STUDIES IN THE STEREOCHEMISTRY OF THE … · means of determination of the zero-field splitting ......

STUDIES IN THE STEREOCHEMISTRY OF THE MANGANESE(II) ION

AND ITS USE AS A STRUCTURAL PROBE

A Thesis submitted for the Degree of

DOCTOR OF PHILOSOPHY

of the University of London

by

ROSHUN BEJON BIRDY, B.Sc.,A.R.C.S.

Department of Chemistry,

Imperial College of Science and Technology,

London SW7

October 1977

2

ASSTRP.CT

9GHz and 36GHz electron spin resonance spectra have been

obtained of some polycrystalline coordination complexes of manganese(II),

and of manganese (II) impurity in complexes of zinc, cadmium, mercury,

magnesium, cobalt and nickel. Spectra are analyzed with particular

attention to the generally neglected 'off-axis' transitions (of

both the AMs ±1 and AM = ±2 types) which are shown to contribute

strongly to powder absorption for the range of D values (0.02 - 1.2 cm-1)

found here. In contrast with previous esr work on Mn(II), computer

calculation of spectral lineshapes is employed extensively, either

as a check on values of D and a derived directly from spectra of low

AH and ID/INI, or, where tnese conditions are not met, as the sole PP

means of determination of the zero-field splitting parameters.

Comparison of the esr parameters with those for related

complexes of known structure provides useful stereochemical information,

which is supplemented, where appropriate, by vibrational and electronic

spectroscopy and X-ray powder diffraction studies of the host materials.

The 2-methylimidazole complexes and Zn(pyz)2X2(X=Ct,Br) have

not been reported previously; the latter constitute unusual examples

of polymeric ZnL2X2 = pyridine or a cognate ligand), most complexes

of this stoicheiometry being tetrahedral monomers.

3

ACKNOWLEDGEMENTS

I am most grateful to Dr. Margaret Goodgame for her capable

supervision of this work. I should also like to thank Dr. J.F. Gibson

and members of his research group for many helpful discussions and

for generously making available computer programs and preprints of

their publications; the assistance of Mr. G.M. Lack in particular

has proved extremely valuable.

The award by the University of London of a Postgraduate Studentship

for the period 1972-1975 is gratefully acknowledged.

No words can express the depth of my gratitude to my parents

for their support and encouragement over the years.

FOR RUSHNA

CONTENTS

Page

Abstract

2

Acknowledgements 3

Abbreviations 6

Chapter 1 Introduction: esr of Mn(II) in the

polycrystalline state 7

Chapter 2 Esr spectra of Mn(II) in hexakis complexes

of 4-methylpyridine N-oxide 18

Chapter 3

Spectral studies of some Mn(II)-doped

chelate complexes

Part I: MI [WII(acac)3 .nH

2 0

46

Chapter 4 Spectral studies of some Mn(II)-doped

chelate complexes

Part II: MII

(dike)2 (LL) and

M11

(LL)3i 2+ 75

Chapter 5

Chapter 6

Chapter 7

Esr and vibrational spectra of some

Mn(II)-doped polymeric complexes 121

Manganese(II) complexes of 2-methylimidazole 181

Experimental 221

References 223

ABBREVIATIONS

acac acety1acetonate

bipy 2, 2'-bipyridyl

monoprotonated 4,41 -bipyridyl

dab monoprotonated dabco(q.v.)

dabco 1,4-diazabicycloE2.2.21 octane

(or triethylenediamine)

dbm dibenzoylrnethanate

dike s-dike tone

1,2Dimeim 1,2-dimethylimidazole

DMF dimethylformamide

DMP 2,2-dimethoxypropane

en ethylenediamine

Et4dien bis(2-diethy1aminoethyl)amine

hfac hexafluoroacetylacetonate

Me5dien bis(2-dimethylaminoethyl)methylamine

2Meim 2-methylimidazole

4MeO-pyo 4-methoxypyridine N-oxide

4Me-pyo 4-methylpyridine N-oxide

OAc acetate

OMPA octamethylpyrophosphoramide

phen 1,10-phenanthroline

pic pyridine-2-carboxylate

py pyridine

pyz pyrazine

pzC pyrazole

6

CHAPTER 1

INTRODUCTION: ESR O MLIIGANESE (II) IN THE POLYCRYSTALLINE STATE

(a) Introduction

Electron spin resonance (esr) is now firmly established as one

of the most important spectroscopic techniques available for the

structural and electronic investigation of coordination complexes of

Kramers ions. When such an ion is placed in a magnetic field, the

degeneracy of each of the spin-doublets comprising the orbital ground

state of the ion is lifted. The extent of the splitting, which depends

on the strength of the applied field as well as on the spin quantum-

number characterizing the zero-field spin-doublet, is commonly of the

order of a few cm1, and the simultaneous application of microwave

radiation of frequency V may therefore induce transitions between the

spin levels at those magnetic fields, Bres,

where the relationship

119 = gBres

is satisfied. h in equation (1.1) is Planck's constant, (3 is the

Bohr magneton and g, the spectroscopic splitting factor.

The technique has been applied to a very large number of

transition metal ions1 of which the D-state ions consitute by far the

largest group. The reasons for this are not far to seek; the relative-

ly simple spectra deriving from transitions within a single Kramers

doublet are easily analyzed and the derived spin Hamiltonian (SPH)

parameters are usually readily amenable to theoretical treatment. In

systems of lower than cubic symmetry, and where S 3 1, both inter-

electronic dipolar interaction and coupling between electron Orbital

and spin angular momenta cause the degeneracy of the spin states

3

to be removed even in the absence of a magnetic field. This 'zerofield-

splitting' is represented in the spin Hamiltonian by a term of the

form S.D.S., where D is a traceless tensor of the second rank. The

spectrum is then characterized by the presence of a large number of

transitions that may be spread over a wide range of magnetic field.

Much esr work has been reported over the last two decades

on high-spin metal ions of configuration d5. In most cases the ions

Fe(III) and Mn(II) have been studied as impurities in binary compounds

(e.g. oxides, halides and chalcogenides)1-10 or in minerals,11-17

and

only recently has attention focused on coordination complexes as host

materials.18-25

Most of these solids have been examined as oriented

single crystals, which method permits not only calculation of the SPH

parameters to a high degree of accuracy but also, in a host of known

crystal structure, determination of the inclination of the SPH tensor

axes to the crystal and molecular axes. However, the difficulty in

many cases of growing satisfactory crystals, particularly of biological

compounds, necessitates the development of accurate and reliable powder

methods; in addition the latter possess certain positive advantages

over single-crystal studies. Thus powder spectra, which often contain

sufficient information for calculation of the more significant SPH

parameters, are much more easily and rapidly obtained than single-crystal

rotation patterns. They are also, in general, more readily interpreted,

particularly where the unit cell has many different orientations of

magnetically equivalent paramagnetic centres, or where the orientation

of the paramagnetic unit is not closely related to the crystal axes.

(b) The spin Hamiltonian for Mn(II)

The theory for the analysis of spectra of high-spin d5 systems

9

1 53 ' 54 is well established-. 1 The spectLa deserbe:d. in this work have

been analyzed using the electronic SPH (1.2) , it being assumed that

the principal axes of the g and D tensors are collinear.

H = grID.S D[S2 - 35/12] + E(S2 - s2) (1.2)

The effect of nuclear hyperfine (lf) interaction may be treated by

adding to (1.2) the term S.A.I.

The presence in Mn(II) of an orbital singlet ground state

(hereafter g.s.) is expected to result in a near-isotropic g-factor

close to the free-spin value, and this has been confirmed experimentally;

throughout this work, therefore, it is assumed that go = 2.00.

The terms a and F, which are quartic in the spin operator, have been

omitted from the Hamiltonian (1.2); this is justified in the case of

Mn(II), for which these quantities are negligibly small in relation to

the quadratic axial and rhombic distortion parameters, the extraction

of which is the primary objective of spectral analysis. D and E are

defined as

D + D D - D xx yy xx yy

D = D E = zz 2 2

DD and D being the elements of the diagonali'zed D-tensor. E xx yy zz

may therefore be loosely regarded as a measure of the bonding anisotropy

in the equatorial plane of the molecule, and D, of the difference in

bonding at the axial and equatorial positions. It should be noted

that the values of D and E are not unique but depend on the choice of

axes;55

the six inequivalent axis systems, corresponding to all possible

permutations of the x,y and z axes, give rise to six sets of values of

D and E for each set of experimental data. In the 'proper' axis system,

the use of which was first recommended by Blutberg56

and which is

employed here, X(= I E/D I ) has its minimum value, E and D are of

10

the same sign, and z is the direction of greatest or least bond

strength or field gradient. Tables which transform expe7:imental data

referred to any axis system have been given by Hall.55

(c) Solution of the SPH matrix

When S = 1 or 3/2, the secular determinant derived from (1.2)

factorizes, giving closed expressions for the energies of the spin levels

in a magnetic field, provided that B is parallel to one of the D-

tensor axes.53 For S = 5/2, this is true only in axial symmetry and

for BIlz, and exact diagonalization of the energy matrix generally

requires the use of computer techniques. However where the Zeeman

energy is either very large or very small compared with the fine structure

energy, approximate solutions for the energy levels may be obtained

from perturbation theory. Several authors have attempted such a cal-

culation for the case gl3B >>D; although there is lack of agree-

3'57-59

ment with regard to some of the higher order terms in the

expression for the reasonance positions, this is of no importance for

the present work in which the first order formula (1.3) has been used,

where applicable, only to obtain initial estimates of the SPH parameters,

further refinement being carried out by exact diagonalization.

(Bres)Ms÷÷ Ms - 1 = Bo - (Ms - 2)

r (3Cos20 - 1)

+ 3ECos 24 Sin2e

(1.3)

Bo in (1.3) is given by

11■21 e and (1) are Euler angles defining 9(3.

the magnetic field direction relative to the tensor axes. It is clear

from (1.3) that, to a first approximation, and within the limits of

applicability of perturbation theory, the ±5/2 ±3/2 transitions

are separated by 8D and the ±3/2 ±1/2 transitions, by 4D, for Bilz.

11

For B iI (i)y and (ii)x the respective separations are (i) 4D + 12E

and 2D + 6E and (ii) 4D - 12E and 2D - 6E.

(d) The spectrum of a powder

The microcrystallites in a powder sample are oriented in

all possible directions to the external magnetic field and all orien-

tations are equally probable; it may then be shown54

that for species

of axial symmetry, the contribution from any transition to the powder

Sine) intensity is proportional to if the transition probability dB/de'

(T.P.) is independent of 0. The Sine factor reflects the relatively

large number of crystallites with principal axes nearly perpendicular

to the field direction. It is obvious that the powder line intensity

is greatest where dB/de is small. In the more general case of ortho-

rhombic symmetry, the condition for observance of a powder line is

( DO ) a,b (4) a,b

= 0

where (a,b) are the coordinates of the critical point.60

The nature

of the critical points is determined by the sign of G61

where

[ L D2B 2 D

2B 2

B G(a,b) = (3114) - (-T?) 1

a,b (1.4) 4i 4

and p = cos 0 . If G(a,b) > 0, then (a,b) is the location of a saddle-

point and B(a,b) corresponds to a divergence in the powder pattern

(i.e. to a 'baseline crossing' in the first derivative spectrum). If

G(a,b) < 0, then B(a,b) is a shoulder in the powder pattern, appearing

as a 'peak' or 'trough' in the first derivative spectrum.

(e) The program ESRS

Where the electronic Zeeman and fine structure terms of (1.2)

are of the same order of magnitude the perturbation approach is not valid

and numerical diagonalization of the matrix is required to calculate

12

electron spin energies and wave functions; this has been performed,

for S=5/2, by Aasa,60

and by Dowsing Lind Gibson.62

The Fortran IV

version of the program ESRS, written by Dowsing62 and extensively

revised by Lack,63 has been employed in the present work. The program

sets up and solves the secular determinant, for any required orientation

of the external wagnetic field, at preset intervals over a specified

field range. (The computational techniques involved are more fully

described in ref. 62.) The manifold of spin levels is then scanned in

order to obtain the fields at which the energy difference between any

two of the levels equals the energy, hy , of the applied microwave

radiation. At these fields the energy matrix is set up again and the

probability of the transition calculated from the eigenvectors of the

relevant levels. The energy levels are numbered 1 to 6, or, where hf

interaction is taken into account, 1 to 36, in decreasing order of

energy; when two levels cross, they exchange labels.) Calculations

of this type, for Bilx, y and z (and assuming a= 0, giso = 2) have

been performed over a wide range of the zfs parameters D and E; in

order to facilitate comparison between experimental and predicted

resonances, the results were presented as a series of plots of

D vs. Bres

, each for a fixed value of X.62

The realization that single-crystal turning-points other than

those along the D-tensor axes could contribute substantially to powder

absornt 60

ion led to the modification of the plots by Aasa; however

the inclusion of the so-called 'off-axis' Iines increases the complexity

of the diagrams to the extent that their usefulness is severely

curtailed. There are other serious objections, detailed in Chapter 6,

to the application of the D-B plot method in powder analysis, and it

is now clear that simulation of at least part of the experimental

13

lineshape is necessary as a check on the correctness of the assigned

SPH parameters, for values of D comparable with the microwave energy.

Failure to recognize this fact has resulted in widespread misinterpre-

:ation of Mn(II) powder spectra in the recent literature.33'64-68

It

may also be mentioned here that the complexity of the spectra frequently

results in their being used, particularly in biological work,64

merely

to monitor the species present, with no attempt at detailed quantitative

interpretation.

(f) Lineshape calculations

The number of published lineshape calculations for. S = 5/2

systems is still very small. Those of Griscom69 and of Bray,70

being

for values of D/hV well within the perturbation limit, are of very

limited applicability and need not be considered further. The work of

Sweeney et al.50

is of considerably greater significance. These authors

calculated line positions and intensities by direct diagonalization of

the SPH matrix for values of IDI/hV between 0.5 and 2, and spherically

averaged spectra were cumulated using Gaussian lineshapes and fixed,

isotropic linewidths. The calculations were limited to the axial case

(E=0); the resonance field positions are then independent of cp and

the expression for the integrated intensity of the absorption spectrum

takes the particularly simple form

n/2 I(H) =

2 k Ef

o Pk (0) GCB-Bk(0)11 Sine de

(1.5).

G[13-Bk(0)] is a Gaussian lineshape centred on Bk(0) with half-width

AB, and Pk(e) is the probability of transition k. For the one system

studied experimentally, viz. [(Ph3 P)2 Cu ]

3 FeIII(S

2C202)3'

the powder

lineshape for IDI = 6.8 GHz (i.e. 0.24 cm 1)at both X- and Q-band

14

frequencies gave quite good agreement with experiment.50

The simulated

spectra of all the axial complexes studied in the present work have

been obtained by an analogous method (see Chapter 5).

In rhombic symmetry a closed-form expression of type (1.5)

is not available, and powder lineshapes must be generated by summation

of 'single-crystal' derivative spectra over all orientations of the

magnetic field. The problem of achieving an approximately uniform

distribution of field directions over unit sphere has been dealt with

by Kopp and Mackey,52

and by Gibson and co-workers,72,73

in the following

manner. Each vertex of a regular icosahedron is equidistant from its

nearest neighbours on the surface of the circumscribing sphere; radii

through these points therefore represent a set of uniformly distributed

fields at a density of 3 per octant of the sphere. By quadrisection

of each triangular face of the icosahedron, a further set of field

directions (defined by radii through the vertices of the smaller triangles)

is generated; these are, to a good approximation, also uniformly

distributed. Although the process of subdivision may be repeated in-

definitely, the program SHAPE9, used in this work, allows for a maximum

of 345 field directions per octant (corresponding to four subdivisions

of each icosahedral face); lineshape calculations at higher field

densities are impracticable owing to restrictions on computer time. In

the present work the use of 93 fields per octant was found to be adequate

for most purposes, and should be assumed throughout except where stated

otherwise. The Euler angles e and (1)for these directions have been

listed by Price.72

Line positions and intensities at these field

orientations having been determined, the program assigns a shape (in

this work, Gaussian) and width to each line prior to accumulation of

the spectra The chief limitation of this lineshape program her in

the assumption of a fixed, isotropic linewidth. The effect of impurities,

15

dislocations and other defects within the crystallites is to alter

slightly the crystal-field (and fine structure) parameters at the sites

in question; it has been shown that, in strong magnetic fields, the

resulting broadening of the lines is greatest for transitions furthest

from the centre of the spectrum, and allowance must be made for this

in any attempt to fit an experimental lineshape with the use of SHAPE9.

A second limitation is the 'noise' in the computed spectrum which arises

from consideration of only a finite number of field orientations. Often

this is apparent only as slight undulations in the baseline which may

be disregarded for the purpose of fitting. However, incomplete spherical

averaging in the region of a highly anisotropic transition may result

in the appearance of a moderately strong 'phantom' band in the simulated

spectrum (see Chapter 5).

Despite these reservations, the program has been applied success-

fully to the study of a number of randomly oriented solids (both

powders and frozen solutions) containing Re(VI),63

Fe(III)72,73

and

Mn(II).38

For the d5 systems studied previously,38,50,72,73 lineshapes

were calculated either over very restricted ranges of the magnetic

field,38,73

or for strictly axial symmetry.50

In the present work,

the results of several such calculations are presented (for values of

D ranging between 0.06 cm-1

and 1.21 cm-1

) over the entire field range

available on the spectrometers, and with no restriction on A.

(g) Mn(II) as a stereochemical probe

Much of the present work is concerned with stereochemical

information derived from study of the esr spectra of Mn(II) impurity

in complexes of Zn(II), Cd(II), Hg(II), Mg(II), Ni(II) and Co(II).

Several features of the half-filled shell of the high-spin d5

16

configuration make the Mn(II) ion uniquely suitable as a structural

probe for divalent: metal ions:

(1) the lack of marked stereochemical preferences due to the

absence of LFSE facilitates adoption by Mn(II) of the

configuration of the host metal ion. This contrasts with

the behaviour of, for example, Cu(II) (see Chapter 4);

(2) the multielectron ion generally gives richer (and potentially

more informative) spectra than do ions with fewer unpaired

electrons;

.(3) the very small spin-orbit coupling with excited states

of the 6A1 g.s. results in long spin-lattice relaxation

times and spectra are readily observable at room temperature

and over a wide temperature range.

In addition, the spin associated with the 100% abundant 55Mn nucleus

(mi = 5/2) interacts with the effective spin of the electrons, producing

the characteristic, nearly isotropic 6-line hyperfine structure. In

the strong field limit the hf lines in the AMs = ±1 transitions are

evenly spaced, but at X-band frequency, measurement of higher order shifts

in the line positions may allow evaluation of the sign of D (Chapters

2, 3 and 4). For D, A <<g$130 the intensity of the 'forbidden'

Ami = ±1 hf lines relative to the Am, = 0 lines is a function of D

(see Chapter 5); this affords a useful means of estimating D when the

zero-field splitting (zfs) is too small to allow resolution of fine

structure transitions. Finally in favourable cases the axial and

rhombic nuclear electric quadrupole coupling constants may be calculated

from the separation of the AmI = ±1 and ±2 doublets.51,74

In conclusion, it is necessary to point out that the present

state of knowledge concerning (i) the fundamental origins of D and E,

and (ii) their relationship with the crystal field parameters, is far

17

from complete. S-state ions pose particular problems in this respect

as the crystal field does not couple directly to the spins and its

contribution to the g.s. splitting is small. Consequently any calculation

of this splitting must take into consideration several esoteric modes

of interaction between ground and excited states which may normally be

neglected for other d-electron configurationsiofthese the relativistic

crystal field75,76

is claimed to be of major importance in the case of

manganese(II). Attention is drawn to an extensive discussion of the

subject in an excellent recent review,77

and to a subsequent paper,78

presenting the most complete ab initio calculations to date of the axial

zfs of manganese(II) (in La2Mg3(103)12- 2

4H20) which affirms the pre-

dominance of the relativistic contribution to D. In connnection with

(ii) above, it may be noted that electronic transitions in manganese(II)

being both spin- and (for a centrosymmetric molecule) parity-forbidden,

the generally poor quality of the electronic reflectance spectra does not

normally allow evaluation of the axial components of the crystal field.

It must be emphasized, however, that imperfect understanding of the

detailed mechanisms of the g.s. splitting does not detract from the

usefulness of the SPH parameters in deducing site-symmetries and in probing

the local environment of the paramagnetic ion.

18

CHLPTER 2

ESR SPECTRA OF MANGANESE (II) IN HEXAKIS

COMPLEXES OF 4-METHYLPYRIDINE N-OXIDE

Introduction

Fifteen years after the first reports79,80

of hexakis transition

metal complexes with aromatic amine N-oxide ligands, interest in these

species continues unabated.81-91

Recent crystallographic studies of

the Ni(II),82

Co(II)83

and Hg(II)84

complexes of pyo confirm that, while

the MO6

chromophore is a regular octahedron, the M-0 and O-N bonds are

non-collinear, as expected for approximately sp2 hybridization at the

oxygen atom. As the pyridine ring is twisted out of the M-0-N plane,

the symmetry of the complex cation is no higher than S6. It has been

suggested83

that, as a result of 7T-overlap between metal d-electrons

and the vacant aromatic orbitals of the ligand, the effective perturbing

charges are not centred at the oxygen atoms. (For example, a simplified

MO or angular overlap model predicts, for the cobalt complex, spectroscopic

and magnetic properties of an octahedron elongated along its 3-fold

axis.83) A wealth of

ER85,86 and nmr

87'88

data does indeed appear to

confirm the presence of substantial back-bonding, as does the frequent

occurrence of charge-transfer absorption extending into the visible

region.80'89 Reduction of the room temperature magnetic moment in the

Co(II) complex from the value associated with octahedral Co(II) testifies

to the presence of a low-symmetry component of the ligand field.90

The

axial zfs in Ni(pyo)6(a04),,determined from susceptibility measurements

over a range of temperature, is 4.3 cm1;91

this greatly exceeds zero-

field splittings in the complexes NiL6(anion)2 (L being an imidazole or

pyrazole),92

and even in the Tutton salt Ni(NH4)2(SO4)2.6H20 where the

in-plane and axial Ni-0 distances of the tetragonal Ni(H20)62+

unit

0 are substantially different (1.9A and 2.15A)

93 D is only -2.24 .cm

-1 94

(see Table 3.13). Diffuse reflectance spectra of the Ni(II) complexes

of the N-oxides show no band-splitting at room temperature86

but it

is significant that the calculated values86

of the Racah parameter B

are unusually high (particular]y in the pyo, 4Me-pyo and 4Me0-pyo

complexes), a feature shared by the trigonal prismatic complexes described

in the following chapter.

None of the above spectroscopic techniques can provide the

accurate measurements of distortion from cubic symmetry which are

essential to any detailed understanding of the origin of this distortion.

Recent esr work43,45

on Mn(II) doped into the hexakis complexes of

Zn(II), Cd(II) and Hg(II) with some substituted pyridine N-oxides is

therefore of considerable interest, as the zfs parameters, which are

apparently readily obtained from X-band powder spectra, exhibit strong

dependence on the py-ring substituent. Rather surprisingly, the reported

values43,45 of D and E show no correlation with the size of the host

metal ion, or with the steric and electronic properties of the

substituents. Within the 4Me-pyo series, for example, the variation

in D was found to be only 12% while X ranged from 0 in the cadmium

complex to 0.23 for the mercury complex with an intermediate value for

the zinc compound.

In the present work the 4Me-pyo complexes are examined at both

X- and Q-band frequencies; several of the zfs parameters deviate

considerably from the values given in refs. 43 and 45.

Results

Previously,43,45 D and E for the 4Me-pyo complexes were obtained

using first-order perturbation theory, as described in Chapter 1. In

the present work, a first estimate obtained in this way was refined by

20

fitting the experimental data to resonance fields calculated by exact

diagonalization of a spin Hamiltonian matrix based on (1.2). Tables

2.1-6 list the observed spectra, together with calculated resonance

fields; the notation a-e for the Ms values follows that of ref. 3.

As only two parameters were to be determined, the fitting

procedure was applied over the outer transitions only; no attempt was

made to analyse the exceedingly complex geff = 2 region. Only the most

prominent of the many transitions observed between 0 and 200 mT in the

X-band spectra (Fig. 2.3) have been included in the tables. Calculated

single-crystal transition probabilities are in every case close to the

• 5:8:9:3:5 ratio required by first order perturbation theory,1

and are not

listed in the tables. Agreement with predicted fields is generally very

good, and markedly better than for the previously assigned43,45

parameters

(Tables 2.1-6), especially at Q-band frequency.

Simulated random-orientation spectra, excluding hyperfine

splitting, were plotted for both sets of parameters using SHAPE9; a

Gaussian line-shape was assumed and the isotropic peak-to-peak linewidth

of the first-derivative spectra taken as 30.0 mT. For the Zn(II) and

Hg(II) complexes, the experimental and calculated traces are compared in

Figs. 2.1 and 2.2. Once again, the parameters found in this work are

seen to give much the better fit.

Woltermann and Wasson43,45

do not list resonance fields for

any of the complexes. Their published spectrum of the Hg(II) complex

is generally similar to ours though there are some intensity differences,

and the lowest field transition seems to be at slightly higher field

in our spectrum. The difference in interpretation arises largely from

their failure to identify the weak bands between 100 mT and 160 mT in

all three complexes (see Fig. 2.3) as spin-forbidden (AMs r ±1)

transitions arising from off-diagonal elements produced by D. When

21

the oscillatory radiofrequency field, B1, is perpendicular to the

externally applied magnetic field B, as in the usual experimental

arrangement, the transitions AMs = ±2, Am

I = 0 have to a first approxi-

mation zero probability along the principal axes;95

for intermediate

directions of the magnetic field, when the T.P.'s are higher,

perturbatiOn theory, even when carried to third order, only approximately

predicts the energies of the spin-levels.4,13,95

Numerical analysis

of half-field transitions in Mn(II) has been attempted previously only

in single crystals of Naa:Mn(II)4,calcite13,14

and (with neglect of

off-diagonal hf matrix elements) tremolite.12

Aasa60

has stated that stationary values of Bcom (which

are associated with the appearance of lines in the powder spectrum)

occur only when the magnetic field lies in one of the principal planes.

Calculations for the Cd(II) and Hg(II) complexes show that the sextets

a and 13 (Fig. 2.3, Tables 2.7-9) centred on ca. 126 and 130 mT, may

be ascribed respectively to the 5-2[-1/2(-5/2, 1/2)-(-)-3/2(1/2, 5/2fl

and 5-3[-3/2(1/2, -1/2) 5/2, 1/2(-3/2, -1/2)J transitions in the

yz plane. Table 2.10 illustrates how T.P.'s for 5-2, in contrast to

those of spin-allowed transitions, are heavily dependent on'the orient-

ation of the magnetic field. (The hf interaction is neglected for these

calculations.) Since each of the 5-2 and 5-3 transitions attains its

maximum probability at roughly the same value of 0 at which the

resonance field is a maximum, powder lines may reasonably be expected

at these fields.

When resonance line positions for the AMs = ±2 transitions

are recalculated with the inclusion of the hf term in the Hamiltonian,

it is seen (Tables 2.8 and 2.9) that several components of 5-2 zx

coincide with those of 5-2 yz, and that the T.P.'s in the two planes

are similar. However, if the former transition contributed significantly

to the powder lines cc then the last component of this sextet (-152mT)

22

should be of lower intensity than the other five, and Fig. 2.3 shows

that this is not the case. One reason for the apparent unimportance

of the zx transition becomes clear on examination of Table 2.10. Beyond

the turning-point of Bres, which occurs at 8 - 30

o in both planes,

further increase in 6 causes the resonance field to fall off sharply

in the zx plane, but much more gradually in yz. Thus Bres for 5-2 yz

remains within 1 mT of its maximum value up to 0 -500 (it may be

noted here that AH for an individual hf line of this transition varies PP

between 1 and 2 mT); for the zx transition the corresponding value of

[ 0 is only -370 L . The behaviour of Bres for 0 < 30° is of little concern,

not only because (as explained in Chapter 1) contributions to the powder

intensity for these orientations of the magnetic field are relatively

unimportant in species of near-axial symmetry, but also because the

calculated T.P.'s are low in this region; at 0 = 10°, for example,

the T.P. of 5-2 yz is an order of magnitude lower than at 30°.]

In the Zn(II) compound, the 100-160 mT region is rather more

complex, and at least three distinct sextets may be recognized. The

set denoted by y (Table 2.7) may originate from a transition which is

nuclear-as well as spin-forbidden, the calculated probability of which

is slightly higher than in the Cd(II) and Hg(II) complexes.

Sextets associated with the transitions ax and b cannot be

y

identified unequivocally in the X-band spectrum of the Zn(II) complex,

and the remaining spin-allowed bands are fit equally well by A = 0.06

and 0.07. For the spin-forbidden lines, the difference between

corresponding resonance fields for the two values of X is on the order of

the linewidth; nevertheless A=0.06 does appear to give better agreement

with experiment (Table 2.7). We are unable to account for a band

centred on 478 mT in the X-band spectrum of this complex; no 'extra'

23

absorption corresponding to this can be detected at Q-band frequency.

Despite its rather low T.P, relative to the AMs = 1 bands,

the observed peak-height of 5-2 yz in all the three complexes is

comparable with that of ez; this is again a result of the dependence

on 0(the single-crystal turning point) of the contribution of a given

transition to the spherically averaged intensity.

Neglect of the spin-forbidden transitions might appear to be

51 validated by the single-crystal study of Zn(Mn)(pyo)6(C904)2, which

gave parameters in fairly good agreement with those from the powder

spectrum.43

However, this result is somewhat fortuitous since, for

the particular parameters concerned (D = 0.038 cm-1

, X = 0), the strongest

spin-forbidden transition (4-2, centred on 187 mT in yz) overlaps

ez(177 mT at 9.5 GHz). No attempt has been made in the present work to

analyse spin-forbidden transitions between 0 and 100 mT, the components

of which are interspersed with those of ez. Additional transitions

between non-adjacent levels are predicted up to -350 mT, but are

presumably buried under the much stronger allowed transitions.

In the Q-band spectra, spin-forbidden lines appear only at

300-500 mT, i.e. well out of the range of the allowed transitions,

and, as the spin-levels involved are characterized by almost pure wave-

functions, their T.P.'s are quite low. In addition, location of

the centre of each group of six hf lines is far easier at the higher

frequency, where the hf components are almost equally spaced. The

determination of A by computing the average of all the hf separations

is also best performed at Q-band. For all throe complexes, the

magnitude of A calculated in this way is 8.9 mT( = 0.00832 cm-1).

This differs from the value (9.5 mT) given by Woltermann and Wasson43,45

by more than the difference between all their compounds, and it seems

24

that such small variations should be viewed with caution when

measured at X-band only.

The relative signs of A and D are readily determined by

comparison (at X--band) of the mean spacing of the hf lines in the

±3/24--4.±5/2 transitions when B is parallel to z. In all three com-

plexes this spacing is greater for the highest observed transition

than for the lowest allowed band; thus if A is negative, as is usually

the case for Mn(II),1

these transitions correspond respectively to

-3/2 4-÷-5/2 and 3/24-4-5/2,1 and D is positive, as in Zn(Mn)(pyo)(a° )2.

51

Recent crystal-field calculations for the d5 configuration indicate

that for a 6-coordinate species of D3d symmetry, a positive value of D

corresponds to trigonal compression.96

Ifs therefore i the effective micro-

symmetry of Zn(Mn)(pyo)6

may he taken as D3d, the distortion here

would appear to be in the opposite sense to that predicted on the AOM

for Co(pyo)(cZ04 )2'83 and found by Mossbauer spectroscopy and magnetic

anisotropy studies in the Fe(II)97

and Ni(II)98

isomorphs.

In contrast to all the other observed transitions, the hf

spacing of ez decreases from left to right across the sextet; similar

observations have been made previously in Mn(II)-doped single crystals

of LiVO3(D -0.008cm-1)

99 and CaV206(D-0.04cm 1),100 and perturbation

expressions for the hf spacing have been given 99,100

which predict that

the effect occurs for Ho/6< IDI<Ho

/2 (at X-band, this corresponds to

values of D between 0.05 and 0.16cm-1). Table 2.11 lists observed and

computer-calculated hf line positions of ez for Hg(Mn)(4Me-pyo)6(CR04)2.

The 'doubling' of the first line (due to the crossing in this region of

the magnetic field of levels 7 and 8) is not clearly seen in the

experimental spectrum, the predicted separation of the two components

being roughly equal to the linewidth.

25

Throughout this work, fine structure resonant fields have been

measured at the midpoint of the third and fourth. lines of each hf group;

in other words, it has been assumed, as in much previous work, that

second- and higher-order hf effects do not shift appreciably the centres

of gravity of the sextets. This is a sufficiently good approximation

where D<...hV or for those polycrystalline spectra in which large line-

,widths preclude the measurement of resonance fields to within 2.0 -

3.0 mT. However for single-crystal studies, or in random-orientation

spectra where linewidths are low, such shifts may be significant. In

view of the disagreement in the literature57

about the identity and sign

of many of the third-order terms in the perturbation expression for

B(Ims, mi > 1 Ms-1, mI>) it is not surprising that no detailed analysis

of these shifts appears to have been attempted.

Tables 2.12 and 2.13 compare X-band resonance fields calculated

for A = 0 with the 'centre of gravity' of the six lines obtained by

complete solution of the 36 x 36 SPH matrix. It is found that shifts

of lines near the centre of the spectrum are close to the expected value

of — 0.6 mT, which is well below the linewidth generally observed in

random-orientation spectra. For az, e z , e

y (but not'a ) and e

x, however,

these shifts may be as high as 2.4 mT. It has been claimed previously42

that the error resulting from measurement of res as the midpoint of the

third and fourth hf lines is largely eliminated, as zfs parameters are

calculated from field separations, rather than absolute line positions;

while this may be true for some transitions, it is to be noted that az

and ez in Tables 2.12 and 2.13 are shifted by -2.0 mT in opposite

directions.

