Proton contact shifts of some magnetically anomalous cobalt (11) complexes

MISSING IMAGE

Material Information

Title:
Proton contact shifts of some magnetically anomalous cobalt (11) complexes
Physical Description:
vii, 73 leaves : ill. ; 28 cm.
Language:
English
Creator:
Vaughn, Joseph Benjamin, 1942-
Publication Date:

Subjects

Subjects / Keywords:
Cobalt -- Magnetic properties   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 70-72).
Statement of Responsibility:
by Joseph Benjamin Vaughn, Jr.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000065652
notis - AAH0866
oclc - 04361972
System ID:
AA00003921:00001


This item is only available as the following downloads:


Full Text















PROTON CONTACT SHIFTS OF SOME MAGNETICALLY
ANOMALOUS COBALT(II) COMPLEXES







By

JOSEPH BENJAMIN VAUGHN, JR.


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





UNIVERSITY OF FLORIDA


1977
















TABLE OF CONTENTS

Page

LIST OF TABLES iii

LIST OF FIGURES iv

ABSTRACT v

INTRODUCTION 1

EXPERIMENTAL 14
Preparations 14
Elemental Analysis 24
Mass Spectrum Analysis 24
Spectrometers 24

RESULTS 26

DISCUSSION 34

APPENDIX 64

BIBLIOGRAPHY 70

BIOGRAPHICAL SKETCH 73















LIST OF TABLES


Table Page

I Selected Infrared Absorptions and
Assignments for the Co(II) and Fe(II)
Compounds Studied .27

II Magnetic Moments and g Values for the
Co(II) and Fe(II) Compounds Studied 30

III Observed and Relative Isotropic Shifts
and Half-Height Peak Widths for Cobalt(II)
Compounds 32

IV INDO Calculated and Experimental Proton
Spin Densities for [Co(GdH)-3+] 39

V INDO Calculated and Experimental Proton
Spin Densities for Cationic and Anionic
Radicals 42

VI Selected INDO Molecular Orbital Symmetries
and Energies of Ligand Radicals 53

VII Proton Chemical Shifts of the Substituted
2-Pyridinalphenylimine Ligands 65

VIII Proton Chemical Shifts of the Substituted
Tris(2-pyridinalphenylimine) Cobalt(II)
Complexes .66

IX Effective Magnetic Moments of the Sub-
stituted Tris(2-pyridinalphenylimine)
Cobalt(II) Compounds 68















LIST OF FIGURES

Figure Page

1. Ligands whose cobalt(II) complexes were
successfully prepared 10

2. Ligands whose cobalt(II) complexes could
not be prepared 11

3. Possible spin transfer mechanisms 37

4. Possible Oh--D correlation diagrams and
electron confi rations for a d7 ion under
D3 symmetry 50

5. Representation of ligand T MO and metal ion d
orbital parity upon rotation about the C2 axes. 52

6. Representation of spin polarization effects
in a fragment of a [M(GdH)r+] complex 56

7. Possible Oh--D2d correlation diagrams and
electron configurations for a d7 ion under
D2d symmetry 60











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



PROTON CONTACT SHIFTS OF SOME
MAGNETICALLY ANOMALOUS COBALT(II) COMPLEXES

By
Joseph Benjamin Vaughn, Jr.

August 1977
Chairman: R. Carl Stoufer
Major Department: Chemistry

A series of cobalt(II) complexes known to lie near the
E( 2G)-- T1( F) cross-over point was prepared. The first

reported proton contact shift investigation of such magnetic-

ally anomalous cobalt(II) complexes was carried out in order

to obtain information about the metal--ligand bonding proper-

ties. These complexes contain ligands with the a-diimine

linkage (glyoxaldihydrazone, GdH; biacetyldihydrazone, BdH;

biacetylbismethylimine, BMI; 2,6-pyridinediacetyldihydrazone,

2,6-PdAdH), which provides relatively large ligand field

splitting. The symmetry of the complexes containing these

symmetric, bidentate ligands is D3 and that of the terdentate

ligand complex is D2d. Under D3 symmetry 2Eg(Oh) transforms

as 2E(D3) and 4Tl(Oh) transforms as 4A(D3) and E(D3). Under
2 2 2
D2d symmetry Eg(Oh) transforms as Al(D2d) and 2Bl(D2d) and
2 4 4 4
Bl(D2d) and Tl(Oh) transforms as B2(D2d) and E(D2d).

Because these terms are isotropic, the dipolar shift








contribution to the contact shifts is neglected. The magni-

tudes of these shifts (100 ppm and less) are typical of those

observed for other complexes.

Spin densities were calculated using a modified INDO

program. The approach commonly used in such calculations is

to treat a single ligand as a radical cation or anion, using

only the "T" symmetry atomic orbitals in some instances, pre-

sumably due to the limited capacity of the available programs.

In contrast, the program used for these calculations is capable

of handling the complete valence basis set (ninety-nine

atomic orbitals) for [Co(GdH)3 ], the simplest complex in

the series. That is, metal penultimate d and ultimate s and

p orbitals and all the ligand valence orbitals were included.

But, meaningful results for even this simple complex were

not obtained, presumably because the number of unpaired spins

(three for high-spin complexes and one for low-spin complexes)

is sufficiently small that the spin densities calculated over

ninety-nine basis orbitals represent only "noise." Thus it

is concluded that such calculations do not yet have the degree

of sophistication necessary to provide meaningful results
++
even for such a simple complex as 1Co(GdH)3 ]. Consequently

each ligand was treated both as a radical anion and a radical

cation, as is commonly done, including the complete valence

basis set. Evaluation of these INDO calculated and experi-

mentally determined spin densities leads to the following con-

clusions.

(1) Because the best fit was obtained for the GdH+1








BdH+1 and BdH1, BMI- and no fit was obtained for 2,6-

PdAdH, it might be concluded that the spin transfer mechanism

is:

(a) L M in the complex [Co(GdH) ].

(b) L M and/or M L in the complex [Co(BdH)4].

(c) M L in the complex [Co(BMI)34.

(d) Undetermined in the complex [Co(2,6-PdAdH)214].
However, such conclusions are somewhat tenuous because of the

possibility of spin polarization, which adds further ambigu-

ity by placing opposing spins simultaneously at the same

position in the ligand.

(2) More than one molecular orbital is involved in the

transfer mechanism in each complex, making definitive conclu-

sions somewhat more difficult. Comparison of the INDO molec-

ular orbital calculations and the experimental data provides

evidence that spin transfer onto the hydrazone amine protons

occurs via a a molecular orbital with an opposing u molecular

orbital contribution, in keeping with the usual assertion

that d7--p7 back bonding is involved in these complexes.

(3) There is increasing lack of agreement within the

series of ligands, between INDO calculated spin densities and

the experimentally determined spin densities,which corresponds

to increasing contribution of the low-spin configuration to

the ground state of the complex. This trend is taken as an

indication that there is a contribution to the contact shifts

attributable to the low-spin configuration, the first such

contribution reported for an octahedrally coordinated d ion.


vii















INTRODUCTION


During the past century or so a great deal of research

has been directed toward the investigation of transition

metal coordination compounds. A significant portion of this

work has involved octahedral cobalt(II) compounds, the ma-

jority of which are high-spin, Veff = 4.7 5.2 BM, having

a 4Tg( F) ground term. There are relatively few octa-
hedral cobalt(II) coordination compounds which are low-spin,

=eff = 1.8 2.2 BM, having a 2E (2G) ground term. Since
g
1956 (40), there have been reported an increasing number of

magnetically anomalous cobalt(II) coordination compounds,

i.e., compounds which exhibit both room temperature magnetic

moments intermediate between high-spin and low-spin moments

and anomalous Curie-Weiss behavior.

Variable temperature magnetic studies led Stoufer and

coworkers (41) to postulate that these anomalous compounds

lie near the "cross-over" point and that Boltzmann distri-

bution over high-spin and low-spin states is attained, i.e.,

AE(2E --4 T1g) kT. They noted that the majority of these

compounds contain ligands, with an a-diimine linkage

(-N=C-C=N-), which provide relatively large ligand field

splitting due in part to their q-acid capabilities. Research

interest in these compounds still exists as evidenced by








the recent work of Simmons and Wilson (39), who investigated

solution and solid state magnetic and spin lifetimes of some

six-coordinate bis(N-R-2,6-pyridinedicarboxaldimiine) cobalt(II)

complexes. For the R=NH(CH3) derivative, they report solution

rates for E(s) 4T(hs) of kI = 3.2 x 106 s-1 and k- =
6 -1 k-1 2
9.1 x 106 s- with corresponding spin state lifetimes of T(2E)

= 3.1 x 10-7 s and ( T) = 1.1 x 10-7 s.

Information about bonding properties in paramagnetic

transition metal complexes as inferred from unpaired spin

distribution is traditionally obtained by electron spin

resonance techniques, where the isotropic couplings are mani-

fest as nuclear multiple structure. The sign of the cou-

pling constants cannot be determined directly from the ESR

experiment. However, the sign is obtained directly from

the NMR contact shift, and thus the sign of the spin density

is gotten directly. Because it is the sign of the spin den-

sity which gives the most useful information about bonding,

the major advantage is with the NMR technique. Furthermore,

the spectral complexity is greater in the ESR experiment.

That is, for a system with n coupling constants, ESR spectra

will exhibit 2n lines; whereas NMR spectra will exhibit only

n lines. However, the two techniques are more complementary

than competitive, since it is rare that both can be usefully

employed for the same system. This is because of the re-

quirements of very short (! 10-1 s) electron-spin relaxation

time, Tle, for observing NMR spectra and long Tle's (> 10-9 s)

for observing ESR spectra.









In their review article, Lewis and Morgan (26) make

a general distinction between species containing one un-

paired electron (s = ) and those with more than one (s > 1).

For S = systems, electron-spin relaxation is very ineffi-

cient, yielding well-resolved ESR spectra. Exceptions can

occur in systems possessing low-lying excited states, which

reduce Tle considerably. For systems with S > 1 the presence

of orbital degeneracy in the ground state, low-lying excited

states and/or modulation in solution of zero field splitting

can produce rapid electron-spin relaxation and therefore

well-resolved NMR spectra.

Because Tle is typically long for complexes in which

S = it has been predicted (42) that low-spin, 2E octa-

hedral cobalt(II) complexes will not yield observable NMR

contact shift spectra. And to date no NMR data have been

reported for any low-spin, octahedral, d7 ion (42). This is

due, in part, to the fact that there are no proton-containing,

low-spin, octahedral cobalt(II) complexes known. It may be

that the Eg species present in anomalous cobalt(II) com-

plexes investigated in this work will have Tie shortened

sufficiently to observe NMR spectra by the presence of the

low-lying T1g excited state near the cross-over point videe

infra).

