kx

s- r- -..r..
M-f'ul‘n- « wm~

-. .t
”4- ,

_.

‘ .... . u
<n ...-.o- nu.

us . _—< L,

4.

-vvn§m . 2:”-

. ._;.V "5' - 65.17..-. .H.

Q-

 

 

«luv. u» .
N .- .

”317511;

.\,, -<

 

a

Sr!

m_n.k- ...- .w- -.._. .

- --..-
M
~.

,4 < ,
_A-‘~.' ..3

.p--

k

[‘.
y

‘ "izft?

" ”H.

'7

IN
I",

n
I O

x .
3!?

"E "
'9}! A

-h
l .

‘7".
‘i

-

 

4? <. A...
‘a. k $“4‘C-
. . 45-90 —. .. .
- .ur—hon‘onrsw‘nj‘avfiwl
~ -» -

vw

"AI-ru-

r
3:

ma.» '0‘ .
.1 .

"‘w n.- mum.
.1 . a

. . "Mk...“ _..

y”.

«A -~ ,m-

)1.“

. ’7

Alfiv‘.‘
z 3;

.,,. 7:"-7'V‘
m-
.-

 

0'.

~ H
H»...—

np -.—.-v
Ana. . E
.g-al-y

no
u

,.

o.

M
.H..

.u .r—O.

q...

,— ‘Itv' u.

”at:

 

This is to certify that the

thesis entitled
5 min/ago mt Cimwcaiuuabaix‘awafi amj mama
W03 ‘fWfiW/YYWO Wam'afl *0 W
Mawma (MLZNUL TCIWQ

presented by

has been accepted towards fulfillment
of the requirements for

M degree in Mia

WM

Major professor

Date AQWMZWL 5/ ’79?

0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

 

 

 

 

 

 

Llafinv

Michigan State
University

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1100 mm.“

SYNTHESIS AND CHARACTERIZATION
OF BINARY MATERIALS COMPOSED OF
TRANSITION METALS COORDINATED TO THE
ORGANIC ACCEPTOR TCNQ

By

Shannon A. O’Kane

A THESIS
Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree Of

MASTER OF SCIENCE
Department of Chemistry

1 999

In the
molecule 7,7,8
research in the
metal TIT-TC
however, that '
With a transiti
cyanide acce;
P013mers in v
DCNQI in
b'mlo‘ltlinone
men/1mm
and TCNQFI
Characmmm
divalent ms
ASH) with T(
eilrlier W0rke

Solve“ in die]

ABSTRACT

SYNTHESIS AND CHARACTERIZATION OF BINARY
MATERIALS COMPOSED OF TRANSITION METALS
COORDINATED TO THE ORGANIC ACCEPTOR TCN Q

By
Shannon A. O’Kane

In the early 1960’s, the surprising discovery Of conducting salts based on the
molecule 7,7,8,8-tetracyanoquinodimethane, (T CN Q), paved the way for an explosion of
research in the area of “all organic” conductors, the prototype of which is the organic
metal TTF-TCNQ where 'I'I‘F is tetrathiafulvalene. It has only been in recent years,
however, that we and others have replaced the donor portion of these charge-transfer salts
with a transition metal that is coordinated directly to the nitrile groups of an organo—
cyanide acceptor. A foremost illustration of this point is the Cu(DCNQD2 series of
polymers in which mixed-valence Cu(I,II) ions are coordinated to the dicyano units of
DCNQI in a three-dimensional fashion (DCNQI = N ,N ’-dicyano-1,4-
benzoquinonediimine). The interesting properties that were observed for the binary
metal/DCNQI system prompted our group to investigate similar materials with TCNQ
and TCNQF4. This thesis details a series of in-depth studies involving the syntheses,
characterization and structural studies of binary TCN Q materials of A30) and the
divalent ions of the first-row transition metals, Mn, Fe, Co, and Ni. The chemistry of
Ag(I) with TCNQF4 is also described. The main issue in this research, which had eluded
earlier workers in the field, is the role of polymorphism and coordinated/interstitial

solvent in dictating the purity and properties of M(TCN Q) and M(TCNQ)2 phases.

achi:

Prof

toh

deci

like

bell

Se\

prt
prc
es]
inc
frt

CX

CI

ACKNOWLEDGMENTS

There are several people who played an integral role in helping me
achieve this degree. First and foremost, I would like to thank my advisor,
Professor Kim Dunbar, for encouraging my desire to learn and helping me strive
to become a better chemist. I would also like to thank her for understanding my
decision to obtain this degree. To my “magnetic” co-worker, Rodolphe, I would
like to express my appreciation for performing the magnetic measurements, and
being another “advisor” who cares about the achievement of his fellow students.
Several previous TCNQ project group members namely, Robert Heintz, Hanhua
Zhao, Xiang Ouyang, and Guilio Grandinetti, were major contributors to this
project. I would like to thank Hanhua Zhao especially for introducing me to the
project and teaching me proper synthetic techniques. Many people in the lab,
especially Paul and Matt, were very helpful in solving everyday problems and
inquiries. This thesis is in memory of Professor Jerry Cowen, our collaborator
from the Department of Physics, who contributed valuable background
experiments as well as information about the complex magnetic properties of the
TCNQ compounds. Most importantly, though, was the emotional support and
encouragement that I received from my fiancé, Jeremy, and my friend, Amanda.

They kept me laughing, but also focused on the task at hand.

iii

 

 

LIST OF T.
LIST OF F.
LIST OF A

CHAPTER
lntroductio
TCNQ. An
Definition
Organic Cr
Molecular-
Complexe:
Conductin
Reference:

CHAPTEI
Agflcxc
Introducti.
Expenmer

A.

TABLE OF CONTENTS

LIST OF TABLES ...................................................................... vii
LIST OF FIGURES ...................................................................... viii
LIST OF ABBREVIATIONS ......................................................... xii
CHAPTER I. INTRODUCTION

Introduction .............................................................................. 2
TCN Q, An Electron Acceptor ......................................................... 3
Definition of Terms ..................................................................... 5
Organic Conductors ..................................................................... 7
Molecular-based Magnets .............................................................. 9
Complexes That Contain Covalently Linked Donors and Acceptors ............ 10
Conducting Metal/T CN Q Materials ................................................... 11
References ................................................................................ 18

CHAPTER II. NOVEL POLYMERIC CRYSTALLINE PHASES OF
Ag(TCNQ) AND Ag(TCNQF4): STRUCTURES AND MAGNETIC STUDIES

Introduction ............................................................................... 25
Experimental .............................................................................. 28
A. Synthesis
Bulk Synthesis of Ag(TCNQ) Phase 1(1) .......................... 29
Bulk Synthesis of Ag(TCNQ) Phase II, (2), from Ag metal. 29
Synthesis of [n-BU4N][TCNQ] (3) .................................. 30
Synthesis of [n-Bu4NIITCNQF4] (4) ............................... 30
Bulk Synthesis of Ag(TCNQF4) (5) ................................ 31
B. Structural Studies
X—ray Structure Determination of [n-Bu4N][TCNQF4] (4) ...... 32
X-ray Structure Determination of Ag(TCNQF4) (5) .............. 32
C. Physical Measurements .................................................... 33
Results and Discussion .................................................................. 34
A. Synthesis of Bulk Phases .................................................. 35
B. Infrared Spectroscopy ..................................................... 37
C. Structural Studies
Field Emission Scanning Electron Microscopy (FESEM) ........ 39
Powder X-ray Diffraction of Chemically Prepared Phases ....... 43
Single Crystal Diffraction Studies .................................... 47
D. Magnetic Behavior .......................................................... 50
Conclusions ............................................................................... 53
References ................................................................................. 55

iv

CIIMER F
pARA'tIAG
FAMILY
Introduction
Experiments
A. C
B. S

C. Si

D. Pi
Results andL
A. S)
B. In}
C. St:

D. M:

Conclusions. . .
References. . ..

CHAPTER IV.

THE Mni’TCNr
ImrOduction. . . .
Expefimental, , ,

A- Gen:
B. Sm
h
F
C

.N
C. Sepai
Mm T

L

T(

Hi

[I

 

CHAPTER III. SYNTHESIS AND CHARACTERIZATION OF
PARAMAGNETIC METAIJTCNQ COMPOUNDS: THE M(TCNQ)2
FAMILY
Introduction ...............................................................................
Experimental ..............................................................................
A. General Considerations ....................................................
B. Synthesis
Preparation of Li(TCNQ) .............................................
Preparation of [n-Bu4N][TCNQ] .....................................
Synthesis of M(TCNQ)2, where M = Mn (1), Fe (2), and CO (3)
Synthesis of Ni(TCNQ)2 (4) ..........................................
C. Single Crystal Growth
Slow Diffusion of Reactants in Solvent .............................
Incorporation of Gels ...................................................
A Typical Setup .........................................................
D. Physical Measurements ....................................................
Results and Discussion ..................................................................
A. Synthetic Analysis ..........................................................
B. Infrared Spectroscopy ......................................................
C. Structural Studies
Field Emission Scanning Electron Microscopy, FESEM .........
Powder X-ray Diffraction .............................................
D. Magnetic Properties
Magnetic Data Analysis ...............................................
Heat Capacity Measurements .........................................
Conclusions ...............................................................................
References .................................................................................

CHAPTER IV. IDENTIFICATION OF TWO PARAMAGNETIC PHASES
OF [Mn(TCNQ)2(L)2] (L = H20, MeOH) AND THEIR CONVERSION TO
THE MnCTCNQ); MAGNET.
Introduction ...............................................................................
Experimental ..............................................................................
A. General Considerations ....................................................
B. Synthesis of Bulk Methanol Phases
MDCTCNQ)2(MCOH)2 (1) .............................................
Fe(TCNQ)2(MeOH)2 (2) ...............................................
C0(TCNQ)2(MCOH)2 (3) ..............................................
NI(TCNQ)2(MCOH)2 (4) ...............................................
C. Separation of the Two Components in Bulk Samples of
MDCTCNQ)2(MCOH)2, (1)-
Low Temperature Reaction to yield [Mn(TCNQ)(TCNQ-

TCNQ)0,5(MCOH)2]OO, (la) ............................................
High Temperature Reactions to yield

[MnCTCNQ-TCNQXMeOHMm, (lb) ................................

62

65
65
66
67

68
69
7O
7 1
7 l
71
77

81
85

87
91
93
95

100
104
104

105
106

107
107

108

108

 

 

R:

 

Conclusions. .
References. . .

 

D. Isolation of a Second Phase of Mn(TCNQ)2(H20)2 (5) That Contains

the o—dimer Structural Unit (5b) ........................................... 109
E. Interconversion of Mn(TCNQ)2 (6) and Mn(TCNQ)2(X)2 polymers
where X = MeOH (1) or H20 (5) .......................................... 109

F. Attempted Conversion of Fe(TCNQ)2(MeOH)2 (2) to Fe(TCNQ)2 (7) 110
G. Attempted Conversion of Ni(TCNQ)2(MeOH)2 (4) to Ni(TCNQ)2 (8) 110

Results and Discussion .................................................................... 111
A. Synthesis ...................................................................... l 11
B. Infrared Spectroscopy ....................................................... 116
C. X-ray Powder Diffraction ................................................... 126
Conclusions ................................................................................. 131
References .................................................................................. 134

vi

 

Table I.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

l
.

J

1

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

LIST OF TABLES

IR data for Cu(TCNQ), Ag(TCNQ), Ag(TCNQF4), [n-BmNIITCNQ],
and [n-Bu4N][TCNQF4] ........................................................ 38

Magnetic parameters for the Ag(T CN Q) family and Ag(TCNQF4).

Cb is the Curie constant of the impurity, and should be 0.375 for one

spin S = 1/2. In parenthesis is given the molar percentage of spin V2

present in the sample ............................................................ 51

Solubility studies of various gels where S = soluble and I = insoluble.
Heat was used with gels that did not readily dissolve ...................... 70

A study of the effect of reaction time on the composition of M(TCNQ)2
phases using elemental analysis as a test for purity ......................... 74

Infrared spectroscopic data in the v(CEN), v(C=C), and 5(C-H) regions

for TCNQ" compounds .......................................................... 78
Conversion reactions of Mn(TCNQ)2 and Mn(TCNQ)2(X)2 compounds,

X = MeOH and H20 ............................................................. 114
A comparison of M(TCNQ)2(MeOH)2 bulk products with varying

reaction conditions ............................................................... 118

vii

LIST OF FIGURES

Figure 1. Illustration of a variety of electron acceptors .............................. 4
Figure 2. Different stacking modes of TCNQ rings ................................... 12
Figure 3. SEM images of Cu(TCNQ) phase I (top) and phase 11 (bottom)

illustrating the morphological differences .................................. 14
Figure 4. X-ray powder diffraction patterns of the conversion of Cu(TCNQ)

phase I to phase II at varying time intervals ............................... 15
Figure 5. The coordination environment of Ag(T CN Q) ............................ 16

Figure 6. Organonitrile acceptor molecules used as ligands in metallopolymers. 25

Figure 7. Schematic drawing showing the change of resistance over time for
Cu(TCNQ) films ............................................................... 27

Figure 8. FESEM photographs of Ag(TCNQ) phase I at various magnifications. 40

Figure 9. FESEM photographs of Ag(TCNQ) phase II at various magnifications. 41

Figure 10. FESEM images of Ag(TCNQF4) and [n-Bu4N][TCNQF4] .............. 42

Figure 11. Powder diffraction patterns for the two phases of Ag(TCNQ) and a
simulation of electrochemically prepared Ag(TCN Q) from the

literature ....................................................................... 44

Figure 12. Powder diffraction patterns illustrating the attempted conversion of
Ag(TCN Q) phase I to phase II ............................................... 45

Figure 13. Comparison of the powder diffraction patterns of Ag(TCNQF4)
and the simulated powder diffraction pattern from the Ag(TCNQF4)

crystal structure ................................................................ 46
Figure 14. Four-fold coordination within Ag(TCNQF4) .............................. 48
Figure 15. Ligand rt-stacking along the b axis in Ag(TCNQF4) .................... 49

Figure 16. Temperature plots of the magnetic susceptibility for
Ag(TCNQ) phase I (at 5000 G). Phase II (at 1000 G), and

viii

Ag(TCNQF4) (at 1000 G) .................................................... 51

Figure l7.Temperature dependence of the magnetic susceptibility for
[n-Bu4N][TCNQ] (at 5000 G). The bump between 40 and 50K is
due to the presence of oxygen. The black line is the best fit obtained
with a dimer model of spin V2 ................................................ 52

Figure 18. Temperature dependence of the magnetic susceptibility for
[n-BllaNIITCNQFd (at 1000G). The black line is the best fit

obtained with a dimer model of spin 1/2 ................................... 52
Figure 19. Schematic diagrams of organic donors and acceptors used in

charge-transfer materials ..................................................... 62
Figure 20. Slow diffusion of (a) solvent layers and (b) solvent/gel layers 68

Figure 21. Plots of magnetization versus temperature for four Mn(TCNQ)2
compounds with various reactant and solvent concentrations .......... 75

Figure 22. Plots of magnetization versus temperature indicating the

“aging process” of the Mn(TCNQ)2 material ............................. 76
Figure 23. IR spectrum of a bulk Ni(TCNQ)2 sample indicating the presence

of a o—dimer containing phase(s) at 803 cm"1 ............................. 80
Figure 24. Spherical particles of NiCTCNQ)2 .......................................... 82
Figure 25. TEM photographs of compounds 1 — 4 .................................... 83

Figure 26. Particle size trend for M(TCNQ)2 compounds at 10k magnification... 84
Figure 27. X-ray powder diffraction patterns for isostructural compounds 1 — 4.. 86
Figure 28. Plots of xmolw" and llxmolm" versus temperature for Mn(TCNQ)2. . 88

Figure 29. Fitting of llxmol‘m versus temperature for Mn(TCNQ)2 to ascertain the
exact critical temperature, Tc ................................................ 88

Figure 30. Plot of peg versus temperature for Mn(TCNQ)2 ........................... 89

Figure 31. Plots of x’mol‘m versus temperature illustrating the frequency
dependence of the M(TCNQ)2 compounds ................................. 90

Figure 32. Plot of magnetization versus field illustrating the hysteresis observed
for Mn(TCNQ)2 ............................................................... 91

ix

Figure 33. Plot of heat capacity measurements for Fe(TCNQ)2 ..................... 92
Figure 34. Different binding modes of TCN Q ......................................... 101

Figure 35. The (a) single unit and (b) packing diagram of [Mn(TCNQ)(TCNQ-
TCNQ)0,5(MeOH)2]..o (la), illustrating the u4-[TCNQ-TCNQ]2’ and
bidentate cis-[tt-TCNQT binding modes ................................. 102

Figure 36. Two diagrams of the [MnCTCNQ—TCNQ)(MeOH)4]ao (lb) structure,
which contain only the tu—[TCNQ-TCNQF’ binding mode ............ 103

Figure 37. The structure of [Mn(TCNQ)2(H20)2]® (5a), which contains the
syn-[u-TCNQT binding mode ............................................... 103

Figure 38. Infrared spectra in the v(CEN) region Show that both crystallographic
forms are present in bulk Mn(TCNQ)2(MeOH)2 product ................ 119

Figure 39. Comparison of the v(CEN) regions for experimental sample obtained
in the present Studies (left) and for X-ray crystals (right)
of [Mn(TCNQ)(TCNQ-TCNQ)05(MeOH)2]ao ............................ 120

Figure 40. Comparison of the v(CEN) regions for powder (left) and crystals
(right) obtained in the chemistry of [Mn(TCNQ-TCNQ)(MCOH)4]oo... 121

Figure 41. A comparison of the infrared spectra of the starting material (top),
MnCTCNQ)2(H20)2, and the converted phase (bottom). Both the
v(CEN) and 8(C-H) regions clearly indicate the differences in
structure ......................................................................... 122

Figure 42. A comparison of the Spectra of the starting and final product
in the conversion of (a) Mn(TCNQ)2(MeOH)2 (1) to (b) Mn(TCNQ)2
and (c) a comparison to bonafide sample of Mn(TCNQ)2 ............... 124

Figure 43. Illustration of the resulting o-dimer products (2b, 4b) obtained after
heating suspensions of Mn(TCNQ)2(MeOH)2. M = Fe, (2), (top)
and Ni, (4), (bottom) ........................................................... 125

Figure 44. IR spectrum for the heated Ni product (4b), (Figure 42), recorded after
exposure to X-rays. The 5(C-H) frequency of 824 cm‘1 is present for
TCNQ" ........................................................................... 126

Figure 45. X-ray powder diffraction patterns for the heated material (lb) (top)
and crystals of [MnCTCNQ-TCNQ)(MeOH)4]w ........................... 127

Figure 46. A comparison of the X-ray powder diffraction patterns for the heated

phase (5b) (top) and the original bulk phase, Mn(l‘CNQ)z(HzO)2 (5)
(bottom) ........................................................................... 128

Figure 47. X-ray powder diffraction patterns measured before (top) and after
(middle) the conversion reaction of Mn(TCNQ)2(MeOH)2 to
MnCTCNQ)2. As a reference, a powder diffraction pattern Of a
bona fide Mn(TCNQ)2 is displayed at the bottom .......................... 130

xi

LIST OF ABBREVIATIONS

br

em

°C

mmol

mol

Angstrom
bending mode
Beta

broad

bridging ligand
Wavenumber
degree Celsius
frequency (stretching)
gram

hour

insoluble
infrared
medium
milliliter
millimole

mole

pi

strong

soluble

volume

weak

xii

w/w weight by weight

X MeOH or H20

xiii

Chapter I:

Introduction

Introduction

The recognition of the superconductivity of Hg in 1911, led McCoy and Moore to
predict the future synthesis of “synthetic metals”.l It was theorized that these organic
materials would exhibit metallic behavior, namely, an increase in electrical conductivity
with decreasing temperature. Upon comparison to their metal counterparts researchers
envisioned that the main advantages of these materials would be ease of synthesis at
lower temperatures and the ability to be tailored for specific properties which alone
makes them worthy of study. The beginning of a realization of this dream occurred in the
early 1960’s when Melby and coworkers2 discovered the semiconducting properties of
alkali salts of the electron acceptor, TCNQ, 7,7,8,8- tetracyanoquinodimethane, in its
reduced form.

During the past four decades a great deal of research has focused on the synthesis
of organic superconductors3 and charge-transfer salts.4 Only recently have chemists
broadened their scope of interest to include the incorporation of paramagnetic transition
metal centers for the purpose of obtaining interesting magnetic5 and conducting6
properties. After several decades of exhaustive exploration of organic donor and acceptor
derivatives, a new trend has emerged in the synthesis of new conductors. The
incorporation of an inorganic component provides a new use for the electron acceptor,
namely, as a ligand that coordinates to the metal. The resulting materials that form by
direct bonding of metals and organic acceptors have opened up a vast new synthetic arena

and brought magnetism to the forefront of research interest.

TCN Q, An Electron Acceptor

Several characteristics of 7,7,8,8-tetracyanoquinodimethane, TCNQ, render it an
ideal ligand for use in the synthesis of magnetic and conducting materials.7 TCNQ is a
powerful electron acceptor and it forms a very stable radical anion. This is due to the
presence of four cyano withdrawing groups, the molecule’s planarity, and its high
symmetry. TCNQ can coordinate to metals through the nitrogen lone pairs on the cyano
groups or through the rt-system, although the latter is rare. Various coordination modes
have been identified which indicate that TCNQ is a very versatile bridging ligand capable
of producing one- , two-, and three-dimensional solids. One of the most important
characteristics of TCNQ is its ability to participate in tit—stacking. The delocalization of
the electrons through the overlapping rt-systems has been Shown to increase the charge-
transfer in these materials in one-dimension. Furthermore, the accessibility of three
oxidation states of the ligand, namely, TCNQO, TCNQ'l and TCNQ'Z, allow for diversity
in synthesis and properties due to the possibility of forming mixed-valence compounds.
Finally, the interaction of the S = V2 spin state of the reduced form of this ligand with the
spin on the metals result in the generation of interesting conducting and magnetic
properties.

The synthesis of alkali metal or other organic cation salts of TCNQ resulted in
identification of two distinct structures.2 The first is a 1:1 composition with the formula
MTCNQ', M = alkali metal. The second incorporates neutral TCN Q within the structure
and is therefore a 1:2 composition with the formula M‘TCNQ'(TCNQ). The
representative formulae indicate that a complete charge-transfer to the acceptor has

occurred within both structures. The formula differences led to marked differences in

electrical conductivity; the former exhibits a high electrical resitivity of 104-109 ohm cm,
whereas in the salt that contains neutral TCNQ, the resistivity is lowered to 001-100 ohm
cm.4 In comparison, a metallic conductor exhibits a resistivity of 106 ohm cm.

Other electron acceptors have been also used to prepare charge-transfer
compounds; these are currently being used to synthesize conducting and magnetic
materials in our laboratories. Schematic diagrams of TCNE (tetracyanoethylene),
DCNQI (N ,N ’-dicyano-1,4-benzoquinonediimine), TCNQ and TCNQ derivatives are

shown in Figure 1. A close relative to TCNQ is TCNE, which can undergo reduction by

MC“ CN

N
TCNE DCNQI
F
N N N N
TCNQ Tenor,
Br on.
N N N N
N N N N
Br Moo
Tonnar2 TCNQ(OMe)
2

Figure 1. Illustration of a variety of electron acceptors.

one or two electrons, and is known to coordinate to metals through the four cyano groups.
Another new class of electron acceptors dicyanoquinodiimines, DCNQI, are similar in
their redox properties to TCNQ, but only two coordination sites are available.6 Further
expansion of the field of electron acceptors has occurred through derivatization of the
TCNQ molecule. In 1975, derivatives of TCNQ were synthesized by Wheland; these are
based on the addition of electron withdrawing halogens (e.g. TCNQF4 or TCNQ(Br)2) or

electron donating groups (e.g. TCNQ(OMe)2).8

Definition of Terms

Conducting charge-transfer salts are composed of highly ordered donors and
acceptors in which at least one component is in the form of a radical species.
Conductivity, or delocalization of electrons, iS the result of favorable intermolecular
orbital interactions in the solid state. The occupancy of the conduction band is, however,
also critical for determining this property. Band theory states that a transfer of electrons
from the valence (filled) to the conduction (empty) band must occur in order for
conductivity to be favorable. Therefore, if there is a sufficiently large energy gap
between the bands, the material is an insulator due to the lack of energy available to
promote electrons. As the gap decreases, electrons are able to be thermally excited to the
HOMO band (conduction) and the material is an intrinsic semiconductor. Metallic
behavior requires partially filled bands in which electrons are promoted to infinitely
higher energy levels. The Fermi level is the highest occupied state and the mobility of

electrons at this level dictates the physical properties of the metal.9

Since conductivity is temperature dependent, the critical temperature is important
in describing the properties of a charge-transfer complex. Measuring a conductor’s
resistivity, p, is another method of characterization used to describe the ability for current
to pass through a substance. Metals have a low resistivity (e. g.10{’ ohm cm) whereas an
insulator exhibits a high resistivity ~ 1010 ohm cm.4 Organic and inorganic
semiconductors fall between those of metals and insulators, having resitivities ranging
from 104 to 108 ohm cm.4

The physical property of magnetism, like conductivity, is based on the intrinsic
behavior of the material. There are five possible types of magnetic behavior observed in
solids, and their appreciation is important for a general understanding of the factors
involved in magnetic design. A solid that lacks unpaired electrons is termed diamagnetic,
while paramagnetism is observed in solids that have unpaired electrons that interact
weakly with a magnetic field. The other three main types of magnetic behavior,
ferromagnetism, antiferromagnetism, and ferrimagnetism, are different than the latter in
that they are characterized by strong interactions and by cooperative (or bulk) properties.
A ferromagnetic ordering occurs when all of the Spins in a solid are aligned in a parallel
fashion (or). If the spins in the solid are all anitparallel (4%) throughout the material,
the ordering is classified as being of the antiferromagnetic type. Finally, the phenomenon
by which spins order antiferromagnetically but with a net spin (4%) remaining is termed
ferrimagnetic.

The coupling constant J, describes the isotropic interaction between two spins S]
and S2, and is defined by the spin Hamiltonian, H = -2J(Sl-S2).ll The energy separation

between the singlet and the triplet states is J. In ferromagnetically coupled spins, J>0

(triplet ground state), while with antiferromagnetically coupled spins, I <0 (singlet ground
state). If the orbitals containing the unpaired electron(s) are orthogonal to each other and
Hund’s rule keeps the spins parallel, then ferromagnetic coupling between the spins
occurs. If there is direct orbital overlap, however, the antiparallel alignment will be
favored. In order to achieve a ferromagnetic interaction, spins in orthogonal orbitals
must be in the same spatial region (adjacent sites).12 Ferrimagnetism is achieved when
the local antiferromagnetic interactions do not have Spins of equal magnitude.

When considering magnetic materials, three factors are useful in describing a
magnet.13 The first is the critical temperature, Tc, which is the temperature at which a
magnetic ordering occurs within a system. The second is the saturation magetization or
the amount of permanent magnetization that a material can store. Finally, a coercive field
defines the magnetic field that must be applied to restore the magnet to its original
condition. The latter is important for instance in the resistance of data loss that is stored

in magnetic materials.

Organic Conductors

The discovery in the early 1960’s that TCNQ salts are organic semiconductors led
to almost four decades of intense investigation of electron donors and acceptors.2 The
ultimate goal is the synthesis of highly conducting, especially superconducting, materials.
Alkali TCNQ salts such as K+ are remarkable in that they exhibit room temperature
conductivities in the range, 5 X 104 Q '1 cm", which is a huge increase (4 -10 orders of

magnitude) over previous organic materials.14 After this discovery there was a renewed

interest in TCNQ salts. It was discovered that the high conductivity is due to the Stacking
of the planar molecules, with segregated stacks of cations and anions. The overlap of the
1: systems increases the charge-transfer properties with conductivity occurring only along
the direction of the Stack.

The synthesis of 'TTF in 1970 introduced a new class of organic donors that
contain electron rich sulfur atoms.15 "6 Coppens found that 'ITF radicals arrange in stacks
similar to TCNQ,l7 and that 'TTF salts's'm‘18 are highly conductive as well I
(O'=O.2 (2" cm'1 at 300K). The combination of TI'F and TCNQ seemed to be an
obvious method for forming a highly conducting charge-transfer material due to the small
difference in redox potentials between the two molecules. Indeed, this proved to be true,
as it was found that a solid with partially oxidized 'I'I'F and partially reduced TCNQ ions
formed an organic metal that exhibited a metallic conductivity (0') of 0’ = 104 S)" cm“1
near 60 K.19 Below this temperature, a phase transition from a metal to a semiconducting
state occurs.