When corrections of the magnitude indicated by the results of

Tables 2.12 and 2.13 were made to the raw experimental data (of

Tables 2.3 and 2.5) there was marked improvement in the fit of the

26

experimental lino---positions to the fields predicted for the previously

assigned zfs parameters. The residual error in the experimental fine

structure resonant fields may well be sufficiently small to allow deter-

mination of la-F1 and giso by application of the least-squares fitting

program MNES;38

this has not been attempted here, however,.

It need hardly be pointed out that where computer calculation

of higher order shifts is not feasible, errors in measurement of the

fine structure fields are greatly reduced by the simple expedient of

taking Bres

as the midpoint of the first and sixth, rather than the

third and fourth, hf lines.

Discussion

Esr spectroscopy provides a fast and accurate means of

measuring the g.s. splitting of the Mn(II) impurity in these Group IIB

complexes. Even in the transition metal analogues, only single-crystal

magnetic anisotropy measurements yield data of comparable (and usually

lower) accuracy.

The detailed mechanism giving rise to sizeable zfs in the

absence of crystallographically detectable steric distortion remains

obscure. As has been pointed out previously,43,45

the esr results

suggest that the electronic properties of the 4-substituents are of

crucial importance. The present work shows that errors in the computed

zfs parameters of refs. 43 and 45 may be as high as 10 mT; nevertheless

variation with Z of the axial zfs in M(Mn)(4Z-pyo)6(CE04)2 is sufficiently

great that the order of ID I may reliably be taken as Z CH3>H -NO2>CN.

It is interesting that the values of the nephelauxetic ratio 13 in the

Ni(II) complexes of the first three ligands lie in the same order,86

and

may therefore reflect the extent of departure from Oh microsymmetry,

as discussed in Chapter 3, rather than the degree of covalence in the

M-0 bonds.

27

There is some evidence that hexakis complexes with, for example,

pyrid-2-one92

and dialkyl and diaryl sulfoxides92,97

are similarly

distorted, and esr investigation of the Mn(II) or Mn(II)-doped complexes

would appear to be worthwhile.

It has been suggested51

that axially located counterions in

Zn(Mn)(pyo)6(a04)2 might be partly responsible for the large zfs.

Subsequent X-ray diffraction work has shown, however, that the isomorphous

Ni(II),82

Co(II)83

and Hg(II)84

complexes have the 'fluorite' structure,

the anions occupying the tetrahedral holes in a ccp array of cations;

thus the eight nearest neighbour anions associated with each cation

are symmetrically disposed about it, and would have little effect on

the g.s. splitting of the Mn(II) ion. It may also be pointed out,

pace the same authors,51

that reduction from 0h of the local symmetry

of the cation does not of itself lead to g.s. splitting of the observed

magnitude; in the hexakis antipyrine complexes referred to in

Chapter 3,101-103

and in Cd(imidazole)6(NO3)2'104

the overall symmetry

of the complex cation is also S6, yet the measured values of D

44,105

are in every case less than one-tenth of those in the pyo complexes.

The redetermined D-values of the 4Me-pyo complexes are found

to lie in the inverse order of the ionic radii of the host metal ions,

a result which has been obtained previously for at least two different

series of isomorphous complexes doped with Mn(II)44

or Gd(III).106

In contrast to the behaviour of the rare earth trifluoride series,106

however, the relation between D and host size is non-linear, and it is

likely that Cd(4Me-pyo)6(a04)2 is structurally distinct from the Zn(II)

and Hg(II) analogues, as is the case in the parent (pyo) complexes.82

In view of the rather large uncertainties in the value of X in refs.

43 and 45, discussion of the variation of this parameter with Z is

unwarranted. In the present complexes X is independent of the host

28

metal; the reason for its departure from zero is not clear.

Preparation of the Complexes

The yellow, crystalline complexes precipitated immediately on

addition of ethanolic solutions of the hydrated metal perchlorates

(containing — 1 mole % of Mn(Ct04)2. 61120) to a slight stoicheiometric

excess of the ligand (mole ratio ca. 1:7) in the same solvent. All

three solids are light-sensitive and spectra were therefore run on

freshly prepared samples.

29

Table 2.1 X-band spectrum (mT) of Zn(Mn)(4Me-pyo)6(Ce04)2

Observed

0)=9.525GHz )

Calc. for

D=0.063cm-1

,

A=0.06

Caic. for

D=0.063cm-1,

A=0.07

Calc.b for

1 D=0.072cm ,

A=0.188

72 ms 71.2 ez 71.5 ez 69.2 e

127 ms

130 ms

a

a

128.1 ay

197 ms 201.2 a 197.2 a 192.1 d Y y z

205.6 dz 205.8 dz 225.2 b

Y

254.s 250.0 ax 254.0 a

256.6 b Y

255.1 b y

strong

overlapping

absorption

436.3 dy

475 ms 474.8 bz 474.7 bz

478 ms 489.2 bz

506 w 505.6 e 509.5 e Y Y

585.9 e Y

608 w 610.2 az 610.2 a 649.5 az

a Assigned as spin-forbidden transition. See Table 2.7.

b Parameters from ref. 43.

30

Table 2.2 Q--hand spectrum (ruT) of

Zn(Mn) (4Me-PY0)6 (cP-04 ) 2

Observed

(V=35.72CHz)

Calc. for

D=0.063cm-1 ,

X=0.06

Caic. for

D=0.063cm-1 ,

X=0.07

Calc.a for

D=0.072cm 1,

X=0.188

968.5 ez

1004 m 1006.2 ez 1006.2 ez 1038.8 ay

1117 ms 1120.5 a 1116.4 a 1121.2 d Y y z

1140 ms 1141.0 dz 1141.0 dz 1152.5 by

1171 w 1169.8 ax 1173.9 a

1190 s 1193.1 b 1191.1 b Y Y

1216 s 1216.1 bx 1218.1 bx 1215.7 ax

1324 s 1325.8 dx 1323.7 dx

1354 s 1351.1 dY

1353.2 dY

1349.3 ex

1389 w 1390.1 ex 1386.2 ex 1392.0 dy

1411 ms 1410.9 bz 1410.8 bz 1429.2 b

1441 ms 1437.9 e 1441.9 e 1519.6 e Y

y y

1550 m 1545.9 az 1545.9 az 1584.9 az

aParameters from ref. 43

31

Table 2.3 X-band spectrum (mT) of

Cd (Mn) (47'0-pY0) 6 (C'e04) 2•

Observed (V=9.522GHz) Caic. for

-1 D = 0.061cm ,

X=0.06

Calc.b for

-1 D=0.064cm r

X=0

Measured at

mid-point of

3rd and 4th hf lines

Corrected for

2nd and higher

order hf shiftsa

82 m 80 79.5 ez 64.3 ez

127 m c

130 m c

202 m 203 204.7 a Y

203.1 d z

+ v. weak 209.8 dz 224.1 a x,y

signal

251 w 252.0 ax

258 ms 258 258.9 b Y

264.3 b x,y

276.1 bx

313 m 314.0 cx

316.0 c x,y

320 s 321.5 c Y

340.2 cz

464 w 466 456.0 ex 383.5 d x,y

468 ms 469 470.0 bz

477.2 bz •

498 w 500 500.2 e Y

484.9 e x,y

597 w 599 601.5 az 614.3 a

aSee Tables 2.12 and 2.13 and text.

bParameters from ref. 43

dAssigned as spin-forbidden transition. See Table 2.8.

32

Table 2.4 Q-band spectrum (mT) of

Cd(Mn)(4Me-PY0)6(a04)2

Observed

(V=35.73GHz)

Calc. for

D=0.061cm 1,

X=0.06

Calc.a for

1 D=0.064cm ,

X=0

1014 m 1014.9 ez 1002.0 ez

1124 ms 1125.5 ay 1139.1 a x,y

1145 ms 1145.5 dz 1143.3 dz

1173.2 ax

1194 s 1196.0 b

1217 s 1218.4 bx 1203.6 b x,y

1324.6 dx

1346 s 1349.1 dy 1339.6 d x,y

1386.6 ex

1401 ms 1406.8 bz 1413.3 b

1430 ms 1432.8 e 1416.4 e Y x,y

1535 m 1537.5 az 1550.3 a

parameters from ref. 43.

33

Table 2.5 X-band spectrum (mT) of

Hg(Mn)(414e-pyo)_6(Ct04)2

Observed (v=9.531Glz)

Calc. for Calc.b for

Measured at Corrected for -1 -1

mid-point of 2nd and higher -_, D=0.059cm , D=0.067cm ,

3rd and 4th hf lines order hf shifts 1=0.06 1=0.23

90 w 87 88.3 ez 60.8 ez

126 w

131 w

c

c

121.9 ay

206 m 207 208.6 ay 202.5 dz

214.3 dz

224.4 by

254.5 ax

260 m 261 261.6 by

278.7 b

401.8 ex

443.2 d Y

452 m 453 452.4 ex

466 m 466.5 bz

477.2 bz

494 ms 496 495.2 e Y

589 w 592 593.3 az 586.5 e Y

628.6 az

as Tables 2.12 and 2.13.

bParameters from ref. 45.

cAssigned as spin-forbidden transition. See Table 2.9.

34

Table 2.6 p-band spectrum (mT) of

1.19(Mn)(4Me-pyo),(3(a04)2

Observed

(V=35.76GHz)

Cale. for

D=0.059cm-1,

X=0.06

Calc.a for

D=0,067cm 1,

X=0.23

991.6 ez

1029 m 1024.8 ez 1038.0 ay

1134 ms 1131.5 ay

1133.2 dz

1154 ins 1151.0 dz 1153.9 by

1177.7 ax

1201 s 1199.9 by

1223 s 1221.6 bx 1239.3 ax

1323 s 1324.4 dx 1327.6 ex

1348 s 1348.1 d Y

1384.2 ex 1395.1 dy

1404 ms 1403.7 bz 1419.6 bz

1430 ms 1428.8 ey

1532 m 1530.2 a 1521.8 e z Y

1565.1 az

aParameters from ref. 45.

Table 2.7 Ms 1 transitions (110-160mT) in the X-band spectrum of Zn(Mn)(4Me-yo)6(a0d2

Observed Calculated for 0..0.053c m-1, X..0.06{0.071, 70.-8.9=7 '

(v..9.525G8z) 5-2 transition, yz plane 5-3 transition, yz plane 5-2 transition, yr plane 5-2 transition, yz plans i 5-2 trzns:tion,

0-40°, 0-90°, Ary-0 0-70°. -90°, ami-o -to°, ::,,-Do',tx, -1 .1 -4e, -90' , .-1 "3-='"F, st-0°. 0 ;,

o107..Ss . 108.0(108.8)(1.1) a

6109.8m [-5/2]b 110.2(110.8} (0.3) 110.7(111.5) (0.1)

Y112.4m [-5/23 [-5/2.-3/23 112.5(113.6) (0.1) 112.5(111.0) (0.5)

[-7,./2 ,*-5/23 [-1/2]

al15.0ma 115.5{116.41(0.9)

6118.1w [-3/2] 118.4(119.0) (0.3) 118.5{119.3) (0.1)

Y120.Cms,br E3/23 [-3/2"'-1/2] 120.7{121.5) (0.2) 12:-...2-',12O.71 f0.3)

[1/2-j

e123.3ms 123.7{124.6)(0.9)

6126.4w [-1/2] 127.1(127.7)(0.2) 127.0{127.9) (0.1)

y130.0ms,br [H1/2] [-1/2.-.- 1/2] 129.3(130.1) (0.1) 131.1(130.1) (0.9)

[ 1/2-, -1/21

al32.1ms 132.6{133.5)(1.0)

6135.5w [1/2] 136.4{137.0)(0.3) 136.3(137.2) (0.05)

y139.3ms,br [1/2] [1/2 . 3/2] 138.6(139.4) (0.1) 140.7(139.6) ;9.9)

[3/2 - 1/2] [5/2]

al41.9ms 142.3(143.1) (1.1)

5145.0w [3/23 146.0(146.6) (0.3) 146.4(147.3) (0.02)

Y149.7ms,sh [3/2) [3/2 .5/2] 142.6(149.4) (0.1)

0152.2s 152.7(153.5) (1.2) [5/2 -, 3/2]

6155.3m [5/2] . 155.9(156.5) (0.4)

[5/23

aT.P. in parentheses

bm in square brackets.

Table 2.8&Ms 1 transitions (100-160 mT) in the X-band spectrum of Cd(Mn)(4Me-pyo)6(a04)2

Observed

(v=9.522GHz)

% Calculated for D=0.061cm

-1 , A=0.06, A=-8.9mT

5-2 transition, yz plane

8=30°, 4=90°; Ami = 0

5-2 transition, zx plane

8=30°, (1)=00; AmI = 0


8=80°, (I)=90°; Am1 = 0

105.0(0.76)

al07.3 m 108.2(0.91)a [-3/21

[-5/2.1b

5109.8 m 110.4(0.10)

113.3(0.73) [-5/2]

a115.0 m 115.3(0.85) [-1/2]

[-3/2]

13117.6 w 118.1(0.08)

122.2(0.74) [-3/2]

a122.8 m 123.2(0.83) [1/2]

[-1/2]

126.4 w 126.5(0.08)

[-1/2]

continued over/

Table 2.8 Continued

Observed A Calculated for D=0.061cm 1, A=0.06, A=-8.9mT

(V=9.522GHz) 5-2 transition, yz plane

Q=30°, (1)=90°,-AmI = 0


0=30°, (1)---e ; Amt = 0


e=soo, q =90o; dmI = 0

al31.9 m 132.1(0.85) 131.6(0.78)

[1/2] [3/2]

_ fil35.2 w 135.6(0.08)

[1/2]

al41.4 m 141.8(0.89) 141.2(0.86)

[3/2] 1.5/2]

13145.1 w 145.3(0.09)

[3/]

a151.8 m 152.3(0.93)

[5/2]

155.4 m 155.5(0.10)

[5/21


b i mI in sauare brackets

Table 2..9&M001 transitions (110-160mT) in the X-band spectrum of Hg(Mn)(4Ms-nyo)6(CY-04)2

Observed Calculated for D=0.059cm- , X=0.06, A=-8.9mT

(V=9.531GHz) 5-2 transition, yz plane

0=30°, cb=90°; Am, = 0


0=30°, 4=00; Am, = 0


0=80°, ,t=90°; Am, = 0

113.9(0.67)

aca.114.0 s - [-1/2]

115.4(0.80)a

13117.6 w 1-3/21 3 118.3(0.08)

[--3/1

al22.8 s 123.4(0.78) 122.8(0.67)

13126.4 w [-1/21 [1/2] 126.7(0.07)

[-1/2]

al31.9 s 132.2(0.79) 132.1(0.71)

13135.3 w [1/21 [3/2] 135.8(0.08)

[1/2]

continued over/

Table 2.9 Continued

Observed

(V=9.531GHz)

I Calculated for D=0.059cm-1 A=0.06, A=-8.9mT


8=30°, (I)=90°; Am, = 0


0=300, (1)=00; Am, = 0


8=80°, ,*=90°; Am, = 0

al41.5 s

8145.1 w

al51.6 s

5155.2 w.

141.9(0.83)

[3/1

152.4(0.90)

15/2]

141.8(0.78)

[5/2] 145.4(0.08)

[3/2]

155.7(0.08)

[5/2]

aTP in parentheses

bmI in square brackets

40

Table 2.10 Calculated resonance fields (A-0) for the 5-2

transition in the yz and zx planes; in

Hg(Mn)(4Me-pyo) (Ct0i)2

yz plane (q)=90°) zx plane (c)-0)

0(o) Bres(mT) T.P. B res(mT) T.P.

10 122.8 0.03 117.1 0.27

20 130.0 0.44 120.1 0.59

30 131.4 0.88 120.4 0.81

40 131.1 1.06 119.2 0.85

50 130.3 1.02 115.8 0.69

60 125.9 0.66 99.7 0.34

70 94,0 0.29 50.8 0.22

Table 2.11 The ez transition in Hg(Mn)(4Me-pyo)r-o(Ct04)2

Observed

Bres

(mT)

Calculated

B res(mT) T.P. Levelsa

67.51 5.00 8-1 68.4

69.59 5.00 7-1

77.7 77.15 4.98 8-2

86.0 85.68 4.97 9-3

94.3 93.87 4.97 10-4

102.1 101.72 4.97 11-5

109.6 109.25 4.98 12-6

aThe levels here are numbered 1-36, of which 1-6 and 7-12 comprise the spin states 5/2 and 3/2.

41

Table 2.12 - Calculated resoaance fields -1,

'N=0.06, g =2 1s07

Transition

Bres

(A=0)

Bres (A=-8.9mT)

Mid pt. of 1st

and 6th hf lines

Mid_pt. of 3rd

and 4th hf lines

ez 88.3 88.3/89.4 89.8

ay

203.6 208.5 208.0

ex

452.4 451.9 450.6

bz 466.5 466.4 466.0

e 495.2 494.7 493.4 Y

az 593.3 592.6 590.9

Table 2.13 Calculated resonance fields (mT) for D=0.061cm-1 ,

A=0.06, gz=2.00, gx-gy 2.01

Transition

Bres

Bres (A=-8.9mT)

(A=0) Mid pt. of 1st

and 6th hf lines

Mid pt. of 3rd

and 4th hf lines

ez 79.1 79.9 80.2/80.8

aY

203.7 203.5 203.0

by

257.6 257.4 257.0

dY

394.7 394.6 394.5

eY

497.7 497.2 495.9

az 601.5 600.9 599.1

42

Table 2.14 Amaytical Results

Compound Found Calculated

%C %H %N %C %H %N

Zn(4Me-pyo)6(Ct04)2

Cd(4Me-pyo)6(Ct04)2

Hg(4M0-pY0)6(Ce04)2

47.7

44.7

41.2

4.6

4.5

4.0

9.2

8.9

8.1

47.1

44.8

41.0

4.6

4.4

4.0

9.1

8.7

8.0

43

(a)

16247 AIM rnT

Fig. 2.1 Esr spectra (9.525GHz, 0-750mT) of Zn(Mn)(4Me-pyo)p,e04)2: (b) experimental; (a) and (c) computer-simulated lineshapfs(giim=30mT) for D=0.063cm-1, A=0.06, and D=0.072cm-1, 1=0.188 resnectively.

400 B/mT

200 0 600

(a)

(b)

X10

44

(c)

Fig. 2.2 Esr spectra (9.531GHz, 0-600mT) of Hg(Mn)(4Me-pyo)6(a04)2;

(b) experimental; (a) and (c) computer-simulated lineshapes

(AHpp=30mT) for D=0.059cm-1, 1=0.06 and D=0.067, a=0.23 respectively.

.1.111 1111111■11.711

0C oc 0 c.c

1111111111,1 ti!,111

50 100

B/mT Fig. 2.3 0-180mT region of the X-band spectrum of lig(Mn)(4Me-pyo)6(a01)2 showing ez and the AMs41

transitions a and (3 (see text).

150

46

CUTPTER 3

SPECTRAL STUDIES OF SOME MANGANESE(TT)-DOPED CHELATE COMPLEXES

PART I: MI TI

ii(acac) 3I .nH2 0

Introduction

In the vast majority of known complexes of type MII(bidentate)

3n+

the microsymmetry about the metal ion is octahedral (giving overall

D3d

symmetry); the coordination polyhedron, when viewed down the

3-fold axis (Fig. 3.1(a)) comprises two perfectly staggered equilateral

triangles of side s and at a distance h apart where s/h = /T72 = 1.22

(Fig. 3.1(b)). The angle (I) describing the twist of the triangular

faces is 600; in this configuration the electrostatic repulsion

between the donor atoms is minimized. Within the last few years, however,

several complexes, notably the tris-dithiolenes and -diselenolenes,

have been found to be trigonal prismatic, with cp = O.107-109

In a

third class of tris-chelates, most of which again involve S-donor

ligands, c lies between these two extremes while the 3-fold symmetry

is retained.110

Several attempts have been made recently to relate

the molecular geometries of these complexes to the ligand dimensions,

where these may be taken as rigid. Stiefel and Brown110 - noted a

correspondence between s/h and (I) for a number of the complexes, and

a Fig. 3.1

47

the near-linear dependence (for (I) >- 200) of (I) on the 'normalized

bite' b/a*, observed for all but the trigonal prismatic dithiolenes

and their analogues, has been rationalized on both geometric111

and electrostatic112 grounds.

Trigonal prismatic and 'intermediate' complexes with N- and

0-donor ligands are as yet of very limited occurrence; the work of

Ho1m113

and Fleischer114

on MN 'clathrochelates' of first row 6

transition metal ions appears to be a unique attempt to study the

influence of 4 on the electronic properties of the central metal ion.

The structural constraints imposed by a variety of hexadentate

macrocyclic ligands allow the preparation of a series of complexes

in which (I) varies over almost the entire range while,in every other respect,

the metal ion environment is virtually unchanged. However extensive

physical investigation113,114

revealed few systematic spectral trends

with stereochemistry.

Table 3.1 lists all known trigonal-prismatic complexes involving

0-donor ligands. (i) is clearly of the ldithiolene' type; the

contraction of the interior angles of the chelate ring is claimed115

to indicate the presence of the interligand bonding interaction between

donor atoms which is a distinguishing feature of the tris-dithiolenes107

In (ii) and (iii) the coordination geometry may be influenced by

the steric requirements of the phenyl rings in the first case, and

the methylene hydrogens adjacent to the donor atoms in the second.

The stabilization of the unusual stereochemistry of the Cd(II) ion in

(v) (and, by analogy, of Na(I) in (vi)) appears to arise from the

high lattice energy of the chain polymeric structure in which alternating

uni- and divalent metal ions are asymmetrically bridged by the ligand;

in the absence of a simple, monoatomic counterion {as in

b represents the intraligand L-L distance and a, the M-L bond-length.

48

1, 121 7122 123 DMANH UMg(hfac)ij, Bu N[Co(acaqi and Ag (NO),[4(acac);11 . H2O

, 4 - 3 3

the complex anion is octahedral. It is interesting that valence-band

theory,124

as well as the results of ab initio and extended Huckel MO

calculations,125

predict metal-ligand interaction energies to be of

similar magnitude in both D3h

and 0h, the usual strong preference

for the latter geometry arising from relatively small repulsive forces

between the surrounding atoms in the octahedron. It is probable that

in (v) and (vi) simultaneous coordination of at least five of the six

donor atoms to both the adjacent metal ions reduces interligand repulsion

to the extent that the less common geometry is favoured. Certainly

there seems to be no steric objection to the attainment of octahedral

coordination of Cd(II) in (vi), as far as can be ascertained from

examination of a framework molecular model of a single chain.

The present work on K[Cd(Mn)(acac)31.H20 is, to the best of

our knowledge, the first esr study of a high-spin d5 ion in a trigonal

prismatic environment. Similar complexes of Ni(II), Co(II) and Zn(II),

as their sodium and potassium salts, are also examined with a view

to assessing the relative importance to the stereochemistry of

M'(II) of such factors as the counterion, the effective ligand bite

and the ligand-field stabilization energy of M'(II). Methods of

preparation for some of the complexes126

and the IR spectra of the

potassium salts127

have been described previously.

Esr spectra of the Mn(TI)-doped complexes

The spectral profiles of Na m'(Ma)(acac)3

(M' = Co,Ni, Zn)

closely resemble those of the 4Me-pyo complexes of the previous

DM:A.1\TH = bis(dimethylamino) naphthalene hydride

49

chapter (Tables 3.2-5, Fig. 3.2). Linewidths in the Co(II) and

Zn(II) complexes are smaller than in the Ni(II) host, for which

satisfactory resolution of hf lines is not obtained at Q-band frequency.

From the AMs = ±1 lines, the axial zfs parameter is calculated

to be - -0.06cm-1

, the sign being determined from measurements of hf

spacing, as described previously; A is 0. The sharp, spin-forbidden

bands between 100 and 160 mT were analyzed as in Chapter 2. Change

in the sign of D reverses the order of the spin levels, so that the

4-2 transition of Table 3.3 corresponds to 5-3 in the 4Me-pyo complexes.

The experimental fields agree quite well with those calculated for 4-2

and 5-2 with AmI = 0, transitions with AmI > 0 generally having

probabilities well below 0.05. If, moreover, the last three lines of

6-3 be taken into account, the observed intensities (with the exception

of the strong line at 114 mT) may be rationalized on the basis of

the calculated T.P.'s.

The behaviour of 6-3 (or 4-1) at IDI = 0.063cm-1 shows an

interesting dependence on the sign of the zfs. For negative D,

the AmI = 0 lines are strongest, although their T.P.'s (- 0.5) are

well below the calculated value (- 1.0) in the absence of nuclear

hf interaction. when D is positive, however, transitions which

are formally nuclear-allowed have negligible probability, while the

= ±1' lines become weakly allowed (T.P.- 0.2).

The very rich X-band spectra of the Mn(II)-doped cadmates

were not analyzed in detail. At Q-band frequency D/hy is close to

0.1 and first-order perturbation formulae yield values of D and A

which are correct to within a few per cent and may be refined as

described above (see Tables 3.6 and 3.7). The 'best' values were

used to calculate X-band lineshapes; the result for the potassium

salt is compared with the experimental spectrum in Fig. 3.3.

50

The very small zero-field splittings in the potasSium salts

(with the exception of the cadmium complex) preclude accurate evaluation

of D (see Fig. 3.4).

Unfortunately, all attempts to dope manganese(II) into the

afore-mentioned octahedral tris acetylacetonates of cobalt(II) and

nickel(II) wore unsuccessful.

Electronic Spectra

Reflectance spectra of the cobalt and nickel complexes were

obtained in the region 4-3OkK (Table 3.9); beyond 2OkK only very weak

shoulders were observed on strong continuous absorption, and their

positions are not tabulated.

(a) The Ni(II) Complexes

Two bands, neither of which is discernibly split, are

observed in the regions associated with the v1 (3A

2g3T2g)

and

v2 (3A2g 3

Tlg(F)) transitions in 0h

symmetry; this interpretation,

however, leads to values of B well in excess of that in the free ion.

Similar observations have been made in the clathrochelates,113,128

the

calculated values of B being close to that. for Ni(bipy)2+

(- 0.7kK)

in complexes of near-octahedral microsymmetry, and apparently

increasing with departure Of q) from 60; in the same complexes, the

bands at 9-13kK move to higher energy from the TP to the TAP

configuration, providing an empirical indicator of structural change.113

To assist in a more rigorous interpretation of these results, the

angular overlap model was applied113 to the calculation of energy-

level diagrams for high-spin d7

and d8 configurations in D

3h symmetry.

Oribtal energies for values of the ratio A1/A2 between 0.4 and 0.8

were plotted vs. A2 /B, Al and A2 being the one-electron orbital

-

51

energy differences E(e')-E(al) and E(e.")-E(e') respectively (Fig. 3.5).

C was assumed equal to 4B, and the bite-angle a and polar angle e were

taken as 77° and 51.5°. An attempt to fit the observed spectrum of

ENTi(PccBF)] BF4* (mean(I) = 1.6°) in this scheme met with limited

3 n 3 3 n success. Transitions Transitions to Al, A

2 and E are expected to occur

3 ' 3 below 4kK, but identification of the 9kK and 11kK bands with A2 E

and E (both assignments being confirmed by single-crystal work)

led to quite reasonable values for B(0.8kK) and A2(4kK); however the

n formally forbidden transitions 3A

2 -4-3A2

3(B), E (P) were found to occur

about 3kK higher than predicted.

Analogous assignment of the 9 and 16 kK maxima of solid

M1Mi(acac)] also gives values of B(0.7 - 0.8 kK) which appear more

plausible than those derived by analysis in 0h symmetry, and

A2(0.36 - 0.45 kK) is of a magnitude comparable with that in the

clathrochelates.113 (The uncertainty in the parameters derives from

uncertainty as to whether the higher energy band represents transition

3 ' 3 n to A2

or E or to both.) However, examination of the spectrum of

octahedral [Ni(acac)23

129,130 (which might be regarded as a 'reference'

compound for trigonally twisted tris-acac complexes, as the tris-

(a-diimines) are to the clathrochelates) suggests that the simple

point-charge treatment is less appropriate here than in the case of the

tris-(a-diimine) cations; analysis in Oh symmetry again results in

an unacceptably high value of B, a more realistic figure being obtained

when the spectrum is fitted to the D3h energy level diagram. Evidently,

failure to fit the spectra of M[Ni(acac)] on an octahedral model does

not necessarily imply departure from octahedral symmetry of the MOc

chromophore.

* Details of nomenclature and structure for the clathrochelates may be found in ref. 113.

52

ahe transmission spectrum of 1.3u44Ni(acac),1 in chloroform

soluti.on has been reported131 (Table 3.10); although nmr studies

indicated some dissociation into Bue(acac) and DNi(acaccjx, the

- 131 predominant species in solution was believed to be Ni(acac)3'

From the data of Table 3.10 it is seen that the spectrum is distinct

from that of Ni(acac)2 or solid Ni(acac)3, (Table 3.9) but shows

some resemblance, particularly at low energy, to that of

Ni(acac)2 (H

2 0)

2'132 the presence of which might well result from

the use of imperfectly dried solvent. A curious feature of this

solution spectrum is the absence of 'V31 , such as appears at 251K

in the dihydrate.

It would be of interest to compare the reflectance spectra

of the sodium and potassium salts with that of Ag3(NO3)2[Ni(acac)31.H20

in which the presence of octahedralmicrosymmetry about the Ni(II)

123 ion has been demonstrated by X-ray analysis. For the present,

all that can be stated with certainty is that the positions of the

9kK bands in the solids MENi(acac)3] are such as to imply, on the

basis of spectral results for the clathrochelates, an increase in the

value of q) from the sodium to the potassium salt.

(b) The Co(II) Complexes

Interpretation of the spectra in Oh symmetry leads to the

values of B in Table 3.9. As in the Ni(II) complexes, the band near

9kK moves to higher energy in the potassium salt; a similar shift

in the Co(II) clathrochelates has been related to the increase in

the value of (I) along the series.113

For [Co(PccB6]BF4 (0 - 1.5°)

and [Cc((py)3tach)](Ce.04)2 (of which the Zn isomorph has (1)- 4.5°)

the spectra have been assigned in D3h symmetry, although the narrow-

ness of the low-energy band was recognized as being incompatible

53

with the expectc.:). multiple transitions.113

It should he noted, however,

that the bands at 9.38 and 20.4 kK of Co((ny)3tach) , when fitted

on the octahedral model, give quite reasonable values of the ligand field

field parameters (Dq = 1.06kK, B = 0.8kK). As with the Ni(II)

complexes, therefore, the electronic spectra of the tris acacs provide

little unambiguous structural information.

Infrared Spectra (4000 - 250cm )

The IR spectra of the complexes are listed in Table 3.11;

thepreviously reported spectra of of the potassium salts of the

Co(II) and Ni(II) complexes are in good. agreement with the present

results.

The position, intensity and appearance of \(C=0) (- 1580cm-1

)

is markedly similar in all the complexes and, together with the absence

of any band near 1700 cm-1

, this is taken as indicating bidentate

coordination (and therefore presumably chelation) by the ligands

in every case.

In D3 symmetry, three IR-active frequences are predicted for

most of the stretching vibrations.133

As has been noted previously,134

the high degree of electron delocalization in the chelated acac

ring leads to extensive coupling of vibrational modes; however, two

bands, which in the present complexes lie in the 415-430 and

538-560cm-1

regions, are metal-sensitive, and that at higher frequency

has been shown,134

by means of 180-induced isotopic shifts in the

MIII(acac)3 complexes to be a relatively 'pure' M-0 stretching mode.

In addition the band at 240-280 which, like the above bands, follows

the Irving-Williams order must also contain a high proportion of

(M-0) character. The frequencies of all three of these vibrations

are invariably greater in the sodium than in the corresponding

54

-1 potassium salt. The opposite is true of the 645-655cm band, but

the fraction of \X4-0) character is believed134 to be smaller here

than in the bands between 400 and 600cm-1.

Other noteworthy features are (i) the splitting of the

out-of-plane (C-11) deformation in NaECd(acac);] and in all the

potassium salts, and (ii) the sharp band at 3630cm , indicative of

coordinated, non-hydrogen-bonded water, in K[Cd(acac)3.. .1120 which

contrasts with the very broad weak absorption centred on 3400 cm1

in the other hydrated species.

X-ray powder diffraction photographs of the two cadmium complexes

are distinct from each other and from those of the remaining complexes.

The sodium salts of the zinc, cobalt and nickel complexes are

isomorphic, and the marked resemblance between the powder patterns

of K[Co(acac)31and K[Ni(acac)3] indicates their structural similarity.

Discussion

All the esr spectra can be interpreted in terms of a 6S

ground state, and the nuclear hf interaction is of the magnitude

expected for magnetically dilute Mn(II). Although, therefore, the

trigonal prismatic K[Cd(acac)]. H2O shares some of the structural

characteristics of the tris-dithiolenes (e.g. s/h == 1, 1.28.‘ b/a< 1.33) ,

it shows none of the highly unusual magnetic properties108

(such as,

for species with 's 1,, near-isotropic g-tensors close to the free-

electron value and anisotropic A's much smaller than those calculated

by theory) of those complexes.

The compression ratio s/h in KCCd(acac)31 .H20, calculated from the

available crystallographic data119

and the expressions of Stiefel and

Brown,110

is 1.04, as compared with 1.22 for a regular octahedron.