Nuclear magnetic resonance spectra of paramagnetic

transition metal complexes typically exhibit very large

(200 ppm and larger for protons) contact shifts. Contact

shifts are measured relative to a diamagnetic standard,









usually the free ligand or, better yet, a homologous dia-

magnetic metal complex. And, the shifts may be related to

the transfer of spin either to or from the ligand. The

ligand is usually a closed shell diamagnetic molecule, where-

as the metal ion may possess some unpaired electrons in in-

completely occupied d or f orbitals. If any of the metal

orbitals containing the unpaired electron(s) is involved in

M-L bonding, the unpaired electron(s) may be transferred to

a new molecular orbital (MO) of the complex. Consequently

the unpaired spin may be partially delocalized onto the

ligand. McConnell and Chestnut (28) showed that for ligands

possessing nuclei with I # 0, this delocalized spin may

interact with these nuclei via the Fermi contact interaction

(8). For a complex possessing m spins, the Hamiltonian for

the Fermi interaction is given by R = A S"mI. A is the

hyperfine coupling constant for the contact interaction

between the electron spin, S, and the nuclear spin, I, and

is given by (36)


A = e gN 4e N -1
where p(r) is the spin density operator at the nucleus, 4 is

the electronic wave function for the molecule, is the

time averaged value of the z component of the electron spin,

and the other symbols have their usual meanings. Thus Fermi

coupling constants arise only from spin in s orbitals, and

A is a linear function of that spin density. The variation

in coupling constants for the various nuclei may lead to









the determination of the distribution of unpaired spin density

in the ligand, and thus to the characterization of the MO

containing the unpaired spin. Consequently a better under-

standing of electronic structure in transition metal complexes

may be achieved.

Both the sign and magnitude of A can be determined from

contact shift spectra. A is related to the contact shift

Av by (10,28)

2 2
Av Ag S(S + 1) (2)
AV e e (2)
v gN N 3 kT

if the Curie law is obeyed. It is clear from equations (1)

and (2) that a downfield shift in resonance position corre-

sponds to a positive value for A and therefore to positive

spin on the corresponding nucleus.

La Mar has outlined the three general areas of considera-

tion for the elucidation of the observed coupling constants

in terms of the M-L bonding: "(i) The relationship between

the observed coupling constant for a given nucleus and the

probability (density) of finding the unpaired spin in an

atomic orbital at the nucleus of interest and/or at a neigh-

boring nucleus, particularly as to how the coupling constant

reflects the type (a or 7) of ligand molecular orbital con-

taining the spin; (ii) the evaluation of the mechanisms by which

the unpaired spin on the metal is transferred to the ligand;

and (iii) the qualitative and, where possible, the quantitative

relationships between the fraction of total spin transferred

and the M-L covalency"(23:87).









There are several possible problems which may introduce

uncertainties into the interpretation of contact shift data.

Three major ones are: (i) Separation of the dipolar or

pseudocontact shifts from the measured isotropic shifts. These

dipolar shifts occur only in complexes which are magnetically

anisotropic, and are produced by the dipolar interaction

between the electron's magnetic moment, W, and the nuclear

magnetic moment h yN I. The dipolar shift averages to zero

by tumbling in solution unless the electronic magnetic moment

is anisotropic. It can, under appropriate conditions, be

separated from the isotropic shift by (36)


Adip (3 cos 2 1
--- (g3 g1) (3)
r

where D is a positive quantity whose value depends on the

temperature and various relaxation times (17,21,22,29) and the

other symbols have their usual meaning. Due to uncertainties

in D, most published calculations of absolute dipolar shifts

are open to question (7). However, if the g value is iso-

tropic or nearly so (gll gl = 0) dipolar shifts can be neg-

lected. (ii) Electron correlation effects can cause spin

density to be induced on L in the M-L fragment, even if the

metal orbital containing the spin is symmetry restricted

from interacting with any orbital on L. These effects can be

accounted for by use of an appropriate quantum mechanical

model for calculated spin densities. (iii) The possibility

of spin-polarization effects can also cause uncertainties in

determining the mechanism of spin delocalization. Again the









appropriate choice of quantum mechanical model can help over-

come this problem.

The mechanism by which the unpaired spin is delocalized

onto the ligand is related to the symmetry of the MO containing

it. These MO's are predominantly metal d orbitals with smaller

contributions from ligand orbitals which may be filled i or-

titals, empty I orbitals, a orbitals, or a mixture of all

three. Symmetry considerations sometimes restrict the possi-

bilities, in complexes of high symmetry for instance. But -

in general there is no easy method available for sorting out

the various contributions. Quantitative evaluation of M-L

covalency is no simple task, and the present development of

the evaluation of spin-delocalization mechanisms does not

permit a definitive presentation (23). "Thus the theoretical

complexities of relating transferred spin density by any

single mechanism to covalency, plus the practical problems

of multiple delocalization mechanisms, suggest that caution

should be exercised in any quantitative interpretations"

(23:120).

The ideal procedure for theoretical analysis would be

a complete, many-electron calculation on the entire complex

with a concomitant breakdown of the isotropic shifts into

their component parts. This would be a formidable undertaking

even for a simple complex. The nonideal approach commonly

employed is to treat a single ligand molecule as a free radical

anion or free radical cation depending on the delocalization

mechanism involved. The spin densities are then calculated









from the MO's containing the unpaired electrons and a corre-

lation is attempted between the calculated spin densities

and the relative spin densities gotten from the observed

contact shifts videe infra). In some cases, experimental

results satisfactorily fit calculated spin densities assuming

only one ligand MO is involved in the delocalization. In

other cases, it appears more than one ligand MO is involved.

In any case, this extreme approximation leaves much to be

desired regarding any quantitative conclusions. One obvious

objection to this approach is that the metal orbital contri-

butions are excluded in forming the MO's. The metal is con-

sidered to be simply an electron source or sink for the

ligand. This obviously assumes the relative energies of the

ligand MO's are unaltered by interaction with the metal.

Certainly this situation is seldom achieved.

There are several theoretical methods available for

treating spin-density calculations in metal complexes. In

this work, the INDO method was chosen because it was readily

available and it is more realistic than the others for the

following reasons (23,34,36):

(i) It is an all valence electron method.

(ii) It is capable of calculating negative spin densities

arising from electron correlation since it uses unrestricted

wave functions. The extended Hickel method, for example, can

calculate only zero or positive densities.

(iii) It accounts for spin polarization by including one-

center exchange integrals.








On the negative side, INDO calculations have "black-box"

aspects which tend to hinder a qualitative understanding of

the resultant calculations.

To date no calculations reported have included the metal

ion orbitals and all the ligand valence orbitals in the basis

set. This is because the present theoretical development is

simply not capable of handling such a large basis set. However,

one complex investigated in this work [Co(GdH) ] does have

a sufficiently small basis set (ninety-nine atomic orbitals)

to include the metal ion orbitals and all the ligand orbitals.

The purpose of the present work is to use NMR proton

contact shifts to investigate the metal-ligand (M-L) bonding

properties of a series of anomalous cobalt(II) compounds.

The bis(2,6-PdAdH) cobalt(II) complex and tris cobalt(II)

complexes of the other ligands in Figure 1 were prepared for

this purpose. Ligands with X = -CN and -NO2 were also pre-

pared, however, they did not coordinate to the metal ion. The

"tris" cobalt(II) complexes of GdH, BdH, BMI and the bis-

(2,6-PdAdH) cobalt(II) complex have been prepared previously

and are known to lie near the cross-over point with a E
g
ground term (5,10,37,41,44). The 2-pyridinalphenylimine

complexes were prepared with the expectation that increasing

the 7-withdrawing capability of the substituent would give a

homologous series of complexes which would exhibit magnetic

moments over the range of high-spin and low-spin values,

the ultimate goal for this series being to relate contact

shift changes to the Hammett acidity function. To accomplish

this goal the ligands in Figure 2 were prepared or their prep-

aration attempted.











H\ H
C-C
H2N-N N- NH2

Glyoxaldihydrazone
(GdH)


H3C / CH3
C- C
H C-N \N-CH3

Biacetylbismethylimine
(BMI)










H

SN-NH2

2-Pyridinalhydrazone
(2-PAH)


H 3C C
HNC
H21'uL-Nd


CH3
-C N
N NH2


Biacetyldihydrazone
(BdH)


C
H2N-N'


CH3
CN_ NH2


2,6-Pyridinediacetyl-
dihydrazone
(2,6-PdAdH)






X = H, CH3, NH2, C1

L '- -1


CN ) X

2-Pyridinalphenylimine
(2-PPI)


Figure 1. Ligands whose cobalt(II) complexes were success-
fully prepared





















H
\r /
4:-C\


H3 C-N


N- CH3


H3C\ C -C/ CH3

X N N X


X = H, NO2


X = H, NO2


R, R
SC-C H
X- N-N N N X


R = H, CH3; X = H, NO2





Figure 2. Ligands whose cobalt(II) complexes could not
be prepared

















H
N N

N N


HX
cX


X = H, OH, OCH3,


NO2


CH3


X = H, CH3, NO2











H3C' CH3
N C
X-- N N x
I i
H H


X = H, Br, CN









The rationale underlying this choice of ligands was to

prepare a homologous series of ligands (containing the a-di-

imine linkage) whose cobalt(II) complexes should lie near

the cross-over point. The substituents, X, have known o-and

n-acid effects. It is known that the r-acid character of

the ligand is important in the formation of the octahedral,

low-spin cobalt(II) complexes. Therefore, it was expected

that a series of ligands with increasing 7-acid effects would

produce a homologous series of complexes some of whose

members have a 4Tg ground term and some of whose members

have a 2E ground term (i.e., span both high-spin and low-

spin sides of the cross-over point). And, it was expected

that the resultant ri-acid capability could be related to the

change in proton contact shifts.















EXPERIMENTAL


In the preparations described below, melting points of

previously prepared compounds are all within experimental

error of those reported. The yields for all the compounds

were rather poor, and varied considerably upon repeating the

preparations. These poor yields were a result of washing or

recrystallizing the soluble product many times in order to

obtain a pure product. Consequently yields are not reported.

For several compounds, mass spectra parent peaks and their

relative intensities are reported in lieu of elemental analy-

sis.



Preparations



Biacetyldihydrazone, BdH. BdH was prepared by the method

of Fisher (10). M. p., 1590C.

Anal.: Calcd. for C4H10N4: C, 42.11; H, 8.77; N, 49.12.

Found: C, 42.23; H, 8.91; N, 49.03.



Glyoxaldihydrazone, GdH. GdH was prepared by the method

of Fisher (9). However, better yields were obtained using

only 40 ml of hydrazine rather than 75 ml. M. p., 990C.

Anal.: Calcd. for C2H6N4: C, 27.91; H, 6.98; N, 65.12.

Found: C, 27.78; H, 7.11; N, 64.85.









2,6-Pyridinediacetyldihydrazone, 2,6-PdAdH. A solution

of 50 ml of methanol and 4.0 ml (0.12 mole) of hydrazine was

prepared and placed in an ice bath. Another solution of

8.1380 g (0.0498 mole) of 2,6-diacetylpyridine dissolved in

50 ml hot methanol was added dropwise with stirring. The

white powdery product was washed with cold methanol and dried

in vacuo over P4 010 M. p., 1790C.

Anal.: Calcd. for C9H13N5: C, 56.54; H, 6.81; N, 36.65.