Derivatization of the 'I'I'F donor contributed to an increase in the number of
conducting salts that exhibit charge-transfer properties. TSF-TCN Q, (T SF =
tetraselenafulvalene) was the first derivative synthesized by replacing the sulfur atoms
with selenium, and, as expected, the high metallic conductivity surpassed that of 'ITF-
TCNQ by two hundred ohms. Furthermore, the conducting state is significantly more
stable (> 40K).20 Other derivatives with selenium, HMTSF-TCNQ21 (HMTSF,
hexamethylene-tetraselenafulvalene) and tetraselenafulvalene (TMT SF) salts”, also Show
metallic behavior. In particular, the charge transfer salt, (TMTSF)2[PF6], is the first

organic metal to exhibit superconductivity at Tc = 0.9K ( P = 12 kbar). The use of

pressure suppresses the metal—to—insulator transition that usually occurs, thus allowing
metallic behavior to occur at higher critical temperatures.23

In the early 1980’s, a new electron donor, namely
bis(ethylenedithio)tetrathiafulvalene (BEDT-TTF) was discovered to form two-
dimensional charge-transfer salts making them distinct from other 'ITF derivatives. This
dimensionality is due to short inter-stacking sulfur interactions in two-dimensions, which
surprinsingly leads to superconducting behavior.24 Two examples of these materials are
(BEDT-‘I'I'F)Re0424a and Cu[N(CN)2]X where X: Br or Cl.24d The last example is
particularly notable as compounds which superconduct at the highest critical temperature
known for organic superconductors (Tc = 11.6 K at ambient pressure and 12.5 K at 0.3

kbar respectively).

Molecular-based Magnets

The introduction of a paramgnetic metal component into charge-transfer materials
allows for exploration in the promising field of molecule-based magnets. Over the past
thirty years, notable breakthroughs have been reported by Olivier Kahn, who synthesized
the first molecular, ferrimagnetic chain based on Mn(II)Cu(II) centers,25 and by Miller et
al., who discovered the first organic—based molecular magnet, namely
[Fe(Cp"‘)2]”‘[TCNE]".26 The latter forms a mixed stack arrangement and exhibits
ferromagnetism below 4.8 K. Shortly after this discovery, other first row transition metal
Cp* donors were used to prepare ferromagnetic charge-transfer materials when combined

with acceptors such as TCNQ, bis(dithiolato)-metalates, and C4(CN)¢5.28 One example,

[(CsMe5)2Mn] [T CNQ], exhibits ferromagnetism at a Tc of 6.2 K which was considered to
be a breakthrough at the time.28f In the ten years since this discovery, however, rapid
developments have occurred in the area of higher Tc molecule-based magnets. In fact,
there are now room temperature and above room temperature magnets based on

molecular precursors.

Complexes That Contain Covalently Linked Donors and Acceptors

A different approach to the design of magnetic and conducting materials is the
assembly of open-shell metal centers and organic radicals into inorganic/organic “hybrid”
structures.5"‘6'29 The o—coordination opens up the possibility for acceptors, such as TCNQ
and TCNE, to undergo charge-transfer reactions with electron rich organometallic species
e.g. (C5R5)(CO)2Mn(THF) where R = H, CH3.30 One mode of binding, specifically the
0'411 mode, was first identified in the complexes [Ru(PPh3)2(TCNQ)] and
[Cu(pdto)(TCNQ)]2 (pdto = 1,8-di-z-pyridyl-3,6—dithiaoctane).31 Other examples include
[Ru2(CO)5(tt-L)2('r|'-TCNX)](TCNX) (X = E,Q) which contain both inner and outer
sphere TCN Q or TCNE.32

One of the intriguing findings that has emerged from the coordination of
organocyanide acceptors with transition metals is an accidental discovery made by
Manriquez and Miller et al. The reaction of V(C6115)2 and TCNE was found to yield an
insoluble, amorphous solid formulated as V(TCNE)2°1/2(CH2C12). Reported to be a
room-temperature ferrimagnet, this compound is thought to contain o—coordinated

reduced TCNE as deduced by infrared Spectroscopy.5m Numerous examples of

10

complexes that contain o-coordinated TCNE have appeared in the literature,3°’3233

several of which have been structurally characterized by X-ray crystallography.34 The
majority of these products exhibit either o-n' or 011-112 coordination modes within a
linear chain structural motif. Two such examples are the ferromagnetic salt,
[MnTPP] [T CNE] (TPP = meSO-tetra-phenylporphinato) and [M(hfacac)2][TCNE] (M =
Co, Cu; hfacac = hexafluoroacetylacetonate). Also, a two—dimensional structure was
reported to consist of o-u-n4-TCNE units linking Rh2(02CCF3)4 through the axial Rh
sites, although these compounds are not of any magnetic interest.34f

An excellent illustration of the design of highly conducting covalent solids with
o— as well as drt-prt interactions between metal and organic constituents is the class of
compounds, Cu(R,R’-DCNQI)2 (DCNQI = N,N’ —dicyanoquinonediimine, R,R’ = halide,
alkyl group or alkoxide groups).6 X-ray studies of these compounds reveal DCNQI
molecules that form one-dimensional columns connected to each other by tetrahedral
mixed-valence Cu+I2+ cations. The high conductivity that results is mainly attributed to

the DCNQI-DCNQI stacking within these structures.

Conducting Metal/TCNQ Materials

The DCNQI compounds are not the only structures composed solely of metal
cations and electron acceptor anions. Alkali metal/'1‘ CNQ complexes35 exhibit electrical
conductivities on the order of 10'2 to 10'5 (2" cm”.36 This class of compounds has been

structurally characterized and the direction of conductivity is ascertained to be parallel to

the Stacks of TCNQ molecules. Surprisingly, there are noticeable structural differences

ll

 

 

'mthisfan

modes of r

 

each cubic C
Structures the

OCIZhedraI coo
only Six nitrile

distances betwe

 

in this family of compounds. One of the most important is the three TCNQ stacking

modes of ring overlap, which are shown in Figure 2. Four TCN Q molecules surround

NC _ CN
N NC ', __ C" CN
' -
NC — CN

(c) ring - external bond

Figure 2. Different stacking modes of TCNQ rings.

each cubic Cs+ cation in Cs2(TCNQ)3,35“ while in the RbCI‘CNQ) and K(TCNQ)
structures the metal cation is surrounded by eight nitrile groups.35M For Na(TCNQ), an
octahedral coordination is exhibited around the metal with the cation being surrounded by

35c

only six nitrile groups. Another structural differences includes variation of stacking

distances between TCNQ molecules which range from 3.2 to 3.6 A.35 These differences

 

 

   
  

conducting
TCNQ' ani.
a lzl ratio 0

manna

 

fieldtoader
Alelectrode
critical thres
Which it rein
high resista'
Original met
shown in I
pheroeieen

molecule ll

[Cu

The

mfonnah.

can be directly correlated to the reported differences in electrical conductivities for the
alkali metal/TCN Q salts.

More to the point of the present work, the material Cu(TCNQ),” whose
conducting properties have been the subject of intense debate, exhibits O-coordination of
TCNQ' anions to the metal center. As the formula indicates, Cu(TCNQ) is composed of
a 1:1 ratio of metal and ligand. The compound was first prepared as a film by dipping Cu
metal into a solution of neutral TCNQ. It was reported that the application of an electric
field to a device consisting of a thin film of Cu(TCNQ) sandwiched between a Cu and an
Al electrode induces a switching of the material from a high to a low resistance state at a
critical threshold potential?" Furthermore, this device displays a memory effect in
which it remains in a low resistant state for a period of time before returning to a state of
high resistance.37M Variations in film design affect the conducting properties. The
original mechanism pr0posed by Potember et al., the discoverers of this phenomenon, is
Shown in Eq 1. Numerous techniques including Rarnan, infrared, auger, and X-ray
photoelectron spectroscopies have been used to probe the electronic state of the TCNQ

molecule in the films.

[CUTTCNQ'Hn e ICu"Ix + [TCNQolr + [Cu(TCNQ)].-. (Eq. 1)

The scientific findings surrounding this controversy led our group to suspect that
subtle changes in film preparation were responsible for the differences in the electrical
properties of the Cu(T CN Q) films. SEM images provided by Hipps et al. revealed that a
transformation from a needle morphology to a more mosaic plate-like phase was evident

at longer dipping times?" These results suggested to us that there are two polymorphs of

13

 

 

 

Cu(TCNQ
data as a

found that
redissolve '
displayed ii
and spectro

illustrate thr

 

transfonnatl

figure 3.

Cu(TCNQ), thus we began an extensive Study to ascertain structural and spectroscopic
data as a means to explain the conducting behavior. Using bulk synthetic methods, it was
found that needles of Cu(TCNQ) phase I form at short reaction times (2 min) and then
redissolve to form platelets of Cu(TCNQ) phase 11.38 SEM images of these phases are
displayed in Figure 3. These compounds have been characterized by several structural
and spectroscopic methods (XPS, 1R). X-ray powder diffraction was also used to
illustrate the sequential conversion of phase I to phase H over several hours. The

transformation is shown in Figure 4.

1 [III]

 

Figure 3. SEM images of Cu(TCNQ) phase I (top) and phase 11 (bottom)
illustrating the morphological differences.38

14

‘ l 44 l A k A A J A A A A l L A l —A l

Cu(TCNQ) phase I

   

 

 

 

 

 

 

 

 

lo 20 30 .40 so
Cu(TCNQ) in MeCN

at room temp. for 6h. _

I .

L-- . T f f j - _._ .4. '

I 2‘0 ab co 5

Cu(TCNQ) in MeCN

I at room temp. for 46h. -

It {

t'o""2h‘ii's‘of"‘4‘oi so

" Cu(TCNQ) in MeCN L

at room temp. for 76h. _

 

 

 

 

LL‘A
I'Yv‘vvv‘w

I‘MAIA

v—vIu—vw

 

 

‘llllllhllljujjl

- vv‘erITv-VI

 

 

So

 

Figure 4. X-ray powder diffraction patterns of the conversion of Cu(TCNQ)
phase I to phase H at varying time intervals.38 (Hanhua Zhao)

15

 

 

 

molecules.
than that
paramagnet

polymorph

During the course of the controversial studies of Cu(T CN Q), a similar compound,
Ag(TCNQ), was structurally characterized and subjected to magnetic measurements.39
Crystals grown by electrochemical means exhibit a structure similar to phase I of
Cu(TCNQ), in that a pseudo four—fold coordination of TCN Q is exhibited about the metal
ion. The distorted tetrahedral coordination, depicted in Figure 5, is similar to that found
in CS2(TCNQ)3. Along the a axis, there are two independent parallel stacks of TCNQ
molecules. The conductivity for Ag(TCNQ) is 3.6 X 10-4 0’1cm'l, which is much lower
than that of Cu(TCNQ). The magnetic behavior was reported to follow Curie
paramagnetism. Only one phase of Ag(TCNQ) has been reported, although

polymorphism should be possible due to its structural similarity to Cu(TCN Q).

  

2;,

 

Figure 5. The coordination environment of Ag(TCNQ).39

l6

The focus of this project is the design and synthesis of “hybrid” polymeric
materials using transition metals and TCNQ with the intention of studying their magnetic
properties. In chapter 11, the synthesis and characterization of the two phases of
Ag(TCNQ), one of them previously known, as well as the structural characterization of
the novel Ag(TCNQFa) material will be presented. Chapter III details the further
expansion of the use of the TCNQ“ ligand with divalent first row transition metals (My:
Mn, Fe, Co, and Ni). These were targeted due to the presence of multiple unpaired
electrons being available to interact with the S = 1/2 spin state of TCNQ'. The final
chapter, Chapter IV, focuses on the development of procedures for the isolation of the
two components of these bulk products. The previously reported bulk product,

Mn(TCNQ)2(MeOH)2 was investigated in an effort to separate the two crystalline phases,

[Mn(TCNQ)(TCNQ-TCNQ)o_5(MeOH)2].., and an “all O-dimer phase”, [Mn(TCNQ—

TCNQ)(MeOH)4l... Additionally, the bulk phase Mn(TCNQ)2(H20)2 phase was checked

for multiple phases and it was found that a second phase is produced in addition to the
reported one. Furthermore, the possible synthetic conversion of three similar bulk
products Mn(TCNQ)2(MeOH)2 and Mn(TCNQ)2(H20)2 to the magnetic material,

Mn(TCNQ)2 will be discussed.

l7

References:

1.
2.

McCoy, H. N.; Moore, W. C. J. Am. Chem. Soc. 1911, 33, 273.

Acker, D. S.; Harder, R. J .; Hertler, W. R.; Mahler, W.; Melby, L. R.; Bensin, R.

E.; Mochel, W. E. J. Am. Chem. Soc. 1960, 82, 6408.

(a) Jerome, D. Science 1991, 252, 1509. (b) Torrance, J. B. Acc. Chem. Res. 1979,

12, 79. (c) Bryce, M. R. Chem. Soc. Rev. 1991, 20, 355. ((1) Williams, J. M.; Schultz,
A. J .; Geiser, U.; Carlson, K. D.; Kini, A. M.; Wang, H. H.; Kwok, W. - K.; Wangbo,
M. - H.; Schirber, J. E. Science, 1991, 252, 1501.

Melby, L. R.; Harder, R. J .; Hertler, W. R.; Mahler, W.; Benson, R. E.; Mochel, W.
E. J. Am. Chem. Soc. 1962, 84, 3374.

(a) Manriquez, J. M.; Yee, G. T.; McLean, S.; Epstein, A. J.; Miller, J. S. Science
1991, 252, 1415. (b) Tamaki, H.; Zhuang, Z. J.; Matsumoto, N.; Kida, S.; Koikawa,
M.; Achiwa, N.; Hashimoto, Y.; Okawa, H. J. Am. Chem. Soc. 1992, 114, 6974. (c)
Stumfi, H. 0.; Pei, Y.; Kahn, 0.; Sletten, J.; Renard, J. P. J. Am. Chem. Soc. 1993,
115, 6738. (d) Inoue, K.; Iwarnura, H. J. J. Am. Chem. Soc. 1994, 116, 3173. (e)
Ohba, M.; Maruono, N.; Okawa, H.; Enoki, T.; Latour, J.- M. J. Am. Chem. Soc.
1994, 116, 11566. (f) Kahn, O. In Molecular Magnetism: From Molecular Assemblies
to the Devices; NATO ASI Series E321; Coronado, E.; Delhaes, P.; Gatteschi, D.;
Miller, J. S., Eds; Kluwer: Dordrecht, 1996; p. 243-288. (g) Decurtins, S.; Schmalle,
H. W.; Schneuwly, P.; Zheng, L. - M.; Ensling, J .; Hauser, A. Inorg. Chem. 1995, 34,
5501. (h) Miyasaka, H.; Matsumoto, N.; Okawa, H.; Re, N.; Gallo, E.; Floriani, C.
Angew. Chem, Int. Ed. Engl. 1995, 34, 1446. (i) Ohba, M.; Okawa, H.; Ito, T.; Ohto,
A. J. J. Am. Chem. Soc., Chem. Commun. 1995, 1545. (j) Michaut, C.; Ouahab, L.;
Bergerat, P.; Kahn, 0.; Bousseksou, A. J. Am. Chem. Soc. 1996, 118, 3610. (k) de
Munno, G.; Poerio, T.; Viau, G.; Julve, M.; Loret, F.; Journaux, Y.; Riviere, E. Chem.
Commun. 1996, 2587.

(a) Aumiiller, A.; Erk, P.; Klebe, G.; Hiinig, 8.; von Schiitz, J .; Werner, H. Angew.
Chem, Int. Ed. Engl. 1986, 25, 740. (b) Aumiiller, A.; Erk, P.; Hiinig, S. Mol. Cryst.
Liq. Cryst. 1988, I56, 215. (c) Aumiiller, A.; Erk, P.; Hiinig, S. Mol. Cryst. Liq.

18

10.
11.

12.

13.
14.
15.
16.

17.
18.
19.

20.
21.

Cryst. Inc. Nonlin. Opt. 1988, 156, 215. (d) Erk, P.; Gross, H. - J .; Hiinig, U. L.;
Meixner, H.; Werner, H. - P.; von Schiitz, J. U.; Wolr, H. C. Angew. Chem, Int. Ed.
Engl. 1989, 28, 1245. (e) Kato, R.; Kobayashi, H.; Kobayashi, A. J. Am. Chem. Soc.
1989, 111, 5224. (f) Sinzger, K.; Hiinig, S.; Jopp, M.; Bauer, D.; Beitsch, W.; von
Schiltz, J. U.; Wolf, H. C.; Kremer, R. K.; Metzenthin, T.; Bau, R.; Khan, S. 1.;
Lindbaum, A.; Lengauer, C. L.; Tillmanns, E. J. Am. Chem. Soc. 1989, 115, 7696.
Aumiiller, A.; Erk, P.; Hiinig, S.; Hadicke, E.; Peters, K.; von Schnering, H. G.
Chem. Ber. 1991, 124, 2001.

Kaim, W.; Moscherosch, M. Coor. Chem. Rev. 1994, 129, 157.

Wheland, R. C.; Martin, E. L. J. Org. Chem. 1975, 40, 3101.

Bryce, M. R. Chem. Soc. Rev. 1991, 20, 355.

Bardeen, J .; Cooper, L. N.; Schrieffer, J. R. Phys. Rev. 1957, 108, 1175.

There are different conventions used in the description of magnetic interaction. See,
for example, (a) Carlin, R. L. Magnetochemistry, Springer-Verlag: Berlin
Heidelberg, Chapter 5, 1986. (b) Mattis, D. C. The Theory of Magnetism, Springer-
Verlag: New York, 1981; Vol. 1. The convention in this thesis follows the
convention used in reference (a).

(a) Kahn, 0. Structure and Bonding, 1987, 68, 89. (b) Kahn, O. Angew. Chem, Int.
Ed. 1985, 24, 834.

Miller, J. S.; Epstein, A. J. Chemistry & Industry 1996, 49.

Torrance, J. B. J. Am. Chem. Soc. 1979, 12, 79.

Wudl, F.; Smith, G. M.; Hufnagel, E. J. J. Chem. Soc., Chem. Commun. 1970, 1453.
Htinig, S.; Kiesslich, G.; Scheutzow, D.; Zahradnik, R.; Carsky, P. Int. J. Sulfur
Chem. Part C. 1971, 109.

Coppens, P. J. Chem. Soc., Chem. Commun. 1971, 889.

Wudl, F.; Wobschall, D.; Hefnagel, E. J. J. Am. Chem. Soc. 1972, 94, 670.
(a)Ferraris, J.; Cowan, D. 0.; Walatka, V. V.; Perlstein, J. H. J. Am. Chem. Soc.
1973, 95, 948. (b)Coleman, L. B.; Cohen, M. J .; Sandman, D. J .; Yamagishi, F. G.;
Garito, A. F.; Heger, A. J. J. Solid State Commun. 1973, 12, 1125.

Engler, E. M.; Patel, V. V. J. Am. Chem. Soc. 1974, 96, 7376.

Bloch, A. N.; Cowan, D. 0.; Benchgaard, K.; Pyke, R. E.; Banks, R. H. Phys. Rev.

19

Lett. 1975, 34, 1561.

22. (a) Jerome, D.; Manzand, A.; Ribault, M.; Benchgaard, K. J. Phys. Lett. 1980, L95.
(b)Benchgaard, K.; J acobsen, C. S.; Mortensen, K.; Pendersen, H. J.; Thorup, N.
Solid State Commun. 1980, 33, 119.

23. (a) Wudl, E; Thomas, G. A.; Nalewajek, D.; Hauser, J. J .; Lee, P. A.; Poehler, T. 0.
Phys. Rev. Lett. 1980, 45, 829. (b) Scott, J. C.; Pendersen, H. J.; Benchgaard, K.

Phys. Rev. Lett. 1980, 45, 2125.

24. (a) Parkin, S. S. P.; Engler, E. M.; Schumaker, R. R.; Laiger, R.; Lee, V. Y.; Scott, J.
C.; Greene, R. L. Phys. Rev. Lett. 1983, 50, 270. (b) Yagubskii, E. B.; Schegolev, I.
F.; Laukhin, V. N .; Karatsovnik, P. A.; Karatsovnik, M. V.; Zvarykina, A. V.;
Buravov, L. I. J.E.T.P. Lett. (Engl. Trans.) 1984, 39, 12. (c) Schirber, J. B.; Azevado,
L. J .; Kwak, J. K.; Venturini, E. L.; Leung, P. C. W.; Beno, M. A.; Wang, H. H.;
Williams, J. M. Phys. Rev. B. 1986, 33, 1987. (d) Williams, J. M.; Kini, A. M.;
Wang, H. H.; Carlson, K. D.; Geiser, U.; Montgomery, L. K.; Pyrka, G. J .; Watkins,
D. M.; Kommers, J. K.; Boryschuk, S. J .; Crouch, A. V.; Kwok, W. K.; Schirber, J.
E.; Overmeyer, D. L.; Jung, D.; Whangbo, M. - H. Inorg. Chem 1990, 29, 3274.

25. Pei, Y.; Verdaguer, M.; Sletten, J .; Renard, J. P. J. Am. Chem. Soc. 1986, 108, 7428.

26. Miller, J. S.; Calabrese, J. C.; Epstein, A. J.; Bigelow, W.; Zhang, J. H.; Reiff, W. M.
J. Chem. Soc., Chem. Commun. 1986, 1026.

27. (a) Miller, J. S.; Epstein, A. J. J. Am. Chem Soc. 1987, 109, 3850. (b) Kollrnar, C.;
Kahn, 0. Acc. Chem Res. 1993, 26, 259.

28. (a) Miller, J. S.; Zhang, J. H.; Reiff, W. M. J. Am. Chem. Soc. 1987, 109, 4584. (b)
Miller, J. S.; Epstein, A. J .; Reiff, W. M. Science 1988, 240, 40. (c) Miller, J. S.;
Epstein, A. J .; Reiff, W. M. Acc. Chem. Res. 1988, 21, 114. (d) Miller, J. S.;
Calabrese, J. C.; Epstein, A. J. Inorg. Chem 1989, 28, 4230. (e) Miller, J. S.;
Glatzhofer, D. T.; O’Hare, D. M.; Reiff, W. M.; Chakraborty, A.; Epstein, A. J.
Inorg. Chem. 1989, 28, 2930. (f) Broderick, W. B.; Thompson, J. A.; Day, E. P.;
Hoffman, B. M. Science 1990, 249, 401.

29. (a) Cayton, R. H.; Chisholm, M. H.; Darrington, F. D. Angew. Chem, Int. Ed. Engl.
1990, 29, 1481. (b) Stoner, T. C.; Dallinger, R. R; Hopkins, M. D.; J. Am. Chem.
Soc. 1990, 112, 5651. (c) Bartley, S. L.; Dunbar, K. R. Angew. Chem, Int. Ed. Engl.

/‘ I-l-
.A _ .r

30.

31.

32.

33.

35.

1991, 103, 447. (d) Cayton, R. H.; Chisholm, M. H.; Huffman, J. C.; Lobkovsky, E.
B. J. Am Chem Soc. 1991, 113, 8709. (e) Huckett, S. C.; Anington, C. A.; Burns, C.
J.; Clark, D. L.; Swanson, B. I. Synt. Met. 1991, 41 ~43, 2769. (t) Dunbar, K. R. J.
Cluster Science 1994, 5, 125. (g) Dunbar, K. R.; Ouyang, X. Mol. Cryst. Liq. Cryst.
1995, 273, 21.

(a) Gross, R.; Kaim, W. Angew. Chem, Int. Ed. Engl. 1987, 26, 251. (b) Olbrich-
Deussner, B.; Gross, R.; Kaim, W. J. Organomet. Chem. 1989, 366, 155. (c) Gross-
Lannert, R.; Kaim, W.; Olbrich-Deussner, B. Inorg. Chem. 1990, 29, 5046. (d)
Schewenderski, B.; Kaim, W.; Olbrich-Deussner, B.; Roth, T. J. Organomet. Chem.
1992, 440, 145.

(a) Ballester, L.; Barrel, M. C.; Gutierrez, A.; J iménez-Aparicio, R.; Martinez-Muyo,
J. M.; Perpir‘ian, M. F.; Monge, M. A.; Ruiz-Valero, C. J. Chem. Soc., Chem.
Commun. 1991, 1396. (b) Humphrey, D. G.; Fallon, G. D.; Murray, K. S. J. Chem.
Soc., Chem Commun. 1998, 1356.

(a) Bell, S. B.; Field, J. S.; Haines, R. J .; Moscherosch, M.; Matheis, W.; Kaim, W.
Inorg. Chem 1992, 31, 3269. (b) Bell, S. B.; Field, 1. S.; Haines, R. J.; Sundermeyer,
J. J. Organomet. Chem. 1992, 427, C1.

(a) Demerseman, B.; Pankowske, M.; Bouquet, G; Bigorgne, M.; J. Organomet.
Chem 1976, 117, C10. (b) Ittel, S. D.; Tolman, C. A.; Krusic, P. J .; English, A. D.;
lesson, J. P. Inorg. Chem. 1978, 17, 3432. (c) Booth, B. L.; McAuliffe, C. A.;
Stanley, G. L.; J. Chem. Soc. Dalton Trans. 1982, 535. (d) Rockenbauer, A.; Speier,
G.; Szabo, L. Inorg. Chim. Acta 1992, 201, 5.

(a) Ewan, A.; McQueen, D.; Blake, A. J.; Stephenson, T. A.; Schroder, M.;
Yellowlees, L. J. J. Chem Soc., Chem. Commun. 1988, 1533. (b) Braunwarth, H.;
Huttner, G.; Zsolnai, L. J. Organomet. Chem. 1989, 372, C23. (c) Bunn, A. G.;
Carroll, P. J.; Wayland, B. B. Inorg. Chem. 1992, 31 , 1297. (d) Miller J. S.;
Calabrese, J. C.; McLean, R. S.; Epstein, A. J. Adv. Mater. 1992, 4, 498. (e) Yee, G.
T.; Calabrese, J. C.; Vazquez, C.; Miller, J. S. Inorg. Chem. 1993, 32, 377. (f)
Cotton, F. A.; Kim, Y.; Ren, T. Ployhedron 1993, 12, 607.

(a) Fritchie, C. 1., Jr.; Arthur, P. Acta Crystallogr. 1966, 21 , 139. (b) Hoekstra, A.;
Spoelder, T.; Vos, A. Acta Crystallogr. 1972, 328, 14. (c) Konno, M.; Saito, Y. Acta

21

36.
37.

Crystallogr. 1974, B30, 1294. (d) Murakarni, M.; Yoshimura, S. Bull. Chem. Soc.
Jpn. 1975, 48, 157. (e) Konno, M.; Saito, Y. Acta Crystallogr. 1975, B31, 2007. (f)
Konno, M.; Ishii, T.; Saito, Y. Acta Crystallogr. 1977, B33, 763.

Siemons, W. J.; Bierstedt, P. E.; Kepler, R. G. J. Chem Phys. 1963, 39, 3523.