That elongation of this magnitude does not in itself contribute

55

significantly Lo the zfs is convincingly illustrated by the esr data

for the three isomorphous hexakisantipyrine complexes44

listed in

Table 3.12. The increase in magnitude of D from the Pb(II) to the

Mg(II) complex reflects the decrease in cation size of the host lattice,

and this clearly more than compensates for the effect on the zfs of

the concomitant progression from trigonally elongated to regular

octahedralgeometry. It might be asked whether the potassium ions,

one of which lies directly on the three-fold axis of the complex anion,

contribute directly to the observed trigonal distortion. Sharma has

calculated the zfs arising from the presence of two unit positive

charges on opposite sides of a 3-fold axis of cubic symmetry, and at

a distance of 4ao(=2.12A) from the Mn(II) ion, as -0.0268 cm-1.135

As

0 119 the K-Cd distance in K[Cd(acac) 1.H2 0 is 3.94A, the contribution

to D from this source would appear to be negligible.

There seems little doubt, then, that trigonal prismatic

coordination in the anion is responsible for the remarkably high axial

distortion in the cadmium complexes; D is well in excess of any

reported previously for a Mn06 chromophore. The small rhombic distortion

is to be expected in view of inequalities in the Cd-O bond-lengths

(which vary in the potassium salt between 2.247 and 2.325 R) and in

the angles subtended at the metal by the donor atoms.119

Values of D for Mn(II) in the mutually isomorphous sodium

salts of the Zn(II), Co(II) and Ni(II) complexes are also among the

highest observed in 0-donor complexes with six identical donor atoms;

comparable zfs's have been recorded only in M(4Me-pyo)6(a04)2

(Chapter 2) and in Zn(H20)6Se04, where the axial distortion is ascribed

to unequal participation in hydrogen-bonding of the six aquo ligands,

the MO6

octahedron being quite regular.41

Such an effect cannot be

ruled out in the present sodium salts, all of which are hydrated, but

56

the 'trigonal twist' explanation is preferred, for the following

reasons. (1) Crystallographic studies of the clathrochelates

reveal113 that for complexes of a given ligand, (60

o-4) generally in-

creases in the order Fe<Ni*.Co;Zn, reflecting the decreasing ligand

field stabilization of the octahedron relative to the trigonal prism.

The magnitudes of D (Zn = Co>Ni) in the sodium salts of the present

study follow this sequence which is the reverse of that expected

from considerations of the size of the host metal ion. The effect,

if real, represents a rare instance of the detection of a subtle

structural difference between isomorphous complexes by a simple

spectroscopic technique, and without recourse to single-crystal studies.

Unfortunately zfs's in the potassium salts are too small to

permit observation of trends in D with changes of host metal.

(2) Esr spectra have been obtained at 4.2K for the solvates of

Na[Ni(acac)3] with p-dioxan and benzene.

136 Comparison of the D-values

with those of other NiL62+

complexes and their Mn(II) analogues

(Table 3.13) suggests that axial distortion in the solvates is

substantially lower than in the hemihydrate examined here. A trial

structure136 for a crystal of the p-dioxan adduct reveals the presence

of chains of alternating complex anions and solvent molecules running

parallel to the three-fold axis of the crystal; significantly, the

sodium ions are not situated on this axis.

Taken in conjunction with the electronic spectral results,

(1) and (2) constitute reasonable evidence that (1) <60°, at least

in the sodium salts.

The esr results then suggest that the twist-angles in the

present complexes represent a balance between retention of the chain

structure and the tendency, dictated by the size of the ion, of M'(II)

to approach an oetahedralconfiguration. The relative ability of the

57

alkali metal ions to form chclates with acac is known to be

137 Li=Na>K. The structure of MLCd(acac) (b/a = 1.28) appears to be 3-

independent of the counterion. In Na[Ms (acac);1 (M'=Co, Ni, Zn) the

larger effective bite may allow a closer approach to octahedral

geometry, within the polymeric framework, than is possible in the

Cd(II) complexes, while in the potassium salts the reduced energy

of interaction of the counterion with the ligand could result in

attenuation of the MI-0 'links' of the chain and a further increase

in .1) which may, however, remain well below the octahedral limit.


The complexes were prepared as in ref. 126 by reaction of the

hydrated metal acetates with sodium acetylacetonate in aqueous

methanol. No pure solid of formula K[Zn(acac)3] could be isolated.

58

Table 3.1 TrigonEtl Prismatic Complexes of 0-donor. Ligands

Complex Ref.

(i) Mo(9,10-phenanthrenequinone)3 115

(ii) [CH2{P(0)Ph2} 3 NaBr, 3H20 116

(iii) [ Coiif ,Con I (OCH2CH2NH2) 312 (OA c)

2 117

(iv) RbHU2. (U = 5-bromo-3-hydroxy-6-

methyluracil.) 118

(v) K[Cd(acac) 3 I. H2O 119

(vi) Rb2[(hfac)3 Na] 120

59

Table 3.2 X-band. esr spectrum (niT) of

NarM(Mn) (acac) 3] . nH20 (M = Co, Zn)

Observed

(V=9.532GHz)

Calculated for D = - 0.063 cm-1, X = 0

B Axis, levels

68

121

126

224

265

316

384

475

482

612

70.6

a

a

205.6

225.9

265.8

317.0

383.7

475.4

483.0

610.4

az

bz

e x,y

d x,y

c x,y

b x,y

dz

a x,y

ez

aAssigned as spin-forbidden transitions. See Table 3.3

Table 3.3 AMS#1 transitions (100-160mT) in the X-band spectrum of

NaEM(Mn) (acac)3 .nH2 0 (M=Co,Zn)

Observed Calculated for D=-0.063cm 1, A=0, A=-9.1mT

(v=9.532GHz) 5-2 transition

8=30°,Ami=0

4-2 transition

8=75°,Ami=0

6-3 transition

8=90o,Ami=0

101.6 s 101.8 (0.66)a r l L5/2jb 101.6 (0.54) T1/2 a

105.7 m 106.2 (0.19) E-5/2]

109.3 m 109.5 (0.33) [3/2]

114.0 s 114.1 (0.19) E-3/2]

117.6 m 117.9 (0.28) 1-1/21

120.2 (0.56) [3/ 1:1

122.4 (0.19) 1-1/21

126.9 m 127.1 (0.34) [1/2]

132.1 s 131.3 (0.19) [1/2]

133.2 (0.32) [5/21 a

continued over/

Table 3.3 Continued

Observed Calculated for D=-C.063um 1 , X=0, A=-9.1mT

(v=9.532GHz) 5-2 transition 4-2 transition 6-3 transition

8=30°, AmI=0 6=75o / AmI=0 6=90o,Am I=0

137.2 m 137.1 (0.49) [3/2]

140.9 (0.20) [3/2]

c

148.1 s 148.1 (0.78) [5/2]

151.2 m 151.1 (0.21) [5/2]


bmI in square brackets

Strong but poorly resolved absorption

dSome 'mixing' of nuclear levels occurs

62

Table 3.4 C-bandesrpct.:strum (mT) of

Na[M(in) (2z1c):3].n11 0 (M = Co, Zn)

Observed

(V=35.44GHz)

Calculated for D = -0.0630m-1, X = 0

B Axis, levels

993

1131

1190

1328

1403

1534

996

1131

1135

1195

1328

1401

1404

1536

az

bz

e x,y

d x,y

b x,y

dz

a x,y

ez

Table 3.5 X-Band esr spectruma (mT) of

Na[Ni(Mn)(acac)3]. 0.5H20

Observed

(v=9.527GHz)

Calculated for D = -0.60cm 1 , X = 0

B Axis, levels

229

268

316

474

230.0

269.0

318.6

475.9

e

d

c

a

x,y

x,y

x,y

x,y

aTransitions for Bliz not resolved.

63

Table 3.6 Q-band esr s'oetrum (mT) of

of K[Cd(Mn) (acac):1 2.F320

Observed Calculated for D = 0.113cm , X= 0.06

()=35.85GHz) Axis,

796 w 797

1005 ms 1009

1036 ms 1039

1085 ms 1098

1124 s 1129

1167 s 1168

1354 s 1361

1405 s 1408

1487 ms 1489

1517 ms 1522

1572 ms 1574

1764 w 1765

levelsa

az

ey

b

ex

dy

dx

bx

by

ax

dz

a

ez

aAssignments. are for negative D.

64

Table 3.7 0-band. esr spectrum (mT) of

NarCd(v.n) (acac) j

Observed

(V=35.815GHz)

Calculated for D = 0.109cm- , X= 0.03

B Axis, levelsa

828 w 813 az

1035 ms 1038 e

1046 bz

1085 ms 1081 ex

1138 s 1143 d Y

1159 s 1162 dx

1374 s 1369 bx

1390 s 1391 by

1501 ms 1501 ax

1513 d z

1532 ms 1541 a Y

1746 ez

aAssignments are for negative D.

65

Table 3.8 Zfs parameters of Mn(II) in the complexes

Complex D(cm-1) X

K[cd(acac)3].H20 0.113 0.06

Na[Cd(acac)] 0.109 0.03

NaEZn(acac)31 .H20 0.063 0.0

Na[Co(acac):;].0.5H20 0.063 0.0

NaLNi(acac) 3 • ] 0.5H20 0.060 0.0

K[Co(acac):0 <0.02 a

K[Ni(acac)2:1 (0.02 a

aIndeterminate.

Table 3.9 Reflectance spectral data (kK) for the

Ni(II) and Co(II) complexes

Complex V1 V2 Dq B

in 0h symmetry

K[Ni(acac)3] 9.52 16.13 0.952 1.275

Na[Ni(acac)3]0.5H20 9.30 16.26 0.930 2.262

K[Co(acac) 3] 8.93 20.00 1.010 0.816

Na[Co(acac)3]0.5H20 8.77 19.60 0.991 0.798

66

Table 3.10 Electronic spectra of some

Ni(II) complexes with acac. (Data from refs. 129-132)

Complex Physical state Spectrum (kK)

Pgi(acac)]3 solid or toluene

solution

8.5, 15

Ni(acac)2(H20)2 solid 8.9, 13.4 sh, 15.4, 25.0

Bu4N5,1i(acac)] CHCZ3 solution 9.0, 13.2 sh, 15.6, 34.1 sh, 37.6

Potassium salts Assignmenta

Ni Cd

3075 w -3175 br

1585 br 1580 br

1510 br

1255 s

1198 m 1195 m

1015 s 1009 s

920 s 917 shl

913 s f

782 m 795 m

762 m 767 m

v(C=0)

V(C=C)

V(C-C)

(C-CH3)

6(C-H) i.p.

CH3 rock

V(C-0) + v(C-CH,)

6(C-H ) o.o .p.

1520, 1505J' s 1245 s

M '

Table 3.11 IR spectra (4000-250cm1) of the complexes

Sodium Salts

Zn. Co Ni Cd Co

3075 w

1585 br

1510 br

-3400 w,br

3075 w

1580 br

1508 br

-3400 w

3075 w

1580 br

1510 br

-3360 v.br

3065 w

1585 br

1515 br

3070 w

1585 br

1510 br

1256 s 1256 s 1258 s 1250 s 1254 s

1192 m 1194 s 1195 s 1197 m 1196 m

1010 s 1010 s 1012 s 1010 s 1014 s

916 s 916 s 919 s 918 s 920 s

790 m 780 m 770 s 769 s 770 s

760m 765 m

continued over/

Table 3.11 Continued

Sodium salts Potassium salts Assignmenta

M' Zn Co Ni Cd Co Ni Cd

655 sh

1 650 m

540 m

425 m

403 m

<240

652 m,br

550 m

425 m

400 m

-250 v.br,m

660 sh

652 m

560 s

430 m

405 m

280 s

655 sh

645 m

540 m

420 m

395 m

665 w

654 m}

550 m

417 m

400 m

660 sh

/ 655 m

559 m

420 m

405 m

269 s

650 sh

} 645 m

538 m

415 m

395 m

Ring def. +

v(n-0)

v(M-0)

v(M-0)4(C-cH 3

5(0-M-0) i.p.

v (M-0)

)

aAs in refs. 133, 134.

69

Table 3.12 Structural and esr data for some Ile:;akisantir)vrine

complexes (from refs. 44 and 103)

-4 -1

0 Complex s/ha D(10

Mn(II)

cm ) of

impurity

Ionic radius(A)

metal ion

of host

Mgyet04)2 1.23 -44.4 0,66

CaL6(CZ04)2 1.17 -32.0 0.99

PbL6(CZ04)2 1.09 -8.7 1.20

aCalculated from the values of h/x (x = centroid to apex distance of a

triangular face) cited in ref. 103.

Table 3.13 Esr parameters (cm-1) of some ML62+ complexes

(M=Mn(II), Ni(II))

Complex DMn DNi Ref.

2 + a M(imidazole)

6 <0.006 0.46 105, 92

M(H20)6+ as

the Tutton salts,

0.023b -2.24 95, 94

M(NH4)2(SO4)2.6H20

M(PY0)6(a04)2 0.038b 4.3 51, 91

Na[Ni(acac)31.p-dioxan - -1.9 136

aAs impurity in the Cd(II) host bAs impurity in the Zn(II) host.

70

Table 3.14 Analytical Results


%C 9.11 %C %H

K[Cd(acac)].H20 38.9 5.1 38.6 5.0

Na[Cd(acac)31 40.2 5.4 40.0 5.1

Na[Zn(acae)31.H20 44.8 5.4 44.6 5.7

Na[Co(acac)31.0.5H 0 46.7 5.5 46.4 5.7

Na[Ni(acac)3].0.5H 0 46.8 5.9 46.4 5.7

K[Co (acac) 3] 45.2 5.6 45.6 5.4

K[Ni(acac)3] 45.2 5.4 45.6 5.4

100 J00 500 mT

Fig. 3.2 Esr spectrum (v=9.532GHz,0-650mT)

of NarZn(Mh)(acac)3'1.H

1

X6

A

Y

AI

72

(a)

(b)

$1

I OZ •07 1t4.90 i1.17 A6.6I 33.06 tat .11 <1,r .91 a.7 614.30 51 617 .-13T17671 FIELD mT

Fig. 3.3 Esr spectrum (V=9.531GHz, 0-700mT) of KECd(Mn)(acac)].H20:

(a) experimental; (b) calculated for D=0.113cm 1,

A=0.06, GIi =3OmT. PP

Fig. 3.4 Esr spectrum (V=9.532GHz, 0-700mT) of KF0o (Mn) (acac) 3]

100 300 500 mT

3.0 a. 77°

20

1.0

0.5

(a) 3.0

2.5

0

_ - -- --------------

-0.4 0 10° zo° 50° 40° 50° 60°

D3r 0 03

0

s.90°

/F0 3 4 5 6 7

Ira= I I f A1

lsa 761

1/B

(ID

74

(b)

(c)

Fig. 3.5 (from ref. 113). (a) Dependence of one-electron d-orbital energies (diagonalized) for the complex M(L-L)3(a=77,900) upon the twist angle q): eas=+0.1e ;

e7s=-0.1e0.. (b) and (c) Energy-level diagrams for :7177 (high-spin) and d8 conrigurations in D3h symmetry for C/B=4.0 and selected Al/A2 ratios: ----, 0.4, , 0.6; ----, 0.8.

%5

CHAPTER 4

SPECTRAL STUDIES OF SO ;E MANGANESE(II)-DOPLD CHELATE COMPLEXES

IT PART II: MII (dike) (LL) AND [m (T L)7

Introduction

This chapter describes esr and IR studies of some complexes

ZI of type M (Mn) (dike)2(LL) = Mg, Zn or Cd) where LL represents

two identical monodentate ligands (H20 or DtF) or a single bidentate

heterocyclic diimine (phen or bipy). With the exception of

Cd(acac)2phen.H

20, complexes of the latter group are almost certainly

cis-octahedral, as has been found for Cu(hfac)2bipy by X-ray diffraction

analysis of a single crystal;138

a trans configuration for (bridging)

bipy has been postulated in a few cases but unequivocal evidence for

this mode of ligation is not available at present. Monodentate

donors, on the other hand, may give cis- or trans-adducts with M(dike)2,

several examples of each type being known.139-143

In theory, the two

configurations may be differentiated by vibrational spectroscopy,144

but the application of this technique to complexes with polyatomic,

multidentate ligands where the group frequency concept may not hold even

approximately has been called into question.145,146

In a recent single-crystal esr study of trans-Mg(Mn)(acac)2(H20)2,51

X was found to approach its maximum value of 0.33 although the in-plane

oxygen atoms form a nearly perfect square;139

this was ascribed

to the charge accumulation along the x-axis (see Fig. 4.1) that results

from metal-ligand dff-piT overlap. It is possible that the equal spacing

of the Kramers doublets arises solely from the D2h symmetry of the

molecule, as discussed by Griffith,147

rather than from any anisotropy

of the in-plane electronic distribution. However the metal ion is also

in a site of D2h symmetry in the halide- and hydrazine- bridged chain

76

24 polymers of Chapter 5, and in the complexes Mn(biuret),X,(X=CZ.Br,I),

all of which have A close to zero. Very recently the D-tensor has

been found to be axially symmetric in the complexes ML2X2 (L = pdma,

pdads*; X = Br, I) of which the IR and electronic spectra were

interpreted on the basis of a trans arrangement of the ligands.66 .

The

extent of M -4>ligand back-donation in complexes with Gp. VD donor

atoms is in dispute,148

but it is noteworthy that the nuclear hf

splitting in the pdma and pdads complexes (-9.7 and -10.7 mT respectively 6

is appreciably higher than in Mg(Mn)(acac)2 (ll2 0)2 (-9.15mT)51

or

Zn(Mn)(pic)2 (H2 0)2. 2H2 0 (-8.6mT),37

indicating relatively little

delocalization of d-orbital spin-density on to the ligands.

It is interesting to determine whether the 'fully rhombic' D-tensor

found in Mg(Mn)(acac)2(H20)2 is of general occurrence in trans-M(dike)2(LL)

and, more importantly, whether identification of a molecule of this

type as the cis or trans isomer is possible from the SPH parameters of

the Mn(II)-doped solid; to this end, several complexes of types

M06

and M04N2 (M=Zn, Cd, Mg) have been examined. Of the adducts with

monodentate ligands, Mg(dbm)2(DMF)2 is known to have the cis structure;

the bond angles at the magnesium(II) ion vary between 86° and 95°, and

0 the Mg-0 bondlengths to dbm and DMF are 2.056 and 2.095A.

143 The

preparation of the diaquo adduct of Mg(hfac)2 was attempted for comparison

of the zfs parameters with those of the acac analogue. Attempts to

dope Mn(II) into Zn(hfac)2(LL) (LL = bipy, phen) and Mg(hfac)2en were

only partially successful, the esr signals obtained being too weak to

allow extraction of the zfs parameters. Esr results for some tris

complexes of phen, bipy and en are also reported. Despite the use of

pdma = o-phenylenebis(dimethylarsine) pdads = 0-phenylene(dimethylarsine)(dimethylstibine)

77

a wide range of counterions and solvent systems, certain of the tris

complexes could not be obtained pure owing to coprecipitation of the

sparingly soluble his complexes. Zfs parameters for two of the latter,

Cd(en)2C.e2'2H20 and Cd(bipy)2 (NO3 )2' have been determined.

Several esr studies of octahedral cis and trans chelate complexes

of Cr(III) (d3) have been reported recently;149-152 in much of the

150,152 powder and frozen-solution work, however, spectra have been

analyzed taking into account only resonances for B in the principal

axis directions, and at the D-values concerned (0.1 - 1.0cm 1) serious

errors may result from neglect or misassignment of off-axis transitions.

Despite their obvious relevance to the present work, therefore, the

results of the chromium(III) studies will not be considered here.

IR Spectra (4000-250cm-1 ) of the phen and bipy Adducts

The mid-IR spectra of the phen and bipy adducts are listed in

153 Tables 4.1-3; the magnesium complexes have been examined previously,

but detailed results were not reported.

In the zinc and cadmium complexes, the acac absorptions appear in

virtually the same positions as in the tris-acetylacetonates of the

preceding chapter, confirming that, with the exception noted below,

the 13-diketone molecules chelate in every case. Comparison of the spectra

of the magnesium adducts and of Mg(hfac)2.3H20 with that of free hfac

leads to a similar conclusion for these complexes.

Extensive spectral studies of the free and coordinated

diimines146,153-156

have led to identification of the features

diagnostic of the chelated ligands. The relevant observations in

the present complexes are as follows. (i) Out-of-plane hydrogen

-1 deformations occurring between 700 and 850 cm in the free ligands

78

are split and/or shifted in te complexos.153-155

h

The spectrum of

Zn (acac) 2bipy is distinguished by the multiplicity of bands in this

region; this is consistent with the esr results which indicate the

presence of more than one site for the metal ion. (ii) The 'new' band

at-1315cm1 in the bipy adducts has been rioted previously in several

55 tris-bipy complexes.

154,1 (iii) The band at -1135cm

-1, ascribed

to the ring-hydrogen in-plane bending vibration in both bipy and

phen,155

is shifted to higher energy in the zinc and cadmium complexes.

(In the magnesium complexes, this region is obscured by the broad,

strong C-F stretching vibrations.) (iv) The ring bending mode at

- 620cm1

in the free diimines also moves to higher energy on

complexation.156

An interesting feature of the magnesium complexes, apparently

overlooked in the earlier study,153

is the splitting of V(C=0)

in the bipy adduct, both components of the doublet being of equal

intensity. Splitting of comparable magnitude has been observed for

V(C-0) in Zn(hfac)2bipy and Zn(hfac)2(py)2.142

A structural study

of the latter complex142

confirms that the hfac ligand is asymmetrically

coordinated; the axial Zn-0 distance is longer than the in-plane

0 bondlength by 0.09A, and the Zn-N distance is intermediate between the

two values. In Mg(hfac)2phen the splitting of v(C=0) is less pro-

nounced than in the bipy adduct, while Mg(hfac)2.3H20, like

Zn(hfac)2py142

and Zn(hfac)2(H20)2

142 has a single sharp peak in this

region.

In the acac complexes, v(C=0) coincides with the diimine C=C

and C=N vibrational frequencies, so that the mode of chelation of the

diketone cannot be inferred with certainty from the m data.

The presence of a band at 1675 cm-1

, assigned to 'free'

C=0, in the spectrum of Cd(acac)2phen.H20 suggests monodentate

coordination by at least one of the diketone molecules. Another

79

distinctive feab.)..,i'e of this spectrum is the breadth of the 5t0 cm-1

absorption, the half-width b •.incl roughly twice as large as in the other

zinc(II) and cadmium(II) adducts; in view of the previous assignment

of this band. as nearly pure V(M-0), this may be taken as additional

evidence for monodentate chelation of one of the acac molecules. The

librational modes of coordinated water are expected to give rise to

two bands between 450 and 900 cm-1 157 and it is noteworthy that both

Cd(acac)2phen.H90 and Mg(hfac)2

.3H20 show weak background absorption

at 500-700 cm1 which is absent in the anhydrous complexes.

With the exception of the above-mentioned V(M-dike) at *540cm1,

it has not proved possible - in this work to identify metal-ligand

stretching vibrations. The band at 415-430 cm-1

, assigned to

V(M-0)-1-6(C-CH)in MII(acac)

31 overlaps with a diimine ring torsion

mode in the phen and bipy adducts. M-N stretching frequencies are

expected to be in the main below 200 cm-1,158-160

and their identification

is difficult on account of the presence in this region of ligand modes

activated by coordination.

Esr Spectra of the Mn(II)-doped Complexes

Values of D in the complexes lie between 0.02cm-1

and 0.16 cm-1.

In some cases the sign of D has been determined from hf measurements;

where this is not possible the assignments a-c in the tables are made

on the assumption that D is positive. Where the zfs is small and

linewidths are large, there is severe overlapping between adjacent

bands even at the extremities of the spectrum. Under these conditions,

consideration of the a, b, d and e transitions alone may not lead to

unambiguous assignment of zfs parameters. Thus for Zn(Mn)(acac)2phen

(Tables 4.4 and 4.5, Fig. 4.4(a)) the positions of the three pairs

of bands flanking the central transition are consistent with

80

(i) D = 0.025cm ,X = 0.03, or (ii) D = 0.028cm 1, A = 0.26, or, less

plausibly, (iii) D = 0.039 cm 1 X = 0.10; for (ii) and (iii) it is

necessary to assume that ez and az are too weak to be observed in the

experimental spectrum. Analysis of the central (4-3) transition at

X-band was attempted in order to distinguish between (i) and (ii). The

powder pattern for this transition (which is isotropic as far as terms

linear in the fine structure parameters) has been calculated by second-

order perturbation theory, assuming an isotropic g-tensor and the

absence of nuclear hf interaction.61,161

The results of the calculations

appear in Tables 4.6 and 4.7. It is to be noted that the positions of

some of the low-field singularities depend on whether X is less

than or greater-than 1/9. We would draw attention to the error in

ref. 161 which gives the location of the low-field divergence for

1/9 < A <1/3 as -8(1-3X)2; the correct expression is -8[1-(3X)23,

i.e. -8(1-912). In measuring the resonance fields in the experimental

spectrum it has been assumed that second-order hyperfine effects do not

shift the centre of gravity of a sextet; it has already been shown

in Chapter 2 that such shifts for the 4-3 transition are only lmT or

less at X-band. Table 4.7 also gives the values of B computed by

exact diagonalization of the SPH matrix at the critical points, which

agree well with the results of the perturbation calculations. The

geff

= 2 absorption pattern in the experimental spectrum corresponds

A closely to that predicted for D = 0.028 cm 1 , A = 0.26. This

assignment is confirmed by the position of the only observed spin-

forbidden band, centred on 122.9 mT and ascribed without difficulty

to the 5-2 transition in yz. The predicted fields (corrected for

higher order hf effects) for (i), (ii) and (iii) are 118.3, 124.2 and

127.3 mT. Table 4.8 compares the experimental and calculated (for (i)

and (ii)) positions of the hf lines.

/lc:curate paraeters , not obteinaele ilg(Mn)nfec)

as the linewidths nf the outer L.1 ,;t1. arc considcrably larger than

in Zn(Nn)(acac),)rhen analysis of the spectra is further complicated.

by the presence of a strong background signal, due possibly to

absorption by aggregated manganese (II) ions ez

and az occur at -1154

and -1410 mT at 35.84 GHz, giving a D value of ca. 0.03cm-1; this

is confirmed by the presence of a very weak signal centred on 470 mT

at X-band. Location of a and ey is more difficult, but estimates for

A of 0.17 or 0.22, obtained from the high-field absorption

(1379/1387 mT) at p-band, were used to calculate the powder pattern of

the 4-3 transition at 9.522 GHz. As Table 4.9 shows, the lower A value

gives results in moderately good agreement with the experimental data.

Mg(Mn)(dbm)2(DMF)2, on the other hand, gives exceedingly well-

defined spectra over the entire range of absorption (L\Hpp in the geff = 2

region is < 1.0 mT) and the zfs parameters are obtained to a high degree

of accuracy (Tables 4.10 and 4.11, Fig. 4.5). The spectra of

Mg(Mn)(hfac)2.3H20 (Table 4.12), Cd(Mn)(en)2a.2.2H20 (Table 4.13),

Cd(Mn)(bipy)2(NO3)2 (Tables 4.14 and 4.15) and Cd(Mn)(bipy)3(NO3)2

_(Tables 4.16 and 4.17, Fig. 4.6) are analysed without difficulty and

require no comment.

Resonance fields for Mg(Mn)(hfac)2bipy at X-band frequency are

not tabulated as the bands are poorly resolved, except in the case

of a moderately intense transition at 160.9 mT (geff = 4.23)

(see Fig. 4.7). This behaviour parallels that of powdered

Zn(Mn)(Ph3P0)2CZ2'

and snectral analysis at 9-band (Table 4.18)

confirms that, as in the tetrahedral complex, A is close to 1/3.

However at the low D value in the magnesium complex, the origin

of the X-band geff = 4 line is not isotropic absorption associated with

the middle Kramers doublet, but a transition between heavily 'mixed'

U2

levels [1/2, -5/2 (--1/2) -4- 3/2] occurring, for A = 0, at -161 and

-166 mT in the zx and vz planes. The Q-hand parameters were used for

reproduction of the X-band lineshape (Fig. 4.7) and although the

experimental spectrum is generally well fitted, the absence of the

predicted zero-field absorption is conspicuous. However the only

feasible alternative, D = 0.138 cm-1 and X = 0.05 (obtained by assigning

the 941 and 1620 mT signals at Q-band to a and e ), is much less

satisfactory on both criteria (X-band lineshape and location of

resonances at Q-band)and the first assignment is taken as correct. In

Cd(Mn) (acac)2phen.H20,with zfs of similar magnitude (Table 4.19), the

X-band zero-field absorption is present although less intense than

predicted (Fig. 4.8).

The spectral profile at X-band of Zn(Mn)(acac)2bipy (Fig. 4.9)

resembles those of NaPn(Mn)(acac)3].H20 (Chapter 3) and the hexakis

4Me-pyo complexes of Chapter 2, placing D in the range 0.055 - 0.075cm-

However, many more lines are observed than can be accounted for by a

single set of SPH parameters; also the almost complete lack of

structure on the strong central band contrasts with the well resolved

hf lines of the corresponding transition in Na [ Zn(Mn)(acac)J.H2O.

Both features are consistent with the presence of manganese ions in

more than one environment. Severe overlapping of some adjacent lines

at X-band precludes accurate measurement of all resonance fields.

At Q-band, however, at least four pairs of transitions at the

extremities of the spectrum are well resolved; their positions and

intensities appear to indicate roughly equal distribution of

manganese ions between two lattice sites characterized by the parameters

(i) D = 0.061 cm 1 , X = 0.08 and D = 0.071 cm

-1 ,X = 0.09 or

(ii) D = 0.061 cm1, X = 0.16 and D = 0.071 cm

-1, A = 0.03 (Table 4.20).

X-band spectra were simulated, assuming equal numbers of manganese ions

83

ir the two environments, for both (1) and (ii) ; each closely resembles

the eNperimental lineshape (Fig. 4.9). A well-defined 'half-field'

spectrum was observed in the usual region (100 - 150 mT) but owing to

the very large number of spin-forbidden lines predicted for two sets

of SPH parameters, it was not possible to make a clear distinction between

(i) and (ii) on this basis.

Zfs parameters for Cd(Mn)(acac)2bipy were calculated from the

Q-band spectrum (Table 4.21, Fig. 4.10) and used to simulate the

X-bamd lineshape (Fig. 4.11).

Discussion

The zfs parameters of the complexes are listed in Table 4.22,

together with previously obtained data for some other chelates.

As expected, the magnitude of the zfs of Mn(II) in cis-Mg(dbm)2(DMF) 2

is much smaller than in trans-Mg(acac)2(H20)2.51

The fine structure

parameters for the cis complex are remarkably close to the values

observed in cis-octahedral Zn(0Ac)2(H20)246'48

(see Fig. 4.2) despite

the much higher angular distortion of the coordination polyhedron in

the latter complex, where the angles 0-M-0 range between 61o and 103

0.162

Both these complexes, as well as Zn(hfac)2py2,142

crystallize in

space group C2/c, with Z = 4, and the molecules are required to have

C2 symmetry; the two-fold axis coincides with gx in Cu(hfac)2bipy 138

and Zn(Cu)(hfac)2(py)2163 and, for the parameters listed in refs. 46

and 48, with Dy (and presumably gy) in Zn(Mn)(0Ac)2(H20)2. For the

latter complex, D and E as given in refs. 46 and 48 are of opposite

sign, and are evidently not referred to a 'proper' coordinate axis system.

We note, however, that simple interchange of x and y, which converts

34

the axis system to that of Zn(Cu)(hfac)2(py)2, reverses the sign of E,

leaving D unchanged;55

this is the 'proper' system of axes (Fig. 4.3),

in which, for this complex, D and F are both positive. The parameters

of Zn(OAc)2.3H20 (Table 4.22) are so similar to those of the dihydrate

as to suggest identical molecular structures for the complexes.

Several attempts to prepare the complex Mg(Mn)(hfac)2.2H20

resulted each time in well-defined, pale yellow crystals analyzing

as the trihydrate; this material showed no significant weight loss

after mild heating (70°C) in vacuo for three hours. (Prolonged heating

has been shown153

to cause sublimation and decomposition of the dihydrate.)

The IR spectrum of the trihydrate suggests coordination of at least

one water molecule, and as the axial zfs lies within the range of values

given by the MO6

chromophores of Table 4.22, it is reasonable to

assume octahedral coordination for the hfac complex. Furthermore,

both D and E are substantially lower than in trans-Mg(Mn)(acac)2(H20)251

and the rhombic distortion is of the same order of magnitude as in

Mg(Mn)(dbm)2(DMF)2 and Zn(Mn)(0Ac)2(H20)2, indicating a cis arrangement.

D is negative, as in Mg(dbm)2(DMF)2,and this is readily explicable

on the basis of the same axis system as in Fig. 4.3, since the

Mg-OH2 bond-length (which is 2.15A in Mg(acac)2(H20)2

139) is expected

to exceed Mg-hfac (2.06 in Mg(hfac)-3 164

) by -0.1A; this is also

the probable reason for the larger D compared with Mg(Mn)(dbm)2(DMF)2.

On present evidence, then, it appears that in this class of molecules

with C2 symmetry, a positive value of D is associated with elongation

in the z-direction, and vice versa; unfortunately, no theoretical

prediction of the sign of D in this symmetry is available for comparison

with the experimental results.

