Citation
Chemical and physical characterization of hybrid organic-inorganic low-dimensional coodination polymers

Material Information

Title:
Chemical and physical characterization of hybrid organic-inorganic low-dimensional coodination polymers
Creator:
Woodward, Jonathan David
Publication Date:
Language:
English
Physical Description:
xi, 269 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Atoms ( jstor )
Azides ( jstor )
Dimers ( jstor )
Ions ( jstor )
Ligands ( jstor )
Magnetic fields ( jstor )
Magnetism ( jstor )
Magnetization ( jstor )
Molecular structure ( jstor )
Molecules ( jstor )
Chemistry thesis, Ph. D ( lcsh )
Dissertations, Academic -- Chemistry -- UF ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2003.
Bibliography:
Includes bibliographical references.
General Note:
Printout.
General Note:
Vita.
Statement of Responsibility:
by Jonathan David Woodward.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030315788 ( ALEPH )
52751663 ( OCLC )

Downloads

This item has the following downloads:


Full Text









CHEMICAL AND PHYSICAL CHARACTERIZATION OF HYBRID
ORGANIC-INORGANIC LOW-DIMENSIONAL COORDINATION POLYMERS












By

JONATHAN DAVID WOODWARD


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2003




























To My Parents and Grandparents














ACKNOWLEDGMENTS

I would very much like to thank my research advisor, Dr. Daniel Talham, for his

support, guidance, encouragement, novel ideas, and advice during my time in his research

group at the University of Florida. Despite some difficult times early on, Dan was

willing to work with me. He encouraged me to continue my efforts and press on so that I

could complete my degree. From Dan I have learned how to approach a research

problem from an objective standpoint, how to interpret data, and, most importantly, how

to present results in an acceptable manner.

I would also very much like to thank Dr. Renal Backov, a postdoctoral fellow in

our group for the past 2 years, for his help, support, and encouragement. Renal was

involved in many aspects of my graduate work and this dissertation would not be possible

without his efforts. I would like to thank Dr. Khalil Abboud for his work in solving

numerous crystal structures and for teaching me all about crystallography. I would also

like to thank Dr. C. Russell Bowers, and his graduate student, Bhavin Adhyaru, for the

use of the NMR equipment in the host-guest investigations. I would like to thank Bhavin

in particular for the rather large amount of time spent in completing NMR experiments as

well as writing and editing. I would like to thank Dr. Mark Meisel and his group

members (Dr. Hitoshi Ohnuki, Dr. Brian Watson, A. Nicole Morgan, Ju-Hyun Park, and

Diktys Stratakis) for performing bulk magnetic measurements on many of my samples

over the past 5 years. I would like to acknowledge Dr. John Reynolds, Dr. Gus Palenik,

Dr. Mark Meisel, and Dr. David Richardson for being part of my Ph.D. committee.








I would like to thank the past and present members of the Talham research group

(Dr. Brian Ward, Dr. Gail Fanucci, Jeff Culp, Eduardo Perez-Cordero, Chen Liu, David

Zipse, and Sarah Lane) for their support during my time here. I would also like to thank

Brian for help in getting me started in my research; Gail for teaching me how to operate

the lab machinery, including the computers; and Eduardo for teaching me some advanced

math in order to model magnetic data. I would also like to especially thank Jeff for

putting up with me during the time spent writing my thesis and his help with numerous

things for the past five years.

I want to also give special thanks to my parents, Ann and Dennis Woodward and

my grandmother, Virginia Prescott, for the support and encouragement throughout my

entire life whenever I needed it. Without them, I certainly would have not made it this

far.














TABLE OF CONTENTS
page

ACKNOW LEDGMENTS ................................................................................................. iii

ABSTRACT....................................................................................................................... ix

CHAPTER

1 INTRODUCTION .......................................................................................................... I

Low-Dimensional Materials........................................................................................... 2
Introduction to Molecular Magnetism ............................................................................ 3
Diamagnetism and Paramagnetism ............................................................................. 3
Basic Relationships..................................................................................................... 4
Non-Interacting Spin Systems and the Curie Law...................................................... 6
Interacting Spin Systems and the Curie-W eiss Law................................................... 7
Magnetic Exchange..................................................................................................... 9
Anisotropy................................................................................................................. 11
Dimers.......................................................................................................................16
Chains .......................................................................................................................20
L adders...................................................................................................................... 24
Self-Assembly of Supramolecular Architectures.......................................................... 27
Strategies for Building Supramolecular Architectures ............................................. 27
Factors Affecting the Structure of Supramolecular Architectures............................ 31
Porous Network Supramolecular Architectures........................................................ 33
Scope of the Dissertation.......................................................................................... 38

2 STRUCTURAL, THERMAL, AND MAGNETIC PROPERTIES INVESTIGATION
OF THREE TRANSITION METAL-4,4'-BIPYRIDINE COORDINATION
POLYMERS: [Ni(4,4'-bipy)3(H20)2](C104)2" .4(4,4'-bipy)-3(H20),
[Co(4,4'-bipy)3(H20)2](C104)2" .4(4,4'-bipy)-3(H20), and
[Cu(4,4'-bipy)3(DM SO)2](C104)2-2(4,4'-bipy) ............................................................ 39

Introduction................................................................................................................... 39
Experimental Section.................................................................................................... 41
Materials ................................................................................................................... 41
Synthesis of [Ni(4,4'-bipy)3(H20)2](C104)2 1.4(4,4'-bipy)-3(H20)......................... 42
Synthesis of [Co(4,4'-bipy)3(H20)2](C104)2.1.4(4,4'-bipy)-3(H20)........................ 42
Synthesis of [Cu(4,4'-bipy)3(DM SO)2](C104)2-2(4,4'-bipy).................................... 42
X-Ray Structure Determination................................................................................ 43








Thermal Analysis ............................................................... ......... .......... .................44
Magnetic Measurements...........................................................................................45
Results and Discussion ............................................................................................46
Compound Synthesis ................................................................................ ......... 46
Description of the Structures ................................................................................ .... 47
Structure of [Ni(4,4'-bipy)3(H20)2](C104)2-1.4(4,4'-bipy)'3(H20)...................47
Structure of [Co(4,4'-bipy)3(H20)2](C104)2 1.4(4,4'-bipy)-3(H20)..................53
Structure of [Cu(4,4'-bipy)3(DM SO)2](C104)2-2(4,4'-bipy)...........................59
Thermal Properties ................................................................................... .. .......... 63
M agnetic Properties ............................................................ ...................................... 72
Magnetic properties of [Ni(4,4'-bipy)3(H20)2](C104)2 1.4(4,4'-bipy)-3(H20) ....72
Magnetic properties of [Co(4,4'-bipy)3(H20)2](C104)2-1.4(4,4'-bipy)-3(H20) .... 77
Magnetic properties of [Cu(4,4'-bipy)3(DMSO)2](C104)2-2(4,4'-bipy) ........... 79
Conclusions ................................................................................................................... 82

3 A 31P MAS NMR INVESTIGATION OF THE HOST-GUEST PROPERTIES OF
TWO POROUS NETWORK SOLIDS,
[Ni(4,4'bipy)3(H20)2](C104)2- 1.4(4,4'bipy)-3(H20) and
[Co(4,4'-bipy)3(H20)2](C104)2.1.4(4,4'-bipy)-3(H20)................................................. 83

Introduction ...................................................................................................................83
NMR Spectroscopy .............................................................................................. 84
Experimental Section........................................................................................... .... 90
Materials...................................................................................................................90
Sample Preparation................................ ................................................................90
Gas Chromatography Analysis............................................................ ..................... 91
NM R Spectroscopy ....................................................................................... ............ 92
Powder X-Ray Diffraction........................................................................................ 92
Results and Discussion ................................................................................................. 92
Sample Preparation.................................... ........................................................ ....... 92
Guest Loss from Network Solids....................................................................... ....... 93
Guest Exchange Investigations of Compound 1 Involving TMPO .......................... 94
Gas chromatography results ............................................................... ................... 94
NMR results ...................................................................................................... 96
X-Ray results.................... ...................................................................................101
Guest Exchange in Compound 1 Involving Variable Quantities of TMPO ....... 106
Gas chromatography results ................................................................................106
NMR results ..................................... ................................................................... 107
X-Ray results ...................................................................................................... 110
Guest Exchange Investigations of Compound 1 Involving TEPO and TPPO........ 113
Gas chromatography results ... .................... .................................113
NMR results ........................................................................................................115
X-Ray results ...................................................................................................... 117
Guest Exchange in Compound 2 Involving TMPO, TEPO, and TPPO ................. 120
Gas chromatography results ................................................................................ 120
NM R results ........................................................................................................ 120








X-Ray results ...................................................................................................... 123
Conclusions................................................................................................................. 126

4 STRUCTURAL AND MAGNETIC CHARACTERIZATION OF A SERIES OF
AZIDO-BRIDGED COPPER(II) COORDINATION POLYMERS.......................... 127

Introduction................................................................................................................. 127
Azide Bridging M odes............................................................................................ 127
Superexchange Properties of Azide Bridges........................................................... 129
Experimental Section.................................................................................................. 137
Materials ................................................................................................................. 137
Synthesis of [Cu2(PhPyPy)2-p/-(N3)2(N3)2]............................................................. 137
Synthesis of [Cu2(terpy)2-p-(N3)4] [Cu2-/t-(N3)2(N3)2] ............................................ 137
Synthesis of [Cu2(terpy)2-/d-(N3)2(N3)2] [Cu3-/--(N3)4(N3)2] ................................... 138
Physical Characterization..................................................... ...................................138
X-Ray Structure Determination.............................................................................. 138
Magnetic Measurements......................................................................................... 139
Results and Discussion ............................................................................................... 140
Description of the Structures ..................................................................................140
Structure of [Cu2(PhPyPy)2-,p-(N3)2(N3)2] ..................................................141
Structure of [Cu2(terpy)2-/-(N3)4] [Cu2-/)u-(N3)2(N3)2]................................. 144
Structure of [Cu2(terpy)2-/t-(N3)2(N3)2] [Cu3-/p-(N3)4(N3)2]................................ 149
Electron Paramagnetic Resonance.......................................................................... 158
M agnetic Properties of [Cu2(PhPyPy)2-p-(N3)2(N3)2] ............................................ 159
Magnetic data...................................................................................................... 159
Magnetic model and fits...................................................................................... 161
Interpretation of the magnetic data..................................................................... 167
Rationalizing the sign and magnitude of the coupling constants........................ 168
Magnetic Properties of [Cu2(terpy)2-,/-(N3)4][Cu2-.-(N3)2(N3)2] ........................... 169
Magnetic data...................................................................................................... 169
Magnetic model and fits...................................................................................... 171
Interpretation of the magnetic data..................................................................... 180
Rationalizing the sign and magnitude of the coupling constants........................ 181
Magnetic Properties of [Cu2(terpy)2-/u-(N3)2(N3)2] [Cu3- -(N3)4(N3)2] ..................186
Magnetic data...................................................................................................... 186
M agnetic model and fits...................................................................................... 190
Interpretation of the magnetic data..................................................................... 193
Rationalizing the sign and magnitude of the coupling constants........................ 194
Conclusions................................................................................................................... 198

5 CONCLUSIONS......................................................................................................... 199










APPENDIX

A CRYSTAL STRUCTURES OF SELECTED LOW-DIMENSIONAL SOLIDS......202

B ATOMIC COORDINATES AND BOND ANGLES AND DISTANCES ............... 223

LIST O F REFEREN CES................................................................................................. 254

BIOGRAPHICAL SKETCH........................................................................................... 269














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

CHEMICAL AND PHYSICAL CHARACTERIZATION OF HYBRID
ORGANIC-INORGANIC LOW-DIMENSIONAL COORDINATION POLYMERS

By

Jonathan David Woodward

May 2003

Chair: Daniel R. Talham
Department: Chemistry

This dissertation presents experimental results from the synthesis and structural,

chemical, and physical characterization of representative low-dimensional coordination

polymers. The structural, thermal, and magnetic properties of a series of clathrated

porous network solids, [Ni(4,4'-bipy)3(H20)2](C104)2-1.4(4,4'-bipy)-3(H20), 1,

[Co(4,4'-bipy)3(H20)2](C104)2.1.4(4,4'-bipy)-3(H20), 2, and

[Cu(4,4'-bipy)3(DMSO)2](C104)2-2(4,4'-bipy) 3, are described first. These materials

consist of chains of transition metal ions (Ni(II), Co(II), and Cu(II)) bridged by

4,4'-bipyridine spacer ligands. The chains pack to form two-dimensional, non-

interpenetrated sheets with hydrophobic, rectangular cavities within the framework. The

sheets, in turn, pack to form three-dimensional structures with oblique channels

containing enclathrated guest molecules and counterions extending throughout the solid.

These enclathrated guests molecules are easily lost, suggesting that the samples are








thermally unstable. The magnetic properties of 1, 2, and 3 are similar in the sense that

only very weak exchange interactions are present between the metal centers.

The host-guest properties of 1 and 2 were investigated. Gas chromatography

experiments determined that both hosts exchange clathrated bipy molecules with

trialkylphosphine oxide probe molecules, TMPO, TEPO, and TPPO. While the uptake of

TMPO by both 1 and 2 is essentially complete, steric constraints are believed to limit the

uptake of TEPO and TPPO by the host. The trialkylphosphine oxides interact with acid

sites within the host as determined by 31P MAS NMR spectroscopy. TMPO interacts

with both coordinated water molecules (strong acid sites) and lattice waters (weak acid

sites) in compound 1 and coordinates directly to the metal centers in compound 2.

However, TEPO and TPPO seem to attack only the weaker acid sites within the hosts.

X-ray diffraction patterns show that the loss of bipy and uptake of the probe causes

significant structural rearrangements in 1 and only mild structural changes in 2.

However, the nature of these guest-exchanged products is unknown. These experiments

showed that solid-state NMR spectroscopy can be used to investigate host-guest

interactions.

Finally, the structural and magnetic properties of three azido-bridged copper(II)

ladder-like coordination polymers, [Cu2(PhPyPy)2-/P-(N3)2(N3)2], 4,

[Cu2(terpy)2-/u-(N3)4Cu2-/-(N3)2(N3)2], 5, and

[Cu2(terpy)2-p-(N3)2(N3)2Cu3-/l-(N3)4(N3)2], 6, are discussed. Compound 4 structurally

resembles ladder-like chains of weakly interacting end-on azido bridged copper(II)

dimers. Magnetically, compound 4 consists of antiferromagnetic chains of

ferromagnetically coupled S = V2 dimers. Compound 5 consists of ladder-like copper(II)








coordination polymers with double and single end-on azido bridges. Magnetically,

compound 5 consists of ladder-like stacks of weakly interacting tetramers with a

dominant and unusual antiferromagnetic exchange mediated through end-on azido

bridges. Compound 6 structurally resembles ladder-like chains of weakly interacting

copper(II) pentamers featuring both single and double end-on azide bridges.

Magnetically, compound 6 consists of antiferromagnetic stacks of pentamers with two

paramagnetic S = Vsites and ferromagnetically coupled trimers.













CHAPTER 1
INTRODUCTION

The original goal of the graduate research pertaining to this thesis focused on the

design, self-assembly, and physical characterization of molecular coordination polymer

ladder systems. This project was only partially successful; several new ladder-like

materials were isolated as a result of serendipity rather than rational design and did not

incorporate the desired structural and physical properties. However, in addition to the

ladder-like structures mentioned above, several new low-dimensional materials were

obtained from the course of this graduate research. The work presented in this

dissertation is concerned primarily with investigating the structural, chemical, and

physical properties of selected examples of these low-dimensional materials.

This chapter briefly provides the background information necessary to the

research detailed in the following chapters. A basic concepts of molecular magnetism as

applied to low-dimensional materials is introduced first. Included are discussions

concerning the magnetic properties of both paramagnetic and exchanged coupled

systems, including discrete oligomers and extended chains and ladders, and the theories

that model the magnetic behavior of such materials. A general introduction to the field of

supramolecular chemistry is then provided. Discussed are general strategies for building

supramolecular architectures, particularly via the self-assembly of simple molecular and

ionic components, general structural types of the resulting assemblies commonly

encountered, and practical applications of such materials.








Low-Dimensional Materials

The study of low-dimensional materials has been a rapidly expanding area of

solid-state chemistry. Low-dimensional materials have structural and physical properties

that are anisotropic in one or two dimensions. 1 3 These materials are often referred to as

"quasi" low-dimensional systems since, despite being part of a three-dimensional solid,

characteristic structural and physical properties exist principally within one or two

dimensions. One-dimensional chains and two-dimensional sheets (Figure 1-1) are the

most commonly encountered types of low-dimensional materials. The key structural

features of low-dimensional materials are strong electrostatic or covalent bonds along

chains or within sheets with weak, Van der Waals interactions between the chains or

sheets. As a result of this anisotropic bonding, electronic, magnetic, and transport

properties exist along chains or within sheets, or are enhanced compared to higher

dimensional analogs, but are small or negligible between the chains or sheets.4











A B C D

Figure 1-1. Examples of low-dimensional materials. A) One-dimensional chain.
B) Two-leg ladder. C) Three-leg ladder. D) Two-dimensional sheet.








Introduction to Molecular Magnetism

Diamagnetism and Paramagnetism

When a material is subjected to an external, homogeneous field, H, a

magnetization, M, is induced within the sample.5,6 The quantities Mand H, are related

by

am
X= -- (1-1)
DH

where X is the magnetic susceptibility. The susceptibility quantitatively measures the

response of a material to an applied magnetic field. If the magnitude of the field is small,

then to a good approximation, the magnetization is a linear function of the field, and the

susceptibility can be expressed as

M
=- = (1-2)
H

The total susceptibility, XT, is the sum of two components, a diamagnetic contribution,

Xdia, and a paramagnetic term, Xpara

XTZr = Xpara, + Xd,a (1-3)

Diamagnetism is a property of all matter originating from the interaction of paired

electrons with an applied magnetic field; since all materials have paired spins, all

materials have diamagnetic contributions to the total susceptibility. When a diamagnetic

material is placed within an external field, it is repelled since the sample produces a flux

opposed to the applied field and moves toward regions of lower field strength.5

Diamagnetic susceptibilities are typically small and negative (on the order of

-10-6 emu mol'-1 to -10-4 emu mol-1) and are independent of field strength and


temperature.5








Paramagnetism, on the other hand, is a property exhibited only by materials with

unpaired electrons, such as transition metal complexes, rare earth compounds, and

organic free radicals. When a paramagnetic material is placed within an external field, it

is attracted to the field because of the interaction of the unpaired spins with the applied

field and thus moves toward regions of higher field strength.5 In general, low fields and

high temperatures will tend to randomize the directions of these spins, resulting in small

or zero net magnetic moments. Low temperatures and high fields, however, will align

the spins with the field, resulting in a net moment. Paramagnetic susceptibilities are

typically positive and much larger in magnitude than the corresponding diamagnetic

susceptibilities, often in the order of 104 to 102 emu mol1 or more.5 Diamagnetic

susceptibilities can be estimated from Pascal's constants and subtracted from the total

susceptibility (or ignored, if small compared to the paramagnetic contribution) to obtain

the paramagnetic susceptibility.5',6 Unlike diamagnetic susceptibilities, paramagnetic

susceptibilities exhibit temperature dependence, often in very complex manners.

Basic Relationships

In classical physics, the magnetization, M, of a sample results from a variation of

its energy, E, in response to an applied magnetic field, B, through


M=- (1-4)


and

B= -uH (1-5)

where B is the magnetic induction field and p is the permeability.5,6 When p> 1, the

sample is paramagnetic and when u < 1 the sample is diamagnetic. Similarly, in quantum








mechanics, a microscopic magnetization, A, of an energy level, E,,, in the presence of an

external field is


(1-6)


/lin


where n = 1,2, 3, and so on. The macroscopic magnetization is then a weighted sum of

all microscopic magnetizations in the sample, given by the Boltzmann distribution law


NA NE (- ) exp(k-')


2exp( f)
PkBT (1-7)

where NA is Avogadro's number and kB is the Boltzmann constant and the denominator is

the partition function, Z. The magnetic susceptibility is the variation of the magnetization

with the external field

aM
X =P(--)
X=o (1-8)

where /M is the permeability of free space. At the limits of high temperature and low

field, Equations 1-7 and 1-8 can be simplified to give new relations that are no longer

functions of the derivatives E,,A/B. Expanding the energy levels, En, in a power series of

B, gives


E, = E) + E)B + E2 ...B2


(1-9)


where E(' terms are the Zeeman coefficients.7a Retaining only linear terms, substituting

back into Equation 1-8, and simplifying under the assumption that at zero field, the

magnetization is zero, the susceptibility vs. temperature is








r(1)2 V(0)
En B 2) T
NE(k--T-- 2E.)exp(- k )
k- n T B IT T
lexp( )
S kBT (1-10)

Equation 1-10 is the Van Vieck susceptibility.

Non-Interacting Spin Systems and the Curie Law

The simplest type of paramagnetism is that of an ideal paramagnet, a system

composed of non-interacting, randomly orientated spin centers (Figure 1-4A) 7b. In an

ideal paramagnet, the magnetic susceptibility is inversely proportional to temperature.

The Curie law is a simple relation that describes the variation of the susceptibility with

absolute temperature for an ideal paramagnet and is given by

C=- (1-11)
T

where C is the Curie constant given by

C 2 =S(S +)
3k B (1-12)

and NA is Avogadro's number, g is the g-Land6 value, /Bis the Bohr Magneton, and S is

the total spin of the system. A plot of the susceptibility versus temperature for a

paramagnetic material that follows the Curie Law is a simple hyperbola. Other

mathematical manipulations of the Curie law are useful as well. A plot of the inverse

susceptibility versus temperature for a paramagnet is a straight line where the slope is the

Curie constant and the x-intercept is zero

(- T
_= -T (1-13)
C








while a plot of the product of susceptibility and temperature vs. temperature is a

horizontal line

%T=C (1-14)

Plots of the X, j1, and XT vs. T for an arbitrary S = material that obeys the Curie law,

with g = 2.10, are shown in Figure 1-2.

Interacting Spin Systems and the Curie-Weiss Law

In many paramagnetic materials, the unpaired electrons on the spin centers can

interact with one another and the magnetic behavior is no longer ideal. The Curie-Weiss

law, a semi-empirical modification to the Curie law, is a "first approximation" to model

the magnetic behavior of materials with weak interactions between the spin centers, and

is given by

C
Z .........(1-15)
(T-0)

In general, the interactions between the spins that cause these deviations are referred to as

ferromagnetic and antiferromagnetic correlations. In the Curie-Weiss law, these

interactions are described by 0, the Weiss constant. When 0 > 0, the interactions are

ferromagnetic, when 0 < 0, the interactions are antiferromagnetic, and when 0 = 0, the

sample is paramagnetic. The Curie-Weiss law is valid for materials that undergo long-

range order above the ordering temperature (i.e. T>> T). Plots of the x, X1, and XTvs. T

for antiferromagnetically (0 = -15 K) and ferromagnetically (0 = 15 K) coupled, S= V2

materials (g = 2.10) that obey the Curie-Weiss law are shown in Figure 1-2. Note that

when 0 > 0, the inverse susceptibility has a positive y-intercept and XT increases as T

decreases. When 0 < 0, the y-intercept is negative and the XT product decreases when T











decreases. The Curie-Weiss law is only a simple, general, empirical correction for


describing deviations from ideal paramagnetism by accounting for the interactions of


unpaired spins for magnetic materials especially when the structure is unknown. More

0.10
Paramagnetic
0.08 ---Antifteromagnetic
..... Ferm ~ n
...... Ferromagnetic
x~ 0.06

004

0.02 A

0.00 L
0 50 100 150 200 250 300
T (K)

300



200 .
150 .

100 .'
Paramagnetic
if ane
50 ,- -- Antiferromagnetic
...... Ferromagnetic

0 50 100 150 200 250 300
T (K)



2.0 -- Paramagnetic
-. Antfterromagnetic
1.5 '. -.. Ferromagnetic
""..................... .....
bL 1.0

,"/" C

0.0

0 50 100 150 200 250 300
T (K)



Figure 1-2. Common Curie (solid line) and Curie-Weiss (broken lines) law plots for an
S ='/ magnetic material (g = 2.10). A) X / C vs. T. B) C / vs. T. C) T / C vs. T. The
dotted lines represent ferromagnetic coupling (0= 15 K) and the dashed lines represent
antiferromagnetic coupling (0= -15 K). The susceptibility is normalized by division by
the Curie constant, C.








complex magnetic systems, such as dimers and chains, require more complicated models

to describe magnetic behavior.

Magnetic Exchange

Magnetic exchange interactions are quantum effects that describe the interactions

among unpaired spins in magnetic materials. Exchange interactions originate from a

combination of the Pauli exclusion principle and electronic repulsions.8 Two principal

types of exchange can be distinguished: direct exchange and indirect exchange.9 Direct

exchange interactions result from the through-space overlap of spin orbitals, as in direct

metal to metal bonding (Figure 1-3). These interactions are typically weak since

electronic repulsive forces are large due to the close proximity of the unpaired electrons

to one another. 10 For indirect exchange, the unpaired spins are coupled via nonmagnetic

intermediaries, such as bridging diamagnetic atoms or molecules or itinerant electrons of

conducting solids. The first case is referred to as a superexchange interaction (Figure

1-3) and the second case is referred to as RKKY exchange. Superexchange interactions

are electronic (and not magnetic) interactions and are usually much larger in magnitude

than the corresponding direct exchange, because intermediary groups increase the

distance between metal centers, thus reducing electronic repulsions. 10 Pairwise

exchange interactions can be expressed mathematically by a spin-Hamiltonian equation:

SH=-2Z Jg, Sj (1-16)
'>j

where the sum is taken over all nearest neighbor interactions between spins, S, and Sj,

with the magnitude of those interactions given by J, the coupling constant.5 If J> 0 and

the unpaired electrons interact such that the spins align in a parallel fashion, the exchange









is referred to as ferromagnetic and the magnetic ground state is triplet (Figure 1-4B)7b.

IfJ < 0, the unpaired spins align in an antiparallel manner, the exchange is referred to as

antiferromagnetic and the magnetic ground state is singlet (Figure 1-4C)7b. The sign and

magnitude of the coupling constant can depend on many factors, such as the number of

unpaired electrons exchanged, structural parameters of the magnetic entity, orbital

overlap of the spin centers and nonmagnetic intermediary, and electronic properties of the

intermediary.

Other arrangements of unpaired spins within a system are also possible.

Ferrimagnetism occurs in systems incorporating two alternating effective spins (Figure

1-4d)7b. For example, in a chain containing alternating S = V2 and S = 1 spin centers, the

unpaired electrons align antiparallel. However, because the effective spin values are

different, the moments do not cancel each other out completely, resulting in a net

magnetization.9 In a canted antiferromagnetic system (Figure 1-4e)7b, the magnetic

moment vectors of nearest neighbors are tilted. A small, but finite, magnetization results

because the moments are not fully antiparallel and cancel each other out.9

A distinction should be made between magnetic exchange and magnetic ordering.

Magnetic exchange describes short-range correlations (the local interactions between
B
unpaired electrons). On the other hand, in a transition to long-range magnetic order, the

unpaired spins over a relatively large domain in a magnetic material will spontaneously

align in the absence of an externally applied field at some critical temperature.9

Alignment of the spins can be parallel or antiparallel corresponding to ferromagnetic or

antiferromagnetic ordering, respectively. The critical temperature for ferromagnetic

ordering is the Curie temperature, Tc, and for antiferromagnetic ordering, the Neel








temperature, TN.5,11 The size of this domain is the magnetic correlation length, E, the

distance over which unpaired spins are ordered; the divergence of at a critical

temperature is necessary for long-range ordering to occur in a magnetic material.2,11 In

an ordered ferromagnetic state, the unpaired spins align in a parallel manner and a net

magnetization results (Figure 1-5). In an antiferromagnetically ordered state, the

unpaired spins align in an antiparallel fashion and no net magnetization results (Figure

1-5).




M M A
0 % ..............


x



M M B



x
X


Figure 1-3. Magnetic exchange pathways. A) The transition metal (M) magnetic orbitals
involved in through-space, or direct exchange. B) A superexchange pathway, mediated
by a nonmagnetic ligand (X).


Anisotropy

Before providing a brief discussion of anisotropy, the concepts of lattice and spin

dimensionality must be described (Figure 1-6). Lattice dimensionality (d) refers to the





geometrical distribution of the spins in space.9a When d= 0, the system consists of
discreet, zero-dimensional (0-D) oligomers or clusters. When d = 1, the system consists


7


4*4


I V I


Figure 1-4. The spin angular momentum vectors representing various interactions
between unpaired electrons. A) Paramagnetic. B) Ferromagnetic. C) Antiferromagnetic.
D) Ferrimagnetic. E) Canted antiferromagnetic. Adapted from reference 7b.











tt 1' t t i t
At t t






Figure 1-5. The spin angular momentum vectors in ordered magnetic states.
A) Ferromagnetic ordering. B) Antiferromagnetic ordering. Note the ordered
ferromagnetic state will have a net magnetic moment in the direction indicated by the
hollow arrow while the ordered antiferromagnetic state will possess no resultant moment.

of one-dimensional (1-D) infinite chains and when d= 2, the system consists of infinite,
two-dimensional (2-D) sheets. The spin dimensionality, n, refers to the contributions by
the vector components of the spin angular momentum. When n = 1, the spin has only one
component, S., and the system is Ising. When n = 2, the spin has two components, Sx and
Sy, and the system is Planar. When n = 3, the spin has three components, S:, Sx, and Sy,

and the system is Heisenberg.2,9
The concepts of spin and lattice dimensionality are especially important in

determining whether a magnetic material can undergo long-range order.2,12-14 The
magnetic ordering phenomenon can be understood by considering that the transition from
short-range to long-range order in these systems is accompanied by a crossover in the

effective lattice dimensionality or the effective spin dimensionality of the system.9 For

example, although no long-range order for a 2-D Heisenberg system is predicted,15 there









are numerous examples of such systems that in fact undergo such a transition.2 This

discrepancy can be explained through a crossover in the lattice dimensionality. The

ordering occurs as a transition from a two-dimensional to three-dimensional lattice

because, at the critical temperature, the interplanar couplings become important.9

Magnetic exchange interactions are often not isotropic in all three lattice and/or

spin dimensions. For highly anisotropic magnetic materials, the predicted magnetic

behavior will no longer be Heisenberg, but may resemble (for example) Ising and Planar

behavior instead.9 A more general spin-Hamiltonian equation that accounts for exchange

anisotropy is

/ =--(JSX -. jx +JyS,, ySjy + JS, -. S) (1-17)
i>J

When J, = Jy = J., equation represents the isotropic Heisenberg model. If the components

of Jare different, then the exchange is anisotropic. For example, when the exchange is

principally characterized by in-plane components, Jx = Jy, and J. = 0, the Hamiltonian is

referred to as the XY model, but ifJ Jy # J., equation is referred to as the XYZ

model.2,16 Table 1-19a,9b summarizes a few spin-Hamiltonian models for various cases

of differing spin anisotropy and spin dimensionality.9 Note the subtle distinction

between the Z model and the Ising model.9 For the Z model, n = 3 and Jx = Jy = 0 but Sx

S0 and Sy # 0 meaning that, although no exchange interaction occurs between nearest

neighbor spins in the x- and y-directions, only in the z-direction, the total spin angular

momentum still has x, y, and z components. In contrast, the Ising model, n = 1 and Jx = Jy

= 0 and Sx = Sy = 0 meaning that, there is no exchange interaction in the x- and y-








directions, only in the z-direction, and the spin angular momentum only has a z

component.

Table 1-1. Summary of Some Spin-Hamiltonian Models for Various Cases of Spin
Anisotropy and Spin Dimensionality
Spin-Dimensionality Interaction Model
n = 3 Jx = Jy = J Heisenberg
S = S2 = S-2 Jx=Jy; JZ= O XY
Jx = Y = 0 ; Z

n = 2 J. = Jy Planar
Sx = Sy Jx = 0 ; Jy Planar Ising

n = 1 Ising
s82

Adapted from references 9a and 9b.


Anisotropy in the exchange interaction often result from zero-field splitting or

spin-orbit coupling effects.6 In each of these cases, the anisotropy is represented by

additional terms in the Hamiltonian such as

IT n 2 S(S + 1) /z 2\/ o
Z = D[S- S(S + 1)]+ E(S3- S) (1-18)


Hs = 2,L-S (1-19)
i

Equation 1-18 represents the effect of zero-field splitting, where D is the axial or

single-ion anisotropy factor and E is the rhombic or in-plane anisotropic component.

Equation 1-19 is the spin-orbit coupling Hamiltonian where X, is the spin-orbit coupling

parameter and L and S are the orbital and spin angular momentum operators,

respectively. Spin-orbit coupling arises from the coupling of a 2S+'Fground state with an

excited state from the same magnetic center.6 The excited states are well separated in

energy from the ground state and are not appreciably populated at room temperature.

















d=1 d=2 d=3

A B C







n= n=2 n=3
D B F
Figure 1-6. Lattice dimensionality and spin dimensionality. A) A one-dimensional chain
has d = 1. B) A two-dimensional sheet has d = 2. C) A three-dimensional lattice has
d= 3. D) An Ising system has n = 1. E) A Planar system has n =2. F) Heisenberg
system has n =3.


Spin-orbit coupling can lead to g-factor anisotropy or zero-field splitting effects. Zero-

field splitting describes the splitting of the Zeeman components in the absence of an

external field due to the coupling of an S > /2 ground state with excited states.6 Magnetic

anisotropy can arise from other sources as well, such as higher-order exchange

interactions, orbital angular momentum contributions, and low-symmetry ligand fields or

magnetic dipolar fields that couple the moments to certain directions of a crystal.9

However, the assignment of the origins of the anisotropy in the magnetic interactions is

often difficult.

Dimers

A dimer, denoted by a lattice dimensionality of d =0, represents the simplest type

of interacting magnetic system. In a dimer, two spin centers can interact directly through








space or indirectly via intermediary superexchange ligands. Only the second case is

considered here. In general, no long range ordering is possible for dimers and other d = 0

oligomers and clusters (unless a crossover in lattice dimensionality occurs) and thus the

magnetic interactions are only short-range correlations between the spin centers.

Figure 1-7 shows the structure of an S = / dimer, a dinuclear copper(II) moiety

bridged by two diamagnetic ligands (X) capable of mediating a superexchange

interaction between the two spin centers. Ancillary ligands (L) fill the remaining

coordination sites on the metal ions.

If the unpaired electrons interact with one another, then individual spin quantum

numbers for each metal center, SA = SB = /2, are no longer valid.6 The spin states of the


L


U CH~
L.. "*' 'x 0 I ""


L

Figure 1-7. A square pyramidal copper(II) dinuclear complex where the metal centers are
bridged by two diamagnetic ligands, X, and the remaining coordination sites are filled by
ancillary ligands, L.


dimer are now S = 0 and S = 1. In general, the energies of these spin states are not equal,

but separated by an energy gap, J, defined as

J=E(S = 0) E(S= 1). (1-20)



The Hamiltonian for the isotropic exchange of a magnetic dimer is

H =-2AA "St + uBB (gA SA + gB-S) (1-21)








where SA and SB are the spin angular momentum operators representing the unpaired

electrons on each metal center and J is called the isotropic exchange parameter that

quantitatively accounts for the energy of the superexchange interaction. The second term

is the Zeeman perturbation. Figure 1-8 shows the relation between the energy levels of

the magnetic spin states as a function of applied field.6 At zero field, the spin-

Hamiltonian splits the two degenerate S = states into the S = 0 and S = 1 states. In the

presence of a magnetic field, the Zeeman term further splits the energy level of the triplet

removing all degeneracy but does not affect the singlet state. When J < 0, the S = 0

singlet state is the magnetic ground state and the exchange is antiferromagnetic. In this

case, the spins are coupled in an antiparallel fashion resulting in no net magnetic moment.

When J> 0, the S = 1 triplet state is the ground state and the exchange is ferromagnetic.

In this case, the spins are coupled in a parallel fashion resulting in a net magnetic

moment. The magnitude of the coupling constant is related to the difference in energy, or

energy gap, between the ground and first excited state. From both the spin and Zeeman

Hamiltonians, the resulting four energy levels for an S = V dimer as a function of external

field are E\ = 0, E2 = J, E3 = J + piBgB, and E4 = J pBgB. The temperature dependence

of the magnetic susceptibility, describing the changes in population of these magnetic

energy levels, is given by

2J
2 2kT
2NRNAI g exp (1-22)
X, lim er ~ ----. 1 J ( 1 -2 2*)
3kBT
1 + 3expk""


A simulation of the X vs. T from Equation 1-22 for an S = /2 dimer with g = 2.2,

J k"j = -50 K, and + 50 K is shown in Figure 1-9. Note that, when the exchange is

antiferromagnetic, a maximum in the susceptibility is observed.











E B= 0 + tBgB
9BgB
A__

S=O- 0
D B
E B=O
S=1- 0
B +B

tS= --j
0- gB
B B


Figure 1-8. The splitting of the magnetic energy levels in an S = V2 dimer. The zero-field
spin states of the dimer correspond to the S = 0 (singlet) and S = I (triplet) states. An
externally applied field splits only the triplet state. A) In an antiferromagnetically
coupled dimmer, the ground state is singlet. B) In a ferromagnetically coupled system
the ground state is triplet. The coupling constant, J, is the energy gap between the ground
state and the nearest excited state.

Dimers are among the most extensively studied magnetic systems.6 In particular,

/j-dioxo bridged copper(II) dimers and/a-diazido copper(II) dimers have received

considerable attention.7,17-21 The systems are convenient for the testing of theoretical
models of magnetic systems. They also provide an understanding of how structural and
electronic parameters affect the resulting magnetic properties. Finally, an understanding
of the magnetostructural correlations for these simple systems provides a basis for
continuing research efforts directed toward the design of novel magnetic materials with
specifically tailored magnetic properties.








Chains

Magnetic chain compounds are one-dimensional systems with a lattice

dimensionality of d = 1. Uniformly spaced magnetically interacting spin centers

represent the simplest class of chains. Since the spin centers are equivalent along the

chain, the nearest neighbor exchange interactions between the spin centers are also

equivalent. Examples of uniform chains include [Cu(ox)]-%H20 (ox = oxalate),

(C6HiI)CuC13,22'23 and [Ni(en)2(N02)](C104) (en = ethylenediamine).24 Figure 1-10

schematically represents an S = /2 uniform chain of Cu(II) ions.

The spin Hamiltonian representing the isotropic nearest neighbor superexchange

between the metal centers over n sites is

H=-J -A, "SA, (1-23)
i=!

When n is infinite, no analytical solution can be calculated in order to determine

the energies of the magnetic spin states and the susceptibility. However, the energies and

susceptibility can be calculated exactly for small chains of finite number of metal centers.

Then, by extrapolating these results to the case of an infinite chain, numerical solutions

for the energies of the magnetic states and susceptibility can be approximated.25,26 The

temperature dependence of the magnetic susceptibility for an S = /2 uniform chain,

extrapolated from a ring of n = 11 spin centers, is

SN^Ag 2 0.25 + 0.074975X + 0.075235x2
x= kT 1.0 + 0.9931x + 0.172135x2 + 0.757825x3 (1-24)

where

X J- (1-25)
kBT
















0.008

0.006

0.004

0.002

0.000


-0.002


.... Ferromagnetic
S--Antiferromagnetic
I
\
\
\
\


L J
0 60 120 180 240 300
T (K)


0 50 100 150 200 250 300
T (K)


Figure 1-9. Temperature dependent magnetic susceptibility plots for an S = V2 dinuclear
complex (g = 2.10) modeled after Equation 1-22. A) ; vs. T. B) f-' vs. T. The dotted
lines represent ferromagnetically coupled dimers (J kB1 = 50 K) and the solid lines
represent antiferromagnetically coupled dimers (J kB' = 50 K).


Cui Cui+- Cu-+2-


Figure 1-10. An exchange coupled, uniform S = V chain of Cu(II) ions.


.1








Note that this equation is valid only for antiferromagnetic exchange along the chains (J <

0) since, as T approaches 0, then ; converges to 0 if n is small and finite. If n is infinite,

; does not converge to 0 but to a finite value since the ground and excited states form a

continuum of energy levels with no energy gap between ground state and next highest

energy level.6 When J> 0, then X diverges as T approaches 0. No corresponding

analytical expression to describe the magnetic behavior of a ferromagnetically coupled

uniform chain has been reported.6

A high temperature series expansion,27 valid for both positive and negative J

values, to describe the magnetic susceptibility for a S = uniform chain is given by

4g2__ 1.0 + 5.7980x + 16.9026x2 + 29.3769x3 13 (
4kBT 1.0+2.7980x +7.0087x2 + 8.6539x3 + 4.5743x4 (1-26)

where

x== (1-27)
kBT

Regardless of the ferromagnetic or antiferromagnetic exchange interactions that

are present along the chains, in principle, the isolated one-dimensional chains

magnetically order only at T = 0. However, in real solids, chains are never completely

isolated but experience interchain interactions, usually much weaker than the dominant

intrachain exchange due to a crossover in lattice dimensionality. At low temperatures,

these interchain interactions become important, the one-dimensional chains effectively

behave as three-dimensional solids, and magnetic ordering occurs at finite temperatures.

A more complicated type of one-dimensional system is the alternating, or zig-zag,

chain. In this system, there are two distinct types of spin centers and, as a result, there is








a regular alternation of the exchange interactions, J and J', representing the nearest

neighbor and next nearest neighbor couplings, respectively. Examples of alternating

chains include Cu(N03)2-5/2H20 28,29 and (ipa)CuCl3 (ipa =

isopropylammonium)30,31. For instance, Figure 1-11 schematically represents an S =

alternating chain of Cu(II) ions.


J aJ
Cu2i. --Cu2-i Cu2i+1-


Figure 1-11. An exchange coupled, alternating S = '/2 chain of Cu(II) ions.


The spin-Hamiltonian representing for an alternating chain system is


H =-J[SA,,.SA +aS A2, AS,. (1-28)
t=1

where a is the alternation parameter such that 0 < a < 1 such that J'= al. Note that

when a = 0, the one-dimensional system behaves as isolated magnetic dimers, and when

a = 1, the system corresponds to a uniform chain. Again, when n is infinite, no analytical

solution can be used to determine the energies of the magnetic spin states and the

susceptibility for the alternating chain. However, analytical solutions can be obtained in

a similar fashion for uniform chain.

The temperature dependence of the magnetic susceptibility for an S = /

alternating chain, extrapolated from a ring of n = 10 spin centers,32 is

Ngg2 A+Bx + Cx2
SkT 1+Dx+Ex2 +Fx3 (1-29)


where









x=- (1-30)
kBT

The coefficients A F are functions of a and are provided elsewhere.6 This equation is

valid only when both exchange interactions are antiferromagnetic J< 0 and both

exchange parameters are within an order of magnitude of one another. In contrast to

uniform chains, when J < 0 and 0 < a < 1, then X converges to 0 as T approaches zero

since an energy gap between the ground and excited magnetic states is present.6

Ladders

Ladders, low-dimensional quantum systems that fall in between one-dimensional

chains and two-dimensional sheets, consist of a finite number of magnetically coupled

chains of spins (Figure 1-1).33,34 In principle, one might expect that a smooth crossover

in physical properties from chains to sheets would result if one assembled chains to

progressively form ladders of increasing width but this is not generally true.33

The spin-Hamiltonian that, in general, represents ladder-like magnetic systems is

n n
H =-Jl -",, -S*+,a -J-". .S0* (1-31)
a=l,2 i=1 i=1

where S,, represents the spin operator at site i (i = 1, 2 ........n) on the leg a (a = 1,2,

.....) of the ladder with n rungs.33 The terms J.L and ./| denote the intra- and interrung

exchange couplings, respectively. In "ideal" ladders, the magnitude of the coupling

along the legs is comparable to the magnitude of the coupling along the rungs, J_ I J11 1.

When JL / J1\ tend to zero, the exchange between the rungs is small compared to the

exchange along the legs and the ladder behaves as a system of isolated chains.

Conversely, when J_ / J11 tend to infinity, the exchange along the legs is small compared








to the exchange along the rungs and the ladder behaves as a system of isolated dimers.

Ideal spin ladders should also be well isolated from one another since appreciable

interladder coupling (J') can cause transitions from the spin liquid ground state to a

magnetically ordered state.33

The magnetic properties of ladders with an even number of legs are drastically

different from those with an odd number of legs.35 Ladders with an even number of legs

(Figure 1-1), such as the inorganic cuprate SrCu203, are characterized by short-range spin

correlations along the legs, a spin-liquid ground state.33,35,36 Even-leg ladders consist

of spin-singlet pairs with a spin-spin correlation length along the legs that show an

exponential decay produced by the presence of a finite spin gap and tends to 0 as T

approaches 0. In contrast, a ladder with an odd number of legs, such as Sr2Cu305,

exhibits power-law decay of spin-spin correlations that tend to finite values as T

approaches 0 that are magnetically ordered due to the presence of gapless spinless

excitations.33,35,36

Ladders with an even number of legs are characterized by a spin gap.33,35 A

spin gap is a finite energy gap between a nonmagnetic ground state and the first excited

triplet state (Figure I-12a). No continuum of excited states exits directly above the

ground state. If the rungs of an antiferromagnetic ladder interact weakly with one

another, i.e., Jt / J\, the ground state has a total S = 0 since the spins on each rung are in a

singlet state.33 To promote the ladder to the lowest excited state with a total S = 1, one

of the pairs of spin singlets of the rungs must be promoted to an S = 1 triplet. A quantum

of energy, called the spin gap energy, AEsg, is required to excite one of the rung singlets

into a triplet state. A frustrated spin state results since the spins are now aligned parallel








along one direction but anti-parallel along the other direction (Figure 1-12b).

Examination of the magnetization of a ladder compound as a function of changing

external applied field can identify a spin gap. An abrupt increase in magnetization at a

particular magnetic field indicates the presence of the energy gap (Figure l-12c).

There has recently been a growing interest in the preparation and study of ladder-

like molecular and solid-state materials. In the field of supramolecular chemistry, ladders

represent one of many familiar structural topologies that are possible from the self-

assembly of simple molecular or ionic nodes and spacers, such as metal ions or

complexes and multifunctional bridging ligands, respectively, under certain

stoichiometric ratios and reaction conditions.37 Furthermore, molecular and solid-state

ladders, as a consequence of their structure, often possess open or enclathrated cavities

and extended channels that exhibit unique inclusion and catalytic phenomena.37 In the

field of low-dimensional materials, ladders represent a structural intermediate between

one-dimensional chains and two-dimensional sheets.33 Ladders represent ideal systems

to investigate the gradual change in physical properties as the dimensionality increases

from ID chains, to quasi ID / 2D systems, to 2D sheets. Furthermore, copper oxide

ladders are part of the structure of many solid-state materials such as (Sr, Ca)Cu203 that,

upon doping with holes, often exhibit superconductivity at liquid nitrogen temperatures

or higher.35,38 These copper oxide ladders are antiferromagnetic and it is believed that

the superconductivity originates and is sustained within this portion of the structure. In

order to better understand the origin and mechanism of high-temperature

superconductivity, it is desirable to synthesize model low-dimensional compounds that

adopt a ladder-like structure similar to those found in the cuprates.








Self-Assembly of Supramolecular Architectures

The recent research efforts devoted to the rational design and crystal engineering

of supramolecular solid-state materials were initially spawned by a concerted interest

toward developing methods for predicting the crystal structures of organic compounds.39

However, ongoing work in this field has continued due to the prospect of developing new

materials with interesting structures and diverse, exploitable properties. One particular

area in supramolecular chemistry has focused on synthesizing hybrid organic / inorganic

materials through the self-assembly of simple, molecular or ionic building blocks.

Compared to purely organic or inorganic analogous systems, these materials often

possess improved thermal, chemical, and mechanical stability and exhibit unique or

enhanced physical properties.40 Hybrid organic / inorganic materials are potentially

useful in a wide variety of applications including low temperature catalysis,41-43

inclusion phenomena,37,44,45 magnetism,40,46-49 electrical conductivity,50-52

photochemistry,53 and second-order nonlinear optical behavior.54-57

Strategies for Building Supramolecular Architectures

Under certain conditions, solution-phase molecular and ionic building blocks can

self-assemble into discreet clusters or oligomers or extended one-, two-, and three-

dimensional solids sustained through various types of chemical interactions such as

coordinate covalent bonding,43,58 electrostatic attractions,59,60 hydrogen bonds,61-65

and t--stacking.66-68 In general, the overall molecular and solid-state structure is

controlled by a combination of the binding constraints, geometrical preferences, and

relative stoichiometric quantities of these building blocks. Therefore, much research has

been devoted to the development of general strategies to better predict, design, and








control the structure of supramolecular architectures with the above guidelines in mind.

By far the simplest and most common strategy applied is the node and spacer method.

The node defines the overall geometry of the structure while multifunctional or

multitopic spacer ligands are tethered to and propagate the geometrical preferences of

node throughout the solid.69,70 These simple, modular components are chosen as

starting materials because their inherent bonding and geometrical propensities allow

some degree of control and predictability over the structure of products.39 Furthermore,

by the careful selection or design of these building blocks, the physical properties of the

solids can be "tuned" or "tweaked" to meet specific needs.39

A number of different approaches derived from the node and spacer strategy can

be applied toward the design of supramolecular materials based on the nature of the

building blocks or the chemical interactions responsible for sustaining the structure. The

most common approach is the generation of hybrid organic / inorganic networks that are

simple extensions of a specific transition metal or metal complex (metal center chelated

by one or more poly-hapto ligands) geometry.39 The metal coordination environment

functions as the node and the spacer ligands are typically bridging ligands. In most cases,

the molecular and solid-state structure of the assembly is sustained through coordinate

covalent bonding and thus the structure is often referred to as a coordination polymer.

A variety of one-, two-, and three-dimensional coordination solids with novel

topologies have been obtained using rigid, multifunctional spacer ligands such as

pyrazine,71,72 4,4'-bipyridine, 44,73-79 4,4'-azobis(pyridine),80, bis(4-

pyridyl)benzene,81 bis-(4-pyridyl)-ethylene,81 2,4,6-tris(4-pyridyl)-l ,3,5-trazine,82-85

1,3,5-tris(4-ethynylbenzonitrile)benzene,86 and 1,3,5-benzenetricarboxylic acid45'87











Excited
Triplet
States


Singlet
Ground
State


#1 If
M--M
I I









M


A I1 --- #

lop-


- AEg


Figure 1-12. Spin Gap. A) The energy gap, AEg, between the singlet ground state and
excited magnetic spin states in a two-leg ladder. B) The excitation of one of the strongly
coupled rungs into a triplet state results in a frustrated spin state. C) The spin gap can
often be identified by an abrupt increase of the field dependent magnetization.


JTAEg








with transition metal cations such as Cu+, Cu2+ Ag+, Zn2+, Cd2, Mn2+, Co 2+, Fe3 and

Ni2. Predictable and structurally well-defined products are often produced from the self-

assembly of rigid spacers with metal cations. In fact, rigid spacers are particularly ideal

for designing and synthesizing porous network solids capable of clathrating small

molecules. On the other hand flexible spacer ligands have not been extensively exploited

except in a few cases.44,'81,88 Flexible spacers may allow the synthesis of hybrid

organic / inorganic solids with structural features not found in those materials with rigid

linking groups.89'90 Unfortunately the incorporation of a higher degree of flexibility

into the building blocks reduces the amount of predefined geometrical information from

the reactant components and a multitude of different structures can arise from identical

metal-ligand combinations or minor experimental variations.91

Hydrogen bonding interactions between nodes and spacers offer an alternative,

but equally powerful, approach to the control of solid-state structures. Molecular

components with complementary hydrogen-binding sites are well known and can be

readily incorporated into building blocks to produce extended polymeric architectures

sustained by interactions either directly between nodes or mediated by spacer ligands6l

Because these interactions are directional and their formation is often reversible,

hydrogen bonded molecular assemblies are ideal for the design of structurally flexible

networks with and adjustable size pores.61,62 Obviously, networks can be also

assembled by the concurrent action of both coordinate covalent and hydrogen bonds as

well, though species like these are relatively less common.92

An alternative approach to the metal node and ligand spacer method involves the

exploitation of exodentate multitopic ligands that act as both the nodes and spacers in the








network architecture. This method is exemplified by the construction of supramolecular

architectures from solely organic molecules and has generated a number of architectures

mimicking the structures of known inorganic minerals.70,'84,'86,'93 However, this

method is also less predictable in terms of rationally developing functional topologies.70

Factors Affecting the Structure of Supramolecular Architectures

A number of different factors can profoundly affect the molecular and solid-state

structure of self-assembled coordination polymers, such as the stoichiometric ratios of the

node and spacer, the different oxidation states and coordination preferences of the metal

ion nodes, the structural and bonding propensities of spacer ligands, and the reaction

solvents. The metal:ligand stoichiometry is one of the most important factors

determining the dimensionality and limiting the possible topological architectures that

can occur in a coordination solid. Consider the possible types of structural motifs, or

supramolecular isomers,94 that result from the combination of various spacer ligands

with metallic moieties. A 1:1 ratio limits the architectures to either one-dimensional

linear or zig-zag chains.64,69,89,95-102 A 1:2 ratio can produce two-dimensional grids

with rectangular cavities if the node is planar or octahedral 40,44,74,75,92,95,103-105 or

three-dimensional diamondoid structures if the metal center is in a tetrahedral or S4

environment.79,106-108 A metal :ligand ratio of 1:1.5 can produce any of six very

different architectures including the molecular ladder,70,88,94,109 brick wall,74,88

Lincoln log,73,110 tongue and groove or bilayer,70,76, 11 herringbone, 112 or three-

dimensional frame113 if the metal moieties adopt only square planar or octahedral

geometries. Additionally a trigonal metal center can generate two-dimensional

hexagonal or honeycomb nets with a three-fold symmetric tritopic spacer such as 1,3,5-








trisubstituted benzenes in a 1:1.5 metal:ligand ratio.86,106 The molecular railroad has

been the only topology observed for a 1:2.5 stoichiometry but, unlike the ladder, the

spacers are present as both bridges and terminal ancillary ligands. 14,115 Three-

dimensional rectangular and interpenetrated grids, are possible with metal:ligand

stoichiometries of 1:3 or smaller. 116-119 However, given the ubiquity of octahedral

coordination environments, it is somewhat surprising that simple three-dimensional

octahedral polymers remain largely unexplored. 117 Figure 1-13 schematically depicts

selected supramolecular architectures built from the self-assembly of nodes (metal ions)

with spacers multifunctionall bridging ligands).

Many of these topologies, such as the two-dimensional sheets, incorporate

cavities and channels as part of the molecular and solid-state structure. In a few cases,

the pores are open 40 but the void space is usually filled either by clathrated guests or

interpenetration of neighboring lattices. This packing diversity between interpenetrated

and noninterpenetrated porous solids can be identified as another form of supramolecular

isomerism.115

The oxidation state and the coordination preferences of transition metals are also

critical in determining the final molecular and solid-state structure. For example, Cu2+

normally prefers an axially distorted octahedral geometry and is known to form two-

dimensional square networks with pyrazine or substituted pyrazine ligands 71,120-122

and interpenetrated two-dimensional networks with 4,4'-bipyridine.43 However, trigonal

and tetrahedral Cu+ cations can form two-and three-dimensional networks when bridged

by bipy, pyrazine, or substituted pyrazine ligands. 107,120,123,124 Ag+ is observed in an








even wider range of coordination environments including linear,41,110 trigonal,41

tetrahedral,41,108,125 square-planar, 126 square pyramidal,126 and octahedral 126 when

bound to pyrazine or bipy.

The presence of coordinated or lattice solvent molecules can dramatically affect

to final structure. For instance, in [Zn(4,4'-bipy)2(H20)2]SiF6,74 the coordination of

solvent water molecules produces an interpenetrated sheet-like structure while a porous

solid, [Zn(4,4'-bipy)2(H20)2]SiF6-xDMF,117 is isolated under nonaqueous conditions. A

similar solid-state structural difference is observed between the double-layered structure

of the solvent inclusion compound [Ag(pyrazine)2][Ag(pyz)5](PF6)3-2S (S = CH2C12,

CHCI3, and CC14)126 and the single-layered structure of the solvent-free compound

[Ag(pyz)2](PF6).125 Counterions can also influence the final structure. In the previous

example, by replacing the PF6" counterions with SbF6-, the characteristic double-layer

structure of Ag-pyraizine-PF6 changes to the three-dimensional noninterpenetrating cubic

framework Ag(pyz)3(SbF6)126 while substituting with BF4" ions changes the layered

sheet-like structure of [Ag(pyz)2](PF6)125 to the interpenetrated three-dimensional

structure [Ag2(pyz)3](BF4)2.41

Porous Network Supramolecular Architectures

The design and construction of hybrid organic / inorganic porous network solids

is a particularly important and emerging area of supramolecular chemistry providing new

generations of functional materials.42,60,127,128 The great importance of porous solids

in inclusion phenomena, such as adsorption / desorption, ion exchange, size and shape-

selective molecular sieving, and catalysis, is due to their ability to reversibly clathrate or








trap species within their cavities and extended channels. 129 These pores possess a

variety of sizes and shapes not observed in analogous inorganic porous solids such as

zeolites and molecular sieves.87 Furthermore, by a careful selection and design of the

chemical components, the size and clathration properties of these pores can be designed

and fine-tuned to meet specific needs while maintaining the overall structural and

functional features of naturally occurring analogs. 44

Like zeolites, most hybrid organic / inorganic coordination polymers bind guests

within pores, cavities, and channels that are part of their lattice framework. In fact, a

number of such coordination networks have been found to exhibit many other desirable

zeolitic properties as well, such as stability and porosity of the framework,130 guest

exchange, 131,132 and selective catalytic activity.44 Recent examples porous solids

include the diamondoid, honeycomb, rectangular grid, ladder, brick wall, and octahedral

frameworks constructed from tetrahedral, trigonal, and octahedral metal templates

(Zn(II),. Cd(II), Ag(I), and Cu(I)) and various multitopic spacer ligands.44,73,133-135

The reversible absorption and desorption of guests without the collapse of cavities

or channels, while well known in materials such as zeolites, is a much less common

phenomenon in molecular porous solids. 136-138 Upon loss of their guests, clathrated

hosts irreversibly lose crystallinity,93 undergo phase changes,139 or alter their

morphology without the simultaneous replacement of substitutes.139,140 Few examples

exist where cavities that have been collapsed due to crystal packing forces from guest

loss are restored upon guest binding, however, it has recently been shown that some

coordination or hydrogen-bonded networks can rapidly exchange inclusions or








counterions while maintaining crystal integrity.42'44,86,141,142 Furthermore, the

construction and access of large pores in either coordination polymers or hydrogen-

bonded assemblies is often mitigated or precluded by self-inclusion or lattice

interpenetration particularly if the void created by the open pores occupies more than 50

% of the crystal by volume.43'107,109,143 Interpenetrated structures are of limited

usefulness in inclusion chemistry but have significant potential in terms of other

enhanced bulk properties, such as improved chemical, mechanical, and thermal

stability.37'109

In order to design and construct functional porous solids with useful inclusion

capabilities, a number of requirements should be met while accounting for the problems

detailed above. The host framework should be rigid and robust, containing large,

accessible cavities or channels capable of reversible guest binding in a stoichiometric

manner.142 Furthermore, the structural integrity of the pores should be maintained in the

absence ofclathrates.138,142 The term "capable", in this context, implies that the van

der Waals surfaces and electrostatic potential surfaces of the host pores and guest should

also be complementary.37'39 The robustness of a channel or cavity is the ability to allow

reversible release and adsorption of guest without the collapse of the host structure.144

In order to address these requirements, one approach focuses on the utilization of

chemical interactions in two or three dimensions to maintain the integrity of the host

structure after removal of the clathrated guests.134 Porous three-dimensional networks

sustained by strong coordinate covalent bonding often retain the vacant cavities and

channels even after removal of the guest molecules without structural changes, such as










II
_M_
Al
MmM
-M- 4 41
I B
C
D M-M-M

F4M~ I I~
M-M--M M-M M
I- M- I I II I



M-M-M-M I I
E
F


II I
-- '-v- ^

H I i M_


LI

I I I I
HI i MM = -II-M





Figure 1-13. Examples of supramolecular architectures from the self-assembly of metal
ions, M, with spacer ligands (lines). A) Linear chains. B) Zig-zag chains.
C) Diamondoid. D) Herringbone sheet. E) Square grid sheet. F) Railroad.
G) Honeycomb. H) Brick wall sheet. I) Ladder.








lattice interpenetration, at ambient temperature.70,'76,'145,146 On the other hand, flexible

pores sustained by weaker hydrogen bonding interactions may change in size in response

to the uptake or loss of guests to ensure that void space is efficiently occupied thus

avoiding self-inclusion.62

Another approach involves the use of interpenetration to produce porous solids

with robust cavities and channels.77,'144 This method may, at first, seem self-defeating

since the interpenetration of neighboring lattice frameworks often completely fills void

space thus preventing the formation of extended cavities and

channels.42,69,73,84,107,108,110 In fact, most strategies aimed at synthesizing porous

solids have been directed toward the inhibition of interpenetrated networks. However, if

the spacer is of sufficient length the self-inclusion may only reduce the size of, but not

completely fill, the pores, leaving small voids for small-molecule inclusion. 144 The

lattice interpenetration can therefore afford rigid, three-dimensional networks with the

desired robust, albeit smaller, channels.

Despite the large amount of research and work effort devoted toward the synthetic

aspect of solid-state supramolecular chemistry in terms of producing functional materials

and elucidating methods for the rational design and fabrication of such materials, studies

of the chemical reactivity of these materials has been lacking.147 In fact most of the

studies have been limited to inclusion properties of clathrated porous solids, such as the

absorption-desorption processes and guest exchange of guest

molecules.45'87,142,144,147 This imbalance of synthetic work compared to chemical

reactivity studies of supramolecular materials is largely due to the fact that, unlike the

well characterized chemical reactivity properties of molecular and ionic species which








are generally soluble in one or more common solvents, most coordination polymers are

insoluble in most common organic and inorganic solvents thus rendering any reactivity

studies difficult at the very least. 147

Scope of the Dissertation

The work presented in this dissertation, Chapters 2, 3, and 4, focuses on, in

general, investigating the structural and physical properties of representative solid-state

materials obtained throughout the course of this graduate research. In Chapter 2, the

structural, thermal, and magnetic properties of a series of clathrated porous network

solids are described in detail. Chapter 3 describes the host-guest properties of these

porous solids determined through gas chromatography, solid-state NMR spectroscopy,

and X-ray diffraction. Chapter 4 details the structural and magnetic properties of a series

of ladder-like azido bridged Cu(II) coordination polymers. Finally, the structures of

selected coordination polymers that did not quite fit within the context of the themes of

the chapters are presented in Appendix A.














