134

 

rHESls

lllllllllllllHillllllllllllllllllllllllllll

3 1293 01834

 

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

thesis entitled

Structure and Magnetism for Alkali Meta] Salts
of Nitrogen Heterocyciic Radical Anions

presented by
Eric Keith Meyer

has been accepted towards fulfillment
of the requirements for

M.S. degree in Chemi strv

 

V (74.

Majorp/ ofessor

Date 2'3 ’A‘fifi“5t 4:77—41

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

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

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

we chlHC/DdoDmpfiS-p.“

 

STRUCTURE AND MAGNETISM FOR ALKALI METAL SALTS OF NITROGEN
HETEROCYCLIC RADICAL ANIONS
By

Eric Keith Meyer

A THESIS
Submitted to
Michigan State University

in partial fulfillment of the requirements
for the degree of

MASTERS OF SCIENCE
Department of Chemistry

1 999

ABSTRACT
STRUCTURE AND MAGNETISM FOR ALKALI METAL SALTS OF NITROGEN
HETEROCYCLES
By

Eric Keith Meyer

In this thesis we report the magnetic and structural properties of air sensitive
radical anion salts Rb*(2,3~bis~(3-methoxyphenyl) quinoxaline)‘ and
M+(dibenzo[a,c]phenazine)' (M = Na, K, Rb). ESR and optical Spectroscopy of the
complexes in solution revealed insights into ion pairing interactions and Spin
delocalization. The relationships between structure and magnetism of these salts in the
solid state were probed by single crystal X-ray analysis and SQUID. The size of the
metal cation dramatically influenced the structural motif of the crystal lattice and the
strength of the magnetic coupling between the paramagnetic centers. The larger alkali
metals were less solvated in the crystal and provide for greater dimensionality. In fact the
Rb+(Dibenzo[a,c]phenazine)' salt crystallized as a solvent free two dimensional network.
Although the desired ferromagnetic coupling was not attained in these systems, trends

between the strength of the magnetic interactions and the size of the metal were observed.

To My Parents

ACKNOWLEDGMENTS

I would like to thank Dr. Jackson for his assistance and support throughout this
research project. I also want to thank members of the Jackson group for their help. I
would also like the thank Dr. Andrew Ichimura and Dr. Rui Huang for the assistance they
gave me with my research.

I give a special thanks to my parents for their love and encouragement. They

motivated me throughout my school years and enforced the value of a good education.

iv

TABLE OF CONTENTS

Chapter 1 Background ............................................................................... 1
1.1 Introduction ............................................................................. 2
1.2 Magnetic Exchange ..................................................................... 3
1.3 F ullerides ............................................................................... 13
1.4 Ion Pairing ............................................................................... 14

Chapter 2 Characterization of the Radical Anion Salts in Solution .......................... 20
2.1 Introduction ............................................................................ 21
2.2 EPR Spectra of the Radical Anions ................................................ 21
2.3 Optical Spectra of the Radical Anions .............................................. 24

Chapter 3 Characterization of the Radical Anion Salts in the Solid State .................. 27
3.1 Introduction ............................................................................ 28

3.2 X-ray Structure and Magnetism of the Rb*(2,3 bis-

(3-rnethoxyphenyl) quinoxaline)‘ (THF) salt ....................................... 28
3.3 X-ray Structure and magnetism of the Na+(dibenzo[a,c]phenazine)'

(MTI-IF)2 salt ........................................................................... 32
3.4 X-ray Structure and magnetism of the K+(dibenzo[a,c]phenazine)'

(MTHF)2,3 salt .......................................................................... 37
3.5 X-ray Structure of the Rb+(dibenzo[a,c]phenazine)' (MTHF)2,3 ................. 43
3.6 Summary ................................................................................ 45

Chapter 4 Experimental ........................................................................... 48

4.1 Ligand Synthesis and Purification .................................................. 49
4.2 Solvent Purification .................................................................. 49
4.3 Synthesis of the Radical Anion Salts .............................................. 50
4.4 Crystallization of the Radical Anion Salts ......................................... 51
4.5 X-ray analysis ......................................................................... 51
4.6 EPR and ENDOR Analysis ......................................................... 52
4.7 Magnetic Susceptibility Analysis ................................................... 52
Appendix ............................................................................................. 53
References ............................................................................................ 62

vi

List of Tables

Table 2.1 Maximum visible absorbtions for the radical anion salts in THF @

-10°C ..................................................................................... 25
Table 3.1 Selected Distances and Angles for Rb+(2,3 bis-(3-methoxy-phenyl)-

quinoxaline)‘ (THF) Salt. Distances are given in A ............................... 30
Table 3.2 Selected Distances and Angles for Na+(Dibenzo[a,c]phenazine)'

(2-MTHF)2. Distances are given in A ............................................... 35
Table 3.3 Selected Distances and Angles for K+(Dibenzo[a,c]phenazine)'

(MTHF)2,3 Distances are given in A .................................................. 40
Table 3.4 Selected Distances and Angles for Rb+(Dibenzo[a,c]phenazine)'

Distances are given in A ............................................................... 45
Table 1A Unit cell parameters and refinement ................................................. 54

Table 2A Atomic Coordinates and Isotropic Displacement Parameters
for Rb+ (2,3 bis-(3 -methoxyphenyl) quinoxaline)’ ................................. 55

Table 3A Atomic Coordinates and Isotropic Displacement Parameters for
Na*(Dibenzo[a,c]phenazine)' (MTI-IF)2 ............................................. 56

Table 4A Atomic Coordinates and Isotropic Displacement Parameters for
K+(Dibenzo[a,c]phenazine)' (MTI-IF)2,3 ............................................ 57

Table 5A Atomic Coordinates and Isotropic Displacement Parameters
for Rb*(Dibenzo[a,c]phenazine)' ...................................................... 61

vii

LIST OF FIGURES

Figure 1.1 Orbital mixing diagram for 5 .......................................................... 6
Figure 1.2 Spin distribution of [2.2]paracyclophanes incorporating diphenylcarbene

units ....................................................................................... 8
Figure 1.3 The structure of superconductive K3C60. ........................................... 14

Figure 1.4 Hyperfine splittings in gauss for the quinoxaline, phenazine, and
dibenzo[a,c]phenazine radical anions produced electrochemically. . . . . . . . . 1 9

Figure 2.1 EPR Spectra of Na*(Dibenzo[a,c]phenazine)' in THF recorded at -5°C ....... 22

Figure 2.2 (A) EPR spectra of K+(Dibenzo[a,c]phenazine)' in THF recorded at
-96.7°C. (B) ENDOR spectrum of K*(Dibenzo[a,c]phenazine)' recorded
at -96.7°C ............................................................................... 23

Figure 2.3 Hyperfine coupling constants (in Gauss) from the ENDOR Spectrum of
K*(dibenzo[a,c]phenazine)' ........................................................... 24

Figure 2.6 Optical spectra for (a) Rb+ (2,3 bis-(3-methoxyphenyl) quinoxaline)’
(B)Na*(dibenzo[a,c]phenazine)‘, and K+(dibenzo[a,c]phenazine)'
(C) Rb+(dibenzo[a,c]phenazine)‘ with and without the presence of C22
cryptand. All optical spectra were done with THF solutions of the
radical anion @ -10°C ................................................................. 25

Figure 3.1 Single crystal X-ray structure of Rb+(2,3bis-(3-methoxyphenyl)-
quinoxaline)’ (THF) with atom labeling using atomic symbols. (A) Part
of a single chain is shown. (B) Crystal packing for the unit cell looking
down the chain axis .................................................................... 29

Figure 3.2 Plot of molar susceptibility x vs temperature and act vs temperature for a
powder sample of the Rb+(2,3 bis-(3-methoxyphenyl) quinoxaline)’ Salt.
The solid line represents a fit to the Curie Weiss Equation for 1-D
magnetic coupling ...................................................................... 32

Figure 3.3 Single crystal X-ray structure of Na*(Dibenzo[a,c]phenazine)’ (MTHF)2
with atom labeling using atomic symbols. (A) A Single dimer is
displayed. (B) Crystal packing for the unit cell with the shortest
distance between dimers depicted by a dashed line. Only seven of twelve
dimers in the unit cell are depicted in this figure for clarity ..................... 33

viii

Figure 3.4 Plot of molar susceptibility x vs temperature and art vs temperature for a
powder sample of the Na+(Dibenzo[a,c]phenazine)' (2-MTHF)1,3 salt.
The solid line represents a fit to the Bleaney-Bowers equation .................. 37

Figure 3.5 X-ray structure of K+(Dibenzo[a,c]phenazine)' (MTHF),,3 with atom
labeling using atomic symbols. Two repeat units are shown for each
chain ...................................................................................... 39

Figure 3.6 Plot of molar susceptibility x vs temperature and xt vs temperature for a
powder sample of the K+(Dibenzo[a,c]phenazine)‘ (2-MTHF),,38alt.
The solid line represents a fit to the Curie Weiss Equation for l-D
magnetic coupling ..................................................................... 42

Figure 3.7 Single crystal X-ray structure of Rb+ (Dibenzo[a,c]phenazine)- with
atom labeling using atomic symbols ................................................. 43

Figure 4.1 A modified H cell ...................................................................... 50

Chapter 1

Background
1.1 Introduction ............................................. 2
1.2 Magnetic Exchange ...................................... 3
1.3 Fullerides ............................. . ................... 13
1.4 Ion Pairing .............................................. 14

1.1 Introduction

In this thesis I report a magnetic and structural analysis for sodium, potassium,
and rubidium salts of dibenzo[a,c]phenazine radical anions and the rubidium salt of 2,3
bis-(3-methoxyphenyl) quinoxaline radical anion. Due to the multidisciplinary nature of
this work, I will devote chapter 1 to discussing magnetic coupling in pure organic
systems and through metal ion bridges. Ion pairing phenomena and their influence on
magnetic behavior for alkali metal salts of organic radical anions will also be covered in
the first chapter. Chapter 2 will focus on characterization of the salts in solution with
EPR and optical spectra being presented. The solid state structural and magnetic
properties of the salts will be covered in chapter 3 along with a summary of results and
conclusions.

Magnetic materials used for applications such as disk storage, sensors, and
switching devices generally consist of metallic centered spin sites assembled into
extended networks of at least two dimensions. Only in recent decades has attention been
focused on using organic based-paramagnetic units for the construction of magnetic
solids. In contrast to metallic magnets which contain (1 or f orbital spin sites, the Spin
density of a pure organic magnet is centered on p or s orbitals.

The typical closed Shell organic molecule has all its electrons spin-paired in an
anti-parallel fashion with all pairs residing in bonding molecular orbitals. In contrast,
organic molecules with an odd number of electrons possess excess spin giving rise to a
permanent magnetic moment. How the spin density is distributed over the organic
framework depends on a number of factors including the presence of conjugation and

electronegative heteroatoms.

The organic radical centers may interact in an ordered solid array to exhibit either
bulk paramagnetic (no interaction), ferromagnetic (spin parallel), or antiferromagnetic
coupling (spin pairing). The type of coupling will depend upon how the magnetic
orbitals of the molecular fragments overlap in the crystal lattice. A brief overview of

magnetic exchange will be given in the next section.

1.2 Magnetic Exchange

When two molecular species, each possessing a single unpaired electron, come
into contact, they interact in either a high or low spin way. What factors govern the
nature and degree of this coupling between paramagnetic subunits? The following
discussion will attempt to address this question.

To understand the nature of the exchange interaction between two electrons in two
orbitals it is necessary to consider the spatial and spin terms of the total electronic
wavefiinction.l The requirement that the electronic wave function must change Sign
under the interchange of two opposite sign electrons gives rise to magnetic exchange. It
follows that the real space wavefunction must be symmetric for antiferromagnetic
coupling and antisymmetric for ferromagnetic coupling. Essentially, an
antiferromagnetic interaction can be viewed as a tendency toward the formation of a
chemical bond (i.e. electrons paired up in the same orbital), while ferromagnetism is like
an antibonding or repulsive interaction between the unpaired electrons. The difference in
energy between the two spin states is the exchange energy, 2k, within a fixed orbital

framework.