Found: C, 56.16; H, 6.98; N, 36.47.



2-Pyridinalhydrazone, 2-PAH. To 3.0 ml (0.09 mole) of

hydrazine in an ice bath, 7.2 ml (0.06 mole) of 2-pyridine-

caroxaldehyde (2-PCA) was added dropwise with stirring. The

golden yellow solution was "dried" in vacuo over P4010 leav-

ing a golden yellow oil. B. p., 1810C.

Anal.: Calcd. for C6H7N3: Mol. wt., 121 d.

Found: Mol. wt., 121 d (46%).



2-Pyridinalphenylimine, 2-PPI. To a solution of 9.1 ml

(0.1 mole) aniline in 10 ml hexane was added dropwise with

stirring 10.7 ml (0.1 mole) of 2-PCA. The bright yellow

powdery precipitate isolated was recrystallized from hot

hexane and dried in vacuo over P4010. M. p., 400C.

Anal.: Calcd. for C12H10N2: Mol. wt., 182 d.

Found: Mol. wt., 182 d (14%).









2-Pyridinal-p-erethylphenylimine, 2-PMI. A solution of

10.7155 g (0.1 mole) of p-aminotoluene dissolved in 50 ml

hexane was prepared by heating on a steam bath, and 10.7 ml

(0.1 mole) of 2-PCA was added slowly with stirring. This

yellow solution was allowed to cool to ambient temperature.

The yellow needle-like crystals which precipitated were

recrystallized from hot hexane and dried in vacuo over P4010

M. p., 550C.

Anal.: Calcd. for C13H12N2: Mol. wt., 196 d.

Found: Mol. wt., 196 d (69%).



2-Pyridinal-p-aminophenylimine, 2-PAI. A solution of

10.8982 g (0.1 mole) of p-phenylenediamine dissolved in

100 ml hot deionized water was prepared. One gram of sodium

hydrosulfite was added to this dark brown solution, and some

black particles were removed by filtration. To this beige

solution, 10.7 ml (0.1 mole) of 2-PCA was added dropwise

with vigorous stirring. The gold powdery precipitate prod-

uced was washed with hot deionized water and dried in vacuo

over P4010. M. p., 1540C.

Anal.: Calcd. for C12H11N2: Mol. wt., 197 d.

Found: Mol. wt., 197 d (100%).



2-Pyridinal-p-chlorophenylimine, 2-PCI. With gentle

heating and stirring, 10.7 ml (0.1 mole) of 2-PCA was added

to 12.7613 g (0.1 mole) of p-chloroaniline, and the solution

was allowed to cool to ambient temperature. The resultant








pale yellow powder was recrystallized from hot hexane and

dried in vacuo over P 010. M. p., 650C.

Anal.: Calcd. for C12H9N2Cl: Mol. wt., 217 d.

Found: Mol. wt., 217 d (41%).



2-Pyridinal-p-cyanophenylimine, 2-PCPI. While heating

7.0 ml (0.06 mole) of 2-PCA on a steam bath, 6.0353 g (0.05

mole) of 4-aminobenzonitrile was added slowly. The resultant

yellow solution was heated for one hour, and then extracted

six times with 100 ml boiling hexane. The hexane extracts

were evaporated and the resultant yellow powdery product

was dried in vacuo over P4010. M. p., 1090C.

Anal.: Calcd. for C12H9N3: Mol. wt., 207 d.

Found: Mol. wt., 207 d (51%).



2-Pyridinal-p-nitrophenylimine, 2-PNI. While heating a

slurry of hot hexane and 13.8017 g (0.1 mole) of p-nitro-

aniline on a steam bath, 10.7 ml (0.1 mole) of 2-PCA was

added. The amount of solid appeared to increase and became

more flocculent. This mixture was heated subsequently for

one hour, washed four times with hexane, and dried in vacuo

over P4010. M. p., 122 C.

Anal.: Calcd. for C12H9N302: Mol. wt., 227 d.

Found: Mol. wt., 227 d (9%).



Tris(biacetyldihydrazone)cobalt(II) tetrafluoroborate,

[Co(BdH)3](BF4)9. To a solution of 1.8247 g (0.01 mole) of









Co(N03)2 in 15 ml deionized water, 3.4175 g (0.03 mole) of

BdH was added with stirring. The resultant deep brown solu-

tion was filtered and 20 g NaBF4 was added with stirring.

The light brown powder that separated was filtered, washed

with deionized water, and dried in vacuo over P4010. M. p.,

2460C dec.

Anal.: Calcd. for [Co(C4H o N4)3] (BF4)2: C, 25.06; H, 5.22;

N, 29.25.

Found: C, 25.23; H, 5.29; N, 29.36.


Tris(biacetyldihydrazone)iron(II) iodide dihyrate,

[Fe(BdH)3]I2-2H90. Excess BdH was added with stirring to a

solution of 3.9219 g (0.01 mole) of Fe(NH4)2(SO4)2.6H20 in

15 ml deoxygenated deionized water. This deep red-brown

solution was filtered, and 45 g NaI was added to the 45 ml

of solution. The red-brown powdery product was washed with

deionized water and dried in vacuo over P4010 M. M. p.,

2050C dec.

Anal.: Calcd. for Fe(C4H10 N4)32-2H20: C, 20.91; H, 4.94;

N, 24.46.

Found: C, 20.61; H, 4.46; N, 24.20.


Tris(glyoxaldihydrazone)cobalt(II) iodide, [Co(GdH)3]T2.

All reagents, solvents and equipment were deoxygenated under

nitrogen for twenty-four hours, and the preparation was also

carried out under nitrogen. A solution of 3.6751 g (0.02

mole) of Co(NO3)2 in 10 ml deionized water was made, and a









white slurry of 5.1116 g (0.06 mole) of GdH in 15 ml deionized

water was added with stirring. The resultant brown slurry

was allowed to stir thirty seconds and then filtered. To

the brown solution, a solution of 25 g of NaI in 25 ml de-

ionized water was added with stirring. Thirty seconds later

a black powder was filtered off and dried in vacuo over P4010.

M. p., 1650C dec.

Anal.: Calcd. for Co(C2H6N4)312: C, 12.62; H, 3.15; N, 29.44.

Found: C, 12.29; H, 3.08; N, 28.71.



Tris(glyoxaldihydrazone)iron(II) iodide, [Fe(GdH)3]I2.

A solution of 1.3675 g (0.01 mole) of FeC12 in 25 ml deionized

deoxygenated water was made under nitrogen. To this solution

2.5817 g (0.03 mole) of GdH was added with stirring. The

resultant intense red-brown solution was filtered, and 25 g

of NaI was added with stirring. The red powdery product was

filtered, washed with deionized water, and dried in vacuo

over P4010. M. p., 2020C dec.

Anal.: Calcd. for Fe(C2H6N4)312: C, 12.68; H, 3.17; N, 29.60.

Found: C, 12.79; H, 3.22; N, 29.80.



Tris(biacetylbismethylimine)cobalt(II) iodide,

[Co(BMI)']I2. This complex was prepared following the method

of Clevenger (5) being extremely careful to deoxygenate all

equipment and reagents with nitrogen. M. p., 2200C dec.

Anal.: Calcd. for Co(C6H12N3)312: C, 33.30; H, 5.55; N, 12.95.

Found: C, 33.13; H, 5.56; N, 12.80.









Tris(biacetylbismethylimine)iron(II) tetrafluoroborate,

[Fe(BMI)_] (BF4)2. This compound was prepared by a method

analogous to that used for preparing Co(BMI)3I2 except that

FeCl2 and NaI were used. M. p., 2680C dec.

Anal.: Calcd. for Fe(C6H12N3)3(BF4)2: C, 38.28; H, 6.37;

N, 14.86.

Found: C, 38.32; H, 6.44; N, 14.91.


Bis(2,6-pyridinediacetyldihydrazone)cobalt(II) iodide,

[Co(2,6-PdAdH)2]I2. A solution of 5.3108 g (0.03 mole) of

Co(N03)2 in 50 ml of deionized water was prepared, and

12.1071 g (0.06 mole) of 2,6-PdAdH was added with stirring.

The resultant brown solution was filtered to remove excess

ligand, and 50 g of NaI was added with stirring. The black

powdery product was then removed by filtering, washed with

cold water, and dried in vacuo over P4010. M. p., 2700C.

Anal.: Calcd. for Co(C9H13N5)2 2: C, 31.09; H, 3.75;

N, 20.15.

Found: C, 31.25; H, 3.76; N, 20.23.


Bis(2,6-pyridinediacetyldihydrazone)iron(II) iodide di-

hydrate, [Fe(2,6-PdAdH)2]I 2H20. This compound was prepared

by a method analogous to that used for preparing

[Co(2,6-PdAdH)2]I2 except that Fe(NH4)2(S04)26H20 was used.

M. p., 3000C.

Anal.: Calcd. for Fe(C9H 3N5)212 2H20: C, 29.81; H, 4.13;

N, 19.36.

Found: C, 30.30; HI, 4.10; N, 19.46.









Tris(2-pyridinalhydrazone)cobalt(II) iodide,

[Co(2-PAH)3 I]. A solution of 3.0 ml (0.06 mole) of hydra-

zine in 10 ml of deionized water was prepared in an ice bath,

and 7.2 ml (0.06 mole) of 2-PCA was added dropwise with

stirring. To this straw yellow solution, 3.6592 g (0.02

mole) of Co(N03)2 dissolved in 10 ml deionized water was

added dropwise with stirring. The resultant deep brown

solution was then heated with stirring to 1000C, allowed to

cool to ambient temperature, and placed in a refrigerator

for 6 hours. A beaker containing the solution was then

scratched with a freshly broken glass rod and allowed to

sit 30 minutes. The resultant tan powder was washed with

cold deionized water and dried in vacuo over P4010. M. p.,

2200C.

Anal.: Calcd. for Co(C6 7N3)312: C, 31.96; H, 3.11; N, 18.65.

Found: C, 31.26; H, 3.11; N, 18.09.



Tris(2-pyridinalphenylimine)cobalt(II) tetrafluoroborate,

[Co(2-PPI)3](BF )2. A solution of 1.8175 g (0.01 mole) of

Co(N03)2 dissolved in 30 ml of 50% ethanol and 50% deionized

water was prepared. To this solution, 5.7518 g (0.03 mole)

of 2-PPI was added with stirring. After adding 100 ml de-

ionized water, this red-brown mixture was allowed to stir

for 24 hours. The aqueous layer was separated from a red oil,

and 6 g of NaBF4 was added with stirring to the aqueous mix-

ture. The resultant red-brown oily precipitate and aqueous

mixture was heated with stirring, and a beige powder was








produced. It was washed with deionized water and dried in

vacuo over P4010. M. p., 2800C dec.

Anal.: Calcd. for Co(C12H10N2)3(BF4)2: C, 55.49; H, 3.85;

N, 10.79.

Found: C, 55.33; H, 3.92; N, 10.88.