(a) Potember, R. S.; Poehler, T. 0.; Cowan, D. 0. Appl. Phys. Lett. 1979, 34, 405.
(b) Potember, R. S.; Poehler, T. 0.; Rappa, A.; Cowan, D. 0.; Bloch, A. N. J. Am.
Chem Soc. 1980, 102, 3659. (c) Potember, R. S.; Poehler, T. 0.; Cowan, D. 0.;
Brant, P.; Carter, F. L.; Bloch, A. N. Chem. Scr. 1981, 17, 219. (d) Kamistos, E. 1.;
Risen, W. M., Jr. Solid State Commun. 1982, 42, 561. (e) Potember, R. S.; Poehler,
T. 0.; Cowan, D. 0.; Carter, F. L.; Brant, P. I. In Molecular Electronic Devices;
Carter, F. L., Ed.; Marcel Dekker: New York, 1982; p. 73. (t) Potember, R. S.;
Poehler, T. 0.; Benson, R. C. Appl. Phys. Lett. 1982, 41, 548. (g) Potember, R. S.;
Poehler, T. 0.; Rappa, A.; Cowan, D. 0.; Bloch, A. N. Synth. Met. 1982, 4, 371. (h)
Kamitsos, E. I.; Risen, W. M. Solid State Commun. 1983, 45, 165. (i) Kamitsos, E.
1.; Risen, w. MJ. Chem. Phys. 1983, 79, 477. (j) Kamitsos, E. 1.; Risen, w. M. Jr. J.
Chem Phys. 1983, 79, 5808. (k) Benson, R. C.; Hoffman, R. C.; Potember, R. S.;
Bourkoff, E.; Poehler, T. 0. Appl. Phys. Lett. 1983, 41, 548. (l) Poehler, T. 0.;
Potember, R. S.; Hoffinan, R.; Benson, R. C.; Mol. Cryst. Liq. Cryst. 1984, 107, 91.
(m) Kamistos, E. I.; Risen, W. M., Jr. Mol. Cryst. Liq. Cryst. 1986, 134, 31. (n)
Potember, R. S.; Poehler, T. 0.; Hoffman, R. C.; Speck, K. R.; Benson, R. C. In
Molecular Electronic Devices 11; Carter, F. L., Ed.; Marcel Dekker: New York,

1987; p. 91 (o) Wakida, S.; Ujihira, Y. J. Appl. Phys. 1988, 27, 1314. (p) Hoffman,
R. C.; Potember, R. S. Appl. Opt. 1989, 28(7), 1417. (q) Duan, H.; Mays, M. D.;
Cowan, D. 0.; Kruger, J. Synth. Met. 1989, 28, C675. (r) Sato, C.; Wakamatsu, S.;
Tadokoro, K.; Ishii, K. J. Appl. Phys. 1990, 68 (12) 6535. (s)Yarnaguchi, S.; Viands,
C. A.; Potember, R. S. J. Vac. Sci. Technol. 1991, 9, 1129. (t) Hua, Z, Y.; Chen, G.
R. Vacuum 1992, 43, 1019. (u) Hoagland, J. J.; Wang, X. D.; Hipps, K. W. Chem.
Mater. 1993, 5, 54. (v) Liu, S. G.; Liu, Y. Q.; Wu, P. J.; Zhu, D. B. Chem. Mater.
1996, 8, 2779. (w) Liu, S. G.; Liu, Y. Q.; Zhu, D. B. Thin Solid Films 1996, 280,
271. (x) Sun, S. Q.; Wu, P. J .; Zhu, D. B. Solid State Commun. 1996, 99, 237. (y)

22

Liu, S. G.; Liu, Y. 0.; Wu, P. J .; Zhu, D. B.; Tian, H.; Chen, K. C. Thin Solid Films
1996, 289, 300.

38. Heintz, R. A.; Zhao, H.; Ouyang, X.; Grandinetti, G.; Cowen, J .; Dunbar, K. R.
Inorg. Chem 1999, 38, 144.

39. Shields, L. J. Chem. Soc., Faraday Trans. 2 1985, 81, l.
40. Zhao, H.; Heintz, R. A.; Ouyang, X.; Dunbar, K. R. Chem. Mater. 1999, 11, 736.

23

Chapter II

Novel Polymeric Crystalline Phases of
Ag(TCNQ) and Ag(TCNQF4):

Structures and Magnetic Studies

24

Introduction

The use of transition metal precursors as building blocks in coordination
compounds allows for impressive Structural diversity and opens up a wealth of potential
applications for porousl, magneticz, and conducting3 solids.” In many cases, knowledge
of the geometry of the ligand and the coordination environment of the metal has allowed
synthetic chemists to design new strategies and to predict the resulting structural
architecture. This concept works well when polymorphism is not an issue, but this is
unfortunately not the case with It-stacked charge-transfer materials. Conducting charge

transfer materials composed of stacks of donor cations and organic radical acceptors such

 

 

TCNE DM-DCNQI
F
N N
N N
TCNQ TCNQF4

Figure 6. Organonitrile acceptor molecules used as ligands in
metallopolymers.

25

as those exhibited in Figure 6 have been under investigation for over forty years, yet there
is still an element of structural serendipity that cannot be denied?"5 In addition to
classical organic or organometallic donor/acceptor salts that involve It-stacks, polymeric
materials composed of a metal ion directly bonded to the organic radical may be
envisioned. Such direct bonding between the metal donor and the organic acceptor
allows for a combination of the unique characteristics of the transition metal (optical,
magnetic, etc.,) with the conducting properties of organic radical stacks.5‘°'h‘”“""° A
foremost example of this approach is the use of 2,5-DM-DCNQI (Figure 6b) with Cu
(LII) in the assembly of tetrahedral metal based, three-dimensional networks.3a The
mixed-valence material, Cu(DM-DCNQDz, exhibits an extraordinarily high metallic
conductivity (5x105 S cm'1 at 3.5 K) which is related to the presence of the 1-D stacked
DCNQI radicals as well as to interactions of the metal with the ligand via metal (1 and
ligand p orbital overlap. The latter interactions serve to increase the dimensionality of the
system and lead to increased conductivities in two-dimensions. The Cu(DCNQI)2 system
is an excellent example of how inorganic/organic hybrid materials of the coordination
polymer type can be competitive in the area of highly conducting materials. 3

The incorporation of “bare” metal ions into networks with organic acceptors
spawned technological, development of binary metal/1‘ CNQ films that behave as switches.
Two of the most well-investigated materials in this regard are Cu(TCNQ) (T CNQ’ =
7,7,8,8-tetra—cyanoquinodimethanide) and Ag(TCNQ) which exhibit electric-field-
induced bistable switching under certain conditions.7‘8 Specifically, the application of an
electric field across a thin film of these materials leads to an observable “switching” from

a high- to a low-resistance state (Figure 7) at a critical threshold potential.8a For over

26

twenty years there has been intense debate in the literature concerning the mechanism of
the switching as well as the role of film growth conditions on the resulting properties.
Earlier this year, our group reported the X-ray structures of two different forms of

Cu(TCNQ) with very different magnetic and conducting

 

 

 

 

 

 

 

 

 

 

 

[ High .Rdogfl'stance Memory Effect
Resistance
Low Resistance
'0‘.
Time

Figure 7. Schematic drawing showing the change of resistance over time
for Cu(TCNQ) films.

properties.9 This unrecognized polymorphism in the Cu system is unfortunate, as it led to
many years of fruitless work on mixtures of two compounds. Clearly, further evaluation
of other metal/TCNQ materials is in order. A search of the literature revealed that the
switching behavior of both Cu and Ag devices has been studied at the same time, but that
Ag(TCNQ) films have not been scrutinized as carefully as the Cu(TCNQ) system. In
terms of structural evidence, however, until recently the Ag analog was better
characterized. In 1985, Shields reported the single-crystal X-ray structure of a
Ag(TCNQ) crystal obtained by electrochemical synthesis.7 No reports of powder data or

comments on the singularity of the electrochemical form of Ag(TCNQ) have appeared

27

since this time, however, therefore it is logical to ask if the reported form of Ag(TCNQ)
is the only stable phase. In this paper we explore the issue of polymorphism in
Ag(TCNQ) and in the corresponding Ag(TCNQFa) material. A combination of infrared
spectroscopy, electron microscopy, X-ray crystallography, and magnetic susceptibility

methods were used to characterize the new compounds.

Experimental

A. Synthesis

Reactions were carried out under a dinitrogen atmosphere unless otherwise
indicated. Acetonitrile was dried over 3 A molecular sieves and distilled under a nitrogen
atmosphere prior to use. Ag(BF4) was purchased from Aldrich Chemical Company, and
TCN Q was purchased from TCI Chemical Company and recrystallized from hot
acetonitrile. The salt [n-BuaN][TCNQ]” and neutral TCNQE.15 were prepared according
to literature methods, while preparation of the reduced anion salts, Li(TCNQFa) and [n-
BuaN] [TCNQFa] were adapted from the original preparations of the corresponding
TCNQ salts. Since Ag cation salts are prone to decomposition to metallic Ag, reaction
flasks were covered with aluminum foil prior to and during reactions to prevent light

induced reduction to Ag+ to Ag.

28

a) Bulk Synthesis of Ag(TCNQ) Phase I, (1).

Ag(BF4) (0.20 g, 1.03 mmol) was dissolved in 20 mL of acetonitrile and filtered
through Celite under a nitrogen atmosphere. Separately, [n-BuaN][TCNQ] (0.405 g,
0.906 mmol) was dissolved in 20 mL of acetonitrile and slowly added to the Ag(BFa)
solution. A blue-purple microcrystalline product precipitated immediately, and, after two
minutes, it was collected by filtration, washed with 10 mL of acetonitrile followed by 20
mL of diethyl ether and dried in vacuo; yield 0.254 mg, 90%. Anal. Calcd for
C12H4N4Ag: C, 46.19; H, 1.29; N, 17.95; Found: C, 45.38; H, 1.32; N, 17.51.

Characteristic IR data (Nujol mull, KBr plates, cm’l): v(CEN), 2199s, 2194sh, 2184s,

2161s; v(C=C), 1507s; 5(C-H), 8243.

b) Bulk Synthesis of Ag(TCNQ) Phase II, (2) from Ag metal.

Ag powder (0.18 g, 1.67 mmol) was suspended in 150 mL of acetonitrile, TCNQ
(1 g, 4.89 mmol) was added, and the mixture was stirred in the dark for 2 days. After
filtration, the reddish purple nricrocrystalline product was washed with acetonitrile and
suspended in a 150 mL TCNQ/acetonitrile solution (0.5 g, 2.44 mmol) to remove any
unreacted Ag metal. After additional days of stirring in the dark, the solution was
decanted to remove any small particles suspended in the solution. The product was
washed repeatedly with acetonitrile, filtered and dried in vacuo; yield 0.45 g, 86.5%.

Anal. Calcd for C12H4N4Ag: C, 46.19; H, 1.29; N, 17.95; Found: C, 45.94; H, 1.38; N,
17.84. Characteristic IR data (Nujol mull, KBr plates, cm'l): v(CEN), 2202Sh, 21855,

21705, 213lsh; v(C=C), 15058; 5(C-H), 820s.

c) Synthesis of [n-BmNHTCNQ] (3).

A quantity of Li(TCNQ) (1.48 g, 0.007 mol) was dissolved in 100 mL of hot
distilled water. In a separate flask an excess of [n-BuaN]I (5.2 g, 0.0140 mol) was
dissolved in 600 mL of boiling distilled water. Upon the addition of Li(TCNQ) to the [n-
BuaN]I solution, an instantaneous reaction ensued with the precipitation of a dark purple
solid. The stirred solution was warmed for one hour, after which time the product was
collected by vacuum filtration. The dark purple powder was washed with several aliquots
of hot water followed by diethyl ether, and dried in vacuo: yield, 3.1 g (99%). The
product was recrystallized as follows. The crude product was dissolved in a minimal
volume of dichloromethane (~100 mL), and the solution was dried with anhydrous
magnesium sulfate. After filtration, the solution was reduced in volume (~30 mL), and an
equivalent volume of hexanes was added to induce crystallization. After the excess
dichloromethane had been evaporated, the dark purple crystalline product was collected
by vacuum filtration, washed with copious quantities of hexanes and dried in vacuo:

yield, 2.96 g, (95%). Anal. Calcd for C23H36N5: C, 75.29; H, 9.03; N, 15.68; Found: C,
73.95; H, 9.52; N, 15.32. Characteristic IR data (Nujol mull, KBr plates, cm'l): v(CEN),

2187 sh, 2181 s, 2160 sh, 2157 s; v(C=C), 1507 s; 5(C-H), 824 m.

(1) Synthesis of [n-BmNHTCNQFa] (4).

[n-BuaN]I (1.69 g, 0.458 mmol) was dissolved in 200 mL of hot distilled water.

Subsequently, Li[TCNQF4] (0.650 g, 0.230 rnmol) was dissolved in 50 mL of distilled

water. Upon addition of the Li[TCNQF4] solution to the [n-BuaN]I solution, a dark
purple microcrystalline product immediately precipitated. After one hour of stirring, the
product was collected by filtration. The crude product was dissolved in methylene
chloride, dried with magnesium sulfate, and the solution was filtered. The volume was
reduced by rotoevaporation, and 40 mL of hexanes was added to induce precipitation of
the product. Crystals of [n-Bu4N][TCNQF4] suitable for X-ray structural determination
were obtained directly from this procedure. To recover more product, the methylene
chloride filtrate was then evaporated which gave an additional crop of dark purple
crystals of [n-BuaN][TCNQF4]. These were collected by filtration, washed with hexanes
and dried in vacuo; yield 1.02 g, 86%. Anal. Calcd for C2gN5H36F4: C, 64.84; H, 7.00; N,

13.50; Found: C, 64.11; H, 7.63; N, 12.36. Characteristic IR data (Nujol mull, KBr

plates, cm"): v(GN), 2197s, 2174s, 2154s, 2134Sh; v(C=C), 1506 sh, 1498 s.

e) Bulk Synthesis of Ag(TCNQFa) (5).

Ag(BFa) (0.135g, 0.693 mmol) was dissolved in 20 mL of acetonitrile and filtered
through Celite, and, separately, [n-BuaN] [TCNQFa] (0.363, 0.70 mmol) was dissolved in
20 mL of acetonitrile. Both solutions were chilled to -20 °C and the solution of [n-
BuaN][TCNQF4] was added to Ag(BFa) and stirred for two minutes. A dark blue purple
microcrystalline powder was collected by filtration, washed with 10 mL of acetonitrile,
followed by 20 mL of diethyl ether, and dried in vacuo; yield 0.238 g (92%). Anal. Calcd

for C12F4NaAg: C, 37.53; H, 0.0; N, 14.95; Found: C, 37.12; H, 0.01; N, 14.92.

31

Characteristic IR data (Nujol mull, KBr plates, cm‘l): v(CEN), 22205, 22135, 21975;

v(C=C), 15015; v(C-F), 1208m, 1154w.

B. Structural Studies

a) X-ray Structural Determination of [n-BuaN][TCNQF4] (4).

A typical needle crystal of dimensions 0.80 X 0.12 x 0.10 mm3 was mounted on
the end of a glass fiber with silicone grease. A full sphere of data was collected at 173:l:(2)
K for a [n-BuaN][TCNQF4] crystal on a Siemens SMART 1K CCD area detector
diffractometer with a graphite monochromated Mo KO: radiation (7tfl = 0.71073 A). The
frames were integrated in the Siemens SAINT software package, and the data were
corrected for absorption using the SADABS program. The data were integrated in a P21/n
monoclinic cell setting a = 7.424405) A, b = 20.212(4) A, c = 19.758(4) A, B =
97.32(3)°. The structure was solved using the SHELXS program, and refinement was
carried out by full matrix least—squares calculations on F2 using the SHELXL - 97
program. All the atoms were refined anisotropically to give final residuals of R1 = 0.0507
and a wR2 of 0.0946 at a resolution of 0.75 A. The final refinement was based on 479
parameters and 35072 reflections, 7113 of which were unique. The highest peak in the

final difference map was 0.224 e'/A3.

b) X-ray Structural Determination of Ag(TCNQFa) (5).
A typical rectangular platelet crystal of dimensions 0.31 x 0.16 x 0.08 mm3 was

mounted on the end of a glass fiber with silicone grease. A hemisphere of data was

32

collected at 1731(2) K for a Ag(TCNQFa) crystal on a Siemens SMART 1K CCD area
detector diffractometer with a graphite monochromated Mo KOt radiation (A.x = 0.71073
A). The frames were integrated with the Siemens SAINT software package, and the data
were corrected for absorption using the SADABS program. The data were integrated in a
C2/c monoclinic cell setting a = 13.429(3) A, b = 6.9331(14) A, c = 25.735(5) A, [3 =
116.66(3)°. The structure was solved using the SHELXS program and refinement was
carried out by full matrix least-squares calculations on F2 using the SHELXL97 program.
All the atoms were refined anisotropically to give final residuals of R1 = 0.0225 and a
wR2 of 0.0586 at a resolution of 0.75 A. The final refinement was based on 191
parameters and 6416 reflections, 2526 of which were unique. The highest peak in the

final difference map was 0.811 e°/A3.

C. Physical Measurements.

Infrared spectra were recorded as Nujol mulls on KBr plates using a Nicolet IR/42
FT-IR spectrometer. Field Emission Scanning Electron Microscopy (FESEM)
measurements were performed on a Hitachi Ltd., Model 8470011 with a Tungsten
filament housed in the W.M. Keck Microfabrication Facility located in the Department of
Physics & Astronomy at Michigan State University. X-ray powder diffraction patterns
were collected on a Rigaku RUZOOB X-ray powder diffractometer with Cu KOt radiation.
The variable temperature magnetic susceptibility data were collected in the range 2-300 K
using a Quantum Design, Model MPMS-5 SQUID magnetometer housed in the Physics

& Astronomy Department at Michigan State University. The data were corrected for the

33

diamagnetism of the sample as evaluated from the Pascal constants“ and for the sample

holder.

Results and Discussion

The controversy over the phenomenon of the bistable switching ability of
Cu(TCNQ) films“, along with our recent verification of two distinct Cu(TCNQ) phases9
prompted us to study other related metal/TCN Q systems. Reports of switching behavior
for devices containing Ag and organic acceptorsgb’g’” such as TCNQF4, 2,5-dimethoxy-
7,7,8,8-tetracyanoquinodimethane (TCNQ(OMe)2), 1 1,1 1,12,12-tetracyano—2,6-
napthoquinodimethane (TNAP), and tetracyanoethylene (T CNE) point to this metal as a
logical choice for synthetic and structural studies. To date, single crystal X—ray data have
been published for one phase of Ag(TCNm phase II which was grown using
electrochemical methods.8 Since the stoichiometry of Ag(TCNQ) is identical to
Cu(TCNQ), namely 1:1, it seemed a likely prospect to exhibit more than one polymorph
as well. Indeed, during our investigation of the bulk synthesis of Ag(TCNQ) another
phase was discovered, referred to as phase I, the existence of which is supported by X-ray
powder diffraction, spectroscopic, and magnetic measurements. The original phase of
Ag(TCNQ) discovered by Shields is designated as Phase 11. Unlike the Cu(TCNQ)
phases, which exhibit very different morphologies, both phases of Ag(TCNQ) exhibit

similar needle morphologies. Powder diffraction reveals that the structures are different

and indexing reveals that phase I is tetragonal as opposed to phase II which is
orthorhombic.

In addition to sorting out the polymorphism of the Ag(TCNQ) materials, we also
extended this chemistry to TCNQF4 (T CNQF4: 2,3,5,6-tetrafluoro-7,7,8,8-tetra-
cyanoquinodimethanide). Although there are several reports of structural”, magnetic",
and spectroscopic” studies of TCNQFa charge transfer salts in the literature, the present
work constitutes the first structurally characterized metal/TCNQF4 polymer. Analysis of
the solid—state structure of Ag(TCNQFa) indicates that it is distinctly different from either
phase of Ag(TCNQ) or the alkali metal TCNQ salts13 . Interestingly, there appears to be
only one phase of Ag(TCNQFa) which one obtains regardless of reaction conditions.

The starting material, [n-BuaN][TCNQF4], was also subjected to X-ray studies in
order to compare the stacking motif of the TCNQF4 anions in this simple organic salt

with the patterns found in the Ag polymers.

A. Synthesis of Bulk Phases

The aims of this study were (1) to determine whether a second Ag(TCNQ)
polymorph exists and (2) to obtain pure samples of the materials for physical properties
measurements. In accord with the literature report, we verified that samples of

Ag(TCNQ) can be prepared in bulk using electrochemical methods with Ag metal. (Eq.
1)

Ag +TCNQ —> Ag(TCNQ) (Eq. 1)
A more direct route is a simple metathesis reaction as shown in Eq. 2. This method is

advantageous for several reasons: the metal and ligand are in the desired oxidation states,

35

the reagents are completely soluble, and the product precipitates upon formation which
allows it to be separated from soluble by—products. Furthermore, this reaction leads to

essentially quantitative yields.

Ag(BF4)+In-BH4NIITCNQI —> Ag(TCNQ) (S) + [N-BWNIIBF4] (139-2)

Variations in reaction times and temperatures do not Significantly alter the outcome of the
reaction in Eq. 2. The favored product is the previously unrecognized phase of
Ag(TCNQ) phase 1(1). With very long reaction times and a large volume of solvent it is
possible to pass from phase I to phase II, but the process is not clean, as large quantities
of Ag metal are also formed in the process. It appears that the conversion involves a
decomposition of phase I to Ag metal and neutral TCN Q, which then react to form phase
II in situ. Isolation of pure phase II is not feasible by this method. The best route to

phase II is, in fact, the direct reaction of Ag(s) and neutral TCNQ. (Eq. 3)

Ag (5) + TCNQ —> Ag(TCNQ) (s) (2) (Eq. 3)

Interestingly, only one phase of Ag(TCNQF4) (4) appears to exist. Reaction conditions
were varied in a manner similar to the Ag(TCN Q) chemistry, but all reactions yielded
only one phase. Furthermore, slow electrolysis of TCNQF4 at a Ag electrode yielded
crystals whose structure is identical to that of the bulk product as judged by powder X-ray

methods.

36

B. Infrared Spectroscopy
Infrared spectroscopy is a useful technique for characterizing TCNQ charge
transfer salts, especially with respect to distinguishing the presence of TCNQ in it’s

neutral versus reduced forms.5f

As expected, the IR spectra of the two phases of
Ag(TCNQ) are similar, owing to their structural Similarities and their basic formulation
as compounds of TCNQ’. Table 1 summarizes the pertinent data for compounds (1 - 5).
In the v(CEtN) region, several stretching modes indicate the presence of TCNQ‘.
Ag(TCNQ) phase I exhibits strong, broad absorptions at 2199, 2194, 2187, and 2162 cm'I
while Ag(TCNQ) phase II shows a similar pattern but at slightly lower frequencies, viz.,
2202, 2185, 2170, 2131 cm". These features are lower in energy than the v(C.=.N) stretch
of neutral TCNQ which occurs at 2222 cm", and are characteristic of TCNQ’.
Ag(TCNQF4) exhibits strong absorptions at 2220, 2212, 2197 cm'l in accord with data for
salts containing TCNQFf.” Neutral TCNQF, occurs at 2227 cm".'5 The v(C=C)
stretching region is characteristic for the TCNQ phenyl ring. The rt-bond delocalization
in the ring results in one strong (C=C) Stretch ranging from 1500-1510 cm’1 for TCNQ’.
Ag(TCNQ) (phases I and II) and Ag(TCNQFa) spectra exhibit v(C=C) features at 1507,
1505, and 1501 cm" respectively. The 8(C-H) bending mode is also a sensitive indicator
of the presence of TCNQ’, TCNQZ', and mixed-valence stacks of TCNQ‘ITCNQ.19
Ag(TCNQ) phases I exhibits a weak absorption at 824 cm’l while phase II exhibits a
lower energy mode at 820 cm". Both frequencies are within the range reported for

reduced TCNQ. The absence of peaks corresponding to bound acetonitrile or [BF4]"

anions indicate that the material is composed solely of Ag-l- and TCNQ‘ ions.

37

Table 1. IR data for Cu(TCNQ). Ag(TCNQ), Ag(TCNQF4),

[n-BuaN][TCNQ], and [n—Bu.N][TCNQF,].

 

 

 

 

TR SPECTRAL DATA
COMPOUND v(CEN) v(C=C) 5(C-H) Refs.
(cm'l) (cm'l) (cm")
Cu(TCNQ)
phase I 2199 5, br; 2172 s 825 s 9
phase II 2211 s, 2172 s 825 s 9
Ag(TCNQ) 2199 5, 2194 Sh, 2184 s, 1507 s 824 5 this work
phase I 2161 s
phase II 2202 sh, 21855, 2170 s, 1505 s 820 5 this work
2131 sh
Ag(TCNQFa) 2220 s, 2213 s, 2197 s 1501 s _ this work
[n-BuaN][TCNQ] 2188 sh, 2180 5, 2162s , 1507 s 824 5 this work
2157 5
[n-BuaN][TCNQF4] 2197 s, 2174 s, 2157 S, 1506 sh, _ this work
2134 5h 1498 s

 

C. Structural Studies

A determination of the solid-state structures of the metal/T CN Q materials is
crucial to fully understanding their unusual switching behavior. Despite several attempts
to obtain single crystals of Ag(TCNQ) phase I (1), small crystal size and/or twinning
prevented a crystal structure analysis.
useful for discerning the differences between phase I and phase II. X-ray quality crystals

of Ag(TCNQFa) (S) and the starting material, [n-BuaN][TCNQF4] (4), was isolated and

the data are presented here.

38

X-ray powder diffraction, however, was very

 

a) Field Emission Scanning Electron Microscopy (FESEM)

Microscopy techniques were very useful for viewing the morphological
differences in the two phases of Cu(TCNQ).9 FESEM images of Ag(TCNQ) phase I and
II (shown in Figures 8 and 9 respectively) indicate that both exhibit needle-like
morphologies, although it is clear that phase II is much more crystalline. The images of
Single crystals of Ag(TCNQFa) and [n-BuaN][TCNQF4] are illustrated in Figure 10. The
rectangular morphology of Ag(TCNQFa) is distinct from the needle morphology of [n-
BuaN][TCNQF4]. Both crystals are large in size and thus diffract well. The single crystal
of Ag(TCNQFa) is a representative size for those obtained through electrocrystallization,
while the crystal of [n-BuaN][TCNQF4] is representative of those obtained in the bulk
synthesis. Unfortunately, one must be careful when choosing crystals for X-ray analysis.
It is apparent, especially with the crystal of [n-BuaN ] [T CNQFa], that a surface
unevenness or growth of “daughter” crystals on the surface causes problems in data

collection. A smooth, nicely shaped crystal is more suitable for structural analysis.

39

llllllll-Illl
;;;I_I_,m

 

Figure 8. FESEM photographs of Ag(TCNQ) phase I at various
magnifications.

40

 

Figure 9. FESEM photographs of Ag(TCNQ) phase II at various
magnifications.

41

 

lillllillll
200um

8.0kV 16.7mm x150 SE(U) 1017/1999 20:21

Ag(TCNQF4)

6.0kV 17.0mm x70 SEIU) 9l14/99 21 :58 1% 429um

 

[n-Bu4NIITCNQF4]

Figure 10. FESEM images of Ag(TCNQF4) and [n-Bu4N][TCNQF4].

42

b) Powder X-ray Diffraction of Chemically Prepared Phases.

X-ray powder diffraction was used extensively in our efforts to identify
Ag(TCNQ) phase I and to distinguish it from phase H. Since the crystal structure of
phase H was already known when we began our studies, it was important for us to
compare both bulk phase products with the simulated powder pattern for the known
material. Figure 11 displays powder diffraction patterns of the bulk phases and a
simulated pattern based on the single crystal X-ray data reported by Shields for the
electrochemically obtained crystals. Samples were analyzed using a step program (12 5
per 0.02° at 20) on a Rigaku RU200B X-ray powder diffractometer with Cu Ka radiation
(Au = 1.54050 A). The patterns were indexed using the Treor90 program.20 This led to
the tetragonal cell with a = b = 12.1422A, c = 17.0498A , and v = 2513.70A’ for
Ag(TCNQ) phase I and an orthorhombic cell with a = 17.4448A, b = 16.5790A , c =
7.2892A, and V = 2108A3 for the bulk Ag(TCNQ) phase H product . The indexed
parameters for Ag(TCN Q) phase H are comparable to the lattice parameters published for
the Ag(TCNQ) crystal (a = 6.975A, b = 16.686A, c = 17.455A, and v = 2031.5Ai‘).7
These results support the fact that the two phases are not the same, and that the bulk
Ag(TCN Q) phase H pattern corresponds to that of the crystal structure.

Attempts to convert phase I of Ag(T CN Q) to phase H were not very successful.
These experiments were originally based on the idea that these phases would be related to
each as found for Cu(TCNQ), viz., that a kinetic product can be converted to a
thermodynamic product with heating in solution, but there is no evidence that this is the

case for Ag(TCNQ). As indicated by the powder data in Figure 12, a mixture is obtained

43

and it appears that phase I (which is only slightly soluble) does not readily convert to

phase H, except by formation of Ag metal and neutral TCN Q.