CHAPTER 2
STRUCTURAL, THERMAL, AND MAGNETIC PROPERTIES INVESTIGATION OF
THREE TRANSITION METAL-4,4'-BIPYRIDINE COORDINATION POLYMERS:
[Ni(4,4'-bipy)3(H20)2](C104)2-1.4(4,4'-bipy)-3(H20),
[Co(4,4'-bipy)3(H20)2](C104)2 1.4(4,4'-bipy)-3(H20), and
[Cu(4,4'-bipy)3(DMSO)2](C104)2-2(4,4'-bipy)

Introduction

The rational design and synthesis of functional organic/inorganic network solids

has recently been the focus of intense research in materials chemistry. Much of the

interest in this field is driven by the wide variety of potential applications such materials

can afford, including host-guest chemistry,44 ion exchange,37 molecular

sieving,37,44,45 catalysis,41-44 non-linear optics,54-57 electrical conductivity,50-52 and

magnetism.40,46-49 Often such materials are designed and built from simple, modular

components such as a metal ions and organic spacer ligands. The spacers are typically

multifunctional ligands capable of coordinating to the metal ions and propagating

structural information dictated by the coordination requirements and geometry of the

metal sites throughout a solid.69 The assembly of these components through

intermolecular forces e.g. donor-acceptor interactions, 148,149 hydrogen bonds,61,63,64

and ir-stacking66 can produce solids whose molecular and crystal structure can be

profoundly affected by a number of factors, most notably the metal:spacer stoichiometry.

For example, transition metal ions combined with linear organic spacers can assemble to

form structures resembling linear 64,69,95,96,100,101 or zigzag chains, 69,78,99








molecular ladders,70,88,94,109 two-dimensional

grids,40,44,47,40,69,74,92,100,104,105,150,151 railroads,1 14,115 or three-dimensional

networks,152,153 if the metal:spacer stoichiometry is 1:1, 1:1.5, 1:2, 1:2.5, or 1:3,

respectively.

One commonly employed spacer is the bifunctional heterocyclic molecule 4,4'-

bipyridine and a number of coordination polymers with different network architectures in

the solid state have been reported incorporating this

ligand.40,44,69,70,101,105,109,114,151 Examples of one-dimensional chains include

[Co(S04)(H20)3(4,4'-bipy)]-2(H20),96 [Ni(Et-XA)2(4,4'-bipy-)]-0.5(EtOH)-(CHCl3) (Et-

XA = ethylcarbonadithiolate),98 [Co(NCS)2(H20)2(4,4'-bipy)]-(4,4'-bipy),95 and

[Mn(hfac)2(4,4'-bipy)] (hfac = hexafluoroacetylacetonato),97 while [Ni(4,4'-

bipy)2.5(H20)2](CIO4)2 1.5(4,4'-bipy)-2(H20)1 14 and [Co(4,4'-bipy)i .5(N03)2] -Guest

(Guest = MeCN or CHCl3)109 form railroad and ladder-like structures, respectively.

Examples of two-dimensional grids or sheets include [Cd(4,4'-

bipy)2(N03)2]'2(C4H6Br2),44 [Cu(4,4'-bipy)(H20)2(FBF3)2]-(4,4'-bipy),92

[Co(NCS)2(4,4'-bipy)2]-2(CH3CH2)20,95 [M(4,4'-bipy)(ox)] (M = Fe(II), Co(II), Ni(II),

and Zn(II) and ox = oxalato),40 and [M(4,4'-bipy)2(H20)2](C104)2-Guest (M = Cu, Zn,

and Cd and guest = enclathrated guest molecule).133

Chapter Summary

This chapter reports the crystal structures, thermal behavior, and magnetic

properties of a series of three linear chain compounds containing the 4,4'-bipyridine

spacer that organize in the solid state to form new non-interpenetrated network solids.








The compounds [Ni(4,4'-bipy)3(H20)2](C104)2 1.4(4,4'-bipy)-3(H20) 1,

[Co(4,4'-bipy)3(H20)2](C104)2"1.4(4,4'-bipy)-3(H20) 2, and [Cu(4,4'-

bipy)3(DMSO)2](C104)2-2(4,4'-bipy) 3, were each prepared by the direct combination of

three moles of 4,4'-bipyridne with one mole of their respective metal ion. Compounds 1

and 2 are isostructural and were crystallized under inert atmosphere, hydrothermal

conditions. The related structure 3 was isolated under ambient laboratory conditions

from dimethyl sulfoxide. All three materials share a common structural motif with one-

dimensional, covalently linked chains interacting via hydrogen bonding and ;,r-stacking

forces to form layered sheets with characteristic hydrophobic, rectangular cavities. These

sheets are packed in a manner that aligns the cavities to form oblique channels occupied

by enclathrated guest molecules and counterions that extend throughout the solid. The

thermal instability of these coordination polymers is associated with the relatively low

temperatures at which the guest molecules are lost. Temperature and field dependent

magnetization measurements revealed weak magnetic coupling between the paramagnetic

metal centers as 4,4'-bipy is a poor mediator of superexchange interactions.97,98,154

Experimental Section

Materials

Copper(II) perchlorate hexahyrdate (98 %), cobalt(II) perchlorate hexahydrate

(98 %), nickel(II) perchlorate hexahydrate (98 %), 4,4'-bipyridine (98 %), and sodium

azide were purchased from Aldrich (Milwaukee, WI). Dimethyl Sulfoxide (99.9 %) was

purchased from Fisher Scientific (Pittsburgh, PA). Ethanol (100 %) was purchased from

Aaper Chemical (Shelbyville, KY). All reagents were used without further purification.








Synthesis of [Ni(4,4'-bipy)3(H20)2](CIO4h'l.4(4,4'-bipy).3(H20)
A solution was prepared by dissolving 731 mg ofNi(C104)2-6 H20 (2.0 x 10"3

mol) in 10 mL of water contained within a Teflon canister. Addition of 934 mg of 4,4'-

bipyridine (6.0 x 10-3 mol) and 2 mL of ethanol to this solution resulted in a blue-green

colored suspension. This canister was sealed within a homemade, stainless steel

hydrothermal vessel, purged with argon gas, and heated to 150 C for 5 days. The

container was subsequently cooled to room temperature over a period of 24 hours without

any specific control over the rate of cooling. The resulting blue-green crystals, obtained

in 88 % yield (based on initial quantity of bipy), were washed with water before further

characterization. The crystals become opaque within a few hours upon removal from the

solvent. Elemental analysis calculated for NiC45H46N9O13CI2: C, 51.44 %; H, 4.42 %; N,

12.00 %. Found: C, 50.88 %; H, 4.47 %; N, 12.14 %.

Synthesis of [Co(4,4'-bipy)3(H20)21(C104)2"1.4(4,4'-bipy)-3(H20)

Using the same procedure described for 1, 731 mg ofCo(C104)2-6 H20 (2.0 x 10-3

mol) was reacted with 934 mg of 4,4'-bipyridine (6.0 x 10-3 mol) in 10 mL of water with

2 mL of ethanol at 150 C for 3 days. The resulting orange crystals, obtained in 92 %

yield (based on initial quantity of bipy), were washed with water before further

characterization. The crystals become opaque within a few hours upon removal from the

solvent. Elemental analysis calculated for CoC45H46N9O13Cl2: C, 51.43 %; H, 4.42 %; N,

11.99 %. Found: C, 50.87 %; H, 4.47 %; N, 12.14 %.

Synthesis of [Cu(4,4'-bipy)3(DMSO)2](C104)2-2(4,4'-bipy)

A solution containing 741 mg of Cu(C104)2-6 H20 (2.0 x 10-3 mol) dissolved in

10 mL of DMSO was combined with a solution containing 934 mg of 4,4'-bipyridine (6.0

x 10-3 mol) dissolved in 10 mL of DMSO. The resulting dark blue colored mixture,








contained within an evaporating dish, initially produced small blue block-like crystals

within two weeks of solvent evaporation. Within and additional four weeks, light blue

colored hexagonal plates of 3 were isolated and washed with DMSO before further

characterization.

X-ray Structure Determination

A blue-green crystal ofl 1 (0.25 x 0.23 x 0.23 mm3), an orange crystal of 2 (0.51 x

0.36 x 0.17 mm3), and a blue crystal of 3 (0.24 x 0.21 x 0.12 mm3) were selected for X-

ray analysis. Each crystal was mounted on a glass fiber under nitrogen gas. The same

data collection procedure was used for each sample. Data were collected at 173 K on a

Siemens SMART PLATFORM equipped with a CCD area detector and a graphite

monochromator utilizing MoKx radiation (X = 0.71073 A). Cell parameters were refined

using up to 8192 reflections. A full sphere of data (1850 frames) was collected using the

co-scan method (0.3 frame width). The first 50 frames were re-measured at the end of

data collection to monitor instrument and crystal stability (maximum correction on I was

< 1%). Absorption corrections by integration were applied based on measured indexed

crystal faces.

The structures were solved by the Direct Methods in SHELXTL6 155 and refined

using full-matrix least squares. Structures 1 and 2 were solved and refined in space group

C2/c while 3 was solved and refined in Cc, which afforded better results. The

stoichiometry is the same in 1 and 2 ([M(4,4'-bipy)3(H20)2](C104)2-1.4(4,4'-

bipy)-3(H20) where M = Ni or Co) but different from 3 ([Cu(4,4'-

bipy)3(DMSO)2](C104)2-2(4,4'-bipy)). The non-H atoms were treated anisotropically,

whereas the hydrogen atoms were calculated in ideal positions by riding on their








respective carbon atoms. The H atoms from the coordinated water molecules were found

and refined. The uncoordinated 4,4'-bipyridine molecules and perchlorate anions are

disordered in each structure. In 1 and 2, two half-bipy guest moieties are each disordered

about a center of inversion. The perchlorate anions are disordered in four parts but the

oxygen atoms on only two major components were found in Difference Fourrier maps

and refined anisotropically. Only the Cl atoms of the minor disordered parts were found

and refined. In 3, a single guest 4,4'-bipy has one pyridyl ring disordered. The S atom of

one coordinated DMSO molecule is disordered as well. The site occupation factors of

the disordered parts were dependently refined to 0.88(1) for the major part and

consequently 0.12(1) for the minor part; S' was refined with an isotropic thermal

parameter. For both perchlorate anions, disorder was found each in two positions and

their site occupation factors were dependently refined to 0.69(1) and 0.31(1) for one

anion, and 0.50(1) for each part of the second disordered anions. A total of 428 and 427

parameters were refined employing F2 in the final cycle using 4328 and 4393 reflections

with I > 2a(I) to yield R, of 7.72 % and 7.79 % and wR2 of 21.58 % and 22.38 % for 1

and 2, respectively. A total of 755 parameters were refined employing F2 in the final

cycle using 10055 reflections with I > 2o(I) to yield Ri and wR2 of 5.62 % and 12.61%,

respectively, for 3.

Thermal Analysis

Thermogravimetric analyses (TGA) of the title compounds were performed on a

computer-controlled Hi Res TGA 2950 thermogravimetric analyzer. Powdered samples

of 1 (4.5740 mg), 2 (3.7460 mg), and 3 (6.3730 mg) were loaded into alumina pans and

heated with a ramp rate of 10 'C/min from room temperature to 600 C. Thermal








desorption mass spectrometry measurements were recorded on a MAT 95 utilizing

electron ionization techniques. Crystalline samples, contained within capillary tubes,

were evacuated, loaded into direct insertion probes, and heated with a ramp rate of

10 C / min from 30 C to 400 C.

Magnetic Measurements

Bulk magnetization measurements were obtained from a standard Quantum

Design MPMS SQUID magnetometer. The samples consisted of randomly oriented

single crystals with a total mass of 32.3 mg for 1, 19.4 mg for 2, and 51.0 mg for 3. A gel

cap and plastic straw were used as the sample holder during the measurements.

Magnetization versus temperature measurements were run from 2 K to 300 K. The

sample was zero-field cooled to 2 K before a measuring field of 1000 G was applied and

the data set was then taken while warming the sample from the lowest temperature.

Magnetization versus field measurements were performed at 2 K from 0 to 50 kG. The

background signals arising from the gel cap and straw were measured independently and

subtracted from the results. The diamagnetic contribution of each sample, estimated from

Pascal's constants (XD = 446 x 10-6 emu mol-1 for 1 and 2 and XD = -516 x 10-6 emu

mol"1 for 3) was also subtracted from the results.5'6

ESR spectra were recorded with a Bruker ER 200D spectrometer modified with a

digital signal channel and digital field controller. Data were collected using a U. S. EPR

SPEC300 data acquisition program and converted to ASCII format using a U. S. EPR

EPRDAP data analysis program.








Results and Discussion

Compound Synthesis

The network coordination polymers 1, 2, and 3 were synthesized by the direct

combination of one mole of M(C104)2-6H20 (M = Ni(II), Co(II), or Cu(II)) with three

moles of 4,4'-bipyridine in solution. Note that the products contain more than three

equivalents of bipy, some present as enclathrated guest molecules in addition to the

coordinated ligands.

Since the bipy ligand is relatively insoluble in water, the reaction conditions

afforded by the hydrothermal technique are essential for the crystallization of 1 and 2

where the metal-bipy suspension is dissolved by the high temperature (150 C) and

pressure conditions within the vessel. In order to prevent the formation of unwanted side

products, such as high oxidation state metal oxides particularly with cobalt, the reaction

mixture was heated for no longer than 5 days, purged with argon gas, and treated with

small quantities of ethanol (acting as a mild reducing agent). The products are sensitive

to loss of solvent and become opaque within a few hours upon exposure to air. For 1 and

2, small changes in the metal-bipy stoichiometry (for example, by using 2.5 or 3.5 moles

ofbipy per mole of Ni(II) or Co(II)) always produced the same products.

Single crystals of 3 were obtained under normal laboratory conditions by

crystallization from DMSO. Similar attempts employing hydrothermal synthesis resulted

in the formation of impure powders. Crystalline products of 3 were obtained only several

weeks due to the slow rate of evaporation of DMSO. Unlike 1 or 2, crystals of 3 do not

appear to be sensitive to loss of solvent. Furthermore, a secondary product crystallizes

from solution in addition to 3. During the first two to three weeks of crystallization,

small blue blocks first appear followed by the crystallization of blue hexagonal plates of








3 after an additional month or more. These blue blocks were determined to be an

extended three-dimensional Cu(II)-4,4'-bipyridine network of the molecular formula

[Cu2(4,4'-bipy)5(DMSO)3(C104)](C104)3-3(DMSO)-(H20) with hydrophobic, rectangular,

enclathrated channels that extend throughout the solid. 156 Reducing the concentration of

bipy (2.5 moles ofbipy per mole of copper), favors the formation of the three-

dimensional network while at higher concentrations (3.0 moles and 3.5 moles of bipy per

mole of copper), relatively equal quantities of 3 and the 3-D network crystallize from

solution.

Description of the Structures

Crystallographic and structural refinement data for 1, 2, and 3 are listed in Table

2-1. Selected bond angles and distances for 1,2, and 3 are given in Tables 2-2, 2-3, and

2-4, respectively. Tables of atomic coordinates and thermal displacement parameters are

provided in Appendix B.

Structure of [Ni(4,4'-bipy)3(H20)2](CI04)2"1.4(4,4'-bipy)-3(H20)

The structure of 1 consists of one-dimensional cationic [Ni(4,4'-bipy)3(H20)2]2

chains that pack to form layered sheets in the solid. The local coordination environment

surrounding a typical Ni(II) ion is shown in Figure 2-1. The metal is six-coordinate and

the coordination sphere consists of four pyridyl nitrogen donors, one from each of four

4,4'-bipyridine ligands and two oxygen atoms from two aqua ligands. The NiN402 unit

locally adopts an axially compressed octahedral geometry. The four nitrogen atoms

define the equatorial plane and the oxygen atoms occupy the axial sites. Both Ni--O

bond distances are equal (2.06 A) but shorter than the Ni-N bonds. Furthermore, both

of the Ni-N bonds from the terminal bipy ligands are equal (Ni-N21, and Ni-N21A =








Table 2-1. Summary of Crystallographic Data for 1, 2, and 3
Empirical Formula C44H46Cl2N9NiO13 C44H46Cl2N9CoO13
Formula Weight 1038.51 1038.73
Space Group Monoclinic, C2/c Monoclinic, C2/c
a, A 17.5696(8) 17.614(2)
b,A 11.4101(5) 11.514(1)
c, A 24.479(1) 24.604(2)
a, deg 90 90
f8, deg 93.065(1) 92.448(2)
y, deg 90 90
V, A3 4900.4(4) 4985.6(9)


C5oH52C12NIoCuOIoS2
1199.62
Monoclinic, Cc
19.0931(9)
11.1949(5)
25.607(1)
90
94.810(1)
90
5454.0(4)
4


T,K
2(Mo KI), A
Pcalc, g cm-3
A cm-'
Ra (Rwb)
a R = IlFol IFc||/l|Fol


173(2) 173(2)
0.71073 0.71073
1.408 1.384
5.76 5.21
0.0709 (0.2069) 0.0736 (0.2173)
bR = X[(Fol- IFcI)w]i/E[IFolw ]


173(2)
0.71073
1.461
6.44
0.0466 (0.1179)


Table 2-2. Selected Bond Lengths [A] and Angles [] for 1a
Ni-01 2.060 01-Ni--OlA 179.82
Ni-OlA 2.060 N 1-Ni-NI lA 180.000
Ni-N 11l 2.127 N21-Ni-N21A 177.20
Ni-N11A 2.184 NIl-Ni-N21 88.60
Ni-N21 2.115 N11I-Ni--01 89.91
Ni-N21A 2.115 N 11A-Ni-N21 91.40
N 11A-Ni--O 1 90.09
N21-Ni--01O 92.44
SSymmetry transformations used to generate equivalent atoms: #1, -x, y, -z+3/2; #2, x,
y+l, z; #3, x, y-l, z; #4, -x-1/2,-y+l/2,-z+2; #5,-x,-y+2,-z+2.


Table 2-3. Selected Bond Lengths [A] and Angles [] for 2'
Co-01 2.061 01-Co-OlA
Co-OlA 2.061 N1--Co-Ni lA
Co--N 11l 2.180 N21-Co--N21A
Co--NIlA 2.232 N11--Co-N21
Co-N21 2.168 N11---Co--O 1
Co-N21A 2.168 Nl1 A-Co-N21
NIlA-Co--Ol1
N21-Co-0- 1


178.68
180.000
176.37
88.19
90.66
91.81
89.34
92.17








Table 2-4. Selected Bond Lengths [A] and Angles [0] for 3'
Cu-Ol 2.396 NI-Cu-Nl' 178.39
Cu-02 2.376 N2-Cu-N3 179.25
Cu-N 1 2.049 01--Cu--02 178.17
Cu-N2 2.025 Nl--Cu-N2 91.09
Cu-N3 2.033 N1l-Cu-N3 89.03
Cu-Nl' 2.060 Nl'-Cu-N2 90.51
S(i )-O 1 1.507 Nl'-Cu-N3 89.36
S(2)-02 1.488 N1--Cu--OI 90.69
S(2')-02 1.347 N2-Cu-Ol 91.49
N3-Cu-Ol 87.78
N1 '-Cu--O1 89.13
N1--Cu--02 90.01
N2-Cu--02 86.81
N3-Cu--02 93.93
N1'-Cu--02 90.21
S1--O(1)--Cu 141.6
S2--O(2)-Cu 145.9
S2'-0O(2)-Cu 154.1


2.12 A) but one of the Ni-N bonds (Ni-N 11A = 2.18 A) from a bridging bipy is longer

than the other such bond (Ni-N 11, 2.13 A).

According to the spectrochemical series, water is a weaker field ligand than bipy

and thus Ni--O bond distances should be longer than Ni-N bonds. The aqua ligands

are expected to be the sites of an axial elongation. However, from the structural data, the

opposite effect is observed. Steric repulsion from the pyridyl rings coordinated to the

metal centers could cause a rather significant lengthening of Ni-N bonds. Note also that

the hydrogen atoms from the coordinated water molecules hydrogen bond to the terminal

nitrogen atoms from monodentate bipy's on adjacent chains thus polarizing the H--O

bonds to a greater extent. The oxygen atom effectively becomes more negatively

charged, and as a consequence, is more attracted to the positively charged metal center

resulting in a smaller than expected Ni--O bond distance.








Two of the bipy ligands, coordinated trans with respect to one another, bridge the

Ni(II) ions to form infinite one-dimensional linear chains that extend along the

crystallographic c-axis. A single chain is depicted in Figure 2-2. The Ni-Ni distance

along a chain is a/2 units (11.41 A). For each bridging bipy ligand, the pyridyl rings are

not coplanar but twisted along the central C-C bond at an angle of 29.7 with respect to

each other. Ignoring the uncoordinated guests, the metal:bipy stoichiometry in 1 is 1:3

since the bipy ligands perpendicular to the chains are monocoordinate.

The Ni-bipy chains are juxtaposed in a side-by-side fashion to form two-

dimensional sheets within the crystallographic bc-plane. A typical sheet is shown in

Figure 2-3. Within each sheet, the chain spacing is b/2 units. In addition to packing

forces, the sheets are held together by a combination of hydrogen bonding interactions

between the protons of the coordinated water molecules and the terminal nitrogen atoms

from the monodentate bipy ligands on adjacent chains (N-H bond distance of 1.81 A)

and offset ;r-stacking between the monocoordinate bipy ligands on adjacent chains. The

face-to-face distance between these overlapping bipy groups is 3.7 A. Note that in

order to maximize the favorable ir-stacking interactions, no twisting of the pyridyl rings

is observed along the central C-C bond for these monodentate bipy ligands. The

characteristic packing motif of the Ni-bipy chains produces rectangular, hydrophobic

cavities within the sheets. Each cavity is defined by four nickel ions at the comers and

along the sides by the faces of the two bridging bipy's and the edges of the two pairs of

r-stacked terminal bipy ligands. The dimensions of the cavities are b/2 x c and, if the

Van der Waals radii of the carbon atoms from the bipy's are approximated as 1.7 A, the

effective size of the cavities is approximately 9.7 A x 10.7 A.








As shown in Figures 2-4 and 2-5, the sheets pack to form a layered solid-state

structure along the crystallographic c-axis. Although the sheets align in registry along the

b-axis, they are offset by V step in both the a- and c directions. The characteristic

packing results in an alignment of the hydrophobic cavities to form oblique channels

extending along the [2 0 -2] direction, as shown in Figure 2-6. The void space within

hydrophobic pores and between the sheets is not empty but occupied by clathrated guest

molecules and counterions acting to prevent the interpenetration of adjacent sheets.

The pores within the framework host are not empty, but occupied by enclathrated

guest molecules and counterions (Figures 2-3 to 2-6). These lattice guests are extensively

disordered throughout the solid thus leading to the rather high final refinement value in

the structural solution. Approximately 1.5 crystallographically inequivalent,

uncoordinated bipy molecules, each disordered about centers of inversion, are present per

asymmetric structural unit. Extensive hydrogen bonding interactions between the lattice

waters and the coordinated waters, perchlorate ions, and bipy guests are present

throughout the channels. The clathrated bipy molecules seem to form a secondary lattice

of organic molecules interpenetrated within the porous network structure of 1.

Coordination polymers with both guest enclathrated and empty cavities and channels

present throughout the solid are known. 40,44,47,69,70,95,109,114

Each rectangular cavity in 1 clathrates an uncoordinated 4,4'-bipy molecule

stabilized by both weak hydrogen bonding and hydrophobic interactions. These guests

are crystallographically centrosymmetric and disordered about a center of inversion. The

guest is positioned approximately at the center of the cavity such that the edges of one

pair of borders (the pair of if-stacked monocoordinate bipy ligands) is directed toward








the faces of the guest and the faces of the other pair of borders (the bipy bridges) is

directed toward the edges of the guest with the edge-to-face distances of ca. 2.5 A 3.0

A. Note that the pyridyl rings of this bipy molecule are coplanar. The interior hydrogen

atoms from the bipy form both two- and three-center hydrogen bonds with one (2.39 A)

and two oxygen atoms (2.29 A and 2.53 A), respectively, from nearby perchlorate

counterions.

Uncoordinated, disordered bipy guests are clathrated between the sheets of 1 and

are stabilized by ir-stacking and hydrogen bonding interactions as well. Each face

directed toward the space between the sheets from the pair of monocoordinate bipy

ligands that comprise part of the borders of the hydrophobic cavities interacts with a

single such bipy guest. These guests stack in an offset parallel fashion with the pair of

ir-stacked monocoordinate bipy ligands that comprise part of the borders of the cavities

with a face-to-face distance of ca. 3.3 A, indicative of if-stacking interactions. Note that

these bipy guests are oriented almost perpendicular with respect to the bipy's clathrated

within the cavities. The terminal nitrogen atoms from these bipy's form hydrogen bonds

(2.48 A) with oxygen atom from nearby perchlorate counterions.

The lattice water molecules are positioned in the vicinity of the bridging bipy

ligands and the perchlorate counterions are in close proximity to the bipy guests located

between the sheets. One of the lattice water molecules simultaneously hydrogen bonds

with a proton from the coordinated water molecule (H--O, 1.99 A), the terminal nitrogen

atom from the bipy molecules between the sheets (N-O, 2.64 2.82 A), the oxygen

atom of a nearby perchlorate counterion (0-0, 2.82 A), and another nearby lattice water








molecule (0-0,2.84 A). This second lattice water molecule also hydrogen bonds with

an oxygen atom from a neighboring perchlorate counterion (2.07 A).

The structure of 1 bears close resemblance to a compound reported by Yaghi and

coworkers, [Ni(4,4'-bipy)2.5(H20)2](C104)2 1.5(4,4'-bipy)-2(H20). 114 Though this

material is a covalent molecular railroad as opposed discrete 1-D chains, both structures

incorporate both bridging and terminal bipy ligands as well as 4,4'-bipy enclathrated

channels. The Ni-N(bridging) distances (2.13 A 2.19 A) in 1 are comparable to those

found in the railroad structure (2.11 A) but are larger than those found in the cis-chain

[Ni(4,4'-bipy)(Et-XA)2]-0.5(EtOH)-(CHCl3) (2.07 A 2.09 A, Et-XA =

ethylcarbonadithiolate)98 and the 2-D covalent grid [Ni(4,4'-bipy)(ox)] (2.09 A, ox =

oxalato).40 The Ni-N(terminal) distances (2.115 A) of 1 are also similar to those found

in [Ni(4,4'-bipy)2.5(H20)2](CI04)2-1.5(4,4'-bipy)-2(H20) (2.15 A). Furthermore, Ni--O

distances (2.06 A) are also close to those in the Yaghi structure (2.08 A) and in [Ni(4,4'-

bipy)(ox)] (Ni-O(oxalato) bonds) (2.05 A). All O-Ni-O, N(bridging)-Ni-

N(bridging), and N(tennrminal)-Ni-N(terminal) angles are 180 and 0--Ni-N and

N(bridging)-Ni-N(terminal) angles are close to 90.

Structure of [Co(4,4'-bipy)3(H20)2](C104)2.1.4(4,4"-bipy).3(H20)

Other than substituting the metal center Co(II) for Ni(II) and small differences in

characteristic bond angles and distances, the structure of 2 is virtually identical to 1 and

thus no pictures or description is given. The structure of 2 resembles the trans-l-D chain

compounds reported by Jacobson and coworkers, [Co(4,4'-bipy)(S04)(H20)2]-2(H20)

and [Co(4,4'-bipy)(Cl)2(DMSO)2] where the solvent molecules and counterions, not

terminal bipy ligands, occupy the non-bridging coordination sites on the metal centers.96




































'6 '17 N21C




N11C






Figure 2-1. The local coordination environment of a typical Ni(II) metal center in
compound 1. All aromatic hydrogen atoms have been omitted for clarity. All non-
hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of
electron density.































Figure 2-2. The structure of compound 1 along the crystallographic c-axis representing a
linear, one-dimensional Ni-bipy chain. All aromatic hydrogen atoms have been omitted
for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to
encompass 30 % of electron density.
































b

C


Figure 2-3. The structure of compound 1 within the crystallographic bc-plane
representing a two-dimensional sheet (left) with the accompanying guests and
counterions (right). For clarity, the aromatic hydrogen atoms and guests in the left-hand
figure have been omitted and the coordinated guests have been omitted in the right-hand
figure.


































a


C

Figure 2-4. The structure of compound 1 within the crystallographic ac-plane showing
the layered two-dimensional sheets (left) with the accompanying guests and counterions
(right). For clarity, the aromatic hydrogen atoms and guests in the left-hand figure have
been omitted and the coordinated bipy ligands and have been omitted in the right-hand
figure.











An


^__-K p-^ -i -_ I^ ^



V "/ .- ^ -.i

ti- / I pt


U



----a


Figure 2-5. The structure of compound 1 within the crystallographic ab-plane showing
the layered two-dimensional sheets (left) with the accompanying guests and counterions
(right). For clarity, the aromatic hydrogen atoms and guestsin the left-hand figure have
been omitted and the coordinated bipy ligands have been omitted in the right-hand figure.






tl! ,s '" H ::' t "- ---"-)-7*'**

*^ ^ 7 -, .1 --- ,- W ;.

./- .l~ .. :,, a b
V.1^..^%^^ Il





Figure 2-6. The hydrophobic channels in compound 1 that extend along the [2 0 -2]
direction with the accompanying guests and counterions. For clarity, the aromatic
hydrogen atoms have been omitted.








The Co-N(bridging) distances (2.18 2.23 A) in 2 are comparable to those in the chains

[Co(S04)(H20)3(4,4'-bipy)]-2(H20) (- 2.17 A)96 as well as in the 2-D covalent grid

[Co(4,4'-bipy)(ox)] (2.15 A, ox = oxalato).40 All Co-N distances in 2 are larger than

the corresponding Ni-N distances in 1. The Co--O distances in 2 (2.06 A) are also

comparable with those in [Co(4,4'-bipy)(S04)(H20)2].2(H20) (2.08 A) and [Co(4,4'-

bipy)(ox)] (2.08 A).

Structure of [Cu(4,4'-bipy)3(DMSO)21(CI04)2-2(4,4'-bipy)

The structure of 3 consists of one-dimensional cationic [Cu(4,4'-

bipy)3(DMSO)2]2+ chains that pack to form layered sheets in the solid. Although 3 is

structurally related to 1 and 2, some important differences are present. The local

coordination environment surrounding a typical Cu(II) ion is shown in Figure 2-7. The

metal is six-coordinate and the coordination sphere consists of four pyridyl nitrogen

donors, one from each of four 4,4'-bipyridine ligands and two oxygen atoms from ligated

DMSO molecules. One DMSO ligand per metal center is disordered about the sulfur

atom. The CuN402 unit locally adopts an axially elongated octahedral geometry. The

four nitrogen atoms define the equatorial plane and the oxygen atoms occupy the axial

sites. Both Cu--O bond distances are equal (2.40 A) and longer than the Cu-N bonds.

Furthermore, the Cu-N bonds from the terminal bipy ligands (2.02 2.03 A) are shorter

than the Cu-N bonds from the bridging bipy ligands (2.05 2.06 A).

Two of the bipy ligands, coordinated trans with respect to one another, bridge the

Cu(II) ions to form infinite one-dimensional linear chains that extend along the

crystallographic c-axis. A single chain is depicted in Figure 2-8. The Cu-Cu distance

along a chain is approximately a/2 units (11.2 A). For each bridging bipy ligand, the








pyridyl rings are not coplanar but twisted along the central C-C bond at an angle of

61.4 with respect to each other, considerably larger than the similar such dihedral angles

observed in 1 and 2. Ignoring the uncoordinated guests, the metal:bipy stoichiometry in 1

is 1:3 since the bipy ligands perpendicular to the chains are monocoordinate.

The Cu-bipy chains are juxtaposed in a side-by-side fashion to form quasi-two-

dimensional sheets within the crystallographic bc-plane. A typical sheet is shown in

Figure 2-9. Within each sheet, the chain spacing is approximately b/2 units. Weak N-H

contacts (2.6 A) between the terminal nitrogen atoms from the bipy ligands with the

nearby hydrogen atoms from the bridging bipy ligands on adjacent chains and S-H

contacts (2.9 A) between the hydrogen atoms near the terminal nitrogen atoms from the

monocoordinate bipy ligands with the neighboring sulfur atoms from DMSO ligands on

adjacent chains are present. Additionally, offset n stacking are observed between the

monocoordinate bipy ligands on adjacent chains. The face-to-face distance between

these overlapping bipy groups is 3.8 A. The pyridyl rings on the monocoordinate bipy

ligands are twisted 11.9 0 along the central C-C bond with respect to one another thus

likely reducing the effectiveness of the stabilizing ;r interactions between the bipy pairs.

The characteristic packing motif of the Cu-bipy chains produces rectangular,

hydrophobic cavities within the sheets. Each cavity is defined by four copper ions at the

comers and along the sides by the faces of the two bridging bipy's and the edges of the

two pairs of 7r-stacked terminal bipy ligands. The dimensions of the cavities are b/2 x c

and, if the Van der Waals radii of the carbon atoms from the bipy's are approximated as

1.7 A, the effective size of the cavities is approximately 9.5 A x 11.3 A, slightly larger

than the cavities present in 1 and 2.








As shown in Figures 2-10 and 2-11, the sheets pack to form a layered solid-state

structure along the crystallographic c-axis. Although the sheets align in registry along the

b-axis, they are offset by '/2 step in both a and c. The characteristic packing results in an

alignment of the hydrophobic cavities to form oblique channels extending along the

[2 0 -2] direction, as shown in Figure 2-12, just as in 1 and 2. Note the hydrophobic

interactions between the nearest neighbor methyl groups from DMSO ligands between

the sheets.

The pores within the framework host are not empty, but occupied by enclathrated

guest molecules and counterions as shown in Figures 2-9 to 2-12. The lattice guests are

relatively less disordered in 3 compared to 1 and 2 thus resulting in a better structural

refinement. Two crystallographically inequivalent, uncoordinated bipy molecules are

present per asymmetric structural unit. Hydrogen bonding interactions between the

perchlorate ions, bipy guests, and the host are observed but weaker and less prevalent

than those similar interactions present in 1 and 2. Again, the clathrated bipy molecules

seem to form a secondary lattice of organic molecules interpenetrated within the porous

network structure of 3.

Each rectangular cavity of 3 clathrates an uncoordinated 4,4'-bipy molecule

stabilized by hydrophobic interactions with the host framework. One pyridyl ring per

guest is disordered about the central C-C bond. Unlike in 1 and 2, the guest is not

positioned at the center of the cavity. The edges of one pair of borders (the pair of

ir-stacked monocoordinate bipy ligands) is directed toward the faces of the guest and the

faces of the other pair of borders (the bipy bridges) is directed toward the edges of the

guest with the edge-to-face distances of ca. 2.9 A 3.8 A. Note that the pyridyl rings of








this bipy molecule are slightly twisted 12.6 with respect to each other along the central

C-C bond.

Uncoordinated, bipy guests are clathrated between the sheets of 3 and are

stabilized by iT-stacking and hydrogen bonding interactions as well. Unlike in 1 and 2,

these guests are not disordered. Each face directed toward the space between the sheets

from the pair of monocoordinate bipy ligands that comprise part of the borders of the

hydrophobic cavities interacts with a single such bipy guest. These guests stack in an

offset parallel fashion with the pair of ir-stacked monocoordinate bipy ligands that

comprise part of the borders of the cavities with a face-to-face distance of ca. 3.7 A,

indicative of ir-stacking interactions. Note that these bipy guests are oriented almost

perpendicular with respect to the bipy's clathrated within the cavities. The terminal

nitrogen atoms from these bipy's form weak N-H contacts (2.6 A) with hydrogen atoms

from nearby bridging bipy ligands that are part of the framework.

The perchlorate counterions are in close proximity to the bipy guests located

between the sheets. The perchlorate oxygen atoms are observed to weakly interact with

hydrogen atoms from the bridging bipy ligands (0-H contacts of 2.4 A), both interior

hydrogen atoms from the monocoordinate bipy ligands (0-H contacts of 2.4 2.6 A),

and the hydrogen atoms from DMSO ligands (0-H contacts of 2.5 A).

Unlike in 1 or 2, the Cu-0-O bonds are longer that the Cu-N bonds in 3. The

DMSO molecules are weaker field ligands compared to the bipy ligands and, in

combination with the Jahn-Teller effect, the solvent molecules are located on the axially

distorted coordination sites. The Cu-N(bridging) distances (2.03 2.06 A) are

comparable to those in the 1-D chains [Cu(4,4'-bipy)(S04)(H20)3]-2(H20) (2.05








A) 69 but smaller than the corresponding distances (1.99 A) in [Cu(4,4'-

bipy)(FBF3)2(H20)2]-(4,4'-bipy). 92 All Cu-N distances in 3 are smaller than the

corresponding Ni-N and Co-N distances in 1 and 2, respectively. The Cu-0-O

distances (2.06 A) are comparable with the those in [Cu(4,4'-bipy)(S04)(H20)3]-2(H20)

(1.95 A 2.2 A). All 0-Cu-O, N(bridging)-Cu-N(bridging), and N(terminal)-

Cu-N(terminal) angles are close to 180 but 0--Cu-N and N(bridging)-Cu-

N(terminal) angles deviate from 90, more so than in 1 and 2, where N(bridging) and

N(terminal) denote the nitrogen atoms from bridging and terminal bipy ligands,

respectively.

Thermal Properties

Compounds 1 and 2 are unstable in air due to loss of enclathrated guest

molecules. The compounds can be kept in humid environment or under solvent, but both

compounds discolor and change texture if left in laboratory atmosphere within a few

hours. Thermogravimetric analysis (TGA) and thermal desorption mass spectrometry

show the stepwise loss of water followed by uncoordinated bipyridine. From Figure 2-

13, mass loss between room temperature and 56 C corresponds to three moles of water

corresponding to guest water molecules in the chemical formula. Water continues to be

released up to near 150 0C. Above 85 C, bipyridine is lost continuously up to

approximately 250 C. The TGA plot of 2, shown in Figure 2-14 undergoes similar, but

somewhat more complex, guest molecule loss process. Note that the onset temperature

for the guest loss in 1 is significantly lower (by approximately 30 0) compared to 2.

In contrast to 1 and 2, the copper compound 3 is stable in air at room temperature.

Without solvent molecule guests, it does not experience the same decomposition process.









From Figure 2-15, the first mass decrease observed by TGA occurs above 100 C and

corresponds to loss of guest bipyridine molecules.









Ni





32' B~

N1 A





N'A

Figure 2-7. The local coordination environment of a typical Cu(II) metal center in
compound 3. All hydrogen atoms have been omitted for clarity. All non-hydrogen atoms
are represented by thermal ellipsoids drawn to encompass 30 % of electron density.


































Figure 2-8. The structure of compound 3 along the crystallographic c-axis representing a
linear, one-dimensional Cu-bipy chain. All hydrogen atoms have been omitted for
clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to
encompass 30 % of electron density.





























Figure 2-9. Structure of compound 3 within the crystallographic bc-plane
representing a two-dimensional sheet (left) with the accompanying guests and
counterions (right). For clarity, the guests have been omitted in the left-hand figure and
the coordinated bipy ligands have been omitted in the right-hand figure.
conein (right. For clrttegusshv en omttdnh lef-han fiur and




tcoordnatedin (bipy).Fo clandy eguss have been omitted in the righ t-hand figure.an


























.j Il
T{_, .- "* .*y' '-" T ... ^ ^ ."..^ V -

\ ''. ~X.._ -'y_ \\
-- .,,. .. ..- ., "n \ r -:- ^- :'!^ 1



a




Figure 2-10. The structure of compound 3 within the crystallographic ac-plane showing
the layered two-dimensional sheets (left) with the accompanying guests and counterions
(right). For clarity, the guests have been omitted in the left-hand figure and the
coordinated bipy ligands have been omitted in the right-hand figure.



























j~jjb



M--a

Figure 2-11. The structure of compound 3 within the crystallographic ab-plane showing
the layered two-dimensional sheets (left) with the accompanying guests and counterions
(right). For clarity, the guests have been omitted in the left-hand figure and the
coordinated bipy ligands have been omitted in the right-hand figure.

























Figure 2-12. The hydrophobic channels within compound 3 that extend along the
[2 0 -2] direction with the accompanying guests and counterions. For clarity, the
hydrogen atoms have been omitted.
















100


80

i60

I 40

20


0


0 C

4 O


2


0I


I ,--. -2
0 100 200 300 400 500 600


Temperature (*C)
Figure 2-13. TGA thermogram of compound 1 depicting the observed mass loss and
negative values of the first derivative (% / C).


0 100 200 300 400 500 600


4 -
al

2 5


0

-2


Temperature (*C)


Figure 2-14. TGA thermogram of compound 2 depicting the observed mass loss and
negative values of the first derivative (% / 0 C).


















Mass Loss





Derivative of
Mass Loss


6 -
U

4 S





0*
2

________o.


0 100 200 300 400 500 600
0 100 200 300 400 500 600


Temperature (C)


Figure 2-15. TGA thermogram of compound 3 depicting the observed mass loss and
negative values of the first derivative (% / C).








Magnetic Properties

The molar magnetic susceptibility, %M, and inverse molar susceptibility, at 1000 G

over the temperature range of 2 K 300 K for 1 and 3 are plotted in Figures 2-16 and 2-

19, respectively. The molar magnetization, MM, at 2 K over the field range of 0 50 kG

for 1 and 3 are plotted in Figures 2-17 and 2-20, respectively. The molar magnetic

susceptibility at 1000 G over the temperature range of 2 K-50 K for 2 is plotted in Figure

2-18.

Magnetic Properties of [Ni(4,4'-bipy)3(H20)2](CIO4)2-1.4(4,4'-bipy)-3(H20)

The room temperature (i.e. 300 K) susceptibility (XM = 4.2 x 10-3 emu mol'1) of 1

correlates well with the value expected for uncoupled, S = 1 metal centers

(ZM = 4.2 x 10-3 emu mol'1). Recall the structure of 1 consists of linear chains of 4,4'-

bipyridine bridged Ni(II) ions. Since 1 is comprised of chains of Heisenberg S = 1 spins

that experience single-ion, D, and in-plane, E, anisotropies, an S = 1, one-dimensional

chain model incorporating both zero-field splitting and exchange parameters is

appropriate to fit the magnetic data. Only exchange interactions along the chains were

considered; coupling between the chains was ignored since interchain bonding is

noncovalent. The Heisenberg model is an appropriate starting point as octahedral Ni(II)

complexes have nearly isotropic g-factors (g = 2.25).5,157 The Hamiltonian may be

written as

H =-JIS, -S,, + D (S;)2 + EZ[(S)2 -(Sy)2+ gUBB. S9 (2-1)
i i t i

where J is the exchange interaction, D is the single-ion anisotropy, E is the in-plane

anisotropy, and B is the applied magnetic field.158 From this Hamiltonian, the parallel,

%X\, and perpendicular, Xj, susceptibilities (neglecting the E parameter) are








S2NAD 1 J(2-2)
S= z 3k T 3kBT )

_2N2( D+8zj}
2NAg 4 (1 D+8zJ (2-3)
Zl x 3kBT 6kBT

and the susceptibility representing the average of contributions that are parallel and

perpendicular to the chains, is
1
X.e =(XII + 2X) (2-4)


since magnetic measurements were performed on randomly oriented samples.1

A theoretical treatment of the magnetization in the high field limit of the

Hamiltonian given in Equation (2-1) has not yet been reported, however in the limit that

J= 0 and E = 0, expressions for the high field dependence of the magnetization do exist.1

The expressions representing the magnetization perpendicular, ML, and parallel, M\\, to

the chains (neglecting the E and J parameters) are
E, E,
2NABj1 gju -e UkT+e
M = 2, 2B' 2 D Eo E, (2-5)
4gI +D2 e-U + e-U + e Ur

where

Eoi+4g2/'lpB29+
E= 24 + D (2-6)
2

and
E2 E3
NAg t-e +e k} (2-7)
M_ E, E,
l+e kT +e U


where








E2 = g^q! + D and E3 = -g^p,! + D (2-8)

and the magnetization representing the average of contributions that are parallel and

perpendicular to the chains, is

Mae =(Mi +2M1) (2-9)


for powdered samples.

The temperature dependence of the inverse susceptibility for 1 between 5 K and

50 K were fit by a simple Curie-Weiss law model and the results of the fit yield a Weiss

temperature 0= 2 K, and from the Curie constant, g = 2.20. The temperature

dependence of the low field susceptibility was fit using the expressions in Equations 2-2,

2-3, and 2-4, and the results of this analysis, when using g = 2.20, are given by the solid

line in Figure 2-16, where J/kB = 0.9 K, D/kB = 4.7 K, and E = 0. The small, negative

coupling constant suggests the presence of a weak antiferromagnetic exchange along the

chains. Since these measurements were performed on randomly oriented samples, the

sign of D cannot be unambiguously determined.6 The DI/V ratio of- 5 suggests

indicates that 1 is an example of a large DIVJ system with strong planar anisotropy.158-

161 Naturally, the fitting procedure would be significantly improved if the data extended

to sufficiently low temperature so as to reveal the maximum of the susceptibility,

observed at T = 2 K.

The field dependence of the low temperature magnetization in Figure 2-17 was fit

using the expressions in Equations 2-5 to 2-9. In the limit that J= 0 and E = 0, the model

provides a prediction that closely resembles the experimental data when g = 2.20 and

D/kB ; 7 K. The analysis of the low field susceptibility indicates that DIVJ] 5, so the








approximation that J= 0 is clearly not justified. Nevertheless, this analysis does provide

an upper bound for the value of D and is consistent with the low field analysis. For the

purpose of comparison, the S = 1 Brillouin function, which is appropriate for zero

exchange coupling effects, is plotted in Figure 10 as the dotted line. Once again, the data

clearly suggest the presence of finite magnetic coupling in 1.

The delocalized n system of 4,4'-bipyridine should allow the ligand to effectively

mediate superexchange interactions when covalently bridging paramagnetic centers.5

Furthermore, the coupling is expected to be antiferromagnetic as explained by a spin-

polarization mechanism for the propagation of exchange interactions. 154 The magnitude

of the coupling is small (J/kB = 0.9 K), but is consistent with the coupling constants

measured for other similar bipy bridged complexes [Ni(Et-XA)2(4,4'-bipy-)]

*0.5(EtOH)-(CHC13) (Et-XA = ethylcarbonadithiolate),98 [Mn(hfac)2(4,4'-bipy)] (hfac =

hexafluoroacetylacetonato),97 [Cu2(tren)2(4,4'bipy)](BPH4)4,162 [Cu2(dien)2(4,4'-

bipy)(C104)2](C104)2,163 and [Mn-/-(4,4'-bipy)(4,4'-bipy)(NCS)2(H20)2]n.164 Since a

typical sp2-sp2 C-C single bond is 1.50 A and a C=C double bond is 1.35 A, the

carbon-carbon bond between the pyridyl rings (1.49 A) for each coordinated bipy

ligand in 1 is principally of single bond character. The z orbital conjugation between

these pyridyl rings is small and the rings twist along this central C-C single bond in

order to minimize the steric repulsions felt by the interior hydrogen atoms. 164 This

twisting disrupts the exchange along the r-orbital pathway and thus hindering the

effectiveness of bipy to propagate a superexchange interaction between paramagnetic

centers.


















0.2



S0.10



001-


300



200
,--,N



100



0


0 100 200 300
T(K)

Figure 2-16. The molar magnetic susceptibility, ZM, and inverse susceptibility, IZm, at 1
kG from 2 K to 300 K for compound 1 are shown as open boxes and open circles,
respectively. The data have been corrected for background signals arising from the
sample container and diamagnetic contributions. The fit of to the Heisenberg S= 1
chain model from Equations 2-2, 2-3, and 2-4 is shown by the solid line.



12 .5


10.0 -
0o

S7.5

50
s 5 -0 /.^

2 n a/at2K
2 5 -Fit to S = Chain Model
-Fit to S I Bnllouinm Function

00
0.0 ,y .
0 10 20 30 40 50
B (kG)

Figure 2-17. The molar magnetization (MM) at 2 K from 0 to 50 kG for compound 1 is
shown as open boxes. The data have been corrected for background signals arising from
the sample container. The fit ofMto the S = 1 chain model from Equations 2-5 to 2-9
and the S = 1 Brillouin function is shown by the solid line and dotted line, respectively.


X x.0 at 1 kG
-- Fit to S = IHeisenberg Chain Model 0
l0 l/x at I kG







Magnetic Properties of [Co(4,4'-bipy)3(H20)2] (CIO4)2-1.4(4,4'-bipy).3(H20)
A complete analysis of the magnetic data for 2 is more complicated compared to
analogous Ni(II) and Cu(II) systems since, in Co(II) systems, spin-orbit coupling effects
are important, a thermal dependence of the spin quantum number is observed
(depopulation of S = 3/2 state to S = V2 state at low temperatures), and significant

anisotropy in the g-factors is present.1,5,6 The Co(II) ions of 2 form chains of Ising S =
V2 spins that assume an S = '/2 state at temperatures below 30 K. The S = V2 Ising model is
an appropriate starting point since octahedral Co(II) complexes have highly anisotropic

g-factors. 1, 5,7 Again, only exchange interactions along the chains were considered. The

general Hamiltonian may be written as

H=-2Jm aS.; +/S .S+ + y Y.SJ + gpuB. (2-10)
i^j i

where, in the Ising model, a = 1 and 3 = y = 0, zero-field splitting terms have been

neglected, and the remaining terms have the usual meaning. From this Hamiltonian, the

parallel and perpendicular susceptibilities are

2 2 (211 J
g==t-exp e (2-11)
4kBT kBT

X X -NAg"p"uR Itanh(l J- I -)+ J1 sech' 2 I) (2-12)
81JI + kl ec) k l kjZ)

and the susceptibility representing the average of contributions that are parallel, 21 and

perpendicular, XL, to the chains, is

Z.ave (ZII + 2 ) (2-13)









since magnetic measurements were performed on randomly oriented samples. 1,7

Unlike Ni(II) systems, significant anisotropy with Co(II) spins is observed and a

complete analysis of the magnetic behavior requires data from single crystals in different

orientations with respect to their measuring magnetic field. 1,5,6 Since the reported

experiments were performed on powder specimens, both perpendicular and parallel g-

tensor components were unavailable. Thus, any meaningful fits to the low-temperature

susceptibility data in the absence of known values of both J and g\\ and g_-Lvalues are

difficult. However, simulations of the low-temperature susceptibility from the Ising

model from Equations 2-11,2-12, and 2-13, shown in Figure 2-18, reproduced the data

relatively well. The coupling constant was fixed at J= 0.4 K, similar to the value

obtained from fits to the susceptibility data of 1 (since 2 is isostructural to 1). Three sets

of g-values (glI = 8 and g-L = 1.5, g1\ = 6and gL = 3.5, and gl = 4 and gjL = 4.25) were

obtained for octahedral Co(II) ions from a universal curve reported by Carlin.5 The

fitting expression was also corrected for uncoupled, S = V impurities and fixed at a high

limit of 10 %. The model incorporating all three sets of g-values reproduced the low-

temperature susceptibility well, with the best results obtained for glI = 8and g-L = 1.5.

Increasing or decreasing the magnitude of the coupling constant had resulted in poor

simulations of the data. Thus, while it was not possible to unambiguously estimate the g-

value components, the simulations provided a rough estimate of the sign and magnitude

of the exchange parameter. Qualitatively speaking, the cobalt analog, 2, also shows little

evidence of magnetic exchange just as in 1 and 2.








Magnetic Properties of [Cu(4,4'-bipy)3(DMSO)2](C1O4)2-2(4,4'-bipy)

For the case of the Cu(II) ions of 3, the room temperature (i.e. 300 K)

susceptibility (ZM = 1.2 x 10-3 emu mol-1) correlates well with the value expected for

uncoupled, S = /2 metal centers (XM = 1.3 x 10-3 emu mol-1). The temperature

dependence of the molar magnetic susceptibility at 1 kG, shown in Figure 2-19, was fit

well by the Curie law using a g-value of 2.06, as determined from room temperature ESR

measurements. The field dependence of the molar magnetization up to 50 kG at 2 K,

shown in Figure 2-20, was also fit well by the S = V Brillouin function, with a g-value of

2.06, which describes non-interacting magnetic spins.5,6 Consequently, 3 behaves

essentially as a chain of non-interacting, S = 2 metal centers that are uncoupled even at 2

K. The larger dihedral angles between the pyridyl rings on the bridging bipy ligands may

disrupt the superexchange pathway to a greater extent than in 1 or 2 and could account

for the lack of any observed exchange interaction.
















15


12


S0.9


S-0.6


031-


001 1 i
0 4 8 12 16 20
T (K)

Figure 2-18. The molar magnetic susceptibility, "M, at 1 kG from 2 K to 300 K for
compound 2 is shown as open boxes. The data have been corrected for background
signals arising from the sample container and diamagnetic contributions. The simulations
of to the S = '/2 Ising model are from Equations 2-11,2-12, and 2-13 where Jis fixed at
- 0.4 K while gi = 1.5 and g, = 8 for the solid line, gi = 3.5 and giI =6 for the dotted line,
and gL = 4.25 and g\ = 4 for the dashed line.


0.20


0.15


S0.10


0,05


0.00


50 100 150 200 250 300
T (K)


1000


800


600 -


400


200


0


Figure 2-19. The molar magnetic susceptibility, %M, and inverse susceptibility, 1IZM, at
1 kG from 2 K to 300 K for compound 3 are shown as open boxes and open circles,
respectively. The data have been corrected for background signals arising from the
sample container and diamagnetic contributions. The fit of to the S = V/2 Curie Law is
shown by the solid line.


0 0 ^at I kG
I--. g- = .5 amg,= 8
o g- =4.25dg -4


F






81







6

5

4 4

2

o f 2at2
S1 --Fit to S= 1/2 Brillouin Function

0
p I I, .
0 10 20 30 40 50
B (kG)

Figure 2-20. The molar magnetization (MM) at 2 K from 0 to 50 kG for compound 3 is
shown as open boxes. The data have been corrected for background signals arising from
the sample container. The fit of M to the S = V2 Brillouin function is shown by the solid
line.








Conclusions

Three new hybrid organic-inorganic coordination polymers have been isolated

and characterized. These materials consist of chains of transition metal ions (Ni(II),

Co(II), and Cu(II)) bridged by 4,4'-bipyridine spacer ligands. The chains pack to form

two-dimensional, non-interpenetrated sheets with hydrophobic, rectangular cavities

present within the framework. The sheets, in turn, pack to form a three-dimensional

structure with oblique channels containing enclathrated guest molecules and counterions

extending throughout the solid. These enclathrated guests are easily lost suggesting that

the samples are thermally unstable. In general terms, the magnetic properties of 1, 2, and

3 are similar in the sense that weak exchange interactions, J, are present between the

metal centers.

Coordination polymers with hydrophobic cavities and channels extending

throughout the solid-state structure have received much attention due to their ability to

act as molecular sieves with size and shape specificity and catalytic substrates.37,44,45

Compounds 1,2, and 3 are clearly examples of network solids with cavities and channels

that prefer to enclathrate hydrophobic guests (uncoordinated bipy molecules). The ability

of these coordination polymer hosts to exchange guests is described in Chapter 3.














CHAPTER 3
A 31P MAS NMR INVESTIGATION OF THE HOST-GUEST PROPERTIES OF TWO
POROUS NETWORK SOLIDS,
[Ni(4,4'-bipy)3(H20)2](C104)2 1.4(4,4 '-bipy)-3(H20) and
[Co(4,4'-bipy)3(H20)2](C104)2- 1.4(4,4'-bipy)-3(H20)

Introduction

The design and construction of hybrid organic-inorganic porous network solids

through the self-assembly of simple, molecular and ionic components is an emerging area

of supramolecular chemistry that can provide new generations of functional materials.

The importance of porous solids in inclusion phenomena, e.g. adsorption / desorption, ion

exchange, and size and shape-selective molecular sieving, as well as catalysis, is due, in

part, to their ability to reversibly clathrate or trap species within their cavities and

extended channels.129 Like zeolites, many hybrid organic / inorganic solids clathrate

guests within pores, cavities, and channels that are part of their lattice framework. These

pores often possess a variety of sizes and shapes not observed in analogous inorganic

porous solids such as zeolites and molecular sieves, thus potentially yielding novel and

unique inclusion capabilities.87 A number of porous coordination networks have been

found to exhibit many other desirable zeolitic properties as well, such as stability and

porosity of the framework, guest exchange, and selective catalytic activity.42,44,93,142

By careful selection and design of the chemical components, the size and clathration

properties of these pores can be fine-tuned to meet specific needs while maintaining the

overall structural and functional features found in naturally occurring analogs.37








Despite the large amount of research devoted toward the synthetic aspect of solid-

state supramolecular chemistry in terms of producing functional materials and elucidating

methods for the rational design and fabrication of such materials, studies of the chemical

reactivity of these materials has been lacking.147 In fact most of these reactivity studies

have been limited to the investigation of the inclusion properties of porous solids, such as

the guest exchange and adsorption-desorption processes of small

molecules.45,87,142,144,147 The imbalance of synthetic work compared to reactivity

studies of supramolecular materials is largely due to the fact that, unlike the well-

characterized chemical reactivity properties of molecular and ionic species that are

generally soluble in many common solvents, most coordination polymers are insoluble in

most organic and inorganic solvents thus rendering any reactivity studies difficult. 147

Yaghi, et al, has reported selective guest binding and removal of alcohols and

ketones from a three-dimensional porous Zn(II)-Benzenetricarboxylate network and

aromatic molecules from a layered porous Co(II)-Benzenetricarboxlyate network without

collapsing the host.45',87 Endo, et al. has described the reversible non-selective guest

binding and removal of ketones, esters, hydrocarbons, and haloalkanes, as solids, liquids,

and gases, within a zeolitic anthracene-bis(resorcinol) layered, hydrogen-bonded

network. 142 Kondo, et al. has reported the adsorption and desorption properties of small

gas molecules, e.g. methane, within microporous interpenetrated M-4,4'-azopyridine (M

= Mn(II), Cd(II), and Co(II)) coordination networks without breaking apart the host.144

NMR Spectroscopy

Magic angle spinning (MAS) NMR spectroscopy is a very useful technique for

investigating chemical interactions within solid-state materials and can be applied toward








determining the presence of enclathrated guests within porous solids. 165 Additionally,

pore sizes can be estimated as well as observing the dynamic behavior of the guests (i.e.

chemical exchange), the nature of the binding sites (provided a coupling constant can be

measured), and the relative strength of any specific chemisorption and physisorption

interaction (i.e. acid-base, coordination, hydrogen bonding, and adsorption). 165

One approach to studying the interactions of guest molecules within hosts

involves the use of small probe molecules that can be inserted within the lattice of a

solid.166 Using probe molecules containing spin V/2 nuclei (e.g. 'H, 13C, 'SN 19F, 31P, and

129Xe) are useful because the corresponding NMR signals are not complicated by

quadrupolar interactions. 165 In the past, several NMR probes, including 2- 13C

acetone,167-171 4-13C mesityl oxide,172-175 and 15N pyridine,176-180 have been used

to specifically identify and quantify acidic sites in solid acids such as zeolites and

amorphous silica-alumina. Unfortunately, the small magnetogyric ratios, low natural

abundances, and relatively limited chemical shift ranges of '15N (0.4 % abundant) and 13C

(1.1% abundant) requires either extensive signal averaging or the use of isotopically

enriched materials in order to obtain high resolution NMR spectra. 166,181 To avoid

both of these problems, a more suitable probe molecule would incorporate a more

sensitive NMR active nucleus such as phosphorus-31.182 The large gyromagnetic ratio

and near 100 % natural abundance of 31P results in a significant increase in NMR signal

intensity compared to either 13C or 15N without the use of expensive enriched

samples. 181,183,184 Furthermore, the large isotropic chemical shift range and full

chemical shift anisotropy can provide useful information regarding the chemical








environments of the 31 P nucleus. 181,185,186 Therefore, phosphorous-containing probe

molecules are well suited for probing the nature and strength of host-guest interactions.

Recently, 31P MAS NMR spectroscopy has been applied specifically toward for

the identification and quantification of Lewis and Bronsted sites on the surfaces and

within solid acids.166,182,187-190 In the past, trialkylphosphines (TAP), and in

particular, trimethylphosphine (TMP), have been the probe molecules of choice for the

characterization of solid acids by 31p MAS NMR spectroscopy. 166,182'190-194 The

coordination of a TAP molecule to an acidic site results in a characteristic chemical shift

that is strongly dependent on the Bronsted or Lewis nature of that site. 186 In fact, 31p

NMR spectroscopy can not only distinguish between Bronsted-complexed TMP from

Lewis-bound TMP but also resolve peaks due to varying local environments of the

coordinated probe molecules.166'186 The basicity of trimethylphosphine (pKa = 5.3 in

water) results in the formation of a protonated base (TMPIH+) upon coordination to any

Bronsted site with a characteristic chemical shift of about 3 ppm (referenced to 80 % o-

phosphoric acid) that is largely invariant with the strength of the acid site. 181,195 Thus,

TMP has been extensively used to determine the presence and quantity of Bronsted sites

in solid acids. Lewis bound TMP exhibits a considerable upfield shift relative to the

Bronsted site. 181,186 However, the similar chemical shift of Lewis bound and

physisorbed TMP causes difficulty in using this probe to unambiguously determine the

presence and population of Lewis sites in a solid.181,186 Rapid chemical exchange

dynamics between bound and free TMP molecules at room temperature can also lead to

uncertainty in the identification and quantification of acid sites. 181 Furthermore, TMP is








a highly flammable and air-sensitive liquid at room temperature and preparing solid acid

standards for quantitative measurements is difficult. 181,185

Like trialkylphosphines, trialkylphosphine oxides (TAPO's) are also suitable

basic probe molecules for studying interactions within solid acids as they are able to

distinguish Bronsted sites from Lewis sites. 166,181,183,186,188 The intrinsic basicity of

trialkylphosphine oxides is on the same order of magnitude as trimethylphosphine. 186

However, the removal of the phosphorous atom from the basic site leads to a wide range

of chemical shifts that vary with the strength of the acid site.186 Unlike the

corresponding TAP's, TAPO's are solids at room temperature, not susceptible to

oxidation, and can thus be introduced into a host through solution-state

chemistry. 181,182 The most commonly used trialkylphosphine oxides are

trimethylphosphine oxide (TMPO) and triethylphosphine oxide (TEPO). 186 The ability

of these materials to measure acidity in both solution and in solids is known and TEPO,

for example, has been used as a probe for determining solvent acidity.196 Unfortunately,

there are few published reports describing the use of trialkylphosphine oxides as probe

molecules to study interactions with porous solids. 181

The chemical shift of an NMR peak resulting from the interaction of a basic probe

molecule with an acid site in a host is primarily due to local changes in the magnetic field

around the NMR active nucleus originating from electronic changes in the probe

molecule as a whole.186,197,198 An acid-base interaction can be viewed as a transfer of

electron density from the base to the acid site thus creating overlap between the LUMO,

the electron deficient orbital on the acid (OA), and the HOMO (qB), one of the orbitals on








the base that contains a lone pair of electrons (Figure 3-1).165,186 As the adduct bond

strength increases, the electron density "flows" away from the basic oxygen atom. The

net loss of electron density from the NMR active nucleus due to the formation of an

acid/base adduct (PAB ) can be observed in the phosphorous-31 NMR.165,186

E
S 'AB +'AB A'AB
..... -\ .....[ /
"; "1 ; -- t ^/ "B'** i
A4'A


(DAB 'AB 'DAB
EA >> E#B EA > EB EOA EB


Increasing Srength of
Acid-Base Interaction

Figure 3-1. An acid-base reaction from a molecular orbital perspective. The electron
deficient acid, OA, reacts with the electron rich base, B, to form an adduct, PAB. As the
strength of the acid-base interaction increases, the adduct progressively becomes more
covalent in character. Adapted from reference 165.