Tris(2-pyridinal-p-methylphenylimine)cobalt(II) tetra-

fluoroborate, [Co(2-PMI)3](BF,)2. This compound was prepared

by a method analogous to that used for [Co(2-PPI)3](BF4)2.

M. p., 2830C.

Anal.: Calcd. for Co(C3H2N2)3 (BF4)2: C, 57.03; H, 4.39;

N, 10.24.

Found: C, 56.76; H, 4.45; N, 10.33.


Tris(2-pyridinal-p-aminophenylimine)cobalt(II) tetra-

fluoroborate, [Co(2-PAI)3](BF,)2. All reagents, solvents,

and equipment were deoxygenated and all procedures were carried

out under nitrogen in order to obtain this compound. A solu-

tion of 1.8271 g (0.01 mole) of Co(N03)2 in 25 ml deionized

water was prepared, and 6.1028 g (0.03 mole) of 2-PAI was

added and the mixture was allowed to stir for 15 minutes.

This red-brown mixture was then filtered, and 5 g of NaBF4

was added with stirring to the aqueous solution. The result-

ant red-orange powdery solid was washed with deionized water

and dried in vacuo over P4010. M. p., 1700C dec.

Anal.: Calcd. for Co(C12H11N3)3(BF4)2: C, 52.46; H, 4.01;

N, 15.30.

Found: C, 52.18; H, 4.11; N, 15.22.









Tris(2-pyridinal-p-chlorophenylimine)cobalt(II) tetra-

fluoroborate, [Co(2-PCI)3](BF,)2. To a solution of 1.8294 g

(0.01 mole) of Co(N03)2 in 15 ml of deionized water, 6.5026 g

(0.03 mole) of 2-PCI was added and the mixture was allowed

to stir for 24 hours. The resultant orange powder was dis-

solved in 50 ml of hot ethanol, and 10 g of NaBF4 was added

with stirring. This red-brown solution was allowed to sit

at ambient temperature for 72 hours. A red-brown solid

consisting of needle-like crystals (soluble in dimethylsulf-

oxide, DMSO) and round globs (insoluble in DMSO) was obtained.

Enough DMSO was added to just dissolve the crystals, and

the round globs were removed by filtering. The orange DMSO

solution was warmed to about 750C, and deionized water added

until an orange powder just began to precipitate. This solu-

tion was placed in a refrigerator for 6 hours. The orange

product was dried in vacuo over P4010. M. p., 2450C.

Anal.: Calcd. for Co(C12H N2C1)3(BF4)2: C, 48.98; H, 3.07;

N, 9.53.

Found: C, 48.72; H, 3.16; N, 9.39.



The cobalt(II) complexes of 2-PCPI and 2-PNI could not

be prepared. Repeated attempts to prepare these complexes

in deionized water, ethanol, methanol, tetrahydrofuran,

carbon tetrachloride, chloroform, and dimethylsulfoxide met

with failure. The cobalt complexes of GdH, BdH, BMI, and

2,6-PdAdH are not stable in air. However, the remaining

cobalt complexes and the iron complexes are considerably

more stable.








Elemental Analysis



All elemental analyses were performed by Atlantic Micro-

lab, Inc., Atlanta, Georgia.



Mass Spectrum Analysis


Parent peaks for several ligands were obtained, in lieu

of elemental analysis, using a voltage of 70 eV, on an AEI

MS 30 mass spectrometer equipped with a DS 30 data system.



Spectrometers


A Varian Associates XL-100 NMR spectrometer was used to

obtain the NMR spectra.

Infrared spectra were obtained on a Perkin-Elmer 137

sodium chloride spectrophotometer, and were calibrated with

polystyrene. Pressed potassium bromide samples were used to

obtain infrared spectra.

Electron paramagnetic resonance spectra were obtained

at 80 K using a Varian Associates E-3 EPR spectrometer,

V-4502-14.

The magnetic susceptibilities were determined by the

Gouy method. The equipment has been described previously

(5). The magnetic field was calibrated using mercury(II)

tetrathiocyanatocobaltate(II) (25). The maximum field strength

at a current of 4 A was 6750 + 80 gauss.









Solution susceptibilities were determined using a

Varian Associates A-60 NMR spectrometer, and the method of

Lbliger and Scheffold (27). Spectral grade methanol was

used to prepare solutions of 2 x 10-5 mole/ml.

Theoretical spin density calculations were obtained from

a INDO/1 program modified to include calculations on molecules

containing atoms from the first transition series. The ori-

ginal INDO model is not changed from that given by Pople and

Beveridge (34). This modified program was written and devel-

oped by Michael Zerner and Alan Bacon at the Chemistry Depart-

ment, University of Guelph, Ontario.














RESULTS


The positions of the infrared absorptions and their

corresponding assignments for characterization of the compounds

prepared for this investigation are listed in Table I. These

assignments are based on the ranges given in Bassler and

Silverstein (2). In addition, similar v(C=N) assignments

were made by Fisher and Stoufer (11) for Co(GdH)3Br2, Co(BdH)312,

Fe(GdH)312, and Fe(BdH)312.

Average room temperature magnetic moments, veff, cal-

culated from susceptibility data, and g values calculated

from ESR data are reported in Table II. These are average

moments determined at field strengths of 5649 Hz and 6750 Hz

to insure there was no field strength dependence and thus

no ferromagnetic or antiferromagnetic interactions were

present. Pascals constants for diamagnetic corrections to

the measured susceptibilities were obtained from Lewis and

Wilkins (25) and from Selwood (38). It would have been

preferable to determine peff in solution for all the compounds,

however, only Co(2,6-PdAdH)212 was amenable to this technique,

since the other cobalt(II) compounds decompose too rapidly

in solution. Solution measurements for these would yield

Veff values which would be the average for each compound
and its decomposition products. As reported previously (5,10)










Table I
Selected Infrared Absorptions and Assignments for
the Co(II) and Fe(II) Compounds Studied


Compounds

Co(GdH)3I2


Co(BdH) 3 (BF4)2
















Co(2,6-PdAdH) 22


Absorption (cm- )


3200 s

3100s(sh)

3000 s

1590 m(sh)

1570 s

1410 m


3510

3420

3220

3000

1650

1610

1440

1375



3400

3300

3180

3100

3000

2850


S

s

s

s

m(sh)

m


Assignment

V (N-H)

as (N-H)

s (C-H)

)(C=N)

v(C=N)

6(C-H)


s (N-H)

as (N-H)
as (C-H)

v (C-H)

6(N-H)

v(C=N)

as (C-H)

6s(C-H)


s (N-H)

vas(N-H)

vas(ring C-H)

vs(ring C-H)

vas(methyl C-H)

v (methyl C-H)













Compounds

Co(2,6-PdAdH)2I2


Table I continued



Absorption (cm- )


1620

1580

1440

1375

800


Assignment

6(N-H)

v(C=N)

6 as(methyl C-H)

6s(methyl C-H)

6(ring C-H)


Co(BMI)3I2


2900

2850

1590

1420

1375


Fe(GdH)3I2b













Fe(BdH)312 2H20b


w(sh)

w

m

m

m


3225 s

3100 s

3000s

1580 s

1540 s

1416 w



3350 m

3250 m

3150 m

2900 w


vas(C-H)

v (C-H)

v(C=N)

6 as(C-H)

is(C-H)



vs(N-H)

as(N-H)

Vs(C-H)

v(C=N)

v(C=N)

as (C-H)



us(N-H)

Sa(N-H)
as

vas(C-H)
vs(C-H)
s













Compounds

Fe(BdH)312 2H20


Table I continued



Absorption (cm- )


1620 m

1590 m

1440 m

1375 m


Assignmenta

6 (N- H)

V(C=N)

as (C-H)
as(C-H)
6 (C-H)


Fe(2,6-PdAdH) 212
























Fe(BMI)3(BF4)2


3400

3300

3180

3100

3000

2850

1620

1590

1440

1375

800



2950

2850

1610

1430

1375


m

m

m

m

w(sh)

w

m

m

m

m

in



w

m

w

s

(s)


s (N-H)

v a(N-H)

vas(ring C-H)

us(ring C-H)

Sas(methyl C-H)

us(methyl C-H)

6 (N-H)

v(C=N)

as(methyl C-H)

6s (methyl C-H)

6(ring C-H)



vas(C-H)

us (C-H)

v(C=N)

6a(C-H)

6 (C-H)


aBased on ranges given in Bassler and Silverstein (2).
Similar v(C=N) assignments were made by Fisher and Stoufer (11).










Table II
Magnetic Moments and g Values for the
Co(II) and Fe(II) Compounds Studied


Compoundsa


Co(GdH)3I2

Co(BdH))312

Co(2,6-PdAdH) 212

Co(BMI)3T2

Fe(GdH) 12e

Fe(BdH)3I2 2H20e

Fe(2,6-PdAdH) 2 e

Fe(BMI)3 (BF4)2e


Jeff (300 K, in BM)

4.23


4.65b

2.28b,c

3.30d


0.00

0.61

0.57

0.80


g (80 K)


2.15

2.06

2.09

2.13


rascals constants for
susceptibilities were
and Selwood (38).

The reported value is


diamagnetic corrections to the measured
obtained from Lewis and Wilkins (25)


4.17 BM (40).


Determined by NMR in solution (27).

The reported value is 2.91 BM (41).


eNo EPR signal was observed.
No EPR signal was observed.







these anomalous moments for all the cobalt(II) compounds

span the range between low-spin, 1.7 BM, and high-spin, 5.2

BM, values. It is noteworthy, however, that the 2,6-PdAdH

complex has a Ueff sufficiently low, 2.28 BM, that it may be

considered to be a low-spin complex. The observed nonzero

moments for the "diamagnetic" iron(II) compounds are a result

of temperature independent paramagnetism.

The EPR spectra for these cobalt(II) compounds consisted

of a very broad signal centered near a g value of 4 attributable

to the slightly populated E ( T2g) excited term, and, superim-

posed on it, a relatively narrow signal centered upon a g value

in the range 2.06 2.15, very close to the spin-only value of

2.00229 and the isotropic value of 2 expected for the 2E level.

The iron(II) compounds exhibited no ESR signal as expected.

The EPR data presented above and the NMR data which follow

are correlated in the discussion videe infra). The isotropic

shifts, Av, for the cobalt(II) compounds are reported in

Table III together with the relative shifts and the corre-

sponding half-height widths. These relative isotropic shifts,

Avrel, are normalized with respect to the largest observed

shift, which in each case happens to be the proton-con-

taining moiety bound to the hydrazone or imine nitrogen. As

stated previously the GdH, BdH, and BMI complexes decompose

rather quickly in solution. Therefore, for each it was

necessary to make all instrumental adjustments and run a

trial spectrum on one sample and run the final spectrum

quickly on a freshly prepared sample. Within ten minutes

resonances attributable to the decomposition products

began to appear. Thus it was possible to make unambiguous










Table III
Observed and Relative Isotropic Shifts
and Half-Height Peak Widths for Cobalt(II) Compounds


Av (ppm)



2.29

- 103.66



2.72

86.69



10.14

42.90


vrel (ppm)a



- 0.022


1.00


0.031

1.00


0.236

1.00


Width (Hz)



60


500


30

150


60

150


2,6-PdAdH


C-H

N-H


8.97


- 61.91

- 12.09

- 6.06


- 0.144

1.00

0.195

0.098


110

750

60

80


aRelative isotropic shifts are normalized with respect
to the largest observed shift.