 

Ag(TCNQ) phase I
Bulk Product

 

 

v 1 i v v i v ' v v v v ' v v v v ' v v v v ' v

Ag(TCNQ) phase ll
Bulk Product

 

 

 

 

Ag(TCNQ) phase II
Simulated from crystal

 

 

 

 

5 10 15 20 25 30 35 40 45 so
29

 

Figure 11. Powder patterns for the two phases of Ag(TCNQ) and a
simulation of electrochemically prepared Ag(TCN Q) from the literature.

 

I = Ag(TCNQ) Phase I
II = Ag(TCNQ) Phase II

 

‘HH

 

 

 

 

 

 

. ' ‘
IT‘IUYTTIIUIYTTTI'T'IIfUIH'UTIUT'Il—FUTUIIT'I

10 20 30 40 50
20

Figure 12. Powder diffraction pattern illustrating the attempted conversion
of Ag(TCNQ) phase Ito phase H.

This conclusion is based on the observation that suspensions of Ag(TCNQ) phase I in
acetonitrile that are not heated and which are protected from light do not convert to phase
H. On the other hand, if the solutions of phase I are warmed and exposed to light, phase
H and Ag metal are formed as deduced from powder X-ray data. In light of our findings,
it is not possible to cite either of the two phases as being the kinetic or thermodynamic
product. It may be that low solubility prevents the conversion of phase I to phase H, or it
may simply be that phase H forms only by direct reaction of Ag metal and TCN Q.

As a final comparison of the various structures possible for 1:1 binary M(TCNX)

materials, we examined the powder diffraction pattern of Ag(TCNQFa) illustrated in

Figure 13. A comparison is made to Ag(TCNQ) phase H which reveals that they are not
isostructural compounds. Indeed this was verified by single crystal X-ray methods as

well. The pattern also shows that Ag(TCNQFa) is highly crystalline and diffracts well

even at high angles.

Ag(TCNQFt)
Simulated
i From Crystal

T I T r

10 20 30 40 50
29

 

Ag(TCNQF4)
Bulk Product

 

 

 

 

 

' v v w v ' w v v v I w v v v' I v ‘—

 

 

 

 

 

 

 

 

 

Figure 13. Comparison of the powder diffraction patterns of Ag(TCNQFa)
and the simulated powder diffraction pattern from the Ag(TCNQF4) crystal
structure.

c)Single Crystal Diffraction Studies
(i) Single Crystal Growth

Numerous attempts to grow sufficiently large single crystals of Ag(T CN Q) phase
I (1), including slow diffusion in solvents and gels as well as bulk electrolysis, were
unsuccessful. Tiny crystals were obtained in some cases, but they were invariably
plagued with twinning or weak diffraction problems. On the other hand, crystals of
Ag(TCNQFa) were easily obtained by electrocrystallization in a manner similar to what
was employed by Shields to prepare crystals of Ag(TCNQ) phase H (2). Electrochemical
reduction of TCNQF4 was carried out at room temperature by slow reduction of TCNQF4
at a Ag electrode (1mm in diameter and 15 mm long) in a two-compartment electrolysis
cell filled with 20 mL of acetonitrile, which functioned as the solvent and the supporting
electrolyte. The largest crystals were obtained with 15 mg of TCNQFa and a constant
current of 3 ILA. The light yellow solution turned dark green with concomitant deposition
of crystals on the Ag electrode. After a period of 2 weeks (corresponding to the
consumption of 80% of TCNQFa), crystals were collected and washed with acetonitrile,
and dried in vacuo. X-ray quality crystals of [n-BuaN][TCNQF4] were obtained directly

from the bulk product.

(ii) Structural analysis of Ag(TCNQFa).

The structure was solved using direct methods. A total of 2526 unique reflections
were refined isotropically on F2 to R1(wR2) = 0.0225 (0.0586). Surprisingly, the
Ag(TCNQF) structure is not isostructural to Ag(TCN Q) phase H. A pseudotetrahedral

geometry (Figurel4) is invoked about the silver atoms due to the four-fold coordination

47

 

 

 

 

Figure 14. Four-fold coordination within Ag(TCNQFa).

of the TCNQFa" ligands. The N-Ag—N angles range between 97.84° and l30.96° and
indicate that the structure is highly distorted. This tetrahedral geometry is characteristic of
Ag(TCNQ) phase II7 and Cs2(TCNQ)3'3", and is unlike the coordination of Na“ '32 K+ ‘3‘,
or Rb+ ‘3” in the alkali metal family salts of TCNQ". In Figure 14 it is also evident that
the TCNQFa ligands are almost perfectly eclipsed which results in a high degree of ligand
dimerization. Ag(TCNQFa) is distinct in that adjacent TCNQFa' stacks are not rotated
90° with respect to each other, which is a common feature in binary metal/TCNQ salts7’13.

Instead the two-fold rotation axis generates layers of parallel TCNQF4 anions which form

columnated stacks along the b axis (Figure 15). Within these stacks, a pattern of dimers

48

emerges with two average interplanar distances, namely, 3.11 and 3.33 A. These

distances are not only less then the van der Waals distance of 3.4 A for carbon

 

Figure 15. Ligand rt-stacking along the b axis in Ag(TCNQF4).

atoms, but are the shortest values for any M(TCNQ) salt (M = Ag or alkali metal). It is
important to point out that since this is the first M(TCNQFa) compound to be structurally
characterized, one is limited to making comparisons to TCN Q structures. The only
structure with a comparable short intra-layer stacking distance (between two TCN Q
ligands) is Mn(TCNQ)2(H20)2 which is 3.05 A.21 The short distances in the
Ag(TCNQFa) structure are attributed to the presence of strong electron withdrawing
fluorine groups on the benzene ring which are known to lead to increased dimerization of

TCNQF4 even in solution.

49

(iii) Structural Analysis of [rt-BluNIITCNQFJ.

Crystals of this starting material were obtained from the bulk synthesis in which a
slow precipitation was induced by addition of hexanes. This structure was solved in the
monoclinic space group, P21/n, using direct methods. A total of 7113 unique reflections
were refined isotropically on F2 to R1(wR2) = 0.0507(0.0948). This structure consists of
a diagonal stacking pattern for the TCNQFa‘ molecules which are arranged parallel to the
a axis. Stacking distances of 3.09A and 3.73 A indicate that a dimerization occurs only
within pairs of TCNQF4 anions. This distance is shorter than the Van der Waals distance
of 3.4 A for carbon atoms as well as those reported for other simple ammonium TCNQ
salts.22 This is the first simple ammonium salt reported for TCNQFa, thus a comparison

to other TCNQFa structures is not possible.

D. Magnetic Behavior

According to the solid-state structures which indicate that the TCNQ'l anions in
the two Ag(TCNQ) polymorphs are strongly dimerized, magnetic susceptibility
measurements (Figure 16) reveal only temperature independent paramagnetism of 1.9 10'
3 (phase H) and 1.4 10'3 emu CGS/mol (phase I). A small Curie contribution at low
temperatures is a signature of defects (or S = V2 paramagnetic impurities) in the samples
and is typical of TCNQ'l materials (Table H). Similar behavior was observed for
Ag(TCNQFa) (Figure 16) with a tip. of 0.910'3 emu CGS/mol. In all these Ag
compounds, the strongly dimerized nature of the organic acceptor is such that the thermal

population of the triplet state (S = 1) cannot be observed in the accessible temperature

0.005 n...”.......,.........,...r,
o Ag(TCNQ) phaseH 1

_ 1
(mm 0 Ag(TCNQ) phase I .

 

 

‘5‘ .
FE, 0.003 .- A Ag(TCNQF4) _
o : 3
U r ‘
a omz .................................. .-
8 ............... ,
R eeeeeeeeeeeeeeeeeeed
0.001

 

d
0 lJnlJllllllJlllllllll llllllllL

o 50 100 150 200 250 300
T(K)

Figure 16. Temperature plots of the magnetic susceptibility for Ag(TCNQ)
phase I (at 5000 G) and H (at 1000 G) and Ag(TCNQF) (at 1000 G).

Table 2. Magnetic parameters for the Ag(TCNQ) family and Ag(TCNQF4).
Cb is the Curie constant of the impurity, and Should be 0.375 for one spin S
= V2. In parenthesis is given the molar percentage of spin 1/2 present in the

 

 

sample.
TIP (emu CGS/mol) (290 K) Cb (emu CGS.K/mol)
Ag(TCNQ) phase I 0.0013 (2) 0.0072 (1.9%)
Ag(TCNQ) Phase H 0.0018 (2) 0.0009 (0.2%)
Ag(TCNQ F4) 0.0010 (2) 0.0082 (2.2%)

 

51

5 104
4104
3 10“
2104
l 104
0100
-l 10“
-210"
-310"

X (emu CGS/mol)

 

 

0 50 100 150 200 250 300
T (K)

Figure 17. Temperature dependence of the magnetic susceptibility for [n-
BuaN][TCNQ] (at 5000 G). The bump between 40 and 50 K is due to the
presence of oxygen. The black line is the best fit obtained with a dimer
model of spin = V2.23

 

X (emu CGS/mol)

 

 

 

T(K)

Figure 18. Temperature dependence of the magnetic susceptibility for [n-
BuaN] [TCNQF4] (at 1000 G). The black line is the best fit Obtained with a
dimer model of spin = V2.23

52

range. Interestingly, the salts [n-BuaN] [T CNQ] and [n-BuaN] [T CNQF4] exhibit different
magnetic behavior. (Figures 17 and 18). The magnetic susceptibility studies reveal a
diamagnetic ground state (S=0) which, is also signature of the organic acceptor
dimerization, but in these cases the triplet state (S=l) is thermally accessible with an
energy gap of 1380 K and 152 K for [n-BuaNHTCNQ] and [n-BucNIITCNQFa]
respectively”. In figure 17, the negative values that are constant through most of the

temperature range represent the magnitude of the diamagnetic correction.

Conclusion

This study supports the existence of two distinct polymorphs of Ag(TCNQ). This
conclusion is based on synthetic, Spectroscopic, and structural data. Both phases can be
synthesized in bulk using two separate synthetic methods. This distinguishes them from
the two polymorphs of Cu(TCNQ) which are connected by a direct
kinetic/thermodynamic relationship. A bulk sample of AgTCNQ phase H prepared from
Ag powder and TCNQ is identical to the electrochemically prepared crystal by Shields as
verified by X—ray powder diffraction. The conversion of Ag(TCNQ) phase I to phase H is
possible, but it requires heat and/or light which produces silver metal. It seems unlikely
that a direct conversion of soluble Ag+ ions and TCNQ’l is occurring.

The use of the stronger electron acceptor TCNQF4 has allowed us to expand our
structural database for metal/1‘ CNQ binary phases. The structure of Ag(TCNQFa) is the
first of it’s kind with a metal bound to [TCNQF4]", and the material exhibits much more

strongly dimerized anions than TCNQ’1 salts of alkali metal salts. The stacking

interactions are exceedingly short, which is undoubtedly related to the electron
withdrawing fluorine groups on the It—delocalized ring. A slipped ring-external bond
stacking mode is apparent in the structure of [n-Bu,N][TCNQF4] which is also evident in
other organic cation TCN Q salts. The stacking distance within a TCNQF4 dimer,
however, is shorter than the corresponding distances observed for ammonium cation salts
of TCN Q. The synthesis of other metal-based T CNQF4 compounds will be continued in
the future, along with a further expansion into the use of other organic acceptor ligands

such as dibromo- and dimethoxy- TCN Q.

References

l. (a) Gardner, G. B.; Venkataraman, D.; Moore, J. S.; Lee, S. Nature 1995, 374, 792-
795. (b) Yaghi, O. M.; Li, G.; Li, H. Nature 1995, 378, 703. (c) Yaghi, O. M.; Li,
H. J. Am. Chem. Soc. 1995, 117, 10401. (d) Venkataraman, D.; Gardner, G. B.; Lee,
S.; Moore, J. S. J. Am. Chem. Soc. 1995, 117, 11601. (e) Whiteford, J. A.; Rachlin,
E. M.; Stang, P. J. Angew. Chem, Int. Ed. Engl. 1996, 35, 2524. (f) Yaghi, O. M.;
Li, H. J. Am. Chem. Soc. 1996, 118, 295. (g) Yaghi, O. M.; Li, H.; Groy, T. L. J.
Am. Chem. Soc. 1996, 118, 9096. (h) Olenyuk, B.; Whiteford, J. A.; Stang, P. J. J.
Am. Chem Soc. 1996, 118, 8221.

2. (a) Manriquez, J. M.; Yee, G. T.; McLean, S.; Epstein, A. J.; Miller, J. S. Science
1991, 252, 1415. (b) Tamaki, H.; Zhuang, Z. J .; Matsumoto, N.; Kida, S.; Koikawa,
M.; Achiwa, N.; Hashimoto, Y.; Okawa, H. J. Am. Chem. Soc. 1992, 114, 6974. (c)
Stumpf, H. 0.; Pei, Y.; Kahn, 0.; Sletten, J.; Renard, J. P. J. Am. Chem. Soc. 1993,
115, 6738. (d) Inoue, K.; Iwamura, H. J. J. Am. Chem. Soc. 1994, 116, 3173. (e)
Ohba, M.; Maruono, N.; Okawa, H.; Enoki, T.; Latour, J.-M. J. Am. Chem Soc.
1994, 116, 11566. (f) Kahn, O. In Molecular Magnetism: From Molecular
Assemblies to the Devices; NATO ASI Series E321; Coronado, E.; Delhaes, P.;
Gatteschi, D.; Miller, J.S., Eds; Kluwer: Dordrecht, 1996; pp. 243-288. (g)
Decurtins, S.; Schmalle, H. W.; Schneuwly, P.; Zheng, L.-M.; Ensling, J.; Hauser,
A. Inorg. Chem. 1995, 34, 5501. (h) Miyasaka, H.; Matsumoto, N .; Okawa, H.; Re,
N.; Gallo, E.; Floriani, C. Angew. Chem, Int. Ed. Engl. 1995, 34, 1446. (i) Ohba,
M.; Okawa, H.; Ito, T.; Ohto, A. J. J. Am. Chem. Soc., Chem. Commun. 1995, 1545.
(i) Michaut, C.; Ouahab, L.; Bergerat, P.; Kahn, 0.; Bousseksou, A. J. Am Chem.
Soc. 1996, 118, 3610. (k) de Munno, G.; Poerio, T.; Viau, G.; Julve, M.; Lloret, F.;
Journaux, Y.; Riviere, E. Chem. Commun. 1996, 2587.

3. (a) Aumiiller, A.; Erk, P.; Klebe, G.; Hiinig, 8.; von Schlitz, J.; Werner, H. Angew.
Chem, Int. Ed. Engl. 1986, 25, 740. (b) Aumiiller, A.; Erk, P.; Hiinig, 8. Mol.
Cryst. Liq. Cryst. 1988, 156, 215. (c) Erk, P.; Gross, H.-J.; Hiinig, U. L.; Meixner,
H.; Werner, H. - P.; von Schlitz, J. U.; Wolr, H. C. Angew. Chem, Int. Ed. Engl.
1989, 28, 1245. (d) Kato, R.; Kobayashi, H.; Kobayashi, A. J. Am. Chem. Soc.
1989, III, 5224. (e) Aumiiller, A.; Erk, P.; Hiinig, S.; Hadicke, E.; Peters, K.; von

55

Schnering, H.G. Chem Ber. 1991, 124, 2001. (f) Sinzger, K.; Hfinig, S.; Jopp, M.;
Bauer, D.; Beitsch, W.; von Schiitz, J. U.; Wolf, H. C.; Kremer, R. K.; Metzenthin,
T.; Bau, R.; Khan, S. I.; Lindbaum, A.; Lengauer, C. L.; Tillmanns, E. J. Am. Chem.
Soc. 1989, 115, 7696.

See for example: (a) Fagan, P J.; Ward, M. D.; Calabrese, J .C. J. Am. Chem. Soc.
1989, III, 1698. (b) Iwamaoto, T. in Inclusion Compounds: Inorganic and Physical
Aspects of Inclusion; Iwamoto, T., Atwood, J. L., Davies, J. E. D., MacNicol, D. D.,
Eds.; Oxford University Press: Oxford, 1991; Vol. 5, Chapter 6, p. 177. (c) Robson,
R.; Abrahms, B. F.; Batten, S. R.; Gable, R. W.; Hoskins, B. F.; Liu, J. in
Supramolecular Architecture, Bein, T., Ed.; American Chemical Society:
Washington, DC, 1992; p. 256. (d) Tannenbaum, R. Chem. Mater. 1994, 6, 550. (e)
Constable, E. C. Prog. Inorg. Chem. 1994, 42, 67. (1) Lu, J.’ Harrison, W. T. A.;
Jacobson, A. J. Angew. Chem, Int. Ed. Engl. 1995, 24, 2557. (g) Dunbar, K. R.;
Heintz, R. A. Prog. Inorg. Chem. 1996, 35, 4449. (i) Whiteford, J. A.; Rachlin, E.
M.; Stang, P. J. Angew. Chem, Int. Ed. Engl. 1996, 35, 2524. (j) Sharma, C. V. K.;
Zaworotko, M. J. Chem. Commun. 1996, 2655. (k) Hirsch, K. A.; Wilson, S. R.;
Moore, J. S. Inorg. Chem. 1997, 36, 2960.

(a) Lacroix, P.; Kahn, 0.; Gliezes, A.; Valade, L.; Cassoux, P.; Nouv. J. Chim.
1985, 643. (b) Gross, R.; Kaim, W.; Angew. Chem, Int. Ed. Engl. 1987, 26, 251. (c)
Humphrey, D. G.; Fallon, G. D.; Murray, K. S. J. Chem. Soc., Chem. Commun. ((1)
Bartley, S. L.; Dunbar, K. R. Angew. Chem, Int. Ed. Engl. 1991, 30, 448. (e)
Ballester, L.; Barral, M.; Gutierrez, A.; Jiménez-Apararicio, R.; Martinez-Muyo, J .;
Perpifian, M.; Monge, M.; Ruiz-Valero, C. J. Chem. Soc., Chem. Commun. 1991,
1396. (f) Comelissen, J. P.; van Diemen, J. H.; Groeneveld, L. R.; Haasnoot, J. G.;
Spek, A. L.; Reedijk, J. Inorg. Chem. 1992, 31, 198. (g) Oshio, H.; Inc, B.; Mogi, I.;
Ito, T. Inorg. Chem. 1993, 33, 5697. (h) Ballester, L.; Barral, M.; Gutierrez, A.;
Perpifian, M.; Monge, M.; Ruiz-Valero, C.; Sénchez-Pélaez, A. Inorg. Chem 1994,
33, 2142. (i) Dunbar, K. R.; Ouyang, X. Mol. Cryst. Liq. Cryst. Sci. Technol. 1995,
273, 21. (i) Oshio, H.; Inc, B.; Ito, T.; Maeda, Y. Bull. Chem. Soc. Jpn. 1995, 68,
889. (k) Dunbar, K. R.; Angew. Chem. 1996, 35, 1659. (l) Decurtins, S.; Dunbar, K.
R.; Gomez-Garcia, C. J.; Mallah, T.; Raptis, R.G.; Talham, D.; Veciana, J. in
Molecular Magnetism: From Molecular Assemblies to the Devices; NATO ASI

Series E321; Coronado, B.; Delhaes, P., Gatteschi, D., Miller J. S., Eds.; Kluwer:
Dordrecht, 1996; 571. (m) Dunbar, K. R.; Ouyang, X. Chem Commun. 1996, 2427.
(n) Zhao, H.; Heintz, R. A.; Rogers, R. D.; Dunbar, K R. J. Am Chem Soc. 1996,
118, 12844. (0) Dunbar, K. R.; Ouyang, X. Inorg. Chem 1996, 35, 7188. (p)
Azcondo, M. T.; Ballester, C. J.; Gutiérrez,, A.; Perpifian, M.; Amador, U.; Ruiz-
Valero, C.; Bellitto, C. J. Chem Soc., Dalton Trans. 1996, 3015.

For an excellent review on the subject of a coordination to TCNX molecules, see:
Kaim, W.; Moscherosch, M. Coord. Chem Rev. 1994, 129, 157.

Structure of Ag(TCNQ): Shields, L. J. Chem. Soc., Faraday Trans. 2 1985, 81 , 1.
(a) Potember, R. S.; Poehler, T. O.; Cowan, D. 0. Appl. Phys. Lett. 1979, 34, 405.
(b) Potember, R. S.; Poehler, T. O.; Rappa, A.; Cowan, D. O.; Bloch, A. N. J. Am.
Chem Soc. 1980, 102, 3659. (c) Potember, R. S.; Poehler, T. O.; Cowan, D. 0.;
Brant, P.; Carter, F. L.; Bloch, A. N. Chem. Scr. 1981, I7, 219. (d) Kamistos, E. 1.;
Risen, W. M., Jr. Solid State Commun. 1982, 42, 561. (e) Potember, R. S.; Poehler,
T. O.; Cowan, D. 0.; Carter, F. L.; Brant, P. I. In Molecular Electronic Devices;
Carter, F. L., Ed.; Marcel Dekker: New York, 1982; p. 73. (f) Potember, R. S.;
Poehler, T. O.; Benson, R. C. Appl. Phys. Lett. 1982, 41 , 548. (g) Potember, R. S.;
Poehler, T. O.; Rappa, A.; Cowen, D. 0.; Bloch, A. N. Synth. Met. 1982, 4, 371. (h)
Kamitsos, E. 1.; Risen, W. M. Solid State Commun. 1983, 45, 165. (i) Kamitsos, E.
1.; Risen, W. MJ. Chem Phys. 1983, 79, 477. (j) Kamitsos, E. I.; Risen, W. M. Jr.
J. Chem Phys. 1983, 79, 5808. (k) Benson, R. C.; Hoffman, R. C.; Potember, R. S.;
Bourkoff, E.; Poehler, T. 0. Appl. Phys. Lett. 1983, 41 , 548. (l) Poehler, T. O.;
Potember, R. S.; Hoffman, R.; Benson, R. C.; Mol. Cryst. Liq. Cryst. 1984, 107, 91.
(m) Kamistos, E. I.; Risen, W. M., Jr. Mol. Cryst. Liq. Cryst. 1986, 134, 31. (n)
Potember, R. S.; Poehler, T. O.; Hoffman, R. C.; Speck, K. R.; Benson, R. C. In
Molecular Electronic Devices II; Carter, F. L., Ed.; Marcel Dekker: New York,
1987; p. 91 (o) Wakida, S.; Ujihira, Y. J. Appl. Phys. 1988, 27, 1314. (p) Hoffman,
R. C.; Potember, R. S. Appl. Opt. 1989, 28 (7), 1417. (q) Duan, H.; Mays, M. D.;
Cowan, D. O.; Kruger, J. Synth. Met. 1989, 28, C675. (r) Sato, C.; Wakamatsu, S.;
Tadokoro, K.; Ishii, K J. Appl. Phys. 1990, 68 (12) 6535. (s)Yarnaguchi, S.;
Viands, C. A.; Potember, R. S. J. Vac. Sci. Technol. 1991, 9, 1129. (t) Hua, Z, Y.;
Chen, G. R. Vacuum 1992, 43, 1019. (u) Hoagland, J. J.; Wang, X. D.; Hipps, K.

57

10.

11.

12.

13.

W. Chem Mater. 1993, 5, 54. (v) Liu, S. G.; Liu, Y. Q.; Wu, P. J.; Zhu, D. B.
Chem. Mater. 1996, 8, 2779. (w) Liu, S. G.; Liu, Y. Q.; Zhu, D. B. Thin Solid Films
1996, 280, 271. (x) Sun, S. Q.; Wu, P. J .; Zhu, D. B. Solid State Commun. 1996, 99,
237. (y) Liu, S. 6.; Liu, Y. Q.; Wu, P. J.; Zhu, D. B.; Tian, H.; Chen, K. C. Thin
Solid Films 1996, 289, 300.

Heintz, R. A.; Zhao, H.; Ouyang, X.; Grandinetti, G.; Cowen, J.; Dunbar, K. R.
Inorg. Chem 1999, 38, 144.

(a) Emge, T. J .; Bryden, W. A.; Wiygul, F. M.; Cowan, D. O.; Kistenmacher, T. J.
J. Chem. Phys. 1982, 77, 3188. (b) Metzger, R. A.; Heirner, N. B.; Gundel, D.; Sixi,
H.; Harms, R.H.; Keller, H.J.; Nothe, D.; Wehe, D. J. Chem. Phys. 1982, 77, 6203.
(c) Wiygul, F. M.; Emge, T.J.; Kistenmacher, T.J. Mol. Cryst. Liq. Cryst. 1982, 90,
163. (d) Emge, T. J.; Cowan, D. O.; Bloch, A. N .; Kistenmacher, T. J. Mol. Cryst.
Liq. Cryst. 1983, 95, 191.

(3) Jones, M. T.; Maruo, T.; Jansen, S.; Roble, J.; Rataiczak, R. D. Mol. Cryst. Liq.
Cryst., 1986, 134, 21. (b) Maruo, T.; Rataiczak, R. D.; Jones, M. T. Mol. Phys.
1991, 73, 1365. (c) Sugano, T.; Fukasawa, T.; Kinoshita, M. Synth. Met. 1991, 41 ,
3281. (d) Agostini, G.; Corvaja, C.; Giacometti, G.; Pasimeni, L. Chem. Phys. 1993,
173, 177. (e) Otsubo, T.; Kono, Y.; Hozo, N.; Miyamoto, H.; Aso, Y.; Ogura, P.;
Tanaka, T.; Sawada, M. Bull. Chem. Soc. Jpn. 1993, 66, 2033. (t) Nishikawa, H.;
Kawakami, K.; Fujiwara, H.; Uehara, T.; Misaki, Y.; Yamabe, T. Synth. Met. 1993,
56, 1983. (g) Nakamura, Y.; Iwamura, H. Bull. Chem. Soc. Jpn. 1993, 66, 3724. (h)
Sugimoto, T.; Tsujii, M.; Murahashi, B.; Makatsuji, H.; Yamauchi, J.; Fujita, H.;
Kai, Y.; Hosoito, N. Mol. Cryst. Liq. Cryst. 1993, 232, 117.

(a) Meneghetti, M.; Girlando, A.; Pecile, C. J. Chem. Phys. 1985, 83, 3134. (b)
Meneghetti, M.; Pecile, C. J. Chem Phys. 1986, 84, 4149. (c) Meneghetti, M.;
Bozio, R.; Pecile, C. Synth. Met. 1987, I9, 451. (d) Meneghetti, M.; Bozio, R. J.
Chem Phys. 1988, 89, 2704.

(a) Fritchie, C. J., Jr.; Arthur, P. Acta Crystallogr. 1966, 21, 139. (b) Hoekstra, A.;
Spoelder, T.; Vos, A. Acta Crystallogr. 1972, 328, 14. (c) Konno, M.; Saito, Y.
Acta Crystallogr. 1974, B30, 1294. (d) Murakami, M.; Yoshimura, S. Bull. Chem
Soc. Jpn. 1975, 48, 157. (e) Konno, M.; Saito, Y. Acta Crystallogr. 1975, B31,
2007. (f) Konno, M.; Ishii, T.; Saito, Y. Acta Crystallogr. 1977, B33, 763.

14.

15.
l6.

17.

18.

19.

20.

21.
22.

Melby, L. R.; Harder, R. J .; Hertler, W. R.; Mahler, W.; Benson, R. B.; Mochel W.
E. J. Am. Chem Soc. 1962, 84, 3374.

Wheland, R. C.; Martin, E. L. J. Org. Chem 1975, 40 (21), 3101.

Boudreaux, E. A.; Mulay, J. N. Theory and Applications of Molecular
Paramagnetism, Eds., J. Wiley & Sons: New York, 1976.