To favor a high Spin (ferromagnetically coupled) species where the Spins dwell in
the same Spatial region they must reside in orthogonal orbitals, i.e. the overlap integral
between the orbitals must be zero or nearly so.2 In-phase or bonding interactions resulting
from a finite overlap integral will cause a HOMO-LUMO gap favoring a singlet ground
state. For a magnetic exchange interaction to exist it is also necessary for the orbitals to
be at least partially co-extensive in space. The degree to which the orbitals share the
same space is related to the exchange integral. The exchange integral K is evaluated by
considering the orbital overlap not taking into account the Sign of the MO coefficients.
Large overlap leads to destabilizing Coulomb repulsions in the singlet state, but electron-
electron repulsion is precluded in the triplet state in accordance with the Pauli Exclusion
Principal. The triplet preference for a hypothetical linear carbene l where the carbon
atom is Sp hybridized with two unhybridized p-orbitals underscores this point. To
minimize repulsion it is necessary for one electron to reside in each p-orbital, and since

these two orbitals are orthogonal the triplet state is favored.

 

Bending the carbene as in 2 lowers the energy of the p-orbital in the plane of the
nuclei by mixing in s-character, while the other p-orbital is not affected.3 Pairing the
electrons into the lower energy orbital is not favored due to the destabilizing electron-

electron repulsion and the triplet state is still preferred in the bent form of CH2. However

if the energy gap between the orbitals becomes great enough due to further bending, a
singlet ground state will be favored as in 3.

Another means of achieving a high spin state between pairs of electrons is
superexchange through degenerate orbitals of a closed Shell molecule or ion. Although
superexchange prevents direct overlap of the SOMO, it does not preclude an exchange
interaction. By interacting differently with the half-filled orbitals, the closed shell
fragment can shrink the HOMO-LUMO gap and stabilize the triplet ground state. The

triplet preference for the biradicals 4 and 5 was explained in these terms.4 The through-

. L]

4 5

 

space interaction of the two radical p-orbitals would lead to a large HOMO-LUMO gap
that would favor spin pairing in the lower energy level. However the filled C-H bonding
orbitals of the CH2 subunits can mix with the symmetric HOMO coincidentally raising its

energy enough to make it nearly degenerate with the antisymmetric LUMO (shown in

Figure 1.1).

 

Figure 1.1 Orbital mixing diagram for S.

Unsaturated organic molecules in which unpaired electrons are delocalized over
the 1t framework fall into a class of molecules referred to as altemant hydrocarbons. A.A
Ovchinnikov developed a rather simple method to determine the spin state of planar
alternate hydrocarbons by grouping atoms into two sets (labeled + and -) such that no two
atoms of the same group are adjacent.’ By comparing the number of + and - atoms the
spin state can be determined. One of the most studied spin coupling units developed in
this class of molecules is “metha through a benzene”. The biradical for m-xylylene 6
shows a substantial preference for a triplet ground state, although an ionic closed shell
resonance structure can be envisioned.6 In contrast, the p-xylylene biradical displays a

singlet ground state.

H
I

HC CH HC N CH HC N CH HZC CH2
20/ 2 2U 2 2 l,\ 2 l \
+/
/ / N

I
H

 

6 7 8 9

Incorporation of neutral heteroatoms into the system may not perturb the system
enough to change the preference for a triplet state. Placing a nitrogen into the benzene
ring as in structure 7 induces only a minimal perturbation of the Singlet-triplet gap with
the triplet state still favored.7 However it was recently shown by Dougherty, using
computational methods, that meta coupling through a pyridinium ring is
antiferromagnetic.8 From a valence bond perspective 8 has five closed shell resonance
structures which lends itself to a Singlet ground state. In MO terms the nitrogen cation
stabilizes one nonbonding MO with respect to the other, widening the HOMO-LUMO
gap. The pyridinium ring 9 is reported to be a ferromagnetic coupling unit since the
positively charged nitrogen is placed at a Site where the non-bonding molecular orbital
coefficients are negligible.

Another class of organic high spin systems are derived from stacking planar
paramagnetic 1t systems. McConnell proposed that high spin coupling would be achieved
if 1: systems pancake such that regions with opposite orbital coefficients overlap.9 For
example, when diphenylcarbene units are incorporated in the [2.2] paracyclophane
Skeleton the pseudo-ortho and pseudo-para isomers display high Spin behavior, while the

meta isomer prefers a low spin state.'0 As Shown in Figure 1.2 the overlap of the benzene

rings for the ortho and para isomers is out-of-phase, but the meta isomer has an operative

in-phase interaction.

 

 

 

 

Figure 1.2 Spin distribution of [2.2]paracyclophanes incorporating diphenylcarbene
units. The Signs of the orbital coefficients in the benzene rings are indicated by +/-.

The major drawback for the development of bulk magnetic materials that apply
these systems as coupling units is their high reactivity and poor stability. The search for
relatively sTable high spin coupling units that can assemble properly into a crystalline
multidimensional magnetic lattice has been difficult. Many organic radicals containing
electronegative heteroatoms can be stabilized through complexation to a metal ion.
Magnetic exchange through paramagnetic and diamagnetic metal ions will be discussed
next. I

Formation of bulk magnets through the assembly of organic radicals with oxygen
and nitrogen functionalities coordinated to paramagnetic metals has stimulated a lot of
interest.11 Their basic structural units comprise organic radical centers bridged by metal
ions and the connectivity must be such that their magnetic orbitals are orthogonal. Unlike
conjugated organic radicals with well defined orbital compositions stacked onto one

another, the nature of the exchange is less clearly defined for organic radicals

communicating through a metal ion. It is often the nature of the metal-ligand bond that

governs the magnetic behavior. For example, with Cu(II) complexes of nitroxyl radicals

ferromagnetic exchange is observed when the ligands are coordinated to the 3dzz orbital

of the copper"c while antiferromagnetic exchange occurs when the ligand bonds to the

3dxy orbitals.“ Ryza Musin proposed that a Slight delocalization of the electron from the

1t"‘ orbital of the nitroxide group to the 3d22 orbital of the paramagnetic Cu(II) ion is

responsible for the magnetic coupling.”

The ability of diamagnetic ions to mediate exchange interactions between organic
radicals via their empty and occupied orbitals has only been recently documented. As
with paramagnetic bridges, the nature of the magnetic coupling is governed by the metal-
ligand interaction. A number of novel systems possessing strong exchange interactions
have been developed using diamagnetic metal bridges. Complexation of quinone radicals
to zinc, reported by BuiLock and Harley-Mason in 1951, is one of the earliest examples
of a radical organic ligand coordination to a diamagnetic metal ion.13 Pierpont and co-
workers have discovered a series of high spin octahedral complexes consisting of
paramagnetic semiquinone ligands chelated to a variety of transition metals." The metal
orbital interacting with the semiquinone 1t* molecular orbital often determines the spin
state for these systems. For instance, Ga(III) and Al(III) metals chelating three 3,6-di-
tert-butyl-l ,2-semiquinoate (DBSQ) radical anions possess an S=3/2 ground state while
the analogous Co(III) trimer displays weak antiferromagnetic exchange. The
ferromagnetic coupling for the Ga(III) and AI(III) trimers originates in the interaction of

an empty metal p-Orbital with the semiquinone 1t*orbital. For the Co(III) complex

coupling through the filled metal 3dr: orbital is operative. A Similar coupling mechanism
is responsible for the near diamagnetism of the square planar M(II) (SQ)2 where M=Ni,
Pd, Pt and SQ = 3,5-DBSQ, 3,6-DBSQ where filled metal d-orbitals and semiquinone 1t*
orbitals are interacting. '5 These results demonstrate the dependence of the exchange
interaction upon the electronic configuration of the diamagnetic ion.

High spin complexes M2*(1,4-di-tert-butyl-1,4-diazabutadiene)2 ‘where M = Mg,
Zn, provide other prominent examples of magnetic coupling through a diamagnetic
metal.16 In these systems the two diazabutadiene units complex to the divalent metal and
are orthogonal to one another. Biradical 10 has one electron delocalized on each N2C2
backbone; providing the proper motif for a high Spin complex. In contrast, the analogous
Ga(III) and Al(III) species 11 has one ligand behaving as a dianion and the other as a
singly reduced radical anion, yielding only a doublet ESR signal. For biradical 10 an
equilibrium between the dipole-dipole coupled species and the ionic monoradicals was
proposed by the authors to explain the loss of high spin coupling at room temperature.
Internal disproportionation of this nature can provide a challenge for the design of bulk
ferromagnetic materials from these materials, especially if the diradical centers come into

close proximity in the crystalline lattice.

10

 

 

10 11

The research presented in this thesis focuses on assembling magnetic materials by
connecting organic paramagnetic centers through alkali metal cation linkages. A few
novel magnetic systems have been developed using this strategy. Alkali metal salts of
fluorenone, which have been shown to self assemble into ion pairs, provide the earliest
examples of high spin coupling through alkali metal ions. Hiroto’S proposal that the
fluorenone radicals are linked through a Na“2 Os‘2 parallelograml7 was later supported by
the X-ray analysis of a [fluorenone‘][Na+(dme)2]2 crystal by Bock.‘8 The Na+ ion is
solvated by two DME molecules allowing for chelation to only two fluorenone oxygens

in order to achieve its preferable coordination sphere of six. The triplet benzophenone

 

salt 12 is thought to possess a similar dimeric structure. It is interesting to note the
absence of magnetic susceptibility data on these salts, but long range order may be

precluded due to the large dimer-dimer distances. Ion pairing of alkali metal cations and

11

organic radical anions is an important concept that will be discussed in the following
section.

Further insights into magnetic coupling through alkali metal cations were
developed from work done in this lab. Using lithium and sodium salts to bring neutral
paramagnetic tripod ethers 13 into communication was largely unsuccessful.19 It was

postulated that the large metal-ligand distance contributed to the weak coupling.

 

13

A more successful approach was to chelate radical anion centers to alkali metal
cations in an attempt to strengthen the interaction between the paramagnetic centers
through stronger electrostatic interactions. Bulk one dimensional ferromagnetism was
reported by Misiolek and Jackson for crystals of the 4-Carboxy-TEMPO radical sodium
salt.20 Coupling between the TEMPO radicals mediated by sodium cations were
apparently responsible for the materials magnetic behavior. The potassium salt also was
reported to display ferromagnetic behavior. Weak antiferromagnetic coupling along
chains of K“[2,3-Bis(2-pyridyl)quinoxaline] (THF)2 radical anion salt was also reported

by Dye and Jackson.21 Multidimensional coupling is precluded in this systems due to the

12

large gaps between the chains. These results encouraged the development of other

similar systems where alkali metal cations would provide an exchange coupling pathway.

1.3 Fullerides

It has been demonstrated that C60 and C70 doped with alkali metals are conductive and that
superconductivity has been observed in K3C60 and Rb3C,0 at low temperature”. K3C60
possesses a Tc value of 19.3K, but the heavily doped KéC60 phase is not
superconductive.” Diffraction data reported by P.W. Stephens of K3C60 Showed that the
potassium atoms reside in tetrahedral and octahedral Sites within the C60 lattice (Figure
1.3).24 This composition has also been identified as the metallic phase, however new
compositions have been reported with possessing higher superconducting transition
temperatues?"27 Some distortion of the C60 framework in comparison to the neutral
species is observed, but the spacing between the 1t systems remains nearly the same. The
shortest distance between a potassium ion and a C,0 carbon atom is 3.27A as compared to
3.06A in intercalated graphite. It is proposed that the conductivity of this salts arise from
the delocalization of one electron per metal into a conduction band consisting of three

degenerate orbitals.

l3

 

Figure 1.3 The structure of superconductive K3C60. The shaded and white circles
represent octahedral and tetrahedral sites respectively.
1.4 Ion Pairing

The magnetic properties of alkali metal salts of organic anions in solution and in
the solid state can be influenced by how strongly the metal associates with the radical
anion. For the design of magnetic solids it is desirable to achieve tight ion pairing
between the metal and the organic radical center. Factors such as solvent polarity, cation
size, and the degree of localization of charge density on the organic unit all play a role
influencing the tight ion pair-solvent separated ion pair equilibrium. EPR has been
effectively used to elucidate alkali metal organic anion interactions in solution. In
systems where the metal comes into close proximity to the organic anion some leakage of
Spin density onto the metal can occur, provided an orbital pathway exists. Spin
delocalization onto the cation is indicated by the presence of alkali metal coupling

constants in the EPR spectrum.