85

It is interesting that the fine structure parameters of the

Mn (II) impurity do provide fairly detailed structural informati on about

the environment of the metal ion in the host lattice. This contrasts

with the situation in Cu(II)-doped species; thus in planar

Pd(Cu)(acac)2165

as well as in the octahedral complexes

trans-Zn(Cu)(pic)2.4H20166

and cis-Zn(Cu)(hfac)2 (py)2

163 the g-tensor

is nearly axially symmetric. Furthermore the relative magnitudes

of the g-tensor components in Zn(Cu)(hfac)2(py)2

indicate a tetragonally

elongated environment for the copper ion163

(as is found in the pure

copper complex142

) while for the host lattice the distortion has been

shown crystallographically to be in the opposite sense.142

Evidently

the copper ion, far from assuming the configuration of the host metal,

imposes its own preferred stereochemistry at the doped site. The

same effect has been observed subsequently in other Cu(II)-doped

systems.

The complexes M(dike)2(LL) (M=Mg, Zn) with phen and bipy are

assumed to have the cis-octahedral geometry of Cu(hfac)2bipy, the IR

spectra being consistent with bidentate chelation of all the ligands.

Cd(acac)2bipy may also possess this structure but in view of the marked

propensity of this metal ion for C.N. 's> 6,170-174

structures involving

bridging acac cannot be ruled out. X-ray powder patterns of the

complexes show no inter-relationships and are therefore structurally

uninformative. Formulation of any of the complexes as EM(LL)2dike][M(dike)1

may be excluded as the anion might be expected to show the strong

geff = 2 esr signal found for K[M(Mn)(acac)3] (M=Ni, Co, Zn); no

such band appears in the spectra of the bipy adducts, and for the

magnesium and zinc complexes with phen, where D < 0.03cm-1, the expected

transitions near 340 mT are not only sharp but in quantitative agreement

83

with prediction for the chosen SPH parameters. In addition the

magnesium complexes at least are non-electrolytes in nitromothane.153

The D-tensor axes in these molecules are assumed to be oriented.

as in the other cis-adducts of C2 symmetry and the correlation, in

the Mg(II) complexes, between the splitting of v(C=0) and toe magnitude

of D appears to confirm this; the enhanced zfs in the bipy adducts

relative to the phen complexes would then seem to imply more heavily

asymmetric chelation of the diketone in the former, although the

reason for this is far from obvious. The sign of D in Zn(mn)(acac)2phen

suggests lengthening of the in-plane (relative to the axial) M-0

bonds, as occurs in the structurally characterized molecule

Zn(hfac)2(py)

2142

For those of the complexes listed in Table 4.22 which are known

or presumed to have C2 symmetry, A generally lies in the range

0.10 - 0.25; on this basis the distortion in Cd(bipy)12+ (which is

the only one of the examined tris complexes to show measurable

zero-field splitting) is considered to take the form of elongation/

compression along a pair of trans Cd-N bonds, rather than along a

3-fold axis of the octahedron. No structural results appear to be

available for these complexes, but it is worth noting that of a series

2+".176 of tris chelates of Cu(II), both Cu(phen)

2+ 175 and Cu(bipy)

3 3 '

as well as Cu(hfac)3' 121

show pronounced 'tetragonal' elongation

of this nature, while in each of the cations Cu(en)2+ 177

and 3

2 Cu(OMPA)3+

178 the six metal--ligand bonds are equal.


Mq(Mn)(dbm)2(DMF)? was prepared as described in ref. 143, and

Mg(Mn)(hfac)2.3H20 by the method given for Mg(acac)2(E,0)2.51

87

Mq(On)(lfac) phen.0.5H 0 was obtained as pale yellow crystals

on mixing chloroform solutions of phen.H20 and Mg(hfac)2.3H20 (-i- 1 mole%

Mn(hfac)2 (H2 0)2 ) in the stoicheiometric ratio; the fibrous bipy

adduct viz prepared similarly but from benzene solution.

Zn(Mn)(acac)2(1,1..)(LL-phen, bipy). Zn(acac)2(H20)2 [I- 1 mole %

Mn(acac)2 2 2 (H 0) and an equimolar amount of the a-diimine were

dissolved in chloroform and petroleum ether (60-80°C) was added to

incipient precipitation. Off-white crystals were obtained on

overnight refrigeration of the solutions.

Cd(acac)2(LL)(LL=phen, bipy). An ethanolic solution of the

a-diimine was added to an equimolar amount of Na[ Cd(Mn)(acac)3 in water.

The finely divided white products were stirred for several minutes

prior to filtration.

Zn(mn)(phen)3(a04)2.A methanolic solution of the hydrated

metal nitrates and phen.H20 (mole ratio 1:7) was added dropwise to a

large excess of NaCt04 in the minimum volume of water; the white solid

precipitated immediately. Microanalytical results indicate the presence

in the product of - 8% 'free' ligand.

Zn(mn)(bipy)3(No3)2.4H20 and Zn(Mn)(bipy)3(SO4)2.8H20 were

prepared from aqueous solutions of the hydrated metal salts. and bipy •

(mole ratio 1:4). The sulphate precipitated immediately on mixing

the reactants; the nitrate was obtained as large yellow plates on

slow evaporation of the solution at room temperature.

Cd(Mn)(bipy)3(NO3)2 was prepared from aqueous solution as in

ref. 179. The initally formed bis complex was filtered off slow

evaporation of the filtrate gave a mixture of the bis and tris species

which were separated manually.

Cd(Mn)(en)3Br2 and Cd(Mn)(en)?(NO3)2 were obtained by dissolving

the appropriate hydrated Cd(II) and Mn(II) salts in neat, freshly

88

distilled en. The bromide precipitated out at room temperature and

the nitrate, on cooling to C C. Cd (Mn) (en) C.f.'. 2E 0 was prepared in

similar fashion.

89

Table 4,1 Select(:.d hands (cm-1

) in the IR spectra of the

phen adducts

phen.H20 Zn(acac)2phen Cd(acac)2- chen.H2 0 Assignments

- 540 m 540 v.br. V(M-0)

620 635 is 636 m phen: ring bend(?) I

727 s 720 m 721 sh phen: ring-H o.o.p.

738 s 731 s 727 s bend

751 m 761 s 764 m 1 phen: ring-H o.o.p. bend and

775 w 782 m 778 s acac: 6(C-H) o.o.p.

830 s 849 s 847 s ,

phen: ring-H o.o.p.

850 s 860 w 861 w I bend

1133 m 1138 v.w. 1146 m phen: ring-H i.p. bend

1500 s 1505 s 1510 s *1 phen: ring stretch C=C+CN and

1558 m acac: v(C=C)+V(C=0)

1585 m 1585 s \?

1590 s,br

1615 w 1605 s

1640 w I

- - 1675 m.sp acac: uncoordinated V(C=0)

3400 m,br 3400 m,br H20:v(0-H)

aphen bands assigned by analogy with bipy complexes

90

Table 4.2 Selectcd bands (cm-1) in the Ih spectra

of the bipy adducts

bipy Zn (acac) 2hipy Cd (acac) 2 - bipv Assignments

- 539 m 533 s V(M-0)

542 m,sh

618 m 625 m 625 in bipy: ring bend

732 m

736 s 739 m

750 s,br 749 m bipy: ring-H o.o.p. bend and

757 m

765 m 765 sh acac: 5(c-H) o.o.p.

771 s 771 s

1135 m 1150 m,br 1152 m,br bipy: ring-H i.p. bend

1170 w 1170 w

- 1310 m 1315 m coordinated bipy (see text)

-

-1552 ,split

1510 s 1510 s acac: v(C=C)

acac: V(C=0)

1575 s 1575 s,br and

1595 s,br 1600 s,br bipy: ring stretch

C=C+C=N

91

Table 4.3 Selected hands (cm-1 i ) in the IR spectra of the

Mg (II) complexes

Mg(hfac)2.3H20 Mg (hfac) 2phen Mg(hfac)2bipy Assignment

630 w 645 w

725 sh

731 s

phen,bipy: ring bend

738 m

741 w

742 m

762 m

phen,bipy: ring-H o.o.p.bend

771

798 m hfac: v(C-CF3)+6(CF3)

795 s

762 s

1500 s,br

810 w

852 m

870 w

1315 m

1508 s

coordinated bipy (see text)

805 s 801 sh

phen,bipy: ring stretch

C=C+C=N

1516 s

1532 s

1559 m

and

1585 w

1595 w

1527 s

1552 in

1565 m

1575 m

1595 m

1605 m

1645 s

hfac: V (C=C)

hfac: V(C0)

1659 s 1657 s

1545 s,br

1570 s

1617 m

1660 s

1668 sh

20: \)(0-H) 3400-3500

v.br.

Table 4.4 X-band esr spectrum (mT) of Zn(Mn)(acac)2phen; a,b, d and e transitions

Observed

(O=9.535Gliz)

Calc. for

,X =0.03 D=0.025cm 1,X =0.03

Calc. for

A A D=0.028cm , =0.26

Calc. for

A D=0.0:9cm-1 A , =0.10

B Axis,levels B Axis,levels j B Axis,levels

--1 123 w

221.9 ez

233 m 233.6 ez 236.1 ay

238.3 ay

278 m 280.2 dz .

284 s 284.8 a Y

286.3 b y 283.5 b y

287.1 dz

294.6 ax

289.6 a

392 s 391.1 ex 391.6 d

Y 388.0 d

v

398 m 394.1 bz 399.4 b z

400.5 eY

403.8 ex

447 m 447.7 a 448.3 e 451.9 e z Y Y

461.0 az

Table 4.5 Q-band esr spectrum (mT) of Zn(Mn)(acac)2phen; a,b,d and e transitions

Observed

(0=35.739GHz)

Calc. for

A D=0.025cm-1

A =0.03

Calc. for

1 A D=0.028cm , A =0.26

Calc. for

1 A D=0.039cm , A =0.10

B Axis,levels B Axis,levels B. Axis,levels

1109.7 e z

1157.0 ez

1166 w 1169.6 e 1170.3 a 1169.3 z ay Y

1193.1 dz

1214 w 1218.9 a Y

1216.4 z d

1220.0 ax

1221 s 1223.1 d 1222.9 1221.1 b z Y

1228.6 ax •

1326.0 ex

1329 s 1330.2 b 1329.5 d 1329.5 d z Y Y

1334 w 1335.5 ey 1336.4 bz 1336.7 ex

1360.1 bz

1383 w 1383.7 az 1383.7 e

Y 1386.3 e

Y I

1396.8 a z 1

1443.8 az i I

1

94

Table 4.5 PowLler patternsa for the c (4-3) transition for >>

B-Bob

for

0< X <1/9

B-Bo for

1/9 <X< 1/3

Critical point

-9(1-0) 2

-9 (1-X) 2

-36X2 sh

16 (1-3X)

16(1+3X)

shc

divC

div

sh

-9 (1+X) 2 sh

-9 (1-X) 2 sh

-8(1-9X2) div

-36X2 sh

16 (1-3X) div

16(1+3X) sh

0=90° 1 x-axis

(1)=0

0=90 1 ( y-axis

(1)=90° j

0=90°

41=90o - 1 _ •y Cos

0=0 z-axis

5-3A Cos20 -

Is xy plane

9-1 1 (--)

X

zx plane

)

J yz plane

9 (1-X)

=0

20 3X+5 Cos- 9 (X+1)

=90°

aAfter refs. 61 and 161.

b Fields in units of 2D2

992

P.2%

csh = shoulder, div = divergence.

Table 4.7 Analysis of c(4-3) transition in Zn(1411)(acac)2phen at X-band (fields in mT)

Observed

(V=9.535GHz)

Perturbation calculations of powder pattern Calculation of B by exact diagonali-

zation of SPH matrix for D=0.028cm 1,

1=0.26

D=0.025cm-1, 1=0.03 D=0.028cm-1, 1=0.26

B Axis B Axis

332.4 sh

335.7 sh

336.2 div

x

y

331.8 sh x 332.2

337.6 div 337.3 sh y 337.9

338.3 div xy p1.(8=90°, cb=58°) 339.0

340.2 sh z 338.8 sh z 339.3

1

343.8 sh or div

347.0 div zx pl.(0=41°, (1)=0)

342.3 div zx pl.(8=37°, q)=0) 342.8

I

348.4 sh yz pl.(6=42°, (P=90°)

a 356.7 sh yz pl. (8=44°,

(1)=90°)

356.7

aStrong but poorly resolved absorption between 350 and 370 mT.

Table 4.8 5-2 transition (mT) in the X-band spectrum of

Zm(Mn) (acac) 2Phen

Observed

(V=9.535GHz)

Calculated

-1 1)=-0.028cm ,

A=0.26

-1 D=+0.025cm ,

A=0.03

104 104.3 99.3

112 111.6 106.4

119 119.8 114.3

127 128.7 122.8

137 138.4 132.1

147 148.9 142.2

96

97

Table 4.9 c(4-3) transition (mT') of Mg Hfl)(hfac),phen at X--band

Observed

(v=9.522GHz)

Calculated

D=0.03cm-1

, X=0.17 D=0.03cm-1

, X=0.22 Field

direction

332.1 sh 331.5 sh 330.7 sh x

334.5b

335.6 sh 336.1 sh y

335.8 div 336.8 div xy pl.

338.4b

339.1 sh 338.6 sh z

345.1 div 343.5 div zx pl. a

356.1 sh 357.7 sh yz pl.

aStrong but poorly resolved absorption to -380 mT.

bNot clear whether sh or div.

98

Table 4.10 Esr spe'tra (mT) of Mg (Mn) (dbm)(DMF a b,d and c

transitions

Observed Calculated for D-=-0.022cm-1

, 1=0.13

V=35.74OGHz v=9.531GHz B B Axis, levels

(Q-band) (X-band)

1183 w 246 w 1183 246.4 az

1212 ms 277 ms 1212 276.6 e Y

1233 s 1230 293.3 bz

305 s - 306.6 d Y

1306 371.0 ax

371 s 1309 371.4 by

1318 s 1324 387.4 dz

1341 ms 404 ms 1343 406.9 ay

1369 w 434 w 1371 434.8 ez

Table 4.11 c transition (mT) of Mg(Mn)(dbm)2(DMF)2 at X-band

Observed

(V=9.531GHz)

Calculated for D=-0.022cm1, 1=0.13

B Axis

335.3 sh 335.9 sh x

337.1 div 337.6 div y, xy pl.

339.9a 339.9 sh z

342.8a 343.6 div zx pl.

348.2 sh yz pl.

aNot clear whether sh or div.

99

Table/1.12 EsTsp-ct-ra(rr2)c" H 0

Observed Calculated for D--0.038cm-1, A=0.18

V-35.840GHz v=9.531GHz B B Axis, levels

Q-band 1 1 X-band

a115 w

a127 w

1114 w 179 w 1118 179.1 az

1157 ms 221 ms 1156 220.7 ey

1199 s 258 s 1199 259.0 hz

1216 vs 274 vs 1217 276.1 dy

325.9 326.5 cx

332.9 333.6 c ,c y xy

338.6 338.8 cz

400.2 vs 397.6 by

1363 s 418 ms 1361 420.8 dz

1406 ms 466 ms 1406 467.9 ay

1443 w 501 w 1443 503.6 ez

a6,m 1 1 1 transitions

100

Table 4.13 3-baud csr spectrum 07n1 of CO(Nn) (on),Ct2,2H0

Observed

(V=9.533GDz)

-I Calculated for D = 0.041cm , A=0.05

B Axis, levels

167 165.0 ez

245 247.0 a Y

252.7 dz

273.9 ax

285 286.3 b Y

296 297.6 bx

369 368.1 dx

383 382.7 d Y

419.7 ex

432 428.2 bz

442 444.9 e Y

512 516.2 az

101

Table 4.:14 X-band esr spectrum (mT) of C(.3.(Mn) (bipy) 2 (NO3) -2

Observed

(V=9.530Gilz)

Calculated for D = 0.041cm-1

, X = 0.19

B Axis, levels

Continuous absorption between 100 and 250mT

252 m 252.6 dz

269 ms 269.9 by

425 ms 426.8 bz

Moderately strong,

broad absorption 480.4 e y

516 w 516.4 az

Table 4.15 0-band esr spectrum (mT) of Cd(Mn)(bipy)2(NO3)2

Observed

(v=35.74GHz)

Calculated for D = 0.041cm- , , = 0.19

B Axis, levels

1100 m 1101.2 ez

1133a m,br 1139.8 ay

1189 s 1188.4 dz

1207 s 1206.5 b Y

1340 s 1344.1 d Y

1357 s 1364.0 bz

1421a m,br 1415.2 P Y

1452 in 1452.3 az

a Owing to high linewidths and overlap with adjacent transitions, the quoted field positions may be in error by as much as 10 mi.

102

Table 4.16 X-band esr spectrum (mT) of Cd (Mn) (bipy) (NO.), 3 2

Observed

(v=9.5350,1z)

Calculated for D = 0.052cmI , X= 0.13

B Axis, levels

125 t,/ 119.6 ez

128 w

195 w 198.2 a Y

229 w 229.6 dz

260 m 260.9 b Y

405 ms 405.5 dy

416.9 ex

449 m 450.8 bz

498 w 499.2 ey

563 vw 563.6 a z

Table 4.17 Q-band esr spectrum (mT) of Cd(Mn)(bipy)3(NO3)2

Observed

()=35.82GHz)

Calculated for D= 0.052cm- ,X = 0.13

B Axis, levels

1061 m 1057.0 ez

1128 m 1126.9 a Y

1171 ms 1168.0 dz

1202 s 1200.2 b Y

1214.9 ax

1350.4 ex

1351 s 1354.4 d Y

1392 ms 1390.7 bz

1441 m 1435.9 e Y

1506 w 1502.5 az

Table 4.18 p-baPd esr spectrum (mT) of Mq(Mn)(hfac)2bipy

Observed

(V35.750GHz)

Calculated for D = 0.079cm-1,A = 0.24

B Axis, levels

454 w

624 w

941 ms 940.5 ez

990 ms 990.5 a Y

1107 s 1106.9 dz

1127 s 1128.8 b Y

1419 s 1417.6 d Y

1444 s 1444.2 bz

1573 ms 1570.5 e Y

1620 ms 1616.4 az

103

104

Table 4.19 Q-band esr spectrum wn of Cd(Mn)(acac) 2phen.H20

Observed

()=35.865(1Hz)

-1 1 Calculated for D=0.078cm , A=0.17

B Axis, levels

946 w 948.0 ez

1031 w 1033.8 a Y

1112 ms 1113.6 dz

1152 s 1151.6 b Y

1207.3 a

1231 s 1231.6 bx

1400 s 1401.7 d Y

1448 s 1447.3 bz

1538 w 1536.3 e Y

1616 w 1615.8 az

105

Table 4.20 Q-band spectrum of Zn

Observed

(V-33.52a1z)

Calculated

- D-0.061cm 1,

1=0.16

D=0.073cm-1,

A=0.03

D-0.061cm-1,

I I A=0.08

-1 D-0.071cm ,

1=0.09

965 w 964.7 ez

964.9 e

1006 w 1008.0 ez

1007.6 e

1075 w 1077.2 a Y

1079.4 a y

1108 m 1108.2 ay 1110.2 ay

1116.6 dz

1116.8 dz

1127 m

1137.9 dz 1136.0 ax 1138.1 d

1164.8 ax a1183 v Lr 1169.2 b

y 1181.4 by

1173.8 by

1168.0 by

1206.9 ax

1194.4 bx 1184.8 ax 1207.7 b

1214.7 bx

a1337 vbr 1339.8 ex

1331.3 dx 1313.1 dx 1316.4 d

1362.7 dY

1345.6 dY

1345.8 ex 1360.4 dy

1371.6 dY

1383.7 ex

1422 m 1399.0 bz 1411.4 e

x 1399.3 bz

1420.6 b

1420.9 bz

1434 m 1438.2 e Y

1433.2 e y

1462 w 1465.3 ey 1465.9 ey

1527 vw 1530.4 az 1530.2 a

1571 w 1573.0 az 1573.1 az

aFlanked by other, less well-resolved lines.

106

Table 4.21 0-band esr spectrum (mT) of Cd(nn)(acac)2bipy

Observed . Calculated for D,-0.1:63cm-1 , A=0.15

(V=35.88GHz)

......_

B Axis, levels

593 m 588.5 ez

803 m 804.5 ay

933 m 932.8 dz

1018 m 1020.4 by

1125.6 ax

1149 in 1152.9 bx

1219 s 1219.5 cx

1249 s 1250.9 . c Y

1331 s 1330.8 dx

+absorption to

1400

1498.2 ex

1504 in 1505.0 d Y

1625 in 1626.9 bz

1802 w 1798.0 e Y

1987 v'i 1981.0 az

107

Tab le: 4 22 2 I 1 -) a P te I s o1 ) in some cb -is I aLc., conyal

Complex D

(cm-1)

X Ref.

Mg(dbm)2(DNF)2 -0.022 0.13 a

Mg(hfac)2.3H20 -0.038 0.18 a

Zn(OAc)2(H20)2 +0.023 0.10 46,48

Zn(0A02.3H20 0.024 0.11 49

Cd(OAc)2.3H20 +0.032 0.10 47

Mg(acac),)(H20)2 +0.060 0.29 51

Zn(pic)2(H20)2.2H20 +0.064 0.28 37

Mg(hfac)2phen 0.03 0.17 a

Mg(hfac)2bipy 0.079 0.24 a

Zn(acac)2phen -0.028 0.26 a

Zn(acac)2bipy 0.061 0.08 a (i) { }

or 0.071 0.09

(ii) 0.061 0.16 f } 0.071 0.03

Cd(acac)2phen.H20 0.078 0.17 a

Cd(acac)2bipy 0.163 0.15 a

Zn(phen)3(a04)2 <0.01 c a

Zn(bipy)3(NO3)2.4H20 <0.01 c a

Zn(bipy)3SO4.8H20 <0.01 a

Cd(bipy)3(NO3)2 0.052 0.13 a

Cd(en)3Br2 <0.02 c a

continued over/

108


Complex D a Ref.

(cm-1)

Cd(en)3(NO3)2 <0.02 c a

Cd(en)2C2.2.2H20 0.041 0.05 a

Cd(bipy)2(NO3), 0.041 0.19 a

aThis work.

.13Where the sign of D has been determined from hf measurements this is indicated in the table. Errors in the values of D and A are estimated at ±0.01cm-1 and ±0.03 for Mg(Mn)(hfac)

2phen, and at

-10.001cm-1 and ±0.01 in the remaining complexes.

c indeterminate.



9.0 %H 96N %C scH %N

Mg(dbm)2(DMF)2 70.7 5.8 4.7 70.1 5.9 4.5

Mg(hfac)2.3H20 24.6 1.6 - 24.4 1.6 -

Mg(hfac)2phen 42.5 2.0 4.4 42.7 1.6 4.5

Mg(hfac)2bipy 40.5 2.3 4.4 40.4 1.7 4.7

Zn(acac)2phen 59.6 5.5 5.8 59.5 5.0 6.3

Zn(acac)2bipy 57.3 5.0 6.9 57.2 5.3 6.7

Cd(acac)2phen.H20 52.0 4.7 5.6 51.9 4.8 5.5

Cd(acac)2bipy 51.8 4.6 5.8 51.5 4.8 6.0

Zn(phen)3(Ct04)2 55.2 3.5 10.7 53.7 3.0 10.4

Zn(bipy)3(NO3)2.4H20 48.8 3.7 15.4 49.4 4.4 15.4

Zn(bipy)3(SO4)2. - 0 47.0 4.2 11.0 46.6 5.2 10.9

continued •over/



%C %H %N %C %H %N

Cd(bipy)3(NO3)2 51.6 3.5 16.2 51.1 3.4 15.9

Cd(en)3Br2 16.8 5.4 19.2 15.9 5.3 18.6

Cd(en)3 (NO3 )2 17.9 5.8 27.3 17.3 5.8 26.9

Cd(en)2a2.2H20 14.2 5.2 16.3 14.2 5.9 16.5

Cd(bipy)2(NO3)2 43.8 2.9 15.3 43.8 2.9 15.3

CIS CH 3 3 I

_--C /':-.----0\ /0

H C Mn / \ CH >Dx

0 C°'°

111

CH3 CH3

Fig. 4.1 The orientation of the D--tensor in trans-Mg(Mh)(acac)2(H20)2.

The Dz vector is perpendicular to the plane of the paper.

After ref. 51 .

0 218

0--?/...( et OH2 Z n

o OH2

0 Fig. 4.2 Bond lengths in Zn(0Ac)(H

2 20) 2

1-1.9. 4.3 The presumed orientation of the D-tensor in cis-MnA4B-)

x10

1,1

100 -300

500 mi Fig. 4.4(a) Esr spectrum (9.527GHz) of Zn(Mn)(acac),phen, 0-600mT.

n

II

11

ii

11!; ;1

ri, II I

I,; ,

i

ii 1 i

'' '

i I r I A ; !"

....../.,........./\...../N.,.../\_...7\___/\,,-- V • I !!11 I ; I! f il

1i II Iil It Il;111 Iil 11

I) II - 1 Al " li .6 i ill I\ l\ i

i

w J \ •

J 250 • 350

Fig. 4.4 (b) Esr spectrum (9.527GHz) of Zn (Mn) (acac)2phen, 250-45OmT.

i ii i

1 /4 Al;/ ..11 1111'.4

/

111;4, 1

11:II

I .11 11 '1 1

1 11 I 1 1 111 I .1:;

` ,Aov litjv I :11i r I I I -

I i

11

'PI.11 11 I nts,..,..„--.....--.......—....-.

''

1 I Cv ,

1 1 il 41 /

; hIl 1 1, i ■

, 0 .1 , 1,, .■ :I II

II 11 ' i 1 11 II 11 il 1 1; It Ill 1

1.11 11 1 1 11 it, 1 1

1 '' :/ P-, 1 1 1 1 1 1 I 1 I

11

11

I I I

1111

111

I I 1

III

41 "

, „ X 25 1111

Ili I II Iq

X25

I I

; I

\I 1

I II 11

11 1 1 11

II

I 1 t tl ItL 1 II , 1 ! I 1 1 1 L_L_Lj_ j_j_it 100 300 500 700 mT

Fig. 4.5 Esr spectrum (9.539GHz, 0-700mT) of Mg(14370(dbm)2(DMF),).

;• t !It'll] t I I l t t Lt ti, 1 I•t .1_111 t tit t ;III •••1 , I. •• 1 •1 •

100 300 500 mT Fig. 4.6 Esr spectrum (9.535GHz, 0-600mT) of Cd(2%Mn)(bipy)3(NO3)2.

I/1.0 Ibi.111 111 -$7 17 7OS•11 a3.44 47 .tt Fyn- II.I.11

(a) 0 A TA 1;i1d11 A '11;j1P 11 .111 kvA

I , ►

116

Fig. 4.7 ESr spectra (9.535GHz, 0-700mT) of Mg(Mn)(hfac)2bipy:

(a) experimental; (b) calculated for D=0.079cm-1

,

A=0.24, Alipp=40mT.

(a)

,_.„.___...-,-----•-".„.

,.----",---v--v----'"--- ,\.‘r

117

(b)

, tl.43 a2 .0 -114.: 7) 4.1.4a 4t?.11 614.t:r 1=4.73 tra ti-7-0:(70 7/1 .t,Z MT FILLO

Fig. 4.8 Esr spectra (9.527GHz, 0-800mT) of Cd(Mn)(acac)obhen.H20:

(a) experimental; (b) calculated for D-0.078cm-1 A=0.17,

.4\_Epp=30mT.

118

00 6 .47 1F2.E7 14.57 2E5.74 217.0 5fi0.61 3b0.04 41.40 42.01 61446 M.70 t17.t2 t FULD

Fig. 4.9 Esr spectra (9.539GHz, 0-700mT) of Zn(Mn)(acac)7bipy:(1) calculated for eaual numbers of Mn(II) ions in sites with (a) D=0.061em-1, A=0.08 and (b) D-0.071em-1, A=0.09; (ii) as for (i), but with (a) D=0.061cm-1, =.0.16 and (b) DF0.071=-1, ?,=0.03. Hpp=40mT foy both (i) and (ii). (iii) Experimental spectrum.

mT

1000 1200

1400 mT

Fig: 4.10 Part of the Q-band esr spectrum (V=35.880GHz, 950-1450mT) of Cd(Mn)(acac)2bipy.

120

(a)

•42 ."72.27 0,4.10 Az Al 7.17.11 64.7 41 114 "St 411.10 4V7f. D 212.26 616.72 i17.22-16.41•aic0.02 77147 ■D.1•1,47"-I-A,M-74--MG.22 7'17.711 11,21.T

M

Fig. 4.11 Esr spectrum (v=9.530GHz, 0-1000mT) of Cd(Mp)(acac),bipy:

(a) experimental; (b) calculated. for D=0.163cm 1 , A=0.15,

30mT. ipp=

(-27)

5,11, n

C1(f)(

. mn

CHAPTER

ESE AND VIBRATIONAL SPECTRA OF SOME

MANGANESE(II)—DOPED POLYMERIC COT'IPLEXES

Introduction

This chapter describes the results of esr studies on some

Mn(II)-doped complexes of zinc, cadmium and mercury halides with the

ligands py, pyz and hydrazine. Most of the complexes are known or

belived to be polymeric in the solid state. Apart from its intrinsic

interest, the stereochemical information obtainable by this technique

may well be of value in the interpretation of exchange effects in the

paramagnetic analogues, the magnitude and sign of the magnetic inter-

action for a given bridging ligand being critically dependent on the

bridge geometry and the stereochemistry at the metal. For example the

dimer Mn2(biquino1y1)2Ct4 (Fig. 5.1) with C -Mn-C =800180

is found

to be ferromagnetic,180

in contrast to Mn(py)2Ct.2 (where, by analogy

with the isomorphous cobalt complex,181

the CZ-Mn-Ct. angle is expected

to be close to 900) for which J is negative.182

Ground state zero-field splittings in the trans N4X/N2X4

C. f 8 "

C1 7) -7,:\\ COO) )c(3)

c(6r C(5) c")

Fig. 5.1 (from ref. 180)

121

122

systems generally exceed by far those encountered hitherto in this

work. Strong transitions between adjacent spin-levels, but with

single-crystal turning-points not along the principal axes, appear

in all the X-band spectra; their intensities are comparable with those

of the 'perpendicular' transitions. These off-axis lines have been

discussed previously by Sweeney et al.,50

who have demonstrated the

importance of lineshape calculations in spectral interpretation when

IDI < 3.5hy. Off-axis transitions have also been identified in the present

work at Q-band frequency, when they are often. distinguished by the

presence of irregular or unresolved hf structure (v.i.).

The halide-bridged polymers ML2X2(L=py or a ring-substituted

pyridine) of first row transition metal ions have been examined thoroughly

by a variety of physical techniques;183-200

only IR and Raman

183-185,201-203 188 and thermogravmetric analyses spectra i

of some of

the Gp.IIB complexes have been reported, although 3-dimensional

X-ray diffraction data are available for a few members of the series204-207

(see Table 5.1). The relevant information afforded by these studies

is summarized below.

The absence in the cadmium complexes of any IR band above 200cm1

assignable to v(Cd-X)(X=Ct,Br) was recognized184,185

as being consistent

with the presence of bridging halide, and recent far-IR(450-20cm-1) and

Raman data202

are fully in accord with expectation for a chain polymer

of line-group isomorphous with the point-group D2h. Cu(py)2Ct2 and

n Hg(py)2Ct2, which crystallize in the same space-group as Cd(py)2C,t.,2,

206

are known to have one pair of M-Ct bonds considerably longer than

° the other (Hg-Ct(1)=2.34A, Hg-C (2)=3.25A

208 ). An early suggestion,

209

based on Fourier analysis of X-ray diffraction data, of similarly

asymmetric halide coordination in the cadmium complex has now been

disproved;206

in this complex, as in Cd(py)2Br2207

(Table 5.1) the

123

metal-ion environment is therefore similar to that in CY-Co(py)2c.

It may be mentioned here that both Mn(by)2a2 and Mn(py)2Br2 are

isomorphou with a-Co(py)2C.E2.206

The zinc complexes are known, from IR184,201

and crystal-

204,205 lographic work, to be monomeric and pseudotetrahedral,the metal

ion being in an environment of C2v

symmetry.

1:1 complexes of pyrazine with halides of the Group IIB metals

were first reported by Stidham et al.,210

who estimated relative chain-

lengths in the polymers using the now discredited211

mid-IR criterion of

Lever;212

more recently, polymeric structures for the zinc complexes

have been postulated on the basis of IR studies to 80cm-1.213

While the

chloride was claimed to be octahedral,213 a pseudotetrahedral arrange-

ment, with bridging pyrazine ligands was indicated for the bromide

and iodide.213

Vibrational data for the cadmium polymers, on the other

hand, revealed both the chloride and bromide to be octahedral,202 and

this structure has been confirmed crystallographically for the former

complex.202

The iodide was tentatively assigned a tetrahedral structure.202

Despite several literature reports of the complexes M(pyz)2CC2

where N is a first-row transition metal,212,214,215

no record exists

of the preparation of analogous complexes of zinc, cadmium or mercury;

it has been claimed,210

in fact, that as long as adequate pyz is

present in the reaction mixture, 1:1 complexes are obtained, irrespective

of the metal-pyrazine ratio. (The only reported complex of a different

stoicheiometry is Zn(pyz)3/2

Br2' believed t7) be the decomposition

product of the initially-formed bis complex.210

) In the present work,

however, preparations with pyz-metal ratios of -2:1 yielded solids giving

non-reproducible esr spectra which clearly indicate occupation of more

than one site by the Mn(II) ion (see Fig. 5.3). The use of a large

excess of one or other reactant permits isolation of solids of stoicheio-

metry Zn(pyz)Ct2.0.5H20 (in which the presence of water is confirmed

181

124

by IR absorption at -3500 and -1700 (To 1) and Zn(pyz)7/4C,e.2

both of which are shown by far-IR and Raman spectroscopy, in addition

to esr, to be polymeric. Elemental analysis for the latter solid is ••

of course as might be expected. for a mixture of the complexes

Zn(pyz)2C,E2 and Zn(pyz)Ce2 in the ratio 3:1, but no trace of the mono-

pyrazine complex can be detected in the esr or vibrational spectra, and

the contaminant may be either Znee2.xH20, which has limited solubility

in acetone, or, bearing in mind the moderately high basicity of the

ligand, some form of hydrous zinc oxide. However spectral analysis

leaves no doubt that Zn(pyz)2a2, structurally similar to the cobalt

complex 216

(Table 5.1), is the major constituent. The bromide analogues

have also been prepared, and the 2:1 complex appears to be quite stable,

which contradicts the earlier finding210

cited above.