Ligand


GdH


C-H

N-H


BdH


C-H

N-H


BMI


C-C-H

N-C-H









assignments based on these observations, the relative peak

intensities, and the fact that all the protons bound to the

hydrazone or imine group would be expected to shift the same

direction. Furthermore, they exhibit the same relative

magnitude (i.e., largest shifts observed in each spectrum)

and all are the broadest peaks in the spectra, which broaden-

ing is due to their proximity to the nitrogen quadrupole.

There was no decomposition problem involved in the

assignments for 2,6-PdAdH. However, the HA and HB isotropic

shifts as assigned by relative intensities are reversed

compared to the expected shifts videe infra). Consequently,

the assignments for HA and HB are based on the following

syllogism. Because the electron-proton dipolar relaxation

mechanism often dominates the line widths in paramagnetic

complexes, the relative line widths for nonequivalent protons

will be determined primarily by their relative values of

R where R is the distance from the central metal ion to

the proton (4,24). Since RH > RH (see the structures in

the introduction), R6 < R6 and the line width for HA should
A B
be less than that for HB.















DISCUSSION


Theoretically it is predicted that proton contact shift

spectra for complexes with s = will not be observable

because Tle is usually too long (42). And as noted in the

introduction no spectra have been reported for low-spin,

octahedrally coordinated d7 ions. As inferred from the

introduction, the purpose of this investigation was, in part,

to determine if proton contact shift spectra are observable

for anomalous cobalt(II) complexes near the cross-over point,

and to determine if the spectra are attributable to the 2E (2G)
g
term. If so, this would mean Tle is sufficiently lowered by

low-lying excited states (42).

There are several approaches which should be considered

in arriving at an interpretation of the isotropic shifts of

these complexes. (i) The site symmetry of these complexes

may be taken to be octahedral. Accordingly, if the concen-

tration of the 4T( F) species is small, and if the 2E --4T1
tration of the species E


interconversion provides a mechanism whereby Tle is sufficient-

ly lowered, then a contact shift should arise from the 2E
g
term. In this case there would be no spin-orbit coupling
9
effects from "E and, because E and A terms are isotropic,
g
there would be no dipolar shift complications to consider (15).

(ii) Under octahedral symmetry the contact shift arises from







the 4T term. This is the origin of all reported cobalt(II)

contact shift spectra (42). In this case there are spin-orbit

coupling and dipolar shift complications. (iii) Both (i)

and (ii) contribute to the observed spectra. In this case,

if the 2E -- T interconversion time is large on the NMR time

scale, two superimposed spectra should result. This was not

observed for these complexes (Table III). (iv) If (i) and

(ii) contribute and the E -- T1 interconversion time is
g 1
small on the NMR time scale, time-averaged spectra should

result. In this case spin-orbit coupling (44) and dipolar

or pseudocontact shifts from the 4T1 term are involved. But,

approaches (i) (iv) are valid explanations only under octa-

hedral symmetry.

Strictly speaking six-coordinate complexes containing

symmetric bidentate ligands are inherently of D3 symmetry;

consequently approaches (i) (iv) are not valid considerations.

Under D3 symmetry 2E (Oh) 2E (D3) and 4Tl(Oh) 4A(D3) +

E(D3). Since all the term levels in D3 symmetry are isotropic

(A and E), the observed isotropic shifts should be due en-

tirely to contact interaction (15,29,43). Therefore pseudo-

contact contributions either should be absent or negligible.

The nonideal approach commonly used to interpret contact

shift data is to treat a single ligand molecule as a free

radical cation or anion depending on the delocalization mech-

anism involved. That is, the delocalization of unpaired spin

is considered to result from L M or M L charge transfer

(23,36). In the discussion which follows, (+) and (+)








represent an electron whose magnetic moment is aligned par-

allel and antiparallel, respectively, to the external magnetic

field.

Figure 3 illustrates the spin transfer mechanisms possible

(23). The ligands contained in these complexes are all dia-

magnetic and thus the ligand MO's, L' either are doubly

occupied or are unoccupied. L M charge transfer is

illustrated in (a) (c). If the metal ion has a singly

occupied d orbital of a symmetry (a) is the only transfer

possible. The spin is transferred-from a ligand a orbital

(OO) to the occupied metal do and paired with the spin in the
do orbital. Since the metal ion unpaired spin will be par-

allel (-) to the external field, this leaves net (+) spin

on the ligand. The metal ion can contain a singly occupied

d" orbital with a low-lying do orbital, (b) and (c). Un-

paired spin (+) and () can be transferred into the do from

the L. However, correlation favors transfer of (t) spin

(Hund's rule of maximum multiplicity) leaving net (+) spin

in eL (32). M L charge transfer is illustrated in Figure

3 (d) (f). For a diamagnetic ligand a vacant L is re-

quired. Vacant o MO's are too high in energy to be im-

portant, so only vacant 7T MO's need be considered. If the

metal d orbital is singly occupied, (d) is the only trans-

fer possible and net (t) spin is delocalized onto the ligand.

However, if the unpaired spin occupies a d orbital, and

the d orbitals are doubly occupied, (e) and (f) spin trans-

fer is possible. In this case, correlation favors transfer

of (+) spin from metal to ligand.



















d 7


(a)


(b)


, --4 d
- __d


(c)


Sd
..-++ 'a -


(f)


Figure 3. Possible spin transfer mechanisms


-- do


L


L


- L


-- L









A theoretical analysis of the [Co(GDH) ]spectrum was

attempted based on a complete, many-electron calculation for

the entire complex. Such a calculation was performed with

the complex considered to be (i) high-spin with sp2 and with

sp3 hybridized amine nitrogen orbitals and (ii) low-

spin with sp2 and with sp3 hybridized amine nitrogen orbitals.

These calculations were accomplished using a modified INDO

program. Since X-ray crystallography data were not available,

the bond lengths used in these calculations are the average

bond lengths given by March (30). This program is capable

of handling the minimum basis set (ninety-nine atomic orbitals)

of the complex, including metal penultimate d and ultimate s

and p orbitals and all of the ligand valence orbitals.

The results of the calculations together with the ex-

perimental results are listed in Table IV. An inspection of

the data will reveal that neither the absolute magnitudes

nor the signs of the theoretical spin densities agree with

the experimentally determined values and signs. Indeed, even

the relative magnitudes do not agree. That is, although the

hydrazone protons are found to have the largest shifts and

therefore the largest spin densities, the INDO calculations

predict the protons attached to the carbons to exhibit the

largest shift. One factor causing this disagreement may be

that these experimentally determined densities were calculated

using Equation (2) which is applicable to complexes having

normal Curie behavior, a condition which does not exist for

this complex videe infra). Another factor may be that,















0
O
0 -r
4-J 4-3
co
)3 -$ 0 -0 N
4> CM D C .N
rJ O 0 0 r4
C 4 0 0 0 0 (U *r2-




b 0 o





4-1 h
00O


L4-4 ,- ,0 Lf Cfr-
0 8 .i1

L > cM CM U
U c CN O Cq C 1 0
0 4J 0 0 0 C C0 aO
U &4 (U nJ
0 "- I I C0 C



















OO 0e nn .
Q o ) o I o N N o
C-4 CN I C N) 0>
*rol S Ir 0
>C 4.




0 C cn CU .C N
*0- > n O



-4 C D O C M 0 0C
co CL 1- r-0 C 0 I r- U



) uo ) P (

o0 m
CI C o no m j-I









P o 0 o 0
0 Lfl 0 iUi 0e














C- 0*i0 o -<4 10

*r- 0 -4 0CD

I-1 M Q1)
-40 CI)


















co x r ap


'-< -1 I I CM MO *i-
tO CO XD r
U-I 0 N


Q co d4
-^c U n )

O r-O O 001
*) C 4 CO CO c 4.

l0 00 C13 0 C
"a C- ) Q0 (










U) 10 t
'O o 0\








I -N CP4








although this modified INDO method is capable of handling this

basis set videe supra), the amount of unpaired spin density

calculated on the protons is sufficiently small (Table IV)

that it can only be considered to be "noise." This is due to

the fact that the tolerance in this program cannot be set

sufficiently low to insure convergence, and thus calculate

meaningful densities for a total of one unpaired electron

(low-spin) or a total of three unpaired electrons (high-spin)

in the complex over ninety-nine basis orbitals. Consequently,

no meaningful importance should be attached to these calculated

densities.

Since INDO calculations did not produce meaningful

results, the common (but nonideal) approach of considering

the ligand to be a radical cation or anion was used even

for this simple complex. However, in this work the entire

valence set of the ligand was used rather than just the 7

symmetry basis functions as is frequently done (see reference

36 and references therein). The metal orbitals are completely

neglected in forming the molecular orbitals for the calcula-

tions; that is, the metal ion has been considered to be

either an electron source or sink. In addition, it is general-

ly accepted that ligands containing an a-diimine linkage

serve not only as a donors but as 7-acids as well (d7--pr

back bonding with the metal ion). That is, there may be M L

and L M spin transfer simultaneously. Obviously it is an

oversimplification to consider the ligand to be either a

cation or an anion under such circumstances. Furthermore,







the effects of one ligand on another in the complex has been

neglected. And the bond lengths of the neutral ligand

molecule were used (30). This certainly is not the ideal

approach, but there is little choice since there is no rea-

sonable way to estimate the magnitude of the change of bond

length.

Table V is a presentation of the INDO spin densities for

the ligand radicals and the corresponding spin densities

calculated from the contact shifts and the relative spin

density of each. An inspection of these densities leads to

the conclusion that, for all of the ligand radicals,the agree-

ment between the relative magnitudes of the calculated spin

densities and, in the case of the cations, the experimental

densities is, in general, fairly good. The BMI1 calculation

produces the best agreement in relative magnitude: prel(INDO)

= 0.166 and 1.00 for CCH3 protons and NCH3 protons, respec-

tively, and prel(exptl) = 0.236 and 1.00 for CCH3 protons and

NCH3 protons, respectively. In comparison, Scarlett et al.

(36) report good agreement for the pyridine N-oxide cation:

prel(INDO) = 1.05, 0.41 and 1.00 for 2-H, 3-H and 4-H,
respectively, and prel(exptl) = 0.81, 0.45 and 1.00 for

2-H, 3-H and 4-H, respectively. Only the GdH radical cation

calculations with sp2 hybridized amine nitrogen orbitals

agree in both sign and relative magnitude videe infra) with

the experimental densities. This may be taken as an indication

of L M charge transfer mechanism of spin delocalization.