14

Unsaturated organic ring systems reduced with alkali metal often display small
metal splittings. Weissman reported the first 23Na coupling constant for a THF solution
of the naphthalene radical anion salt due to spin transfer from the naphthalene n-system
to the sodium S orbital.28 A sodium Splitting of 1.05 gauss was observed at room
temperature and declined steadily with decreasing temperature. From this initial study it
was postulated that the equilibrium between tight ion pairs and solvent separated ones is
temperature dependent.

The presence of heteroatoms on the organic moiety can lead to tight ion pairing in
solution resulting in large metal coupling constants. The sodium splitting for o-
dimesitoylbenzene radical anion is 6.95 gauss in DME at room temperature, one of the
largest reported.29 These ion pairs are considered “tight” while alkali metal cations
associating with hydrocarbon radical anions are often defined as “loose”. Ether solvents
such as THF or glymes can effectively compete for the metals coordination sphere since
the charge on unsaturated hydrocarbons tends to be more delocalized then on molecules
possessing electronegative heteroatoms.

The unusually large alkali metal coupling constants a(”K)=1.2-1.5, a(8’Rb)=0.4-
0.84, a(87Rb)=l4-28, and a('33CS)=7.O-26 gauss reported by Gerson for salts of 1,4 and 2,3
di-tert-butylbuta-l ,3 dienes in DME and THF provides an interesting anomaly.30 It is
rationalized by Gerson that partial encapsulation of the metal by t-butyl groups prevents
metal solvolysis and facilitates tight ion pairing. Molecular modeling seems to support
this hypothesis.

Metal splittings of aromatic nitrogen heterocyclic radical anions often do not

appear in the EPR spectrum although contact ion pairing does occur in these systems. No

15

metal splittings were reported for the sodium and potassium salts of
diphenquuinoxaline3| (compound 14) in DME and the potassium salt of 2,3-BiS(2-
Pyridyl)quinoxaline2‘ (compound 15) in THF. Chelation to the nitrogen lone pair places
the metal s-orbital in the nodal plane of the it system in the diphenquuinoxaline Species

preventing a pathway for spin transfer.

 

14 15

Often the degree of ion pairing in solution can have a large impact on the
assembly in the crystalline lattice Since solvent molecules can compete effectively for the
coordination sphere of the metal. In recent years a number of investigations done on
alkali metal salts of unsaturated ring systems by Book and co-workers has uncovered
clues in cation solvation and aggregation phenomena of these species in the crystal
lattice.32 He recently made this general conclusion,” “By comparison of single crystal
structure data with results from NMR, ESR, or UV/vis measurements, it ofien becomes
obvious that solid state structures of organometallic complexes largely represent the
species in solution.” The crystallization process of alkali metal cations and organic anions
into a lattice can be controlled by a number of factors among which solvation can play a

dominating role. Crystallizing from highly chelating environments often result in solvent

16

wrapped cations, only weakly coordinated to the anion. A dramatic example is the
crystallization of potassium perylene radical anion in triglyme. The potassium metal is
encapsulated in a shell of two triglyme molecules completely separated from the perylene
anion.33 Extensive delocalization of negative charge around this unsaturated 1r
hydrocarbon seems to further compromise contact ion pair formation. Solvation of the
metal to this extent is undesirable from a magnetic standpoint Since the orbitals of
isolated organic anions and metal cations will not overlap significantly.

Since the solvation energies of alkali metal cations tend to decrease with
increasing ionic radii the larger metals tend to chelate less solvent, leaving more sites
available for radical anion ligands to coordinate. This trend is displayed quite
dramatically with alkali metal salts of [TCNE] radical anions.

Crystal structures of [TCNE]' M” complexes where M = Na, K, and CS have been
reported by Bock and co-workers.34 The cesium tetracyanoethylenide salt crystallizes
from DME and hexane into a solvent free 2-D network. Although DME is an
energetically favorable ligand for cation solvation, its “bite” is too small for both oxygens
to properly coordinate to the cesium cation with an ionic radii of 169pm. The result is
two unique cesium ions coordinated to seven and eight nitrogen centers. TCNE radical
anions stack into chains with Cs+ connecting the chains and adjacent ligands. Solvent
free networks possessing small ligand-ligand distances are desirable synthetic targets due
to their potential magnetic properties.

In contrast to the cesium structure, the sodium and potassium salts possess a

single DME chelated to the metal. The structural motifs of the chains are also influenced

17

by the size of the cation. The lattice of the sodium salt consists of it stacked dimers
separated by 300 pm and connected by a DME chelated Na+ ion forming infinite chains.
The potassium structure has layers of TCNE ions separated by 315 and 360pm with the
cations residing in the smaller gap.

Recently crystals were grown using substituted quinoxalines as the radical anion
center. The facilitation of contact ion pair formation through the concentration of
negative charge at the two nitrogen atoms makes them attractive for building
multidimensional magnetic networks. Crystals of sodium and potassium salts of diphenyl
quinoxaline in glyme assemble into one dimensional chains with the metal cations
coordinated directly to the nitrogen lone pair forming M+(solvent) bridges linking the
anions.31 Interestingly no temperature dependent susceptibility data was
reported for the salt. The sodium and potassium salts are rather Similar considering the
36pm difference in their atomic radii. The potassium salts of Bis(2-Pyridyl)quinoxaline
provide a prominent example of how metal solvation can influence the dimensionality of
the crystal lattice. Crystallizing in THF resulted in one dimensional chains of the
quinoxaline connected by K“(THF)2 linkages while in methylamine dimers resulted from
three solvent molecules coordinating to the metal.2| Presumably the stronger chelating
ability of the amine nitrogen and its smaller size makes it better able to compete for the
potassium coordination sphere then THF.

Many aromatic nitrogen heterocyclic radical anions have been characterized in
solution, but not in the solid state. EPR analysis of quinoxaline, phenazine and
dibenzo[a,c]phenazine radical anions along with molecular orbital calculations indicate

that the spin density distribution is fairly localized on the nitrogen atoms.” The proton

l8

and nitrogen hyperfine splitting constants of many other nitrogen heterocyclic radical
anions have also been measured (Figure 1.4).

063 0.27
222 5N“ " 178 495. 0.88”
0:: 'u\
/
N 0

Figure 1.4 Hyperfine splittings in gauss for the quinoxaline, phenazine, and
dibenzo[a,c]phenazine radical anions produced electrochemically in DMF with tetra-n-
butylammonium iodide as a supporting electrolyte.

 

l9

Chapter 2

Characterization of the Radical Anion Salts in Solution

2.1 Introduction ............................................................... 21
2.2 EPR Spectra of the Radical Anions .................................... 21
2.3 Optical Spectra of the Radical Anions ................................. 24

20

2.1 Introduction.

In this chapter I will report the EPR spectrum for the sodium and potassium salts
of the dibenzo[a,c]phenazine radical anion. The hyperfine splitting constants that I report
are taken directly from the ENDOR spectrum of the potassium salt. In the last section of
this chapter I will Show optical Spectra for THF solutions of Rb” ((Bis-2-methoxyphenyl)
quinoxaline) and M+ (dibenzo[a,c]phenazine) (M = Na, K, Rb) and make comparisons to
spectra of other anion salts reported in literature. The influence ion pairing has on the

optical spectrum of these salts in solution will be discussed.

2.2 EPR and ENDOR Spectra of the Radical Anions.

The solution EPR spectrum for the Na+(Dibenzo[a,c]phenazine)' salt in THF is
displayed in Figure 2.1 and solution EPR and ENDOR spectra for the potassium salt are
in Figure 2.2. Due to the complexity of the hyperfine splitting, the coupling constants
shown in Figure 2.3 were taken directly from the ENDOR spectrum by Andrew Ichimura.
Inspection of the Figure reveals that the center of the spectrum is split by the sodium
cation. It is possible that in the lowest energy state of the complex the cation resides
directly over the 1: system. Chelation to the nitrogen lone pair may be energetically
disfavored due to steric congestion caused by the hydrogen atom on the phenanthrene
subunit. In the EPR spectrum of Na+(2,3-diphenquuinoxa1ine)' in dimethoxyethane
reported by Book, metal Splitting was not reported. In this cas,e the rotation of the phenyl
rings can allow the sodium to chelate directly to the nitrogen lone pair minimizing spin

transfer. Dimethoxyethane may also solvate the metal weakening the association

21

between the cation and radical anion. No metal splitting appears in the EPR spectrum for

the potassium salt.

 

 

 

 

I I

3340 3350 3360 3370 3380 3390
Field(Gauss)

.__—r_n___.r)._vk _- . ,__._____.___, .__J

R
l
l
I
l
l
2020 # 1 J r t
l
'I
T

Figure 2.1 EPR spectra of Na+(Dibenzo[a,c]phenazine)' in THF, recorded at -5°C.

22

30000
20000 ~~
10000 .

0 -_
-10000 4
-20000 1—

-30000 . . .
3340 3360 3380 3400 3420

Field (Gauss)

 

T

 

 

 

 

(B)

 

 

 

 

 

 

I 1.65 I

| 2.53 I
3.99

5.34

 

 

q-
.-

 

 

 

m 11 1'2 1‘3 174 175 to 17 18
MHz

Figure 2.2 (A) EPR spectra of K+(Dibenzo[a,c]phenazine)' in THF, recorded at
-96.7°C. (B) ENDOR spectrum of K*(Dibenzo[a,c]phenazine)' , recorded at -96.7°C.

23

0.21

0.59 0.906
' 1.91 5.12 "
MEIR
.Q
N

Figure 2.3 Hyperfrne coupling constants (in Gauss) from the ENDOR spectrum of
K"(dibenzo[a,c]phenazine)'

 

 

2.3 Optical Spectra of the Radical Anions.

The optical spectrum of the radical anion salts in THF (Figure 2.4) were measured
in the range 400-2000nm. Two visible absorptions were Observed for all the salts along
with a broad band in the near IR. The Spectrum for the intensely blue Rb+ (2,3 bis-(3-
methoxyphenyl) quinoxaline) has two visible absorptions at 550nm and 620nm and one
broad band in the near IR centered around 1 100nm. The spectrums for red THF solutions
of the dibenzo[a,c]phenazine radical anions are shown in Figure 2(b) and 2(c). The
optical spectrum reveals a sharp absorption in the 558 to 568nm range; the exact position
depends on the size of the alkali metal cation (Table 2.2). Similar bathochromic shifts
arising from ion pairing have been reported for alkali metal salts of various aromatic
hydrocarbon radical anions and ketones in solution.” 39 These studies suggested that the
shifts to higher wavelengths are a result of a weaker ion pair between the radical anions
and the larger alkali metal cations. The Shorter wavelength visible absorption was
unaffected by the size of the cation and may indicate the presence of aggregates or a

different species.

24

 

 

 

Absorbance

0.6 ——

<14 m

0.2 1~ i“
0 i .

400 900 1400 1900
Wavelength (nm)

 

 

 

 

 

1.8
1.6 «~
1.4 ~—
1.2 «—

(is.
(16 w
(14 w
052“
o if ,
400 900 1400 1900

Absorbance

 

 

 

 

Wavelength (nm)

Figure 2.4 (Top graph) Optical spectra for Na+(dibenzo[a,c]phenazine)' (dark line), and
K+(dibenzo[a,c]phenazine)' (light line). (Bottom graph) Rb+(dibenzo[a,c]phenazine)' with
(dark line) and without the presence of C222 cryptand (light line).

25

To further weaken the interaction between the metal and radical anion C22

cryptand was introduced into the rubidium salt solution. The result is the formation of a

cryptand separated ion pair where the interactions between the anion and metal cation are

significantly reduced (Figure 2.5). As expected the adsorption peak was red shifted

significantly (568nm to 584nm).

 

 

Rat.

Rb+

 

Cryptand [2.2.2]
>

 

 

Figure 2.5 Formation of the Rb+ C22 cryptand separated ion pair.