(a) Lunille, B.; Pecile, C. J. Chem. Phys. 1970, 52, 2375. (b) Bozio, R.; Girlando,
A.; Pecile, C. J. Chem Soc., Faraday Trans. 2 1975, 71, 1237. (c) Van Duyne, R.
P.; Suchanski, M. R.; Lakovits, J. M.; Siedle, A. R.; Parks, K. D.; Cotton, T. M. J.
Am Chem. Soc. 1979, 101, 2832. (d) Chappell, J. S.; Bloch, A. N.; Bryden, A.;
Maxfiled, M.; Poehler, T. O.; Cowan, D. O. J. Am. Chem. Soc. 1981, 103, 2442. (e)
Farges, J. P.; Brau, A; Dupuis, P. Solid State Commun. 1985, 54 (6), 531. (t) Inoue,
M.; Inoue, M. B. J. Chem Soc., Faraday Trans. 2 1985, 81, 539. (g) Inoue, M.;
Inoue, M. B. Inorg. Chem. 1986, 25, 37. (h) Inoue, M.; Inoue, M. B.; Fernando, Q.;
Nebesny, K. W. J. Phys. Chem. 1987, 91, 527. (i) Pukaki, W.; Pawlak, M.; Graja,
A.; Lequan, M.; Lequan, R. M. Inorg. Chem. 1987, 26, 1328.

(a) Meneghetti, M.; Pecile, C. J. Chem. Phys. 1986, 84 (8), 4149. (b) Meneghetti,

M.; Bozio, R.; Pecile, C. Synth. Met. 1987, 19, 451. (c) Meneghetti, M.; Bozio, R.
J. Chem. Phys. 1988, 5, 2704. (d) Chiang, L. Y.; Goshom, D. P. Mol. Cryst. Liq.

Cryst. 1989, 26, 229.

(a) Yashihiro, Y.; Furukawa, Y.; Kobayashi, A.; Tasumi, M.; Kato, R.; Kobayashi,

H. J. Chem. Phys. 1994, 100, 2449. (b) Kobayashi, A.; Kato, R.; Kobayashi, H.;
Mori, T.; Inokuchi, H. Solid State Commun. 1987, 64, 45. (c) Willett, R. D.; Long,
G. Personal communication.

Werner, P.E.J. Appl. Cryst. 1985, I8, 367. TREOR program was adapted to PC and
PROSZKI system at Jaqielloniam University 1989.

Zhao, H.; Heintz, R. A.; Ouyang, X.; Dunbar, K. R. Chem. Mater. 1999, 11, 736.
(a) Kobayashi, H.; Danno, T.; Saito, Y. Acta Crystallogr. 1973, 329, 2693. (b)
Kobayashi, H. Acta. Crystallogr. 1978, B34, 2818. (c) Fihol, A.; Rovira, M.; Hauw,
C.; Gaultier, J.; Chasseau, D.; Dupuis, P. Acta Crystallogr. 1979, B35, 1652. (d)
Shibaeva, R. P.; Kaminskii, V. P.; Simonov, M. A. Cryst. Struct. Commun. 1980, 9,
655.

59

23.

We fit our experimental data using the Bleaney-Bowers formula and a Curie law for
S=1/2, equation 1 where N is the number of TCNQ or TCNQF4 coupled into dimer,
N’ is the number of molecules with an unpaired spin, u“ is the Bohr magneton, and

the g -factor is assumed to be equal to 2 (usual for TCNQ compounds):

I =2Ngzfli l + N’gzfli
P kT 3+exp[—2Jl(kT)] 4kT

 

 

Chapter 111

Synthesis and Characterization of
Paramagnetic Metal/TCNQ Compounds:
The M(TCNQ); Family

61

Introduction

Charge-transfer materials have been under intense investigation for the past four
decades. This research has resulted in the identification of a wide variety of conducting
salts containing organocyanide acceptors such as TCNQ.l The conducting ability of
these materials is attributed to the stacking arrangement of the planar organic radicals
within these charge-transfer materials which allows for a continuous n—pathway for
communication. The nature of the stacking has been noted to dramatically affect the
magnetic and charge-transport properties. For example, the well known ['I'I'F][TCNQ]
salt crystallizes as segregated stacks of donors and acceptors and exhibits high electrical
conductivity, whereas integrated stacks of alternating donors and acceptors exhibit

interesting optical and magnetic properties.2

[>0 thZ>-<::I::J

BEDT-TI'F
C »~ C C ’C
‘>—< ,~ ~
C " C C
Ne, “N It; \\ n”! R'
TCNQ TCNE DCNQI

Figure 19. Schematic diagrams of organic donors and acceptors used in
charge-transfer materials.

62

Recently, researchers have expanded the field of donor-acceptor chemistry by
incorporating molecules of the types depicted in Figure 19 into multi-dimensional
framework solids that contain a paramagnetic inorganic component.3 This approach
appears to be very promising, indeed materials containing BEDT-TTF ions and
paramagnetic metal anions have been found to exhibit highly conducting and even
superconducting properties!4 In terms of magnetic materials, one can envision the direct
coordination of paramagnetic metal cations to open-shell organic acceptor anions such as
TCNQ" (7,7,8,8-tetracyanoquinodirnethane)5, TCNE' (tetracyanoethylene)6, and
DCNQI' (N ,N ’-dicyanoquinodiimine)7. In this way, multi-dimensional inorganic/organic
polymers can be produced by taking advantage of the n—stacking ability of TCNQ, and, at
the same time, imparting new properties based on the presence of unpaired electrons on
the metals. The most well known examples of this strategy are: (1) metallic polymers of
Cu with DCNQI] (2) electrically bistable conducting materials of Cu and Ag with
TCNQ,8 and (3) a room-temperature bulk ferromagnet of V with TCNE.“ In all cases,
the organic molecules in the structures are in the radical anion (-1) or mixed valence (Ol-
1) states.

Although TCN Q radicals are more stable than TCNE, and the compound has been
known since the early 1960’s, little is known about the chemistry of TCNQ as a ligand
for paramagnetic transition metals. Furthermore, only several structures were known
when studies commenced in or laboratories in the early 1990’s. This chapter describes
the syntheses and characterization of M(TCNQ); compounds, prepared from reactions of
first row transition metals, Mn (1), Fe (2), Co (3), and Ni (4) with TCNQ". These

materials are binary compounds composed solely of metals and organic acceptor with no

63

co-ligands and can therefore be classified as true “inorganic/organic hybrid” materials.
Their simplicity makes them attractive candidates for study, but a major disadvantage of
working with these compounds is that they are notoriously difficult to crystallize. Not
surprisingly, structural information, which is of central importance for our complete
understanding of the magnetic properties of these materials, is sorely lacking.
Fortunately, other means of characterization such as powder X-ray diffraction and
infrared spectroscopy have allowed for educated guesses about the nature of these
materials. These details are useful when examining their magnetic properties, which has

been carried out in detail by both d.c. and ac methods.

Experimental Section

A. General Considerations

All reactions were carried out under a nitrogen atmosphere using standard
Schlenk-line techniques. Acetonitrile was dried over 3A molecular sieves and distilled
under a nitrogen atmosphere prior to use. Diethyl ether was dried and distilled over Na/K
amalgam. TCN Q was purchased from TCI Chemical Company and recrystallized from
hot acetonitrile. LiI and [Bu4N]I was purchased from Aldrich Chemical Company; [n-
Bu4N]I was recrystallized from hot water. The divalent metal [BF4]' anion salts,
[M(MeCN)4,5][BF4]2, where M = Mn, Fe, Co, and Ni, were prepared as described in the

literature.10

B. Synthesis
Although the literature outlines general synthetic methods for the syntheses of
both Li(TCNQ) and [n-BmN] [TCNQ],"’ a more detailed description of their preparation

and characterization is provided here for convenience.

(a) Preparation of Li(TCNQ)

A typical bulk preparation requires a 3:2 ratio of LiI (1.50 g, 0.0112 mol) and
TCNQ (1.53 g, 0.0075 mol) respectively. Under a nitrogen atmosphere, LiI was
dissolved in 40 mL of acetonitrile whereas TCNQ was dissolved in 400 mL of boiling
acetonitrile in a l—L beaker. The solution of LiI was transferred via cannula into the
solution of TCNQ and a dark purple precipitate formed immediately. The solution was
cooled to allow for full precipitation of the product. The product was recovered by
vacuum filtration, washed with 50 mL of acetonitrile and 100 mL of diethyl ether, and
dried in vacuo: yield 1.48 g (94%). Anal. Calcd for C12H4N4Li: C, 68.3; H, 1.9; N, 26.5;

Found: C, 68.09; H, 1.87; N, 26.25. Characteristic IR data (Nujol mull, KBr plates, cm'

1): v(CEN), 2211 s, 2196 s, 2180 sh 2154 s; v(C=C), 1507 s; 5(C-H), 826 m.

(b) Preparation of [n-BmNHTCNQ]

A quantity of Li(TCNQ) (1.48 g, 0.007 mol) was dissolved in 100 mL of hot
distilled water, and, in a separate flask, an excess of [n-Bu4N]I (5.2 g, 0.0140 mol) was
dissolved in 600 mL of boiling distilled water. Upon the addition of Li(TCNQ) to [n-
Bu4N]I, a dark purple precipitate formed immediately. The stirred solution was kept

warm for l h after which time the product was collected by vacuum filtration. The

65

resulting dark purple powder was washed with several additions of hot water, followed
by diethyl ether, and finally dried in vacuo: yield, 3.1 g (99%). The product was
recrystallized as follows. The crude product was dissolved in a minimal volume of
dichloromethane (~100 mL), and the solution was dried with anhydrous magnesium
sulfate. After filtration, the solution was reduced in volume to ~30 mL, and an equivalent
volume of hexanes was added to induce crystallization. After the excess
dichloromethane was evaporated, the dark purple crystalline product was collected by
vacuum filtration, washed with copious amounts of hexanes and dried in vacuo: yield,

2.96 g, (95%). Anal. Calcd for C23H36N5: C, 75.29; H, 9.03; N, 15.68; Found: C, 73.95;

H, 9.52; N, 15.32. Characteristic IR data (Nujol mull, KBr plates, cm”): v(CEN), 2187

sh, 2181 s, 2160 sh, 2157 s; v(C=C), 1507 s; 5(C-H), 824 m.

(c) Synthesis of M(TCNQ); compounds, where M = Mn (1), Fe (2), and Co (3).

A quantity of [n-Bu4N][TCNQ] (0.455 g, 1.02 mmol) was dissolved in 20 mL of
acetonitrile to give a green solution, which was transferred via cannula into respective 20
mL volume acetonitrile solutions of [Mn(MeCN)4][BF4]2 (0.200 g, 0.51 mmol),
[Fe(MeCN)6][BF4]2 (0.242 g, 0.51 mmol), and [Co(MeCN)6][BF4]2 (0.244g, 0.51 mmol).
A dark purple precipitate formed immediately in all three cases, with no visible changes
occurring during the twenty-minute reaction time. The precipitates were collected by
filtration under nitrogen, washed with 10 mL of acetonitrile followed by of 20 mL diethyl
ether, and dried in vacuo: yield: (1), 0.223 g, 95%; (2), 0.224 g, 95%; (3), 0.220 g, 93%.
Analytical data: Calcd for C24H3MnN3: C, 62.20; H, 1.74; N, 24.19. Found: C, 61.71; H,

1.90; N, 21.38. Calcd for C24H3FeNg: C, 62.10; H, 1.74; N, 24.14. Found: C, 60.78; H,

66

 

2.26 ; N, 22.54. Calcd for C24H8C0Ng: C, 61.68; H, 1.73; N, 23.98. Found: C, 59.87; H,

2.27; N, 21.49. Characteristic 1R data (Nujol mull, KBr plates, cm'l): Mn(TCNQ)2:

v(&N), 2205 s, 2187 s, 2137 sh; v(C=C), 1505 s; 5(C-H), 826 m. Fe(TCNQ)2: v(CEN),
2217 s, 2187 s, 2142 sh; v(C=C), 1505 s; 5(C-H), 827 m. Co(TCNQ)2: «@N), 2217 s,

2188 s 2137 sh; v(C=C), 1505 s; 5(C-H), 827 m.

(d) Synthesis of M(TCNQ); (4).

A quantity of [n-Bu4N][TCNQ] (0.455 g, 1.02 mmol) was dissolved in 30 mL of
acetonitrile to give a green solution which was slowly added to a 30 mL acetonitrile
solution of [Ni(MeCN)6][BF4]2 (0.244 g, 0.51 mmol). The resulting reaction solution
was allowed to stir for 3 hours in order to bypass a green kinetic intermediate before
going on to form the desired dark purple product. The product was collected by
filtration under nitrogen, washed with acetonitrile (10 mL) followed by diethyl ether (20
mL) and dried in vacuo. Yield: 0.223 g, 94%. Calcd for C24H3NIN32 C, 61.72; H, 1.73;

N, 23.99;. Found: C, 60.02; H, 2.21; N, 23.06. Characteristic IR data (Nujol mull, KBr

plates, cm'l): V(C"=‘N), 2224 s, 2208 s, 2192 s, 2155 sh; v(C=C), 1503 s; 5(C-H), 829 m.

C. Single Crystal Growth

Several methods of crystallization, including slow diffusion of reactants in
solventsu and gels,12 were used in attempts to isolate single crystals of the M(TCNQ);
compounds (1 - 4) (Figure 20). Each method is described in detail along with the
outcome. As a preface, it should be noted that these compounds are exceedingly

insoluble in acetonitrile and thus great care was taken to prevent rapid precipitation of the

67

 

products. Their nearly total insolubility seriously hinders crystallization, as crystal
growth requires time in solution for a crystal to heal defects and to grow to an appreciable

size.

 

_, Solvent with Reactant B

 

‘G

—> Solvent with Reactant A

 

 

 

 

—> Solvent with Reactant B

 

 

b)
I —> Gel/Solvent with Reactant A

Figure 20. Slow diffusion of (a) solvent layers and (b) solvent/gel layers.

(a) Slow Diffusion of Reactants in Solvents

This was the first method used in attempts to crystallize compounds (1 - 4). The
general procedure involves careful layering of a solution of [n-Bu4N ] [TCN Q] over a
layer of the metal containing salt with both of the reagents being dissolved in CH3CN.

This leads to the formation of the product at the interface (Figure 20a).11 Metal solvated

68

solutions were layered on the bottom of the tube because they are often more dense in
solution compared to the TCNQ'I solutions. Several factors were used to slow the
process of product formation. A lower concentration of reactants (50 mg of metal
solvated in 100 mL) obviously reduces the tendency for immediate precipitation of the
product at the interface, and a smaller diameter tube size (6-mm diameter) further slows
the diffusion of the reactants. Different solvents with higher densities than acetonitrile,
such as methylene chloride, were used in the metal component layer to further slow the
diffusion of the layers as well. Finally, once the tubes were layered they were placed in a
refrigerator to take advantage of the temperature effect on diffusion rates of solvents.
Inevitably, powders or films were obtained at the interface in all tubes, underscoring the

extreme insolubility of these compounds.

(b) Incorporation of Gels

As it became clear that slow diffusion of solvent layers was not producing the
desired result, gels were incorporated into the metal layer as a means to severely limit the
diffusion of reactants (Figure 20b).12 A solvent solubility study was performed to
determine the most versatile gel (Table 3). PMMA, (poly-methylmethacrylate), was
determined to be the better than PEO, (poly-ethyleneoxide),'2‘ or Sephadax. ’2‘ Different
gel percentages were used to vary the diffusion rates; the higher the gel concentration, the
more viscous the solution and therefore the slower the diffusion rate. Gel percentages are
based on w/w %;12" for instance, a 9% gel is composed of 9 g PMMA/ 91 g of solvent

totaling 100 g or 100%.

69

Table 3. Solubility studies of various gels where S = soluble and I =
insoluble. Heat was used with gels that did not readily dissolve.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Solvent PEO w/heat PMMA w/ heat Sephadex w/ heat
Methylene

Chloride S — S — I I
Benzene I S S — I I
Tetrahydrofuran I S S — I I
Acetone S — S — I I
Acetonitrile S — S — I I
Methanol I I I I I I
Ethanol I I I I I I
Ether I I I I I I
Hexanes I I I I I I
Toluene I S S — Partially —

(c) A Typical Setup

For a 5% gel, approximately 1 g of PMMA was added to a Schlenk flask and
placed under vacuum. Under nitrogen, an appropriate quantity of Mn[MeCN]4[BF4]2
(0.100 g) was added. A total acetonitrile volume of 20 mL was added under nitrogen,
and the solution was stirred until the gel completely dissolved. Often slight application
of heat was necessary to achieve complete dissolution of the gel. Approximately 2 mL of
the metal solution was added to the tube via a cannula, followed by slow addition of the
same volume of a solution containing 0.227 g of [n-Bu4N][TCNQ]. Gel conditions were
varied in order to optimize crystal growth. The gel percentage was increased to 15 or

20% to increase viscosity, and solvents with a higher density than acetonitrile were used

70

in the metal layer. Also, cold temperatures were used to further slow diffusion rates.
This gel method proved to be the most fruitful because microcrystals were obtained,
albeit small and irregular in shape. Unfortunately, the crystals diffracted only weakly,
thus indexing was not possible. These results, however, do suggest that gels will be
useful for growing crystals of these neutral phase materials. Additional studies in this

area will be continued by other group members.

D. Physical Measurements

Infrared spectra were recorded as Nujol mulls on IGr plates in the range 400-4800
cm'1 using a Nicolet IR/42 FT -IR spectrometer. Powder X-ray data were collected using
a Rigaku RU200B X-ray powder diffractometer with Cu Powder Kat radiation in the
range 5-80° in steps of 002° in 29. Variable temperature magnetic susceptibility data
were collected from 5-300K at fields of 500 or 1000 G using a Quantum Design, Model
MPMS SQUID magnetometer housed in the Physics & Astronomy Department at
Michigan State University. Field emission scanning electron microscopy measurements
(FESEM) were performed on a Hitachi Ltd., Model S-470011 with a Tungsten filament
housed in the W.M. Keck Microfabrication Facility located in the Department of Physics

& Astronomy at Michigan State University.

Results and Discussion

A. Synthetic Analysis

71

r «mm!

Reactions of the acetonitrile starting materials [M(MeCN)x][BF4]2 (M: Mn, x =
4; Fe, Co, Ni, x = 6) with [n-Bu4N][TCNQ] in acetonitrile led to rapid formation of dark

purple crystalline solids (Eq.l).

[M(McCN)4.6][BF4]2 + [n-BU4N][TCNQ] -> M(TCNQ)2 (S) + [n-BlelBFd (EQ- 1)

There is no visible decomposition when these materials are exposed to air, moisture and
light. IR spectroscopy, however, indicated that these materials are hygroscopic when
exposed to ambient lab atmosphere, and SQUID measurements revealed a detectable
decrease in the magnetic susceptibility that can be attributed to air exposure and/or
general decomposition over several months. When samples are handled under a nitrogen
atmosphere, analytical data for these compounds support the empirical formulae
M(TCNQ)2, which is consistent with the charge balance of the divalent metal ions.
Unlike previously reported materials of this type, e.g., M(TCNQ)2(H20)2 (M = Mn, Fe,
Co and Ni) or M(TCNQ)(MeOH)2,3,4 (x = 2, Mn, Fe; x = 3, Co; x = 4, Ni),13 the present
acetonitrile derived materials retain no coordinated solvent molecules according to
elemental analysis. Once formed, these compounds are extremely insoluble in common
solvents except acids which destroy the materials. Interestingly, and somewhat
surprisingly due to their rapid precipitation, M(TCNQ)2 bulk products are
microcrystalline and diffract well in powder X-ray experiments.

The choice to react the fully solvated acetonitrile cations that contain the
“innocent’ [BF4]' anion with salts of the coordinating TCNQ’1 anion is based on the
notion that such metathesis reactions will produce clean, isolable products. By—products

remain in solution and are simply washed away from the metal/1' CNQ solid; this results

72

in reproducibly high yields. The main benefit to preparing M(TCNQ)2 products by this
approach is that a pre-selected oxidation state of the metal ion is introduced into the
material. Other reactions such as oxidation of the metallic form of the element or
reduction of neutral TCNQ with metal iodides do not always produce the same results

(Eq. 2 and 3). Indeed, these reactions are known to lead to multiple products when more

M +NTCNQ —> M(TCNQ)“ (Eq. 2)
3MI +2nTCNQ —> 2M(TCNQ)n + M13 (Eq. 3)

than one oxidation state is electrochemically accessible for the transition metal. Thus, it
is synthetically advantageous to begin with both reactants in their desired oxidation states
to insure that the same phase is reproducibly obtained.

The primary synthetic focus of this project was to obtain M(TCNQ); products in
the highest purity and with the highest ordering temperatures, Tc. To achieve this, it was
necessary to carefully monitor and optimize reaction conditions. Several factors such as
reaction time and concentration were found to affect the overall magnetic behavior of the
resulting products. A time elapsed study, the results of which are listed in Table 4,
indicated that the M(TCNQ); products degrade from a crystalline to an amorphous phase
with increasing reaction times and prolonged contact with the solvent. Elemental
analysis results indicated that the optimal time for the reaction is approximately twenty
minutes. Increased contact with solution produced average empirical formulae of
M(TCNQ)x (1 < x < 1.5) which are further along on the reaction profile than the desired
products, and, in fact, may merely represent a mixture of decomposed phases. Analyses

of various Ni(TCNQ)2 samples indicate that Ni is unique among the transition metal

73

series that we studied, in that the desired product does not form at short reaction times.
Instead, the final product is formed after approximately three hours of stirring in
solution. The conclusion is that increased contact with solution leads to subtle

decomposition reactions at the surfaces of the particles, which are accompanied by loss of

TCNQ.

Table 4. A study of the effect of reaction time on the composition of
M(TCNQ); phases using elemental analysis as a test for purity.

 

 

 

Compound! Reaction Time C% H% N%
Sample #

Mn(TCNQ)2 Theoretical % 62.20 1.74 24.19
1 20 min. 61.84 2.25 21.71
2 30 min. 61.26 1.98 22.41
3 1 hour 20 min. 60.18 2.51 21.68
4 13 hours 59.39 2.38 21.87
5 48 hours 55.81 2.06 20.82
6 8 days 53.80 2.13 20.06
7 30 days 51.83 1.59 20.03
M(TCNQ); Theoretical % 61.72 1.73 23.99
1 5min. 51.90 2.27 19.64
2 2 hours 53.46 2.33 19.72
3 3 hours * 60.02 2.21 23.06
4 24 hours 55.21 2.17 20.51

 

 

Once the optimal reaction times were found, it was necessary to combine these
findings with the optimal concentration. Concentrations of 1.3 mmol in the metal
reactant led to products with the highest Tc value when 20 mL of acetonitrile were used

as the total reaction volume. Figure 21 summarizes the reaction times and concentrations

versus magnetization intensities for samples of four Mn(TCNQ)2. The trend that higher

74

concentrations lead to increased magnetization reflects the fact the M(TCNQ): are
apparently dissolving slightly and decomposing even while they are forming. With

conditions that favor the decomposition there will be an increase in physical defects that

 

 

 

 

2w 1 fiTT I 1 V F 1 I T T T I I ‘ T fiT l U I U I l I U
" 1
t _ """ . . 0.625 mg/mL ’
j. . .- O 2.5 mglmL
200 l" I I I 1.25m mL 2
» ' 0°00 - o 125 mSmL «
r: .'o°"-.°n - ' a
0 I80 3°33°°°o ' "
\E. 150 :_ '5 0° '20 oo - .:
U) _ I; 0° .0 o ' -
U " °° "3 o I d
o '0 o
5 r '30 O .9 o I
E 1% l— .. 0° 0 O a
o P O O O O 4
V r a o D .
c- . . O on O -
X _ ° 3 o .
so - 3w ' «
008 I
D D 1
. 0° I 4
”B
. a] J
1- 03° .1
O I 1 I 1 r l r L l r flaw
0 10 20 30 40 50
T (K)

Figure 21. Plots of magnetization versus temperature for four Mn(TCNQ)2
compounds with various reactant and solvent concentrations.

form within the material. An increase of solvent volume (lower concentrations) leads to
greater degradation of the product even when the same reaction time is constant. This

trend has a limiting concentration, however, beyond which higher reactant concentrations

(i.e. 1.25 mg/mL (u) shown in Figure 21) lead to a lower magnetization value. Finally, as

Figure 21 illustrates, magnetic data reproducibility is not possible even when the same
conditions are used. In one case, the same reaction conditions led to magnetization

values that differed by almost 50 emu CGS/mol! This inconsistency strongly hints at a

75

dependence of the magnetic properties on particle size which is a parameter that is not
easily controlled.

One final comparison (Figure 22) concerns the degradation of the samples over
time. Two Mn(TCNQ)2 samples (illustrated as circles) were analyzed at varying time
intervals after their isolation in the solid-state (illustrated as squares). In general, as time
increased, the magnetic moments decreased. Deliberate exposure to air appeared to
hasten the process of decomposition, but this occurred even in samples that were kept
under nitrogen in a dry box! The question of how and why the degradation occurs is still

not clear, but they are not stable in air or light for long periods of time.

 

 

 

 

350 p T Y I l l I I I v I v I I T rfi 1 r v T T I V I I I 1“
i . MncrcNQ),(rcw rumzswmrinzomr
300 ; o°°°°°°° o magnum) ruwnmwmr
t 0° 0 - WQ),(22daysinn:/N,anduw
A 250 t o o 25 rug/ml in 20 m1
fl 1—
a : 0° "u. o a Mn(TCNQ)2(Mdaysinthedr-ybox)
~ ' ° 12.5 rug/ml in 40 ml
\ . O .. .3
8 200 :- o0 ..0 00050200000 'j
. ‘ n .
U . 0° .° noon 0 .0. no
5 L ‘ 0 o _
E 150 t o o. ODE-IIIII.- o .o O
O , o B .I. o n
V 0 I3 I .
1' .n .
.R 100 l.’ ' 33° . . o .-
b a . . D
05 '. o
50 l— o '. '.° .3
I o 'I. on I
P 0° .-00 :4
o L A I L L J l I 1 I A L a m L 4 J LOOOAnnch‘E
O 10 20 3O 40 50
T (K)

Figure 22. Plots of magnetization versus temperature indicating the “aging
process” of the Mn(TCNQ)2 material.

76

B. Infrared Spectroscopy

This technique proved to be invaluable for the characterization of the M(TCNQ);
compounds, not only for establishing metal-TCNQ binding, but for addressing structural
changes that accompany decomposition. Unfortunately, analysis of the IR data for
TCN Q compounds is often difficult due to the extremely versatile nature of the molecule.
There are three possible oxidation states, namely TCNQO, TCNQ', TCNQZ", multiple
isomers for binding the cyano groups such as 0', bridging or dimerized, and in addition,
different types of interactions of the TCNQ units themselves (i.e. n—stacked and 0'-
dimerized).” Several research groups have reported the unique infrared bands

characteristic of their compounds with the intention of discerning between TCN Q
oxidation states and ligand coordination modes in the mid—IR v(CEN) region.5d‘f""’16

For metal-bound TCN Q, three regions provide important information for

assigning charge and discerning the coordination and stacking modes in TCNQ“—
salts.5"'f"5"16 Values in the v(CEN) in the region 2100 - 2300 cm’1 verify the presence of

TCNQ. The v(C=C) region is characteristic of 1t-delocalization within the TCNQ ring,
and often occurs around 1500 cm’1 for TCNQ". However, there are difficulties with the

interpretation of metal-bound TCNQ compounds in these regions of the spectrum,

because vibrational stretching modes shift upon coordination. A shift to higher

frequencies for v(CEN) occurs if TCN Q acts as a o-donor whereas a shift to lower
frequencies indicates a significant TCN Q to metal 1t-back-bonding.14 Furthermore,
multiple stretches in the v(CEN) region render assignments and correlations to other

compounds difficult. To alleviate this problem, the 5(C-H) bending region has also been

77

used because of its’ high sensitivity to changes in oxidation state and coordination. In
this region, oxidation states are more easily distinguished from each other, for example
the neutral TCNQ bending mode occurs at 864 cm", while TCNQ‘ is in the range 819 -
830 cm". Recently, a unique TCNQ dimerized species, referred to as the o-dimer,
[TCNQ-TCNQ]2’, has been identified by our research group to have a 5(C-H) bend occur
at 802 cm".5f’16 Since this discovery, the bending mode of TCNQ has become an
exceedingly useful tool in the study of metal/1‘ CN Q coordination compounds. For
instance, both n—stacked and o—dimerized forms of reduced TCN Q were found to be
present in [Mn(TCNQ)(TCNQ-TCNQ)05(MeOH)2].. whereas only a o-dimer is found in
the closely related polymeric material, [Mn(TCNQ-TCNQ)2(MeOH)4]...‘3 As the number
of metal TCN Q coordination compounds increases, a comparative analysis of structural
features by infrared spectroscopy becomes possible as in the case of the present set of
materials.

Analyses of the infrared spectral data for M(TCNQ); compounds (1 - 4) indicate

that they are similar in structure. A compilation of infrared data is shown in Table 5.

Table 5. Infrared spectroscopic data in the v(CsN), v(C=C), and 5(C-H)
regions for TCNQ’ compounds.

 

 

 

Compound v(CEN) (cm") v(C=C) (cm") 5(C-H) (cm")
Mn(TCNQ)2 (l) 2205 S, 2187 S, 2137 Sh 1505 S 826 m
Ft:('l"Cl‘lQ)2 (2) 2217 S, 2187 S, 2142 Sh 1505 S 827 m
COCI'CNQ)2 (3) 2217 S, 2188 S 2137 Sh 1505 S 827 m
Ni('TCNQ)2 (4) 2224 s, 2208 S, 2192 S, 2155 sh 1503 S 829 m
TCNQ 2222 s — 864 s
Li(TCNQ) 2211 s, 2196 S, 2180 Sh, 2154 S 1507 s 826 m
[n-Bu,N][TCNQ] 2188 Sh, 2180 S, 21625 , 2157 S 1507 S 824 m

 

 

78

In general, there is a notable absence of features that correspond to bound acetonitrile or

[BF4]' counterions. This is in accord with TCNQ being in its mono-reduced form and
acting as the only ligand for the divalent metals. In the v(CEN) region of the Mn, Fe and

Co compounds there are three stretching modes whereas there are four features in the
spectrum of Ni(TCNQ)2. The slight differences may be due, in part, to the presence of an
intermediate that is known to form in the Ni(TCNQ)2 reaction or to a solid-state splitting
effect not observed in the others. The stretching modes for all four metal compounds are
generally lower in energy than the corresponding mode for neutral TCN Q, viz., 2222 cm'
’, and are characteristic of TCNQ”. The v(C=C) stretching region is characteristic for the
TCNQ phenyl ring. The n-bond delocalization over the ring results in one strong (C=C)
stretching feature that is typically in the range 1500-1510 cm’l for TCNQ". Compounds
(1 - 3) exhibit bands at 1505 cm”, while the (C=C) stretch for 4 is slightly lower in
energy at 1503 cm". The 5(C-H) bending modes of the four compounds were examined
as well. In this region, modes for TCNQ", TCNQ”, and mixed-valence stacks of TCNQ‘
and TCN Q are well-documented.l7 In accord with literature reports for TCNQ’,
compounds 1 - 4 exhibit 5(C-H) frequencies between 826 and 829 cm’1 which is well
within the expected range for mono-reduced TCNQ.

An interesting phenomenon was noted after infrared spectral analysis of
M(TCNQ); products from reactions that were stirred for longer periods of time. The
presence of the unit that we have come to realize is quite prevalent in metal/TCNQ
compounds, namely the o—dimer form [TCNQ-TCNQ]2', was evident. This surprising

occurrence is noted in Figure 23. As indicated in the figure, the feature for TCNQ‘ anion

79

 

 

./ ~
.m/l

I8

13

3 E

Figure 23. IR spectrum of a bulk sample of Ni(TCNQ)2 indicating presence
of a o-dimer containing phase(s) at 803 cm".

22

 

is still present, but it diminishes as the o-dimer structural unit appears as a feature at 803
cm". It appears that M(TCNQ); is being converted over time to a new phase that
contains both reduced and dimerized forms of TCN Q. Alternatively it may be that the
new phase is composed solely of o-dimers represented by the formula, M[TCNQ-
TCNQJZ'. This phenomenon occurs in all of the M(TCNQ)2 materials, and even in
compounds isolated at shorter reaction times; in the latter cases the 802 cm" feature is
detected as only a very small shoulder.

In summary, infrared spectroscopy has proven to be quite useful for helping to

identify the structural components of these M(TCNQ)2 polymers. Although there are

80

slight differences in the IR data in the family, the striking similarities lend support to the

conclusion that they contain the same arrangement and oxidation state of TCNQ.

C. Structural Studies

(a) Field Emission Scanning Electron Microscopy, FESEM

Several important facts emerged from analysis of l - 4 by FESEM techniques.
One point is that compounds are not as crystalline as the copper and silver TCNQ
compounds. Instead of well-defined crystallites, the bulk products contain “spheres”
which are often rough and irregular in shape as shown in Figure 24.

Upon examination at higher magnifications it is possible to see that the surface of
an individual sphere is very rough and uneven and that there are smaller particles or
domains on the surface. In other words, the spherical shapes are secondary and not
primary particles. The observations above can be seen more clearly in images obtained
using transmission electron microscopy (TEM). In Figure 25, the crystallinity and
particle sizes are shown for compounds 1-4. For all compounds except Mn, the particles
tend to form homogenous “clusters” throughout the material.

The high surface areas of these solids may very well affect the magnetic behavior
due to the large number of physical defects. As mentioned earlier, the magnetic
properties are never fully reproducible even when one tries to control the synthetic

parameters; the extremely small particle size could be the reason.

81

\.

 

~16

Figure 24. Spherical particles of Ni(TCNQ)2.

82

 

M(TCNQ)z Co(TCNQ)1

Figure 25. TEM photographs of compounds 1 — 4.

83

lllllllllll
3C0um

1001.\’ 14 7mm x1001’SEiU7 9114193 20:32

lll'l'lll
t

_ 300m

 

Figure 26. Particle size trend for M(TCNQ)2 compounds at 10k
magnification.

84

It was evident in the images of the M(TCNQ); products that there is a particle size
trend. In Figure 26, the four compounds were each imaged at a 10k magnification using
a 10.0 kV accelerating voltage. The largest particles are found in Mn(TCNQ)2 followed
by Ni(TCNQ)2, and finally the Fe and Co analogs whose particles are approximately the
same size. This trend correlates with the clarity of the powder patterns for these
compounds, in that the larger the particle size, the greater the intensity in the powder X-
ray diffraction pattern. The images provide important visual evidence of the high surface
area of these compounds and also reveal a homogeneity that is remarkably good.
Normally, without any attempts to control particle size, one obtains a distribution of
particle sizes. In the M(TCNQ); samples nearly every particle is identical, and there is

no evidence for more than one phase.

(b) Powder X-ray Diffraction

Bulk microcrystalline products of 1 - 4 yield simple powder patterns as illustrated
in Figure 27. Samples were analyzed using a step program (12 5 per 0.02° at 26) on a
Rigaku RU200B X-ray powder diffractometer with Cu Ka radiation ( ha = 1.54050 A),

and the patterns were indexed using the Treor90 program.18 These efforts led to a
tetragonal cell with a = 18.8527, b = 15.4499, c = 8.2539 and V = 2404.14 for
Mn(TCNQ)2 (1). As indicated in the figure, the intensities and peak positions at 29 are
nearly identical, which provide evidence that the compounds are isostructural. Therefore,
the lattice parameters found for Mn(TCNQ)2 should also be similar for the other

members of the family.

85

 

 

 

 

 

 

 

 

 

 

 

 

1
x
h. l l
l. H r I 1
ML”; l‘JllJU’ \Wb' \M"
Fe(TCNQ)2
I Co(TCNQ)2
l
Ni(TCNQ)2
1h203040j50
29

Figure 27. X-ray powder diffraction patterns for isostructural
compounds 1 - 4.

86

D. Magnetic Properties

(a) Magnetic Data Analysis

The temperature-dependent d. c. magnetic susceptibilities (x) for compounds 1 - 4
(M = MnII S = 5/2; Fen S = 2; CoII S = 3/2; Niu S = 1) were measured from 5 to 300 K on
a SQUID magnetometer. Plots of 1mm” , um, and llxmolcorlr versus temperature are
shown in Figures 28-31. All four compounds exhibit globally similar magnetic behavior
therefore only the magnetic properties of Mn(TCNQ)2 will be discussed. In the plots of
Wm and llxmrcomr versus temperature (Figure 28 and 29), the compounds follow
Curie-Weiss behavior at high temperatures. At a certain temperature, there is an increase
in the response (the highest temperature in the series is for Mn at 44 K). Application of
the field causes domains of spins within the material to align in a parallel manner, which
is typically thought of in terms of ferromagnetic ordering. This phenomenon is recorded
in Figure 28 as a sharp increase in magnetic susceptibility at low temperatures. It is
evident in Figure 29 that a positive Weiss constant characteristic of ferromagnetic
behavior is observed. A plot of “eff vs. temperature for Mn(TCNQ)2 (Figure 30) reveals a
room temperature um value of ~ 5.9 us, which is the spin-only value for Mn(II). Analysis
of the dc. measurements point to the conclusion that these compounds are ferrromagnetic
in nature. However, their synthetic instability and non-reproducibility represented in
previous magnetic plots (Figures 21 and 22) indicate that they are not typical
ferromagnets. Physical defects, such as a variation in the percentage of the diamagnetic

o—dimer within these compounds or just the physically small size, can both be

87

 

 

 

 

 

 

 

 

30.0 ' T I Y I r r I u v r r
500 G ,
25.0 FC-ZFC .1
A ll!) , v
E 20.0: 55...; 3% 1 -j
m . E. I
. "‘ mt .
8 15.0 r g » i
a : i40.0 :
8 10.0 i 3.. - ”'"r , i
R T- : 000 jmm . . - LL 1
,. ‘ 0 i) 11!) 150 200 ”0 300 ‘
5.00 C- 3 "'0 i
" l
t S 1
0.00 ,
0 50 100 150 200 250 300
T (K)

Figure 28. Plots of xmolw“ and llxmolcorr versus temperature for Mn(TCNQ)2.

 

 

 

 

 

 

100...Tr..fi.,1.r.‘...4,.rr.l.r..
’ y = -13.586 + 0.34867! R: 0.9988
A 80.0r-
'5 .
a, .
8 60.0?
0
=8, .
§ r
H ..
20.0-
L.
l
000.11 lrrrrLl...rLL..lLanl...u
0 50 100 150 200 250 300

T (K)

Figure 29. Fitting of llxmolm" versus temperature for Mn(TCNQ)2 to
ascertain the exact critical temperature, Tc.

88

 

 

 

 

 

60.0_e...,....,....,.,, 1 Wu
CA SOOG :
50.0”. . FC‘ZFC -.
r-e q

.I ‘
i: I. _
400T: e j
’2» :o t .
1 C . j
b 30.0 s _
3: g <
u 1
:3. 8 .
20.0_ ', .-
100:— ..
....."000eeeeeeeeeeeeeee‘l
0.00.4...1114.lrrlrl....4.rrnnrrrrr‘
0 50 100 150 200 250 300

T(K)

Figure 30. Plot of peg versus temperature for Mn(TCNQ)2.

contributors to spin-glass behavior. Further magnetic studies using the a.c. method
allowed for these compounds to be classified as spin-glass materials based on observation
of a frequency dependence for these compounds. Compounds that exhibit a
ferromagnetic ordering do not show a frequency dependence, in contrast to spin glass
materials. A.C. studies are the principal methods used to distinguish closely related
magnetic behavior.

A plot of x’molw“ versus temperature is shown in Figure 31. Different frequencies
were applied over the temperature range and all data were plotted. For each frequency
study, the maximum was located at slightly different temperatures. In general, as the
frequency is decreased the maxima are located at lower temperatures. The frequency
dependence is a key characteristic for spin glass materials. As the frequency of the field

oscillates very rapidly, the spins within a ferromagnet remain correlated with the field,

89

whereas those of a spin glass become blocked. A spin glass has small domains of
ordering rather than long-range ordering, and in these domains a blocking (or freezing) of
the spins occurs that becomes more and more evident at higher frequencies. As
previously indicated, defects in the compound or small particle sizes could be the main
factor affecting the frequency measurements. Therefore, although d.c. measurements
indicate an apparent ferromagnetic ordering, it is the a.c. measurements that classify the

M(TCNQ); family of compounds as spin glass materials.

 

 

 

3m r‘rVV"v11V'Tzzr—YTTArr'rvrr'rvyr‘.f“'1
. ell ll -
' l O
250’- :15 lg .
b .
O
E 200‘- :il l .
W l 0 e
o I g. .
U 150};
a ' .
E I
3 I“)? I
.R I I
b
50.0: I
P 8
' I
0.mb J‘J‘lnmn‘nlmannlanmnlmAnal;Anal

 

 

 

LA

10 15 20 25 30 35 40
T(K)

Figure 31. Plots of x’mol‘m versus temperature illustrating the frequency
dependence of the M(TCNQ); compounds.

Whether the M(TCNQ); family of compounds are spin glass or ferromagnetic
materials, they are magnets. This conclusion is well illustrated in Figure 32. A plot of
magnetization versus field indicates that a hysteresis or “memory effect” is observed for
Mn(TCNQ)2. If a material exhibits magnetic hysteresis, then it has a memory and
therefore is, by definition, a magnet. Our interests lie primarily with the discovery of
novel magnetic materials, but, when extraordinary phenomenon such as the spin glass

nature of these compounds is observed, scrupulous analysis of the magnetic behavior is

 

 

 

 

warranted.
1000 104 j TV} I T I T T T 1 I I ' I r ffiT 1' I 5 ' I ' I V I T Y ‘I' Y T 1‘; ‘ ‘

r . e -1
t. 2 K :89... . u
. ZFC and FC (10000G) ,.;:= *
T Q. I

A 5.00 103 b e

:6. _ .1

E L -l

\ -4

U) r

O * -

O 0.00 e _.

:1 ’ u

E _

3 t

2 -5.00 103 L . —

“08"... BC 2 20 G
1» ’0'. .
4.00101 efffrmrrrlrrrml r 1; In )4 4 l

—200 -150 -100 -50 0 50 100 150 200
B (Gauss)

Figure 32. Plot of magnetization versus field illustrating the hysteresis
observed for Mn(TCNQ)2.

(b) Heat Capacity Measurements
Heat capacity measurements are useful for corroborating the existence of bulk

magnetic transitions. If the material undergoes a change in magnetic state, it necessarily

91

must undergo a concomitant change in entropy, which is manifested as a peak in the heat
capacity data at the transition temperature. Materials that undergo ferro-, ferri-, or
antiferromagnetic ordering often exhibit temperature dependent heat capacity behavior.
Spin glasses and totally amorphous materials undergo short-range magnetic interactions,
thus the thermodynamic properties of the system are not drastically altered and there is no
maximum in their heat capacity plots. In this vein, we are using heat capacity data for
discerning whether these M(TCNQ); compounds are bulk ferromagnets or actually spin
glass materials.

The heat capacity measurements of compounds 1 - 4 are shown in Figure 33. It is

obvious that there is an absence of a peak indicative of ferromagnetic ordering.

Fe(TC no) 2

100 I I I I I I

 

..
10 - ° ..

 

Heat Capacityrl' (mlcrané')

 

0.1 I l l l l I
BC!) 1 $0 2400 3200 4000 4&0

T2 (K2)

Figure 33. Plot of heat capacity measurements for Fe(TCNQ)2.

92

Conclusions

Our use of paramagnetic transition metal ions as building blocks for the synthesis
of novel TCNQ materials has resulted in the formation of three-dimensional crystalline
materials with interesting magnetic properties. In this family of compounds, the high
number of spins on the metal centers of Mn, Fe, Co and Ni along with the S = 1/2 spin on
the TCNQ anion produces a strong ferromagnetic coupling according to do
measurements. Synthetic studies and magnetic studies in the ac. mode point to the
conclusion that these compounds are not ordinary ferromagnetic materials. When
reaction conditions are varied only slightly, net magnetization and “critical temperature”
values are very different. Attempts to reproduce the magnetic data for a particular
compound proved to be futile, as xmof‘“ versus temperature plots indicated that both net
magnetization as well as critical temperature values were never exactly the same. It
appears that small quantities of impurities are being produced which are causing some of
the inconsistencies in the magnetic measurements. It has been shown by infrared
spectroscopy that the o—dimer unit, [TCNQ-TCNQ12’, begins to form in the solid with
increased reaction times. The formation of this component is occurring even at the
shortest reaction times, therefore it will always be a minor impurity of varying degree in
these samples. The presence of o—dimer [TCNQ-TCNQ?" impurities may be
contributing to the spin glass magnetic behavior; certainly the presence of another
structural unit does not help in terms of leading to a pure material with long-range
ferromagnetic ordering!

The FESEM images of these materials indicate that they consist of rough-edged

spheres with a very large surface area. The fact that such a large volume of the material

93

 

is tied up in surface defect sites is most likely the largest contributing factor to the
anomalous magnetic properties. The dangling bonds are structural defects that disrupt
the two or three—dimensional metal/TCNQ framework. This structural discontinuity
prohibits a ferromagnetic ordering and leads to local areas of competing interactions due
to the breaking of symmetry. This situation is just the sort that Mydosh describes as
leading to spin glass behavior.9a The absence of a maximum in the heat capacity
measurements of these compounds, and the ac. frequency dependence both support the
conclusion that the M(TCNQ); family of compounds are spin glasses.

Future syntheses and characterization of transition metal/TCNQ materials will
likely produce interesting magnetic results as well. The incorporation of d8, divalent Pt
or Pd compounds, which typically form square planar compounds, will alter the
geometrical requirements of the growing polymers, and may lead to a new class of
materials in which a metal ion stack and an organic stack occurs in the same material.
The diamagnetic nature of Pd2+ and Pt“ would allow for the analysis of only the
interactions of the TCN Q S=1/2 spins without complications due to metal ion spins.
Furthermore, the use of other acceptors such as TCNQF4, TCNQ(OMe)2, or TCNQBrz
would be an excellent means of probing the magnetic properties when small changes in
the electronic properties of the ligand are implemented. Most importantly, structural
analysis of new binary metal/TCNQ compounds is necessary, which requires much focus

on new crystallization techniques to inhibit rapid precipitation of these materials.

94

References

l

(a) Acker, D. S.; Harder, R. J .; Hertler, W. R.; Mahler, W.; Melby, L. R.; Bensin, R.
B.; Mochel, W. E. J. Am. Chem. Soc. 1960, 82, 6408. (b) Melby, L. R.; Harder, R. J.;
Hertler, W. R.; Mahler, W.; Benson, R. B.; Mochel W. E. J. Am. Chem. Soc. 1962,
84, 3374. (c) Torrance, J. B. Acc. Chem. Res. 1979, 12, 79. (d) Endres, H. In
Extended Linear Chain Compounds; Miller, J .S. Ed.; Plenum: New York. 1983; Vol.
3, pp. 263-312. (e) Wudl, F. Acc. Chem. Res. 1984, I7, 227. (0 Jerome, D. Science
1991, 252, 1509. (g) Bryce, M. R. Chem. Soc. Rev. 1991, 20, 355. (h) Williams, J.
M.; Schultz, A. J.; Geiser, U.; Carlson, K. D.; Kini, A. M.; Wang, H. H.; Kwok, W. —
K.; Wangbo, M. - H.; Schirber, J. E. Science, 1991, 252, 1501. (i) Ward, M. D.
Electroanal Chem. 1988, 16, 181. 0) Martin, N.; Segura, J. L.; Seoane, C. J. Mater.
Chem. 1997, 7, 1661.

(a) Miller, J. S.; Calabrese, J. C.; Epstein, A. J.; Bigelow, R. W.; Zhang, J. H.; Reiff,
W. M. J. Chem. Soc., Chem. Commun. 1986, 1026. (b) Miller, J. S.; Zhang, J. H.;
Reiff, W. M.; Dixon, D. A.; Preston, L. D.; Reis, A. H.; Gebert, B.; Extine, M.;
Troup, J.; Epstein, A. J .; Ward, M. D. J. Phys. Chem. 1987, 91 , 4344. (c) Broderick,
W. B.; Thompson, J. A.; Day, E. P.; Hoffman, B. M. Science, 1990, 249, 401. (d)
Kollmar, C. Couty, M.; Kahn, O. J. Am. Chem. Soc. 1991, 113, 7994. (e) Miller, J.
S.; Epstein, A. J. Angew. Chem, Int. Ed. Engl. 1994, 33, 385. (f) Murphy, V. J.;
O’Hare, D. Inorg. Chem. 1994, 33, 1833.

(a) Pénicaud, A.; Batail, P.; Perrin, C.; Coulon, C.; Parkin, S. S. P.; Torrence, J. B. J.
Cehm. Soc., Chem. Commun. 1987, 330. (b) Davidson, A.; Boubekeur, K.; Penicaud,
A.; Auban, P.; Lenoir, C.; Batail, P.; Hervé, G. J. Chem. Soc., Chem. Commun. 1989,
1373. (c) Pénicaud, A.; Bourbekeur, K.; Batail, P.; Canadell, B.; Auban - Senzier, P.;
Jerome, D. J. Am. Chem. Soc. 1993, 115, 4101. (d) Coulon, C.; Livage, C.;
Gonzalavez, L.; Boubekeur, K.; Batail, P. J. Phys. I Fr. 1993, 3, l. (e) Coronado, B.;
Gomez - Garcia, C. J. Comments Inorg. Chem. 1995, I 7, 255. (t) G6mez - Garcia, C.
J; Girnénez - Saiz, C. J.; Triki, S.; Coronado, B.; Manqueres, P. L.; Ouahab, L.;
Ducasse, L.; Sourisseau, C.; Delhaes, P. Inorg. Chem. 1995, 34, 4139. (g) Coronado,

95

B.; Delhaus, P.; Galaan - Mascaros, J. R.; Giménez - Saiz, C. J. Synth. Met. 1997, 85,
1647.

(a) Bousseau, M.; Valade, L.; Legros, J. - P.; Cassoux, P.; Garbauskas, M.;
Interrante, L. V. J. Am. Chem. Soc. 1986, 108, 1908. (b) Day, P.; Kurmoo, M.;
Mallah, T.; Marsden, I. R.; Allan, M. R.; Friend, R. H.; Pratt, F. L.; Hayes, W.;
Chasseau, D.; Bravic, G.; Ducasse, L. J. Am. Chem. Soc. 1992, 114, 10722. (c)
Kurmoo, M.; Graham, A. W.; Day, P.; Coles, S. J.; Hursthouse, M. B.; Caulfield, J.
L.; Singleton, J.; Pratt, F. L.; Hayes, W.; Ducasse, L.; Guionneau, P. J. Am. Chem.
Soc. 1995, 117, 12209.; see also references therein.

(a) Lacroix, P.; Kahn, 0.; Gliezes, A.; Valade, L.; Cassoux, P. Nouv. J. Chim. 1985,
643. (b) Gross, R.; Kaim, W. Angew. Chem, Int. Ed. Engl. 1987, 26, 251. (c)
Bartley, S. L.; Dunbar, K. R. Angew. Chem, Int. Ed. Engl. 1991, 30, 448. (d)
Ballester, L.; Barral, M.; Gutierrez, A.; Jiménez-Apararicio, R.; Martinez - Muyo, J.;
Perpifian, M.; Monge, M.; Ruiz - Valero, C. J. Chem. Soc., Chem. Commun. 1991,
1396. (e) Humphrey, D. G.; Fallon, G. D.; Murray, K. S. J. Chem. Soc., Chem.
Commun. 1988, 1356 (f) Comelissen, J. P.; van Diemen, J. H.; Groeneveld, L. R.;
Haasnoot, J. G.; Spek, A. L.; Reedijk, J. Inorg. Chem. 1992, 31 , 198. (g) Oshio, H.;
Ino, E.; Mogi, I.; Ito, T. Inorg. Chem. 1993, 33, 5697. (h) Ballester, L.; Barral, M.;
Gutierrez, A.; Perpifian, M.; Monge, M.; Ruiz-Valero, C.; Sénchez-Pélaez, A. Inorg.
Chem. 1994, 33, 2142. (i) Dunbar, K. R.; Ouyang, X. Mol. Cryst. Liq. Cryst. Sci.
Technol. 1995, 273, 21. (j) Oshio, H.; Ino, B.; Ito, T.; Maeda, Y. Bull. Chem. Soc.
Jpn. 1995, 68, 889. (k) Dunbar, K. R. Angew. Chem 1996, 35, 1659. (l) Decurtins,
S.; Dunbar, K R.; Gomez - Garcia, C. J.; Mallah, T.; Raptis, R. G.; Talham, D.;
Veciana, J. In Molecular Magnetism: From Molecular Assemblies to the Devices;
NATO ASI Series E321; Coronado, B.; Delhaés, P., Gatteschi, D., Miller J. S., Eds.;
Kluwer: Dordrecht, 1996; 571. (m) Dunbar, K. R.; Ouyang, X. Chem. Commun.
1996, 2427. (n) Zhao, H.; Heintz, R. A.; Rogers, R. D.; Dunbar, K. R. J. Am. Chem
Soc. 1996, 118, 12844. (0) Dunbar, K. R.; Ouyang, X. Inorg. Chem. 1996, 35, 7188.
(p) Azcondo, M. T.; Ballester, C. J.; Gutiérrez,, A.; Perpifian, M.; Amador, U.; Ruiz -
Valero, C.; Bellitto, C. J. Chem. Soc., Dalton Trans. 1996, 3015.

96

6

10

(a) Manriquez, J. M.; Yee, G. T.; McLean, S.; Epstein, A. J.; Miller, J. S. Science
1991, 252, 1415. (b) Miller, J. S.; Calabrese, J. C.; McLean, R. S.; Epstein, A. J. Adv.
Mater. 1992, 4, 498. (c) Miller, J. S.; Vazquez, C.; Jones, N. L.; McLean, R. S.;
Epstein, A. J. J. Mater. Chem. 1995, 5, 707. ((1) Miller, J. S.; Calabrese, J. C.;
Vazquez, C.; McLean, R. S.; Epstein, A. J. Adv. Mater. 1994, 6, 217. (e) Bohm, A.;
Vazquez, C.; McLean, R. S.; Calabrese, J. C.; Kalm, S. B.; Manson, J. L.; Epstein,
A.; Miller, J. S. Inorg. Chem. 1996, 35, 3083. (f) Brinckerhoff, W.B.; Morin, B.G.;
Brandon, E. J .; Miller, J. S.; Epstein, A. Appl. Phys. 1996, 79, 6147.

(a) Aumtiller, A.; Erk, P.; Klebe, G.; Hiinig, S.; von Schiitz, J.; Werner, H. Angew.
Chem, Int. Ed. Engl. 1986, 25, 740. (b) Aumfiller, A.; Erk, P.; Hfinig, S. Mol. Cryst.
Liq. Cryst. 1988, 156, 215. (c) Erk, P.; Gross, H. - J.; Htinig, U. L.; Meixner, H.;
Werner, H. - P.; von Schtitz, J. U.; Wolr, H. C. Angew. Chem, Int. Ed. Engl. 1989,
28, 1245. (d) Kato, R.; Kobayashi, H.; Kobayashi, A. J. Am Chem. Soc. 1989, 111,
5224. (e) Aumfiller, A.; Erk, P.; Hiinig, S.; Hadicke, B.; Peters, K; von Schnering,
H.G. Chem. Ber. 1991, 124, 2001. (f) Sinzger, K.; Hfinig, S.; Jopp, M.; Bauer, D.;
Beitsch, W.; von Schtitz, J. U.; Wolf, H. C.; Kremer, R. K; Metzenthin, T.; Bau, R.;
Khan, S. I.; Lindbaum, A.; Lengauer, C. L.; Tillmanns, E. J. Am. Chem. Soc. 1993,
1 I 5, 7696.
(a) Potember, R. S.; Poehler, T. 0.; Cowan, D. 0. Appl. Phys. Lett. 1979, 34, 405. (b)
Potember, R. S.; Poehler, T. 0.; Cowan, D. 0.; Carter, F. L.; Brant, P. In Molecular
Electronic Devices II; Carter, F. L., Ed.; Marcel Dekker: New York, 1982; p. 91 (c)
Hoagland, J. J .; Wang, X.D.; Hipps, K. W. Chem. Mater. 1993, 5, 54. (d) Dunbar, K.
R.; Cowen, J.; Heintz, R. A.; Grandinetti, G.; Ouyang, X.; Zhao, H. Inorg. Chem

1999, 38, 144.

(a) Mydosh, J. A. Spin Glasses: An Experimental Introduction, Taylor &Francis,
London, 1993, pp. 64-73. (h) Chowdhury, D. Spin Glasses and Other Frustrated
Systems, Princeton University Press, New Jersey, 1986. (c) Moorjani, K and Coey, J.
M. D. Magnetic Glasses, Elsevier Ed., New York, 1984.