The isolation of polymeric Zn(pyz)Ct2 is unexpected in view of

the non-existence of zinc analogues of other halide-bridged transition

metal complexes, virtually all known complexes of type ZnL2X2 (L=py, or

ring-substituted py) being tetrahedral. Apparently the increase in

lattice energy resulting from cross-linking of (Zna2)n chains by bridging

pyz groups is sufficiently large to overcome the strong preference of

Zn(II) for tetrahedral coordination.

X-ray powder photographs were obtained for the 1:1 chlorides

and bromides and for the 2:1 zinc complexes; none of the patterns

shows any marked resemblance to any other.

Esr finds particular application in studies of Mn(II)-doped

hydrazine complexes of Group IIB metal ions, as the ligand absorbs

strongly below 250cm-1, making identification of metal-halide vibrations

difficult.217

Only complexes with briding hydrazine are considered here;

it should he repeated,218

nevertheless, that determination of the mode

of ligation from the position of v(N-N), as has been proposed, 219-221

125

is subject to uncertainty. In first the ass,icfnment of the

-1. 222 band at 875cm in liquid hydrazine s N)(N-N) has been challenged,

and secondly the approximation that this vibration is 'pure' may not be

tenable in the complexes, particularly where the ligand is bidentate;218

moreover, the N-N bond-length, as determined from X-ray work is the

same, within experimental error, in N2H4 and in the mono- and bidentate

complexes.218

The complexes M(N2H4)2X2 (M=Zn(II), Mn(II) or Cd(II);

n 223 X=Ct, Br or I) are isomorphous; the the complexes with ZnC-u2

and

n 224 Mne-t2

have been shown, by single-crystal X-ray diffraction, to

comprise chains of hydrazine-bridged MC C2 units parallel to the c-axis

of the crystal. Surprisingly the Mn-Ct. bond length is marginally

shorter than Zn-Ct. (Table 5.1).

In several cases, and notably for a number of mercury complexes,

the attempted preparation of Mn(II)-doped polymers was unsuccessful,

or the poor quality and/or non-reproducibility of the spectra precluded

detailed analysis. The highly asymmetric halide coordination in

208 Hg(py)2X2 might be expected to result in a A-value in excess of those

found in, e.g. Cd(py)2X2 and Cd(pyz)X2(X=Ct„Br); failure to obtain

reproducible esr spectra of Hg(py)2Ct2 is disappointing but not unexpected,

in view of the known instability of this complex.203

Vibrational Spectra of the Complexes 7,n(pyz)X2.nH20 and- Zn(pyz)2X2

(X=CZ,Br).

In all four complexes the apparent lack of coincidences of

IR and Raman bands in the region (1600-400cm-1) of the internal pyrazine

fundamentals suggests that the centre of symmetry of the pyrazine

126

molecule is retained, i.e. the ligand is bidentate. Pyrazine has

no fundamentals below 400 cm .

(i) The 2:1 complexes (Table 5.2)

The marked similarities between the spectra suggest that the

complexes are isostrucLural. Esr studies indicate in each case a

tetragonally distorted N4X2 environment for the metal ion, and the

IR results are in broad agreement with this; the stronger of the

two halide-sensitive bands in the far-IR (at 238 cm-1

in the chloride

and 160 cm 1 in the bromide) lies within the range of values observed

for V(Zn-Xterminal) in octahedral 1L4

X2' and is quite inconsistent

with the presence of, e.g.,. tetrahedral ML2X2 or octahedral ML2X4

.;involving bridging halides and monodentate pyz). The skeletal modes

for Co(pyz)2Ct2,in which the presence of the polymeric sheet structure

has been verified216

by X-ray analysis, have been listed:211

= sheet

alg(R) + e (R) + 2a2u

(IR)

+ b2u(inactive) + 3eu

(IR)

rV(MX) = alg(R) + a2u(IR)

V(MN) = eu(IR).

The factor group is taken as D4h and the pyz molecules are treated

as point masses.211

The occurrence in the zinc complexes of five IR

bands between 250 and 100 cm1 appears to accord well with these

predictions; moreover,

1) all the bands appear at lower frequencies than in the Co(II)

counterpart, and

2) bands assigned in this scheme as v(Zn-N) and 6(NZnN) are,

127

as expected, in virtually the same positions in both the

zinc halide complexes; the band thought to be 6(NZnX) moves

to lower frequency in the bromide although, predictably, the shift

is not as pronounced as for V(Zn-X) .

Raman spectra for this type of complex have not been reported

previously. The only predicted Raman-active stretching frequency is

Vsym (M -X); in fact the two strong bands observed at 195 and 219/228 cm

-1

are virtually independent of halide, the latter absorption actually

occurring at higher energy in the bromide. Despite the apparent in-

adequacy of the foregoing analysis, however, the strong similarity of

the IR spectra to those of the cobalt analogues provides empirical

evidence for the presence of a similar sheet polymeric structure

in the zinc complexes.

(ii) The 1:1 complexes (Tables 5.3, 5.4)

IR spectra to 80 cm-1

of the anhydrous complexes have been

reported previously213

but detailed assignments were not made; the

results are in fair agreement with those given here.

The IR spectrum of the chloride is similar to those of

M(pyz)C,L2186 (M=Mn, Co, Ni) which have been interpreted in terms of

a sheet structure with halide-bridged (MX2)n chains linked by pyz

groups. If the sheet factor group be taken as D2h, the predicted 186

skeletal modes are as follows:

rsheet = ag(R) + blg(R) + b2g(R) + 2blu(IR)

+ 2b2u(IR) + 2b3u(IR)

rV(MX) = a .(R) + b2g(R) + blu(IR) + b3u(IR)

128

(MX2)n o.o.p. = blg

(R) + b2u(TR)

= b2u(IR)

- V (MN)

r46(NmX) = blu(IR) + b

3u(TR)

Five of the observed bands are assigned without difficulty

(Table 5.3) but it is not possible to say whether the moderately

intense absorption at 106 cm-1 in the IR is the o.o.p. chain mode or

one of the IR-active NZnCe. deformations. The Raman-active o.o.p.

(MX2)n mode is apparently too weak for observation. Comparison of the

Raman data with those of the cadmium analogue202 reveals that the band

energies of v (M-Ct.) are similar, while shows the expected shift

sym as

to lower energy in the cadmium complex; the effect has been observed

previously,226 and is ascribed to greater involvement of the metal ion

in the asymmetric stretching motion.

Zn(pyz)Br2 is unique among the present complexes in the

complexity of its low-frequency Raman spectrum: which points to a

site-symmetry lower than in the chloride; it is also noted that the

Zn-ligand stretching frequencies lie within the range found for

tetrahedrally coordinated zinc(II). Since it has been shown above that

the pyrazine ligand is bidentate, the vibrational spectra will be

analyzed in terms of the previously suggested213

chain structure of

(ZnBr2) units linked by pyrazine bridges. The normal modes for

tetrahedral MX2L2

(point group C2v) are:

rmol . 4a

1(R,IR) + a

2(R) + 2b

1(RIR) + 2b

2(R,IR)

= a1 ' (R IR) + b

1 (R,IR)

rV(MX)

V(ML) = al(R,IR) + b2(R,IR)

129

Comparison the isostructural iodide complex has led213

to assignment of the two higher energy bands in the far-IR as Zn-Br

stretches and the remaining two bands, as V(Zn-N). The detailed (but

tentative) assignments in Table 5.4 have been made on the assumption

that the symmetric stretching modes (al) are strongest in the Raman,

while the asymmetric (b1,b2) stretches are most intense in the IR. It

is interesting that Val(Sn-Br) is not observed in the Raman, and, in

contrast to the behaviour202

of ML2X2 where L is monodentate,

Val(Zn-N) occurs at higher energy than Vb2(Zn-N); both features have

been observed preViously202

in Cd(dioxan)I2 which is believed to

possess the same tetrahedral polymeric structure.

Esr Spectra

The esr spectra of the Mn(II)-doped complexes appear in

Figs. 5.2-17 and Tables 5.5-16. In no case has it been possible to

determine the sign of D. At X-band, for values of D> - 0.10 cm-1,

the positions of individual hf lines are not accurately reproduced by

second-order perturbation theory; third-order terms in, for example,

A2D, ED and E

2A must be taken into consideration,

227 and the method

used to determine the sign of D in Chapters 2-4 is here inappropriate.

At Q-band, where D/hV is sufficiently low for the third order terms to

be negligible, shifts in the hf lines are so small as to be well within

the experimental uncertainty in the measured field positions. In a few

complexes of type trans-MA

4B2' the sign of D has been determined

from the temperature dependence of esr line intensities228

or from

heat capacity229

or magnetic susceptibility229

measurements. D is

positive in the tetragonally elongated species Mn(pyrazole)412229

and

and Mg(Mn)-CZ boracite17(where the metal ion is in an 04C-t.2 environment)

130

228,230 , and negative in Mn(II)-doped NHAC,t, where the Mn (II) ion lies

at the centre of a tetragonally compressed octahedron with four chloride

ions in the equatorial plane and water molecules in the axial positions.

The metal ions in most of the polymeric species studied here

were found to occupy sites of axial or near-axial symmetry; since this

is true also of the Me5dien and dab complexes of Chapter 6 it is

convenient to discuss their esr spectra together with those of the

polymers.

In Cd(py)2a2, Cd(4Me-py)2a2, Zn(pyz)a2,0.5H20 and Cd(pyz)CZ2,

the magnitude of the axial distortion is similar to that in Cd(acac)3

and Cd(acac)2bipy and, as in the latter complexes, severe overlapping

of transitions precludes detailed analysis of X-band spectra; however

the parameters extracted from Q-band resonance fields (Tables 5.5-8)

have been used to calculate X-band spectra which are found to agree

well with experiment. (see Fig. 5.2).

In Zn(Mn)(dab)Ct.3 (Fig. 6.17),on the other hand, D is high

enough, and the hf structure sufficiently well resolved, to allow

direct measurement of resonant fields at the lower frequency. The highly

irregular hf pattern on the lowest band (0-7OmT) is noteworthy, as is

the very poor resolution at 500-640 mT. The remaining bands lead

to the assignment D = 0.19cm-1

, A = 0.04, whereupon the anomalous

absorptions are seen to correspond to off-axis turning-points (Table 6.7).

In several of the complexes, in fact, the occurrence of an irregular

hf pattern or of unresolved hf structure has proved to be a valuable

means of identifying off-axis transitions between adjacent spin levels

(see Tables 5.12, 5.13 and 5.15). The effect is probably due to the

appreciable intensity, at intermediate orientations of the magnetic

field, of nuclear forbidden hf lines, which in powder spectra may not

be resolved from the allowed transitions. In species of axial symmetry

131

with g and A isotropic, and for D, A<<cp11 0, the intensity of the

AmI=±1 transitions, 1+1 mI

1M,-1, S"

and 1M m S' I

relative to the Am.=0 lines is given74

by

}2 , (31)Sin20/4g(3B0) 2 {1+(.s (S+1) /3ms (Ms-1).] j X [I (I+1) -m1 (m1-1)]

The AmI=±1 lines are therefore expected. to be strongest at some inter-

mediate value of 0 and to vanish for BIIx, y or z. The formula also

predicts higher relative intensities for the Ami=±1 lines in the

electronic transitions l/24-*-1/2 and ±3/2±-)- ±1/2 than in ±-5/2± ± 3/2;

it has been observedpreviously,60

and is confirmed in this work, that

these last transitions (2-1 and 6-5) do not yield off-axis powder lines.

Perturbation theory affords a satisfactory quantitative treatment of

the nuclear forbidden transitions only when B is very close to a

principal axis; at higher values of 0 orcb, 'mixing' of the nuclear

levels leads to breakdown of the normal selection rules. Table 5.14

gives computer calculated resonance fields for the transition centred

on 1593 mT in the 36 GHz spectrum of Zn(Mn)(pyz)2Br2. It is seen that

within spin levels 4 and 5 transitions to virtually all nuclear states

are allowed, which accounts for the unusual'hfpattern in the powder

spectrum (Fig. 5.15(ii)). It should be borne in mind, however, that

under certain conditions, notably when the energy levels in question

are parallel in their functional relationship with the magnetic field,

even a 'normal' (i.e. along-axis) transition may show an unusual hf

pattern; the 5-4 transitions, for BIlz,in Zn(Mn)(Ph3P0)2C,E2

are a case in point and are discussed in detail in ref. 38.

X-band spectra of the complexes [Ni(Mn)(dab)2Cqa (Fig. 6.18),

Cd(Mn)(py)2Br2 (Fig. 5.8), Cd(Mn)(pyz)Br2 (Fig. 5.9), Zn(Mn)(N2H4)2C-C.2

(Fig. 5.6), Cd(Mn)(N2H4)2C,C.2 (Fig. 5.5) and Zn(Mn)(pyz)2Ct2 (Fig. 5.4)

all show the now familiar50 7-line fine structure characteristic of

'axial' species with D in the range -0.20 - - 0.55 cm-1. It is emphasized,

122

however, that tbe only along-axis transitions of the group are a,b

and f 5.10) all of which correspond to perpendicular to the

principal axis (Table 5.10). The 'spurious' bands (p and q in Fig. 5.10)

in the calculated spectra for many of the non-axial species are an

artefact of the icosahedral approximation, as is clear from examination

of Fig. 5.10(i) and (ii). For the latter calculation, the points on

the sphere which define the directions in which B is to be applied

were chosen. at 1° intervals on one line of arc from pole to equator.

This procedure gives a lineshape superior to that of Fig. 5.10(i),

but its use is of course restricted to the axial case. Another noteworthy

feature is the substantial (-10mT) shift to low field. of band f in the

lower spectrum; this explains why this band is consistently higher

than its experimental counterpart in the calculated lineshapes of

Figs. 5.4, 5.5 and 6.15 while the remaining lines are very well fitted.

The low-lying (<100 mT) 5-4 transition in the yz plane is well

separated from the geff = 6 line at D = 0.19cm-1

but moves upfield

with increasing axial distortion and is virtually coincident with the

latter absorption at D = 0.24cm1; any apparent splitting of this

band at these D-values should not, therefore, be assumed, in.:the

absence of additional evidence, to arise from lowering of the symmetry

from axial to rhombic. The splitting of band f, if this can be

measured accurately, is a more reliable and sensitive index of the

degree of rhombic distortion. At the linewidths observed in this work,

bands b and d are not visibly split for values of A <-0,025. Q-band

spectra confirm the X-band results in every case. The sharp, spin-

forbidden lines above 400 mT are exactly analogous to those at

120-130 mT in the X-band spectra of the complexes M(Mn)(4Me-pyo)6(a04)2;

in both cases D/hv= 0.2.

For values of D between -0.60 cm-1

and -0.75 cm-1, X-band

133

spectra usually show no roze than four hands. Aea.rt from the

strong geff = 6 band, the 4-3 transition in the perpendjcular plane

is prominent and when well resolved, as in Zn(Mn)(pvz)2Br2 (Fig. 5.13)

and M(Mn)(N2H4)2Br2(M=Cd,Hg) (Fig. 5.11) allows immediate determination

of both zfs parameters. In the two hydrazine complexes any splitting

of the geff

= 6 transition due to the very slight rhombicity of the

D-tensor cannot be detected; the x and y components of 4-3, however,

are separated by --40 mT and the centres of gravity of the two sextets,

at 552 and 593 mTi agree well with the calculated fields, 553.0 mT

(BIlx) and 593.4 mT(Elly), for X= 0.006. (In Zn(Mn) (N H ) 2 4 2 -2

(Fig. 5.12) and [Ni(Mn)(dab) 2 Br

3] Br (Fig. 6.19), where the hf

structure is less well defined, X has been taken as zero.) All the

remaining bands represent off-axis turning points. 0-band resonant

fields for Zn(Mn)(pyz)2Br2 are given in Table 5.13.

In the Q-band spectra of M(Mn)(N2H4)2I2 (M=Zn,Cd), assignment

of the weak but fairly well resolved sextets at 140-160 mT to

x,y(4-3) leads to the parameters of Tables 5.15 and 5.16; the presence

of an off-axis band at 600-700 mT is in accord with this interpretation.

An entirely satisfactory analysis of the rather ill-defined X-band

spectra could not be achieved, however; the experimental lineshapes (Figs.

5.16 and 5.17) show little relation to those calculated from Q-band

parameters, but the weak bands observed between 600 and 900 mT at

9.5 GHz are roughly in the positions expected for x,y(4-3). It is

necessary to point out that the assumptions of gis0=2.00 and a

vanishingly small cubic field term a may not be justified in the case

of these iodide complexes; a has been found to be unusually large

in Zn(Mn)(Ph3PO)

2I2 and g markedly anisotropic in the chloride and

38 34,231,232 bromide analogues. Also it is now well established

that

134

g is enhanced in proportion to the deree of covalenne in

ligand bonds, and the value of this paraweter in. d5 systews mi be

expected to increase from the chloride to the iodide in a group of

isostructural halide complexes. As only three or four transitions at

Q-band could be measured with any accuracy in the spectra. of the bis

hydrazine complexes no attempt was made to fit the data to resonance

fields calculated for ial >0 and g >2.

Accurate zfs parameters could not be obtained for the complex

Cd(Mn)(pyz)I2; apart from a strong unsplit geff = 6 band, the only

absorption at X--band occurs at -820 mT. The extreme simplicity of

the spectrum suggests a very high axial zfs, and assignment of the high

field band as x,y(4-3) gives D = 1.23 cm-1, A = O. The Q-band spectrum

shows the three fine structure lines predicted for these parameters

but several other prominent transitions are observed over the entire

field range. The latter might be assumed to arise from a contaminant

of less severely distorted molecular geometry, but the X-band spectrum

provides no evidence for the presence of any such species. Attempts were

also made to fit the data with D = 0.60 cm-1 and 0.35 cm , both of which

give rise to absorption at -SOO mT at X-band, but in each case there are

serious discrepancies between observed and calculated spectra at one or

both frequencies.

The spectrum of manganese(II) in Zn(py)2C-e2 is found to depend

on the preparative method employed; a sample obtained by precipitation .

with ether from a pyridine solution of the metal salts gives (at X-band

frequency) a very broad (Allpp -

50 mT), strong geff

= 2 signal. suggesting

that much of the Mn(II) is not incorporated into the crystal lattice;

however weak but well-defined sextets centred on 143 and 194 mT are also

_., observed, as in the pseudotetrahedral ( - C2v) complexes Zn(Mn)(Ph

3PO)2et2

38

and Zn(Mn)(1,2Dimeim)2a2

282. On the other hand, the complex obtained

135

by addition of the stoicheiometric quantity of py to a boiling

ethanolic solution of the metal halides gives a. single band at 340mT

with well-resolved hyperfine structure. Spectra of the latter type are

also given by manganese (II) in Zn (py) 2Br2, Cd(PY)21-,, and Zn(pyz)Br2.

201,202,213 For all these: complexes, far-IR results indicate a tetrahedral

stereochemistry for the metal ion; in the case of Zn(py)2C12 the

presence of discrete ZnN2CZ2 units of approximately tetrahedral symmetry

has been confirmed by single-crystal X-ray diffraction.204

The effect on the spectrum of increasing the level of manganese(IT)

was studied for several of the polymers. In every case it was found

that, apart from the loss of hf structure and the inevitable broadening

due to spin-spin interaction, the lineshape was virtually unchanged even

for manganese(II) concentrations of 20% or higher.

Discussion

Zfs parameters of the Mn(II)--doped polymers and of some previously

examined 4:1 complexes are listed in Tables 5.17 and 5.18. The order of

the D values in the chloro complexes is approximately M(4Me-py)2Ct2'=

14.(pY)2CZ2-.:::M(PY'z)a..2 <M4Me-py)4a24M(pyz)2a2‹.M(N2H4)2C.,E2

M(5Me-pzt)4C22. (The relatively high value for the pyrazole complex

is in accord with electronic spectral233

and crystallographic234

results

which indicate attenuation of the axial M-Ct bonds due to hydrogen bonding

between the halide ions and the imine nitrogen atoms of the vertically .

oriented pyrazole rings.) The sequence for the bromo complexes is

similar except in that the positions of M(pyz)2X2 and M(5Me-pzi._)4X2 are

interchanged. Where electronic spectra of the nickel analogues are

available, the calculated values of the ligand field parameter

- xy DtL = -(Dq -Da7)] also lie in the above order (Table 5.19). The axial

distortion in Zn(pyz)2Br2 appears to he anomalously high, exceeding even

136

that in Zn(N2H4)2B; it is rot known whether the effect occurs also

in the nickel analogue, the available electronic spectral data212

for

Ni(pyz)2X2(X-CZ,Br) being insufficient to allow unambiguous assignment

of the bands.

The listed hydrazine complex of mercuric bromide is highly

susceptible to loss of ligand and satisfactory analytical results were

not obtained; however the esr spectrum of a. freshly prepared Mn(II)-doped

sample is identical with that of the cadmium complex, contradicting an

earlier proposal,235

based on IR evidence, of a linear Hg(N2H4)22+ cation.

A is found to be very small in all the polymers and it appears that

the g.s. splitting of the Mn(II) guest is not appreciably influenced

by structural factors outside the primary coordination sphere (such as

the conformation of the hydrazine ligand, or the inclination of py or

pyz rings to the M-N-C(3) axis) that might be expected to contribute

to the asymmetry in the xy plane. This result is significant because

it justifies the assumption of D4h symmetry for the interpretation of

spectral data in these complexes; difficulties in the past in fitting

spectral and magnetic results on a D4h model have been imputed

to the supposedly low site symmetry at the metal atom.

Comparison of the D values for Cd(Mn)(py)2X2 and Cd(Mn)(pyz)X2

confirms an earlier observation211 of roughly equal ligand field strengths

for py and bridging pyz; a more unexpected result (and one which

immediately establishes the new complexes Zn(pyz)2X2 as pyrazine- rather

than halide-bridged) is the very large difference in the values of D for

the donor sets trans- N4X2 and -N2X4. The apparently lower axial

distortion in the latter might be rationalized on the assumption that

the principal D-tensor axis lies along the unique metal-ligand bond in

MN4X2 but in the plane of the halide ions in the polymers; in the

latter case, however, a sizeable rhombic distortion would be expected

137

(e.g. I is 0.07 in trans- E.::r(en) 20.21 Ce.HCt.2H20, where Dzz bisects

the open angles in the plane containing the nitrogen atoms149

) and is

not observed. In addition, MOssbauer study of polymeric Fe(H20)2CZ2

239

has confirmed that the direction of maximum electric field gradient

coincides with. the Fe-OH2 bond, and it is reasonable to assume that D

is similarly oriented in the present structurally related complexes.

Pond lengths in Co(py)2Ce2 exceed those in Co(py)4a2 (Table

5.1); if this is generally true for the species MNiX and MN 2X4,

somewhat lower D values for the latter might be expected from electrostatic

considerations. The commonly occurring reduction in D from a Zn(II)

host to its Cd(II) analogue is a manifestation of this effect, as is,

presumably, the observed decrease, both absolute and relative, in

IDZn-1)Cd in the isostructural complexes M(N2H4)2X2

on passing from the

chloride to the iodide. In the chlorides the percentage reduction in D

on altering the host metal ion from Zn(II) to Cd(II) is as high as

17%; for the bromide and iodide the figures are 6% and 2%. However, since

the difference in the ionic radii of Zn(II) and Cd(II) is,-0.2 and

0 bond-lengths are on average less than--0.3A higher in M(py)2X2 than

in M(py)4X2, some additional explanation for the disproportionately

large difference in their D values is called for. It is informative,

in this connection, to compare the esr results with the recently reported240

crystal-field parameters for Ni(py)2X2 (X=CZ,Br) (Table 5.20).

Electronic spectra of the complexes Ni(py)2X2 have been studied,

by several workers;191, 237,240,241 despite some differences of opinion

with regard to band assignments, particularly at low energy, it is

generally agreed that the 3E9 components (in D

4h symmetry) of the

3T7 g

(F)

and 3Tlg(F) levels lie sufficiently close that neglect of the off-

diagonal elements of the crystal-field matrix is not justified, and that

configuration interaction (between the -F and 3P states) must also be

136

taken into consideration for the purpose of evaluating ligand-field

parameters. There is some doubt as to whether the weak band at--6kK

in both spectra represents a d-d transition. The data (i) of Table 5.20

are obtained when this band is assigned as 3B1g

3B2g 2'F);

237 if

it is assumed, on the other hand, that the 3B2g transition is hidden

under the 3E band and the 6kK absorption is ascribed to some other cause g

(such as interactions between the chains, or the presence of Nia;

fragments in the chain) the parameters (ii) result.240

In assignment (i),

z xyl the value of IDq -Dg for Ni(py) exceeds that previously determined 2. 2

for Ni(py)4X2'

Assignment (ii), which is claimed240

to give better agree-

ment between experimental and calculated band positions, leads to the

opposite result, I z Dq -DqXyI then showing the same trend as D in the

manganese-doped complexes. The interesting point about assignment. (ii)

is that the calculated Dq for the bridged halide ions of the polymers

is higher than DqZ of the terminal halides in Ni(py)4X,, despite the much

240 shorter Ni-X distance in the latter complexes; this is attributed

to the lower lone-pair electron density and 1T-donor capability of

bridging, as opposed to terminal, halide. If this is in fact the case,

then, other things being equal, monomeric trans-ML2X4

might be expected

to show axial distortion of a magnitude similar to that in ML4X2, and

appreciably greater than that in polymeric ML2X2. Electronic spectra

-1242 are available for one such system, Rb[Nia4(H20)2J and polymeric

NiC-E2

(H20)2,240 and zfs's of the ground state have been derived from

magnetic susceptibility studies.242,243

It is seen from Table 5.21 that

bond-lengths and angles in the two chromophores are closely comparable.

In accordance with the postulate of enhanced ligand field strength in

bridged halide, the electronic spectral bands V1 and V2 are seen to

occur at somewhat higher energy in the polymer, Dq. being ,--40 cm-1

higher than in Rb21-NiCt4(H20)1; on the other hand, D is about three

times higher than in the rubidium salt, for which no splittings of the

139

d-d bands are resolved, even in the polarized. absorption spectrum at

1.6 K (Table 5.21). In view of these apparently contradictory results

(and bearing in mind also that the demonstrated240,244

presence of

substantial hydrogen bonding in Ni(H20)2CZ2 may cause the anomalously

large axial compression .in this compound) it seemed desirable to attempt

the preparation of other monomeric, tetragonal complexes of type trans-

MN2X4, for spectroscopic comparison with the polymers ML2X2 (L=NH3,py,

aniline).

It has been shown in recent years245-253

that unipositive ligands,

such as mono-protonated or -quarternized diamines or diphosphines,

readily coordinate to metal ions, and that the field strength of the

ligand is not greatly attenuated by the presence of the positive charge,

provided that this is sufficiently remote from the coordination site. It

seemed likely that complexes of the requisite geometry might conveniently

be prepared by the use of a sterically unencumbered ligand of this type,

and, after some preliminary work involving salts of en, pyz, quinoxaline

and piperazine, analytically pure, X-ray isomorphous complexes of

composition ML2Br4 were obtained for M=Co, Ni, Mn and L = 4, 4'-bipyridinium.

The proximity of the band positions in the electronic spectra of the

Ni(II) and Co(II) complexes to those in Ni(py)2Br2191

and

Co(3Br-py)2Br2200 confirms the presence in both cases of an essentially

tetragonal N2Br4 environment (Table 5.22). In the nickel complex,

splittings of the 'octahedral' Vi and V2 transitions are apparently

smaller than in polymeric Ni(py)2Br2 but, for reasons outlined above,

the significance of this observation is uncertain; owing to rather

poor spectral resolution and the inapplicability of the simple weak-field

tetragonal treatment, no attempt has been made to obtain the crystal-

field parameters. It is worth noting, in relation to the unassigned240

6kK band in the spectrum of Ni(py)2Br2, that both Co(bipy11)2Br4 and

140

the nickel complex appear to absorb, weakly in. the 4.5 - 6kK region,

,but the position of the band maximum is obscured by a sharp overlying

4 4 peak [probably an overtone of v(C-EI)] at -6.1 k..K. The .7\2 ,-).Tl (F)

transition in tetrahedral. CoBr42-

occurs at 5.5 kK,254

and traces of

this isomeric species in octahedral Co(bipyH)2Br4

could account for the

observed absorptioni but the presence of a band at similar energy in the

nickel complex is not easily explained, as 3T+

3A2(v2

) of NiBr42-

lies near 7 kK.255

Table 5.22 also includes, for comparison, band energies in

octahedral NiL2X4 (L = 1,1, 1-trimethylhydrazinium; X = Ct,Br),

250 the

detailed molecular geometries of which are not known. No splitting of

v2 is detectable here, in contrast to the pronounced asymmetry of this

band in Ni(py)2X2 and Ni(bipyH)2Br4.

Unfortunately attempts to prepare solid solutions of Mn(bipyH)2Br4

in the nickel and cobalt complexes for esr spectroscopy were unsuccessful,

but the spectrum of the pure manganese complex indicates quite clearly

that the desired monomeric structure is not attained. The single

remarkably broad (AHPP 300 mT) band centred on -300 mT indicates the

presence of magnetic exchange between Mn(II) ions, which could occur

1 in these complexes only via halide bridging. The dimer [Mn2L4X6J

2+ .2X

represented in Fig. 5.18 is regarded as the most attractive of several

structural possibilities, on the ground that this arrangement leaves

just a sufficient number of free bromide ions to 'buffer' the

bipyridinium ligands of adjacent M-M pairs.


Cd(Mn)(py)2C,E2 was obtained on addition of a large excess of

pyridine to a hot suspension of powdered Cdet2.2.5H20 and MnC-e2.4H20

in equal volumes of ethanol and DMP. The precipitated solid was stirred

141

under nitrogen for several minutes. The 4:11,( yv com7) was prepared

similarly.

Cd(Mn)(py)B 2 was obtained on drop; ic;e addition of a hot ethanol

solution of Cd13r2.4H20 and MnBr2.4H20 to a 10-fold excess of py in

ethanol.

Zn(Mn)(pyz)et2.0.5H20. An acetone solution of pyz was added

dropwise to a vigorously stirred suspension of Znet2 and Mnae2

.4112 0

'

in the same solvent, the metal-ligand ratio being 5:1. The finely divided

product was stirred for several minutes, filtered, washed with ethanol

to remove unreacted Zna2, and air-dried. Zn(Mn)(pyz)Br2 was prepared

in similar fashion from n-propanol.

Zn(Mn)(pyz)2Ct2. A suspension of ZnCt2 and Mnet2.4H20 in hot

n-propanol was added slowly to a stirred solution of pyz in the same

solvent (metal-ligand ratio 1:5) and the precipitated solid was filtered

off and air-dried. Zn(Mn)(pyz),Br, was prepared similarly, but from

acetone.

Cd(Mn)(pyz)X2(X=Ce,Br). The appropriate cadmium and manganese

halides were dissolved in boiling ethanol and added to an ethanol solution

containing 2 equivalents of pyz.

M(Mn)(N2-4 2 H ) et

2 (M=Zn, Cd). A solution of the metal chlorides in

hydrazine hydrate was added to stirred ethanol.

Zn(Mn)(N2114)2sI2 was obtained on dropwise addition of hydrazine

hydrate to a boiling solution of ZnBr and MnBr2.4H20 in methanol.

Cd(Mn)(N2114)2Br2. A hot acetone suspension of CdBr2.4H20 and

MnBr2.4H20 was added to excess hydrazine hydrate in acetone. As in the

case of 'Zn(pyz)2Ct21 , the analytical data indicate the metal-ligand

ratio in the product to be 1:1.75; substantial contamination by a

manganese-containing species is also evident from the strength of the

geff

=2 esr signal. However the anisotropic esr spectrum, which is fully

142

reproducible over sever i different proparatioy_ is believed to originate

from the majr)r constitutent of the product, viz. Cd(Mn)(N,H ) Br . 4 2 2

IIg(Mn)(N,H, 1 2)Br,. An ethanol solution of HgBr2 and

MnBr2.4H20 was added to the stoicheiomotric quantity of hydrazine hydrate

in the same solvent. V(N-N) in the freshly prepared product is 952 cm-1

,

as reported by Brodersen.235

Zn(Mn) (112H )212. A solution of Zn17 and MnI2.4H20 in acetone

containing a few drops of DMP was refluxed briefly and an acetone solution

containing a stoicheiometric excess of hydrazine hydrate was added

dropwise.

Cd(Mn)(N2H4)22:2. An ethanol solution of CdI2

and MnI2.4H20

containing a few drops of DMP was refluxed briefly and added to the

stoicheiometric amount of hydrazine hydrate in ethanol.

M(bipyH)2BIA (M=Mn, Co, Zn). (a) Preparation of the ligand. 4,4'-

bipyridyl was dissolved in water containing a few drops of ethanol.