26

Chapter 3

Characterization of the Radical Anion Salts in the Solid State
3.1 Introduction ....................................................................... 28

3.2 X-ray Structure and Magnetism of the Rb+(2,3 bis-

(3-methoxyphenyl) quinoxaline) (THF) salt ................................. 28
3.3 X-ray Structure and magnetism of the Na+(dibenzo[a,c]phenazine)'

(MTHF)2 salt ...................................................................... 32
3.4 X-ray Structure and magnetism of the K+(dibenzo[a,c]phenazine)'

(MTHF)2 salt ...................................................................... 37
3.5 X-ray Structure and magnetism of the Rb+(dibenzo[a,c]phenazine)'

salt ................................................................................. 43
3 .6 Summary .......................................................................... 45

27

3.1 Introduction.

In this chapter I will report the Single crystal X-ray structures and solid state
susceptibility data for Rb +(2,3 bis-(3-methoxyphenyl) quinoxaline) (THF),
Na+(dibenzo[a,c]phenazine)' (MTHF)2, K+(dibenzo[a,c]phenazine)' (MTHF)2,3, and
Rb+(Dibenzo [a,c]phenazine)’ salts. The extremely air sensitive crystals were prepared by
diffusing pentane into THF or MTHF solutions of the radical anion salt at room
temperature under high vacuum. Since they were grown under comparable conditions the
structural diversity Observed with these salts can be attributed to the Size of the alkali
metal cation. The magnetic data provides further insight into structure-magnetic
relationships of organic radical anions bridged by diamagnetic alkali metal cations.

Tables of selected distances and angles will be presented for each structure. A
Table summarizing unit cell dimensions and refinement is located at the beginning of the

appendix. The coordinate Tables are also in the appendix.

3.2 X-ray Structure and Magnetism of the Rb*(2,3 bis-(3-methoxyphenyl)
quinoxaline) (T HF) Salt.

The Rb+(2,3 bis-(3-methoxyphenyl) quinoxaline)’ (THF) salt crystallizes into
polymeric chains of 2,3 bis-(3-methoxyphenyl) quinoxaline radical anions strung together
by Ni-Rb’fiN“ linkages. The chains possess a connectivity pattern of repeating NS-Rbi-
N“(C-C)2-N“-Rb+ rings (Figure 3.1a) with Rb-N-Rb bond angles of 117°. The radical
anions stack parallel to one another along the C direction with singly THF solvated Rb+

ions residing between them. Within each chain the ligands point in the same direction

28

(A) 0'

(B)

 

 

 

 

 

Figure 3.1 Single crystal X-ray structure of Rb+(2,3-biS-(3-methoxyphenyl)quinoxaline)'
(THF) with atom labeling using atomic symbols. (A) Part of a single chain is shown.
(B) Crystal packing for the unit cell looking down the chain axis.

29

along the stacking axis, presumably to optimize packing. The phenyl rings are twisted
39.7° and 453° from the plane of the quinoxaline moiety. Selected angles and distances
are displayed in Table 3.1.

Three nitrogens and one THF oxygen form a distorted square planar environment
around the metal with an average Rb-N distance of 3.1 A. A formally seven fold
coordination sphere of the cation is completed through contacts under 3.5 A to two
adjacent carbon centers on the phenyl ring C(9), C(10) and one carbon adjacent to the
nitrogen on the quinoxaline moity C(ZB). The methoxyphenyl ring (C10B-C18B) caps
off one face of the cation but does not appear itself to be close enough to interact. It is
interesting to note that the methoxy oxygen does not serve as a ligand in the environment
of the metal.

Table 3.1 Selected Distances and Angles for Rb*(2,3 bis-(3-methoxy—
phenyl) quinoxaline) (THF) Salt. Distances are given in A.

 

 

Distances

Rb-N(1A) 3.212(2) N(1)-Rb-C(10) 103.14(5)
Rb-O(1) 2.885(2) O(1)-Rb-C(ZB) 126.88(6)
Rb-N(1 B) 2.973(2) N(1)-Rb-C(2B) 93.13(5)

Rb-N(8) 3.042(2) N(8)-Rb-C(2B) 9566(5)

Rb-C(10) 3.306(2) N(l)-Rb-C(ZB) 23.71(5)

Rb-C(2B) 3.438(2) C(10)-Rb-C(2B) 121.02(S)
Rb-C(9) 3.462(2) O(1)-Rb-C(9) 9806(5)

Rb-C(11) 3.540(2) N(1B)-Rb-C(9) 165.89(5)

N(8)-Rb-C(9) 2308(5)

Angles N(1A)-Rb~C(9) 8800(5)

O(1)-Rb-N(1B) 77.50(5) C(10)-Rb-C(9) 25.41(5)

O(1)-Rb-N(8) 82.87(5) O(l)-Rb-C(11) 7204(6)

N(1B)-Rb-N(8) 160.01(5) N(lB)-Rb-C(11) 124.50(5)
O(1)-Rb-N(1A) 149.85(5) N(8)-Rb-C(11) 5060(5)

N(1A)-Rb-N(IB) 102.59(4) N(1B)-Rb-C(1 l) 126.51(5)
N(8)-Rb-N( 1 A) 92.47(5) C(10)-Rb-C(1 1) 23 .48(5)

O(1)-Rb-C(10) 94.10(6) C(9)-Rb-C(11) 4200(5)

N(1)-Rb-C(10) 140.69(5) O(101)-Rb-C(3B) 109.23(6)
N(8)-Rb-C(10) 4404(5) Rb-N(1)-Rb(B) 116.81(5)

 

30

 

Magnetic susceptibility data for the crystalline salt was recorded over the
temperature range 2-300K. The xt vs t plot is displayed in Figure 3.2 along with a fit to
the one dimensional Curie-Weiss model for magnetic coupling. The Weiss constant for
the salt is found to be -4K indicating antiferromagnetic behavior at low temperature. The
Curie constant is 0.3 Semu/mol which is slightly lower then the expected value of
0.375emu/mol for a crystal of single spin molecules. Some decomposition of the
extremely air sensitive crystals may be responsible. The magnetism appears to be
consistent with the structural information just presented. The weak magnetic interaction
is a consequence of the nearly 5 A distance between the paramagnetic centers and the
poor ability of the Rb+ to act as a coupling unit. Long range magnetic exchange

throughout the crystal lattice is precluded by the large chain-chain distances.

31

0.07
0.06
0.05
0.04
0.03
0.02
0.01

0 . T
0.00 100.00 200.00 300.00

 

x (emu/mol)

 

 

Temperature (K)

0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
O . + .
0 100 200 300
Temperature (K)

 

xT(emu K/mol)

 

 

Figure 3.2 Plot of molar susceptibility x vs temperature and xt vs temperature for a
powder sample of the Rb+(2,3 bis-(3-methoxyphenyl) quinoxaline) Salt. The solid line
represents a fit to the Curie Weiss Equation for 1-D magnetic coupling.

3.3 X-ray Structure and magnetism of the Na”(Dibenzo[a,c]phenazine)’ (MTHF)2
salt.

The Na*(Dibenzo[a,c]phenazine)' (MTHF)2 salt crystallizes into an orthorhombic
unit cell of cyclic dimers possessing a Nai-N“(C-C)2'-N5’-Na+ connectivity pattern (Figure

3.2). Selected distances and angles with deviations are displayed in Table 3.2.

32

 

 

 

 

Figure 3.3 Single crystal X-ray structure of Na+(Dibenzo[a,c]phenazine)' (MTHF)2 with
atom labeling using atomic symbols. (A) A single dimer is displayed. (B) Crystal
packing for the unit cell with the shortest distance between dimers depicted by a dashed
line. Only seven of twelve dimers in the unit cell are depicted in this Figure for clarity.

33

Comparisons will be made to the Na+(2,3-diphenquuinoxaline)‘ (DME) crystal structure
reported by Bock and co-workers throughout this discussion.26

The two dibenzophenazine anions of the dimer displayed in Figure 3.1a,
“pancake” onto one another and are connected through two NS-Nai-NS bridges that are
offset by 0.27 A and -0.27 A from the plane of the four nitrogen atoms. The dimer
exhibits Ci symmetry inverting equivalent sets of atoms through a point in the center (i.e.
N16 into N16A, C14 into C14A). The mean molecular planes of the 1: systems are
parallel with stacking distances ranging from 3.245 A at the nitrogens to 3.416 A at C18.
Two coordination Sites of the sodimn ion are occupied by MTHF oxygens, to complete a
tetrahedral environment around the metal. It is interesting to note that the cations do not
directly interact with the in-plane sp2 lone pair of the quinoxaline nitrogens, but are offset
by 42° from the plane of the quinoxaline moiety. This contrasts to the direct chelation of
the sodium cation to the nitrogen nonbonding electrons of the 2,3-diphenquuinoxaline
radical anion.

The metal appears not to interact significantly with the carbon framework. The
shortest C-Na distance of 3.09 A is considerably longer then those to the electron rich
oxygen and nitrogens that average 2.33 A and 2.43 A respectively. In contrast, two
sodium-carbon contacts of 3.06 A were reported for the 2,3-diphenquuinoxaline salt to
complete a pseudo-octahedral environment around the metal. Spatial arrangements and
the rigidity of the dibenzo[a,c]phenazine framework apparently prevent such close metal-
carbon contacts that would allowing sodium to attain its preferred coordination sphere of

six. The dibenzo[a,c]phenazine skeleton is not precisely planar, but slightly curved at the

34

phenanthrene subunit, with atoms C5 and C12 being situated 0.19 A and 0.18 A from the
mean molecular plane of the phenazine moiety.

Table 3.2. Selected Distances and Angles for Na*(Dibenzo[a,c]phenazine)‘
(2-MTHF)2. Distances are given in A.

 

 

Distances
Na(1)-O(2A) 2.315(4) O(1A)-Na(l)-N(16A) 153.5(2)
Na(1)—O(1A) 2.344(3) O(2A)-Na(1)-N(1) 151.5(2)
Na(1)-N(16A) 2.422(3) O(1A)-Na(1)-N(1) 94.9(1)
Na(l)-N(1) 2.434(3) N(16A)-Na(l)-N(1) 83.8(1)
Na(1)-C(4) 3.094(5) O(2A)-Na(1)-C(4) 94.4(2)
O(1A)-Na(1)-C(4) 98.2(2)
Angles N(16A)-Na(1)-C(4) 58.8(1)
O(2A)-Na(1)-O(1A) 88.4(2) N(1)-Na(l)-C(4) 112.9(1)
O(2A)-Na(1)-N(16) 105.3(2)

 

 

The packing of the crystal showing a unit cell is displayed in Figure 3.3B. An
inversion point lies on each edge of the unit cell with one residing in the center of the
box. The smallest distance between dimers in the unit cell is 3.43A as measured by the
shortest (C-C) distance (shown as a dashed line in Figure 3.1b). All dimers in the cell
are crystallographically equivalent.

The magnetic susceptibility of a powder sample was recorded for the
Na+(dibenzo[a,c]phenazine)' (2-MTHF)2 salt in the temperature range 2-300K. The solid
line through the data points represent a least squares fit of the magnetic susceptibility data
to the Bleaney-Bowers equation for magnetic coupling in isolated dimers.4O The fit
follows the curve at high temperatures with some deviation below 10K. A J/K of -22.2K
indicates antiferromagnetic coupling between the two unpaired Spins. The susceptibility

reaches a maximum at 30°K of 0.136 emu/mol and falls precipitously to 0.0021 emu/mo]

35

at 2K. The equation used to fit the data predicts a value of 3.6 * 10’lo emu/mo! at 2.0K.
It is not surprising to find some paramagnetism in the sample at low temperature.

The magnetic behavior appears to be consistent with the solid state structural data
presented above. Long range order throughout the crystal lattice is precluded by the
empty gaps between the dimers. The SOMO’S between the two rt systems should interact
most strongly at the nitrogen, the site with the largest orbital coefficients and where the 71:
systems come in closest proximity. Since the interaction is in-phase (or bonding) the
nature of the coupling should be antiferromagnetic. It is uncertain to what extent
exchange through sodium is operative, though it has been established through work done

previously in this lab and others that the Na cation can act as a coupling unit.”

36

 

0.016

0.014 -_
g 0.012 .—
-§ 0.01 -
"g 0.008 A.
E 0.006 -_
30.004 -
X

0.002

 

 

 

 

0 100 200 300
Temperature (K)

 

xT (emu K/mol)

 

 

 

 

0 100 200 300
Temperature (K)
Figure 3.4 Plot of molar susceptibility x vs temperature and xT vs temperature for a

powder sample of the Na+(Dibenzo[a,c]phenazine)‘ (2-MTHF),,3 salt. The solid line
represents a fit to the Bleaney-Bowers equation.

3.4 X-ray Structure and Magnetism of the K"(Dibenzo[a,c]phenazine)° (MTHF)2,3
salt.

The K” (dibenzo[a,c]phenazine)’( MTHF) salt crystallizes into the orthorhombic
space group Pbca with six unique anions in the unit cell. Distances and angles are

summarized in Table 3. In the crystal structure the radical anions are strung together by

37

Ni-KNS‘ linkages to form the two crystallographically independent triple braided chains
shown in Figure 3.5. The chains propagate along the c-axes by packing three unique
dibenzophenazine anions into three separate stacks that are connected by potassium
bridges to form a single “thick chain”. The stacking distance is 5 A with potassium
cations residing between the stacks. Within the chain the dibenzo[a,c]phenazine anions
are pointing in the same direction along the c-axis.

The coordination environments of the metals residing at the chain edge differ
significantly from the one in the interior. The four metals on the edge interact with three
nitrogens and one solvent oxygen with average K-N and KO contact distances of 2.86 A
and 2.69 A . The average K-N distances reported for the K+(2,3-Bis(2-Pyridyl) (THF),27
and the K+(2,3-diphenquuinoxaline) (DME)26 salt were respectively 2.9 A and 2.8 A.
The coordination sphere of the metals are completed by contacts to three or four carbon
atoms that are Situated under 3.4 A away. All K-C distances under 3.5 A are displayed in
Table 3.2. Two K—C contacts of 3.12 A and 3.33 A were reported for the

diphenquuinoxaline salt.

38

< E ESE on Bunsen saute .EEEEEOEBEBEVE 8e ease. as message 882% on sets

 

39

 

The environment of the cations residing in the center of the chains (K2 and K5)
are free of solvent, with interactions only to three radical anions. The K2-N8A and K5-
N1C distances are too long to complete a pseudo-square planer arrangement of nitrogens
around the metal. Instead the metals interact with three nitrogens in a slightly distorted T
geometry. All other nitrogen atoms in the crystal interact with two metals with K-N—K
bond angles averaging 123°. For the K”(2,3-bis(2-pyridyl)(THF)2 salt and the K*(2,3-
diphenquuinoxaline)‘(DME) salts the potassium cation coordinated directly to the
nitrogen lone pairs, but in the dimer based K“(2,3Bis(2-Pyridyl)(CH3NH2)3 structure the
K+ is out of plane.

Table 3.3 Selected Distances and Angles for K+(Dibenzo[a,c]phenazine)‘.
(MTHF)2,3 Distances are given in A.

 

 

Distances

K(1)-O(101) 272(2) K(5)-C(2D) 3.45(2)
K(l)-N(8) 2.810(14) K(5)-C(3C) 326(2)
K(1)-N(1) 2.858(14) K(6)-N(1A) 2.797(14)
K(1)-N(8B) 3.022(13) K(6)-N(8A) 2.87(2)
K(1)-C(7B) 323(2) K(6)-N(lB) 2.875(12)
K(1)-C(1 l) 3.31(2) K(6)-O(201) 2.662(10)
K(1)-C(7) 339(2) K(6)-C(2A) 330(2)
K(1)-C(20) 339(2) K(6)-C(3A) 3.49(2)
K(l)-C(6B) 3.42(2) K(6)-C(2B) 326(2)
K(l)-C(2) 3.44(2) K(6)-C(11A) 324(2)
K(l)-C(3) 3.49(2)

K(2)-N(1) 3.158(12) Angles

K(2)—N(8B) 2.81(2) O(101)-K(1)—N(8) 94.3(4)
K(2)-N(1B) 2.850(14) O(101)-K(1)-N(1) 800(4)
K(2)-C(20B) 324(2) N(8)-K(1)-N(1) 166.9(4)
K(2)-C(3) 325(2) O(101)-K(1)-N(8B) 164.6(5)
K(2)—C(6A) 326(2) N(8)-K(l)—N(8B) 89.8(4)
K(2)-C(2) 325(2) N(l)-K(1)-N(SB) 98.8(4)
K(2)-C(1 13) 335(2) K(1)-N(1)-K(2) 120.1(4)
K(2)-C(7A) 3.40(2) K(2)-N(lB)-K(6) 124.8(5)
K(2)-C(7B) 3 .40(2) K(4)-N(1C)-K(5) 1 1 8.5(5)

 

40

Table 3.3 (Cont’d).

 

 

K(2)-C(2B) 3.44(2) K(5)-N(1D)-K(4) 124.0(5)
K(6)-C(3B) 342(2) N(8B)-K(2)—N(IB) 169.3(5)
K(6)-C(22B) 3.49(2) N(8B)-K(2)—N(l) 94.8(4)
K(3)-O(301) 265(2) N(lB)-K(2)-N(1) 87.9(4)
K(3)-N(8E) 2.877(14) N(8B)-K(2)-C(ZOB) 115.9(5)
K(3)-N(8D) 2.904(13) O(301)-K(3)-N(1E) 92.1(4)
K(3)-N(1E) 2.806(14) O(3OI)-K(3)-N(8E) 79.9(4)
K(3)-C(7D) 3.16(2) N(lE)-K(3)-N(8E) 166.3(4)
K(3)-C(11E) 331(2) O(301)—K(3)-N(8D) 164.5(5)
K(3)—C(6D) 336(2) N(1B)-K(3)-N(8D) 91.8(4)
K(3)-C(2OE) 337(2) N(8E)-K(3)-N(8D) 98.9(4)
K(3)-C(7E) 3.45(2) O(401)-K(4)-N(8C) 94.3(5)
K(3)-C(22E) 3.49(2) O(401)-K(4)-N(1D) 149.0(6)
K(3)-C(2E) 341(2) N(8C)-K(4)-N(1D) 96.8(4)
K(4)-N(1C) 289(2) O(40l)-K(4)-N(1C) 80.4(5)
K(4)—N(8C) 277(2) N(8C)-K(4)-N(1C) 159.9(5)
K(4)-N(1D) 2.886(13) N(1D)-K(4)-N(1C) 97.6(4)
K(4)-O(401) 272(2) N(8D)-K(5)-N(1D) 171.2(6)
K(4)-C(20C) 323(2) N(8D)-K(5)«N(8E) 93.6(4)
K(4)-C(7C) 330(2) N(ID)-K(5)-N(8E) 86.4(4)
K(4)-C(3D) 3.44(2) N(8D)-K(5)-N(1C) 84.3(4)
K(4)-C(2D) 319(2) N(1D)-K(5)-N(1C) 94.3(4)
K(5)-N(8D) 283(2) N(8E)-K(5)-N(1C) 170.7(5)
K(5)-N(1D) 283(2) O(201)-K(6)-N(1A) 90.1(4)
K(5)-N(8E) 3.220(12) O(201)-K(6)-N(8A) 81.0(4)
K(5)-N(1C) 3.463(14) N(1A)-K(6)-N(8A) 162.3(4)
K(5)-C(2C) 331(2) O(201)-K(6)-N(IB) 161.5(4)
K(5)-C(7E) 3.264(14) N(1A)-K(6)-N(1B) 963(4)
K(5)-C(6E) 327(2) N(8A)-K(6)-N(IB) 96.9(4)
K(5)-C(20D) 336(2) K(2)—N(8B)-K(1) 124.6(5)
K(5)-C(7D) 339(2) K(5)-N(8D)-K(3) 126.2(5)
K(5)-C(11D) 339(2) K(3)-N(8E)-K(5) 120.0(4)

 

 

Magnetic susceptibility for the crystalline salt was recorded over the temperature
range 2-300K. The xT vs T plot is displayed in Figure 3.6 along with a fit to the one
dimensional Curie-Weiss model for magnetic coupling providing a Curie constant of

0.33emu/mol. The Weiss constant for the salt is found to be -12.6K indicating

41

antiferromagnetic behavior at low temperature. The weak magnetic interaction is a
consequence of the large distance between the ligands and the inability of the K+ to act as
an efficient coupling unit. The K+(2,3-Bis(2-Pyridyl) (TI-IF)2 salt had a similar Weiss
constant of -12°K. Although the one dimensional model appeared to fit the data well the
chains do not represent a pure one dimensional system. Long range 3-D coupling is

precluded by the nearly 5 A distance between the chains.

0.025

 

0.02
0.015
0.01 ~~

x (emu/moi)

0.005 .

 

 

 

0 i .
0.00 100.00 200.00 300.00

 

Temperature (K)

0.35

 

.9
w
i

 

o
Pix)
Nth

 

xT (emu K/mol)
p
G

O
b?
L)..—

 

O

 

r

0 100 200 300
T (K)

Figure 3.6 Plot of molar susceptibility x vs temperature and xt vs temperature for a
powder sample of the K+(Dibenzo[a,c]phenazine)' (2-MTHF),,3salt. The solid line
represents a fit to the Curie Weiss Equation for l-D magnetic coupling.

42

The Rb+ (dibenzo[a,c]phenazine) salt crystallizes in the orthorhombic space
group Pbca with a single unique radical anion and metal cation in the unit cell. Dibenzo
[a,c]phenazine radical anions linked through N&-Rb+-N& bridges assemble into a solvent

free 2-D network consisting Of a N&-Rb+-N&(C-C)2-N&-Rb+ pseudo-five membered ring

     

e
e ('3‘ RbAA

D‘O'O

 

Figure 3.7 Single crystal X-ray structure of Rb+ (Dibenzo[a,c]phenazine)- with atom
labeling using atomic symbols.

43

connectivity pattern. The metals alternate by 0.28 A above and below the least squares
plane of the metals. The anions are alternately turned around the b-axis and are not
packed in a parallel fashion.

The Rb+ resides in a near square planer environmentof four nitrogen atoms with
an average Rb-N contact distance of 3.21 A. This is greater then the average Rb-N
distance of 3.07 A I reported for the Rb+(2,3 bis-(3-methoxy-phenyl) quinoxaline)“ (THF)
salt. The longer Rb-N contacts may be attributed to the steric congestion arizing from
packing four rigid dibenzo[a,c]phenazine ligands around the metal. The nearest carbon
lies 3.36 A away with four others within 3.5 A, however the metal does not reside
directly above the 1: system for optimal metal-carbon interactions.

The sheets are tightly packed together with the nearest distance between them
being 2.45 A as measured by the shortest (H-H) distance. Their packing leads to infinite

cylindrical cavities approximately 2 A between the sheets.

44

Table 3.4. Selected Distances and Angles for Rb+(Dibenzo[a,c]phenazine)'
Distances are given in A.