In contrast, the calculated spin densities for both the BdH

radical cation and anion with sp2 hybridized amine nitrogen









































CQ)



C)
0



OO


4O
CC
OU








)O
-W H
C)CC



ScO





0t 0


LC
0
1
O C




U 4-4




,(-A
*r-
'* 4
(L T


4-J

















*H-
41






'-4
ct
u






0






I









O
4-J
ed















0C








C)




r-
0

I-



















U
-A
U


cn r-
coo

I N


coo
Cn r-










c10


0 -


-o










Cl








0.o
*-l 7H
Q-4 C









CO


rC'
Td r


oo
co
kfl
I






o0

- 0











1 -4











N





C4
NO


0 n
in kD
I r-


,-- co
4 00

NW






U


r
o N


^~\s


cri

CC


NC3





















c?0
00


0r4
C )


CC

00

0 H r


r--4
r--


)0
U- r-l

r--l 0 0






0

O0



,- co o

cU i C





*r-4

cil 0 rn

Q)
r4 r











-4 -4 -
\D
0






O O's












c-4

+
N


CM
C- 0
















N
cn
cCM








0o4




1r--4 C
-n o
emn















--4







Ci

/)



a a f
pqu


00

,--i
o4




















-.. 0
cq















N
Nl
































C14
00












0 coi

cq
r'N






0cl
0c0
00
























C)
1^~ LD


CM

0 0














o4
C0 r-

0 o
r-4








00













o4
0 r-













N4
























0-4 C
r-4


r-4


































0



O
r-H














c3


M o



.O-I
*r- )
'-c-c


Ni
U)

















Q)
;c>



O O N O C) o C)o
C)o CD ojo C) o7 CD








.--
I








CCO C Ll 00 Ll U0
>r u4

4C-C






Cd 4 CO C O CD
CC)


















00 No NO
-A 0 0 c
4( O. ,O .O






















b0 0
l 0 0 C N 0









U


-C 0 0 r 00 Xu0
uC d
Cd 0)
Cd O ,-0






:> ( o Cs o C'. o


r-,I

0




O, 0




Q I





-U/-
CdC)
,--I










u0 0







H+ L) m +j c o7
UU d co H Z]

















c0
>

















*r-
CO
r--i
























-4
0















r-1
X





























r-4
CO
LI






4-1
o








r4

















U
r.-
X










,-





l-a


r--








o


co
00







nI .O 00 C






-A 0N r-4







o
0-40C
Cn iO c00\







c-4 CM C








r-4








cNI o
0- r-


C) r-o o








N,- r N- n- oo
m co c7 m
c D Co CIN
CN N i-

I







IIN





0-
C


Zn4
WC fi < PQ
u ^ 1


> ,--1 co .."
0








u o



-40 0 0

0 r)-- 0
-I CM a' r1











4 o r-4 o D














,1-
Nr) N ) c-


-4fm







ift 0 %- --

N ONO
0 0 00





oC i-4 0 Ca








I IN







uZ-4
I m






r-41


co
00


--" 0 i o

S-1 00










cn 110 00 a'










-t 0 m m'
r-4 0 r-4 0

0 0 0










clM CrM r-

N -
I





S o

00 Dam-t
0 0 r-4 CM
0,00









0t Crm 0 0


on rn 0 CD





I I








'0 C'1 N
<3 M Q
0-1 '-
I~-0



vo m CM
a \1 0 p


r-4
4-i


cL
0C
a









































H
.,-


0

z













(-



- C




i-l


















4-t
a)




r4
,D
/-4




C)

0































C)
or
4-1
(U

-4




I


0


-u
U
-4
(C







(4
r>

-4




CD
i-u







(U
r-
CCs
U0


-4 0 o'


0 C- 0


r-4
4-1

































O
CL























-0



Ho


CCO
0 0


m -1 00 CV) ^D 00 mo
r--l r-l


4J














41
*r)
o


















41
C



















<)



.4
0














4-J
C:
Cl)


















H
0
cn










































C)
4-1












C)
aH
0













































.-l
0)
























CJ
0)


1)



42
aC
*i-



C)
N






C







*H

*r






*Hr

COl

0)



(0



tO


O *Hr
0 1-4 (U



S0 Cl ,




0 0 0 N



-H
4- j Cl)



co ,-1 1


s:: cd






71 0 eH
C l)















r 4
4 4 r- -(















0 U 0 C'

r--l r-4

0r- co
1 0)
C r- 42 N

















T- 0 0



Q)> l


0 C O0

4- Cu C




t Q) 0l
;j 0 CO


42 U U
Ct .)0 r
U4) l U
S- -

(IC t4 C
(a H






(n) (14 ) -4
) O











CC C 0
c -H 0
*r-4 r 1':





0 O

io c d 3
C



C Cn C
CJ r C 4c
ma- r-
01^ (


N CN CO r-
of o r-- L)
N LO CO r-i
I-







CN 0 L C

LO 0 C(n r-






o o o o






(D 0 CO CD
- co cr- _4













cr) c"I
r 2


o
C




0
rl-






0



o




0
N

CC
7e


0

Q)
42




4

a)





-0


r4
4-1
CC

4-u





cn
















*.r- C
o-
CC



UN








orbitals agree in sign and relative magnitude with the experi-

nentally calculated densities. Thus the charge transfer mech-

anism may be either L M and/or M L. Calculated spin

densities for the BII radical anion agree in both sign and

relative magnitude of spin densities with the experimental

densities. This may indicate M L charge transfer mechanism

of spin delocalization. Although the radical anion calculated

densities for 2,6-PdAdH with sp3 hybridized hydrazone nitrogen

agree with the relative magnitude of the experimental densities,

the signs of the densities do not agree, which means there is

no agreement at all videe infra).

The experimentally determined densities listed in Table

V, calculated using Equation (2), apply to the special case

of normal Curie behavior and are derived from (12,18)

AH = Ah z /N N (3)

where is the time-averaged value of the z component of

the electron spin. So, the problem of calculation reduces

to an evaluation of
= g g S(S + 1) H/3kT (4)
Golding (13) evaluated Sz> for complexes of octahedral site

symmetry. This evaluation included spin-orbit coupling and

magnetic field interactions


Z(2J + 1) Sz > e[-E/kT]
SSz> J (5)
zJ -(2J + 1) e[-/kT]
J


where E is expressed as a power series in H and










(Y Y) V

(6)
This reduces to


3kT (A + (7)

where C is the spin-orbit coupling constant of the transition

metal ion. Golding tabulated values for the parameters y, v,

g, A, and B required to evaluate <(S) for octahedrally coor-

dinated metal ions of dn configuration.

Under octahedral symmetry and by including spin-orbit

coupling, the T1 and E terms split into 2U' + E" + E'

and U' terms respectively. At ambient temperature both the

ground term U' E ) and first excited term E' ( j

are occupied (44). Therefore the summation of
Equation (5) must include these occupied terms. Substitution

into and evaluation of Equation (3) requires two hyperfine

coupling constants (12). Thus there are four unknowns AU,

AE,, E, and r and only one equation. By assuming the free

ion value of c and varying E to produce a best fit, Equation

(3) has been used to calculate coupling constants for spin-

state equilibrium systems (13). However under the noncubic

D3 symmetry of these cobalt(II) complexes the situation is

much more complex. In addition, the free ion value of C is

certainly not appropriate. To be sure, a value less than the

free ion value usually 75% of the free ion value should

be used to account for covalency effects (44). And, under









D3 more than two term levels may be occupied at ambient tem-

perature adding further complications. Furthermore, mixing

of terms by spin-orbit coupling must also be considered.

Although matrix elements for spin-orbit coupling under octa-

hedral symmetry have been calculated previously by Griffith

(14), those for D3 symmetry have not been published. Con-

sequently, it was not feasible to determine coupling constants

and thus spin densities from the observed contact shifts.

Fortunately relative contact shifts and therefore relative

spin densities can be determined using Equation (3) with-

out evaluating That is, there is no dependence on the

distribution over more than one term, as evidenced by the

close agreement of relative spin densities for high-spin

and low-spin values recorded in Table V.

The degeneracy of the t2g and eg orbitals is lifted

under D3 symmetry as shown in Figure 4 (33). The t2g T-

bonding orbitals transform as ag and e" under D ; the e

transforms as e. It is seen that a high-spin or a low-spin

configuration is possible for seven d electrons under D3;

thus, the Oh and D3 situations are qualitatively analogous.

That is, anomalous magnetic behavior can arise under either

symmetry. Considerations of the spin transfer mechanisms

represented in Figure 3 and the electron configurations

given in Figure 4 reveal videe infra): (i) If the electron

configuration is either high-spin or low-spin, (+) spin can

be transferred onto the ligand by L MI or 'I -+ L, respectively.

(ii) If the electron configuration is low-spin, (p) spin













-- e -. e
g

-- a1




-- t2g


is {(e)4(a) 2(e) 1





hs {(e)4l 1 )2
(al) (e)}


7T
e



Oh D3







-- -- e -------------- e
g


ST


- g t2


Is {(a 2(e)4(e) }





hs {(al 2 e)3 )2


7T
al


Figure 4. Possible Oh--D3 correlation diagrams and electron
configurations for a d7 ion under D3 symmetry









can be transferred onto the ligand by L M. (iii) If the

electron configuration is either high-spin or low-spin, both

(t) and (i) spin can be transferred onto the ligand by M L.

The ligand orbitals and metal ion orbitals can interact

and thereby transfer unpaired spin only if they are of

appropriate energy and of the same symmetry. The 7 MO's for

a symmetric bidentate ligand can be classified as (+) or (-)

according to their parity, Figure 5, which either remains

unchanged or is reversed upon rotation about the C2 symmetry

axis (33). The (-) MO's can interact only with the metal

e orbitals, while the (+) MO's can interact with either al or

e metal ion orbitals depending on the choice of C2 axis for

the metal ion orbitals. Table VI is a presentation of the

INDO highest bonding and lowest antibonding T MO's, their

symmetries, (+) or (-), and their energies. These are pre-

sented for the ligand radicals which afford the best fit

with experimental densities.

The reader should apply the general discussion above

and refer to Figures 3 and 4 and Table VI for the discussion

of transfer mechanisms which follows. The calculations of
2
INDO spin densities for the radical cation with sp hybridized

amine nitrogen orbitals gave the best fit with the GdH com-

plex. For this cation the unpaired spin is predicted to oc-

cupy i17, whose symmetry upon rotation about the C2 axis is

(+). Therefore it can interact through L M spin transfer

with e(d) or al(d) metal orbitals. Thus a low-spin configu-

ration should result in (+) spin and a high-spin configuration

should result in (t) spin in i'17.




















I -_' / 7

C2


N+)
N


NC
*N. K-


+)


-x



2


dz2 (+)


C2
dxy (-) or (+)
2


Figure 5. Illustration of (+) and (-) symmetries of selected
's and selected metal orbitals upon rotation
about the C2 axis









Table VI
Selected INDO Molecular Orbital Symmetries
and Energies of Ligand Radicals


Ligand Molecular Occupation Energy
Radical Orbitala Number (Hartrees)


GdH+1 (sp2)


BdH-1 (sp2)


BdH+1 (sp2)


BMI-1


- 0.627005

- 0.573568

- 0.066018


16 (-)

17 (+)
18 (-)



23 (+)
24 (-)


25


2'22
423

24

25


222

0'23

'24

J25


(a)b


(-)


(+)

(-)

(+)




(c)b
ab


- 0.072167

0.124016


0.513461


- 0.587260

- 0.552269

- 0.050940

0.058413


- 0.086492

- 0.083043


0.107814

0.496251


(+)


an MO's are classified (+) or (-) depending on whether they
remain unchanged or reverse sign upon rotation about their
C2 axis (32). The MO's are numbered in order of increasing
energy.

bThese are molecular orbitals and are not classified further.








