f

, . .

. .. 5......rh 915..

5.1.: [-3.5 . a:
f v .(u

2 1

..

.. ,.
$23”an ﬂ m a:
i...b.r.uv

gnagﬁueaxnv. $2;
7» r! 91711.... ‘3“;
v e5. 3.; V t
Y a.
g ‘
”w," i "Ad”.tudu

r ,..........u:.. . If. I
Jonah” rt: wry: L. .

fl.~l r
o...

l. .1
.axt‘

£3335

.
£5. .. .—
.....l..x...

. if 3.
”enigma-f
. a : ,3
::,...H....::.
I , .:
f. 7.2.: I. x.:. ,
r1 1... .X y w. . .sdhxw
£3.) .12 . ...
:..C.,I(
. 1
”v N. . 1’ :- 7 v.
u 5.? In. in 3‘ ,
Zinc“. .«nlu r531: $53”: ti.)
. a t r . . Sf . y
”:r c a: 5.“.txvg..1.:.v
a r: 4.5... ..v
. u. 1‘
:Lzrbta!
‘ v-

.::v..

...

.. am...) I ..

PA. 1:) In. .1511! a) IMHsJ

‘— A . :6.
. a x 9..
In

I, .

E63.-

...;2!: .
. 1"1y7!??.¥l
3'1:
.7". ’1:
{33:1

van... . . crﬂrhmmdﬁz
.:sa..w:...»$$. an:
«3 ’33.»? 3.. 3

«iron! :s ,

I

.x. . .2113!

0‘
a

I I."
2. . .3
. igihuwu :
.Jﬂuﬁfng‘af. $1.5. a . .._
in I. I \ .laliahrﬁvxt?
.m. ..
.54. ﬁctittuv:
as. s, f»: .
t , a: 2.33.4535
2.535

 

x. . is."
o. s} x. 39.72.93
. Q .
u m . ‘ ’t
.

I1.)-
;

 

.rh,

. . ‘ .

5:} tut
u

. 5.. :
.7.

 

300]

This is to certify that the
thesis entitled

THE METAL FLUX SYNTHESIS AND CRYSTAL GROWTH OF
BINARY AND TERNARY BORIDES

presented by

Britt Andrew Vanchura II

has been accepted towards fulﬁllment
of the requirements for the

 

MS. degree in Chemistry

    

Major Professor’s Signature
ﬂj 0 6 / Q 00 6

Date

 

MSU is an Aﬁirmative Action/Equal Opportunity Institution

 

 

l "TIBRARY

Michigan State
University

 

 

 

—.-.—.—.-.—--.—.—.-.-.-.—.—.—.-..—A—.-o—o-c-u-o-o-t---o--u-

00-0-.----p-I-c-u-o-n-c-u-n-o-u-o-n--—.-»—.—.-.-

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:!ClRC/DaleDue.indd-p.1

THE METAL FLUX SYNTHESIS AND CRYSTAL GROWTH
OF BINARY AND TERNARY BORIDES

By

Britt Andrew Vanchura II

A THESIS

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

MASTER OF SCIENCE
Department of Chemistry

2006

ABSTRACT

THE METAL FLUX SYNTHESIS AND CRYSTAL GROWTH OF
BINARY AND TERNARY BORIDES

By

Britt Andrew Vanchura II

Borides form a large class of compounds with diverse structural and physical
properties. Traditionally, high-temperature furnaces, arc welders, and radiofrequency
(RF) induction furnaces have been used to reach the elevated synthesis temperatures of
most borides. By using a low-melting point metal as a solvent-like ﬂux, the reaction
temperatures for boride synthesis can be lowered by several hundred degrees. Our group
has recently had success using molten gallium to synthesize the new borides
RE3-XC28ig(Blz)3 (RE = Tb and Br) and B-SIB3, which are only obtainable through flux
synthesis. Literature searches have found that molten copper and aluminum have also
been used successfully as fluxes for the synthesis and crystal growth of many borides.

In the present work we expanded the use of gallium as a metal ﬂux by
synthesizing and growing crystals of borides in the Cr—B, Mo—B, W—B, Mn—B, Re—B,
Ho—Mn—B, and RE-Re—B (RE = Dy, Er, Yb) systems. Molten copper was used to
synthesize borides from the Cr—B, Ru—B, and RE—Ru-B systems. Structural
characterization by single crystal x-ray diffraction was carried out on all borides
produced from the metal ﬂuxes. The structures of the borides and general observations

on the metal ﬂux technique and its applications to boride synthesis are discussed.

Copyright by
Britt Andrew Vanchura H

2006

ACKNOWLEDGEMENTS

I would like to thank my wife Jaime for being the wonderful person that she is.
Nobody could be more supportive, encouraging, and understanding. My daughter
Caroline has been an inspiration. Even on the bad days she’s there to great me with a
“Da!” and hug. I could never thank my parents enough for all that they have done. I
thank God everyday for giving me such a loving family.

I would like to thank my advisor, Mercouri Kanatzidis, for all his help, support,
and advice. The help and support from past and present Kanatzidis group members is
greatly appreciated. The support staff in the chemistry department has been a tremendous

help, even when things were more than a little crazy.

iv

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................................... viii
LIST OF FIGURES ................................................................................................... xiii
LIST OF ABBREVIATIONS .................................................................................... xv
Chapter One The Metal Flux Technique in Boride Synthesis ................................ 1
A. Introduction .................................................................................. 1
B. The Metal Flux Technique ........................................................... 1
C. Structural Features in Borides ...................................................... 5
D. Chemical and Physical Properties of Borides .............................. 7
Chapter References .......................................................................................... 10
Two The Metal Flux Synthesis of Group Six Borides ............................... 13
A. Introduction .................................................................................. 13
B. Experimental ................................................................................ 15
Reagents ....................................................................................... 15
Furnaces ....................................................................................... 15
Synthesis and Isolation ................................................................ l7
Cr5B3 ...................................................................................... 17
033.; ...................................................................................... 17
CrzB3 ...................................................................................... 18
Cl‘Bz ........................................................................................ 19
M08 ....................................................................................... 19
WB ......................................................................................... 20
C. Physical Measurements ................................................................ 20
Energy Dispersive Spectroscopy ................................................. 21
Single Crystal X-ray Crystallography .......................................... 21
D. Structural Description .................................................................. 38
Cl'sB3 ...................................................................................... 38
Cr3B4 ...................................................................................... 38
0ng ...................................................................................... 38
CrBz ........................................................................................ 41
MB (M= Mo, W) ................................................................... 41
E. Results and Discussion ................................................................. 41
F. Conclusions ................................................................................... 44
References .......................................................................................... 46
Chapter Three The Metal Flux Synthesis of Group Seven Borides ........................... 48
A. Introduction .................................................................................. 48
B. Experimental ................................................................................ 49
Reagents ....................................................................................... 49
Furnaces ....................................................................................... 50
Synthesis and Isolation ................................................................ 51

Chapter Four

MnB4 ...................................................................................... 52
R632 ....................................................................................... 52
HoMnB4 ................................................................................. 53
REReB4 (RE = Dy, Er, Yb) ................................................... 54
C. Physical Measurements ................................................................ 55
Energy Dispersive Spectroscopy ................................................. 55
Single Crystal X-ray Crystallography .......................................... 55
D. Structural Description .................................................................. 71
MnB ....................................................................................... 71
WE; ...................................................................................... 71
ReBz ....................................................................................... 74
HoMnB4 and REReB4 (RE = Dy, Er, Yb) ............................. 74
E. Results and Discussion ................................................................. 77
MnB ....................................................................................... 77
MnB4 ...................................................................................... 78
R632 ....................................................................................... 80
HoMnB4 ................................................................................. 80
REReB4 (RE = Dy, Er, Yb) ................................................... 80
General Discussion ...................................................................... 82
F. Conclusions ................................................................................... 83
References .................................................................................... 84
The Metal Flux Synthesis of Ruthenium Borides .............................. 86
A. Introduction .................................................................................. 86
B. Experimental ................................................................................ 87
Reagents ....................................................................................... 87
Furnaces ....................................................................................... 88
Synthesis ...................................................................................... 89
R0733, RilzB3, and RuBz ........................................................ 89
B-RuBz ................................................................................... 89
RERu434 (RE = Y, Ce, Sm, Dy, Yb) ..................................... 9O
RE,‘(Ru4B4)y (RE = Pr, Nd) .................................................... 90
Isolation from ﬂux ....................................................................... 90
C. Physical Measurments .................................................................. 91
Energy Dispersive Spectroscopy ................................................. 91
Single Crystal X-ray Crystallography .......................................... 92
D. Structural Description .................................................................. 110
RU7B3 ...................................................................................... I 10
RU233 ...................................................................................... 1 10
RuBz ....................................................................................... 110
B-RuBz ................................................................................... 114
RERu4B4 (RE = Y, Ce, Sm, Dy, Yb) ..................................... 114
E. Results and Discussion ................................................................. 116
F. Conclusions ................................................................................... 120
References .......................................................................................... 121

MnB ....................................................................................... 51

vi

Chapter Five Gallium-Rich Gallides Isolated as By-Products from Molten Flux

Reactions Containing Boron .......................................................................... 123
A. Introduction .................................................................................. 123

B. Experimental ................................................................................ 124
Reagents ....................................................................................... 124

Furnaces ....................................................................................... 124

Synthesis and Isolation ................................................................ 124

MnGa4,65 ................................................................................. 124

ReGa4_5 ................................................................................... 125

. NIo_57Ga4_43 .............................................................................. 126

C. Physical Measurements ................................................................ 127
Energy Dispersive Spectroscopy ................................................. 127

Single Crystal X-ray Crystallography .......................................... 127

D. Structural Description .................................................................. 136
MGa4+x (M= Mn,Re) .................................................................... 136
Nio_57Ga4_43 .................................................................................... 140

E. Results and Discussion ................................................................. 142

F. Conclusions ................................................................................... 144
References .......................................................................................... 145
Chapter Six Conclusions and Future Work ........................................................... 146
References .......................................................................................... l 50
Appendix RF Furnace Design and Setup ............................................................ 151

vii

Table 2-1.

Table 2-2.

Table 2-3.

Table 2-4.

Table 2-5.

Table 2-6.

Table 2-7.

Table 2-8.

Table 2-9.

Table 2-10.

Table 2-11.

Table 2-12.

Table 2-13.

Table 2-14.

Table 2-15.

Table 2-16.

Table 2-17.

LIST OF TABLES

Crystal data and structure reﬁnement for Crng, 24

Atomic coordinates (x 104) and equivalent isotropic displacement

parameters (Azx 103) for Cr5B3. 25
Anisotropic displacement parameters (Azx 103) for Cr5B3. 25
Bond distances (A) for Cr5B3. 26
Crystal data and structure reﬁnement for 033;. 27

Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (Azx 103) for CI'3B4. 28
Anisotropic displacement parameters (Azx 103) for 033;. 28
Bond distances (A) for Cr3B4. 29
Crystal data and structure reﬁnement for Cr2B3. 30

Atomic coordinates (x 104) and equivalent isotropic displacement

parameters (Azx 103) for Cr2B3. 31
Anisotropic displacement parameters (Azx 103) for CrzB3. 31
Bond distances (A) for Cr2B3. 32
Crystal data and structure reﬁnement for CrBz. 33

Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (Azx 103) for CrBz. 34
Anisotropic displacement parameters (Azx 103) for 032. 34
Bond distances (A) for CrBz. 34

Crystal data and structure reﬁnement for MoB and WB. 35

viii

Table 2-18.

Table 2-19.

Table 2-20.

Table 2-21.

Table 3-1.

Table 3-2.

Table 3-3.

Table 3-4.

Table 3-5.

Table 3-6.

Table 3-7.

Table 3-8.

Table 3-9.

Table 3-10.

Table 3-11.

Table 3-12.

Table 3-13.

Table 3-14.

Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103)for M03 and WB. 36

Anisotropic displacement parameters (Azx 103) for M03 and WB. 36

Bond distances (A) for M08. 37
Bond distances (A) for WB. 37
Crystal data and structure reﬁnement for MnB. 58

Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (Azx 103) for MnB. 59
Anisotropic displacement parameters (Azx 103) for MnB. 59
Bond distances (A) for MnB. 59
Crystal data and structure reﬁnement for MnB4. 60

Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (Azx 103) for MnB4. 61
Anisotropic displacement parameters (A2x 103) for MnB4. 61
Bond distances (A) for MnB4. 61
Crystal data and structure reﬁnement for ReBz. 62

Atomic coordinates (x 104) and equivalent isotropic displacement

parameters (Azx 103) for ReBz. 63
Anisotropic displacement parameters (Azx 103) for ReBz. 63
Bond distances (A) for ReBz. 63
Crystal data and structure reﬁnement for HoMnB4. 64

Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (A2x 103) for HoMnB4. 65

ix

Table 3-15. Anisotropic displacement parameters (Azx 103) for HoMnB4. 65
Table 3-16. Crystal data and structure reﬁnement for REReB4 (RE= Dy, Er) 66
Table 3-17. Crystal data and structure reﬁnement for YbReB4. 67
Table 3-18. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for REReB4 (RE = Dy, Er,Yb). 68
Table 3-19. Anisotropic displacement parameters (Azx 103) for REReB4 (RE = Dy,
Er, Yb). 69

Table 3-20. Bond distances (A) for HoMnB4 and REReB4 (RE = Dy, Er, Yb). 70

Table 4-1. Crystal data and structure reﬁnement for R0733. 95
Table 4-2. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for RU7B3. 96
Table 4-3. Anisotropic displacement parameters (Azx 103) for Ru7B3. 96
Table 4-4. Selected bond distances (A) for Ru7B3. 97
Table 4-5. Crystal data and structure reﬁnement for Rung. 98
Table 4-6. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for RuzB3. 99
Table 4-7. Anisotropic displacement parameters (Azx 103) for Ru2B3. 99
Table 4-8. Selected bond distances (A) for Ru2B3. 99
Table 4-9. Crystal data and structure reﬁnement for RuBz. 100

Table 4-10. Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (Azx 103) for RuBz. 101
Table 4-11. Anisotropic displacement parameters (A2x 103) for RuBz. 101
Table 4-12. Selected bond distances (A) for RuBz. 101

Table 4-13.

Table 4-14.

Table 4-15.

Table 4-16.

Table 4-17.

Table 4-18.

Table 4-19.

Table 4-20.

Table 4-21.

Table 4-22.

Table 5-1.

Table 5-2.

Table 5-3.

Table 5-4.

Table 5-5.

Table 5-5.

Table 5-7.

Crystal data and structure reﬁnement for B-RuBz. 102

Atomic coordinates ( x 104) and equivalent isotropic displacement

parameters (Azx 103) for B-RuBz. 103
Anisotropic displacement parameters (Azx 103) for B-RuBz. 103
Selected bond distances (A) for B-RuBz. 103

Crystal data and structure reﬁnement for RERu4B4 (RE= Y, Ce). 104
Crystal data and structure reﬁnement for RERu4B4 (RE= Sm, Dy). 105
Crystal data and structure reﬁnement for YbRu484. 106
Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (A2x 103) for RERu484 (RE = Y, Ce, Sm, Dy, Yb). 107
Anisotropic displacement parameters (A2x 103) for RERu4B4 (RE = Y,
Ce, Sm, Dy, Yb). 108
Bond distances (A) for RERu4B4 (RE = Y, Ce, Sm, Dy, Yb). 109
Crystal data and structure reﬁnement for MnGa4_65 and ReGa4,5. 130
Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for MnGaMs and ReGa4_5. 131

Anisotropic displacement parameters (Azx 103) for MnGa4+x and

ReGa4+x. 132
Bond distances (A) for MnGa4,65 and ReGa4_5. 133
Crystal data and structure reﬁnement for Nio,57Ga4,43. 134

Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for Nio_57Ga4,43. 135

Anisotropic displacement parameters (Azx 103) for Nio,57Ga4,43. 135

xi

Table 58 Bond distances (A) for Nio,57Ga4_43. 135

xii

Figure 1-1.

Figure 2-1

Figure 2-2.

Figure 2-3.

Figure 2-4.

Figure 3-1.

Figure 3-2.

Figure 3-3.

Figure 3-4.

Figure 3-5.

LIST OF FIGURES

Various geometric arrangements of boron atoms in borides: (a) isolated
atoms (B:M < 0.5), (b) isolated pairs (B:M z 0.67), (c) zigzag chain
(B:M =1), ((1) double chain (B:M z 1.3), (e) tn'ple chain (B:M = 1.5), (f)
two-dimensional network (2.0 S B:M S 2.5). 6

(a) Representation of CrsB3 along the b-axis (b) Boron pairs in CrsB3 (Cr
atoms omitted for clarity). 39
Double and triple chains of boron atoms in (a) 0384 and (b) 0283,
respectively. 40
(a) CrBz structure seen along c-axis (b) Planar-nature of boron layers in
CrBz. 42
Structure of MB (M = M0, W) viewed down the b-axis 43
(a) Structure of CrB-type MnB with boron atom chains parallel to page.
The boron atom chains are perpendicular to the page in (b) FeB-type MnB
and (c) CrB-type MnB. 72
(a) The structure of MnB4 viewed down the a-axis (b) Sheet of RuBz
with similarly puckered hexagonal sheets of B atoms. 73
(a) Structure of ReBz and (b) an individual graphite-like layer of boron
atoms. 75
(a) Structure of YCrB4-type compounds (b) View down a-axis showing
alternating layers of metal atoms and boron atoms. 76

SEM micrograph of a MnB4 crystal. 79

xiii

Figure 3-6.
Figure 3-7.
Figure 3-8.
Figure 4-1.

Figure 4-2.

Figure 4-3.

Figure 4-4.

Figure 4-5.
Figure 4-6.
Figure 5-1.
Figure 5-2.
Figure 5-3.
Figure 5-4.

Figure 5-5.

Figure A-l.

Figure A-2.

SEM micrograph of a cluster of ReBz crystals. 79
SEM micrograph of HoMnB4. 81
SEM micrograph of DyReB4. 81
(a) Structure of Ru783 and (b) the stellaquadrangula building unit. 111
(a) Structure of Ru2B3 and the (b) puckered boron layer and (c) planar
layer of boron atoms. 112
Structure of RuBz crystallizing with the (a) RuBz-type and (b) ReBz-type
structures. 113

Structure of RERu4B4 (RE = Y, Ce, Sm, Dy, Yb) shown down

the b-axis. . 115
SEM micrographs of (a) RuBz and (b) B-RuBz. 117
SEM micrographs of (a) CeRu484 and (b) CeRuB4. 119
Structure of MGa4+x (M = Mn, Re) viewed down the c-axis. 137
View down the c-axis of (a) MnGa45 and (b) PdGas. 138
Connectivity along the c-axis in (a) MGa4+x and (b) PdGas. 139

(a) Structure of Nio.57Ga4,43 and (b) the parent y-brass structure. 141
SEM micrographs of (a) MnGa4 and (b) ReGa4,5. 143
Cross-sectional view of (a) BN susceptor holder and Mo susceptor along
(b) vertical and (0) horizontal axes. 153

General setup for RF furnace. 154

xiv

CCD

SEM

EDS

LIST OF ABBREVIATIONS

Charge Coupled Device
Scanning Electron Microscope
Energy Dispersive Spectroscopy

Radio Frequency

XV

CHAPTER ONE

The Metal Flux Technique in Boride Synthesis

A. Introduction

The metal ﬂux technique is an interesting alternative to traditional solid state
synthesis techniques. Low melting point metals are used as a solvent in this technique to
facilitate greater diffusion of the reagents and to aid in single crystal growth. The
application of the ﬂux technique to the synthesis of borides has had some success, but the
potential seems to be much greater. An examination of the metal flux technique, basic

boride structural features, and properties of note for borides is presented.

B. The Metal Flux Technique

Solid state synthesis oﬁen requires the use of high temperature equipment.
Traditionally, high-temperature furnaces, arc welders, and radio frequency (RF) induction
furnaces have been used to reach the elevated temperatures at which synthesis takes
place. While working at elevated temperatures may be required for any reaction to take
place, it can also be very limiting. A desired ternary phase may be inaccessible in a
system of reagents because more thermodynamically stable binary phases are formed.
Efforts to combat binary formation include frequent grinding and reheating of reagents,
using less-stable binary or ternary phase starting materials, and using very long annealing
steps in heating proﬁles.1 The metal ﬂux technique is a practical alternative to the

traditional solid state techniques.

The metal ﬂux technique makes use of a metal with a relatively low melting
point, in which the reagents of interest have some solubility, to act as a solvent for the
reaction. The melting point of the metal need only be low in relation to the other
reagents in use. For instance, copper, with a melting point of 1092 °C has been used
successfully as a ﬂux for many different systems containing metals with higher melting
points from the d- and f-blocks. The liquid metal ﬂux acts as a solvent and often allows
for greater diffusion of the reagents, which allows the reagents to react at temperatures
well below their melting points. The lower reaction temperature allows crystals to grow
with fewer defects, a boon when the crystals are going to be used in properties
measurements.2

Another advantage to using a metal ﬂux is that the ﬂuxes can getter impurities
from a reaction environment, allowing a clean environment for crystal growth.3 This
advantage is easily overlooked, but ensures a critical oxygen-free environment.
Generally ﬂux reactions are run at lower temperatures, so the reactivity of the reagents
with oxygen is already less than that found in traditional techniques. Because most ﬂux
reactions have a large excess of the ﬂux metal, oxygen, or any other impurity, is much
more likely to react with the ﬂux metal than the intended reagents.

Disadvantages do exist in the use of metal ﬂuxes. At times, ﬁnding an
appropriate metal in which the reagents are soluble can be difﬁcult or impossible.
Because the ﬂux is soluble with so many materials there can be unexpected incorporation
of reaction vessel materials into the reactions. We have found that tantalum and quartz
tubes react too readily with gallium and copper and are destroyed by these ﬂuxes. Yet,

the most challenging part of using the ﬂux may be in how to extract the reaction products

from the solvent-like ﬂux. These problems must be taken into consideration when
attempting a new molten metal ﬂux synthesis.

Our group has implemented the metal ﬂux in exploratory synthesis for many
multinary systems. Many new alumindesf’ 5 silicidesf’ 7 and germanidess’ 9 were found,
which led to interest in extending the technique to systems including boron. Initial
successes included the synthesis of a new binary phase, [3-SiB3,lo and a series of
quaternary phases, REL3CZSig(B12)3 (RE= Y, Tb, Er).ll Further inspection of the literature
showed that metal ﬂuxes have been used successfully for the synthesis and growth of
quite a few borides.”23

The ﬁrst reports of growth of borides from molten metals dealt with mixtures of
iron and aluminum as the ﬂux metals.12 Iron melts near 1500 °C, but aluminum melts at
only 660 °C, and boils near 2500 °C, which allows researchers a very large, and easily
attainable, temperature range in which to carry out experiments. Using aluminum as the
ﬂux metal also allows the use of more accessible oxide starting reagents because the
aluminum can reduce the oxides to elemental reagents while in the ﬂux.13 The advantage
of using these oxides is that the preparation for synthesis can be done out on a bench top,

rather than inside a glovebox. These advantages made aluminum the ﬂux metal of choice

for borides. It has produced crystals of many different borides, ranging from simple

14-17 18

binary borides to temaries that include aluminum.
One of the main problems with using aluminum as the ﬂux metal is that it makes
stable binaries, not only with many metals, but also with boron. Aluminum diboride is a

common side product in ﬂux reactions, and we have found at least three different

aluminum dodecaborides that form from reactions rich in aluminum and boron. There

have been no reports of true ternary borides grown from an aluminum ﬂux that do not
include aluminum in the structure. To make temaries that do not incorporate aluminum,
other metals must be used as the ﬂux.

Copper has been used with some success in the ﬂux growth of binary crystals in
the Nb—B19 and W—B20 systems. In the last decade the use of a copper ﬂux was extended
to ternary RE—Rh—BZland RE—B—C22 systems and quaternary RE—Rh—B—C23 and RE—B—
C—sz systems. An advantage to using copper over aluminum is that there are no known
Cu—B binaries to inhibit other borides from forming. Disadvantages to using copper
include its high melting temperature (1092 °C) and the inability to use oxides as reagents.

Our group has found gallium to be a useful ﬂux metal in the synthesis of RE—B—
Si—C and Si—B compounds. After having difﬁculty isolating aluminum-free ternary
silicides from an aluminum ﬂux, our group moved to gallium and found that gallium
stayed out of the ternary phases. Seeing that aluminum was also incorporated into many
borides, we again moved to gallium thinking results similar to those found with silicides
could follow. A new binary phase, [3-SiB3,10 and series of quaternary phases,
RE|_3C2SIs(Blz)3 (RE= Y, Tb, Er),ll showed the thinking to be correct and provided a
basis for a closer examination into the use of a gallium ﬂux for the synthesis of multinary
borides.

Gallium has several characteristics that make it an interesting ﬂux metal. As in
aluminum, there is a very large difference between the melting (29 °C) and boiling points
(2400 °C) of gallium, so there is seemingly a large range for reaction temperatures.
However, boron has a very low solubility in gallium below about lOOO°C. Thus,

potential components in multinary borides can form stable non-boride phases that crash

out of the reaction mixture before boron has even had a chance to react. By working at
1000 °C and above we looked to determine the applicability of the gallium ﬂux in the
synthesis of a wider range of borides.

If successful, this synthesis technique could prove its usefulness by providing
larger, better-formed crystals than those previously made, allowing for better property
analyses. In addition, the unique chemistry of boron, which is responsible for the great

diversity of known borides, could be extended to previously unknown phases.

C. Structural Features in Borides

Borides are a large and diverse class of compounds. This diversity is readily seen
in the geometric arrangement of the boron atoms in borides, shown in Figure 1-1, which
varies based on the ratio of boron to metal atoms found in the compound.” 25 In metal-
rich borides, in which the boron-to-metal ratio (B:M) is less than 0.5, the boron atoms are
isolated from each other, as shown in Figure l-la, and only metal-metal or metal-boron
bonds are formed. When more boron is incorporated into the structure, increasing B:M to
approximately 0.67, boron-boron bonds begin to form, resulting in isolated boron pairs ,
as shown in Figure 1-1b. A further increase in B:M to 1.0 allows the isolated boron pairs
to link together and form the zigzag boron atom chain shown in Figure 1-1c. These
chains will link together as more boron is added, resulting in double chains (Figure 1-1d)
for B:M near 1.3, and triple chains (Figure 1-1e) for B:M around 1.5. Two-dimensional
networks are formed when B:M increases to between 2.0 and 2.5. These networks most

commonly comprise six-membered rings of boron atoms, as shown in Figure l-lf,

(a) (b)

 

(0

Figure 1-1. Various geometric arrangements of boron atoms in borides: (a) isolated
atoms (B:M < 0.5), (b) isolated pairs (B:M z 0.67), (c) zigzag chain (B:M =1), ((1) double
chain [3:954 2'- 1.3), (e) triple chain (B:M = 1.5), (f) two-dimensional network (2.0 S B:M
S 2.5). ‘

however, rings with ﬁve and seven boron atoms are also well known.”28 Boron-rich
compounds, with B:M values greater than 2.5, contain three-dimensional networks of
boron atoms. The most common polyhedra formed by the boron atoms are octahedra,
icosahedra, and cubo—octahedra.

The rich structural variety of borides is even more impressive when compared to
that found in carbides and nitrides. Both carbides and nitrides have limited structural
variety due in part to geometric factors.29 One such geometric factor is based on Hagg’s
rule, which states that interstitial compounds with a ratio of nonmetal to metal radii less
than 0.59 will adopt a simple structure, such as body-centered cubic, face-centered cubic,
hexagonal close-packed, or hexagonal. Because carbon and nitrogen atoms have rather
small radii compared to most metals, they can occupy the octahedral holes between metal
atoms, creating simple structures, as predicted by Hagg. Boron atoms tend to be larger,
exceeding Hagg’s 0.59-ratio rule, and prefer a trigonal prismatic coordination
environment. The trigonal prismatic coordination environment and larger atom sizes
bring boron atoms within bonding distance to each other, leading to the pairs, chains, etc.

that are present throughout boride structural chemistry.24

D. Chemical and Physical Properties of Borides

Accompanying the structural diversity in borides is an equally wide diversity of
chemical and physical properties. Many borides have a great deal of chemical stability.
Nowhere is this stability more evident than when borides are separated from a molten
metal flux. To remove products from a copper ﬂux, for instance, the ingot containing the

now-solid ﬂux and products is placed into an etching solution, a bath of nitric acid.

Regardless of concentration, the acid has little effect on the boride crystals, but
completely removes the copper. Similar results are seen in aluminum ﬂux reactions,
where the etching solution can be either hydrochloric acid or sodium hydroxide. Again,
minimal damage is done to the boride crystals while completely removing the metal. In
cases where the boride crystals must be dissolved, warm or boiling aqua regia is used
because the dissolution is often too slow at room temperature, leaving the boride crystals
intact and unharmed.10

The strong covalent bonding in borides leads to great thermal stability. This
stability allows borides to be used in many high temperature applications, such as
thermoelectric conversion materials.”32 Promising results have been found for boron-
rich borides, and an entire thermoelectric device was recently made consisting entirely of
boron-rich components.33 The current results on boride thermoelectric components are
not optimized yet, so future work in this area looks to maximize effectiveness of these
devices in harsh environments where other systems fail.

Borides have consistently found use in harsh industrial applications because they
are hard and incompressible materials. The hardness of the compounds comes from the
directional covalent bonds in borides, while the incompressibility is due to electrostatic
repulsion between atoms and can be correlated with the valence electron density of the
compounds.34 The advantage of borides over more traditional industrial products is made
apparent in their application in the cutting of iron-based metals. Boride coated tools are
used instead of diamonds, which have superior hardness, but also react chemically with
the iron in the metal.35 The added chemical and thermal stability of borides give them

potential for many industrial applications.

Superconductivity is not exactly a rare property for a boride to exhibit. However,
when the critical transition temperature (Tc) associated with that superconductivity is
39 K, as in magnesium diboride,36 it is something of which to take note. Magnesium
diboride had been prepared and its structure known for nearly ﬁﬁy years before it was
tested for superconductivity. The T0 of 39 K is the highest known for a bulk sample
outside of the copper oxides, and led to a ﬂurry of research in borides in recent years.
Our group is included in that ﬂurry and we hope that the molten metal ﬂux technique can

help isolate a boride with an even high Tc.

References

1. Prokhorov, A. M.; Lyakishev, N. P.; Burkhanov, G. S.; Dement'ev, V. A., High Purity
Transition-Metal Borides: Promising Materials for Present Day Technology. Inorganic
Materials 1996, 32, (11), 1195-1201.

2. Kanatzidis, M. G.; Pottgen, R.; Jeitschko, W., The Metal Flux- A Preparative Tool for
Intermetallic Compounds. Angewandte Chemie International Edition 2005, 44, (43),
6996-7023.

3. Fisk, Z.; Remeika, J. P., Growth of Single Crystals from Molten Metal Fluxes. In
Handbook on the Physics and Chemistry of the Rare Earths; K.A. Gschneidner, J .;
Eyring, L.,Eds. Elsevier Science: 1989; Vol. 12, pp 53-70.

4. Latturner, S.; Kanatzidis, M. G., Formation of Multinary Intermetallics from
Reduction of Perovskites by Aluminum Flux--M3Au6+xA126Ti (M = Ca, Sr, Yb), a Stuffed
Variant of the BaHgn Type. Inorganic Chemistry 2004, 43, 2-4.

5. Latturner, S. E.; Bilc, D.; Ireland, J. R.; Kannewurf, C. R.; S.D.Mahanti; Kanatzidis,
M. G., REAu3A17 (RE = rare earth): New Ternary Alumindes Grown from Aluminum
Flux. Journal of Solid State Chemistry 2003, 170, (1 ), 48-57.

6. Latturner, S.; Kanatzidis, M. G., REAu4A13Si: The End Member of a New
Homologous Series of Intermetallics Featuring Thick AuA12 layers. Chemical
Communications 2003, (18), 2340-2341.

7. Salvador, J. R.; Malliakas, C.; Gour, J. R.; Kanatzidis, M. G., RE5C04Si14 (RE= Ho,
Er, Tm, Yb): Silicides Grown from Ga Flux Showing Exceptional Resistance to
Chemical and Thermal Attack. Chemistry of Materials 2005, 17, 1636-1645.

8. Wu, X.; Bilc, D.; Mahanti, S. D.; Kanatzidis, M. G., VzAlsGes: First Ternary
Intermetallic in the V-Al-Ge System Accessible in Liquid Aluminum. Chemical
Communications 2004, 1506-1507.

9. Wu, X.; Kanatzidis, M. G., REAuAl4Ge2 and REAuAl4(AuxGe1-x)2 (RE= rare earth
element): Quaternary Intermetallics Grown in Liquid Aluminum. Journal of Solid State
Chemistry 2005, 178, 3233-3242.

10. Salvador, J. R.; Bilc, D.; S.D.Mahanti; Kanatzidis, M. G., Stabilization of B-SiB3
from Liquid Ga: A Boron-Rich Binary Semiconductor Resistant to High-Temperature Air
Oxidation. Angewandte Chemie International Edition 2003, 42, 1929-1932.

11. Salvador, J. R.; Bilc, D.; S.D.Mahanti; Kanatzidis, M. G., Gallium Flux Synthesis of
Tb3-xC2Si3(B|2)3: A Novel Quaternary Boron-Rich Phase Containing B12 Icosahedra.
Angewandte Chemie International Edition 2002, 41, (5), 844-846.

10

12. Bernard, A. US. Patent 3096149, 1963.

13. Okada, S.; Tanaka, T.; Leithe-Jasper, A.; Michiue, Y.; Gurin, V. N., Crystal Growth
and Structure Analysis of a New Scandium Aluminum Boride SCZA1B6. Journal of Solid
State Chemistry 2000, 154, 49-53.

14. Okada, S.; Atoda, T.; Higashi, 1., Structural Investigation of 028;, 033,, and CrB
by Single-Crystal Diffractometry. Journal of Solid State Chemistry 1987, 68, 61-67.

15. Okada, S.; Atoda, T.; Higashi, 1.; Takahashi, Y., Preparation of Single Crystals of
M08; by the Aluminum-ﬂux Technique and Some of Their Properties. Journal of
Materials Science 1987, 22, 2993-2999.

16. Okada, S.; Hamano, K.; Higashi, 1.; Lundstrom, T.; Tergenius, L.-E., Preparation of
Single Crystals of a New Compound, Ta586 by the Aluminum Flux Method. Bulletin of
the Chemical Society of Japan 1990, 63, 687-691.

17. Higashi, I.; Takahashi, Y.; Atoda, T., Crystal Growth of Borides and Carbides of
Transition Metals from Molten Aluminum Solutions. Journal of Crystal Growth 1976,
33, 207-211.

18. Becher, H. J .; Krogmann, K.; Peisker, E., Uber das ternare Borid anAle.
Zeitschriftfu’r anorganische und allgemeine Chemie 2004, 344, (3-4), 140-147.

19. Okada, S.; Hamano, K.; Lundstrom, T.; Higashi, I. In Crystal Growth of the New
Compound Nb 23 3, and the Borides NbB, Nb 586, Nb 334, and Nsz, Using the Copper-Flux
Method, Boron-Rich Solids, Albuquerque, New Mexico, 1986; Emin, D.; Aselagc, T.;
Beckel, C. L.; Howard, 1. A.; Wood, C., Eds. American Institute of Physics:
Albuquerque, New Mexico, 1985; pp. 456-459.

20. Okada, S.; Kudou, K.; Lundstrom, T., Preparations and Some Properties of W28, 8-
WB, and WB; Crystals from High-Temperature Metal Solutions. Japanese Journal of
Applied Physics 1995, 34, (1), 226-231.

21. Shishido, T.; Ye, J .; Sasaki, T.; Note, R.; Obara, K.; Takahashi, T.; Matsumoto, T.;
F ukuda, T., Growth of Single Crystals in the Systems with R-Rh-B and R-Rh-B-C (R=
Rare Earth Element) from Molten Copper Flux. Journal of Solid State Chemistry 1997,
133, 82-87.

22. Zhang, F. X.; Tanaka, T., Single Crystal Growth of Some Rare-Earth Boron-Rich
Compounds in RE-B-C(N) and RE-B-Si Systems. Journal of Crystal Growth 2004, 271, p
159-164.

23. Ye, J .; Shishido, T.; Kimura, T.; Matsumoto, T.; F ukuda, T., End Member of the
Rhodium-Based Quaternary Borocarbides, ErhaBzC. Acta Crystallographica 1998,
C54, 1211-1214.

11

24. Etoumeau, J .; Hagenmuller, P., Structure and Physical Features of the Rare-Earth
Borides. Philosphical Magazine B 1985, 52, (3), 589-610.

25. Lundstrom, T., Structure, Defects and Properties of Some Refractory Borides. Pure
and Applied Chemistry 1985, 57, (10), 1383-1390.

26. Rogl, P., New Ternary Borides with YCrB4-type Structure. Materials Research
Bulletin 1978, 13, 519-523.

27. Kuz'ma., Y. B., Crystal Structure of the compound YCrB4 and its analogs.
Kristallograﬁya 1970, 15, 372-374.

28. Sobczak, R.; Rogl, P., Magnetic Behavior of New Ternary Metal Borides with
YCrB4-type Structure. Journal of Solid State Chemistry 1979, 27, 343-348.

29. Chen, J. G., Carbide and Nitride Overlayers on Early Transition Metal Surfaces:
Preparation, Characterization, and Reactivities. Chemical Reviews 1996, 96, 1447-1498.

30. Mori, T., High Temperature Thermoelectric properties of Ba Icosahedral Cluster-
containing Rare Earth Boride Crystals. Journal of Applied Physics 2005, 97, 093703.

31. Nakayama, T.; Shimizu, J .; Kimura, K., Thermoelectric Properties of Metal-Doped
B-Rhombohedral Boron. Journal of Solid State Chemistry 2000, 154, 13-19.

32. Imai, Y.; Mukaida, M.; Ueda, M.; Watanabe, A., Screening of the Possible Boron-
based n-type thermoelectric conversion materials on the basis of the calculated densities
of states of metal borides and doped B-boron. Intermetallics 2001, 9, 721-734.

33. Takeda, M.; Kurita, Y.; Yokoyama, K.; Miura, T. In Synthesis and High
Temperature Thermoelectric Properties of Alkaline-Earth Metal Hexaborides MB6
(M=Ca,Sr, Ba), 2003 MRS Fall Meeting, Boston, MA.

34. Cumberland, R. W.; Weinberger, M. B.; Gilman, J. J.; Clark, S. M.; Tolbert, S. H.;
Kaner, R. B., Osmium Diboride, An Ultra-Incompressible, Hard Material. Journal of the
American Chemical Society 2005, 127, 7264-7265.

35. Gao, F.; Hou, L.; He, Y., Origin of Superhardness in Icosahedral B12 Materials.
Journal of Physical Chemistry B 2004, 108, 13069-13073.

36. Nagamatsu, J .; Nakagawa, N.; Muranaka, T.; Zenitani, Y.; Akimitsu, J .,
Superconductivity at 39K in Magnesium Diboride. Nature 2001, 410, 63-64.

12

CHAPTER TWO

The Metal Flux Synthesis of Group Six Borides

A. Introduction

The group 6 elements form a very interesting series of binary borides. Chromium
borides seem to vary in structure based on B:M ratio, as outlined earlier in this work. The
boron atoms in these borides go from isolated atoms to isolated pairs, pairs to chains,
chains to double chains, double chains to triple chains, triple chains to 2-D networks, and
2—D to 3-D networks with increasing B:M ratios. Molybdenum and tungsten borides
follow most of the structural patterns according to B:M ratios. However, no molybdenum
or tungsten borides with B:M ratios of 1.5 or 1.67 have been isolated. Interestingly, there
are more borides with B:M 2 2.0 for molybdenum and tungsten than there are for
chromium.

Slight variations in the B:M ratios above 2.0 lead to various 2-D and 3-D boron
networks for the molybdenum and tungsten borides. MoB2.,,,l Mongg,2 Mol.,,B3,3 WBz,4
and W2B52 all have layers with planar hexagonal boron rings. MozBs and W2B5 also have
layers with puckered hexagonal boron rings. The structures of MoBz.x , WBz, MoB4, and
WB., all have a common feature: planar layers of boron hexagons. However, in the
tetraborides, these layers are linked to form a 3-D network. Pairs of boron atoms replace
one-third of the molybdenum or tungsten atoms in the metal atom layers of the diboride

structures, linking the boron atoms of the tetraboride structure into a 3-D networks

13

Further interest into group 6 borides came from previous results with the molten
metal ﬂux technique. A Japanese group of researchers has used aluminum as the ﬂux
metal for the synthesis of borides.“9 Initially, W2B5 was the only binary boride to grow
from the ﬂux reactions.6 Ternary phases incorporating aluminum were found in both the
chromium and molybdenum systems, but no binary group 6 borides were found. Over a
decade later, the same group reported that they could grow binary crystals in both the
chromium—boron7 and molybdenum—boron8 systems by introducing a slow-cooling step
into the heating profile following the soak at maximum temperature. Nearly twenty years
after the initial research into the tungsten-boron system, two other W—B binaries, W2B
and WB, were grown in molten aluminum.9 Also in the new article was the report of
copper ﬂux synthesis of W2B and WB in single-crystalline form.

Armed with the knowledge that both aluminum and copper had worked as ﬂuxes
for group 6 boride syntheses, we wanted to see if we could extend the technique to
include gallium as a ﬂux metal. Extending the use of a copper ﬂux to chromium and
molybdenum borides was also of interest because of its success in forming tungsten
borides. No ternary group 6 borides synthesized from a molten metal flux have been
reported (excluding the aluminum-containing borides). Thus, the chance to apply the
ﬂux technique to synthesize ternary compounds was also an exciting possibility.

Our initial gallium flux reactions targeted multinary group 6 borides, such as
YbCr(Si,B)4, YM03B2, and NinBls. Despite many attempts targeting various ternary
and quaternary borides, the only borides formed where binary phases. Copper ﬂux
reactions had similar results. While not giving up on our goal of synthesizing ternary

borides from a metal ﬂux, we did shift focus slightly, and started targeting binary borides

l4

that had formed out of the multinary systems. In this chapter we describe the synthesis
and structural characterization of binary borides synthesized using gallium and/or copper

ﬂuxes.

B. Experimental

Reagents:

All reagents were used as received without further puriﬁcation: (i) Boron metal,
99% purity, -325 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (ii) Chromium
metal, 99.95% purity, -200 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (iii)
Molybdenum metal, 99.9% purity, Cerac Specialty Inorganics, Milwaukee, WI, (iv)
Tungsten metal, 99.9% purity, Cerac Specialty Inorganics, Milwaukee, WI, (v) Gallium
3-5 mm shot, 99.999% purity, Plasmaterials, Liverrnore, CA, (vi) Copper metal cuttings,
Mallinckrodt Specialty Chemicals Co., Paris, KY, (vii) Dysprosium metal (filed from
ingot) 99.9% purity, Chinese Rare Earth Information Center, Inner Mongolia, China,
(viii) Yttrium metal, 99.9% purity, -40 mesh, Cerac Specialty Inorganics, Milwaukee, WI,
(ix) Nickel metal powder, 98-99% purity, -325 mesh, E.H. Sargent & Company.

Fumaces:

Commercial tube furnaces from Applied Test Systems, Inc. were used to heat
reactions sealed in quartz tubes. Furnace controllers from the Omega Company were
used to program heating profiles for individual reactions. Furnace temperatures were
monitored during reactions by thermocouples connected to the Omega controllers. To
extend the life of the furnaces, the operating temperature for prolonged heating was

limited to 1000 °C.

15

A commercial, hi gh-temperature, vertical tube furnace, with controller, from the
Mellen Company was used for copper ﬂux reactions. Heating proﬁles for individual
reactions were programmed into the furnace controller. Thermocouples connected to the
furnace controller were used to monitor furnace temperatures during reactions. To extend
the life of the furnace, the operating temperature for prolonged heating was limited to
1350 °C.

A Lindberg Blue hi gh-temperature, horizontal tube furnace was used for gallium
and copper ﬂux reactions. Heating proﬁles for individual reactions were programmed
into the accompanying Lindberg UPISO furnace controller. Thermocouples connected to
the furnace controller were used to monitor furnace temperatures during reactions. The
maximum operating temperature for the furnace was 1500 °C.

A radio frequency (RF) furnace was constructed from an Ameritherm XP7.5CE
Induction Heater and fused quartz reaction chamber. A molybdenum cylinder, machined
to act as a reaction crucible holder, was used as a susceptor. A boron nitride cylinder was
machined to hold the molybdenum crucible in the center of the reaction chamber. Details
on the RF furnace setup are given in the appendix of this work.

RF furnace temperatures were controlled manually using the Ameritherm control
unit. Reaction temperatures were monitored using an Ircon SA Series Infrared
Thermometer. During the course of some reactions, ﬁlms deposited inside the reaction
chamber would prevent the IR thermometer from accurately reporting furnace
temperatures. In such circumstances, reaction temperatures were approximated based on

power settings for the Ameritherm control unit and a series of calibration runs conducted

16

by heating only the molybdenum cylinder. Temperatures greater than 1700 0C were
easily reached with this furnace.
Synthesis and Isolation:

Cr5B3: Boron (2 mmol) and chromium (2 mmol) powders were mixed and added
to an alumina crucible, along with dysprosium ﬁlings (1 mmol). Pieces of copper totaling
2.54 g (40 mmol) were then placed in the crucible. After the addition of a loose-ﬁtting
alumina lid, the crucible was placed in the RF furnace. The RF furnace was evacuated
under vacuum to a pressure near 10'3 Torr and ﬂushed with nitrogen. Heating was
initiated with the system under vacuum. The temperature was raised to 1400 °C in
approximately 25 minutes and then held at that temperature for 2 hours, at which time the
power to the furnace was shut off and the furnace was allowed to cool to room
temperature.

After cooling, the crucible was removed from the furnace and placed into a bath
of half-concentrated I'INO3. The etching solution entered the crucible through the loose-
ﬁtting lid. An overnight soak was used to completely remove the copper ﬂux, however,
many product crystals were removed from the ﬂux after a 3- to 4-hour soak. Extended
soaking in the bath was not found to damage the crystals. Silver, plate-like crystals were
isolated in an estimated 20-30% percent yield based on chromium. Elemental analysis
performed by EDS showed that chromium and boron were the only elements present in
the plate-like crystals. The remainder of the collected product comprised elemental
chromium and traces of DyB6.

Cr3B4: Boron (4 mmol) and chromium (1 mmol) powders were mixed and added

to an alumina crucible. Pieces of copper totaling 2.0 g were then placed in the crucible.

l7

The crucible was suspended in a high-temperature tube furnace and heated under a slow
ﬂow of argon. The furnace was heated to 1300 °C in 13 hours and held at that
temperature for 10 hours, at which point the power to the furnace was shut off and the
furnace was allowed to cool to room temperature.

After cooling, the crucible was removed from the furnace and placed into a bath
of half-concentrated I'INO3 in order to remove the copper ﬂux. An overnight soak was
used to completely remove the copper ﬂux, however, many crystals were removed from
the ﬂux after a 3- to 4-hour soak. Extended soaking in the bath was not found to damage
the crystals. Silver, plate-like crystals were isolated in approximately 20 to 40% yields
based on chromium. The remainder of the collected product comprised Cr2B3, CrBz, and
elemental chromium.

Cr2B3: Boron (2 mmol) and chromium (1 mmol) powders were mixed and added
to an alumina crucible. Pieces of gallium shot totaling 1.6 g were then placed in the
crucible. The alumina crucible was then placed in 13 mm diameter quartz tube,
evacuated under vacuum to a pressure near 104 Torr, and then sealed using an oxygen-
acetylene torch. The quartz tube was then heated in a commercial tube furnace under a
heating proﬁle that increased the furnace temperature to 1000 °C in 10 hours, held the
temperature constant for 96 hours, and then allowed the furnace to cool to room
temperature.

After cooling, the quartz tube was opened and the alumina crucible was placed in
a half-concentrated HCl bath in order to remove the excess gallium. Most of the gallium
was removed after an overnight soak, but longer soaks can be used without damage to the

recovered product. Silver needle-shaped crystals were isolated in approximately 40 to

18

60% yields based on chromium. The remainder of the product comprised CrBz and
elemental chromium.

CrBz: Boron (10 mmol) and chromium (1 mmol) powders were mixed and added
to a glassy carbon crucible, along with yttrium (1 mmol). Pieces of gallium shot totaling
2.0 g were then placed in the crucible. The crucible was placed in the RF furnace,
evacuated under vacuum to a pressure near 10'3 Torr, and ﬂushed with argon. Heating
was initiated under vacuum and the temperature was raised to 1300 °C in ten minutes.
After holding at that temperature for 45 minutes, the power to the furnace was shutoff,
and the furnace was allowed to cool to room temperature.

After cooling, the crucible was removed from the furnace and placed into a bath
of 3M 12 in DMF in order to remove the gallium ﬂux. An overnight soak was found to be
insufﬁcient for removing all of the gallium, so soaking was continued for an additional 3
days. Extended soaking in the bath was not found to damage the crystals. Silver,
hexagonal plate-like crystals were isolated in an estimated 30 to 40% yield based on
chromium. Elemental analysis performed by EDS showed that chromium and boron were
the only elements present in these crystals.

MoB: Boron (3 mmol), molybdenum (1 mmol), and yttrium (0.3 mmol) were
added to a graphite crucible. Pieces of gallium shot totaling 1.6 g were then added, along
with a loose-ﬁtting graphite cap, and the crucible was placed in the RF furnace. The RF
furnace was evacuated under vacuum to a pressure near 10'3 Torr and flushed with argon.
Heating was initiated with the system under vacuum. The temperature was raised to

1300 0C in approximately 15 minutes and then held at that temperature for 15 minutes, at

19

which time insufﬁcient cooling water ﬂow shut off the power to the furnace. The furnace
then cooled to room temperature.

After cooling, the crucible was placed in a bath of 3 M12 in DMF in order to
remove the remaining gallium ﬂux. An overnight soak was found to be insufﬁcient for
removing all of the gallium, so soaking continued for an additional 2 days. Extended
soaking in the bath was not found to damage the crystals. Silver, plate-like crystals were
a minor product (estimated 10 to 20% yield based on molybdenum). Purple powder and
crystals of YB6 were present in 20-30% yields based on yttrium. The remainder of the
product collected comprised unreacted boron powder and traces of YB6.

WB: Boron (7.5 mmol), nickel (0.5 mmol), and tungsten (1 mmol) powders were
mixed and added to an alumina crucible along with pieces of gallium shot totaling 1.45 g.
A quartz ﬁlter was placed in the crucible which was then placed in 13mm diameter quartz
tube, evacuated under vacuum to a pressure near 104 Torr, and sealed using an oxygen-
acetylene torch. The tube was then heated in a commercial tube furnace under a heating
proﬁle that increased the furnace temperature to 1000 °C in 10 hours, held the
temperature constant for 48 hours, allowed the furnace to cool to 850 °C in 5 hours, held
the temperature constant for an additional 48 hours, and then allowed the furnace to cool
to 250 0C in 60 hours.

The quartz tube was removed from the furnace at 250 °C and centrifuged to
remove excess gallium. After centrifugation, the tube was opened and the crucible was
placed in a bath of 5 M 12in DMF in order to remove the remaining gallium ﬂux. An
overnight soak was found to be insufﬁcient for removing all of the gallium, so soaking

continued for an additional 2 days. Extended soaking in the bath was not found to

20

damage the crystals. Silver, prismatic crystals were a minor product (estimated 10-20%
yield based on tungsten). Hexagonal plates of Ni2Ga3 were the major product, while
rectangular prismatic crystals of NiGa4 were also a minor product (IO-20% yield based on
nickel). The remainder of the product collected was unreacted boron powder and traces

of a tungsten-gallium phase.

C. Physical Measurements

Energy Dispersive Spectroscopy: Energy dispersive spectroscopy (EDS)
analyses were carried out on selected crystals to determine their chemical composition.
The analyses were performed with a JEOL J SM-6400 scanning electron microscope
(SEM) equipped with Noran Vantage Energy Dispersive Spectroscopy (SizLi) detector
and Norvar window for standardless quantization of elements with Z Z 4. The crystals
were afﬁxed to an alumina sample holder with double-sided carbon tape. The EDS data
were acquired at an accelerating voltage of 10kV with a 30 to 60 second accumulation
time.

Single-Crystal X-ray Crystallography: Single crystals were mounted on a glass
ﬁber with super glue and their intensity data were collected on either a Bruker SMART
platform CCD diffractometer, or a STOE IPDS H diffractometer using Mo K0: radiation
at 50 kV and 40 mA.

Individual frames of Cr3B4 and WB examined with the Bruker machine were
collected with a 03° (1) rotation. The SMART software was used for data collection, and

SAINTlo software was used for data extraction and reduction. After applying analytical

21

absorption corrections, structure solution and reﬁnement were completed using direct
methods and the SHELXTLll suite of programs.

Samples examined with the STOE (Cr5B3, Cr2B3, CrB2, and MoB) had individual
frames collected on a 34 cm image plate with a 60 second exposure time and a 10° a)
rotation. The X-SHAPE'2 and X-RED-32l3 software packages were used for data
extraction and reduction and to apply an analytical absorption correction. Direct methods
and the SHELXTL suite were used to solve and reﬁne the structures.

The crystal structure reﬁnement data for 0ng are listed in Table 2-1. Atomic
parameters found in the previous structure report'4 were used as a starting point for
structural reﬁnement. The atomic positions and isotropic displacement parameters are
listed in Table 2-2. Anisotropic displacement parameters for the chromium and boron
atoms are found in Table 2-3. The bond distances, listed in Table 2-4, match well with
those calculated from previous reports.

The crystal structure reﬁnement data for 033, are listed in Table 2-5. Atomic
parameters found in previous structure reports7'15 were used as a starting point for
structural reﬁnement. The atomic positions and isotropic displacement parameters are
listed in Table 2-6. Anisotropic displacement parameters for the chromium and boron
atoms are found in Table 2-7. The bond distances, listed in Table 2-8, match well with
those found in previous reports.

The crystal structure reﬁnement data for Cr2B3 are listed in Table 2-9. The atomic
positions and isotropic displacement parameters are listed in Table 2-10. Anisotropic

displacement parameters for the chromium and boron atoms are found in Table 2-11.

22

The bond distances, listed in Table 2—12, match well with those calculated from the
previous report on the structure.7

The crystal structure reﬁnement data for CrB2 are listed in Table 2-13. The atomic
positions and isotropic displacement parameters are listed in Table 2-14. Anisotropic
displacement parameters for the chromium and boron atoms are found in Table 2-15. The
bond distances are listed in Table 2-16.

The crystal structure reﬁnement data for MoB and WB are listed in Table 2-17.
Atomic parameters found for MoB were used as a starting point for structural reﬁnement
of WB. The atomic positions and isotropic displacement parameters for both compounds
are listed in Table 2-18. Anisotropic displacement parameters for the all of the atoms are
found in Table 2-19. The structure of WB needs further reﬁnement to pinpoint the
location of the boron atom. The anisotropic displacement parameters for this atom are
quite large and would be reduced with a more accurate atom location. Additional work
with absorption correction for the crystal can also help reduce the anisotropic values. The
bond distances for MoB and WB, listed in Table 2-20 and Table 2-21, respectively, match

well with those calculated from the previous report.2

23

Table 2-1. Crystal data and structure reﬁnement for Cr5B3.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 36.70°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on 13'2

Final R indices [I>2sigma(1)]
R indices (all data)

Extinction coefﬁcient

292.43

293(2) K

0.71073 A

Tetragonal

I4/mcm

a = 5.4334(8) A a: 90°
b = 5.4334(8) A B= 90°
c = 10.031(2) A y = 90°
296.14(9) A3

4

6.559 g/cm3

17.451 mm‘1

540

0.098 x 0.080 x 0.035 mm3

4.06 to 36.70°
-9Sh_<_9,-8_<_k58,-1651516
2235

220 [R(int) = 0.0994]

99.1 %

Full-matrix least-squares on F2
220 / 0/ 16

1.098

R1: 0.0618, wR2 = 0.1592

R1 = 0.0698, wR2 = 0.1661
0.026(7)

 

R1 = 2 ”Fa! - IFcII / 2 IE)! and wR2 = {Z (IF.2 - FRI )2 / z (wF.2)21”2

24

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

parameters (Azx 103) for Cr5B3.

 

 

x y 2 U(GQ)
Cr(l) 3285(1) 1715(1) 1459(1) 7(1)
Cr(2) 0 O 0 7(1)
B(1) 0 0 2500 8(2)
B(2) -1161(12) -3839(12) 0 4(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

Table 2-3. Anisotropic displacement parameters (Azx 103) for Cr5B3.

 

 

ull U22 u33 u23 [113 ul2
Cr(l) 7(1) 7(1) 6(1) 0(1) 0(1) 0(1)
Cr(2) 6(1) 6(1) 7(1) 0 0 0
B(l) 8(3) 8(3) 7(4) 0 0 0
B(2) 5(2) 5(2) 3(3) 0 0 -3(3)

 

The anisotropic displacement factor exponent takes the form: -21t2[ h2 3*2U11 +

+2hka*b*U12]

25

Table 2-4. Bond distances (A) for Cr5B3.
Crl B2 2x 2.1621(3)
Crl B2 1x 2.1916(2)
Crl B1 2x 2.2684(2)
Crl CR1 1x 2.4127(4)
Crl Cr2 2x 2.4892(3)
Cr2 B2 4x 2.1786(3)
Cr2 Crl 8x 2.4892(3)
B l Crl 8x 2.2684(2)
82 B2 1x 1.7865(2)
B2 Crl 4x 2.1621(3)
B2 Cr2 2x 2.1786(3)

B2 Crl 2x 2.1916(2)

26

Table 2-5. Crystal data and structure reﬁnement for Cr3B4.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 29.05°
Reﬁnement method

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

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

Extinction coefﬁcient

199.24

293(2) K

0.71073 A

Orthorhombic

Immm

a = 2.9550(10)A a: 900
b = 2.9823(10) A 8: 90°
c = 13.030(4) A y = 90°
114.83(7) A3

2

5.762 g/cm3

13.527 mm‘1

184

3.14 to 28.21°
-3ShS3,-3Sks3,-l6£ls 16
689

102 [R(int) = 0.0176]

100.0 %

Full-matrix least-squares on F2
102/0/17

1.072

R1 = 0.0145, wR2 = 0.0338

R1 = 0.0183, wR2 = 0.0371
0.0031(5)

 

R1 = z IlFoI - IE“ I z IE1 and wR2 = 12 (11:3 — p.21 )2 7 z (”(3)2110

27

Table 2-6. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for Cr3B4.

 

 

x y 2 U(eq)
Cl‘( 1) 0 0 0 2( 1)
Cr(2) -5000 0 -1862(1) 2(1)
3(1) -5000 -5000 -668(3) 3(1)
B(2) 5000 -5000 - 1393(3) 4(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

Table 2-7. Anisotropic displacement parameters (Azx 103) for Cr3B4.

 

 

ull u22 1133 U23 ul3 ul2
Cr(l) 2(1) 3(1) 1(1) 0 0 O
Cr(2) 1(1) 3(1) 2(1) 0 0 0
B(l) 0(2) 5(2) 4(2) 0 O 0
B(2) 5(2) 1(2) 5(2) 0 0 O

 

The anisotropic displacement factor exponent takes the form: -27t2[ h2 a"‘2Ull +

+2hka*b*U12]

28

Table 2-8. Bond distances (A) for Cr3B4.
Crl 82 8x 2.2699(3)
Crl B1 4x 2.3353(3)
Cr2 B2 2x 2.1410(3)
Cr2 B1 4x 2.1824(3)
Cr2 B1 1x 2.2707(5)
B1 B2 2x 1.7381(3)
B1 Cr2 4x 2.1824(3)
B l Cr2 1x 2.2707(5)
Bl Crl 2x 2.3353(3)
B2 B1 2x 1.7381(3)
82 B2 1x 1.7561(4)
B2 Cr2 2x 2.1410(3)

B2 Crl 4x 2.2699(3)

29

Table 2-9. Crystal data and structure reﬁnement for Cr2B3.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 29.13°
Reﬁnement method

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

Final R indices [I>2sigma(1)]
R indices (all data)

Extinction coefﬁcient

136.43
293(2) K

0.71073 A

Orthorhombic

Cmcm

a = 3.0059(6) A a: 90°
b = 18.084(4) A [3: 90°
c = 2.9436(6) A y = 90°
160.01(6) A3

4

5.663 g/cm3

12.949 mm-1

252

2.25 to 29.13°
45hs4,-245ks24,-45153
737

145 [R(int) = 0.0286]

98.6 %

Full-matrix least-squares on F2
145/0/22

1.257

R1 = 0.0316, wR2 = 0.0503

R1 = 0.0320, wR2 = 0.0503
0.066(6)

 

R1 = z ||F0| — |F.|| / z |F0| and wR2 = [z (|F.,2 — F.2l )2 / Z (wFo2)2]“2

30

Table 2-10. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for Cr2B3.

 

 

x y z U(eq)
Cr( 1) 0 4276(1) 2500 2(1)
Cr(2) 0 7036(1) 2500 2(1)
B( l) 0 8280(5) 2500 5(2)
B(2) 0 1 191(5) 2500 4(2)
B(3) 0 235(5) 2500 3(2)

 

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

Table 2-11. Anisotropic displacement parameters (Azx 103) for Cr2B3.

 

 

ull {)2 [)33 [)23 ul3 U12
Cr(l) 4(1) 1(1) 2(1) 0 0 0
Cr(2) 2(1) 1(1) 2(1) 0 0 0
B(2) 0(5) 4(4) 8(5) 0 0 O
B(3) 3(5) 2(4) 4(4) 0 0 O
B(l) 8(6) 6(4) 2(4) 0 0 0

 

The anisotropic displacement factor exponent takes the form: ~27t2[ h2 a"‘2Ull +

+2hka*b*U12]

31

Table 2-12. Bond distances (A) for Cr2B3.
Crl B2 4x 2.2654(3)
Crl B3 4x 2.2817(3)
Crl B3 2x 2.2936(3)
Crl B1 2x 2.3433(3)
Cr2 B2 2x 2.1462(3)
Cr2 B 1 4x 2.1808(3)
Cr2 B1 1x 2.2548(4)
B1 B2 2x 1.7555(3)
B1 Cr2 4x 2.1808(3)
Bl Cr2 1x 2.2548(4)
B1 Crl 2x 2.3433(3)
B2 B3 1x 1.7247(3)
B2 B 1 2x 1.7555(3)
B2 Cr2 2x 2.1462(3)
B2 Crl 4x 2.2654(3)
B3 B3 2x 1.6990(3)
B3 B2 1x 1.7247(3)
B3 Crl 4x 2.2817(3)

B3 Crl 2x 2.2936(3)

32

Table 2-13. Crystal data and structure reﬁnement for CrB2.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 29.02°
Reﬁnement method

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

Final R indices [I>281gma(1)]
R indices (all data)

Extinction coefﬁcient

73.62
100(2) K

0.71073 A

Hexagonal

P6/mmm

a = 2.9604(4) A a: 90°
b = 2.9604(4) A B= 90°
c = 3.0574(6) A y = 120°
23.205(6) A3

2

10.536 g/cm3

22.350 mm'1

68

6.66 to 29.02°
45h54,—35k53,—45154
219

24 [R(int) = 0.0289]

100.0 %

Full-matrix least-squares on F’2
24/0/6

1.296

R1 = 0.0163, wR2 = 0.0402

R1 = 0.0163, wR2 = 0.0402
1.0(5)

 

R1 = Z ||F..| - |F.|| / Z |F0| and wR2 = [z (|F.,2 — F..2| )2 / z (wF02)2]'/2

33

Table 2-14. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for CrB2.

 

 

x y z U(eq)
Cr(l) 0 0 5000 1(1)
13(1) -3333 3333 0 1(1)

 

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

Table 2-15. Anisotropic displacement parameters (Azx 103) for CrB2.

 

 

U11 u22 u33 U23 ul3 {112
Cr(l) 1(1) 1(1) 0(1) 0 0 1(1)
3(1) 1(1) 1(1) 0(2) 0 0 0(1)

 

The anisotropic displacement factor exponent takes the form: -27t2[ h2 a"‘2U11 +

+2hka*b*U12]

Table 2-16. Bond distances (A) for CrB2.

Crl B 1 12x 2.2943(3)
B1 Bl 3x 1.7069(2)

B1 Crl 6x 2.2943(2)

34

Table 2-17. Crystal data and structure reﬁnement for MoB and WB

Formula
Formula weight
Crystal system
Space group

Unit cell dimensions

Volume

Z, Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size, mm

Theta range for data collection

Index ranges

Reﬂections collected
Independent reﬂections
Completeness to 0mm,
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F’2

Final R indices [I>20(I)]

R indices (all data)

Extinction coefﬁcient

MoB
106.75
Tetragonal
141/amd
a = 3.1103(4) A
b = 3.1103(4) A
= 16.899(3) A
163.48(4) A3
8, 8.674 g/cm3
14.713 mm'1
376
0.004 x 0.052 x 0.096
4.73 to 29.11°
4 5 h 5 3.
4 5 k 5 3,
-21 5 1 5 21
679
76 [R(int) = 0.0454]
100.0 %

WB
194.66
Tetragonal
141/amd
a = 3.1138(8) A
b = 3.1138(8) A

= 16.929(9) A
164.14(10) A3
8, 15.755 g/cm3
139.427 mm1
632
0.004 x 0.072 x 0.106
4.71 to 28.30°
-35h53
-35k53
-21 5 1 5 21
524
65 [R(int) = 0.0435]
91.5%

Full-matrix least-squares on E2

76/ 0 / 9
1.075

R1 = 0.0206,
wR2 = 0.0456
R1 = 0.0308,
wR2 = 0.0469
0.013(2)

65/ 0/ 10
1.384

R1 = 0.0663,
wR2 = 0.1680
R1 = 0.0669,
wR2 = 0.1688
0.015(5)

 

R1 = Z ||F.,| — |F..|| / z |F0| and wR2 = [2 (pa,2 — Ffl )2 / z (6)5521"?

35

Table 2-18. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103)for MoB and WB.

 

 

x y 2 U(EQ)
M0(1) 0 2500 721(1) 3(1)
B(2) -5000 2500 -309(6) 4(2)
W( 1) 0 2500 724(1) 4(2)
B( 1) -5000 -2500 300(40) 40(30)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

Table 2-19. Anisotropic displacement parameters (Azx 103) for MoB and WB.

 

 

U11 U22 U33 U23 U13 U12
Mo(l) 2(1) 2(1) 6(1) 0 0 1(1)
3(1) 4(5) 3(5) 6(4) 0 0 0(1)
W(1) 6(2) 5(2) 0(2) 0 0 0
3(1) 0(40) 130(80) 0(30) 0 0 0

 

The anisotropic displacement factor exponent takes the form: -27t2[ h2 a*2Ull +

+2hka*b*U12]

36

Table 2-20. Bond distances (A) for MoB.

M01 B2 4x 2.3070(2)
Mol B2 2x 2.3353(3)
Mol B2 1x 2.4826(5)
B2 B2 2x 1.8739(2)
BZ M01 4x 2.3070(2)
B2 Mol 2x 2.3353(3)

B2 Mol 1x 2.4826(5)

Table 2.21. Bond distances (A) for WB.

W1 Bl 4x 2.3153(4)
W1 Bl 2x 2.3307(7)
W1 B1 1x 2.4979(13)
B1 B1 2x 1.8604(4)
B1 W1 4x 2.3153(4)
Bl W1 2x 2.3307(7)

Bl W1 lx 2.4979(13)

37

D. Structural Description

Cr5B3: The structure of Cr5B3 (Figure 2-1a) was ﬁrst reported by Bertaut and
Blum,l4 and came to be the representative of its own structure type. A common
characteristic for the structure type is the formation of dumbbell-like pairs of atoms,
shown in Figure 2-1b. An investigation of the coordination environments of the atoms
reveals elongated octahedra of boron atoms around Crl and antiprisms of chromium
atoms around B1. The trigonal prismatic coordination commonly found for boron atoms
is seen in the chromium atoms around B2.

Cr3B4: The structure of Cr3B4 belongs to the T a3B4-structure type and was ﬁrst
described by researchers in Sweden.15 The structure type is characterized by double
chains of boron atoms that link together to form a series of side-sharing hexagons
(Figure 2-2a). The structure can also be described, as in the work by Okada, et al.,7 by
stacking sheets of AlB2-type structure. A V2, 1/2, 0 unit cell translation of the A1B2
structure builds the Cr3B4 structure. The sheets forming the structure are then
interconnected through Cr-Cr bonds.

Cr2B3: A description of the Cr2B3 structure, which belongs to the V2B3-structure
type, was given by Okada, et al., in the ﬁrst account on the structure of the compound.7
As with Cr3B4, the structure can be built by V2, 1/2, 0 unit cell translations of an A1B2-type
structure. In this compound, triple chains of boron atoms link together and form two
columns of hexagons (Figure 2-2b). The previous structural description by Okada also
points out that the interatomic distances between chromium atoms (2.62 A) are larger
than the distance in the elemental metal (2.498 A), suggesting that Cr—B and B—B bonds

hold the sheets in the structure together.

38

 

 

 

 

 

Figure 2-1 (a) Representation of Cr5B3 along the b-axis. (b) Boron pairs in Cr5B3

(Cr atoms omitted for clarity).

39

 

 

 

 

 

 

c
c
“3‘ 3:3”? 8}: ~
c c c c :g'
4 c
t

(b)

 

Figure 2-2. Double and triple chains of boron atoms in (a) Cr3B4 and (b) Cr2B3,
respectively.

40

CrB2: Like most transition metal diborides, CrB2 crystallizes with the A1B2-
structure type. This structure-type comprises lamellar sheets of boron atoms in a
graphite-like arrangement and hexagonally close-packed metal atoms. The structure
of CrB2 is shown in Figure 2-3a and the planar nature of the graphite-like boron
layers are emphasized in Figure 2-3b.

MB (M=Mo, W): The monoborides of molybdenum and tungsten both crystallize in
the MOB-structure type (Figure 2-4). The structure, ﬁrst described by Kiessling,2
features inﬁnite zigzag chains of boron atoms running parallel to the a-axis in one sheet
and parallel to the b-axis in the next. The metal atoms form trigonal prisms around the

boron atoms.

E. Results and Discussion

As mentioned, the molten metal ﬂux technique has been used to synthesize
group 6 borides with copper and aluminum as the ﬂux metals, so we were interested in
extending the technique to include gallium. The synthesis and growth of Cr2B3, CrB2,
MoB, and WB from a gallium ﬂux indicates initial success in our extension of the
technique. These results are encouraging, but they have been limited to binary borides.

The chromium borides can be formed from systems containing just chromium,
boron, and gallium, so the inﬂuence of varying the chromium to boron ratio on the
resulting borides was brieﬂy investigated. One of the most interesting things found was
that CrB2 was the only boride formed at the most boron-rich ratios. Though it was

expected that CrB4 would form at some point, that was not the case, even when the B:Cr

41

\ £5

   
  
    
    

\ aces".
a‘nmnmm~
“V‘V "“\'

    

waxes-er -

 

° W
ﬁ. nu! ‘ 32!". 1;?{zliji{?1—z :":;T-‘ :7:."'. ' ‘u-I- ‘— ‘Cl’
“La 1811 _ * ca
391712." i.‘.‘_ -._ 1... , .

.
I - _
A ‘1’ an ‘-
‘I' u
u I" ' r I. I! ' _-'
- if ,I‘." -‘,-_ 2.1. . - -—
v r ‘ s
' r I'—
r V l I" I r I _
' """‘-"z‘~._I. ..«-‘- ‘- '-

Figure 2-3. (a) CrB2 structure seen along c-axis (b) Planar-nature of boron layers in
CrB2.

42

     

 

 

“'4‘.

6%..

b‘i/OU‘O“

M
WB
L.

      

 

Q‘bca
shown

 

ratio reached 10:1. It was also slightly troubling that more borides with lower B:Cr ratios
were not found. An expanded study looking at more B:Cr ratios should help determine
which borides can be synthesized from the gallium ﬂux. The results of the previous
aluminum ﬂux studies also imply that cooling rates for the reactions must also be studied.

Our limited attempts to synthesize ternary borides from a gallium ﬂux were
unsuccessful. However, monoborides of the heavier group 6 members from these
reactions aimed at ternary species were synthesized. Subsequent attempts to isolate these
borides from systems containing only the metal, boron, and gallium have not been
successful. However, further investigations varying the B:M ratios and heating proﬁles
are needed for a complete study. Varying the heating proﬁles and reagent ratios may also
help in forming ternary group 6 borides when other metals are included in the reactions.

Crystals of Cr5B3, Cr3B4, and Cr2B3 have been successfully grown from reactions
utilizing a molten copper ﬂux. The latter two borides were grown from systems
containing only chromium, boron, and copper, so there is potential that other borides will
be found from similar reactions that vary the B:Cr ratio and heating proﬁles. While we
were unable to isolate borides of the heavier group 6 elements or ternary borides with any
group 6 members from copper flux reactions, further studies may be made more

successful by varying the reagent stoichiometry and/or heating proﬁles.

F. Conclusions

The molten metal ﬂux technique has successfully been applied to the synthesis of
Group VI borides. Both gallium and copper have been used as the ﬂux metal in the

synthesis of Cr2B3. Molybdenum monoboride, tungsten monoboride, and chromium

44

diboride have been synthesized from gallium flux reactions. 0ng and Q33, have been
synthesized from reactions utilizing molten copper. We have not synthesized any ternary
Group VI borides from either gallium or copper, but our investigations into these systems
are rather limited. We believe that further investigations into the ﬂux reactions are

merited and necessary based on these initial results.

45

References

1. Klesnar, H.; Aselage, T. L.; Morosin, B.; Kwei, G. H.; Lawson, A. C., The Diboride
Compounds of Molybdenum: MoB2.x and M02B5-y. Journal of Alloys and Compounds
1996, 241, 180-186.

2. Kiessling, R., The Crystal Structure of Molybdenum and Tungsten Borides. Acta
Chemica Scandinavica 1947, 1, 893-916.

3. Lundstrom, T.; Rosenberg, I., The Crystal Structure of the Molybdenum Boride
M01433. Journal of Solid State Chemistry 1973, 6, 299-305.

4. Lundstrom, T., The Structure of RU283 and WB2”, as determined by single-crystal
diffractometry, and some notes on the W-B system. Arkivfbr Kemi 1969, 30, 115-127.

5. Romans, P. A.; Krug, M. P., Composition and Crystallographic Data for the Highest
Boride of Tungsten: Acta Crystallographica 1966, 20, 313-315.

6. Hi gashi, 1.; Takahashi, Y.; Atoda, T., Crystal Growth of Borides and Carbides of
Transition Metals from Molten Aluminum Solutions. Journal of Crystal Growth 1976,
33, 207-211.

7. Okada, S.; Atoda, T.; Higashi, 1., Structural Investigation of Cr2B3, Cr3B4, and CrB by
Single-Crystal Diffractometry. Journal of Solid State Chemistry 1987, 68, 61-67.

8. Okada, S.; Atoda, T.; Higashi, I.; Takahashi, Y., Preparation of Single Crystals of
MoB2 by the Aluminum-flux Technique and Some of Their Properties. Journal of
Materials Science 1987, 22, 2993-2999.

9. Okada, S.; Kudou, K.; Lundstrom, T., Preparations and Some Properties of W2B,
5-WB, and WB2 Crystals from Hi gh-Temperature Metal Solutions. Japanese Journal of
Applied Physics 1995, 34, (1), 226-231.

10. Sheldrick, G. M. SAINT, Version 4; Siemens Analytical X-Ray Instruments, Inc.:
Madison, WI.

11. Sheldrick, G. M. SHELXTL Structure Determination Program, Version 5.0; Siemens
Analytical X-Ray Instruments, Inc.: Madison, WI, 1995.

12. X-SHAPE, Version 2.05; STOE & Cie GmbH: 2004.
13. X-RED 32, Version 1.10; STOE & Cie GmbH: 2004.

14. Bertaut, F.; Blum, P., Etude des borures de chrome. Comptes Rendus
Hebdomadaires des Seances de l'Academie des Sciences 1953, 236, 1055-1056.

46

15. Elfstrom, M., The Crystal Structure of Cr3B4. Acta Chemica Scandinavica 1961, 15,
(5), 1178.

47

CHAPTER THREE

The Metal Flux Synthesis of Group Seven Borides

A. Introduction

There are many similarities between the borides of chromium and the borides of
manganese. Both metals have borides that display the interesting series of boron atom
conﬁgurations from isolated boron atoms to a three-dimensional network found only in
CrB4 and MnB4.l Several reports have even mentioned a manganese monoboride that
crystallizes with the CrB-type structure, instead of its normal FeB-type structure.2‘ 3
However, several manganese analogs are “missing” when compared to the chromium
system, such as Mn5B3 and Mn2B3. We became interested in whether it was possible to
use the metal ﬂux to isolate these “missing” borides.

The literature for ﬂux syntheses of manganese borides returned only a ternary
phase that incorporated aluminum from the ﬂux, Mn2AlB2.4 This was a bit surprising, due
to the success of the ﬂux technique with the group 6 metalss'8 During the literature
search we also checked for reports on ﬂux synthesis of rhenium borides. Again, we
found no previous work done in the area, but the overall lack of work done in the Re-B
system did not make this overly surprising.

The Re-B system has produced Re3B,9 ReB2,10 and Re7B3,ll and only Re3B had

structural characterization performed on a single crystal. Rather large single crystals of

48

ReB2 were recently grown using the ﬂoating zone method, however, the growth
temperature was near 2400°C, well above temperatures the equipment in our labs can
attain consistently. The lower temperatures used in our ﬂux reactions would be very
attractive for producing crystals of any of these compounds.

Single crystals are useful when looking for anisotropic behavior of properties in a
compound. The magnetism in Mn3B4 is a ﬁne example of this anisotropy, as
ferromagnetism is detected in the ac-plane and antiferromagnetism is found between
layers along the b-axis.”’ 13 Discovering and studying similar phenomena in other
compounds would be greatly simpliﬁed by using single crystals. The nature of the

14, 15 may also be better understood when single

superconductivity in rhenium borides
crystals are used in the physical measurements. The use of single crystals limits the
possibility of minor phases negatively inﬂuencing results, while these minor phases may
go undetected in powdered samples.

We report here our attempts to use the molten metal ﬂux technique in the
synthesis of group 7 borides. A molten gallium ﬂux has been used to synthesize three
binary borides, MnB, MnB4, and ReBz, and four ternary borides, HoMnB4, DyReB4,

ErReB4, and YbReB4. Our attempts to use alternate ﬂux metals in the syntheses of these

compounds are also discussed.

B. Experimental Section

Reagents:
All reagents were used as received without further puriﬁcation: (i) Boron metal,

99% purity, -325 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (ii) Rhenium

49

powder, 99.99% purity, Strem Chemicals, Newburyport, MA, (iii) Gadolinium metal
(ﬁled from ingot) 99.9% purity, Chinese Rare Earth Information Center, Inner Mongolia,
China, (iv) Dysprosium metal (ﬁled from ingot) 99.9% purity, Chinese Rare Earth
Information Center, Inner Mongolia, China, (v) Erbium metal (ﬁled from ingot) 99.9%
purity, Chinese Rare Earth Information Center, Inner Mongolia, China, (vi) Gallium 3-5
mm shot, 99.999% purity, Plasmaterials, Liverrnore, CA, (vii) Manganese powder, 99.9%
purity, -50 mesh, Aldrich Chemical Company, Milwaukee, WI, (viii) Tantalum powder,
99.9% purity, Cerac Specialty Inorganics, Milwaukee, WI.

Furnaces:

A Lindberg Blue high-temperature, horizontal tube furnace was used with a
continuous argon ﬂow of approximately 0.1 L/min. Heating proﬁles for individual
reactions were programmed into the accompanying Lindberg UP150 furnace controller.
Thermocouples connected to the furnace controller were used to monitor ﬁimace
temperatures during reactions. The maximum operating temperature for the furnace was
1500 °C.

A radio frequency (RF) furnace was constructed from an Ameritherm XP7.5CE
Induction Heater and fused quartz reaction chamber. A molybdenum cylinder, machined
to act as a reaction crucible holder, was used as a susceptor. A boron nitride cylinder was
machined to hold the molybdenum crucible in the center of the reaction chamber. Details
on the RF furnace setup are given in the appendix of this work.

RF furnace temperatures were controlled manually using the Ameritherm control
unit. Reaction temperatures were monitored using an Ircon SA Series Infrared

Thermometer. During the course of some reactions, ﬁlms deposited inside the reaction

50

chamber would prevent the IR thermometer from accurately reporting furnace
temperatures. In such circumstances, reaction temperatures were approximated based on
power settings for the Ameritherm control unit and a series of calibration runs conducted
by heating only the molybdenum cylinder. Temperatures greater than 1700 °C were
easily reached with this furnace.

Synthesis and Isolation:

MnB: Boron (4 mmol), tantalum (1 mmol), and manganese (2 mmol) powders
were mixed and added to an alumina crucible along with pieces of gallium shot totaling
1.67 g. A quartz ﬁlter was placed in the crucible which was then placed in 13 mm
diameter quartz tube, evacuated under vacuum to a pressure near 10'4 Torr, and sealed
using an oxygen-acetylene torch. The tube was then heated in a commercial tube furnace
under a heating proﬁle that increased the furnace temperature to 1000 °C in 10 hours,
held the temperature constant for 48 hours, allowed the furnace to cool to 850 °C in 5
hours, held the temperature constant for an additional 48 hours, and then allowed the
furnace to cool to 250 °C in 60 hours.

The quartz tube was removed from the furnace at 250 °C and centrifuged to
remove excess gallium. After centrifugation, the tube was opened and the crucible was
placed in a bath of 5M 12 in DMF in order to remove the remaining gallium ﬂux. An
overnight soak was found to be insufﬁcient for removing all of the gallium, so soaking
continued for an additional 2 days. Extended soaking in the bath was not found to
damage the crystals. Silver, prismatic crystals were produced in about 40% yields, based
on manganese. Visual distinction between MnB and MnSi crystals is nearly impossible,

so EDS analysis was used to differentiate crystals. Elemental analysis performed by EDS

51

using an SEM showed that manganese and boron were the only elements present in the
crystal used for diffraction study. Besides the MnSi, the collected product also contained
MnGa4 and MnGa4_5.

MnB4: Boron and manganese powder were mixed and added to alumina
crucibles in millimolar quantities in boron-to-manganese ratios between 2:1 and 5:1.
Pieces of gallitnn shot totaling between 1.5 g and 2.5 g were then placed in the crucibles.
The alumina crucibles were then placed in 13 mm diameter quartz tubes and evacuated
under vacuum to a pressure near 10'4 Torr. The quartz tubes were then sealed using an
oxygen-acetylene torch. The quartz tubes were placed in mullite tubes, then heated in
commercial tube furnaces under heating proﬁles that increased the furnace temperature to
1000 °C in 10 hours, held the temperature constant for 48 hours, allowed the furnace to
cool to 850 0C in 5 hours, held the temperature constant for an additional 48 hours, and
then allowed the furnace to cool to room temperature in 60 hours.

After cooling, the quartz tubes were opened and the alumina crucibles were
placed in a half-concentrated HCl bath in order to remove the excess gallium. Most of
the gallium was removed after an overnight soak, but longer soaks were used without
damage to the recovered product. Silver, plate-like crystals with ridges (later found to be
twinned crystals) were produced in 30-40% yields. Elemental analysis performed by
EDS using an SEM showed that manganese and boron were the only elements present in
the crystals. The remainder of the product comprised MnGa4 and MnGa4,5.

ReBz: Boron and rhenium powder were mixed and added to alumina crucibles in
millimolar quantities in boron-to-rhenium ratios between 2:1 and 5:1. Pieces of gallium

shot totaling between 1.5 g and 2.5 g were then placed in the crucibles. The alumina

52

crucibles were then placed in 13 mm diameter quartz tubes and evacuated under vacuum
to a pressure near 10'4 Torr. The quartz tubes were then sealed using an oxygen-
acetylene torch. The quartz tubes were placed in mullite tubes, then heated in
commercial tube furnaces under heating proﬁles that increased the furnace temperature to
1000 °C in 10 hours, held the temperature constant for 48 hours, allowed the furnace to
cool to 850 °C in 5 hours, held the temperature constant for an additional 48 hours, and
then allowed the furnace to cool to room temperature in 60 hours.

After cooling, the quartz tubes were opened and the alumina crucibles were
placed in a half-concentrated HCl bath in order to remove the excess gallium. Most of
the gallium was removed aﬁer an overnight soak, but longer soaks were used without
damage to the recovered product. Silver, hexagonal plates are produced in 40-50%
yields, based on rhenium. Elemental analysis performed by EDS using a SEM showed
that rhenium and boron were the only elements present in the crystals. The remainder of
the product comprised ReGa4_5.

HoMnB4: Boron (3.5 mmol) and manganese (0.5 mmol) powders were mixed
and added to an alumina crucible along with holmium (1.5 mmol) ﬁlings. Pieces of
gallium shot totaling 1.5 g were added to the crucible before it was placed in a 13 mm
diameter quartz tube and evacuated under vacuum to a pressure near 10'4 Torr. The
quartz tube was then sealed using an oxygen-acetylene torch and placed in a mullite tube.
Heating took place in a commercial tube furnace under a heating proﬁles that increased
the furnace temperature to 1000 °C in 10 hours, held the temperature constant for 48

hours, allowed the furnace to cool to 850 °C in 5 hours, held the temperature constant for

53

an additional 48 hours, and then allowed the furnace to cool to room temperature in 60
hours.

After cooling, the quartz tubes were opened and the alumina crucibles were
placed in a half-concentrated HCl bath in order to remove the excess gallium. Most of
the gallium was removed after an overnight soak, but longer soaks were used without
damage to the recovered product. Silver, plate-like crystals are a minor product (20-30%
yields, based on manganese). Elemental analysis performed by Energy Dispersive
Spectroscopy (EDS) using a Scanning Electron Microscope (SEM) showed that
holmium, manganese, and boron were the only elements present in the crystals. The
remainder of the product comprised MnGa4, MnGars, traces of HoB6, and unreacted
boron.

REReB4 (RE= Dy, Er, Yb): Boron and rhenium powders were added to alumina
crucibles, along with ﬁlings from an ingot of a rare-earth metal, in millimolar quantities
in a 1.5:] :4 ratio of RE:Re:B. Pieces of gallium shot totaling 2 g were placed atop the
reagents in the crucibles. The alumina crucibles were then placed in an alumina boat,
which was in turn place inside an alumina tube. The alumina tube was placed in a high-
temperature tube furnace and heated under a slow ﬂow of argon. The furnace was heated
to 1400 °C in 8 hours and held at that temperature for 10 hours, then cooled to 1200 °C in
5 hours and held at that temperature for 10 hours, and ﬁnally cooled to 500 °C in 36
hours, at which point the power to the furnace was shut off and the furnace was allowed
to cool to room temperature.

After cooling, the crucibles were removed from the boats in the tube and placed in

a half-concentrated HCl bath in order to remove excess the gallium. An overnight soak

54

was adequate to remove the gallium. DyReB4 and ErReB4 were collected in estimated
30-50% yields based on rhenium. YbReB4 was collected in an estimated 20-40% yield
based on rhenium. Side products in the DyReB4 reaction included ReGa4_5 and DyB6.
Side products in the ErReB4 reaction included ReGa4_5 and ErB6. Side products in the

YbReB4 reaction included ReGa4,5 and YbB6.

C. Physical Measurements

Energy Dispersive Spectroscopy: Energy dispersive spectroscopy (EDS)
analyses were carried out on selected crystals to determine their chemical composition.
The analyses were performed with a J EOL J SM-6400 scanning electron microscope
(SEM) equipped with Noran Vantage Energy Dispersive Spectroscopy (SizLi) detector
and Norvar window for standardless quantization of elements with Z Z 4. The crystals
were afﬁxed to an alumina sample holder with double-sided carbon tape. The EDS data
were acquired at an accelerating voltage of 10 kV with a 30 to 60 second accumulation
time.

Single-Crystal X-ray Crystallography: Single crystals were mounted on a glass
ﬁber with super glue and their intensity data were collected on either a Bruker SMART
platform CCD diffractometer, or a STOE IPDS II diffractometer using Mo K01 radiation
at 50 kV and 40 mA.

Individual frames of REReB4 (RE= Dy, Er, Yb) examined with the Bruker
machine were collected with a 03° 0) rotation. The SMART software was used for data

collection, and SAINTl6 software was used for data extraction and reduction. After

55

applying analytical absorption corrections, structure solution and reﬁnement were
completed using direct methods and the SHELXTLl7 suite of programs.

Samples examined with the STOE (MnB, MnB4, ReB2, and HoMnB4) had
individual frames collected on a 34 cm image plate with a 60 second exposure time and a
10° 0) rotation. The X-SHAPEl8 and X-RED19 soﬁware packages were used for data
extraction and reduction and to apply an analytical absorption correction. Direct methods
and the SHELXTL suite were used to solve and reﬁne the structures.

The crystal structure reﬁnement data for MnB are found in Table 3-1. Atomic
parameters previously reported5 for CrB were used as a starting point for the reﬁnement
of the structure. The atomic positions and isotropic displacement parameters are listed in
Table 3-2. AnisotrOpic displacement parameters for the atoms are found in Table 3-3.
The bond lengths in MnB and a count of those bonds are listed in Table 3-4.

The reﬁnement data for MnB4 are found in Table 3-5. The atomic parameters
from CrB420 were used as a starting point for reﬁnement of the structure. Parameters

21. 22 apply to the monoclinic space group

from previous reports on the structure of MnB4
C2/m (a = 5.5029, b = 5.3669, c = 2.9487, [3 = 122.71) and are not useful in the
orthorhombic Immm space group. Atomic positions and isotropic parameters found in
this study of W., are given in Table 3-6. Anisotropic displacement parameters for the
atoms are listed in Table 3-7. The values for these parameters are quite high and may be
the result of the twinning in the crystal or a modulated structure. The bond lengths in
MnB4 and a count of those bonds in the unit cell are listed in Table 3-8.

The crystal structure reﬁnement data for ReB2 are listed in Table 3-9. The atomic

parameters reported in the ﬁrst structural reportlo were used as a starting point for the

56

reﬁnement of the structure from our data. The atomic positions and isotropic
displacement parameters we have found are listed in Table 3-10. Anisotropic
displacement parameters for the atoms are listed in Table 3-11. Bond lengths and a count
of those bonds in the structure are found in Table 3-12.

The crystal structure reﬁnement data for HoMnB4 are listed in Table 3-13. The
atomic positions and isotropic displacement parameters for HoMnB4 are listed in Table
3-14. The manganese atoms were distributed over split positions, so anisotropic
parameters were not carried out for these atoms. Anisotropic displacement parameters
for the remaining atoms are listed in Table 3-15. The anisotropic displacement
parameters are high, especially for the boron atoms. Inadequate absorption correction,
modulated structure, and a twinned crystal may all play a role in creating these high
values. Given the similar crystal morphology between HoMnB4 and MnB4, crystal
twinning is likely present.

The reﬁnement data for REReB4 (RE = Dy, Er) are listed in Table 3-16. The
reﬁnement data for YbReB4 are listed in Table 3-17. The atomic parameters for YCrB423
were used as a starting point for the reﬁnement of our data, as all of these compounds are
members of the YCrB4 structure type. The atomic parameters and isotropic displacement
parameters for REReB4 (RE = Dy, Er, Yb) are listed in Table 3-18. The anisotropic
displacement parameters for REReB4 (RE = Dy, Er, Yb) are listed in Table 3-19.
Inadequate absorption corrections are most likely responsible for the low values of the
parameters for the heavy atoms in all of these compounds. The bond distances for

HoMnB4 and REReB4 (RE = Dy, Er, Yb) are listed in Table 3-20.

57

Table 3-1. Crystal data and structure reﬁnement for MnB.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefficient

F (000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 34.65°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F 2

Final R indices [I>28igma(1)]
R indices (all data)

Extinction coefﬁcient

65.75

293(2) K

0.71073 A

Orthorhombic

Cmcm

a=3.0113(6)A or= 90°
b = 7.6733(15) A B= 90°
c = 2.9610(6) A y = 90°

6842(2) A3

4

6.383 g/cm3

17.695 mm“

120

5.31 to 34.65°
-45h54,-125k512,-45154

444

92 [R(int) = 0.0299]

91.1 %

Full-matrix least-squares on F2

92/ 0/ 10

1.230

R1 = 0.0135, wR2 = 0.0324

R1 = 0.0135, WR2 = 0.0324
035(4)

 

R1 = z ”Fol — mu /2 |F0| and wR2 = [2 (IF,2 — FCZI )2 / Z (wFo2)2]"2

58

Table 3-2. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for MnB.

 

 

x y z U(eq)
Mn(l) 5000 1438(1) 2500 4(1)
8(1) 0 663(3) -2500 5(1)

 

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

Table 3-3. Anisotropic displacement parameters (Azx 103) for MnB.

 

 

U11 U22 u33 U23 U13 ulz
Mn(1) 4(1) 4(1) 4(1) 0 0 0
8(1) 5(1) 5(1) 4(1) 0 0 O

 

The anisotropic displacement factor exponent takes the form: -21r2[ h2 a"'2Ull + + 2 h
k a“ b* 012 ].

Table 3-4. Bond distances (A) for MnB.
Mnl B1 4x 2.1937(3)
Mnl Bl 2x 2.2059(3)
Mnl B1 1x 2.2245(4)

Bl Bl 2X 1.7966(3)
Bl Mnl 4x 2.1937(3)
Bl Mnl 2x 2.2059(3)

B1 Mnl 1x 2.2245(4)

59

Table 3-5. Crystal data and structure reﬁnement for MnB4,

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 36.51 °
Reﬁnement method

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

Final R indices [I>231gma(l)]
R indices (all data)

Extinction coefﬁcient

98.18

293(2) K

0.71073 A

Orthorhombic

Im

a = 2.9380(6) A or= 90°
b = 4.6104(9) A B= 90°
c = 5.3528(11) A y = 90°
7251(3) A3

2

4.497 g/cm3

8.403 mm'1

90

5.84 to 36.51°
45h$4,-7Sk$7,-8Sl$8
589

119 [R(int) = 0.0512]

96.7 %

Full-matrix least-squares on F2
119 / 0/ 11

1.255

R1 = 0.0355, wR2 = 0.0854

R1 = 0.0355, wR2 = 0.0854
0.61(13)

 

R1 = Z “Fol - IFcII /Z IFoI and Wm = 12 (IF: — Fczl )2/z(wF.2)21“2

60

Table 3-6. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for MnB4.

 

 

x y 2 U(CQ)
Mn(l) 0 0 0 8(1)
3(1) 0 1995(13) 3440(11) 28(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

Table 3-7. Anisotropic displacement parameters (A2x 103) for MnB4_

 

 

ull U22 U33 U23 [113 [112
Mn(l) 22(1) 2(1) 1(1) 0 0 0
13(1) 3(1) 45(3) 35(2) -38(2) 0 0

 

The anisotropic displacement factor exponent takes the form: -21t2[h2 a"'2Ull + + 2 h
k at b“ [112 ]

Table 3-8. Bond distances (A) for MnB4,
Mnl Bl 4X 2.0580(3)
Mnl Bl 8x 2.1853(3)

Bl 131 lx 1.6706(3)
131 Bl 1x 1.8391(4)

Bl Bl 2x 1.8403(3)

61

Table 3-9. Crystal data and structure reﬁnement for ReB2,

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 35.9l°
Reﬁnement method

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

Final R indices [I>231gma(1)]
R indices (all data)

Extinction coefﬁcient

207.82

293(2) K

0.71073 A

Hexagonal

P63/mmc

a = 2.8891(4) A or= 90°
b = 2.8891(4) A [3: 90°
c = 7.4606(15) A 'y = 120°
53.930(15) A3

2

12.798 g/cm3

11 1.675 M'1

170

0.140 x 0.140 x 0.010 mm3

8.17 to 35.91°

-1 ShS4,-4Sk$4,-8SIS 11
270

66 [R(int) = 0.0842]

89.9 %

F ull-matrix least-squares on F2
66 / 0 / 7

1.352

R1 = 0.0218, wR2 = 0.0509

R1 = 0.0263, wR2 = 0.0517
0.091(17)

 

RI = z llFol — chll / z w and sz = 12 0F: - F31 )2 / z (wFffrm

62

Table 3-10. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for ReB2.

 

 

x y z U(eq)
Re(1) 6667 3333 2500 2(1)
3(1) 6667 3333 5460(20) 1(2)

 

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

Table 3-11. Anisotropic displacement parameters (A2x 103) for ReB2,

 

 

ull U22 u33 U23 1113 ulz
Re(1) 1(1) 1(1) 5(1) 0 0 1(1)
8(1) 1(3) 1(3) 0(4) 0 0 1(2)

 

The anisotropic displacement factor exponent takes the form: -2n2[ h2 a"‘2Ull + + 2 h
k a* b* U12 ]

Table 3-12. Bond distances (A) for ReB2,

Rel Bl 2x 2.2094(4)
Rel Bl 6x 2.2573(3)
81 81 3x 1.8045(2)
Bl Rel 1x 2.2094(4)
Bl Rel 3x 2.2573(3)

63

Table 3-13. Crystal data and structure reﬁnement for HoMnB4,

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 36.07°
Reﬁnement method

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

Final R indices [I>2sigma(1)]
R indices (all data)

Extinction coefﬁcient

263.1 1

293(2) K

0.71073 A

Orthorhombic

Pbam

a = 5.8946(12) A or= 90o
b=11.411(2)A 8=90°
c = 3.4041(7) A y= 90°
228.96(8) A3

4

7.633 g/cm3

39.388 mm"

448

4.97 to 36.07°
-9Sh$9,-17Sk$18,-551$4
3176

606 [R(int) = 0.1861]

97.9 %

Full-matrix least-squares on F 2
606 / 0 / 38

1.1 18

R1 = 0.0949, wR2 = 0.2203

R1 = 0.1477, wR2 = 0.2535
0.033(8)

 

RI = 2: “For — 1F.“ / z 181 and wR2 = 12 0F} — F21 )2 / z (mem

64

Table 3-14. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for HoMnB4_

 

 

x y z U(eq) Occupancy

Mn(l) 871 1(10) 5846(5) -5000 10(1) 0.86728
Mn(15) 8400(60) 6470(30) -5000 10(1) 0.13272
Ho(1) 6278(3) 3495(1) 5000 14(1)

B(2) 8700(100) 4520(30) 0 19(8)

B(3) 6480(1 10) 5350(50) 0 35(15)

B(4) 71 10(70) 6830(30) 0 14(6)

B(5) 10200(50) 693 0(40) 0 18(7)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

Table 3-15. Anisotropic displacement parameters (Azx 103) for HoMnB4,

 

 

ull U22 u33 U23 {113 ulz
Ho(l) 16(1) 15(1) 10(1) 0 0 0(1)
B(2) 50(20) 5(1 1) 2(9) 0 0 19(18)
B(3) 60(30) 40(20) 4(1 1) 0 0 60(3 0)
B(4) 31(18) 6(11) 6(11) 0 0 8(13)
3(5) 0(9) 30(19) 24(18) 0 0 10(1 1)

 

The anisotropic displacement factor exponent takes the form: -21'C2[ h2 a"'2Ull + + 2 h

ka* b* U12]

65

Table 3-16. Crystal data and structure reﬁnement for REReB4 (RE= Dy, Er).

F orrnula
Formula weight
Crystal system
Space group

Unit cell dimensions

Volume

Z, Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size, mm

Theta range for data collection

Index ranges

Reﬂections collected
Independent reﬂections
Completeness to 0mm,
Reﬁnement method

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

Final R indices [1>20'(1)]

R indices (all data)

Extinction coefﬁcient

DyReB4

391.94
Orthorhombic

Pbam

a = 5.9710(12) A

b = 11.545(2) A

c = 3.5980(7) A
248.03(9) A3

4, 10.496 g/cm3
78.259 mm'1

644

0.044 x 0.044 x 0.064
3.53 to 28.06°

-7 S h S 7,

-14 S k S 15,

-4 S l S 4

2410

340 [R(int) = 0.0490]
97.7 %

ErReB4

396.70
Orthorhombic

Pbam

a = 5.9490(12) A

b = 11.508(2) A

c = 3.5820(7) A
245.23(8) A3

4, 10.745 g/cm3
82.905 mrn'l

652

0.038 x 0.046 x 0.050
3.54 to 28.08°

-7 S h _<_ 7.

-14 S k S 15,

-4 S l S 4

2459

335 [R(int) = 0.0559]
97.1 %

F ull-matrix least-squares on F2

340 / 0 / 38
1.064

R1 = 0.0275,
wR2 = 0.0788
R1 = 0.0278,
wR2 = 0.0790
0.0324 (18)

335 / 0 / 38
1.131

R1 = 0.0395,
wR2 = 0.0887
R1 = 0.0413,
wR2 = 0.0898
0.00138(1 1)

 

R1 = Z llFol — 1F.” / 2 IF.) and sz = 1: (IF: — FEI )2 / z (wFozflm

66

Table 3-17. Crystal data and structure reﬁnement for YbReB4,

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Crystal size

Theta range for data collection
Index ranges .

Reﬂections collected
Independent reﬂections
Completeness to theta = 28.28°
Reﬁnement method

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

Final R indices [I>25igma(1)]
R indices (all data)

Extinction coefﬁcient

402.48

273(2) K

0.71073 A

Orthorhombic

Pbam

a = 5.9273(9) A or= 90°
b = 11.4497(18) A 8= 90°
c = 3.5617(6) A y = 90°
241 .72(7) A3

4

11.060 g/cm3

88.080 mm'1

660

0.052 x 0.052 x 0.104 mm3

3.56 to 28.28°
-7ShS7,-14SkS 15,-45154
2363

333 [R(int) = 0.0605]

96.2 %

F ull-matrix least-squares on F 2
333 / 0 / 38

1.143

R1 = 0.0431, wR2 = 0.1076

R1 = 0.0448, wR2 = 0.1088
0.0134(13)

 

R1 = 2 117.1 - 1F.” /2 IF.) and sz = 12 0F: - F}: )2/Z(wFoz)21”2

67

Table 3-18. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for REReB4 (RE = Dy, Er,Yb).

 

 

x y z U(eq)
Re(l) 1328(1) 4133(1) 0 2(1)
Dy(2) 1264(1) 1504(1) 0 2(1)
B(l) 2860(30) 3183(14) -5000 3(3)
B(2) -240(3 0) 3087(15) -5000 6(3)
B(3) -1 130(3 0) 4520(17) -5000 7(3)
B(4) 363 0(3 0) 4703(16) -5000 6(3)
Re(l) 1335(1) 4131(1) 0 0(1)
Er(2) 1267(2) 1502(1) 0 0(1)
B( l) 2860(40) 3 140(20) 5000 1(4)
B(2) 3850(40) 490(20) 5000 2(5)
B(3) 4770(40) 1920(20) 5000 1(5)
B(4) 365 0(40) 4700(20) 5000 3(5)
Yb(1) 1271(2) 1501(1) 0 0(1)
Re(2) 1339(2) 4132(1) 0 2(1)
B(l) 3610(50) 4690(20) 5000 4(6)
B(2) 1 120(50) 5510(30) 5000 3(6)
B(3) 283 0(5 0) 3 140(30) 5000 4(6)
B(4) -170(60) 3070(30) 5000 9(6)

 

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

68

Table 3-19. Anisotropic displacement parameters (Azx 103) for REReB4 (RE = Dy, Er,

 

 

Yb).

U1 l U22 u33 U23 ul3 ulz
Re(l) 2(1) 0(1) 4(1) 0 0 0(1)
Dy(2) 1(1) 0(1) 4(1) 0 0 0(1)
8(1) 4(7) 0(8) 6(8) 0 0 -3(6)
B(2) 4(7) 0(7) 1 3(8) 0 0 0(6)
B(3) 5(7) 4(8) 1 1(9) 0 0 5(6)
B(4) 2(7) 6(9) 1 1(8) 0 0 0(6)
Re(l) 0(1) 0(1) 0(1) 0 0 1(1)
Er(2) 0(1) 0(1) 0(1) 0 0 0(1)
B(l) 3(11) 0(11) 0(11) 0 0 0(9)
B(2) 1(11) 5(11) 0(11) 0 0 -5(9)
B(3) 0(10) 0(11) 3(11) 0 0 5(8)
B(4) 0(10) 0(1 1) 8(12) 0 0 1(8)
Yb(l) 0(1) 1(1) 0(1) 0 0 0(1)
Re(2) 2(1) 4(1) 0(1) 0 0 1(1)
3(1) 11(13) 0(12) 0(12) 0 0 15(11)
B(2) 7(14) 0(12) 3(13) 0 0 0(10)
B(3) 11(14) 0(12) 0(13) 0 0 2(11)
B(4) 18(16) 0(13) 10(15) 0 0 -4(12)

 

The anisotropic displacement factor exponent takes the form: -21t2[ h2 a"'2UH + + 2 h

kalt bar: U12]

69

Table 3-20. Bond distances (A) for HoMnB4 and REReB4 (RE = Dy, Er, Yb).

 

 

HoMnB4 DyReB4 ErReB4 YbReB4
RE(1)—M(1) 276(3) 3.0359(11) 3.0230(13) 3.0115(14)
M(1)—M(1) 2.457(11) 2.5537(11) 2.5546(17) 2.545(2)
M(1)—B(1) 228(2) 2.298(1 1) 2.309(16) 232(2)
M(1)—B(2) 222(5) 2.361(11) 2.364(15) 2.340(19)
M(1)—B(3) 225(2) 2.366(10) 2.357(16) 2.290(19)
M(1)—B(4) 228(3) 2.357(10) 2.352(15) 233(2)
RE(1)—B(1) 251(4) 2.737(12) 2.734(18) 274(2)
RE(1)—B(2) 272(4) 2.718(13) 2.632(19) 262(2)
RE(1)—B(3) 265(3) 2.655(13) 2.699(19) 274(2)
RE(1)—B(4) 261(3) 2.750(14) 2.739(19) 267(3)
B(1)—B(2) 161(7) 186(3) 180(4) 180(7)
B(1)—B(4) 178(5) 181(2) 186(4) 183(4)
B(2)—B(3) 183(5) 174(3) 174(4) 174(4)
B(3)—B(3) 1.92(15) 175(4) 184(4) 177(6)
B(3)—B(4) 173(7) 174(2) 174(3) 172(5)
B(4)—B(4) 1.89(12) 178(3) 175(5) 182(5)

 

70

D. Structural Description
MnB: The structure of MnB, shown in Figure 3-la, is composed of trigonal

prisms of manganese atoms centered by a zigzag chain of boron atoms. The most
interesting feature of the structure is the arrangement of the prisms and chains. Normally,
MnB crystallizes with the F eB-type structure (Figure 3-lb), where the prisms share
corners. We have isolated a variant phase of MnB that crystallizes in the CrB structure-
type (Figure 3-1c) where the prisms in each layer share trigonal faces. The layers in the
CrB-type MnB are more distinct, but overall, the two structure types are very similar.
Actually, the volume of this MnB phase is only 0.3% less than the volume of the F eB-
type.

As mentioned in the introduction to this chapter, CrB-type MnB has been
reported.2’ 3 However, the powder pattern in each of the previous reports also showed
lines attributed to FeB-type MnB. We saw no evidence of disorder in our diffraction data
suggesting a mix of the FeB and CrB-structure types, as seen by Kanaizuka.2 Manganese
was the only metal atom detected by the reﬁnement of the atom occupancy during
reﬁnement. EDS analyses on the single crystals conﬁrm this ﬁnding.

MnB4: The structure of MnB4 reported here is analogous to that ﬁrst found for
CrB4,20 and may be the “semi-MnB4” structure described by Burdett and Canadell.l The
deﬁning feature in the structure of MnB4 (Figure 3-2a) is a three-dimensional network of
boron atoms. The network comprises boron atom rectangles lying parallel to the bc-plane
that bond to boron atoms in other rectangles, extending the network in all three
dimensions. The manganese atoms sit in the middle of channels formed by six

neighboring rectangles.

71

 

.—¥
_1,

.3 :1“

\

 

 

 

 

 

 

 

 

 

Figure 3-1. (a) Structure of CrB-type MnB with boron atom chains parallel to page.
The boron atom chains are perpendicular to the page in (b) F eB-type MnB and (c)
CrB-type MnB.

72

....d .1
.%

 

 

-axis (b) Sheet of RuB

-2. (a) The structure of n

An alternate way to consider the structure is in terms of layers. From this
perspective, the structure is made of puckered hexagonal layers of boron atoms, with
layers of manganese atoms centering the boron hexagons. The boron layers are puckered
in the boat-type conﬁrmation, which is very similar to the boron layers found in RuB2
(Figure 3-2b). The interlayer bonds, which would need to break to completely reduce the
M., structure into separate RuB2-type layers, are longer than the intralayer bonds by
about 8%.

ReB2: Rhenium diboride has a structure, seen in Figure 3-3a, built along the
c-axis by puckered hexagonal layers of boron atoms and metal atom layers where
rhenium atoms sit directly above the boron atoms in the l, 3, and 5 positions of the
hexagon in one layer and above the atoms in the 2, 4, and 6 positions in the next boron
layer. The layers in ReB2 (Figure 3-3b) display hexagons in a chair conformation, in
contrast to the planar hexagons in AlB2-type structures and the boat conformation for
boron hexagons in RuB2-type structures. The distance between rhenium atoms in the
c-direction is larger than the bonding distance for rhenium—rhenium bonds. Similarly, the
distances between successive boron layers are larger than boron—boron bonding
distances, suggesting that the structure is held together along the c-axis only through
rhenium—boron bonds.

HoMnB4 and REReB4 (RE = Dy, Er, Yb): The group 7 ternary borides we have
synthesized all crystallize in the YCrB4 structure type. This structure type (Figure 3-4a)
has an interesting 2-D network of boron atoms. The boron atoms form both 5- and 7-
membered rings that sit in one planar layer. The metal atoms in the structure reside in

their own layer, halfway between boron layers. The rare earth atoms sit below the

74

 

 

 

 

 

 

 

 

 

 

(b)

Figure 3-3. (a) Structure of ReB2 and (b) an individual graphite-like layer of boron
atoms.

75

 

 

 

 

 

Figure 3-4. (a) Structure of YCrB4-type compounds (b) View down a-axis showing
alternating layers of metal atoms and boron atoms.

76

centers of the 7-membered rings, while the transition metals sit below the centers of the
5-membered rings.

Generally, when the B:M ratio reaches 2.0, a layered structure results. The boron
atoms in these layers typically have a hexagonal arrangement, planar or puckered
conformation depending on the metal. The rare earth atoms and transition metals in the
YCrB4 structure type have such different radii that the rings must also be of different
sizes. The boron atoms accommodate the metals by moving from 6-membered rings to

the 5- and 7-membered rings.

E. Results and Discussion

All of the compounds described in this chapter have been previously synthesized
using traditional solid state methods. However, single crystals have not been previously
isolated for all of these compounds. Our group has used the molten metal ﬂux technique
in the synthesis and crystallization of all these compounds. Gallium was the only molten
metal from which the group 7 borides would grow. Reactions using aluminum, copper,
indium, or tin as the ﬂux metal did not yield borides as a product. The products from
these reactions were generally binary phases of either manganese or rhenium and the ﬂux
metal. Our success using a gallium ﬂux should be seen as a starting point for further
research, because our synthesis procedures were not optimized.

MnB: One of the main problems encountered in the synthesis of MnB was in
picking the crystals out from the entire reaction product. The majority of the product
from the initial MnB reaction was actually MnSi. The alumina crucibles used for this

reaction had a cement bottom and we concluded that the gallium ﬂux pulled silicon out of

77

the cement and into the reaction. The crystals for MnB and MnSi look nearly identical,
so EDS on the crystal is necessary to ensure that only manganese and boron are present in
the crystal. Reactions with identical conditions were carried out again, except in purely
alumina crucibles. These reactions produced some crystals, but still in low yields. In
these reactions, the major products were manganese gallides, which will be discussed in a
later chapter of this work. Crystals of the gallides can also look like the MnB phase, so
again it was difficult to separate boride crystals from the product mixtures.

MnB4: The crystals of M4, while easier to identify, do not form single
crystals. We believe that the crystals we have isolated are a series of thin plates stacked
atop each other. This stacking leads to the ridged appearance of the crystals (Figure 3-5).
For the “single crystal” diffraction study a specimen with the smoothest faces possible
was selected. The most intense peaks were used to index the unit cell. While this
technique was in no way an ideal method, the results do give some structural insight.

As noted, the crystal were indexed in an orthorhombic cell to match CrB4, rather
than the monoclinic cell used in previous studies on MnB.1.2"22 The previous studies
found a 03° deviation from a 90° angle for the B angle when the cell was transformed to
the orthorhombic 1 setting. In our crystal, the largest deviation from 90° was
approximately 015°, changing slightly when different orientations of the cell were used.
These ﬁndings are inconclusive because of the aforementioned imperfections of the
specimen. The quality of the data from previous studies has also been questioned.1
Further research is needed to see whether MnB4 really crystallizes in the orthorhombic 1

setting, or if our metal ﬂux synthesis is trapping a metastable phase of the compound.

78

20 1,1111

 

Figure 3-6. SEM micrograph of a cluster of ReBz crystals.

79

ReB2: At a lower ﬂux reaction temperature (lOOO°C) ReB2 crystals are thin
hexagonal plates. The plates turn into hexagonal prisms at a higher ﬂux reaction
temperature (1400°C), with the prisms narrowing toward each end. The hexagonal plates
will often grow together, forming clusters of plates (Figure 3-6). Single plates are often
available, but they can be manually broken from these clusters if needed. An optimized
synthesis would help increase yields of the plates, making physical property
measurements easier. Property measurements have not been carried out because of the
time consuming separation of enough crystals for measurement.

HoMnB4: The synthesis of HoMnB4 is very similar to W., and faces similar
obstacles to single crystal isolation. Crystals of HoMnB4 (Figure 3-7) are thin plates and
appear to also have the ridges seen in the WB., crystals. The diffraction data did not
show any obvious signs of twinned or intergrown crystals, but the quality of the
reﬁnement data is not as high as is expected for a true single crystal. Reactions
speciﬁcally targeting this compound are needed to obtain better crystals, not only for
better diffraction studies, but also for other physical properties measurements, especially
magnetic measurements.

REReB4 (RE = Dy, Er, Yb): The ternary rhenium borides were ﬁrst seen in
reactions heated to 1000°C, but the crystals were very small and poorly formed.
Reactions were repeated at 1400°C and the crystals grew larger and better formed despite
shorter soaking and cooling times. A crystal of DyReB4 is shown in Figure 3-8.
Reactions including other rare earth elements yielded similar looking products, but a

short scan on the single crystal machine revealed that the rare earth tetraboride had

80

 

7. SEM micrograph of HoMnB4.

Figure 3

 

Figure 3-8. SEM micrograph of DyReB4.

81

formed. Rhenium diboride and rhenium gallides, which are discussed later in this work,
also crystallized from these reactions.

The reﬁnements on the single crystals of these ternary rhenium borides are
consistently accurate. Problems do arise are in the displacement parameters, both
isotropic and anisotropic. These problems were a bit unexpected since the crystal faces
were measured, indexed, and then used to make an analytical absorption correction. Low
temperature diffraction studies may be needed to reduce the thermal motion of the atoms.

General Discussion: The gallium ﬂux is a useful medium in the synthesis of
group 7 borides. The exploratory nature of our research in this group changed after
ﬁnding MnB and M.,, focusing more on repeating and optimizing reactions aimed at
these products. While some improvements were made, there is still a great deal of
synthesis reﬁnement to be done. Further research should also be carried out, searching
for the “missing” Mnng and Mn2B3 phases mentioned in the introduction to this chapter.
Ternary systems with the rare earths should also continue to yield results. Even if the
structures have been reported before, very few have had any physical properties
measured. The large crystals grown from molten gallium would be ideal specimens to
investigate.

The only binary rhenium boride we focused on was ReB2, so further research into
this system needs to consider Re3B and Re7B3. Exploratory work should also be carried
out to see if the ﬂux can isolate other phases that traditional syntheses can not. This work
should also carry-over into the synthesis of ternary systems, especially since it has been
successfully performed with rare earth metals. As with manganese, compounds have

been structurally characterized, but their physical properties remain unmeasured.

82

F. Conclusions

In this chapter we successfully used a molten gallium as a solvent to synthesize
and grow crystals of MnB, MnB4, ReB2, HoMnB4, DyReB4, ErReB4, and YbReB4.
While the structural characterization of some compounds needs additional work, the
overall picture is encouraging. The synthesis of ternary borides is especially
encouraging, since the technique was unsuccessful with group 6 borides. These results
set a foundation for continued research into group 7 borides using the metal ﬂux

technique.

83

References

1. Burdett, J. K.; Canadell, E., GB, and MnB4: Electronic Structures of Two Unusual
Systems Containing the Tetragonal Carbon Net. Inorganic Chemistry 1988, 27, 4437-
4444.

2. Kanaizuka, T., Phase Diagram of Pseudobinary CrB-MnB and MnB-FeB Systems:
Crystal Structure of the Low Temperature Modiﬁcation of FeB. Journal of Solid State
Chemistry 1982, 41, 195-204.

3. Papesch, G.; Nowotny, H.; Benesovsky, F., Untersuchungen in den Systemen: Chrom-
Bor-Kohlenstoff, Mangan-Bor-Kohlenstoff und Mangan-Germanium-Kohlenstoff.
Monatsheftefur Chemie 1973, 104, 933-942.

4. Becher, H. J .; Krogrnann, K.; Peisker, E., Uber das terniire Borid Mn2AlB2. Zeitschrift
ﬁlr anorganische und allgemeine Chemie 1966, 344, (3-4), 140-147.

5. Okada, S.; Atoda, T.; Higashi, 1., Structural Investigation of Cr2B3, Cr3B4, and CrB by
Single-Crystal Diffractometry. Journal of Solid State Chemistry 1987, 68, 61-67.

6. Okada, S.; Atoda, T.; Higashi, 1.; Takahashi, Y., Preparation of Single Crystals of
MoB2 by the Aluminum-ﬂux Technique and Some of Their Properties. Journal of
Materials Science 1987, 22, 2993-2999.

7. Higashi, 1.; Takahashi, Y.; Atoda, T., Crystal Growth of Borides and Carbides of
Transition Metals from Molten Aluminum Solutions. Journal of Crystal Growth 1976,
33, 207-211.

8. Okada, S.; Kudou, K.; Lundstrﬁm, T., Preparations and Some Properties of W2B, 8-
WB, and WB2 Crystals from High-Temperature Metal Solutions. Japanese Journal of
Applied Physics 1995, 34, (1), 226-231.

9. Aronsson, B.; Backman, M.; Rundqvist, S., The Crystal Structure of Re3B. Acta
Chemica Scandinavica 1960, 14, 1001-1005.

10. Placa, S. L.; Post, B., The Crystal Structure of Rhenium Diboride. Acta
Crystallographica 1962, 15, 97-99.

11. Aronsson, B.; Stenberg, E.; Aselius, J ., Borides of Rhenium and the Platinum Metals.
Acta Chemica Scandinavica 1960, 14, 733-741.

12. Ishii, T.; Shimada, M.; Koizumi, M.; Sakakibara, T.; Date, M., Magnetic Properties
of Mn3B4 Under High Magnetic Field. Solid State Communications 1984, 51, (2), 103-
105.

13. Ishii, T.; Shimada, M.; Koizumi, M., Synthesis and Magnetic Properties of (Mn).
xCr,‘)3B4 and (Mn1-xMox)3B4. Inorganic Chemistry 1982, 21, 1670-1674.

84

14. Kawano, A.; Mizuta, Y.; Takagiwa, H.; Muranaka, T.; Akimitsu, J ., The
Superconductivity in Re-B System. Journal of the Physical Society of Japan 2003, 72,
(7), 1724-1728.

15. Strukova, G. K.; Degtyareva, V. F.; Shovkun, D. V.; Zverev, V. N.; Kiiko, V. M.;
Ionov, A. M.; Chaika, A. N., Superconductivity in the Re-B System. eprint arXiv:cond-
mat/0105293 2001.

16. Sheldrick, G. M. SAINT, Version 4; Siemens Analytical X-Ray Instruments, Inc.:
Madison, WI.

17. Sheldrick, G. M. SHELXT L Structure Determination Program, Version 5.0; Siemens
Analytical X-Ray Instruments, Inc.: Madison, WI, 1995.

18. X-SHAPE, Version 2.05; STOE & Cie GmbH: 2004.
19. X-RED 32, Version 1.10; STOE & Cie GmbH: 2004.

20. Andersson, S.; Lundstrdm, T., The Crystal Structure of CrB4. Acta Chemica
Scandinavica 1968, 22, 3103-3110.

21. Andersson, S., A Note on the Crystal Structure of mm. Acta Chemica
Scandinavica 1969, 23, (2), 687-688.

22. Andersson, S.; Carlsson, J .-O., The Crystal Structure of M.,. Acta Chemica
Scandinavica 1970, 24, (5), 1791-1799.

23. Kuz'ma., Y. B., Crystal Structure of the compound YCrB4 and its analogs.
Kristallograﬁya 1970, 15, 372-374.

85

CHAPTER FOUR

The Metal Flux Synthesis of Ruthenium Borides

A. Introduction

The structure and composition of many binary borides have been reported for the
past half-century. However, few of the physical properties of these compounds have ever
been reported. The discovery of superconductivity in MgB2 in 20011 has renewed
interest in binary borides of all kinds. This ﬁnding led to a study of group 8 borides by
the Kaner group in 2005, who ﬁrst reported on the hardness and incompressibility of
osmium diboride.2 While the structures of OsB2 and RuB2 were ﬁrst reported in 1962,3
property studies have been limited to a single report on their superconductivity in 19754
and a few reports on the electronic structures of the compoundss'7 in the past decade.
Chapnik8 actually called for further study on these unique compounds, and we were
interested in preparing single crystals to investigate the unknown physical properties of
group 8 binary borides.

Physical measurements carried out on single crystals are advantageous in
eliminating minor phases in powdered samples and allow researchers to search for
anisotropy in the properties of the compound. Single crystals of the ruthenium borides
have been isolated from arc-melted pellets, but we were interested in using the molten
metal ﬂux technique to grow higher quality crystals. Our success in synthesizing and

crystallizing borides from groups 6 and 7 and reports of ternary rhodium borides grown

86

from ﬂux9 led us to believe that the technique could also work for ruthenium borides.
The reports of ternary rhodium borides from ﬂux synthesis also inspired us to investigate
ternary ruthenium borides as well.

In this chapter we report on the synthesis and crystal growth of the known
compounds RU7B3, RU2B3, RuB2, and RERu4B4 (RE = Y, Ce, Sm, Dy, and Yb). A new
phase of ruthenium diboride, B-RuB2, crystallizing in the ReB2-type structure is
presented. A brief report on two ternary borides, Pr,,(Ru4B4)y and Ndx(Ru4B4)y, with

modulated structures follows as well.

B. Experimental

Reagents:

All reagents were used as received without further puriﬁcation: (i) Boron metal,
99% purity, -325 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (ii) Ruthenium
metal, 99.95% pruity, -325 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (iii)
Copper shot, 99% purity, -3 +14 mesh, Aldrich Chemical Company, Milwaukee, WI, (iv)
Yttrium metal, 99.9% purity, -40 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (v)
Dysprosium metal (ﬁled from ingot) 99.9% purity, Chinese Rare Earth Information
Center, Inner Mongolia, China, (vi) Cerium metal (ﬁled from ingot) 99.9% purity,
Chinese Rare Earth Information Center, Inner Mongolia, China, (vii) Samarium metal
(ﬁled from ingot) 99.9% purity, Chinese Rare Earth Information Center, Inner Mongolia,
China, (viii) Ytterbium metal (ﬁled from ingot) 99.9% purity, Chinese Rare Earth
Information Center, Inner Mongolia, China, (ix) Erbium metal (ﬁled from ingot) 99.9%

purity, Chinese Rare Earth Information Center, Inner Mongolia, China, (x) Lanthanum

87

metal (ﬁled from ingot) 99.9% purity, Chinese Rare Earth Information Center, Inner
Mongolia, China, (xi) Erbium metal (ﬁled from ingot) 99.9% purity, Chinese Rare Earth
Information Center, Inner Mongolia, China, (xii) Praseodymiurn metal (ﬁled from ingot)
99.9% purity, Chinese Rare Earth Information Center, Inner Mongolia, China, (xiii)
Neodymium metal (ﬁled from ingot) 99.9% purity, Chinese Rare Earth Information
Center, Inner Mongolia, China, (xiv) Holmium metal (ﬁled from ingot) 99.9% purity,
Chinese Rare Earth Information Center, Inner Mongolia, China, (xv) Boron metal, 99.9%
purity, typically 1 um or less, Cerac Specialty Inorganics, Milwaukee, WI.

Furnaces:

A Lindberg Blue high-temperature, horizontal tube furnace was used with a
continuous argon ﬂow of approximately 0.1 L/min. Heating proﬁles for individual
reactions were programmed into the accompanying Lindberg UP150 fumace controller.
Thermocouples connected to the furnace controller were used to monitor fumace
temperatures during reactions. The maximum operating temperature for this furnace was
1500 °C.

A radio frequency (RF) furnace was constructed from an Ameritherm XP7.5CE
Induction Heater and fused quartz reaction chamber. A molybdenum cylinder, machined
to act as a reaction crucible holder, was used as a susceptor. A boron nitride cylinder was
machined to hold the molybdenum crucible in the center of the reaction chamber. Details
on the RF fumace setup are given in the appendix of this work.

RF furnace temperatures were controlled manually using the Ameritherm control
unit. Reaction temperatures were monitored using an Ircon SA Series Infrared

Thermometer. During the course of some reactions, ﬁlms deposited inside the reaction

88

chamber would prevent the IR thermometer ﬁ'om accurately reporting furnace
temperatures. In such circumstances, reaction temperatures were approximated based on
power settings for the Ameritherm control unit and a series of calibration runs conducted
by heating only the molybdenum cylinder. Temperatures greater than 1700 °C were
easily reached with this furnace.

Synthesis:

Ru7B3, RuzB3, and RuB2: Boron and ruthenium powders were added in
millimolar quantities, with B:M ratios from 0.25 to 4, to alumina crucibles and topped
with 40 mmol of copper shot. The crucibles were capped individually with alumina lids
and then place into rectangular alumina boats. The boats were loaded into the mullite
tube of a commercial high temperature furnace preheated to 500 °C. The furnace was
then heated to 1450 °C in 10 hours, held for 48 hours at that temperature, and then cooled
to 1100 °C in 24 hours. Upon reaching 1100°C, the furnace was turned off, allowing the
products to cool to 500 °C (a preset hold temperature for the ﬁirnace), at which time the
alumina boats were removed.

B-RuB2: Boron (10 mmol) and ruthenium (1 mmol) powders were added to an
alumina crucible along with holmium (2 mmol) ﬁlings. Copper shot (40 mmol) was then
added to the crucible, as was an alumina lid. The RF furnace was evacuated under
vacuum and ﬂushed with N2. Heating was initiated with the RF fumace under dynamic
vacuum. Over 30 minutes the furnace temperature was raised to 1400 °C. After holding
for 30 minutes at 1400 °C, the power to the furnace was shutoff due to mechanical
problems. The furnace cooled to room temperature over the course of 1 hour, at which

time the reaction crucible was removed.

89

RERu4B4 (RE = Y, Ce, Sm, Dy, Yb): Boron (1—4 mmol) and ruthenium
(1 mmol) powders and rare earth metal (0.5—l mmol) ﬁlings were added to alumina
crucibles. After adding copper shot (40 mmol) and alumina lids to the crucibles, the
crucibles were placed in the RF furnace. The RF furnace was evacuated under vacuum
and ﬂushed with N2. Heating was initiated with the RF furnace under dynamic vacuum.
Over 30 minutes the furnace temperature was raised to 1350 °C or 1400 °C. After
holding for 60 minutes at the maximum temperature, the power to the furnace was
shutoff and the furnace was allowed to cool to room temperature.

RE,,(Ru..B.1)y (RE = Pr, Nd): Boron (3 mmol), ruthenium (1 mmol), and the rare
earth metal (2.5 mmol) were added to an alumina crucible. After adding copper shot
(40 mmol) and lids, the crucibles were placed in an alumina boat. The boat was loaded
into the mullite tube of a commercial high temperature furnace preheated to 500 °C. The
furnace was then heated to 1450 °C in 8 hours, held for 33 hours at that temperature, and
then the furnace was turned off, allowing the products to cool to 500 °C (a preset hold
temperature for the furnace), at which time the alumina boat was removed.

Isolation from flux:

When fully cooled to room temperature, crucibles were placed in individual baths
of half-concentrated HNO3 in order to remove the copper ﬂux. An overnight soak was
used to completely remove the copper ﬂux, however, many crystals were removed from
the ﬂux after a 2-3 hour soak. Extended soaking in the bath was not found to damage the
crystals. The crystals were ﬁltered from the acid bath, washed with H20, and dried with
ethanol. The well-faceted crystals were easily distinguishable from the spherical

elemental ruthenium that remained in most reactions.

90

The products from reactions aimed at binary borides all contained unreacted
elemental ruthenium, with the amount of residual ruthenium decreasing as the B:M ratio
increased. The yield of Ru7B3 from reactions with a B:M ratio of 0.25, 0.33, 0.5, and 1.0
was between 70% and 80%, based on boron. Ru7B3 was found as a minor product in the
reactions with B:M ratios of 2.0, 3.0, and 4.0. RU2B3 was the major product from the
B:M =2.0 reaction, with an estimated yield of 60%, based on ruthenium. RuB2 was a
minor product in the B:M = 2.0 reaction, but was the major product in the B:M = 3.0 and
4.0 reactions. The yield of RuB2 was approximately 75% in the B:M = 3.0 reaction and
nearly 90% in the B:M = 4.0 reaction, based on ruthenium.

The new B-RuB2 phase was collected in an estimated 25% yield, based on
ruthenium. The ordinary RuB2 phase was present in approximately the same ratio. The
remainder of the collected product was unreacted ruthenium and boron.

The RERu4B4 compounds were collected in an estimated 50% yield, based on
ruthenium. The side products in the reactions producing these compounds were RERuB4,
traces of REB5 and Ru7B3, and elemental ruthenium. YRu3B2 was a minor side product

in the reaction producing YRu484.

C. Physical Measurements

Energy Dispersive Spectroscopy:

Energy dispersive spectroscopy (EDS) analyses were carried out on selected
crystals to determine their chemical composition. The analyses were performed with a
JEOL ISM-6400 scanning electron microscope (SEM) equipped with Oxford Inca

Energy Dispersive Spectroscopy (SizLi) detector and Norvar window for standardless

91

quantization of elements with Z 2 4. The crystals were afﬁxed to an alumina sample
holder with double-sided carbon tape. The EDS data were acquired at an accelerating
voltage of 10kV with a 30 to 60 second accumulation time.

Single-Crystal X-ray Crystallography:

Single crystals were mounted on a glass ﬁber with super glue and their intensity
data were collected on either a Bruker SMART platform CCD diffractometer, or a STOE
IPDS Il diffractometer using Mo K01 radiation at 50 kV and 40 mA.

Individual frames of R0733, Ru2B3,RERu4B4 (RE = Y, Ce, Sm, Yb) and
RE,,(Ru4B4)y (RE = Pr, Nd) examined with the Bruker machine were collected with a 03°
0) rotation. The SMART software was used for data collection, and SAINT software was
used for data extraction and reduction. After applying analytical absorption corrections,
structure solution and reﬁnement were completed using direct methods and the
SHELXTL suite of programs.

Samples examined with the STOE (RuB2, and DyRu4B4) had individual ﬁames
collected on a 34 cm image plate with a 60 second exposure time and a 10° to rotation.
The X-SHAPE and X-RED-32 software packages were used for data extraction and
reduction and to apply an analytical absorption correction. Direct methods and the
SHELXTL suite were used to solve and reﬁne the structures.

The crystal structure reﬁnement data for Ru7B3 are found in Table 4-1.
Previously reported10 atomic parameters were used as a starting point for the reﬁnement
of the structure. The atomic positions and isotropic displacement parameters are listed in
Table 4-2. Anisotropic displacement parameters for the atoms are found in Table 4-3.

Selected bond lengths in RU7B3 and a count of those bonds are listed in Table 4-4.

92

The crystal structure reﬁnement data for Ru2B3 are found in Table 4-5.
Previously reported” atomic parameters were used as a starting point for the reﬁnement
of the structure. The atomic positions and isotropic displacement parameters are listed in
Table 4-6. Anisotropic displacement parameters for the atoms are found in Table 4-7.
Selected bond lengths in Ru2B3 and a count of those bonds are listed in Table 4-8.

The crystal structure reﬁnement data for RuB2 are found in Table 4-9. Previously
reported3 atomic parameters were used as a starting point for the reﬁnement of the
structure. The atomic positions and isotropic displacement parameters are listed in Table
4-10. Anisotropic displacement parameters for the atoms are found in Table 4-11.
Selected bond lengths in Ru2B3 and a count of those bonds are listed in Table 4-12.

The crystal structure reﬁnement data for B-RuB2 are found in Table 4-13. Atomic
parameters for ReB2l2 were used as a starting point for the reﬁnement of the structure.
The atomic positions and isotropic displacement parameters are listed in Table 4-14.
Anisotropic displacement parameters for the atoms are found in Table 4-15. Selected
bond lengths in Ru2B3 and a count of those bonds are listed in Table 4-16.

The crystal structure reﬁnement data for YRu4B4 and CeRu4B4 are found in Table
4-17, in Table 4-18 for SmRu484 and DyRu4B4, and in Table 4-19 for YbRu434.
Previously reported atomic parameters for LuRu4B4l3 were used as a starting point for the
reﬁnement of the structure. The atomic positions and isotropic displacement parameters
are listed in Table 4-20. Anisotropic displacement parameters for the atoms are found in
Table 4-21. The bond distances for RERu4B4 (RE = Y, Ce, Sm, Dy, Yb) are listed in

Table 4-22.

93

Initial indexing of the single crystal data for RE,((Ru.1B4)y (RE = Pr, Nd):
suggested that these compounds crystallized with the LaRmB4 structure.14 However,
during reﬁnement, it became obvious that this suggestion was incorrect. When examining
the reciprocal lattice, two distinct lattices became visible along the c-axis. One lattice
corresponds to a c-axis length of 3.67 A and the other corresponds to a length of 4.04 A.
While we found no reports of incommensurate ruthenium borides, several reports on
15, 16

incommensurate iron borides with similar cell (and substructure) parameters exist.

Further data collection is necessary for complete structural characterization.

94

Table 4-1. Crystal data and structure reﬁnement for Ru7B3.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 28.23°
Reﬁnement method

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

Final R indices [I>2sigma(1)]
R indices (all data)

Absolute structure parameter
Extinction coefﬁcient

Largest diff. peak and hole

739.92

293(2) K

0.71073 A

Hexagonal

P63mc

a = 7.4570(11) A or: 900
b = 7.4570(11) A B= 90°
c = 4.7140(9) A y= 120o
227.01(6) A3

2

10.825 g/cm3

22.466 mm“

646

0.052 x 0.052 x 0.326 mm3

3.15 to 28.23°
-9ShS9,-9_<_I(S9,-6_<_IS6
2328

226 [R(int) = 0.0316]

99.2 %

F ull-matrix least-squares on F 2
226/ 1 / 23

1.273

R1 = 0.0116, WR2 = 0.0288

R1 = 0.0116, WR2 = 0.0288
0.1(4)

0.0254(8)

0.512 and -0.573 e.A'3

 

R1 = 2 MRI — chll /2 IFol and wR2 = [z (1F.2 - F31 )2 / 5; (WF02)2]V2

95

Table 4-2. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for RU7B3.

 

 

x y z U(eq)
Ru( 1) 3333 6667 9527(3) 6(1)
Ru(2) 1226(1) 8774(1) 7754(1) 5(1)
Ru(3) 5429(1) 4571(1) 9623(1) 7(1)
B(l) 1891(6) 8109(6) 11911(l8) 9(2)

 

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

Table 4-3. Anisotropic displacement parameters (Azx 103) for Ru7B3.

 

 

ull U22 [)33 U23 ul3 ulz
Ru(l) 6(1) 6(1) 6(1) 0 0 3(1)
Ru(2) 4(1) 4(1) 6(1) 0(1) 0(1) 2(1)
Ru(3) 7(1) 7(1) 7(1) 0(1) 0(1) 3(1)
13(1) 6(2) 6(2) 12(4) 1(2) -1(2) -1(3)

 

The anisotropic displacement factor exponent takes the form: -21t2[ h2 a"‘2Ull + + 2 h
k a” b" U12 ]

96

Table 4-4. Selected bond distances (A) for Ru7B3.
Rul Bl 3x 2.1756(2)
Ru2 B1 1x 2.1395(3)
Ru2 B1 2x 2.1825(2)
Ru3 B 1 2x 2.1637(2)
Rul Ru3 3x 2.7075(4)
Rul Ru3 3x 2.8105(4)
Rul Ru2 3x 2.8472(3)
Rul Ru3 3x 2.8852(4)
Ru2 Ru2 2x 2.7427(4)
Ru2 Ru3 2x 2.7913(3)
Ru2 Ru2 4x 2.8395(4)
Ru2 Rul 1x 2.8472(3)
Ru2 Ru3 2x 2.8539(4)
Ru3 Ru3 2x 2.6049(4)
Ru3 Rul lx 2.7075(3)
Ru3 Ru3 2x 2.7680(4)
Ru3 Ru2 2x 2.7913(3)
Ru3 Rul 1x 2.8105(4)
Ru3 Ru2 2x 2.8539(4)

Ru3 Rul 1x 2.8852(4)

97

Table 4-5. Crystal data and structure reﬁnement for Ru2B3.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 27.76°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F 2

Final R indices [I>23igma(l)]
R indices (all data)

Extinction coefﬁcient

Largest diff. peak and hole

234.57

293(2) K

0.71073 A

Hexagonal

P63/mmc

a = 2.9029(4) A or= 90°
b = 2.9029(4) A 8: 90°
c=12.814(3)A y=120°
93.51(3) A3

2

8.331 g/cm3

15.613 mm'1

206

0.038 x 0.084 x 0.084 mm3

3.18 to 27.76°
-3ShS3,-3Sk£3,-16SIS 16
899

65 [R(int) = 0.0327]

95.4 %

Full-matrix least-squares on F2
65 / 0/ 10

1.211

R1 = 0.0148, WR2 = 0.0353

R1 = 0.0163, WR2 = 0.0366
0.169(13)

0.751 and -0.503 e.A'3

 

RI = z llFol — chll / z W and we = [2 (IF: — 18.21 )2 / 2 maﬁa”

98

Table 4-6. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (Azx 103) for Ru2B3.

 

 

X y Z U(CQ)
Ru(l) 6667 3333 1400(1) 3(1)
B(l) 6667 3333 9697(7) 4(2)
B(2) 6667 3333 7500 5(3)

 

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

Table 4-7. Anisotropic displacement parameters (Azx 103) for Ru2B3.

 

 

ull U22 u33 U23 ul3 ul2
Ru(l) 3(1) 3(1) 1(1) 0 0 2(1)
8(1) 5(3) 5(3) 3(4) 0 0 2(1)
B(2) 5(4) 5(4) 5(6) 0 0 2(2)

 

The anisotropic displacement factor exponent takes the form: -21t2[ h2 a*2U1 ‘ + + 2 h
k a* b* U12 ]
Table 4-8. Selected bond distances (A) for Ru2B3.

Rul 131 1x 2.1821(4)

Rul Bl 3x 2.1871(3)

Rul BZ 3x 2.1901(3)

B1 Bl 3x 1.8474(2)

Bl Rul 1x 2.1821(4)

Bl Rul 3X 2.1871(3)

82 Rul 6x 2.1901(3)

99

Table 4-9. Crystal data and structure reﬁnement for RuB2.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 36.3 8°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F7-

Final R indices [I>28igma(1)]
R indices (all data)

Extinction coefﬁcient

Largest diff. peak and hole

122.69

293(2) K

0.71073 A

Orthorhombic

Pmmn

a = 4.6322(9) A or= 90°
b = 2.8542(6) A 8: 90°
c = 4.0292(8) A y = 90°
53.271(19) A3

2

7.649 g/cm3

13.716 mrn'l

108

0.031 x 0.040 x 0.048 mm3

5.06 to 36.38°
-7ShS7,-4$k$4,-6SIS6
863

165 [R(int) = 0.0595]

100.0 %

F ull-matrix least-squares on F2
165 / 0/ 12

1.210

R1 = 0.0157, WR2 = 0.0341

R1 = 0.0165, WR2 = 0.0344
0.200(16)

1.204 and -1.l41e.A'3

 

R1 = 2 “FA — 1F.” / z |F0| and wR2 = [z (IF,2 — 18.21 )2 / z (wF,2)2]'”

100

Table 4-10. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (Azx 103) for RuB2.

 

 

x y 2 U(CCI)
Ru(l) 2500 2500 6512(1) 1(1)
3(1) 5550(8) -2500 8615(6) 4(1)

 

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

Table 4-11. Anisotropic displacement parameters (Azx 103) for RuB2.

 

 

U” U22 U33 U23 U13 U12
Ru(l) 1(1) 1(1) 1(1) 0 0 0
B(l) 4(1) 4(1) 3(1) 0 0(1) 0

 

The anisotropic displacement factor exponent takes the form: -21t2[ h7- a*2U” + + 2 h
k a* b* 1112 ]

Table 4-12. Selected bond distances (A) for RuB2.

Rul 81 2x 2.1612(4)
Rul 81 4x 2.1796(3)
Rul 81 2x 2.2547(4)
81 81 1x 1.8067(4)
81 81 2x 1.8818(3)
81 Rul 1x 2.1612(4)
81 Rul 2x 2.1796(3)
81 Rul 1x 2.2547(4)

101

Table 4-13. Crystal data and structure reﬁnement for B-RuB2.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 27.73°
Reﬁnement method

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

Final R indices [I>2sigma(1)]
R indices (all data)

Extinction coefﬁcient

Largest diff. peak and hole

122.69
273(2) K

0.71073 A

Hexagonal

P63/mmc

a = 2.9142(4) A or= 900
b = 2.9142(4) A 0: 90°
C = 7.283(2) A y = 120°
53.56509) A3

2

7.607 g/cm3

13.640 mm"

108

0.008 x 0.084 x 0.090 mm3

5.60 to 27.73°
-3Sh$3,-3Sk$3,-9SIS9
475

38 [R(int) = 0.0276]

94.7 °/o

Full-matrix least-squares on F 2
38 / 0 / 7

0.936

R1 = 0.0239, WR2 = 0.0820

R1 = 0.0276, WR2 = 0.0880
007(4)

0.789 and -1.243 e.A'3

 

R1 = 2 “F01 — chlI / Z W and wR2 = [2 (IF,2 — Fﬂ )2 / 2 008,621"2

102

Table 4-14. Atomic coordinates (x 104) and equivalent isotropic displacement
parameters (A2x 103) for B-RuB2.

 

 

x y z U(eq)
Ru(l) 3333 6667 7500 5(1)
B(2) 6667 3333 5590(30) 8(4)

 

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

Table 4-15. Anisotropic displacement parameters (Azx 103) for B-RuB2.

 

 

ull U22 1133 U23 ul3 ulZ
Ru(l) 4(1) 4(1) 6(1) 0 0 2(1)
B(2) 8(7) 8(7) 7(8) 0 0 4(3)

 

The anisotropic displacement factor exponent takes the form: -21t2[ h2 a*2U” + + 2 h
k a* 8* U12 ]

Table 4-16. Selected bond distances (A) for B-RuB2.

Rul B2 6x 2.1838(3)
Rul B2 2x 2.2493(7)
B2 B2 3x 1.8882(2)
B2 Rul 3x 2.1838(3)
BZ Rul 1x 2.2493(7)

103

Table 4-17. Crystal data and structure reﬁnement for RERu4B4 (RE= Y, Ce).

Formula
Formula weight
Temperature
Crystal system
Space group

Unit cell dimensions

Volume

Z, Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size, mm

Theta range for data collection

Index ranges

Reﬂections collected
Independent reﬂections
Completeness to 0max
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F 2

Final R indices [I>20(I)]

R indices (all data)

Extinction coefficient

Largest diff. peak & hole

YRu4B4

536.43

297(2) K

Tetragonal

141/acd

a = 7.4525(14) A

c = 14.988(6) A
832.4(4) A3

8, 8.560 g/cm3
27.561 mm'1

1880

0.050 x 0.150 x 0.150
4.73 to 28.3l°

-9 S h _<_ 9.

-9 S k S 9.

-18 S l S 19

3793

253 [R(int) = 0.0413]
96.9 %

CeRu4B4

587.64

297(2) K

Tetragonal

141/acd

a = 7.4744(4) A

c = 15.0642(19) A
841 .59(12) A3

8, 9.276 g/cm3
24.417 mm'1

2032

0.028 x 0.066 x 0.066
4.71 to 28.30°

-9 S h S 9.

-9 S k S 9.

-19 S l S 18

3898

257 [R(int) = 0.0317]
97.0 %

F ull-matrix least-squares on F2

253 / 0 / 23
1.337

R1 = 0.0271,
wR2 = 0.1355
R1 = 0.0274,
wR2 = 0.1358
0.0053(8)

2.084 & -l .244 e.A'3

257 / 0 / 23

1.013

R1 = 0.0185,

wR2 = 0.0599

R1 = 0.0193,

wR2 = 0.0604
0.00236(18)

1.106 & -0.916 e.A'3

 

RI = z 181 — chll / z 181 and sz = [z (“.02 - r121 )2 I 2: 012.2121“

104

Table 4-18. Crystal data and structure reﬁnement for RERu4B4 (RE= Sm, Dy).

Formula
Formula weight
Temperature
Crystal system
Space group

Unit cell dimensions

Volume

Z, Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size, mm

Theta range for data collection

Index ranges

Reﬂections collected
Independent reﬂections
Completeness to emax
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 & hole

SmRu4B4

597.87

297(2) K

Tetragonal

I4./acd

a = 7.4801(4) A

c = 15.0375(l8) A
841.38(12) A3

8, 9.440 g/cm3
27.761 mm‘1

2064

0.050 x 0.150 x 0.150
4.71 to 28.97°

-10 S h S 10,

-10 S k S 9.

-20 S l S 19

4297

278[R(int) = 0.1325]
97.5 %

DyRu4B4

610.02

100(2) K

Tetragonal

141/acd

a = 7.4195(10) A

c = 14.930(3) A
821.9(2) A3

8, 9.860 g/cm3
32.109 mm"1

2096

0.070 x 0.079 x 0.080
4.75 to 36.59°

-12 S h _<_ 12,

-9 S k S 12,

-24 S l S 20

6128

511 [R(int) = 0.0760]
99.0 %

F ull-matrix least-squares on F2

278 / 0 / 23

1.202

R] =0.0304,

wR2 =0.0928

R1 = 0.0304,

wR2 = 0.0928
0.0038(4)

2.288 & -2.497 e.A‘3

511 / 0 / 23

1.317

R1 = 0.0450,

wR2 = 0.1397

R1 = 0.0491,

wR2 = 0.1733
0.0039(6)

6.458 & -12.129 e.A'3

 

RI = 2 118.1 - 1F.” / z rm and wR2 = [>3 (IRE — FA )2 / 2 682121":

105

Table 4-19. Crystal data and structure reﬁnement for YbRu4B4.

Formula weight
Temperature
Wavelength
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F(000)

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 29. 15°
Reﬁnement method

Data / restraints / parameters
Goodness-of-ﬁt on F 2

Final R indices [I>2sigma(1)]
R indices (all data)

Extinction coefﬁcient

Largest diff. peak and hole

620.56

297(2) K

0.71073 A

Tetragonal

141/acd

a = 7.418500) A or= 90°
b = 7.4185(10) A 0= 90°
c = 14.9570) A y = 90°
823.2(2) A3

8

10.015 g/cm3

36.626 M'1

2128

4.75 to 29.15°
-10$h59,-9Sk59,-20$1$19
4234

272 [R(int) = 0.0369]

96.1 °/o

F ull-matrix least-squares on F 2
272 / 0 / 23

0.945

R1 = 0.0183, WR2 = 0.0425

R1 = 0.0188, WR2 = 0.0427
0.0054(2)

0.957 and -0.969 e.A'3

 

R1 = Z IIFol — IFcll / Z IE1 and wR2 = [z (IF,2 — F31 )2 / 2 (1111252)"2

106

Table 4-20. Atomic coordinates ( x 104) and equivalent isotropic displacement
parameters (A2x 103) for RERmB4 (RE = Y, Ce, Sm, Dy, Yb).

 

 

x y 2 U(CCI)
Y(1) 0 2500 1250 8(1)
Ru(l) -3519(1) 3654(1) 1867(1) 6(1)
B(l) -3329(12) 3920(13) 433(7) 9(2)
Ce(l) 0 2500 1250 4(1)
Ru(l) -3520(1) 1355(1) 1864(1) 3(1)
B(l) -3533(11) 4174(10) 2068(6) 5(1)
Sm(1) 0 2500 1250 2(1)
Ru(l) 3844(1) 1033(1) 633(1) 2(1)
3(1) 1675(14) -1058(13) 438(8) 3(2)
Dy(l) 0 2500 1250 5(1)
Ru(l) 3846(1) 3982(1) 632(1) 4(1)
B( l) 6662(15) 3889(15) 425(8) 4(2)
Yb(l) 0 2500 1250 3(1)
Ru(l) 3504(1) 1347(1) 1868(1) 2(1)
B(l) 1410(10) 836(11) 2941(6) 4(1)

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

107

Table 4-21. Anisotropic displacement parameters (Azx 103) for RERmB4 (RE = Y, Ce,

 

 

Sm, Dy, Yb).

ull U22 {)3 U23 [113 1112
Y(1) 8(1) 8(1) 8(1) 0 0 2(1)
Ru(l) 5(1) 6(1) 6(1) 0(1) -10) 0(1)
3(1) 0(4) 18(5) 9(4) 13(3) -8(4) 4(3)
Ce(1) 4(1) 4(1) 3(1) 0 0 -1(1)
Ru(l) 4(1) 3(1) 2(1) 0(1) -1(1) 0(1)
B( 1) 9(4) 1(3) 3(3) 0(3) -10(3) 1(3)
Sm(1) 3(1) 3(1) 0(1) 0 0 0(1)
Ru(l) 3(1) 3(1) 1(1) -10) 0(1) 0(1)
8(1) 3(4) 5(4) 1(3) 6(3) -1(3) 1(3)
Dy(l) 3(1) 3(1) 7(1) 0 0 1(1)
Ru(l) 3(1) 3(1) 5(1) 0(1) 0(1) 0(1)
B( 1) 5(4) 7(4) 0(4) 8(3) 3(3) 2(3)
Yb(1) 4(1) 4(1) 3(1) 0 0 1(1)
Ru(l) 3(1) 2(1) 1(1) 0(1) 1(1) 0(1)
3(1) 4(3) 3(3) 5(3) 3(3) 4(3) -4(3)

 

The anisotropic displacement factor exponent takes the form: -21t2[ h2 a"‘2Ull + + 2 h

ka“ 8* 1112]

108

Table 4-22. Bond distances (A) for RERu4B4 (RE = Y, Ce, Sm, Dy, Yb).

 

 

YRu484 CeRu4B4 SmRu484 DyRu4B4 YbRu4B4
RE—Ru 2.9109(9) 2.9170(6) 2.9315(7) 2.8981(9) 2.8884(9)
RE—B 2.962(7) 2.983(7) 2.973(10) 2.952(1 1) 2.9467(9)
3.006(10) 3.031(9) 3.029(12) 2.972(13) 3.0025(9)
Ru—B 2.127(10) 2.130(8) 2.138(10) 2.114(11) 2.1102(9)
2.163(11) 2.176(9) 2.165(12) 2.165(11) 2.1462(9)
2.216(10) 2.258(8) 2.229(10) 2.185(11) 2.2075(9)
2.273(10) 2.259(8) 2.273(1 1) 2.265(10) 2.2622(9)
2.269(10) 2.259(8) 2.279(1 l) 2.284(13) 2.2655(9)
Ru—Ru 2.7089(11) 2.7087(9) 2.7103(1 l) 2.6989(14) 2.7100(9)
2.7612(13) 2.7879(11) 2.7687(14) 2.7484(16) 2.7511(9)
2.7979(12) 2.7980(10) 2.7943(13) 2.7869(15) 2.8024(9)
2.9770(14) 2.9876(1 1) 2.9316(7) 2.8981(9) 2.9529(9)
3.1047(12) 3.1178(10) 3.1246(13) 3.0910(16) 3.0824(9)
B—B 1.79(2) 1.793(18) 1.81(2) 1.78(2) 1.810(10)

 

109

D. Structural Description

R0783: The structure of Ru7B3 (Figure 4-la) comprises ruthenium tetrahedra and
octahedra and isolated boron atoms.lo The structure can also be described with a more
complex building unit known as a stellaquadrangula.l7 The stellaquadrangula (Figure
4-1b) comprises a central tetrahedron sharing each face with another capping tetrahedron.
The central tetrahedron in the stellaquadrangula has a corner pointing into the page, and
the fourth capping tetrahedron has a corner coming out of the page. In Ru7B3, the
stellaquadrangula share comers with the octahedra to form a 3-D network of ruthenium
atoms. The boron atoms reside in cavities in the framework.

Ru2B3: A structural description of Ru283 was ﬁrst reported in 1962. The
structure (F igurer 4-2a) is made of alternating metal and boron atom layers. The metal
atoms are hexagonally close-packed. The boron atoms form two distinct layers, one with
puckered hexagonal rings (Figure 4-2b), and the other with a hexagonally packed layer of
non-bonding atoms (Figure 4-2c). The structure is quite different from the Cr2B3
structure reported in Chapter 2. In the Ru2B3 structure, the boron atoms form 2-D
networks instead of the triple chains seen in Cr2B3 and expected based on the B:M ratio
of 1.5. A previous report'8 on Ru2B3 suggests that the larger ruthenitun atoms would
force unreasonably short boron-boron bonds should the compound crystallize with the
Cr2B3 structure. The puckered layers in Ru2B3 distance the boron atoms from each other
sufﬁciently.

RuB2: The structure of RuB2, ﬁrst reported in 1962,3 contains corrugated layers
of boron atoms (Figure 4-3a). The corrugated boron rings are limited to RuB2 and OsB2

among diborides. The ruthenium atoms in the structure appear to be two separate layers,

110

 

 

 

 

(b)

Figure 4-1. (a) Structure of Ru7B3 and (b) the stellaquadrangula building unit.

111

‘Ru
.8

 

 

 

 

 

 

 

 

g \ B\2—B2 2.9029A
‘ O

 

 

(b) (c)

Figure 4-2. (a) Structure of Ru2B3 and the (b) puckered boron layer and (c) planar
layer of boron atoms.

112

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(b)

Figure 4-3. Structure of RuBz crystallizing with the (a) RuB2-type and (b) ReB2-
type structures.

113

as they alternately center boron rings above and below one metal atom layer. Recent
quantum mechanical studies suggest that the corrugated layers are a result of limited
electron donation from the ruthenium atoms to the boron layers.7

B-RuB2: This phase of RuB2 crystallizes in the ReB2-structure type12 described in
Chapter 3. In this newly discovered phase, the boron atoms are in puckered hexagonal
rings, with the “chair” conformation. In the known RuB2, the layers are more corrugated,
with the boron atoms sitting in the “boat” conformation of a hexagonal ring. B-RuB2,
shown in Figure 4-3b, crystallizes in the hexagonal space group P63/mmc, rather than the
orthorhombic Pmmn space group of the known RuB2.

RERu484 (RE= Y, Ce, Sm, Dy, Yb): All of these compounds crystallize in the
LuRmB4 structure type13 (Figure 4-4). The ruthenium atoms in these compounds form
tetrahedra that bond to each other in a zigzag fashion. A distorted NaCl lattice is formed
as the centers of the ruthenium tetrahedra reside on the anion sites and the rare earth

atoms reside on the cation sites. Boron atoms are found in pairs throughout the voids of

this distorted face-centered cubic lattice.

114

 

 

 

 

 

 

Figure 44. Structure of RERu4B4 (RE = Y, Ce, Sm, Dy, Yb) shown down the b-axis.

115

E. Results and Discussion

After initially synthesizing RuB2 from a copper ﬂux reaction in the RF ﬁrrnace,
we decided to do a study on the ruthenium borides that would grow from a copper ﬂux at
varying BzRu ratios. Seven reactions with varying BzRu ratios from 0.25 to 4.0. Ru7B3
crystallized from all metal rich reactions and traces of the compound were still found in
reactions with BzRu ratios up to 2.0. Ru2B3 was more limited, crystallizing only when
the BzRu ratio was 2.0. Crystals of RuB2 were found beginning in the reaction with a
BzRu ratio of 2.0 and became the major product for B:M = 3.0 and B:M = 4.0 reactions.
Elemental ruthenimn was found in all products, though the amount decreased as the BzRu
ratio increased.

Surprisingly, no Ru) (B3 or RuB 1,) crystals were found during the study.
Reactions targeted speciﬁcally at these reported phases yielded mostly Ru7B3 and Ru2B3.
Further research using more varied BzRu ratios and different heating proﬁles should be
conducted before the ﬂux synthesis of more ruthenium boride phases can be discounted.
To see further varying of B:M ratios is necessary, we need only look at the synthesis of
B-RuB2. Our study on binary ruthenium borides would have produced this phase had we
extended the BzRu ratio to 10.0.

The initial synthesis of B-RuB2 was quite unexpected, as the reaction was aimed
at producing an analog to the TthBro phases found by the Jeitschko group.l9 Finding
that RuB2 can crystallize with the ReB2 structure should not be quite as surprising. There
are a couple of reports of molybdenum- and tungsten-substituted RuB2 and OsB2
crystallizing in the ReB2 structure.” 2' Despite several reports on the stability of the

RuB2 structure versus the A1B2-type structure} 7 we found no reports comparing the

116

 

(a)

 

100 11m

0’)

Figure 4-5. SEM micrographs of (a) RuB2 and (b) B-RIIBz.

117

RuB2 and ReB2 structures. It would be interesting to conduct a study on both RuB2
phases to see if the structure really depends on how many electrons the ruthenium atoms
donate to the boron layers. Crystals of RuB2 and B-RuB2 are shown in Figures 4-5a and
4-5b, respectively.

Synthesis of the RERmB4 compounds was fairly easy for most rare earths with
the aid of the ﬂux. In addition to the compounds reported here, LaRu4B4 and ErRu4B4
were also synthesized. Most reactions containing a rare earth metal, ruthenium and
boron, regardless of the stoichiometry, would produce a RERU4B4 compound.
Unfortunately, RERuB4 compounds also crystallized along with the RERu4B4
compounds, sometimes with very similar crystal morphology. Figure 4-6 clearly displays
the similar morphology of CeRu4B4 and CeRuB4 crystals.

The synthesis of incommensurate ruthenium borides was very interesting. The
previous report on LuRu4B4 compoundsI3 did not mention any problems in synthesizing
Pr or Nd members of the series. Because we were able to synthesize the Ce member of
the series, we feel the size of the rare earth should not be a factor in determining whether
or not the structure is incommensurate. Heating proﬁles do seem to play a decisive role
in determining which phase is produced. The reaction temperature producing
incommensurate phases was at least 50°C greater than in reactions producing the
commensurate phases. The reaction time for the reactions yielding incommensurate
phases was also quite extended when compared to the reaction times for reactions
yielding the commensurate phases. It should be noted, however, that reactions at lower
temperatures and run for a shorter duration failed to produce any ternary borides for

either Pr or Nd.

118

 

-1_OO 1117—1

(5)

Figure 4-6. SEM micrographs of (a) CeRu434 and (b) CeRuB4.

119

F. Conclusions

Presented here are the details of the copper ﬂux synthesis and crystallographic
studies on single crystals of the known binary ruthenium borides Ru7B3, Ru2B3, and
RuB2. A new binary phase, B-RuB2, was synthesized and crystals of the compound were
grown from a copper ﬂux. The crystallographic data and atomic positions for B-RuBz are
reported here for the ﬁrst time. We successfully targeted and synthesized YRu434,
CeRu484, SmRu4B4, DyRu4B4, and YbRu4B4. Single crystals of these compounds, all
known compounds of the LuRu4B4-structure type, were grown from a copper ﬂux. The
ﬁrst report of incommensurate ternary ruthenium borides is also presented here.

We have learned that varying reaction parameters can have quite an effect on
compounds isolated from the ﬂux. For instance, by increasing the BzRu ratio to 10.0 a
new phase of ruthenium diboride, B-RuB2 was isolated. Additionally, increased ﬂux
reaction temperatures also play a role in the formation of incommensurate phases of
PrRu4B4 and NdRu4B4. Further investigations into the synthesis adjustments that create
these variations should be able to pinpoint a stronger cause and effect relationship to the

molten ﬂux synthesis techniques.

120

References

1. Nagamatsu, J.; Nakagawa, N.; Muranaka, T.; Zenitani, Y.; Akimitsu, J.,
Superconductivity at 39K in Magnesium Diboride. Nature 2001, 410, 63-64.

2. Cumberland, R. W.; Weinberger, M. B.;Gi1man, J. J.; Clark, S. M.; Tolbert, S. H.;
Kaner, R. B., Osmium Diboride, An Ultra-Incompressible, Hard Material. Journal of the
American Chemical Society 2005, 127, 7264-7265.

3. Roof, R. B.; Kemper, C. P., New Orthorhombic Phase in the Ru-B and Os-B Systems.
Journal of Chemical Physics 1962, 37, (7), 1473-1476.

4. Vandenberg, J. M.; Matthias, B. T.; Corenzwit, E.; Barz, H., Superconductivity of
Some Binary and Ternary Transition-Metal Borides. Materials Research Bulletin 1975,
10, 889-894.

5. Ivanovskii, A. L.; Medvedeva, N. 1., Electronic Structure and Energies of Interatomic
Interactions in Technetium, Rhenium, Ruthenium, and Osmium Diborides. Russian
Journal od Inorganic Chemistry 1999, 44, (10), 1633-1641.

6. Chiodo, S.; Gotsis, H. J .; Russo, N.; Sicilia, E., 0832 and RuB2, ultra-incompressible,
hard materials: First-principles electronic structure calculations. Chemical Physics Letters
2006, 425, 311-314.

7. Pallas, A.; Larsson, K., Structure Determination of the 4d Metal Diborides: A
Quantum Mechanical Study. Journal of Physical Chemistry B 2006, 110, 5367-5371.

8. Chapnik, I. M., On the Possibility of Semiconducting Properties in Diborides ReB2,
TcB2, OsB2, and RuB2. Physica Status Solidi 1977, A41, K71-K73.

9. Shishido, T.; Ye, J .; Sasaki, T.; Note, R.; Obara, K.; Takahashi, T.; Matsumoto, T.;
F ukuda, T., Growth of Single Crystals in the Systems with R-Rh-B and r-Rh-B-C (R=
Rare Earth Element) from Molten Copper Flux. Journal of Solid State Chemistry 1997,
133, 82-87.

10. Aronsson, B., The Crystal Structure of Ru7B3. Acta Chemica Scandinavica 1959, 13,
109-114.

11. Lundstrﬁm, T., The Structure of Ru2B3 and WB20, as determined by single-crystal
diffractometry, and some notes on the W-B system. Arkivfbr Kemi 1969, 30, 115-127.

12. Placa, S. L.; Post, B., The Crystal Structure of Rhenium Diboride. Acta
Crystallographica 1962, 15, 97-99.

13. Johnston, D. C., Superconductivity in a New ternary Structure Class of Boride
Compounds. Solid State Communications 1977, 24, 699-702.

121

l4. Grilttner, A.; Yvon, K., Lanthanum Ruthenium Boride, LaRu4B4. Acta
Crystallographica 1979, B35, 451-453.

15. Bezinge, A.; Yvon, K.; Braun, H. F .; Muller, J .; Nissen, H.-U., Ndee4B4: A
composition-modulated compound with incommensurate composite crystal structure.
Physical Review B 1987, 36, (3), 1406-1414.

16. Gubich, I. B.; Zavalii, P. Y.; Kuz'ma, Y. B., H020(Fe4B4)17 Boride- A New
Representatuve of Disproportionate Structures. Inorganic Materials 1993, 29, (2), 301-
303.

17. Nyman, H.; Andersson, S., The Stella Quadrangula as a Structure Building Unit.
Acta Crystallographica 1979, A35, 934-937.

18. Aronsson, B., The Crystal Structure of RuB2, OsB2, and IrBL35 and Some General
Comments on the Crystal Shemistry of Borides in the Composition Range MeB-MeBg.
Acta Chemica Scandinavica 1963, 17, 2036-2050.

19. Konrad, T.; Jeitschko, W., The Thorium transition Metal Borides Th2TBro (T = Fe,
Co, Ni) with a Structure Very Similar to that of CaB(,. Zeitschriftfur Naturforschung
1995, 50b, 1195-1199.

20. Rogl, P.; Nowotny, H.; Benesovsky, F ., Komplexboride mit ReB2-Struktur.
Monatsheftefu'r Chemie 1970, 101, 27-31.

21. Rogl, P.; Rudy, B., New Complex Borides with ReB2- and M02IrB2-type Structure.
Journal of Solid State Chemistry 1978, 24, 175-181.

122

Chapter Five

Gallium-Rich Gallides Isolated as By-Products from Molten

Flux Reactions Containing Boron

A. Introduction

One of the main obstacles to overcome in the molten ﬂux technique is the
formation of binary phases containing one reagent and the ﬂux metal." 2 These binaries
are often so stable that it is nearly impossible to form the intended products. During the
course of our research on synthesizing borides from a metal ﬂux we frequently
encountered this problem due to the limited solubility of boron in the ﬂux metals. The
other reagents in the reaction would react with the ﬂux metal before boron had a chance
to dissolve in the ﬂux. This was especially true when gallium was used as the ﬂux metal.

We have had a good deal of success in synthesizing borides from a gallium ﬂux,
producing Cr2B3, CrB2, MoB, WB, MnB, MnB4, ReB2, and HoMnB4. However, even in
reactions producing these borides, gallides were a side product, if not the main product.
Many of the gallides we have encountered throughout our research, such as V2Ga5,3
MGa4 (M = Cr, Mn),4 MGa3 (M = Fe, Co, Ru),5 and Ni2Ga3,6 have been known for some
time and have been well studied. Given the amount of research done on gallium systems,
we were fairly surprised to ﬁnd several binary gallides that had little, if any structural

characterization. In this chapter we will discuss the synthesis, single crystal growth, and

123

 

characterization of gallium-rich compounds from the Mn—Ga, Re—Ga, and Ni—Ga

systems.

B. Experimental

Reagents:
All reagents were used as received without further puriﬁcation: (i) Boron metal,

99% purity, -325 mesh, Cerac Specialty Inorganics, Milwaukee, WI, (ii) Rhenium
powder, 99.99% purity, Strem Chemicals, Newburyport, MA, (iii) Manganese powder,
99.9% purity, -50 mesh, Aldrich Chemical Company, Milwaukee, WI, (iv) Gallium 3-5
mm shot, 99.999% purity, Plasmaterials, Liverrnore, CA, (v) Tungsten metal, 99.9%
purity, Cerac Specialty Inorganics, Milwaukee, WI, (vi) Nickel metal powder, 98-99%
purity, -325 mesh, E.H. Sargent & Company.

Furnaces:

Commercial tube furnaces from Applied Test Systems, Inc. were used to heat
reactions sealed in quartz tubes. Furnace controllers from the Omega Company were
used to program heating proﬁles for individual reactions. Furnace temperatures were
monitored during reactions by thermocouples connected to the Omega controllers. To
extend the life of the fumaces, the operating temperature for prolonged heating was

limited to 1000 °C.

Synthesis and Isolation:

MnGa4,65: Boron and manganese powder were mixed and added to alumina
crucibles in millimolar quantities in boron-to-manganese ratios between 2:1 and 5:1.
Pieces of gallium shot totaling between 1.5 g and 2.5 g were then placed in the crucibles.

The alumina crucibles were then placed in 13 mm diameter quartz tubes and evacuated

124

under vacutun to a pressure near 10'4 Torr. The quartz tubes were then sealed using an
oxygen-acetylene torch. The quartz tubes were placed in mullite tubes, then heated in
commercial tube furnaces under heating proﬁles that increased the furnace temperature to
1000 °C in 10 hours, held the temperature constant for 48 hours, allowed the furnace to
cool to 850 °C in 5 hours, held the temperature constant for an additional 48 hours, and
then allowed the furnace to cool to room temperature in 60 hours.

After cooling, the quartz tubes were opened and the alumina crucibles were
placed in a half-concentrated HCl bath in order to remove the excess gallium. Most of
the gallium was removed after an overnight soak, but longer soaks were used without
damage to the recovered product. Silver, octahedral crystals were recovered in 50-60%
yields, based on manganese. Elemental analysis performed by Energy Dispersive
Spectroscopy (EDS) using a Scanning Electron Microscope (SEM) showed that
manganese and gallium were the only elements present in the crystals. M4 and MnGa4
were side products in these reactions. Elemental boron also remained in the collected
products

ReGa4,5: Boron and rhenium powder were mixed and added to alumina crucibles
in millimolar quantities in boron-to-rhenium ratios between 2:1 and 5:1. Pieces of
gallium shot totaling between 1.5 g and 2.5 g were then placed in the crucibles. The
alumina crucibles were then placed in 13 mm diameter quartz tubes and evacuated under
vacuum to a pressure near 10'4 Torr. The quartz tubes were then sealed using an oxygen-
acetylene torch. The quartz tubes were placed in mullite tubes and then heated in
commercial tube furnaces under heating proﬁles that increased the furnace temperature to

1000 °C in 10 hours, held the temperature constant for 48 hours, allowed the furnace to

125

 

cool to 850 °C in 5 hours, held the temperature constant for an additional 48 hours, and
then allowed the furnace to cool to room temperature in 60 hours.

After cooling, the quartz tubes were opened and the alumina crucibles were
placed in a half-concentrated HCl bath in order to remove the excess gallium. Most of
the gallium was removed after an overnight soak, but longer soaks were used without
damage to the recovered product. Silvery, octahedral crystals were recovered in 50-60%
yields. Elemental analysis performed by EDS using an SEM showed that rhenium and
gallium were the only elements present in the crystals. ReB2 was a side product in these
reactions.

N1057Ga4,43: Boron (7.5 mmol), nickel (0.5 mmol), and tungsten (1 mmol)
powders were mixed and added to an alumina crucible along with pieces of gallium shot
totaling 1.45 g. A quartz ﬁlter was placed in the crucible which was then placed in a 13
mm diameter quartz tube, evacuated under vacuum to a pressure near 10'4 Torr, and
sealed using an oxygen-acetylene torch. The tube was then heated in a commercial tube
furnace under a heating proﬁle that increased the furnace temperature to 1000 °C in 10
hours, held the temperature constant for 48 hours, allowed the furnace to cool to 850 °C
in 5 hours, held the temperature constant for an additional 48 hours, and then allowed the
furnace to cool to 250 °C in 60 hours.

The quartz tube was removed from the furnace at 250 °C and centrifuged to
remove excess gallium. After centrifugation, the tube was opened and the crucible was
placed in a bath of 5M 12 in DMF in order to remove the remaining gallium ﬂux. An
overnight soak was found to be insufficient for removing all of the gallium, so soaking
was continued for an additional 2 days. A subsequent investigation still found remaining

gallium, so an overnight soak in half-concentrated HCl was used to ﬁnally remove all

126

gallium. Silver, rectangular prisms are a minor product (IO-20% yield). Ni2Ga3 and WB
were side products in this reaction. Elemental boron was also present in the collected

product.

C. Physical Measurements

Energy Dispersive Spectroscopy: Energy dispersive spectroscopy (EDS)
analyses were carried out on selected crystals to determine their chemical composition.
The analyses were performed with a JEOL J SM-6400 scanning electron microscope
(SEM) equipped with Noran Vantage Energy Dispersive Spectroscopy (SizLi) detector
and Norvar window for standardless quantization of elements with Z Z 4. The crystals
were afﬁxed to an alumina sample holder with double-sided carbon tape. The EDS data
were acquired at an accelerating voltage of 10 kV with a 30 to 60 second accumulation
time.

Single-Crystal X-ray Crystallography: Single crystals were mounted on a glass
ﬁber with super glue and their intensity data were collected on either a Bruker SMART
platform CCD diffractometer, or a STOE IPDS II diffractometer using Mo K01 radiation
at 50 kV and 40 mA.

Samples examined with the STOE (MnGa4,65 and ReGa4_5) had individual frames
collected on a 34 cm image plate with a 60 second exposure time and a 10° to rotation.
The X-SHAPE and X-RED-32 software packages7’ 8 were used for data extraction and
reduction and to apply an analytical absorption correction. Direct methods and the

SHELXTL suite were used to solve and reﬁne the structures.

127

Individual frames of Nio_57Ga4,43 examined with the Bruker machine were
collected with a 03° a) rotation. The SMART software was used for data collection, and
SAINT software9 was used for data extraction and reduction. After applying analytical
absorption corrections, structure solution and reﬁnement were completed using direct
methods and the SHELXTL suite of programs.10

'1’ '2 suggested that the space group for MnGa6 was either Ccc2 or

Previous reports
Cccm. The value of IE2—1| was 0.863 for MnGa4_55 and 0.972 for ReGa45, indicating a
centrosymmetric structure, so Cccm was chosen. During initial indexing of the cell we
noticed weaker satellite reﬂections in the reciprocal lattice, as described previously by
another group.ll We have not yet attempted to include these satellites for a solution of
the supercell structure.

During reﬁnement we encountered a problem in placing gallium atom G35. The
most stable position for the atom leads to Ga—Ga bonds of unacceptably short lengths
(1.58 A). The occupancy of the Ga5 atom was reﬁned to nearly 50%, making the short
“bond” distance more acceptable. The real atom is assumed to be in one of the Ga5
locations in half of the structure, and in the other site in the remaining half of the
structure. Despite these adjustments, the anisotropic displacement parameters for Ga5 are
quite large. Gallium atom Ga4 also has large anisotropic thermal parameters, but these
are most likely a result of the problems with Ga5. The crystal structure reﬁnement data
for MnGa4,65 and ReGa45 are listed in Table 5-1. The atomic positions and isotropic
displacement parameters are listed in Table 5-2. Anisotropic displacement parameters for

the atoms are listed in Table 5-3. Bonding distances for MnGa4,65 and ReGa4_5 are listed

in Table 5-4.

128

The crystal structure solution and reﬁnement for Nio,57Ga4,43 were based on
parameters from a previous structural report from powder diffraction data.‘3 The
reﬁnement data for our study are listed in Table 5-5. During reﬁnement we found that
the Nil site had a mixed occupancy of nickel and gallium atoms. Occupancy reﬁnements
on this crystal showed the site contained 57% nickel, though results likely vary from
crystal to crystal. Atomic parameters and isotropic displacement parameters are listed in
Table 5-6. Anisotropic reﬁnement was performed on the non-mixed sites, and the results

are listed in Table 5-7. Bonding distances for Nio,57Ga4,43 are listed in Table 5-8.

129

 

Table 5-1. Crystal data and structure reﬁnement for MnGa4_65 and ReGa4,5.

Formula
Formula weight
Crystal system
Space group

Unit cell dimensions

Volume

Z, Density (calculated)
Absorption coefﬁcient

F(000)

Crystal size, mm

Theta range for data collection

Index ranges

Reﬂections collected
Independent reﬂections
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

MnGa4.6s
368.68

Orthorhombic

Cccm
a = 8.8323(18) A
b = 8.9523(18) A
c = 9.939(2) A
785.8(3) A3
8, 6.232 g/cm3
33.309 mm‘1
1316
0.144 x 0.179 x 0.185
3.83 to 34.54°
0 S h S 14.
0 S k S 14.
0 S l S 15
5655
875[R(int) = 0.2593]
99.4 %

ReGa4.s
499.94

a = 9.007708) A

b = 9.0618(18) A

c = 10.158(2) A
829.2(3) A3

8, 8.009 g/cm3
57.695 mm‘1

1716

0.157 x 0.164 x 0.260
4.50 to 36.00°

0 S h S 14,

0 S k S 14,

0 S 1 S 16

3898

1027[R(int) = 0.1536]
99.2 °/o

F ull-matrix least-squares on F2

875 / 0 / 35
1.068

R1 = 0.1020,
wR2 = 0.1590
R1 = 0.0274,
wR2 = 0.1358
0.0026(9)

1027 / 0 / 35
1.079

R1 = 0.0825,
wR2 = 0.1866
R1 = 0.1451,
wR2 = 0.2032
0.0022(3)

 

R1 = Z ”Fol — Well /2 IR! and wR2 = 12: (IF.2 - all >2 / 2 017.2121”

130

Table 5-2. Atomic coordinates (x 104) and equivalent isotropic displacement parameters

(A2x 103) for MnGa4_65 and ReGa4_5.

 

 

x y z U(eq) Occupancy
Mn(1) 2500 2500 3458(3) 4(1)
Ga(l) 1164(4) 4293(4) 5000 12(1)
Ga(2) 71 1(4) 1 150(4) 5000 14(1)
Ga(3) 2407(4) O 2500 14(1)
Ga(4) 0 2588(5) 2500 50(2)
Ga(S) 2500 2500 817(1 1) 48(3) 0.65
Re(l) 2500 2500 3461(1) 9(1)
Ga(l) 1187(5) 4369(4) 5000 15(1)
Ga(2) 623(5) 1 182(5) 5000 18(1)
Ga(3) 2383(5) 0 2500 20(1)
Ga(4) 0 2607(6) 2500 33(1)
Ga(S) 2500 2500 751(7) 33(2) 0.50

 

U(eq) is deﬁned as one third of the trace of the orthogonalized U11 tensor.

131

Table 5-3. Anisotropic displacement parameters (Azx 103) for MnGa4,65 and ReGa4,5.

 

 

ull U22 u33 U23 ul3 [J12
Mn(l) 7(1) 2(1) 3(1) 0 0 -1(1)
Ga(l) 12(1) 13(1) 10(1) 0 0 10(1)
Ga(2) 12(1) 16(1) 15(2) 0 0 -11(1)
Ga(3) 16(1) 8(1) 17(1) -10(1) 0 0
Ga(4) 56(3) 22(2) 71(4) 0 -57(3) 0
Ga(5) 99(9) 18(4) 28(5) 0 0 -16(5)
Re(l) 17(1) 8(1) 3(1) 0 0 -l(1)
Ga(l) 16(2) 12(2) 16(2) 0 0 8(1)
Ga(2) 1 8(2) 1 6(2) 19(2) 0 0 -9(2)
Ga(3) 24(2) 15(1) 21(1) -11(1) 0 0
Ga(4) 42(2) 21 (2) 35(2) 0 -28(2) 0
Ga(5) 60(7) 29(5) 9(3) 0 0 -8(6)

 

The anisotropic displacement factor exponent takes the form: -27r2[ h2 a"'2Ull + + 2 h
k a* b* U12 ]

132

Table 5-4. Bond distances (A) for MnGa4_65 and ReGa45.

 

 

MnGaw ReGa4.s
M—M 3.066(7) 3.1276(19)
M—Ga(1) 2.513(3) 2.591(3)
M—Ga(2) 2.512(3) 2.594(3)
M—Ga(3) 2.4334(14) 2.4689(6)
M—Ga(4) 2.4057(14) 2.4562(6)
Ga(l)—Ga(1) 2.415(6) 2.424(7)
Ga(1)—Ga(2) 2.788(5) 2.917(6)
Ga(l)—Ga(3) 2.858(2) 2.905(3)
Ga(2)—Ga(2) 2.412(6) 2.419(8)
Ga(2)—Ga(4) 2.868(3) 2.904(3)
Ga(3)—Ga(5) 2.796(6) 2.881(5)
Ga(4)—Ga(5) 2.772(6) 2.870(5)
Ga(5)—Ga(5) 1.620) 1.52605)

 

133

Table 5-5. Crystal data and structure reﬁnement for NiO.57Ga443.

Empirical formula
Formula weight
Temperature
Crystal system
Space group

Unit cell dimensions

Volume

Z

Density (calculated)
Absorption coefﬁcient

F (000)

Crystal size

Theta range for data collection
Index ranges

Reﬂections collected
Independent reﬂections
Completeness to theta = 28.26°
Reﬁnement method

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

Final R indices [I>2sigma(1)]
R indices (all data)

Absolute structure parameter
Extinction coefﬁcient

Largest diff. peak and hole

Ga4.43 NiO.57

342.30

296(2) K

Cubic

123

a = 8.424000) A a= 90°
b = 8.4240(10) A 8: 90°
c = 8.4240(10) A y = 90°
597.80(12) A3

8

7.607 g/cm3

42.607 mm"

1226

0.082 x 0.072 x 0.072 mm3

3.42 to 28.26°.
-11ShS11,-10SkS11,-11S1S10
2475

245 [R(int) = 0.0504]

96.8 %

F ull-matrix least-squares on F 2
245 / 0/ 18

1.093

R1 = 0.0329, wR2 = 0.0739

R1 = 0.0445, WR2 = 0.0780
0.5(4)

0.0022(3)

1.534 and -2.056 e.A'3

 

R1 = z ||Fo| — 1F.“ / 2 IE] and wR2 = [z (1F,2 — F31 )2 / z (wFfﬂm

134

Table 5-6. Atomic coordinates (x 104) and equivalent isotropic displacement parameters

(AZX 103) for N10,57Ga4_43.

 

 

 

 

 

x y z U(eq) Occupancy

Ni(1) -1617(5) 1617(5) 1617(5) 3(1) 0.57
Ga(ll) ' -1617(5) 1617(5) 1617(5) 3(1) 0.43
Ga(l) 1610(5) 1610(5) 1610(5) 4(1)
Ga(2) 0 3506(2) 0 1 1(1)
Ga(3) 0 5000 2499(2) 1 1(1)
U(eq) is deﬁned as one third of the trace of the orthogonalized Uij tensor.
Table 5-7. Anisotropic displacement parameters (Azx 103) for NiO.57Ga443.

ull U22 u33 023 ul3 ulz
Ga(l) 4(1) 4(1) 4(1) 1(1) 1(1) 1(1)
Ga(2) 12(1) 10(1) 12(1) 0 9(2) 0
Ga(3) 15(1) 15(1) 3(1) 0 0 -3(3)

 

The anisotropic displacement factor exponent takes the form: -21r2[ h2 a"‘2Ull + + 2 h

ka“ 8* U12]

Table 5-8. Bond distances (A) for Nio,5yGa4,43.

Ni—Ga(1 )
Ni—Ga(1 )
Ga( 1 )——Ga(2)
Ga(2)—Ga(2)
Ga(2)—Ga(3)
Ga(2)—Ga(3)
Ga(3)—Ga(3)

2.7184(16)
2.587(3)
2.496(2)
2.518(3)

2.4540(17)

2.4532(17)
2.9783(4)

135

D. Structural Description

MGa4,5 (M = Mn, Re): The MGa4+x compounds have a structure (Figure 5-1)
that is caught between the PtHg4 structure of the MGa4 compounds (Figure 5-2a) and the
PdGas structure (Figure 5-2b). In the PtHg4 structure, the metals reside in the center of
gallium cubes. As more gallium is incorporated into the structure, the metal-centered
cubes distort into metal-centered square antiprisms connecting along the c-axis (Figure 5-
3a,b). The squares of the antiprisms in the MGa4,5 structure are not quite offset by 45
degrees, as they are in the PdGas structure. This distortion may not be exactly the same
for every antiprism, and may lead to the satellite reﬂections we found in the reciprocal
lattice and the superstructure reﬂections mentioned by Girgis and Schulz.ll

In the MGa4_5 structure, gallium atoms Gal and Ga2 form intra-square bonds
(2.788 A) in one of the antiprism squares. The atoms in the other square, atoms Ga3 and
Ga4, are separated beyond Ga—Ga bonding distance (3.1445 A). In the PdGas structure,
all intra-square gallium atoms are separated from each other beyond Ga—Ga bonding
distance (3.472 A). There is inter-square Ga—Ga bonding in both the MGa4+x (Gal—Ga3
2.858 A, Gal—Ga4 2.858 A, and Ga2—Ga4 2.868 A) and PdGas structures (Ga2—Ga2
2.858 A). The antiprisms in PdGa5 connect to form a rectangular prism that is centered
by a gallium atom. In the MnGa4,5 structure, the rectangular prism is elongated and
centered by a pair of gallium atoms. However, during reﬁnement of the MnGa4_5
structure we found that the Ga5 atoms forming this pair are approximately half-occupied.
Attempts to place a fully occupied atom at the center of this pair have not been

successful.

136

mm
L0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(b)

-axis of (a) MnGa4 and (b) PdGas.

 

 

 

Figure 5-3. Connectivity along the c-axis in (a) MGa4,5 and (b) PdGas.

139

NiosyGa443: The NiO.57Ga443 structure is based on the y-brass structure.13 The
structures are composed of a series of body-centered cubes and chains of intersecting
rhombi. NiO.57Ga4,43 (Figure 5-la) and y-brass (Figure 5-1b) crystallize in different space
groups, 123 and 143m respectively, because of the position the nickel atoms occupy in
NiO.57Ga443. As shown in Figure 5-1a, the nickel atoms (Nil) occupy two comers of the
central body-centered cube in the Nio_57Ga4,43 structure, while gallium atoms (Gal)
occupy the other two corners . This lowers the degree of the rotation axis along the b-
axis from a 4-fold axis to 2-fold axis. This in turn lowers the space group symmetry of
Nio_57Ga4,43 from 143m to 123.

Through occupancy reﬁnements we found that the Nil position was split between
nickel atoms and gallium atoms. If the Gal position had the same mixed occupancy of
the Nil position, the 143m symmetry of the parent y—brass structure would be preserved.
The 143m symmetry could also be preserved if the mixed occupancy site were in the
Ga2 position. However, our occupancy reﬁnements for both of these positions found
them to be completely occupied by gallium atoms.

The original structural report of NiGa4 describes the structure as a vacancy

controlled y-brass phase.'3 The vacancies in the structure arise because the Ga3 atoms in

NiGa4 would reside in the 12d position of 143m, instead of the 24g position that the Zn2
atom occupies in the y—brass structure. This reduces the number of atoms in the unit cell

from 52 to 40.

140

 

 

E. Results and Discussion

As mentioned in Chapter Three, manganese gallides have been isolated from
every reaction targeting a manganese boride. Crystals of MnGa4 are often rectangular
plates or prisms (Figure 5-5a), so they can be mistaken for crystals of boride phases we
have seen. While we have seen some needle-like and larger prismatic crystals of
MnGa4,65, the crystals of this phase often have an octahedral morphology. These
octahedral crystals are indistinguishable from crystals of silicon, another common,
unintended side product in many of our gallium ﬂux reactions. We were quite lucky to
choose a gallide crystal when picking octahedral crystals on which to perform EDS
analysis, otherwise, we may have completely overlooked the compound.

After fortuitously ﬁnding MnGaMs, we knew to check the octahedral crystals
from rhenium-containing reactions, hoping to ﬁnd the ReGa4,5 analog. We were a bit
surprised to ﬁnd that the rectangular plates and prisms (Figure 5-5b) formed in these
reactions were not ReGa4, because the rhenium- and manganese-containing reactions
were carried out under identical conditions. The prisms and plates were the same ReGa4,5
phase as the octahedral crystals.

The synthesis of Nio_57Ga4,43 was also a chance occurrence. Our synthesis
targeting NiW2B15 happened to have a heating proﬁle that ended with a hold at 250 °C in
order to centrifuge the reaction tube and remove the excess gallium. Our literature search
on NiGa, later revealed that the compound forms near 250 °C, though single crystals had
not yet been isolated.“ '3 This may explain why we have not found this product in other
reactions, as the centrifugation step was removed from most of the later reaction

parameters. The absence of NiGa4 from other reaction products indicates that the

142

20 um

 

(b)

Figure 5-5. SEM micrographs of (a) MnGa4 and (b) ReGa4,5.

143

reaction temperature is the most important reaction parameter, and not necessarily the

stoichiometry.

The goal of this project was to synthesize borides using the metal ﬂux technique.
Therefore, as unintended products to these reactions, the gallides reported here are an
interesting side-note to the main focus of this research. We have not yet initiated
reactions targeting only these compounds, however such reactions will likely be carried
out in the future. Possible physical and electronic properties measurements should
provide even more inspiration to conduct further research on this interesting group of

gallides.

F. Conclusions

We have synthesized and grown single crystals of MnGaMs, ReGa45, and
Nio,5yGa4,43 from gallium ﬂux reactions originally targeting borides. The ﬁrst structural
reﬁnement on MnGa4,65 and ReGa4,5 shows that the compounds have an intermediate
structure between the PtHg4 and PdGas structures. Single crystal x-ray difﬁaction
analysis of a Nio_5yGa4_43 crystal conﬁrms the results of previous work by J inkui and

Sishen on powder data.

144

References

l. Fisk, Z.; Remeika, J. P., Growth of Single Crystals from Molten Metal Fluxes. In
Handbook on the Physics and Chemistry of the Rare Earths, ed.; K.A. Gschneidner, J .;
Eyring, L., Eds. Elsevier Science: 1989; Vol. 12, pp 53-70.

2. Kanatzidis, M. G.; P6ttgen, R.; Jeitschko, W., The Metal F lux- A Preparative Tool for
Intermetallic Compounds. Angewandte Chemie International Edition 2005, 44, (43),
6996-7023.

3. Lobring, K. C.; Check, C. B; Zhang, J .; Li, S.; Zheng, C.; Rogacki, K., Single Crystal
Growth, Bonding Analysis and Superconductivity of V2Ga5. Journal of Alloys and
Compounds 2002, 347, 72-78.

4. Haussermann, U.; Viklund, P.; Bostrtim, M.; Sirnak, S. I., Bonding and Physical
Properties of Hume-Rothery Compounds with the PtHg4 Structure. Physical Review B
2001, 63, (12), 125118.

5. Haussermann, U.; Bostrtim, M.; Viklund, P.; Rapp, O.; Bjtirnangen, T., FeGa3 and
RuGa3-Semiconducting Intermetallic compounds. Journal of Solid State Chemistry 2002,
165, 94-99.

6. Feschotte, P.; Eggimann, P., Les Systemes Binaries Cobalt-Gallium et Nickel-Gallium
- Etude Comparee. Journal of Less-Common Metals 1974, 63, 15-30.

7. X-SHAPE, Version 2.05; STOE & Cie GmbH: 2004.
8. X-RED 32, Version 1.10; STOE & Cie GmbH: 2004.

9. Sheldrick, G. M. SAINT, Version 4; Siemens Analytical X-Ray Instruments, Inc.:
Madison, WI.

10. Sheldrick, G. M. SHELXT L Structure Determination Program, Version 5.0; Siemens
Analytical X-Ray Instruments, Inc.: Madison, WI, 1995.

11. Girgis, K.; Schulz, H., A Gallium-rich Mn-Ga Compound. Naturwissenschaften
1971, 58, (2), 95.

12. Meissner, H. G.; Schubert, K.; Ananthraman, T. R., The Constitution and Structure
of Manganese-Galliurn Alloys. Proceedings of the Indian Academy of Sciences 1965,
56A, (16), 340-367.

13. Jingkui, L.; Sishen, X., The Structure of NiGa4 Crystal-a New vacancy Controlled y-
Brass Phase. Scientia Sinica 1983, 26A, (12), 1305-1313.

145

CHAPTER SIX

Conclusions and Future Work

When we began this research, we knew that the metal ﬂux technique had been
used with some degree of success in boride synthesis. Much of the preliminary work on
metal ﬂux boride synthesis covered the use of aluminum as the ﬂux metal. Our goals
were to extend the ﬂux technique to include gallium, and to expand on the work using
copper as a ﬂux for boride synthesis. Not only have we synthesized borides using metal
ﬂuxes, but we have also learned some of the limitations to using either gallium or copper
as a ﬂux.

While we have tried to use gallium to synthesize borides for most of the transition
metals, our success has been limited to groups 6 and 7. Crystals of Cr2B3, CrB2, MoB,
and WB all grew from reactions targeting ternary borides. We have not yet synthesized a
ternary boride from group 6; however, we have grown crystals of group 7 ternary borides.
Gallium was used to grow crystals of HoMnB4, DyReB4, ErReB4, and YbReB4. Binary
group 7 borides synthesized in ﬂux reactions are MnB, MnB4, and ReB2.

Further research on gallium ﬂux synthesis of group 6 and 7 borides should begin
with a focus on reproducing all known binaries for each metal in the groups. This may
seem like a trivial task at ﬁrst, but our synthesis of new phases for MnB and RuB2

illustrate the possibilities when using a ﬂux to grow known compounds. Exploratory

146

 

syntheses of “missing” analogs to known phases can also be carried out during this study.
After the binary systems are well analyzed, the ternary systems can be targeted.

The remaining YCrB4-type borides for manganese and rhenium should be among
the ﬁrst ternary targets. Commensurate and incommensurate RETM4B4 phases are
reported for both manganese and rhenium,I so the inﬂuence of the ﬂux on which phases
formed in this stoichiometry would be an interesting study. Exploratory synthesis in
ternary systems should also be carried out because of the ﬂux’s ability to trap metastable
phases that are inaccessible using other synthesis techniques.

When gallium ﬂux reactions were used with metals outside of groups 6 and 7,
they tended to include the ﬁrst row metals. Further research with gallium should include
more of the second- and third-row transition metals. Reactions with rhodium, iridium,
and platinum would be particularly interesting, since several borides from these metals
also include gallium in their structures.2 Preliminary reactions with rhodium and iridium
have produced thin plates too fragile to study with EDS or single-crystal diffraction.

Rhodium and iridium are also the next metals we would like to investigate using a
copper ﬂux. Several ternary rhodium borides have been synthesized using a copper ﬂux,3
but similar work with iridium has not yet been reported. Research targeting just binary
borides could be as ﬁ'uitful as our work with ruthenium. A potential complication is that
copper forms ternary borides with rhodium and iridium.4

Further research into the ruthenium system still needs to be conducted. Despite
synthesizing Ru7B3, Ru2B3, and two phases of RuB2, we were unable to isolate Ru) 133 or

RuB“. Our studies with the copper ﬂux varied the BzRu ratio in 1.0 increments. Ru) .B3

and RuB” have empirical formulas close enough to other phases that a 1.0 difference in

147

 

the B:M ratios between reactions can easily bypass favorable reaction stoichiometries for
these phases. Future studies should decrease the difference between B:M ratios of
reactions to 0.5 or 0.25.

To study properties of the new B-RuB2 phase, a little crystal engineering needs to
be done. We looked into having microhardness and incompressibility measurements
conducted on this phase, but our crystals were too small. Magnetic measurements can
also be performed once more crystals are isolated. Electronic structure calculations
should be carried out to determine the stability of the ReB2-structure versus the RuB2-
structure for ruthenium and other metal diborides.

The primary goal in further copper ﬂux syntheses of ternary ruthenium borides
should be completing the series of LuRu4B4-type compounds. It will be interesting to see
if the entire series can be synthesized from a copper ﬂux. Reactions temperatures need to
be varied to determine whether Pr and Nd always form the modulated phases from a ﬂux,
or if the commensurate phases can be isolated under different reaction parameters.
Substitution of other metals for ruthenium presents possible extension of our work.

After synthesizing the LuRu4B4-type compounds, more of their physical
properties need to be measured. The magnetic properties of these compounds have
already been determined from bulk samples, but single crystals will allow us to examine
the system more thoroughly, searching for anisotropy in the magnetic ordering. It will
also be interesting to compare the results of these studies with a study on LaRu4B4, to see
if the different connectivity in the Ru, tetrahedra inﬂuences the magnetic properties.

Copper ﬂux growth of binary niobium borides has been reported,5 and we have

grown needle-like crystals of Nb3Co4B7, These results indicate that more group 5 borides

148

should be targeted in copper ﬂux reactions. A few reactions targeting vanadium or
tantaltnn analogs to Nb3Co4B7 have not been successful. Initial attempts to synthesize
binary vanadium borides have been unsuccessful, yielding only large crystals of V-doped
boron. Initial attempts to synthesize other ternary niobium borides have only produced
Nb3B4.

One of the most challenging tasks in metal ﬂux synthesis of boride is ﬁnding a
suitable ﬂux metal for the late ﬁrst-row transition metals. We were fairly surprised that
the copper ﬂux did not produce iron or cobalt borides, considering that both ruthenium
and rhodium borides grow from that metal ﬂux. Aluminum, gallium, indium, and tin ,
though unsuccessful so far, are the metals most likely to produce borides. Canﬁeld has
also used Ni2B to grow quaternary nickel borocarbides,‘5 however the isolation calls for
decanting the ﬂux while above 1000 °C.

The results presented in this work set a foundation for continued research into the
metal ﬂux synthesis of borides. The above suggestions for future work are just a brief

glimpse of what could be done in this ﬁeld.

149

 

References

1. Zavalii, P. Y.; Mikhalenko, S. I.; Kuz'ma, Y. B., New incommensurate structures of
the borides Pr4)(Mn4B4)35 and Pr7(Re4B4)6. Journal of Alloys and Compounds 1994, 203,
55-59.

2. Klitnter, W.; Jung, W., Gaglr4B - ein Gallium-Iridiumborid mit isolierten, annﬁhemd
quadratisch planaren Ir4B-Gruppen in einer vom CaF2-Typ abgeleiteten Struktur.
Zeitschriftﬁir anorganische und allgemeine Chemie 1995, 621, 197-200.

3. Shishido, T.; Ye, J.; Sasaki, T.; Note, R.; Obara, K.; Takahashi, T.; Matsumoto, T.;
F ukuda, T., Growth of Single Crystals in the Systems with R-Rh-B and r-Rh-B-C (R=
Rare Earth Element) from Molten Copper Flux. Journal of Solid State Chemistry 1997,
133, 82-87.

4. Klunter, W.; Jung, W., The Copper Iriditun Boride Cu2Ir4B3 with a Layered Structure
Derived from teh ZnIr4B3 Type. Zeitschriftfur anorganische und allgemeine Chemie
2000, 626, 502-505.

5. Okada, S.; Hamano, K.; Lundstrém, T.; Higashi, I. In Crystal Growth of the New
Compound Nb 2B 3, and the Borides NbB, Nb 586, Nb 334, and NbB2, Using the Copper-Flux
Method, Boron-Rich Solids, Albuquerque, New Mexico, 1986, 1985; Emin, D.; Aselage,
T.; Beckel, C. L.; Howard, 1. A.; Wood, C., Eds. American Institute of Physics:
Albuquerque, New Mexico, 1985; pp 456-459.

6. Canﬁeld, P. C.; Fisher, 1. R., High-Temperature Solution Growth of Intermetallic
Single Crystals and Quasicrystals. Journal of Crystal Growth 2001, 225, 155-161.

150

 

Appendix

RF Furnace Design and Setup

The RF furnace has been a vital instrument for high-temperature synthesis in our
group for a number of years. However, several issues repeatedly arose and led to a
redesign of the furnace. A brief examination of the previous design and the associated
problems help illustrate the beneﬁts of the new design.

In the previous design, a thin sheet of molybdenum foil was wrapped around a
reaction crucible to act as a susceptor to the magnetic ﬁelds generated by the copper
induction coil. The crucible and susceptor were then balanced in a small mullite tube
packed with insulating wool. The tube was placed into a pyrex reaction chamber, which
was also packed with insulating wool. A ﬂow of argon was introduced into the reaction
chamber and upon turning on the power to the induction coil heating commenced.

The main problem with the previous design was that the crucible and susceptor
would frequently tip over and contact the reaction chamber. The reaction chamber would
then crack because of the extreme heat of the susceptor. The chamber also cracked a few
times because the insulating wool was packed too tightly and expanded slightly on
heating. Another small problem was that the molybdenum foil became brittle after
several reactions, and new pieces had to be used often. These problems and the inability

to run reactions under vacuum led to the new furnace design.

151

The new RF furnace has a reaction chamber made of fused quartz, which is more
thermally stable than the previous pyrex chamber. A boron nitride cylinder (Figure A-la)
has replaced the insulating wool and mullite tube from the previous furnace. The
cylinder is centered along the central axis of the reaction chamber by a small dimple in
the ﬂoor of the chamber. Instead of using molybdenum foil as a susceptor, we machined
a crucible holder from a molybdenum cylinder (Figure A-lb,c) capable of holding
crucibles of different sizes. The crucible holder sits on the BN cylinder and is more
stable than the molybdenum foil. Many reactions have been run and the new crucible
holder / susceptor is still working quite well.

In order to run reactions under vacuum, we installed a three-way valve that opens
the reaction chamber to a gas ﬂow or a vacuum pump. With this design the chamber can
be ﬁlled and purged several times before the reaction is initiated. T-joints have been
added to the gas lines, allowing for argon, nitrogen, or a mix of the two to be used for
ﬂushing the chamber or during reactions. Running reactions under dynamic vacuum
slows the deposition of ﬁhns on the inside of the reaction chambers. This allows the
thermometer to have a clear view of the crucible and susceptor, resulting in more
accurate temperature monitoring.

While the design of several components has changed, the overall setup of the RF
furnace has not changed (Figure A-2). The reaction chamber is placed in the induction
coil, with the susceptor located in the middle of the induction coil. The induction coil is
connected to a remote power station, which is controlled by the main control unit. The
reaction temperatures are monitored with an IR thermometer aimed between the copper

tubing and reported on a furnace controller connected to the thermometer.

152

 

 

 

(a)

(b) (c)

Figure A-l. Cross-sectional view of (a) BN susceptor holder and Mo susceptor along
(b) vertical and (c) horizontal axes.

153

E 533 wczoou H

 

 

Seam
wagon 8053—
EoﬁtoE<

UDJSE

 

 

 

 

 

 

“ED .8250
EcoﬁtoE<

 

 

 

y

 

 

 

 

\xlmJJ

2m

\II}

62

0

 

DUDDDU

maﬁa :U

 

 

 

 

\\?

 

 

 

 

 

E565,
.14
v NZ

 

.0355 “a .8.“ 958 .2266 .N.< v.53..—

 

 

 

5

5628885.

33.—m5
:09:

v So .653 @580

 

 

 

 

 

 

 

Soéaom
2883th

 

154

 

LIB

1ll]ll]]]]ll]]]l]]l1111111