If the basic probe molecule interacts with a Bronsted acid site, such as water, the

resulting adduct bond is characterized by some degree of proton sharing between the acid

and the base. 186 Consider the hydrogen bonding of the oxygen atom from a

trialkylphosphine oxide with the proton of an acidic water molecule. Increasing the

strength of the acid site transfers this shared proton to the base. The (water) 0-H and

(TAPO) P-0-O bond order both decrease while the 0-H bond order increases in the

resulting TAPO-HaO20 adduct. The phosphorous atom therefore becomes deshielded

shifting the corresponding NMR signal downfield. The stronger acid-base interactions,








the greater the 31P resonance shifts downfield. However, once the proton is completely

transferred to the base, no further downfield shift of the NMR signal will be observed.

Instead, for any stronger acid sites, the same chemical shift will be seen in the

spectrum. 186

The coordination of a base to a Lewis acid, such as a metal ion, will also result in

the formation of an adduct bond.186 The strength of the acid site will therefore be

directly related to the strength of the adduct bond. Again, consider the effect of

coordination of a trialkylphosphine oxide probe molecule directly to a metal center. As

the acid strength increases, the electron density gradually transfers from the oxygen atom

to the metal ion to a greater extent. The P-O bond order decreases while the M-O

bond order increases. The phosphorous atom becomes more deshielded and the

corresponding NMR resonance progressively shifts downfield. At very high acid

strengths no further change in the chemical shift is expected, unless oxygen atom transfer

occurs.186 In this case, acid site is oxidized thus reducing the trialkylphosphine oxide to

the corresponding trialkylphosphine.

Recently, Rakiewicz et al. characterized the acid sites of amorphous silica-

alumina, zeolites HY, dealuminated HY, and USY, and gamma-alumnina with

trimethylphosphine oxide. 181 The authors were able to identify and differentiate Lewis

and Bronsted sites and then measure the population of those sites. The trend in assigning

chemical shift ranges for Lewis or Bronsted bound phosphine oxides includes work by

Lunsford and Baltusis. 166,187,188,193 Despite consistency in identifying and

characterizing Bronsted sites, each author has assigned a different region to the chemical

shifts for the Lewis bound probe.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EC36FFLTA_94UY3Z INGEST_TIME 2014-06-24T20:35:40Z PACKAGE AA00022323_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

CHEMICAL AND PHYSICAL CHARACTERIZATION OF HYBRID ORGANIC-INORGANIC LOW-DIMENSIONAL COORDINATION POLYMERS By JONATHAN DAVID WOODWARD A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2003

PAGE 2

To My Parents and Grandparents

PAGE 3

ACKNOWLEDGMENTS I would very much like to thank my research advisor, Dr Daniel Talham for his support guidance encouragement novel ideas, and advice during my time in his research group at the University of Florida. Despite some difficult times early on, Dan was willing to work with me. He encouraged me to continue my efforts and press on so that I could complete my degree. From Dan I have learned how to approach a research problem from an objective standpoint how to interpret data and most importantly how to present results in an acceptable manner. I would also very much like to thank Dr. Renal Backov a postdoctoral fellow in our group for the past 2 years for his help support, and encouragement. Renal was involved in many aspects of my graduate work and this dissertation would not be possible without his efforts I would like to thank Dr Khalil Abboud for his work in solving numerous crystal structures and for teaching me all about crystallography I would also like to thank Dr C Russell Bowers and his graduate student, Bhavin Adhyaru for the use of the NMR equipment in the host-guest investigations. I would like to thank Bhavin in particular for the rather large amount of time spent in completing NMR experiments as well as writing and editing. I would like to thank Dr Mark Meisel and his group members (Dr Hitoshi Ohnuki Dr. Brian Watson A. Nicole Morgan Ju-Hyun Park and Diktys Stratakis) for performing bulk magnetic measurements on many of my samples over the past 5 years. I would like to acknowledge Dr. John Reynolds Dr. Gus Palenik Dr. Mark Meisel and Dr David Richardson for being part of my Ph D committee lll

PAGE 4

I would like to thank the past and present members of the Talham research group (Dr. Brian Ward Dr. Gail Fanucci Jeff Culp Eduardo Perez-Cordero, Chen Liu David Zipse, and Sarah Lane) for their support during my time here. I would also like to thank Brian for help in getting me started in my research; Gail for teaching me how to operate the lab machinery including the computers; and Eduardo for teaching me some advanced math in order to model magnetic data. I would also like to especially thank Jeff for putting up with me during the time spent writing my thesis and his help with numerous things for the past five years I want to also give special thanks to my parents, Ann and Dennis Woodward and my grandmother Virginia Prescott, for the support and encouragement throughout my entire life whenever I needed it. Without them, I certainly would have not made it this far. IV

PAGE 5

TABLE OF CONTENTS ACKNOWLEDGMENTS ........... ....... ......................... ............... ........ ..................... ........ iii ABSTRACT ....................................................................................................................... ix CHAPTER 1 INTRODUCTION . ......... ...................... ... .......................... ......... ... ................... . ........ 1 Low-Dimensional Materials ........................................................................................... 2 Introduction to Molecular Magnetism ........................................................................... .3 Diamagnetism and Paramagnetism ............... ...... .. ................................ ... .......... ..... . .3 Basic Relationships . ... . ..... . ...... .. .. . . .................... ..... ....... .... ... ... ....... .. ......... ......... 4 Non-Interacting Spin Systems and the Curie Law ........................ ..... ..... ..... ............. 6 Interacting Spin Systems and the Curie-Weiss Law ................................................... 7 Magnetic Exchange ....... ............................................... ... ... ........... .. ....................... 9 Anisotropy ... . .. .. ..... .... .. .... ... .. . ............................... ........ . ...... .... .. ....... ........ ... .. 11 Dimers ....................................................................................................................... 16 Chains ....................................................................................................................... 20 Ladders ............... . . .......... ...... ... . .. ......... .............................. ............. .. ........... .. ... 24 Self-Assembly of Supramolecular Architectures .. ..... ... ... .. ... .. .. . . . .. .. ............... ... 27 Strategies for Building Supramolecular Architectures .. ........... . ... ... ......... ..... ...... .. .27 Factors Affecting the Structure ofSupramolecular Architectures ....... .. .. .. .. ........ .. .31 Porous Network Supramolecular Architectures ........................................................ 33 Scope of the Dissertation ..... .. . ... . . . ... ........ ... ...... ...... .......... ........ .. .......... . .. .. .. 3 8 2 STRUCTURAL THERMAL AND MAGNETIC PROPERTIES INVESTIGATION OF THREE TRANSITION MET AL-4,4 '-BIPYRIDINE COORDINATION POLYMERS: [Ni( 4 4 '-bipy)J(H2O)2](ClO4)2 1.4( 4,4 '-bipy) 3(H2O) [Co(4,4' -bipy)J(H2O)2](ClO4)2 l .4(4,4' -bipy)(H2O), and [Cu( 4 4 '-bipy)J(DMSO)2](ClO4)2( 4,4 '-bipy) ............................................................ 39 Introduction ................................................................................................................... 39 Experimental Section ........................... .. .. . ... ... ..... .... ..... ........................... ... ........... ... 41 Materials .... .. .... ....................... ....... ......... .... ....................................... ............... . ... 41 Synthesis of [Ni( 4,4 '-bipy)3(H2O)2](ClO4)2 l .4( 4,4 '-bipy) 3(H 2 O) .......... ..... ......... .42 Synthesis of [Co( 4 4 -bipy)J(H2O) 2 ](ClO 4 ) 2 l .4( 4 4 '-bipy) 3(H 2 O) .... ............ . ... . .42 Synthesis of [Cu( 4 4 '-bipy)J(DMSO)2](ClO 4 ) 2 2( 4,4 '-bipy) ....................... ... . ... ... .42 X-Ray Structure Determination .... .............................. ..... ..... ............ .. ..... ..... ......... .43 V

PAGE 6

Thermal Analysis ...................................................................................................... 44 Magnetic Measurements ........................................................................................... 45 Results and Discussion .................................................................... . ..... ...................... 46 Compound Synthesis ................................................................................................ 46 Description of the Structures .................................................................................... 4 7 Structure of [Ni(4,4' -bipy) 3 (H2O)2](ClO4)2 l .4(4,4' -bipy)(H2O) ..................... .47 Structure of [Co( 4,4, -bipy)3(H2O)2](ClO4)2 l .4( 4,4, -bipy) 3(H2O) ..................... 53 Structure of [Cu( 4,4, -bipy)3(DMSO)2](ClO4)2( 4,4, -bipy) ........ ... ..................... 59 Thermal Properties .................................................................................................... 63 Magnetic Properties .................................................................................................. 72 Magnetic properties of [Ni(4,4' -bipy)J(H2O)2](ClO4)2" 1.4(4,4' -bipy)(H2O) .... 72 Magnetic properties of [Co( 4,4 '-bipy)3(H2O)2](ClO4)2 l .4( 4,4 '-bipy) 3(H2O) .... 77 Magnetic properties of [Cu(4,4' -bipy)3(DMSO)2](ClO4)2(4,4' -bipy) ............... 79 Conclusions ................................................................................................................... 82 3 A 31 P MAS NMR INVESTIGATION OF THE HOST-GUEST PROPERTIES OF TWO POROUS NETWORK SOLIDS, [Ni( 4,4 'bipy)3(H2O)2](ClO4)2 l .4( 4,4 'bipy) 3(H2O) and [Co(4,4' -bipy)3(H2O)2](ClO4)2 l .4(4,4' -bipy)(H2O) ................................................. 83 Introduction ................................................................................................................... 83 NMR Spectroscopy ................................................................................................... 84 Experimental Section .................................................................................................... 90 Materials .......................................................................... ....... .................................. 90 Sample Preparation ................................................................................................... 90 Gas Chromatography Analysis ................................................................................. 91 NMR Spectroscopy ................................................................................................... 92 Powder X-Ray Diffraction ........................................................................................ 92 Results and Discussion ................................................................................................. 92 Sample Preparation ................................................................................................... 92 Guest Loss from Network Solids .............................................................................. 93 Guest Exchange Investigations of Compound 1 Involving TMPO .......................... 94 Gas chromatography results .................................................................................. 94 NMR results .......................................................................................................... 96 X-Ray results ...................................................................................................... 101 Guest Exchange in Compound 1 Involving Variable Quantities of TMPO ....... 106 Gas chromatography results ................................................................................ 106 NMR results ........................................................................................................ 107 X-Ray results . .................................................................................................... 110 Guest Exchange Investigations of Compound 1 Involving TEPO and TPPO ........ 113 Gas chromatography results ................................................................................ 113 NMR results ........................................................................................................ 115 X-Ray results .................................................... ......... ......................................... 117 Guest Exchange in Compound 2 Involving TMPO, TEPO, and TPPO ................ .120 Gas chromatography results ................................................................................ 120 NMR results ........................................................................................................ 120 VI

PAGE 7

X-Ray results ...................................................................................................... 123 Conclusions ................................................................................................................. 126 4 STRUCTURAL AND MAGNETIC CHARACTERIZATION OF A SERIES OF AZIDO-BRIDGED COPPER(II) COORDINATION POL YMERS .......................... 127 Introduction ................................................................................................................. 127 Azide Bridging Modes ............................................................................................ 127 Superexchange Properties of Azide Bridges .......................................................... .129 Experimental Section .................................................................................................. 13 7 Materials ................................................................................................................. 13 7 Synthesis of [ Cu2(PhPyPy )2--(N 3)2(N 3)2] ............................................................. 13 7 Synthesis of [Cu2(terpy)2--(N3)4][Cu2--(N3)2(N3)2] ............................................ 137 Synthesis of [Cu2(terpy)2--(N3)2(N3)2][Cu3--(N3)4(N3)2] ................................... 138 Physical Characterization ........................................................................................ 138 X-Ray Structure Determination .............................................................................. 138 Magnetic Measurements ......................................................................................... 139 Results and Discussion ............................................................................................... 140 Description of the Structures .................................................................................. 140 Structure of [Cu2(PhPyPy)2--(N3)2(N3)2] .......................................................... 141 Structure of [Cu2(terpy)2--(N3)4][Cu2--(N3)2(N3)2] ........................................ 144 Structure of [Cu2(terpy)2--(N3)2(N3)2][Cu3--(N3)4(N3)2] ................................ 149 Electron Paramagnetic Resonance .......................................................................... 158 Magnetic Properties of [Cu2(PhPyPy)2--(N3)2(N3)2] ............................................ 159 Magnetic data ...................................................................................................... 159 Magnetic model and fits ...................................................................................... 161 Interpretation of the magnetic data ..................................................................... 167 Rationalizing the sign and magnitude of the coupling constants ........................ 168 Magnetic Properties of [Cu2(terpy)2--(N3)4J(Cu2-,U-(N3)2(N3)2] ........................... 169 Magnetic data ...................................................................................................... 169 Magnetic model and fits ..................................................................................... .1 71 Interpretation of the magnetic data ..................................................................... 180 Rationalizing the sign and magnitude of the coupling constants ........................ 181 Magnetic Properties of [Cu2(terpy)2--(N3)2(N3)2][Cu3--(N3)4(N3)2] .................. 186 Magnetic data ...................................................................................................... 186 Magnetic model and fits ...................................................................................... 190 Interpretation of the magnetic data ..................................................................... 193 Rationalizing the sign and magnitude of the coupling constants ........................ 194 Conclusions ................................................................................................................... 198 5 CONCLUSIONS ......................................................................................................... 199 Vll

PAGE 8

APPENDIX A CRYSTAL STRUCTURES OF SELECTED LOW-DIMENSIONAL SOLIDS ..... .202 B ATOMIC COORDINATES AND BOND ANGLES AND DISTANCES . . ........... 223 LIST OF REFERENCES ............... ................. ........................... .... . . . ... . . ....... . . ... .. 254 BIOGRAPHICAL SKETCH . ..... ..... . ... . . . ... ....................................... ..... . . ... ... . . .. .. 269 Vlll

PAGE 9

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CHEMICAL AND PHYSICAL CHARACTERIZATION OF HYBRID ORGANIC-INORGANIC LOW-DIMENSIONAL COORDINATION POLYMERS By Jonathan David Woodward May 2003 Chair: Daniel R. Talham Department: Chemistry This dissertation presents experimental results from the synthesis and structural, chemical, and physical characterization of representative low-dimensional coordination polymers. The structural, thermal, and magnetic properties of a series of clathrated porous network solids, [Ni(4,4' -bipy)J(H2O)2](ClO4)2 l.4(4,4' -bipy)(H2O), 1, [Co( 4,4 '-bipy)3(H2O)2](ClO4)2 1.4( 4,4 '-bipy) 3(H2O), 2, and [Cu(4,4'-bipy)3(DMSO)2](ClO4)2(4,4'-bipy) 3, are described first. These materials consist of chains of transition metal ions (Ni(II), Co(II), and Cu(II)) bridged by 4,4' -bipyridine spacer ligands. The chains pack to form two-dimensional, non interpenetrated sheets with hydrophobic, rectangular cavities within the framework. The sheets, in turn, pack to form three-dimensional structures with oblique channels containing enclathrated guest molecules and counterions extending throughout the solid. These enclathrated guests molecules are easily lost, suggesting that the samples are lX

PAGE 10

thermally unstable. The magnetic properties of 1, 2, and 3 are similar in the sense that only very weak exchange interactions are present between the metal centers. The host-guest properties of 1 and 2 were investigated. Gas chromatography experiments determined that both hosts exchange clathrated bipy molecules with trialkylphosphine oxide probe molecules, TMPO, TEPO, and TPPO. While the uptake of TMPO by both 1 and 2 is essentially complete, steric constraints are believed to limit the uptake of TEPO and TPPO by the host. The trialkylphosphine oxides interact with acid sites within the host as determined by 31 P MAS NMR spectroscopy. TMPO interacts with both coordinated water molecules (strong acid sites) and lattice waters (weak acid sites) in compound 1 and coordinates directly to the metal centers in compound 2. However, TEPO and TPPO seem to attack only the weaker acid sites within the hosts. X-ray diffraction patterns show that the loss of bipy and uptake of the probe causes significant structural rearrangements in 1 and only mild structural changes in 2. However, the nature of these guest-exchanged products is unknown. These experiments showed that solid-state NMR spectroscopy can be used to investigate host-guest interactions. Finally, the structural and magnetic properties of three azido-bridged copper(II) ladder-like coordination polymers, [Cu2(PhPyPy)2--(N3)2(N3)2], 4, [Cu2(terpy)2--(N3)4Cu2--(N3)2(N3)2], 5, and [Cu2(terpy)i--(N3)2(N3)2Cu3--(N3)4(N3)2], 6, are discussed. Compound 4 structurally resembles ladder-like chains of weakly interacting end-on azido bridged copper(II) dimers. Magnetically, compound 4 consists of antiferromagnetic chains of ferromagnetically coupled S = dimers. Compound 5 consists of ladder-like copper(II) X

PAGE 11

coordination polymers with double and single end-on azido bridges. Magnetically compound 5 consists of ladder-like stacks of weakly interacting tetramers with a dominant and unusual antiferromagnetic exchange mediated through end-on azido bridges. Compound 6 structurally resembles ladder-like chains of weakly interacting copper(II) pentamers featuring both single and double end-on azide bridges Magnetically compound 6 consists of antiferromagnetic stacks of pentamers with two paramagnetic S = sites and ferromagnetically coupled trimers. Xl

PAGE 12

CHAPTER 1 INTRODUCTION The original goal of the graduate research pertaining to this thesis focused on the design, self-assembly, and physical characterization of molecular coordination polymer ladder systems. This project was only partially successful; several new ladder-like materials were isolated as a result of serendipity rather than rational design and did not incorporate the desired structural and physical properties. However, in addition to the ladder-like structures mentioned above, several new low-dimensional materials were obtained from the course of this graduate research. The work presented in this dissertation is concerned primarily with investigating the structural, chemical, and physical properties of selected examples of these low-dimensional materials. This chapter briefly provides the background information necessary to the research detailed in the following chapters. A basic concepts of molecular magnetism as applied to low-dimensional materials is introduced first. Included are discussions concerning the magnetic properties of both paramagnetic and exchanged coupled systems, including discrete oligomers and extended chains and ladders, and the theories that model the magnetic behavior of such materials. A general introduction to the field of supramolecular chemistry is then provided. Discussed are general strategies for building supramolecular architectures, particularly via the self-assembly of simple molecular and ionic components, general structural types of the resulting assemblies commonly encountered, and practical applications of such materials. 1

PAGE 13

2 Low-Dimensional Materials The study of low-dimensional materials has been a rapidly expanding area of solid-state chemistry. Low-dimensional materials have structural and physical properties that are anisotropic in one or two dimensions_ 1-3 These materials are often referred to as quasi low-dimensional systems since despite being part of a three-dimensional solid characteristic structural and physical properties exist principally within one or two dimensions. One-dimensional chains and two-dimensional sheets (Figure 1-1 ) are the most commonly encountered types of low-dimensional materials. The key structural features of low-dimensional materials are strong electrostatic or covalent bonds along chains or within sheets with weak Van der Waals interactions between the chains or sheets. As a result of this anisotropic bonding, electronic magnetic and transport properties exist along chains or within sheets or are enhanced compared to higher dimensional analogs but are small or negligible between the chains or sheets.4 A B C D Figure 1-1. Examples of low-dimensional materials. A) One-dimensional chain B) Two-leg ladder. C) Three-leg ladder. D) Two-dimensional sheet.

PAGE 14

3 Introduction to Molecular Magnetism Diamagnetism and Paramagnetism When a material is subjected to an external, homogeneous field, H a magnetization, M is induced within the sample.5,6 The quantities Mand H, are related by BM x= an (1-1) where x is the magnetic susceptibility. The susceptibility quantitatively measures the response of a material to an applied magnetic field. If the magnitude of the field is small then to a good approximation the magnetization is a linear function of the field, and the susceptibility can be expressed as M x=n (1-2) The total susceptibility, x r is the sum of two components a diamagnetic contribution, Xd i a, and a paramagnetic term Xpara X T = X para + X dia (1-3) Diamagnetism is a property of all matter originating from the interaction of paired electrons with an applied magnetic field; since all materials have paired spins all materials have diamagnetic contributions to the total susceptibility When a diamagnetic material is placed within an external field it is repelled since the sample produces a flux opposed to the applied field and moves toward regions of lower field strength. 5 Diamagnetic susceptibilities are typically small and negative (on the order of -10 6 emu mor 1 to -10 4 emu mor 1 ) and are independent of field strength and temperature. 5

PAGE 15

4 Paramagnetism, on the other hand, is a property exhibited only by materials with unpaired electrons, such as transition metal complexes, rare earth compounds, and organic free radicals. When a paramagnetic material is placed within an external field, it is attracted to the field because of the interaction of the unpaired spins with the applied field and thus moves toward regions of higher field strength. 5 In general, low fields and high temperatures will tend to randomize the directions of these spins, resulting in small or zero net magnetic moments. Low temperatures and high fields, however, will align the spins with the field, resulting in a net moment. Paramagnetic susceptibilities are typically positive and much larger in magnitude than the corresponding diamagnetic susceptibilities, often in the order of 10 4 to 10 2 emu mor' or more.5 Diamagnetic susceptibilities can be estimated from Pascal's constants and subtracted from the total susceptibility ( or ignored, if small compared to the paramagnetic contribution) to obtain the paramagnetic susceptibility.5,6 Unlike diamagnetic susceptibilities, paramagnetic susceptibilities exhibit temperature dependence, often in very complex manners. Basic Relationships In classical physics, the magnetization, M, of a sample results from a variation of its energy, E, in response to an applied magnetic field, B, through and M=-aE BB B=H (1-4) (1-5) where B is the magnetic induction field and is the permeability. 5 ,6 When > 1, the sample is paramagnetic and when< 1 the sample is diamagnetic. Similarly, in quantum

PAGE 16

5 mechanics, a microscopic magnetization, ,,, of an energy level, En, in the presence of an external field is (1-6) where n = 1, 2, 3, and so on. The macroscopic magnetization is then a weighted sum of all microscopic magnetizations in the sample, given by the Boltzmann distribution law NA~)aEn)exp(-En) M = n BB k 8 T -E Iexp( n) n k 8 T (1-7) where NA is Avogadro's number and k 8 is the Boltzmann constant and the denominator is the partition function, Z. The magnetic susceptibility is the variation of the magnetization with the external field (1-8) where J1(J is the permeability of free space. At the limits of high temperature and low field, Equations 17 and 1-8 can be simplified to give new relations that are no longer functions of the derivatives BEnlBB. Expanding the energy levels, En, in a power series of B, gives (1-9) where E!i ) terms are the Zeeman coefficients. 7a Retaining only linear terms, substituting back into Equation 1-8, and simplifying under the assumption that at zero field, the magnetization is zero, the susceptibility vs. temperature is

PAGE 17

6 X = -E (O) Iexp( n ) n k 8 T Equation 1-10 is the Van Vleck susceptibility. Non-Interacting Spin Systems and the Curie Law (1-10) The simplest type of paramagnetism is that of an ideal paramagnet, a system composed of non-interacting, randomly orientated spin centers (Figure 1-4 A) 7b. In an ideal paramagnet, the magnetic susceptibility is inversely proportional to temperature. The Curie law is a simple relation that describes the variation of the susceptibility with absolute temperature for an ideal paramagnet and is given by C x=T where C is the Curie constant given by C= NAg 2 !S(S+l) 3kB (1-11) (1-12) and NA is Avogadro's number, g is the g-Lande value, sis the Bohr Magneton, and Sis the total spin of the system. A plot of the susceptibility versus temperature for a paramagnetic material that follows the Curie Law is a simple hyperbola. Other mathematical manipulations of the Curie law are useful as well. A plot of the inverse susceptibility versus temperature for a paramagnet is a straight line where the slope is the Curie constant and the x-intercept is zero -I T X = C (1-13)

PAGE 18

7 while a plot of the product of susceptibility and temperature vs. temperature is a horizontal line xT=C (1-14) Plots of the X, x-1, and xTvs. Tfor an arbitrary S= material that obeys the Curie law, with g = 2.10, are shown in Figure 1-2. Interacting Spin Systems and the Curie-Weiss Law In many paramagnetic materials, the unpaired electrons on the spin centers can interact with one another and the magnetic behavior is no longer ideal. The Curie-Weiss law, a semi-empirical modification to the Curie law, is a "first approximation" to model the magnetic behavior of materials with weak interactions between the spin centers, and is given by C X = (T-0) (1-15) In general, the interactions between the spins that cause these deviations are referred to as ferromagnetic and antiferromagnetic correlations. In the Curie-Weiss law, these interactions are described by 0, the Weiss constant. When 0 > 0, the interactions are ferromagnetic, when 0< 0, the interactions are antiferromagnetic, and when 0 = 0, the sample is paramagnetic. The Curie-Weiss law is valid for materials that undergo long range order above the ordering temperature (i.e. T>> Tc)Plots of the X, x1 and zTvs. T for antiferromagnetically ( 0 = -15 K) and ferromagnetically ( 0 = 15 K) coupled, S = materials (g = 2.10) that obey the Curie-Weiss law are shown in Figure 1-2. Note that when 0 > 0, the inverse susceptibility has a positive y-intercept and xT increases as T decreases. When 0 < 0, they-intercept is negative and the xT product decreases when T

PAGE 19

8 decreases The Curie-Weiss law is only a simple, general, empirical correction for describing deviations from ideal paramagnetism by accounting for the interactions of unpaired spins for magnetic materials especially when the structure is unknown. More 0 08 -"" 0 06 I l) '. "< 0 04 I 0 02 0 00 0 300 250 200 g 150 100 50 ., 0 0 2 0 1 5 t) 1 0 ... 0 5 0 0 \ . \ \ .. 50 B ., ., ., ., ., 50 Paramagnetic Antiterromagnetic Ferromagnetic A 100 150 200 250 300 T(K) ., ., ., ., ., ., ., ., ., ., ., ., ., ., ., Paramagnetic Antiferromagnetic Ferromagnetic 100 150 200 250 300 T (K) Paramagnetic Antiferromagnetic Ferromagnetic ----------------C 0 50 100 150 200 250 300 T(K) Figure 1-2. Common Curie (solid line) and Curie-Weiss (broken lines) law plots for an S = magnetic material (g = 2.10). A) XI C vs T. B) CI X vs T. C) zT IC vs. T. The dotted lines represent ferromagnetic coupling (0= 15 K) and the dashed lines represent antiferromagnetic coupling (0= -15 K). The susceptibility is normalized by division by the Curie constant C.

PAGE 20

9 complex magnetic systems such as dimers and chains require more complicated models to describe magnetic behavior. Magnetic Exchange Magnetic exchange interactions are quantum effects that describe the interactions among unpaired spins in magnetic materials. Exchange interactions originate from a combination of the Pauli exclusion principle and electronic repulsions. 8 Two principal types of exchange can be distinguished: direct exchange and indirect exchange. 9 Direct exchange interactions result from the through-space overlap of spin orbitals as in direct metal to metal bonding (Figure 1-3). These interactions are typically weak since electronic repulsive forces are large due to the close proximity of the unpaired electrons to one another 10 For indirect exchange the unpaired spins are coupled via nonmagnetic intermediaries such as bridging diamagnetic atoms or molecules or itinerant electrons of conducting solids. The fust case is referred to as a superexchange interaction (Figure 1-3) and the second case is referred to as RKKY exchange. Superexchange interact i ons are electronic (and not magnetic) interactions and are usually much larger in magnitude than the corresponding direct exchange, because intermediary groups increase the distance between metal centers thus reducing electronic repulsions 10 Pairwise exchange interactions can be expressed mathematically by a spin-Hamiltonian equation : iI = -iIJi, i s 1 ( 1-16) i > J where the sum is taken over all nearest neighbor interactions between spins S ,and S 1 with the magnitude of those interactions given by J the coupling constant. 5 If J > 0 and the unpaired electrons interact such that the spins align in a parallel fashion the exchange

PAGE 21

10 is referred to as ferromagnetic and the magnetic ground state is triplet (Figure 1-4B)7b. If J < 0, the unpaired spins align in an antiparallel manner, the exchange is referred to as antiferromagnetic and the magnetic ground state is singlet (Figure 1-4C) 7b. The sign and magnitude of the coupling constant can depend on many factors, such as the number of unpaired electrons exchanged, structural parameters of the magnetic entity, orbital overlap of the spin centers and nonmagnetic intermediary, and electronic properties of the intermediary. Other arrangements of unpaired spins within a system are also possible. Ferrimagnetism occurs in systems incorporating two alternating effective spins (Figure l-4d)7b_ For example, in a chain containing alternating S= and S= 1 spin centers, the unpaired electrons align antiparallel. However, because the effective spin values are different, the moments do not cancel each other out completely, resulting in a net magnetization.9 In a canted antiferromagnetic system (Figure 1-4e)7b, the magnetic moment vectors of nearest neighbors are tilted. A small, but finite, magnetization results because the moments are not fully anti parallel and cancel each other out. 9 A distinction should be made between magnetic exchange and magnetic ordering. Magnetic exchange describes short-range correlations (the local interactions between unpaired electrons). On the other hand, in a transition to long-range magnetic order, the unpaired spins over a relatively large domain in a magnetic material will spontaneously align in the absence of an externally applied field at some critical temperature. 9 Alignment of the spins can be parallel or anti parallel corresponding to ferromagnetic or antiferromagnetic ordering, respectively. The critical temperature for ferromagnetic ordering is the Curie temperature, Tc, and for antiferromagnetic ordering, the Neel

PAGE 22

11 temperature, T N ,5,11 The size ofthis domain is the magnetic correlation length, i; the distance over which unpaired spins are ordered; the divergence of i; at a critical temperature is necessary for long-range ordering to occur in a magnetic material.2,11 In an ordered ferromagnetic state, the unpaired spins align in a parallel manner and a net magnetization results (Figure 1-5). In an antiferromagnetically ordered state, the unpaired spins align in an antiparallel fashion and no net magnetization results (Figure 1-5). 0 .............. 0 ()M ~ --------
PAGE 23

12 geometrical distribution of the spins in space. 9a When d = 0 the system consists of discreet zero-dimensional (0-D) oligomers or clusters. When d = 1 the system consists A t t t t B t + t + C t + t D !\!\ E Figure 1-4. The spin angular momentum vectors representing various interactions between unpaired electrons A) Paramagnetic. B) Ferromagnetic. C) Antiferromagnetic. D) Ferrimagnetic. E) Canted antiferromagnetic. Adapted from reference 7b.

PAGE 24

tttttt A tttttt II tttttt Li tttttt 13 B Figure 1-5 The spin angular momentum vectors in ordered magnetic states A) Ferromagnetic ordering B) Antiferromagnetic ordering Note the ordered ferromagnetic state will have a net magnetic moment in the direction indicated by the hollow arrow while the ordered antiferromagnetic state will possess no resultant moment. of one-dimensional (1-D) infinite chains and when d = 2 the system consists of infinite two-dimensional ( 2-D) sheets. The spin dimensionality n refers to the contributions by the vector components of the spin angular momentum. When n = 1 the spin has only one component S z, and the system is Ising When n = 2 the spin has two components S x and S y, and the system is Planar. When n = 3 the spin has three components, S z, S x, and S y, and the system is Heisenberg. 2 9 The concepts of spin and lattice dimensionality are especially important in determining whether a magnetic material can undergo long-range order.2 12-14 The magnetic ordering phenomenon can be understood by considering that the transition from short-range to long-range order in these systems is accompanied by a crossover in the effective lattice dimensionality or the effective spin dimensionality of the system.9 For example although no long-range order for a 2-D Heisenberg system is predicted 15 there

PAGE 25

14 are numerous e x amples of such systems that in fact undergo such a transition. 2 This discrepancy can be explained through a crossover in the lattice dimensionality The ordering occurs as a transition from a two-dimensional to three-dimensional lattice because at the critical temperature the interplanar couplings become important. 9 Magnetic exchange interactions are often not isotropic in all three lattice and/or spin dimensions. For highly anisotropic magnetic materials, the predicted magnetic behavior will no longer be Heisenberg but may resemble (for example) Ising and Planar behavior instead 9 A more general spin-Hamiltonian equation that accounts for exchange anisotropy is iI = -I(J x S ix s jx + J y S i y s jy + J z S iz s jz ) ( 1 -17) i > j When J x = J y = J =, e quation represents the isotropic Heisenberg model. If the components of J are different then the exchange is anisotropic. For example, when the exchange is principally characterized by in-plane components J x = J y, and J z = 0 the Hamiltonian is referred to as the XY model but if J x -:t:J y -:t:J z, equation is referred to as the XYZ model.2 16 Table l-19a 9b summarizes a few spin-Hamiltonian models for various cases of differing spin anisotropy and spin dimensionality 9 Note the subtle distinction between the Z model and the Ising model.9 For the Z model n = 3 and J x = J y = 0 but S x -:t:0 and S y -:t:0 meaning that although no exchange interaction occurs between nearest neighbor spins in the xand y-directions only in the z -direction the total spin angular momentum still has x y, and z components. In contrast the Ising model, n = 1 and J x = J y = 0 and S x = S y = 0 meaning that there is no exchange interaction in the xand y

PAGE 26

15 directions, only in the z-direction, and the spin angular momentum only has a z component. Table 1-1. Summary of Some Spin-Hamiltonian Models for Various Cases of Spin Anisotropy and Spin Dimensionality Spin-Dimensionality Interaction n=2 S/=S/ n=l s/ Adapted from references 9a and 9b. Model Heisenberg XY z Planar Planar Ising Ising Anisotropy in the exchange interaction often result from zero-field splitting or spin-orbit coupling effects.6 In each of these cases, the anisotropy is represented by additional terms in the Hamiltonian such as fl = n[s2 S(S + 1)] + E(S2 -Sl) ZFS z 3 x y (1-18) (1-19) Equation 1-18 represents the effect of zero-field splitting, where D is the axial or single-ion anisotropy factor and Eis the rhombic or in-plane anisotropic component. Equation 1-19 is the spin-orbit coupling Hamiltonian where A; is the spin-orbit coupling parameter and L and S are the orbital and spin angular momentum operators, respectively. Spin-orbit coupling arises from the coupling of a 2S+ 1 Fground state with an excited state from the same magnetic center.6 The excited states are well separated in energy from the ground state and are not appreciably populated at room temperature.

PAGE 27

16 d = I d = 2 d = 3 A B C t + n = l n = 2 n = 3 D E F Figure 1-6. Lattice dimensionality and spin dimensionality. A) A one-dimensional chain has d = 1. B) A two-dimensional sheet has d = 2. C) A three-dimensional lattice has d= 3. D) An Ising system has n = I. E) A Planar system has n = 2. F) Heisenberg system has n = 3. Spin-orbit coupling can lead tog-factor anisotropy or zero-field splitting effects. Zero field splitting describes the splitting of the Zeeman components in the absence of an external field due to the coupling of an S > ground state with excited states. 6 Magnetic anisotropy can arise from other sources as well, such as higher-order exchange interactions, orbital angular momentum contributions, and low-symmetry ligand fields or magnetic dipolar fields that couple the moments to certain directions of a crystal.9 However, the assignment of the origins of the anisotropy in the magnetic interactions is often difficult. Dimers A dimer, denoted by a lattice dimensionality of d = 0, represents the simplest type of interacting magnetic system. In a dimer, two spin centers can interact directly through

PAGE 28

17 space or indirectly via intermediary superexchange ligands. Only the second case is considered here. In general, no long range ordering is possible for dimers and other d = 0 oligomers and clusters (unless a crossover in lattice dimensionality occurs) and thus the magnetic interactions are only short-range correlations between the spin centers. Figure 17 shows the structure of an S = dimer, a dinuclear copper(II) moiety bridged by two diamagnetic ligands (X) capable of mediating a superexchange interaction between the two spin centers. Ancillary ligands (L) fill the remaining coordination sites on the metal ions. If the unpaired electrons interact with one another, then individual spin quantum numbers for each metal center, SA= S 8 =,are no longer valid.6 The spin states of the L L,1/X"'/L Cu Cu L/~/l"L L Figure 17. A square pyramidal copper(II) dinuclear complex where the metal centers are bridged by two diamagnetic ligands, X, and the remaining coordination sites are filled by ancillary ligands, L. dimer are now S = 0 and S = 1. In general, the energies of these spin states are not equal, but separated by an energy gap, J, defined as J = E(S = 0) E(S = 1 ). (1-20) The Hamiltonian for the isotropic exchange of a magnetic dimer is (1-21)

PAGE 29

18 where SA and Ss are the spin angular momentum operators representing the unpaired electrons on each metal center and J is called the isotropic exchange parameter that quantitatively accounts for the energy of the superexchange interaction. The second term is the Zeeman perturbation. Figure 1-8 shows the relation between the energy levels of the magnetic spin states as a function of applied field. 6 At zero field, the spin Hamiltonian splits the two degenerate S = states into the S = 0 and S = 1 states. In the presence of a magnetic field, the Zeeman term further splits the energy level of the triplet removing all degeneracy but does not affect the singlet state. When J < 0, the S = 0 singlet state is the magnetic ground state and the exchange is antiferromagnetic. In this case, the spins are coupled in an antiparallel fashion resulting in no net magnetic moment. When J> 0, the S = 1 triplet state is the ground state and the exchange is ferromagnetic. In this case, the spins are coupled in a parallel fashion resulting in a net magnetic moment. The magnitude of the coupling constant is related to the difference in energy, or energy gap, between the ground and first excited state. From both the spin and Zeeman Hamiltonians, the resulting four energy levels for an S = dimer as a function of external field are E1 = 0, E2 = J, E3 = J + sgB, and E4 = J sgB. The temperature dependence of the magnetic susceptibility, describing the changes in population of these magnetic energy levels, is given by 21 2NRNA;g2 expkBT X Dimer = 3 k T ----21B 1 + 3expkBT A simulation of the x vs. T from Equation 1-22 for an S = dimer with g = 2.2, J ks1 = -50 K, and+ 50 K is shown in Figure 1-9. Note that, when the exchange is antiferromagnetic, a maximum in the susceptibility is observed. (1-22)

PAGE 30

19 E B = 0 ~+ sgB S=l-~O A t-J -agB E B S=O--0 )Ill B B=O 0 S=O-t+J S=l~;agB -sgB )Ill B Figure 1-8. The splitting of the magnetic energy levels in an S =dimer. The zero-field spin states of the dimer correspond to the S = 0 (singlet) and S = 1 (triplet) states. An externally applied field splits only the triplet state. A) In an antiferromagnetically coupled dimmer, the ground state is singlet. B) In a ferromagnetically coupled system the ground state is triplet. The coupling constant, J, is the energy gap between the ground state and the nearest excited state. Dimers are among the most extensively studied magnetic systems.6 In particular, -dioxo bridged copper(II) dimers and -diazido copper(II) dimers have received considerable attention.7.17-21 The systems are convenient for the testing of theoretical models of magnetic systems. They also provide an understanding of how structural and electronic parameters affect the resulting magnetic properties. Finally, an understanding of the magnetostructural correlations for these simple systems provides a basis for continuing research efforts directed toward the design of novel magnetic materials with specifically tailored magnetic properties.

PAGE 31

20 Chains Magnetic chain compounds are one-dimensional systems with a lattice dimensionality of d = 1. Uniformly spaced magnetically interacting spin centers represent the simplest class of chains. Since the spin centers are equivalent along the chain, the nearest neighbor exchange interactions between the spin centers are also equivalent. Examples of uniform chains include [Cu(ox)]-H2O (ox= oxalate), (C6H11)CuCh,22,23 and [Ni(en) 2 (NO 2 )](ClO 4 ) (en= ethylenediarnine).24 Figure 1-10 schematically represents an S = uniform chain of Cu(II) ions. The spin Hamiltonian representing the isotropic nearest neighbor superexchange between the metal centers over n sites is (1-23) i=I When n is infinite, no analytical solution can be calculated in order to determine the energies of the magnetic spin states and the susceptibility. However, the energies and susceptibility can be calculated exactly for small chains of finite number of metal centers. Then, by extrapolating these results to the case of an infinite chain, numerical solutions for the energies of the magnetic states and susceptibility can be approximated.25,26 The temperature dependence of the magnetic susceptibility for an S = uniform chain, extrapolated from a ring of n = 11 spin centers, is N A g 2 ; 0.25 + 0.074975x + 0.075235x 2 x=~C.-~-----------=-----~ kBT 1.0 + 0.993lx + 0.l 72135x 2 + 0.757825x 3 (1-24) where (1-25)

PAGE 32

21 0 008 I Ferromagnetic I I --Antiferromagnetic \ 0 006 ,,..... \ \ 0 004 \ A ' ... l-'( 0 002 -------0 000 -0 002 0 60 120 180 240 300 T (K) 0 40 0 35 0 30 ---------,,..... .... 0 25 :::.:: 0 20 B 0 15 ..... l-'( 0 10 Ferromagnetic 0 05 --Antiferromagnetic 0 00 0 50 100 150 200 250 300 T (K) Figure 1-9. Temperature dependent magnetic susceptibility plots for an S = dinuclear complex (g = 2.10) modeled after Equation 1-22. A) zvs. T. B) z1 vs. T. The dotted lines represent ferromagnetically coupled dimers (J kn1 = 50 K) and the solid lines represent antiferromagnetically coupled dimers (J kn 1 = 50 K). J J ---Cu;--Cu;+ 1 ---Cui+ 2 --Figure 1-10. An exchange coupled, uniform S = chain of Cu(II) ions.

PAGE 33

22 Note that this equation is valid only for antiferromagnetic exchange along the chains (J < 0) since, as T approaches 0, then x converges to 0 if n is small and finite. If n is infinite, x does not converge to 0 but to a finite value since the ground and excited states form a continuum of energy levels with no energy gap between ground state and next highest energy level.6 When J> 0, then x diverges as T approaches 0. No corresponding analytical expression to describe the magnetic behavior of a ferromagnetically coupled uniform chain has been reported. 6 A high temperature series expansion, 27 valid for both positive and negative J values, to describe the magnetic susceptibility for a S = uniform chain is given by N A g 2 [ 1.0 + 5. 7980x + 16.9026x 2 + 29.3769x 3 ] X = 4k 8 T 1.0 + 2.7980x + 7.0087x 2 + 8.6539x 3 + 4.5743x 4 (1-26) where (1-27) Regardless of the ferromagnetic or antiferromagnetic exchange interactions that are present along the chains, in principle, the isolated one-dimensional chains magnetically order only at T= 0. However, in real solids, chains are never completely isolated but experience interchain interactions, usually much weaker than the dominant intrachain exchange due to a crossover in lattice dimensionality. At low temperatures, these interchain interactions become important, the one-dimensional chains effectively behave as three-dimensional solids, and magnetic ordering occurs at finite temperatures. A more complicated type of one-dimensional system is the alternating, or zig-zag, chain. In this system, there are two distinct types of spin centers and, as a result there is

PAGE 34

23 a regular alternation of the exchange interactions, J and J', representing the nearest neighbor and next nearest neighbor couplings, respectively. Examples of alternating chains include Cu(NO 3 ) 2 /2H 2 O 28,29 and (ipa)CuCh (ipa = isopropylammonium)30,31. For instance, Figure 1-11 schematically represents an S = alternating chain of Cu(II) ions. J aJ --Cu 2 ;_ 1 --Cu 2 ;--Cu 2 ;+ 1 -Figure 1-11. An exchange coupled, alternating S = chain of Cu(II) ions. The spin-Hamiltonian representing for an alternating chain system is (1-28) where a is the alternation parameter such that 0 S a S 1 such that J' = al. Note that when a= 0, the one-dimensional system behaves as isolated magnetic dimers, and when a= 1, the system corresponds to a uniform chain. Again, when n is infinite, no analytical solution can be used to determine the energies of the magnetic spin states and the susceptibility for the alternating chain. However, analytical solutions can be obtained in a similar fashion for uniform chain. The temperature dependence of the magnetic susceptibility for an S = alternating chain extrapolated from a ring of n = 10 spin centers 32 is where (1-29)

PAGE 35

24 (1-30) The coefficients A F are functions of a and are provided elsewhere. 6 This equation is valid only when both exchange interactions are antiferromagnetic JS O and both exchange parameters are within an order of magnitude of one another. In contrast to uniform chains, when J < 0 and O S a < 1 then x converges to O as T approaches zero since an energy gap between the ground and excited magnetic states is present. 6 Ladders Ladders, low-dimensional quantum systems that fall in between one-dimensional chains and two-dimensional sheets, consist of a finite number of magnetically coupled chains of spins (Figure 1-1).33 ,34 In principle, one might expect that a smooth crossover in physical properties from chains to sheets would result if one assembled chains to progressively form ladders of increasing width but this is not generally true. 3 3 The spin-Hamiltonian that, in general, represents ladder-like magnetic systems is (1-31) a= l 2 i = I i= I where S ;, a represents the spin operator at site i (i = 1, 2 ... .... n) on the leg a (a = 1,2 ... . ) of the ladder with n rungs 33 The terms J .1 and Jj J denote the intraand interrung exchange couplings respectively. In "ideal" ladders, the magnitude of the coupling along the legs is comparable to the magnitude of the coupling along the rungs J .1 I Jj 1 ~ 1 When J.1 I Jj 1 tend to zero the exchange between the rungs is small compared to the exchange along the legs and the ladder behaves as a system of isolated chains. Conversely, when J .1 I Jj 1 tend to infinity, the exchange along the legs is small compared

PAGE 36

25 to the exchange along the rungs and the ladder behaves as a system of isolated dimers. Ideal spin ladders should also be well isolated from one another since appreciable interladder coupling (J') can cause transitions from the spin liquid ground state to a magnetically ordered state.33 The magnetic properties of ladders with an even number of legs are drastically different from those with an odd number of legs.35 Ladders with an even number of legs (Figure 1-1 ), such as the inorganic cuprate SrCu2O3, are characterized by short-range spin correlations along the legs, a spin-liquid ground state.33,35,36 Even-leg ladders consist of spin-singlet pairs with a spin-spin correlation length along the legs that show an exponential decay produced by the presence of a finite spin gap and tends to 0 as T approaches 0. In contrast, a ladder with an odd number oflegs, such as Sr2Cu3Os, exhibits power-law decay of spin-spin correlations that tend to finite values as T approaches 0 that are magnetically ordered due to the presence of gapless spinless excitations. 3 3 ,3 5 ,36 Ladders with an even number of legs are characterized by a spin gap.33,35 A spin gap is a finite energy gap between a nonmagnetic ground state and the first excited triplet state (Figure l-l 2a). No continuum of excited states exits directly above the ground state. If the rungs of an antiferromagnetic ladder interact weakly with one another, i.e J.1. I Jj1, the ground state has a total S = 0 since the spins on each rung are in a singlet state.33 To promote the ladder to the lowest excited state with a total S = 1, one of the pairs of spin singlets of the rungs must be promoted to an S = 1 triplet. A quantum of energy, called the spin gap energy, AEsg, is required to excite one of the rung singlets into a triplet state. A frustrated spin state results since the spins are now aligned parallel

PAGE 37

26 along one direction but anti-parallel along the other direction (Figure 1l 2b ). Examination of the magnetization of a ladder compound as a function of changing external applied field can identify a spin gap. An abrupt increase in magnetization at a particular magnetic field indicates the presence of the energy gap (Figure 1-12c). There has recently been a growing interest in the preparation and study of ladder like molecular and solid-state materials. In the field of supramolecular chemistry, ladders represent one of many familiar structural topologies that are possible from the self assembly of simple molecular or ionic nodes and spacers, such as metal ions or complexes and multifunctional bridging ligands, respectively, under certain stoichiometric ratios and reaction conditions.37 Furthermore, molecular and solid-state ladders, as a consequence of their structure, often possess open or enclathrated cavities and extended channels that exhibit unique inclusion and catalytic phenomena.37 In the field of low-dimensional materials, ladders represent a structural intermediate between one-dimensional chains and two-dimensional sheets. 3 3 Ladders represent ideal systems to investigate the gradual change in physical properties as the dimensionality increases from 1 D chains, to quasi 1 D / 2D systems, to 2D sheets. Furthermore, copper oxide ladders are part of the structure of many solid-state materials such as (Sr, Ca)Cu 2 O 3 that, upon doping with holes, often exhibit superconductivity at liquid nitrogen temperatures or higher.35,38 These copper oxide ladders are antiferromagnetic and it is believed that the superconductivity originates and is sustained within this portion of the structure. In order to better understand the origin and mechanism of high-temperature superconductivity, it is desirable to synthesize model low-dimensional compounds that adopt a ladder-like structure similar to those found in the cuprates.

PAGE 38

27 Self-Assembly of Supramolecular Architectures The recent research efforts devoted to the rational design and crystal engineering of suprarnolecular solid-state materials were initially spawned by a concerted interest toward developing methods for predicting the crystal structures of organic compounds.39 However ongoing work in this field has continued due to the prospect of developing new materials with interesting structures and diverse exploitable properties One particular area in suprarnolecular chemistry has focused on synthesizing hybrid organic / inorganic materials through the self-assembly of simple molecular or ionic building blocks. Compared to purely organic or inorganic analogous systems these materials often possess improved thermal chemical and mechanical stability and exhibit unique or enhanced physical properties 40 Hybrid organic/ inorganic materials are potentially useful in a wide variety of applications including low temperature catalysis 41-43 inclusion phenomena,37 44 45 magnetism 40 46-49 electrical conductivity 50-52 photochemistry 53 and second-order nonlinear optical behavior.54-57 Strategies for Building Supramolecular Architectures Under certain conditions solution-phase molecular and ionic building blocks can self-assemble into discreet clusters or oligomers or extended one, two-, and three dimensional solids sustained through various types of chemical interactions such as coordinate covalent bonding 43 58 electrostatic attractions 59,60 hydrogen bonds 61-65 and Jl'-stacking. 66-68 In general the overall molecular and solid-state structure is controlled by a combination of the binding constraints geometrical preferences and relative stoichiometric quantities of these building blocks. Therefore much research has been devoted to the development of general strategies to better predict design and

PAGE 39

28 control the structure of supramolecular architectures with the above guidelines in mind. By far the simplest and most common strategy applied is the node and spacer method. The node defines the overall geometry of the structure while multifunctional or multitopic spacer ligands are tethered to and propagate the geometrical preferences of node throughout the solid.69,70 These simple, modular components are chosen as starting materials because their inherent bonding and geometrical propensities allow some degree of control and predictability over the structure of products.39 Furthermore, by the careful selection or design of these building blocks, the physical properties of the solids can be ''tuned" or ''tweaked" to meet specific needs.39 A number of different approaches derived from the node and spacer strategy can be applied toward the design of supramolecular materials based on the nature of the building blocks or the chemical interactions responsible for sustaining the structure. The most common approach is the generation of hybrid organic I inorganic networks that are simple extensions of a specific transition metal or metal complex (metal center chelated by one or more poly-hapto ligands) geometry.39 The metal coordination environment functions as the node and the spacer ligands are typically bridging ligands. In most cases, the molecular and solid-state structure of the assembly is sustained through coordinate covalent bonding and thus the structure is often referred to as a coordination polymer. A variety of one-, two-, and three-dimensional coordination solids with novel topologies have been obtained using rigid, multifunctional spacer ligands such as pyrazine,71, 72 4,4' -bipyridine, 44,7379 4,4 '-azobis(pyridine), 80, bis( 4pyridyl)benzene, 81 bis-( 4-pyridyl)-ethylene, 81 2,4,6-tris( 4-pyridyl)-1,3,5-trazine, 82-85 1,3,5-tris( 4-ethynylbenzonitrile )benzene, 86 and 1,3,5-benzenetricarboxylic acid45,87

PAGE 40

29 Excited Triplet States A #f Singlet Ground State -B M C ~LIB g B Figure 1-12. Spin Gap. A) The energy gap, L1Eg, between the singlet ground state and excited magnetic spin states in a two-leg ladder. B) The excitation of one of the strongly coupled rungs into a triplet state results in a frustrated spin state. C) The spin gap can often be identified by an abrupt increase of the field dependent magnetization.

PAGE 41

30 with transition metal cations such as Cu +, Cu 2 +, Ag + Zn 2 +, Cd 2 +, Mn 2 + Co 2 +, Fe 3 +, and Ni2 + Predictable and structurally well-defined products are often produced from the self assembly of rigid spacers with metal cations. In fact rigid spacers are particularly ideal for designing and synthesizing porous network solids capable of clathrating small molecules. On the other hand flexible spacer ligands have not been extensively exploited except in a few cases.44 81,88 Flexible spacers may allow the synthesis of hybrid organic I inorganic solids with structural features not found in those materials with rigid linking groups.89 90 Unfortunately the incorporation of a higher degree of flexibility into the building blocks reduces the amount of predefined geometrical information from the reactant components and a multitude of different structures can arise from identical metal-ligand combinations or minor experimental variations.91 Hydrogen bonding interactions between nodes and spacers offer an alternative but equally powerful, approach to the control of solid-state structures. Molecular components with complementary hydrogen-binding sites are well known and can be readily incorporated into building blocks to produce extended polymeric architectures sustained by interactions either directly between nodes or mediated by spacer ligands61 Because these interactions are directional and their formation is often reversible hydrogen bonded molecular assemblies are ideal for the design of structurally flexible networks with and adjustable size pores. 61,62 Obviously networks can be also assembled by the concurrent action of both coordinate covalent and hydrogen bonds as well, though species like these are relatively less common.92 An alternative approach to the metal node and ligand spacer method involves the exploitation of exodentate multitopic ligands that act as both the nodes and spacers in the

PAGE 42

31 network architecture. This method is exemplified by the construction of supramolecular architectures from solely organic molecules and has generated a number of architectures mimicking the structures of known inorganic minerals.70,84,86,93 However, this method is also less predictable in terms of rationally developing functional topologies. 70 Factors Affecting the Structure of Supramolecular Architectures A number of different factors can profoundly affect the molecular and solid-state structure of self-assembled coordination polymers, such as the stoichiometric ratios of the node and spacer, the different oxidation states and coordination preferences of the metal ion nodes, the structural and bonding propensities of spacer ligands, and the reaction solvents. The metal:ligand stoichiometry is one of the most important factors determining the dimensionality and limiting the possible topological architectures that can occur in a coordination solid. Consider the possible types of structural motifs, or supramolecular isomers,94 that result from the combination of various spacer ligands with metallic moieties. A 1 : 1 ratio limits the architectures to either one-dimensional linear or zig-zag chains. 64,69 ,89 ,95-102 A 1 :2 ratio can produce two-dimensional grids with rectangular cavities if the node is planar or octahedral 40,44,74,75,92,95,103-105 or three-dimensional diamondoid structures if the metal center is in a tetrahedral or S4 environment.79,106-108 A metal:ligand ratio of 1 :1.5 can produce any of six very different architectures including the molecular ladder,70,88,94,109 brick wall,74,88 Lincoln log,73,110 tongue and groove or bilayer,70,76,111 herringbone,112 or three dimensional framel 13 if the metal moieties adopt only square planar or octahedral geometries. Additionally a trigonal metal center can generate two-dimensional hexagonal or honeycomb nets with a three-fold symmetric tritopic spacer such as 1,3,5

PAGE 43

32 trisubstituted benzenes in a 1: 1.5 metal:ligand ratio. 86,106 The molecular railroad has been the only topology observed for a 1 :2.5 stoichiometry but, unlike the ladder, the spacers are present as both bridges and terminal ancillary ligands.114, 115 Three dimensional rectangular and interpenetrated grids, are possible with metal:ligand stoichiometries of 1 :3 or smaller.116-119 However, given the ubiquity of octahedral coordination environments, it is somewhat surprising that simple three-dimensional octahedral polymers remain largely unexplored.117 Figure 1-13 schematically depicts selected supramolecular architectures built from the self-assembly of nodes (metal ions) with spacers (multifunctional bridging ligands). Many of these topologies, such as the two-dimensional sheets, incorporate cavities and channels as part of the molecular and solid-state structure. In a few cases, the pores are open 40 but the void space is usually filled either by clathrated guests or interpenetration of neighboring lattices. This packing diversity between interpenetrated and noninterpenetrated porous solids can be identified as another form of supramolecular isomerism.115 The oxidation state and the coordination preferences of transition metals are also critical in determining the final molecular and solid-state structure. For example, Cu 2 + normally prefers an axially distorted octahedral geometry and is known to form two dimensional square networks with pyrazine or substituted pyrazine ligands 71,120-122 and interpenetrated two-dimensional networks with 4,4 '-bipyridine.43 However, trigonal and tetrahedral Cu + cations can form two-and three-dimensional networks when bridged by bipy, pyrazine, or substituted pyrazine ligands.107,120,123,124 Ag+ is observed in an

PAGE 44

33 even wider range of coordination environments including linear,41, 110 trigona1,41 tetrahedral,41,108,125 square-planar,126 square pyramidal,126 and octahedral126 when bound to pyrazine or bipy. The presence of coordinated or lattice solvent molecules can dramatically affect to final structure. For instance, in [Zn(4,4'-bipy)2(H2O)2]SiF6,74 the coordination of solvent water molecules produces an interpenetrated sheet-like structure while a porous solid, [Zn( 4, 4 '-bipy ) 2 (H 2 O ) 2 ] SiF 6 xD MF, 11 7 is isolated under nonaqueous conditions. A similar solid-state structural difference is observed between the double-layered structure of the solvent inclusion compound [Ag(pyrazine)2][Ag(pyz)s](PF6)3-2S (S = CH2Cli, CHCh, and CCI.i)126 and the single-layered structure of the solvent-free compound [Ag(pyz) 2 ](PF 6 )_125 Counterions can also influence the final structure. In the previous example, by replacing the PF 6 counterions with SbF6, the characteristic double-layer structure of Ag-pyraizine-PF 6 changes to the three-dimensional noninterpenetrating cubic framework Ag(pyz) 3 (SbF 6 )126 while substituting with BF 4 ions changes the layered sheet-like structure of [ Ag(pyz ) 2 ] (PF 6 ) 125 to the interpenetrated three-dimensional structure [Ag2(pyz) 3 ](BF4)2.41 Porous Network Supramolecular Architectures The design and construction of hybrid organic / inorganic porous network solids is a particularly important and emerging area of supramolecular chemistry providing new generations of functional materials.42,60,127,128 The great importance of porous solids in inclusion phenomena, such as adsorption / desorption, ion exchange, size and shape selective molecular sieving, and catalysis, is due to their ability to reversibly clathrate or

PAGE 45

34 trap species within their cavities and extended channels.129 These pores possess a variety of sizes and shapes not observed in analogous inorganic porous solids such as zeolites and molecular sieves.87 Furthermore, by a careful selection and design of the chemical components, the size and clathration properties of these pores can be designed and fine-tuned to meet specific needs while maintaining the overall structural and functional features of naturally occurring analogs. 44 Like zeolites, most hybrid organic / inorganic coordination polymers bind guests within pores, cavities, and channels that are part of their lattice framework. In fact, a number of such coordination networks have been found to exhibit many other desirable zeolitic properties as well, such as stability and porosity of the framework, 130 guest exchange,131,132 and selective catalytic activity.44 Recent examples porous solids include the diamondoid, honeycomb, rectangular grid, ladder, brick wall, and octahedral frameworks constructed from tetrahedral, trigonal, and octahedral metal templates (Zn(II),. Cd(II), Ag(I), and Cu(I)) and various multitopic spacer ligands.44,73,133-135 The reversible absorption and desorption of guests without the collapse of cavities or channels, while well known in materials such as zeolites, is a much less common phenomenon in molecular porous solids.136-138 Upon loss of their guests, clathrated hosts irreversibly lose crystallinity,93 undergo phase changes, 139 or alter their morphology without the simultaneous replacement of substitutes.13 9,140 Few examples exist where cavities that have been collapsed due to crystal packing forces from guest loss are restored upon guest binding, however, it has recently been shown that some coordination or hydrogen-bonded networks can rapidly exchange inclusions or

PAGE 46

35 counterions while maintaining crystal integrity.42,44,86,141,142 Furthermore, the construction and access of large pores in either coordination polymers or hydrogen bonded assemblies is often mitigated or precluded by self-inclusion or lattice interpenetration particularly if the void created by the open pores occupies more than 50 % of the crystal by volume.43,107,109,143 Interpenetrated structures are oflimited usefulness in inclusion chemistry but have significant potential in terms of other enhanced bulk properties, such as improved chemical, mechanical, and thermal stability. 3 7, 109 In order to design and construct functional porous solids with useful inclusion capabilities, a number of requirements should be met while accounting for the problems detailed above. The host framework should be rigid and robust, containing large, accessible cavities or channels capable of reversible guest binding in a stoichiometric manner_ 14 2 Furthermore, the structural integrity of the pores should be maintained in the absence of clathrates.138,142 The term "capable", in this context, implies that the van der Waals surfaces and electrostatic potential surfaces of the host pores and guest should also be complementary.37,39 The robustness of a channel or cavity is the ability to allow reversible release and adsorption of guest without the collapse of the host structure.144 In order to address these requirements, one approach focuses on the utilization of chemical interactions in two or three dimensions to maintain the integrity of the host structure after removal of the clathrated guests.134 Porous three-dimensional networks sustained by strong coordinate covalent bonding often retain the vacant cavities and channels even after removal of the guest molecules without structural changes, such as

PAGE 47

36 I I ff'" -MI M-M I A -MI I M-M v~l -MI I B C I I I 0 r-M-r-r ,-M-,-, i r-M-,-, ,-M-r M-,-, ,-M-r-M -r-r-r _i_i_i_ 1 ,-M-r-M -,-,-rM---M-M-M H F I I -r, _i_i_ -,-,E Jv G I I M M '-./ L l L / I I I I I I I ,-r -M-r-M-,--r-M-r-M-r -M-,-M-r-M-1-i ,-, I Figure 1-13. Examples of supramolecular architectures from the self-assembly of metal ions, M, with spacer ligands (lines). A) Linear chains. B) Zig-zag chains. C) Diamondoid. D) Herringbone sheet. E) Square grid sheet. F) Railroad. G) Honeycomb. H) Brick wall sheet. I) Ladder.

PAGE 48

37 lattice interpenetration, at ambient temperature.70,76,145,146 On the other hand, flexible pores sustained by weaker hydrogen bonding interactions may change in size in response to the uptake or loss of guests to ensure that void space is efficiently occupied thus avoiding self-inclusion. 62 Another approach involves the use of interpenetration to produce porous solids with robust cavities and channels.77,144 This method may, at first, seem self-defeating since the interpenetration of neighboring lattice frameworks often completely fills void space thus preventing the formation of extended cavities and channels.42,69,73,84,107,108,110 In fact, most strategies aimed at synthesizing porous solids have been directed toward the inhibition of interpenetrated networks. However, if the spacer is of sufficient length the self-inclusion may only reduce the size of, but not completely fill, the pores, leaving small voids for small-molecule inclusion.144 The lattice interpenetration can therefore afford rigid, three-dimensional networks with the desired robust, albeit smaller, channels. Despite the large amount of research and work effort devoted toward the synthetic aspect of solid-state supramolecular chemistry in terms of producing functional materials and elucidating methods for the rational design and fabrication of such materials, studies of the chemical reactivity of these materials has been lacking.14 7 In fact most of the studies have been limited to inclusion properties of clathrated porous solids, such as the absorption-desorption processes and guest exchange of guest molecules.45,87,142,144,147 This imbalance of synthetic work compared to chemical reactivity studies of supramolecular materials is largely due to the fact that, unlike the well characterized chemical reactivity properties of molecular and ionic species which

PAGE 49

38 are generally soluble in one or more common solvents, most coordination polymers are insoluble in most common organic and inorganic solvents thus rendering any reactivity studies difficult at the very least.14 7 Scope of the Dissertation The work presented in this dissertation, Chapters 2, 3, and 4, focuses on, in general, investigating the structural and physical properties of representative solid-state materials obtained throughout the course of this graduate research. In Chapter 2, the structural, thermal, and magnetic properties of a series of clathrated porous network solids are described in detail. Chapter 3 describes the host-guest properties of these porous solids determined through gas chromatography, solid-state NMR spectroscopy, and X-ray diffraction. Chapter 4 details the structural and magnetic properties of a series of ladder-like azido bridged Cu(II) coordination polymers. Finally, the structures of selected coordination polymers that did not quite fit within the context of the themes of the chapters are presented in Appendix A.

PAGE 50

CHAPTER2 STRUCTURAL, THERMAL, AND MAGNETIC PROPERTIES INVESTIGATION OF THREE TRANSITION METAL-4,4'-BIPYRIDINE COORDINATION POLYMERS: [Ni( 4,4 '-bipy)J(H2O)2](ClO4)2 1.4( 4,4 '-bipy) 3(H2O), [Co( 4,4 '-bipy)3(H2O)2](ClO4)2 1.4( 4,4 '-bipy) 3(H2O), and [Cu( 4,4' -bipy)3(DMSO)2](ClO4)2( 4,4' -bipy) Introduction The rational design and synthesis of functional organic/inorganic network solids has recently been the focus of intense research in materials chemistry. Much of the interest in this field is driven by the wide variety of potential applications such materials can afford, including host-guest chemistry,44 ion exchange,37 molecular sieving,37,44,45 catalysis,41-44 non-linear optics,54-57 electrical conductivity,50-52 and magnetism.40,46-49 Often such materials are designed and built from simple, modular components such as a metal ions and organic spacer ligands. The spacers are typically multifunctional ligands capable of coordinating to the metal ions and propagating structural information dictated by the coordination requirements and geometry of the metal sites throughout a solid. 69 The assembly of these components through intermolecular forces e.g. donor-acceptor interactions,148,149 hydrogen bonds,61,63,64 and .1r-stacking66 can produce solids whose molecular and crystal structure can be profoundly affected by a number of factors, most notably the metal:spacer stoichiometry. For example, transition metal ions combined with linear organic spacers can assemble to form structures resembling linear 64,69,95,96,100,101 or zigzag chains, 69,78,99 39

PAGE 51

40 molecular ladders 70,88,94,109 two-dimensional gr ids 40,44,47,40,69,74,92,100,104,105,150,151 railroads 114,115 or three-dimensional networks,152,153 if the metal:spacer stoichiometry is 1 :1, 1 :1.5, 1 :2, 1 :2.5, or 1 :3, respectively. One commonly employed spacer is the bifunctional heterocyclic molecule 4,4' bipyridine and a number of coordination polymers with different network architectures in the solid state have been reported incorporating this ligand.40,44,69,70,101,105,109,114,151 Examples of one-dimensional chains include [Co(SO4)(H2O)3(4,4' -bipy)]-2(H20),96 [Ni(Et-XA)2(4,4' -bipy-)]-0.5(EtOH)-(CHCh) (Et XA = ethylcarbonadithiolate),98 [Co(NCS)2(H 2 O) 2 (4,4' -bipy)]-(4,4' -bipy),95 and [Mn(hfac ) 2 ( 4,4 '-bipy)] (hfac = hexafluoroacetylacetonato ), 97 while [Ni( 4,4 bipy)2_ 5 (H 2 O) 2 ](ClO4)2 l.5( 4,4' -bipy)(H20)11 4 and [Co( 4,4' -bipy)u(NO3)2]-Guest (Guest= MeCN or CHCh)109 form railroad and ladder-like structures, respectively. Examples of two-dimensional grids or sheets include [Cd(4,4'[Co(NCS)2(4,4'-bipy)2]-2(CH3CH2)2O,95 [M(4,4'-bipy)(ox)] (M = Fe(II), Co(II), Ni(II), and Zn(II) and ox= oxalato),40 and [M(4,4'-bipy) 2 (H 2 O) 2 ](ClO 4 )2"Guest (M = Cu, Zn, and Cd and guest= enclathrated guest molecule).133 Chapter Summary This chapter reports the crystal structures, thermal behavior, and magnetic properties of a series of three linear chain compounds containing the 4,4 '-bi pyridine spacer that organize in the solid state to form new non-interpenetrated network solids.

PAGE 52

41 The compounds [Ni( 4,4 '-bipy)J(H2O)2](ClO4)2 1.4( 4,4 '-bipy)-3(H2O) 1, [Co( 4,4 '-bipy)J(H2O)2](ClO4)2 1.4( 4,4 '-bipy) 3(H2O) 2, and [Cu( 4,4 bipy)3(DMSO) 2 ](ClO 4 ) 2 (4,4'-bipy) 3, were each prepared by the direct combination of three moles of 4,4 '-bipyridne with one mole of their respective metal ion. Compounds 1 and 2 are isostructural and were crystallized under inert atmosphere, hydrothermal conditions. The related structure 3 was isolated under ambient laboratory conditions from dimethyl sulfoxide. All three materials share a common structural motif with one dimensional, covalently linked chains interacting via hydrogen bonding and .1l"-stacking forces to form layered sheets with characteristic hydrophobic, rectangular cavities. These sheets are packed in a manner that aligns the cavities to form oblique channels occupied by enclathrated guest molecules and counterions that extend throughout the solid. The thermal instability of these coordination polymers is associated with the relatively low temperatures at which the guest molecules are lost. Temperature and field dependent magnetization measurements revealed weak magnetic coupling between the paramagnetic metal centers as 4,4 '-bipy is a poor mediator of superexchange interactions. 97,98, 154 Experimental Section Materials Copper(II) perchlorate hexahyrdate (98 %), cobalt(II) perchlorate hexahydrate (98 % ), nickel(II) perchlorate hexahydrate (98 % ), 4,4 '-bi pyridine (98 % ), and sodium azide were purchased from Aldrich (Milwaukee, WI). Dimethyl Sulfoxide (99.9 %) was purchased from Fisher Scientific (Pittsburgh, PA). Ethanol (100 %) was purchased from Aaper Chemical (Shelbyville, KY). All reagents were used without further purification.

PAGE 53

42 Synthesis of [Ni( 4,4 '-bipy)3(H20)2](CI04)21.4( 4,4 '-bipy)(H20) A solution was prepared by dissolving 731 mg of Ni(ClO 4 )2 H 2 O (2.0 x 103 mol) in 10 mL of water contained within a Teflon canister. Addition of 934 mg of 4,4' bipyridine (6.0 x 103 mol) and 2 mL of ethanol to this solution resulted in a blue-green colored suspension. This canister was sealed within a homemade, stainless steel hydrothermal vessel, purged with argon gas, and heated to 150 C for 5 days. The container was subsequently cooled to room temperature over a period of 24 hours without any specific control over the rate of cooling. The resulting blue-green crystals, obtained in 88 % yield (based on initial quantity of bipy), were washed with water before further characterization. The crystals become opaque within a few hours upon removal from the solvent. Elemental analysis calculated for NiC4sfu~9O13Ch: C, 51.44 %; H, 4.42 %; N, 12.00 %. Found: C, 50.88 %; H, 4.47 %; N, 12.14 %. Synthesis of [ Co( 4,4 '-bipy )J(H20)2] ( CI04)2 1.4( 4,4 '-bipy 3(H20) Using the same procedure described for 1, 731 mg ofCo(ClO4)2 H2O (2.0 x 10 3 mol) was reacted with 934 mg of 4,4' -bipyridine (6.0 x 10 3 mol) in 10 mL of water with 2 mL of ethanol at 150 C for 3 days. The resulting orange crystals, obtained in 92 % yield (based on initial quantity of bipy ), were washed with water before further characterization. The crystals become opaque within a few hours upon removal from the solvent. Elemental analysis calculated for CoC4sfu6N9O13Ch: C, 51.43 %; H, 4.42 %; N, 11.99 %. Found: C, 50.87 %; H, 4.47 %; N, 12.14 %. Synthesis of [ Cu( 4,4, -bipy )3(D MSO)i] ( CI04)2 2( 4,4 '-bipy) A solution containing 741 mg of Cu(ClO 4 )i-6 H 2 O (2.0 x 10 3 mol) dissolved in 10 mL ofDMSO was combined with a solution containing 934 mg of 4,4' -bipyridine (6.0 x 10 3 mol) dissolved in 10 mL ofDMSO. The resulting dark blue colored mixture,

PAGE 54

43 contained within an evaporating dish, initially produced small blue block-like crystals within two weeks of solvent evaporation. Within and additional four weeks, light blue colored hexagonal plates of 3 were isolated and washed with DMSO before further characterization. X-ray Structure Determination A blue-green crystal of 1 (0.25 x 0.23 x 0.23 mm\ an orange crystal of2 (0.51 x 0.36 x 0.17 mm\ and a blue crystal of 3 (0.24 x 0.21 x 0.12 mm 3 ) were selected for ray analysis. Each crystal was mounted on a glass fiber under nitrogen gas. The same data collection procedure was used for each sample. Data were collected at 173 K on a Siemens SMART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing MoKa radiation (A= 0.71073 A). Cell parameters were refined using up to 8192 reflections. A full sphere of data (1850 frames) was collected using the ro-scan method (0.3 frame width). The first 50 frames were re-measured at the end of data collection to monitor instrument and crystal stability (maximum correction on I was < 1 % ). Absorption corrections by integration were applied based on measured indexed crystal faces. The structures were solved by the Direct Methods in SHELXTL6 155 and refined using full-matrix least squares. Structures 1 and 2 were solved and refined in space group C2/c while 3 was solved and refined in Cc, which afforded better results. The stoichiometry is the same in 1 and 2 ([M(4,4' -bipy)J(H2O) 2 ](ClO 4 ) 2 l.4(4,4' bipy) 3(H2O) where M = Ni or Co) but different from 3 ([Cu( 4,4 bipy)J(DMSO)2](ClO4)2(4,4'-bipy)). The non-H atoms were treated anisotropically, whereas the hydrogen atoms were calculated in ideal positions by riding on their

PAGE 55

44 respective carbon atoms. The H atoms from the coordinated water molecules were found and refined. The uncoordinated 4,4 '-bipyridine molecules and perchlorate anions are disordered in each structure. In 1 and 2, two half-bipy guest moieties are each disordered about a center of inversion. The perchlorate anions are disordered in four parts but the oxygen atoms on only two major components were found in Difference Fourrier maps and refined anisotropically. Only the Cl atoms of the minor disordered parts were found and refined. In 3, a single guest 4,4 '-bipy has one pyridyl ring disordered. The S atom of one coordinated DMSO molecule is disordered as well. The site occupation factors of the disordered parts were dependently refined to 0.88(1) for the major part and consequently 0.12(1) for the minor part; S' was refined with an isotropic thermal parameter. For both perchlorate anions, disorder was found each in two positions and their site occupation factors were dependently refined to 0.69(1) and 0.31 (1) for one anion, and 0.50(1) for each part of the second disordered anions. A total of 428 and 427 parameters were refined employing F 2 in the final cycle using 4328 and 4393 reflections with I> 2cr(I) to yield R1 of 7.72 % and 7.79 % and wR 2 of 21.58 % and 22.38 % for 1 and 2, respectively. A total of 755 parameters were refined employing F 2 in the final cycle using 10055 reflections with I> 2cr(I) to yield R 1 and wR 2 of 5.62 % and 12.61 %, respectively, for 3. Thermal Analysis Thermogravimetric analyses (TGA) of the title compounds were performed on a computer-controlled Hi Res TGA 2950 thermogravimetric analyzer. Powdered samples of 1 (4.5740 mg), 2 (3.7460 mg), and 3 (6.3730 mg) were loaded into alumina pans and heated with a ramp rate of 10 C/min from room temperature to 600 C. Thermal

PAGE 56

45 desorption mass spectrometry measurements were recorded on a MAT 95 utilizing electron ionization techniques. Crystalline samples contained within capillary tubes were evacuated loaded into direct insertion probes and heated with a ramp rate of 10 C / min from 30 C to 400 C. Magnetic Measurements Bulle magnetization measurements were obtained from a standard Quantum Design MPMS SQUID magnetometer. The samples consisted of randomly oriented single crystals with a total mass of 32.3 mg for 1 19.4 mg for 2, and 51 0 mg for 3. A gel cap and plastic straw were used as the sample holder during the measurements. Magnetization versus temperature measurements were run from 2 K to 300 K The sample was zero-field cooled to 2 K before a measuring field of 1000 G was applied and the data set was then taken while warming the sample from the lowest temperature Magnetization versus field measurements were performed at 2 K from 0 to 50 kG The background signals arising from the gel cap and straw were measured independently and subtracted from the results. The diamagnetic contribution of each sample estimated from Pascal s constants ( x.n = 446 x 106 emu mol 1 for 1 and 2 and Xo = 516 x 106 emu mor 1 for 3) was also subtracted from the results.5,6 ESR spectra were recorded with a Bruker ER 200D spectrometer modified with a digital signal channel and digital field controller. Data were collected using a U S. EPR SPEC300 data acquisition program and converted to ASCII format using a U.S. EPR EPRDAP data analysis program.

PAGE 57

46 Results and Discussion Compound Synthesis The network coordination polymers 1, 2, and 3 were synthesized by the direct combination of one mole ofM(ClO4)2H2O (M = Ni(II), Co(II), or Cu(II)) with three moles of 4,4 '-bipyridine in solution. Note that the products contain more than three equivalents of bipy, some present as enclathrated guest molecules in addition to the coordinated ligands. Since the bipy ligand is relatively insoluble in water, the reaction conditions afforded by the hydrothermal technique are essential for the crystallization of 1 and 2 where the metal-bipy suspension is dissolved by the high temperature (150 C) and pressure conditions within the vessel. In order to prevent the formation of unwanted side products, such as high oxidation state metal oxides particularly with cobalt, the reaction mixture was heated for no longer than 5 days, purged with argon gas, and treated with small quantities of ethanol (acting as a mild reducing agent). The products are sensitive to loss of solvent and become opaque within a few hours upon exposure to air. For 1 and 2, small changes in the metal-bipy stoichiometry (for example, by using 2.5 or 3.5 moles ofbipy per mole ofNi(II) or Co(II)) always produced the same products. Single crystals of 3 were obtained under normal laboratory conditions by crystallization from DMSO. Similar attempts employing hydrothermal synthesis resulted in the formation of impure powders. Crystalline products of 3 were obtained only several weeks due to the slow rate of evaporation of D MSO. Unlike 1 or 2, crystals of 3 do not appear to be sensitive to loss of solvent. Furthermore, a secondary product crystallizes from solution in addition to 3. During the first two to three weeks of crystallization, small blue blocks first appear followed by the crystallization of blue hexagonal plates of

PAGE 58

47 3 after an additional month or more. These blue blocks were determined to be an extended three-dimensional Cu(II)-4 4 -bipyridine network of the molecular formula [Cu 2 ( 4 4 '-bipy) 5 (DMSO)3(ClO4)](ClO4)33(DMSO)-(H 2 O) with hydrophobic rectangular enclathrated channels that extend throughout the solid_ 156 Reducing the concentration of bipy (2.5 moles ofbipy per mole of copper), favors the formation of the three dimensional network while at higher concentrations (3 0 moles and 3.5 moles of bipy per mole of copper) relatively equal quantities of 3 and the 3-D network crystallize from solution. Description of the Structures Crystallographic and structural refinement data for 1 2 and 3 are listed in Table 2-1. Selected bond angles and distances for 1 2 and 3 are given in Tables 2-2 2-3 and 2-4, respectively. Tables of atomic coordinates and thermal displacement parameters are provided in Appendix B. Structure of [Ni( 4,4 '-bipy )3(H20)2]( CI04)2 1.4( 4,4 '-bipy 3(H20) The structure of 1 consists of one-dimensional cationic (Ni( 4 4 '-bipy) 3 (H 2 O) 2 ] 2 + chains that pack to form layered sheets in the solid. The local coordination environment surrounding a typical Ni(II) ion is shown in Figure 2-1. The metal is six-coordinate and the coordination sphere consists of four pyridyl nitrogen donors one from each of four 4 4 '-bipyridine ligands and two oxygen atoms from two aqua ligands. The NiN 4 O 2 unit locally adopts an axially compressed octahedral geometry The four nitrogen atoms define the equatorial plane and the oxygen atoms occupy the axial sites. Both Ni--0 bond distances are equal (2.06 A) but shorter than the Ni-N bonds. Furthermore both of the Ni-N bonds from the terminal bipy ligands are equal (Ni-N21 and Ni N21A =

PAGE 59

48 Table 2-1. Summary of Crystallographic Data for 1, 2, and 3 Formula Weight 1038.51 1038.73 Space Group Monoclinic, C2/c Monoclinic, C2/c a, A 17.5696(8) 17.614(2) b, A 11.4101(5) 11.514(1) c, A 24.479(1) 24.604(2) a, deg 90 90 /J, deg 93.065(1) 92.448(2) x~ w w V, A 3 4900.4(4) 4985.6(9) Z 4 4 T, K 173(2) 173(2) l(Mo Ka), A 0.71073 0.71073 Peale, g cm3 1.408 1.384 cm1 5.76 5.21 R 8 CRwb) 0.0709 (0.2069) 0.0736 (0.2173) Table 2-2. Selected Bond Lengths [A] and Angles [ 0 ] for 1 Ni-01 2.060 O1-Ni-OIA Ni-OlA 2.060 NI 1-Ni-NllA Ni-Nll Ni-NllA Ni-N21 Ni-N21A 2.127 2.184 2.115 2.115 N21-Ni-N21A Nll-Ni-N21 Nll-Ni-01 NI 1A-Ni-N21 NllA-Ni-01 N21-Ni-01 1199.62 Monoclinic, Cc 19.0931(9) 11.1949(5) 25.607(1) 90 94.810(1) 90 5454.0(4) 4 173(2) 0.71073 1.461 6.44 0.0466 (0.1179) 179.82 180.000 177.20 88.60 89.91 91.40 90.09 92.44 a Symmetry transformations used to generate equivalent atoms: #1, -x, y, -z+3/2; #2, x, y+l, z; #3, x, y-1, z; #4, -x-1/2, -y+l/2, -z+2; #5, -x, -y+2, -z+2. Table 2-3. Selected Bond Lengths [A] and Angles ( 0 ] for 2 Co-01 2.061 O1-Co-OlA Co-OIA 2.061 NI 1-Co-Nl IA Co-NI I 2.180 N21-Co-N21A Co-NllA Co-N21 Co-N21A 2.232 2.168 2.168 Nll-Co-N21 Nll-Co-01 Nl 1A-Co-N21 NllA-Co-01 N21-Co-01 178.68 180.000 176.37 88.19 90.66 91.81 89.34 92.17

PAGE 60

49 Table 2-4. Selected Bond Lengths [A] and Angles [ 0 ] for 3a Cu-01 2.396 Nl-Cu-Nl' Cu-O2 2.376 N2-Cu-N3 Cu-NI 2.049 Ol-Cu-02 Cu-N2 2.025 Nl-Cu-N2 Cu-N3 2.033 Nl-Cu-N3 Cu-NI' 2.060 Nl'-Cu-N2 S(l}--01 1.507 NI '-Cu-N3 S(2}-02 1.488 Nl-Cu-01 S(2'}-02 1.347 N2-Cu-01 N3-Cu-Ol Nl'-Cu-01 Nl-Cu-02 N2-Cu-02 N3-Cu-02 Nl'-Cu-02 S 1-0(1 }-Cu S2-0(2}--Cu S2'-0(2}--Cu 178.39 179.25 178.17 91.09 89.03 90.51 89.36 90.69 91.49 87.78 89.13 90.01 86.81 93.93 90.21 141.6 145 9 154.1 2.12 A) but one of the Ni-N bonds (Ni-NI IA= 2.18 A) from a bridging bipy is longer than the other such bond (Ni-NI 1, 2.13 A). According to the spectrochemical series, water is a weaker field ligand than bipy and thus Ni-0 bond distances should be longer than Ni-N bonds. The aqua ligands are expected to be the sites of an axial elongation. However, from the structural data the opposite effect is observed. Steric repulsion from the pyridyl rings coordinated to the metal centers could cause a rather significant lengthening of Ni-N bonds. Note also that the hydrogen atoms from the coordinated water molecules hydrogen bond to the terminal nitrogen atoms from monodentate bipy's on adjacent chains thus polarizing the H-0 bonds to a greater extent. The oxygen atom effectively becomes more negatively charged, and as a consequence, is more attracted to the positively charged metal center resulting in a smaller than expected Ni-0 bond distance.

PAGE 61

50 Two of the bipy ligands, coordinated trans with respect to one another, bridge the Ni(II) ions to form infinite one-dimensional linear chains that extend along the crystallographic c-axis. A single chain is depicted in Figure 2-2. The Ni-Ni distance along a chain is a/2 units (11.41 A). For each bridging bipy ligand the pyridyl rings are not coplanar but twisted along the central C--C bond at an angle of29.7 with respect to each other. Ignoring the uncoordinated guests, the metal:bipy stoichiometry in 1 is 1 :3 since the bipy ligands perpendicular to the chains are monocoordinate. The Ni-bipy chains are juxtaposed in a side-by-side fashion to form two dimensional sheets within the crystallographic be-plane. A typical sheet is shown in Figure 2-3. Within each sheet, the chain spacing is b/2 units. In addition to packing forces, the sheets are held together by a combination of hydrogen bonding interactions between the protons of the coordinated water molecules and the terminal nitrogen atoms from the monodentate bipy ligands on adjacent chains (N-H bond distance of 1.81 A) and offset JZ'-stacking between the monocoordinate bipy ligands on adjacent chains. The face-to-face distance between these overlapping bipy groups is~ 3.7 A. Note that in order to maximize the favorable JZ'-stacking interactions, no twisting of the pyridyl rings is observed along the central C--C bond for these monodentate bipy ligands. The characteristic packing motif of the Ni-bipy chains produces rectangular, hydrophobic cavities within the sheets. Each cavity is defined by four nickel ions at the comers and along the sides by the faces of the two bridging bipy s and the edges of the two pairs of JZ'-stacked terminal bipy ligands. The dimensions of the cavities are b/2 x c and, if the Van der Waals radii of the carbon atoms from the bipy s are approximated as 1.7 A, the effective size of the cavities is approximately 9.7 Ax 10.7 A.

PAGE 62

51 As shown in Figures 2-4 and 2-5, the sheets pack to form a layered solid-state structure along the crystallographic c-axis. Although the sheets align in registry along the b-axis, they are offset by step in both the aand c directions. The characteristic packing results in an alignment of the hydrophobic cavities to form oblique channels extending along the [2 0 -2] direction, as shown in Figure 2-6. The void space within hydrophobic pores and between the sheets is not empty but occupied by clathrated guest molecules and counterions acting to prevent the interpenetration of adjacent sheets. The pores within the framework host are not empty, but occupied by enclathrated guest molecules and counterions (Figures 2-3 to 2-6). These lattice guests are extensively disordered throughout the solid thus leading to the rather high final refinement value in the structural solution. Approximately 1.5 crystallographically inequivalent, uncoordinated bipy molecules, each disordered about centers of inversion, are present per asymmetric structural unit. Extensive hydrogen bonding interactions between the lattice waters and the coordinated waters, perchlorate ions, and bipy guests are present throughout the channels. The clathrated bipy molecules seem to form a secondary lattice of organic molecules interpenetrated within the porous network structure of 1. Coordination polymers with both guest enclathrated and empty cavities and channels present throughout the solid are known. 40,44,47,69,70,95,l09,l l4 Each rectangular cavity in 1 clathrates an uncoordinated 4,4 '-bipy molecule stabilized by both weak hydrogen bonding and hydrophobic interactions. These guests are crystallographically centrosymmetric and disordered about a center of inversion. The guest is positioned approximately at the center of the cavity such that the edges of one pair of borders (the pair of JZ"-stacked monocoordinate bipy ligands) is directed toward

PAGE 63

52 the faces of the guest and the faces of the other pair of borders (the bipy bridges) is directed toward the edges of the guest with the edge-to-face distances of ca. 2.5 A3.0 A. Note that the pyridyl rings of this bipy molecule are coplanar. The interior hydrogen atoms from the bipy form both twoand three-center hydrogen bonds with one (2.39 A) and two oxygen atoms (2.29 A and 2.53 A), respectively, from nearby perchlorate counterions. Uncoordinated, disordered bipy guests are clathrated between the sheets of 1 and are stabilized by Jr-stacking and hydrogen bonding interactions as well. Each face directed toward the space between the sheets from the pair of monocoordinate bipy ligands that comprise part of the borders of the hydrophobic cavities interacts with a single such bipy guest. These guests stack in an offset parallel fashion with the pair of Jr-stacked monocoordinate bipy ligands that comprise part of the borders of the cavities with a face-to-face distance of ca. 3.3 A, indicative of Jr-stacking interactions. Note that these bipy guests are oriented almost perpendicular with respect to the bipy's clathrated within the cavities. The terminal nitrogen atoms from these bipy' s form hydrogen bonds (2.48 A) with oxygen atom from nearby perchlorate counterions. The lattice water molecules are positioned in the vicinity of the bridging bipy ligands and the perchlorate counterions are in close proximity to the bipy guests located between the sheets. One of the lattice water molecules simultaneously hydrogen bonds with a proton from the coordinated water molecule (H--0, 1.99 A), the terminal nitrogen atom from the bipy molecules between the sheets (N--0, 2.642.82 A), the oxygen atom of a nearby perchlorate counterion (0--0, 2.82 A), and another nearby lattice water

PAGE 64

53 molecule (0-0, 2.84 A). This second lattice water molecule also hydrogen bonds with an oxygen atom from a neighboring perchlorate counterion (2.07 A). The structure of 1 bears close resemblance to a compound reported by Y aghi and coworkers, [Ni(4,4' -bipy) 2 5 (H2O)2](ClO4)2 l.5(4,4' -bipy)(H2O).11 4 Though this material is a covalent molecular railroad as opposed discrete 1-D chains, both structures incorporate both bridging and terminal bipy ligands as well as 4,4' -bipy enclathrated channels. The Ni-N(bridging) distances (2.13 A-2.19 A) in 1 are comparable to those found in the railroad structure (2.11 A) but are larger than those found in the cis-chain [Ni(4,4' -bipy)(Et-XA)2]-0.5(EtOH)(CHCh) (2.07 A2.09 A, Et-XA = ethylcarbonadithiolate)98 and the 2-D covalent grid [Ni(4,4'-bipy)(ox)] (2.09 A, ox= oxalato ).40 The Ni-N(terminal) distances (2.115 A) of 1 are also similar to those found in [Ni(4,4'-bipy)2_s(H2O)2](ClO4)2.5(4,4'-bipy)-2(H2O) (2.15 A). Furthermore, Ni--0 distances (2.06 A) are also close to those in the Yaghi structure (2.08 A) and in [Ni(4,4' bipy)(ox)] (Ni--O(oxalato) bonds) (2.05 A). All O--Ni--0, N(bridging)-Ni N(bridging), and N(terminal)-Ni-N(terminal) angles are 180 and 0--Ni-N and N(bridging)-Ni-N(terminal) angles are close to 90. Structure of [ Co( 4,4 '-bipy )3(H 2 0)2]( CI0 4 ) 2 1.4( 4,4 '-bipy 3(H20) Other than substituting the metal center Co(II) for Ni(II) and small differences in characteristic bond angles and distances, the structure of 2 is virtually identical to 1 and thus no pictures or description is given. The structure of 2 resembles the trans-1-D chain compounds reported by Jacobson and coworkers, [Co( 4,4 '-bipy)(SO 4 )(H 2 O) 2 ]-2(H 2 O) and [Co( 4,4 '-bipy)(Cl)2(DMSO)2] where the solvent molecules and counterions, not terminal bipy ligands, occupy the non-bridging coordination sites on the metal centers.96

PAGE 65

54 N218 N21C N11C Figure 2-1 The local coordination environment of a typical Ni(II) metal center in compound 1. All aromatic hydrogen atoms have been omitted for clarity All non hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 66

55 Figure 2-2. The structure of compound 1 along the crystallographic c-axis representing a linear, one-dimensional Ni-bipy chain All aromatic hydrogen atoms have been omitted for clarity All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 67

56 Figure 2-3. The structure of compound 1 within the crystallographic be-plane representing a two-dimensional sheet (left) with the accompanying guests and counterions (right). For clarity the aromatic hydrogen atoms and guests in the left-hand figure have been omitted and the coordinated guests have been omitted in the right-hand figure.

PAGE 68

57 Figure 2-4. The structure of compound 1 within the crystallographic ac-plane showing the layered two-dimensional sheets (left) with the accompanying guests and counterions (right). For clarity the aromatic hydrogen atoms and guests in the left-hand figure have been omitted and the coordinated bipy ligands and have been omitted in the right-hand figure.

PAGE 69

58 Figure 2-5. The structure of compound 1 within the crystallographic ab-plane showing the layered two-dimensional sheets (left) with the accompanying guests and counterions (right). For clarity, the aromatic hydrogen atoms and guestsin the left-hand figure have been omitted and the coordinated bipy ligands have been omitted in the right-hand figure. Figure 2-6. The hydrophobic channels in compound 1 that extend along the [2 0 -2] direction with the accompanying guests and counterions. For clarity, the aromatic hydrogen atoms have been omitted.

PAGE 70

59 The Co-N(bridging) distances (2.18-2.23 A) in 2 are comparable to those in the chains [Co(SO 4 )(H 2 O)J( 4,4 '-bipy)]-2(H 2 O) ( ~ 2.17 A)96 as well as in the 2-D covalent grid [Co(4,4'-bipy)(ox)] (2.15 A, ox= oxalato).40 All Co-N distances in 2 are larger than the corresponding Ni-N distances in 1. The Co-0 distances in 2 (2.06 A) are also comparable with those in [Co(4,4'-bipy)(SO 4 )(H2O)z]-2(H 2 O)(2.08 A) and [Co(4,4' bipy)(ox)] (2.08 A). Structure of [ Cu( 4,4 '-bipy )J(D MS0)2] ( CI04)2 2( 4,4 '-bipy) The structure of 3 consists of one-dimensional cationic [Cu( 4,4 bipy) 3 (DMSO) 2 ] 2 + chains that pack to form layered sheets in the solid. Although 3 is structurally related to 1 and 2, some important differences are present. The local coordination environment surrounding a typical Cu(II) ion is shown in Figure 27. The metal is six-coordinate and the coordination sphere consists of four pyridyl nitrogen donors, one from each of four 4,4 '-bipyridine ligands and two oxygen atoms from ligated DMSO molecules. One DMSO ligand per metal center is disordered about the sulfur atom. The CuN 4 O 2 unit locally adopts an axially elongated octahedral geometry. The four nitrogen atoms define the equatorial plane and the oxygen atoms occupy the axial sites. Both Cu--0 bond distances are equal (2.40 A) and longer than the Cu-N bonds. Furthermore, the Cu-N bonds from the terminal bipy ligands (2.022.03 A) are shorter than the Cu-N bonds from the bridging bipy ligands (2.05 2.06 A). Two of the bipy ligands, coordinated trans with respect to one another, bridge the Cu(II) ions to form infinite one-dimensional linear chains that extend along the crystallographic c-axis. A single chain is depicted in Figure 2-8. The Cu-Cu distance along a chain is approximately a/2 units (11.2 A). For each bridging bipy ligand, the

PAGE 71

60 pyridyl rings are not coplanar but twisted along the central C-C bond at an angle of 61.4 with respect to each other, considerably larger than the similar such dihedral angles observed in 1 and 2. Ignoring the uncoordinated guests, the metal:bipy stoichiometry in 1 is 1 :3 since the bipy ligands perpendicular to the chains are monocoordinate. The Cu-bipy chains are juxtaposed in a side-by-side fashion to form quasi-two dimensional sheets within the crystallographic be-plane. A typical sheet is shown in Figure 2-9. Within each sheet, the chain spacing is approximately b/2 units. Weak N-H contacts (2.6 A) between the terminal nitrogen atoms from the bipy ligands with the nearby hydrogen atoms from the bridging bipy ligands on adjacent chains and S-H contacts (2.9 A) between the hydrogen atoms near the terminal nitrogen atoms from the monocoordinate bipy ligands with the neighboring sulfur atoms from DMSO ligands on adjacent chains are present. Additionally, offset 1t stacking are observed between the monocoordinate bipy ligands on adjacent chains. The face-to-face distance between these overlapping bipy groups is~ 3.8 A. The pyridyl rings on the monocoordinate bipy ligands are twisted 11.9 along the central C-C bond with respect to one another thus likely reducing the effectiveness of the stabilizing .1rinteractions between the bipy pairs. The characteristic packing motif of the Cu-bipy chains produces rectangular, hydrophobic cavities within the sheets. Each cavity is defined by four copper ions at the comers and along the sides by the faces of the two bridging bipy's and the edges of the two pairs of .1r-stacked terminal bipy ligands. The dimensions of the cavities are b/2 x c and, if the Van der Waals radii of the carbon atoms from the bipy's are approximated as 1.7 A, the effective size of the cavities is approximately 9.5 Ax 11.3 A, slightly larger than the cavities present in 1 and 2.

PAGE 72

61 As shown in Figures 2-10 and 2-11, the sheets pack to form a layered solid-state structure along the crystallographic c-axis. Although the sheets align in registry along the b-axis, they are offset by step in both a and c. The characteristic packing results in an alignment of the hydrophobic cavities to form oblique channels extending along the [2 0 -2] direction, as shown in Figure 2-12, just as in 1 and 2. Note the hydrophobic interactions between the nearest neighbor methyl groups from DMSO ligands between the sheets. The pores within the framework host are not empty, but occupied by enclathrated guest molecules and counterions as shown in Figures 2-9 to 2-12. The lattice guests are relatively less disordered in 3 compared to 1 and 2 thus resulting in a better structural refinement. Two crystallographically inequivalent, uncoordinated bipy molecules are present per asymmetric structural unit. Hydrogen bonding interactions between the perchlorate ions, bipy guests, and the host are observed but weaker and less prevalent than those similar interactions present in 1 and 2. Again, the clathrated bipy molecules seem to form a secondary lattice of organic molecules interpenetrated within the porous network structure of 3. Each rectangular cavity of 3 clathrates an uncoordinated 4,4 '-bipy molecule stabilized by hydrophobic interactions with the host framework. One pyridyl ring per guest is disordered about the central C-C bond. Unlike in 1 and 2, the guest is not positioned at the center of the cavity. The edges of one pair of borders (the pair of Jl'-stacked monocoordinate bipy ligands) is directed toward the faces of the guest and the faces of the other pair of borders (the bipy bridges) is directed toward the edges of the guest with the edge-to-face distances of ca. 2.9 A3.8 A. Note that the pyridyl rings of

PAGE 73

62 this bipy molecule are slightly twisted 12.6 with respect to each other along the central C---C bond. Uncoordinated, bipy guests are clathrated between the sheets of 3 and are stabilized by Jl'-stacking and hydrogen bonding interactions as well. Unlike in 1 and 2, these guests are not disordered. Each face directed toward the space between the sheets from the pair of monocoordinate bipy ligands that comprise part of the borders of the hydrophobic cavities interacts with a single such bipy guest. These guests stack in an offset parallel fashion with the pair of Jl'-stacked monocoordinate bipy ligands that comprise part of the borders of the cavities with a face-to-face distance of ca. 3.7 A, indicative of Jl'-stacking interactions. Note that these bipy guests are oriented almost perpendicular with respect to the bipy' s clathrated within the cavities. The terminal nitrogen atoms from these bipy s form weak N-H contacts (2.6 A) with hydrogen atoms from nearby bridging bipy ligands that are part of the framework. The perchlorate counterions are in close proximity to the bipy guests located between the sheets. The perchlorate oxygen atoms are observed to weakly interact with hydrogen atoms from the bridging bipy ligands (0-H contacts of 2.4 A), both interior hydrogen atoms from the monocoordinate bipy ligands (0-H contacts of2.42.6 A), and the hydrogen atoms from DMSO ligands (0-H contacts of 2.5 A) Unlike in 1 or 2, the Cu-0 bonds are longer that the Cu-N bonds in 3. The DMSO molecules are weaker field ligands compared to the bipy ligands and, in combination with the Jahn-Teller effect, the solvent molecules are located on the axially distorted coordination sites. The Cu-N(bridging) distances (2.03 -2.06 A) are comparable to those in the 1-D chains [Cu(4,4'-bipy)(SO 4 )(H 2 O) 3 ](H 2 O) (2.05

PAGE 74

63 A) 69 but smaller than the corresponding distances (1.99 A) in [Cu( 4,4 bipy)(FBF 3 ) 2 (H 2 O)2]-(4,4' -bipy). 92 All Cu-N distances in 3 are smaller than the corresponding Ni-N and Co--N distances in 1 and 2, respectively. The Cu---0 distances (2.06 A) are comparable with the those in [Cu(4,4'-bipy)(SO4)(H2O)3]-2(H2O) (1.95 A2.2 A). All O-Cu---0, N(bridging)-Cu-N(bridging), and N(terminal}Cu-N(terminal) angles are close to 180 but 0-Cu-N and N(bridging)-Cu N(terminal) angles deviate from 90, more so than in 1 and 2, where N(bridging) and N(terminal) denote the nitrogen atoms from bridging and terminal bipy ligands, respectively. Thermal Properties Compounds 1 and 2 are unstable in air due to loss of enclathrated guest molecules. The compounds can be kept in humid environment or under solvent, but both compounds discolor and change texture if left in laboratory atmosphere within a few hours. Thermogravimetric analysis (TGA) and thermal desorption mass spectrometry show the stepwise loss of water followed by uncoordinated bi pyridine. From Figure 213, mass loss between room temperature and 56 C corresponds to three moles of water corresponding to guest water molecules in the chemical formula. Water continues to be released up to near 150 C. Above 85 C, bipyridine is lost continuously up to approximately 250 C. The TGA plot of 2, shown in Figure 2-14 undergoes similar, but somewhat more complex, guest molecule loss process. Note that the onset temperature for the guest loss in 1 is significantly lower (by approximately 30 compared to 2. In contrast to 1 and 2, the copper compound 3 is stable in air at room temperature. Without solvent molecule guests, it does not experience the same decomposition process.

PAGE 75

64 From Figure 2-15, the first mass decrease observed by TGA occurs above 100 C and corresponds to loss of guest bipyridine molecules. Figure 27. The local coordination environment of a typical Cu(II) metal center in compound 3. All hydrogen atoms have been omitted for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 76

65 Figure 2-8. The structure of compound 3 along the crystallographic c-axis representing a linear, one-dimensional Cu-bipy chain. All hydrogen atoms have been omitted for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 77

66 Figure 2-9 The structure of compound 3 within the crystallographic be-plane representing a two-dimensional sheet (left) with the accompanying guests and counterions (right). For clarity, the guests have been omitted in the left-hand figure and the coordinated bipy ligands have been omitted in the right-hand figure.

PAGE 78

67 Figure 2-10. The structure of compound 3 within the crystallographic ac-plane showing the layered two-dimensional sheets (left) with the accompanying guests and counterions (right). For clarity the guests have been omitted in the left-hand figure and the coordinated bipy ligands have been omitted in the right-hand figure.

PAGE 79

68 Figure 2-11. The structure of compound 3 within the crystallographic ab-plane showing the layered two-dimensional sheets (left) with the accompanying guests and counterions (right). For clarity, the guests have been omitted in the left-hand figure and the coordinated bipy ligands have been omitted in the right-hand figure.

PAGE 80

69 a b v_c Figure 2-12. The hydrophobic channels within compound 3 that extend along the [2 0 -2] direction with the accompanying guests and counterions. For clarity the hydrogen atoms have been omitted.

PAGE 81

100 80 i 60 ._. "' "' Ill 40 20 0 0 Mass Loss Derivative of Mass Loss 70 8 6 4 2 0 --~--------------- -2 100 200 300 400 500 600 Temperature ( 0 C) ,u e..... -;;!!. 0 -"' "' Ill ... Q ., > .:: Ill > ;: ., Q Figure 2-13. TGA therrnogram of compound 1 depicting the observed mass loss and negative values of the first derivative(%/ C). 8 100 Mass Loss 80 6 --u e..... 60 -;;!!. 4 0 .._. "' 0 .. --Derivative of Ill "' "' 40 Mass Loss Ill ... 2 Q ., > 20 .:: Ill 0 .:: .. ., 0 Q -2 0 100 200 300 400 500 600 Temperature ( 0 C) Figure 2-14. TGA thermogram of compound 2 depicting the observed mass loss and negative values of the first derivative(%/ C).

PAGE 82

71 8 10 0 M as s L o ss 8 0 6 .--. u 60 .--. 4 "' "' = '-' Deri v ati ve o f :; "' M ass L oss ... "' 4 0 0: 2 0 :; .. 2:: .... = 20 0 > ;: .. Q 0 -2 0 100 200 300 400 500 600 Temperature ( 0 C) Figure 2-15. TGA thermo gram of compound 3 depicting the observed mass loss and negative values of the first derivative (% / C).

PAGE 83

72 Magnetic Properties The molar magnetic susceptibility, XM, and inverse molar susceptibility, at 1000 G over the temperature range of 2 K 300 K for 1 and 3 are plotted in Figures 2-16 and 219, respectively. The molar magnetization, MM, at 2 K over the field range of O 50 kG for 1 and 3 are plotted in Figures 2-17 and 2-20, respectively The molar magnetic susceptibility at 1000 Gover the temperature range of 2 K-50 K for 2 is plotted in Figure 2-18. Magnetic Properties of [Ni( 4,4 '-bipy)3(H20)2](CI04)2 1.4(4,4 '-bipy)(H20) The room temperature (i.e. 300 K) susceptibility (xM = 4.2 x 103 emu mor 1 ) of 1 correlates well with the value expected for uncoupled, S = 1 metal centers (xM = 4.2 x 103 emu mor 1 ). Recall the structure of 1 consists of linear chains of 4,4 bi pyridine bridged Ni(II) ions. Since 1 is comprised of chains of Heisenberg S = 1 spins that experience single-ion, D, and in-plane, E, anisotropies, an S = 1, one-dimensional chain model incorporating both zero-field splitting and exchange parameters is appropriate to fit the magnetic data. Only exchange interactions along the chains were considered; coupling between the chains was ignored since interchain bonding is noncovalent. The Heisenberg model is an appropriate starting point as octahedral Ni(II) complexes have nearly isotropic g-factors (g = 2.25).5,157 The Hamiltonian may be written as where J is the exchange interaction, Dis the single-ion anisotropy, Eis the in-plane anisotropy, and B is the applied magnetic field_ 15 8 From this Hamiltonian, the parallel, X II, and perpendicular, X.1, susceptibilities (neglecting the E parameter) are

PAGE 84

73 _ 2NAglf ; (1D-4z.JJ X n X z 3k T 3k T B B and the susceptibility representing the average of contributions that are parallel and perpendicular to the chains, is since magnetic measurements were performed on randomly oriented samples. l (2-2) (2-3) (2-4) A theoretical treatment of the magnetization in the high field limit of the Hamiltonian given in Equation (2-1) has not yet been reported, however in the limit that J = 0 and E = 0, expressions for the high field dependence of the magnetization do exist. l The expressions representing the magnetization perpendicular, M.1, and parallel, Mi 1 to the chains (neglecting the E and J parameters) are where and where Eo E1 2N B g2 -e-kT +e-kT MAl.l.B l. / 4gl.2B2 Bl.22 + Dl _!!_ Eo E1 v e kT + e kT + e kT 4g1 Bl + Dl E = l. B l. +D 0,1 2 E1 E3 N g {-e kT + e kT } M, A II B II E 2 E 3 1 + e kT + e kT (2-5) (2-6) (2-7)

PAGE 85

74 and the magnetization representing the average of contributions that are parallel and perpendicular to the chains, is for powdered samples. 1 M ove = (M 11 +2M _j_ ) 3 (2-8) (2-9) The temperature dependence of the inverse susceptibility for 1 between 5 K and 50 K were fit by a simple Curie-Weiss law model and the results of the fit yield a Weiss temperature 0= 2 K, and from the Curie constant, g = 2.20. The temperature dependence of the low field susceptibility was fit using the expressions in Equations 2-2, 2-3, and 2-4, and the results of this analysis, when using g = 2.20, are given by the solid line in Figure 2-16, whereJ/k 8 = 0.9 K, Dlk 8 = 4.7 K, and E = 0. The small, negative coupling constant suggests the presence of a weak antiferromagnetic exchange along the chains. Since these measurements were performed on randomly oriented samples, the sign of D cannot be unambiguously determined.6 The DIIJI ratio of~ 5 suggests indicates that 1 is an example of a large DIIJJ system with strong planar anisotropy.158161 Naturally the fitting procedure would be significantly improved if the data extended to sufficiently low temperature so as to reveal the maximum of the susceptibility, observed at T = 2 K. The field dependence of the low temperature magnetization in Figure 2-17 was fit using the expressions in Equations 2-5 to 2-9. In the limit that J = 0 and E = 0, the model provides a prediction that closely resembles the experimental data when g = 2.20 and D/ks R: 7 K. The analysis of the low field susceptibility indicates that DIIJJ R: 5, so the

PAGE 86

75 approximation that J = 0 is clearly not justified. Nevertheless, this analysis does provide an upper bound for the value of D and is consistent with the low field analysis. For the purpose of comparison, the S = 1 Brillouin function, which is appropriate for zero exchange coupling effects, is plotted in Figure 10 as the dotted line. Once again, the data clearly suggest the presence of finite magnetic coupling in 1. The delocalized 1! system of 4,4' -bi pyridine should allow the ligand to effectively mediate superexchange interactions when covalently bridging paramagnetic centers. 5 Furthermore, the coupling is expected to be antiferromagnetic as explained by a spinpolarization mechanism for the propagation of exchange interactions.154 The magnitude of the coupling is small (Jlk 8 = 0.9 K), but is consistent with the coupling constants measured for other similar bipy bridged complexes [Ni(Et-XA)2(4,4'-bipy-)] .5(EtOH)(CHCh) (Et-XA = ethylcarbonadithiolate),98 [Mn(hfac) 2 (4,4' -bipy)] (hfac = hexafluoroacetylacetonato ),97 [Cu 2 (tren) 2 ( 4,4'bipy)](BPH.i) 4 162 [Cu 2 (dien)2( 4,4' bipy)(ClO4)2](ClO4)2, 163 and [Mn--( 4,4, -bipy)( 4,4 '-bipy)(NCS) 2 (H 2 O) 2 ]n.164 Since a typical sp 2 -sp 2 C-C single bond is 1.50 A and a C=C double bond is 1.35 A, the carbon-carbon bond between the pyridyl rings (1.49 A) for each coordinated bipy ligand in 1 is principally of single bond character. The l!orbital conjugation between these pyridyl rings is small and the rings twist along this central C-C single bond in order to minimize the steric repulsions felt by the interior hydrogen atoms.164 This twisting disrupts the exchange along the 1!-orbital pathway and thus hindering the effectiveness ofbipy to propagate a superexchange interaction between paramagnetic centers.

PAGE 87

0 3 ,...._ 0 2 -~ e s 0 1 :i: >< 0 0 76 ~--...-...-..--..---.--.--,---,---.--.--,--,300 200 0 0 100 200 300 T(K) Figure 2-16. The molar magnetic susceptibility, XM, and inverse susceptibility, 1/XM, at 1 kG from 2 K to 300 K for compound 1 are shown as open boxes and open circles, respectively. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. The fit of x to the Heisenberg S = 1 chain model from Equations 2-2, 2-3, and 2-4 is shown by the solid line. 12 5 10 0 ., ,...._ -l 7 5 C, 5 0 ..,0 ...., .__, "f 2.5 Af .,a l 2 K Fil lo S = 1 Chain Model Fit to S = 1 Brillouin Function 0 0 0 10 20 30 40 50 B (kG) Figure 2-17. The molar magnetization (MM) at 2 K from Oto 50 kG for compound 1 is shown as open boxes. The data have been corrected for background signals arising from the sample container. The fit of M to the S = I chain model from Equations 2-5 to 2-9 and the S = I Brillouin function is shown by the solid line and dotted line, respectively.

PAGE 88

77 Magnetic Properties of [ Co( 4,4 '-bipy h(H20)2] ( CI04)2 1.4( 4,4 '-bipy 3(H20) A complete analysis of the magnetic data for 2 is more complicated compared to analogous Ni(II) and Cu(II) systems since, in Co(II) systems, spin-orbit coupling effects are important, a thermal dependence of the spin quantum number is observed ( depopulation of S = 3/2 state to S = state at low temperatures), and significant anisotropy in the g-factors is present.1,5,6 The Co(II) ions of 2 form chains of Ising S = spins that assume an S = state at temperatures below 30 K. The S = Ising model is an appropriate starting point since octahedral Co(II) complexes have highly anisotropic g-factors.1,5, 7 Again, only exchange interactions along the chains were considered. The general Hamiltonian may be written as iI = -2fiaSt -s; + ps: -s; + ,s rs; + gBB. rs; (2-10) iS. j where, in the Ising model, a.= 1 and p = y = 0, zero-field splitting terms have been neglected, and the remaining terms have the usual meaning. From this Hamiltonian, the parallel and perpendicular susceptibilities are (2-11) X = X = NAg~; [tanh(l{j_)+l{j_sech 2 (1{j_)] .1 x, y 81 J I k T k T k T B B B (2-12) and the susceptibility representing the average of contributions that are parallel, Xii and perpendicular, X.L, to the chains, is (2-13)

PAGE 89

78 since magnetic measurements were performed on randomly oriented samples. I, 7 Unlike Ni(II) systems, significant anisotropy with Co(II) spins is observed and a complete analysis of the magnetic behavior requires data from single crystals in different orientations with respect to their measuring magnetic field.1,5,6 Since the reported experiments were performed on powder specimens, both perpendicular and parallel tensor components were unavailable. Thus, any meaningful fits to the low-temperature susceptibility data in the absence of known values of both J and g ll and g.1-values are difficult. However, simulations of the low-temperature susceptibility from the Ising model from Equations 2-11, 2-12, and 2-13, shown in Figure 2-18, reproduced the data relatively well. The coupling constant was fixed at J = 0.4 K, similar to the value obtained from fits to the susceptibility data of 1 (since 2 is isostructural to 1). Three sets of g-values (g ll = 8 and g.1 = 1.5, g ll = 6and g.1 = 3.5, and g ll = 4 and g.1 = 4.25) were obtained for octahedral Co(II) ions from a universal curve reported by Carlin. 5 The fitting expression was also corrected for uncoupled, S = impurities and fixed at a high limit of 10 %. The model incorporating all three sets of g-values reproduced the low temperature susceptibility well, with the best results obtained for g ll = 8and g.1 = 1.5. Increasing or decreasing the magnitude of the coupling constant had resulted in poor simulations of the data. Thus, while it was not possible to unambiguously estimate the value components, the simulations provided a rough estimate of the sign and magnitude of the exchange parameter. Qualitatively speaking, the cobalt analog, 2, also shows little evidence of magnetic exchange just as in 1 and 2.

PAGE 90

79 Magnetic Properties of [Cu(4,4'-bipy)3(DMS0)2](CI04)2(4,4'-bipy) For the case of the Cu(II) ions of 3, the room temperature (i.e. 300 K) susceptibility (xM = 1.2 x 103 emu mor 1 ) correlates well with the value expected for uncoupled, S = metal centers (xM = 1.3 x 103 emu mor 1 ). The temperature dependence of the molar magnetic susceptibility at 1 kG, shown in Figure 2-19, was fit well by the Curie law using a g-value of 2.06, as determined from room temperature ESR measurements. The field dependence of the molar magnetization up to 50 kG at 2 K, shown in Figure 2-20, was also fit well by the S =Brillouin function, with a g-value of 2.06, which describes non-interacting magnetic spins.5,6 Consequently, 3 behaves essentially as a chain of non-interacting, S = metal centers that are uncoupled even at 2 K. The larger dihedral angles between the pyridyl rings on the bridging bipy ligands may disrupt the superexchange pathway to a greater extent than in 1 or 2 and could account for the lack of any observed exchange interaction.

PAGE 91

1.5 1.2 _,...... l 0 9 s 0 6 :E ....: 0.3 80 o z.,a t I kG -g~ 1.5 andg,... S -g,.,. 3 5 andg,_. 6 g,.,.. 4 25 andg,... 4 0 0 ....._.....____.______,_.,____._____,__..___.___.__._____. 0 4 8 12 16 20 T(K) Figure 2-18. The molar magnetic susceptibility, XM, at 1 kG from 2 K to 300 K for compound 2 is shown as open boxes. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. The simulations of x to the S = Ising model are from Equations 2-11, 2-12, and 2-13 where J is fixed at 0.4 K while g.1 = 1.5 and g 11= 8 for the solid line, g.1 = 3.5 and g 11= 6 for the dotted line, and g.1 = 4.25 and g11= 4 for the dashed line. 1()00 0 20 z_., al I kG 0 -Fil 10S= 112 Curie Law 0 800 0 llz.,at I kG 00 0 0.15 00 600 _,...... 0 _,...... 00 e 0 10 00 400 0 ::, 00 ,, e 00 ::. .., 00 '-' :E 00 >< 0 05 0 200 0 50 100 150 200 250 300 T(K) Figure 2-19. The molar magnetic susceptibility, XM, and inverse susceptibility llXM, at 1 kG from 2 K to 300 K for compound 3 are shown as open boxes and open circles, respectively. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. The fit of x to the S = Curie Law is shown by the solid line.

PAGE 92

81 6 s '7,-.. 0 4 5 0 ::, 3 "'o 2 ..... '-' -:! I M., at 2 K Fit to S = l /2 Brillouin Function 0 0 10 20 30 40 so B(kG) Figure 2-20. The molar magnetization (MM) at 2 K from 0 to 50 kG for compound 3 is shown as open boxes. The data have been corrected for background signals arising from the sample container. The fit of M to the S = Brillouin function is shown by the solid line.

PAGE 93

82 Conclusions Three new hybrid organic-inorganic coordination polymers have been isolated and characterized. These materials consist of chains of transition metal ions (Ni(II), Co(II), and Cu(II)) bridged by 4,4 '-bipyridine spacer ligands. The chains pack to form two-dimensional, non-interpenetrated sheets with hydrophobic, rectangular cavities present within the framework. The sheets, in turn, pack to form a three-dimensional structure with oblique channels containing enclathrated guest molecules and counterions extending throughout the solid. These enclathrated guests are easily lost suggesting that the samples are thermally unstable. In general terms, the magnetic properties of 1, 2, and 3 are similar in the sense that weak exchange interactions, J, are present between the metal centers. Coordination polymers with hydrophobic cavities and channels extending throughout the solid-state structure have received much attention due to their ability to act as molecular sieves with size and shape specificity and catalytic substrates.37,44,45 Compounds 1, 2, and 3 are clearly examples of network solids with cavities and channels that prefer to enclathrate hydrophobic guests (uncoordinated bipy molecules). The ability of these coordination polymer hosts to exchange guests is described in Chapter 3.

PAGE 94

CHAPTER3 A 31 P MAS NMR INVESTIGATION OF THE HOST-GUEST PROPERTIES OF TWO POROUS NETWORK SOLIDS [Ni( 4,4 '-bipy)3(H2O)2](ClO4)2 1.4( 4,4, -bipy) 3(H2O) and [Co(4 4' -bipy)3(H2O)2](ClO4)2 1.4(4,4 -bipy)-3(H2O) Introduction The design and construction of hybrid organic-inorganic porous network solids through the self-assembly of simple, molecular and ionic components is an emerging area of supramolecular chemistry that can provide new generations of functional materials. The importance of porous solids in inclusion phenomena, e.g adsorption/ desorption ion exchange, and size and shape-selective molecular sieving as well as catalysis is due in part, to their ability to reversibly clathrate or trap species within their cavities and extended channels.129 Like zeolites many hybrid organic/ inorganic solids clathrate guests within pores cavities and channels that are part of their lattice framework. These pores often possess a variety of sizes and shapes not observed in analogous inorganic porous solids such as zeolites and molecular sieves, thus potentially yielding novel and unique inclusion capabilities 87 A number of porous coordination networks have been found to exhibit many other desirable zeolitic properties as well such as stability and porosity of the framework guest exchange, and selective catalytic activity.42,44 93,142 By careful selection and design of the chemical components, the size and clathration properties of these pores can be fine-tuned to meet specific needs while maintaining the overall structural and functional features found in naturally occurring analogs. 3 7 83

PAGE 95

84 Despite the large amount of research devoted toward the synthetic aspect of solid state supramolecular chemistry in terms of producing functional materials and elucidating methods for the rational design and fabrication of such materials, studies of the chemical reactivity of these materials has been lacking.14 7 In fact most of these reactivity studies have been limited to the investigation of the inclusion properties of porous solids, such as the guest exchange and adsorption-desorption processes of small molecules.45,87,142,144,147 The imbalance of synthetic work compared to reactivity studies of supramolecular materials is largely due to the fact that, unlike the well characterized chemical reactivity properties of molecular and ionic species that are generally soluble in many common solvents, most coordination polymers are insoluble in most organic and inorganic solvents thus rendering any reactivity studies difficult.14 7 Yaghi, et al, has reported selective guest binding and removal of alcohols and ketones from a three-dimensional porous Zn(II)-Benzenetricarboxylate network and aromatic molecules from a layered porous Co(II)-Benzenetricarboxlyate network without collapsing the host.45,87 Endo, et al has described the reversible non-selective guest binding and removal of ketones, esters, hydrocarbons, and haloalkanes, as solids, liquids, and gases, within a zeolitic anthracene-bis(resorcinol) layered, hydrogen-bonded network_ 142 Kondo, et al. has reported the adsorption and desorption properties of small gas molecules, e.g. methane, within microporous interpenetrated M-4,4 '-azopyridine (M = Mn(II), Cd(II), and Co(II)) coordination networks without breaking apart the host.144 NMR Spectroscopy Magic angle spinning (MAS) NMR spectroscopy is a very useful technique for investigating chemical interactions within solid-state materials and can be applied toward

PAGE 96

85 determining the presence of enclathrated guests within porous solids.165 Additionally, pore sizes can be estimated as well as observing the dynamic behavior of the guests (i e. chemical exchange), the nature of the binding sites (provided a coupling constant can be measured), and the relative strength of any specific chemisorption and physisorption interaction (i.e. acid-base, coordination, hydrogen bonding, and adsorption).165 One approach to studying the interactions of guest molecules within hosts involves the use of small probe molecules that can be inserted within the lattice of a solid.166 Using probe molecules containing spin nuclei (e.g. 1 H, 13 C, 1 5N 19 F, 31 P, and 129 Xe) are useful because the corresponding NMR signals are not complicated by quadrupolar interactions.165 In the past, several NMR probes, including 213 C acetone, 167-171 413 C mesityl oxide, 172-175 and 1 5N pyridine, 176-180 have been used to specifically identify and quantify acidic sites in solid acids such as zeolites and amorphous silica-alumina. Unfortunately, the small magnetogyric ratios, low natural abundances, and relatively limited chemical shift ranges of 1 5N (0.4 % abundant) and 13 C (1.1 % abundant) requires either extensive signal averaging or the use of isotopically enriched materials in order to obtain high resolution NMR spectra_ 166,181 To avoid both of these problems, a more suitable probe molecule would incorporate a more sensitive NMR active nucleus such as phosphorus-31.182 The large gyromagnetic ratio and near 100 % natural abundance of 31 P results in a significant increase in NMR signal intensity compared to either 13 C or 1 5N without the use of expensive enriched samples.181, 183,184 Furthermore, the large isotropic chemical shift range and full chemical shift anisotropy can provide useful information regarding the chemical

PAGE 97

86 environments of the 31 P nucleus.181,185,186 Therefore, phosphorous-containing probe molecules are well suited for probing the nature and strength of host-guest interactions. Recently, 31 P MAS NMR spectroscopy has been applied specifically toward for the identification and quantification of Lewis and Brnnsted sites on the surfaces and within solid acids.166,182,187-190 In the past, trialkylphosphines (TAP), and in particular, trimethylphosphine (TMP), have been the probe molecules of choice for the characterization of solid acids by 31 P MAS NMR spectroscopy_ 166,182, 190-194 The coordination of a TAP molecule to an acidic site results in a characteristic chemical shift that is strongly dependent on the Brnnsted or Lewis nature of that site. 186 In fact, 31 P NMR spectroscopy can not only distinguish between Brnnsted-complexed TMP from Lewis-bound TMP but also resolve peaks due to varying local environments of the coordinated probe molecules.166,186 The basicity oftrimethylphosphine (pKa = 5.3 in water) results in the formation of a protonated base (TMPir) upon coordination to any Brnnsted site with a characteristic chemical shift of about3 ppm (referenced to 80 % aphosphoric acid) that is largely invariant with the strength of the acid site_ 181,195 Thus, TMP has been extensively used to determine the presence and quantity of Brnnsted sites in solid acids. Lewis bound TMP exhibits a considerable upfield shift relative to the Brnnsted site.181,186 However, the similar chemical shift of Lewis bound and physisorbed TMP causes difficulty in using this probe to unambiguously determine the presence and population of Lewis sites in a solid_ 181,186 Rapid chemical exchange dynamics between bound and free TMP molecules at room temperature can also lead to uncertainty in the identification and quantification of acid sites_ 181 Furthermore, TMP is

PAGE 98

87 a highly flammable and air-sensitive liquid at room temperature and preparing solid acid standards for quantitative measurements is difficult 181 185 Like trialkylphosphines trialkylphosphine oxides (T APO s) are also suitable basic probe molecules for studying interactions within solid acids as they are able to distinguish Br0nsted sites from Lewis sites.166 181 183 186,188 The intrinsic basicity of trialkylphosphine oxides is on the same order of magnitude as trimethylphosphine.186 However the removal of the phosphorous atom from the basic site leads to a wide range of chemical shifts that vary with the strength of the acid site.186 Unlike the corresponding TAP s TAPO's are solids at room temperature, not susceptible to oxidation and can thus be introduced into a host through solution-state chemistry 181,182 The most commonly used trialkylphosphine oxides are trimethylphosphine oxide (TMPO) and triethylphosphine oxide (TEP0).186 The ability of these materials to measure acidity in both solution and in solids is known and TEPO, for example, has been used as a probe for determining solvent acidity.196 Unfortunately there are few published reports describing the use of trialkylphosphine oxides as probe molecules to study interactions with porous solids.181 The chemical shift of an NMR peak resulting from the interaction of a basic probe molecule with an acid site in a host is primarily due to local changes in the magnetic field around the NMR active nucleus originating from electronic changes in the probe molecule as a whole_ 186 197 198 An acid-base interaction can be viewed as a transfer of electron density from the base to the acid site thus creating overlap between the LUMO the electron deficient orbital on the acid (A) and the HOMO (s) one of the orbitals on

PAGE 99

88 the base that contains a lone pair of electrons (Figure 3-1 ).165, 186 As the adduct bond strength increases, the electron density "flows" away from the basic oxygen atom. The net loss of electron density from the NMR active nucleus due to the formation of an acid/base adduct ( A 8 ) can be observed in the phosphorous-31 NMR.165, 186 E ;-........ ! lpA JL/ 1' / 'Ps ... r Increasing Srength of Acid-Base Interaction Figure 3-1. An acid-base reaction from a molecular orbital perspective. The electron deficient acid, s, to form an adduct, AB As the strength of the acid-base interaction increases, the adduct progressively becomes more covalent in character. Adapted from reference 165. If the basic probe molecule interacts with a Brnnsted acid site, such as water, the resulting adduct bond is characterized by some degree of proton sharing between the acid and the base.186 Consider the hydrogen bonding of the oxygen atom from a trialkylphosphine oxide with the proton of an acidic water molecule. Increasing the strength of the acid site transfers this shared proton to the base. The (water) 0-H and (TAPO) P--0 bond order both decrease while the 0-H bond order increases in the resulting T APO-H20 adduct. The phosphorous atom therefore becomes deshielded shifting the corresponding NMR signal downfield. The stronger acid-base interactions,

PAGE 100

89 the greater the 31 P resonance shifts downfield. However, once the proton is completely transferred to the base, no further downfield shift of the NMR signal will be observed. Instead, for any stronger acid sites, the same chemical shift will be seen in the spectrum.186 The coordination of a base to a Lewis acid, such as a metal ion, will also result in the formation of an adduct bond.186 The strength of the acid site will therefore be directly related to the strength of the adduct bond. Again, consider the effect of coordination of a trialkylphosphine oxide probe molecule directly to a metal center. As the acid strength increases, the electron density gradually transfers from the oxygen atom to the metal ion to a greater extent. The P-0 bond order decreases while the M-0 bond order increases. The phosphorous atom becomes more deshielded and the corresponding NMR resonance progressively shifts downfield. At very high acid strengths no further change in the chemical shift is expected, unless oxygen atom transfer occurs.186 In this case acid site is oxidized thus reducing the trialkylphosphine oxide to the corresponding trialkylphosphine. Recently, Rakiewicz et al characterized the acid sites of amorphous silica alumina, zeolites HY, dealuminated HY, and USY, and gamma-alunina with trimethylphosphine oxide.181 The authors were able to identify and differentiate Lewis and Brnnsted sites and then measure the population of those sites. The trend in assigning chemical shift ranges for Lewis or Brnnsted bound phosphine oxides includes work by Lunsford and Baltusis.166, 187,188,193 Despite consistency in identifying and characterizing Brnnsted sites, each author has assigned a different region to the chemical shifts for the Lewis bound probe.

PAGE 101

90 Chapter Summary The purpose ofthis chapter is to investigate the host-guest properties of two porous network structures, [Ni( 4,4 '-bipy)3(H2O)2](ClO4)2 1.4( 4,4 '-bipy) 3(H2O), 1, and [Co(4,4'-bipy)J(H 2 O) 2 ](ClO 4 ) 2 .4(4,4'-bipy)(H2O), 2, introduced in Chapter 2. In particular, solid-state NMR spectroscopy is shown to be an effective probe of host-guest interactions. Specifically, this work focuses on exchanging pre-existing guests clathrated within the hosts with trialkylphosphine oxide probe molecules. Additionally, 31 P MAS NMR spectroscopy is shown to be an effective tool for investigating host-guest interactions by observing the interactions of these probe molecules within the hosts. Finally, the structural changes within the host that result from the guest exchange is examined by powder X-ray diffraction. Experimental Section Materials Trimethylphosphine oxide (100 %) was purchased from Alpha Aesar. Triethylposphine oxide (98.0 %) and tri-n-propylphosphine oxide (99.9 %) were purchased from Strem Chemicals. Hexanes (99 .9 % ) and ethyl acetate (99 .9 % ) were purchased from Fisher Chemicals. Details regarding the preparation of crystalline samples of the network solids, 1 and 2, are described in Chapter 2. All reagents were used without any further purification. Sample Preparation A generalized procedure was employed for the guest exchange reactions involving the host network solids 1 and 2. All reactions were performed under ambient laboratory conditions. A known amount of TMPO was dissolved in 50.0 mL of a 50 / 50 % vol. mixture of hexane and ethyl acetate, while TEPO and TPPO were dissolved in

PAGE 102

91 50.0 mL of hexane. Crystalline samples of the network solids were ground into fine powders and suspended within the solution containing dissolved probe molecule. The standard initial loading concentration, referred to as the insertion ratio, used in the preparation of all samples was one mole of probe molecule per mole of host. This suspension was then stirred for approximately two hours. The resulting powder was coHected by gravity filtration and a portion of the remaining solvent was saved for gas chromatography (GC) analysis. The powder was then suspended in pure solvent without any dissolved probe molecule and stirred for an additional hour. Again, the product was collected by gravity filtration and a portion of the remaining solution was saved for GC analysis. The powder was air dried before any further characterization. In addition, samples of 1 with variable initial loading concentrations of TMPO were also prepared by an analogous method. Insertion ratios varying from 0.01 moles to 2.0 moles of TMPO per mole of host were utilized. Gas Chromatography Analysis All gas chromatography experiments were performed using a Hewlett Packard 5890 Series II equipped with a hydrogen flame ionization detector. Component concentrations from both the insertion and rinsing solutions described above were determined relative to an internal standard, cyclohexanone. A 0.5 L portion of the hexane solution was injected onto the column (25 m, 0.33 m film thickness, 0.20 mm ID HP-5 (Crosslinked 5 % PH ME Siloxane)). The column, initially at 80 C, was immediately ramped to a temperature of 250 C at a rate of 25 C / min and held for 5 mm.

PAGE 103

92 NMR Spectroscopy All 31 P NMR experiments were acquired on a Bruker Avance 400 MHz NMR Spectrometer operating at 161.976 MHz. A 4.0 mm triple resonance MAS probe was used for conventional work and a 2.5 mm double resonance MAS probe for fast-spinning experiments. A 90 x pulse width of 4.0 sand a sweep width of 140 kHz were used in the acquisition of all spectra. An inversion recovery experiment on the host network solids containing the TAPO probe molecules showed that the spin-lattice relaxation times for the 31 P sites were small (20-30 ms), hence a 100 ms recycle delay time was used. All resonances were referenced to 85 % a-phosphoric acid. Powder X-Ray Diffraction All powder X-ray diffraction experiments were performed at the Major Analytical Instrumentation Center, University of Florida, using a Philips APD 3720 X-ray powder diffractometer with the CuKa line, A-= 1.54 A, as the X-ray source. Finely powdered samples of both the network solids and the network solids containing the probe molecules were mounted onto glass slides with an amyl acetate/ Collodion adhesive mixture. All samples were scanned at a rate of 0.2 per minute from 4 to 20 Results and Discussion Sample Preparation Solution-phase guest exchange reactions were utilized in order to facilitate the uptake of the probes within the hosts. Crystalline samples of 1 and 2 were ground into fine powders in order to provide a maximum surface area for the probe molecules to enter the host. Non-coordinating, chemically inert solvents (hexane and ethyl acetate) were used as reaction media to ensure that only the probe molecules are taken into the host. Upon completion of the guest exchange reactions, the products were washed with solvent

PAGE 104

93 to remove any TAPO s physisorbed onto the surfaces of the host. No additional loss of bipy and or probe was observed suggesting that the probes must be bound strongly to the host. The insertion ratio is the molar quantity of trialkylphosphine oxide probe taken in per mole of host (or host metal ion). A standard insertion ratio of one mole ofTAPO per mole of host employed in most guest exchange reactions was chosen to ensure that a sufficient quantity of the phosphorus-containing probes are taken within the host to be easily be detected by 31 P NMR spectroscopy without extensive signal averaging The addition of excess quantities of probe was avoided to minimize the structural rearrangements resulting from the guest exchange. In order to investigate the possibility of the trialkylphosphine oxide probe molecules populating different or multiple acid sites as well as monitor structural changes that occur as function of quantity of guests exchanged within the hosts samples of 1 were prepared with variable insertion ratios of TMPO. The initial loadings cover the range of0.01 moles to 2.0 moles ofTMPO per mole of 1. Guest Loss From the Network Solids For all guest exchange experiments the uptake of the trialkylphosphine oxides by 1 and 2 is always accompanied by the loss of 4,4' -bipyridine. The displaced bipy molecules are presumably the clathrated guests not the coordinated ligands. In order t o determine the origin of the bipy loss, i.e ., from the TAPO or the solvent, blank reactions those without the presence of the probes were performed in parallel to the guest exchange reactions. As shown in Figures 3-3 and 3-15 negligible bipy loss was observed from the control experiments indicating that the probe molecules, not the solvent displace the bipy from the hosts.

PAGE 105

94 The X-ray powder diffraction patterns of 1 and 2 before and after the blank reactions (without the probe molecules) are shown in Figure 3-2. The Bragg peaks of the powder patterns do not shift, appear, or disappear to any appreciable extent. The solvent does not cause any bipy to be released from the host and no appreciable structural rearrangements occur within the porous solids. However, the solvent causes the hosts to become less crystalline however, as many peaks are broadened decrease in intensity due to a small loss of structural coherence, particularly in 2. Guest Exchange Investigations of Compound I Involving TMPO Gas Chromatography Results The gas chromatography experiments monitoring the guest exchange of bipy with TMPO from 1 are summarized in column bar graph format in Figure 3-3. The insertion ratio corresponds to one mole ofTMPO per mole of host. The TMPO uptake by 1 is essentially complete, i.e., the host takes in essentially all of the initial available quantity of the probe. Furthermore, the displacement of a significant quantity of bipy accompanies the TMPO uptake suggesting that the probe is actually taken within the host as opposed to physisorbing or chemisorbing onto the surfaces of crystallites of 1. A schematic of the exchange of bipy for TMPO within the layered, porous host 1 is summarized schematically in Figure 3-4. The exchange of guests within 1 (and 2) is essentially governed by two competing factors: the interactions between TMPO and the host versus the bipyridine and the host and steric and size constraints of the probes, guests, and pores. Since TMPO is potentially both a Brnnsted and Lewis base, the principal driving force for its uptake may be the formation of strong acid-base and donor-acceptor adducts between the probes and the host, however other stabilizing interactions may lead to absorption of the probe as

PAGE 106

95 1400 -A H o s tl 1200 B H ost I + Hexane W as h '-' 1000 .c ..... 800 B A 11) .E 11) .6 600 400 A 200 0 4 6 8 10 12 14 16 18 20 Angle / 20 (Degrees) 4000 3500 A H os t 2 B H ost 2 + Hexane W as h 3000 2500 B .c .. fll 1 2000 B 1500 i 1000 A 500 0 4 6 8 10 12 14 16 18 20 Angle / 20 (Degrees) Figure 3-2. Powder X-ray diffraction patterns before (black) and after (red) washing with hexane. A) Compound 1. B) Compound 2.

PAGE 107

96 well. From the crystal structure of 1, these potential acid sites correspond to coordinated and lattice water molecules and metal ions. The size and steric bulk of TMPO ( and especially TEPO and TPPO with the longer alkyl chains), the limited size and space present within the pores and between the sheets, and the presence of clathrated guests may act to hinder the probe uptake by the host. Since perchlorate counterions, lattice waters, and bipy molecules occupy much of the void space within the pores and between the sheets, evidently insufficient space is available within the host to accommodate both TMPO and the existing guests. As a result, the TMPO displaces the bipy guests. The interactions between the TMPO and the hosts must therefore be stronger and more favorable than the packing forces and host-guest interactions holding the bipy's within the host and the probes. Aiding and driving the guest exchange are the stabilizing solvation interactions between the hexane and the displaced hydrophobic bipy molecules. NMRResults In order to investigate the interactions between the probe and the host after the guest exchange, experiments were performed to determine the 31 P MAS NMR chemical shift range for TMPO upon successive dilution water since the phosphine oxide presumably interacts with water molecules present in 1. The shaded regions in the NMR spectra represent this range. The 31 P NMR spectrum for crystalline TMPO is shown in Figure 3-5 A. One peak at 39.7 ppm is observed, corresponding to crystalline, non-interacting TMPO. Figure 3-5 B shows the static 31 P NMR spectrum upon adding a small aliquot of water 41.2 ppm and 45.9 ppm. As shown in Figures 3-5 C and 3-5 D, the peak at 45.9 ppm progressively loses intensity while a new peak at 51.0 ppm appears and grows in intensity as the TMPO solution ( containing approximately 79 % and 71 % probe by mass,

PAGE 108

97 Uptake ofT APO Probe(% b y mass) "0 Displacement ofBi p y (per mo l ofNi(II)) ti) "' 1. 00 aS ti) 1.0 'ii 0 77 >, 0 8 !:0 6 1 i::o --0 6 ti) .:.: <,:I ..., Q., 0.4 0 ti) .0 0 2 0 0 00 1 53 .... p... 0 0 Host I TMPO TEPO TPPO Sample Figure 3-3 The gas chromatography results monitoring the guest exchange reactions in Host 1. The left-hand columns denote the fraction of probe taken in by the host and the right hand columns represent the relative quantity of clathrated bipy released. TM.PO Probe 0 II p + /cj'\.Me Me B i py Bost+Bipy Bost+ Probe ... + Figure 3-4. The guest exchange of 4 4 '-Bipyridine with TMPO in Hosts 1. respectively) is further diluted. In the high dilution limit (ca 100 L of water) corresponding to approximately 43 % TMPO by mass, all intermediate resonances at 45.9 ppm have disappeared and only a single peak at 51.0 ppm remains (Figure 3-5 E). The downfield shifting of the TMPO 31 P resonance is consistent with that reportect 165 166,181 183 186 188 For aqueous solutions ofTMPO the peak with the highest chemical shift

PAGE 109

98 represents the strongest possible acid-base interaction with free water and corresponds to the hydrated solute. The peaks that appear in the intermediate regions may, in fact, be comprised of two distinct resonances, one corresponding to the hydrated TMPO, and the other from crystalline TMPO. The chemical exchange at room temperature between these two species is fast on the NMR time-scale so the two peaks coalesce and a single resonance is observed. At higher TMPO concentrations, the corresponding NMR peaks are dominated by undissolved fraction and thus only partial downfield shifting is observed. As the TMPO is diluted with water, the hydrated TMPO peak begins to dominate and the average of the wet and dry peaks shift downfield. When there is no crystalline TMPO remaining and the entire sample is hydrated, no further downfield shifting is be observed. The downfield shifting of the 31 P resonances as a function of dilution of trialkylphosphine oxides with water can provide a means for assigning NMR peaks when TMPO is used to probe interactions with 1. After the exchange of bipy with TMPO, 31 P resonances appearing in this shaded region indicate that the TMPO is interacting with an acid site of strength similar to water. Peaks shifted downfield with respect the former are then likely due to the probe interacting with stronger acid sites. Figure 3-6 schematically depicts some of the possible physisorption interactions, e g. the probe adsorbing to the surface or between the layers, and chemisorption interactions, e.g., hydrogen bonding to lattice waters or aqua ligands or direct coordination to the metal centers. Note that this scheme does not imply any specific orientation or position of the probe with respect to the host or metal centers.165 Figure 37 shows the 31 P MAS NMR spectrum of 1 after the exchange of bipy with one mole ofTMPO per mole of host. A spinning speed of 8.0 kHz effectively

PAGE 110

99 separates the spinning sidebands, identified with asterisks, from the characteristic peaks corresponding to the various chemical environments of the 31 P nucleus. Two broadened peaks are observed between the chemical shift range of 40.0 ppm and 80.0 ppm. The resonance at 49 .6 ppm appears within the shaded region for TMPO interacting with water. The peak is therefore assigned to the probe molecules binding to acid sites of comparable strength to water. Based on the crystal structure of 1, these weak acid sites may be uncoordinated lattice water molecules. The peak at 68.0 ppm corresponds to TMPO interacting with an acid site stronger than unbound water molecules since this peak is clearly shifted downfield with respect to the former resonance. From the crystal structure of 1, the strong acid sites are most likely the coordinated water molecules. The assignment of these peaks is consistent with those of Baltusis and Rakiewicz in their NMR studies of solid acid zeolites and amorphous silica-alumina substrates with trialkylphosphine oxides.166,181,188 Furthermore, the NMR spectrum in does not indicate the presence of surface bound, physisorbed TMPO to crystallites of 1, expected at 39.7 ppm.166,181,188 Evidently the TMPO does not coordinate directly to the Ni(II) ions. Otherwise, significant broadening and shifting of the resonances would result, perhaps to the extent that the spectrum would not be observed at alt 199 This conclusion is further justified since the guest exchange reaction causes no observable change the color of 1. The direct coordination of the TMPO to the metal centers necessarily involves the replacement of the aqua and /or bipy ligands. However, the GC experiments did not indicate the presence of displaced water. The limited size and space present within the pores and between the sheets likely hinders ligand replacement and thus rendering the metal ions

PAGE 111

(AJ (BJ (CJ (DJ (EJ ss so 45 40 100 TMJ>o Dry ssTMpo ByAfass Figu,,, 3-5. The 31 p AfAs /VMR s With Water. A.) Dry TMPo. B) 8 ~~JMPo_as a resuJr of disso/vii,g a,,d dilutiog D) 71 % TMPo solution E) 43 % 'T'A rho 1 so/utJon. C) 79 % TMPo solution. 0 1 vw so Ution. 35 .P.Plll

PAGE 112

101 less accessible acid sites compared to the water molecules for TMPO. In fact, the presence of relatively well-defined peaks suggests that the phosphorous nuclei are not in close proximity to the metal centers at all. Based on the insertion ratio of one mole of TMPO per mole of host, the TMPO should preferentially bind to the strongest available acid sites, one or both of the aqua ligands. Then, in the NMR spectrum, only one or two peaks at approximately 70 ppm corresponding to these strong-acid interactions is expected. However, the NMR spectrum also shows evidence of the probe interacting with both weak acid sites (the lattice waters) along with the strong acid sites. Evidently, due to the above-mentioned size and steric constraints of the probe, clathrates, and pores, the TMPO is unable to access all of the strong acid sites and distribute themselves uniformly throughout the host. Consequently, the probe molecules interact with more accessible weaker acid sites, the lattice water molecules, as well. X-Ray Results Figures 3-8 A and 3-8 B show the powder X-ray diffraction patterns of 1 before and after the exchange of bipy with one mole ofTMPO per mole of host, respectively. The powder patterns are clearly different from one another indicating the guest exchange produces significant structural rearrangements within 1. Recall that the characteristic porous network structure of 1 is sustained, in large part, by stabilizing interactions between the host and the guest bipy molecules. TMPO is topologically and chemically different than bipy and is not expected to fulfill the same role as a guest. As a result the structural changes occur in order to accommodate the presence of the TMPO as well as fill the void spaces left by the vacated bipy guests. The host also becomes less crystalline due to a decrease in the structural coherence as many Bragg peaks have broadened and

PAGE 113

TMPO 0 II /p" + Me J Me Mc O:::~n~tll~e~S~u:::r~f~ce /Me -O= P--Me Me Within the Host 0 II p /e j "' Me Me 102 Host ~isorption Lattice Water Site /e H-O=P-Me 0/ "'H Metal Coordination Site Me Coordinated Water Site -M--1 /\ H H i 0 II /1"'Me I Me Me Figure 3-6. The possible interactions ofTMPO with Hosts 1 and 2. Physisorption interactions on the surface or within the host are shown on the left. Chemical interactions with the host, such as acid-base interactions with the acidic water molecules or direct coordination to the metal centers, are shown on the right.

PAGE 114

(C) Host 1 +TPPO (B) Host 1 +TEPO (A) Host 1 +TMPO 150 103 * 100 50 O ppm Figure 3-7. The 31 P MAS NMR Spectra of Host 1 after the guest exchange of 4,4' Bipyridine. A) Exchange with TMPO. B) Exchange with TEPO. C) Exchange with TPPO. The spinning sidebands are denoted by asterisks(*).

PAGE 115

104 lost intensity. Because of the relatively amorphous nature of the product, the nature of the structural changes cannot be readily determined. However, the characteristic porous network structure of 1 has not completely collapsed. If this were the case, observable color or solubility changes might occur and essentially no diffraction would be observed. A few Bragg peaks present in the original host material remain in the TMPO sample but other reflections have either disappeared or moved as a result of the guest exchange. Note the non-correspondence of many Bragg peaks from the 7 to 9 and 14 to 20 regions. For example, the three broadened, low-intensity peaks at 7.3 7.8 and 8.0 observed after the guest exchange are not present in the original host structure. Additionally, the small, broadened peak at 14.3 and intense, unresolved peak at 19.0 both disappear after the guest exchange, being replaced by a group of coalesced peaks between 14.5 and 15.5 and two intense, unresolved reflections at 19.5 and 20.0 respectively. However, a few Bragg peaks seem to be common to both samples. The two intense peaks at 12.5 and 18.8 as well as the two moderately intense reflections at 16.0 and 18.3 are preserved after the guest exchange, although somewhat broadened and less intense. The most striking feature is the movement or disappearance of the [2 0 O] reflection at 10.0 corresponding to the interplanar spacing. In response to the bipy loss and probe uptake, the sheets should move and a corresponding change in position of the [2 0 O] peak is expected. Unfortunately, this peak cannot be readily identified the product because the ordered layering of sheets has been disrupted and randomized. The X-ray diffraction results seem to suggest two possibilities concerning the nature of the guest-exchanged host. One possibility is that only the overall porous network structure of 1 essentially remains intact after the guest exchange but only the

PAGE 116

105 layering of the sheets has become disrupted and randomized within the solid. As a result, a loss of crystallinity and some structural rearrangements occur within the host due to the non-uniform dispersal of the probe throughout the host causing significant structural disorder in the system. The other possibility is that the guest-exchanged product may be a mixture of two distinct materials. One fraction retains characteristics of the original host and the other portion corresponds to a new, unidentified material. Because the guest exchange does not occur uniformly throughout the host, likely due to the steric constraints described above, the probe molecules are not distributed homogeneously throughout the host and only some of the bipy guests are lost. Therefore, some portion of the original material is unaffected while the other fraction undergoes significant structural rearrangements. It is unknown to what extent the original host structure has been preserved in the affected fraction. 3200 -A H o st I BH os t I +TMPO 2800 -C Ho s t l + TEPO 8 D DH o st I +TPPO $ 2400 c 2000 C -~ ..s 1600 1200 B -~ 800 400 A 0 4 6 8 10 12 14 16 18 20 Angle / 20 (Degrees) Figure 3-8 Powder X-ray diffraction patterns of Host 1 before and after the guest exchange of 4,4 '-Bipyridine. A) No probe. B) Exchange with TMPO. C) Exchange with TEPO. D) Exchange with TPPO.

PAGE 117

106 Guest Exchange in Compound 1 Involving Variable TMPO Concentrations Gas Chromatography Results In order to further examine the changes in population of the various acid sites by TMPO, samples were prepared by exchanging bipy with variable quantities of the probe from 1. Specifically, insertion ratios ranging from 0.01 moles to 2.0 moles of TMPO per mole of host were employed. The gas chromatography experiments monitoring the guest exchange from 1 as a function ofTMPO concentration are depicted in Figure 3-9. Throughout the concentration range of 0.01 moles to 2.0 moles ofTMPO per mole of host, the probe uptake by 1 is essentially complete and is accompanied by a displacement ofbipy. From the plot in Figure 3-9, the TMPO uptake by the host is relatively linear. However, the quantity of bipy released initially increases but quickly begins to level out at a loading of 0.1 moles of probe per mole of host and then remains relatively constant. Then, at a concentration of 1.0 mole of probe per mole of host, the quantity of bipy released gradually increases again. Even at relatively low probe concentrations (0.1 moles ofTMPO per mole of host and higher) evidently sufficient quantities of TMPO are present within the pores to hinder any further bipy loss. If these escape routs for bipy are blocked, then the entry paths for the probe should also be hindered preventing the probe from further penetrating the host. Since the strong acid sites deep within the host are now no longer accessible but all of the available TMPO is still taken within 1, the probes must then be populating weaker, but more accessible, acid sites within the host. At high probe concentrations (1.0 mole per mole of host), the uptake of excess quantities of probe facilitates the displacement of more bipy. At low probe loadings, significant quantities of bipy are released despite the small amounts of probe taken within the host. This result is surprising since the uptake of small

PAGE 118

107 amounts of the probe is expected to displace correspondingly small amounts ofbipy. Thus the interactions between 1 and the TMPO are much stronger and more favorable than the preexisting host-bipy interactions. At higher probe loadings, the opposite effect is observed; the amount of bipy lost relative to the TMPO taken in by the host is now small. NMRResults Figures 3-10 and 3-11 depict the 31 P MAS NMR spectra of 1 as a function of the TMPO loading. Figure 3-10 shows selected NMR spectra over the entire measured 1.8 rJ> 1.6 rJ> o--. ...:l t; 1.4 ;;,,.. 0 p,. ..c: 1.2 .... Ill 0 s 1.0 0 ..... 0 8 (Ii rJ> ..... 0 p,. ::, 0 6 0 00 0 0 0 4 /:l., s ::E -0 .2 E-< 0.0 -0 2 o TMPO Uptake fl Bip y Loss 0 0 6. 0 6. 6. 6. 6. 0 (;J 0.0 0 5 0 0 6. 1.0 1.5 TMPO Insertion Ratio (mmol TMPO I mmol host) 0 2.0 Figure 3-9. Exchange of 4,4' -Bipyridine with variable quantities ofTMPO within Host 1. The loss of 4,4' -Bipyridine is shown by the open triangles and the TMPO uptake is represented as the open circles. loading range of 0.01 moles to 2.0 moles of probe per mole of host. Figure 3-11 focuses on a specific loading range of 0.05 moles to 1.0 mole of probe per mole of host to illustrate change in the spectra as the quantity of TMPO taken in by 1 increases

PAGE 119

108 ( discussed below). At low TMPO loadings (0.01 moles to 0.1 moles per mole host), poor signal-to noise ratio in the NMR spectra is observed due to the sparse presence of the probe (Figure 3-10). In each spectrum two peaks are observed, one at 49 .6 ppm and another at 68.0 ppm corresponding to the interaction ofTMPO with multiple types of acid sites. These spectra closely resemble the NMR spectrum in Figure X described above in both the presence and positions of the two peaks. As explained earlier, the peak at 49.6 ppm is assigned to the TMPO interaction with weak acid sites, presumably the unbound, lattice water molecules in 1. The peak at 68.0 ppm corresponds to TMPO interacting with stronger acid sites, the aqua ligands. Note that the peaks in these spectra are of comparable intensity suggesting that, within this concentration regime, the population of both the weak and strong acid sites accessible by the probe is relatively equal. The presence of two peaks within this low concentration regime is similar to the spectra observed at the higher loading of one mole of probe per mole of host. The TMPO, present in very limited quantities, was expected to preferentially bind to the strongest available acid sites and consequently produce only the single NMR peak at 68.0 ppm. However, even at the lowest TMPO loadings, the steric constraints prevent the probe from accessing all of the strongest acid sites and thus are not uniformly distributed throughout the host. As a result, in addition to binding to the stronger acid sites, the probe molecules interact with weaker, but more accessible, acid sites as well thus giving rise to the additional NMR peak at 49 .6 ppm. At intermediate TMPO loadings (0.5 moles to 1.0 moles per mole of host), the signal-to-noise ratio in the NMR spectra has improved significantly due to the greater

PAGE 120

109 amount of probe present (Figure 3-11 ). As the concentration of TMPO increases from 0.5 moles to 0.6 moles per mole of host, a rather dramatic change occurs in the corresponding NMR spectra. The peak at 68.0 ppm, assigned to TMPO binding to the strong acid sites (aqua ligands) abruptly decreases in intensity (on an absolute intensity scale) while a similar, sudden increase in intensity is also observed from the peak at 49.6 ppm, corresponding to the probe interacting with the weak acid sites (the lattice water molecules). The change in intensity is related to a change in population of the various binding sites by TMPO in the host. The TMPO suddenly begins populating the weaker acid sites in favor of the stronger acid sites. This event could be due to structural rearrangements occurring as a result of the guest exchange. If the pores or sheets are opened in a manner that exposes the lattice water molecules, then these weaker acid sites may be more accessible to the TMPO thus accounting for the changes in population of the acid sites. At the high TMPO loadings (1.0 moles 2.0 moles per mole of host) the intensity of the peak at 49.6 ppm corresponding to TMPO binding to unbound acid sites (the lattice water molecules) levels off while the peak at 68.0 ppm corresponding to the probe interacting with stronger acid sites (the aqua ligands) progressively decreases in intensity (Figure 3-10). As more TMPO is taken within the host, a small shoulder at 42.2 ppm appears and is assigned to TMPO physisorbed to surfaces of 1. Furthermore, two new peaks at 74.4 ppm and 79.4 ppm are observed as well, assigned to TMPO interacting with chemically inequivalent strong acid sites, again the aqua ligands. Recall from the crystal structure, the two protons on the water ligands are not equivalent since one hydrogen atom bonds to a nearby lattice water molecule and the other interacts with the terminal

PAGE 121

110 nitrogen atom from a neighboring monocoordinate bipy ligand. As more TMPO is taken in, the more accessible unbound water sites become saturated. In response, the remaining available probes begin to physisorb to the surfaces of the host as well as populate stronger, albeit less accessible, acid sites. X-ray Results The powder X-ray diffraction patterns of 1 before and after the exchange of bipy with variable quantities of TMPO are shown in Figure 3-12 (A-J). Comparing the diffraction patterns for each guest-exchanged product to the original material indicates that significant structural rearrangements have occurred within the host. From the loading range of0.01 moles to 0.6 moles of probe per mole of host, the diffraction patterns appear the same indicating that the guest exchange produces essentially the same material (Figures 3-12 B-3-12 F). The products are amorphous because many of the diffraction peaks are broadened and have lost intensity. There is virtually no correspondence of the Bragg peaks from the guest-exchanged samples with the original material. In particular, the interplanar [2 0 O] reflection has disappeared and the layered structure of 1 has been disrupted. The structural rearrangements occur in response to fill the void spaces left by large quantity of bipy released from 1 compared to the small amount of TMPO taken in. Insufficient quantities of TMPO are available to fulfill the role as guest and occupy the void spaces left by the vacated bipy. Additionally, the non-uniform, random distribution of the TMPO produces structural disorder within the host leading to the broadened, low intensity reflections in the diffraction pattern. A structural change occurs between the loadings of 0.6 moles and 0.7 moles of TMPO per mole of host (Figures 3-12 F and 3-12G). Although the structure of the product resembles that of the samples from the lower probe loadings, an increase in the

PAGE 122

111 n =2.0 n=l.5 n = 1.0 n =0.5 n = 0.1 150 100 so * n moITMPO/ 1 mol Host 1 * 0 50 ppm Figure 3-10. The 31 P MAS NMR Spectra of Host 1 after the guest exchange of 4,4' Bipyridine with variable quantities ofTMPO. The spinning sidebands are denoted by asterisks (*).

PAGE 123

112 n = 1.0 n=0.9 n =0.8 n=0.7 n =0.6 n=0.5 150 100 50 n molTMPO/ 1 mol Host 1 * * * 0 50 ppm Figure 3-11. The 31 P MAS NMR Spectra of Host 1 after the guest exchange of 4,4' Bipyridine with variable quantities ofTMPO, focusing between the concentration range of 0.5 moles to 1.0 moles ofTMPO per mole of host. The spinning sidebands are denoted by asterisks (*).

PAGE 124

113 structural coherence is observed and the product is significantly more crystalline. The nature of this transformation is, however, uncertain. The host releases less bipy compared to the lower probe loadings and sufficient TMPO is available that it can more efficiently fill void spaces left by the displaced bipy and the structural rearrangements to fill the resulting void spaces thus reducing the severity of the structural rearrangements by the host. The increased crystallinity could also be due to a more homogeneous distribution of the TMPO throughout the host. Note that the increase in crystallinity of the sample occurs at the same concentration range that the change in population of the acid sites by the probe in the NMR spectra. The significance of this correlation is, however, unclear. At the high TMPO loadings (1.0 moles to 2.0 moles per mole of host), no further structural rearrangements are observed and the product remains relatively crystalline (Figures 3-12 H3-121). However, as the TMPO loading increases to 2.0 moles per mole of host, the guest-exchanged sample becomes amorphous again which is consistent with an increased quantity of bipy lost by 1. Guest Exchange Investigations of Compound 1 Involving TEPO and TPPO Gas Chromatography Results The gas chromatography experiments monitoring the guest exchange of bipy with TEPO and TPPO within 1 are summarized in column bar graph format in Figure 3-3. Recall that 1 takes in essentially all of the initial available quantity of TMPO. In contrast, 1 takes in approximately 75 % of the available quantity ofTEPO and only a small quantity (15 %) ofTPPO. Despite the (slightly) greater predicted acidity of both TEPO and TPPO, the relatively long alkyl chains act to inhibit their uptake by the 1 and 2. The TMPO "fits" within the pores and between the sheets better than TEPO and TPPO. Only

PAGE 125

32 00 2 800 ,.__, 2400 ~ 2 000 1600 ..9 -! .s 1 2 00 800 400 0 114 a small quantity of bipy is displaced by TEPO and TPPO suggesting that this probe may not be penetrating the host as well as TMPO but instead collects at or near the surfaces of crystallites of 1. AHost l 32 00 AHost l BTMP00 0 1 Eq ETMPO0 SEq C TMPO 0 05 Eq 2 800 F TMPO 0.6 Eq D DTMPO0 l Eq G GTMP00 7Eq 2 400 ,.__, ~ 2000 C j F ..9 1 6 00 B 1 2 00 .c E "' 800 A 400 A 0 4 6 8 10 1 2 14 16 18 2 0 4 6 8 10 12 14 1 6 18 20 Angle / 20 (Degrees) Angle / 20 (De g rees ) 32 00 AHost l -HTMPO I. 0Eq 2 800 J I TMPO 1. 5 Eq KTMP02 0E q 2 400 ,.__, 2 000 I ;;; 5 1 6 00 i:: 1 2 00 H .c "' 0 800 C:G 400 0 4 6 8 10 12 14 1 6 18 2 0 An g le / 20 (Degrees ) Figure 3-12. Powder X-ray diffraction patterns before and after the guest exchange of 4 4' -Bipyridine with variable quantities ofTMPO per mole of Host 1. The quantity of probe taken in corresponds to A) No TMPO. B) 0.01 moles of TMPO. C) 0 05 moles of TMPO. D) 0.1 moles ofTMPO E) 0.5 moles ofTMPO F) 0.6 moles ofTMPO. G) 0.7 moles ofTMPO. H) 1.0 mole ofTMPO. I) 1.5 moles ofTMPO. J) 2 0 moles of TMPO.

PAGE 126

115 NMRResults The NMR spectra of crystalline and dilute aqueous samples of TEPO and TPPO, shown in Figures 3-13 and 3-14, respectively, exhibit a similar shifting of dry and wet peaks, as does TMP0.165,166,183,186,188 Note the presence of peaks corresponding to both crystalline and hydrated TEPO, shown in Figure 3-13 A, due to the hygroscopic nature of TEPO and TPPO. The NMR spectrum of 1 after the guest exchange ofbipy with 0.75 moles of TEPO per mole of host is shown in Figure 3-7 B. The spectrum shows a single broad peak at 58.9 ppm. This peak is assigned to TEPO interacting uncoordinated lattice water molecules of 1 since this resonance appears in the shaded region corresponding to TEPO interacting with free water. The lack of any additional peaks downfield with respect to the 58.9 ppm resonance indicates that TEPO is not populating any stronger acid sites, such as coordinated water molecules. This result is not readily explained and is not consistent with the findings of Baltusis and Rakiewicz. l 66,181,188 The TEPO is unable to access to stronger acid sites within the host possibly due to steric constraints. The probe therefore only interacts with lattice water molecules within the host or surface physisorbed water if the TEPO only collects at the surface. Furthermore, there is no evidence for direct coordination of TEPO to the metal centers due to the presence of relatively narrow, well-defined peaks in the NMR spectrum as well as the lack of any observable color change in the host resulting from the guest exchange. In fact, the relatively mild broadening of the peaks indicates that the magnetic interaction between the 31 P nucleus and the paramagnetic Ni(II) ions is weak suggesting that TEPO is not in close proximity to the metal centers.

PAGE 127

116 The NMR spectrum of 1 after the guest exchange ofbipy with 0.15 moles of TPPO per mole of host is shown in Figure 37 C. The initial quantity of the probe corresponds to 1 mole of TPPO per mole of host. The relatively low signal-to-noise ratio correlates with the GC measurements confirming that 1 takes in only a small quantity of TPPO. The NMR spectrum shows two broad peaks at 52.7 ppm and 105.3 ppm. Since the upfield peak appears in the shaded region corresponding to TPPO interacting with unbound water, this resonance is assigned to TPPO interacting with lattice water molecules present in 1. This peak could be due to the interaction ofTPPO with surface bound physisorbed water as well if the probe is not taken within the host. This interaction could also be due to TPPO interacting with surface physisorbed water as well. The downfield peak could be due to TPPO interacting with a coordinated water molecules at or near the surfaces of 1, however, the assignment ofthis peak is not clear since a similar downfield peak was not observed with TEPO. If the probe is in close proximity to the metal ions, this resonance could also result from through-space interactions between the dipolar 31 P nuclei and the paramagnetic Ni(II) ions and is expected to experience considerable upfield or downfield shifting depending on the orientation of the phosphorous nuclei with respect to the metal centers. Again, there is no conclusive evidence for direct coordination of TPPO to the metal centers. Therefore, the assignment of the downfield resonance at 105.3 ppm to a strong acid-base interaction between the probe and the host rather than a magnetic effect is more reasonable. The resonances corresponding to both TEPO and TPPO interacting with unbound, lattice water (58.9 ppm and 52.7 ppm, respectively and are shifted downfield compared to the corresponding TMPO peak (49.6 ppm). Since TEPO and TPPO are both stronger

PAGE 128

117 bases than TMPO, the former probes interact more strongly with the unbound water sites resulting in the corresponding 31 P NMR resonances to shift down.field to a greater extent. However, note that the TEPO resonance is significantly shifted down.field more than the TPPO peak, despite the greater expected basicity ofTPPO. The reason for this discrepancy is unclear though. X-Ray Results Figures 3-8 C and 3-8 D show the powder X-ray diffraction patterns for 1 after the exchange ofbipy with 0.75 moles ofTEPO and 0.15 moles ofTPPO per mole of host, respectively. The powder patterns of the TEPO and TPPO samples are quite similar indicating the guest exchange causes the same structural changes and produces essentially the same product. However a comparison of the powder patterns, particularly the 14 to 20 range shows that the guest-exchanged TMPO product is different from the TEPO and TPPO samples. Comparing the powder patterns of the TEPO and TPPO samples with the corresponding pattern of the original host indicates that the guest exchange produces significant structural rearrangements as well as reducing the crystallinity of 1 The extent of the structural changes is much greater than expected considering the relatively small quantity of bipy released compared to the TMPO material. Had TEPO and TPPO simply physisorbed or chemisorbed to the surfaces of crystallites of 1 the corresponding powder diffraction patterns would more closely resemble that of the original host material A few Bragg peaks present in the original host material remain in the TEPO and TPPO samples but other reflections have either disappeared or moved as a result of the guest exchange Just as in the TMPO sample the nature of the structural changes that occur in 1 as a result of the bipy loss and TEPO and TPPO insertion cannot be readily determined.

PAGE 129

118 (A) (B) (C) (D) (E) 65 60 55 50 45 TEPO Dry 88% TEPO By Mass 78%TEPO By Mass 65%TEPO By Mass 56%TEPO By Mass 40 ppm Figure 3-13. The 31 P MAS NMR Spectra ofTEPO as a result of dissolving and diluting with water. A) Dry TEPO. B) 88 % TEPO solution. C) 78 % TEPO solution. D) 65 % TEPO solution E) 56 % TEPO solution

PAGE 130

119 (A) (B) (C) (D) (E) 60 55 so 45 40 TPPO Dry 88%TPPO By Mass 71 %TPPO By Mass 60%TPPO By Mass 52 %TPPO By Mass 35 ppm Figure 3-14 The 3 1 P MAS NMR Spectra ofTPPO as a result of dissolving and diluting with water A) Dry TPPO. B) 88 % TPPO solution. C) 71 % TPPO solution. D) 60 % TPPO solut i on. E) 52 % TPPO solution

PAGE 131

120 Host-Guest Exchange of Compound 2 Involving TMPO, TEPO, and TPPO Gas Chromatography Results The gas chromatography experiments monitoring the guest exchange of bipy with TMPO, TEPO, and TPPO within 2 are summarized in column bar graph format in Figure 3-15. Just as in 1, 2 takes in essentially all of the available quantity ofTMPO but releases more bipy than 1. Furthermore, significantly more TEPO and TPPO are taken within 2 than 1 and the accompanying bipy loss proceeds to a greater extent as well. In fact, the TEPO uptake is by 2 is nearly complete. Cobalt(II) is more oxophilic and susceptible to oxidation than nickel(II) and this propensity is exhibited in the guest exchange reactions. Given the same initial available quantities of the T APO' s, the uptake of the oxygen containing probe molecules by 2 is much greater compared to 1. Therefore, despite the steric and size constraints that apparently hinder the uptake of TEPO and TPPO by 1, these obstacles are overcome in 2 due to the oxophilic nature of the host. NMRResults The large quadrupole moment of the Co(II) nucleus along with the coupling of the dipolar interaction of the 31 P nucleus from the trialkylphosphine oxides with the magnetic moment due to the unpaired electrons of the metal centers produces significant broadening and shifting of the characteristic resonances in 2.199 In order to unambiguously identify and resolve the peaks from spinning sidebands, high spinning speeds (>10 kHz) were required. Because of these magnetic interactions, the assignment of the 31 P peaks in the NMR spectra are difficult because the positions of the resonances

PAGE 132

121 are now determined not only by the chemical environment but by the relative orientation of the phosphorous nucleus with respect to the Co(II) ions as wen 199 -Uptake ofT APO Pr obe ( 0 o b y m ass) "'O D is p laceme nt o fBip y (per m o l ofCo(lI)) (1.) "' LOO 0 99 (1.) I.O (1.) :>.. 0 8 p,. ..... i:!l --0 6 (1.) ..... p,. 0.4 ;:J (1.) .c 0 2 0 .... 0 00 1 8 0 0 Host 2 TMPO TEPO TPPO Sample Figure 3-15. The gas chromatography results monitoring the guest exchange reactions in Host 2. The left-hand columns denote the fraction of probe taken in by the host and the right hand columns represent the relative quantity of clathrated bipy released. The NMR spectrum of2 after the guest exchange ofbipy with 1.0 mole ofTMPO per mole of host is shown in Figure 3-16 A. A single broad resonance was identified at 26.9 ppm. This peak is shifted upfield considerably compared to resonances observed for both crystalline and wet TMPO as well as TMPO interacting with 1. This peak is tentatively assigned to direct TMPO coordination to the Co(II) ions. However it is unclear whether this resonance is a chemical shift effect due solely to the TMPO coordination, to the coupling of the dipolar moments of the 31 P nucleus with the magnetic moments from the metal centers or a combination of both effects.199 The uptake of TMPO by 2 results in an observable color change from orange to pink suggesting a slight change in the ligand field about the Co(II) ions although the geometry still remains

PAGE 133

122 octahedral. Due to the highly oxophilic nature of the Co(II) ions, the TMPO may have replaced one or both of the aqua ligands coordinated to the metal centers although the GC measurements did not indicate the presence of water in solution. This replacement would account for the color change after the guest exchange. It should also be noted that no peaks corresponding to TMPO interacting with either coordinated lattice water molecules are observed in the NMR spectrum, even at high spinning speeds. Evidently, the TMPO prefers to coordinate directly to metal centers rather than interact with the water molecules. In contrast, TMPO interacts with the lattice waters and aqua ligands rather that the Ni(II) ions in 1. The profound broadening of the peak is likely due to the close proximity to and strong interaction of the TMPO with the Co(II) nucleus.199 The NMR spectra of 2 after the guest exchange of bipy with 1.0 mole of TEPO and 0.75 moles ofTPPO per mole of host are shown in Figures 3-16 Band 3-16 C, respectively. In the case of2 with TEPO, two chemical shifts were observed at high spinning speeds, one at 62.4 ppm and another at 103 .1 ppm. These peaks are assigned to the interaction of TEPO with unbound and coordinated water sites, respectively. In the case of2 with TPPO, a relatively low signal-to-noise ratio spectrum was observed just as with TPPO in 1. A single broad resonance is observed at 124.3 ppm attributed to probe interacting only with bound water. However, again these assignments are tentative because broadening and possible chemical shift effects between the dipole moment of 31 P and the unpaired electrons of Co(II) ions.199 Since the peaks are observed at frequencies consistent with those observed with TEPO and TMPO in 1 where interactions corresponding to bound and unbound water (the shaded region) were observed, these peak assignments are not unreasonable. Unlike TMPO, the uptake of TEPO and TPPO

PAGE 134

123 by 2 produces no observable color changes in the samples suggesting no appreciable change in the ligand field of the Co(II) ions. Steric constraints may prevent ligand replacement by TEPO and TPPO resulting in direct metal coordination. However, the same spectral broadening from the interactions between the 31 P nuclei and the Co(II) ions is observed suggesting that the TEPO and TPPO are in close proximity to the metal centers. X-Ray Diffraction Results Figures 3-17 A and 3-17B show the powder X-ray diffraction patterns of 2 before and after the exchange of bipy with one mole of TMPO, respectively. Recall from above that washing the host with solvent caused 2 to become less crystalline. A similar loss of crystallinity is observed after the guest exchange reaction as well. Note that the powder pattern of the guest-exchanged product resembles the corresponding pattern of the original material rather closely. However, the reflections in the 7 to 10 range decrease in intensity, move, and disappear after the guest exchange, particularly the interplanar (2 0 O] peak at 10 so the structure of the new product is not exactly the same as the original host. Some disorder within the host is apparent and the layered sheet structure becomes disrupted. Despite the larger quantity of bipy displaced and the (aqua) ligand replacement that occurs in the TMPO sample, the structural rearrangements are rather mild in contrast to the significant structural changes in analogous samples of 1. Although TMPO is chemically and topologically different from bipy, evidently TMPO fulfills the same role of guest as does bipy because the structures before and after the guest exchange are quite similar. Unlike 1, 2 is capable of exchanging bipy with TMPO without producing dramatic structural rearrangements within the host.

PAGE 135

E C. C. r-,.. -.;;I" N T""" E C. C. T""" C") 0 * I I I I I I I I 200 100 124 E C. C. m (0 N * I I 0 (C) Host2 +TPPO (B) Host 2 +TEPO (A)Host 1 +TMPO I i I ppm Figure 3-16 The 31 P MAS NMR Spectra of Host 2 after the guest exchange of 4 4 Bipyridine. A) Exchange with TMPO. B) Exchange with TEPO. C) Exchange with TPPO The spinning sidebands are denoted by asterisks(*)

PAGE 136

125 Figures 3-17 C and 3-17 D show the powder X-ray diffraction patterns for 2 after the exchange ofbipy with 1.0 moles ofTEPO and 0.75 moles ofTPPO per mole of host, respectively. The powder patterns of the TEPO and TPPO samples are quite similar, indicating the guest exchange causes the same structural changes and produces essentially the same product. The extent of the structural changes is consistent with the relatively large quantity of bipy released and probe taken in. A comparison of the powder patterns, particularly the 15 to 20 range, shows that the guest-exchanged has peaks common to both the original sample 2 and the analogous TEPO and TPPO samples of 1. The exchange of bipy with TEPO and TPPO seem to cause similar structural rearrangements in both 1 and 2. Again, the nature of these structural changes is not completely clear. 3600 3200 2800 2400 '-' 0 .... 2000 VJ j 1600 .E 1200 -~ di 800 400 0 D C B A 4 6 -A Host 2 Blank BHost2+TMPO -CHost2+TEPO 8 IO 1 2 14 16 I 8 20 Angle/ 20 (Degrees) Figure 3-17. Powder X-ray diffraction patterns of Host 2 before and after the guest exchange of 4,4'-Bipyridine. A) No probe. B) Exchange with TMPO. C) Exchange with TEPO. D) Exchange with TPPO.

PAGE 137

126 Conclusions This chapter describes the host-guest properties of two porous network structures, [Ni( 4,4 '-bipy)3(H20)2](Cl04)2 l.4( 4,4, -bipy) 3(H20), 1, and [Co(4,4' -bipy)3(H20)2](Cl04)2 l.4(4,4' -bipy)(H20), 2. Gas chromatography experiments have determined that the both hosts exchange clathrated bipy molecules with trialkylphosphine oxide probe molecules, TMPO, TEPO, and TEPO. While the uptake of TMPO by both 1 and 2 is essentially complete, steric constraints are believed to limit the uptake ofTEPO and TPPO by the host. The trialkylphosphine oxides interact with acid sites within the host as determined by 31 P MAS NMR spectroscopy. TMPO interacts with both coordinated water molecules (strong acid sites) and weak acid site (lattice water) in 1 and coordinates directly to the metal centers in 2 but TEPO and TPPO seem to attack only the weaker acid sites within the hosts. X-ray diffraction patterns indicate the loss of bipy and uptake of probe causes significant structural rearrangements within the 1 but only mild structural changes are observed in 2 however the nature of these guest-exchanged products is unknown. These experiments have shown that solid-state NMR spectroscopy can be used to investigate host-guest interactions.

PAGE 138

CHAPTER4 STRUCTURAL AND MAGNETIC PROPERTIES OF A SERIES OF AZIDO BRIDGED COPPER(II) LADDER-LIKE COORDINATION POLYMERS Introduction The rational design and synthesis of polynuclear transition metal complexes with characteristic structural and physical properties has in recent years been an area of great interest in materials science and solid-state chemistry. These materials can be often built through the self-assembly of simple molecular or ionic components, e g. transition metal ions or complexes with multifunctional bridging ligands in solution These bridging ligands act to propagate the structural and geometrical properties of the metal center which can heavily influence the topology of the overall molecular or solid-state structure. These topologies include oligomers one-dimensional chains two-dimensional sheets and three-dimensional networks. The overall structures can also be affected by reaction and crystallization conditions, most notably the metal:bridging ligand stoichiometry and as such, much effort has dedicated to elucidating and cataloging various synthetic strategies for building such molecular and extended solid-state materials. These materials often have useful bulk electronic, magnetic, and optical properties. Azide Bridging Modes Among the most common bridging ligands employed in the design of molecular and solid-state materials are halides and pseudohalides ( thiocyanate cyanate and azide ) The azide (N3 J anion is a versatile ligand known to adopt a wide variety of coordination modes in transition metal complexes. Aside from coordinating in a monodentate 127

PAGE 139

128 fashion200-203 ligating through only one nitrogen atom this ligand can exhibit a variety of bridging modes as well. In general two types of bridging modes are known for the azido ligand: the -(1 1) or end-on mode204-209 and -(1 3) or end-to-end mode210-215 illustrated schematically in Figure 4-1. In the end-on mode the bridge is established through a single nitrogen donor while in the end-to-end mode both peripheral nitrogen atoms participate in the bridge. The bis(bidentate) bridging modes ( 1 1 ) and -(1,3)) where the metal centers are doubly bridged are the most commonly encountered but single end-on216 and end-to-end bridges217 although more scarce are known as well. Numerous examples of discreet and extended transition metal structures sustained by azide-bridge are known including oligomers (dimers trimers tetramers etc. ), clusters chains and sheets.218-225 Of these materials azido-bridged dimers of Cu(II) are by far the most common 206 226-229 Other bridging modes are possible, such as the -(1 1,1)230 where a single nitrogen donor bridges three metal centers and the -(1 1 3)231-235 a combination of the end-on and end-to-end bridge. Furthermore, the simultaneous presence of two or more different azide coordination modes within the same structure is possible.236 For example complexes with the combination of both terminal and end-to-end bridged azido ligands are rare23 7 but structures with both terminal and end-on bridged azido groups are more common238-241 and samples with both end-on and end-to-end azido groups present are very rare.226 242-245

PAGE 140

129 N II N II N II N / "'N=N=N / "' M M N II N /~ M M C M M "'/ N "' / N=N=N II N B II N M D A II / N N II II N N II M/i"M N / "'M M M E F Figure 4-1 Schematic depicting the various bridging modes adopted by the azide ligand. A) the -(1,1) or double end-on bridge. B) The -(1 3) or double end-to-end bridge. C) The -(1 1) or single end-on bridge. D) The -(1,3) or single end-to-end bridge. E) The -(1 1, 1) bridge. F) The -(1 1 3) bridge. Superexchange Properties of Azide Bridges The azide ion is a well-known mediator of superexchange interactions between metal centers with unpaired spins. A diverse array of magnetic properties is possible due to the various types of bridging modes exhibited by the azide ion. 6 The end-on bridging mode typically mediates ferromagnetic exchange interactions between the metal centers.6 17,237 246-249 In many cases, the exchange is so strong, that the energy gap between the spin triplet ground state and the spin singlet excited state is so large that the excited S = 0 state is at best only weakly populated at high temperature. 250 251 In contrast the end-to-end bridging mode typically mediates antiferromagnetic exchange interactions.6 17,222,237,246,252 253 In some cases, the singlet state is stabilized with

PAGE 141

130 respect to the excited triplet state to the extent that the corresponding compounds appear diamagnetic at room temperature.214,254 In symmetric end-to-end azide bridges where the three nitrogen atoms are contained within the plane of the metal centers, the exchange is typically strongly antiferromagnetic.6,205 For example, if the end-to-end azide ligands coordinate to the equatorial sites of two Cu(II) ions, the metal centers are strongly antiferromagnetically coupled. This is due to the significant unpaired spin density associated with the d/-/ orbitals for both octahedral and square pyramidal coordinated Cu(II) ions. The appreciable delocalization of the electron density to the bridging azide ligand coordinated to both equatorial Cu(II) ions renders this type of bridge an efficient mediator of superexchange interactions. In complexes with asymmetric azide bridges (with both long and short Cu-N bonds), and non-coplanar azide bridges (all three atoms of the N3ion are not contained within the plane of the metal centers), the exchange between the spin centers is typically weak.6,205,255 Obviously, poor orbital overlap and hence weak exchange results from long Cu-N bonds. In Cu(II) complexes, non-coplanar bridges often result from the azide ion coordinating to the axial sites on one metal center and the equatorial sites on the other ion. If the Cu(II) ions are octahedral or square pyramidal, the exchange is negligibly small while the coupling is weakly antiferromagnetic in trigonal bipyramidal geometries.205,255 The unpaired spins in octahedral and square pyramidal Cu(II) ions principally reside in the d/-/ atomic orbitals and low unpaired electron density is associated with the d} orbital. There is a good delocalization of the electron density to the bridging azide ligand coordinated to the equatorial Cu(II) ions but the delocalization

PAGE 142

131 to the azide bridge coordinated to the axial Cu(II) ions is poor since the dz 2 orbital is occupied by paired spins. Therefore, the magnitude of the exchange is weak. The antiferromagnetic coupling is greater between the axial-equatorial coordination sites in the trigonal bipyramidal geometry of Cu(II) ions due to the increased unpaired spin density associated with the d/ orbitals due to admixture from d/-/ orbitals. 205,255 It is known that both the magnitude and sign of superexchange interactions mediated by end-on azide bridges are strongly dependent on characteristic structural bonding parameters of the magnetic M-N 3 -M unit, including M-N bond distance, the M-N-M bridging angle, and the planarity of the M-N 3 -M moiety.6,250,251 Under most circumstances, an end-on azide bridge mediates a ferromagnetic exchange between the metal centers. In order to rationalize this behavior, numerous theoretical approaches have been developed to determine the relative energies between the magnetic orbitals, the frontier molecular orbitals involved in the superexchange processes.6,205,250,251 In fact, the relative energies of these magnetic orbitals heavily influence both the sign and magnitude of the exchange. For example, in -1, 1-diazido copper(II) complexes, the sign and magnitude of the exchange interaction is highly dependent on the Cu-N-Cu bridging angle.6,251 If the bridging angle is small (100 105 ), the net exchange is ferromagnetic and the magnitude of the exchange increases as the bridging angle closes.21,205,238,256 However, if the bridging angle is large(> 108.5 ), the net exchange is typically antiferromagnetic.238,256 Consider the magnetic orbitals of a doubly end-on azido bridged copper(II) dimer (Figure 4-2).6,17,236,246,270 The magnetic orbitals,
PAGE 143

132 atomic orbitals that contain the unpaired electrons and the symmetry adapted a-type HOMO's (highest occupied molecular orbitals) from the principally p-type azide atomic orbitals. The "symmetric" and "asymmetric" character refers to the symmetry of the orbitals with respect to a mirror plane perpendicular to the molecular plane. Note that the magnetic orbitals are principally of antibonding character. The energy variation of the As' a-orbitals are equal and negative and thus the two magnetic orbitals are degenerate; strict orthogonality of the orbitals is observed. Actually, the angle is slightly higher than 90 due to admixture from ligand s orbitals close in energy to the p orbitals. Therefore, at low bridging angles, since the magnetic orbitals are degenerate, a triplet ground state is stabilized, and the ferromagnetic contribution to the total exchange dominates. As the bridging angle increases, the energy of
PAGE 144

133 has been used to rationalize the signs of coupling constants for other structural and electronic perturbations as well, such as in-plane distortions of the magnetically important Cu-(N 3 ) 2 -Cu moiety, the effects of ancillary ligands, and changing the electronegativity of the bridge. 6 Another approach to rationalizing the signs of the coupling constants between the end-to-end and end-on azide bridges is the spin polarization mechanism (Figure 43). 205,257 The role of the low-energy, doubly occupied ligand orbitals is considered in the spin polarization approach. Each 7l"g orbital of the azide ion describes two paired electrons localized at the two terminal nitrogen atoms. At each instant, one of the two electrons is "spin up" or has an a. spin, and the other is "spin down ", or has a p spin. For end-on azide ligands bridging octahedral and square pyramidal Cu(II) ions, a lrg electron on the bridging nitrogen atom is partially delocalized toward the two metal d/_/ orbitals. The polarization of the unpaired spins on the metal centers produces a parallel alignment thus giving rise to ferromagnetic coupling. In contrast, for the end-to-end case, an a. spin electron is partially delocalized toward one of the metal centers and the p spin is partially delocalized toward the other metal center. The polarization of the unpaired electrons on the metal centers results in an antiparallel alignment thus giving rise to the antiferromagnetic coupling. Chapter Summary The purpose of this chapter is to investigate the structural and magnetic properties of three azido-bridged copper(II) ladder-like coordination polymers, [Cu 2 (PhPyPy) 2 (N3)2(N3)2], 4, [Cu2(terpy)2--(N3)4Cu2--(N3)2(N3)2] 5, and [Cu2(terpy)2(N3)2(N3)2Cu3--(N3)4(N3)2], 6 The first part of the chapter examines in detail both the

PAGE 145

134 molecular and solid-state structure of the title compounds. The second part of the chapter describes the magnetic properties of the title compounds. Bulk magnetization and susceptibility data as well as the appropriate theoretical models needed to describe the magnetic behavior are discussed. Furthermore, the magnetic exchange interactions, mediated by the various azide ligands bridging the Cu(II) metal centers, are examined and rationalized on the basis of characteristic structural-bonding parameters.

PAGE 146

EleV -12.2 T -12.3 I 135 AS, s, 108.5 -12.4 I ,__ _______ 90 100 110 Angle (degrees) A B Figure 4-2. Correlating the sign of exchange interactions with the metal-azide bridging angle. A) Magnetic orbitals for planar -1,1-diazido copper(II) dimers. B) Graph depicting the energy variation of the
PAGE 147

136 A B C Figure 4-3. Schematic depicting the spin polarization mechanism A) The D g orbital of an azide ion. B) The unpaired spins on the metal centers are polarized in a parallel fashion in the end-on bridging mode, leading to ferromagnetic exchange interactions. C) The spins are polarized in an antiparallel manner in the end-to-end bridging mode leading to the antiferromagnetic coupling (bottom right). Adapted from references 205 and 257.

PAGE 148

137 Experimental Section Materials Copper(II) perchlorate hexahydrate (98 % ), 2,2 ':6,2' '-terpyridine (98 % ), 4-(3phenylpropyl)pyridine (97 %) and sodium azide (99 %) were purchaced from Aldrich Chemical Co. Dimethyl Sulfoxide (99.9 %) was purchaced from Fisher Scientific (Pittsburgh, PA). All reagents were used without further purification. Throughout this chapter, DMSO refers to dimethyl sulfoxide, terpy refers to 2,2':6,2"terpyridine, and PhPrPy refers to 4-(3-phenylpropyl)pyridine. Unless stated otherwise, all reactions and crystallizations were performed under ambient laboratory conditions. Synthesis of [Cu2(PhPyPy)2--(N3)2(N3)2] A solution containing 105 mg of Phprpy (0.532 mmol) dissolved in 20 mL of DMSO was added, drop wise, to a solution containing 800 mg of Cu(ClO 4 )iH 2 O (2.16 mrnol) dissolved in 20 mL of DMSO resulting in a blue colored solution. A solution containing 344 mg ofNaN3 (5.29 mmol) dissolved 20 rnL ofDMSO was added, drop wise, to the above copper-terpy solution producing a dark green colored solution. After approximately one week, dark green-black needles of 4 crystallized from solution, werecollected by vacuum filtration, and washed with ethanol (yield 61 % based on PhPrPy being the limiting reagent). Analysis calculated for Cu2C1~1sN 1 4: C, 48.75 %; H, 4.39 %; N, 28.44 %. Found: C, 47.92 %; H, 4.29 %; N, 28.26 %. Synthesis of [Cu2(terpy)2--(N3)4Cu2--(N3)2(N3)2] A solution containing 80 mg of terpy (0.343 mmol) dissolved in 10 mL of DMSO was added, drop wise, to a solution containing 800 mg of Cu(ClO 4 )i-6H 2 O (2.16 mmol) dissolved in 10 mL of DMSO resulting in a blue colored solution. A solution containing

PAGE 149

138 501 mg ofNaN 3 (7.71 mmol) dissolved 20 mL ofDMSO was added drop wise to the above copper-terpy solution producing a dark green colored solution. After approximately one week dark green-black needles of 5 crystallized from solution were collected by vacuum filtration, and washed with ethanol (yield 64 % based on terpy being the limiting reagent). Analysis calculated for C1.4C30H 22 N 3 0: C 34.09 % ; H 2 10 % ; N, 39.76 %. Found: C 34.00 %; H 1.99 %; N, 39.46 %. Synthesis of [Cu2(terpy)2--(N3)2(N3)2Cu3--(N3)4(N3)2) A solution containing 80 mg ofterpy (0.343 mmol) dissolved in 10 mL ofDMSO was added drop wise, to a solution containing 800 mg of Cu(ClO4)2H2O (2 16 mmol) dissolved in 10 mL ofDMSO resulting in a blue colored solution. A solution containing 344 mg ofNaN3 (5.29 mmol) dissolved 20 mL of DMSO was added drop wise to the above copper-terpy solution producing a dark green colored solution. After approximately one week dark green-black needles of 6 crystallized from solution were collected by vacuum filtration, and washed with ethanol (yield 58 % based on terpy being the limiting reagent) Analysis calculated for CusC30H2 2 N36: C 29.91 %; H 1.84 % ; N 41.89 %. Found: C 30.03 % ; H 1.71 %; N, 41.64 %. Physical Characterization Elemental analyses were performed by the University of Florida Spectroscopic Services Laboratory X-ray Structure Determination Black needles of 4 (0.45 x 0.12 x 0.04 mm\ 5 (0.46 x 0.10 x 0.03 mm 3), and 6 (0.17 x 0.07 x 0.02 mm 3 ) were selected for X-ray analysis Each crystal was mounted on a glass fiber under nitrogen gas The same data collection was used for each sample. Data were collected at 173 Kon a Siemens SMART PLATFORM equipped with a CCD

PAGE 150

139 area detector and a graphite monochromator utilizing Mo.Ka radiation (A. = 0.71073 A). Cell parameters were refined using 7838 reflections. A hemisphere of data (1381 frames) was collected using the ro-scan method (0.3 frame width) The first 50 frames were re measured at the end of data collection to monitor instrwnent and crystal stability (maximum correction ofl was< 1 %). Absorption corrections by integration were applied based on measured indexed crystal faces All structures were solved by the Direct Methods in SHELLXTL6 and refined using full-matrix squares.155 The non-H atoms were treated anisotropically whereas the hydrogen atoms were calculated in ideal positions by riding on their respective carbon atoms. For 4 the asymmetric unit consists of a half-dimer. Part of the pyridine moiety and part of the propyl fragment -C6(H2)C7(H2), are both disordered ; Nl of the pyridyl group is not disordered The disorder was refined in two parts. Their site occupation factors were dependently refined to 0 56(1) for the major part and consequently 0.44(1) for the minor part. A total of 264 parameters were refined using F 2 in the final cycle using 3477 reflections with I> 2cr(I) to yield R 1 = 3.18 % and wR 2 = 6.99 %. For 5 the asymmetric unit consists of a half-tetramer. A total of 289 parameters were refined using F 2 in the final cycle using 4330 reflections with I> 2cr(I) to yield R 1 = 3.86 % and wR 2 = 9 .69 %. For 6 the asymmetric unit consists of a half-pentamer. A total of 323 parameters were refined using F 2 in the final cycle using 4771 reflections with I > 2cr(I) to yield R1 = 4.39 % and wR 2 = 8.37 % Magnetic Measurements Bulk magnetization measurements performed in Professor Mark Meisel s laboratory in the Department of Physics University of Florida, were obtained from a

PAGE 151

140 standard Quantum Design MPMS SQUID magnetometer. The samples consisted of randomly oriented single crystals with a total mass of 40.8 mg for 4, 96.3 mg for 5, and 57.7 mg for 6. A polyethylene canister and plastic straw were used as the sample holder during the measurements. Magnetization versus temperature measurements were run from 2 K to 300 K The sample was zero-field cooled to 2 K before a measuring field of 100 G was applied and the data set was then taken while warming the sample from the lowest temperature. Magnetization versus field measurements were performed at 2 K over the range of Oto 50 kG. The background signals arising from the canister and straw were measured independently and subtracted from the results. The diamagnetic contribution from each sample, estimated from Pascal's constants (Xo = 345.52 x 106 emu mol -I for 4, xo = 452.08 x 106 emu mol for 5, and x,o = 491.10 x 1 o6 emu mol 1 for 6) was also subtracted from the results.5,6 ESR spectra were recorded in a Bruker ER 200D spectrometer modified with a digital signal channel and digital field controller on polycrystalline samples contained within evacuated quartz sample tubes. Data were collected using a U.S. EPR SPEC300 data acquisition program and converted to ASCII format using a U.S. EPR EPRDAP data analysis program. The temperature was controlled by and Oxford Instruments ITC 503 temperature-controller and an ESR 900 continuous flow liquid helium cryostat. Results and Discussion Description of the Structures Crystallographic and structural refinement data for 4, 5, and 6 are listed in Table 4-1. Tables of atomic coordinates and thermal displacement parameters and bond angles and distances for 4 5, and 6 are are provided in the Appendix B.

PAGE 152

141 Table 4-1. Summary of Crystallographic Data for Compounds 4, 5, and 6. Sample 4 5 6 Empirical Formula C2sH30Cu2N14 C1sH11Cu2N1s C1sHllCu2 sN1s Formula Weight 689.74 528.47 602.27 Space Group Monoclinic P2(1 )/n Monoclinic, P2(1 )/n Triclinic, P-1 a, A 5.2066(2) 14.4872(8) 6.6035(6) b, A 10.7847(4) 7.1430(4) 12 660(1) c, A 27.069(1) 18.454(1) 13.110(1) a, deg 90 90 88.682(2) p, deg 91.620(1) 95.719(1) 76.278(2) y de' 90 90 82.819(2) V A 1519.4(1) 1900.2(2) 1056.4(1) Z 2 4 2 T K 173(2) 173(2) 173(2) A(Mo Ka) A 0.71073 0.71073 0.71073 Peale g cm 3 1.508 1.847 1.893 mm I 1.445 2.280 2.552 R 1 8 (wR 2 b) 0.0318 (0 0699) 0.0386 (0.0969) 0.0439 (0 0837 ) aR, = ~) I I F o 11 F e I I) / ~) F 0 I bwR 2 = [~)w(F ; -F c 2 )2] / ~)w(F ; )2]] 112 S = [~)w(F 0 2 -F :)2 ] / (n-p)] 112 w = 1 /[cr 2 (F ; ) + (0.0370 p)2 + 0.31 p ] p = [max(.F; / ,0) + 2 ~ 2 ] / 3 The structure of 4 consists of neutral chain-like stacks of [Cu2(PhPrPy) 2 --(l N3)2(N3) 2 ] dinuclear units. The centrosymrnetric dimers shown in Figure 4-4 are comprised Cu(II) ions bridged by two end-on azido ligands. The Cu-Cu distance is 3.10 A while the Cu-N-Cu bridging angle is 101.9 The Cu(N 3 ) 2 Cu units are planar; no folding of the CuN2 planes about an in-plane axis joining the two bridging atoms (N2 and N2A) is observed The local coordination environment of each metal center is distorted octahedral. The equatorial plane is defined by four nitrogen atoms two from the intradimer end-on azide bridges (N2 and N2A) one from the monocoordinate azide ligand (N5), and one from the pyridyl donor (NI) of the organic ligand 4-(3phenylpropyl)pyridine. All equatorial Cu N bond distances (1.97 A to 1.99 A ) are similar. The bridging azide ligands are nearly linear (178.5) while the terminal azide

PAGE 153

142 groups are slightly bent (176.4) at the central nitrogen atom. The axial coordination results from weak Cu N contacts with the monocoordinate azide ligands on adjacent dimers within a chain ; each metal center interacts with the terminal and ligated nitrogen atoms from two different azide ions. The Cu N bond to the terminal nitrogen atom (2.62 A) is much shorter than the Cu-N bond to the ligated nitrogen atom (3.14 A) on the monocoordinate azide ligand on adjacent dimers. As a result of Jahn-Teller distortions the axial Cu-N bonds are considerably longer than the equatorial bonds. The coordinated PhPrPy molecules are disordered in two parts about the 1 3-propyl fragments. Due to packing forces and bonding constraints of the propyl fragments the phenyl groups of the organic ligands are not coplanar with the pyridyl moieties but twisted upwards ( or downwards) 113 .1 with respect to one another. These pyridyl moieties are themselves not coplanar with the Cu(N 3 )2Cu moieties but twisted 69 5 out of the plane of the dimers The dinuclear units stack atop one another to form chains that extend along the crystallographic a-axis. A single chain of 4 is shown in Figure 4-4. Within each chain adjacent dimers are related by a single translation along the crystallographic a-axis and translation along the b-axis but stack in registry along the c-axis. The interdimer bonding consists of both double asymmetric end-to-end and end-on azide bridges. The terminal nitrogen atoms from the monocoordinate azide ligands weakly interact with the metal centers on adjacent dimers to form the asymmetric end-to-end bridges with long Cu-N contacts of2 62 A. The CuA-N5-N6 and CuB-N7-N6 (CuB on an adjacent dimer) bridging angles are 126.4 and 105.8 respectively while the dihedral angle between the two planes defined by CuB-N5-N6--N7 and N5-N6--N7-CuA

PAGE 154

143 is 94.3 The ligated nitrogen atoms from the same monodentate azide ligands also weakly interact with the metal centers on adjacent dimers within a chain thus producing the asymmetric end-on bridges with the long Cu N contacts of 3 .14 A. The CuAN 5--CuAB (CuAB is a copper ion on an adjacent dimer) bridging angle is 90.6 Note that 4 can be viewed as an alternating chain because of the regular alternation of the intradimer end-on azide bridging with the interdimer asymmetric end-on and end-to-end azide brides. However since the interdimer interactions are significantly weaker than the intradimer bonding 4 is better described as stacks of weakly interacting dimers. Although both the phenyl and pyridyl rings from the organic ligands within each chain are stack in an offset parallel fashion, 1rinteractions are negligible due to the large distance(~ 5 A) between the molecules. Twin N-H contacts from the weak interactions between the terminal nitrogen atoms from the monocoordinate azide ligands and the pyridyl hydrogen atoms of the organic ligands are observed both within (2.44 A) and between (2.59 A) the ladder-like chains The packing diagrams of 4 are shown in Figure 4-5. Note that the packing of the chains in the crystallographic be-plane resembles a herringbone motif. Adjacent chains are related by a single translation along the crystallographic b-axis, but translation along the c-axis. The chains pack in registry along the b-axis. Along the c-axis the chains are arranged in a zig-zag fashion and positioned approximately 75 with respect to one another. Since the phenyl rings between the chains are not oriented in a face-to-face coplanar fashion no Jr-stacking interactions between the organic fragments are present. Instead Jr-facial hydrogen bonding( ~ 3 A) is observed between hydrogen atoms of the propyl fragments and the faces of the phenyl rings on adjacent chains

PAGE 155

144 Structure 4 bears close resemblance to [Cu(4-Etpy)(N3)2h (4-Etpy = 4ethylpyridine), a ladder-like chain of weakly interacting end-on azido bridged Cu(II) dimers. In fact, the ladder-like molecular structure of the latter compound is almost identical to 4 other than the differing the organic ligands as well as solid-state packing.258 All Cu-N bond distances and Cu-N-Cu and N-Cu-N bond angles are comparable to those reported for similar end-on azido bridged Cu(II) complexes. 23 7 ,24 7-249 Structure of [ Cu2( terpy )2--(N3)4Cu2--(N 3)2(N 3)2) The structure of 5 consists of neutral, ladder-like azido-bridged copper(II) coordination polymers that extend along the crystallographic b-axis. The "rungs" consist of centrosymmetric Cu2(N3)4 dinuclear units positioned parallel to the crystallographic axis. Figure 4-6 depicts a portion of a typical ladder. The rung copper ions adopt a distorted square pyramidal geometry. The basal plane is defined by two nitrogen atoms from the double end-on azide bridges (N7 and N7 A), a nitrogen atom (NI 3) from a single end-to-end azide bridge to the adjacent [Cu(terpy)(N3)2] groups, and a nitrogen atom (NI 0) from a monocoordinate azide ligand. The apical site is occupied by the ligated nitrogen atom (N4) from a single end-on azide bridge to the adjacent (Cu(terpy)(N 3 ) 2 ] units. An apical elongation is observed since the weak Cu2-N4 contacts (2.37 A) are significantly longer than the basal bonds. In the basal plane, the Cu2-N bonds to the double end-on azide bridges (1.99 A and 2.04 A) are slightly longer than the Cu-N bonds to the single end-to-end azide bridge (1.98 A) and the monodentate azide ligand Cu-NlO (1.98 A) bonds. The average values of the N(apical)-Cu-N(basal) angles is 95.1 while the trans-basal N7 A-Cu-NI O and NI 3-Cu-N7 angles are 177.1 and

PAGE 156

145 A B Figure 4-4. [Cu2(PhPyPy)2--(N3)2(N3)2]. A) The -diazido dicopper units. B) A ladder-like stack of dimer. All hydrogen atoms have been omitted from A) for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 157

146 b LC A B a C LC Figure 4-5. [Cu2(PhPyPy)2--(N3)2(N3)2]. A) The structure within the crystallographic be-plane. B) The structure within the crystallographic ab-plane. C) The structure within the crystallographic ac-plane. For clarity, the phenylpropyl fragments and pyridyl hydrogen atoms from the organic ligand have been omitted from B) and the organic ligands have been removed altogether from C).

PAGE 158

147 154.2 respectively. The Cu(II) ion is displaced 0.187 A from mean basal plane Cu2N7-N10-N13-N16 directed toward the apical ligand. The monocoordinate azide ligand is slightly bent (176.5 at the central nitrogen atom. The Cu(II) ions within the Cu 2 (N 3 ) 4 moieties are bridged by the nitrogen atoms (N7 and N7A) of two -1,1-azido (end-on) ligands. The bridge is slightly asymmetric since the Cu-N7 and Cu-N7 A bond distances, 1.99 A and 2.04 A respectively, are not quite equal. The bridging azide ligands are nearly linear (178.0 ). The Cu2-N7Cu2A bridging angle is 103 .3 and the Cu-Cu distance within the dimeric unit is 3 .17 A. The cyclic Cu2(bridging)N2 units are planar and no folding of the CuN2 planes about an in-plane axis joining the bridging atoms is observed. The "legs" of 5 consist of stacks of [Cu(terpy)(N3) 2 ] monomeric units that extend along the crystallographic b-axis. These moieties are not directly connected to each other but coordinated to the Cu2(N3) 4 dimers. In the [Cu(terpy)(N3)2] units, the Cu(II) ions adopt a distorted square pyramidal geometry. The rigid nature of the terpy ligand as well as the bridging azide groups causes the deviation from the ideal square pyramidal behavior. The basal plane is defined by three pyridyl nitrogen donors from the terpyridine ligand (Nl, N2, and N3) and a nitrogen atom (N4) from a single end-on azide bridge to the rungs. The apical site is occupied by a single end-to-end azide bride (N15) to the rungs. An axial elongation is observed since the Cul-Nl5 bond (2.36 A) is significantly longer than the basal Cu-N bonds. In the basal plane, the Cul-N bonds from both the bridging azide ligand and the central pyridyl moiety (1.94 A) of the terpy are slightly shorter than the bonds to the peripheral pyridyl groups (2.04 A). The average value of the N(apical)-Cul-N(basal) angles is 94.4 while the trans-basal NI-Cul

PAGE 159

148 N3 and N2-Cul-N4 angles are 159.5 and 163.2 respectively. The Cu(II) ion is displaced 0.13 A from mean basal plane Cul-Nl-N2-N3-N4 directed toward the apical ligand. Due to the steric influence of nearby azide ions, the pyridyl rings of the terpy ligands are not coplanar, but bowed. Both the single end-on (176.5 and single end-to-end (177.1 bridging azide ions are slightly bent at the central nitrogen atoms. Furthermore, the terpy ligand is not planar, but slightly bowed. Due to the large distance as well as the steric interference of the terminal azido ligands, no Jl"-stacking interactions between the terpy ligands within the ladders are present. The Cu(terpy)(N3)2 moieties, the "legs", are connected to the Cu2(N3)4 groups, the "rungs", by both single end-to-end azido (N13, N14, and Nl5) and single end-on azido (N4, N5, and N6) bridges. Therefore, 5 can be considered a true coordination polymer ladder because of the strong coordinate covalent bonds between the legs and the rungs as opposed to the a ladder-like structures defined by stacks of weakly interacting oligomers. A single ladder of 5 is shown in Figure 4-6. The single end-on bridge is asymmetric since the Cul-N4 bond, (1.94 A), is shorter than the Cu2-N4 bond (2.37 A). The Cul-N4-Cu2 bridging angle is 107.1 The single end-to-end bridge is asymmetric as well since the Cul-N15A bond (2.36 A) is longer than the Cu2B-N13 bond (1.98 A). The Cu2B-Nl3-N14 and Nl4--Nl5-Cul bond angles are 124.2 and 114.6 respectively. The dihedral angle between the Cu2B-N13-Nl4--Nl5 and Nl5Nl4--Nl5-Cul planes is 89.8 The packing diagrams of 5 are shown in Figure 47. The ladders pack to form sheets within the crystallographic ab-plane. The ladders within a sheet are juxtaposed in registry; adjacent ladders are related by a single translation along the a-axis. The sheets

PAGE 160

149 of ladders are layered along the crystallographic c-axis. Note that the different orientation of the ladders in adjacent sheets. A ladder in an adjacent sheet is related by both a net glide plane as well as translation along both the aand c-axes. Within the sheets, the terpy ligands between adjacent ladders overlap in an offset parallel fashion with a face-to-face distance of 3.4 A. This distance is well within the range for significant .1r-stacking interactions. However, the parallel overlap between the adjacent terpy ligands is hindered due to the steric interference of the terminal azide ligands from the ladder rungs. Weak N-H contacts within and between the sheets of ladders are present. The central nitrogen atoms from the monocoordinate azide ligands that are part of the ladder "rungs" form N-H contacts (2.60 A) with the hydrogen atoms from the peripheral terpy pyridyl moieties on adjacent ladders within the same sheet. The terminal nitrogen atoms from the double azide bridges that are part of the ladder "rungs" form N-H contacts with the hydrogen atoms from both the central (2.56.A) and peripheral (2.54 A) terpy pyridyl moieties on adjacent ladders between the sheets as well. All Cu-N bond distances and Cu-N-Cu and N-Cu-N bond angles in 5 are comparable to those reported for similar end-on azido bridged Cu(II) complexes. 216,237,247-249,255 Structure of [Cu2(terpy)i--(N3)2(N3)2Cu3--(N3)4(N3)2) The structure of 6 consists of neutral, ladder-like stacks of [Cus(terpy)2(N3)10] pentanuclear units. Each pentamer, one of which is shown in Figure 4-8, possesses -1

PAGE 161

150 A N13A B Figure 4-6. [Cu 2 (terpy)i--(N3)4Cu2--(N3)2(N3)2]. A) A fragment of the ladder-like structure. B) A single ladder. All hydrogen atoms have been omitted from A) for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 162

151 C L. Figure 4-7. [Cu2(terpy)2--(N3)4Cu2--(N3)2(N3)2]. A) The structure within the crystallographic ac-plane. B) The structure within the crystallographic be-plane. C) The structure within the crystallographic ab-plane. For clarity, the terpy ligands have been omittecJ from B).

PAGE 163

152 crystallographic symmetry. The pentamers are comprised of two terminal [Cu(terpy)(N 3 ) 2 ] moieties and a single central [Cu3(N3)6] fragment. In the monomeric [Cu(terpy)(N 3 ) 2 ] units, the Cu(II) ions adopt a distorted square pyramidal geometry. The rigid nature of the terpy ligand as well as the bridging azide groups causes the deviation from the ideal square pyramidal behavior. The basal plane is defined by three pyridyl nitrogen donors from the terpyridine ligand (NI, N2, and N3) and a nitrogen atom from a monocoordinate azide ion (N4). The apical site is occupied by a single end-on azide bridging ligand (N7). An axial elongation is observed since the Cul-N7 bond (2.29 A) is significantly longer than the basal Cu-N bonds. In the basal plane, the Cul-N bonds from both the terminal azide ligand and the central pyridyl moiety (1.95 A) of the terpy are slightly shorter than the bonds to the peripheral pyridyl groups (2.03 A). The average value of the N(apical)-Cul-N(basal) angles is 98.0 while the trans-basal Nl-Cul-N3 and N2-Cul-N4 angles are 159.7 and 150.8 respectively. The Cu(II) ion is displaced 0.21 A from mean basal plane Cul-Nl-N2-N3-N4 directed toward the apical ligand. Due to the steric influence of nearby azide ions, the pyridyl rings of the terpy ligands are not coplanar, but bowed. Both the monocoordinate (174.6 and bridging (174.6 azide ions are slightly bent at the central nitrogen atoms. The terminal [Cu(terpy)(N3)2] units are connected to the central [Cu3(N 3 ) 6 ] moieties through single end-on azide bridges. The bridge is asymmetric since the Cul N7 bond, (2.288 A), is longer than the Cu2-N7 bond (l.956 A). The Cul-N4-Cu2 bridging angle is 117.8 The central [Cu3(N3)6] fragment is itself comprised of two types of metal centers. The two peripheral copper(II) ions (Cu2 and Cu2A) are connected to the central Cu(II)

PAGE 164

153 ions (Cu3) through double end-on azide bridges thus forming two Cu(-N3)2Cu units. The peripheral to central metal (Cu2-Cu3) distance is 3.09 A while the Cu2-N-Cu3 bridging angle is 101.1 These bridging azides are nearly linear (179 5 ). The N3)2Cu units are planar ; no folding of the CuN2 planes about an in-plane axis joining the two bridging nitrogen atoms (NB and Nl6) is observed. The local geometry of the central copper ion is distorted octahedral. The equatorial plane is defined by four nitrogen atoms from the four azide bridges (NB, Nl6 NBA, and Nl6A). All equatorial Cu3-N bond distances (1 97 A to 1.99 A) are similar. The axial positions are formed by weak Cu3-N contacts (2 65 A) with the terminal nitrogen atoms of monocoordinate azide ligands from the [Cu(terpy)(N3)2] moieties on adjacent pentamers. As a result of Jahn-Teller distortions the axial Cu3-N bonds are considerably longer than the equatorial bonds. The peripheral copper ions adopt a distorted square pyramidal geometry. The basal plane is defined by two nitrogen atoms from the double end-on azide bridges (N13 and N16) to the central metal center (Cu3) a nitrogen atom (N7) from a single end-on azide bridge to the terminal [Cu(terpy)(N 3 ) 2 ] groups, and a nitrogen atom from a monocoordinate azide ligand (NlO). The apical site is occupied by the ligated nitrogen atom of the monodentate azide ligand from the [Cu(terpy)(N 3 )2] moiety on an adjacent pentamer. An apical elongation is observed since the weak Cu2-N4 contacts (2.64 A) are significantly longer than the basal bonds. In the basal plane, the Cu2 N bonds to the double end-on azide bridges are unequal (Cu2-NB of2 03 A and Cu2Nl6 of2.00 A) but longer than the bonds to the single end-on azide bridge (Cu2 N7 of 1.97 A) and the monodentate azide ligand Cu-NI O (1.94 A) bonds. The average values of the N(apical)-Cu-N(basal) angles is 93.4 while the trans-basal Nl0-Cu-Nl6

PAGE 165

154 and Nl3---Cu-N7 angles are 167.5 and 164.5 respectively. The Cu(II) ion is displaced 0.09 A from mean basal plane Cu2-N7-Nl0-N13-N16 directed toward the apical ligand. The monocoordinate azide ligand is slightly bent (176.5 at the central nitrogen atom. The pentanuclear units stack atop one another to form ladder-like chains within the [0 -1 0] plane that extend along the crystallographic a-axis. A single ladder of 6 is shown in Figure 4-8. Within each ladder, adjacent pentamers are related by a single translation along the a-axis and stack in registry along both the band c-axes. The weaker interpentamer bonding consists of double asymmetric end-on and single asymmetric end-to-end azide bridges. The terminal nitrogen atoms from the monocoordinate azide ligands from the [Cu(terpy)(N3)2] fragments of each pentamer interact with the central Cu3 ions from the [Cu3(N3) 6 ] moieties on adjacent pentamers to form the single end-to-end bridges. The single end-to-end bridge is asymmetric since the Cul-N4 bond (1.95 A) is shorter than the Cu3-N6 bond (2.65 A) on an adjacent pentamer. The Cu3B-N6-N5 and N5-N4---Cul bridging angles are 102.4 and 131.9 respectively, and the dihedral angle between the Cu3B-N6-N5-N4 and N6N5-N4---Cul planes is 111.7 The ligated nitrogen atoms from the monodentate azide ligands on both the [Cu(terpy)(N3)2] and [Cu3(N3)6] fragments interact with the metal centers on adjacent pentamers to form the double end-on azide bridges as well. The double end-on bridge is asymmetric since the Cul-N4 bond, (1.95 A), is shorter than the Cu2-N4 bond (2.64 A) on an adjacent pentamer. The Cul-N4---Cu2 bridging angle is 102.3 The shortest Cu-Cu distance (Cul---Cu2B) between the pentamers is 3 .60 A. Since the interpentamer bonding is significantly weaker than the intrapentamer

PAGE 166

155 interactions the molecular structure of 6 is best described as ladder-like stacks of weakly interacting pentarners. Although the terpy's within each ladder stack in an offset parallel fashion Jl"interactions are negligible due to the large distance(~ 6 5 A) and steric interference of bridging azide groups between the organic ligands. The packing diagrams of 6 are shown in Figure 4-9 The ladders pack to form layered sheets in the solid state. The sheets are contained within the [O 1 1] plane The ladders within a sheet are juxtaposed in registry ; adjacent ladders are related by translations along the [O 1 -1] direction. The sheets however do not pack in registry with respect to one another, and adjacent ladders related by single translations along either the crystallographic bor c-axes. Weak N-H contacts within and between the sheets of ladders are present. The terminal nitrogen atoms from both the monocoordinate azide ligands from the [Cu(terpy)(N 3 ) 2 ] units form twin N H contacts (2.47 2.53 A ) with two hydrogen atoms from two terpy pyridyl moieties on adjacent ladders within the same sheet. The terminal nitrogen atoms of the double azide bridges from the [Cu(terpy)(N3) 2 ] units also form twin N-H bonds (2.42 2.46 A) with two hydrogen atoms from two terpy pyridyl moieties on adjacent ladders between the sheets Due to the lack of any significant parallel face-to-face overlap as well as the steric interference of bridging azide ligands, no Jl"-stacking interactions are observed between the terpy ligands on adjacent ladders within the sheets. All Cu-N bond distances and Cu-N-Cu and N-Cu-N bond angles in 6 are comparable to those reported for similar end-on azido bridged Cu(II) complexes 216,237 247-249

PAGE 167

156 A B Figure 4-8. [Cu2(terpy)2--(N3)2(N3)2Cu3--(N3)4(N3)2]. A) A pentamer. B) A ladder like stack of pentamers All hydrogen atoms have been omitted from the A) for clarity All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 168

157 a C B A b b a C C b Figure 4-9 [Cu2(terpy) 2 --(N3)2(N3)2Cu3--(N3)4(N3)2]. A) The structure within the crystallographic be-plane. B) The structure within the crystallographic ab-plane. C) The structure along the [O 1 1] direction. For clarity, the terpy ligands have been removed from B).

PAGE 169

158 Electron Paramagnetic Resonance Selected X-Band EPR spectra of powder samples of 4 over the temperature range 7 K 300 K are shown in Figure 4-10. The EPR spectrum exhibits a broadened singlet (1441 G peak-to-peak) at room temperature corresponding to an average g-value of gavg = 2.00. The broadened spectra lead to large uncertainty in the measuring the g-values thus accounting for the rather low calculated value. The normalized, integrated area of each EPR signal was found to increase as the temperature decreases. The spectra do not change significantly upon cooling to 7 K, remaining relatively broad indicating that dipolar effects are important while magnetic exchange interactions are relatively weak. No half-field forbidden transition between AMs = 2 states were detected over the temperature range further indicating the coupling in 4 are weak.157 Selected X-Band EPR spectra of powder samples of 5 over the temperature range 5 K 300 K are shown in Figure 4-10. Sample 5 exhibits a broad (832 G peak-to-peak) and symmetric X-band EPR signal at room temperature. A weak signal at B = 1500 G, possibly the half-field signal in the g = 4 region corresponding to "forbidden" transitions within the AMs = 2 states, indicates low-dimensional magnetic exchange interactions between the Cu(II) ions. As the temperature is decreased, the signal in the g = 2 region narrows and the g = 4 peak disappears. This progressive narrowing of the EPR spectra indicates the presence of relatively strong magnetic exchange interactions present within 5. However, below 50 K, the signal in the g = 2 region evolves into a weak axial spectrum with gu = 2.07 and g-1. = 2.23. This effect appears to be intrinsic to the sample; the presence of a small quantity of uncoupled, paramagnetic impurity within 5 would not account for the abrupt change in the spectra. The narrowing is likely a thermal effect due

PAGE 170

159 to the onset of strong exchange interactions or a decrease in the relaxation time of the unpaired spins. The g-values are consistent with the distorted square pyramidal geometry and indicate a basically dx 2 -/ ground state for the Cu(II) ions.157 Selected X-Band EPR spectra of powder samples of 6 over the temperature range 5 K 300 K are shown in Figure 4-10. Sample 6 exhibits a broad (832 G peak-to-peak) EPR signal in the gavg = 2.14 region at room temperature. A weak signal at B = l 500 G may be the half-field signal in the g = 4 region corresponding to "forbidden" transitions within the '1Ms = 2 states but disappears at lower temperatures. The integrated area of the peaks increases as the temperature is lowered and a slight, but apparent, narrowing is observed, however, the spectrum does not vary significantly upon cooling indicating the presence of relatively weak magnetic exchange interactions.157 Magnetic Properties of [Cu2(PhPrPy)2--(l,l-N3)2(N3)2] Magnetic Data A plot of the zero-field cooled magnetization (open boxes) and the field-cooled magnetization ( open circles) for 4 is shown in Figure 4-11. The two sets of data superimpose one another indicating that 4 does not experience any long-range magnetic order down to 2 K. Plots of both the magnetic susceptibility CxM) vs. temperature and the inverse susceptibility vs. temperature per mole of dimer for 4 are shown in Figures 4-11 and 4-12, respectively, as open squares. For 4, the susceptibility increases steadily as the temperature is lowered and no maximum is observed. The high temperature inverse susceptibility data (from 250 K to 300 K) were fit to a straight line and extrapolated to the x-axis giving T= 17.3 K suggesting that the dominant magnetic exchange in 4 is ferromagnetic. A plot of the temperature dependence of the product of the susceptibility

PAGE 171

5000 4000 3000 2000 1000 0 -1000 -2000 0 160 A B 8000 6000 4000 2000 0 -2000 .__......______,__...____.__...___.___._....___. 1000 2000 3000 4000 5000 6000 7000 B (G) 1000 6000 5000 4000 3000 2000 1000 0 -1000 1000 2000 3000 4000 5000 B (G) 2000 3000 4000 5000 B (G) C Figure 4-10. X-Band, Variable Temperature ESR Spectra. A) Compound 4 at (A) 7 K, (B) 10 K, (C) 50 K (D) 150 K, and (E) 298 K. B) Compound 5 at (A) 5 K, (B) 10 K, (C) 50 K, (D) 150 K, and (E) 298 K. C) Compound 6 at (A) 5 K, (B) 10 K, (C) 50 K (D) 150 K, and (E) 298 K.

PAGE 172

161 and temperature CxMn per mole of dimer is shown in Figure 4-12 as open squares The %MT value at room temperature (0.423 emu K mor 1 of Cu(II) ions) is very close to the expected spin-only value for a collection of non-interacting S = paramagnetic centers (xMT= 0.413 emu K mol" 1 ). As the temperature is lowered, the %MT value increases until reaching a maximum value of 1.00 emu K mor 1 at 10 K. The increasing XMT value with decreasing temperature indicates the presence of a ferromagnetic coupling. Below 10 K however, the XMT value decreases, quickly approaching zero indicating that an additional weak, antiferromagnetic exchange interaction is also present. The field dependence of the magnetization (MM) per mole of dimer for 4 is shown in Figure 4-13 as open squares. The magnetization initially increases with increasing field, but begins to level out at 20 kG asymptotically approaching a constant value of approximately 11 000 emu G mol" 1 Below 10 kG, the magnetization varies linearly with the applied field with no apparent change in slope observed. Magnetic Model and Fits Recall the structure of 4 consists ofladder-like chains of azido-bridged Cu(II) dimers. The various types of superexchange pathways between the Cu(II) ions in 4 are schematically depicted in Figure 4-13. The intradimer coupling (J) is expected to be the dominant magnetic exchange interaction since the double end-on azide ligands bridge the square pyramidal copper ions through the equatorial coordination sites. The interdimer exchange is expected to be weak due to the long Cu-N contacts associated with the asymmetric azide bridges as well as the end-on and end-to-end azide ligands bridging the axial-equatorial coordination sites of the metal centers. Note that there are two possible interdimer superexchange pathways through the asymmetric end-to-end bridges (J') and

PAGE 173

162 3 0 -~-..----.--..-----r---.----.--"T,-...---, g 5' 2 5 a C 2 0 ., C C .., o a a 1.5 .. \ 1.0 0 5 -r 0 0 0 0.4 E 0 3 = j 0 2 g_ 0 1 ::; l'< 0 0 I 0 10 C a Ze r o-Fie ld Coo l e d Ma g n e tizati on o Field-Cooled Magnetization D D D D 20 30 40 T (K) o Experime nt a l x,, vs T Data Fit to S = 1/2 Dimer Model D 0 50 100 150 200 250 300 T(K) I 50 A B Figure 4-11. Magnetic data for compound 4 at 100 G from 2 K to 300 K. A) Field cooled (MF c ) and zero-field cooled (MzF c ) magnetization vs. temperature. B) Molar susceptibility, X M, per dimer. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. The fit of x to the S =dimer model from Equations 4-3 to 4-5 is show by the solid line in B).

PAGE 174

_,,...... ::I 5 0 E '-' b Cl ... 8. -: ::;; X 163 400 350 Experimental x,;' vs T Data High T Linear Fit fa'trapo lated 300 to TAxis 250 200 ISO 100 so 0 ~..__...L... ........ -.L__.__,J...._..__...L... ........ _....__.__,J.___., 1.00 0 92 0 88 0 84 0 SO 100 ISO 200 250 300 Q 0 so T (K) Experimental x,,T vs. T Data -Fit toS = 1/2 Dimer Model Q 100 ISO 200 250 300 T(K) A B Figure 4-12. Magnetic data for compound 4 at 100 G from 2 K to 300 K. A) Inverse molar susceptibility, x1 per dimer versus temperature. B) The product of the molar susceptibility and temperature, xT, per dimer at versus temperature. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. A linear fit to the high temperature data in A) is shown by the solid line. The fit of xT to the S = dimer model from Equations 4-3 to 4-5 is show by the solid line in B)

PAGE 175

12 ,-._ ) 10 c:, 8 "' 6 0 ...... '--' II) 4 Cl 8.. 2 i 0 0 10 164 Ex p e r i m e n ta l Mj 1 vs 8 Da t a -Fi tt oS= l/2Dim e rModel Fi t t o S = I B rill o uin Fun ctio n 20 30 40 B (kG) A 50 J B Figure 4-13. A) The molar magnetization (M) per dimer at 2 K from Oto 50 kG for compound 4. The data have been corrected for background signals arising from the sample container. The fit of M to the S = dimer model from Equation 4-2 is shown as the solid line and the fit to the S = 1 Brillouin function is shown as the dotted line. B) A fragment consisting of two dinuclear units of compound 4 depicting the possible superexchange pathways. The copper(II) ions are represented by the shaded circles.

PAGE 176

165 the asymmetric end-on bridges (J' '). Since the Cu-N contacts associated with the asymmetric end-on bridges (3.14 A) are significantly longer than the corresponding Cu N contacts with the asymmetric end-to-end bridges (2.62 A), only the weak exchange through the latter pathway is considered significant. Any magnetic interactions between the ladder-like chains are assumed to be negligible since only weak N-H(aromatic) contacts and hydrophobic interactions between the organic ligands are observed. Therefore, the magnetic data were analyzed by assuming that 4 consists of isolated chains comprised weakly interacting S = dimers. The exchange interactions of a Cu(II) dimer can be described by the spin Hamiltonian with the corresponding the Zeeman term A A A A A H=-2J(S 1 2 8 B(g 1 1 +g 2 2 ) (4-1) which assumes isotropic exchange interactions, J. For simplicity, g-values are assumed to be isotropic. The weak, interdimer exchange is approximated by a molecular field term (0) applied to the Hamiltonian. The zero field energy levels of the spin-Hamiltonian are E1 = 0 (S = 0) and E 2 = J (S = 1).6 The magnetization vs. field is (4-2) where the parameter NR normalizes the number of moles of sample between O and 1 and Mo corrects the zero-field magnetization to 0. The Van Vleck magnetic susceptibility vs. temperature, %Dimer, is (4-3)

PAGE 177

166 where Na is the temperature independent paramagnetism. The mean field corrected susceptibility, x', is x'= X Dime,( T ~ 0 ) (4-4) and the total susceptibility,%, corrected for a fraction of uncoupled, monomeric impurity, p, is (4-5) The best fit for both the x and xT data of 4 to Equations 4-3 to 4-5 was obtained using non-linear regression analysis with NR, g, J, p, 0, and xTo as parameters. The results are shown as the solid lines in Figures 4-11 and 4-12 with NR = 0.943, g = 2.09, 2J = 73.255 K 1.921 K, 0 = 0.206 K 0.010 K, p= 0.01, and xTo = 1.6 x 10 4 emu K mor 1 and x, 0 = 1. 7 x 10 4 emu mor 1 The impurity fraction was fixed since the fitting procedure consistently produced large, negative values when p was allowed to vary. The fitting ultimately therefore converged giving NR and p values that do not correlate. The agreement factor R = 3 x 105 is very small indicating the model fits the data well, where R is defined as R = ~)(xMT) obsd -(xMTt'cd]2 ~) (x M T)b sd ]2 (4-6) The best fit for the magnetization data of 4 to Equation 4-2 was obtained with NR, g, J, and Mo as parameters. The results are shown as the solid line in Figure 4-13 with NR = 0.94, g = 2.09, 2J = 71.066 K and Mo = -1.107 emu G mor 1 Although the agreement factor is large, expected for fits to M vs. B data, the NR, g, and J values are consistent

PAGE 178

167 with those obtained from the xT data and the fit to the data is relatively good. An additional simulation of the S = 1 Brillouin function given as is shown in Figure 4-13 as the dashed line, with NR g and Mo, obtained from the fits to the magnetization data as fixed parameters. The average isotropic g-value obtained from the bulk magnetic data is consistent with square pyramidal Cu(II) ions. The magnetic data and corresponding fits indicate that the dominant intradimer exchange interaction is ferromagnetic while the interdimer exchange is weak and antiferromagnetic Interpretation of the Magnetic Data The xT plot in Figure 4-12 depicts the thermal population of the triplet (S = 1 ) ground state at low (T = 10 K) temperature At high temperature 4 behaves paramagnetically i e ., the unpaired spins within each dimer along the ladder-like chains are uncoupled and essentially randomized resulting in a small net magnetic moment. As Tis lowered the intradimer coupling takes effect, as insufficient thermal energy is now present to populate excited states. An increase in the net magnetic moment is observed due to the onset of the ferromagnetic interactions. At T = 10 K the triplet ground state is fully populated and 4 now consists of stacks of S = 1 dimers. At even lower temperatures the weak interdimer coupling takes effect and a sudden drop in the net magnetic moment is observed due to antiferromagnetic coupling of the dimers. The magnetization plot in Figure 4-13 depicts the population of the S = 1 triplet state at high

PAGE 179

168 field. At 2 K due to the strong ferromagnetic intradimer coupling 4 consists of S = 1 dimers and thus the magnetization closely follows the S = 1 Brillouin function In contrast if the coupling was weak 4 would be comprised of weakly coupled S = dimers and the magnetization would be expected to follow the product of two S = Brillouin functions ( not shown in the plot). Rationalizing the Sign and Magnitude of the Coupling Constants Recall that the intradimer exchange interactions are mediated by end-on azide ligands bridging the two Cu(II) ions Both the magnetic data and corresponding fits indicate that the intradimer coupling J is strongly ferromagnetic. The calculated coupling constant is consistent with other similar -diazido bridged copper(II) complexes 6,17 237 246-249 The positive sign of the coupling constant is also consistent with the small Cu N-Cu bridging angle of 101.9 The large magnitude of the coupling constant is reasonable for the small bridging angle the short Cu-{bridging)N bonds the planar Cu--{N) 2 -Cu moiety, and the double end-on azido ligands that bridge equatorial-equatorial coordination sites (both associated with the d /-/ orbitals with significant unpaired spin density) on both Cu(II) ions (Figure 4-14).255 The interdimer superexchange pathway consists of asymmetric double end-on and single end-to-end bridging azide ligands Both the magnetic data and corresponding fits indicate that the interdimer coupling is weakly antiferromagnetic. The small antiferromagnetic coupling is consistent with asymmetric end-on and end-to-end azide ligands bridging square pyramidal Cu(II) ions in other similar complexes as well.205 255 The long weak Cu N bonds between the dimers are associated with poor orbital overlap between the two exchanging metal centers thus reducing the magnitude o f the

PAGE 180

169 exchange interaction. The azide ligands bridging the apical-basal coordination sites of the square pyramidal Cu(II) ions between the dimers also accounts for the small interdimer coupling as well (Figure 4-14). The interdimer exchange is between the axial equatorial coordination sites of the metal centers between the dimers, in which case the unpaired spins from the metal centers reside in d/ and d/-/ atomic orbitals respectively.255 Low unpaired electron density is associated with the d/ orbital in octahedral and square pyramidal coordinated Cu(II) ions. There is a good delocalization of the electron density to the bridging azide ligand coordinated to the equatorial Cu(II) ions since the unpaired spin is primarily located in a d/-/ orbital but the delocalization to the azide bridge coordinated to the axial Cu(ll) ions is poor since the d/ orbital is occupied by paired spins.6,205,255 As a result, the superexchange pathway via azide ligands bridging axial and equatorial sites between Cu(II) ions is inefficient. Therefore, 4 consists of antiferromagnetic chains of ferromagnetically coupled S = dimers. Magnetic Properties of [Cu2(terpy)2--(N3)4Cu2--(N3}z(N3)2) Magnetic Data A plot of the zero-field cooled magnetization (open boxes) and the field-cooled magnetization (open circles) is shown in Figure 4-15. The two sets of data superimpose one another indicating that 5 does not experience any long-range magnetic order down to 2 K. Plots of both the magnetic susceptibility (xM) vs. temperature and the inverse susceptibility vs. temperature per mole of tetramer for 5 are shown in Figures 4-15 and 4-16, respectively, as open squares. For 5, the susceptibility increases steadily as the temperature is lowered and no maximum is observed. The high temperature inverse susceptibility data (from 140 K to 300 K) were fit to a straight line and extrapolated to

PAGE 181

170 N N J \ Figure 4-14. The metal d-orbitals involved in the superexchange interactions present in compound 4. Within the dimer, the end-on azide ions bridge both equatorial sites of the metal centers and thus a strong intradimer coupling is expected. The long Cu-N contacts between the dimers as well as the axial-equatorial coordination by the azide ligands results in a weak interdimer coupling.

PAGE 182

171 the x-axis giving T= 46.32 K suggesting that the dominant magnetic exchange in 5 is antiferromagnetic in nature. A plot of the temperature dependence of the product of the susceptibility and temperature (xM1) is shown in Figure 4-16 as open squares. The %MT value at room temperature (0.393 emu K mor 1 Cu(II) ions) is very close to the expected spin-only value for a collection of non-interacting S = paramagnetic centers (xMT = 0.413 emu K mor 1 ). As the temperature is lowered to 50 K, the %MT value decreases until reaching a plateau where the %MT value (0.380 emu K mor 1 Cu(II) ions) remains relatively constant. The decreasing %MT value with decreasing temperature indicates the presence of an antiferromagnetic coupling. Below 5 K, the %MT value decreases again approaching zero suggesting that an additional, weak antiferromagnetic exchange is present. The field dependence of the molar magnetization (MM) of 5 is shown in Figure 4-17 as open squares. The magnetization increases with increasing field but begins to level out at 40 kG asymptotically approaching a constant value of approximately 11 000 emu G mor 1 Below 10 kG, the magnetization varies linearly with the applied field with no apparent change in slope observed. Magnetic Model and Fits Recall the structure of 5 consists of ladder-like coordination polymers of azidobridged Cu(II) ions. The various types of superexchange pathways between the Cu(II) ions in 5 are schematically depicted in Figure 4-18. The rung or intradimer coupling (J.1) within the Cu2(N3)4 units is expected to be the dominant magnetic exchange interaction since the double end-on azide ligands bridge the square pyramidal copper ions through the equatorial coordination sites. Two exchange pathways are possible through the legs: the single end-to-end (Ji!) and single end-on (Jj1< 2 ) bridges between the rungs and the

PAGE 183

3 5 ,-_ 0 3 0 ::I s 2 5 cu .., 'o 2 0 ..... "-' u "' 1.5 ::E "C s:I 1.0 co u "' 0 5 N ::E 0 0 0 0.40 ----0 35 0 0.30 E ::s 5 0 25 "-' cu 0 20 0.1 S cu r-< 8. 0. 10 .-< 0 05 0 00 8 8 8 10 172 a Zero-Fie ld Cooled Magnetization o Field Cooled Magnetization Ooo Oo 20 D D D D D 30 40 T (K) Experimental x., vs. T Data -Fillo Tetramer Model with 2 J's D 0 so 100 150 200 250 300 T(K) A so B Figure 4-15. Magnetic data for compound 5 at 100 G from 2 K to 300 K. A) Field cooled (MFC) and zero-field cooled (MzFc) magnetization versus temperature. B) Molar susceptibility, XM per tetramer. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. The fit of x to the tetramer model with two coupling constants from Equations 4-11 to 4-15 is show by the solid line in B).

PAGE 184

173 250 ..--..--.--.---r--.---,-"-"T"--r---.---,r--.--..--.......... -, S 200 o E x p e rim e ntal .r,; vs. T Data High T Linear Fit E x trnpolated t o T Axis Cl) 0 5 150 .... Cl) s Cd .... Cl) E-< -:.z s ;:3 5 .._,, .... I Cl) E-< .... Cl) 0.. E-< ::E l'< 100 50 0 ....,....__ ___ _._.....___.____._____.____.___.__..__..__..__...._____.__. -50 1.5 1.4 1.3 1.2 1.1 1.0 0 9 0 8 0 7 0 so 100 1 so 200 250 300 T (K) Experimental .r, 1 vs T Dal -Fit to Tctramer Model wilh 2 J's 0 so 100 1 so 200 250 300 T (K) A B Figure 4-16. Magnetic data for compound 5 at 100 G from 2 K to 300 K. A) Inverse molar susceptibility, x1 per tetramer versus temperature. B) The product of the molar susceptibility and temperature, xT, per tetramer at versus temperature. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. A linear fit to the high temperature data in A) is shown by the solid line. The fit of xT to the tetramer model with two coupling constants from Equations 4-11 to 4-15 is show by the solid line in B).

PAGE 185

12 ,,..._ -~ 8 10 C, ::I 8 'b ..... '-' 6 b J 4 b 0.. 2 ;l 0 24 ,,..._ :.... 20 0 8 C, ::I 16 8 Q) ""o 12 ..... '-' ... 8 r--< ... 4 Q) Q. i" 0 0 10 0 500 174 Experimen tal M" vs B Data -F it to Tetramer Model "ith 2.f s Simulation 2 x S 1 /2 Brillouin F unction Simulation S ~ I Brillouin Function 20 30 40 B (kG) 50 Experimental MM vs B Data Simulation to Tetramer Model with Two Coupling Constants 1000 B (kG) 1500 2000 A B Figure 4-17. Magnetic data for compound 5. A) The molar magnetization (M) per tetramer at 2 K from Oto 50 kG. The data have been corrected for background signals arising from the sample container. The fit of M to the tetramer model with two coupling constants is shown as the solid line and the fit to the S = 1 Brillouin function is shown as the dotted line. B) A simulation of the M data to 2000 kG by the tetramer model. In the simulation, the upper critical field for the first plateau is 68 kG. The critical fields for the second plateau are Hc 1 = 1300 kG and Hc 2 = 1460 kG.

PAGE 186

175 Cu(terpy)(N 3 )2 units. Due to the long Cu-N contacts associated with the single azide bridges as well as the end-on and end-to-end azide ligands bridging the axial-equatorial coordination sites of the metal centers between the legs and rungs the exchange is expected to be weak A ladder model would best reproduce the magnetic behavior of 5. Assuming pair wise coupling of the unpaired electrons the appropriate spin Hamiltonian ladder model based on the structure from Figure 4-18 for describing the magnetic data is given by: i= I i= I i= I N / 4 N / 4 B t~::s. 1.J g,. s + Is. 12 g2 -s} ( 4-8 ) i=l i= l which assumes isotropic exchange interactions.271 The first term in the spin Hamiltonian represents the nearest neighbor exchange within and between the rungs and the second term is the Zeeman perturbation. The coordination environment of the copper ions within the rungs is distinct from that in the legs so two different g-values are needed to account for this difference. However no attempt to fit the experimental magnetic data based on this Hamiltonian was made due to the complexity of such processes. Furthermore no explicit expressions for the thermal and field dependence of magnetization and susceptibility for such a Hamiltonian to our knowledge exist in the literature. Therefore modeling the magnetic data according to this Hamiltonian was abandoned. By making certain assumptions regarding the relative magnitudes of the e x change interactions mediated by the various azide bridges a linear tetramer model with two

PAGE 187

176 coupling constants was employed to model the magnetic data.259 The tetramer consists of the central Cu2(N3) 4 moiety bridged to the two terminal Cu(terpy)(N3)4 groups via the 9 5 1 4 12 8 J23 or Jpara l J12 or Jperp zJ' or Jperp2 ,vvvvvv Figure 4-18. A fragment of the ladder-like structure of compound 5 depicting the possible superexchange pathways. The copper(II) ions, shown in the circles enclosed by the vertical lines, are numbered to correspond to Equation 4-8.

PAGE 188

177 single end-to-end azide ligands. Again, the exchange mediated by the single end-to-end azide bridges (J 12 ) is expected to be smaller, but not negligible, compared to the exchange mediated by the double end-on bridges (J 2 3). However, the intertetramer exchange mediated by the single end-on azide ligands is assumed to be very small, if not negligible in comparison. Despite the short Cu-N bonds, the Cu-N-Cu bridging angle (107 1 is very close to the angle of accidental orthogonality (108.5 ), where the ferroand antiferromagnetic contributions to the total exchange nullify each other. The single end on azide bridge is the magnetic ''weak" link in the ladder. The magnetic data is therefore analyzed assuming 5 consists of weakly interacting tetramers. Interactions between the ladders, such as exchange mediated by the overlapping terpy ligands, have been neglected. The exchange interactions of a linear, tetranuclear Cu(II) oligomers with two coupling constants can be described by the spin-Hamiltonian ,-. ,._ l'I. A ,._ I'\ A H=-2J 12 (S 1 2 +S 3 -S 4 )-2J 23 (S 2 3 )+ ,._ A A A (4-9) sB(g1S1 +g2S2 +g3S3 +g4S4) which assumes an isotropic exchange interactions where J23 describes the intradimer exchange interactions within the Cu 2 (N 3 ) 4 groups mediated by the double end-on azide bridges and J12 = J34 accounts for the exchange mediated by the single end-to-end azide bridges between the "rungs" and the Cu(terpy)(N3)2 moieties. For simplicity, g-values are assumed to be isotropic. The next-nearest neighbor intratetramer couplings, i.e. J 13 J14, andJ24, have been neglected. The exchange mediated by the single end-on azide bridges between the "rungs" and the terminal Cu(terpy)(N 3 )2 units is small and is

PAGE 189

178 approximated by a molecular field term ( 0) applied to the Hamiltonian. The zero-field energy levels of the spin-Hamiltonian are259 (4-10) The Van Vleck magnetic susceptibility vs. temperature is where Z, the partition function, is and parameters have the usual meaning. The mean field corrected susceptibility is z'= Xretram e ,(T ~ 0 ) (4-13) and the total susceptibility, corrected for a fraction of uncoupled, monomeric impurity, is N 2 2 X = x'(lp) + ~: :B p + Xo B (4-14) The best fit for both the x and xT data of 5 to Equations 4-10 to 4-14 was obtained using non-linear regression analysis with g, J12, J23, p, 0, and xTo or Xo as parameters. The

PAGE 190

179 results of the best fit are shown as the solid lines in Figures 4-15 and 4-16 with NR:::;: 0.952, g = 2.05, 21 12 = 10.697 K 3.408 K, 2123 = -108.283 K 1.188 K, 0= 0.174 0.165 K, p= 3.482 x 10-6, xTo = -5.550 x 106 emu K mor 1 and X,o = 6.92 x 106 emu mor 1 A similar fit was also obtained when 0 is negative. Given the present susceptibility data, little significance should be associated with the result that 0 is positive since in the absence of extremely low temperature data, the sign cannot be unambiguously determined. Although the agreement factor R = 0.329 is somewhat large, the model fits the data well. The best fit for 5 to the corresponding field dependent magnetization expression derived from equation 4-9 was obtained with NR, g, J12, J23, and Mo as parameters. 2 60 The results are shown as the solid line in Figure 4-17 with NR = 0.935, g = 2.11, 2112 = 8.310 K, 21 23 = -101.627 K, and Mo = -1.255 emu G mor 1 Although the agreement factor is large, expected for fits to M vs. B data, the NR, g, and J12, and J23 values are consistent with those obtained from the zT data and the fit to the data is relatively good. The magnetization data were also compared to simulations of both the S:::;: 1 Brillouin function and the S:::;: Brillouin function multiplied by two, shown in Figure 4-17 as the dotted and dashed lines, respectively. The simulations were performed with the fixed parameters, NR, g, and Mo, obtained from the fits to the susceptibility data. The average, isotropic g-value agrees well with the corresponding average value obtained from room temperature ESR measurements (g = 2.12). The data indicate the exchange between the double end-on bridges within the ladder "rungs" is strongly antiferromagnetic, an unexpected result. The exchange between the single end-to-end bridges is weak and ferromagnetic, and the exchange between the single end-on bridges

PAGE 191

180 in very weak. Thus the utilization of the tetramer model to analyze the magnetic data of 5 was justified. Interpretation of the Magnetic Data The zT(I') plot in Figure 4-16 depicts the thermal population of an excited triplet state (S = 1) at 10 K, then the S = 2 state at 50 K. Similarly, the M(B) plot in Figure 4-17 depicts the population of the S = 1 triplet state at 7 K. A simulation of the magnetization high field, up to 2000 kG (Figure 4-17), shows two plateaus indicating the population and saturation of the S = 1 state with an upper critical field of 68 kG followed by the population and saturation of the S = 2 state with a lower critical field of Hc1 = 1300 kG and an upper critical field ofHc2 = 1460 kG. The behavior of the variable field magnetization and variable temperature susceptibility can be correlated with Figure 4-19 Figure 4-19 schematically depicts two tetramers of 5 with the signs of the corresponding coupling constants within and between the tetramers, represented by the small arrows, determined from the fitting results obtained above. The schematic shows the changes in the total spin state of each tetramer as a function of increasing temperature and external field strength. At the lowest temperature, a singlet ground state is observed. Each tetramer has no net spin (S = 0) and the weak, intertetramer exchange, ( 0) is important and couples the tetramers ferromagnetically. However, as explained above, the intertetramer coupling could be antiferromagnetic, but the overall message conveyed by the diagram does not change. As the temperature is raised slightly, the intertetramer coupling is broken, but not the intratetramer exchange interactions, so the total spin of each tetramer is still zero, and a singlet state is still observed. As the temperature is increased to IO K, the plateau in the zTplot corresponds to an S = I excited state. This

PAGE 192

181 can be achieved if the weak, ferromagnetic coupling, J12, mediated by the single end-on azide bridges within each tetramer is broken and the spins at the ends of the oligomers align parallel with respect to each other. As the temperature is increased above 50 K, the xT product increases again, now approaching an S = 2 excited state. Sufficient thermal energy is available to break the strong, antiferromagnetic coupling, J23, mediated by the double end-on azide bridges within each tetramer. Now all of the spins per tetramer align in a parallel fashion. A similar effect is observed as an external field is applied to 5, although no change in the magnetization is observed at low field due to the very weak intertetramer coupling. From the M(B) plot in Figure 4-17, between 0 and 50 kG the data did not fit the S = 1 Brillouin function, but rather the product of two S = Brillouin functions indicating that the 112 coupling between the terminal metal centers with the rungs is weak since the first plateau in the magnetization is related to the exchange constant between these two types of magnetic centers. Rationalizing the Sign and Magnitude of the Coupling Constants Recall the exchange interactions within the dinuclear Cu 2 (N 3 ) 4 moieties are mediated by end-on azide ligands bridging the Cu(II) ions. Both the magnetic data and corresponding fits indicate that the intradimer coupling, J 23 is strongly antiferromagnetic. The negative sign of the coupling constant is not consistent with the small Cu-N-Cu bridging angle of 103.3 where a large ferromagnetic coupling is expected. The sign of the coupling constant is also not consistent with other similar -diazido bridged copper(II) complexes as there appear to be no examples with antiferromagnetic coupling when the bridging angle is less than 108.5 .6,17 237,246-249 However, the magnitude of the coupling constant is strong due to the short Cu-(bridging)N bonds, the planarity

PAGE 193

182 of the cyclic Cu-(Nh-Cu moiety, and the double end-on azido ligands that bridge basal-basal coordination sites (both associated with the d/-/ orbitals with significant unpaired spin density) on both Cu(II) ions (Figure 4-20).255 t)t-k'>' t"t-~(~ S=O t --.,,,,. ,,, t t t "' .. ~... ... ,., ... ,, .. t .,, .,.,.,, t t ~,;,. ,,.mo, '.,.,, t ., .. S=2 Increasing BandT Ill( S=l S= 0 Figure 4-19. The change in the total S per tetramer as a function of temperature or applied field in compound 5. The nearest neighbor interand intratetramer exchange interactions, represented by the small black arrows, were obtained from best fits to the bulk magnetic susceptibility data.

PAGE 194

183 The reason for this unexpectedly large antiferromagnetic intradimer exchange in 5 is not clear. Perhaps strong bonding interactions to neighboring rungs via the Cu(terpy)(N 3 ) 2 moieties or other electronic factors are reducing the ferromagnetic component to the point that the antiferromagnetic contribution to the total exchange dominates.21 Although these terminal units are only weakly coupled to the rungs, the presence of the terminal moieties may influence the relative energies of the magnetic orbitals within the dimers. The normally degenerate magnetic orbitals of the dimers must experience a dramatic shifting of the magnetic energy levels producing a large energy gap between the magnetic orbitals. Hence, a singlet ground state is strongly stabilized thus giving rise to the dominant antiferromagnetic coupling. Molecular orbital and singlet-triplet energy gap calculations may be required to confirm or reject this explanation. The exchange between both the single end-to-end (J12) and single end-on ( 0) azide bridges in 5 was determined to be small in magnitude. Both types of Cu(II) ions exhibit square pyramidal geometry. In the single end-to-end azide bride, one nitrogen atom is linked to the apical position of the Cu(terpy)(N 3 ) 2 group and the other nitrogen atom from the same bridge is coordinated to the basal sites of neighboring Cu 2 (N 3 ) 2 moiety (Figure 4-20). Conversely, in the single end-on azide bridge, the nitrogen atom is linked to both the basal position of the Cu(terpy)(N 3 )2moiety and the apical site of neighboring Cu2(N3)2 group (Figure 4-20). When the superexchange interaction is between the apical-basal coordination sites, the coupling is very small in magnitude. For square pyramidal and octahedrally coordinated Cu(II) ions, there is a good delocalization of the electron density to the bridging azide ligand coordinated to the basal Cu(II) ions

PAGE 195

184 since the unpaired spin density is primarily located in a d/-/ atomic orbital but the delocalization to the azide bridge from the apical Cu(II) ions is poor since the d/ orbital is occupied by paired spins and thus the unpaired spin density is low.6,190,255 The longer apical Cu-N distance compared to the analogous equatorial Cu-N distance also results in a weaker exchange interaction as well. Recall that the coupling through the single end-to-end bridge was ferromagnetic. The small Cu-N-N bridging angles (124.2 and 114.6 tend to favor antiferromagnetic coupling. However, the large Cu-N3--Cu torsion angle (89.8 ), along with the apical-basal bridging mode of the single end-on azide ligand and the square pyramidal geometry of the Cu(II) metal centers may orient the magnetic orbitals an almost orthogonal fashion to nullify the antiferromagnetic contribution of the total superexchange resulting in a net weak ferromagnetic coupling through the single end-toend bridges.255 For the single end-on bridges, the coupling was found to be very weakly ferromagnetic.216 As explained above, in the absence of low temperature data, the sign cannot be unambiguously determined. The Cu-N--Cu bridging angle (107.1 is very close to, but slightly lower than, the angle of accidental orthogonality (108.5 ), where the ferromagnetic and antiferromagnetic contributions cancel out one another and virtually no net exchange interaction is observed. The assignment of the intertetramer coupling to the exchange through the single end-on bridges rather than the single end-to-end bridges is justified. Thus, 5 consists of weakly interacting antiferromagnetically coupled tetramers and the reasoning for analyzing the magnetic data using a tetramer model as opposed to the ladder model is justified.

PAGE 196

185 Figure 4-20. The metal d-orbitals involved in the superexchange interactions present in compound 5. Within the dinuclear units that comprise the ladder rungs the end-on azide ions bridge both equatorial sites of the metal centers and thus a strong intradimer coupling is observed. The axial-equatorial coordination by the single end-on and end-to end azide ligands between the rungs and legs results in a weak coupling.

PAGE 197

186 Magnetic Properties of (Cu2(terpy)z--(N3)z(N3)2Cu3--(N3)4(N3)z) Magnetic Data A plot of the zero-field cooled magnetization (open boxes) and the field-cooled magnetization (open circles) is shown in Figure 4-21. The two sets of data superimpose one another indicating that 6 does not experience any long-range magnetic order down to 2 K. Plots of both the magnetic susceptibility (xM) vs. temperature and the inverse susceptibility vs. temperature for 6 are shown in Figures 4-21 and 4-22, respectively, as open squares. The susceptibility increases steadily as the temperature is lowered and no maximum is observed. The high temperature inverse susceptibility data (from 250 K to 300 K) were fit to a straight line and extrapolated to the x-axis with T= 40.72 K suggesting that the dominant magnetic exchange in 5 is ferromagnetic in nature. A plot of the temperature dependence of the product of the susceptibility and temperature (x M T) is shown in Figure 4-22 as open squares. The %MT value at room temperature (0.336 emu K mor 1 Cu(II) ions) is slightly lower that the expected spin-only value for a collection of non-interacting S =paramagnetic centers (xMT= 0.413 emu K mor 1 ). As the temperature is lowered, the %MT value increases until reaching a maximum value of 2.48 emu K mor 1 at 6.5 K. The increase is not smooth as two changes in slope of the data in this region are observed. The increasing of the xT product with decreasing temperature indicates the presence of a ferromagnetic coupling. Below 6.5 K however, the %MT value decreases approaching zero suggesting that a weak antiferromagnetic exchange interaction is present as well. Note that although 1 (I') is reported per mole of pentamer, x(I') and xT(I') are given as per mole of trimer; this discrepancy is addressed below. The field dependence of the molar magnetization (MM) per mole of pentamer for 6 is shown in

PAGE 198

s ,___ c., ::s 4 El Cl) ,., 3 'o ..... .._., u .... :E 2 '"O = (,:I u l ... N :E 0 0 0 6 0 5 ,___ 0 .4 ::s 5 0 3 .._., t e 0.2 r' & 0.1 X 0 0 0 0 0 0 10 187 Zero-Fi eld Co ol e d M ag n e t iza ti o n F i e ld Coo l e d M ag neti za t io n D D D D 20 30 40 T (K) Ex p er im e ntal z"' vs. T Data -Fi t to Trimer Model wi th One J D 0 SO I 00 I SO 200 250 300 T (K) A so B Figure 4-21. Magnetic data for compound 6 at 100 G from 2 K to 300 K. A) Field cooled (M Fc ) and zero-field cooled (M zpc ) magnetization versus temperature. B) Molar susceptibility, X M, per trimer The data have been corrected for background signals arising from the sample container and diamagnetic contributions. The fit of x to the uniform S = trimer model from Equations 4-18 to 4-20 is show by the solid line in B).

PAGE 199

,,...... 200 ::, 180 8 Cl) 0 160 E 140 .._, ... Cl) 1 20 E I .... 100 Cl) p.. 80 ... Cl) p. 60 ::. X 40 --20 l. 7S _,......_ l l.SO ::, 5 1.2S .._, i, -~ r-' i, p. 1.00 r-' ::. X 0 7S 188 D Experimenta l x,;' vs T Data High T Linea r Fit Extrapo l ated D 0 to T Axis so 100 ISO 200 2SO 300 T (K) Experimental x .T vs T Data -F it l o Trime r Model with one J 100 ISO 200 250 300 T(K) A B Figure 4-22. Magnetic data for compound 6 at 100 G from 2 K to 300 K. A) Inverse molar susceptibility z1 per petamer versus temperature. B) The product of the molar susceptibility and temperature zT per trimer at versus temperature. The data have been corrected for background signals arising from the sample container and diamagnetic contributions. A linear fit to the high temperature data in A) is shown by the solid line. The fit of zT to the uniform S =trimer model from Equations 4-18 to 4-20 is show by the solid line in B).

PAGE 200

1.0 ... Q) E 0 8 s C Q) 0.6 a.. ... / 1 0 4 :i 0.2 0.0 0 10 189 ,, -6 .. / .. / / D o Experimental M., IM.,. vs B Data Simulation S = 5/2 Brillouin Function Simulation 5 x S = 1 /2 Brillouin Function S i mulation S = 3/2 Brillouin Function + 2 x S = 1 /2 Brillouin Function 20 30 40 50 B (kG) A B Figure 4-23. A) The molar magnetization (M) per trimer at 2 K from Oto 50 kG for compound 6, normalized to the saturation magnetization, Msat The data have been corrected for background signals arising from the sample container. The simulation of M to the S = 5/2 Brillouin function, 5 x S = Brillouin function, and S = 3/2 Brillouin + 2 x S = Brillouin function is shown is shown as the solid line, the dotted line, and the dashed line respectively. B) A fragment consisting of two pentamers of compound 6 depicting the possible superexchange pathways. The copper(II) ions are shown as the shaded circles

PAGE 201

190 Figure 4-23 as open squares. The magnetization initially increases with increasing field but begins to level out at 30 kG asymptotically approaching a constant value of approximately 26 000 emu G mor 1 Below 10 kG, the magnetization varies linearly with the applied field with no apparent change in slope observed. Magnetic Model and Fits Recall the structure of 6 consists of ladder-like stacks of azido-bridged Cu(II) pentamers. Three distinct types of exchange pathways can be distinguished, depicted schematically in Figure 4-23. The central trinuclear Cu3(N3)6 units of each pentamer are expected to mediate the dominant superexchange interactions (J23) in 6 since the double end-on azide ligands bridge the square pyramidal copper ions through the basal coordination sites. However, the coupling (J 1 2) between the central Cu3(N3)6 trimers to the terminal Cu(terpy)(N3)2 moieties is expected to be, at best, weak due to the single end-on azide ligands bridging the apical-basal coordination sites of the metal centers. Due to the long Cu-N contacts associated with the asymmetric end-to-end and end-on azide bridges, the corresponding interpentamer exchange (J') is also expected to be weak. Any magnetic interactions between the ladders-like stacks are assumed to be negligible since only weak N-H(aromatic) con~cts and hydrophobic interactions are present between the organic ligands. The spin-Hamiltonian, with the corresponding Zeeman term, that describes the nearest-neighbor exchange within linear, pentanuclear Cu(II) oligomers with two distinct superexchange pathways resembling 6 is iI =-2J 12 (s 1 -s 2 +S 4 -S 5 )-2J 23 (s 2 3 +S 3 -S 4 )+ BB(i1S1 +i2S2 +i3-S3 +g4S4 +g5S5) (4-15)

PAGE 202

191 which asswnes an isotropic exchange interactions, 112 and 123, and neglects the next nearest neighbor intratrimer coupling between the end atoms, i.e. l13,l14,l2s,l36, etc. However, calculating the eigenvalues from Equation 4-16 is not trivial and the resulting analytical expressions for the field dependence of the magnetization and temperature dependence of the susceptibility are rather complex. The exchange between the terminal Cu(terpy)(N 3 )2 units and the central Cu 3 (N3)6 trimers, mediated by the single end-on azide bridges, is small compared to the intratrimer coupling. Thus the terminal units are asswned to behave in a paramagnetic fashion since they are magnetically isolated from the central trimers. The susceptibility of these two S = uncoupled units was calculated and subtracted from the total susceptibility of 6 to obtain the susceptibility due to the coupled trimers and the interactions between the trimers. Thus, 112 is neglected and 6 is asswned to consist of stacks of weakly interacting trimers. The exchange interactions of uniform, linear trinuclear Cu(II) oligomers can be well described by the spin-Hamiltonian with corresponding including the Zeeman term A A A A A = ,.. =A A H =-21(S1 2 +S2 SJ+ BB(gl SI+ g2 S2 + g3 S3) (4-16) which asswnes an isotropic exchange interaction, 1, within the central Cu 3 (N 3 ) 6 trimers and neglects the next-nearest neighbor intratrimer coupling between the end atoms, i e 113. For simplicity, g-values are assumed to be isotropic. The weak, intertrimer exchange is approximated by a molecular field term ( 0) applied to the Hamiltonian. The zero-field energy levels of the spin-Hamiltonian are E 1 = 2J (S = ), E 2 = 0 (S = ), and E 3 = -1 (S = 3/2).261 The Van Vleck susceptibility vs. temperature is

PAGE 203

192 (4-17) where the parameters have the usual meaning. The mean field corrected susceptibility is X = X Trimer ( T ) T-0 (4-18) and the total susceptibility, corrected for a fraction of uncoupled, monomeric impurity is (4-19) The best fit for both the x and xT data of 6 to Equations 4-17 to 4-19 was obtained using non-linear regression analysis with g, J, p, 0, and xTo or X,o as parameters. The results are shown as the solid lines in Figures 4-21 and 4-22 with NR = 1.007, g = 2.11, 2J= 14.598 K 0.556 K, 0= -1.526 K 0.034 K, p= 9.33 x 10-3, xTo = 9.3 x 10 4 emu K mor' and xo = 7. 7 x 10 4 emu mor'. The agreement factor R = 1.31 x 103 is very small indicating the model fits the data well. Because of the complexity of expressions describing the field dependence of magnetization for a five-spin system, no corresponding fits to the field dependent magnetization data were attempted. However, the data were compared to simulations of the S = 5/2 Brillouin function, S = Brillouin function multiplied by five, and the product of the S = 3/2 Brillouin function and two the S = Brillouin function, are shown in Figure 4-23 as the solid, dotted, and dashed lines, respectively. The simulations were performed with the fixed parameters, NR, g, and Mo, obtained from the fits to the susceptibility data.

PAGE 204

193 The average, isotropic g-value agrees well with the corresponding average value obtained from room temperature ESR measurements (g = 2.14). The magnetic data and corresponding fits indicate that the dominant exchange interaction within the central Cu3(N 3 ) 6 trimers in 6 is weakly ferromagnetic and the interpentamer coupling is weakly antiferromagnetic. Furthermore, the assumption that 6 consists of weakly interacting trimers by neglecting the weak interpentamer exchange, 112, between the central Cu3(N3)6 trimers and the terminal Cu(terpy)(N 3 ) 4 units, is justified since the corresponding trimer susceptibility expressions fit the magnetic data well. Interpretation of the Magnetic Data The xT plot in Figure 4-22 depicts the thermal population of the S = 5/2 ground state at low temperature (T= 6.5 K). At high temperature, 6 behaves paramagnetically, i.e., the unpaired spins within each pentamer along the ladder-like chains are uncoupled and essentially randomized resulting in a small net magnetic moment. As Tis lowered, the intrapentamer coupling takes effect, as insufficient thermal energy is present to populate excited states. An increase in the xT product is observed due to the onset of the ferromagnetic exchange interactions. Two changes in slope of the data in this region possibly indicate the onset of two distinct coupling constants. At T= 6.5 K, the S= 5/2 ground state (where all five spin vectors per pentamer align in a parallel fashion) is fully populated and 6 consists of stacks of ferromagnetically coupled pentamers. At even lower temperatures, a sudden decrease in the xT product indicates the weak interpentamer interactions take effect. The magnetization data in Figure 4-23 depicts the population of the S = 5/2 state at high field. Since all of the exchange interactions, 1 12 123, and 0, are relatively weak, 6 resembles a linear array of five weakly interacting S =

PAGE 205

194 metal centers since the magnetization data follow the S = Brillouin function. If both interpentamer coupling constants, J 12 and J 23 were strong, the magnetization data would follow the S = 5/2 Brillouin function. If the intratrimer exchange, J, within central Cu 3 (N 3 ) 6 trimers was strong but the exchange, J12, between the two terminal Cu(terpy)(N 3 ) 4 units was weak, then the magnetization data would be reproduced by the S = 3/2 Brillouin function + the S = Brillouin function multiplied by two. Rationalizing the Sign and Magnitude of the Coupling Constants Recall the exchange interactions within the trinuclear Cu3(N3)6 moieties are mediated by end-on azide ligands bridging the Cu(ID ions. Both the magnetic data and corresponding fits indicate that the intratrimer coupling, J, is weakly ferromagnetic. The positive sign of the coupling constant is consistent with the small Cu-N-Cu bridging angle of 101. 7 However, the magnitude of the coupling constant is much smaller than expected and is not consistent with other similar -diazido bridged copper(II) complexes. 6,17,237,246-249 Toe small bridging angle, short Cu-(bridging)N bonds, the planar Cu-(N)2-Cu moiety, and the double end-on azido ligands that bridge basal-basal coordination sites (both associated with the d/-/ orbitals with significant unpaired spin density) on both Cu(ID ions (Figure 4-24) suggest that the magnitude of the coupling should be greater.255 The reason for this unexpectedly small ferromagnetic exchange is not clear but it does correlate with the unexplained antiferromagnetic intradimer coupling observed in 5. Perhaps structural-bonding or electronic factors are reducing the ferromagnetic component and thus enhancing the antiferromagnetic contributions to the total exchange within the trinuclear Cu3(N3)6 moieties. One possible cause could be the presence of the

PAGE 206

195 monomeric Cu(terpy)(N 3 ) 2 units linked to the ends of the trimers in 6.21 Note that similar Cu(terpy)(N 3 ) 2 groups are present in 5 as well. Although these terminal units are essentially magnetically isolated from the coupled central trimers, the presence of the terminal moieties may influence the relative energies of the magnetic orbitals within the trimers. The normally degenerate magnetic orbitals of the trimers could be experiencing an increased energy gap thus reducing the ferromagnetic contribution and increasing the antiferromagnetic effects. In the case of 6, however, the shifting of energies is not sufficient to stabilize a singlet state, the antiferromagnetic contribution does not dominate, and the net exchange is still, albeit weakly, ferromagnetic. Molecular orbital and singlet-triplet energy gap calculations may be required to confirm or reject this explanation. Recall that, for the purposes of fitting the magnetic data, the exchange between the single end-on azide bridges, J12, between the Cu(terpy)(N3)i units and the Cu3(N3) 6 moieties was neglected. Therefore, the sign and magnitude of the coupling constant was not determined. However, based on the structural data, the magnitude of the exchange is expected to be weak. In the single end-on azide brides, the same nitrogen atom is linked to the basal position of the Cu(terpy)(N3)2 moiety and to the apical sites of Cu 2 (N 3 ) 2 group (Figure 4-24).6,205,255 There is a good delocalization of the electron density to the bridging azide ligand coordinated to the basal Cu(II) ions since the unpaired spin is primarily located in a d}-/ orbital but the delocalization to the azide bridge coordinated to the apical Cu(II) ions is poor since the d/ orbital is occupied by paired spins. The longer apical Cu-N distance compared to the analogous basal Cu-N distance also results in a weaker exchange interaction as well. Computer simulations of zT vs. T based

PAGE 207

196 with analytical susceptibility expressions derived from Equation ( 4-16), utilizing the 123 value obtained from the above fits, suggests that 112 is weakly ferromagnetic. The xT vs. T plot in Figure 4-22 is uniquely reproduced only when both coupling constants are positive and 1 23 > > 1 1 2. When 112 is negative and small, for example, a broadened maximum in xT at low temperature, rather than the observed sharp peak, results. Attempts to rationalize the sign were based on similar magnetostructural correlations for single end-to-end bridges. Despite the large Cu-N-Cu bridging angle (117 .8 that is expected to favor antiferromagnetic coupling, the azide ligands bridging the approximately perpendicular oriented apical-basal coordination sites of the square pyramidal Cu(II) ions between the metal centers the may orient the magnetic orbitals an orthogonal or nearly orthogonal fashion to nullify the antiferromagnetic contribution of the total superexchange resulting in a net weak, ferromagnetic coupling.205,216,255 The interpentamer superexchange pathway consists of asymmetric, double end-on and single end-to-end bridging azide ligands. Both the magnetic data and corresponding fits indicate that the interpentamer coupling is weakly antiferromagnetic. The negative sign of the interpentamer coupling is consistent with the interdimer exchange observed in 4 since the structural parameters of the asymmetric azide bridges in both samples are similar. The small, antiferromagnetic coupling is also consistent with asymmetric end-on and end-to-end azide ligands bridging square pyramidal Cu(II) ions in other similar complexes as welt.205,255 The long, weak Cu-N bonds creates large distances between the metal centers and thus orbital overlap between the two exchanging metal centers is poor resulting in the small coupling constant. The azide ligands bridging the apical-basal coordination sites of the square pyramidal Cu(II) ions between the pentamers

PAGE 208

197 also accounts for the small interpentamer coupling as well (Figure 4-24).205 2 55 Therefore 6 consists of antiferromagnetic stacks of pentamers with two paramagnetic S = sites and ferromagnetically coupled trimers. /N~N '-.._ ~ / '-..,. N / N N I I N / I / Figure 4-24. The metal d-orbitals involved in the superexchange interactions present in compound 6. Within the central trinuclear units the end-on azide ions bridge both equatorial sites of the metal centers and thus a strong intratimer coupling is expected The axial-equatorial coordination by the single end-on azide ligands to the terminal Cu terpy moieties is expected to provide a poor superexchange pathway resulting in a weak coupling. The long Cu-N contacts between the pentamers as well as the axial equatorial coordination by the azide ligands results in a weak interpentamer coupling.

PAGE 209

198 Conclusions This chapter summarizes the experiments performed in order to investigate the structural and magnetic properties of three azido-bridged copper(II) ladder-like coordination polymers, [Cu2(PhPyPy)2--(N3)2(N3)2], 4, [Cu2(terpy)2--(N3)4Cu2(N3)2(N3)2], 5, and [Cu2(terpy)2--(N3)2(N3)2Cu3--(N3)4(N3)2], 6. Compound 4 structurally resembles ladder-like chains of weakly interacting end-on azido bridged copper(II) dimers. Magnetically, 4 consists of antiferromagnetic chains of ferromagnetically coupled S = dimers. Compound 5 consists of ladder-like copper(II) coordination polymers with double and single end-on and single end-on azido bridges. Magnetically, 5 consists of ladder-like stacks of weakly interacting tetramers with a dominant antiferromagnetic exchange, but this coupling cannot be easily rationalized on the basis of present structural-bonding parameters. Compound 6 structurally resembles ladder-like chains of weakly interacting copper(II) pentamers featuring both single and double end-on azide bridges. Magnetically, 6 consists of antiferromagnetic stacks of pentamers with two paramagnetic S = sites and ferromagnetically coupled trimers.

PAGE 210

CHAPTERS CONCLUSIONS This dissertation describes the synthesis and structural, chemical, and magnetic properties oflow-dimensional coordination polymers. The original goal ofthis project focused on the design self-assembly, and magnetic properties of molecular coordination polymer ladders Instead, several new quasi-one-, one/two-, and two-dimensional materials were isolated as a result of serendipity rather than rational design but did not incorporate the desired structural and physical properties of ladder-like systems. Consequently, the work presented here is concerned primarily with the characterization of selected examples of these low-dimensional materials obtained in the course of this graduate research. Chapters 2 and 3 describe the structural, thermal, host-guest, and magnetic properties of a series of two-dimensional, layered, network solids. Chapter 4 details the structural and magnetic properties of a series of ladder-like azido bridged Cu(II) coordination polymers. The structural, thermal, and magnetic properties of a series of clathrated porous network solids, [Ni( 4,4 '-bipy)J(H2O)2](ClO4)2 1.4( 4,4 '-bipy)-3(H2O), 1, [Co(4,4' -bipy)3(H2O)2](ClO4)2 1.4(4,4' -bipy)(H 2 O), 2, and [Cu( 4,4 '-bipy)3(DMSO)2](ClO4)2 2( 4 4 '-bipy) 3, are in Chapter 2. These materials consist of chains of transition metal ions (Ni(II), Co(II), and Cu(II)) bridged by 4,4 '-bipyridine spacer ligands. The chains pack to form two-dimensional non interpenetrated sheets with hydrophobic, rectangular cavities within the framework. The 199

PAGE 211

200 sheets, in turn, pack to form layered, three-dimensional solids with oblique, extended channels containing enclathrated guest molecules and counterions. The guest molecules are easily lost, suggesting that the samples are thermally unstable. The magnetic properties of 1, 2, and 3 are similar in the sense that only very weak exchange interactions are present between the metal centers. The host-guest properties of 1 and 2 are investigated in Chapter 3. Gas chromatography experiments determined that both hosts exchange clathrated bipy molecules with trialkylphosphine oxide probe molecules, TMPO (trimethlyphosphine oxide), TEPO (triethylphosphine oxide), and TPPO (tripropylphospine oxide). While the uptake of TMPO by both 1 and 2 is essentially complete, steric constraints are believed to limit the uptake of TEPO and TPPO by the host. The trialkylphosphine oxides interact with acid sites within the host as determined by 31 P MAS NMR spectroscopy. TMPO interacts with both coordinated water molecules (strong acid sites) and lattice waters (weak acid sites) in compound 1 and coordinates directly to the metal centers in compound 2. However, TEPO and TPPO seem to attack only the weaker acid sites within the hosts. X-ray diffraction patterns show that the loss ofbipy and uptake of the probe causes significant structural rearrangements in 1 and only relatively mild structural changes in 2. However, the nature of these guest-exchanged products is unknown. These experiments show that solid-state NMR spectroscopy can be used to investigate host guest interactions. Finally, the structural and magnetic properties of three azido-bridged copper(ID ladder-like coordination polymers, [Cu2(PhPyPy) 2 --(N 3 ) 2 (N 3 )2], 4, [Cu2(terpy)2--(N3)4Cu2--(N3)2(N3)2], 5, and

PAGE 212

201 [Cu 2 (terpy) 2 --(N 3 )2(N3)2Cu3--(N3)4(N3)2] 6, are discussed in Chapter 4. Compound 4 structurally resembles ladder-like chains of weakly interacting end-on azido bridged copper(II) dimers. Magnetically compound 4 consists of antiferromagnetic chains of ferromagnetically coupled S = dimers. Compound 5 consists of ladder-like copper(II) coordination polymers with double and single end-on azido bridges. Magnetically compound 5 consists ofladder-like stacks of weakly interacting tetramers with a dominant and unusual antiferromagnetic exchange mediated through end-on azido bridges. Compound 6 structurally resembles ladder-like chains of weakly interacting copper(II) pentamers featuring both single and double end-on azide bridges. Magnetically compound 6 consists of antiferromagnetic stacks of pen tamers with two paramagnetic S = sites and ferromagnetically coupled trimers. The study of low-dimensional materials has been a rapidly expanding area of solid-state chemistry. Because of the anisotropic bonding present, these systems often exhibit unique or enhanced chemical and physical properties compared to higherdimensional systems4 such as low temperature catalysis 41-43 inclusion phenomena,37 44 45 magnetism,40 46-49 electrical conductivity 50-52 photochemistry 53 and second-order nonlinear optical behavior.54-57 This dissertation describes the chemical (host-guest) properties of selected porous layered two dimensional sheets as well as the physical (magnetic) properties ofladder-like coordination polymers. These samples represent new examples of low-dimensional materials and an understanding of their characteristic structural chemical, and physical properties can better contribute to solid-state chemistry.

PAGE 213

APPENDIX A CRYSTAL STRUCTURES OF SELECTED LOW-DIMENSIONAL SOLIDS Comment A few selected structures, obtained throughout the course of the graduate research, are presented in this appendix. As discussed at the beginning of the introduction, many products were obtained from the various synthetic attempts employed to self-assemble ladder-like coordination polymers that did not meet with success. Since these materials did not quite fit within the context of the preceding chapters, a few are included in this appendix for completeness. Discussed are the structures of [Ni(terpy)(H2O)][Ni(CN) 4 )], 7, a one-dimensional linear chain comprised of alternating [Ni(terpy)(H 2 O)] 2 + cations bridged by trans-[Ni(CN) 4 )]2anions. Also presented are [Zn( 4,4 '-bipy)(DMSO)4]n(ClO4)2n, 8, a linear one-dimensional chain with trans-4,4 bipyridine bridging ligands and [Cu(4,4'-bipy)(DMSO) 4 ]n(ClO 4 )2n, 9, a zig-zag chain with cis-4,4 '-bi pyridine bridges. Finally [Cd( 4,4 '-bipy)3(H 2 O) 2 ](ClO 4 ) 2 2(DMSO), 10, a layered sheet structure with clathrated hydrophobic pores, similar to structures 1, 2, and 3, is detailed. Experimental Section Materials The starting materials nickel(II) perchlorate hexahydrate (98 % ), zinc(II) perchlorate hexahydrate (98 % ), copper(II) perchlorate hexahydrate (98 % ), and cadmium(II) perchlorate hexahydrate (98 %), 4,4' -bipyridine (98 %) and 2,2':6',2" 202

PAGE 214

203 terpyridine (98 % ) were purchased from Aldrich Chemical Co. Potassium cyanide (99.9 %), dimethyl sulfoxide (99.9 %), and ammonia (15 M) were purchased from Fisher Scientific. All reagents were used without further purification. Synthesis of [Ni(terpy)(H20)][Ni(CN)4)) The combination ofNi(ClO 4 ) 2 H 2 O (366 mg, 1 mmol) with terpy (234 mg, 1 mmol) and KCN (97.5 mg, 1.5 mmol) in water (50 mL) produced a tan precipitate. Addition of NH 3 solution (20 mL, 15 M) followed by 50 mL of 100 % ethanol with stirring dissolved the precipitate resulting in a yellow-colored solution. The reaction mixture was filtered into a 500 mL Erlenmeyer flask, capped with paraffin (punctured with small holes), and set aside for crystallization. Within about four months, small brown blocks appeared in solution. These blocks were determined to be [Ni(terpy) 2 ](ClO 4 ) 2 2 O, a known monomeric bis(terpy)nickel(II) complex.262 Small gray needles of 7 were obtained after an additional month of solvent evaporation. Synthesis of (Zn( 4,4 '-bipy)(DMS0)4)n(CI04)2n A solution containing 741 mg of Zn(ClO 4 )2"6 H 2 O (2.0 x 103 mol) dissolved in 10 mL ofDMSO was combined with a solution containing 467 mg of 4,4'-bipyridine (4.0 x 103 mol) dissolved in 10 mL of DMSO. The resulting clear mixture, contained within an evaporating dish, produced small plates of 8 within two weeks of solvent evaporation. Synthesis of [Cu(4,4'-bipy)(DMS0)4)n(CI04)2n A solution containing 741 mg of Cu(ClO 4 )2"6 H 2 O (2.0 x 103 mol) dissolved in 10 mL of DMSO was combined with a solution containing 311 mg of 4,4 '-bi pyridine (2.0 x 103 mol) dissolved in 10 mL ofDMSO. The resulting light blue colored mixture, contained within an evaporating dish, produced small blue blocks of 9 within two to three weeks of solvent evaporation.

PAGE 215

204 Synthesis of [Cd(4,4'-bipy)3(H20)2](CI04)2(DMSO) A solution containing 741 mg of Cd(ClO4)2 H2O (2.0 x 103 mol) dissolved in 10 rnL of DMSO was combined with a solution containing 934 mg of 4,4' -bipyridine (6.0 x 103 mol) dissolved in 10 mL ofDMSO. The resulting clear mixture, contained within an evaporating dish, produced large blocks within a month of solvent evaporation. X-ray Structure Determination Gray needles of 7 (0.17 x 0.09 x 0.09 mm3), clear plates of 8 (0.04 x 0.20 x 0.23 mm3), blue blocks of 9 (0.32 x 0.19 x 0.19 mm3), and clear blocks of 10 (0.38 x 0.38 x 0.38 mm 3 ) were selected for X-ray analysis. Each crystal was mounted on a glass fiber under nitrogen gas. The same data collection was used for each sample. Data were collected at 173 K on a Siemens SMART PLATFORM equipped with a CCD area detector and a graphite monochromator utilizing MoKa radiation (A= 0.71073 A). Cell parameters were refined using 3713 reflections for 7, 4817 reflections for 8 and 9, and 8192 reflections for 10. A hemisphere of data (1381 frames) was collected using the ro scan method (0.3 frame width). The first 50 frames were re-measured at the end of data collection to monitor instrument and crystal stability (maximum correction of I was< 1 % ). Absorption corrections by integration were applied based on measured indexed crystal faces. All structures were solved by the Direct Methods in SHELLXTL6 and refined using full-matrix squares.155 The non-H atoms were treated anisotropically, whereas the hydrogen atoms were calculated in ideal positions by riding on their respective carbon atoms. For 7, a total of264 parameters were refined using F 2 in the final cycle using 3713 reflections with I> 2cr(I) to yield R 1 = 3.00 % and wR 2 = 6.75 %. For 8, a total of

PAGE 216

205 363 parameters were refined using F 2 in the final cycle using 6069 reflections with I> 2cr(I) to yield R 1 = 2.74 % and wR2 = 6.71 %. For 9, a total of 199 parameters were refined using F 2 in the final cycle using 3097 reflections with I > 2cr(I) to yield R 1 = 2.90 % and wR 2 = 6.98 %. For 10, the non-H atoms were treated anisotropically, whereas the hydrogen atoms were calculated in ideal positions by riding on their respective carbon atoms except for the the hydrogen atoms from the two water molecules, which were found from a Difference Fourier map and refined freely. The asymmetric unit consists of [Cd(4,4'-bipy)J(H2O)2] units, two perchlorate anions and two DMSO molecules A sulfur atom from one of the DMSO molecules was disordered and refined in two parts. Their site occupation factors were dependently refined to 0.85(1) for the major part, and consequently 0.15( 1) for the minor part. A total of 540 parameters were refined using F 2 in the final cycle using 8601 reflections with I> 2cr(D to yield R 1 = 3.72 % and wR 2 = 9.90%. Description of the Structures Crystallographic and structural refinement data for 7 and 8 are listed in Table A-1 and for 9 and 10 in Table A-2. Tables of atomic coordinates and thermal displacement parameters are provided in Appendix B. Structure of [Ni(terpy)(H20))[Ni(CN)4)) The structure of 7 consists of infinite, one-dimensional linear chains comprised of alternating [Ni(terpy)(H2O)]2 + cations and [Ni(CN) 4 ] 2 anions that extend along the [I 0-1] direction. Figure A-1 shows a typical chain fragment. In the [Ni(terpy)(H 2 O)] 2 + cations, the local coordination environment of the metal centers is distorted octahedral. The equatorial plane is defined by three nitrogen atoms (NS, N6 and N7) from the terpy

PAGE 217

206 Table A-1. Summary of Crystallographic Data for Compounds 7 and 8 Sample 7 8 Empirical Formula C19H13N1Ni2O C1sH32ChN2O12S~n Formula Weight 472.78 732.97 Space Group Monoclinic, Cc Triclinic, P-1 a, A 15.5712(8) 8.3491(4) b, A 11.5546(6) 9.3154(4) c, A 11.2428(6) 19.4783(8) a, deg 90 86.677(1) p,deg 97.076(1) 89.337(1) y,deg 90 89.421(1) V A 3 2007.4(2) 1512.2(1) z 4 2 T,K 173(2) 173(2) A(Mo Ka), A 0.71073 0.71073 Peale, g cm 3 1.564 1.610 mml 1.900 1.323 Ria (wRl) 0.0300 (0.0675) 0.0274 (0.0671) a RI= ~)II Fa 1-1 Fe II) /LI Fa I bwR 2 = [L[w(Fi -F;)2]/L[w(Fi)2]] 1 1 2 S = [L[w(Fi -F;)2]/(n-p)]1' 2 w = 1/[cr 2 (F;) + (0.0370 p)2 + 0.31 p ],p = [max(Fo 2 ,0) + 2 F/]/3

PAGE 218

207 Table A-2. Summary of Crystallographic Data for Compounds 9 and 10 Sample 9 10 Empirical Formula C1sH32Cl2CuN2O12S4 C3J-l4oCdCliN6O12S2 Formula Weight 731.14 972.14 Space Group Trigonal, P3(2)21 Orthorhombic, Pna2(1) a,A 10.6492(5) 23.563(2) b, A 10.6492(5) 12.7457(8) c,A 23.036(2) 13.6674(8) a, deg 90 90 p, deg 90 90 y,deg 120 90 V A 3 2262.4(2) 4104.7(4) z 3 4 T,K 173(2) 173(2) ).,(Mo Ka), A 0.71073 0.71073 Peale, g cm 3 1.610 1.573 mm 1 1.236 0.831 R1 3 (wR/) 0.0290 (0.0698) 0.0372 (0.0990) 3 R1 = 1)11 Fo 1-1 Fe JI) (LI Fo I bwR2 = [~)w(F; -F})2]/~)w(F;)2]f 2 S = [~)w(F; -Fc 2 )2]/(n-p)f 2 w = 1/(cr 2 (F;) + (0.0370* p)2 + 0.31 p],p = [max(Fo 2 ,0) + 2 ~ 2 ]/3

PAGE 219

208 ligand and one oxygen atom (01) from the aqua ligand. The Ni2-N5 and Ni2-N7 bonds (2.11 A2.13 A) from the peripheral pyridyl moieties of the terpy ligands are longer than the Ni2-N6 bond (2.01 A) from the central pyridyl group and the Ni2-01 bond (2.06 A). The axial coordination sites are filled by the nitrogen atoms from the bridging cyano groups from adjacent [Ni(CN) 4 )]2anions. Since the axial Ni2-N2 and Ni2-N4 bonds (2.08 A-2.09 A) are of comparable length to the Ni2-01 bond and Ni2-N6 bond to the central pyridyl moiety of the terpy, the axis of elongation is N7Ni2-N5 from the peripheral pyridyl nitrogen donors of the terpy. The deviation ofNi2 from ideal octahedral geometry is partly due to the geometric requirements of the sterically bulky terpy ligands. The Ni-0 (water) and Ni-N (terpy) bond angles and distances agree with those for other similar Ni-terpy complexes.263-265 In the planar [Ni(CN) 4 ] 2 anions, the bridging cyano ligands are positioned trans to one another. The mean deviation of the Ni(II) ions from the least squares plane, Nil Cl-C2-C3-C4, is 0.0037 A. The Ni-Cl and Ni-C4 bonds (1.85 A) from one pair of terminal and bridging cyano groups are shorter than the other Ni-C2 and Ni-C3 pair (1.89 A) while, in contrast, the Cl-Nl and C4-N4 bonds (1.17 A) are longer than the C2-N2 and C3-N3 bonds (1.12 A). Interactions with nearby terpy ligands force the anion to twist, thereby relieving the repulsion felt between the terpy and the terminal cyano groups. This twist forces angles Ni2-N2-C2 of 172.2 and Nil-C2-N2 of 176.2 to deviate significantly from the ideal geometry. This repulsion can also be seen in the opening of angle Cl-Nil-C2 to 91.1 The Ni-C and C-N bond distances, from both the bridging and terminal cyano groups, in the [Ni(CN) 4 ]2group are typical of those found in other similar trans-bridging tetracyanonickelate chains. 266,267

PAGE 220

209 A single chain of 7 is shown in Figure A-1. Within the chains the distance between two octahedral Ni(II) ions is 10.149 A while the distance between octahedral and planar metal centers is 5.075 A. Note the regular alternation in regards to the relative positions of both the [Ni(CN) 4 ] 2 groups and terpy ligands along the chain. Adjacent [Ni(CN) 4 ] 2 groups are oriented nearly perpendicular (86.3 with respect to one another. The terpy ligands alternate along the chains in an antiparallel fashion with respect to one another. The packing diagrams of 7 are shown in Figure A-2. Adjacent chains are related by translation along the crystallographic a, b-, and c-axes. In the solid, the chains pack to maximize both JZ"-stacking and hydrogen bonding interactions between the chains. The offset JZ"-stacking interactions originate from the overlap of terpy ligands between neighboring chains with a face-to-face distance of~ 3.5 A. The protons from the aqua ligands form twin hydrogen bonds (1.88 A 1.90 A) with the terminal nitrogen atoms from monocoordinate cyano group on two adjacent chains. Ordoenac, et al has published a review of detailing the synthesis crystal structures, and magnetic properties of one-dimensional tetracyanonickelate complexes. 268 Structure of [Zn(4,4'-bipy)(DMS0)4]n(CI04)2n The structure of8 consists of cationic one-dimensional [Zn(4,4 -bipy)(DMSO ) 4 ]2 + chains and ClO4anions The local coordination environment surrounding a typical Zn(II) ion is depicted in Figure A-3. The metal center is six-coordinate and the coordination sphere consists of two pyridyl nitrogen donors from two 4,4 -bi pyridine

PAGE 221

210 A B Figure A-1. [Ni(terpy)(H20)][Ni(CN)4)]. A) A typical chain fragment. B) A one dimensional linear chain. All aromatic hydrogen atoms have been omitted from the top figure for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 222

211 A B C C C b a_J Figure A-2. [Ni(terpy)(H20)][Ni(CN)4)]. A) The structure of 7 along the [1 0-1] direction. B) The structure within the crystallographic ac-plane. C) The structure within the crystallographic ab-plane. For clarity, the aromatic hydrogen atoms have been omitted.

PAGE 223

212 ligands and four oxygen atoms from ligated DMSO molecules. The Zn-01, Zn-OIA, Zn -NI, and Zn -NIA bond distances are all similar (2.13 A) but slightly shorter than the Zn -02 and Zn -02A bonds (2.15 A). The local coordination geometry of ZnN2O4 adopts an axially elongated octahedral geometry with two pyridyl nitrogen atoms (Znl and ZnlA) and two DMSO oxygen atoms (01 and OlA) occupying the equatorial positions while the remaining two DMSO oxygen atoms (02 and O2A) fill the axial sites. The bipy ligands are coordinated to trans to one another other. The pyridyl rings the bipy ligands are not coplanar but twisted along the central C-C bond at an angle of 38 0 with respect to each another. All Zn-0 and Zn-N bond angles and distances are consistent with those reported for other similar Zn-bipy complexes.133 269 The molecular structure of 8 consists of infinite one-dimensional chains that extend along the [1 -1 1] direction. The trans-coordination of the bridging bipy ligands results in a linear configuration of the chains, shown in Figure A-3 The distance between any two adjacent Zn(II) ions within a chain is 11.32 A. Packing diagrams of8 are shown in Figures A-4. In the solid adjacent chains are related by a single translation along the a-axis and of a "step along both band c. The methyl groups from the DMSO ligands on adjacent chains interact with each other in a side-by-side fashion. The perchlorate counterions reside in the void spaces present between the chains in close proximity to methyl groups of the DMSO ligands. However no S-0 contacts between the perchlorate oxygen atoms and the DMSO sulfur atoms are observed.

PAGE 224

213 A B Figure A-3. [Zn(4,4 -bipy)(DMSO)4]n(ClO4)2nA) A typical chain fragment. B) A one dimensional linear chain. All hydrogen atoms have been omitted from both figures for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 225

214 b A B b C C a C Figure A-4. [Zn(4,4' -bipy)(DMSO) 4 ]n(ClO 4 )2n. A) The structure within the crystallographic be-plane. B) The structure along the [1 -1 l] direction. C) The structure within the crystallographic ac-plane. For clarity, the hydrogen atoms have been omitted.

PAGE 226

215 Structure of [Cu(4,4'-bipy)(DMS0)4)n(CI04)2n The structure of 9 consists of cationic one-dimensional [Cu( 4,4 bipy)(DMSO) 4 ] 2 + chains and ClO 4 anions. The local coordination environment surrounding a typical Cu(II) ion is depicted in Figure A-5. The metal center is six coordinate and the coordination sphere consists of two pyridyl nitrogen donors from two 4,4' -bipyridine ligands and four oxygen atoms from ligated DMSO molecules. The Cu O2, Cu-02A, Cu-NI, and Cu-NIA bond distances are all similar (2.0I A) but shorter than the Cu-OI and Cu-OIA bonds (2.30 A). The CuN2O4 unit therefore adopts an axially elongated geometry with two pyridyl nitrogen atoms (NI and NIA) and two DMSO oxygen atoms (02 and O2A) occupying the equatorial positions while the remaining two DMSO oxygen atoms (01 and OIA) fill the axial sites. Not only are the bipy ligands coordinated cis to one another but are positioned at nearly right angles with respect to one another as well (90.9 ). The pyridyl rings of the bipy ligands are not coplanar but twisted along the central C-C bond at an angle of 17 .1 with respect to each another. All Cu-0 and Cu-N bond angles and distances are consistent with those reported for other similar Cu-bipy complexes.92,69,I04 The molecular structure of 9 consists of infinite, one-dimensional chains that extend along the crystallographic c-axis. The cis-coordination of the bipy ligands that bridge the metal centers results in a zig-zag configuration of the chains, shown in Figure A-5. The distance between any two adjacent Cu(II) ions within a chain is approximately c/2 units. Packing diagrams of9 are shown in Figures A-6. In the solid, adjacent chains are related by single translations along both the aand b-axes. The chains pack in registry with respect to one another along both the aand b-axes. The methyl groups

PAGE 227

216 from the DMSO ligands on adjacent chains interact with each other in a side-by-side fashion. The perchlorate counterions reside in the void spaces present between the chains in close proximity to the methyl groups of the DMSO ligands. The oxygen atoms of the perchlorate anions interact with the sulfur atoms from the DMSO ligands with S--Cl contacts of 3 .22 A. Structure of [Cd(4,4'-bipy)3(H 2 0)2](CI04)z-2(DMSO) The structure of 10 consists of one-dimensional cationic [Cd(4,4' -bipy)3(H 2 O) 2 ] 2 + chains that pack to form layered sheets in the solid. The local coordination environment surrounding a typical Cd(II) ion is depicted in Figure A-7. The metal center is six coordinate and the coordination sphere consists of four pyridyl nitrogen donors, one from four 4,4 '-bipyridine ligands and two oxygen atoms from aqua ligands. The four nitrogen atoms define the equatorial plane and the oxygen atoms occupy the axial sites. Both Cd--0 bond distances are unequal (Cd--01 = 2.30 A and Cd--02 = 2.32 A) and one Cd-{bridging)N bond (Cd-N2' = 2.38 A) is longer than the other three Cd-N bonds (Cd-Nl, Cd-N2, and Cd-N3 = 2.34 A). Since the Cd--0 bonds are shorter than the Cd-N bonds, the CdN4O2 unit adopts an axially compressed geometry with the compression axis O2--Cd--O 1 from the aqua ligands. All Cu--0 and Cu-N bond angles and distances are consistent with those reported for other similar Cu-bipy complexes. 44,133 Two of the bipy ligands, coordinated trans with respect to one another, bridge the Cd(II) ions to form infinite one-dimensional linear chains that extend along the crystallographic a-axis. A single chain is depicted in Figure A-7. The Cd-Cd distance

PAGE 228

217 A ... B Figure A-5. [Cu(4,4 -bipy)(DMSO)4]n(ClO 4 )2n. A) A typical chain fragment. B) A one dimensional zig-zag chain. All hydrogen atoms have been omitted for clarity. All non hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 229

218 A C Figure A-6. [Cu(4 4 -bipy)(DMS0)4]n(CI04)2nA) The structure within the crystallographic be-plane. B) The structure within the crystallographic ac-plane. C) The structure within the crystallographic ab-plane. For clarity the hydrogen atoms have been omitted

PAGE 230

219 along a chain is a/2 units (11.78 A). From each bridging bipy ligand, the pyridyl rings are not coplanar but twisted along the central C-C bond at an angle of 45.1 with respect to each other. The metal:bipy stoichiometry in 10 is 1 :3 since the bipy ligands perpendicular to the chains are monocoordinate. A twist angle of 29 .4 is observed along the central C-C bond between the pyridyl ligands in these monodentate bipy's as well. The plane defined by Cd(II) and the four pyridyl nitrogen atoms (Cd-Nl-N2-N2' N3) is not coplanar with analogous adjacent planes along the chains but twisted 42.8 with respect to one another. Thus the monocoordinate bipy ligands are not aligned in registry but twisted in an alternating criss-cross fashion along the chains, as depicted by a top view perspective of a chain fragment in Figure A-7. The Cd-bipy chains are juxtaposed in a side-by-side fashion to form two dimensional sheets within the crystallographic ab-plane. A typical sheet is shown in Figure A-8. Within each sheet, the chain are spaced b units apart. In addition to packing forces, the sheets are sustained through a combination of hydrogen bonding interactions between the protons of the coordinated water molecules with the terminal nitrogen atoms from the monodentate bipy ligands on adjacent chains (N-H contacts 2.81 A), and offset .1l"-stacking between the monocoordinate bipy ligands on adjacent chains. The face-to face distance between these overlapping bipy groups from adjacent chains is about 3.7 A. The characteristic packing motif of the Cd-bipy chains produces hydrophobic, rectangular cavities within the sheets. Each cavity is defined by four cadmium ions at the comers and along the sides by the faces of two bridging bipy' s along the chains and the edges of four bipy' s perpendicular to the chains. The dimensions of the rectangles are a/2 x band, if the Van der Waals radii of the carbon atoms from the bipy ligands are

PAGE 231

220 approximated as 1.7 A, the effective size of the cavities is approximately 10.1 Ax 11.oA. As shown in Figure A-8, the sheets pack along the crystallographic c-axis to form a layered solid-state structure. Note the criss cross arrangement of the bipy pairs perpendicular to the chains extending along the a-axis. The sheets are not packed in registry but related by translation along the a-axis and translation along both the and c-axes. This packing motif aligns the hydrophobic cavities within the sheets to form oblique channels that extend along the [4 2 2] direction throughout the solid. The void space within the hydrophobic pores and between the sheets is not empty but occupied by guest molecules and counterions acting to prevent self-inclusion from neighboring sheets. Specifically, two perchlorate anions and two DMSO molecules occupy each rectangular portion of a channel. Each face of the sheet is associated with one perchlorate and one DMSO per Cd(II) ion in the vicinity of the hydrophobic cavities. The perchlorate counterions are positioned near the center of the cavity while the DMSO molecules are positioned near the comers. One DMSO guest is disordered about the S atom. The oxygen atoms of each DMSO guest forms hydrogen bonds with nearby protons from the aqua ligands (1.6 A 1.9 A), not interacting with the bipy nitrogens, and the hydrogen atoms with neighboring monocoordinate bipy ligands (2.4 A 2.5 A). The oxygen atoms from each perchlorate interact with the hydrogen atoms from nearby bridging and monocoordinate bipy ligands (2.4 A 2.5 A) and hydrogen atoms from neighboring DMSO ligands (2.3 A).

PAGE 232

221 "" A B C Figure A-7. [Cd(4,4' -bipy)J(H2O)2](ClO4)2(DMSO). A) The local coordination sphere of Cd(II). B) A one-dimensional zig-zag chain viewed from the top. C) A chain viewed from the side. All hydrogen atoms except the aqua protons have been omitted from the both figures for clarity. All non-hydrogen atoms are represented by thermal ellipsoids drawn to encompass 30 % of electron density.

PAGE 233

A C 222 ~ '---l ..v el. ~ r ..,. ~ Jr' Figure A-8. [Cd(4,4' -bipy)J(H20)2](Cl0 4 )2(DMSO). The structure within the crystallographic ab-plane. B) The structure within the crystallographic be-plane, with the guests shown separately. C) The structure within the crystallographic ac-plane. For clarity, the hydrogen atoms except for water protons have been omitted.

PAGE 234

APPENDIXB TABLES OF ATOMIC COORDINATES AND BOND ANGLES AND DISTANCES Table B-1. Atomic coordinates (x 10 4 ) and equivalent isotropic displacement parameters (A 2 x 10 3 ) for compound 1. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Ni 0 4210(1) 7500 21(1) 01 -1071(2) 4207(2) 7113(1) 27(1) NI I 0 2345(3) 7500 26(1) Nl l' 0 -3877(3) 7500 23(1) Cll 636(2) 1733(3) 7426(2) 33(1) C12 659(2) 528(3) 7423(2) 36(1) C13 0 -107(4) 7500 31(1) Cl4 0 -1411(4) 7500 28(1) Cl5 643(2) -2054(3) 7670(2) 31(1) Cl6 620(2) -3258(3) 7664(2) 29(1) N21 -448(2) 4164(2) 8285(1) 24(1) N21' -1216(2) 4755(3) 11094(1) 41(1) C21 -146(2) 3430(3) 8665(2) 30(1) C22 -320(2) 3463(3) 9207(2) 32(1) C23 -831 (2) 4295(3) 9385(1) 26(1) C24 -1165(2) 5023(4) 8987(2) 35(1) C25 -960(2) 4929(3) 8450(2) 31(1) C26 -985(2) 4432(3) 9975(1) 29(1) C27 -652(2) 3705(4) 10374(2) 36(1) C28 -774(2) 3906(4) 10921(2) 38(1) C29 -1559(3) 5424(5) 10709(2) 53(1) C30 -1457(3) 5302(5) 10154(2) 45(1) N31 -2649(6) 3081(9) 8550(4) 41(2) N31' -2541(6) 2261(11) 11421(4) 48(2) 223

PAGE 235

224 Table B-1. Continued Atom X y z Ueq C31 -2255(7) 2137(11) 8769(5) 38(3) C32 -2230(7) 1928(10) 9328(5) 40(3) C33 -2570(6) 2690(9) 9691(4) 25(2) C34 -2952(7) 3620(13) 9448(5) 43(4) C35 -2977(6) 3760(11) 8918(4) 42(3) C36 -2568(7) 2561(11) 10289(5) 42(3) C37 -2119(10) 1666(16) 10544(7) 66(6) C38 -2151(10) 1554(16) 11158(7) 73(5) C39 -2938(11) 3048(15) 11165(7) 77(6) C40 -2968(9) 3203(12) 10611(6) 57(4) Cll -1640(5) 7980(8) 8699(4) 92(3) 011 -1890(15) 7709(17) 9214(8) 123(10) 012 -1718(10) 9273(15) 8635(8) 80(5) 013 -2110(20) 7530(40) 8387(12) 210(20) 014 -768(10) 7752(14) 8811(10) 81(6) Cl2 -807(13) 8933(17) 9188(7) 125(5) Cl3 -1262(6) 8528(7) 8884(3) 86(2) 031 -390(30) 8390(30) 8814(13) 220(20) 032 -1350(60) 8570(90) 9445(15) 520(80) 033 -1180(20) 9650(20) 8953(18) 203(19) 034 -1934(19) 7360(20) 8257(9) 131(13) Cl4 -323(9) 9430(14) 9215(5) 105(4) N41 1201(10) 11160(18) 9041(8) 120(8) C41 1382(11) 11340(18) 9594(8) 110(9) C42 918(15) 10880(20) 9984(6) 124(11) C43 272(14) 10240(20) 9820(8) 890(190) C44 90(11) 10060(20) 9267(9) 100(8) C45 555(11) 10520(17) 8877(6) 91(7) 02 -2469(19) 3870(20) 7553(15) 73(9) 02' -2401(11) 3780(20) 7508(7) 110(9) 03 -2677(19) 6120(30) 8024(14) 148(10)

PAGE 236

225 Table B-2. Atomic coordinates (x 10 4 ) and equivalent isotropic displacement parameters (A 2 x 10 3 ) for compound 2. Ueq is defined as one third of the trace of the orthogonalized Uijtensor. Atom X y z Ueq Co 0 4224(1) 7500 22(1) 01 -1068(2) 4244(2) 7125(1) 28(1) Nl 1 0 2331(3) 7500 28(1) Nll' 0 -3838(3) 7500 24(1) Cl 1 635(2) 1727(3) 7420(2) 34(1) Cl2 658(2) 529(3) 7416(2) 38(1) CB 0 -99(4) 7500 33(1) Cl4 0 -1394(4) 7500 30(1) Cl5 645(2) -2022(3) 7658(2) 33(1) Cl6 623(2) -3223(3) 7652(1) 29(1) N21 -461(2) 4164(2) 8303(1) 24(1) N21' -1217(2) 4766(3) 11106(1) 41(1) C21 -160(2) 3439(3) 8679(2) 32(1) C22 -332(2) 3476(3) 9223(1) 33(1) C23 -835(2) 4303(3) 9403(1) 27(1) C24 -1169(2) 5022(4) 9009(2) 36(1) C25 -968(2) 4926(4) 8473(2) 34(1) C26 -989(2) 4438(3) 9991(1) 28(1) C27 -651(2) 3723(3) 10385(2) 37(1) C28 -776(3) 3929(4) 10932(2) 39(1) C29 -1560(3) 5432(5) 10725(2) 57(1) C30 -1463(3) 5305(5) 10171(2) 48(1) N31 -2633(6) 3059(9) 8558(4) 44(2) N31' -2575(6) 2283(10) 11416(4) 47(2) C31 -2250(7) 2104(10) 8776(4) 39(2) C32 -2233(6) 1897(8) 9333(4) 38(2) C33 -2577(5) 2693(9) 9688(3) 25(2) C34 -2926(6) 3632(12) 9445(5) 45(3) C35 -2930(7) 3747(12) 8911(5) 49(3)

PAGE 237

226 Table B-2. Continued Atom X y z Ueq C36 -2576(6) 2557(10) 10285(4) 37(3) C37 -2103(8) 1721(15) 10545(6) 61(4) C38 -2147(10) 1610(16) 11148(7) 73(5) C39 -2994(10) 3053(13) 11162(6) 72(4) C40 -3011(8) 3193(11) 10596(5) 58(3) Cll -1616(3) 8005(5) 8717(2) 88(2) 011 -1949(8) 7641(12) 9210(5) 98(4) 012 -1697(8) 9217(10) 8637(6) 88(4) 013 -1952(9) 7453(17) 8269(5) 95(5) 014 -789(7) 7766(10) 8781(7) 78(4) Cl2 -829(13) 8874(15) 9178(7) 118(5) 021 -270(40) 8460(50) 8831(15) 210(30) 022 -1240(50) 8960(100) 9491(16) 360(70) 023 -1080(20) 9690(20) 9001(18) 140(15) 024 -170(30) 8800(30) 9646(13) 163(19) Cl3 -1216(9) 8585(10) 8957(7) 103(3) Cl4 -343(10) 9341(14) 9176(5) 114( 4) N41 1241(8) 11180(13) 9013(6) 121(6) C41 1414(8) 11350(14) 9564(6) 124(8) C42 945(9) 10887(15) 9949(5) 120(8) C43 304(8) 10255(14) 9784(5) 109(6) C44 131(7) 10086(13) 9233(6) 95(5) C45 600(9) 10548(14) 8848(4) 114(7) 02 -2404(12) 3910(60) 7570(20) 81(7) 02' -2427(18) 3690(90) 7490(30) 86(11) 03 -2703(16) 6130(30) 8026(11) 125(8) 03' -2250(20) 7200(30) 8357(15) 121(13)

PAGE 238

227 Table B-3. Atomic coordinates (x 10 4 ) and equivalent isotropic displacement parameters (A 2 x 10 3 ) for compound 3. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Cu 1845(1) 5923(1) 2992(1) 21(1) SI 3582(1) 6193(1) 2464(1) 33(1) Cl 4075(4) 7430(6) 2715(2) 72(2) C2 4237(3) 5081(6) 2527(3) 61(2) S2 58(1) 6173(1) 3442(1) 28(1) S2' -51(6) 5567(9) 2961(4) 45(3) C3 -464(3) 4871(5) 3456(3) 53(2) C4 -585(3) 7070(6) 3072(3) 61(2) 01 3075(2) 5913(3) 2867(1) 33(1) 02 618(2) 5907(3) 3089(1) 38(1) NI 1840(2) 7753(2) 3001(2) 20(1) NI' 1860(3) 14083(2) 3005(2) 22(1) N2 1588(2) 5906(3) 2209(1) 24(1) N2' 780(2) 5868(4) -565(2) 36(1) N3 2116(2) 5935(3) 3777(1) 22(1) N3' 3049(3) 5757(4) 6516(2) 44(1) Cll 2404(2) 8365(4) 3213(2) 25(1) C12 2428(2) 9600(4) 3223(2) 24(1) C13 1844(3) 10250(2) 3024(2) 20(1) C14 1262(2) 9624(4) 2818(2) 23(1) C15 1281(2) 8376(4) 2804(2) 23(1) C16 1853(3) 11582(2) 3021(2) 20(1) C17 1387(2) 12234(4) 3296(2) 25(1) C18 1410(2) 13473(4) 3284(2) 25(1) C19 2306(2) 13451(4) 2734(2) 26(1) C20 2332(2) 12213(4) 2741(2) 25(1) C21 1879(2) 6696(4) 1895(2) 27(1) C22 1733(2) 6718(4) 1357(2) 30(1) C23 1280(2) 5888(4) 1116(2) 25(1)

PAGE 239

228 Table B-3. Continued Atom X y z Ueq C24 982(3) 5069(4) 1436(2) 33(1) C25 1149(3) 5108(4) 1975(2) 32(1) C26 1109(3) 5889(4) 536(2) 27(1) C27 1486(2) 6558(4) 204(2) 34(1) C28 1309(3) 6522(4) -337(2) 37(1) C29 422(3) 5218(5) -242(2) 38(1) C30 558(3) 5204(4) 298(2) 34(1) C31 2676(2) 5308(4) 3980(2) 25(1) C32 2880(2) 5290(4) 4510(2) 28(1) C33 2509(2) 5926(3) 4859(2) 23(1) C34 1932(2) 6590(4) 4642(2) 26(1) C35 1755(2) 6571(4) 4111(2) 25(1) C36 2697(2) 5876(4) 5431(2) 26(1) C37 3056(3) 4901(4) 5657(2) 31(1) C38 3214(3) 4889(5) 6199(2) 39(1) C39 2721(3) 6699(5) 6297(2) 45(1) C40 2526(3) 6805(4) 5762(2) 35(1) N4 40(2) 2178(5) 6813(2) 49(1) N4' -1176(3) 2717(5) 4107(2) 62(2) C41 225(4) 1445(6) 6458(2) 64(2) C42 2(4) 1496(6) 5922(2) 61(2) C43 -452(3) 2386(5) 5742(2) 41(1) C44 -650(3) 3181(7) 6117(2) 67(2) C45 -384(3) 3039(7) 6631(2) 69(2) C46 -707(3) 2499(5) 5179(2) 44(1) C47 -401(3) 1830(6) 4791(2) 55(1) C48 -660(4) 2006(6) 4278(2) 63(2) C49 -1472(3) 3337(6) 4479(2) 59(2) C50 -1253(3) 3263(5) 5006(2) 49(1) N5 263(3) -2253(6) 6719(2) 68(2) N5' 1040(7) -175(11) 4267(5) 63(3)

PAGE 240

229 Table B-3. Continued. Atom X y z Ueq N5" 296(8) -482(14) 4023(6) 66(4) C51 -206(3) -2415(5) 6318(2) 57(1) C52 -113(3) -2100(5) 5820(2) 51(1) C53 508(3) -1560(5) 5693(2) 46(1) C54 1023(4) -1422(6) 6117(2) 63(1) C55 865(4) -1788(6) 6613(3) 68(2) C56 728(5) -1145(7) 5201(3) 36(2) C57 327(5) -1444(8) 4739(4) 48(2) C58 497(7) -969(11) 4274(5) 60(3) C59 1425(5) 63(8) 4654(4) 51(2) C60 1301(5) -392(8) 5148(4) 47(2) C56' 384(7) -1182(9) 5092(4) 35(2) C57' -206(8) -1384(12) 4749(5) 54(3) C58' -222(9) -1036(14) 4222(6) 69(4) C59' 828(14) -400(20) 4276(8) 63(6) C60' 964(7) -613(12) 4868(5) 50(3) Cll -2149(2) -4676(2) 6638(1) 41(1) 011 -1880(3) -3501(5) 6564(3) 62(2) 012 -1712(10) -5445(17) 7006(9) 182(10) 013 -2839(4) -4567(7) 6819(3) 60(2) 014 -2202(4) -5305(6) 6150(2) 70(2) Cl2 -1794(6) -4586(6) 6838(2) 66(2) 021 -1679(13) -5310(20) 7039(7) 51(4) 022 -2650(19) -4920(30) 6800(14) 116(12) 023 -1880(20) -3660(30) 7112(15) 186(13) 024 -1546(18) -4550(30) 6327(13) 165(11) Cl3 -1562(3) -2906(5) 4436(2) 47(1) 031 -1748(12) -2457(16) 4896(5) 92(7) 032 -2048(11) -3575(17) 4123(7) 138(7) 033 -1298(6) -1987(9) 4135(4) 80(4) 034 -1033(6) -3696(12) 4653(7) 93(4)

PAGE 241

230 Table B-3. Continued Atom X y z Ueq Cl4 -1834(4) -3239(5) 4414(2) 45(1) 041 -2438(5) -4018(10) 4386(4) 73(3) 042 -1929(8) -2424(10) 3996(4) 91(5) 043 -1934(10) -2597(18) 4860(7) 97(7) 044 -1241(7) -3935(9) 4321(6) 78(4)

PAGE 242

231 Table B-4. Atomic coordinates (x 10 4 ) and equivalent isotropic displacement parameters (A 2 x 10 3 ) for compound 4. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Cu 7896(1) 6007(1) 45(1) 34(1) NI 6698(3) 7368(2) -391(1) 35(1) N2 9489(3) 4598(2) 422(1) 30(1) N3 9590(3) 4524(2) 869(1) 35(1) N4 9643(5) 4436(2) 1286(1) 60(1) N5 5113(4) 6119(2) 514(1) 48(1) N6 3307(3) 6785(2) 484(1) 31(1) N7 1475(4) 7389(2) 467(1) 48(1) Cl 5310(30) 7176(14) -772(7) 44(4) C2 4485(16) 8226(7) -1068(3) 39(2) C3 4967(12) 9424(5) -919(3) 34(1) C4 6390(17) 9570(7) -486(3) 44(2) C5 7240(20) 8587(10) -224(3) 36(2) Cl' 5490(30) 7070(20) -852(9) 37(3) C2' 4511(19) 7828(9) -1197(4) 36(2) C3' 4985(12) 9079(8) -1134(3) 33(1) C4' 6330(20) 9429(8) -704(4) 42(2) C5' 7080(40) 8505(17) -376(4) 49(3) C6 4038(9) 10564(4) -1207(2) 45(1) C7 1965(8) 10328(5) -1600(2) 43(1) C6' 4070(10) 9992(5) -1525(2) 42(2) CT 2139(14) 10914(8) -1344(3) 59(2) C8 987(5) 11616(3) -1803(1) 62(1) C9 3019(4) 12330(2) -2066(1) 44(1) ClO 3887(5) 11945(2) -2520(1) 50(1) Cll 5797(5) 12570(3) -2758(1) 61(1) C12 6878(6) 13601(3) -2552(1) 67(1) C13 6061(7) 14015(3) -2104(1) 75(1) C14 4146(6) 13378(3) -1861(1) 62(1)

PAGE 243

232 Table B-5. Bond lengths [A] and angles [ 0 ] for compound 4. Bond Distance / Angle Cu-NS 1.9568(18) Cu-Nl 1.9747(17) Cu-N2#1 1.9934(16) Cu-N2 1.9968(15) Nl-C5' 1.243(18) Nl-Cl 1.258(18) Nl-C5 1.416(10) Nl-Cl' 1.42(2) N2-N3 1.212(2) N2-Cu#l 1.9934(16) N3-N4 1.134(2) N5-N6 1.184(2) N6-N7 1.155(2) Cl-C2 1.447(16) C2-C3 1.374(7) C3-C4 1.378(7) C3-C6 1.526(7) C4-C5 1.343(15) Cl'-C2' 1.33(2) C2'-C3' 1.382(9) C3'-C4' 1.394(9) C3'-C6' 1.514(8) C4'-C5' 1.38(2) C6-C7 1.515(7) C7-C8 1.574(5) C6'-C7' 1.506(9) C7'-C8 1.561(7) C8-C9 1.504(4) C9-C14 1.382(4) C9-CI0 1.385(3) Cl l-C12 1.359(4)

PAGE 244

233 Table B-5. Continued Bond Distance / Angle Cl 0-Cl 1 1.377(4) C13-C14 1.391(5) C12-C13 1.372(5) N5-Cu-Nl 96.60(7) N5-Cu-N2#1 164.38(8) Nl-Cu-N2#1 94.20(7) N5-Cu-N2 91.20(7) Nl-Cu-N2 172.19(7) N2#1-Cu-N2 78.08(7) C5'-Nl-Cl 106.1(9) C5'-Nl-C5 16.8(7) Cl-Nl-C5 121.2(8) C5'-Nl-Cl' 108.7(10) Cl-Nl-Cl' 9.0(18) C5-Nl-Cl' 124.8(9) C5'-Nl-Cu 131.7(6) Cl-NI-Cu 122.1(7) C5-Nl-Cu 116.3(4) Cl'-Nl-Cu 118.8(9) N3-N2-Cu#l 126.92(13) N3-N2-Cu 124.37(14) Cu#l-N2-Cu 101.92(7) N4-N3-N2 178.5(2) N6-N5-Cu 126.36(15) N7-N6-N5 176.4(2) Nl-Cl-C2 118.7(12) C3-C2-Cl 121.6(9) C2-C3-C4 116.5(6) C2-C3-C6 123.7(6) C4-C3-C6 119.8(6) C5-C4-C3 121.3(7)

PAGE 245

234 Table B-5. Continued Bond Distance / Angle C4-C5-Nl 120.3(7) C2'-Cl'-Nl 129.0(19) Cl'-C2'-C3' 116 7(12) C2'-C3'-C6' 119.8(7) C4'-C3'-C6' 123.4(7) C5'-C4'-C3' 117.9(9) Nl-C5'-C4' 130.3(9) C7-C6-C3 115.6(4) C6-C7-C8 108.3(4) C7'-C6'-C3' 113.6(5) C6'-C7'-C8 107.8(6) C9-C8-C7' 111.4(3) C9-C8-C7 113.1(2) C7'-C8-C7 34.9(3) C14-C9-C10 117.2(2) C14-C9-C8 121.7(3) C10-C9-C8 121.1 (2) Cl 1-Cl0-C9 121.8(2) C12-Cl 1-C10 120.3(3) Cl 1-C12-C13 119.6(3) C12-Cl3-Cl4 120.1(3) C9-C14-C13 121.0(3) Symmetry transformations used to generate equivalent atoms: #1 -x+2,-y+l,-z

PAGE 246

235 Table B-6. Atomic coordinates (x 1 o4) and equivalent isotropic displacement parameters (A2 x 103) for compound 5. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Cul 12615(1) 12105(1) -367(1) 19(1) Cu2 10997(1) 15636(1) -182(1) 18(1) Nl 12092(2) 10955(4) 516(1) 19(1) N2 13741(2) 11852(4) 288(1) 18(1) N3 13563(2) 13029(4) -1027(1) 24(1) Cl 11213(2) 10406(5) 555(2) 23(1) C2 10943(2) 9465(5) 1154(2) 27(1) C3 11596(2) 9101(5) 1736(2) 27(1) C4 12510(2) 9688(5) 1706(2) 23(1) C5 12732(2) 10600(4) 1088(2) 18(1) C6 13684(2) 11212(4) 965(2) 18(1) C7 14470(2) 11134(5) 1461(2) 22(1) C8 15312(2) 11678(5) 1226(2) 24(1) C9 15369(2) 12292(5) 517(2) 23(1) Cl0 14550(2) 12387(4) 55(2) 18(1) Cl 1 14445(2) 12977(4) -716(2) 20(1) C12 15199(2) 13365(5) -1113(2) 26(1) C13 15035(3) 13752(5) -1846(2) 32(1) C14 14133(3) 13767(6) -2166(2) 40(1) C15 13414(3) 13433(6) -1742(2) 35(1) N4 11421(2) 12944(4) -827(1) 25(1) N5 11165(2) 12931(4) -1464(2) 26(1) N6 10873(3) 12928(7) -2065(2) 56(1) N7 10345(2) 14590(4) 628(1) 22(1) N8 10585(2) 14218(4) 1262(1) 20(1) N9 10788(2) 13836(5) 1855(2) 35(1) NlO 12285(2) 15744(5) 281(2) 29(1) Nll 12507(2) 15360(4) 907(1) 21(1) N12 12765(2) 15001(5) 1499(2) 31(1)

PAGE 247

Table B-6. Continued Atom N13 Nl4 Nl5 X 11207(2) 11908(2) 12574(2) 236 y 17508(4) 18360(4) 19240(4) z -938(2) -960(1) -1004(2) Ueq 31(1) 21(1) 28(1)

PAGE 248

237 Table B-7. Bond lengths [A] and angles [ 0 ] for compound 5. Bond Distance / Angle Cul-N2 1.940(2) Cul-N4 1.943(3) Cul-N3 2.035(3) Cul-NI 2.037(3) Cul-N15#1 2.358(3) Cu2-N10 1.975(3) Cu2-N13 1.977(3) Cu2-N7 1.992(3) Cu2-N7#2 2.040(3) Cu2-N4 2.374(3) Nl-Cl 1.341(4) Nl-C5 1.358(4) N2-C6 1.342(4) N2-C10 1.343(4) N3-Cl 1 1.348(4) N3-C15 1.348(4) Cl-C2 1.383(5) C2-C3 1.383(5) C3-C4 1.395(5) C4-C5 1.380(4) C5-C6 1.485(4) C6-C7 1.388(4) C7-C8 1.390(5) C8-C9 1.390(5) C9-C10 1.393(4) Cl0-Cll 1.477(4) Cl 1-C12 1.402(4) C12-C13 1.379(5) Cl3-Cl4 1.380(5) C14-C15 1.384(5) N4-N5 1.197(4)

PAGE 249

238 Table B-7. Continued Bond Distance / Angle N5-N6 1.147(4) N7-N8 1.217(4) N7-Cu2#2 2.040(3) N8-N9 1.137(4) NlO-Nl 1 1.200(4) Nl 1-N12 1.149(4) N13-N14 1.188(4) N14-N15 1.161(4) N15-Cul#3 2.358(3) N2-Cul-N4 163.23(11) N2-Cul-N3 80.00(11) N4-Cul-N3 105.28(11) N2-Cul-Nl 79.67(10) N4-Cul-Nl 95.05(11) N3-Cul-Nl 159.47(11) N2-Cul-Nl5#1 101.95(10) N4-Cul-N15#1 94.19(11) N3-Cul-N15#1 88.18(11) Nl-Cul-N15#1 93.20(10) NI 0-Cu2-Nl 3 94.22(12) N10-Cu2-N7 100.87(11) N13-Cu2-N7 154.23(13) NI0-Cu2-N7#2 177.09(11) NI 3-Cu2-N7#2 88.65(11) N7-Cu2-N7#2 76.61(11) N10-Cu2-N4 88.16(11) N13-Cu2-N4 97.41(12) N7-Cu2-N4 103.79(11) N7#2-Cu2-N4 91.02(10) Cl-NI-CS 118.8(3) Cl-NI-Cul 126.3(2)

PAGE 250

239 Table B-7. Continued Bond Distance / Angle C5-Nl-Cul 114.69(19) C6-N2-C10 122.1(3) C6-N2-Cul 119.1(2) C10-N2-Cul 118.7(2) Cl 1-N3-C15 118.5(3) Cl l-N3-Cul 113.8(2) C15-N3-Cul 127.1(2) Nl-Cl-C2 122.1(3) Cl-C2-C3 119.1 (3) C2-C3-C4 119.3(3) C5-C4-C3 118.4(3) Nl-C5-C4 122.2(3) Nl-C5-C6 113.5(3) C4-C5-C6 124.2(3) N2-C6-C7 120.3(3) N2-C6-C5 112.8(3) C7-C6-C5 126.9(3) C6-C7-C8 118.1(3) C9-C8-C7 121.2(3) C8-C9-C10 117.7(3) N2-C10-C9 120.5(3) N2-C10-Cl 1 112.4(3) C9-C10-Cl 1 127.1(3) N3-Cl 1-C12 121.8(3) N3-Cl 1-C10 114.8(3) C12-Cl 1-C10 123.3(3) Cl3-C12-Cl 1 119.1(3) C12-Cl3-Cl4 119.0(3) Cl3-Cl4-C15 119.4(3) N3-C15-C14 122.2(3) N5-N4-Cul 127.0(2)

PAGE 251

240 Table B-7. Continued Bond Distance / Angle N5-N4-Cu2 115.3(2) Cul-N4-Cu2 107.11(12) N6-N5-N4 176.5(4) N8-N7-Cu2 133.8(2) N8-N7-Cu2#2 122.7(2) Cu2-N7-Cu2#2 103.39(11) N9-N8-N7 178.0(3) Nl 1-N10-Cu2 123.9(2) N12-Nl l-Nl0 176.5(3) NI 4-Nl 3-Cu2 124.2(2) N15-N14-N13 177.1(3) N14-N15-Cu1#3 114.5(2) Symmetry transformations used to generate equivalent atoms: #1 x,y-1,z #2 -x+2 -y+3,-z #3 x,y+ 1,z

PAGE 252

241 Table B-8. Atomic coordinates (x 104) and equivalent isotropic displacement parameters (A2 x 103) for compound 6. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Cul -1356(1) 2807(1) 7795(1) 19(1) Cu2 4031(1) 3351(1) 6766(1) 21(1) Cu3 5000 5000 5000 20(1) NI 91(5) 2147(3) 6368(2) 19(1) N2 -1149(5) 1305(3) 8176(3) 18(1) N3 -2633(5) 2919(3) 9365(3) 20(1) N4 -2961(6) 4103(3) 7435(3) 25(1) NS -2540(5) 4750(3) 6756(3) 23(1) N6 -2302(6) 5387(3) 6112(3) 34(1) N7 1720(5) 3318(3) 8011(3) 24(1) N8 2076(6) 3111(4) 8865(3) 36(1) N9 2329(7) 2908(6) 9673(4) 80(2) Nl0 5307(6) 1890(3) 6814(3) 33(1) Nll 4623(5) 1215(3) 7402(3) 25(1) Nl2 4059(6) 531(3) 7943(3) 38(1) Nl3 5784(6) 3491(3) 5280(3) 23(1) Nl4 6346(5) 2775(3) 4639(3) 23(1) Nl5 6869(6) 2102(3) 4042(3) 38(1) Nl6 3067(6) 4836(3) 6392(3) 25(1) N17 1796(6) 5496(3) 6950(3) 24(1) N18 594(6) 6111(4) 7466(3) 39(1) Cl 720(6) 2662(4) 5471(3) 21(1) C2 1847(7) 2131(4) 4555(3) 26(1) C3 2383(7) 1048(4) 4581(3) 25(1) C4 1716(6) 512(4) 5509(3) 24(1) cs 542(6) 1073(3) 6390(3) 19(1) C6 -319(6) 585(3) 7419(3) 19(1) C7 -333(7) -496(4) 7636(3) 24(1) C8 -1218(6) -795(4) 8652(3) 25(1)

PAGE 253

242 Table B-8. Continued Atom X y z Ueq C9 -2085(6) -33(4) 9421(3) 24(1) ClO -2022(6) 1030(4) 9151(3) 21(1) Cll -2812(6) 1968(4) 9859(3) 20(1) C12 -3623(7) 1921(4) 10932(3) 29(1) CB -4213(7) 2849(4) 11505(3) 32(1) C14 -3975(7) 3816(4) 11008(3) 30(1) C15 -3166(6) 3816(4) 9923(3) 24(1)

PAGE 254

243 Table B-9. Bond lengths [A] and angles [ 0 ] for compound 6 Bond Distance I Angle Cul-N4 1.950(4) Cul-N2 1.951(4) Cul-N3 2.032(3) Cul-NI 2.034(3) Cul-N7 2.288(3) Cu2-NIO 1.941(4) Cu2-N7 1.956(3) Cu2-Nl6 1.996(4) Cu2-Nl3 2.033(3) Cu3-Nl3 1.966(4) Cu3-Nl3#1 1.966(4) Cu3-Nl6#1 1.987(3) Cu3-Nl6 1.987(3) NI-Cl 1.334(5) NI-CS 1.356(5) N2-Cl0 1.330(5) N2-C6 1.333(5) N3-Cl5 1.329(5) N3-Cl I 1.359(5) N4-N5 1.203(5) N5-N6 1.151(5) N7-N8 1.214(5) N8-N9 1.130(5) Nl0-Nll 1.196(5) NI 1-Nl2 1.148(5) Nl3-N14 1.213(5) N14-Nl5 1.133(5) Nl6-Nl7 1.225(5) NI 7-N18 1.144(5) Cl-C2 1.392(6) Cl-HlA 0.9500

PAGE 255

244 Table B-9. Continued Bond Distance I Angle C2-C3 1.374(6) C2-H2A 0.9500 C3-C4 1.385(6) C3-H3A 0.9500 C4-C5 1.383(6) C4-H4A 0.9500 C5-C6 1.486(5) C6-C7 1.392(6) C7-C8 1.386(6) C7-H7A 0.9500 C8-C9 1.381(6) C8-H8A 0.9500 C9-C10 1.386(6) C9-H9A 0.9500 ClO-Cl 1 1.486(6) Cl 1-C12 1.384(5) C12-Cl3 1.372(6) C12-H12A 0.9500 Cl3-Cl4 1.383(6) Cl3-Hl3A 0.9500 C14-C15 1.396(6) C14-H14A 0.9500 C15-H15A 0.9500 N4-Cul-N2 150.78(15) N4-Cul-N3 95.54(14) N2-Cul-N3 79.45(14) N4-Cul-Nl 103.01(14) N2-Cul-Nl 80.32(14) N3-Cul-Nl 159.69(15) N4-Cul-N7 105.46(15) N2-Cul-N7 103.43(14)

PAGE 256

245 Table B-9. Continued Bond Distance / Angle N3-Cul-N7 91.27(13) Nl-Cul-N7 91.69(13) N10-Cu2-N7 97.93(16) N10-Cu2-N16 167.54(15) N7-Cu2-N16 93.83(15) N10-Cu2-N13 90.18(15) N7-Cu2-N13 164.50(15) N16-Cu2-N13 77.43(14) N13-Cu3-N13#1 180 000(1) Nl 3-Cu3-Nl 6# 1 100.78(15) Nl3#1-Cu3-Nl6#1 79.22(15) N13-Cu3-N16 79.22(15) N13#1-Cu3-N16 100.78(15) N16#1-Cu3-N16 180.000(1) Cl-Nl-C5 119.4(3) Cl-Nl-Cul 126.9(3) C5-Nl-Cul 113.6(3) C10-N2-C6 122.1(4) Cl0-N2-Cul 119.2(3) C6-N2,.Cul 118.3(3) C15-N3-Cl 1 119.7(3) C15-N3-Cul 125.7(3) Cl l-N3-Cul 114.5(3) N5-N4-Cul 131.9(3) N6-N5-N4 174.6(4) N8-N7-Cu2 120.4(3) N8-N7-Cul 114.4(3) Cu2-N7-Cul 117.84(16) N9-N8-N7 177.3(5) Nl l-N10-Cu2 126.4(3) N12-Nl l-N10 176.5(5)

PAGE 257

246 Table B-9. Continued Bond Distance / Angle N14-N13-Cu3 126.8(3) N14-Nl3-Cu2 125.1(3) Cu3-Nl3-Cu2 101.14(16) N15-N14-Nl3 179.7(5) NI 7-N16-Cu3 130.1(3) NI 7-N16-Cu2 127.6(3) Cu3-N16-Cu2 101.68(16) N18-Nl 7-N16 179.4(5) Nl-Cl-C2 121.9(4) Nl-Cl-HIA 119.0 C2-Cl-HIA 119.0 C3-C2-Cl 119.1(4) C3-C2-H2A 120.5 Cl-C2-H2A 120.5 C2-C3-C4 118.9(4) C2-C3-H3A 120.5 C4-C3-H3A 120.5 C5-C4-C3 119.7(4) C5-C4-H4A 120.1 C3-C4-H4A 120.1 Nl-C5-C4 120.9(4) Nl-C5-C6 114.4(3) C4-C5-C6 124.7(4) N2-C6-C7 120.1(4) N2-C6-C5 112.9(4) C7-C6-C5 127.0(4) C8-C7-C6 118.4(4) C8-C7-H7A 120.8 C6-C7-H7A 120.8 C9-C8-C7 120.4(4) C9-C8-H8A 119.8

PAGE 258

247 Table B-9. Continued Bond Distance / Angle C7-C8-H8A 119.8 C8-C9-C10 118.3(4) C8-C9-H9A 120.8 C10-C9-H9A 120.8 N2-C10-C9 120.7(4) N2-C10-Cl 1 112.4( 4) C9-C10-Cl 1 127 0(4) N3-Cl 1-C12 120.9(4) N3-Cl 1-ClO 114.0(3) C12-Cl 1-C10 125.1 ( 4) Cl3-Cl2-Cl 1 119.4(4) Cl3-C12-Hl2A 120.3 Cl 1-C12-Hl2A 120.3 C12-Cl3-Cl4 119.8(4) C12-Cl3-Hl3A 120.1 C14-Cl3-Hl3A 120.1 Cl3-Cl4-C15 118.3(4) C 13-C 14-Hl 4A 120.8 C15-C14-H14A 120.8 N3-C15-C14 121.9(4) N3-C15-Hl5A 119.l C14-C15-Hl5A 119.l Symmetry transformations used to generate equivalent atoms: #1 -x+l,-y+l,-z+l

PAGE 259

248 Table B-10. Atomic coordinates (x 1 o4) and equivalent isotropic displacement parameters (A2x 103) for compound 7. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Nil 2102(3) 2497(1) 1094(3) 24(1) Cl 2733(6) 1395(9) 2016(9) 32(2) C2 2936(8) 2450(9) 9(12) 38(3) C3 1467(6) 3607(9) 112(9) 32(2) C4 1258(6) 2542(7) 2113(9) 19(2) Nl 3132(7) 704(9) 2607(9) 60(3) N2 3507(7) 2471 (6) -505(8) 28(2) N3 1104(6) 4278(9) -448(8) 50(3) N4 731(7) 2586(7) 2773(8) 33(2) Ni2 -395(3) 2538(1) 3596(3) 22(1) 01 -439(7) 4322(3) 3532(13) 76(1) N5 -1161(5) 2151(7) 1937(6) 24(2) N6 -403(7) 801(2) 3595(10) 24(1) N7 390(6) 2165(9) 5221(7) 30(2) C5 -1543(6) 2942(10) 1131 (8) 38(3) C6 -2076(9) 2553(10) 127(10) 40(3) C7 -2246(7) 1468(11) -21(9) 51(3) C8 -1847(6) 616(11) 760(8) 37(2) C9 -1314(6) 957(8) 1764(9) 30(2) ClO -868(5) 253(8) 2649(7) 21(2) Cl 1 -861(6) -1008(9) 2683(9) 35(2) C12 -406(10) -1567(3) 3547(14) 38(1) C13 86(6) -929(10) 4528(8) 32(2) C14 57(5) 208(10) 4448(8) 33(2) C15 535(6) 1064(9) 5410(7) 25(2) C16 1101(6) 598(11) 6373(8) 37(2) C17 1488(7) 1344(9) 7212(9) 35(2) C18 1337(8) 2572(10) 7031(10) 42(3) C19 770(6) 2880(10) 6024(8) 33(2)

PAGE 260

249 Table B-11 Atomic coordinates (x 1 o4) and equivalent isotropic displacement parameters (A2x 103) for compound 8. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Znl 0 0 5000 19(1) Zn2 5000 5000 0 17(1) SI -3616(1) -507(1) 4249(1) 30(1) S2 1849(1) -2671(1) 4249(1) 25(1) S3 6697(1) 7394(1) 838(1) 26(1) S4 7942(1) 2890(1) 591(1) 24(1) 01 -2389(2) 280(1) 4645(1) 25(1) 02 330(2) -2040(1) 4554(1) 27(1) 03 5203(2) 6977(1) 457(1) 23(1) 04 7270(2) 4371(1) 387(1) 26(1) NI 865(2) 1156(2) 4099(1) 22(1) N2 3805(2) 4137(2) 907(1) 19(1) Cl 2323(2) 849(2) 3843(1) 24(1) C2 2909(2) 1459(2) 3232(1) 25(1) C3 1976(2) 2453(2) 2850(1) 21(1) C4 490(2) 2820(2) 3127(1) 23(1) C5 -18(2) 2148(2) 3743(1) 23(1) C6 2566(2) 3055(2) 2175(1) 20(1) C7 3426(2) 2178(2) 1742(1) 23(1) C8 3999(2) 2751(2) 1120(1) 22(1) C9 2946(2) 4975(2) 1315(1) 21(1) CIO 2298(2) 4477(2) 1942(1) 22(1) Cll -3620(3) -2316(2) 4594(1) 42(1) C12 -2741(3) -795(3) 3427(1) 54(1) C13 1959(3) -4482(2) 4577(1) 45(1) C14 1357(3) -2991(3) 3386(1) 46(1) C15 6439(4) 9276(2) 909(2) 62(1) C16 ,, 6347(3) .. 6847(3) 1717(1) 44(1) C17 9901(2) 2856(2) 225(1) 37(1)

PAGE 261

250 Table B-11. Continued Atom X y z Ueq C18 8465(3) 2970(3) 1468(1) 45(1) Cll 10998(1) 8872(1) 1614(1) 38(1) 05 11073(4) 9581(2) 952(1) 95(1) 06 10123(3) 9732(2) 2070(1) 78(1) 07 10251(2) 7502(2) 1577(1) 54(1) 08 12578(2) 8652(2) 1880(1) 58(1) Cl2 6159(1) 14432(1) 3407(1) 36(1) 09 5785(3) 13667(3) 4038(1) 78(1) 010 6277(2) 13464(2) 2870(1) 70(1) 011 4931(2) 15480(3) 3262(1) 72(1) 012 7650(2) 15143(2) 3475(1) 63(1)

PAGE 262

251 Table B-12. Atomic coordinates (x 104) and equivalent isotropic displacement parameters (A2x 103) for compound 9. Ueq is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Cu 4310(1) 4310(1) 5000 17(1) SI 5759(1) 3350(1) 6143(1) 29(1) S2 5984(1) 7555(1) 5279(1) 24(1) 01 5521(2) 3337(2) 5496(1) 27(1) 02 5145(2) 6023(2) 5539(1) 24(1) Nl 2571(2) 3398(2) 5526(1) 19(1) Cl 2175(2) 2118(2) 5781(1) 21(1) C2 1163(2) 1560(2) 6223(1) 21(1) C3 516(2) 2338(2) 6418(1) 18(1) C4 864(2) 3618(2) 6125(1) 20(1) cs 1890(2) 4118(2) 5686(1) 20(1) C6 7598(3) 4702(3) 6264(1) 50(1) C7 5907(4) 1785(3) 6283(1) 45(1) C8 4849(3) 8323(3) 5384(1) 34(1) C9 7325(3) 8570(3) 5814(1) 39(1) 03 1154(3) -3266(2) 5675(1) 50(1) Cl 1398(1) -1865(1) 5523(1) 40(1) 04 102(3) -1815(4) 5576(2) 77(1) 05 2548(6) -801(6) 5876(3) 68(2) 06 1859(4) -1602(5) 4924(2) 86(1) Cl' 1165(9) -1867(8) 5764(4) 41(2) 04' 930(30) -1390(30) 5204(10) 66(6) 05' 2310(60) -800(60) 5950(20) 59(11) 06' 180(20) -1960(20) 6123(8) 51(5)

PAGE 263

252 Table B-13. Atomic coordinates (x 104) and equivalent isotropic displacement parameters (A2x 103) for compound 10. U(eq) is defined as one third of the trace of the orthogonalized Uij tensor. Atom X y z Ueq Cd 3237(1) 7487(1) -719(1) 20(1) 01 3202(1) 8129(3) 869(2) 27(1) 02 3202(1) 6812(2) -2281(2) 27(1) Nl 3275(1) 5772(2) -96(3) 26(1) N2 4228(1) 7546(2) -728(7) 24(1) N3 3207(1) 9194(2) -1346(3) 24(1) N4 3264(2) 335(3) 987(3) 47(1) N5 7229(1) 7572(2) -715(5) 23(1) N6 3115(2) 14620(3) -2463(3) 43(1) Cl 2835(2) 5331(3) 365(3) 30(1) C2 2821(2) 4281(3) 635(3) 32(1) C3 3291(2) 3645(3) 420(3) 28(1) C4 3745(2) 4108(3) -26(3) 30(1) cs 3730(2) 5163(3) -275(3) 31(1) C6 3278(2) 2498(3) 647(5) 34(2) C7 2768(2) 1950(3) 638(3) 36(1) C8 2787(2) 867(4) 821(4) 43(1) C9 3746(2) 871(3) 1000(3) 45(1) ClO 3774(2) 1950(3) 831(3) 37(1) Cll 4524(2) 7738(3) 82(3) 28(1) C12 5111(2) 7748(3) 116(3) 27(1) Cl3 5410(1) 7564(3) -753(6) 22(1) C14 5106(2) 7403(3) -1595(3) 28(1) C15 4519(2) 7381(3) -1564(3) 28(1) C16 6041(1) 7578(3) -757(5) 23(1) C17 6339(2) 7047(3) -11 (3) 30(1) C18 6926(2) 7064(3) -23(3) 29(1) C19 6936(2) 8070(3) -1414(3) 29(1) C20 6348(2) 8096(3) -1452(3) 29(1)

PAGE 264

253 Table B-13. Continued Atom X y z Ueq C21 3659(2) 9820(3) -1247(3) 27(1) C22 3660(2) 10868(3) -1507(3) 31(1) C23 3165(2) 11318(3) -1876(3) 25(1) C24 2701(2) 10668(3) -2009(3) 32(1) C25 2739(2) 9621(3) -1743(3) 31(1) C26 3150(2) 12453(3) -2106(5) 29(1) C27 2637(2) 13012(3) -2042(3) 36(1) C28 2646(2) 14090(3) -2217(4) 41(1) C29 3589(2) 14075(3) -2546(3) 42(1) C30 3630(2) 13003(3) -2378(3) 36(1) Cll 358(1) 646(1) 731 (1) 57(1) Cl2 4831(1) 9548(1) 2718(1) 48(1) 010 138(5) 756(8) 1621 (6) 208(5) 011 198(3) -191 (6) 252(8) 195(5) 012 103(3) 1546(6) 237(7) 146(3) 013 938(2) 809(4) 697(5) 102(2) 020 4936(2) 10514(3) 2197(4) 86(1) 021 4952(2) 9666(5) 3722(3) 85(2) 022 5196(2) 8769(3) 2319(3) 73(1) 023 4253(2) 9235(4) 2568(4) 82(2) S1 4028(1) 6041(1) 2120(1) 38(1) S1' 3958(4) 6384(7) 2831(7) 50(3) S2 3934(1) 8500(1) -4177(1) 52(1) 030 3536(1) 6740(2) 2290(3) 45(1) 040 3405(2) 8286(3) -3613(3) 48(1) C31 3898(3) 4890(4) 2804(5) 64(2) C32 4608(3) 6480(4) 2793(6) 79(2) C41 3951(3) 9897(4) -4358(5) 62(2) C42 3761(3) 8154(6) -5374(5) 78(2)

PAGE 265

LIST OF REFERENCES (1) Hatfield, W. M.; Estes, W E.; Marsh, W. E.; Pickens, M. W.; Harr, L. W. T.; Weller, R. R. In Extended Linear Chain Compounds; Miller, J. S., Ed.; Plenum Press: New York, 1983; pp 43-142. (2) de Jongh, L. J. Magn,etic Properties of Layered Transition Metal Compounds; Kluwer Academic Publishers: Dordrecht, 1990. (3) Rouxel, J. Acc Chem Res 1992, 25,328. (4) Ward, B. H. Electronic and Magn,etic Phenomena of Inorganic and Organic Low Dimensional Solids University of Florida, 1997. (5) Carlin, R. L. Magn,etochemistry; 1st ed.; Springer-Verlag: Berlin, 1986. (6) Kahn, 0. Molecular Magn,etism; VCH: New York, 1993 (7) a) Kahn, O. ; Galy, J.; Journaux, J.; Jaud, J.; Morgenstem-Badarau I. J Am Chem. Soc 1982, 104, 2165-2176. b) Miller, J. S.; Epstein, A. J. MRS Bulletin 2000 II, 21-28. (8) Coronado E.; Tsukerblat B. S.; Gorges R Eds.; Molecular Magn,etism : From Molecular Assemblies to the Devices; Kluwer Academic Publishers: Boston 1996 ; Vol. 321. (9) a) Fanucci G. E. Metal Phosphonate Solids and Langmuir-Blodgett Films : Structural and Physical Chatacterization University of Florida, 1999. b) Boersma, F. ; de Jonge W. J.M.; Kopinga K Phys Rev. B 1981 23, 186-197. (10) Jezowska-Trezbiatowska, B.; Wojciechowski, W. Theory of Bridge Bonding and the Structure of Binuclear Coordination Compounds; 1st ed.; Marcel Dekker: New York, 1971; Vol. 6. (11) Palacio E. Ed.; Molecular Magn,etism : From Molecuar Assemblies to the Devices; Kluwer Academic Publishers: Boston, 1996; Vol. 321. (12) de Jongh, L. J.; Miedema, A. R. Adv. Phys. 1974 24, 1-260. (13) Onsager L. Ph y s Rev 1944, 65 117-149. 254

PAGE 266

255 (14) Pokrovsky, V. L.; Uimin, G. V., Eds.; Magnetic Properties of Layered Transition Metal Compounds; Kluwer Academic Publishers: Dordrecht, 1990, pp 53-103. (15) Block, F. Z. Phys. 1930, 61 206. (16) Boersma, F. L.; de Jongh, W. J. L. Phys. Rev B 1981, 23, 186-214. (17) Bencini, A.; Gatteschi, D. lnorg. Chim. Acta. 1978, 31, 11-18. (18) Crawford, V H.; Richardson, H. W.; Wasson, J. R.; Hodgson, D J.; Hatfield, W. E. lnorg Chem. 1976, 15, 2107-2110. (19) Jeter, D. Y.; Lewis, D. L.; Hempel, J.C.; Hodgson, D. J.; Hatfield, W. E. lnorg. Chem 1972,JJ, 1958-1960 (20) Willet, R. D.; Gatteschi, D.; Kahn, 0. Magneto-Structural Correlations in Exchanged Coupled Systems; Reidel Publishing Company: Dordrecht, 1985 (21) Hay, P. J.; Thibeault, J.C.; Hoffmann, R. J Am. Chem. Soc. 1975, 97, 4884-4899. (22) Michalowicz, A.; Girerd, J. J.; Goulon, J. lnorg Chem. 1979, 18, 3004. (23) Girerd, J. J.; Kahn O.; Verdaguer M. lnorg. Chem 1980 19,274. (24) Meyer, A.; Gleizes, A.; Girerd, J. J.; Verdaguer, M.; Kahn, O. Jnorg. Chem. 1982, 21, 1729. (25) Bonner, J. C.; Fisher, M. E. Phys Rev. A 1964, 135, 640. (26) Orbach, R. L. Phys. Rev 1958, 112, 3 09. (27) Baker, G. A.; Rushbrooke G. S.; Gilbert, H. E. Phys Rev. 1964, 135 A1272. (28) Diederix, K. M.; Blote, H W. J.; Bray, J. W.; Jacobs, I. S J Appl. Phys. 1979, 50, 1810. (29) Bonner, J.C.; Friedberg, S. A.; Kobayashi, H.; Meier, D. L.; Blote, H. W. J. Phys. Rev 1983, B27, 248. (30) Roberts, S. A.; Bloomquist, D. R.; Willet, R. D .; Dodgen, H. W. J. Am. Chem. Soc. 1981, 103, 2603. (31) Harr, L. W T.; Hatfield, W. E. lnorg. Chem. 1985, 24, 1022. (32) Hong, C. S.; Do, Y. Angew Chem. Int. Ed 1999, 38, 193. (33) Rovira, C. Structure and Bonding 2001, JOO, 163-187.

PAGE 267

256 (34) Rovira, C. Chem. Eur. J 2000, 6, 1723. (35) Dagotto, E.; Rice, T. M. Science 1996, 271, 618-623. (36) Azuma, M.; Hiroi, Z.; Takano, M.; Ishida, K.; Kitaoka, Y. Phys. Rev Lett. 1994 73, 3463. (37) Yaghi, O.; Li, H.; Davis, C.; Richardson, D ; Groy, T. L. Acc. Chem. Res. 1998, 32, 474-484. (38) Maekawa, S. Science 1996, 273, 1515. (39) Zaworotk:o, M. J. Chem. Commun. 2001, 1-9. (40) Lu, J. Y.; Lawandy, M.A.; Li, J.; Yuen, T.; Lin, C. L. Inorg. Chem. 1999, 38, 26952704. (41) Carlucci, L.; Ciani, G.; Prosperio, D. M.; Sironi, A. J Am. Chem. Soc. 1995, 117, 4562. (42) Hoskins, B. F.; Robson, R. J Am. Chem Soc 1990, 112, 1546-1554. (43) Robson, R.; Abrahams, B. F.; Batten, S. R.; Gable, R. W.; Hoskins, B. F.; Liu, J. Supramolecular Architecture; American Chemical Society: Washington D. C., 1995. (44) Fujita, M.; Kwon, Y. J.; Washizu, S.; Ogura, K. Inorg. Chem. 1994, 116, 11511152. (45) Yaghi, 0. M.; Li, G.; Li, H. Nature 1995, 378, 703-706. (46) Inoue, M.; Kubo, M. Coord. Chem. Rev. 1976, 21, 1. (47) Kawata, S.; Kitagawa, S.; Kondo, M.; Furuchi, I.; Munakata, M. Angew Chem Int. Ed. Engl. 1994, 33, 1759-1761. (48) Entley, W.R.; Girolami, G. S. Inorg. Chem. 1994, 33, 5165. (49) Manson, J. L.; Campana, C.; Miller, J. S. Chem. Commun. 1998, 251. (50) Manavalan, R.; Pata!, M. M. J Makromol. Sci Chem. 1983, Al 9, 951. (51) Kobel, W.; Hanack, M. Inorg. Chem. 1986, 25, 103. (52) Sinzger, K.; Hunig, S.; Jopp, M.; Bauer, D.; Bietsch, W.; Schutz, J. U. v.; Wolf, H. C.; Kremer, H. C.; Metzenthin, T.; Bau, R.; Khan, S. I.; Lindbaum, A.; Langauer, C. L.; Tillmanns, E. J Am. Chem Soc. 1993, 115, 7696.

PAGE 268

257 (53) Masciocchi, N.; Cairati D ; Carlucci, L. ; Mezza, G. ; Ciani, G.; Sireni A. J Chem. Soc. Dalton Trans 1996 2739. (54) Evans 0. R. ; Xiong R. G .; Wang Z.; Wong G. K. ; Lin W. Angew Chem Int Ed. Eng 1999 38 536. (55) Janiak, C. ; Scharmann T. G. ; Albrecht, P. ; Marlow F. ; MacDonald R. J Am. Chem Soc 1996 118 6307. (56) Lin O ; Evans 0. R. ; Xiong, R G.; Wang, Z. J Am Chem. Soc. 1998 120 13272. (57) Zhang H. ; Wang X.; Teo B. K. J Am. Chem Soc 1996, 118. (58) Lehn J.M Supramolecular Chemistry : Concepts and Perspectives ; VCH : Weinheim 1995 (59) Ward M. D.; Fagan P. J. ; Calabrese, J.C. ; Johnson D. C. J Am. Chem. Soc 1989, 111 1719-1732. (60) Fagan P .; Ward, M. D Sci Am 1992, 7 48. (61) Garcia-Tellado F. ; Geib S. J. ; Goswami S. ; Hamilton A. D. J Am Chem Soc 1991,113 9265-9269 (62) Russell V. A .; Evans C. C. ; Li W .; Ward, M. D Science 1997 273 575-579. (63) Aakeroy C B .; Seddon K. R. Chem. Soc. Rev. 1993, 397-407 (64) Groeneman R. H.; MacGillivray, L. H. ; Atwood, J L. lnorg Chem 1999 38 208209. (65) Melendez R. E .; Sharma, C. V K.; Zaworotko, M. J.; Bauer C.; Rogers R. D. Angew Chem. Int Ed. Engl 1996 32, 2213-2215 (66) Lehn, J. M. Angew Chem. Int. Ed Engl. 1998 27 89. (67) Hunter C. A. Chem. Rev. 1994 101-109. (68) Desiraju G. R .; Gavezzotti A. J Chem Soc. Chem Commun. 1989 621. (69) Hagrman, D.; Hammond, R. P.; Haushalter R.; Zubieta J. Chem. Mater. 1998 JO, 2091-2100 (70) Power K N. ; Hennigar T. L. ; Zaworotko M. J. New J Chem 1998 177-181. (71) Haynes J. S.; Retting, S. J.; Sams, J. R ; Thompson R. C.; Trotter J. Can. J Chem 1987 65 420.

PAGE 269

258 (72) Real, J. A.; DeMunno, G.; Munoz M. C.; Julve, M. lnorg. Chem 1991, 30, 2701. (73) Robinson, F.; Zaworotko, M. J. J Chem Soc Chem. Commun. 1995 2413-2414. (74) Gable, R. W.; Hoskins, B. F.; Robson, R. J Chem Soc ., Chem Commun 1990, 1677. (75) Hagrman, D.; Zubieta, C.; Rose, D. J .; Zubieta, J.; Haushalter, R. C. Angew. Chem. Int. Ed. Engl 1997 36 873-876. (76) Kondo, M.; Yoshitomi, T.; Seki, K.; Matsuzaka, H.; Kitagawa, S. Angew. Chem. Int Ed. Engl. 1997, 36 1725-1727. (77) Yaghi, 0. M.; Li, G. Angew. Chem Int. Ed. Engl. 1995, 34, 207-209. (78) Lu, J. ; Crisci, G.; Niu T.; Jacobson, A. J. lnorg. Chem. 1997, 36, 5140 (79) Yaghi, 0. M.; Li, H. J Am. Chem. Soc. 1995, 117, 10401-10402. (80) Carlucci L.; Ciani, G.; Proserpio, D. M. New. J Chem. 1998, 1998, 1319 (81) Blake, A. J.; Champness, N. R.; Chung, S.S. M.; Li W S.; Schoder, M. Chem. Commun. 1997, 1675-1676 (82) Abrahams, B. F.; Batten S. R. ; Hamit, H .; Hoskins B. F.; Robson R. Angew Chem. Int. Ed. Engl. 1996, 35 1690-1692. (83) Batten, S. R.; Hoskins, B. F.; Robison, R. Angew. Chem. Int. Ed. Engl. 1995, 34, 820-822. (84) Batten, S. R.; Hoskins, B. F.; Robson, R. J. Am. Chem. Soc. 1995, 117 5385-5386. (85) Abrahams, B. F.; Batten, S. R.; Hamit, H.; Hoskins, B. F.; Robson, R. Chem. Commun. 1996 1313-1314. (86) Gardner, G. B. ; Venkataraman, D.; Moore, J. S.; Lee, S. Nature 1995, 374, 792-795. (87) Yaghi, 0. M.; Davis, C. E.; Li G.; Li, H J. Am. Chem. Soc. 1997, 119 2861-2868. (88) Fujita, M.; Kwon, Y. J. ; Sasaki Y. O.; Yamaguchi, K.; Ogura, K. J. Am. Chem. Soc. 1995 117 7287. (89) Wang, Z.; Xiong R. G.; Foxman B. M.; Wilson S. R. ; Lin W. lnorg Chem. 1999 38, 1523-1528. (90) Mayr, A .; Mao L. F. lnorg. Chem. 1998, 37 5776-5780.

PAGE 270

259 (91) Tabellion, F. M.; Seidel, S. R.; Arif, A. M.; Stang, P. J. Angew. Chem. Int. Ed. 2001,40, 1529-1532. (92) Blake, A. J.; Hill, S. J.; Hubberstey, P.; Li, W. S. J Chem. Soc., Dalton Trans. 1997, 913-914. (93) Venkataraman, D.; Gardiner, G. B.; Lee, S.; Moore, J. S. J Am. Chem. Soc. 1995, 117, 11600-11601. (94) Hennigar, T. L.; MacQuarrie, D. C.; Losier, P.; Rogers, R. D.; Zaworotko, M. J. Angew. Chem. Int. Ed. Engl. 1997, 36, 972. (95) Lu, J.; Paliwala, T.; Lim, S. C.; Yu, C.; Niu, T.; Jacobson, A. J. Inorg. Chem. 1997, 36,923. (96) Lu, J.; Yu, C.; Niu, T.; Paliwala, T.; Crisci, G.; Somosa, F.; Jacobson, A. J. lnorg. Chem 1998, 37, 4637-4640. (97) Shen, H.; Liao, D.; Jiang, Z.; Yan, S.; Sun, B.; Wang, G.; Yao, X.; Wang, H. Polyhedron 1998, 17, 1953-1957. (98) Xiong, R. G.; Liu, C. M.; You, X. Z. Polyhedron 1997, 16, 2667-2671. (99) Chen, C.; Xu, D.; Xu, Y.; Cheng, C. Acta. Cryst. Sect. C 1992, C48, 1231. (100) Abrahams, B. F.; Hoskins, B. F.; Winter, G. Aust. J Chem. 1990, 43, 1759. (101) Hagrman, D.; Zubieta, C.; Rose, D. J.; Zubieta, J.; Haushalter, R. C. Angew. Chem. Int. Ed. Engl. 1997, 36, 8. (102) Lopez, S.; Keller, S. W. lnorg. Chem. 1999, 1999. (103) Chen, X. M.; Tong, M. L.; Luo, Y. J.; Chen, Z. N. Aust. J Chem. 1996, 49, 835. (104) Tong, M. L.; Chen, X. L.; Yu, X. L.; Mak, T. C. W. J Chem Soc. Dalton Trans 1998, 5-6. (105) Biradha, K.; Domasevitch, K. V.; Moulton, B.; Seward, C.; Zaworotko, M. J. Chem. Commun. 1999, 1327-1328. (106) Hayashi, M.; Miyamoto, Y.; Inoue, T.; Oguni, N. J Chem. Soc., Chem. Commun. 1992, 1752. (107) MacGillivray, L. R.; Subramanian, S.; Zaworotko, M. J. J Chem. Soc., Chem. Commun 1994, 1325-1326. (108) Carlucci, L.; Giani, G.; Proserpio, D. M.; Sironi, A. J Chem. Soc., Chem. Commun. 1994, 2755-2756.

PAGE 271

260 (109) Losier, P.; Zaworotko, M. J. Angew Chem. Int. Ed. Engl. 1996, 35, 2779-2782. (110) Yaghi, 0. M.; Li, H. J Am. Chem. Soc. 1996, 118, 295-296. (111) Blake, A. J.; Charnpness, N. R.; Khlobystov, A.; Lemenovskii, D. A.; Li, W. S.; Schoder, M. Chem Commun. 1997. (112) Withersby, M.A.; Blake, A. J.; Charnpness, N. R.; Cooke, P.A.; Hubberstey, P.; Schroder, M. New. J Chem. 1999, 23,573. (113) Gudbjartson, H.; Biradha, K.; Poirier, K. M.; Zaworotko, M. J. J Am. Chem Soc. 1999,121,2599-2600. (114) Yahgi, 0. M.; Li, H.; Groy, T. L. lnorg Chem 1997, 36, 4292-4293. (115) Domasevitch, K. V.; Enright, G.D.; Moulton, B.; Zaworotko, M. J. Journal of Solid State Chemistry 2000, 152, 280-285. (116) Soma, T.; Yuge, H.; Iwamoto, T. Angew. Chem. Int. Ed. Engl. 1994, 33, 1655. (117) Subramanian, S.; Zaworotko, M. J. Angew. Chem. Int. Ed. Engl. 1995, 34, 21272129. (118) Buser, H.J.; Schwarzenback, D.; Petter, W. lnorg. Chem. 1977, 16, 2704. (119) Abrahams, B. F.; Hardie, M. J.; Hoskins, B. F.; Robson, R.; Sutherland, E. E. J Chem. Soc., Chem. Commun 1994, 1049. (120) Halasyarnani P.; Heier, K. R.; Willis, M. J.; Stem, C. L.; Poppelmeier, K. R. Z. Anorg. Alig. Chem. 1996, 622,479. (121) Darriet, J.; Haddad, M. S.; Duesler, E. N.; Hendrickson, D. N. lnorg. Chem 1979, 18, 2679. (122) Otieno, T.; Rettig, S. J.; Thompson, R. C.; Trotter, J. lnorg. Chem 1993, 32, 4384. (123) Otieno, T.; Rettig, S. J.; Thompson, R. C.; Trotter, J. lnorg. Chem 1993, 32, 1607. (124) Kitagawa, S.; Kawata, S.; Kondo, M.; Nozaka, Y.; Munakata, M. Bull. Chem. Soc. Jpn. 1993,66,3387. (125) Carlucci, L.; Ciani, G.; Proserpio D. M.; Sironi, A. lnorg. Chem. 1995, 34, 5698. (126) Carlucci, L.; Ciani, G.; Proserpio, D. M.; Sironi, A. Angew. Chem. Int Ed. Engl 1995, 34, 1895. (127) Lu, J.; Harrison, W. T. A.; Jacobson, A. J. Angew Chem. Int. Ed. Engl. 1995, 34, 2557-2559.

PAGE 272

261 (128) Venkataraman D.; Lee, S. ; Moore J. S. ; Zhang, P.; Hirsch, K. A.; Gardner, G. B.; Covey A. C.; Prentice C. L. Chem Mater 1996, 8 2030-2040. (129) Dyer A. An Introduction to Zeolite Molecular Sieves; Wiley: Chichester 1988. (130) Yaghi 0 M.; Sun Z.; Richardson D. A .; Groy T. L. J Am. Chem Soc 1994 116 807-808. (131) Park, K. M. ; Iwamoto, T.J Chem Soc Chem. Commun. 1992, 72-74. (132) Kim, J. ; Whang D.; Lee J. I. ; Kim K. J Chem Soc Chem Commun. 1993 1400-1402. (133) Tong, M. L. ; Ye B. H.; Cai, J. W ; Chen X. M.; Ng, S. W. lnorg Chem 1998 37 2645-2650. (134) Michaelides A. ; Kiritsis V. ; Skoulika S.; Aubry A. Angew. Chem. Int. Ed. Engl. 1993,32 1495-1497. (135) Munukata, M ; Ning, G. L.; Kuroda-Sowa T.; Maekawa, M.; Suenaga, Y. ; Horino, T. lnorg. Chem 1998, 37, 5651-5656. (136) Powell, H. M.; Wetters B. D. P. Chem Ind. 1955, 256-7. (137) Allison S. A. ; Barrer R. M. J Chem Soc. A 1967 1717-1723. (138) MacNicol D. D. Inclusion Compounds; Academic Press: London, 1984; Vol. 2. (139) MacAdie H. G Can. J Chem 1962, 40, 2195-2203. (140) Knight, H.B. ; Witnauer, L. P ; Coleman, J.E.; Noble W.R. J. ; Swem D. Anal Chem. 1952 24 1331-1334. (141) Wang X.; Simard M.; Wuest J. D. J Am Chem. Soc. 1994 116 12119-12120 (142) Endo, K.; Sawaki, T.; Koyanagi, M.; Kobayashi K.; Masuda, H. ; Aoyama, Y. J Am Chem. Soc 1995 117 8341-8352. (143) Zaworotko M. J. Chem Soc Rev. 1994 283. (144) Kondo M. ; Shimamura M ; Noro, S. I.; Minakoshi S.; Asami A.; Seki K.; Kitagawa, S. Chem Mater. 2000 12 1288-1299. (145) Kondo, M. ; Okubo T.; Asami A.; Noro, S.; Kitagawa S.; Ishii T ; Matsuzaka, M.; Seki, K. Angew Chem Int. Ed. Engl. 1999, 38, 140-143. (146) Li, H. ; Eddaoudi, M.; O'Keeffe, M .; Yaghi, 0. M. Nature 1999 402 276-279

PAGE 273

262 (147) Pan, L.; Ching, N.; Huang, X.; Li, J. Inorg. Chem. 2000 39, 5333-5340. (148) Hillier, A. C.; Ward, M. D. Science 1994, 263, 1261. (149) Russel, V. A.; Etter, M. C.; Ward, M. D. J Am Chem Soc. 1994, 116, 1941. (150) Chen, X. M. ; Tong M. L.; Luo Y. J.; Chen, Z. N. Aust. J Chem 1996 49,835. (151) MacGillivray, L. R.; Groeneman, R.H.; Atwood, J. L. J Am. Chem. Soc 1998, 120, 2676-2677 (152) Decurtins, S.; Schmalle, H. W.; Pellaux, R.; Schneuwly, P.; Hauser A. Inorg. Chem 1996, 35, 1451-1460. (153) Kiritsis, V.; Michaelides A.; Skoulika, S.; Golhen, S.; Ouahab L. Inorg. Chem 1998, 37, 3407-3410. (154) McCleverty, J. A.; Ward, M. D .A cc Chem Res. 1998, 31, 842-851. (155) Sheldrick, G. M.; 6 ed.; Bruker-AXS: Madison, Wisconsin, USA 2000. (156) Woodward, J. D.; Backov R.; Abboud, K. A.; Talham, D.R. Unpublished Results. 1998 (157) Abragam, A. ; Bleaney B. Electron Paramagnetic Resonance of Transition Ions ; Oxford University Press: London, 1970. (158) Papanicolaou, N .; Spathis, P. N. Physical Review B 1995, 52, 16001-16011. (159) Orendac, M.; Orendacova, A.; Cemak, J.; Feher, A.; Signore, P. J.C.; Meisel, M. W.; Merah, S.; Verdaguer, M. Physical Review B 1995, 52, 3435-3440. I (160) Orendac, M. ; Zvyagin, S. ; Orendacova, A.; Sieling, M .; Luthi, B.; Feher, A.; Meisel, M. W. Physical Review B 1999, 60, 4170-4175. (161) Orendac, M.; Cizmar, E.; Orendac, M.; Cemak, J.; Feher, A.; Meisel, M. W.; Abboud, K. A.; Zvyagin S ; Sieling, M.; Rieth, T.; Luthi, B. Phys Rev. B 2000 61, 3323. (162) Haddad, M. S.; Hendrickson, D. N.; Cannady, J.P.; Drago, R. S.; Bieksza, D S. J Am. Chem. Soc. 1979, 101,898. (163) Julve, M.; Verdauger M.; Faus, J.; Tinti, F.; Moratal, J.; Monge A.; Guiterrez Puebla, E. Inorg Chem. 1987 26, 3520. (164) Li, M. X.; Xie, G. Y.; Gu Y. D. Polyhedron 1995, 14, 1235-1239. (165) Osegovic, J. The Prospect of Using Triethylphosphine Oxide to Observe Intercalation Compounds 1999; pp 1-4.

PAGE 274

263 (166) Baltusis, L.; Frye, J. S.; Maciel, G. E. J Am Chem. Soc. 1987, 109, 40-46. (167) Biaglo, A. I.; Gorte, R. J.; Kokotailo, G. T.; White, D. Journal of Catalysis 1994, 148. (168) Biaglow, A. I.; Gorte, R. J.; White, D. Journal of Catalysis 1994, 150, 221-224. (169) Biaglow, A. I.; Sepa, J.; Gorte, R. J.; White, D. Journal of Catalysis 1995, 151, 373-384. (170) Fletcher, W. P.; Gilbert, C. S.; Biaglow, A. I. Catalysis Letters 1997, 47. (171) Farcasiu, D.; Ghenciu, A. J Org. Chem. 1991, 56, 6050. (172) Farcasiu, D.; Ghenciu, A. Journal of Catalysis 1992, 134, 126. (173) Farcasiu, D.; Ghenciu, A. J Am. Chem. Soc. 1993, 115, 10901. (174) Xu, T.; Munson, E. J.; Haw, J. F. J Am. Chem. Soc. 1994, 116, 1962. (175) Joerg, S. D. Extensions of the Electrostatic-Covalent and Unified Salvation Models. University of Florida, 1998. (176) Ripmeester, J. A. J Am. Chem. Soc 1983, 105, 2925-2927. (177) Maciel, G. E.; Haw, J. F.; Chuang, I. S.; Hawkins, B. L.; Early, T. E.; McKay, D. R.; Petrakis, L. J Am. Chem Soc. 1983, 105, 5529-5535. (178) Haw, J. F.; Chuang, I. S.; Hawkins, B. L.; Maciel, G. E. J Am Chem. Soc. 1983, 105, 72067207. (179) Majors, P. D.; Ellis, P. D. J Am. Chem. Soc. 1987, 109, 1648-1653. (180) Xu, T.; Koh, N.; Drago, R. S.; Nicholas, J.B.; Haw, J. F. J Am Chem Soc. 1997, 119, 12231-12239. (181) Rakiewicz, E. F.; Peters, A. W.; Wormsbecher, R. F.; Sutovich, K. J.; Mueller, K. T. J Phys Chem. B 1998, 102, 2890-2896. (182) Zalewski, D. J.; Chu, P. J.; Tutunjian, P. N.; Lunsford, J. H. Langmuir 1989, 5, 1026-1030. (183) Osegovic, J.P.; Drago, R. S. Journal of Catalysis 1999, 182, 1-4. (184) Majors, P. D.; Raidy, T. E.; Ellis, P. D.; Inners, R.R. J Am. Chem. Soc. 1986, 108, 8123-8129.

PAGE 275

264 (185) Sutovich, K. J.; Peters, A. W.; Rakiewicz, E. F.; Wormsbecher, R. F.; Mattingly, S. M.; Mueller, K. T. Journal of Catalysis 1999, 182, 155-158. (186) Osegovic, J. New Methods for Characterizing Solid Acidity. University of Florida, 1999. (187) Lunsford, J. H.; Rothwell, W. P.; Shen, W. X. J Am Chem. Soc. 1985, 107, 15401547. (188) Baltusis, L.; Frye, J. S.; Maciel, G. S. J Am. Chem Soc 1986, 108, 7119-7120. (189) Sheng, T. C.; Gay, I. D. Journal of Catalysis 1994, 145. (190) Sang, H.; Chu, H. Y.; Lunsford, J. H. Catalysis Letters 1994, 26. (191) Kolling, 0. W. Transactions of the Kansas Academy of Science 1984, 84, 115-118. (192) Riddle, F. L.; Fowkes, F. M. Polymer Preprents 1988, 29, 188-189. (193) Lunsford, J. H.; Tutunjian, P. N.; Chu, P. J.; Zalewski, D. J. J Phys Chem 1989, 93. (194) Rothwell, W. P.; Shen, W. X.; Lunsford, J. H. J Am. Chem. Soc. 1983, 106, 2452. (195) Lunsford, J. Top Cata/. 1997, 4, 91-98. (196) Meyer, U.; Gutmann, V.; Gerger, W. Monatsh. Chem 1975, 106. (197) Riddle, F. L.; Fowkes, F. M. J Am. Chem. Soc. 1990, 112, 3259-3264 (198) Pearson, R. M. Journal of Catalysis 1971, 23,288. (199) Bertini, I. NMR of Paramagnetic Molecules in Biological Systems; Banjamin/Cummings Publishing Company: Menlo Park, California, 1986. (200) Rojo, T.; Larramendi, J. I. R.; Duran, I.; Mesa, J. L.; Via, J.; Arriortua, M. I. Polyhedron 1990, 9, 2693. (201) Gober, M.A. S.; Wang, R. J.; Mark, T. C. W. Polyhedron 1992, 11,829. (202) Real, J. A.; Ruiz, R.; Faus, J.; Lloret, F.; Julve, M.; Jounaux, Y.; Philoche Levisalles, M.; Bois, C. J Chem. Soc., Dalton Trans 1994, 3769. (203) Mautner, F. A.; Gober, M.A. S. Polyhedron 1996, 15, 1133. (204) Drew, M. G. B.; Nelson, J.; Esho, F.; McKee, V.; Nelson, S. M. J Chem. Soc., Dalton Trans 1982, 1837.

PAGE 276

265 (205) Sikorav, S. ; Bkouche-Waksman, I.; Kahn, O. lnorg. Chem. 1984, 23, 490-495. (206) Goher, M. A. S.; Mak, T. C. W. Inorg Chim Acta 1984, 89, 119. (207) Baily, N. A. ; Fennon, D. E.; Mclean, C.R. lnorg. Chim Acta. 1991, 179, 1. (208) Mautner, F. A.; Gober, M.A. S. Polyhedron 1993, 12, 2823-2829. (209) Goher, M.A. S.; Al-Salem, N. A.; Mautner, F. A. Polyhedron 1996 15, 4513. (210) Felthouse, T R.; Hendrickson, D. N. lnorg. Chem 1978, 17,444. (211) Bkouche-Waksman, I.; Siorav, S.; Khan, 0. J Cryst. Spectrosc. Res. 1983, 13, 303. (212) Angus, Y.; Louis, R.; Gisselbrecht, J.P.; Weiss, R. J Am. Chem. Soc 1984, 106, 93. (213) Bkouche-Waksman, I.; Boillot, M. L.; Khan O.; Sikorav, S. lnorg. Chem. 1984, 23, 4454. (214) McKee, V.; Dagdigian, J. V.; Bau, R.; Reed, C. A. J Am Chem Soc 1981, 103, 7000. (215) Sorrell, T. N.; O'Connor, C. J.; Huchinson J.P.; Zubieta, J. J Am. Chem Soc 1985, 107 4199. (216) Mukherjee, P. S.; Maji, T. K.; Mostafa, G.; Mallah, T.; Chaudhuri, N. R. lnorg. Chem. 2000, 39, 5147-5150. (217) Mukherjee, P. S. ; Dalai, S.; Mostafa, G.; Lu, T. H.; Rentschler, E.; Chaudhuri, N. R. New J Chem. 2001, 25, 1203. (218) Ribas, J.; Monfort, M.; Diaz, C.; Bastos, C.; Solans, X. lnorg Chem. 1993 32 3557. (219) Ribas, J.; Monfort, M.; Costa, R. ; Solans, X. lnorg. Chem. 1993 32, 695. (220) Escuer, A.; Vicente, R.; Ribas, J.; Fallah, M. S. E.; Solans, X.; Font-Bardia, M. lnorg. Chem. 1994, 33, 1842 (221) Vicente, R.; Escuer, A.; Ribas, J.; Fallah, M. S. E.; Solans X.; Font-Bardia, M. lnorg Chem 1995, 34, 1278. (222) Ribas, J.; Monfort M.; Ghosh, B. K.; Cortes, R.; Solans, X.; Font-Bardia, M. lnorg. Chem. 1996, 35, 864. (223) Monfort M.; Ribas, J.; Solans, X. J Chem Soc ., Chem. Commun 1993, 350.

PAGE 277

266 (224) Monfort, M.; Resino, I.; Ribas, J.; Stoeckli-Evans, H. Angew Chem. Int. Ed 2000, 39. (225) Mautner, F. A.; Cortes, R.; Lezema, L.; Rojo, T. Angew. Chem Int Ed Engl. 1996, 35, 96. (226) Goher, M.A. S.; Mak, T. C. W. lnorg. Chim. Acta 1984, 85, 117. (227) Mak, T. C. W.; Goher, M.A. S.; Mautner, F. A. lnorg Chim Acta 1986, 115, 17. (228) Goher, M.A. S.; Mautner, F. A. Naturforsch 1991, 46(b), 687. (229) Goher, M.A. S.; Mautner, F. A.; Al-Salem, N. A. Polyhedron 1997, 16, 2239. (230) Halcrow, M. C.; Sun, J. S.; Huffman, J.C.; Christou, G. lnorg. Chem. 1995, 34, 4167. (231) Maji, T. K.; Mukherjee, P. S.; Koner, S.; Mostafa, G ; Tuchagues, J.P.; Chaudhuri, N. R. lnorg. Chim. Acta 2001, 314, 111-116. (232) Goher, M.A. S.; Mak, T. C. W. lnorg. Chim Acta 1985, 99,223. (233) Munno, G.D.; Real, J. A.; Julve, M.; Munoz, C. M. lnorg Chim. Acta 1993, 211, 227. (234) Goher, M.A. S.; Wang, R. J.; Mak, T. C. W. J Mal. Struct. 1991, 243, 179. (235) Escuer, A.; Font-Bardia, M.; Penalba, E.; Solans, X.; Vicente, R. Polyhedron 1998, 18,211. (236) Thompson, L. K.; Tandon, S.S.; Lloret, F.; Cano, J.; Julve, M. lnorg Chem. 1997, 36, 3301-3306. (237) Chaudhuri, P ; Guttmann, M.; Ventur, D.; Weighardt, K.; Nuber, B.; Weiss, J. J Chem Soc., Chem. Commun 1985. (238) Tandon, S.S.; Thompson, L. K.; Manuel, M. E.; Bridson, J. N. lnorg. Chem. 1994, 33, 5555-5570. (239) Mak, T. C. W ; Goher, M.A. S. lnorg. Chim Acta 1986, 115, 17. (240) Escuer, A.; Vicente, R.; Ribas, J. J Magn. Magn. Mater 1992, 110, 181. (241) Arriortua, M. I.; Cortes, A. R.; Lezema, L.; Rojo, T.; Solans, X. lnorg. Chim Acta 1990, 174,263. (242) Cortes, R.; Lezama, L.; Pizarro, J. L.; Arriortua, M. I.; Solans, X.; Rojo, T. Angew. Chem. Int. Ed Engl. 1994, 33, 2488-2489.

PAGE 278

267 (243) Ribas, J.; Monfort, M.; Solans, X.; Drillon, M. lnorg. Chem. 1994, 33, 742. (244) Tandon, S.S.; Thompson, L. K.; Miller, D. 0. Chem. Commun 1995, 1907. (245) Mautner, F. A.; Goher, M.A. S. Polyhedron 1992, 11, 2537. (246) Kahn, O. Jnorg. Chim. Acta. 1982, 62, 3-14. (247) Cortes, R.; Pizarro, J. L.; Lezama, L.; Arriortua, M. I.; Rojo, T. Inorg. Chem 1994, 33, 2697. (248) Han, S.; Manson, J. L.; Kim, J.; Miller, J. S. lnorg. Chem. 2000, 39, 4189. (249) Barandika, M. G.; Cortes, R.; Lezama, L.; Urtiaga, M. K.; Arriortua, M. I.; Rojo, T. J Chem. Soc. Dalton Trans. 1999, 2971. (250) Comannond, J.; Plumere, P.; Lehn, J.M.; Angus, Y.; Louis, R.; Weiss, R.; Khan, O.; Morgesten-Badarau, I. J Am. Chem. Soc 1982, 104, 6330. (251) Kahn, O.; Sikorav, S.; Gouteron, J.; Jeannin, S.; Jeannin, Y. lnorg. Chem 1983, 22, 2877-2883. (252) Pierpont, C. G.; Hendrickson, D. N.; Duggan, D. M.; Wagner, F.; Barefield, E. K. Inorg. Chem. 1975, 14, 604-610. (253) Escuer, A.; Vicente, R.; Ribas, J.; Fallah, M. S. E.; Solans, X. lnorg. Chem. 1993, 32, 1033. (254) Angus, Y.; Louis, R.; Weiss, R. J Am. Chem. Soc. 1979, 101, 3381. (255) Mukherjee, P. S.; Maji, T. K.; Escuer, A.; Vincente, R.; Ribas, J.; Rosair, G.; Mautner, F. A.; Chadhuri, N. R. Eur. J Inorg. Chem 2002, 943-949. (256) Thompson, L. K.; Tandon, S. S. lnorg Chem 1995, 34, 2356-2366. (257) Wyatt, J. F.; Hillier, I. H.; Saunders, V. R.; Connor, J. A.; Barber, M. J Chem. Phys. 1971, 54, 5311. (258) Goher, M.A. S.; Escuer, A.; Mautner, F. A.; Al-Salem, N. A. Polyhedron 2001, 20, 2971-2977. (259) Rubenacker G. V.; Drumheller, J.E.; Emerson, K.; Willett, R. D. Journal of Magnetism and Magnetic Materials 1986, 54-57, 1483-1484. (260) Woodward, J. D.; Backov, R.; Meisel, M. W.; Tatham, D.R. Unpublished Results. 2000.

PAGE 279

268 (261) Grigereit, T. E.; Ramakrishna, B. L.; Place, H.; Willett, R. D.; Pellacani, G. C.; Manfredini, T.; Menabue, L.; Bonamartini-Corradi, A.; Battaglia, L. P. Inorg. Chem. 1984,26,2235-2243. (262) Baker, A. T. ; Craig, D. C.; Rae, A. D. Aust. J Chem 1995, 48, 1373-1378. (263) Cortes, R.; Arriortua, M. I.; Rojo, T.; Solans, X.; Beltran, D. Polyhedron 1986, 5, 1987-1990. (264) Cortes, R.; Arriortua, M. I.; Rojo, T.; Solans, X.; Miravitlles, C.; Beltran, D. Acta Cryst. 1985, C41, 1733-1736. (265) Iwamoto, T.; Miyoshi, T.; Sasaki, Y. Acta. Cryst. 1974, B30, 292-295. (266) Cemak, J.; Chomic, J ; Domiano P.; Ori O.; Andreett, G.D. Acta. Cryst. 1991 C46, 2103-2107. (267) Yuge, H.; Iwamoto, T. J Chem Soc ., Dalton Trans. 1994, 1237-1242. (268) Cemak, J.; Orendac, M.; Potocnak, I.; Chomic, J.; Orendacova, A.; Skorsepa, J.; Feher, A. Coordination Chemistry Reviews 2002 224, 51-66 (269) Sen, S.; Mitra, S.; Kundu, P.; Saha, M. K.; Kruger, C.; Bruckmann, J. Polyhedron 1997, 16 2475-2481. (270) Hay, P. J.; Thibeault, J.C.; Hoffmann, R. J Am Chem Soc 1975, 97, 4884-4899. (271) Watson, B. C. Quantum Transitions in Antiferromagnets and Liquid Helium-3 University of Florida, 2000.

PAGE 280

BIOGRAPHICAL SKETCH Jonathan David Woodward was born in Birmingham, Alabama on April 22, 1974. He has lived in Birmingham for most of his life where, in 1992, he graduated from Pelham High School, Pelham, Alabama. In 1996, he graduated cum laude from the University of Montevallo, Montevallo, Alabama, with a Bachelor of Science in Chemistry. It wasn't until late in his junior year of college that he realized that he wanted to go to graduate school and that a career in medicine was really feasible. Later that same year, he began graduate school at the University of Florida (Gainesville, Florida) to work toward a Ph.D. in Chemistry under the supervision of Dr. Daniel R. Talham. Despite living in Gainesville for the better part of 5 years, he still remains a "gator -hater ." Upon completing his doctorate, Jonathan's immediate plans include a postdoctoral fellowship at the University of Tennessee. Eventually he plans to pursue an industrial research career. 269

PAGE 281

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy DfilU~~IT Professor of Chemistry I certify -that I have read this -study and-that in my opinion it confo1 ms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Phi~loso _....__.tJuzi~ avidifiiichardson Professor of Chemistry I certify-that I have Tead this -study and-that in my opinion it confonns to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy ~~ Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify "that I have read this -study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy Mark W. Meisel Professor of Physics