The calculations of the INDO spin densities for the

radical cation and anion with sp2 hybridized amine nitrogen

orbitals give the best fit with the BdH complex. The anion

calculation predicts the unpaired spin to occupy 1i24, its

symmetry upon rotation about the C2 axis is (-). Therefore

it can interact through M L spin transfer only with the

al(d) metal orbital. Consequently, a low-spin configuration

should result in (+) spin and a high-spin configuration

should result in (t) spin in q924. In addition, it may well

be that 25 is energetically accessible. If it were acces-

sible it could accept (+) spin and/or (4) spin from a e (d)

orbital. The BdH cation calculation of INDO spin densities

predicts the unpaired spin to occupy 123' whose symmetry

upon rotation about the C2 axis is (+). Therefore, it can

interact through L M spin transfer with e(d) or a (d) metal

orbitals. Thus a low-spin configuration should result in

(+) spin and a high-spin configuration should result in (t)

spin in 23.

The radical anion calculation of INDO spin densities

gives the best fit with the BMI complex. The unpaired spin

is predicted to occupy P24, whose symmetry upon rotation

about the C2 axis is (-). Therefore it can interact through

M L spin transfer only with the al(d) metal orbitals. Thus

a low-spin configuration should result in (+) spin and a high-

spin configuration should result in (t) spin in r24'

A discussion of the 2,6-PdAdH complex is deferred,

because the complex is of D2d symmetry rather than D3 symmetry,

it will be discussed separately videe infra).









Because it also places opposing spins at the same posi-

tion in the ligand, spin polarization adds further ambiguity

to the interpretation of the contact shifts in these five-

member (or any odd member) chelate ring complexes. If, for

example, the spin is delocalized into a a orbital, this places

metal spin into s orbitals of all atoms. Also, the bulk of

the delocalized spin resides in the lone pair of the coor-

dinated nitrogen atom causing a "slight unpairing of all

filled a bonds" by spin polarization effects (23). This is

readily seen for GdH by structures such as those in Figure 6.

If (,) spin is considered to be placed in a MO at the

coordinated nitrogen, the other electrons associated with

that nitrogen atom will also align with (t) spin preferentially

(Hund's rule of maximum multiplicity). The electron associated

with the adjacent atom must be paired, (+), with the nitrogen

atom electron. All the electrons associated with the adjacent

atom must also align (+) (Hund's rule), and so on through

the molecule. Thus spin polarization places (+) spin and

(+) spin at each position simultaneously. These considerations,

when applied to the other ligands, lead to the same con-

clusions for each of the other ligands. Thus the ambiguity.

The foregoing discussion certainly indicates there are

many difficulties in reaching definitive conclusions about

the contact shifts of these compounds. However, considera-

tion of how the lack of agreement between INDO calculated

and experimentally determined spin densities relates to.low-

spin contributions may be more enlightening. In each instance






56








i- +Ht



C+ + C


H+ fH



















+C+ tC

Ht +H


N+ tN' WN4 tNN
4- 1 41

Ht ....M +H









Figure 6. Representation of spin polarization effects in a
fragment of a [M(GdH)++] complex









for which the signs and relative magnitudes of the calculated

densities agree with the experimentally determined densities

(Table V), the largest density is on the protons of the

amine or imine moiety; this density is always positive. The

smaller density is on the remaining protons of the ligand;

but the agreement in relative magnitude is only fair. This

may indicate contribution from both high-spin and low-spin

species. As can be deduced from the transfer mechanisms in

Figure 3 and the electronic configurations in Figure 4, con-

tribution from both configurations would place (t) spin

and (+) spin at the same proton. For example, if the elec-

tronic configuration is Is {(e) (a) 2(e) 1, the unpaired

spin occupies the e MO and will align parallel with the

external magnetic field. The transfer must be L -* M. Con-

sequently, (-) spin will remain on the ligand. By comparison,

if the electronic configuration is hs {(e) (a) (e) 2}, the

unpaired spins occupy the a1 and e and will also align

parallel with the magnetic field. The transfer mechanism

may be either L M (L e ) or M L (a -T IT). If it is

L M1, (+) spin will remain on the ligand. Consequently, if

there is low-spin contribution to the contact shift, the

effect of the opposing spins on the ligand would be a smaller

shift and therefore a smaller relative density than that

calculated.

The preceding syllogism can be applied to the contact

shifts of these cobalt(II) compounds. Susceptibility studies

yielding intermediate, room temperature, magnetic moments of









4.65 BM and 4.23 BM for the BdH and GdIH complexes, respec-

tively, are taken as an indication of relative contributions

from both high-spin and low-spin configurations. The low-

spin contribution is about the same in the GdH complex as in

the BdH complex and is presumably reflected in the similar

disagreement between the INDO calculated and the experi-

mentally determined relative magnitudes of spin density.

Consequently, it might be inferred that the effect of a con-

tact shift contribution attributable to the low-spin (2E)

configuration is observed in these isotropic shifts. By

comparison, the relative magnitudes of the experimentally

determined densities for the BMI complex are much larger -

not smaller than those calculated by the INDO method. The

intermediate, room temperature, magnetic moment of 3.30 BM

of the BMI complex also indicates contribution from both

high-spin and low-spin configurations, and indicates a greater

relative contribution from the low-spin configuration than

in the BdH and GdH complexes. It may be that in the BMI

complex the high-spin and low-spin contributions-reenforce

rather than oppose one another. This mutual reenforcement

of high-spin and low-spin contribution may also account for

the larger shift of the BMI methyl protons compared to the

BdH methyl protons.

It should be reemphasized at this point that with the

present stage of development of theory it is not possible








to distinguish with certainty between contributions from

the 2E term and spin polarization involved in these iso-

tropic shifts. As pointed out in the introduction contact

shifts of cobalt(II) complexes are previously known to arise

from the high-spin configuration ( T under Oh) only.

The 2,6-PdAdH complex was not considered in the pre-

ceding discussion because its symmetry is D2d rather than

D Under D2d symmetry 2E (Oh) + A + Bl(Dd) and T (Oh

- B2 + E(D2d) (6). As under D3 all terms contribute zero

orbital angular momentum (A, B, and E). Consequently, the

isotropic shifts are due entirely to contact interaction.

Therefore pseudocontact contributions should be negligible.

The possible d orbital correlation diagrams and electronic

configurations for D2d are shown in Figure 7. Based upon

the room temperature magnetic moment of 2.28 BM, a low-spin

configuration should apply. This value is effectively

within the range generally accepted for six-coordinate low-

spin cobalt(II) complexes (1). Nonetheless, a high-spin

configuration is also feasible under D2d, a condition which

can give rise to a degree of anomalous magnetic behavior as

is possible under octahedral symmetry.

None of the INDO spin density calculations agrees in

either sign or relative magnitude with the experimental den-

sities of the 2,6-PdAdH complex. Consequently no assertion

can be made as to whether the spin transfer mechanism is

M L and/or L M. The complete lack of agreement for the

2,6-PdAdH complex completes the trend discussed above for
















- -e
9


al + b


s ((e)4(b2) (al or bl) }



hs {(e)4(b2) l(al) 1(b)}


--e




D2d


- -e


al + b1


- e


-t2g.


Is {(b2)2(e)4(a or b1 I



hs {(b2) 2e) 3 (a)1 (b)


D2d


Figure 7. Possible 0h--D2 correlation diagrams and electron
configurations or a d7 ion under D2d symmetry


- b2


- 2g.








the BdH, GdH and BMI complexes. That is, increasing dis-

agreement between INDO calculated spin densities and experi-

mentally determined spin densities may reflect a greater

low-spin contribution in each complex. The low-spin con-

tribution is smaller in the BdH complex and the agreement

is better. By comparison, the low-spin contribution is

greater in the 2,6-PdAdH complex and there is no agreement

between the INDO calculated and experimentally determined

spin densities.

It appears that the low-spin configuration may indeed

contribute to the contact shifts of these cobalt(II) com-

plexes. It is also of interest to note that the hydrazone

amine and imine proton resonances are shifted downfield,

and are the largest shifts observed in each complex. There-

fore it is reasonable to conclude that they are dominated

by the same delocalization mechanism. Since delocalization

into a a 10 leads irrevocably to a downfield shift (23), it

is tempting to propose that delocalization into a a MO

dominates at these positions with an opposing 1 MO contribu-

tion which makes an upfield contribution to the contact shift.

This postulate is in keeping with that of Stoufer and co-

workers (41) that f back bonding plays an important role in

these anomalous cobalt(II) complexes.

Although attempts were made to prepare and investigate

the proton contact shifts of cobalt(II) complexes of all the

ligands listed in the introduction, these attempts did not

meet with success in some cases. Specifically, it was not

possible to prepare the complexes of the ligands in Figure 2.







The tris-glyoxalbismethylimine cobalt(II) complex decomposes

so rapidly it can not be isolated in pure form. The complexes

of the remaining ligands could not be prepared, probably be-

cause of steric and/or electronic effects (20). That is,

either intramolecular steric strain or the electron with-

drawing capabilities of the substituents was sufficiently

great to prevent coordination of the ligand to the metal ion.

In conclusion, although the modified INDO program used

was capable of handling the minimum basis set of the simple

tris-glyoxaldihydrazone cobalt(II) complex, the spin den-

sities calculated were so small as to be of little value.

Consequently the commonly used approach of treating each

ligand as a radical anion and/or cation was used for this

work. However the various ambiguities which arise from this

treatment (spin polarization, delocalization into odd number

chelate ring complexes, uncertainty as to relative energies

of ligand orbitals upon interaction with the metal ion)

require caution in making very definitive conclusions. There

is evidence, however, that T back bonding plays an important

role, and that there is contribution to the contact shifts

attributable to the low-spin, 2E(D3) derived from 2E (h)'

configuration.

This investigation of contact shifts of magnetically

anomalous cobalt(II) complexes has lead to essentially the

same conclusion reached by Bertini and Getteschi (3). "The

enthusiasm of researchers in the proton contact field has

recently attenuated owing to the lack of a simple model

correlating the observed shifts to the nature of the





63

increasing sophistication have been proposed and laborious

calculations have been performed which gained in some cases

partial success but never found general applicability. In

many instances however these partial models have been the

origin of lively controversies."















APPENDIX


Chemical shift assignments for the proton magnetic res-

onance spectra of the 2-pyridinalphenylimine ligands are

listed in Table VII. They are based on the assignments

made by Matsubayashi, Okunaka, and Tanaka, which were made

for substituted 2-pyridinalphenylimines, in which the 4-

phenyl substituents are OCH3, CH3, and C1 (31). The follow-

ing numbering system applies.