 

Distances
Rb(1)-N(1) 3.110(2) N(16B)-Rb(l)-C(18B) 41 .97(6)
Rb(1)-N(16I) 3.150(2) N(1)-Rb(1)-C(21A) 7939(6)
Rb(1)-N(1A) 3.256(2) N(16I)-Rb(1)—C(21A) 145.18(6)
Rb(1)-N(l6B) 3.320(2) N(1A)-Rb(l)-C(21A) 4228(6)
Rb(1)-C(13I) 3.360(2) N(1)-Rb(1)-C(22) 22.73(5)
Rb(1)-C(1 8B) 3.402(3) N(16I)-Rb(1)-C(22) 9638(6)
Rb(1)-C(21A) 3.405(3) N(1A)-Rb(1)-C(22) 8037(5)
Rb(1)-C(22A) 3.425(2) N(16B)-Rb(1)-C(22) 174.01(5)
Rb(1)-C(151) 3.482(2) N(16B)-Rb(1)-C(21A) 8977(6)
Rb(1)-C(14I) 3.524(2) N(1)—Rb(1)-C(22A) 9504(6)
Rb(1)-C(22) 3.553(2) N(16I)-Rb(1)-C(22A) 164.03(5)
Rb(1)-C(17B) 3.562(2) N(1A)-Rb(l)-C(22A) 23.72(5)
N(16B)-Rb(1)-C(22A) 80.73(6)
Angles N(1)-Rb(l)-C(17B) 138.50(5)

N(1)-Rb(1)-N(16I) 8992(5) N(16I)-Rb(l)-C(17B) 7784(5)
N(l)-Rb(1)-N(1A) 8969(4) N(1A)-Rb(l)-C(l7B) 107.48(5)
N(16I)-Rb(1)-N(1A) 172.06(5) N(16B)-Rb(l)-C(17B) 2274(5)
N(1)-Rb(1)-N(16B) 159.78(5) N(1)-Rb(1)-C(151) 8437(5)
N(16I)-Rb(1)-N(r613) 8939(4) N(16I)-Rb(l)—C(151) 2321(5)
N(1A)-Rb(1)-N(16B) 9370(5) N(1A)-Rb(l)-C(151) 148.97(5)
N(l)-Rb(1)-C(131) 123.26(6) N(16B)-Rb(l)-C(151) 102.30(5)
N(16I)-Rb(l)-C(131) 5089(6) N(1)-Rb(1)-C(14I) 99.66(5)
N(16B)-Rb(1)-C(131) 7056(5) N(16I)-Rb(l)-C(14I) 4205(5)
N(1)-Rb(l)-C(18B) 117.84(6) N(1A)-Rb(1)-C(14I) 130.33(5)
N(16I)-Rb(l)-C(18B) 85.71(6) N(16B)-Rb(1)-C(14I) 9330(5)
N(1A)-Rb(1)-C(18B) 101.46(6) Rb(1)-N(l)-Rb(BB) 125.16(6)

 

 

 

3.5. Summary

The Na+(Dibenzo[a,c]phenazine) (MTHF)2, K*(Dibezo[a,c]phenazine) (2-
MTHF)2,3, Rb”(Dibenzo[a,c]phenazine)' salts exhibit remarkable structural differences
based on variations in the ionic radii of the metal. Many questions surrounding the topic
of how and why molecules assemble into a given crystal lattice remain unanswered. The

size of the alkali metal and its ability to coordinate to solvent no doubt plays an important

45

role in dictating the dimensionality of these structures. The sodium cation interacts with
only two radical anions forming simple dimers with the remaining two coordination sites
effectively “solvated.” In stark contrast the large solvent free rubidium cation can
interact with four radical anions to result in a two dimensional network.

Crystallizing in a weakly chelating and sterically demanding solvent such as
MTHF may lead to higher dimensional systems then solvents such as gyhne and diglyme
that tend to more effectively compete for the coordination sphere of the metal. The
potassium cation in the K+(diphenyquinoxaline)' (DME) salt was chelated to two
dimethoxyethane oxygens and two radical anions to form a one dimensional chain. In
contrast the metal in the K+(dibenzo[a,c]phenazine)‘(MTI-IF)2,3 salt could only interact
with three radical anions.

It is not completely certain to what extent the carbon fiamework participates in
the coordination environment of the metals. The metals do not reside directly above the
carbon fiamework to interact strongly with the 1: system as do the alkali metal salts of the
perylene radical anion. In that particular case C-Na contacts under 2.9 A were reported.
The strongest electrostatic interaction will occur on the nitrogen where most of the
negative charge is localized. In all the crystal structures I reported each nitrogen is
chelated directly to two alkali metal cations with M+-N-M+ bond angles near 120°. The
crystal structure for the diphenquuinoxaline salts had the metals in the sz N plane.

The structural motif of these systems provided only for antiferromagnetic
interactions. In the rubidium and potassium salts the paramagnetic centers were to far

apart to attain significant orbital overlap, giving rise to small Curie constants. However

46

the sodium salt possessed comparatively strong magnetic coupling due to the shorter
distance between the radical centers. Unfortunately, the greater solvation of the sodium
cation limited the dimensionality of the system to isolated dimers. This appears to
underscore the challenge of reducing cation solvation in the crystal lattice and at the same

time providing for short distances between radical anions.

47

Chapter 4

Experimental
4.1 Ligand Synthesis and Purification .................................... 49
4.2 Solvent Purification ..................................................... 49
4.3 Synthesis of the Radical Anion Salts ................................. 50
4.4 Crystallization of the Radical Anion Salts ........................... 50
4.5 X-ray analysis ........................................................... 51
4.6 EPR and ENDOR Analysis ........................................... 52
4.7 Magnetic Susceptibility Analysis .................................... 52

48

4.1 Ligand Synthesis and Purification

Dibenzo[a,c]phenazine was synthesized by adding an acetic acid solution of
phenanthroquinone to a refluxing acetic acid solution of o-phenylenediamine. The
product precipitated almost immediately following the addition. The solid was filtered
and purified by recrystallizing three times in CHzClz/CHC13. The bis-2-methoxyphenyl

quinoxaline was already prepared and purified by Lawrence P. Szaj ek.

4.2 Solvent Purification

Reagent grade tetrahydrofuran (300ml) was pre-dried by refluxing over NaK in
the presence of benzophenone for about 24 hours. About two thirds of the blue solution
was distilled and placed into a 300ml bottle fitted with a Teflon stopcock. The bottle was
placed on the vacuum line (10" torr) and three fieeze-pump-thaw cycles were carried out
before distilling the solution over NaK. More freeze-pump-thaw cycles were done until
the solution turned an aqua blue.

Anhydrous 2-Methyltetrahydrofuran (200ml) was poured into a bottle fitted with
a Teflon stopcock under a nitrogen atmosphere. The solution underwent multiple freeze-
pump-thaw cycles at 10" torr until the bubbling stopped. It was then distilled over NaK
where a final freeze-pump—thaw cycle was carried out.

Pentane (300ml) was pre-dried by boiling over Na for about Six hours. The
solution was distilled and placed in a 300mL bottle where it was freeze-pump-thawed

three times on a high vacuum line(10") torr. The solvent was distilled over NaK.

49

4.3 Synthesis of the Radical Anion Salts

M'(Dibenzo[a,c]phenazine)' (M = Na, K, Rb): Dibenzo[a,c]phenazine (40mg,
0.143mmol) was introduced into chamber 1 of the H-cell shown below, fitted with 3mm
EPR quartz tubes or an optical cell with a 1mm path length in place of one of the tubes.
In a helium dry box a slight excess of alkali metal was added to chamber two. The cell
was then attached to a vacuum line at 10'5 torr where the alkali metal was sublimed into a
shiny metal mirror by heating with a torch. Chamber 1 was submerged into liquid
nitrogen, into which dry THF (30ml) was distilled. The cell was warmed to dissolve the
ligand and then cooled down in a dry ice/isopropanol bath to -60°C. The THF solution
was added to the sodium metal mirror resulting in a red reaction mixture of the radical

anion salt. Most of the mirror was consumed.

Ground Glass
Stopper

3m 17'

Teflon Stopcock

   

 

 

Crystallization or
EPR Tubes

Chamber 1

 

Figure 4.1 A modified H cell.

4.4 Crystallization of the Radical Anion Salts.
M*(Dibenzo[a,c]phenazine)' (M = Na, K, Rb) The modified H cell in Figure 4.1
was fitted with two 18mm Pyrex tubes. Synthesis of the radical anion salt was carried out

in 30 ml of MTHF using the same procedure discussed above with 100mg of

50

dibenzo[a,c]phenazine and a small excess of metal. The red MTHF solution was poured
into the crystallization tube and about one quarter of the solvent removed through
evaporation. The solution was submerged in liquid nitrogen and about an equal volume
of pentane was distilled in. The Pyrex tube was sealed off with a torch. The solution was
allowed to thaw slowly by first completely thawing the pentane layer then the MTHF
layer. This was done to minimize the disturbance around the solvent interface. The two
solvents were allowed to diffuse for 2-3 weeks with black lustrous needles forming for
the sodium and potassium salts. Black blocks were formed of the rubidium salt.

Rb *(2,3 bis-(3-methoxyphenyl) quinoxaline) The exact same procedure as
above was used except 150 mg of 2,3 bis-(3-methoxyphenyl) quinoxaline was dissolved
in 30ml of THF. About ‘/4 of the solvent was evaporated before being capped with
pentane. The solvents were allowed to diffuse together for 1 week. Lustrous dark blue

needles resulted.

4.5 X-ray analysis.

The crystallization tubes were opened up in a nitrogen glove box and the solvent
was decanted off. The crystals were washed out with octane and poured onto a copper
block at about -30°C. A Single crystal was removed from the octane and coated with
viscous oil. It was then mounted on a glass fiber. The crystal structures were solved by
Rui Huang using direct methods (SHELXS-93) and refined by full matrix least-squares
procedures using TEXSAN. All atoms were refined anisotropically. The hydrogen atoms

were not refined, except for the Rb+(Dibenz[a,c]phenazine)' structure.

51

4.6 EPR and ENDOR Analysis.

EPR and ENDOR spectra were recorded on a Bruker ESP300E Spectrometer.
Typical parameters were: modulation frequency of 12.5kHz, Modulation amplitude
0.290gauss, Power = 2.5] mW, and receiver gain = 100000. All samples were prepared

in THF at a concentration near 0.001moler

4.7 Magnetic Susceptibility Analysis.

The temperature dependence on the magnetic susceptibility for the crushed
polycrystalline samples were determined with a Quantum Design MPMSZ SQUID
magnetometer. Crystals of the radical anions were placed inside a plastic bag in a helium
dry box (>1ppm of oxygen). The temperature dependence of the susceptibility was
measured over the temperature range 2-300K. The susceptibility of the plastic bag was
measured separately and its diamagnetism was subtracted from the signal. Pascal’s

constants were used to account for the diamagnetism within the sample.

52

Appendix

X-ray Data

53

 

 

88¢ ammo 8222 £85 23

335 6:6 :85 $35 :3

RR 3 E ”to new 3.686%”— 383885

as: 223 83a 83. 38:8 88833

a: mad 2 2 .82» Ease

: :38». E: E: 82% 83 k .o§_o>

6:33 $23.3. $33.8 593.8 3. .6

$53. a €352 535: $8». a < .3

538.3 533 €892 528.” < a

8.5 no.5 «RE «RE @580 3an

0583855 0583855 05:20:35.5 0582855 8298 Euro

3 m: m: m: 90363863869

$33 328 E 2.84 Ems? 33:53

5.26:5 5.33.8 +£Ez~£so ”aftermzszsmso osmdtaszsoemso 33388 336262
-3583 samtzé 3338.0. ”$52-3 3353.32 235.52.358.23 be

 

 

.83 2580ch Eran Babe 29mm 5.2 033—.

54

Table 2A Atomic Coordinates and Isotropic Displacement Parameters
for Rb+ (2,3 bis-(3-methoxyphenyl) quinoxaline) (THF).

 

 

 

x y 2 Wm)
Rb 1662(1) 218(1) 1856(1) 32(1)
N(1) 5318(2) 199(1) 1996(1) 21(1)
C(2) 5922(3) -393(1) 1963(1) 20(1)
C(3) 5129(3) -870(1) 2226(1) 25(1)
C(4) 5657(3) -1476(1) 2176(1) 29(1)
C(5) 6994(3) -1615(1) 1865(1) 27(1)
C(6) 7828(3) -1149(1) 1608(1) 23(1)
C(7) 7343(3) -533(1) 1658(1) 19(1)
N(8) 8255(2) -76(1) 1426(1) 20(1)
C(9) 7637(3) 504(1) 1466(1) 18(1)
C(10) 8746(3) 981(1) 1242(1) 19(1)
C(11) 9616(3) 851(1) 789(1) 21(1)
C(12) 10715(3) 1281(1) 587(1) 23(1)
O(13) 11602(2) 1198(1) 147(1) 32(1)
C(14) 11196(3) 682(1) -l64(l) 34(1)
C(15) 10982(3) 1835(1) 832(1) 28(1)
C(16) 10153(3) 1961(1) 1281(1) 27(1)
C(18) 9051(3) 1539(1) 1483(1) 24(1)
C(19) 6150(3) 634(1) 1720(1) 19(1)
C(20) 5329(3) 1248(1) 1719(1) 20(1)
C(21) 4482(3) 1451(1) 2151(1) 26(1)
C(22) 3778(3) 2034(1) 2161(1) 31(1)
C(23) 3917(3) 2422(1) 1750(1) 28(1)
C(24) 4705(3) 2216(1) 1314(1) 24(1)
O(25) 4744(2) 2627(1) 917(1) 35(1)
C(26) 5592(5) 2440(2) 468(1) 63(1)
C(27) 5371(3) 1631(1) 1290(1) 22(1)
0001) 2781(2) 430(1) 966(1) 43(1)
C(102) 4182(3) -328(1) 647(1) 37(1)
C(103) 4148(4) -850(2) 271(1) 47(1)
C(104) 3497(5) 4375(2) 586(2) 74(1)
C(105) 2638(4) -1080(1) 1031(1) 50(1)

 

55

 

Table 3A Atomic Coordinates and Isotropic Displacement Parameters
for Na+(Dibenzo[a,c]phenazine)’ (MTHF)2.

 

 

 

x y 2 Wm)
Na(l) 1019(1) 3138(1) 30(1) 49(1)
N(l) -345(2) 3674(2) -139(1) 38(1)
C(2) -805(2) 3975(2) 355(1) 36(1)
C(3) -806(2) 3470(2) 955(2) 42(1)
C(4) -381(2) 2640(3) 1000(2) 53(1)
C(5) -344(3) 2179(3) 1578(2) 70(1)
C(6) -717(3) 2528(4) 21 18(2) 78(2)
C(7) -1 152(3) 3305(4) 2068(2) 68(1)
C(8) -1215(2) 3804(3) 1493(2) 51(1)
C(9) -1696(2) 4610(3) 1440(2) 56(1)
C(10) -2143(3) 4974(4) 1961(2) 79(2)
ca 1) -2630(4) 5714(5) 1876(4) 1 10(3)
C(12) 2701(4) 3859(4) -1294(4) 103(2)
C(13) 2275(3) 4182(3) 472(3) 76(1)
C(14) 1750(2) 4936(2) -850(2) 50(1)
C(15) 1274(2) 5232(2) -306(2) 39(1)
N(16) 1300(2) 4719(2) 233(1) 44(1)
C(17) -881(2) 4941(2) -749(2) 42(1)
C(18) -912(3) 5400(3) -1346(2) 63(1)
C(19) -495(3) 5096(4) -1864(2) 73(1)
C(20) -22(3) 4318(4) -1817(2) 69(1)
C(21) 24(2) 3858(3) -1243(2) 55(1)
C(22) -408(2) 4155(2) -701(1) 40(1)
0(101) 886(3) 1868(2) -633(2) 89(1)
0(201) 1967(3) 2347(3) 616(2) 1 18(2)
C(103) 1327(6) 1818(7) -1735(4) 141(4)
C(104) 119(11) 906(10) -1291(8) 284(12)
C(106) 841(10) 510(10) -1251(7) 202(6)
C(105) 96(6) 1496(5) -768(4) 126(3)
C(202) 1773(7) 1221(7) 682(7) 160(4)
C(102) 1451(8) 1523(8) -1070(6) 170(5)
C(203) 2047(8) 938(8) 1269(9) 191(6)
C(205) 2543(12) 2240(31) 929(13) 484(27)
C(204) 2739(5) 1834(8) 1483(4) 138(4)
C(206) 2907(8) 3229(9) 1005(12) 25 1 (1 1)

 

56

 

Table 4A Atomic Coordinates and Isotropic Displacement Parameters
for K*(Dibenzo[a,c]phenazine)‘ (MTHF)2,3.

 

 

 

x y 2 WW)
K(l) 9623(5) 2903(1) 1 16(1) 34(1)
N( 1) 3227(1 8) 3070(4) 226(4) 21 (4)
N( 1 A) 3368(18) 5524(4) 138(4) 21(4)
N(lB) 11154(18) 4300(4) 242(4) 16(3)
N(l C) 1 337(20) -984(4) 2920(4) 27(4)
N( 1 D) 61 1 7(20) -241 (4) 2939(4) 29(4)
N(l E) 1 105(1 8) 1492(4) 3042(4) 25(4)
K(2) 4665(5) 4011(1) 194(1) 36(2)
C(2) 3940(26) 3058(5) -65(5) 29(5)
C(2A) 4112(26) 5285(5) -71(5) 28(5)
C(2B) 10397(25) 4156(5) -22(4) 17(5)
C(2C) 529(27) -954(5) 3185(6) 35(6)
C(2D) 5341 (26) -1 12(5) 3186(5) 30(5)
C(ZE) 311(24) 1272(5) 3267(5) 20(5)
K(3) 4482(5) 1151(1) 3015(1) 34(1)
C(3) 3176(24) 3313(5) -280(5) 23(5)
C(3A) 3469(26) 5280(5) -358(5) 27(5)
C(33) 10956(27) 431 1(5) -303(5) 32(5)
C(3C) l 150(28) -688(6) 3430(6) 39(6)
C(3D) 5997(25) -222(5) 3468(5) 29(5)
C(33) 860(24) 1282(5) 3561 (5) 21(5)
K(4) 5002(5) -1107(1) 3044(1) 31(1)
C(4) 3899(27) 3296(6) -578(5) 35(6)
C(4A) 4136(28) 5038(5) -614(5) 28(5)
C(43) 10239(27) 4154(6) -577(5) 32(5)
C(4C) 477(27) -692(5) 3714(5) 35(5)
C(41)) 5337(27) -63(6) 3739(5) 35(5)
C(4E) 190(25) 1044(5) 3789(5) 24(5)
K(S) -372(6) 34(1) 2975(1) 42(2)
C(5) 5249(24) 3032(5) -653(4) 23(5)
C(5A) 5502(28) 4768(6) -528(5) 35(6)
C(53) 8844(28) 3877(6) -563(5) 36(5)
C(5C) -1012(31) -957(6) 3787(5) 43(6)
C(SD) 3898(27) 207(6) 3734(5) 29(5)
C(SE) -1231(24) 759(5) 3719(5) 24(5)
K(6) 10035(5) 5164(1) 149(1) 36(1)
C(6) 6023(26) 2781(5) 421(5) 29(5)
C(6A) 6203(26) 4753(5) -244(5) 31(5)
C(63) 8122(25) 3751(5) -291(5) 28(5)
C(6C) -1581(28) -1225(6) 3561(5) 34(6)
C(61)) 3198(26) 332(5) 3447(5) 24(5)

 

57

 

Table 4A (Cont’d).

 

C(6E)
C(7)
C(7A)
C(7B)
C(7C)
C(7D)
C(7E)
N(8)
N(8A)
N(8B)
N(8C)
N(8D)
N(8E)
C(9)
C(9A)
C(9B)
C(9C)
C(9D)
C(9E)
C(10)
C(IOA)
C(IOB)
C(10C)
C(lOD)
C(IOE)
C(1 1)
C(1 1A)
C(1 lB)
C(1 1C)
C(1 1D)
C(1 1E)
C(12)
C(12A)
C(IZB)
C(12C)
C(12D)
C(12E)
C(13)
C(13A)
C(IBB)
C(13C)
C(131))
C(13E)

 

4934(29)
5474(24)
5590(26)
8922(26)
-924(27)
3838(22)
4 173(21)
6216(18)
6406(21)
8180(20)
4712(20)
3143(19)
4880(17)
5508(24)
5776(24)
8922(25)
-888(25)
3899(23)
4 139(26)
6343(25)
6617(28)
8208(24)
4579(26)
3109(24)
4931(24)
7959(27)
8213(27)
6573(25)
-3209(28)
1540(22)
-3501(27)
8739(27)
9188(25)
5777(24)
-3899(29)
731(22)
4352(26)
8107(26)
8633(25)
6538(26)
-3088(29)
1460(28)
-3641(27)

753(6)
2797(5)
5026(5)
3903(5)
4226(6)
186(5)
101 1(4)
2576(4)
5026(4)
3775(4)
4472(4)
298(4)
988(4)
2571(5)
5303(5)
3904(5)
4477(5)
139(5)
1244(5)
2337(5)
5365(6)
3807(5)
4759(5)
237(5)
1265(5)
2128(5)
5137(5)
3590(5)
4968(6)
455(5)
1043(5)
1906(5)
5223(5)
3499(5)
-221 1(6)
523(4)

1 101(5)
1905(5)
5546(5)
3663(5)
-2260(6)
379(6)
1363(6)

3434(5)
435(5)
49(5)
-6(5)
3264(5)
3183(4)
3186(4)
92(4)
268(4)
260(4)
3040(4)
2914(4)
2910(3)
367(4)
465(4)
507(5)
2784(5)
2645(4)
2707(5)
606(5)
737(5)
800(5)
2554(5)
2364(5)
2408(5)
532(5)
805(5)
803(5)
2604(6)
2355(4)
2361(5)
769(5)
1053(4)
1083(5)
2366(6)
2091(4)
2087(5)
1061(5)
1259(5)
1343(5)
2100(5)
181 1(5)
1844(5)

37(6)
21(5)
24(5)
25(5)
28(5)
15(4)
9(4)

20(4)
30(4)
27(4)
32(4)
22(4)
17(3)
19(4)
18(4)
21(4)
30(5)
15(4)
29(5)
26(5)
36(6)
15(4)
27(5)
20(5)
17(4)
30(5)
29(5)
27(5)
38(6)
12(4)
38(5)
34(5)
23(5)
21(5)
32(6)
14(4)
31(5)
29(5)
26(5)
31(5)
37(6)
36(5)
38(5)

 

58

 

Table 4A (Cont’d).

 

 

C(14)
C(14A)
C(14B)
C(14C)
C(14D)
C(14E)
C(15)

C(15A)
C(ISB)
C(15C)
C(15D)
C(15E)
C(16)

C(16A)
C(16B)
C(16C)
C(16D)
C(16E)
C(17)

C(17A)
C(17B)
C(17C)
C(17D)
C(17E)
C(18)

C(18A)
C(1813)
C(18C)
C(18D)
C(18E)
C(19)

C(19A)
C(19B)
C(19C)
C(19D)
C(19E)
C(20)

C(20A)
C(ZOB)
C(20C)
C(201))
C(20E)
C(21)

6545(26)
7045(28)
8132(24)
4532(27)
3077(26)
-1981(27)
5680(23)
6050(24)
8998(26)
-748(25)
3869(25)
4200(28)
3894(23)
4359(23)
10698(25)
1031(26)
5598(22)
554(26)
3169(25)
3603(25)
1 1612(26)
1945(24)
6427(24)
1433(27)
1494(27)
1976(25)
13127(28)
3635(27)
8075(24)
3083(28)
797(25)

1 173(24)
13823(26)
4194(30)
8705(27)
3749(31)
1593(23)
1897(24)
12959(25)
3196(25)
7800(24)
2797(22)
3254(23)

21 15(5)
5728(5)
3852(5)
-206l(5)
157(5)
1562(5)
2340(5)
5647(5)
3948(5)
4810(5)
99(5)
1527(6)
2521(5)
5851(5)
4158(5)
4614(5)
4 18(4)
1738(5)
2492(5)
6108(5)
4244(5)
4689(5)
-220(5)
1941(5)
2675(5)
6310(5)
4450(5)
4486(6)
438(5)
2142(5)
2938(5)
6274(5)
4598(5)
4 179(6)
-576(6)
2134(6)
3004(5)
6007(5)
4522(5)
4091(5)
487(5)
1938(5)
2773(5)

1 120(5)
1234(5)
1339(4)
2045(5)
1818(5)
1917(5)
903(4)

971(4)

1062(5)
2271(5)
2101(5)
2194(5)
962(4)

918(4)

1059(5)
2223(5)
21 12(4)
2259(5)
1237(5)
1 149(5)
1304(5)
1967(4)
1850(4)
2003(5)
1288(5)
1097(5)
1309(5)
1924(5)
1868(4)
2078(5)
1068(5)
812(5)

1039(5)
2129(5)
2148(5)
2371(6)
795(5)

581(5)

756(5)

2390(5)
2397(5)
2590(4)
720(4)

25(5)
35(6)
20(4)
26(5)
26(5)
29(6)
18(4)
21(5)
18(5)
24(5)
23(5)
35(6)
18(4)
22(4)
27(5)
27(5)
10(4)
26(5)
27(5)
29(5)
27(5)
22(5)
20(5)
34(5)
36(5)
29(5)
32(6)
39(6)
21(4)
34(5)
23(5)
22(5)
26(5)
42(6)
37(5)
44(6)
21(4)
22(5)
29(5)
24(5)
19(5)
15(4)
18(4)

 

59

 

 

Table 4A (Cont’d).

 

C(21A)
C(21B)
C(21C)
C(21D)
C(21E)
C(22)

C(22A)
C(22B)
C(22C)
C(221))
C(22E)
0(101)
C(102)
C(103)
C(104)
C(105)
C(106)
0(201)
C(202)
C(203)
C(204)
C(205)
C(206)
0(301)
C(302)
C(303)
C(304)
C(305)
C(306)
0(401)
C(402)
C(403)
C(404)
C(405)
C(406)

 

3532(23)
1 1301(23)
1595(25)
6240(24)
1208(25)
3976(22)
4178(25)
10474(24)
636(26)
5383(23)
372(23)

1 1073(21)
1 1709(62)
10620(40)
12281(39)
12830(63)
12106(32)
8745(13)
8076(40)
891 1(37)
7655(42)
7179(71)
7793(27)
5862(20)
6519(36)
5345(34)
7235(32)
7876(44)
6802(29)
1368(26)
2562(68)
2193(48)
2276(34)
1 160(66)
1302(48)

5800(5)
4288(5)
4308(5)
-226(5)
1716(5)
2821(4)
5540(5)
4171(5)
4262(5)
409(5)
1481(5)
2215(4)
2169(13)
2352(8)
1722(9)
1636(15)
1890(7)
5861(3)
5866(9)
561 1(8)
6308(9)
6480(16)
6181(5)
1848(4)
1847(8)
1693(7)
2306(7)
2406(10)
2158(6)
3303(6)
2927(16)
2868(10)
2694(7)
3035(16)
3448(1 1)

650(4)
772(5)
2436(5)
2397(5)
2526(5)
435(4)
404(5)
494(4)
2720(5)
2676(4)
2771(4)
460(4)
456(1 1)
-631(7)
489(8)
446(12)
92(6)

4 10(3)
412(8)
-628(7)
482(8)
440(14)
71(5)
3254(4)
3570(7)
3783(6)
3603(6)
3279(8)
3064(6)
3424(5)
3274(13)
3062(10)
3644(7)
3834(12)
3774(9)

22(5)
19(4)
21(5)
17(5)
20(5)
7(4)
24(5)
19(4)
24(5)
15(4)
16(4)
51(5)
166(17)
81(9)
92(10)
182(19)
59(7)
36(2)
95(9)
71(8)
97(1 1)
83(23)
33(5)
60(5)
67(8)
66(8)
56(7)
101(1 1)
48(6)
103(7)
164(23)
1 18(12)
70(8)
199(21)
121(12)

 

60

 

 

Table 5A Atomic Coordinates and Isotropic Displacement Parameters

for Rb+(Dibenzo[a,c]phenazine)'.

 

 

 

x y 2 U(CCI)
Rb(l) 1209(1) 1474(1) 2729(1) 33(1)
N(l) -469(1) 278(3) 2199(1) 22(1)
C(2) -648(1) 1230(3) 1735(1) 19(1)
C(3) -110(1) 1009(3) 1243(1) 20(1)
C(4) 551(2) -170(3) ' 1248(1) 25( 1)
C(5) 1094(2) -301(4) 797(1) 28(1)
C(6) 996(2) 73 8(4) 333(1) 29(1)
C(7) 342(2) 1862(3) 313(1) 24(1)
C(8) -225(1) 2031(3) 761(1) 20(1)
C(9) -920(2) 3206(3) 740(1) 20(1)
C(10) -1 124(2) 4123(4) 250(1) 27(1)
C(11) -1 802(2) 5175(3) 230(1) 30(1)
C(12) -2311(2) 5376(3) 701(1) 29(1)
C(13) -2139(2) 4509(3) 1189(1) 24(1)
C(14) -1445(1) 3402(3) 1218(1) 19(1)
C(15) -1296(1) 2424(3) 1725(1) 18(1)
N(16) -1808(1) 2716(3) 2178(1) 22(1)
C(17) -l635(2) 1777(3) 2650(1) 21(1)
C(18) -2119(2) 2015(3) 3140(1) 26(1)
C(19) -1956(2) 1103(3) 3624(1) 28(1)
C(20) -1300(2) -36(3) 3640(1) 28(1)
C(21) -806(2) -288(3) 3165(1) 24(1)
C(22) -971(2) 586(3) 2661(1) 21(1)

 

61

 

 

10.

11.

12.

13.

14.

References

. Miller, J. S.; Epstein, A. J. Angew. Chem. Int. Eng]. 1994, 33, 385-415.

Dougherty, D. A. Acc. Chem. Res. 1991, 24, 88-94.
Hoffmann, R.; Van Dire, G. W.; Zeiss, G. D. J. Am. Chem. Soc. 1968, 90, 1485-1499

(a) Hoffmann, R. Acc. Chem. Res. 1971, 4, 1-9. (b) Hoffmann, R. J. Am. Chem. Soc.
1968, 90, 1475-1485-1499.

Ovchinnikov, A. A. Theor. Chim Acta 1978, 47, 297-301

(a) Borden, W. T.; Davidson, E. R.; Feller, D.; Kato, S. D.; Morokuma, K J. Am.
Chem. Soc. 1983, 105, 1791-1795. (b) Borden, W. T. ; Fort, R.C., Jr.; Hrovat, D. A.;
Lahti, P. M. J. Am. Chem. Soc. 1992, 114, 7549-7552.

Clites, J. A.; Dougherty, D. A.; Jacobs, S. J.; Murray, M.; Shultz, D. A.; Silverman, S.
K.; West, A. P. Mol. Cryst. Liq. Cryst. 1993, 232, 289-304.

Dougherty, D. A.; Silverman, S. K.; West, A. P., Jr. J. Am. Chem. Soc. 1996, 118,
1452-1463.

McConnell, H. M. J. Chem. Phys. 1963, 39, 1910.
(a) Iwamura, H.; Izuoka, A.; Sugawara, T. J. Am Chem. Soc. 1985, 107, 1786-1787.

(a) Buchachenko, A.L. Usp. Khim. 1990, 59, 529-546. (b) Epstein, A. J .; Miller, J. S.;
Reiff, W. M. Chem. Rev. 1988, 88, 201-235. (c) Gatteschi, D.; Laugier, J.; Rey,
P.; Zanchini, C. Inorg. Chem. 1987, 26, 938-943. ((1) Caneschi, A.; Gatteschi, D.;
Grand, A.; Laugier, J.; Pardi, L.; Rey, P. Inorg. Chem. 1988, 27, 1031-1035 (e)
Gelman, A. B.; Ikorsky, V. N.; Ovcharenko, V. I. Zh. Strukt. Khim. 1989, 30, 142-
145.

Malinouskaya, S. A.; Musin, N. R.; Schastnev, P. V. Inorg. Chem. 1992, 31 4118-
4121.

Bu’Lock, J.D.; Harley-Mason, J. J. Chem. Soc. 1951, 2248-2253

Conklin, B. J.; Lange, C. W.; Pierpont, C. G. Inorg Chem. 1994, 33, 1276-1283. (b)
McGarvey, B. R.; Ozarowski, A.; Peppe, C.; Tuck, D. G. J. Am. Chem. Soc. 1991,
113, 3288-3293.

62

15. (a) Fox, G. A.; Pierpont, C. G. Inorg. Chem. 1992, 31, 3718. (b) Abakumov, G. A.;
Bubnov, M. P.; Cherkasov, V. K.; Ellert, O.G.; Rakitin, U. V.; Saf’yanov, U. N;
Struchkov, Y. T.; Zakharov, L. N.;. Isv. Akad. Nauk SSSR 1992, 2315.

16. Gardier, M. G.; Hanson, G. R.; Henderson, M. J .; Lee, F. C.; Raston, C. L. Inorg
Chem. 1994, 33, 2456-2461.

17 Hirota, N. J. Am. Chem. Soc. 1967, 89, 32-41.

18. Book, H.; Fenske, D.; Goosmann, H.; Hemnann, H. Angew. Chem. Int. Ed Engl.
1988, 27, 1067-1069.

19. Bakulin, A.; Gopalan, R; Jackson, J .E.; Stoudt, S.J. Inorg. Chem. 1996, 35, 6614-
6621.

20. Huang, R.; Jackson, J. E.; Kahr, B.; Misiolek, A. J. Am. Chem. Soc., Chem.
Commun., 1996, 2119-2120.

21. Dye, J. L.; Huang, R.H.; Huang, S. Z. Ichimura, A. S.; Jackson, J. E.; Szajek, P. L.;
Wagner, M. J.; Xie, Q. J. Phys. Chem B. 1998, 102, 11029-11034.

22. Glarum, S.H.; Haddon, R. C.; Hebard, A. F.; Murphy, D. W.; Palstra, T. T. M.;
Rosseinsky, M. J. Nature 1991, 350, 600-601.

23. Dabbagh, G.; Eick, R.H.; Fleming, R. M.; Glarum, S.H.; Haddon, R. C.; Hebard, A.
F.; Kortan, A. R.; Makhija, A. V.; Miller, B.; Muller, A. J.; Murphy, D. W.;
Rosamilia, J. M.; Rosseinsky, M. J.; Tycko, R.; Zaharak, S. M. Nature 1991, 350,
320-322.

24. Diederich, F.; Holczer, K.; Huang, S.-M.; Kaner, R.; Mihaly, L.; Stephens, P. W.
Nature 1991, 351, 632-634.

25. Coustel, N.; Cox, D. E.; Fischer, J. E.; Kycia, S.; McCauley Jr., J. P.; McGhie, A,
R.; Romanow, W. J .; Smith, A. B.; Zhou, 0. Nature 1991, 351, 462-464.

26. Chen, C.-C.; Kelty, S. P.; Lieber, C. M. Nature 1991, 352, 223-224.

27. Ebbesen, T. W.; Kubo, Y.; Kuroshima, S.; Mizuki, J .; Saito, S.; Tanigaki, K. Nature
1991, 353, 222-223.

28. Atherton, N. M.; Weissman, S. I. J. Am. Chem. Soc. 1961, 83, 1330-1334.

29. Herold, B. J .; Neiva Correia, A. F.; dos Santos Veiga, J. J. Am. Chem. Soc. 1965, 87
2661-2665.

63

 

30. Fischer, D.; Gerson, F.; Merstetter, P.; Mlynek, C. J. Am. Chem. Soc. 1998, 120,
4815-4824.

31. Andreas, J.; Bock, H.; Nather, C.; Ruppert, K. Helv. Chim Acta 1994, 77, 1505-1519.

32. For reviews on alkali metal organic compounds see: (a) Arad, C.; Bock, H.; Gobel,
1.; Havlas, Z.; Meuret, J .; Herrmann, H.-F.; Nather, C.; Nick, 8.; Rauschenback, A.;
Ruppert, K.; Seitz.; Solouki, B.;Vaupe1, T. Angew. Chem, Int. Ed. Engl. 1992, 31,
550-581 . (b) Schade, C.; Schleyer, P. v. R. Adv. Organomet. Chem. 1987, 27, 169-
186 .

33. Bock, H.; Hauk, T.; Havlas, Z.; Nather, C. Organomet. 1998, 17, 4707-4715.

34. Bock, H.; Ruppert, K. Inorg. Chem. 1992, 31, 5094-5099.

35. Habich A.; Hausser, K., A.; Franzen, Z. Naturforsch., Teil A 1961, 16, 836-845.

36. Maki, A. H.; Stone, E. W. J. Chem. Phys. 1963, 39, 1635-1642.

37. Sevenster, A. L.; Tabrer, B. J. J. Org. Mag. Res. 1984, 22, 521-526.

38. Carter, H. V.; McClelland, B. J.; Warhurst, E. Trans. Faraday Soc. 1960, 56, 455-
458.

39. Buschov, K. H. J.; Dieleman, J.; Hoijtink, G. J. J. Chem. Phys. 1965, 42, 1993-1999

40. Bleaney, B.; Bowers, K. D. Proc. R. Soc. London, Ser. A. 1952, 214, 451.