Heintz, R.A.; Smith, J. A.; Szalay, P. S.; Weisgerber, A.; Dunbar, KR . Inorg. Synth.

in press.

97

11

12

13
14
15

16

17

18

(a) Holden, A. and Singer P. Crystals and Crystal Growing, Anchor Books-
Doubleday, New York, 1960. (b) Laudise, R. A. The Growth of Single Crystals,
Solid State Physixl Electronics Series, N. Holonyak Jr. Ed., Prentice-Hall, Inc., 1970.
(c) Henisch, H. K. Crystal Growth in Gels, Pennsylvania State University Press,
1970. (d) Sulb, S. L. J. Chem. Ed. 1985, 62, 81. Hulliger, J. Angew. Chem Int. Ed.
Engl. 1994, 33, 143.

(a) Desiraju, G. R.; Curtin, D. Y.; Paul, I. C. J. Am. Chem. Soc. 1977, 99, 6148. (b)
Arend, H.; Connelly, J. J. J. Cry. Growth 1982, 56, 642. (c) Robert, M. C.;
Lefaucheux, F. J. Cry. Growth 1988, 90, 358. (d) Doxsee, K. M.; Chang, R. C.;
Chen, E. J. Am Chem. Soc. 1988, 120, 585. (e) Yaghi, 0.; Li, G.; Li, H. Chem.
Mater. 1997, 9, 1074.

Zhao, H.; Heintz, R. A.; Ouyang, X.; Dunbar, K. R. Chem Mater. 1999, 11, 736.
Kaim, W.; Moscherosch, M. Coord. Chem. Rev. 1994, 129, 157.

(a) Lunille, B.; Pecile, C. J. Chem. Phys. 1970, 52, 2375. (b) Bozio, R.; Girlando, A.;
Pecile, C. J. Chem. Soc., Faraday Trans. 2 1975, 71 , 1237. (c) Van Duyne, R. P.;
Suchanski, M. R.; Lakovits, J. M.; Siedle, A. R.; Parks, K. D.; Cotton, T. M. J. Am.
Chem. Soc. 1979, 101, 2832. (d) Chappell, J. S.; Bloch, A. N.; Bryden, A.; Maxfield,
M.; Poehler, T. 0.; Cowan, D. O. J. Am Chem. Soc. 1981, 103, 2442. (e) Farges, J.
P.; Brau, A; Dupuis, P. Solid State Commun. 1985, 54 (6), 531. (f) Inoue, M.; Inoue,
M. B. J. Chem. Soc., Faraday Trans. 2 1985, 81, 539. (g) Inoue, M.; Inoue, M. B.
Inorg. Chem. 1986, 25, 37. (h) Inoue, M.; Inoue, M. B.; Fernando, Q.; Nebesny, K.
W. J. Phys. Chem. 1987, 91, 527.

Pukaki, W.; Pawlak, M.; Graja, A.; Lequan, M.; Lequan, R. M. Inorg. Chem. 1987,
26, 1328.

(a) Yashihiro, Y.; Furukawa, Y.; Kobayashi, A.; Tasurni, M.; Kato, R.; Kobayashi,
H. J. Chem. Phys. 1994, 100, 2449. (b) Kobayashi, A.; Kato, R.; Kobayashi, H.;
Mori, T.; Inokuchi, H. Solid State Commun. 1987, 64, 45. (c) Willett, R. D.; Long,
G. Personal communication.

Werner, P. E. J. Appl. Cryst. 1985, 18, 367. TREOR program was adapted to PC and
PROSZKI system at Jaqielloniam University 1989.

98

 

Chapter IV:
Identification of Two Paramagnetic

Phases of [Mn(TCNQ)2(L)2] (L = H20,
MeOH) and their Conversion to the
Mn(TCNQ)2 Magnet

99

Introduction

The design of polymeric materials that contain a paramagnetic metal ion linked by
an electron acceptor such as TCNQl has resulted in a new area of hybrid
inorganic/organic compounds2 with interesting magnetic and conducting properties. The
most well studied example is Cu with DCNQI acceptors,3 but other recently reported
examples of transition metals combined with TCNQ4 and TCNES exhibit interesting
properties as well. For example, Cu(TCN Q) has been shown to undergo an electric-field
induced bistable switching from a high to a low resistance state at a certain threshold
potential.“ Reactions TCNE with the first row transition elements have produced

ferromagnets with higher critical temperatures than any known synthetic magnets. In
particular, the TCNE compound with the proposed formula, V('I‘CNE)x ' y(CH2C12),5"

exhibits room temperature ferromagnetism, a goal many researchers have been trying to
attain for the past forty years.

The synthesis of the aforementioned materials is not without it’s difficulties and
frustrations. Their high degree of insolubility in common solvents leads to extreme
difficulty in crystallizing these compounds. This is unfortunate because knowledge of
structure is crucial for interpreting the physical properties. Furthermore, as has been
demonstrated for Cu(TCNQ) and Ag(TCNQ), polymorphism is an issue that further
hinders understanding of these materials. The recognition of two phases of CuTCNQ and
subsequent structural analysis helped to unravel the reproducibility problems with
encountered in the switching experiments with this material. The confirmation of the

existence of multiple polymorphs obviously requires careful analysis, but most

100

 

importantly, the study and possible application of these compounds requires separation
of the phases into pure components.

The subject of this chapter is the isolation of the separate crystalline phases that
have been identified in bulk samples of Mn(TCNQ)2(MeOH)2 (l) and
Mn(TCNQ)2(H20)2 (5). The use slow diffusion techniques used to crystallize products of
the methanol reaction led to the realization that two distinct structures were possible;
these are designated (1a) and (lb) because they are both components of the bulk material
(1). Crystallization of Mn(TCNQ)2(H20)2 (5) has resulted in only one phase to date,
which is hereafter designated as (5a). According to X-ray structural analyses, the three
types of TCNQ coordinated ligands found in these compounds are depicted schematically

in Figure 34.6

cis-[u-TCNQT \M

N. N
c>_ : _ <0
N’c C?"

syn-[u-TCNQI‘

Figure 34. Different binding modes of TCNQ.

101

One of the two polymorphs of bulk Mn(TCNQ)2(MeOH)2 (1), namely,
[Mn('I‘CNQ)(TCNQ-TCNQ)0.5(MeOH)2]a, (1a), exhibits both the bidentate cis-[tt-
TCNQ]‘ binding mode and the tu-[TCNQ-TCNQf' unit, which is the o-dimer form of
('I‘CNQ')2. The other component, [Mn(TCNQ-TCNQ)2(MeOH)4]w, (1b), contains only

the tu-[TCNQ-TCNQF' binding mode. Diagrams in Figure 35 and 36 produced from the

X-ray coordinates show these different binding modes within the structures). The other

 

(b)

Figure 35. The (a) single unit and (b) packing diagram of
[Mn(TCNQ)(TCNQ-TCNQ)0_5(MeOH)2].,, (1a), illustrating the 114-[TCNQ-
TCNQ]2' and bidentate cis-[tt— TCNQ]‘ binding modes.

102

 

Figure 36. Two diagrams of the [Mn(TCNQ-TCNQ)(MeOH)4],o (lb)
structure, which contain only the pa-[TCNQ-TCNQF' binding mode.

 

 

 

 

 

Figure 37. The structure of [Mn(TCNQ)2(H20)2],o (5a) which contains the
syn-[u-TCNQT binding mode.

103

structurally characterized polymeric material, [Mn(TCNQ)2(H20)2],o, (5a), contains only

the syn-[u-TCNQT ligand as shown in the diagram in Figure 37 which is also based on
X-ray coordinates.

Unlike the bulk methanol product, only one polymorph (5a) has been structurally
characterized for Mn(TCNQ)2(H20)2 (5), but the similarity of these compounds indicate
that there may be another phase present. Also, the other metal/TCNQ materials with M =
Fe, Co, and Ni are suspected of being able to form different solid-state structures. The
topic of this chapter is the development of a general synthetic method for the bulk
isolation and purification of pure materials of Mn2+ with TCNQ'l ligands. In particular,
we are interested in phases that contain the o-dimer m-[TCNQ—TCNQF’ binding mode,
as there is evidence that these materials are on the structural evolution pathway to the

desolvated magnets M(TCNQ)2 .

Experimental

A. General Considerations

All reactions were carried out under a nitrogen atmosphere using standard
Schlenk-line techniques. Methanol was dried over magnesium methoxide whereas
diethyl ether and THF were dried using Na/K amalgam; all three were freshly distilled
under a nitrogen atmosphere prior to use. The reagent TCNQ was purchased from TCI
Chemical Company and recrystallized from hot acetonitrile before use. Lil was
purchased from Aldrich Chemical Company and used to prepare LiTCNQ and [n-

BuaN][TCNQ] according to the synthesis outlined in chapter 111. The divalent metal

104

salts, [M(MeCN)4,6][BF4]2, where M = Mn, Fe, Co, and [Ni(MeCN)6][PF6]2, were
prepared as described in the literature.8 The bulk polymeric products,
Mn(TCNQ)2(MeOH)2 (l) and Mn(TCNQ)2(H20)2 (5), were prepared according to
methods established in our laboratories.6 Infrared spectra were recorded as Nujol mulls
in the range 400-4000 cm'1 on a Nicolet IR/42 spectrometer. Powder X-ray diffraction

data were collected on a Rigaku RU2OOB powder diffractometer with CuKa radiation in

the range 5 - 60°.

B. Synthesis of Bulk Methanol Phases.

(a) Preparation of Mn(TCNQ)2(MeOH)2, (1)

Synthesis #1: A MeOH solution (20 mL) of [n-BuaN][TCNQ] (1.19 g, 2.55 mmol) was
slowly added to a MeOH solution (10 mL) of [Mn(MeCN)4][BF4]2 (0.500 g, 1.3 mmol).
The resulting reaction solution was stirred for 20 min to give a dark purple precipitate,
which was collected by filtration, washed with 20 mL of methanol followed by 20 mL of

diethyl ether, and dried in vacuo. Yield: 0.320 g, 61%. Characteristic IR data (Nujol mull,
KBr plates, cm"): v(CEN), 2205 s, 2180 s, 2141 s; v(C=C), 1520 s; 8(C-H), 826 m, 802

sh.

Synthesis #2: A MeOH solution (20 mL) of [n-BuaN][TCNQ] (0.246 g, 0.550 mmol)
was slowly added to a MeOH solution (10 mL) of [Mn(MeCN)4][BF4]2 (0.108 g, 0.275
mmol). The resulting reaction solution was stirred for 20 min to give a dark purple
precipitate, which was collected by filtration, washed with 20 mL of methanol followed

by 20 mL of diethyl ether, and dried in vacuo. Yield: 0.060 g, 41%. Characteristic IR

105

data (Nujol mull, KBr plates, cm'l): v(CEN), 2205 s, 2194 s, 2180 s, 2150 s; v(C=C),

1509 s; 5(C-H), 826 m, 803 w.

Synthesis #3: A MeOH solution (25 mL) of [n-BuaN] [T CNQ] (0.910 g, 2.04 mmol) was
slowly added to a MeOH solution (5 mL) of [Mn(MeCN)4][BF4]2 (0.400 g, 1.02 mmol).
The resulting reaction solution was stirred for 25 min to give a dark purple precipitate,
which was collected by filtration, washed with 10 mL of methanol followed by 20 mL of

diethyl ether, and dried in vacuo. Yield: 0.134 g, 25%. Characteristic IR data (Nujol mull,
KBr plates, cm’l): v(SN), 2206 s, 2188 s, 2150 s, 2133 s; v(C=C), 1507 s; 5(C-H), 826

m, 802 s.

(b) Preparation of Fe(TCNQ)2(MeOI-I)2, (2)

Synthesis #1: A MeOH solution (60 mL) of [n-BuaNHTCNQ] (0.760 g, 1.70 mmol) was
slowly added to a MeOH solution (20 mL) of [Fe(MeCN)6][BF4]2 (0.405 g, 0.851 mmol).
The resulting reaction solution was stirred for 20 min to give a dark purple precipitate
that was collected by filtration, washed with 20 mL of methanol followed by 20 mL of.

diethyl ether, and dried in vacuo. Yield: 0.326 g, 62%. Characteristic IR data (Nujol mull,
KBr plates, cm’l): v(&N), 2203 s, 2186 s, 2155 s, 2124 s; v(C=C), 1505 s; 5(C-H), 826

s, 802 3.

Synthesis #2: A MeOH solution (40 mL) of [n-Bu4N][TCNQ] (1.43 g, 3.19 mmol) was
slowly added to a MeOH solution (10 mL) of [Fe(MeCN)6][BF4]2 (0.760 g, 1.59 mmol).
The resulting reaction solution was stirred for 20 min to give a dark purple precipitate
that was collected by filtration over a period of 45 minutes, washed with 20 mL of

methanol followed by 20 mL of diethyl ether, and dried in vacuo. Yield: 0.806 g, 96%.

106

Characteristic IR data (Nujol mull, KBr plates, cm"): v(CEN), 2203 s, 2187 s, 2155 s,

2130 s; v(C=C), 1505 s; 5(C-H), 826 s, 802 s.

(c) Preparation of Co(TCNQ)2(MeOH)z, (3)

A MeOH solution (20 mL) of [n-Bu4N][TCNQ] (0.187 g, 0.418 mmol) was
slowly added to a MeOH solution (10 mL) of [Co(MeCN)6][BF4]2 (0.100 g, 0.209
mmol). The resulting reaction solution was stirred for 20 min to give a dark purple
precipitate that was collected by filtration, washed with 20 mL of methanol followed by

20 mL of diethyl ether, and dried in vacuo. Yield: 0.091 g, 73%. Characteristic IR data
(Nujol mull, KBr plates, cm'l): v(&N), 2215 s, 2193 s, 2170 s, 2130 sh; v(C=C), 1503 s;

5(C-H), 824 s, 802 sh.

(d) Preparation of NiCI‘CNQ);(MeOH)2, (4)

Synthesis #1: A MeOH solution (20 mL) of [n-Bu4N][TCNQ] (0.754 g, 1.69 mmol) was
slowly added to a MeOH solution (10 mL) of [Ni(MeCN)6][PF6]2 ( 0.404 g, 0.844 mmol).
The resulting reaction solution was stirred for 20 min to give a dark purple precipitate
that was collected immediately by filtration, washed with 20 mL of methanol followed by

20 mL of diethyl ether, and dried in vacuo. Yield: 0.250 g, 47%. Characteristic IR data
(Nujol mull, KBr plates, cm"): v(CEN), 2201 s, 2178 s, 2138 sh; v(C=C), 1505 s; 5(C-

H), 826 s, 802 sh.
Synthesis #2: A MeOH solution (30 mL) of [n-BuaNHTCNQ] (1.14 g, 2.55 mmol) was
slowly added to a MeOH solution (10 mL) of [Ni(MeCN)6][PF6]2 (0.76 g, 1.27 mmol).

The resulting reaction solution was stirred for 20 min to give a dark purple precipitate

107

that was collected by filtration over a period of two hours, washed with 20 mL of

methanol followed by 20 mL of diethyl ether, and dried in vacuo. Yield: 0.616 g, 91%.
Characteristic IR data (Nujol mull, KBr plates, cm'l): v(CEN), 2201 s, 2178 s, 2138 sh;

v(C=C), 1505 s; 5(C-H), 826 s, 802 s.

B. Separation of the Two Components in Bulk Mn(TCNQ)2(MeOH)2, (l).
(a) Low Temperature Reaction to yield [Mn(TCNQXTCNQ-TCNthMeOth,
(la)

Separate MeOH solutions of [n-Bu4N][TCNQ] (0.455 g, 1.018 mmol) (40 mL)
and of [Mn(MeCN)4][BF4]2 (0.200 g, 0.509 mmol) (10 mL) were chilled in an ice bath
for 1 hour. The [n-BuaNHTCNQ] solution was added to the solution of
[Mn(MeCN)4][BF4]2 and allowed to stand at low temperature for 30 minutes. The dark
purple precipitate was collected by filtration, washed with 20 mL of cold methanol
followed by 20 mL of diethyl ether, and dried in vacuo. Yield: 0.117 g, 44%.

Characteristic IR data (Nujol mull, KBr plates, cm'l): v(CEN), 2214 s, 2193 s, 2167 s,

2152 s; v(C=C), 1505 s; 5(C-H), 826 s, 803 sh.

(b) Higher Temperature Reactions to yield [Mn(TCNQ-TCNQXMeOHMOO, (lb)

Synthesis #1: A quantity of THF (20 mL) was added to a Schlenk flask which contained
Mn(TCNQ)2(MeOH)2 (0.150 g, 0.272 mmol). The reaction was stirred for 1.5 h at 60°C,
then removed from heat and allowed to stand for 30 minutes. The resulting pale blue-

green product was collected by filtration under nitrogen, washed with 20 mL of THF

followed by 20 mL of diethyl ether and dried in vacuo. Yield: 0.120 g, 80%.

108

 

Characteristic IR data (Nujol mull, KBr plates, cm"): v(CEN), 2197 s, 2124 s; v(C=C),

1507 s; 5(C-H), 802 5.

Synthesis #2: A quantity of MeOH (20 mL) was added under nitrogen to a Schlenk flask
which contained Mn(TCNQ)2(MeOH)2 (0.150 g, 0.272 mmol). The reaction was stirred
for 4 hours at 70°C then removed from the heat and allowed to stand for 30 minutes. The
resulting pale blue product was collected by filtration under nitrogen, washed with 20 mL

of MeOH followed by 20 mL of diethyl ether, and dried in vacuo. Yield: 0.123 g, 82%.
Characteristic IR data (Nujol mull, KBr plates, cm'l): v(CEN), 2197 5, 2124s; v(C=C),

1507 s; 8(C-H), 802 s.

C. Isolation of a Second Phase of Mn(TCNQ)2(H20)2 (5) That Contains the o—dimer
Structural Unit (5b).

A quantity of THF (20 mL) was added to a bulk sample of Mn(TCNQ)2(H20)2 (0.150 g,
0.300 mmol) under nitrogen. The reaction was stirred for 2 days at 70°C then removed
from the heat and allowed to stand for 30 minutes. The resulting pale blue-green product
was collected by filtration under nitrogen, washed with 20 mL of THF followed by 20
mL of diethyl ether and dried in vacuo. Yield: 0.139 g, 93%. Characteristic IR data

(Nujol mull, KBr plates, cm'l): v(CEN), 2195 s, 2124 s; v(C=C), 1509 s; 5(C-H), 828 w,

802 m.

D. Interconversion of Mn(TCNQ)2 (6) and Mn(TCNQ)2“); polymers where X =

MeOH (1) or H20 (5).

109

General procedure: For each reaction, a 0.150 g sample of the starting material was
added to a 100 mL Schlenk flask after which time 20 mL of a particular solvent was
added under nitrogen. The flask was heated to approximately 60 °C, and stirred for the
designated time. The solvent was removed, the product was washed with 20 mL of

diethyl ether and dried in vacuo.

E. Attempted Conversion of FeCl‘CNQ)2(MeOH)¢ (2) to F eCl‘CNQ); (7).
A quantity of FeCl‘CNQ)2(MeOH)2 (0.150 g, 0.271 mmol) was added to a 100 mL

Schlenk flask, and 20 mL of acetonitrile were added under nitrogen. The flask was
heated to 52°C in an oil bath and stirred for 1.5 h, after which time the solution was
allowed to cool for 30 minutes. After the solvent was removed by cannula, the product

was washed with 20 mL of diethyl ether and dried in vacuo. Yield: 0.117 g, 78%.
Characteristic IR data (Nujol mull, KBr plates, cm"): v(CEN), 2193 sh, 2120 s; v(C=C),

1505 s; 5(C-H), 825 sh, 803 s.

F. Attempted Conversion of NiCTCNQhMeOH); (4) to M(TCNQ); (8)

In this reaction, Ni(TCNQ)2(MeOH)2 (0.120 g, 0.216 mmol) was added to a 100
mL Schlenk flask and 20 mL of acetonitrile were added under nitrogen. The flask was
heated to approximately 65°C and stirred for 1.5 h after which time the solution was

allowed to cool for 30 minutes. The solvent was evaporated and the product was washed

with 20 mL of diethyl ether and finally dried in vacuo. Yield: 0.120 g, 96%.
Characteristic IR data (Nujol mull, KBr plates, cm'l): v(CEN), 2208 s, 2198 sh, 2174 s,

2155 8, 2063s; v(C=C), 1507 s; 5(C-H), 825 s, 800 rn.

llO

Results and Discussion

A. Synthesis
(a) Preparation of the Bulk Methanol Compounds, (1-4).

The bulk compounds M(TCNQ)2(MeOH)2, where M = Mn (1), Fe (2), Co (3), and
Ni (4), were used as starting materials for various transformations in the present studies.
The reported syntheses worked according to the earlier reports,6 but it was noted that
these reactions produce mixtures of products. We realized that we were obtaining
different ratios of two crystal types that contain different forms of coordinated TCNQ",
namely the bidentate cis-[tt-TCNQr and pr-rICNQ-TCNQf‘ modes as outlined in the

experimental section.

(b) Isolation of the two components (1a and 1b) of Mn(TCNQ)2(MeOH)2 (1).

Armed with the knowledge that there are two components in the bulk methanol
phases, we tackled the next step, namely to determine synthetic methods for isolating
them as pure solids. One obvious method, if possible, is to capitalize on any differences
in thermal stability of the phases. The material Mn(TCNQ)2(MeOH)2 (l) was the
prototype for these studies since two crystalline phases (la and lb) had already been
structurally characterized. During the course of investigating reactions that lead to

Mn(TCNQ)2(MeOH)2, it was found that dilute conditions and short reaction times favor

the kinetic phase, [Mn(TCNQ)(TCNQ-TCNQ)0,5(MeOH)2]¢, (la) whereas higher
concentrations and longer reaction times favor the thermodynamic phase, [Mn(TCNQ-

TCNQ)(MeOH)4]w (lb). The isolation of (la) [Mn(TCNQ)(TCNQ-TCNQ)0,5(MeOH)2]w

111

(la) was performed at low temperatures to prevent rapid conversion to the second phase
(lb) in situ. A reaction time of twenty minutes is considered optimal, as it produces the
most crystalline product. After the designated reaction time, the filtration was performed
quickly to prevent the solution from warming and thus possibly forming the

thermodynamic phase. Isolation of the thermodynamic phase, [Mn(TCNQ-

TCNQ)(MeOH)4]w (lb) was achieved by heating a stirring suspension of the bulk

product (1) for a long period of time. This reaction is best performed in MeOH to insure
retention of axial MeOH ligands and to provide the additional methanol that is needed to
fill the interstices in (lb). As determined by TGA analysis, the bulk product (1) only
contains an average of two methanol ligands per metal, i.e. the formula is
Mn(TCNQ)2(MeOH)2,6 therefore an excess of solvent is necessary to provide the
remaining methanol groups in (lb).

One final point that is worth noting when attempting to distinguish between the two
phases is product color. Crystals of [Mn(TCNQ)(TCNQ-TCNQ)o,5(MeOH)2]m (la) are
dark purple which is similar to the color of the bulk methanol product (1), but crystals of

[Mn(TCNQ-TCNQXMeOHMm (1b) are a pale blue color. The same pale blue color

(which is unusual for materials that contain reduced TCN Q) was observed for the bulk
products obtained from the higher temperature reactions in THF and MeOH. In this way,
color can be used to help determine whether the separation of the two phases has been
achieved before one works up the reaction and analyzes the product spectroscopically.
The explanation for the pale blue color of (lb) is that the TCNQ radicals are completely

dimerized to form a new C-C bond, thus the intense 1t- 1!?“ transition of the radical

TCNQ' is no longer observed.

112

(c) Investigation of bulk Mn(TCNQ)2(H20)2 (5) which contains o-dimers (5b).

The crystallization and characterization of one phase formed in the aqueous
chemistry of an+ with TCNQ" has been determined.6 Crystals of [Mn(TCNQ)2(H20)2],0
(5), are different from either of the two methanol products (la and lb). The structure of

[Mn(TCNQ)2(H20)2]co (5a) is composed of TCNQ anions acting as ligands but with no

o—dimer being present. This however, does not mean that alternative phases with o-
dimers cannot exist, and, in fact, we noted with consternation in our earlier work that the
simulated powder pattern of [Mn(TCNQ)2(H20)2]w (5a) crystals do not perfectly match
the experimental powder data on the bulk sample (5).6 We therefore sought to isolate a
new phase or phases of Mn(TCNQ)2(H20)2 (5). In a similar manner to the methanol
chemistry, a suspension of the bulk product was heated in a solution of THF to quickly
ascertain if a o—dimer phase would form. As indicated by IR, a new phase containing 0-

dimers formed upon heating.

((1) Conversions of bulk Mn(TCNQ)2(X)2 compounds, (X = MeOH or H2O) to
Mn(TCNQ)2 (6)-

The similarity of Mn(TCNQ)2 and Mn(TCNQ)2(X)2 compounds, where X =
MeOH or H2O, prompted us to ask whether one can inter-convert these compounds by
adding or removing solvent molecules. The results of these reactions are summarized in
Table 6. The IR spectra of the materials were compared to the spectral features for the
starting materials as well as to the pure crystalline phases established by single crystal X-

ray methods. In the water-based material, removal of axial bound water was

113

Table 6. Conversion reactions of Mn(TCNQ)2 and Mn(TCNQ)2(X)2

compounds, X = MeOH and H20.

 

 

 

 

 

Reaction/Expected product Reaction Result! IR Data
Conditions
1)Mn(TCNQ)2(MeOH)2 + MeCN 9 Mn(TCNQ)2 Heat at 57 °C Mn(TCNQ)2
150 mg 20mL 7/16/98 and stir 1.5 h, 1) IR- 2203 s, 2180 5,
Yield: 78% purple product under N2 2137 m, 1503 s, 826
2) Product after exposure to X-rays w
2) IR- 2207 s, 2187 s,
2139 s, 1503 s, 826 m
Mn(TCNQ)2(MeOH)2 + MeCN -> Mn(TCNQ)2 No heat and Mn(TCNQ)2
150 mg 20 mL 10/14/98 stir, 5 days IR- 2204 s, 2186 8,
Yield: 80% purple product under N2 2163 s, 2140 m, 1502
s, 824 m, 804 m
Mn(TCNQ)2(H20)2 + MeCN -) Mn(TCNQ)2 Heat at 57 °C Mn(TCNQ)2(H20)2
150mg 20mL 7/17/98 Room temp IR- 2195 m, 2180 m,
Yield: 99% purple product and stir for 5 2157 m, 2054 w,
h under N2 1507 m 828 m
Mn(TCNQ)2(H20)2 + MCCN ‘9 Mn(TCNQ)2 Heat at 57 °C Mn(TCNQ)2(H20)2
150 mg 20 mL 7/21/98 and stir for 7 h IR- 2195 m, 2154 m,
Yield: 91% purple product under N2 2053 w, 1505 m, 826

m, 816 sh

 

Mn(TCNQ) 2 + H20 '9 Mn(TCNQ)2(H20)2

Heat at 57 °C

Mn(TCNQ)2

 

 

 

 

 

 

 

 

 

 

 

100 mg 20mL 7/31/98 and stir for IR- 2205 s, 2187 5,
Yield: 83% purple product 30h, under N2 2131 sh, 826 w
Mn(TCNQ)2(H20)2 (150 mg) + MeOH (20mL) Heat at 57 °c Mn(TCNQ)2(H20)2
-> Mn(TCNQ)2(MeOH)2 and stir for IR- 2212 m, 2187 m,
Yield: 99% purple product 7/21/98 15 h under N2 2167 m, 2155 sh, 827

m, 802 w
Mn(TCNQ)2(H20)2 (100 mg) + MeOH (20mL) Reflux and Mn(TCNQ)2(H20)2
-) Mn(TCNQ)2(MeOH)2 stir for 24h, IR- 22111 m, 2187
Yield: 80% purple product 7/31/98 under N2 m, 2166 m, 2152 sh,
826 m, 802 w
Compound IR data
Mn(TCNQ)2(MeOH)2 (1) v(C=—=N), 2214 s, 2191 s, 2176 s, 2159 s, 2139
sh; v(C=C), 1520 s; 5(C-H), 827 m, 804 w
[Mn(TCNQ)(TCNQTCNQ)os(MeOH)2]s v(aN), 2216 s, 2187 5,2168 m, 2156 sh;
(13) 5(C-H), 825 m
[Mn(TCNQ-TCNQXMCOHMm (1b) v(CEN), 2202 s, 2139 s; 6(C-H), 806 m
Mn(TCNQ)2(HzO)2 (5) v(CEN), 2211 s, 2194 s, 2168 m;
v(C=C),1508 s; 5(C-H), 825 m
Mn(TCNQ)2 (6) v(C'='N), 2205 s, 2187 s, 2137 sh;
v(C=C), 1505 s; 8(C-H), 826 m

 

 

114

 

unsuccessful, despite the use of elevated temperatures and longer reflux times. This is
disconcerting because the powder X-ray pattern of [Mn(TCNQ)2(H20)2]co (5a) is more
similar to that of the magnetic phase, Mn(TCNQ)2 (6), than any of the others, so we
naturally thought the interconversion of these two be the easiest to perform. Even
attempts to remove the water by heating in solid state were to no avail, as they occurred
only with decomposition of the compound. On the other hand, the conversion of
[Mn(TCNQ)2(MeOH)2]00 to the solvent-free form, Mn(TCNQ)2, was successful after
being heated in acetonitrile for 1.5 hours. This is the first time that a solvent was used to
facilitate this transformation. Earlier it had been observed in our laboratories that
[Mn(TCNQ-TCNQXMeOHMm crystals converted to Mn(TCNQ)2 upon exposure to X-
ray radiation or with heating in the solid-state.6

The successful conversion of bulk Mn(TCNQ)2(MeOH)2 (l) to Mn(TCNQ)2 (6)
was attempted with the isostructural materials, FeCI‘CNQ)2(MeOH)2 (2) and
Ni('I‘CNQ)2(MeOH)2 (4). In these cases, however, similar reaction times and
temperatures produced only a mixture of the original phase and a o-dimer phase instead
of the magnetic phase. In case of Fe, the o-dimer (2b) was the major product, while in
the Ni case, both the original and o—dimer phases (4b) are present in roughly equal
concentrations as indicated by IR spectral features. Although they were not successful at
producing the desolvated magnetic phases, these reactions provide further evidence that

the o-dimer form is important in the bulk methanol products of the other metals as well.

115

B. Infrared Spectroscopy

Infrared spectroscopy is the tool that was used to discern the formation, identity
and conversion of the metal/TCNQ materials. Both the v(C‘='N) and v(C=C) regions are

useful in confirming the presence of the reduced form of TCNQ and its coordination to a
metal, but the 5(C-H) region is the primary region used to distinguish TCNQ’ vs.
['I‘CNQ-TCNQ]2'.9 In the v(CEN) and v(C=C) regions, compounds containing all types
of TCNQ moieties exhibit similar energies which render specific assignments difficult.
In the 5(C-H) region, however, reduced TCNQ exhibits a bend that appears at ~824 cm’l
while the TCNQ o-dimer, [TCNQ-TCNQ]Z", exhibits a characteristic bend at ~802 cm".

Therefore, compounds that contain either type or mixtures of both types of TCNQ units

can be readily identified.

(a) Analyses of the bulk methanol compounds, (1 - 4).

The bulk methanol TCN Q starting materials, M(TCNQ)2(MeOH)2 where M = Mn
(1), Fe (2), Co (3), and Ni (4), prepared for the present studies exhibit some striking
differences in the IR spectra from the previously published results. It was, therefore a
natural question to ask whether the bulk products of Fe, Co and Ni also contain two
components of the types [M(TCNQ)(TCNQ-TCNQ)o5(MeOH)2]m (2 - 4a) and
[M(TCNQ-TCNQXMeOHhLO (2 - 4b) encountered in the Mn chemistry. If so, we
reasoned that changes in the reaction conditions should also affect the relative
distributions of these phases in the corresponding bulk products (2 - 4). Several

reactions were attempted for Fe and Ni in which the concentrations and reaction times

were increased in a manner similar to what was found to be successful with Mn. These

116

efforts resulted in equal mixtures of the two components or a slight favoring of the “all 0'-
dimer” phase (2b and 4b) as indicated by IR spectroscopy. Clearly, these reactions lead
to the same two products in the bulk materials as have been firmly established by single
crystal X-ray for the Mn derivatives.

A comparison of the newly synthesized materials with the published data is
summarized in Table 7. In the v(CEN) region, there are more signals in the spectra of the

compounds prepared in the earlier studies, which is most likely due to the use of mild
reaction conditions that favored two products. Most important in these comparisons is
the overall trend evident from the recorded frequencies in the 5(C-H) region. When one
increases the concentrations, there is an increase in percentage of o—dimer present in the
mixture. This is indicated by an intense (medium or strong) peak that appears at 802 cm"
that is absent in the published data for samples prepared under conditions that are more
dilute. The increased ratio of the o—dimer feature with longer times, higher
concentrations, and elevated temperatures evidence that it is a stable form of TCNQ'l

under certain conditions.

117

Table 7. A comparison of M(TCNQ)2(MeOH)2 bulk products with varying
reaction conditions.

 

 

 

Reaction
Conditions: IR
region
Time and
Compound concentration v(C'='N) v(C=C) 5(C-H) Ref.
(mmol/mL)
Mn(TCNQ)2 20 min. 2214 s, 2191 s, — 827 m, 804 w 6
-(MeOH)2 0.0125 2176 s, 2159 s, 2139 sh
Synthesis #1 20 min. 2205 s, 2180 s, 2141 s 1520 s 826 m, 802 sh this
0.0433 work
Synthesis #2 20 min. 2205 s, 2194 s, 2180 s, 1509 s 826 m, 803 m this
0.0092 2150 5 work
Synthesis #3 25 min 2206 s, 2188 s, 2150 s, 1507 s 826 m, 802 3 this
0.0680 2133 s work
Fe(TCNQ)2 20 min. 2216 s, 2187 s, 2167 m, -- 827 m, 802 w 6
-(MeOH)2 0.0125 2154 sh
Synthesis #1 20 min. 2203 s, 2186 s, 2155 s, 1505 s 826 m, 802 In this
0.0106 2124 5 work
Synthesis #2 20 min. + 2203 s, 2187 s, 2155 s, 1505 s 826 m, 802 m this
45 min. filter 2130 3 work
0.0318
Co(TCNQ)2 20 min. 2213 s, 2189 s, 2169 m, — 825 m, 804 w 6
-(MeOH)2 0.0125 2156 sh
Synthesis #1 20 min. 2215 s, 2193 s, 2170 s, 1503 s 824 m, 802 sh this
0.0070 2130 sh work
Ni(TCNQ)2 20 min. 2224 s, 2193 s, 2170 m, — 825 m, 802 w 6
-(MeOH)2 0.0125 2158 sh
Synthesis #1 20 min. 2201 s, 2178 s, 2138 sh 1505 s 826 m, 802 sh this
0.0281 work
Synthesis #2 20 min. 2201 s, 2178 s, 2138 sh 1505 s 826 m, 802 In this
0.0318 work

 

118

 

(b) Confirmation of the isolation of phases (1a) and (lb) from bulk
Mn(TCNQ)2(MeOH)2. (1).

The bulk separation of the two components of the Mn(TCNQ)2(MeOH)2 mixtures
was performed by designing separation methods that focus on their relationship as kinetic
and thermodynamic products. As Figure 38 indicates, the two phases of

Mn(TCNQ)2(MeOH)2 (1) are a physical mixture of the bulk product. Therefore, infrared

/1

 
  
 

[Mn(TCNQ-TCNQ)(MeOH)4]..
2202s, 2139s

 

[Mn(TCNQ)(TCNQ-TCNQ)os(MeOH)2]..
2216s, 2187s, 2168m,
2156sh

 
   

Mn(TCNQ)2(MeOH)2
bulk product

22145, 21918, 2176s,
2159s, 2139sh

 

 

 

Figure 38. Infrared spectra in the v(CEN) region showing that both
crystallographic forms are present in bulk Mn(TCNQ)2(MeOH)2 product.

119

spectral data of these three samples must be examined in order to check whether a
particular sample contains one or both of these compounds. Spectra of the products from
different experiments were carefully compared to original published data for the bulk

product and two crystalline phases. Both isolated products exhibit frequencies in the

v(CEN) region that can be related directly to those found in bona fide crystals of the

respective components (Figures 39 and 40). Furthermore, in the 6(C-H) region, the
isolated product that corresponds to [Mn(TCNQ)(TCNQ-TCNQ)0_5(MeOH)2]w, (la),
exhibits frequencies at 826 cm'1 (s), and 803 cm'1 (w) while the product that corresponds
to [Mn(TCNQ-TCNQXMeOHmw, (1b), exhibits a single intense feature at 802 cm".

These data firmly support the conclusion that isolation of the two components of

Mn(TCNQ)2(MeOH)2 (l) was successful.

 

 

 

35.38%
sass ~~~~
N—t—r—r NNNN
NNNN

2220 2120 2220 2120

Figure 39. Comparison of the v(CEN) regions for an experimental sample
obtained in the present studies (left) and for X-ray crystals (right) of
[Mn(TCNQ)(TCNQ-TCNQ)0,5(MeOH)2],,.

120

 

 

 

 

 

2124 a-=====:;_
2202
2139

l—'-_'l 1"—"'|

2220 2000 2220 2000

Figure 40. Comparison of the v(CE-N) regions for powder (left) and crystals
(right) obtained in the chemistry of [Mn(TCNQ-TCNQ)(MeOH)4],,.

(c) Identification of a second phase of Mn(TCNQ)2(H20)2, (5).

A second, previously unrecognized, phase of Mn(TCNQ)2(H20)2 (5) was
identified using infi'ared spectroscopy as well. Conversion of the bulk product, which
shows a mixture of the two forms of reduced TCNQ (5a), (5(C-H) : 828 cm“(s), 803 cm‘1
(w)), to one that only contains the o—dimer form (5b), formulated as [Mn(TCNQ-
TCNQ)(I-I2O)x]w, was attempted. We were encouraged to undertake this experiment by
the presence of a low-energy feature in the 8 (C-H) region where the o-dimer occurs, viz.,

near 800 cm". Spectra in the v(CEN) and 5(C-H) regions for the starting material and

the isolated product indicate that distinct differences exist (Figure 41). The o—dimer

121

product contains only two broad features in the v(CEN) region, which is characteristic of

the all a-dimer methanol phase. Furthermore, the intense 5(C-H) bend at 802 cm"

 

 

 

 

 

 

 

 

 

 

 

 

confirms its presence.
8
ea
3&3
Nv-t—t
NNN
338
coco
323
flfi
NN
I I l I
2410 am “III «I:

Wuhan-hen. cm"

Figure 41. A comparison of the infrared spectra of the starting material
(top), Mn(TCNQ)2(H20)2, and the converted phase (bottom). Both v(CEN)
and 8(C-H) modes clearly indicate the differences in structure.

122

(c) Spectral evidence for the conversion of Mn(TCNQ)2(MeOH)2, (1) to
Mn(TCNQ)2, (6).

The infrared data for the interconversions of Mn(TCNQ)2. Mn(TCNQ)2(MeOH)2,
and Mn(TCNQ)2(H20)2 are listed in Table 6. As the data clearly show, the only
successful conversion is that of Mn(TCNQ)2(MeOH)2 to Mn(TCNQ)2 in acetonitrile. A
spectral comparison is shown in Figure 42 for both the starting material,
Mn(TCNQ)2(MeOH)2 (1), the heated material, and an authentic sample of Mn(TCNQ)2
(6). Attempted conversions of the isostructural materials, Fe(TCNQ)2(MeOH)2 (2) and
Ni(TCNQ)2(MeOH)2 (4), to Mn(TCNQ)2 in acetonitrile resulted in the formation of

mixtures of the original phase and o-dimer phase or formation of only the o—dimer phase
as evidenced by the data in Figure 43, in which the v(GN) and 5(C-H) modes are

displayed for both compounds.

Finally, the instability of the C-C bond that connects the two TCNQ’s in the o-
dimer ligand was confirmed by IR spectroscopy performed on samples after various
treatments. Samples containing this form of coordinated TCNQ'l were exposed to an
intense X-ray beam and then reanalyzed by infrared spectroscopy. Results consistently
indicated that the o-dimer phase was irreversibly destroyed into separate TCNQ‘
moieties in the solid state. As indicated in Figure 44, the spectra recorded after X-ray
exposure indicate the presence of only reduced TCNQ, which displays a 8(C-H) mode at
825 cm]. This finding substantiates our earlier discovery of the thermal conversion of
Mn(TCNQ—TCNQ)(MeOH)4, (l), to the magnetic phase, Mn(TCNQ)2, (6), in the solid-

state.

123

.1

 

(a) starting with
bulk mixture

"‘3
826

 

 

 

1

all
8
a (b) after heating
‘ in MeCN

2205
2120
2143

 

 

 

 

 

 

g
N
(c) Authentic
phase
5 Si
N N
24in 20:0 to}... in

Figure 42. A comparison of the spectra of the starting material and final
product in the conversion of (a) Mn(TCNQ)2(MeOH)2 (1) to (b)
Mn(TCNQ)2 and a comparison to a bona fide sample of Mn(TCNQ)2.

124

 

 

 

 

 

 

 

 

Wanna-Lon, cm"

Figure 43. Illustration of the resulting o-dimer products (2b, 4b) obtained
after heating suspensions of M(TCNQ)2(MeOH)2, M = Fe, (2), (top) and Ni,

(4), (bottom).

125

 

 

 

 

 

 

 

 

Figure 44. IR spectrum for the heated Ni product (4b), (Figure 42), recorded
after exposure to X-rays. The 5(C-H) frequency at 824 cm"1 is present for
TCNQL

C. X-ray Powder Diffraction

(a) Confirmation of the successful isolation of (1a) and (1b) from
Mn(TCNQ)2(MeOH)2 (1)-

Compounds that contain the o—dimer component are difficult to analyze using X-
ray powder diffraction because the material changes during the course of the reaction;
consequently, the resulting powder pattern is not representative of the compound being
analyzed. At low temperatures, however, sample decomposition can be avoided during

the powder diffraction experiment, and an accurate powder diffraction pattern can be

obtained. Since decomposition of the o—dimer compounds occurs upon exposure to the

126

X-ray beam, this low temperature technique was used to confirm the isolation of the
heated sample corresponding to the formula, [Mn(TCNQeTCNQ)2(MeOH)4]oo (1b). The
resulting powder pattern was compared to the one obtained for crystals of [Mn(TCNQ-
TCNQ)(MeOH)4],o in Figure 45. From this comparison, one can clearly see that heating

a bulk methanol sample causes a conversion that is similar to the crystal phase but

obviously mixed with something else.

 

Heated product

' f“ 9‘ .. Cu

 

 

 

 

1.1- 1.31 .7 rs- 1.1 .1 30 l» .1 Tr. J~ 1.“ —:I 1 1.1a a ~Ii- I, 11 1.61:1

[Mn(TCNQ-TCNQ)(MeOH)4]ee
from Single-Crystal

 

 

 

 

 

 

 

 

Ag ' 1.11 1.11.11. “1.2.2.21... in...
10 20 30 40 50
20

Figure 44. X-ray powder diffraction patterns for the heated material (1b)
(top) and crystals of [Mn(TCNQ-TCNQ)(MeOH).;].,o (bottom).

127

A comparison of powder diffraction patterns for the heated Mn(TCNQ)2(H20)2
(5) phase and the original bulk sample is illustrated in Figure 46. The relative intensities
and positions of the peaks indicate that a new phase has formed, (5b). Therefore,

identification of a second phase using infrared and X-ray powder diffraction techniques

 

was possible.
' 1
' l
1 Cu
1 1 1 1 l
‘ I ‘ .1 i
. .‘ 11 f
: ’ ,“1 ; ’1 I " = :1 1
1 ' 1 i i 1 f1 1X E H 1
1 3 ‘1‘: 1‘ 1 1‘1: ?0 , l 13 fl
1 . ‘ '1, ‘ 1 . ‘ 1",, '13-], 31,1‘ 11 A _'.‘ f
1 ,2 1...... 11111111 01.11.12 WW1
1‘ W. ,. ,, ~1;1 . ., , ‘V- T.,! ,_ ‘

1 1
, . 2.1

 

 

 

 

 

 

II.IIIEIIIxlxl'l'llllvllvll.II'.|II|IIIIIII'.I'|'I|'II

5. 10. 15. 20. 25. 30. 35. 40. 45. 50. 55.

20

Figure 46. A comparison of the X—ray powder diffraction patterns for the
heated phase (5b) (top) and the original bulk phase, Mn(TCNQ)2(H20)2 (5)
(bottom).

128

Finally, powder diffraction X-ray data further confirmed the conversion of (1) to
the magnetic phase, (6). At the solid state/solution interface, the samples were analyzed
before and after conversion and clearly show that a conversion to Mn(TCNQ)2 is taking
place. Figure 47 illustrates the three powder patterns: bulk Mn(TCNQ)20VIeOH)2 (l), the
heated material, and an authentic powder pattern of Mn(TCNQ)2 (6). From these data, it
is clear that the heated sample has converted to a phase that is similar to Mn(TCNQ)2.
These powder patterns as well as infrared data confirm that the conversion of (l) to (6) is
possible, and also that the conversion of the o—dimer phase to the magnetic phase by

exposure to an X-ray beam has occurred.

129

 

Mn(TCNQ)2(MeOH)2
Before Heating

 

 

 

 

V T T j V ' I I 1 V V T 1' V f r V I V V

 

Heated Product

   
 

 

 

 

 

 

Mn(TCNQ)2

 

 

 

 

 

.VTT-rv'wv I 1Yrv—v-v'v'v'V—v—v 'yyv 'e‘yvvviv,r_'-w'v‘vyv‘v-.vv-‘v1.1rvrj’vrv '17rVYT‘TYr'rVV1'VVV'Y

5 . '10 . :5. 129'. 25. 30 . 3.5- 40 .

20

Figure 47. X-ray powder patterns measured before (top) and after (middle)
the conversion reaction of Mn(TCNQ)2(MeOH)2 to Mn(TCNQ)2. As a
reference, a powder pattern of bona fide Mn(TCNQ)2 is displayed at the
bottom.

130

Conclusions

A number of the known metal-TCN Q polymer complexes have been recognized
to exist as multiple phases, but it is likely to be a problem for most of these materials.
Unfortunately, many of these will remain undiscovered as it is not a simple task to
identify and characterize multiple components of insoluble materials. We were fortunate
that two phases (1a and 1b) could be crystallized from slow diffusion of the reactants of
the bulk Mn(TCNQ)2(MeOH)2 compound (1). Characterization by infrared spectroscopy
and X-ray powder diffraction of these crystals provided a basis for the analysis of related
compounds. After detailed studies of the formation of the two components of bulk
samples of Mn(TCNQ)2(MeOH)2 (1), it was obvious to us that [Mn(TCNQ)(TCNQ-

TCNQ)0.5(MeOH)2],,o (1a) is the kinetic product and [Mn(TCNQ—TCNQ)2(MeOH)4]co

(lb) is the thermodynamic product.7 The former is produced at short reaction times, and,
according to infrared spectroscopy, it is the major component in the bulk methanol
product (1).6 The latter, however, has been shown experimentally to form over time,
often with increased application of heat.

An interesting phenomenon was observed during analysis of [Mn(TCNQ-
TCNQ)2(MeOH)4]w (lb) crystals by X-ray powder diffraction. Exposure to the X-ray

source caused decomposition through the severing of the weakly bonded C-C 0' bond in
the tu-[TCNQ-TCNQF' ligand. Furthermore, there is a concomitant removal of axially

coordinated methanol groups in these crystals. Even more surprising is the fact that the
two TCNQ. radicals do not recombine in the solid state, but instead form a phase similar

to the magnetic material prepared directly from [Mn(MeCN)4(BF4)2] and [n-

Bu4N][TCNQ]. We have now verified by X—ray powder diffraction and infrared

131

spectroscopy that crystals of [Mn(TCNQ-TCNQ)2(MeOH)4]w (lb) convert to the

magnetic phase, Mn(TCNQ)2 with exposure to heat or to an intense X-ray source.6

Since it is the thermodynamic phase, [Mn(TCNQ-TCNQ)2(MeOH)4]w (lb), that is

readily converted to the magnetic phase, obtaining pure samples of it from the bulk
material Mn(TCNQ)2(MeOH)2 (I) became an extremely important goal. The use of slow
diffusion techniques to grow single crystals of (lb) takes several months, therefore
isolation via bulk methods was targeted instead. The successful separation of the two
crystalline phases (la and lb) was performed by taking advantage of the differences in
their thermal stabilities and/or solubilities. The kinetic product (la) was isolated at low
temperatures and short reaction times, whereas the thermodynamic phase, (lb), appears
to be favored with heating and longer reaction times.

A detailed examination of the synthetic conditions leading to Mn(TCNQ)2(HzO)2
(5), a second phase, led to the discovery of composed of o—dimers (5b) which was
verified by infrared data and X-ray powder diffraction. The first phase, (53), crystallized
several years ago in our laboratories contains the syn-[u-TCNQT coordination mode.
Confirmation of the presence of two phases in the isostructural methanol products, (2-4),
was also successful in that the presence of O-dimer was detected, but conversion to the
respective unsolvated, magnetic phases did not occur. It may be that the conditions were
not forceful enough to effect the changes and loss of solvent, but this remains to be
verified by further studies.

As the rapid expansion of binary metal/organocyanide acceptor compounds
continues with the use of TCNQ derivatives and other more exotic acceptors, the

identification of multiple polymorphs in compounds will doubtless be an important

132

priority. It is crucial that researchers obtain pure samples of each phase, or at the very
least, ascertain how each component contributes to the general properties of the bulk
product. If this is not done, then much wasted effort in trying to understand magnetic and

other physical data will ensue.

133

l!

References

l

(a) Acker, D. S.; Harder, R. J .; Hertler, W. R.; Mahler, W.; Melby, L. R.; Bensin, R.
B.; Mochel, W. E. J. Am. Chem. Soc. 1960, 82, 6408. (b) Melby, L. R.; Harder, R. J .;
Hertler, W. R.; Mahler, W.; Benson, R. B.; Mochel, W. E. J. Am. Chem. Soc. 1962,
84, 3374. (c) Torrance, J. B. Acc. Chem. Res. 1979, 12, 79. (d) Endres, H. In
Extended Linear Chain Compounds; Miller, J.S. Ed.; Plenum: New York. 1983; Vol.
3, pp. 263-312. (e) Wudl, F. Acc. Chem. Res. 1984, 17, 227. (f) Jerome, D. Science
1991, 252, 1509. (g) Bryce, M. R. Chem. Soc. Rev. 1991, 20, 355. (h) Williams, J.
M.; Schultz, A. J .; Geiser, U.; Carlson, K. D.; Kini, A. M.; Wang, H. H.; Kwok, W.-
K.; Wangbo, M. - H.; Schirber, J. E. Science, 1991, 252, 1501. (i) Ward, M. D.
Electroanal. Chem. 1988, I6, 181. (j) Martin, N.; Segura, J. L.; Seoane, C. J. Mater.
Chem. 1997, 7, 1661.

(a) Pénicaud, A.; Batail, P.; Perrin, C.; Coulon, C.; Parkin, S. S. P.; Torrence, J. B. J.
Cehm. Soc., Chem. Commun. 1987, 330. (b) Davidson, A.; Boubekeur, K.; Pénicaud,
A.; Auban, P.; Lenoir, C.; Batail, P.; Hervé, G. J. Chem. Soc., Chem. Commun. 1989,
1373. (c) Pénicaud, A.; Bourbekeur, K.; Batail, P.; Canadell, B.; Auban-Senzier, P.;
Jerome, D. J. Am. Chem. Soc. 1993, 115, 4101. (d) Coulon, C.; Livage, C.;
Gonzalavez, L.; Boubekeur—K.; Batail, P. J. Phys. I Fr. 1993, 3, 1. (e) Coronado, B.;
Gomez-Garcia, C. J. Comments Inorg. Chem. 1995, 17, 255. (f) Gomez-Garcia, C. J;
Giménez-Saiz, C. J .; Triki, S.; Coronado, B.; Manqueres, P. L.; Ouahab, L.; Ducasse,
L.; Sourisseau, C.; Delhaes, P. Inorg. Chem. 1995, 34, 4139. (g) Coronado, B.;
Delhaus, P.; Galaan-Mascaros, J. R.; Giménez - Saiz, C. J. Synth. Met. 1997, 85,
1647.

(a) Aumiiller, A.; Erk, P.; Klebe, G.; Hiinig, S.; von Schiitz, J.; Werner, H. Angew.
Chem, Int. Ed. Engl. 1986, 25, 740. (b) Aumiiller, A.; Erk, P.; Hiinig, S. Mol. Cryst.
Liq. Cryst. 1988, 156, 215. (c) Erk, P.; Gross, H. - J.; Hiinig, U. L.; Meixner, H.;
Werner, H. - P.; von Schlitz, J. U.; Wolr, H. C. Angew. Chem, Int. Ed. Engl. 1989,
28, 1245. (d) Kato, R.; Kobayashi, H.; Kobayashi, A. J. Am. Chem. Soc. 1989, 111,
5224. (e) Aumiiller, A.; Erk, P.; Hiinig, S.; Hadicke, E.; Peters, K.; von Schnering,
H.G. Chem. Ber. 1991, 124, 2001. (f) Sinzger, K.; Hiinig, S.; Jopp, M.; Bauer, D.;

134

Beitsch, W.; von Schiitz, J. U.; Wolf, H. C.; Kremer, R. K; Metzenthin, T.; Bau, R.;
Khan, S. I.; Lindbaum, A.; Lengauer, C. L.; Tillmanns, E. J. Am. Chem. Soc. 1993,
115, 7696.
(a) Potember, R. S.; Poehler, T. 0.; Cowan, D. 0.; Appl. Phys. Lett. 1979, 34, 405.
(b) Potember, R. S.; Poehler, T. 0.; Cowan, D. 0.; Carter, F. L.; Brant, P. In
Molecular Electronic Devices II; Carter, F.L., Ed.; Marcel Dekker: New York, 1982;
p. 91 (c) Hoagland, J. J .; Wang, X. D.; Hipps, K. W. Chem. Mater. 1993, 5, 54. (d)
Dunbar, K. R.; Cowen, J.; Heintz, R. A.; Grandinetti, G.; Ouyang, X.; Zhao, H.
Inorg. Chem. 1999,38, 144.
(a) Manriquez, J. M.; Yee, G. T.; McLean, S.; Epstein, A. J.; Miller, J. S. Science
1991, 252, 1415. (b) Miller, J. S.; Calabrese, J. C.; McLean, R. S.; Epstein, A. J. Adv.
Mater. 1992, 4, 498. (c) Miller, J. S.; Vazquez, C.; Jones, N. L.; McLean, R. S.;
Epstein, A. J. J. Mater. Chem. 1995, 5, 707. ((1) Miller, J. S.; Calabrese, J. C.;
Vazquez, C.; McLean, R. S.; Epstein, A. J .; Adv. Mater. 1994, 6, 217. (e) Bohm, A.;
Vazquez, C.; McLean, R. S.; Calabrese, J. C.; Kalm, S. B.; Manson, J. L.; Epstein, A.
J .; Miller, J. S. Inorg. Chem. 1996, 35, 3083. (f) Brinckerhoff, W. B.; Morin, B.G.;
Brandon, E. J .; Miller, J. S.; Epstein, A. J. Appl. Phys. 1996, 79, 6147.
Zhao, H.; Heintz, R. A.; Ouyang, X.; Dunbar, K. R. Chem. Mater. 1999, II, 736.
Dunitz, J. D.; Bernstein, J. Acc. Chem. Res. 1995, 28, 193.
Heintz, R. A.; Smith, J. A.; Szalay, P. S.; Weisgerber, A.; Dunbar, K. R. Inorg.
Synth. in press.

(a) Pukacki, W.; Pawlak, M.; Graja, A.; Lequan, M.; Lequan, R. M. Inorg. Chem.
1987, 26, 1328. (b) Comelissen, J. P.; van Diemen, J. H.; Groeneveld, L. R.;
Haasnoot, J. G.; Spek, A. L.; Reedijk, J. Inorg. Chem. 1992, 31 , 198.

135