4
5 r 3 a
6 1' YN X

H X
a





Table VIII is a list of the chemical shifts of the tris-

(2-pyridinalphenylimine) complexes. The purpose in preparing

these complexes was to obtain and investigate contact shifts

of a homologous series of complexes which span both sides

of the cross-over point. Because this situation did not ob-

tain, Table IX, the compounds are of little value for this

purpose. Furthermore the symmetry of these complexes is no

higher than D3 (face isomer) and may be C1 (edge isomer).










Table VII
Proton Chemical Shifts of the Substituted
2-Pyridinalphenylimine Ligands


Position H NH2 CH3 Cl 2-PAH

H3 8.15 8.15 8.17 8.16 7.96

H4 7.83 7.85 7.85 7.86 7.73

H5 7.45 7.39 7.41 7.46 7.23

H6 8.66 8.67 8.69 8.70 8.63

H 8.56 8.62 8.61 8.58 8.13

H 7.30 7.25 7.23 7.31 a

H 7.30 6.71 7.23 7.38

H 7.30 4.22 2.34
x


Note: Chemical shifts are
at 30 C.


in ppm's relative to TMS


in CD3CN


aThere are no phenyl protons in the 2-PAH ligand. However,
the NH2 chemical shift in d6-DMSO was found to be 4.13 ppm
from TMS.










Table VIII
Proton Chemical Shifts of the Substituted
Tris(2-pyridinalphenylimine) Cobalt(II) Complexes


[Co(2-PAH)]3


















[Co(2-PAI) ]


7.1
7.6
15.7
51.2
61.7
64.2
81.7
107.8
128.8
163.7
174.7
182.7


53.4
42.0
32.3
- 29.8
- 1.9
3.9
5.4
7.3
9.5
14.1
15.1
15.4
50.7
53.3
58.2
67.4
68.7
72.2
92.3
144.7


[Co(2-PPI) ]


- 51.5
- 40.3
- 25.5
- 19.6
- 14.8
- 11.0
- 8.0
- 6.5
4.2
6.8
8.9
10.4
11.0
16.0
16.6
17.6
18.7
51.5
58.1
67.3
70.3
72.1
76.6
87.8
111.5
122.2










Table VIII continued



[Co(2-PCI)3 53.2 [Co(2-PMI) ] 51.1
40.4 38.6
28.8 28.1
21.1 20.7
6.3 6.2
8.3 8.2
11.1 10.0
15.1 10.6
16.1 11.6
17.9 16.0
19.4 16.4
51.8 16.9
59.3 17.7
69.6 51.5
72.7 57.1
78.1 58.0
89.5 64.4
112.0 70.8
125.0 74.5
77.5
97.5
115.7
Note: Chemical shifts are given in ppm from TMS in CD3CN
at 300 C.










Table IX
Effective Magnetic Moments of the Substituted
Tris(2-pyridinalphenylimine) Cobalt(II) Compounds


Compounds Oeff (BM) 'eff(soln) (B8)a

Co(2-PAH)3(BF4)2 b 4.61
Co(2-PPI) (BF4)2 4.68 4.60

Co(2-PAI) 3(BF4) 4.64 4.09
Co(2-PMI)3 (BF4)2 b 4.45

Co(2-PCI)3(BF4)2 4.42 c


aDetermined by the method of L6liger and Sheffold (27).

Undetermined

CCo(2-PCI)3(BF )2 was not sufficiently soluble to determine
Ueff in solution.






69

In addition, a mixture of the two isomers is obtained; there-

fore, two superimposed spectra resulted. Interpretation of

contact shifts of a mixture of these complexes of such low

symmetry would be even more difficult than for D3 complexes.

Consequently, neither assignments of the contact shifts nor

a discussion of their significance was attempted.















BIBLIOGRAPHY


1. E. K. Barefield and D. H. Busch, Quart. Rev., 22,
457 (1968).

2. G. C. Bassler and Robert M. Silverstein, Spectrometric
Identification of Organic Compounds (John Wiley & Sons,
Inc., New York, 1968), pp. 64 78.

3. I. Bertini and D. Gatteschi, Inorg. Chem., 12, 2740
(1973).

4. N. Bloembergen and L. 0. Morgan, J. Chem. Phys., 34,
842 (1961).

5. E. A. Clevenger, M. S. Thesis, The University of Florida,
Gainesville, 1961.

6. F. A. Cotton, Chemical Applications of Group Theory
(Interscience Publishers, New York, 1967), p. 285.

7. D. R. Eaton, NHR of Paramagnetic Molecules (G. N. La Mar,
W. DeW. Horrocks, Jr., and R. H. Holm, Eds., Academic
Press, New York, 1973), pp. 179 214.

8. E. Fermi, Z. Phys., 60, 320 (1930).

9. H. M. Fisher, M. S. Thesis, The University of Florida,
Gainesville, 1964.

10. H. M. Fisher, Ph. D. Dissertation, The University of
Florida, Gainesville, 1966.

11. H. M. Fisher and R. C. Stoufer, Inorg. Chem., 5, 1172
(1966).

12. R. M. Golding, Molec. Phys., 8, 561 (1964).

13. R. M. Golding, W. C. Tennant, J. P. M. Bailey, and A.
Hudson, J. Chem. Phys., 48, 764 (1968).

14. J. S. Griffith, The Theory of Transition Metal Ions
(Cambridge University Press, London, 1961), pp. 106-
134.









15. W. DeW. Horrocks, Jr., NMR of Paramagnetic Molecules
(G. N. La Mar, W. DeW. Horrocks, Jr., and R. H. Holm,
Eds., Academic Press, New York, 1973), pp. 128 175.

16. C. M. Hurd and P. Coodin, J. Phys. Chem. Solids, 28,
523 (1967).

17. J. P. Jesson, J. Chem. Phys., 47, 579 (1967).

18. J. P. Jesson, NMR of Paramagnetic Molecules (G. N. La Mar,
W. DeW. Horrocks, Jr., and R. H. Holm, Eds., Academic
Press, New York, 1973), pp. 2 51.

19. F. Keffer, T. Oguchi, W. O'Sullivan, and J. Yamashita,
Phys. Rev., 115, 1553 (1959).

20. P. Krumholz, J. Am. Chem. Soc., 75, 2163 (1953).

21. G. N. La Mar, J. Chem. Phys., 43, 1085 (1965).

22. G. N. La Mar, Inorg. Chem., 10, 2633 (1971).

23. G. N. La Mar, NMR of Paramagnetic Molecules (G. N. La Mar,
W. DeW. Horrocks, Jr., and R. H. Holm, Eds., Academic
Press, New York, 1973), pp. 86 123.

24. G. N. La Mar and C. R. Van Hecke, J. Am. Chem. Soc., 91,
3442 (1969).

25. J. Lewis and R. G. Wilkins, Eds., Modern Coordination
Chemistry (Interscience Publishers, Ind., New York, 1960),
pp. 229 300, pp. 400 454.

26. W. B. Lewis and L. O. Morgan, Transition Metal Chemistry
(R. L. Carlin, Ed., Marcel Dekker, Inc., New York, 1968),
Vol. 4, pp. 33 108.

27. J. Lbliger and R. Scheffold, J. Chem. Ed., 49, 646 (1972).

28. H. M. McConnell and D. B. Ches':nut, J. Chem. Phys., 28,
107 (1958).

29. H. M. McConnell and R. E. Robertson, J. Chem. Phys., 29,
1361 (1958).

30. J. March, Advanced Organic Chemistry: Reactions, Mech-
anisms, and Structure (McGraw-Hill Book Co., New York,
1968), p. 22.

31. G. E. Matsubayashi, M. Okunaka, and T. Tanaka, J. Organo-
metal. Chem., 56, 215 (1973).

32. L. E. Orgel, Discuss. Faraday Soc., 26, 92 (1958).








33. L. E. Orgel, J. Chem. Soc., 3683 (1961).

34. J. A. Pople and D. L. Beveridge, Approximate Molecular
Orbital Theory (McGraw-Hill Book Co., New York, 1970),
pp. 58 83.

35. M. A. Robinson and D. H. Busch, Inorg. Chem., 2, 1171
(1963).

36. M. J. Scarlett, A. T. Casey, and R. A. Craig, Aust. J.
Chem., 23, 1333 (1970).

37. J. G. Schmidt, W. S. Brey, Jr., and R. C. Stoufer, Inorg.
Chem., 6, 268 (1967).

38. P. W. Selwood, Magnetochemistry (Interscience Publishers,
Inc., New York, 1956), p. 78.

39. M. G. Simmons and L. J. Wilson, Inorg. Chem., 16, 126
(1977).

40. R. C. Stoufer and D. H. Busch, J. Am. Chem. Soc., 78,
6016 (1956).

41. R. C. Stoufer, D. W. Smith, E. A. Clevenger, and T. E.
Norris, Inorg. Chem., 5, 1167 (1966).

42. T. J. Swift, NMR of Paramagnetic Molecules (G. N. La Mar,
W. DeW. Horrocks, Jr., and R. H. Holm, Eds., Academic
Press, New York, 1973), pp. 53 81.

43. M. Wicholas, Inorg. Chem., 10, 1086 (1971).

44. D. L. Williams, D. W. Smith, and R. C. Stoufer, Inorg.
Chem., 6, 590 (1967).















BIOGRAPHICAL SKETCH


Joseph B. Vaughn, Jr, was born April 21, 1942, in Elfers,

Florida. He graduated from Paxon High School in Jacksonville,

Florida, in June, 1960. After a year of work and frugality,

he entered the University of Florida in 1961 and completed

his freshman year. After three more years of work, frugality,

and military service, he returned to the University of Florida

in September, 1964, and received the Bachelor of Science

degree with Honors in Chemistry in June, 1968.

Graduate study began in September, 1968, at the University

of Florida. From September, 1968, until June, 1973, he held

a teaching assistantship, and in September, 1973, no longer

eligible for an assistantship, be began work for the Pharmacy

Department at the University of Florida. From December, 1973,

until June, 1976, he held a medical technician position at

the V. A. Hospital in Gainesville, Florida. He also held an

Instructor's position, June, 1976, to September, 1976, with

Santa Fe Community College in Gainesville, Florida. Returning

to his graduate work at the University of Florida in September,

1976, he received the Ph. D. degree in August, 1977.











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 Philos phy.




R. Carl Stoufer, Chairman
Associate Professor'of Ch istry







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.




Wallace S. Brey, Jr.
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.




Gerhard M. Schmid
Associate 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.




Richard D. Dresdner
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.




Paul Urone
Professor of Enviromental
Engineering Sciences







This dissertation was submitted to the Graduate Faculty
of the Department of Chemistry in the College of Arts and
Sciences and to the Graduate Council, and was accepted as
partial fulfillment of the requirements for the degree of
Doctor of Philosophy.

August 1977


Dean, Graduate School





























































UNIVERSITY OF FLORIDA


3 1262 08553 5226




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 EUGDCUCIL_99AZBR INGEST_TIME 2012-03-09T00:15:59Z PACKAGE AA00003921_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES