 0 0 -1(1)
Gd 8(1) 7(1) 5(1) 0 0 -1(1)
Fe(l) 4(1) 2(1) 5(1) 0 0 0
4(1) 4(1) 4(1) 0 0 0
7(1) 3(1) 3(1) 0 0 0
5(1) 3(1) 4(1) 0 0 0
Mn(1) 6(1) 4(1) 4(1) 0 0 0
6(1) 4(1) 4(1) 0 0 0
6(1) 2(1) 7(1) 0 0 0
0(1) 1(1) 0(1) 0 0 0

 

The anisotropic displacement factor exponent takes the form: -2n2[h2a*2U11 + +

2hka*b*U12]

222

Table 9—6. (continued) anisotropic displacement parameters (A2 x 103) for

RE,,M2,,,A17,,,Si8 (RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

 

U11 U22 U33 U23 U13 U12

 

Si(l) 3(1) 5(1) 6(1) 0 0 0(1
3(1) 7(1) 5(1) 0 0 0(1)
6(1) 6(1) 6(1) 0 0 -1(1)
5(1) 5(1) 6(1) 0 0 0(1)
5(1) 6(1) 6(1) 0 0 0(1)
6(1) 8(1) 6(1) 0 0 0(1)
6(1) 3(1) 10(1) 0 0 0(1)
0(1) 5(1) 2(1) 0 0 -1(1)

Si(2) 5(1) 4(1) 4(1) 0 0 0
3(1) 6(1) 5(1) 0 0 0
8(1) 4(1) 4(1) 0 0 0
6(1) 3(1) 3(1) 0 0 0
8(1) 4(1) 4(1) 0 0 0
8(1) 6(1) 4(1) 0 0 0
9(1) 3(1) 7(1) 0 0 0
2(1) 2(1) 0(1) 0 0 0

 

 

The anisotropic displacement factor exponent takes the form: -2n2[h2a*2U11 + +

2hka*b*U12]

223

Table 9-6. (continued) anisotropic displacement parameters (A2 x 103) for

RE,M,,,AI,_,818 (RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

 

U11 U22 U33 U23 U13 U12

 

Si(3) 6(1) 4(1) 12(1) 0 0 0
3(1) 6(1) 9(1) 0 0 0
7(1) 4(1) 10(1) 0 0 0
4(1) 4(1) 8(1) 0 0 0
7(1) 5(1) 12(1) 0 0 0
9(1) 6(1) 9(1) 0 0 0
7(1) 2(1) 17(1) 0 0 0
3(1) 2(1) 13(2) 0 0 0

Al(l) 4(1) 5(2) 34(2) 0 0 0
3(1) 6(1) 18(1) 0 0 0
7(1) 5(1) 23(2) 0 0 0
4(1) 6(1) 17(1) 0 0 0
7(1) 7(1) 31(2) 0 0 0
7(1) 8(2) 25(2) 0 0 0
6(1) 5(1) 55(2) 0 0 0
1(2) 5(2) 61(3) 0 0 0

¥

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2U11 + +

2hka*b*U12]

224

 

Table 9-6. (continued) anisotropic displacement parameters (A2 x 103) for

RE,M,,,A1,_,81, (RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

 

U11 U22 U33 U23 U13 U12

 

Al(2) 2(1) 7(1) 9(1) 0 0 0
4(1) 7(1) 8(1) 0 0 0
5(1) 7(1) 11(1) 0 0 0
3(1) 6(1) 8(1) 0 0 0
6(1) 5(1) 7(1) 0 0 0
7(1) 8(1) 9(1) 0 0 0
4(1) 6(1) 9(1) 0 0 0
0(1) 5(1) 7(2) 0 0 0
Al(3) 6(1) 6(1) 6(1) 0 0 0
6(1) 7(1) 5(1) 0 0 0
7(1) 7(1) 5(1) 0 0 0
4(1) 5(1) 9(1) 0 0 0
6(1) 5(1) 7(1) 0 0 0
8(1) 8(1) 7(1) 0 0 0
6(1) 5(1) 7(2) 0 0 0
1(1) 5(1) 0(2) 0 0 0

‘

The anisotropic displacement factor exponent takes the form: ~2n2[h2a*2U11 + +

2hka*b*U12]

225

Table 9-6. (continued) anisotropic displacement parameters (A2 x 103) for

RE,,M2,,,A17_,,S18 (RE=Ce, Pr, Nd, Sm, Gd; M: Fe, Mn).

 

U11 U22 U33 U23 U12

 

M(1) 5(2) 6(2) 9(2) 0
9(2) 8( 1) 8(1) 0
9(2) 9(2) 9(2) 0
8(1) 7(1) 9(2) 0
9(2) 6(1) 8(2) 0
10(2) 9(2) 10(2) 0
7(2) 3(2) 9(2) 0
4(2) 6(2) 5(2) 0

oooooooo c
p—-
b.)

OOOOOOOO

 

The anisotropic displacement factor exponent takes the form: -2n2[h2a*2Ull + +

2hka*b*U12]

226

Table 97 Selected Bond Lengths (A) for Sm,Fe,,,A1,_,31,.

 

Bond Distance (A)
Sm-Si(l) 3.0484(10)
Sm-Si(2) 3.0541(8)
Sm-Si(3) 3.0640(7)
Sm—Si(1) 3.1017(10)
Sm-A1( 1) 3.2595(5)
Sm—M(l) 3.2963(5)
Sm~Al(2) 3.3276(10)
Sm-A1(3) 3.360
Sm-Al(3) 3.389
Fe-Si(2) 2.2751(9)
Fe-Si(1) 2.3739(14)
Fe-Al(2) 2.5690(13)
Fe-A1(3) 2.6151(14)
M(1)-Si(3) 2.6171(12)
M(1)-Al(1) 2.8546(14)
Si(1)-Si(l) 2.364(2)
Si(1)-Al(3) 2.5658(8)
Si(1)-Al(l) 2.641(2)
Si(2)-Si(2) 2.377(3)
Si(2)-Al(3) 2.643(2)
Si(2)-Al(2) 2.7122(13)
Si(3)-Al(2) 2.481(3)
Si(3)-Al(l) 2.5833(19)
Al(2)-Al(2) 2.669(4)
227

compounds possess an incomplete f shell and as describe elsewhere“'7“'”, the 4f orbitals
are closely localized within the rare earth electron core therefore the 4f electrons interact
much more strongly with each other (i.e.: electronic repulsion) than with the electrons of
the neighboring atoms. In contrast, the remaining electrons belonging to the rare earth
valence (5d, 6s, 6p) orbitals strongly mix with those of the surrounding atoms.
Consequently, the 4f orbitals will be not included in the electronic structure calculations.
One way to avoid the f orbitals in the calculations is to use Y atom to mimic the rare
earth elements because Y possesses no f shell and its orbital energies are similar to those
of the Ce”, Pr17 and Sml6 atoms. Therefore the electronic structure performed on the
theoretical compounds Y4Fe2Al7Si8, using the program package CAESAR 1.0”, will be
representative of the different compounds presented in this paper. The atomic orbital
parameters, i.e. energies and exponents, used in the Slater-type wave functions are chosen
from one the three data bases available in the CEASAR package and are reported in
Table 9-22”"5"°'”. The non-diagonal Hamiltonian matrix elements were computed with
the modified Wolfsberg- Helmohltz formula”. The number of electrons corresponding
to the Fermi level will be 162 for the hypothetical Y compound as well as for Ce analog

and 154 for the Nd, Pr and Sm homologues.

Charge Transport Analysis.
DC electrical conductivity and thermopower measurements were completed on
Selected single crystals. Conductivity measurements were performed with the usual four-
Probe technique.19 Thermopower measurements using a slow AC technique as described

elsewhere20 were conducted.

228

 

TABLE 98 Exponents and parameters used in the extended Hiickel calculations. The

contraction coefficients used in the double-C expansion are c1 and c2.

 

 

 

 

Atoms Orbitals Hii (eV) C] c, C2 c2 ’ Ref.
Y 5s -6.81 1.6 0.55 0.95 0.5686 20
5p -3.76
4d -6.10
Fe 48 -9.1 1.9 1.0 21
4p -5.32 1.9 1.0
3d -12.60 5.35 0.55 2.0 0.626
Si 38 -17.299 1.383 1.0 22
3p -9.20 1.383 1.0
A1 38 -12.30 1.167 1.0 23
3p -6.50 1.167 1.0
229

 

Magnetic Characterization.

Magnetic susceptibilities for RE,,Fe2.,,Al7,,,Si8 (RE=Ce, Pr, Sm, Nd) were
measured as a function of temperature using a MPMS Quantum Design SQUID
magnetometer. Temperature dependence measurements on polycrystalline samples were
conducted under increasing temperature within a 200 G field.

Magnetic susceptibilities for RE,,Mn2+,,Al7.,,Si8 (RE=Ce, Nd) were conducted using
a MPMS Quantum Design SQUID magnetometer. Temperature dependence
measurements on polycrystalline samples were conducted under increasing temperature

within a 1000 G field while Field dependent measurements where conducted at 3K.
Diamagnetic corrections where made to all data for core diamagnetism while a
second correction was made to the Mn analogs measured to correct for contributions

from temperature independent paramgentism seen in the measurements.

Mdssbauer Spectroscopy.

Mossbauer measurements were performed in transmission geometry between 4.2
and 300K using a 57Co(Rh) source and conventional methods for RE,,Fe2,,,Al7,,,Si8 (RE =
Ce and Pr). Isomer shifts are reported relative to iron metal at room temperature (300K).
The least—squares fit gave a slightly asymmetric doublet with the intensity of the low
Speed line lower than that of the high speed line. The asymmetry is more pronounced in
the case of the Pr compound, and it is constant versus temperature. Taking these results

into account, and the fact that the line widths of the Lorentzian lines are the expected

230

 

ones for our experimental set up and equal for the two lines of the quadruple doublet, we

can conclude that the observed asymmetry is, most likely, due to texture effects.

3. Results and Discussion.

Synthesis.

Molten Al has proven to be an excellent solvent in which to explore the reactions
of various metals with itself and Si. Due to the eutectic nature of the Si/Al system7 large
amount of unreacted Si can be dissolved and remain highly reactive in molten Al. This
dissolved Si then creates an excellent opportunity for reaction with other elements within
the flux, even if these elements are present in only a small percentage. The combinations
of rare earth elements, transition metals, and silicon in molten Al conveniently then gives
ternary and quaternary silicide compounds often exhibiting interesting new structural
arrangements. Under such reaction conditions, with excess molten Al, large crystals of
quaternary compounds can be grown. An example of this is the synthesis of large crystals
of RE4Fe2+xAl7-xSi3 (RE=Ce, Pr, Nd, Sm) which can be grown in a needle-like
morphology, see Figure 9-1, in molten A] metal.

The Al flux also permits the use of oxide starting materials, decreasing costs and
allowing sample preparation in an oxygen atmosphere. When an oxide precursor is used,
the Al acts not only as a flux but also as a reducing agent. The molten Al reduces the
metal oxide species converting the starting oxide materials into dispersed fine particles of
reactive metal species within the molten flux. These fine particles are then much more

I‘eactive then the bulk starting materials. In addition, the reduction of the metal oxide and

the subsequent formation of A1203 produces a large driving force within the reaction

231

media for more difficult reactions to use. This force has been seen to allow reactions to
occur at lower temperatures than would occur with bulk starting materials as described in
the synthesis of szNi(NixSi1-,)Al4Si6. In the synthesis of the title compounds these
avenues were not explored, however, since their synthesis was achieved at low
temperature in both methods. This concept could be important however in planning the
synthesis of more difficult species where the driving force for the reaction is small or
where the range of formation temperatures is much smaller then the title compounds
appears to be.

Through the current research many new compounds of the class
RE,,TMy(A1/Ga),,(Si/Ge)m can be formed at temperatures below 1000° C while in molten
metal Al or Ga. In A1 fluxes, as with the title compound, appears to acts as a reactive
solvent causing Al incorporation in the final structure producing aluminides. The other
three components are variable and can be present or absent depending on'the reaction
conditions, forming everything from binaries to quaternary compounds.21 There is no
reason, however, that quinary or higher compounds cannot be made with the same basic
ideas as the compounds above simply by adding more elements to the starting reaction
mixtures. This behavior differs from what has been seen in Ga flux were the Ga may or

may not act as a reactive fiux depending on the specific reaction.22

Structural Description.
The compounds RE,,Fe2,,,,Al7.,Si8 (RE=Ce, Pr, Nd, Sm) and RE,,Mn2,,,,Al7_,,Si8 (RE
= Ce, Pr, Nd, Gd) adopt a new structure type in the space group Cmmm (No. 65). They

exhibit a complex three-dimensional M-Al-Si framework with small channels running

232

parallel the c axis which contain the rare earth ions. The overall structure is shown in
Figure 9-2. The rare-earth atoms within the channels are in a 14 coordinate environment
shown in Figure 9-3i. For simplicity only distances in the Sm,,Fe2+,,Al7,,,Si8 analog will
actually be reported.

Before the structure can be discussed in any detail the positions of Al and Si have
to be assigned. As mentioned earlier the atomic assignments of Si and Al positions in
these compounds were made based on bond distances. This approach was shown to be
reliable in the Al and Si assignments for similar intermetallic compounds.7M Based on
their relative atomic sizes bonds involving Si are shorter than the corresponding Al
bonds. In the Sm,,Fe2,,,,A17,,,Si8 analog the Sm-Si distances range from 3.048(1) to 3.102(1)
A where as the Sm-Al distances range from 3.259(1) to 3.328(1) A. These assignments
also agree with the Fe distances of 2.275(1) to 2.374(2) A for the Fe-Si bonds and
2.569(2) to 2.615(2) A for the Fe-Al distances. These Fe-Si distances are short
considering sum of the Van der Waals radii (1.26 for Fe and 1.32 for Si) likely signifying
significant bonding interactions between the Fe and neighboring Si atoms. Similar bond
distances are seen in other iron silicides such as ZrFe,,Si2 and szFe4Si9.23 The Al and Si
assignments can then be made with reasonable certainty in RE,,Fe2,,,Al-,,,,Si8 (RE=Ce, Pr,
Nd, Sm) and RE,,Mn2,,,Al7,,,Si8 (RE = Ce, Pr, Nd, Gd).

In order to describe this especially complex structure it is useful to break it into
conceptual layers. One can then illustrate how the layers are assembled and what other
elements they bind as the layers stack, forming the overall three dimensional structure.

Although the structure is by no means lamellar, a basic repeating layer along the ab-plane

233

 

 

Figure 9-1. SEM image of crystals of Sm,,Fe2+,,Al7_,,Si8 exhibiting typical needle-like

morphology.

234

is shown in Figure 9-4. This perfectly flat slab, made of Al, Si, and M atoms, repeats
along the c-axis at the unit cell repeat length of ~4.1-4.2 A. There are both single Al and
Si atoms as well as Al-Al and Si-Si dimers. The latter Al(2)-Al(2), at 2.669(4) A, and
Si(1)-Si(l), at 2.364(2) A, are parallel and bind to the individual atoms, M, Al(l), and
Si(3), to complete the layer. As the layers stack along the c-axis, they are joined with
extra atoms found in the interlayer space, namely Al(3), Si(2) and the mostly A]
containing disorder position M(1). The Si(2) atoms are found as dimers at a distance of
2.377 A. As the extra interlayer atoms bond, the M atoms in the original layer (see Figure
9-4) expand their original coordination sphere from four to eight, bonding to both Al and
Si atoms. This eight-coordinate M arrangement can be viewed as a result of two
interpenetrating distorted tetrahedra of Si and Al atoms, see Figure 9-3h. The local
coordination geometry of the mixed occupancy site M(1) is best described as a distorted
square prismatic since the M(1) atom is sandwiched between two identical Si2A12 rhombi,
see Figure 9-3g. The M(1)-Al(l) and M(1)-Si(3) bond distances are 2.855(1) A and
2.617(1) A, indicative of Al-Al and Al-Si bonding.

The three pure Al sites of the structure exhibit several different bonding
environments. Al(l) exhibits a slightly distorted square planar arrangement bonding to
four separate Si atoms, while Al(2) exhibits a much greater distorted square planar
arrangement bonding to itself, two Fe atoms, and a Si(3) atom. The last Al atom, Al(3),
exhibits a higher coordination number at 7. The arrangement is best described as a
distorted capped trigonal prism arrangement, seen in Figure 9-3c. The Si atoms exhibit

both four- and five-coordinate environments. Si(l) and Si(3) exhibit distorted tetragonal

235

pyramidal arrangements shown in Figures 9-3d and f. Si(1) bonding involves Fe, Al(l),
two Al(3) atoms and itself to form a dimer at 2.36 A, while Si(3) bonds to one Al(2), two
Al(l) atoms, and two bonds to the mixed Al/Fe site M(1). The last atom, Si(2), exhibits a
highly distorted square planar arrangement bonding to two Fe atoms, Al(3) and also to

itself forming the Si(2)-Si(2) dimer.

Electronic Structure.

To better understand the electrical and magnetic properties of RE,,Fe2+,,Al7,,,Si8
and RE,,Mn2,,,Al7,,Si8 electronic band structure calculations were performed on the ideal
theoretical compound Y4Fe2Al7Si8. These calculations were able to predict accurately the
physical properties of the disordered materials as well as to obtain additional insight into
the charge distribution. The electronic structure is characteristic of a metal for both
valence electron counts 162 (Y and Ce compounds) and 154 (Nd, Pr and Sm
compounds), see figure 9-5. This predicted metallic behavior was confirmed by
resistivity measurements on the actual Fe compounds. The calculation also indicates a
certain degree of charge transfer from the Si and/or Al atoms to the Y atoms and to the Fe
atoms in order to partially fill its 4p orbitals. Above the Fermi level, however, the bands
are predominantly Y d character but slightly hybridized with the 3p orbitals of Si and Al

atoms, consistent with an oxidized Y atom.

236

. -. ‘ \\\‘?M§A\~ --
iii/MNK I 0A An- “‘1

. . (”\VW’ ., .‘V. 1.0)?“
- \ '/1 1s\\\ I/'I1.- .z‘ /“

,. In . \. '}\\\~\;f

 

.REOSiOFe @AI 1"

V

Figure 9-2. Structure of RE‘Fe2,xAl7_,Si3 (RE: Ce, Nd, Pr, Sm) and RE4MnmAlMSi8 (RE:

Ce, Pr, Nd, Gd) viewed down the Z axis.

237

  
   
 

 

Si(1) Si(1)

0 99> 9)

Fe Al(3)
Al(3)

 

    

Al(2)

Si(3)

Al(1) M(”1111(1) “(1)

Al(1)

b) Si(3) Si(3)
Al(1)
M(1)
Si(3) Si(3)

Al(1)

 

M(1)

O
O
O
0 Fe
ORE

Figure 9-3. Coordination environments in RE,,Fe2,,,Al7,,,Si8 and RE,,Mn2,,,Al,,,Si8 of the

individual atoms out to 3 A, except for the RE atom which is shown to 3.5 A. Only the Fe

environments are shown for simplicity.

238

$3
0

' 6' o.“
0.1.91.
. y . y..

Figure 9-4. Structure of the repeating layer along the c axis. Only the layer of the Fe

analog's is shown for simplicity.

239

 

I I I I I F T I T

— Total DOS
- - - Fe 3d
‘ ....... Si 3p

0)
N

01
O
1

Density of States
N OJ
0') CD
7 “I

 

 

 

 

 

 

 

14 ..
L- -
2 — . ‘~-. V“: .. _
--'1'"'*'r"";'-‘““"i:“"’ 1 ‘1' " ' " E ‘ ' ’1‘1“"""'"'1' 2.; "1“ " 3"}
-24.0 -20.0 -16.0 -12.0 -8.0 -4.0 0.0
Energy (eV)

Figure 9-5 Total DOS for Y,,Fe2Al,Si8 and contributions to the DOS of Fe 3d orbitals and

Si 3p orbitals. The Fermi level is drawn by a dashed line for 162 valence electrons. (taken

from reference 1)

240

 

Interestingly attempts to assign formal charges, based on the Mulliken
populations for both valence electron counts, describe Fe as the most reduce element with
a formal charge of —2 (electron densities 11.14 and 11.36 for 154 and 162 valence
electrons, respectively). In a rough picture, the Si atoms are neutral but a closer look
shows, indeed, different formal charges for the Si(1), Si(2) and Si(3) atoms which are 0, 0
and -0.5, respectively. The neutral silicon atoms are consistent with the presence of
homo-atomic Si(1)-Si(1) and Si(2)-Si(2) bonds. In the crystal structure, Si(3) is
completely surrounded by Al atoms (and Fe atoms on the disordered site) and the
difference in electronegativity between these atoms could be responsible for weakly
reducing Si(3) (Pauling’s electronegativity: 1.9 for Si and 1.6 for Al). As expected, the
two most electropositive elements, Y and A1, are oxidized and their formal charges are +2
and +1, respectively, with no difference between the four types of Al. While we do not
place much emphasis on the calculated numbers, it seems more probable that Y holds a
+3 charge, which will be in agreement with the high temperature magnetic data. The
corresponding electron densities are reported in Table 9-9 for both valence electron
counts of 154 and 162. Most likely, the real charge on the atoms will be different from
the assigned formal charge. As mentioned above, the assumption, however, that the iron
fills its 3d shell resulting in a formal charge of —2 is quite reasonable and qualitatively the

electron density distribution seems plausible as seen elsewhere.24 In addition, the fact that

241

 

 

Table 9-9. Calculated electron densities for the Y4Fe2Al7Si8 compound as a function of

the electron count.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Valence Electron Density

Electrons Y(1) Fe(1) Si(1) Si(2) Si(3) Al(l) Al(2) Al(3) Al(22)

154 1.53 11.14 4.18 4.19 4.53 2.16 1.94 2.02 2.17

162 1.78 11.36 4.46 4.45 4.74 2.26 1.99 2.07 2.26
242

these compounds are diamagnetic upholds the idea of a dlo system which is also

supported by the Mtissbauer spectroscopy (see below).

Charge Transport Properties.

The compounds exhibit metallic conductivity with resistivity values ranging from
80 uQ-cm for the Pr compound to 150 119-cm for the Sm analog at room temperature, see
Figure 9-6. Both the low resistivity values and the temperature dependence are in
agreement with metallic behavior.

Thermopower values, on the order of a few uV/K are also consistent with metallic
behavior, see Figure 9-7. Sm,,Fe2+,,Al7_,Si8 exhibits positive thermopower values, i.e. p-
type behavior, with a general increase in value with decreasing temperature. The Pr
analog exhibits similar behavior with a gradual increase in values as temperature
decreases. The Pr analog differs, however, in that the measured values starts negative and
passes through zero, switching from n-type to p-type behavior, at about 50 K, as seen in
Figure 9-7a. The Ce compound deviates markedly from the other two analogs by showing
much larger thermopower values which increases with decreasing temperature. The
values reach a maximum, at about 100 K, of +34 v/k and drops to lower values as the
temperatures tends to 0 K. The presence of such maxima in the thermopower of metallic
systems has been attributed in the past to Kondo effects.25 Crystals of the Mn analog

have all been too small for conductivity and thermopower measurements.

243

 

100000

 

ConductIVIty (S/cm)
. M;
/p
0
CD

 

 

 

 

10000 f‘x \
; Sm
r- l 1 In 1 L 1; L 1 #1 m 1 1 mm L
0 50 100 150 200 250 300 350

Temperature (K)

Figure 9-6. Resistivity measurements of Ce4Fe2+xAl7.,Sig, Pr,,Fe2,,,Al7,,,Si8 and Sm4Fe2+xAl7,

x818

244

93

 

 

 

 

 

 

 

 

12,0

§ 10%

3 821°

9 6:2:

o 4' .036; - __

8 2. “w

‘5' 0:- ‘x..._

a) _

.c -2P Pr

I‘— 4: 1 1 1 1‘1\
0 50 100 150 200 250 300

Temperature (K)

b)

S

>

3-

L.

(D

3

O

O.

O

E

Q)

.C

l— :
0‘11111111111 .1

 

 

 

0 50 1001150200 25015300

Temperature (K)

Figure 9-7. Thermopower Measurements of (a) RE,,Fe2,,,Al7,,,Si8 (RE 2 Pr and Sm) and (b)

Ce,,Fe2,,,Al7,,Si8 showing small absolute values indicative of metallic compounds.

245

Magnetic Properties.

The magnetic properties of the series RE,,Fe2,,,Al7,,Si8 (RE: Ce, Pr, Nd, Sm) were
examined as a function of temperature. The low temperature magnetic behavior is
dominated by a minor ferromagnetic impurity in the Fe systems which has not been
identified. Therefore no comment will be made about the low temperature magnetic
behavior of these compounds.

The high temperature (T>60 K) magnetic data for all the Fe analogs except the
Sm analog conforms to Curie-Weiss law behavior.26 Graphs of these relationships,
standardized to one rare earth atom per formula unit, are shown in Figures 9-8 and 9-9.
From the Curie-Weiss behavior a 1.18“ for each rare earth atom can be calculated,
assuming a diamagnetic state of the other atoms in the structure. For Pr,,Fe2,,,Al7_,,Si8 an
experimental value of 3.63 B.M. was obtained which is very near the accepted
experimental value of 3.40 B.M for a Pr3+ ion.27 The Ce and Nd analogs exhibit similar
behavior with 11,,“ values of 2.52 and 3.31 B.M. which are very close to the accepted
values of 2.28 and 3.50 B.M. for these RE3+ ions. The measured magnetic moments
clearly indicate that the rare earth atoms are in a RE 3* state while the Fe atoms do not
contribute to the overall magnetism. This absence of magnetic contribution from the Fe
atoms is caused by extensive hybridization of the Fe (1 orbitals with neighboring Si atoms
and also possibly a diamagnetic behavior for the Fe atoms due to the nearly complete
filling of the Fe (1 orbital manifold.

Similar magnetic behavior has been observed in other ternary rare earth iron

compounds involving Al or Si. As for example in REFeZAlm and ErzFe3Si5.28 In these

246

compounds only the rare earth atoms contribute to the observed magnetic moments. In
compounds containing only diamagnetic rare earth elements, as in YFezAlw or La FezAllo,
Pauli paramagnetism is observed, due to their metallic nature. This further supports the
lack of substantial magnetism from the Fe atoms. Sm,,Fe2,,,Al7,,Si8 does not exhibit Curie-
Weiss behavior at any temperatures, see Figure 9-9. This deviation is common in Sm“
magnetism which can be exceedingly complex due to the thermal population of closely
lying excited magnetic states.40

The magnetic data for Ce,,Mn2,,,Al7_,,Si8 is shown in Figure 9-10. At low
temperatures Ce,,Mn3,,,Al7,,Si8 exhibits a sharp ferromagnetic transition at about 14 K
while at temperatures slightly higher a gradual decrease in the magnetization is seen
which does not conform to Curie-Weiss behavior. This is very similar to the behavior of
the Ce,Fe2,,Al7,,Si8 analog, discussed above, and may be due to a valence fluctuation for
the Ce analog in this temperature range. At temperatures over 100 K the compound
exhibits Curie-Weiss behavior with a linear response in the inverse susceptibility. From
this data a pm of ~2.42 B.M. can be extracted which is very close to the expected value
for the Ce atoms in a 3+ oxidation state (2.52 B.M.) again mimicking the Fe analog. Also
in supporting that idea that the Fe and Mn analogs are analogous is the fact that they
exhibit roughly the same emu per mole values at high temperatures for the two Ce
analogs. The field dependence measurements for Ce,,Mn2,,,Al7,,Si8 shows very general
increase with increasing fields accounting for only 0.2 B.M. at the maximum measured
field of 55000 G.

Nd,,Mn2,,,Al7,,,Si8 showed an antiferromagentic transition at about 3K, Figure 9-11,

with Curie-Weiss behavior at temperatures greater than 3 K. A it,“ can be found from the

247

high temperature data of ~3.06 B.M. which as in the Fe analogs is close to the value
expected for the Nd in the 3+ oxidation state of 3.50 B.M. When Field dependence was
examined a sharp increase in response is seen up to about 10000 G then a slower
response as a function of field accounting for only 1.3 B.M. at the maximum field of
55000 G.

These results are very exciting causing one to imagine that the Mn atoms in the
structure are actually in a reduced state with a filled d shell as was seen in the Fe atoms.
And well this may be the case here a more definite means of assigning the charge of the
Mn atoms is needed before a definitive answer can be made. Typically Mn atoms in
systems such as these are reported to exhibit magnetic moments however work with Al
quasi-crystals and their approximants where Al atoms outnumber the Mn atoms have
shown that the Mn atoms do not exhibit a magnetic moment.29 The major difference
between the Mn analog and the Fe compounds is that elemental Fe is more
electronegative than Al (1.83 Fe and 1.61 Al) whereas the Mn is actually less
electronegative than Al (1.55 Mn). For these reason it is not easy to imagine how the Mn
atoms can fill their (1 shells even though the magnetic data seems to imply that such is
true. One needs to realize that the data can also be explained assuming that the d shell of
the Mn compounds are not actually filled but instead exhibiting some other magnetically

silent state.

Mossbauer Spectroscopy.

In order to further probe the electronic state of iron in RE4Fe2,xAl,,,Sig, Mdssbauer

spectroscopic measurements of the Ce and Pr analogs were carried out in the temperature

248

600

 

[rffr
.

500

400

300

200

1/x m
(mole RE ion/ emu)

100

 

 

'rlfijTerrT—[fiYTrIIITI
b—

1 l 4;; A l l L J 1 1 l A L L j l L l 1 l

50 100 150 200 250 300
Temperature (K)

O
1.

 

O

 

b) 200

_A
U'l
O
I
.

O
O

‘T—I Y
O

50— '

(mole RE ion/ emu)

 

 

obmmrr11111111441411111111.111
0 50 100 150 200 250 300

Temperature (K)

Figure 9-8. Inverse magnetic susceptibility data for (a) Ce,,Fe2,,Al7_,,Si8 and (b)

Pr,,Fe2,,,,Al7,,,Si8 standardized to one RE per formula unit.

249

£1

 

200 ~ 0

150 r

100

1/x m
(mole RE ion/ emu)

(I!
O

 

 

50 100 150 200 250 300

 

60 F . '
t
58 r .

56

1/x m
(mole RE ion/ emu)

1
54F.
i
F

52 a
50 I .1 l 1144 L L 1 LL Lil L .1._l_ L11 1 A I l l l l I l L 1 l l
0

50 100 150 200 250 300 350

 

 

Temperature (K)

Figure 9-9. Inverse magnetic susceptibility data for (a) Nd,,Fe2,,,Al7_,Si8 and (b)

Sm,,Fe2,,,,Al,,,,Si8 standardized to one RE per formula unit.

250

X
m
(emu / mole RE ion)

m

llx
(mole RE ion / emu)

Magnetization
mole RE ion)

(BM.

007

006

0.05 :

0.04 :

003

002

001

700

500

300

200

100

013

(12

0.1 :

 

Yr'TVY‘Y

0".
o . .
. o

YTrIY'I

1_ L

 

 

A l 1 L

O

1 + L 1 1
40 60
Temperature (K)

 

 

 

 

144 LAJ A 1“ IJLILJ 1 A l A A L A

100 150 200 250 300
Temperature (K)

 

 

 

l A A

-410‘

l 4

410‘

1 l m 1 1 1 4 1
-2 10‘ 0
Field (G)

 

1 1 1 1
210‘ 6104

Figure 9-10. Magnetic behavior of Ce,,Mn2,,,Al7,,Si8 (top) susceptibility as a function of

temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field standardized to one RE per formula unit.

251

m

X
(emu/mole RE ion)

1/xm
(mole RE ion / emu)

Maynetization
mole RE ion)

(BM.

Figure 9-11. Magnetic behavior of Nd,,Mn2,,,Al7,,,Si8 (top) susceptibility as a function of
temperature (middle) inverse susceptibility as a function of temperature and (bottom)

magnetization as a function of field standardized to one RE per formula unit.

0.6

0.5

0.4

0.3

0.2 ’

0.1

300

.1 -~ to N
0 or 0 or
o o o o

01
O

.1"

-1.5 '
-6104

 

YYITYIYIYYTIITTWT

TWI

 

 

 

10
Temperature (K)

 

 

A 1 A J A A l A I l A

100

 

1 AL 1114 A A l A A A A

150 200 250 300 350
Temperature (K)

 

I T

TIWIYII'

Vrr“

p—

y.

 

F 00.0000".

1 A l

l J A

I

I

x’

 

I 1 l 1 A n L l

 

-4 10‘

1 1 11111 11
-210‘ 0 210‘ 410‘ 610‘

Field (G)

region between 4.2 and 300 K. Essentially a classical quadruple doublet was observed at
all temperatures with no significant changes in the line widths of the quadruple doublets.
A typical spectrum taken at 20 K is shown in Figure 9-12. The spectra were analyzed by
least squares fitting the quadruple doublet with Lorentzial lines. The best fitted values of
the Mossbauer parameters are given in Table 9-24. The spectra show no evidence for
magnetic splitting due to magnetic ordering of the Fe atoms. This is consistent with the
results of the aforementioned susceptibility measurements, which indicate that only the
RE-ions carry magnetic moments.

TheMossbauer spectra suggest that there is no Fe 3d band magnetism, nor Fe
Curie-Weiss paramagnetism. The onset of 3d band magnetism as a result of Fe-Fe
exchange interactions could hardly be expected in these compounds since Fe-Fe distances
are relatively large (4.080 A) and the number of Fe nearest neighbors is only 2. Further
evidence for the non-magnetic character of Fe is provided by the positive values of the
isomer shift of 0.21 mm/sec at room temperature. This positive increase of the isomer
shift relative to the value of metallic Fe relates to a decrease of the density of the s
electrons at the Fe nucleus. This can be due either to a decrease in the number or to an
enhanced screening of the 4s electrons in these compounds. In other words the Fe
atomsseem to be more reduced than those in Fe metal which is consistent with the
expectation of electron flow from the electropositive RE and Al atoms to the more
electronegative Fe atoms. Similar changes of the isomer shifts were found in other Fe/Si
compounds and alloys in which Fe has essentially zero moment. These cases include
REFeZSis, REzFe3Si5, RE: rare earth and B-FeSis, MFeSi (M: Ti, Mn, Nb) and

amorphous alloys as for instance, Fe3OSi70.

253

The crystallographic data presented above lend support to strong hybridization via
Fe-Si interactions. The Fe-Si distances are rather short 2.30-2.6 A as compared to 2.40A
which represents the sum of Goldschmidt radii of Fe and Si. Fe-Si electron hybridization
is therefore likely destabilizing the Fe atom orbitals and thus giving rise to two identical

sub-bands, with strong d-character one with spin-up an one with spin down.

4 Conclusions.

Using liquid Al as a solvent the quaternary compounds containing RE/Fe/Al/Si
and RE/Mn/Al/Si were discovered as part of a new isostructural series with the general
formula of RE4Fe2,,Al7,,Si3 (RE: Ce, Nd, Pr, Sm) and RE,,Mn2,,,Al7,,,Si8 (RE: Ce, Pr, Nd,
. Gd). The compounds posses a novel structural arrangement built of flat layers stacking
along the c axis. They also are strongly metallic properties with conductivity values of
about 10,000 S/cm at room temperature. Ce4Fe2,,Al7,,Sig, unlike the other analogs, shows
a complex charge transport behavior that may be due to Kondo effects. A unique site in
the structure exhibits mixed Fe/Al occupancy and may be key to the stability of the
structure type. The most interesting property of these materials, however, is the highly
reduced Fe atoms, likely caused by electron transfer from the more electropositive atoms
in the structure. The Mn analog likely shows similar electronic properties though direct
measurements of the Mn atoms themselves has not been completed. The discovery of
novel complex compositions and structure types in relatively refractory aluminum
silicides once again underscores the value of liquid Al in exploratory synthesis of this

little known class of materials.

254

Table 9-10. Mossbauer Parameters, collected at 20 K, of RE,,Fe2,,,Al7_,,Si8
(RE: Ce, Pr).

 

 

 

 

5(mm/s) AEq(mm/sec) F/2(mm/sec) IL/IR
Ce 0.31 0.42 0.14 0.94
Pr 0.315 0.45 0.13 0.90

 

 

 

 

 

1L. = Intensity of Left Line
I R Intensity of Right Line

 

 

     
   
  
  

13.9.28.

.3

J.

 

8.35

A l

 

RELATIVE TRANSMISSION (%)
§

 

 

 

fT V—r I I V l V I V I V r V ‘

-4 -2 0 2
VELOCITY (nun/s)

Figure 9-12. Typical Mossbauer spectra of Pr,,Fe2+,,Al7_,,Si8 and Ce,,Fe2+,,Al7,,,Si8 taken at 20

K.

255

 

References

 

‘ Sieve, B.; Kanatzidis, M.G. work in progress

2 Wang Z.; Dunlap R.A.; Foldeaki M. J. Mar. Sci., 1994 29, 5333-5336.

3 Perez, R.; Juarezislas, J.A.; Martinez, L. Mat. Sci. Eng. A, 1994 182, 837—840.

4 Rico, M.M.; Sort, J .; Surinach, S.;Munoz, J.S.; Greneche, J.M.; Alcazar, G.A.P.; Baro,
M.D. Mater. Sci. Forum 2002 386-3, 497-502

5 da Rocha, F.S.; Fragul, G.L.F.; Brandao, D.E.; Granada, C.M.; da Silva, C.M.; Gomes,
A.A. Eur Phys J B 2002 25(3), 307-311

6 Srivastava, A.K.; tha, S.N.; Ranganathan, S. J. Mater. Sci. 2001 36, 3335-3341

7 Sieve, B.; Sportouch, S.; Chen, X.Z.; Cowan, J .A.; Brazis, P.; Kannewurf, C.R.;
Papaefthymiou, V.; Kanatzidis, M.G. Chem. Mater, 2001 13, 273-283

8 Sheldrick, G.M. 1995, SHELXL. Structural Determination Programs, Version 5.0.
Siemens Analytical X-ray Instruments Inc., Madison, WI.

9 SAINT, version 4, Siemens Analytical X-ray Instruments Inc., Madison WI.

1° Sheldrick, G.M. University of Gdttingen, Germany, to be published.

” Ortiz, I.V.; Hoffmann, R. Inorg. Chem. 1985 24, 2095-2104.

12 Y. C. Tian, T. Hughbanks Inorg. Chem. 1993 32, 400-405

‘3 “Crystal and Electronic Structure analysis Using CAESAR, PrimeColor Software, Inc.,
NC, USA1998”, J .Ren, W. Liang, M.-H. Whangbo

1“ Hii values are taken from CAESAR package data base

‘5 R.H. Summerville, R. Hoffmann, J. Am. Chem. Soc. 1976 98, 7240

'6 N. Trong Anh, M. Elian, R. Hoffmann, J. Am. Chem. Soc. 1978 100, 110-116

‘7 A.B. Anderson, R. Hoffmann, J. Chem. Phys. 1974 60, 4271

256

 

‘8 J. Ammeter, H.-B. Bi’Irgi, J. Tibeault, R. Hoffmann, J. Am. Chem. Soc. 1978 100, 3686-
3692

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

20 Marcy, H.O.; Marks, T.J.; Kannewurf, C.R. IEEE Trans Instrum. Meas. 1990 39, 756-
760. A

2' Sieve, B.; Kanatzidis, M.G. work in progress

22 Zhuravleva, M. Chen, X.C. Wang, X.; Schultz, A.; Ireland, J .; Kannewurf, C.R.
Kanatzidis, M.G. Chem. Mater. submitted

23 a) ZrFe4Si2: Yarmolyuk, Ya; P.;Lysenko, L.A.; Gladyshevskii, E.I. Dopo Akad Nauk
Ukr Rsr Ser A: F iz Mat Tekh Nauki 1975 37, 279 (b) szFe4Si9: Gladyshevskij, E.I.;
Bodak , O.I. Ukranskii F izicheskii Zhurnal (russian edition), 1978 23(1), 77-82.

2“ G. V. Vajenine, R. Hoffmann, J. Am. Chem. Soc. 1998 120, 4200-4208

25 (a) CePdJBX: Houshiar, M.; Adroja, D.T.; Rainford, B.D. Physica B 1996 268-270. (b)
Ce2T3X9 : Buschinger, B.; Trovarelli, O.; Weiden, M.; Geibel, C.; Steglich, F. J. Alloys
Camp. 1998 633-636.

26 At this time no diamagnetic correction has been applied to the data due to the impurity
presence.

27 Bourdreaux, E.A.; Mulay, L.N. in Theory and Applications of Molecular
Paramagnetism, John Wiley and Sons, New York 1976.

28 (a)REFe2A110: Tiede, Verena M.T.; Ebel, Thomas; Jeitschko, W.J. J. Mater. Chem,
1998 8(1), 125-130. (b) ErZFe3Si5: Moodenbaugh, A.R.; Cox, DE, Phys. Rev., 1984 B29,

109.

257

 

 

29 (a) Hippert, F; Simonet, V.; dc Laissardiere, G.T.; Audier, M.; Calvayrac, Y. J. Phys.
1999 11, 10419-10435 (b) Liu, F.; Khanna S.N.; Magaud, L.; Jena, P.; Decoulon, V.;

Reuse, F.; Jaswal, 8.8.; He, X.G.; Cyrotlackman, F. Phys. Rev. B 1993 48 1295-1298

258

Chapter 10.
Synthesis, Structure and Physical Characterization of

REFe4Algsi6 (RE=Tb, Er, Gd, Dy, Ho)

1. Introduction

Only a few studies of quaternary iron intermetallic systems exist in the literature,
mostly concentrating on psuedo-temary compounds such as szFe,,,_,,Co,,Al3l or the
quasi-crystalline phases that form in alloys of Al/Cu/Co/Fe.2 Recently we published a
true quaternary phase, RE,,Fez,,,Al7_,,Si8 (RE: Ce, Pr, Nd, Sm), which was synthesized
using molten Al flux methods.3 Continuing with the idea of synthesizing quaternary Fe
systems we decided to conduct reactions investigating the late rare earth metals. These
reactions yielded a new structural series of compounds REFe4A19Si6 (RE = Tb, Er, Gd,
Dy, Ho).

This phase exhibits a highly complex three dimensional Fe/Al/Si framework
containing the RE ions. Unlike in the RE,,Fe2,,,Al7_,,Si8 systems the structure of the
compounds does not contain 1D channels in which the RE ions reside instead the RE ions
are completely enclosed by the non-rare earth framework. The structure is related to the
NdRh,,Al,5_4 structure type that has been studied and published by Jeitschko though

significant differences are present in the ordering of the structure, see below.4 In this

259

chapter we present the synthesis, structure, magnetic properties, and Messbauer

spectroscopy of a second quaternary Fe system, REFe4A19Si6 (RE = Tb, Er, Gd, Dy, Ho).

2. Experimental Section

Synthesis.

Reagents.

Gd, Tb, Dy, Ho, Er, 99.9%, -40 mesh, Cerac, Milwaukee, WI; Fe, 99.99%, fine powder,

Aldrich Chemical; A1 99.5%, -20 mesh, Cerac Inc.; Si, 99.96%, -325 mesh, Cerac Inc..

Synthetic Method

Method 1: In a N2 filled glove box RE (RE=Tb, Er, Gd, Dy, Ho), Fe, A1 1.11 Si
were mixed in a 1:4:20:6 ratio and placed inside of an A1203 crucible. The crucible was
placed inside a fused silica tube and flame sealed under vacuum (<10“ Torr) then heated
to 850° C in 20 hours and maintained there for 4 days. The sample was cooled to 500° C
in 3 days and quickly cooled to 50° C in 12 hours. After heating the crucible Was
submerged in 5 M N aOH to remove the excess Al matrix. This procedure yielded two
types of products silvery cube-like crystals and black powder both of which were
identified as REFe4Al9Si6 by comparison of experimental powder X-ray diffraction
patterns to a calculated pattern from the solved crystal structure. For the synthesis of

TbFe,,AlgSi6 yields based on the Tb metal were on the order of 95%.

Method 2: In a N2 filled glove box Tb, Fe, Al, and Si powders were rrrixed in a

1:4:20:6 ratio. This mixture was cold pressed and loaded into an arc welder and melted

260

 

on a water cooled Cu plate, under a Zr gettered Ar atmosphere for approximately 30
seconds until a good melt was observed. The samples were then flipped and re-melted
several times to ensure homogeneity. The excess Al was then removed by submersion of
the arc melted pellet inSM N aOH for 24 hours. After the isolation an X-ray powder
diffraction pattern was taken and matched to the calculated pattern to verify phase purity.
Yields based on Tb for this reaction are ~ 70%. Attempts to form this phase without the

excess Al present thus far have failed.

Physical Measurements.
EDS Analysis.

Quantitative microprobe analysis of the compounds was performed with a JEOL
J SM-6400 Scanning Electron Microscope (SEM) equipped with Noran Vantage Energy
Disperse Spectroscopy (EDS) detector. Data were acquired using an accelerating voltage
of 25 kV and 100 sec accumulation time. Standards where recorded under the same
experimental conditions to yield correction factors to be applied to the data. After
calibration the title compounds gave an elemental ratio of about "REFelsAlgSi," for both
compounds, which agrees well with the values from the refined X-ray structure. Crystals
selected from the different synthetic methods were measured showing no significant

differences in elemental ratio.

Single Crystal X-ray Crystallography.

Single crystal X-ray diffraction data for REFe4A198i6 (RE=Tb, Er, Gd, Dy, Ho)

were collected at 298 K on a Siemens Platform CCD diffractometer on using Mo K01

261

 

(A=0.71069) radiation. The SMART software5 was used for the data acquisition and the
program SAINT6 was used for the data extraction and reduction. An empirical absorption
correction using SADABS was applied to the data and the structure was solved in the
SHELXL package of programs. The crystallographic and refinement data are listed in
Tables 10-1 through 10-3. The fractional atomic positions and displacement parameters
(U values) are listed in Tables 10-4 and 10-5. Selected bond distances for TbFe4A19Si6 are
listed in Table 10-6.

After the structure was solved several questions still remained outstanding about
the Al/Si assignments in the structure because of the fact that 7 out of the 10 total
positions were either a Si or Al atom. To resolve these questions a single crystal neutron
analysis was conducted at Argonne National Lab. This refinement confirmed the
assignment of all atoms except the atom in the 4c position. This position was originally
assigned as a Si atom however the neutron refinement confirmed it was in fact an Al

atom.

262

Table 10-1. Crystal data and structure refinement for REFe4A19Si6 (RE = Tb, Er).

Formula
Formula weight
Temperature (K)
Wavelength (A)
Space group

Unit cell dimensions (A)

Volume (A3)

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range (°)

Index ranges

Ref. collected / unique

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20'(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

R1=£||Fo| - IFCIIIZIFOI, wR2=[Z(w|F02 - Fc2|)2/Z(wF02)2]”2

TbFe4A19Si6
793.68
293(2)

0.71073

P42/nmc

a = 8.7184(11)
c = 15.171(3)
1153.1(3)

4, 4.572 Mg/m3
12.238 mm'l
1480.00
0.04x0.03x0.03 mm
2.69 to 28.30
-11<=h<=11
-11<=k<=11
-20<=l<=19
10686/812

812 / 0 I 60
1.273
R1=0.0202
wR2=0.0505
R1=0.0219
wR2=0.0510
0.0052(2)
1.362 and -0.685 e-/A3

263

ErFe,,Al,,Si6
802.02

293(2)

0.71073
P42/nmc

a = 8.6882(12)
c = 15.115(3)

1 141.0(3)

4, 4.669 Mg/rn3
13.53 mm'1
1492.00
0.30x0.23x0.17 mm
4.28 to 31.08
-12<=h<=12

-1 1<=k<=12
-21<=l<=21
12800/ 1001
996 / 0 / 60
2.798
R1=0.0362
wR2=0.0867
R1=0.0380
wR2=0.0869
0.0011(2)
0.951 and -1.098 e-/A3

Table 10-2. Crystal data and structure refinement for REFe,Al,,Si,S (RE = Gd, Dy).

Formula
Formula weight
Temperature (K)
Wavelength (A)
Space group

Unit cell dimensions (A)

Volume (A3)

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range (°)

Index ranges

Ref. collected / unique

R.

101

Data / restraints / parameters

Goodness-of—fit on F2

Final R indices [I>20'(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

GdFe,,Al.,Si6
791.81

293(2)

0.71073
P42/nmc

a = 8.7491(18)
c = 15.254(4)
1167.7(5)

4, 4.504 Mg/m3
11.710 mm'l
1476.00

0.06 x 0.12 x 0.12 mm
2.67 to 28.28
-11<=h<=1l
-11<=k<=11
-20<=l<=19
10708 / 808
0.0426

808 / 0 / 60
1.095

R1 = 0.0260
wR2 = 0.0674
R1 = 0.0272
wR2 = 0.0681
0.0044(3)
2.658 and -1934 e—/A3

R1=Z||F°| - IFCIIIXIFOI, wR2=[X(w|F°2 - Fc2|)2/Z(WF02)2]”2

264

DyFe4A198i6
802.02

293(2)

0.71073
P42/nmc

a = 8.705(3)

c = 15.194(7)
1151.5(7)

4, 4.598 Mg/nn3
12.604 mm’l
1484.00

0.03 x 0.06 x 0.06
2.68 to 28.34
-10<=h<=11
-11<=k<=11
-20<=l<=19
10496 / 792
0.0899

792 / 0 / 60
1.180

R1 = 0.0354
wR2 = 0.0936
R1 = 0.0362
wR2 = 0.0941
0.0033(3)

3.260 and -2539 e-/A3

Formula
Formula weight
Temperature (K)
Wavelength (A)
Space group

Unit cell dimensions (A)

Volume (A3)

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range (°)

Index ranges

Ref. collected / unique

R

int

Data / restraints / parameters

Goodness-of-fit on F2

Final R indices [I>2o(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak and hole

Table 10-3. Crystal data and structure refinement for HoFe4AlgSi6.

HoFe4AlgSi,5
799.46

293(2)

0.71073
P42/nmc

a = 8.732(5)

c =15.217(11)
1160.3(12)

4, 4.577 Mg/m3
12.886 mrn'l
1488.00

0.08 x 0.10 x 0.18 mm
2.68 to 28.23
-11<=h<=11
-11<=k<=11

_ -18<=l<=19

10527 / 795

0.1915

795 / 0 / 60

1.024

R1 = 0.0411

wR2 = 0.0906

R1 = 0.0472

wR2 = 0.0925
0.0028(3)

3.550 and -2.363 e-IA3

Rl=2‘3!le - chlllleJ. wR2=12(wIF.2 - Fe2|)2/Z(WF02)ZIW

265

 

Table 10-4. Atomic coordinates ( x 10“) for REFe4A19Si6 (RE=Tb, Er, Gd, Dy, Ho).

 

 

Wykoff x y
Position
Tb 4d 2500 2500 1642(1)
Er 2500 2500 1641(1)
Gd 2500 2500 1645(1)
Dy 2500 2500 1645(1)
Ho 2500 2500 1644(1)
Fe( 1) 8g 2500 4768(1) -4( 1)
2500 4760(1) -1(1)
2500 4782(1) -2( 1)
2500 4780(1) 1(1)
p 2500 4777(2) 2(1)
Fe(2) 8f 5006(1) 4994(1) 2500
5010(1) 4990(1) 2500
4999( 1) 5001(1) 2500
5000( 1) 5000(1) 2500
4999(1) 5001(1) 2500
Al(l) 16h 655(1) 2500 3388(1)
656(2) 2500 3387(1)
652(2) 2500 3392(1)
652(3) 2500 3392(1)
652(3) 2500 3394(2)

 

266

Table 10-4. (continued) atomic coordinates ( x 10“) for REFe,,A19Si6 (RE=Tb, Er, Gd, Dy,

 

 

Ho).
Wykoff x y
Position

Al(2) 8g 5019(1) 4553(1) 813(1)
5024(2) 4536(2) 812(1)
5009(2) 4593(2) 819(1)
5009(2) 4592(3) 820(1)
5011(2) 4587(3) 819(1)

Al(3) 8g 952(1) 7500 3503(1)
950(2) 7500 3505(1)
951(2) 7500 3491(1)
948(3) 7500 3489(2)
941(3) 7500 3491(2)

Al(4) 4c 2500 7500 -199(1)
2500 7500 -200(2)
2500 7500 -194(2)
2500 7500 -192(3)
2500 7500 -199(3)

Si(1) 8g 5939(1) 2500 1922(1)
5938(2) 2500 1925(1)
5941(2) 2500 1922(1)
5941(3) 2500 1921(2)
5937(3) 2500 1924(2)

 

267

 

Table 10-4. (continued) atomic coordinates ( x 104) for REFe4Alg,Si6 (RE=Tb, Er, Gd, Dy,

 

 

Ho).
Wykoff x y
Position
Si(2) 8g 2500 4880(1) 3125(1)
2500 4879(2) 3122(1)
2500 4875(2) 3122(1)
2500 4872(3) 3122(2)
2500 4872(3) 3120(2)
Si(3) 8g 3878(1) 2500 -282(1)
3877(2) 2500 -284(1)
3884(2) 2500 -275( 1)
3885(3) 2500 -271(2)
3889(3) 2500 -274(2)

 

268

Table 10-5. Anisotropic displacement parameters (A2 x 103)3 for REFe4A19516(RE=Tb,

 

 

Er, Gd, Dy, Ho).
U11 U22 U33 U23 U13 U12
Tb 13(1) 12(1) 9(1) 0 0 0
Br 22(1) 21(1) 16(1) 0 0 0
Gd 9(1) 7(1) 6(1) 0 0 0
Dy 6(1) 4(1) 5(1) 0 0 0
Ho 2(1) 1(1) 15(1) 0 0 0
Fe(l) 14(1) 16(1) 9(1) -1(1) 0 0
22(1) 23(1) 15(1) -1(1) 0 0
10(1) 17(1) 6(1) -1(1) 0 0
6(1) 13(1) 3(1) -2(1) 0 0
2(1) 10(1) 14(1) -1(1) 0 0
Fe(2) 12(1) 12(1) 10(1) 0(1) 0(1) 0(1)
21(1) 21(1) 16(1) 0(1) 0(1) 0(1)
8(1) 8(1) 7(1) 0(1) 0(1) 0(1)
4(1) 4(1) 5(1) 0(1) 0(1) 0(1)
1(1) 1(1) 16(1) 0(1) 0(1) 0(1)
Al(l) 16(1) 14(1) 10(1) 0 1(1) 0
25(1) 23(1) 16(1) 0 1(1) 0
16(1) 10(1) 7(1) 0 2(1) 0
13(1) 5(1) 4(1) 0 1(1) 0
8(2) 4(1) 15(2) 0 2(1) 0

 

 

aT he anisotropic displacement factor exponent takes the form: -2 1:2 [ h2 a*2 U11 + + 2
h k 3* b* U12 ]

269

Table 10-5. (continued) anisotropic displacement parameters (A2 x 103)8 for REFe,,A19Si,.3

 

 

(RE=Tb, Er, Gd, Dy, Ho).
U11 U22 U33 U23 U13 U12
Al(2) 14(1) 20(1) 9(1) 2(1) —1(1) —3(1)
23(1) 28(1) 16(1) 2(1) -l(1) -3(1)
12(1) 21(1) 7(1) 4(1) -2(1) -5(1)
8(1) 17(1) 5(1) 4(1) -2(1) -5(1)
5(1) 12(1) 15(1) 3(1) -2(1) -6(1)
Al(3) 13(1) 18(1) 12(1) 0 1(1) 0
22(1) 27(1) 18(1) 0 1(1) 0
9(1) 18(1) 12(1) 0 1(1) 0
5(1) 12(1) 11(1) 0 1(1) 0
2(1) 10(2) 19(2) 0 1(1) 0
Al(4) 21(1) 15(1) 19(1) 0 0 0
31(1) 22(1) 24(1) 0 0 0
28(2) 14(1) 25(2) 0 0 0
25(2) 10(2) 23(2) 0 0 0
27(3) 7(2) 35(3) 0 0 0
Si(1) 18(1) 15(1) 16(1) 0 -3(1) 0
26(1) 23(1) 22(1) 0 -2(1) 0
15(1) 11(1) 15(1) 0 4(1) 0
11(1) 7(1) 11(1) 0 -5(1) 0
9(1) 3(1) 24(2) 0 -4(1) 0

 

a'The anisotropic displacement factor exponent takes the form: -2 n2 [ h2 a*2 U11 + + 2
h k a* b* U12 ]

270

Table 10-5. (continued) anisotropic displacement parameters (A2 x 103)a for REFe4A195i6

(RE=Tb, Er, Gd, Dy, Ho).

 

 

U11 U22 U33 U23 U13 U12
Si(2) 13(1) 15(1) 11(1) 0(1) 0 0
22(1) 23(1) 18(1) 0(1) 0 0
9(1) 12(1) 10(1) 0(1) 0 0
5(1) 9(1) 8(1) 1(1) 0 0
3(1) 3(1) 19(2) -l(l) 0 0
Si(3) 22(1) 14(1) 13(1) 0 1(1) 0
28(1) 23(1) 20(1) 0 2(1) 0
30(1) 10(1) 12(1) 0 0(1) 0
25(1) 7(1) 11(1) 0 -1(1) 0
21(2) 4(1) 22(2) 0 -1(1) 0

 

3T he anisotropic displacement factor exponent takes the form: -2 7:2 [ h2 a*2 U11 + + 2

hka*b*U12]

271

Table 10-6. Selected Bond Distances ( A) for TbFe4Al9Si6.

 

 

Bond Bond distances (A) Bond Bond distances (A)
Tb-Al(3) 3.0176( 13) Al(1)-Al(2) 2.6589(11)
Tb-Si(1) 3.0286(12) Al(1)-Al(2) 2.6589(11)
Tb-Al(l) 3.0982(12) Al(1)-Si(3) 2.8460(14)
Tb-Al(2) 3.1002(9) Al(2)-Si( 1) 2.5845(12)
Fe(1)-Si(3) 2.3514(8) Al(2)-Al(2) 2.5877(17)
Fe(1)-Al(4) 2.4005(7) Al(2)-Si(3) 2.6375(12)
Fe(1)-Al(1) 2.4670(13) Al(2)-Si(3) 2.8590(11)
Fe(1)-Al(3) 2.5012(14) Al(2)-Al(4) 2.9584(12)
Fe(1)-Al(2) 2.5292(9) Al(3)-Si(2) 2.6404(17)
Fe(2)-Si(2) 2.3841(6) Al(3)-Al(3) 2.699(2)
Fe(2)-Si( 1) 2.4813(7) Al(3)-Si(2) 2.7142(12)
Fe(2)-Al(2) 2.5876(10) Al(3)-Al(2) 2.7182(12)
Fe(2)-Al(1) 2.6219(7) Al(3)-Al(4) 2.905(2)
Fe(2)-Al(3) 2.7903(9) Si(1)—Si(1) 2.6026(19)
Fe(2)-Tb 3.3458(4) Si(1)-Si(1) 2.721(2)
Al(1)-Si( l) 2.6225(17) Si(2)-Si(3) 2.5693(16)
Al(l)-Si(2) 2.6553(11) Si(3)-Si(3) 2.403(2)

272

 

Magnetic Characterization.

Though'single crystals were selected for the magnetic studies a small magnetic
impurity was present on the crystals, likely as powder on the surface. To remove this
impurity the crystals were sonicated in a 10% HCl solution for 1 min. Magnetic
susceptibility for REFe,,A19Si6 (RE: Tb, Er) was measured as a function of both , 1‘
temperature and field using a MPMS Quantum Design SQUID magnetometer on

powdered single crystal samples. An initial study of field dependence was conducted to

 

find a suitable field for the variable temperature studies. The measurements for
temperature dependence on all samples were then conducted under increasing
temperature using a‘500 G applied field, while field dependent measurements, conducted
at 5K, were carried out between i 55000 G, A diamagnetic correction was applied to the
data to account for core diamagnetism though no correction was made for the sample
container because the correction factor was well over an order of magnitude smaller than

the sample signal itself.

Mossbauer Spectroscopy.

Mossbauer spectra of ErFe4A19Si6were taken between 18 and 300 K, in
transmission geometry, using a constant acceleration spectrometer equipped with a
57Co(Rh) source. The spectrometer was calibrated with or-Fe and isomer shift values are
reported relative to this. The experimental data were analyzed with a least-squares

minimization routine using a sum of Lorentzian lines.

273

3. Results and Discussion.
Synthesis.

Tb or Er metal powder, iron and silicon react in excess molten A1 as described in
to yield pure REFe,,AlgSi6 product as small cube-like crystals.7 These faceted crystals can
range from micrometer to several millimeters in size often with several crystals growing
together as a polycrystalline ingot. Unlike in previous cases8 where arc-melting and flux
methods have lead to different products, for the synthesis of REFe4A195i6 both methods
do indeed lead to the same final product. The main difference however is are melting
leads to only microcrystalline samples whereas the Al flux leads to well grown mm-sized
crystals. Interestingly when the reaction is conducted with larger pieces of rare earth
metal (>1 mm sized) as the starting material in Al flux the reaction leads to the synthesis
of other non-desired phases including Tb2A13Si2” and a variety of Fe/Al/Si phases. The
reasons for this are likely due to differences in rare earth availability in the Al solution.

The REFe,,Al9Si6 phases appears to form predominantly with the late rare earth
metals while earlier rare earth metals tend to form quaternary phases of the
REFez+xAl7+xSi8 type (see chapter 9) under similar reaction conditions. To form pure
phases of each however the heating profiles and reactant ratios need to be adjusted for
each specific phase. Crystals of ErFeyAlgSi6 appears to be stabile at ambient temperatures
for indefinite time periods. However at elevated temperatures (~1000° C) the compounds

decomposes in less than a day forming a polycrystalline unit of merged crystals.

274

Structural Description.

The compounds REFe,,AlgSi6 (RE=Tb, Er) crystallize in the tetragonal space
group P42/nmc with the NdRh,,Al,5_4 structure type, see Figure 10-1. Substantial
deviations from the parent type, however, exist due to ordering of atoms at the z = 0 layer
which is highly disordered in NdRh4Alm, see Figure 10-2. This ordering is likely a result
of Al replacement of Si in the layer. The structure contains highly corrugated layers of
merged cyclohexane-like Al6 rings in the chair confirmation, shown in Figure 10-3a.
These corrugated layers stack with the direction of the corrugation being mutually
perpendicular as one moves along the c axis. The rings are comprised of Al(l), Al(2), and
Al(3) while the Al(4) atom sits between the corrugations of the layer, seen as the single
non-bonded atom in the figure. When the Fe atoms are placed in the framework they
occupy two positions in the structure the first being the center of the A16 rings, the dark
blue Fe atoms in Figure 10-3b, while the second position is inserted between the
corrugated layers, see the light blue Fe atoms in the figure. When the Si atoms are added
in a subsequent step, Figure 10-3c, the non rare earth framework is complete. Finally the
rare earth atoms are added, Figure 10-3d, to complete the complex dense 3 dimensional
structure.

The atoms in the structure exhibit a wide variety of coordination environments.
The 4 Al sites exhibit several different bonding environments. Al(l) and Al(2) exhibit
bicapped distorted trigonal pyramidal arrangements, Figure 10-4A and 10-4B. Al(3)
exhibits a 5 coordinate square pyramidal arrangement, Figure 10-4C and the final Al,
Al(4), exhibits a 2 coordinate bend arrangement, Figure 10-4D. Fe(l) has an unusual 9

coordinate arrangement that is best described as the Fe bonding to each member of a 6

275

membered ring along with bonding to 3 other atoms outside the ring itself, Figure 10-4e.
Fe(2) exhibits a distorted trigonal prismatic arrangement, see Figure 10-4F. The Si
positions range from 3 coordinate for Si(1) to 5 coordinate environments for Si(2) and
Si(3). Si(1) shows a trigonal pyramidal arrangement, Figure 10-4G, where Si(1) sits at the
apex. Si(2), Figure 10-4H, bonds to 5 atoms in a unusual 5 coordinate pentagonal
pyramidal arrangement where Si(2) is the apex atom of the arrangement. Si(3)’s
environment shown in Figure 10-41 is similar to Si(2) though not as distorted. The rare
earth ion exhibits a 20 coordinate environment, Figure 10-5. The coordination
polyhedron of the rare earth is a 33 faced unit consisted of 30 trigonal and 3 square faces.
Of note here is that all the atoms in the unit cell are contained within the RE coordination

environment except Al(4) which is at a distance of 4.879 A from the RE ion.

Magnetic Properties.

Magnetic susceptibility data for REFe4Al9Si,5 (RE = Tb and Er) show Curie-Weiss
behavior in the measured temperature range, shown in Figure10-6. The Tb compound
exhibits a 11.111 calculated from the Curie-Weiss behavior, of 10.36 1113 and a Weiss
constant of —17 K. This um value is close to the theoretical magnetic moment for Tb3+ of
9.5 1139 indicating that the Tb is in a 3+ state while the Fe does not contribute to the
overall magnetic susceptibility. For ErFe4AlgSi6 similar behavior is seen with an observed
it,“ of 10.03 1113 close to the predicted Er"+ value of 9.74 113, Figure 6b. The Weiss
constant for the Er analog is - 22 K indicating the presence of a weak antiferromagnetic
interaction at high temperatures even though a true transition is not seen. The field

dependence measured at 5 K shows a gradual increase in magnetization up to 55,000 G,

 

 

      

 
  

s5

'Isa’. {I
t?‘

on»)
Dowager

      
    
 
  
 

.115

  
  
 

 

. Fel

‘ F82 ~‘\q\"‘~'/’z & $w“
. Al "’I\\l \\ I“"I\\\ A”)
as owwswwww

1.4? one 5" all!
“V 9'0.» . 51"0

Figure 10-1. Structure of REFe4Algsi,5 (RE = Tb, Er) viewed down the c axis to show the

tetragonal symmetry of the structure.

277

e
F
m
h
R
ma

 

UH/‘\H H H /‘M.
b. i?

.- "V :1 Aw.

k? 0. D8 kw... :D,
.06..
D: > :D

   

 

II‘VO

 

a)

Figure 10-2. The z=0 layers in (a) NdRh4A115.4 showing a highly disordered Al atom

arrangement and in (b) REFe4A19Si,5 showing an ordered Al and Si layer

278

Figure 10—3. Structure of REFe4AL,Si6 (RE = Tb, Er) shown being built in stages as each
elemental type is added to the structure. First showing the (a) corrugated Al layers
(b) then addition of the Fe atoms and the (c) Si atoms to the structure and finally

the (d) RE atoms are added. The unit cell outline is represented by the thick black

lines.

 

 

 

 

 

 

 

 

 

 

 

 

8'
5w

4!!
IA‘
all
: ‘d '1
392‘“?

Mt":-

"A’" Q‘
~ 'A'.’ M
Vs?“
«th

" "x‘
,

 

 

 

Fm2 AR
“ J§,$2

     

 

SH 91

   

AI3 A13
Si2
. Al3 .Si3 Al3
S1
G)sn.l. H) as.
. A|4 .Sl1 Al1 I2 Al1

     

Al2 Al2 Ali

Figure 10-4. Local coordination environments of the individual non rare earth atomic
positions in REFetAlgSi6 out to 3 A. Bonds are drawn for interatomic distances
less than 2.4 A for Fe-Si and Si-Si, 2.7 A for Fe-Al and Si-Al, and 2.9 A for Al-

A].

280

 

_ Al(2) Al(3) Fe(2)

 

Al(3) Fe(2)

Figure 10-5. Local coordination environment of rare earth atomic position in

REFe4AIQSi6 drawn out to 3.5 A.

281

though a decrease in the slope of the response is seen at fields greater then 20,000 G for
both compounds, Figures 10-7a and b.

These properties are very similar to the previously reported compound
RE4Fe2+xAIGSi7_x.‘ Both types of compounds exhibit magnetism purely based on the rare
earth with the Fe not contributing to the overall behavior of the material. This is
intriguing given the relatively high FezRE ratio in REFe‘.Al9Si6 unlike in the previous
RE4Fe2+,‘Al.,Si7,x studies. Instead of simple transfer of electrons from the RE ions in this
structure the diamagnetic behavior of Fe is also likely due to the presence of excess Al
atoms which are comparatively highly electropositive and strongly donate electrons
filling the Fe (1 orbitals. This is consistent with the results of the Mossbauer spectroscopic

analysis.

Mossbauer Spectroscopy.

The measured spectra profile for ErFe4A19$i6 does not change in the temperature
range of 18 to 300K. It can be analyzed with two superimposing paramagnetic doublets
corresponding to the two independent Fe positions in the structure. Figure 10-8 shows a
typical spectrum taken at 18K while the hyperfine parameters are shown in Table 10-7.
No evidence for magnetic ordering of Fe atoms is seen in the spectra. The isomer shifts
are centered at the positive values of 0.37 and 0.21 mm/sec again this behavior is similar
to the previously reported values for RE4Fe2+xAl6Si7,x, which exhibited a value of 0.21
mm/sec for the single Fe position. The isomer shift values are also very close to those

observed for other Fe/Si compounds and alloys10 and point out that the iron atoms are in

282

D)
v
m
01

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

25
2 A20} . '
A a ’
£2 . x6315: . °
° 1.5a = a: -
E . a , .
E \ '. E 10- '
X 3 :. v o.
E 1 :: 5’0.
q; o if .- - . . . .
V I‘ so 100 150 200 250 300 350
. Temperature (k)
0.5~ o
a.“
' o
0 rag--i.f:t?-!m’19.9.9.9.011.I... . ..
0 50 100 150 200 250 300 350
Temperature (K)
b) 2
30,
i
25} . 0
1.5 3 l . °
0 '- % 15- o .
R \ . 10:- o .
3 .0 ’.°'
E .0 5 Lon-nu.- ----
33, \ 50 100 150 200 250 300 350
0 . 5 .. Temperature (K)
I '5‘
C '-
O ....1. f:?.?.’l’.9.’.’..t!...4...¥
0 50 100 150 200 250 300 350

Figure 10-6. Temperature dependent magnetic susceptibility for (a) TbFe4AlgSi6 and

Temperature (K)

(b)ErFe4AlgSi6. Insets show the inverse susceptibility for each compound.

283

 

.‘f‘

 

 

 

 

 

 

 

 

 

 

6
a) r ...C
C .0
a) 4* 0'
— > e
O .
2 _ O
\ 2,” 0
8A
'52 o— ’
fie :
4—0 . .
8 '2r
CD F e
(U * o
2 -4; .0

:.0....
_6.4..i.4..J.4..1-.41;.-1;-L4
-610 -410 -210 0 210 410 610

Field(G)

b)

6F

I ..o"
.02 4r .-
‘23 r o
\ 2? .
8’7 ' '
€2- 0: I
NE i
2.": e
8 '2.— e
o: . .
(U
2 -4L .'

:00...
_6b4.LJ.4#.1.4..1...1.4.1:‘..4
-610 -410 -210 0 210 410 610

Field(G)

Figure 10-7. Field dependent magnetic susceptibility for (a) TbFe4A198i6 and

(b)ErFe4A198i6.

284

more reduced state than Fe metal itself. This shift means that there is a decrease of the s-
electron density on Fe nucleus either due to a decrease in the number of s-electrons or
due to an enhanced screening of the 4s electrons by addition of electrons to the outer
orbitals. In other words the Fe atoms seem to be more reduced than those in Fe metal
which is consistent with the expectation of electron flow from the electropositive RE and
Al atoms to the more electronegative Fe atoms. This could explain the diamagnetic

behavior of the Fe apparent in the magnetic data.

4. Conclusions.

Using liquid Al as a solvent the new compounds REFe4Al.,Si6 (RE = Tb, Er, Gd,
Dy, Ho) were discovered and studied in terms of structure, magnetism and M'o'ssbauer
spectroscopy. The structure is related to the NdRh4Alm structure type, however, in
REFe4A198i6 (RE = Tb, Er, Gd, Dy, Ho) the highly disordered layer of NdRh4Al 15.4 has
been replaced by a well ordered arrangement. This could be caused by the substitution of
Si atoms for Al atoms in the disorder layer.

Magnetically the RE atoms ions in the system exhibit behavior typical of a 3+
oxidation state while the Fe atoms appear to be magnetically silent. This diamagnetic
result is particularly interesting as the Mossbauer spectroscopy suggests that the Fe atoms
appear to be more reduced than those in Fe metal. The discovery of these new and
interesting compounds again underscores the importance of and opportunities available

with molten metal fluxes in exploratory synthesis.

285

 

 

 

 

 

Table 10-7. Mossbauer Parameters, Collected at 18 K, of REFe4A19Si6 (RE = Tb, Er).

 

 

 

 

 

 

 

 

 

 

 

 

5(mm/s) AEq(mm/s) 172 (mm/s) Area (%)
Site 1 0.37 0.46 0.13 56
Site 2 0.21 0.34 0.13 44
1 00
1 00 _

99.8:
99.6}
99.4:
99.2_

99

 

Relative Transmission (%)

98.8%-

98.6*L“‘"*AA141114;...11.141..-.1--..1....
'2 -i.5 -1 -0.5 0 0.5 1 1.5 2

Velocity (mm/s)

 

 

 

Figure 10-8. Typical Mossbauer spectra of ErFe4Al.,Si6 taken at 18 K. Velocity

corresponds to the isomer shift (IS) which is referenced to a-Fe.

286

 

References

 

' Wang 2.; Dunlap R.A.; Foldeaki M. J. Mat. Sci., 1994 29, 5333-5336.

2 Perez, R.; Juarezislas, J.A.; Martinez, L. Mat. Sci. Eng. A, 1994 182, 837-840.

3 Sieve, B.; Sportouch, S.; Chen, X.Z.; Cowan, J.A.; Brazis, P.; Kannewurf, C.R.;
Papaefthymiou, V.; Kanatzidis, M.G. Chem. Mater., 200, 13, 273-283

4 Fehrmann, B.; Jeitschko, w. J. Alloys Compd., 2000 298, 153-159

5 SMART. Data Collection Software for the SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

6 SAINT. Data Processing Software for SMART System. Siemens Analytical X-Ray
Instruments Inc., 1995

7 It should be noted here that a small amount of an unknown magnetic impurity is
detected when these samples are studied magnetically however this impurity is not seen
by other methods such as powder X-ray diffraction. See the magnetic procedure for more
information and the removal if this small amount of impurity.

8 Sieve, B.; Kanatzidis M.G. unpublished results

9 Bourdreaux, E.A.; Mulay, L.N. in Theory and Applications of Molecular
Paramagnetism, John Wiley and Sons, New York 1976.

‘0 Noakes. D.R.l Sheny, G.K.; Niarchos, D; Umarji, A.M.; Aldred, A.T. Phys. Rev. B.

1983 27(7), 4317

287

 

5h

Chapter 11.

Cubic Quaternary Aluminum Silicides REsRulel49Si9(Al,Si,2_x)

(x~4) (RE = Pr, Nd, Sm, Gd, Tb, or Er) from Molten
Aluminum. Empty (Al,Si)12 Cuboctahedral Clusters and the
Assignment of the Al/Si Distribution with Neutron Diffraction

1. Introduction.

After exploring first row transition metals and finding a wide variety of differing
structural arrangements we decided to explore a second row transition metal, namely Ru,
in relation to its first row counterpart Fe. Along with structural relationships,
the valence state of the Ru ions in the structures is of great interest. The Fe ions have
previously shown to be non-magnetic and more reduced than elemental iron, shown
through magnetism and Mossbauer studies, so it is therefore of interest if the Ru ions will
show similar behavior to these Fe phases. Along with structural and'property
relationships to the Fe phases, Ru compounds are themselves of interest as very few
quaternary phases exist containing Al.

In this chapter, we report the synthesis, structure, and magnetic properties of
several new quaternary ruthenium aluminum silicides REgRU12A14gsI9(AIXSi12-x) (RE = Pr,
Nd, Sm, Gd, Tb, or Er) with a new structural type synthesized in Al flux. Our group has
recently published the synthesis of the Sm and Pr analogs along with properties of these
analogs in the literature.1 These compounds crystallize in the cubic space group Pm-3m

in a very complex structure comprised of smaller building units merging together. The

288

 

bf“

 

magnetic behavior is characteristic of a system with the RE ion in a 3+ oxidation state
and the transition metal not exhibiting a magnetic moments similar to the Fe compounds
formed in molten Al fluxes. Another point of interest is the assignment of Al and Si

atoms in the structure, these atoms are nearly impossible to determine on the basis of X-

 

ray diffraction only. In REgRulel498i9(Al,Si12,) the assignment based on bond distances
was not possible and the ambiguity was resolved by a neutron diffraction study on the Pr -"‘

analog.

 

2. Experimental Section.

Synthesis.

Reagents.

The following reagents were used as obtained: Sm, 99.9%, metal chips, Research
Chemicals, Phoenix, AZ; Pr, Nd, Er, Tb, Gd, 99.9%, -40 mesh, Cerac, Milwaukee, WI;
Pr203, 99.99%, fine powder, Sylvania, Towanday, PA; Sm203, 99.99% fine powder,
Rhone—Poulenc, Princeton NJ; Ru, Cerac, -325 mesh, 99.95% Milwaukee, WI; Si,

99.96%, -325 mesh, Cerac, Milwaukee, WI; and A1, 99.5% 20 mesh, Cerac, Milwaukee,

 

WI.

Synthetic Method.

Method 1. Sm203, Ru, Si, and Al metal were mixed in a molar ratio of
Sm203:Ru:Al:Si = l:1:40:8 and placed into an alumina container which was then flame

sealed in a fused silica tube under high vacuum (<10“‘ torr). The sample was heated to

289

‘—

810 °C, kept at that temperature for eight days, cooled to 300 °C in four days, and finally

cooled to room temperature quickly.

Method 2. In a Nz-filled dry box, RE (RE = Pr, Nd, Sm, Gd, Tb, or Er), Ru, Si,
and A] were mixed in a molar ratio of RE:Ru:Al:Si = 6.5 :10:100:8 and transferred into an
alumina tube which was sealed in a fused silica tube as described above. The sample was
first heated to 1000 °C for 15 hours, cooled to 860 °C in 10 hours, kept at 860 °C for 4
days, and finally cooled to 500 °C in three days. N aOH (aq.) solution was then used to

remove excess Al flux in the product (see below).

Method 3. Though method 2 produced many crystals for PrgRulelei9(Al,Sim),
these were too small for the neutron analysis and grew together as a polycrystalline ingot.
To form large single crystals, suitable for neutron diffraction, an adjusted heating profile
than was used. The reaction mixture was heated at 1000° C for 15 hrs followed by
cooling to 500° C in 96 hours, slightly longer than the original elemental synthesis. From
this procedure, an impure product was obtained but large single crystals of
PrgRu12Al498i9(Al,Si 12.,,) were present, largest ones exhibiting dimensions of about 2 mm
per side. The major impurity phase found in this synthesis was large crystals of
PrRuzAlw.2

In all synthetic methods, 5M NaOH (aq.) solution was used to remove excess Al
flux from the product. Purity of the final product was confirmed through comparison of
the experimental X-ray diffraction powder patterns, taken of the bulk product, to

theoretical patterns calculated from the refined single crystal data. Pure phases of all

290

 

analogs, REgRu12A1498i9(Al,Si,2_,) (RE = Pr, Nd, Sm, Gd, Tb, or Er), were obtained using

method 2.

Physical Measurements.

 

EDS analysis.
Quantitative microprobe analysis of the compounds were performed with a JEOL t“
J SM-35C Scanning Electron Microscope (SEM) equipped with Noran Vantage Energy
Dispersive Spectroscopy (EDS) detector. Data were acquired by using an accelerating
voltage of 25 kV and 100 sec accumulation time. Crystals selected from the different
synthetic methods were each measured showing no significant differences in their final
elemental ratios. The EDS analysis however consistently yielded low values for Al and Si
as seen in the past with other compounds.18 To correct for the error a standard was

measured under the analysis conditions yielding an appropriate correction factor, which

 

was then used to adjust the measured ratios of Al and Si. Analysis on several crystals was
then conducted and corrected to yield the correct elemental ratios. These analyses yielded

a consistent elemental ratio of REgRuBAlein for both compounds.

Single Crystal X-ray Crystallography.

Single crystal X-ray diffraction data of REgRu12A149Si9(Al,Si,2_,) (RE: Sm, Pr)
were collected at room temperature using a Rigaku 4 circle diffractometer with Mo Kor (A
= 0.71073 A) radiation. An empirical absorption correction based on ‘I’ scans was applied
to the data. The structure was solved by direct methods (SHELXS863) within the

TEXSAN4 crystallographic software package and was then further refined with the

291

SHELXLS package of programs. Data collection parameters, atomic positions,
anisotropic thermal parameters information is provided in Tables 11-1, 11-4, 11-5.
Selected bond distances for the Sm analog are shown in Table 11-6.

Single crystal X-ray diffraction data of REsRu12A149Si9(Al,Si 12,,) (RE: Nd, Gd, Er,
Tb) were collected at room temperature using a Siemans Platform CCD diffractometer
using Mo KOt (x = 0.71073 A) radiation. The SMART software6 was used for the data
acquisition and the program SAINT 7 was used for the data extraction and reduction. And
empirical absorption correction using SADABS8 was applied to the data and the structure
was solved in the SHELXL package of programs.9 Data collection parameters, atomic
positions, anisotropic thermal parameters information is provided in Tables 11-2 through
11-5.

As method earlier in any Al and Si-containing compound, the Al and Si positions
are difficult to distinguish by X-ray diffraction alone due to their similar X-ray absorption
cross sections. The final X-ray structure gave the lowest R values, best temperature
factors, and most reasonable composition compared to the EDS analysis. Bond distances
also agree well with those in the literature except the Si(1)-Al(3) distances of 2.789(3)
and 2.828(2) for M: Sm and Pr, respectfully, in the AlSi6 octahedra (see below structural
description) which are also reasonable for Al-Al bonds. The Si(1)-Sm distance (2.9894
A), however, suggests that the assignment of Si is correct since Al-Sm distance should be
slightly longer with values around 3.25 to 3.32 A. Nevertheless, ambiguity still remained
so single crystal neutron diffraction was used to confirm the identity of the Al and Si
positions. When the neutron refinement was conducted all the atomic assignments were

confirmed except the assignment of A1 at the M(1) position. The neutron refinement was

292

Table 11-1. Crystal data and structure refinement for REsRunAl498i9(Al,Si12,,)

(RE = Sm, Pr).

Formula

Formula weight
Temperature
Wavelength

Space group

Unit cell dimensions
Volume

Z, Calculated density
Absorption coefficient (11)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected/ unique
R.

int

Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)]“

R indices (all data) "

Largest diff. peak and hole

Rl=ZI|FOI - IFCII/ZIFOI. wR2=[E(w|F,2 - Fc2|)2/Z(wF02)2]“2

Sm,3Ru12Als3Sil7
3918.67

539.28

0.71073 A
Pm-3m (# 221)
a =11.510(1)A

Pr,3Ru,2A153Sil7
3842.99

529.84

0.71073 A
Pm-3m (# 221)
a = 11.553(2) A

1524.9(3) A3 1541.9(4) A3

1, 4.267 Mg/m3 1, 4.138 Mg/m3

11.518 mrn'l 10.094 mrn‘l

1943 1919

0.10 x 0.14 x 0.30 mm 0.20 x 0.20 x 0.16 mm

2.50 to 29.98° 2.50 to 30.110

-l6ShSl6 -16ShS16

~11SkSll -11S1(S11

-9 S l S 9 -9 S l S 9

3561 /568 3588 / 511

0.0313 0.0519

99.6 % 99.6 %
Full-matrix least-squares on F2

506/0/31 509/0/31

1.666 1.379

R1=0.035 R1=0.0178

wR2=0.083 WR2=0.0402

R1=0.045 R1=0.0282

wR2=0.124 wR2=0.1 l 15

1.949 and -2.036 e‘/A'3

293

0.762 and -o.939 e‘/A'3

Table 11-2. Crystal data and structure refinement for REgRulel4gSi9(Al,Silz,x)

(RE = Nd, Gd).

Formula

Formula weight
Temperature
Wavelength

Space group

Unit cell dimensions
Volume

2, Calculated density
Absorption coefficient (11.)
F(000)

Crystal size

0 range-

Limiting indices

Ref. collected/ unique
R(int)
Completeness to 0m,

Refinement method

Data / restraints / parameters

Goodness-of-fit on F2
Final R indices [I>20(I)] “

R indices (all data)3

Largest diff. peak and hole

R1=2llFol - chll/ZIFJ. wR2=[Z(WIF.2 -I=.1>2e<wF.2>21”2

NdsRui2A153Sir7

Gd,Ru,,A1,,s117

3869.63 3973.71
293(2) K 293(2) K
0.71073 A 0.71073 A
Pm-3m (#221) Pm-3m (#221)
a = 11.543(5) a = 11.472(3)
1538.0(10) A3 1510.00) A3
1, 4.178 Mg/m3 1, 4.369 Mg/m3
10.536 mm'l 12.639 mm‘1
1927 1959
0.017 x 0.02 x 0.02 0.06 x 0.06 x 0.06 mm
1.76 to 28.36° 1.78 to 28.28°
-15<=h<=14 -14<=h<=15
-15<=k<=15 ~15<=k<=15
-15<=l<=15 -15<=l<=15
15293 /445 14612 I435
0.0954 0.0547
99.8% 99.3 %
Full-matrix least-squares on F2
445/0/30 435/0/31
1.056 1.374
R1 = 0.0207 R1 = 0.0250,
wR2 = 0.0525 wR2 = 0.0703
R1 = 0.0265 R1 = 0.0253
wR2 = 0.0547 wR2 = 0.0705

1.065 and -1.764 e.A'3

294

1.399 and 0912 e.A‘3

 

Table 11-3. Crystal data and structure refinement for REgRu12A1498i9(Alei13_,)

(RE = Br, Tb).

Formula

Formula weight
Temperature
Wavelength

Space group

Unit cell dimensions
Volume

Z, Calculated density
Absorption coefficient (1.1)
F(000)

Crystal size

0 range

Limiting indices

Ref. collected / unique
R(int)
Completeness to 0m,

Refinement method

Er,Ru,,A1,,Si,,
3987.15

293(2) K

0.71073 A

Pm-3m (#221)

a = 11.439507) A
1497.0(4) A3

1, 4.423 Mg/m3
15.099 mm"

1991

0.02 x 0.3 x 0.3 mm
1.78 to 28.24°
~14<=h<=l4
-l4<=k<=14
-15<=l<=15

15035 /428

0.0395

99.8 %

TbBRulelflSin
4064.95

293(2) K

0.71073 A
Pm-3m (#221)

a = 11.463403) A
1506.4(3) A3

1, 4.481 Mg/m3
10.943 mm1

1837

0.02 x 0.02 x 0.03 mm
1.78 to 28.24°
-l4<=h<=14
-15<=k<=15
-l4<=l<=14
15069 / 430
0.0429

99.3 %

Full-matrix least-squares on F2

Data / restraints / parameters 428 / 0 / 32

Goodness-of—fit on F2

Final R indices [I>20(I)] “

R indices (all data)3

Largest diff. peak and hole

R1=2n120| - Iran/21F), sz=[Z(wlF.2 - Ff

0.373

R1 = 0.0140

WR2 = 0.0305

R1 = 0.0144

WR2 = 0.0307

1.204 and -0.657 e.A‘3

 

295

)2/Z(WF02)2] 1/2.

430 / 0 / 31

1.338

R1 = 0.0127

wR2 = 0.0289

R1 = 0.0129

WR2 = 0.0290

2.086 and -0.468 e.A'3

 

Table 11—4. Atomic coordinates ( x104) for REgRulel49Si,(Al,Si12.x) (RE: Sm, Pr, Nd,

 

 

Gd,Er,Tb).
Wyckoff x y 2
Position
Sm 8g 3221(1) 3221(1) 3221(1)
Pr 3211(1) 3211(1) 3211(1)
Nd 3214(1) 3214(1) 3214(1)
Gd 3225(1) 3225(1) 3225(1)
Er 3238(1) 3238(1) 3238(1)
Tb 3230(1) 3230(1) 3230(1)
Ru 121 3056(1) 0 3056(1)
3054(1) 0 3054(1)
3056(1) 0 3056(1)
3059(1) 0 3059(1)
3059(1) 0 3059(1)
3061(1) 0 3061(1)
M(1) 121 1571(2) 0 1571(2)
1568(1) 0 1568(1)
1568(1) 0 1568(1)
1573(2) 0 1573(2)
1568(1) 0 1568(1)
1571(1) 0 1571(1)

 

296

 

 

Table 11-4. (continued) atomic coordinates ( x 10“) for REBRu12A149Si9(Al,Si,2,,)
(RE: Sm, Pr, Nd, Gd, Er, Tb).

 

 

Wyckoff x y 2
Position

Si( 1) 6f 5000 2552(2) 5000
5000 2552(2) 5000
5000 2560(2) 5000
5000 2582(3) 5000
5000 2625(2) 5000
5000 2590(2) , 5000

Si(2) 3d 0 0 5000
0 0 5000
0 0 5000
0 0 5000
0 0 5000
0 0 5000

Al(l) 241 5000 1130(2) 3223(1)
5000 1122(1) 3217(1)
5000 1125(1) 3218(1)
5000 1 132(2) 3224(2)
5000 1147(1) 3228(1)
5000 1138(1) 3226(1)

 

297

 

 

Table 11-4. (continued) atomic coordinates ( x104) for REsRu12A1498i9(Al,Si,2_,)
(RE: Sm, Pr, Nd, Gd, Er, Tb).

 

 

Wyckoff x y 2
Position
Al(2) 24m 1277(1) 3618(2) 1277(1)
1270(1) 3615(1) 1270(1)
1271(1) 3616(1) 1271(1)
1280(1) 3623(2) 1280(1)
1286(1) 3622(1) 1286(1)
1282(1) 3622(1) 1282(1)
Al(3) lb 5000 5000 5000
5000 5000 5000
5000 5000 5000
5000 5000 5000
5000 5000 5000
5000 5000 5000

 

298

 

 

Table 11-5. Anisotropic displacement parameters (A2 x 103) for REsRu12A149Si9(Al,Si12_,)
(RE: Sm, Pr, Nd, Gd, Er, Tb).

 

U11 U22 U33 U23 U13 U12

 

 

Sm 7(1) 7(1) 7(1) 1(1) 1(1) 1(1)
Pr 6(1) 6(1) 6(1) 1(1) 1(1) 1(1)
Nd 8(1) 8(1) 8(1) 1(1) 1(1) 1(1)
Gd 7(1) 7(1) 7(1) 1(1) 1(1) 1(1
Er 8(1) 8(1) 8(1) 1(1) , 1(1) 1(1)
Tb 9(1) 9(1) 9(1) 1(1) 1(1) 1(1)
Ru 8(1) 70) 8(1) 0 1(1) 0
5(1) 5(1) 5(1) 0 0(1) 0
7(1) 7(1) 7(1) 0 0(1) 0
6(1) 6(1) 6(1) 0 1(1) 0
6(1) 7(1) 6(1) 0 1(1) 0
7(1) 7(1) 7(1) 0 1(1) 0
M(1) 19(1) 7(1) 19(1) 0 -6(1) 0
17(1) 4(1) 17(1) 0 -6(1) 0
19(1) 5(1) 19(1) 0 -70) 0
16(1) 7(1) 16(1) 0 -7(1) 0
8(1) 10(1) 8(1) 0 0 0
19(1) 8(1) 19(1) 0 -7(1) 0

 

The anisotropic displacement factor exponent takes the form: ~27t2[h2a*2U11 + +

2hka*b*U12]

 

Table 11-5. (continued) anisotropic displacement parameters (A2 x 103) for

RE,Ru,,A1,,Si,(A1,Si,,,,) (RE: Sm, Pr, Nd, Gd, Er, Tb).

 

U11 U22 U33 U23 U13 U12

 

Si(1) 7(1) 90) 7(1) 0 0 0
5(1) 7(1) 5(1) 0 0 0
7(1) 9(1) 7(1) 0 0 0
70) 7(2) 7(1) 0 0 0
8(1) 10(1) 8(1) 0 0 0
8(1) 10(1) 8(1) 0 0 0

Si(2) 9(1) 9(1) 8(2) 0 0 0
8(1) 8(1) 6(1) 0 0 0
10(1) 10(1) 8(2) 0 0 0
8(2) 8(2) 8(2) 0 0 0
9(1) 9(1) 9(2) 0 0 0
9(1) 9(1) 8(1) 0 0 0
Al(l) 7(1) 9(1) 10(1) -2(1) 0 0
5(1) 7(1) 9(1) -2(1) 0 0
8(1) 10(1) 11(1) —2(1) 0 0
8(1) 9(1) 9(1) -2(1) 0 0
8(1) 9(1) 10(1) -l(1) 0 0
9(1) 10(1) 11(1) -2(1) 0 0

 

 

The anisotropic displacement factor exponent takes the form: ~2n2[h2a*2Ull + +

2hka*b*U12]

300

Table 11-5. (continued) anisotropic displacement parameters (A2 x 103) for

RE,Ru,,A1,,s19(A1,s1,,,,) (RE: Sm, Pr, Nd, Gd, Er, Tb).

 

U11 U22 U33 U23 U13 U12

 

 

 

Al(2) 8(1) 13(1) 8(1) 1(1) 0(1) 1(1)
7(1) 11(1) 7(1) 0(1) 1(1) 0(1)
10(1) 14(1) 10(1) 0(1) 1(1) 0(1)
8(1) 12(1) 8(1) 0(1) 0(1) 0(1)
9(1) 12(1) 9(1) 0(1) 2(1) 0(1)
9(1) 14(1) 9(1) 0(1) 1(1) 0(1)

 

Al(3) 7(1) _ 7(1) 7(1) 0 0 0
12(1) 12(1) 12(1) 0 O 0
15(2) 15(2) 15(2) 0 0 0
6(2) 6(2) 6(2) 0 0 0
18(1) 6(1) 18(1) 0 -6(1) 0
2(1) 2(1) 2(1) 0 0 0

 

The anisotropic displacement factor exponent takes the form: -27t2[h2a*2Ull + +

2hka*b*U12]

 

301

Table 116 Selected bond distances for Sngulel49Si9(A1,Sim)

Bond Distance (A)

 

Sm-Si(l) 2.9894(9)
Sm -Al(1) 3.160(2)
Sm -Al(2) 3.197(2)
Ru-Al(1) 2.596(1)
Ru-Al(2) 2.6021(6)
Ru-M(1) 2.416(3)
Si(1)-Al(1) 2.638(3)
Si(1)-Al(3) 2.789(3)
Si(2)-Al(2) 2.617(2)
Al(1)-Al(1) 2.601(4)
Al(l)-Al(2) 2.752(2)
Al(2)-M(1) 2.798(2)
M(1)-M(1) 2.557(3)

 

302

 

able to detect a mixture of Al and Si on the position and refine a relative occupancy of
each. This assignment of disorder between Al and Si is not possible with x-ray

diffraction due to the similar scattering powers Al and Si with x-rays.

Charge Transport Analysis.

DC electrical conductivity and thermopower measurements were completed on
selected single crystals of SmBRulel49Si9(Al,Sim). Conductivity measurements were
performed with the conventional four-probe technique.lo Thermopower measurements

were made using a slow AC technique as described elsewhere.”

Magnetic Characterization.

Magnetic susceptibility for REgRu12A1498i9(Al,Si,2,,) (RE = Pr, Nd, Sm, Gd, Tb,
or Er) were measured as a function of both temperature and field using a MPMS
Quantum Design SQUID magnetometer. An initial study of field dependence was
conducted to find a field suitable for the temperature dependence studies. Temperature
dependence measurements on polycrystalline samples were then conducted under
increasing temperature within a 500 G field. Field dependent measurements were
conducted at 5K in fields between $55000 G except for the Nd analog which was

measured in fields of i1000 G.

303

3. Results and Discussion.
Synthesis.

Molten aluminum metal has again proven to be a very good medium for the
synthesis of multinary aluminum containing silicides. The title compounds grew in large
well shaped shiny black crystals with a cubic morphology, see Figure 11-1. The oxide
precursor method gave very low yield of product (2-15%) while using elemental reactants - 1‘“

yielded a larger amount of crystals though often twinned and smaller in size. The oxide

 

method however does often produce single crystals with very well defined morphologies.
To obtain pure phases of REgRu12A14gSi9(Al,Si12,,) (x~4) (RE =‘ Pr, Nd, Sm, Gd, Tb, or
Er), it was necessary to use stoichiometric RE, Ru, and Si with excess Al metal, which in
the end it can be removed in N aOH (aq.) solution. The advantage of using oxides,
however, becomes even more apparent when ternary and quaternary oxides are used,
which bring to the reaction a stoichiometrically fixed combination of elements, which
end up in the final aluminum silicide compound. When reactions are conducted off ideal
conditions several impurities are seen with the major impurity being PrRuzAlm. It should
be noted however for safety reasons that a small unknown phase, which appears to be a
Ru/Al/Si ternary phase, is also present which appears to be explosive upon contact with
air. This phase also appears to be the reason synthesis attempts with arc-melting have

failed to date due to the reaction pellets exploding under the arc during heating.

304

 

 

 

Figure 11-1. SEM image of crystals of PrgRulel498i9(Al,Sin_,) exhibiting typical

morphology.

305

Structural Description.

The compounds REgRu12A1493i9(Al,Si 12,x) (x~4) (RE = Pr, Nd, Sm, Gd, Tb, or Er)
adopt a new, rather elaborate structure type with Pm-3m symmetry, see Figure 11-2. This
complex three dimensional arrangement of atoms can be more easily understood if it is
conceptually separated into different section occupying different areas of the unit cell: an
octahedron AlSi6 at the cell center, SizRu4Al8 clusters at each face center, SiAl8 cubes at
the center of the cell edges, and M12 clusters at the cell comers. Each of these individual
units are shown separately in Figure 11-3 highlighting both the unit itself and its position
in the unit cell. For simplicity in the discussion of the structure only the bond distances
involved in the Sm analog will be highlighted.

The AlSi6 octahedron, at the cell center, contains an Al(3) atom which is coordinated by
six Si(1) atoms at distances of 2.789 (see Figures 11-3a and 11-4). The AlSi,5 octahedron
shares its six corners with six SizRu,,Al8 clusters, which are located at the face center
position of the cubic cell, Figure 11-4. The SizRu4Al8 cluster consists of an aluminum 8
square prism capped with four Ru and two Si(2) atoms, Figure 11-3b. While the Al(l)-
Al(l) bond distances are 2.601 and 2.591 A in Sngu12A1498i9(Al,Sim), the Ru-Al(1)
and Si(1)-Al(1) bond distances are 2.596 and 2.638 A for the compound. This basic unit
has been seen in several other compounds recently discovered in our group, using molten
Al as a flux, as a merged unit forming NiAl,,Si2 layers.18 While sharing Si sites with two
separate AlSi6 octahedra along one axis, the SizRu,,Al8 cluster also connects, along the
perpendicular plane, four SiA18 cubes, comprised of Si(2) and eight Al(2) atoms, which
occupy the center of each of the 12 cell edges, Figures ll-3c. The SiAl8 cube's and the

SizRu4Al8 cluster's bonding is facilitated by Al-Al and Al-Ru bonds, see Figure ll-Sa.

306

 

 

The SiAl8 units also bond two M12 clusters along the a-axis. The Ml2 clusters, shown in
3d, comprised of the mixed Al/Si site (see the neutron refinement section above), sit at
each comer of the cubic cell to complete the basic three dimensional structure. The M(1)-
M(1) bond distances in the M12 cluster are 2.557 and 2.561 A for Sngu12A149819(Al,Si,2,
x).

The M12 cluster typeis also found LiAl;,,12 but in LiAl3 the A112 clusters
encapsulate one Li atom. The inside diameter is 5.671 A with Li-Al and Al-Al bond
distances of 2.836 A. The M12 unit in REgRulelmSi9 has an inside diameter of 5.114 A
which appears to small to allow any other atom present inside the cluster. A boron analog
exists in the form of the Bl2 cubo-octahedron found in borides such as ZrB12 and others
with large electronegative metal species.13 The B12 icosahedron found in AlBlz“is
different from the M12 cluster we are discussing here due to the fact that it exhibits the
icosahedron not the cubo-octahedron form of B12. The M12 clusters in
REgRU|2A149819(A1x8112_x) are easily recognized as they are enveloped in a cavity of Ru
that helps isolate them from the remaining Al framework.

The rare earth atoms then sit in the space between those building units in an 12
coordinate bonding arrangement which is best described as an distorted
anticuboctahedron, shown in Figure 11-6A. The Ru atoms in the structure exhibit a
distorted monocapped cube arrangement bonding to four Al(l) atoms, four Al(2) atoms,
and an M(1)atom, Figure 11-6b. Si(1) exhibits a square pyramidal arrangement, bonding
to four Al(l) atoms and an Al(3) atom (Figure 11-6c), while Si(2) exhibits a cube
arrangement bonding to eight Al(2) atoms (Figures 11-3c and 11-6d). The Al/Si disorder

position M(1) exhibits a monocapped cube arrangement seen in Figure ll-6e.

307

The pure Al positions exhibits environments varying from a simple octahedron to
a highly distorted bicapped trigonal pyramidal arrangement. Al(l) exhibits the distorted
bicapped trigonal pyramidal arrangement, shown in Figure ll-6f, with bonding to two Ru
atoms, two Al(2) atoms, an Al(l) atom and a Si(1) atom. The Al(2) environment, Figure
ll-6g, is a bicapped square pyramidal arrangement with bonding involving two Ru
atoms, two Al(l) atoms, two M(1) atoms, and a Si(2) atom. The last Al atom, Al(3),
exhibits an octahedral arrangement bonding to six Si(1) atoms, see Figures 11-3a and 11-

6h.

308

_ n _ . ‘

. — .—
.- . ~

.......... — —

  
  
  

    

l8

"1184. 16314114
-~:5-'5!%Zé§/£21
'filfislggeaqe'fi

     

   
 
 

  

 

  
   

   
  

      

  
  

   

  
 

XL SHE.RU.A|.Si.M

1’

Figure ll-2.Structure of REgRulel49$ig(Al,Sim) (one units cell shown)

309

 

 

 

 

‘i”'—'—'"

 

I._ dun-- .0— .u---“ .0- _. O.

l
\

x
l

 

 

 

 

 

Figure 11-3. Structural units of REgRu12A1498i9(Al,Si12,,) (A) AlSi6 unit at body center

position, (B) Ru2A14Si unit at center of cell faces (two units of Ru2A14Si removed for

clarity), (C) SiAl8 units at center of cell edges, and (D) the M12 unit at the cell comers.

310

 

 

 

"
Ill \

A K;
- '1
‘
Int
1' “

 

L.

O M Gnu .910) $1110) ®A1(4)

Figure 11-4. View of the AlSi,S octahedron at the cell center sharing Si atoms with the

RuzAl4Si clusters at the face centers.

311

 

 

 

 

 

 

 

SizRU4AI3
,1—7‘

@ Si(2) @ Al(2) % Al(1) Ru 0 Si(1)

 

 

 

 

M(1) cluster

Figure 11-5. (A) The SiAl8 cube and the Ru2A14Si clusters bonding through Al-Al and

Al-Ru bonds (B) and a SiAl8 cube linked to a M12 cluster in a through Al-Al bonds.

312

 

 

 

 

Figure 11-6. Coordination environments of the various atoms in the unit cell.

313

 

Charge Transport Properties.

Thermopower measurements on single crystals of Sngu12Al4gSi9(Al,Si,2.x) were
carried out over a temperature range of 50-300 K while the electrical conductivity
measurements were conducted between 5-300 K. The exhibited thermopower values are
small, with a value of about -4 uV/K at room temperature (see Figure 11-7A), which is
indicative of the expected metallic behavior of these compounds. The gradual decrease in
thermopower values as temperatures approaches zero is also consistent withthis
expectation.

The compound exhibits a conductivity value of about 4000 S/cm at room
temperature, see Figure ll-7B. And as expected a general increase in conductivity values
are seen as temperatures decrease. The high conductivity values and the temperature
dependence of the conductivity values are in agreement with the expected metallic

behavior for the compound as seen in the thermopower measurements.

Magnetic Properties.

Sngu[2A149S19(Al,2_,Si,) exhibits a ferromagnetic transition at Tc = 10 K, shown
in Figure 11-8A, and exhibits not to display Curie-Weiss behavior at any temperature,
Figure ll-8b. A small linear section of the inverse plot is seen after 200 K though when a
ueff is found a value of 1.95 B.M much larger than the expected value of 0.84. This
absence of Curie-Weiss behavior is characteristic behavior in many Sm3+ and Eu3+
systems where the spacing of the multiplet levels above the ground state is not large
compared to kT causing thermal population of excited magnetic states.15 The

magnetization vs. field curve at T = 5 K for the Sm compound, Figure ll-8c, exhibits a

314

 

 

 

Thermopower (11V/K)

S

 

 

 

 

 

14LL1L1_LAI k111LkLLLL4

50 1 00 150 200 250 300

Temperature (K)

 

6000

Conductivity (S/cm)

5500 i
5000 E
4500 E
4000 I
3500 E
3000 E

2500 '

 

 

2000

 

4
11A 1 L L l l L LL44 l l l l l L A A IL A l A

 

50 100 150 200 250 300

Temperature (K)

Figure 11-7. (A) Thermopower values for Sngu12A1493i9(Al,Sim). (B) Conductivity

values of Sngulel498i9(Al,Si12,). Values shown have been corrected for the

contribution of the Au electrodes used to make contacts to the samples.

315

 

hysteresis with a coercive field of approximately 6500 G. This hysterisis is evident until
5000 G where the curves remerge, due to changes in the curves slope starting at 2000 G.
The magnetization then exhibits a linear relation with field out to the maximum measured
value of 55000 G. No sign of magnetic saturation due to field aligning of magnetic spins
is evident in the field dependence measurements.

The Pr analog exhibits a ferromagnetic transition at 20 K and exhibits Curie-
Weiss behavior above 50 K with a Weiss conStant of 8 K, Figure 11-9a and b. From
Curie-Weiss behavior a it,“ of 3.40 B.M. can be assigned to each Pr atoms which is very
close to the expected value of 3.58 B.M. for a Pr“.40 This matching 11,“ value is calculated
assuming diamagnetic behavior for the other atoms in the structure, highlighting the
absence of substantial magnetic moment on the Ru atoms.”22 The field dependence curve
for the PrgRulelagSi9(Al12_,Si,), Figure ll-9c, exhibits a similar change in slope at 2000
G but does not exhibit a significant coercive field. After the change in slope the
magnetization increases linearly as a function of field, as seen in the Sm analog. In
addition, no sign of magnetic saturation is seen in PrgRu12A149Si9(Al,2,,Si,) similar to the
behavior seen in the Sm analog.

ngRulelagSi9(Al,2_,Si,) exhibits a ferromagnetic transition at roughly 5K and

Curie-Weiss behavior at temperatures greater than 50 K. The effective moment from a fit
of the CW behavior is 3.38 B.M. which signifies that the Nd is also in the 3+ oxidation
state in this compound as the accepted value for Nd3+ is 3.50 B.M.. The Weiss constant
for the fit is -5.48 K. When the field dependence is examined the compound shows a
general increase in moment up to the maximum field measured of 1000 G. The slope of

the response changes slightly however at around 100 G so that past this field the moment

316

 

b)

1/x
( mole RE ion / emu)

O
V

Magnetization
(B.M. / mole RE ion)

(emu / mole RE ion)

08
07
06
05
0A»
03
02
OJ

1000

800

600

400

200

-200

O
N

.0
_L

.0
—L

I
P
—L

I
P
_L

-02

-6 10‘

 

I

l

I

 

é

 

0-

 

2 0 3 0
Temperature (K)

40 50

 

 

l l 1 J

 

l 4

 

100 150 200
Temperature (K)

250

300 350

 

 

l 1 l I

 

l

 

-410‘ -210‘ 0
Field (G)

2104

4

410‘ 610

Figure 118 (a) Magnetic susceptibility vs. temperature of Sngu12A1498i9(Al,2,,Si,) (b)

inverse magnetic susceptibility and (c) response as a function of field.

317

 

(emu / mole RE ion)

13)

1/ x
(mole RE ion / emu)

8

mole RE ion)

Maqnetiization

(B.M.

0.5

0.4

0.3

0.2

0.1

250

200

150

100

P
OUAIIAAL lLAJllALAll llAlLA‘ALlLAA

OA
03
02
GA
0
~04
-02
-03

-OA

610‘

 

 

 

h.

Y

4 0 6 0
Temperature (K)

 

80 100

 

I

r
O

 

 

150 200
Temperature (K)

100 250 300 350

 

 

 

_A A L l

_L l _A_ 4‘. A 1 A L A l 1 1 L A

 

-210‘ 0 210‘ 410‘ 610‘

Field (G)

-410‘

Figure 11-9. (a) Magnetic susceptibility vs. temperature of PrgRu12A149S19(AlmSix) (b)

inverse magnetic susceptibility and (c) response as a function of field.

318

 

 

increases slower then before this transition. Gngu 12A149319(A112,XSIX) also exhibits a
ferromagnetic transition, shown in Figure 1 1-11, at about 30 K while at high
temperatures exhibits Curie-Weiss behavior. From the behavior a ueff of 8.48 and a Weiss

constant of 31 K is found. This pm is very close to the expected value of 7.91

corresponding to a Gd in the 3+ oxidation state.

ErgRu12A14()Si9(Al 12_,Si,), Figure 11-12, also exhibits Curie-Weiss behavior at
high temperatures with an effective moment of 9.55 BM which is again very close to that
expected if the Eris in a 3+ oxidation state and the Ru atoms do not contribute to the
overall magnetism of the compound. From the fit of the Curie-Weiss behavior a Weiss
constant of -2.90 K is found for the Er analog. At low temperatures a no transition can be
seen. The field dependence behavior for ErgRu12A149Si9(Al,Si12,) shows a dramatic
increase in magnetization as a function of field up to 20,000 G where the increase
significantly decreases with increasing field. Saturation of the magnetic moment is not
seen in the sample up to the maximum field measured of 55,000 G with only 6 of the
total 9.4BM possible seen aligned at this maximum field.

The Tb analog exhibits an antiferromagnetic transition at about 3K, see Figure 11-
13. The value for tie“ is found as 9.66 BM while the Weiss constant is -2.66 K in this
system. This p,“ value is comparable to the Tb3+ value of 9.50 which is expected as seen
in the other analogs. The field dependence behavior for TbgRu12A149$i9(Al,Siu,,) shows a
dramatic increase in magnetization as a function of field up to 20,000 G where the
increase significantly lessens with increasing field. Saturation of the magnetic moment is

not seen in the sample up to the maximum field measured of 55,000 G with only 6 of the

total 9.50 BM possible seen aligned at the maximum field.

319

 

 

_A
O

on

Y ‘I’ V 1_T 1'

(emu / mole RE ion)

250

200

150

1/x

(mole RE ion / emu)

100

0'!
O

F3
N)

F3
_L

Magnetization
a

BM. / mole RE ion)

03 ’
-1500

02 '

 

T

V ‘I

Y T 1 Fr" 1

 

..
1 1 ,.n““ l

LAlLAnn‘a

 

 

1:

10 20 30

Temperature (K)

 

Yj

VYYjIVY

I

 

.L

LllLLlJlllllll

100 150 200
Temperature (K)

ILJLI

250

1 l A L

50

l L A I

[All

350

 

300

 

VYYIIWV‘ITIV

'I

 

4 A l

I l I I l A l J_l_ I. l I l l l l l l l

 

llLLll

 

-1000 -500 0 500
Field (G)

1000 1500

Figure 11-10. (a) Magnetic susceptibility vs. temperature of ngRu12A1498i9(Al,2,,Si,) (b)

inverse magnetic susceptibility and (c) response as a function of field.

320

 

 

 

X
(emu / mole)

26;.

)-
10L

 

 

O L+L.AA“‘AAALALALAAA AAAIAJLJAILAL
vvvvvv v v v v v v v v v

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

 

35

 

25:— '

20'— '

YIVT"

15

10

YTTIFTI
O

1/x
(emu / mole RE ion)

 

 

7177]!
O

o.‘JJAIlAlltlAlAlLJJL4LLLIJJLLLJALJ

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

Figure 11-1 1. (a) Magnetic susceptibility vs. temperature of Gngu12A1498i9(Al12-,Si,) (b)

inverse magnetic susceptibility and (c) response as a function of field.

321

 

 

2.5 L

1.5 L

x
(emu / mole RE ion)

rT
O

0.5 _

 

 

h
0 A A A A J. A 1 A a l A A A L 1 4 A A A l A A LA 4.

0 10 20 30 40 50
Temperature (K)

 

 

30_
as; .
20E- o
15) .'

10’ '

5_ '
r-
y-
b .
O AlAlAAAAI4LALIAALAILLAJJAAAJLAAAL

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

1/x

(mole RE ion / emu)

 

 

 

mole RE ion)
., . --
O
O

Magnetization
rb
O

(BM.

-4 ._ .
r . . . .
-6 4 9 A l J; L A 1 AL A J l A A 1 4L 1 .L 1 J

-610‘ .410‘ -210‘ 0 210‘ 410‘ 610‘
Field(G)

 

 

Figure 11-12. (a) Magnetic susceptibility vs. temperature of ErgRu12A149$i9(Al 12,,,Si,) (b)

inverse magnetic susceptibility and (c) response as a function of field.

Collective magnetic properties for this series are shown in Table. 11-20. Each
analog appears to exhibit Curie-Weiss behavior at high temperatures. Each compound's
behavior is indicative of these systems exhibiting the RE ions in 3+ oxidation states while
the Ru atoms, as in other compounds studied, appear to be effectively diamagnetic. At
low temperatures a variety of interatomic transitions, both ferromagnetic and
antiferromagnetic, appear except in the samples of ErgRu12A149Si9(Al12_,Si,) that does not
show a low temperature transition. Also of note is that the type of interaction appears to
go from ferromagnetic in nature to absent to antiferromagnetic as one decreases the size
of the RE ion in the structure. This is likely due to the RKKY interactions that oscillate
from positive coupling constants to negative values as a function of distance between

magnetic centers.“5

4. Conclusions.

Liquid Al has again proven to be an effective solvent in the synthesis of
compounds containing second row transition metals such as Ru (highlighted here) as was
seen previously with first row transition metals. The ferromagnetic compounds of the
series REgRu12A149Si9(Aln_,Si,) form well defined crystals in Al readily between 800 and
10000 C much lower required in traditional synthetic methods. In terms of relations to the
Fe analog studied the Ru compounds also exhibit 3+ oxidation states for the RE ions and
effectively diamagnetic behavior for the Ru atoms. This absence of magnetic moment on

the transition metal is a recurring theme in the chemistry where the transition metal is one

323

 

 

x
(emu / mole RE ion)

1/x
(mole RE ion / emu)

Magnetization
mole RE ion)

(BM.

1.4

 

1.2 E0. '

0.8
0.6
0.4 ,

0.2 —

 

 

O _J__L A A _L _A _A A A 1 A A A A J I A A A l 4 .A L A

0 10 20 30 40 50
Temperature (K)

 

 

 

 

P
b—-
.
b—
)-
l‘
P
F
5 P
V
p
o JJAIAAAAJLLLA[ALAIIAAAJIAAAAIAJSLJ

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

 

 

 

 

6
O
.0
4: .0.
0
: O
O
2*: .
0
0~ I
. O
’ O
-2: .
. O
.4: o.
O
I..0'
- 1.11.111...11P111144..1
-610‘ -410‘ -210‘ 0 210‘ 410‘ 610‘
Field(G)

Figure 11-13. (a) Magnetic susceptibility vs. temperature of TbgRU12A149819(A112_XSIX) (b)

inverse magnetic susceptibility and (c) response as a function of field.

324

Table 11-7. Temperature dependent magnetic behavior of REgRu12A149Si9(Al,Si,2_,) (RE:

Pr, Nd, Sm, Gd, Tb, or Er)

 

 

RE Hen 11.1.... 9 (K) Tc/T n... (K)
Pr 3.40 3.5 8 8 20

Nd 3.38 3.62 -5 5

Sm * 0.84 * 10

Gd 8.48 7.94 31 30

Er 9.55 9.57 -2 none

Tb 9.66 9.72 -2 3

* The Sm analog does not appear to exhibit Curie-Weiss behavior at any temperature

325

of the more electronegative elements in the structure. The major difference however in
the Fe/Ru comparison occurs in the overall structures of the compounds. While the
oxidation states of the components are likely very similar the overall structure varies
greatly in both design and symmetry between the Fe and Ru systems unlike the other
family explored, namely the structural relationships between Ni and Pd (see chapters 3

and 8).

326

 

References

 

1 Sieve, B.; Chen, X.Z.; Henning, R.; Brazis, p.; Kannewurf, C.R.; Schultz, J.A.;
Kanatzidis, M.G. J. Am. Chem. Soc., 2001 29, 7040-7047

2 Tiede, V.M.T.; Ebel, T.; Jeitschko, W.J. J. Mater. Chem, 1998 8(1), 125-130.

3 Sheldrick, G.M., in "Crystallographic Computing 3" (G.M. Sheldrick, C.C. Kruger, R.
Doddard, Eds), pp. 175-189. Oxford Univ. Press, Oxford, UK, 1985.

4 "TEXSAN -- TEXRAY Structure Analysis Package", Molecular Structure Corporation,
The Woodlands, TX, 1992.

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

6 SMART. Data Collection Software for the SMART System. Siemans AnalyticalX—ray
Instruments Inc., 1995

7 SAINT. Data Processing Software for SMART System. Siemans Analytical X-ray
Instruments Inc., 1995

8 Sheldrick, G.M. University of Gottingen, Germany. Manuscript to be Published

9 Sheldrick, G.M. SHELXL. Structural Determination Programs, Version 5.0; Siemans
Analytical X-Ray Instruments Inc.: Madison, WI, 1995

'0 Lyding, J .W.; Marcy, H.O.; Marks, T.J.; Kannewurf, C.R. IEEE Trans. Meas. 1988 37,
76-80.

” Marcy, H.O.; Marks, T.J.; Kannewurf, C.R. IEEE Trans Instrum. Meas. 1990 39, 756-
760.

‘7 Yoshi-yama, T.; Hasebe, K.; and Mannami, M-H. Journal of the Physical Society of

Japan 1968 25, 908.

327

 

 

‘3 Greenwood, N .N .; Eamshaw, A. Chemistry of the Elements Pergamon Press, Oxford,
1984 p. 168

” Higashi, I.; Sakurai, T.; and Atoda. T. J. Solid State Chem. 1977 20, 67-77.

‘5 Bourdreaux, E.A.; Mulay, L.N. in Theory and Applications of Molecular
Paramagnetism, John Wiley and Sons, New York 1976.

‘6 Meyers, H.P. Introductory Solid State Physics, Taylor & Francis, Bristol, PA 1997.

328

 

 

Chapter 12

Conclusions and Future Work

Throughout this research molten Al has proven to be an excellent solvent in the
synthesis of quaternary phases containing tetrelides. The results indicate that large
advances in the study of complex phases are achievable utilizing molten metal fluxes in
terms of both the discovery of new phases as well as the large crystal formation of known
compounds. These advances, possibly due to the superior properties of molten metal
fluxes compared to traditional high temperature synthetic methods, include a decreased
dependence on excessive reaction temperatures. When these benefits of flux synthesis are
combined with the ability to explore the chemistry of most elements of the periodic table
a small inkling of the possibilities for the future are realized. The research will be largely
synthetic in nature but the results will affect both basic and applied chemistry since the
ideas and principles can be applied equally to both synthesis of unknown phases as well
as the growth of large crystals of target compounds.

This work has largely concentrated on the formation of new quaternary rare earth
transition metal aluminum tetrelide phases using molten Al as a solvent. Though many
new and interesting compounds have been discovered, the results represent only one
small area of the total possibilities for research. The benefits of molten metal solvents are

by no means limited to liquid Al. Many other metal fluxes can also be explored including

329

 

elemental solvent with relatively low melting points such as Pb,l Ga2 and Bi3. One can
also easily imagine the use of binary or ternary compounds with similar low melting
points such as NizB4 as solvents for similar reactions. It should be noted that work with
Ga fluxes is currently being studied by other students in the group and thus far have
shown very promising results.

Along with varying solvent type one can explore the chemistry of differing
elements than those studied here within Al solvents. Work here has concentrated on the
first row transition metals; however, recent work using third row transition metals has
shown great promise. The overall structures of the new compounds vary but similar
structural units are often seen. These phases are often interesting as new compounds but
also yield an opportunity to compare and contrast the chemistry of third row elements
with that of first row transition metals in unusual synthetic conditions such as molten
metal solvent systems.

Another area worthy of examination is the use of complex starting materials.
Traditionally the use of such materials as reactants has been to introduce structural units
to the reaction and build a structure based on the merging of these pieces. With molten A1
this simple merging of units is not likely possible due the high temperatures and the
reducing power of the molten Al. While the insertion of complex building units is
unlikely, the use of binary or ternary phases, including oxides, does provide an excellent
means to deliver exact elemental ratios in a reactive state into the reaction mixture.

With all these areas for exploration many more avenues could be invstigated than
time allowed in this thesis. One of the more pressing examples would be to examine

crystal growth processes in terms of intermediate formation. This would provide an

330

 

excellent opportunity to examine the behavior of reactants in molten metals as well as
provide insight into the formation pathways that can lead to such complex final products.
Another experiment would involve studying the reactions of early transition metals
matched with later ones to form quaternary systems, excluding the rare earth ions. Along
with these types of reactions being more applicable to industrial needs they may provide
insight into the competitive nature of electron transfer between the different transition
metals as well as the interaction with main group elements such as Al, Si and Ge. It
would also be very interesting to explore reaction systems containing B as the tetrelide
component. Elemental boron acts significantly different in terms of general chemical and
physical properties than the other tetrelides used. This difference may lead to unusual
behavior in the final reaction products both in terms of structures and electrical transport

properties.

331

References

 

' Okada S, Kudou K, Miyamoto M, Hikichi Y, Lundstrom Nippon Kagaku Kaishi 1993 5,

681-684

2 Zhuravleva, M. Chen, X.C. Wang, X.; Schultz, A.; Ireland, J .; Kannewurf, C.R.
Kanatzidis, M.G. Chem. Mater. in press

3 Canfield, P.C.; Fisk, Z. Philos. Mag. B 1992 65, 1117-1123

4 Canfield, P.; Gammel P.L.; Bishop D.J. Physics Today 1998 51, 40-46

332

 

 

. u .
, .. ..“".
5.34.12.2' . W H , . httutiuflfitnprpuunflibr 111111.) 1* t I

      
  
 

         
       
        
   

 

EVUNI VSER IV LIB

11111111111111 ‘1

l
‘l ( l

 
 
        

     

["‘l———_‘ - - I
.. . . U... . .-