Hydrobromic acid (60%) was added in slight excess of the amount required

for the formation of bipy.2HBr, which crystallized out on evaporation

of the solution. The dibromide was treated with an equimolar amount of

bipy in hot water; the monoprotonated base crystallized out on ccoling

the solution. (b) Preparation of the complexes. The hydrated metal

bromides were dissolved in methanol/DMP and the solution was refluxed

briefly and added slowly to the stoicheiometric amount of the ligand

in the same solvent.

143

Table 5.1 Bonfl lengths in some MN X2 -

MN2-4 X -type com_ .plexes

Complex M-0) M-X(:) Ref.

a-co(py)2cZ2 256

Type A 2.11 (i) 2.49

(ii) 2.52

Type B 2.18 ,

_

(i) 2.51

(ii) 2.50

Co(py)40e2 1.99 2.32 257

Co(pyz)2Ct2 2.18 2.41 216

Cd(py)2a2 2.35 2.65 206

Cd(py)2Br2 2.12 2.76 207

Mn(N H )- Ct 2 4 2--2 2.27 2.57 224

Zn(N2H4)20E2 2.15 2.58 223

Table 5.2 Low-frequency vibrational spectra (cm1) of M(pyz)2X2

Complex

V (M-X)

IR spectruma

V (M-N) (5(NMN) S(Nmx)

Raman spectrum

a2u eu

a2u

+ eu

eu 1

Co(pyz)2a2b 260 s 235 s 162 s 135 s 125 s

Zn(pyz)2C.E2

Zn(pyz)2Br2

238 s

160 m

196 m

196 m

145 s 126 m

(142) 128 m

114 m

103 m

219s,br; 195s,br; 128s,br;

228m, 195s,131s, 112s, 77s

71s, 51sh I

aTentative assignments only (see text)

bRoom temperature data from ref. 211.

Table 5.3 Low-frequency vibrational spectra (cm ) of Zn(pyz)0Z2.0.5H20 and Cd(pyz)C.E2

Complex V (M-N)

b2u

v(M-CZ)

ag + b

2g blu

+ b3u

Other bands •

Zn(pyz)Ct2.0.5H20 IR 255 sh 212s,br; 172s,br 106m

R 218s, 181s

Cd(pyz)CC2a

IR

R

217 s

215s; 153m

180s, 120s 195sh, 80w

aData (300K) from ref. 202.

145

-1 Table 5.4' Low"frquoncy vibrational. spectra. - (cm ) of

F--- Raman IR Tentative assignment

4.9 s

61 s

68 sh

84 s

92 s

108 s

122 s

130 sh

184 w

217 s 210 m b2 V(Zn-N)

226 vs 240 sh a1 V(Zn-N)

257 m a V(Zn-Br) 1

278 w 270 ms b1 V(Zn-Br)

146

Table 5.5 Q-band esr spectrum (mT) of Cd(MniM72C,

, Observed

N=35.875C-7.,z)

CalcuLated for D=0.125cm-I

, A=0.02

B T.P. Field direction Levels

753 m 744.0 5.0 z 2-1

1012 ms 1012.7 8.1 z 3-2

1013.4 5.1 y 6-5

1047 ms 1046.2 5.1 x 6-5

1124 s 1125.6 8.0 y 5-4

1140 s 1139.9 8.0 x 5-4

1381 s 1383.1 8.0 x 3-2

1398 s 1400.5 8.0 y 3-2

1542.0 5.1 x 2-1

1546 ms

1547.1 8.1 z 5-4

1571 ms 1573.0 5.1 y 2-1

1813 w 1814.9 5.0 z 6-5

147

TalA.e 5,6 0-1-,:-And. spectrum (m= P) of Cc.i 1.431) ( ,17,147‘--r

....

Observed

(V-35.86GHz)

Calculated for D=0.118cm-1 , X=0.01 ]


461 w 464.0 0.1 0=30°, (P=0 5-2

466.0 0.1 0=300, 4)=,0o 5-2

686.0 0.5 0=600, 4)=0 4-2

683 w 692.0 0.5 0=60°, cb=90° 4-2

773 m 775.5 5.0 z 2-1

1036 ms 1036.5 5.1 y 6-5

1053 ms 1052.0 5.1 x 6-5

1138 s 1139.6 8.0 y 5-4

1146 s 1146.4 8.0 x 5-4

1381 s 1384.3 8.0 x 3-2

1390 s 1392.5 8.0 y 3-2

1531 ms 1533.7 8.0 z 5-4

1535.8 5.1 x 2-1

1549 ms 1546.8 5.1 y 2-1

1787 w 1786.5 5.0 z 6-5

148

Table 5.7 (-barr'l esr spectrum mT) of Zn.pn nw—z)C' 0

Observed Calculated for D.143=-1 =0 , 1-0.03

(V=35.820GHz) B T.P. Field direction Levels

696 w 667.1 5.0 z 2-1

969.6 5.1 y 6-5 970 m

973.3 8.0 z 3-2

1022 m 1026.0 5.1 x 6-5

1098 s 1098.7 8.0 y 5-4

1126 s 1122.7 8.0 x 5-4

1395 s 1388.7 8.0 x 3-2

1417 s 1418.9 8.0 y 3-2

1567 m 1572.4 5.1 x 2-1

1585.7 8.0 z 5-4

1624 m 1625.3 5.1 y 2-1

1892.1 5.0 z 6-5

149

Table Q-hand es r spectrum (mT) of Cd(n)(pyz)Ct_

Observed Calculated for D=0.111cm-1

, X=0.01

(9=35.735 447) B T.P. Field direction Levels

459.3 0.1 0=30°, ch =O 5-2 456 w

461.6 0.1 0=30°, (:)=90° 5-2

679.4 0.4 6=60°, (i)=0 4-2 681 w

684.9 0.5 0=60°, 4)=900 4-2

800 one 801.1 5.0 z 2-1

1000 1038.8 8.0 z 3-2

to 1045.5 5.1 y 6-5

1080 m 1 1060.1 5.1 x 6-5

1141 s 1143.9 8.0 y 5-4

1150 s 1150.4 8.0 x 5-4

1371 s 3374.9 8.0 x 3-2

1380 s 1382.6 8.0 y 3-2

1514 m 1514.3 8.0 z 5-4

1515.9 5.1 x 2-1

1531 m 1529.7 5.1 y 2-1

v.w. absorption 1752.0 5.0 z 6-5

centred on 1750

N

150

Table 5.9 0-hand esr spectrur,). (MT) of Zn

Observed

(V='-35.0255Az)

-1 Calculated for D-0.24om , A r=0.005


250.7 5.0 z 3-1

254 m

467 m 471.0 1.0 0=30°, (1)=0 5-2

479 m 475.0 1.0 0=30°, 4)=90° 5-2

483 m

769 w 765.7 7.9 z 3-2

834 m 836.6 5.3 y 6-5

851 m 852.0 5.3 x 6-5

990 ms 992.4 7.8 y 5-4

997.9 7.8 x 5-4

1187 s 1188.2 8.5 x 4-3

1190.8 8.5 y 4-3

1279 w 1279.7 8.9 z 4-3

1438.0 7.7 x 3-2

/ 1450 vs 1447.1 7.7 y 3-2

1796 w,br 1793.8 7.9 z 5-4

1815.4 5.2 x 2-1

1833 w 1829.9 5.2 y 2-1

Table 5.10 Assignment of resonance fields (mT) at e=9.521GHz for D=0.24cm 1 , A=0

Band designation

in Fig. 5.10

fExtremum in cal-

culated spectrum

of Fig. 5.10(ii)

Calculated single

crystal turning-

point

T.P. Field direction Levels

a -125.0a 126.0 5.2 x,y 6-5

b 366.2(min) 355.4 5.4 x,y 4-3

c 469.1(min) 458.5(min) 5.4 0=140 5-4

d 654.3(min) 644.0(max) 5.6 0=63° 3-2

e 720.1(min) 706.3(min) 4.2 0=28° 4-3

f 888.8(min) 877.8 7.0 x,y 2-1

g 983.4(min) 971.3(max) 4.4 0=8° 5-4

a Field measured at point where trace crosses base-line.

152

Table 5.11 Q-band esr sccoLrum (mT) of Zn (En) W,I1,)„CX_

Observed

(V-35.065Gliz)

Calculated for D--0.29cm1 , A,--0.015


236 in 234.3 0.5 0=60°, (1)=90° 6-3

-450 in 474.5 1.9 0=40°, (1)=90° 5-2

481.6 1.1 0=60°, q)-90° 5-3

659 w 660.3 8.0 z 3-2

747 m 752.5 5.4 y 6-5

806 in 807.1 5.5 x 6-5

930 s 933.8 7.9 y 5-4

-950 m 950.8 7.9 x 5-4

1154 in 1156.1 8.4 x 4-3

1166 ms 1168.8 8.4 y 4-3

strong overlapping 1424.9 7.5 x 3-2

absorption 1459.2 7.6 y 3-2 j(

1400-1600

1937 w,br 1902.2 8.0 z 5-4

1965.7 5.3 y 2-1

153

Table 5.12 0-band OSY cmeci—um (Tm) of Cd(Mn)(7,,,)

Observed -1 Calculated for D=0.30cm , Xr-0

(V=35.840011z) B T.P. Field direction Levels

231 m 230.7 0.5 0=65° 6-3

460 m 464.3 2.0 0=35° 5-2

643 w 637.0 7.9 z 3-2

700-800 m 766.9 5.5 y,x 6-5

. 807.0 2.3 0=55° 4-2

928 s 931.2 7.9 17,x 5-4

1153 vs 1156.5 8.3 y,x 4-3

1280 w 1280.2 8.9 z 4-3

1439 m 1438.0 7.5 y,x 3-2

-1525 ma 1527.9 7.4 0=35° 4-3

1541.8 7.0 0=70° 3-2

. 1922.8 7.9 z 5-4

1941 w

1961.9 5.3 y,x 3-1

structure absent. ahf

15,1

7, „ 2 2 Table 5.13 spectrrim (iy,T) of Zn(na)(PYz)

Observed Calculated for D=C.75cm A=0.0

(V=36.000GHz) T.P.

-148a s 154.0

314 w 320.4

493 s 501.0

1253 s 1259.2

1285 m 1286.0

1593bw 1590.9

2.0

8.0

5.4

5.8

9.0

5.8

Field direction Levels

0=28° 5-4

6-4

x,y 6-5

x,y 4-3

o 6-4

8=15° 5-4

a Irregular hif structure: peaks at 133 br,s; 143 sp,s;

148 w; 153 sp,s; 167 sp,s; 177 sp,m.

Peaks at 1579, 1588, 1600 and 1614.

155

Table 5.14 Calcnlated resonance f.'lelds (inT) for transltions_

between sp.in levelo 4 and 5(e=15°) driZn(En)jp Br,.

D-0.75em-1, )■=0, A=-8.7mT,

B /es

T.Y. hf levels B res

T.P. hf levels

1562.9 0.35 30-19 1592.3 0.90 28-23

1566.6 1.43 30-20 1593.0 0.51 27-21

1570.2 1.35 29-19 1593.7 0.94 26-19

1570.4 2.23 30-21 1595.8 1.48 28-24

1574.0 1.67 29-20 1596.5 0.98 27-22

1574.1 1.65 30-22 1597.5 1.70 26-20

1577.8 2.16 28-19 1601.2 0.38 26-21

1577.8 0.57 30-23 1602.2 0.24 25-19

1581.3 0.94 29-22 1604.7 0.30 26-22

1581.5 0.14 28-20 1606.1 1.10 25-20

1584.9 1.67 29-23 1603.2 1.92 27-24

1585.1 1.11 28-21 1608.0 1.38 26-23

1585.6 1.88 27-19 1610.0 1.87 25-21

1588.6 0.53 29-24 1611.0 0.96 26-24

1588.7 0.33 28-22 1613.7 1.43 25-22

1589.3 0.59 27-20 1617.0 0.48 25-23

I

156

'.1: (m:17) of Zn (Mr') (NH ) 2--4- 22

Observed

(v=35.73GHx)

_ Calculated for D-1.21cm-1 , A=0.019

I


413 s,ra 415.7 4.9 y 6-5

489 s,r 491.2 6.1 x 6-5

-634 m,p 653.5 2.3 0=20°, q)=90° 5-4

669.6 0.67 e-10°, 4)=-0 5-4

b 1404.9 4.9 x 4-3

1597 m,r 1599.2 5.0 y 4-3

ar-well resolved, 6-line hf structure

IT=poorly defined hf structure

bgeff

=2 region obscured by broad, very strong absorption.

Table 5.16 -band esr spectrum (mT) of Cd(Mn)(N H ) 242I2

Observed Calculated for 1),,--1.18cm-1

, A=0.01

(V--35.725GHz) B T.P. Field direction Levels

430 m,ra 432.5 5.25 y 6-5

468 s,r 472.4 5.79 x 6-5

710 m,p b

geff

=2

region obscured by

broad, v.strong

absorption

1437 m,r 1436.4 4.97 x 4-3

)1532 m,r 1535.5 4.99 y 4-3

aSee footnote a to Table 5.15

'DOff-axis transition; see Table 5.15

157

Table 5,17 Zfs parameters of In (II) in some meric comT.)lexes

Complex Zfs parameters

-1 D(cm ) A Table Figure

(i) MN2X4

complexes

Cd(py)2Ct2 0.125 0.02 5.5 5.2

Cd(4Nc-py)2C,e2 0.118 0.01 5.6

Cd(py)2Br2 0.30 0 5.12 5.8

Zn(pyz)Ct2.0.5H20 0.143 0.03 5.7 5.3

Cd(pyz)Ct2 0.111 0.01 5.8

Cd(pyz)Br2 0.31 0 5.9

(ii) MN4X2 complexes

Zn(pyz)2Cf2 0.24 0.005 5.9 5.4

Zn(pyz)2Br2 0.75 0 5.13, 5.14 5.13-15

Zn(N2H4)2CZ2 0.29 0.015 5.11 5.6,5.7

Cd(N2H4)2a2 0.24 0.015 5.5

Zn(N2H4)2Br2 0.71 0 5.12

Cd(N2H4)2Br2 0.67 0.006

Hg(N2H4)2Br2 0.67 0.006 5.11

Zn(N2H4)212 1.21 0.019 5.15 5.17

Cd(N2H4)2

12 1.18 0.010 5.16 5.16

158

5.1P Z s param? of Mn(TT) in some tetraonnal ccrfplex

Mn X -----1--2

Complex 1

D(cm ) X

aCo(4Me-py)4CZ2

-0.20 -0

aCo(4Me-py)4

Br2 -0.62 -0

bCo(4Me-py)4I2 0.87 -0

aMn(3Me-pzt)4Ct2 0.25 - 0.30 -0

aMn(3Me-przt)4Br2

c

-0.62 -0

Mn(pzt)4I2

dZn3B7X013(cubic phase

boracites)

(i) X = Ct

+0.98

+0.275

0

0

(ii) X = Br +0.415 0

(iii) X = I +0.625 0

aApproximate zfs parameters obtained by comparison of published spectra

(refs. 33 and 226) with lineshapes calculated in this work.

bRef. 226.

CRef. 229. Spectra of the pyrazole complexes are stated to be very

similar to those of 3(5)-methylpyrazole (ref. 33).

dRef. 17. The metal and halide ions occupy the interstices of an

uninterrupted three-dimensional network of boron and oxygen atoms.

Each metal ion has trans--octahedral coordination with four (approximately 258

planar) oxygen and two halide neighbours at distances of -2R and -3A.

Table 5.19 Valu.es of the licland , c.,,I!r Pt in some

Ni (II) complexeZ1,

Complex Dt(cm 1) Ref.

Ni(FY)2CZ2 -89 240

Ni(4Me-py)4Ct2 +300 191

Ni(pz,E)4C,C2 +337 233

Ni(N2H4)2CZ2 +390 217

Ni(py)2Br2 -90 240

Ni(4Me-py)4Br2 +410 191

Ni(pz.e.)4Br2 +418 233

Ni(N2H4)2Br2

+430 217

aExcept for Ni(py)2X2, Pt has been calculated from the splitting

of the first spin-allowed spectral band using the relation

35 74— Dt = [E(

3E ) - E(

3B2g)j . Owing to neglect of configuration

g

interaction and off-diagonal crystal field matrix elements, the listed

values are probably overestimates, but do reflect the order in

which the true Dt values lie in the MN4X2 complexes.

Table 5.20 Ligand field strengths (cm-1) in some

octahedral nickel cont:plenes. (Data from

refs. 237, 240)

Complex Dci DqxY IDqz-DqxY1

Ni(py)2a2: (i) 1210 629 581

(ii) 946 790 156

Ni(py)2Br2: (i) 1200 594 606

(ii) 928 770 158

Ni (PY ) 4cz2 680 1173 493

Ni(py)4Br2 600 1149 549

6 0

Table 5.21 Comv)arion of. the trans-Nla_ (2 0) cbr,-,mr)ol,o7= •

in Rb,[NiCt (2 0) 1 and. NiCt (2- .0) 2' 4 -2 2- 2 -2 2

(Data from references 240, 242-244)

Bond lengths:

Ni - CZ(g)

Ni - 0&

Bond angles:

0 - Ni - CZ(1)

0 - Ni - a.(2)

a(1)- Ni - a(2)

Electronic

spectrum (k. K.)

D(cm-1)

Rb2ENCZ4(H20)2 NiCZ2 2

(HO) 2

(1) 2.431 (2) 2.438

2.077

.

89.1

89.5

90.1

7.3

12.3

}a

22.7

....,0.3

(1) 2.387, (2) 2.410

2.089

87.9

89.6

87.6

b c 7.7 .

12.6,13.2

23.2

}

----'0.98

a1.6

oK

bBroad, asymmetric.

c77oK

161

Table 5.22 Reflectance spectra (kK) of some Mli X4 chromophores

Complex Spectrum Ref.

Ni(bipyH)2Br4 d 6.9 sh, 8.1 v.br, m, 11.8 sh, 13.5 m, a b

Ni(py)2Br2 5.8, 8.0, 11.6 sh, 13.7 m a 191

Ni(TNH)2Br4c 6.15, 7.14, 11.95, 18.4 sh, 21.6 250

Ni(py)2Ce2 6.0, 8.4, 12.2 sh, 13.9 22.2 sh, 24.1 191

Ni(TME)2Ct4c 6.15, 7.3, 12.6, 19.05 sh, 22.6, a 250

Co(bipyH)2Br4 d 8.3 m, 13.3 m, sh, 15.4 sp,s 18.2 sp,s a b

Co(3Br-py)2Br2 5.6s, 8.2 m, 12.6, 15.2 s, 17.4, 18.3 s, 21.5 200

ad-d spectrum obscured by charge transfer band

°This work

cTMH = 1,1, 1-trimethylhydrazinium

dLow-energy (4.5-6kK) absorption present; see text.

163

TF)le 5.23 Analytical Results

Compound

%(.7

Found

%H %N %C

Calculated

%H 96N

Cd(py)2a2 34.9 2.7 8.3 35.2 3.0 8.2

Cd(py)2Br2 27.8 2.2 6.6 27.9 2.3 6.5

Zn(pyz)Ct.2.0.5H20 21.5 1.9 12.2 21.3 2.2 12.4

Cd(pyz)Ct2 19.2 1.8 10.8 18.2 1.5 10.6

Zn(pyz)Br2 15.5 1.3 9.2 15.7 1.3 9.2

Cd(pyz)Br2 13.7 1.3 7.7 13.6 1.1 8.0

Zn(pyz)2C.E2 30.0 2.5 18.4 32.4 2.7 18.9

Zn(pyz)2Br2 25.2 2.0 14.9 24.9 2.1 14.5

Zn(N2H4)2Ce2 - 4.1 27.3 - 4.0 28.0

Cd(N,H4)2Ct2 - 3.3 21.9 - 3.3 22.7

Zn(N2H4)2Br2 - 2.9 19.4 - 2.8 19.4

Cd(N2H4)2Br2 - 2.3 14.9 - 2.4 16.7

Zn(N2H4)212 - 1.9 14.3 - 2.1 14.6

Cd(N2H4)2I2 - 1.7 13.5 - 1.9 13.0

Mn(bipyH)2Br4 34.8 2.7 8.0 34.9 2.6 8.1

Co(bipyH)2Br4 34.7 2.9 8.1 34.7 2.6 8.1

Ni(bipyH)2Br4 35.0 2.8 8.3 34.7 2.6 8.1

( 1 )

164

.CO 61.0 1Ui.a W.50 11)6.14 Al .17 L.•01 31 111,0 4137.111 61,3T-V17711 snort 6..CY 1}O.011 1.71.6t 10/t i51.t mT

Fig. 5.2 Esr spectra (9.530GHz, 0-1000 mT) of Cd(Mn)(py)2CZ2:

(i) experimental; (ii) calculated for D=0.125cm 1,

F111.0

X=0.02, Alipp=40mT

165

..J.■04110121•411..

1000 0 500 B/mT

Fig. 5.3 X-band esr spectra of Mn(II) in the pyrazine complexes of

Zna2: (i) Zn(pyz)Ct2.0.51120; (iii) Zn(pyz)2a2;

(ii) product of a preparation us.l.ng a Zna2/pyz

ratio of 1:2. The non-reproducible spectrum shows

features of both (i) and (iii).

166

00 sl.0 14241 ti130 15.7♦ [bat iC.•0t AO•C■ 41.40 441f.no 6 I .1L, 1400 117.12 010 720.C3 151.41 i/t.,, 0'74 OS UN .09 mT

Fig. 5.4 Esr spectra (9.530GHz, O-l000mT) of Zn(Mn)(pyz)2a2:

(i) experimentali(ii) calculated for D=0.24cm-1

,

A=0.005 , Q Hpp=40mT

11)7

ao alas flnArr 114.p) lte.o- th.17 34,1 41t.o—Ttiffl e eio,u ete.as 71.tx 6,s4' 17 mT

Fig. 5.5 Esr spectra (9.521GHz, 0-1000mT) of Cd(5%/4n)(N H ) a :

(i) experimental; (ii) calculated for D=0.24cm-1,

X=0.015, AHpp=4OmT.

168

(1)

....—..—"\\.‘,,,.........

axv, 0114) M7o. iemosv.41 s11.1 011.4* 7.044 7Sim et.os 04.m; is.sa en. mT

Fig. 5.6 Esr spectra (9.512GHz, 0-1000mT) of Zn(Mn)(NnH A),,a,;

(i) experimental; (ii) calculated for. D=0.29cm-1

,

A=0.015, AHpp=40mT.

169

(i )

t■•71 3i•I3 64.11 t5.611 6.17 OM, 10.n 11.7t is.s7 tin •!: CT

Fig. 5.7 Esr spectra (35.865GHz, 0-2000mT) of Zn(Mn)(N2H4)2a2:

-1 (i) experimental; (ii) calculated for D=0.29cm ,

X=0.015, AHpp-40mT.

170

\I

IC 0'7.50 1'7S•00 Oee .50 .00 6S1.na abs . so eri.sc 6<o•00 01. rho PM -00 741.0 I I 0 . V: .17.ca s<II.00 Lciz.ra I coo .cro 7147.6,0 1175. .1.0 11:27E-0.-735). m T

Fig. 5.8 Esr spectra (9.533GHz, 0-1000mT) of Cd(2096Mn)(pv)2Br2:

experimental; -- calculated for D=0.30cm

A=0.0, AHpo=40mT.

(i )

171

51.43 1C2.57 154.30 R5.74 25747 252.61 5n.114 41.0 4c2.91 624.25 E65oe 627.:2 666.25 720.09 771.52 U2.55 8'7441f 77 FIELD

Fig. 5.9 Esr spectra (9.520GHz, 0-1000mT) of Cd(596Mn)(pyz)Br2:

(i) experimental; (ii) 0=0.31cm-1 , X=0.0,

AHpp=40mT.

172

0 71.0 1 ,32 tA'44 tn.n ,b0.n wm

Fig. 5.10 X-band esr spectra calculated for D=C124cm-1, X=0,

AHpp=4OmT: (i) using 93 angles per octant in the

icosahed.ral approximation; (ii) with magnetic field

directions chosen at 1° intervals alone one line of

arc from pole to equator.

X25

o'.a 711.00 140.00 2112.00 Z.00 4h.O' 41°4° iS" 62FTEL.0.6'"D' 7',7.1a M.00 a D•00 Oc.c zxx zr,.x .C3 '114.°2 31n." 744.14' mT

Fig. 5.11 Esr spectra (V7-9.532GHz. , 0-1400mT) of Hg (21/411n) (N2H4) 2Br_.2 : (i) exp,er_imo-en,717ppe • Cd (2%Mn) (N H Br is identical) ; (ii) calculated for .G06, AH

0.00 /411.00 10410 210.00 220.00 -74040 420.00 420.00 01440 surinoTisst.at 00.00 1,42.03 10.00 2.1.00 1141,02 ilro.rn ita.r3 vre?..= mT Fig. 5.12 Esr spectra (v=9.510GH-; 0-1400mT) of Zn(5%/.1n) (1.\1-,H,) Br - (i) ,xoerimental; (ii) calculated for 4. - . _

D=0. 71 cm 2=0 H -35mT.

A=0 ,

640 /i•IC. 1.5.04 Z17.2.0 6o.vo A2.m) 61.= nb743 coma frr4g FLCLD

Eldt-SO S.;10,30 atm.sa 14,..r..a mt.= Isv.ta rrit

Fig. 5.13 Esr spectra (9.525GHz, 0-1400mT) of Zn(Mn) (pvz) ,-)Br-).; (i) experimental; (ii) calculated for D=O.75cm

(1 )

176

cT .CD MIT 7..33 66.53 7J.17 60.53 ice.cs f.5.17 Aem 00411 1■13.03 I 1E40

Fig. 5.14 Esr spectra (36.00GHz, 50-1800mT) of Zn(Mn)(pyz)2Br2:

(i) experimental; (ii) calculated for. D=0.75cm-1,

X=0, AHpc,-50mT.

177

1614 1579

177

133

• Fig. 5.15 Off-axis transitions centred on (i) 143mT and

1593mT in Lhe 0-band spectrum of Zn(Mn)(pyz1 2Br2,

showing irregular hf structure.

178

.0.1 61.0 11247 11441 2.16.74 211.7 0. 841 010.0. 411.0 .L2FILL0 61,36 a6.71 ISO .22 6E045 7ZO.CG 1'11.52 I72 41 04.31 5,0.55 677.4 rrif

Fig. 5.16 Esr spectra (9.531GHz, 0-1000mT)of Cd(596Mn)(N211)2I2:

-1 (i) experimental; (ii) calculated for D=1.18cm ,

A=0.01, AHpp=4OmT. No microwave absorption was detected

between 1000 and 1350mT.

17D

.DO 11.41 ItZ•l? Ill..) 11644 [0.11 11141 11“).04 .41•C:roi 171.15 II-67711 Z7.17-41$.1th eha.01, Zt.GS 0.1 .36 111"7.6S ;Nal ITT URL

Fig. 5.17 Esr spectra (9.530GHz, 0-l000mT) of Zn(5%Mn)(N2N)21.2:

(i) experimental; (ii) calculated for D=1.21cm-1 ,X=0.019,

L\Hpp=4OmT. No microwave absorption was detected

between :MOO anci 1350mT

0

0

X\ /X\

m/

x

Nx/ Nx

180

Fig, 5.18 Proposed structure for M(bipyH)2X.04,--Mn,Co,Ni; X=Br)

181

CH2TTER 6

Mn.NOTIESE(II) COMPLEXES OF 2-PIETITYTJTjD=LE

Introduction.

While complexes of zinc(II), cobalt(II), nickel(II) and copper (II)

with 2-methylimidazole have been the subject of detailed study,259,260

few manganese(II) complexes have been examined, owing, presumably,

to the difficulty of characterizing them by conventional spectroscopic

_ techniques. Very recently, the crystal structure was reported

261 of

Mn(2Meim)3Ct2, which comprises distorted trigonal-bipyramidal molecules

(Fig. 6.1, Table 6.1) linked by hydrogen bonds into a three-dimensional

network. As some other complexes with the same ligand[Ni(2Mcim)4X2

259 (X=d1, Br) and, arguably, the B-form of Co(2Meim)

4(NO

3)2 2591 are

known to be 5-coordinate, and since several recently reported instances

of 5-coordination in manganese(II)180,262-264

show this to be of more

frequent occurrence (even in the absence of ligands specifically

tailored265,266

to induce pentacoordination) than was previously suspected,

a study of 2Meim complexes of other manganese(II) salts was undertaken.

The complexes prepared here are Mn(2Meim)3X2 (X=CE,Br), Mn(2Meim)4X2

(X=Br, I) and Mn(2Meim)4(NO3)2.H2O.

As reported earlier,261

interaction of Mna2.4H20 and 2Meim in

ethanol yields Mn(2Meim)3a2 irrespective of the ratio of starting

materials. Mn(2-Meim)3Br2' on the other hand, is difficult to obtain

pure; most preparations employing ligand-metal ratios of 3:1 or less

give solids of variable composition which are identified from IR and

esr spectra as mixtures of the 3:1 and 4:1 complexes. A 6:1 mole ratio

of reactants is required to suppress completely the formation of

Mn(2Meim)3Br2. No MnI2 complex of stoicheiometry other than 4:1 could

182

be prepared, either directly or by vacuum pyrolysis of Mn(2Mcim)4I2.

In view of the struotural insight afforded. by the esr r3sults

on Mn(2Meim)3X2, it appeared worthwhile to initiate an investigation into

the esr of other Mn(II)-doped 5-coordinate molecules of donor-sets N3X2 and

N2X3, and data are presented for M(Nn)(Mesdien)Ct) (M=Mn,Zn,Co) and

[

Ni (Mn) (dab) 2 X3IX, X-C,E,Br. It should be mentioned here that electronic

spectra of 5-coordinate nickel(II) and cobalt(II) complexes, although

usually distinct from those of 4- and 6-coordinate species, do not

in general give unambiguous indication of stereochemistry, the energy

level diagrams of the two ideal configurations being very similar; 248,

for the same reason magnetic susceptibility measurements on polycrystalline

samples are of limited diagnostic value. It is of more than usual

interest, therefore, to examine whether the more sensitive technique of

esr assists in deducing, even approximately, the molecular geometry in

such cases.

IR Spectra

(a) 4000 - 400 cm 1

Ligand vibrations in the mid-IR region are for the most part

insensitive to complex stereochemistry; however a broad band of moderate

intensity at 1675 cm-1

in the free imidazole is shifted to 1685 cm1

in the 4:1 complexes, and to 1625 cm-1 in the 3:1 complexes. A complete

vibrational analysis of 2Meim is not yet available, but bands in this

region in heterocyclic aromatics are associated with ring breathing

vibrations (see Chapter 4). The C-H o.o.p. deformations in such molecules

are expected between 700 and 900 cm-1

; the free ligand shows a strong

doublet (740, 755 cm-1

) which also appears at slightly lower frequency

(732, 741 cm1), but with little modification in appearance, in both

the 4:1 complexes. The 3:1 complexes have at least three bands here

183

and, in addition, show a heavily structured band between 860 and 880 cm;

the corresponding absorption in the 4:1 complex occurs at 842 cm-1

and

is strong and exceedingly sharp. (The ligand itself is transparent

between 800 and 900 cm-1.) In both the 750 and 870 cm

1regions the

pattern of the absorption in the 3:1 complexes differs in detail from

the chloride to the bromide.

(b) --":400 cm-1

As has been reported previously,259,269

the richness of the

ligand spectrum in the low-energy region seriously hampers identification

and assignment of metal-ligand stretching vibrations in the complexes.

In the present study, IR spectra /were run to 200 cm-1

only. The results

appear in Table 6.2.

(i) 4:1 complexes

The predicted vibrations in D41 symmetry are

V(M-N): Alg(R)lg (P.) + E (IR)

V(M-X): Alg(R) + A

2u(IR)

For both the 4:1 halides, the IR-active Mn-X stretch is expected below

200 cm-1.226 A strong band near 220 cm

-1 17-22e

in the IR, which appears at 230 cm-1 in the 3:1 complexes, is clearly

associated with a metal-nitrogen stretching vibration [cf.V(Mn-N) at

210-235 cm-1

in tetragonal MnL42 (L = subsituted pyridine);

226

2591 V (Ni-N) at 243 cm

-1 in square-pyramidal Ni(2Meim)4X2 (X=CZ,Br) j .

In an earlier study,226 this has been assigned as the Alg

mode (which

is formally IR-inactive) but the 2Meim complexes show no complementary

. absorption in the Raman; the weak shoulder at 228 cm

-1 is present in

all four complexes and is probably not a skeletal vibration.

The Raman spectrum of the bromide is distinguished from that of

the iodide by the presence of strong absorption at 91 cm-1

and 131 cm-1

,

184

but even the higher energy band lies outside the usual limits

(145 - 200 cm '1) of v(Mn-Dr) in tetragonal complexes of this stoicheiometry;

it is noted that in the complexes Mn(3Me-py)4X2(X- 226

Br,I) as well as

in Mn(imidazole)4Ct2226 no metal-halide stretches could be detected in

the Raman.

(ii) 3:1 complexes

The IR absorption at 230 cm 1 , assigned as V(Mn-N), is comparable

with the value of 245 cm1 for the corresponding vibration in

, 2,270 Zn(terpy) the Raman band near 180 cm

-1 in both the 2Meim complexes Ct

may be of similar origin. NA1(Zn-N) in tetrahedral Zn(2Meim)4(NO3)2

has been found at 213 cm-1259

)

. A strong IR band at 216 cm

1 in the chloride is most likely

v(Mn-Ct), but it is not clear which of the two Mn-Ct bonds is involved;

for this one complex IR measurements were extended below 200 cm-1

but

no band was observed, in a rather poorly defined spectrum; which could

be ascribed unequivocally to a metal-ligand stretch. The low energy

of the 216 cm-1 vibration (as compared with, e.g., 287 cm -1 and 278 cm

-1

for v(M-Ct) in Zn(terpy)Ct2,270

and 300 and 280 cm-1

in Mn(paphy)Ct2*271

is entirely consistent with the unusually long Mn-Ct bond lengths in

the 2Meim complex.

Esr spectra

In previous chapters, we have dealt with spectra of Mn(II) in

diamagnetic lattices in which the concentration of the paramagnet did

not exceed 10%. In most cases the linewidths were sufficiently narrow

for unambiguous identification of individual transitions, which greatly

facilitated assignment of SPH parameters. However where a suitable

* paphy = pyridine-2-aldehyde 2'-pyridylhydrazone

185

matrix for isomorphous dilution is nor available, or the complex, in

question does not dissolve without decomposition in non-polar solvents,

it may be necessary to examine the magnetically concentrated material,

for which esr linewidths may be as large as 200 mT.

Previous attempts to interpret such spectra have relied for the

most part on fitting the observed absorptions to the D-B plots of Dowsing

and Gibson,62

but the inadequacies of this method, even where the zfs

parameters are required only to a fair degree of accuracy, are now

recognized. The omission from the plots of lines representing single-

crystal off-axis turning points has already been pointed out. Even for

the x,y,z transitions, however, estimation of resonance fields from

first-derivative spectra becomes very difficult when the separation

between adjacent singularities in the powder pattern is comparable with

the linewidth, as is well illustrated by the data of Table 6.3(v.i.).

In addition, D-B plots give no indication of the intensities of the lines

in the spherically averaged spectrum; only the transition probability

at the turning-point is known, and this is less important, generally,

than the orientation of the magnetic field and the rate of bending at

the turning-point.50

In this work, extraction of the SPH parameters of the 2Meim

complexes was achieved solely by comparison of the experimental spectra

with lineshapes calculated for a wide range of values of D and X. The

limitations of this calculation have been described in Chapter 1; in

general, however, the results of spectral simulation, employing SHAPES,

for complexes in which the SPH parameters have been determined

independently, e.g. by examination of the magnetically dilute soecies

(as in Chapters 2-4), are encouraging, and appear to justify the use of .

this method as the sole means of analysis when none other is available.

Further evidence for the validity of the calculation is presented in

186

Fig. 6.2. The SP N parameters of Mn(Ph,D3' 20) Ct 2 in the

isomorphous

zinc(II) complex have been obtained by single-crystal work;38

random-

orientation spectra calculated therefrom are seen to reproduce accurately

all the important features in the experimental spectrum of pure

Mn(Ph3P0)2Ct2 (Fig. 6.2). The figure shows, incidentally, that at the large

linewidths in the magnetically concentrated complex, the spectrum

calculated for the optimum parameters gx = 2.03, g = gz = 2.02, ,

a = 0.000 6 cm-1 38

is virtually indistinguishable from that for

giso = 2.0, a = 0.

For each set of SPH parameters X-band spectra were simulated,

in most cases, over two ranges of magnetic field (0-1000 mT and

500 - 1500 mT); this was found convenient not only for ease of comparison

with experimental spectra (the maximum range available on our spectrometer

is 1000 mT) but also because, as pointed out earlier, crystal imperfections

may result in linewidths being considerably greater at the higher fields.1

226 It has been found previously that for D>-0.2 cm

-1, the position of

the most intense absorption may yield a rough estimate of X. This is

confirmed on examination of the series of 9.51GHz spectra of Fig. 6.3

all of which were calculated for D = 0.30 cm-1. At the chosen linewidth

of 90 mT the band in question lies near 140 mT in axial symmetry, and

moves to higher field with increasing X, the apparent peak-to-peak

distance increasing simultaneously. Both the position and the width

of the band reach a maximum around X- 0.15 and then decline until at

X - 0.30 the characteristic sharp geff = 4.3 signal (Bre = 160 mT) is

obtained.

The X-band spectrum of Mn(2-Meim)3Ce2 (Figs. 6.4, 6.5) shows

absorption right up to the highest accessible field (1350 mT) which

suggests that D>-0.2 cm The position of the strongest band in the

187

spectrum, which lies close to g f, = 4, may then be taken indicate indlate

that ?:0.23. From lineshapes calculated using 27 field directions

-1 , per octant it was found that the parameters D = 0.25 cm , A = 0.22

fitted the experimental spectrum reasonably well; on repeating the

calculations with 93 field directions a slightly higher D (0.28 cm-1)

was required. The main bands in the experimental spectrum and their

assignments are given in Table 6.3. Dipolar broadening and the effects

of differentiation combine to make prominent features of the spectrum

complex functions of both the Hamiltonian parameters and the broadening

linewidths,161,272,273

and it is not surprising that the calculated

fields in the second column of the table do not generally coincide with

the peak positions and baseline crossings in the experimental spectrum.

While most of the bands arise from the normal 'along-axis' transitions,

the moderately intense absorption at 964 mT and the shoulder to high

field are associated with transitions occurring for the magnetic field

in the yz plane. The simulated lineshape for the same parameters at

Q-band (Fig. 6.6) is in satisfying agreement with experiment and Table

6.4 indicates that even at this frequency, the contribution of off-axis

transitions is not inconsiderable.

Attempts to study a dilute frozen solution of the complex were

thwarted by its insolubility in non-polar solvents and by its apparent

decomposition in ethanol, and even in dichloromethane (or benzene)

containing the minimum alcohol for solution. These latter solutions

gave, at 80 K, only a single, strong, partially split sextet at geff = 2.

The complex Mn(2-Meim)3Br2 gives a reproducible spectrum

(Tables 6.5 and 6.6, Figs. 6.7 - 10) which is distinct from that of

an equimolar mixture of Mn(2Meim)4Br2 and Mn(1,2Dimeim)2Br2. The

position of the most intense absorption at X-band frequency suggests

that 0 <X 0.10. Preliminary attempts to reproduce the lineshape,

188

using 27 field directjons per octant, were uncuccessful untilit was

realized that, for the particular magnitude of the zfs in this molecule,

increasing the number of fields to 93 per octant actually leads to

reversal of the relative intensities of the two high-field maxima

(780 and 1060 mT) in the calculated. spectra (see Fig. 6.8). This observation

naturally casts doubt on the reliability of even the more accurate

simulation, but in fact spectra calculated for D = 0.41 cm1,X = 0.05

agree moderately well with the experimental results at both microwave

frequencies (Figs. 6.7 - 10). There are some discrepancies at Q-band

notably in regard to the relative intensities of the three band maxima

between 100 and 1100 mT, the presence of the 'double-hump' between

310 and 410 mT in the experimental spectrum, and the shapes of the

750 and 1460 mT maxima. These last features at least are quite sensitive

to small departures of 1 from 0.05, but no definite improvement in fit

could be effected by such adjustments and, moreover, the above parameters

give undoubtedly the best fit to the X-band data.

The strongest absorption in the X-band spectrum of

Mn(2Meim)4(NO3)2.H20 (Fig. 6.11) crosses the baseline close to 200 mT.

Evidently D<-0.2 cm-1 and / or I approaches the 'rhombic' limit;

the absence of absorption beyond -1120 mT confirms the first possibility.

The positions of the minima at 620 and 830 mT and their intensities

relative to the main band are fairly accurately reproduced in lineshapes

calculated for D = 0.18 - 0.20 cm], 0.03<X 0.09; increasing A beyond

-0.10 results, as expected, in attenuated peak heights at high field.

Although the high field lines are rather insensitive to A within the

above limits, the observed splitting of the main band indicates that

A A 0.08.

In both the 4:1 bromide (Fig. 6.12, 13) and iodide complexes

the relatively narrow aeff

= 6 line stands in contrast to the breadth -

139

of the higher field absorptions, and marks these molecules as axially

symmetric. Whereas the axial zfs is easily determined for the bromide,

the high field lines in the iodide spectrum are broadened beyond

detection at both microwave frequencies, and no estimate of D is

possible.

Spectra of the Me5dien and dab complexes (Figs. 6.14-19, Table 6.7)

have been discussed in Chapter 5.

Electronic spectra of the complexes are so weak and poorly

resolved as to be totally uninformative. That of Mn(2-Meim)3C2

2 shows

marginally better definition, with a relatively sharp maximum at

23.3 kK and a broad band ca. 26.3 kK; these are tentatively assigned

as 6A1 -*

4A1(4E) and

4T2(

4D).

226

Discussion

Both IR and esr results point to the 3:1 complexes being structurally

distinct from the 4:1 complexes and from each other (they are not

X-ray isomorphous). Several possible structures may be envisaged for

this stoicheiometry, but even prior to the solution of the crystal

structure of the chloro complex, the spectroscopic results gave clear

indication of the existence of discrete molecular units, MnN3C-C2, of

rather low symmetry. Thus

(1) oligomeric or polymeric structures involving single, double or

triple halide briding would be expected to give broad, symmetric

geff

= 2 signals of the type observed in octahedral MnL2C-C.2

(L = substituted py, quinoline, imidazole);226

(2) the axial zfs in Mn(2Meim) Ce. is considerably higher than has

been observed in pseudotetrahedral MnL2C- 2. [Typical values of D

in this stereochemistry are 0.15 (± 0.02)cm-1

for Zn(Mn)(2Meim)2C, ,c2274

- and 0.17 cm

-1 in Zn(Mr)(Ph3PO)

2C-e2'

8j

In addition V(Mn-Ct) occurs

some 100 cm-1 lower than is generally observed in tetrahedral

coordination,226 and the possibility of a trigonal pyramidal

i+ cation [Mn(2Meim)

3C.t1 may also be excluded on this evidence.[ A

chromophore of C3v

symmetry should also give a 1 value close to zero,

as has been found in Zn(Mn)(dab)CZ3 (Table 6.7).]

(3) The apparently high rhombic distortion argues against a regular

trigonal bipyramidal (D3h) arrangement of the ligands.

In Mn(2Meim)3Br2, the presence of weaker tetragonal perturbation

than in the 4:1 complex, taken together with the apparently low in-

plane anisotropy. (compared with that in Mn(2Meim)3Ct.)is compatible

with a regular trigonal bipyramidal arrangement of ligands or with

+ the presence of the pseudotetrahedral cation [Mn(2Meim)3Brj . The absence

of a Mn-Br IR stretching frequency above 200 cm-1 favours the first

possibility.

The distortion parameters of Mn(2Meim)4Br2

are very similar

to those in analogous complexes with alkyl substituted pyridines as

the equatorial ligands;226 unlike square-pyramidal ENi(2Meim)4

Brj Br,259

therefore, the manganese complex appears to have a regular, centrosymmetric

structure.

In the case of the nitrate complex the esr parameters are

uninformative owing to a paucity of data on related systems. However, as

1 2+ D greatly exceeds that in trans- ECo(Mn)(3Et-py) 4 (H2 4

0),J , 226

and is

also higher than in any of the mixed-ligand (MN204 ) chelates examined

in Chapter 4, coordination of one or more oxygen atoms is indidated;

259 thus the complex differs from its tetrahedral (unsolvated) zinc analogue.

Comparison of the X-ray powder pattern with those of the A and B forms

of Co (2Meim)4 (NO3 )2

259 may prove instructive, as would a study of the

191

esr of manganese(IT) impurity in one or both of those complexes.

2 The geometry of Co(Ma

5dien)Gc.

, 2' 75 like that of Mn(2Meim)

3CZ

2r261

may be viewed es a heavily distorted trigonal bipyramid with one chloride lig-

and each in axial and equatorial positions, and with axial M-N and

M-Ct. bond-lengths exceeding equatorial. Corresponding band angles in

the two molecules are seen from Table 6.1 to be very similar. It is

Most interesting, therefore, to find that A approaches zero in

M(Mn)(Me5dien)Ct2. The identity of the host complex is not in doubt;

elemental analyses for the zinc(II) and cobalt(II) complexes and the

electronic spectrum of the latter are in excellent agreement with those

reported earlier.266

The IR spectra have not been recorded previously;

however, the presence in the manganese complex of strong, broad absorption

centred on 255 and 280 cm-1

(the latter value being intermediate between

those commonly observed for V(Mn-Ct) in octahedral milN4a2 and tetrahedral

275 MnN22

) is quite consistent with the structure revealed by X-ray analysis.

The possibility that the Mn(II) ion may not be isomorphously substituted

in the host is definitely excluded on the ground of similarity of

D and A in the doped complexes to the values found in the pure manganese

complex. There is no suggestion whatever, either in the literature or

from present observation, that the complex might be polymorphic; on

the contrary it would appear that the particular steric requirements

of the ligand are strongly conducive to formation of 5-coordinate complexes

of the desired geometry. Thus Me5dien reacts with several divalent

chlorides of first-row transition metals to give the isomorphous series

M(Me5dien)Ct2

(M=Mn, Fe, Co, Zn);266

although the nickel and copper

analogues give a different X-ray powder pattern, there is spectroscopic

evidence for 5-coordination in these complexes as well.266 This contrasts

with the behaviour of the related ligand Et4dien, for which only cobalt(II)

and nickel(II) complexes have been reported;276

the first is 5-coordinate

277

192

and its gemrtry resembles that of Co(Me5dien)Ct2 (. a detailed

structural npru of the molecules is given in. Ref. 275) but the

low-spin nickel compound is square in the solid state.276

It is of

obvious interest to obtain esr spectra of Mn(II) in Co(Et4dien)Ct2,

but attempts to prepare the doped material proved unsuccessful. A

complex of stoicheiometry Mn(Et4dien)Ctw which is not X-ray isomorphous

with the cobalt complex, gives the broad geff = 2 esr signal indicative

of strong interaction between the magnetic centres. The far-IR spectrum

shows, in addition to bands in roughly the same positions as reported

for Mn(Me dien)Ct2' strong absorption at 310 cm

-1.

The observation of an 'axial' spectrum for a molecule lacking

a symmetry axis is unusual although not unprecedented. A A value of

0.0087 has been reported278

for NiCt2.4H20, in which the chloride ions

occupy cis positions in a severely distorted octahedron, NiCt204, lacking

even a 2-fold axis; the molecular--quantization axes found from

susceptibility data bear no obvious relation to the components of the

molecular unit.

A is close to 0.3 in Zn(terpy)Ct2 which has been described as

a very distorted trigonal bipyramid of approximately C2 symmetry,279

both halide ions occupying equatorial sites; the molecule is so far

removed from either of the idealized geometries, however, that a

description in terms of a square pyramid is at least equally valid.280

Unfortunately the spectrum is too poorly resolved to allow more precise

determination of the parameters. The magnitude of D in [Ni(dab)2X

3".])C

(X=a, Br), believed251

to possess D3h

symmetry, might be expected

to exceed that in tetragonal MN4X2'

and this appears to be the case in

the chloro complex, which possesses the highest axial zfs of any

mixed N-Ct chromophore examined in this work, but the situation in the

bromide is less clear. Any detailed rationalization of the esr data

must await the result of structural studies.

193


Mn(2fleim),Ct, was obtT:ined as well-defined white crystals on

mixing ethanol solutions of the ligand and MnCt2.4H20. This was

invariably the sole product of several preparations employing ligand-

metal ratios ranging between 2:1 and 6:1.

Mn(2Meim)3Br2. 2Meim in DMP containing the minimum ethanol

for solution was added to MnBr2.4H20 in the same solvent, the metal-ligand

ratio being 1:2. Mn(2Meim) Br2 was obtained when the MnBr2.4H20 solution

was added. to the ligand with the reactants in the ratio 1:6.

Mn(2Meim)_412. A DMA'/ethanol solution of Mn12.4H20 was refluxed

briefly and added to the stoicheiometric quantity of ligand in the

same solvent. The solution yielded pale green crystals on refrigeration.

Mn(2Meim).4(NO3)2.H20 was precipitated as a finely divided

white solid when deoxygenated ether was added to an ethanol solution

containing Mn(NO3)2.4H20 and 2Meim in the ratio 1:6.

Mn(Me5dien)Ct2 and the manganese-doped zinc(II) and cobalt(II)

complexes were prepared by the method of ref. 266.

The complexes Zn(Mn)(dab)Ct3' {Ni (Mn) (dab) 2 X3

ix (x=a,Br). and

Zn(Mn)(tExTy)Ct2 were prepared as in refs. 249, 251 and 281 by G. Brun,

who also obtained the esr spectra.

194

C 15

C2sri3O

2 I

',,,,i' '■:1,

C 2 6 (J)

NJ] irr

7,,rt 3

Clok

Ci 2

Fig. 6.1 The molecular structure of Nn(2Meim)3Ct2

(from ref. 261)

Tahle 6.1 Interatonic distances ces (A) and. angles ( ) in

Mn(2Meim)3Ct2 and Co (Me dien)Cf

Mn(2Meim)3Ct2

M - Ct(1) 2.53

M CE(2) 2.39

M - N(11) 2.25

M - N(21) 2.19

M - N(31) 2.20

Ce.(1)-M - N(11) 168

N(21) 119

C(3)-M - N(31) 110

N(21)-M - N(31) 131

Co(Me5dien)Ct2

2.33

2.29

2.27

2.11

2.13

162

112

108

135

aData from refs. 261 and 275. The numbering scheme for the atoms

is as in Fig. 6.1

Table 6.2 Low-energy vibrational spectra (cm-1) of the 2Meim complexes

Mn(2Meim)3CZ2

Mn(2Meim)3Br2

Mn(2Meim).4Br2

Mn(2Meim)4I2

a Ligand bands.

IR: 376 sa, 286 sh, 277 sh, 271 sh, 271 sa, 230 m, 216 s.

R: 387 ma, 275 m

a, 228 sh,w, 184 s, 90 s.

IR: 378 sa, 270 s,br, 230 ms.

R: 386 msa, 276 msa, 228 sh, w, 180 ms, 150 msa, 70 vs.

IR: 378 sa, 286 sh, 277 sh, 270 s

a, 220 s.

R: 381 msa, 276 sa, 228 sh, 165 ma, 131 m, 91 s, 72 m.

IR: 377 sa, 280 sh, 270 sa, 217 s,br.

R: 383 ma, 276 ma, 228 sh,w, 165 sa, 76 s.

196

Table 6.3 X-band spectrum (mT) of Ma(2Meilm) Ct. (v=0.61 C=Hz)

Turning-points of ex- Calculated for D-0.28cm-1,

. _ . •

A =0.22

perimental spectrum B T.P. Field direction Levels

133 (max.) 139.9 3.85 z 4-3

184.4 5.09 x 4-3

223 (min.)

320 (max.)

339.2 8.01 x 6-5

365 (min.)

-450 (max.)

608.1 4.82 y 4-3

635.7 7.64 x 2-1

650 (min.)

706.5 1.14 0=14°, y5=90° 4-3

717.8 2.96 z 4-3

857 (max.)

926.1 4.65 0=39°, q5=90° 4-3 964 (min.)

992.7 5.12 0=71°, 4)=90° 3-2

-1100 1137.1 5.01 0=14°, 4)=90° 5-4

(saddle pt.)

1220 (max.)

1311 (min.)

1333.1 5.85 Y 2-1

197

Table 6.4 0-band Fn-ctl-um (mT) of Mn(2Meim),CP n (v=35.91(11z) -

Turning-points of ex-

perimental spectrum

-1 Calculated for D=0.28cm , )',=0.22


114 (max.) 138.3 4.98 z 3-1

148.1 1.40 x 4-1

239.5 1.89 y 6-4 283 (min.)

300.9 1.33 z 5-2

411.1 5.13 Y 6-5 426 (max.)

543.0 1.87 0=49o, (1)=90

o 5-2

617.9 3.19 0=36°, (1)=0 4-2 626 (min.)

710.9 2.75 y 4-2

. 717.4 7.82 z 3-2

813 (max.) 818.4 7.87 NT , 5-4

954.9 2.34 z 5-3

1063.6 7.70 x 5-4

1069 (max.) 1080.6 8.21 x 4-3

1183.2 5.54 x 6-5

1212.8 7.53 x 3-2

1246.4 8.49 y 4-3

1283.2 8.71 z 4-3

198

Table 6.5 X-band spectrum (mT) 14n (2Maim) ,Br (N0-9.51 Cliz) 2- —

Turning-points of ex-

perimental spectrum

-1 Calculo:ted for D=0.41cm , X 0.05


88 (max.) 97.4. 3.67 Y 65

148.4 6.84 x 6-5

189 (min.)

303 (max.)

355.3 4.68 x 4-3

388.2 9.34 z 6-5

422 (min.)

490.9 8.83 z 6-5

. 536.8 4.76 y 4-3

785 (max.)

953 (min.)

1057 (max.)

1108.2 7.60 x 2-1

1239 (min.)

1241.8 7.89 z 6-5

1344.3 7.02 Y 2-1

1394.1 5.61 z 6-5

1 9 '13

Tablr,. 6 .6 (..)-J";j1.11t1 (ml) of Mr-1 ( 2-11c:bp) (v =35 . 72CH %

'Turning-points of ex-

perimental spectrum

-1 Calculated for D=0.41cm , X.,0.05

13 T.P. Field direction Levels

171.2 1.23 x 6-3

192 (min.) 181.4 2.23 Y 6-4

318 (max.) 351.1 6.42 z 5-3

368 (min.)

413.0 3.46 z 4-2 422 (max.) .

463.8 5.41 z 4-2

524.5 3.93 z 5-3

539.8 5.59 y 6-5

579 (min.) 629.2 3.46 x 3-2

711.9 3.21 y 3-2 723 (max.)

769.3 5.83 x 6-5

801.2 7.74 y 5-4

843.1 7.64 x 5-4 894 (min.)

1025 (max.)

1074.8 7.52 x 4-3

1174.6 7.56 y 4-3

1183.5 6.23 x 3-2 1202 (min.)

1242.1 7.55 z 5-3

1329.2 5.52 z 4-3

1460 (max.) 1392.3 6.51 y 3-2

1784 (min.) 2091.2 5.60 x 2-1

2150.4 7.99 z 5-4

200

Table 6.7 X-barld spectrum (mT of 7a (r (dab)CO3

Observed -1 Calculated for D=0.19cm , X=0.01

-9.17501z) B T.P. Field direction ......_

Levels

0-70 sa 38.8 2.8 0=40°, (---90° 5-4

43.5 1.1 0=20°, 1)=0 5-4

107.5 s 106.9 5.0 y 6-5

152.0 s 154.8 6.46 x 6-5

291.0 m 292.4 5.74 x 4-3

347.0 m 347.8 5.80 y 4-3

404.0 mb

broad,medium 577.1 4.6 0=30°, (1)=0 4-3

intensity

absorption 613.1 4.8 0=30°, q5=90° 4-3

500.0-640.0

711.5 w 713.8 6.27 x 2-1

794.0 w 801.1 6.03 y 2-1

aResolved but irregular hf structure

bPoorly defined hf structure

Table '?.8 Zfs narameters in the complex ..s

Complex D -1 cm A

Mn (21•Ieim) 0.28 0.22

Mn (2Meim) i _Br2 0.41 0.05

Mn (2Meim) 4Br7 0.63 0.0

Mn (2Meim) 412 a 0.0

Mn (2Meim) 4 (NO3) 2.H20 0.18 0.09

M (Mn) (Me5dien) a2 0.24 0.01

(M=Co, Zn)

Mn (Me5dien) a2 -0.23 0.05 ± 0.02

[Ni (Mn) (dab) 2Ct..3] a 0.31 0.03

[Ni (Mn) (dab) 213171 Br 0.60 0

Zn (Mn) (dab) a, 0.19 0.04

Zn (Mn) (terpy)Ct2 a ^-0.3

201

aIndeterminate


Compound

%C

Found

%H %N %C

Calculated

%H %N

Mn(2Meim)3CZ2 38.7 5.2 22.5 38.7 4.9 22.6

Mn(2Meim)3Br2 31.3 4.2 18.8 31.3 3.9 1 8.2

Mn(2Meim)4Br2 35.6 4.7, 21.2 35.4 4.5 20.6

Mn(2Meim)4I2 30.4 3.9 17.3 30.2 3.8 17.6

Mn(2Meim)4 (NO3 2 ),.H2 0 36.8 4.7 27.0 36.6 5.0 26.7

Mn(Me5 dien)Ct.2 36.3 7.7 14.5 36.2 7.8 14.1

Co(Me5dien)a2 36.2 7.6 13.7 35.7 7.7 13.9

Zn(Me5 dien)a

2 35.4 7.4 13.7 34.9 7.5 13.6

Mn(Etdien)CZ, 42.8 8.6 12.1 42.2 8.6 12.3 4 4

203

Cfl IL! • 57 114.50 1.4.74 197.17 64 Ti .C4 /11.46 42 1L91

LO 41 . 947-75-77 Al 617.14 C 96• 64 190 CI 1./1 .6 t int7 .17 SSTS.; --/6.6-7-191■77-971 mT

Fig. 6.2 Esr spectra (9.5330Hz, 0-l000mT) of Mn(Ph3P0)2Ct2:

(i) experimental; (ii) calculated for g =2.03,

- -1

g =2.02, a=0.0006cm-1 , D=0.172cm , -y z

A =70mT; (iii) calculated for q. =2.00, a=0;

D,A and. AHpp as above.

204

".--11=-114.37 --i14.14 :41.17 l':17.41 34.34 411.4D L4.43 lb:., 2.1.0 :t1.L1 3;044

B/mT

B/mT

Fig. 6.3 Esr spectra (9.51GHz, 0-450mT) calculated for

D=0.30cm-1, LC' =90mT: (i) A=0.05, (ii) 1=0.10 PP

(iii) 1=0.15, (iv) 1=0.20, (v) 1=0.25, (vi) 1=0.30.

1

/

11 1

1 1

\ ,.. / , ■ ...- 1 \ -, 4..

/ ....

1 / ■

•

61.43 1'22.67 154.33 235.74 2'67417 36341 352.34 41.43 462.31 614.35 656.73 517.22 663.56 72.7.27 *1.62 642.53 04.3: C.5.35 677.211 11;_2.73 FIELD

Fig. 6.4 Esr spectra (9.51GHz, O-l000mT) of Mn(2Meim) a • 3 2'

D=0.28cm 1, A=0.22, AHTDD=90mT.

experimental, ----, calculated for

CC247 544.90 75.74 757.17 Er;.5.51 58C44 911 •45 SSZFILLD 1314.35 11:55.79 1117.'221'155.65 1.220.0.7 171.52 1322.93 1374.n 1C.:543 1477 •ZS 14.75.71 M1

Fig. 6.5 Esr spectra- (9.51GHz, 500-1350mT) of Mil(2Meim)3a2: ex-oerimental, , calculated

for D=0.28cm-1, X=0.22, AHpry=170mT.

+. .- - --- -- ... .. ...... ., / 4\ /

/ \

...,. / / /

\ I' /

..., / s, /.......

.......- ..". \ / \ / ss ...,

........

1'4.ES 217.71 27G.E0 2:41.42 204.27 43.13 61140 570.04 C2S.5S GL2.45 747.40 00,5.20 $i.11 S23.87 Eb2.32 11CC.Z3 1.11.24 1077.12 M EIELO

Fig. 6.6 Esr spectra (35.91GHz, 100-1277mT) of Mn(2Meim)3a2: ----, experimental; A=0.22, AH =170mT. PP

N.) 0 , calculated for D=0.28cm-1

e.43 2b7.17 =244 344 41;.43 4C2.91 ;14.;5 E:;;;.:37.22 771.CZ CZ:4; &74.2:: C:4.03 07.::; nqi

Fli.LD

Fig. 6.7 Esr spectra (9.51GHz, 0-1000mT) of Mn(24eim)3Br2: ----, experimental, , calculated for

D=0.41cm-1

, X=0.05, AHpp=90mT. CD

\ ./.,

\ - ,

\ -,

. . -... ---' -... ..-- —.—...... ---•

551.43 6:LL: • 37 64 • 90 'iC11 44. 7:71.-1-7- zo-or 8.1 I -42 LE-05/4- 111-7-42. I LEZ • 611 : 1L Z3 •ES IL:71 •S2 22 • ES 1?•74 411 1 ' 14 77 • ZS IL,"_'?,•7? M I

Fig. 6.8 Esr spectra (9.51GHz, 500-1350mT) of Mn(2Meim)38r2: , experimental; calculated for

D=0.41cm1, X=0.05, Alipp=180mT, using 27 fields per octant; , calculated for the above

parameters, using 93 fields per octant.

1E0 .C3 2.2.3.03 9C7.C3 37a.c3 44540 ::14.C3 G33.00 C'S2.:0 72Z.Ca 7S3.C3 Ch.= £23.C3 CO.= LEV.7.:4 6:;540 1204.00 :273.Z0 1'442.C3 mT FIELD

Fig. 6.9 Esr spectra (35.72GHz, 100-1100mT) of Mn(2Meim)3Br2: ----, experimental, , calculated for

D=0.41cm 1, X=0.05, AHPP =150mT.

-o mT

1:07.CO 117.1.= ::24S.GO 1'314.= 7::.3:7;.CC 14s2.c3 I2t.CC I.C3 16S5-00 .1.723.:61 11:-.1.0C I2C.I 24.0 FIELD

Fig. 6.10 Esr spectra (35.72GHz, 90-1900mT) of Mn(2Meim)3Br2: experimental; -, calculated

for D=0.41cm, X=0.05, AH P =150mT. P

212

C3 62.43 7-2 037-717.33 2,5.1.1 13741 232.61 252.04 411.46 452.3 FIE

1 LD C14.35 555.13 C17.7: -512.25 7 P.C2 111.52 6:'2.53 0 4.39 mT

Fig. 6.11 Esr spectra (9.51GHz, 0-900mT) of 4n(2Meim)4(NO,)2. H2O:

-1 (i) experimental, (ii) calculated for D=0.18cm , X=0.09,

L.H.1)=.60mT.

• / ■ / -------- •-

1 1

1

I I

I

I

5:•43 " .2:S.14 257.17 SSS.S1 Z;S:144 431.4.1 4SZ.SI e14..SS St-4.73 Cl7.22 7?1•SZ S22 - SS' mT

1E1.0

Fig. 6.12 Esr spectra (9.51GHz, 0-1000mT)of 4n(2Meim)4Br2: ----, experimental; , calculated for

D=0.63cm-1 , A=0, AHpp=90mT.

1.43 EC2.C? ?CZ...74 ?`"7 17 eoo 04 '11 42 '" 1.11444 ICC11•7.1 1:117.21: 11:SC•Cs 1'220.C: 1271.L2 1:171.•12 1:::?•CC 11:::;1•711 M T 1-'1E

Fig. 6.13 Esr spectra (9.51GHz, 500-1350mT) of Mn(2Meim)4131-7. , experimental, , calculated

for D=0.63cm 1, A=0' AHP2 =200mT.

61.43 162-27 164.3:1 2C5•74 323.61 310.04 411.48 462.9: • 71445 065.78 2..7•.:2 663.E5' 722.03 7171.52 en.ss 674.3Z. S:".:1•63 977.2'5 1523.73 FIELD T

Fig. 6.14 Esr spectra (9.51GHz, 0-1000mT) of Mn(Me5dien)Ct2: ----, experimental; , calculated for

D=0.23cm 1, X=0.05, AHPP =90mT. ts,) H L,T1

I

1

/ 1 1

/.', .......'' 1

/ \

k / 1 / \

.........s....,

\ / ./ .....

\ / \

\

\ / \ ss / \ .., -.-. .... ■ ....

/ '....

51-43 1:Z.G7 1C4..Z3 VZ5.74 20.17 3C8.S1 2C,C.C4 411.48 462.S1 Sia.35 GCS-78 e17.22 6C0.55 1Z.:-Cg 771.22 5:.2.25 13‘74.Z2 21".-27 677.25 CZ.7.2-70 M I FIELD

Fig. 6.15 Esr spectra (9.51GHz, 0-1000mT) of Co(Mn)(Mesdien)a2: experimental, ----, calculated

1 for D=0.24cm , A=0.01, AHpr=40mT.

217

(i) ti

.00 1).17 4400 01.00 71).01 0,00 11740 0..0.11 147.77 11)4.CC 11.00 eT • FICLD

Fig. 6.16 Esr spectra (35.720Hz, 100-1900mT) of Co(Mn)(Me5dien)a2:

(i) experimental; (ii) calculated for D=0.24,X=0.01,

AH =40mT. PP

j

CO IL LS W2ar, 676:4 61,V7 5btf.t .. ainzz aha.ao m. ", nir FIELD

Fig. 6.17 Esr spectra (9.175GHz, 0-1000mT) of Zn(Mn).(dab)Ct3:

(i) experimental, (ii) calculated for D=0.19cm,

A=0.04cm 1, AHPP -40mT.

216

A It

/ ,

1 1 i I 1

1 1

\ i I I

....-?./ \

-%.

\ ./ -... / .... -.. '''...- -

\ \ \ / \

..- ..-- ....-...

1 ......... / \ -..

\/ \

\ / \ /

,.....,

•-ssf

I /\ / I / `-` \ I

■

■

1b2.s7 11;429 2C.:-.74 2.17.,7 293.61 3ca.o4 411.49 4.7.5 E.Li.73 617.22 esc•cz. .7zO.C9 •27:. sz:1-1,".; rriT

Fig. 6.18 Esr spectra (9.50GHz, 0-1000mT) of Ni(Mn)(dab)20qa: ----, experimental, ----, calculated

for D=0.31cm , A=0.03, AHpp=40mT.

51.43 12.87 154.5C 2C5•74 257.17 3,78.61 280.24 4/1.43 462F._101.:Lo 614.25 565.” 517.22 655.55 7'2z•ca 771.52 822.51 5174.39 522.53 3:/7.25 M T

Fig. 6.19 Esr spectra (9.18GHz, 0-1000mT) of 51i(Mn)(dab)2Br;b3r: ----, experimental; , calculated

for D=0.60cm1, X=0, An —50mT. ND pp

221

CHAPTER 7

EXPLRIT-IENTAL

(a) Preparation of the Complexes

Preparative details are given at the end of each chapter.

Solvents were deoxygenated before use by bubbling 'white-spot' nitrogen

gas through them for at least 15 minutes. Products were vacuum dried,

except where stated otherwise.

The nominal concentration of manganese(II) in most of the doped

solids is 1 mole96; where the manganese(II) level is higher, this is

indicated in the captions to the appropriate spectra.

(b) Physical Measurements

Analyses for C, H and N were carried out by Mr. K.I. Jones of

the Microanalytical Laboratory, Imperial College.

X-band esr spectra were obtained on a Varian Ell spectrometer

with 100kHz field modulation. The Q-band instrument comprises a V-4561

microwave bridge and cavity and a Newport Instruments type F magnet

powered by a C905 rotary generator. Spectra were obtained on polycrystalline

samples at room temperature and analysed using the computer programs

ESRS and SHAPES) described in Chapter 1. Programs ADD, for summation

of calculated spectra (see Fig. 4.9), and HYP, which calculates hf line

positions, were written by G.M. Lack. All calculations were performed

on the CDC 6400 computer at Imperial College Computer Centre.

IR spectra (4000-200 cm 1) were run on Perkin-Elmer 457 and

325 grating spectrophotometers with the samples as nujol or

hexachlorohutadiene mulls between caesium iodide plates. Far-IR spectra

(200-40 cm-1)* were obtained on a Beckman R.I.I.C. FS-720 interferomcter:

am grateful to Dr. I. Jeeves for obtaining these spectra.

222

using Vaseline mulls and polythene 7)1ELtes. Raman spectra of the

powdered sol:'_ds were run on a Cary model 81 spectrophotometer with

a CRL 52MG argon-krypton laser.

Diffuse reflectance spectra. were obtained on Beckman DK2i and

Cary 14 instruments.

X-ray powder diffraction photographs were obtained by

Miss R.S. Osborn (of the Crystallography Department, Imperial College)

with a Guinier camera using Cu(Ka) radiation.

223

REFERENCES

1. A. Abragam and B. Bleaney, 'Electron Paramagnetic Resonance of

Transition Ions', Clarendon Press, Oxford, 1970.

2. T.P.P. Hall, W. Hayes and F.I.B. Williams, Proc. Phys. Soc.,

78, 883 (1961).

3. A. Chatelain and R.A. Weeks, J. Chem. Phys., 52, 3758 (1970).

4. D.W. Posener, J. Mag. Res., 9, 128 (1973)•

5. R.S. Title, Phys. Rev., 131, 2503 (1963).

6. F.J. Owens, J. Chem. Phys., 57, 2349 (1972).

7. J.R. Gabriel, D.F. Johnston and M.J.D. Powell, Proc. Roy. Soc.,

A 264, 503 (1961).

8. A.J. Shuskus, Phys. Rev., 127, 1529 (1962).

9. A. Igelmund and A. Hausmann, Z. Physik B, 21, 65 (1975).

10. S.A. Al'tshuler and B.M. Kozyrev, 'Electron Paramagnetic Resonance',

Academic Press Inc. (London) Ltd., 1964.

11. A. Chatelain and R.A. Weeks, J. Chem. Phys., 58, 3722 (1973).

12. R.M. Golding, R.H. Newman, A.D. Rae and W.C. Tennant, J. Chem.

Phys., 57, 1912 (1972).

13. R.M. Golding and W.C. Tennant, Mot. Phys., 28, 167 (1974).

14. W.C. Tennant, J. Mag. Res., 14, 152 (1974).

15. J.S. Shaffer, H.A. Farach and C.P. Poole, Jr., Phys. Rev., B13,

1869 (1976).

16. U. Krauss and G. Lehmann, Z. Naturforsch., A30, 28 (1975).

17. J.-P. Rivera, H. Bill, J. Weber, R. Lacroix, G. Hochstrasser and

H. Schmid, Solid State Commun., 14, 21 (1974).

18. J. Chandrasekhar, N.C. Kumar and S. Subramanian, J. Mag. Res., 18,

129 (1975).

19. D.C. Bradley, R.G. Copperthwaite, S.A. Cotton, K.D. Sales and

J.F. Gibson, J. Chem. Soc. Dalton Trans., 191 (1973).

224

20. J.C.M. Henning and P.F. Bangers, J. Phys. Cbem. Solids, 27,

745 (1966).

21. M. Schlaak and A. Weiss, Z. Naturforsch., 28a, 1932 (1973).

22. R.D. Dowsing, J.F. Gibson, M. Goodgame and P.J. Hayward,

J. Chem. Soc.(A), 187 (1969).

23. R.D. Dowsing, J.F. Gibson, D.M.L. Goodgame, M. Coodgame and

P.J. Hayward, J. Chem. Soc. (A), 1242 (1969).

24. R.D. Dowsing, J.F. Gibson, M. Goodgame and P.J. Hayward,

J. Chem. Soc. (A), 1133 (1970).

25. D.M.L. Goodgame, M. Goodgame and p.a. Hayward, J. Chem. Soc. (A),

1352 (1970).

26. M. Goodgame and P.J. Hayward, J. Chem. Soc. (A), 3406 (1971).

27. S.A. Cotton and J.F. Gibson, J. Chem. Soc. (A), 2105 (1970).






33. R.D. Dowsing, B. Nieuwenhuijse and J. Reedijk, Inorg. Chim.

Acta, 5, 301 (1971).

34. G.L. McPherson and R.C. Koch, J. Chem. Phys. 60, 1424 (1974).

35. G.L. McPherson and K.O. Devaney, Inorg. Chem. 16, 1565 (1977).

36. G.M. Cole and B.B. Garrett, Inorg. Chem., 13, 2680 (1974).

37. H.J. Stoller, H.H. Gunthard and G. Rist, J. Chem. Phys., 57,

4651 (1972).

38. D. Vivien and J.F. Gibson, J. Chem. Soc., Faraday Trans. 2,

71, 1640 (1975).

39. R.S. Saraswat and G.C. Upreti, J. Mag. Res., 20, 39 (1975).

40. R.S. Saraswat and G.C. Upreti, J. Phys. C. (Solid State), 10, 2233 (1977).

225

41 M.R. John and R. Myers, Phys. Rev., 513, 1006 (1976).

42. G.M. WolLermAnn and J.R. Wasson, Chem. Phys. Lett., 16,92 (1972).

43. G.M. Woltermann and J.R. Wasson, Inorg. Chem. 12, 2366 (1973).

44. G.M. Woltermann and J.R. Wasson, J. Phys. Chem. , 77, 945 (1973).

45. G.M. Woltermann and J.R. Wasson, J. Phys. Chem., 78, 45 (1973).

46. R. Janakiraman and G.C. Upreti, J. Phys. Chem. Solids, 31 ,

1419 (1970).

47. G.C. Upreti, J. Mag. Res., 13, 336 (1974).

48. G.C. Upreti, J. Mag. Res. 18, 287 (1975).

49. H. Kumagai, K. Ono, I. Hayashi and K. Kambe, Phys. Rev.,

87, 374 (1952).

50. W.V. Sweeney, D. Coucouvanis and R.E. Coffman, J. Chem. Phys.,

59, 369 (1973).

51. R.L. Carlin and C.J. O'Connor, Inorg. Chem., 14, 291(1975).

52. R.B. Birdy and M. Goodgame, J. Chem. Soc. Dalton Trans., 461 (1977).

53. J.W. Orton, 'Electron Paramagnetic Resonance% Iliffe, 1968.

54. J.E. Wertz and J.R. Bolton, 'Electron Spin Resonance: Elementary

Theory and Practical Applications', McGraw-Hill Book Company,

1972.

55. P.L. Hall, B.R. Angel and J.P.E. Jones, J. Mag. Res., 15, 64 (1974).

56. W.E. Blumberg, in 'Magnetic Resonance in Biological Systems',

A. Ehrenberg, B.E. Malmstrom and T. Vanngard, Eds.; Pergamon

Press, Ltd., London, 1967.

57. H.W. de Wijn and R.F. van Balderen, J. Chem. Phys., 46, 1381 (1967).

58. W. Low, 'Paramagnetic Resonance in Solids', Academic Press Inc.,

New York, 1960.

59. G.L. Bir, Soviet Phys. Solid State, 5, 1628 (1964).

60. R. Aasa, J. Chem. Phys., 52, 3919 (1970).

226

61. J.F. isaugher, P.C. Taylor, T. Oja and P.J, Bray, J. Chem.

Phys., 50, 4914 (1969).

62. R.D. Dowsing and J.F. Gibson, J. Chem. Phys., 50, 294 (1969).

63. G.M. Lack, Ph.D. thesis, University of London (to be submiLted).

64. J.J. Villafranca, D.E. Ash and F.C. Wedler, Biochem., 15, 544 (1976).

65. P.S. Bryan and J.C. Dabrowiak, Inorg. Chem., 14, 296 (1975).

66. M.H. Jones, W. Levason, C.A. McAuliffe and M.J. Parrott,

J. Chem. Soc. Dalton Trans., 1642 (1976).

67. J.P. Barbier, R.P. Hugel, P.J. van der Put and J. Reedijk,

Rec. Tray. Chim. Pays-Has, 95; 213 (1976).

68. W.C. Potter and L.T. Taylor, Inorg. Chem., 15, 1329 (1976).

69. D.L. Griscom and R.E. Griscom, J. Chem. Phys., 47, 2711 (1967).

70. P.C. Taylor and P.J. Bray, J. Phys. Chem. Solids, 33, 43 (1972).

71. M. Kopp and J.H. Mackey, J. Computational Physics, 3, 539 (1969).

72. E.M. Price, Ph.D. thesis, University of London (1972).

73. R.M. Butterworth, Ph.D. thesis, University of London (1976).

74. B. Bleaney and R.S. Rubins, Proc. Phys. Soc. (London), 77, 103

(1961); corrigendum, 78, 778 (1961)-

75. B.G. Wybourne, J. Chem. Phys., 43, 4506 (1965).

76. A. van Heuvelen, J. Chem. Phys., 46, 4903 (1967).

77. D.J. Newman and W. Urban, Adv. Phys., 24, 793 (1975).

78. R. Chatterjee, M.R. Smith, and H.A. Buckmaster, Can. J. Phys.,

54, 1224 (1976).

79. J.V. Quagliano, G. Franz, J. Fujita, D.J. Phillips, J.A. Walmsley,

and S.Y. Tvree,J. Amer. Chem. Soc.,83, 3770 (1961).

80. R.L. Carlin, J. Amer. Chem. Soc., 83, 3773 (1961).

81. N.M. Karayannis, L.L. Pytlewski and C.M. Mikulski, Coorci. Chem.

Rev., 11, 93 (1973).

227

82. A.D. van Ingen Schcnau, Acta Cryst., Sect. B; 30, 1686 (1974).

83. T.J. Berdendahl and J.S. Wood, Inord. Chem., 14, 338 (1975).

84. D.L. Kepert, D. Taylor and A.H. White, J. Chem. Soc. Dalton Trans.,

670 (1973).

85. L.C. Nathan and R.O. Ragsdale, Inorg. Chim. Acta. 10, 177 (1974).

86. D.W. Herlocker, R.S. Drago and V.I. Meek, Inorg. Chem., 5, 2009

(1966).

87. W.D. Perry, R.S. Drago, D.W. Herlocker, G.W. Pagenkopf

and K. Czworniak, Inorg. Chem. 10, 1087 (1971).

"88. I. Bertini, D. Gatteschi and A. Scozzafava, Inorg. Chim Acta. 6,

185 (1972).

89. W. Byers, B. Fa-Chun Chou, A.B.P. Lever and R.V. Parish,

J. Amer. Chem. Soc., 91, 1329 (1969).

90. W. Byers, A.B.P. Lever and R.V. Parish, Inorg. Chem., 7, 1835 (1968).

91. R.L. Carlin, C.J. O'Connor and S.N. Bhatia, J. Am. Chem. Soc.,

98, 3523 (1976).

92. J. Reedijk and B. Nieuwenhuijse, Rec. Tray. Chim. Pays-Bas,

91, 533 (1972).

93. P.L. Mukherjee, Z. Kristallogr., 91, 504 (1935).

94. J.H.E. Griffiths and J.. Owen, Proc. Roy. Soc. Ser. A., 213,

459 (1952).

95. B. Bleaney and D.J.E. Ingram, Proc. Roy. Soc. (IQ, 205, 336 (1951).

96. J.C. Hempel, J. Chem. Phys., 64, 4307 (1976).

97. J.R. Sams and T.B. Tsin, Inorg. Chem., 14, 1573 (1975).

98. D.J. Mackey and S.V. Evans, J. Chem. Soc. Dalton Trans.,

2C04 (1976).

99. G.L. Bir, L.V. Dmitrieva, and S.A. Kizhaev, Soviet Phys. Solid

State, 14,18 (1972).

228

300. V. S. Grunin, Z.N. loon, V.A. Tol'fe, and T.B. PaMrina,

Soy. Phys. Solid State, 17, 2065 (1976).

101. M. Vijayan and N.A. Viswamitra, Acta. Cryst., 21, 522 (1966).

102. M. Vijayan and M.A. Viswamitra, Acta Cryst., 23, 1000 (1967).

103. M. Vijayan and M.A. Viswamitra, Acta Cryst., Sect. B, 24, 1067 (1968).

104. A.D. Mighell and A. Santoro,Acta Cryst., B27, 2089 (1971).

105. R. Birdy and M. Goodgame, unpublished work.

106. V.K. Sharma, J. Chem. Phys., 54, 496 (1971).

107. H.B. Gray, R. Eisenberg and E.I. Stiefel, Adv. Chem. Series,

62, 641 (1967).

108. J.A. McCleverty, Progr. in Inorg. Chem., 10, 49 (1968).

109. R. Eisenberg, Progr. in Inorg. Chem., 12, 295 (1970).

110. E.I. Stiefel and G.F. Brown, Inorg. Chem., 11, 434 (1972).

111. A. Avdeef and J.P. Fackler, Inorg. Chem., 14, 2002 (1975).

112. D.L. Kepert, Inorg. Chem., 11, 1561 (1972).

113. E. Larsen, G.N. La Mar, B.E. Wagner, J.E. Parks and R.H. Holm,

Inorg. Chem., 11, 2652 (1972).

114. E.B. Fleischer, A.E. Gebala, D.R. Swift and P.A. Tasker,

Inorg. Chem., 11, 2775 (1972).

115. C.G. Pierpont and R.M. Buchanan, J. Am. Chem. Soc., 97, 4912 (1975).

116. W. Hewertson, B.T. Kilbourn and R.H.B. Mais, Chem. Commun., 952 (1970).

117. J.A. Bertrand and E.G. Vassian, J. Am. Chem. Soc., 91, 2394 (1969).

118. M.R. Truter and B.L. Vickery, J. Chem. Soc. (A), 2077 (1971).

119. T.M. Greaney, C.L. Raston and A.H. White, J. Chem. Soc. Dalton

Trans. 876 (1975).

120. D.E. Fenton, C. Nave and M.R. Truter, J. Chem. Soc. Dalton

Trans., 2188 (1973).

121. M.R. Truter and B.L. Vickery, J. Chem. Soc. Dalton Trans.,

395 (1972).

229

122. B. Granoff, Ph.D. thesis, Princeton University, 1966.

123. W'.H. Watson and C.T. Lin, Inorg. Chem. , 5, 1074 (1966).

124. R. Hultgren, Phys. Rev., 40, 891 (1932).

125. R. Hoffmann, J.M. Howell and A.R. Rossi, J. Amer. Chem. Soc.,

98, 2484 (1976).

126. F.P. Dwyer and A.M. Sargeson, J. Proc. Roy. Soc. New South Wales,

40, 29 (1956).

127. R.D. Hancock and D.A. Thornton, J. Sth.African Chemical Institute,

23, 71 (1970).

128. W.O. Gillum, R.A. Wentworth and R.F. Childers, Inorg. Chem.,

9, 1825 (1970).

129. D.P. Graddon, Coord. Chem. Rev., 4, 1 (1969).

130. G. Maki, J. Chem. Phys., 29, 162 (1958).

131. W. de W. Horrocks, R.H. Fischer, J.R. Putchison and G.N. La Mar,

J. Am. Chem. Soc., 88, 2436 (1966).

132. F.A. Cotton and J.P. Fackler, Jr.,J. Amer. Chem. Soc., 83,

2818 (1961).

133. K. Nakamoto, P.J: McCarthy and A.E. Martell, J. Amer. Chem.

Soc., 83, 1272 (1961).

134. S. Pinchas,B.L. Silver and I. Laulicht, J. Chem. Phys., 46,

1506 (1967).

135. R.R. Sharma,Phys. Rev. B, 3, 76 (1971).

136. M. Peter, Phys. Rev., 116, 1432 (1959).

137. E.A. Noe and M. Raban, Chem. Commun., 165 (1976).

138. M.V. Veidis, G.H. Schreiber, T.E. Gough and G.J. Palenik,

J. Am. Chem. Soc., 91, 1859 (1969).

139. B. Morosin, Acta Cryst., 22, 315 (1967).

140. H. Montgomery and E.C. Lingafelter, Acta Cryst., 17,

1481 (1964).

230

141. J.T. He:_h gen and J.P. Fackle„ Jr., J. Amer, Chem. Soc.,

87 j 2821 (1965).

142. J. Pradilla-Sorzano and J.P. Fackler, Inorg. Chem., 12, 1174 (1973).

143. F.J. Hollander, D.H. Templeton, and A. Zalkin, Acta Cryst.,

Sect. 13, 29, 1289 (1973).

144. K. Nakamoto, 'Infrared Spectra of Inorganic and Coordination

Compounds', 2nd ed., Wiley, 1970.

145. J.G. Gibson and E.D. McKenzie, J. Chem.Soc. (A), 2637 (1969).

146. A.J. Carty, K.R. Dymock and P.M. Boorman, Can. J. Chem.,

48, 3524 (1970).

147. J.S. Griffith, Mol. Phys., 8, 213, 217 (1964).

148. See, for example, C.A. McAuliffe (ed.), 'Transition Metal

Complexes of Phosphorous, Arsenic and Antimony Ligands

Macmillan, 1973.

149. B.R. McGarvey, J. Chem. Phys., 41, 3743 (1964).

150. J.C. Hempel, L.O. Morgan and W.B. Lewis, Inorg. Chem., 9,

2064 (1970).

151. W.T. Andriessen, Inorg. Chem., 14, 792 (1975).

152. W.T.M. Andriessen and M.P. Groenewege, Inorg. Chem.,

15, 621 (1976).

153. D.E. Fenton, J. Chem. Soc. (A), 3481 (1971).

154. A.A. Schilt and R.C. Taylor, J. Inorg. Nucl. Chem., 9,

211 (1959).

155. E. Konig and E. Lindner, Spectrochim. Acta, 28A, 1393 (1972).

156. A.J. Carty, Can. J. Chem., 46, 3779 (1968).

157. I. Nakagawa and T. Shimanouchi, Spectrochim. Acta, 20, 429 (1964).

158. R.J.H. Clark and C.S. Williams, Spectrochim. Acta, 23A, 1055 (1967).

159. B. Hutchinson, J. Takemoto and K. Nakamoto, J. Am. Chem.

Soc., 92, 3335 (1970).

231

160. Y. Saito, J. Takemotc, F. HutchinL;on and K. Nakamoto,

Inorg. Chem., 11, 2003 (1972).

161. P.C. Taylor, J.F. Baugher and H.M. Kriz, Chem. Rev., 75,

203 (1975).

162. J.N. van Niekerk, F.R.L. Schoening and J.H. Talbot, Acta

Cryst., 6, 720 (1953).

163. J. Pradilla-Sorzano and J.P. Fackler, Inorg. Chem., 12, 1182 (1973).

164. D.E. Fenton, M.P. Truter and B.L. Vickery, Chem- Commun.,

93 (1971).

165. A.H. Maki and B.R. McGarvey, J. Chem. Phys., 29, 31 (1958).

166. J. Ammeter, G. Rist and Hs. H. Gunthard, J. Chem. Phys., 57,

3852 (1972).

167. P.G. Hodgson and B.R. Penfold, J. Chem. Soc. Dalton Trans., 1870

(1974).

168. P.C. Healy, G.M. Mockler, D.P. Freyberg and E. Sinn, J. Chem.

Soc. Dalton Trans., 691 (1975).

169. M.R. Lowrey and J.R. Pilbrow, J. Phys.(C), 10, 439 (1977).

170. D.L. Kepert, A.H. White and A.C. Willis, J. Chem. Soc.

Dalton Trans., 1342 (1977).

171. N. Milinski and B. Ribar, Cryst. Structure Commun., 3,

757 (1974).

172. B. Matkovic, B. Ribar, B. Zelenko and S.W. Peterson, Acta Cryst.,

21, 719 (1966).

173. M.L. Post and J. Trotter, J. Chem. Soc. Dalton Trans., 1922

(1974).

174. M.L. Post and J. Trotter, J. Chem. Soc. Dalton Trans.,

674 (1973).

175. O.P. Anderson, J. Chem. Soc. Dalton Trans., 1237 (1973).

176. O.P. Anderson, J. Chem. Soc. Dalton Trans., 2597 (1972).

232

177. D.E. Cullen and E.C. Lipgafelter, Inorg. Chem., 9, 1858 (_x.970).

178. M.D. Joesten, M.S. Hussain and P.G. Lenhert, Inorg, Chem.,

9, 51 (1970).

179. F.M. Jaeger and J.A. van Dijk, Z. anorg. al1gem. Chem., 227,

273 (1936).

180. E. Sinn, J. Chem. Soc. Dalton Trans., 162 (1976).

181. J.D. Dunitz, Acta Cryst., 10, 307 (1957).

182. P.M. Richards, R.K. Quinn and B. Morosin, J. Chem. Phys.,

59, 4474 (1973).

183. S. Akyuz, A.B. Dempster, J.F.D. Davies and K.T. Holmes, J. Chem.

Soc. Dalton Trans., 1746 (1976).

184. R.J.H. Clark and C.S. Williams, Inorg. Chem., 4, 350 (1965).

185. G.E. Coates and D. Ridley, J. Chem. Soc., 166 (1964).

186. M. Goldstein and W.D. Unsworth, Inorg. Chim. Acta, 4, 342 (1970).

187. M. Goldstein and W.D. Unsworth, Spectrochim. Acta Part A, 28,

1107 (1972).

188. G. Beech, C.T. Mortimer and E.G. Tyler, J. Chem. Soc. (A),

512 (1969).

189. S.M. Nelson and T.M. Shepherd, J. Chem. Soc., 3276 (1965).

190. D.M.L. Goodgame, M. Goodgame and M.J. Weeks, J. Chem. Soc.,

5194 (1964).

191. • D.M.L. Goodgame, M. Goodgame, M.A. Hitchman and M.J. Weeks,

J. Chem. Soc. (A), 1769 (1966).

192. D.H. Brown, R.H. Nuttall and D.W.A. Sharp, J. Inorg. Nucl. Chem.,

25, 1067 (1963).

193. J.R. Allan, D.H. Brown, R.H. Nuttall and D.W.A. Sharp,

J. Inorg. Nucl. Chem., 26, 1895 (1964).

194. J.R. Allan, D.H. Brown, R.H. Nuttall and D.W.A. Sharp, J. Inorg.

Nucl. Chem., 27, 1529 (1965).

233

195. S. Buffagrii, L.M. Vallarino and J.V- Quaglia.no, Inorg. Chem.,

3, 480 (1964).

196. S. Buffacjni, L.M. Vallarino and J.V. Ouaglino, inorg. Chem.,

3, 671 (1964).

197. J.R. Allan, D.H. Brown, R.H. Nuttall and D.W.A. Sharp, J. Chem.

Soc. (A), 1031 (1966).


Nucl. Chem., 27, 1865 (1965).


Nucl. Chem., 27, 1305 (1965).

200. D.E. Billing and A.E. Underhill, J. Chem. Soc. (A), 29 (1968).

201. Y. Saito, M. Cordes and K. Nakamoto, Spectrochim. Acta, A28,

1459 (1972).

202. M. Goldstein and W.D. Unsworth, J. Mol. Struct., 14, 451 (1972).

203. R.M. Barr, M. Goldstein and W.D. Unsworth, J. Cryst. Mol.

Struct., 4, 165 (1974).

204(i) Y.A. Sokolova, L.O. Atovmyan and M.A. Porai-Koshits, J. Struct.

Chem., 7, 794 (1965).

(ii) W.L. Steffen and G.J. Palenik, Acta Cryst., B32, 298 (1976).

205. J.F. Le Querler, M.M. Sorel and A. Leclaire, Acta Cryst., B33,

2299 (1977).

206. H. Paulus, Z. anorg. Chem., 369, 38 (1969).

207. T.I. Malinovskii and Y.A. Simonov, Dokl. Akad. Nauk,

147, 96 (1962).

208. D. Grdenic and J. Krstanovic, Ark. Kim., 27, 143 (1955).

209. R. Zanetti, Gazzetta, 90, 1428 (1960).

210. H.D. Stidham and J.A. Chandler, J. Inorg. Nucl. Chem., 27,

397 (1965).

234

211_ M. Goldstein, F.B. Taylor and U.D. Unsworth, J. Chem. Soc.

DaltonTrans., 410 (1972).

212. A.B.P. Lever, J. Lewis and R.S. Hyhoim, J. Chem. Soc., 1235 (1962);

5042,(1963); 4761 (1964).

213. J.R. Ferraro, W. Wozniak and G. Rosh, Ricerca. sci., 38, 433 (1968)-

214. A. Tenhunen, Suom. Kemistilehti B, 45, 298 (1972).

215. G. Beech and C.T. Mortimer, J. Chem. Soc.(A), 1115 (1967).

21.6. P.W. Carreck, M. Goldstein, E.M. McPartlin and W.D. Unsworth,

Chem. Commun., 1634 (1971).

217. C. Furlani, G. Mattogno, A. Monad and. F. Tarli, Inorg. Chim.

Acta, 4, 187 (1970).

218. F. Bottomley, Quart. Rev., 24, 61.7 (1970).

219. L. Sacconi and A. Sabatini, J. Inorg. Nucl. Chem., 25, 1389 (1963).

220. D. Nicholls and R. Swindellst J. Inorg. Nucl. Chem., 30, 2211 (1968).

221. A. Braibanti, F. Dallavalle, M.A. Pellinghelli and E. Leporati

Inorg. Chem., 7, 1430 (1968).

222. J.R. Durig, S.F. Bush and E.E. Mercer, J. Chem. Phys., 44,

4238 (1966).

223. A. Ferrari, A. Braibanti and G. Bigliardi, Acta Cryst., 16,

498 (1963).

224. A. Ferrari, A. Braibanti, G. Bigliardi and F. Dallavalle,

Z. Krist., 119, 284 (1963).

225. D.R. Glasson and S.J. Gregg, J. Chem. Soc., 1493 (1953).

226. P.J. Hayward, Ph.D. thesis, University of London (1968).

227. V.J. Folen, Phys. Rev., 139, A1961 (1965).

228. M.M. Zaripov and G.K. Chirkin, Soviet Physics Solid State, 7,

2390 (1966).

229. J. Reedijk, F.W. Klaaijsen and H.T. Witteveen, J. Chem. Soc.,

Faraday Trans. 2, 69, 1537 (1973).

235

230. V. Lupei, A. LI;pei and F. D.)1-nr-iza, J. 1.1g. Res., 19, 337 (1975).

231. L.D. Kanc1-1, Y.C.G. Passcgi and T. Birch, 3. Phys. Chem.

Sono's, 30, 321 (1969).

232. I. Fidone and K.W.H. Stevens, Proc. Phys. Soc. (London) 73,

116 (1959).

233. C.W. Reimann, 3. Phys. Chem. 74, 561 (1970).

234. C.W. Reimann, A.D. Mighell and F.A. Mauer, Acta Cryst. 823,

135 (1967).

235. K. Bradersen, Z. anorg. chem., 290, 24 (1957).

236. S.V. Waggett, Ph.D. thesis, University of London, 1972.

237. D.A. Rowley and R.S. Drage, Inorg. Chem., 7, 795 (1968).

238. R.B. Bentley, M. Gerlach, J. Lewis and P.N. Quested,

J. Chem. Soc.(A), 3751 (1971).

239. A. Narath, Phys. Rev. 139A, 1221 (1965).

240. R.L. Chiang and B.S. Drago, Inorg. Chem., 10, 453 (1971).

241. A.B.P. Lever, S.M. Nelson and T.M. Shepherd, Inorg. Chem. 4,

811 (1965).

242. G.E. Shankle and J.B. Bates, J. Chem. Phys., 64, 2539 (1976).

243. C.H.W. Swuste, A.C. Botterman, J. Miilenaar and W.J.M. de Jonge,

J. Chem. Phys., 66, 5021 (1977).

244. B. Morosin, Acta Cryst., 23, 630 (1967).

245. C.K. Prout, and H.M. Powell, J. Chem. Soc., 4177 (1961).

246. J.V. Quagliano, J.T. Summers, S. Kida and L.M. Vallarino,

Inorg. Chem., 3, 1557 (1964).

247. A.K. Banerjee, L.M. Vallarino and J.V. Quagliano, Coord. Chem.

Rev., 1, 239 (1966).

248. V.L. Goedken, J.V. Quagliano and L.M. Vallarino, Inorg. Chem.,

8, 2331 (1969).

249. J.V. Quagliano, A.K. Banerjee, V.L. Goedken and L.M. Vallarino,

J. Am. Chem. soc., 92, 482 (1970).

236

250. V.L. Goedken, L.M. Vallarino and J.V. Ouagliano, Inorg.

Chem., 10, 2682 (1971).

251. L.M. Vallarino, V.L. Goedken and J.V. Quagliano, Inorg. Chem.,

11, 1466 (1972).

252. L.M. Vallarino, V.L. Goedken and J.V. Quagliano, Inorg. Chem.,

12, 102 (1973).

253. W.D. Perry, J.V. Quagliano and L.M. Vallarino, Inorg. Chim. Acta,

7, 175 (1973).

254. F.A. Cotton, D.M.L. Goodgame and M. Goodgame, J. Amer. Chem. Soc.1

83, 4690 (1961).

255. D.M.L. Goodgame, M. Goodgame and F.A. Cotton, J. Amer. Cnem. Soc.,

83, 4161 (1961).

256. P.J. Clarke and H.J. Milledge, Acta Cryst., B31, 1543 (1975).

257. M.A. Porai-Koshits, Trudy Inst. Krist., Akad. Nauk SSSR, 10,

117 (1954).

258. T. Ito, N. Morimotoand R. Sadanaga, Acta Cryst., 4, 310 (1951).

259. D.M.L. Goodgame, M. Goodgame and G.W. Rayner-Canham, Inorg.

Chim. Adta, 3, 399 (1969).

260. W.J. Eilbeck, F. Holmes, C.E. Taylor and A.E. Underhill,

J. Chem. Soc. (A), 128 (1968).

261. F.L. Phillips, F.M. Shreeve and A.C. Skapski, Acta Cryst., B32,

687 (1976).

262. C. Pelizzi and G. Pelizzi, Acta Cryst., Sect. B30, 2421 (1974).

263. B. Gonzalez, J. Kouba, S. Yee, C.A. Reed, J.F. Kirner and W.R. Scheidt,

J. Amer. Chem. Soc., 97, 3247 (1975).

264. R.J. Butcher and E. Sinn, J. Chem. Soc. Dalton Trans., 1186 (1976).

265. M. Di Vaira and P.L. Orioli,Inorg. Chem., 6, 955 (1967).

266. M. Ciampolini and G.P. Speroni, inorg. Chem. 5, 45 (1966).

237

267. J.K. St;-Jick, P.W.R. CorfieLd and D.)7, iice.k, Inorg. Chem.,

12, 166F3 (1973).

268. R. MoraF,si, I. Beitini and L. Sacconi, Coord. Chem. Rev., 11,

343 (1973).

269. G.W. Rayner Canham, Ph.D. thesis, University of London, 1969.

270. C. Postrous, J.R. Ferraro and W. Wozniak, Inorg. Chem., 6, 2030

(1967).

271. D.J. Bryson and R.H. Nuttall, Inorg. Nucl. Chem. Lett., 5,

347 (1969).

272. B.F. Williams and R.R. Hewitt, Phys. Rev.1 146, 286 (1966).

273. R. Lefebvre and J. Maruani, J. Chem. Phys., 42, 1493 (1963).

274. C.C. Briars and M. Goodgame, unpublished work.

275. M. Di Vaira and P.L. Orioli,Inorg. Chem. 8, 2729 (1969).

276. Z. Dori and H.B. Gray, J. Amer. Chem. Soc., 88, 1394 (1966).

277. Z. Dori, R. Eisenberg and H.B. Gray, Inorg. Chem., 6, 483 (1967).

278. J.N. McElearney, D.B. Losee, S. Merchant and R.L. Carlin, Phys.

Rev. B7, 3314 (1973).

279. F.W.B. Einstein and B.R. Penfold, Leta Cryst., 20, 924 (1966).

280. M. Gerloch, J. Chem. Soc. (A) , 1317 (1966).

281. G. Morgan and F.H. Burstall, J. Chem. Soc., 1649 (1937).

282. G.W. Rayner Canham and M. Goodgame, unpublished work.

Date post:	26-May-2018
Category:	Documents
Upload:	truongminh
View:	214 times
Download:	0 times

STUDIES IN THE STEREOCHEMISTRY OF THE … · means of determination of the zero-field splitting ......

Documents