Citation
Preparation and characterization of copper (II) complexes of the condensation products of 2,6-diacetylpyridine and 2,6-pyridinedicarboxaldehyde with l,8-diaminonaphthalene

Material Information

Title:
Preparation and characterization of copper (II) complexes of the condensation products of 2,6-diacetylpyridine and 2,6-pyridinedicarboxaldehyde with l,8-diaminonaphthalene
Creator:
Romanik, Barbara Judith, 1942-
Publication Date:
Language:
English
Physical Description:
x, 86 leaves : ill. ; 28cm.

Subjects

Subjects / Keywords:
Absorption spectra ( jstor )
Atoms ( jstor )
Chlorides ( jstor )
Infrared spectrum ( jstor )
Ions ( jstor )
Ligands ( jstor )
Magnetism ( jstor )
Metal ions ( jstor )
Solvents ( jstor )
Trimers ( jstor )
Chemistry thesis Ph. D
Complex compounds ( lcsh )
Dissertations, Academic -- Chemistry -- UF
Ligand field theory ( lcsh )
Metal ions ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 82-85.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Barbara Judith Romanik.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
025211452 ( ALEPH )
02664352 ( OCLC )

Downloads

This item has the following downloads:


Full Text












PREPARATION AND CHARACTERIZATION OF COPPER(II) COMPLEXES
OF THE CONDENSATION PRODUCTS OF
2,6-DIACETYLPYRIDINE AND 2,6-PYRIDINEDICARBOXALDEHYDE WITH 1, 8-DIAMINONAPHTHALENE












By

BARBARA JUDITH ROMANIK


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











UNIVERSITY OF FLORIDA


1975





































To my parents and grandparents,

for love and encouragement















ACKNOWLEDGEMENTS

The author wishes to express her sincere appreciation to Dr. R.C.

Stoufer, Chairman of the author's Supervisory Committee, and to the other members of her Supervisory Committee.

Special thanks to Dr. R.W. King for his aid in obtaining the mass spectral data given in the manuscript.

Much appreciation is expressed to the typist, Miss Oonagh Kater.















TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS ................................................ iii

LIST OF TABLES .................................................. v

LIST OF FIGURES ................................................. vi

KEY TO SYMBOLS USED IN TEXT ..................................... ...viii

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

INTRODUCTION .......................................................... 1

EXPERIMENTAL ........................................................ 8

Reagents ................................................... 8
Preparation of Starting Materials .............................. 8
Procedures ...................................................... 9
Apparatus .................................................. 15

RESULT AND DISCUSSION ............................................... 18

Ligands .................................................... 18
Complexes .................................................. 22
Infrared Spectra ............................................... 25
Electronic Spectra ............................................. 38
Magnetic Properties ............................................ 56
Electron Spin Resonance Studies ............................... 67
Further Characterization Attempts ............................. 75

SUMMARY ......................................................... 77

APPENDIX ........................................................ 79

BIBLIOGRAPHY .................................................... 82

BIOGRAPHICAL SKETCH ................................................. 86















LIST OF TABLES


Table Page

1. Near Infrared, Visible and Ultraviolet Spectral
Data .................... .................................. 41

2. Average Magnetic Susceptibilities and Moments ............. 59

3. Temperature Dependence of Average Magnetic
Susceptibilities and Moments .............................. 61

4. Temperature Dependence of_ Values and Line Widths
and Singlet-Triplet Separations ........................... 70

1-A. Mass Spectral Cracking Pattern of Impure Trimer of
TMTC ...................................................... 80

2-A. Mass Spectral Cracking Pattern of Impure DDnTC ............. 80
i
3-A. Octahedral Ionic Radii of Metal Ions in Oxide
Salts ...................................................... 81














LIST OF FIGURES


Figure

1. Structural formula of TMTC ........................

2. Structural formula of DDnTC .......................

3. Structural formula of TMCD ........................

4. Structural formula of Cu(DAN)2(N03)2

5. Infrared spectra of DDnTC and its precursors ......

6. Infrared spectra of impure trimer of TMTC and its
precursors ........................................

7. Infrared spectrum of Cu(D2A ?O3)2 4H20...........


Page

5 6 7 26 27


Infrared spectrum of Cu (DA)C ' 2H20 32
2 22I4 2. .....................
Infrared spectrum of Co(P2A2)Cl2-4H20 ....................... 33

Infrared spectrum of Cu2(P2A2)C14 4H20 ...................... 34

Infrared spectrum of Cu(D2A2)(C104)2-4H20 ...................... 35

Infrared spectrum of Cu(P2A2)(ClO4)2-2H0...................... 36

Infrared spectrum of Cu(DAN) 2 (NO3)2 ......................... 37

Electronic spectrum of impure trimer of TMTC ................... 39

Electronic spectrum of DDnTC .................................... 40

The splitting of the 2D term of the copper(II) ion in ligand fields of different symmetries .......................... 45


Electronic spectrum Diffuse reflectance Electronic spectrum Diffuse reflectance Electronic spectrum


of Cu(D2A2) (NO3)2" 4H20 ..................

spectrum of Cu2 (D2A2)C14" 2H20 ...........

of Cu(D2A2) (ClO4)2 4H20 ..................

spectrum of Cu2 (P2A2)C4- 4H29 ...........

of Cu(P2A2) (ClO4)2- 2H20 .................








LIST OF FIGURES (Continued)


Figure Page 22. Diffuse reflectance spectrum of Cu(DAN) )2(N03)2 ............... 53

23. Diffuse reflectance spectrum of Co(P2A2)CI2*4H20 ............ 55

24. Temperature dependence of inverse susceptibility
per copper atom in Cu2 (P2A2)CI4 ..............................63

25. Temperature dependence of inverse susceptibility
per copper atom in Cu2 (P2A2)C14-4H20 ........................ 64

26. Temperature dependence of inverse susceptibility
per copper atom in Cu2(D2A2)C14 .............................. 65


vii















KEY TO THE SYMBOLS USED IN TEXT DDnTC 13,9:28-24-dinitrilo-9H,24H-dinaphtho[1,8-bc:l',8'-no][1,5,13,17]tetraazacyclotetracosine

TMTC 8,14,23,29-tetramethyl-13,9:28-24-dinitrilo-9H,24H-dinaphthoE
[1,8-bc:l',8'-no][1,5,13,17]tetraazacyclotetracosine

TMCD 6,12,19,25-tetramethyl-7,11:20,24-dinitrilodibenzo[b,n]E
[1,4,12,15]tetraazacyclodocosine DAP 2,6-diacetylpyridine PDC 2,6-pyridinedicarboxaldehyde DAN 1,8-diaminonaphthalene D2A2 condensate of two moles DAP and two moles DAN P2A2 condensate of two moles PDC and two moles DAN DMSO dimethylsulfoxide DMF N,N'-dimethylformamide NH3 ammonia, en ethylenediamine CDC13 chloroform, deuterated o-phen 1,10-phenanthroline


viii















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


PREPARATION AND CHARACTERIZATION OF COPPER(II) COMPLEXES
OF THE CONDENSATION PRODUCTS OF 2,6-DIACETYLPYRIDINE AND 2,6-PYRIDINEDICARBOXALDEHYDE WITH 1, 8-DIAMINONAPHTHALENE

By

Barbara Judith Romanik

August, 1975

Chairman: R. Carl Stoufer
Major Department: Chemistry

Six new complexes proposed to contain the macrocyclic ligands, 13,9:28,24-dinitrilo-9H,24H-dinaphtho[l,8-bc:l',8'-no][l,5,13,17],.tetraazacyclotetracosine (DDnTC) and 8,14,23,29-tetramethyl-13,9:28,24dinitrilo-9H,24H-dinaphtho[l,8-bc:l' ,8'-no][l,5,13,17]tetraazacyclotetracosine (TMTC), have been prepared via the template method. The former ligand was produced by the Schiff base condensation of 1,8diaminonaphthalene and 2,6-pyridinedicarboxaldehyde with the salts Cu(CI04)2-6H 20, CuCl2, and CoCl2. The latter was prepared by the Schiff base condensation of 1,8-diaminonaphthalene and 2,6-diacetylpyridine with the salts Cu(NO3)2"xH20, CuCI and Cu(CIO4)2 6H20. Isolation of the free macrocycle base could not be accomplished by precipitation of the metals with sulfide ion.

The complexes were characterized by elemental analysis, infrared, ultraviolet, visible and electron spin resonance spectra, and magnetic susceptibility determinations. The results of these studies support the ix









formulation of each of the complexes as metal ion(s) surrounded by a planar, quadra- or hexadentate ligand with counterions or water either very loosely held in the axial positions or present as part of the crystal lattice.

The copper(II) chloride and perchlorate complexes presented subnormal room temperature magnetic moments. Temperature-dependent magnetic susceptibility data were obtained for Cu2(TMTC)C14, Cu2(DDnTC)C14, and Cu2(DDnTC)C14"4H20. The complex Cu2(TMTC)C14 presented a transition temperature in the range of 1970 to 2090 K, following the Curie-Weiss law above and below this plateau.

Electron spin resonance data of those compounds possessing anomalous magnetic moments gave evidence of spin-spin exchange of either intra- or intermolecular character. A structure determination would be desired to aid in the elucidation of the mechanism of this exchange, i.e., direct metal-metal interaction or super-exchange. Unfortunately, extreme insolubilities of the complexes have prevented growth of single crystals.















INTRODUCTION

In the past fifteen years the possibility of using synthetic macrocycles as simple models for complicated biological systems has been actively pursued Consequently the area of coordination chemistry involving the synthesis of transition metal complexes incorporating these macrocyclic ligands has grown significantly.

A macrocyclic ligand is one in which the donor atoms are contained as integral parts of a closed ring. The majority of the macrocyclic ligands reported to date are quadridentate ligands. Examples of penta-, hexa-, and octadentate macrocycles are rare.2-5 In most instances, the donor atoms are located approximately in a planar array about the central metal ion.

The complete characterization of a variety of model macrocyclic complexes should lead to a better understanding of naturally occurring analogues, e.g., porphyrins, corrins, which themselves are highly conjugated and which may contain metal ions confined within a cyclic, approximately planar array of four donor atoms.

Initially, the study of these new synthetic systems described herein was undertaken in an attempt to prepare and to characterize complexes containing metal-metal bondsf-9 There are but a very few compounds of this kind which have been prepared and these are incompletely characterized. Macrocyclic ligands with significantly large "holes" lend themselves to incorporating more than one metal ion within the ring, thus increasing the tendency for metal-metal interaction.








Metal-metal bonding is now noted as a widespread phenomenon which is both interesting and important. Of the numerous compounds now known in which metal-metal bonding is thought to occur, most examples contain metals of the second and third transition series. Too, the majority of these compounds contain metal clusters in which the metal ions are in a low oxidation state, bridged by oxo-, halo-, or carbonyl groups. Although direct metal to metal bonding is rare for first transition series metal ions, copper complexes have been reported to involve this type of inter10,11

Macrocyclic ligands and complexes are generally uncommonly stable both in the thermodynamic and the kinetic sense; they are inert toward dissociation, even in strong mineral acid. Because of the ring structure, a simple dissociative step cannot occur; the ring has no "end." It is not possible to extend metal-donor distance sufficiently to constitute bond rupture without additional bond rupture involving the ligand or extensive rearrangement within the coordination sphere. Thus, activation energies are considered to be very large. It is suspected that the more flexible unconjugated rings fold before the first bond between the donor atom and metal ion is broken. It is relatively simple to move a donor atom away from the metal ion when the ring is in a folded configuration. The in-plane ligand field strengths characteristic of macrocyclic ligands have been shown to be somewhat greater than expected on the basis of kinds of donor atoms,12 leading to a more thermodynamically stable product. Accordingly, it is without surprise that one discovers that, although various methods have been attempted for the displacement of macrocyclic ligand from the metal ion, e.g., cyanide ion13 sulfide ion14 and even solvated electrons5 rarely is the free macrocycle base released; either








the macrocyclic complex remains intact or the ligand is destroyed completely.

Many synthetic macrocyclic complexes have been prepared containing nitrogen as the donor atoms. Many of these are derived via a Schiff base condensation of a carbonyl-containing compound with an amine moiety. Such condensations are known to proceed by way of nucleophilic attack of the amine nitrogen on the carbon atom of the carbonyl group to yield a carbinolamine intermediate6 Normally, this reaction is acid-catalyzed. Thus coordination of the carbonyl oxygen to a positive center favors condensation by making the carbonyl carbon atom more susceptible to nucleophilic attack. Schiff base-containing macrocycles have been widely studied by Curtis17 and Busch18,19 and their co-workers.

Another type of macrocyclic complex is that obtained by the reaction of coordinated mercaptides with alkyl and aryl dihalides. Numerous examples of these complexes have been prepared by Busch and co-workers?0-22

Although a large-number of macrocyclic complexes have been prepared, free macrocyclic ligands are rarely isolated because of contamination by acyclic impurities. Accordingly macrocyclic ligands are prepared in the presence of a metal ion, such as Ni(II), Co(II), or Cu(Il).. The metal ion serves to orient the reacting species in such a way that cyclization occurs with minimal side reactions. This influence, exhibited by the metal ion, is known as the "coordinate template effect."

The particular systems chosen for these investigations result from the Schiff base condensation of 2,6-pyridinedicarboxaldehyde (PDC) with 1,8-diaminonaphthalene (DAN) and 2,6-diacetylpyridine (DAP) with the same diamine. In each instance a 20-membered macrocyclic ring is possibly formed, i.e., a potential hexadentate ligand. The macrocycles do not









satisfy the HUckel (4n + 2) criterion but do exhibit a large degree of conjugation (Figures 1 and 2). Often, in the case of large annulenes and their metal complexes, the HUckel rule is not stringent. In view of this fact and Framework Molecular Model representations, this author has concluded that the donor atoms can be planar and the R-system of the ligand delocalized.

The ligands of this study were conceived as modifications of

6,12,19,25-tetramethyl-7,11:20,24-dinitrilodibenzo[b,m][l,4,12,15]tetra6
azacyclodocosine, hereafter TMCD (Figure 3) in the attempt to promote metal-metal bonding in a slightly larger ring, 20-membered rather than 18-membered. Also, most synthetic macrocycles have been formulated from alkyl amines rather than aromatic amines. For broader scope and application to natural ystems, continued synthesis and study should envelop the incorporation of aromatic amines in the model systems.
















H3CT"


0
C N C- CH3 N NO

N NO,


II C,


N
0


II\-J C~cH3


6,1423,2 9-TETR A METHYL-13,9:2,24DINITRILO-9H,24HDINAPHTHO F, 6-bc:1'8,
[1,5,13,17] T ETRA A ZACYCLOTETRACOSI NE


Fig. 1 Structural formula of TMTC















N N
ON NO

H" N CH
0


13, 9:26, 24-DIN ITR I LO-9H, 24HDINAPHTHO [,6-bc:1 "8'-n_ [1,5,13,17]O
T ETRA A ZACYC LOTETRACOSNE


Fig. 2 Structural formula of DDnTC











CH3


H3C


'CH


6,12,19,2 5-TE TRA M ETHYL-711:20,24-DI NITRI LOD IBENZO[,m], 4,12,1 5] TETRAAZACYCLODOCOSINE


Fig. 3 Structural formula of TMCD















EXPERIMENTAL

Reagents

Unless otherwise specified all chemicals were commercially available as reagent grade and were used without further purification.

2,6-Pyridinedimethanol - Aldrich Chemical Co., Inc.; used as received; mp, 111-114o C.

2,6-Diacetylpyridine - Aldrich Chemical Co., Inc.; used as received (97%); mp, 78-79� C.

2,2'-Dimethoxypropane - Eastman Kodak Co,; used as received (practical); bp, 76-80� C.

Preparation of Starting Materials

Copper(II) perchlorate hexahydrate. The salt was prepared by reacting an aqueous slurry of reagent grade copper(ll) carbonate with a slight excess of 70% perchloric acid, concentrating the resulting solution on a hot plate, cooling, filtering and drying over P 4010 in vacuo. The salt contained some residual perchloric acid reflected by the acidity of its aqueous solutions.

2,6-Pyridinedicarboxaldehyde. The dialdehyde was prepared by employing a modification of the method of Papadopoulos et al.23 Sixty grams (0.69 mol) of freshly prepared manganese dioxide was suspended in 500 ml of chloroform containing 5.7 g (0.041 mol) of 2,6-pyridinedimethanol. The mixture was stirred at reflux for 5 hr. filtered, and the oxide washed with five 100-ml portions of ether. The filtrates were combined and evaporated under a stream of nitrogen gas. The residue was taken up in a minimum of solvent, 80% benzene-20% ethyl acetate. This solution

8








was placed on a silica gel (60-200 mesh) column and eluted with 500 ml of the solvent mixture; the middle portion (350 ml) was collected after discarding the first 125 ml. The desired fraction was blown dry under a stream of nitrogen gas to yield a white crystalline product. Yield:

1.75 g (32%); mp 115-122� C; lit. mp, 1240 a24

Fresh manganese dioxide. The oxide was prepared by a modification of the procedure of Sondheimer et al.25 A solution of 70 g (0.44 mol) potassium permanganate in 700 ml water was made acidic with 25 ml concentrated sulfuric acid. To this hot solution was added 100 g (0.59 mol) manganese sulfate monohydrate in 400 ml of water. The mixture was stirred for 2 hr. while maintaining a temperature of 900 C. An excess of permanganate was also maintained throughout the reaction.

The resulting solid was filtered by suction and washed five times with water by decantation. The solid was dried subsequently in an oven at 135' C for 24 hrs, pulverized, and stored over calcium chloride or P4 010

1,8-Diaminonaphthalene (Aldrich Chemical Co., Inc., 97%). The

diamine was recrystallized from a saturated solution of 95% ethanol-at room temperature. An excess of distilled water was added with stirring. The resulting mixture was an opaque flesh color. Formation of red-brown crystals was allowed overnight in the refrigerator. The product was collected and washed with water on a BUchner funnel. This procedure was repeated 2-3 times. The solid was dried over P 4010 in vacuo at room temperature. mp, 64-650 C; lit, mp, 660 C.

Procedures

Cu(D2A2)(NO3)2-4H20.* In a 500-ml flask, equipped with a reflux




*D2A2 = condensate of two moles DAP and two moles DAN








condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 2.0 g (0.012 mol) 2,6-diacetylpyridine. To this was added a filtered solution of 2.8 g (0.012 mol) hydrated copper(II) nitrate in 100 ml absolute methanol. Upon bringing the solution to reflux the color changed from light blue to light green. Added to the hot solution dropwise was 1.9 g (0.012 mol) 1,8-diaminonaphthalene in 100 ml absolute methanol. The mixture was refluxed for 48 hr. then cooled and filtered. The product was collected on a sintered glass filter, washed with several small portions of solvent and ether and dried over P 4010 in vacuo. Yield: 2.5 g (30%)

The complex is obtained as black powder containing some shiny black flakes. It is soluble in dimethylformamide, dimethylsulfoxide, and methanol to a limited degree. The complex was found to absorb atmospheric moisture reversibly, i.e., the water can be removed entirely in a drying pistol at 1000 over P 4010 in vacuo. After 24 hr in the drying pistol, the complex loses an average of 9.3% by weight. A comparable gain of weight in the atmosphere requires a 72 hr exposure.
Anal. Calcd for Cu(C38H30N6)(N3)24H 20 :C,-.54.94; H, 4.58; N, 13.49; Cu, 7.71. Found: C, 54.93; H, 4.23; N, 12.43; Cu, 8.29.

Cu2(D2A2)C 4*2H 20. In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying rube, containing 100 ml absolute methanol, was dissolved 1.63 g (0.010 mol) 2,6-diacetylpyridine. To this was added a solution of 2.70 g (0.020 mol) anhydrous copper(II) chloride in 100 ml absolute methanol. Upon bringing this solution to reflux, 1.98 g (0.012 mol) 1,8-diaminonaphthalene in 100 ml absolute methanol was added dropwise. After 48 hr reflux the hot mixture was filtered. The product was collected on a sintered glass filter, washed with several small








portions of solvent and ether and dried over P 4010 in vacuo. Yield:

4.4 g (100%)

The complex is obtained as a dark brown powder. It is slightly

soluble in pyridine, dimethylformamide, and dimethylsulfoxide. The complex was found to absorb atmospheric moisture. After 24 hr in a drying pistol at 1000 over P 4010 in vacuo the complex loses 6.3% by weight. A comparable gain of weight in the atmosphere requires the same amount of time.

Anal. Calcd for Cu2(C38H30N6)CI4 2H 20: C, 52.11; H, 3.89; N, 9.60; Cu, 14.51. Found: C, 51.90; H, 3.82; N, 9.42; Cu, 13.01.

Cu(D2A2)(Cl4)2-4H20. In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 1.63 g (0.010 mol) 2,6-diacetylpyridine. To this was added a solution of 7.41 g (0.020 mol) hydrated copper(II) perchlorate in 100 ml solvent. Upon bringing this solution to reflux, 1.58 g (0.010 mol) 1,8-diaminonaphthalene in 100 ml solvent was added dropwise. A dark solid formed within 24 hr. After 48 hr of reflux 150 ml solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of methanol and ether and dried over P 4010 in vacuo. Yield: 1.9 g (21%)

The complex is obtained as a black powder. It is slightly soluble in dimethylformamide, pyridine, dimethylsulfoxide, and acetonitrile. Over a period of several days the complex gains 7.66% in weight on exposure to the atmosphere. This moisture may be removed at 1000 over P 4010 in vacuo.

Anal. Calcd for Cu(C38H30N6)(ClO4)2-4H20: C, 50.39; H, 4.20; N, 9.28; Cu, 7.07. Found: C, 50.58; H, 3.68; N, 9.42; Cu, 7.12.








Co(P2A2)CI2"4H 0.* In a 500-ml flask, equipped with a reflux conC(2A2)C 2 2

denser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 1.35 g (0.010 mol) 2,6-pyridinedicarboxaldehyde. To this was added a solution of 2.60 g (0.020 mol) anhydrous cobalt(II) chloride in 100 ml of solvent. Upon bringing this solution to reflux, 1.58 g (0.010 mol) 1,8-diaminonaphthalene in 100 ml solvent was added dropwise. The mixture changed from a deep blue-violet color to black during the reaction. After 48 hr of reflux 150 ml of solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of methanol and ether and dried over P 4010 in vacuo. Yield: 2.0 g (56%)

The complex is obtained as a black powder. It is slightly soluble in dimethylformamide, pyridine, and dimethylsulfoxide. The complex is considered not to be hygroscopic.

Anal. Calcd for Co(C34H22N6)CI2 4H20: C, 56.98; H, 4.19; N, 11.73; Co, 8.24. Found: C, 57.30; H, 3.55; N, 11.50; Co, 8.33.

Cu2(P2A2)Cl 44H2 0. In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 absolute methanol, was dissolved 1.21 g (0.009 mol) 2,6-pyridinedicarboxaldehyde. To this was added a solution of 5.40 g (0.040 mol) anhydrous copper(II) chloride in 100 ml solvent. To the refluxing solution was added dropwise 1.58 g (0.010 mol) 1,8-diaminonaphthalene in 100 ml solvent. The mixture immediately changed from bright green to dark brown in color. After 48 hr of reflux, 150 ml solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small



*P2 A2 = condensate of two moles PDC and two moles DAN








portions of methanol and ether and dried over P 4010 in vacuo. Yield:

2.8 g (73%)

The complex is obtained as a black powder. It is slightly soluble in dimethylformamide and pyridine. The complex absorbs atmospheric moisture. After 24 hr in a drying pistol at 1000 over P 4010 in vacuo the complex loses 13.4% by weight.

Anal. Calcd for Cu(C34H22N6)C4-4H20: C, 47.72; H, 3.51; N, 9.82; Cu, 14.85. Found: C, 48.36, H, 3.17; N, 9.57; Cu, 14.85.

Cu(P2A2)(C104)2"2H20). In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 1.35 g (0.010 mol) 2,6-pyridinedicarboxaldehyde. To this was added a solution of 7.41 g (0.020 mol) hydrated copper(II) perchlorate in 100 ml of solvent. To the refluxing, light blue solution was added dropwise 1.58 g (0.010 mol) 1,8-diaminonaphthalene in 100 ml of solvent. The mixture slowly turned very dark brown in color. After 24 hr of reflux a dark solid was observed. After 48 hr 150 ml of solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of methanol and ether and dried over P 4010 in vacuo. Yield: 4.2 g (100%)

The complex is obtained as a black powder. It is slightly soluble in dimethylformamide, pyridine, and dimethylsulfoxide. Over a period of several days the complex gains 5.28% by weight while exposed to the atmosphere.

Anal. Calcd for Cu(C34H22N6)(C104)2-2H20: C, 50.18; H, 3.20; N, 10.33; Cu, 7.87. Found: C, 51.02; H, 3.10; N, 9.53, Cu, 8.04.

Cu(DAN)2(NO3)2. In a 250-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 75 ml absolute methanol, was







dissolved 1.58 g (0.010 mol) 1,8-diaminonaphthalene. To this was added a filtered solution of 1.17 g (0.005 mol) hydrated copper(II) nitrate in 75 ml solvent. The mixture refluxed for 2 hr and was filtered hot. The filtrate was deep purple in color. The product was collected on a sintered glass filter, washed with several small portions of solvent and ether and :dried over P4010 in vacuo. Yield: 2.3 g (46%)

The complex is obtained as a black-green powder. It is slightly soluble in acetone, dimethylformamide, pyridine and dimethylsulfoxide. The complex is not considered to be hygroscopic; over a period of a week, one sample gained 0.59% by weight.

Anal. Calcd for Cu(C10H10N2)(N03)2: C, 47.62; H, 3.97; N, 16.67; Cu, 12.70. Found: C, 47.70; H, 4.09; N, 16,73; Cu, 12.68.

Attempted preparation of DDnTC. In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 200 ml absolute methanol, was dissolved 1.35 g (0.010 mol) 2,6-pyridinedicarboxaldehyde. The solution was brought to reflux and 1.98 g (0.012 mol) 1,8-diaminonaphthalene in 100 ml solvent was added dropwise. After 6 hr of reflux, a brick-red solid was observed. The mixture was refluxed for 48 hr, after which 100 ml of solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of solvent and ether; and dried over P4010 in vacuo. Yield: 1.0 g (36%)

The ligand is obtained as a red-orange powder which darkens to winered on standing. It is slightly soluble in ethyl acetate, acetone, chloroform, methylene chloride, dimethylformamide, pyridine, and dimethylsulfoxide.

Anal. Calcd for C34H22H6*2H20: C, 74.18; R, 4.73; N, 15.27. Found: C, 75.12; H, 5.00; N, 15.18.








Attempted preparation of TMTC. In a 1-liter flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 400 ml n-butanol, was dissolved 6.52 g (0.040 mol) 2,6-diacetylpyridine. This solution was brought to reflux and 2 drops of concentrated sulfuric acid added. A solution of 7.90 g (0.050 mol) 1,8-diaminonaphthalene in 400 ml n-butanol was added dropwise to the hot mixture. After 48 hr of reflux 300-400 ml of solvent was distilled off. No solid was observed in the dark brown reaction mixture. A small portion (10-15 ml) of the mixture was combined with 75 ml hexane and allowed to stand for 30 min. A light tan-grey solid was collected on a BUchner funnel, washed with a small portion of hexane, and sucked dry. mp, 105-6o C with decomposition.

The solid is slightly soluble in chloroform, dimethylformamide, methanol, methylene chloride, and ethanol.

Anal. Calcd for C38H30N6: C, 80.00; H, 5.26; N, 17.74. Found: C, 76.57; H, 6.10; N, 14.41.

Elemental analysis. All CHN microanalyses were performed by Atlantic Microlab, Inc., Atlanta, Ga.

Metal analysis. Metal analyses were obtained using a Perkin-Elmer

Model 290-B atomic absorption spectrometer. All samples were analyzed in aqueous solution after digesting almost to dryness in 20 ml 1:1 mixture of concentrated nitric and perchloric acids. Calibration curves were obtained by employing aqueous solutions prepared from certified atomic absorption reference solutions (Fisher Scientific Co.).

Apparatus

Magnet. The magnetic susceptibilities were determined by the Guoy method.26 The magnet used was a Varian Associates Model V-40041 equipped with four-inch cylindrical pole pieces, separated by an air gap of 2-1/4 inches. A Varian Associates Model V-2501-A current regulator was used to





16

provide a constant current (+ ixl0-3 amp). The maximum field strength attained was 6.75 x 103 gauss. The magnetic field was calibrated by using mercury(II) tetrathiocyanatocobaltate(II).27 The current regulator was powered by a Varian Associates Model V-2300-A power supply. Temperatures between 950 and 350� K could be maintained within +0.1 degree as determined by a platinum resistance thermometer.

Cryostat and temperature control. The cryostat and temperature con28,29
trol apparatus used were of the basic design of Figgis and Nyholm. Temperatures between 950 and 350* K could be maintained with less than +0.1 degree fluctuation.

Sample tube. The sample tube was made of a cylindrical piece of

quartz approximately 3.5 mm inside diameter and 17 cm in length and sealed at one end. Approximately 16 cm was used for containing the sample volume. It was suspended in the cryostat from a semi-micro balance by a gold chain attached to a tapered Teflon plug, The diamagnetic correction of the tube was measured as-a function of temperature.

Balance. A Mettler Model B-6 semi-micro balance of 0.01 mg sensitivity was used to measure the force exerted by the magnetic field upon the sample.

Spectrometers. The solution visible and ultraviolet electronic

spectra were obtained by using a Cary Model 15 recording spectrophotometer. The solid state diffuse reflectance spectra were obtained by using a Cary Model 1411 Diffuse Reflectance Accessory in conjunction with a Cary Model 14 recording spectrophotometer. Magnesium carbonate was employed as reference material.

Infrared spectra were obtained using a Perkin-Elmer Corp. Model 137B NaCl prism and 237B grating spectrophotometers. Also employed was a Beckman Model IR-10 grating spectrophotometer. All spectra were calibrated








with polystyrene. The pressed KBr pellet technique was used.
Hnmr spectra were measured on Varian Associates Models A60-A and XL-100 nuclear magnetic resonance spectrometers (TMS reference).

Electron spin resonance spectra were obtained on microcrystalline samples using the Varian Associates Model E-3 recording spectrometer.

Mass spectra were obtained on AEI Scientific Apparatus Model MS-30

double-beam, double-focusing mass spectrometer equipped with a DS-30 data system. Each solid sample was run by direct introduction probe. Probe temperatures ranged from 2000 to 340� C. Dr. R.W. King kindly performed these spectral analyses.

Melting point apparatus. A Thomas Hoover "Uni-Melt" capillary melting point apparatus was used; the temperatures are uncorrected.

Conductance apparatus. Conductances were measured using an Industrial Instruments, Inc., Model RC-18 Conductivity Bridge and a cell with a constant of 1.464 cm-. A constant temperature of 250 C � 0.02 was maintained by the use of a water bath, regulated by a Sargent Thermonitor, Model SW. The units of conductance obtained were mho cm-1, specific


conductance.















RESULTS AND DISCUSSION

Ligands

Attempts were made to prepare the macrocycles DDnTC and TMTC using procedures generally followed for such ligands.1,18-20 Reaction mixtures were dilute in reactants in order to avoid the formation of oligomeric side-products. Several dehydrating and azeotroping solvents were used to ensure completion of reaction, i.e., the removal of product water. Variations in reactant mole ratios (other than 1:1) were employed also.

The initial attempt to produce TMTC proved to be the most encouraging; however, cyclization did not occur, Following the procedure of Stotz,6 a slight excess of diamine was added to a dilute refluxing solution of DAP in n-butyl alcohol. In view of the fact that Schiff base condensations are frequently acid catalyzed30 and because of the lesser relative reactivity of a ketone as compared to an aldehyde, several drops of concentrated sulfuric acid were used as catalyst. Because of the insolubility of macrocycles in most organic solvents, solid product was expected to begin forming early in the course of reaction. However, after 48 hr of reflux no solid was observed. Upon addition of excess hexane to a small portion of reaction mixture, a product precipitated which proved to be the impure acyclic trimer consisting of two moles of DAN condensed with one mole of DAP.

The mass spectrum of this impure product which decomposes at 105-i06* C contained the parent ion at m/e 443, the molecular weight of the trimer. Other lighter fragments in the mass spectrum could not be identified








(Table 1-A). The electronic and infrared spectra support the conclusion that the product isolated contained some trimer (vide infra).

Using 2,2'-dimethoxypropane as solvent and two drops of concentrated sulfuric acid, a small portion of greyish, light brown solid was obtained. This material decomposed above 1900 C. Elemental analysis was inconsistent with either the formation of the trimer or the macrocycle.

The reactants were also refluxed in n-butyl alcohol using ammonium chloride as catalyst. The ammonium ion is a much weaker acid than the hydronium ion and should not react readily with the diamine. A small amount of dark brown solid was formed but elemental analysis data do not agree with the calculated values for either the trimer or the macrocycle. The color of the reaction mixture was identical to that of the n-butyl alcohol preparation.

Thin-layer chromatography on alumina and silica gel did not show any separation of components in the above solids. Various polar to nonpolar organic solvents were employed as eluants.

A small amount of the impure trimer was dissolved and refluxed in n-butyl alcohol. Several drops of concentrated sulfuric acid were added with slow addition of a dilute DAP solution in the hope that cyclization would occur. But, after 23 hr of reflux, no solid had formed.

An attempt was made to deuterate the impure trimer to show evidence of a primary amine. A sample of the trimer was dissolved in a small portion of methanol and several drops of D 20 added. This solution was evaporated in a vacuum desiccator and a KBr pellet of the product was prepared. The infrared spectrum showed no observable N-D or 0-D absorptions.

The 1H nmr spectrum of a saturated solution of the impure trimer in CDC13 was attempted; however, solute concentration is so low that no








resonances could be observed. The spectrum of saturated solution of the trimer in d6-DMSO gave resonances at 66.55-7.15 (broad multiplet) and 61.55 (singlet). The former resonance is indicative of the overlap of pyridine and naphthalene protons and the latter of the methyl groups. Integration revealed a ratio of 2:1, respectively, as compared to the 15:6 ratio expected for the pure trimer.

In the attempt to produce the free ligand, DDnTC, methanol provedto be the best reaction medium. No acid catalyst was used in the reactions. N-butyl alcohol and 2,2'-dimethoxypropane were also employed as reaction media. In all cases the solid formed was red-brown to brick-red in color. Elemental analyses showed consistently low carbon and nitrogen percentages except for the product separated from methanolic mixtures.

Similarities of the compounds include not only limited solubilities

but also relative high decomposition temperature. At atmospheric pressure the melting point is above 3000 C. Upon examination of the sample after the mass spectrum was taken, it was noted that some melting and sublimation had occurred (340* C and 10-6 torr). A small portion of the impure ligand dissolved in hot chloroform and was recrystalltzed by dropwise addition of cold cyclohexane. This procedure did not improve the purity of the macrocycle.

An attempt was made to prepare the hydrogen chloride adduct in the hope of obtaining a single product. A small amount of the impure ligand was dissolved in hot chloroform and a moderate stream of dry hydrogen chloride was passed through the solution for 15 min. After 1 min the deep red solution turned brown and solid formed. The dried product was a tan powder which had an infrared spectrum only very slightly different than that of the free ligand, i.e., a broader absorption between 2600 and 3300 cm-I.







However, again elemental analysis showed the presence of a mixture of products. The adduct placed in distilled water changed the pH gradually from 6 to 2 over a period of five days; a similar change of pH was noted for the TMCD adduct.6 The excess solid in contact with the saturated solution did not change color during this time.

The mass spectrum of the original reaction product is characteristic of a mixture rather than the free ligand sought (DDnTC). A small peak was observed at m/e 514, which would represent the parent ion of the free macrocycle; but, a more intense peak at higher m/e was also noted. The only other identifiable peak was at m/e 410. This would be the parent ion minus a (C5H3N)CHN fragment (Table 2-A).

The 1H nmr of the impure ligand was attempted in CDC13 and d6-DMSO. The saturated d6-DMSO solution gave absorptions at 66.6-7.05 (broad multiplet) and at 65.55 (broad singlet). The former is indicative of the overlap of pyridine and naphthalene protons, Integration revealed a ratio. of 26:3, respectively, a disproportionate ratio in view of the possible overlap and presence of impurities. The ligand was not sufficiently soluble in CDC13 to give any observable absorptions.

Electronic and infrared spectral data support the inclusion of some macrocycle in the impure product (vide infra). Thin-layer chromatography shows no separation into components on either alumina or silica gel even though various-polar and nonpolar organic solvents were used as eluants.

The unsuccessful attempts to isolate the free macrocycles are not

1 14,81
considered unusual. '418,19 Because the initial intent was to investigate the complexes of these macrocycles, the impure "free" ligands were not further characterized.

Because the attempts were unsuccessful in isolating the free macrocycle








and because of the lack of proof (vide infra) for macrocyclic structure in the complexes, D2A2 is used as the general notation for the condensation product of two moles of DAP with two moles of DAN and P2A2 for two moles of PDC condensed with two moles of DAN.

Complexes

For the purpose of brevity the seven complexes will be hereafter denoted as follows (see also Figures 1 and 2):

Compound Number

Cu(D2A2) (NO3)2" 4H20 I Cu2 (D2A2)Cl4 2H2 0 II

Cu(D2A2) (ClO4) 2" 4H20 III Co(P2A2)Cl2" 4H20 IV Cu2(P2A2)C14"4H20 V Cu(P2A2) (ClO4)2 21 20 VI Cu (DAN) 2 (NO3) 2 VII

Synthesis of each complex followed the straightf6rward template method.' All attempts made to produce each of the complexes via the direct method were unsuccessful. First row transition metals, in general, form very stable amine complexes, hence the order of addition of reactants to the reaction mixture is important. Accordingly, amine solutions were added to either ketone- or aldehyde-metal salt solutions. Anhydrous metbanolproved to be a satisfactory solvent in that all starting materials possessed high solubility in it and the products were of low solubility.

The strategy of the syntheses was to choose counterions which generally do not coordinate with the central ion and yet give reacting salts which are soluble in organic solvents. For these reasons the








nitrate, chloride and perchlorate salts of several metals were used. The only isolable compounds which appeared to be macrocyclic complexes were those of Cu(II) and one of Co(II), although salts of nickel(II), manganese(II) and zinc(II) also were employed in the attempts. The analytical results for the nickel(II) and manganese(II) products did not fit the formation of complexes. In the zinc-containing system, the only compound isolated appeared to be the bis(diamine) complex. These results are not surprising in view of the selectivity of the "hole" of the macrocycle in relation to metal ion size31 (Table 3-A). Also many first row transition metals form thermodynamically stable amine complexes.

An unusual phenomenon for some macrocycles is the difficulty encountered in obtaining elemental analyses which agree precisely with reasonable formulation of the complexes.32 Because of the unusual stability of these complexes, accurate elemental percentages may be impossible to obtain. Complex I repeatedly gave low nitrogen analyses (ca. 1%). Stotz6 also noted this in his related complex prepared by the in situ method (ca. 1.5% low). But these complexes are nitrates. Therefore, it may be that, since both of these compounds contain nitrogen in the counterion as well as in the ligand, analytical procedures presently used may be inadequate for accurate nitrogen determinations.

. Complex II presents consistently low copper analyses (ca. 1.5%).

Copper standards were checked several times; furthermore, numerous copper determinations, of good precision, were made. This difficulty was considered to be a characteristic of the complex because analyses for carbon, hydrogen, and nitrogen are in good agreement with the theoretical amounts. At this point the author cannot suggest a reasonable solution to the problem.








Several other complexes show small inconsistencies in carbon and hydrogen analyses. But, these problems appear to go hand-in-hand with compounds of very high stability.2 Because stabilities of macrocycle complexes are known to be abnormally high,l more than 8 hr was required for the complete digestion of samples in preparation for metal analyses. The extreme stability of all seven complexes is further noted in their inactivity towards hydrogen sulfide in dilute pyridine solutions. Metal sulfide did not precipitate although mixtures were allowed to stand for a period of several days.

Another factor which may lead to the troublesome results in elemental analyses is that all the complexes, with the exception of IV, will dehydrate and hydrate reversibly. During microanalytical procedures rapid weight gain fwas reported by Atlantic Microlab, Inc.; this was substantiated by the author after water was reversibly removed at elevated temperatures over P 4010 in vacuo and gained under atmospheric conditions. Small amounts of samples were placed in tared vessels and subsequently dried, weighed, exposed to atmospheric moisture and weighed again. Anhydrous products appear to be formed in absolute methanol which hydrate, Water in analyzed Water gained by Compound product anhydrous form

I 4H20 4H20

II 2H20 3H20

III 4H20 4H20

IV 4H2 0

V 4H20 4H20
VI 2H20 3H20


with ease, on work-up in the atmosphere. One cannot definitively differentiate between water in lattice sites, water coordinated to the metal








ion, or water included as carbinolamine linkage(s).33 The 1H nmr spectra are not obtainable because of low solubilities and the presence of paramagnetic ions. Infrared spectra are inconclusive because 0-H absorptions of the three possibilities occur in the same high energy infrared region and because metal-oxygen absorptions are generally present in broad regions where aromatic peaks also occur.34 All measurements were taken on either completely hydrated or dehydrated samples.

Since it is highly probable that the metal ions in these complexes

are coordinated in an approximately planar arrangement of nitrogen atoms, it was desired to prepare a simpler model complex for comparison (Figure 4). Compound VII was selected for this purpose. The preparation of VII was completely straightforward and rapid; it is not hygroscopic. Spectral comparisons will be noted in the next sections. The preparation of the analogous cobalt(II) complex was attempted using chloride, nitrate, and perchlorate as counterions. The bis(diamine) complex of Co(II) was not successfully prepared in these trials, which were performed in absolute methanol.

Infrared Spectra

The infrared spectra of 2,6-pyridinedicarboxaldehyde,

1,8-diaminonaphthalene, and impure DDnTC are presented in Figure 5. The noteworthy differences between the infrared spectra of the ligand and its precursors are the disappearance of the characteristic carbonyl band of the dialdehyde at 1700 cm- and of the primary amine bands of the diamine lying near 3300 cm-I and 900 cm-I. With the exception of the characteristic carbonyl and primary amine absorptions, the spectrum of the macrocycle is essentially a composite of the spectra of the parent compounds.












H H
I 12

o N
NJ
onI
H H
2 2


(NO3)2


BIS(1,8 -DIAMINONAPHTHALENE)COPPER(-F) NITRATE


Fig. 4 .Structural formula of Cu(DAN)2(N03)2












PDC


DDnTC


00 2000 i500 1200 1000 900 800 Frequency (cm-1)

Fig. 5 Infrared spectra of DDnTC and its precursors


Ud 41I 4J










DJ


4000 ?0


LN








The broad band at 1600 cm-I has been assigned to the imine

stretching mode, and 0-H deformation, and higher frequency ring vibrations. An imine stretching mode in this region is characteristic of an imine conjugated with an aromatic system.35 Isolated imines absorb at higher energies, ca. 1670 cm-I.

Unsubstituted, uncoordinated pyridine has characteristic absorptions lying at 1580, 1570 and 1485 cml.36 The two bands at higher energy arise from the interaction between the C=C and C=N vibrations of the ring.

The infrared spectra of 2,6-diacetylpyridine, 1,8-diaminonaphthalene, and the impure trimer of TMTC are presented in Figure 6. It is apparent that primary amine groups are contained in the trimer because of the presences of the band at 3250 cm-I. The medium-intense absorption at 2900 cm-I, which also appears in the Sadtler spectrum, may be due to an impurity in the DAN. The relative intensity of this absorption decreases after recrystallization of DAN; the freshly recrystallized diamine oxides very rapidly even in the absence of light. The carbonyl band of the diketone at 1700 cm-I has all but disappeared. The strong absorption at 1600 cm-I contains the imine stretching mode. With these exceptions, the spectrum of the trimer is a composite of the absorption bands of its precursors.

The infrared spectra of all the complexes investigated contain bands which are much broader than those of the respective precursors. This is typical of metal complexes prepared as wafers. All samples exhibit a disappearance of the primary amine bands at 3300 and 900 cm-I and of the carbonyl band at 1700 cm-1. Water vibration, be it lattice or coordinated, appears as a broad band centered at 3100-3400 cm-l.34 Each












DAP


Trimer of TMTC

Cd


4-,

Ca
14








DAN















4000 s000 2000 j500 )Z jOO 100 Q,0 80O Frequency (cm-1) Fig. 6 Infrared spectra of impure trimer of TNTC and its precursors








displays a strong, broad absorption centered at approximately 1600 cm-I, containing the imine stretching mode, 0-H deformation, and high frequency ring vibrations.

Complex I (Figure 7) displays bands at 1390 and 825 cm-I, both

typical of ionic nitrate.37 The former is attributed to the asymmetric stretching mode of N03- and the latter to the deformation mode of the ion.

Complexes II, IV, and V (Figure 8-10), metal chloride salts, have infrared spectra which are virtually identical. The imine and aromatic absorption, which are broad, lie at the expected frequencies.

Complexes III and VI (Figure 11 and 12) are both perchlorate salts. The intense, broad, semistructured band centered at 1100 cm-I and a peak at approximately 625 cm-1 are typical of ionic perchlorate.38 The former absorption represents the asymmetric stretching modes, and the latter an asymmetric bending mode. Also, an infrared-forbidden band at 930 cm-1 (symmetric stretching mode) is often observed as a weak absorption because of the reduction of the T symmetry of the anion within the lattice; it is not observed in these compounds. The broad and composite imine and aromatic absorption envelope is the same as that noted in the spectra discussed earlier.

Complex VII (Figure 13) was synthesized as a comparison to the above complexes. Primary amine bands are evident at 3300, 3100 and 1635 cm-I. A strong absorption at 1390 cm-1 is characteristic of ionic nitrate. Other bands are similar to the absorptions of the precursor.

A Framework Molecular Model representation of each macrocycle presents the pyridine nitrogens at significantly greater distance from the metal ion in comparison to the imine nitrogens. Upon complexation for a
































C
.
4-J .4

0 0 'U














4000 3000


Fig. 7 Infrared spectrum of Cu(D'A)(N03)2"4H20 (I)


2000 1500 1200 1000 900 800 700

Frequency (cm-I)







































3000 2000 1500 1200 1 Frequency (cm-l)
Fig. 8 Infrared spectrum of Cu2(D2A2)CI4.2H20 (II)


000


800


)0






































1500


1200


900 800 700


Frequency (cm-1)


Fig. 9 Infrared spectrum of Co(P2A 2)C2.4H20 (IV)


3000





































4000 3000 2000 1500 1200 1000 900 800 700 Frequency (cm-I)

Fig. 10 Infrared spectrum of Cu2(P2A2 )C14.4H20 (V)




















N




E-4




L







4000


900 800 700


3000 2000 1500 1200 1000 Frequency (cm-1)

Fig. 11 Infrared spectrum of Cu(D2A )(C104)2"4H20 (III)




































4000 3000 2000 1500 1200 1000 900 800 700 Frequency (cm-l)

Fig. 12 Infrared spectrum of Cu(P2A2 )(C104)2"2H20 (VI)



























CU

4

-W















4000 3000 2000 1500 1200 10 Frequency (cm-1) Fig. 13 Infrared spectrum of Cu(DAN)2(N03)2 (VII)







-1
given metal ion, the pyridine absorptions at 604 and 405 cm are sensitive to the stereochemistry of a complex. These absorptions are shifted to higher frequencies (ca. 50 cm- ) upon coordination.7 Both bands are weak in uncoordinated pyridine but have proved tobe of diagnostic value 33
in determining coordination sites. Spectra of each complex revealed no absorptions at the expected frequencies.

If the nitrogens of the pyridine rings are uncoordinated, a strong Lewis acid such as hydrogen chloride should form an adduct at these sites. A stream of dry hydrogen chloride was passed through a weighed sample of Complex I, followed by a gentle flow of nitrogen gas. Upon weighing it was found that the sample had gained a small amount of weight but elemental analysis revealed a nonintegral number of moles of hydrogen chloride added, viz., between three and four. Although the number of moles of hydrogen chloride added is puzzling, the fact that the hydrogen chloride is absorbed is certainly taken to mean that there are free base sites available. The identification of these base sites could not be accomplished because the infrared spectrum of the adduct was essentially identical to that of the precursor.

Electronic Spectra

The electronic spectra of the proposed macrocycles are presented in Figures 14 and 15 with absorption maxima- listed in Table 1. The spectrum of impure DDnTC in acetonitrile is a composite of its precursors. Absorptions between 50.0 and 40.0 kK encompass the range of the E and K 39
bands observed in benzenoid compounds. These absorptions are assigned to n to N* transitions. The band at 28.6 kK, attributed to naphthalene, has shifted to slightly lower energy as compared to pure DAN; this would be expected with an electron-donating substituent and the fusion of an additional ring.


















d 1A
44

0


.J









I I

200 300 400
Wavelength-(nm) Fig.. 14 Electronic spectrum of impure trimer of TMTC

























o
-4
0


















200 300 400
Wavelength, ,.(rm) Figj 15 Electronic spectrum of DDnTC








Table 1


Near Infrared, Visible and Ultraviolet Spec und Solvent Wavelength, nm

CH3 CN 200 225


235 350

203 (sh)

237 350

Strong
absorption below
220 260 267 211 237 275 231 334 270 500 363 480

225(sh)

270 600 360 480


Compo DDnTC


tral Data Frequency, kK

50.0 44.4 42.5 28.6 49.3 42.2 28.6



45.4 38.5 37.5 47.4 42.2 36.4 43.3 29.9 37.0 20.0 27.5 20.8 44.4 37.0 16.7 27.8 20.8


Trimer of
TMTC


MeOH


PDC


MeOR


DAP


MeOH


DAN


MeOH


MeOH Solid Solid


III


CH3 CN Solid Solid








Table 1 (continued)


Compound

V VI


Solvent Solid CH 3CN


Solid


Solid


Wavelength, nm

500

215(sh)

230 273 475 600 320 490

525(sh)

570 610


Frequency, kU

20.0 46.5 43.5 35.4 21.1 16.7 31.3 20.4 19.0 17.5 16.4








The impure trimer of TMTC in methanol also shows similarities in its spectrum as compared with the precursors. Intense bands lying at 49.3 and 42.2 kK are typical of E and K bands in aromatic m6ieties. -The naphthalene absorption is again observed at 28.6 kK. Bathochromic shifts are expected when a substituent is electron-donating or capable of conjugation.39 Both of the proposed macrocycles as well as the impure trimer have increased conjugation in comparison to the reactants.

The electronic spectra of the copper(II) complexes in this study, in general, present over a range of energy from near infrared to the ultraviolet, only one band; in this region diffuse reflectance spectra were required because of the extremely low solubilities of the complexes. Each of the bands has much the same appearance, indicating some common factor in origin., Copper(II) complexes are known to exhibit a wide range of possible stereochemistries. Considering the ability of the two macrocycles to coordinate four to six nitrogens, it would be likely to limit the possible stereochemistries to octahedral, square coplanar, tetrahedral, or square pyramidal, including anion coordination. Because macrocycles similar to DDnTC and TMTC exhibit the tendency to remain planar upon complexation, the author will eliminate the tetrahedral arrangement (vide infra). Hathaway40 has summarized the correlation of electronic properties and stereochemistry of numerous {CuN4-6} chromophores. He has shown that square coplanar {CuN4} species generally absorb between 18 and 20 kK. The range of absorption of species studied is between 17 and 21 kK. The [Cu(en)2] 2+ cation is often used as a model complex.41 The approximately tetragonal ligand environment of [Cu(en)2 ]2+ is D4h. Its complexes possess a magnetic moment between 1.80 and 1.90 B. M. This is typical of an orbitally non-degenerate B-type ground state.








It is rare that copper(II) complexes are found in pure octahedral symmetry. In a cubic field a d9 ion possesses an 2E ground term which
-- g
is highly susceptible to Jahn-Teller configurational instability.42 Upon tetragonal distortion to an approximate D4h symmetry the 2E term
g
splits into 2B and 2A terms, and the 2T term into 2B and 2E.
ig lg 2g 2g g
terms. The broad absorption envelopes seen in the copper(II) complexes following can, indeed, be related to symmetry and the d-d transitions centered on the metal ion.

In principle the maximum number of d-d transitons in the electronic spectrum of the copper(II) complex, {CuN4-6}, can be as many as four (Figure 16); in practice, very few complexes give any indication of more than two bands. Many spectra show a main band with a low frequency shoulder only partially resolved. The effect of coordination number upon energies of d-d transitions of copper(ll) complexes is complicated. In practice, it is the degree of tetragonal distortion which has the major effect in determining the energy of bands in the d-d spectrum of a copper(II) complex. As tetragonal distortion increases from regular octahedral to tetrahedral stereochemistry towards the square coplanar one, the center of gravity of d-d transitions moves to higher energy.40

In such cases of tetragonal distortion the precise energy levels in D4h synmetryare uncertain. Depending upon the degree of tetragonal distortion present, three possible energy level sequences may arise:

a) B < A < B < 2E (octahedral)
a)2lg <2lg <22g < g

b) Blg < B2g < A 19 < E

c) B < B g< E < A (square planar)
c)2lg <22g < g 2lg

The 2B + 2 g transition is believed to be the lowest energy transition and possibly reflects the degree of distortion as the shoulder on the low












2Eg
I


2T2g


\ 2B2g


\ \ 2B2g


2Blg


2D


2Alg 2Ag_


\ 2Eg
\ j


\ 2Blg 2A


Free ion


Oh
(octahedral)


D4h
(square planar) or tetragonal


Fig. 16 The splitting of the 2D term of the copper(l) ion in ligand fields of different
symmetries.


D2h
(rhombic)


2 B3g








frequency side of the absorption envelope. Polarization spectra of single crystals have tentatively verified this.41

Thus, the broad absorptions presented by the copper(II) complexes herein are actually envelopes including three absorptions of similar energies, indicative of the {CuN 4 chromophore in the presence of a tetragonal distortion.

Complex I presents a broad maximum at 20.0 kK (Figure 17). The

higher energy absorptions of the macrocycle have shifted to a less intense band lying at 37.0 kK. Bathochromic shifts of ligand absorptions are commonly observed as metal ions complex and participate in the i-system.33 The DAN absorption, shifted to lower energy, is probably hidden by the broad envelope centered at 20.0 kK.

Complex II has extremely low solubility in solvents transparent in the ultraviolet region. The reflectance spectrum of II (Figure 18) presents two broad absorptions lying at 27.5 kK and 20.8 kK; the former is attributed to charge-transfer absorption and the latter to metal d-d transitions.

The solution spectrum of complex III (Figure 19) shares some similarity with the impure trimer in the ultraviolet region in that bands lying at 42.2 and 28.6 kK have been shifted to longer wavelength upon complexation. The broad envelope centered at 16.7 kK can be assigned to metal d-d transitions.

Like its analogue, II, complex V (Figure 20) has low solubility in solvents transparent in the ultraviolet region. The only feature in the visible region is the broad absorption attributed to d-d transitions on the metal ion.

The solution spectrum of complex VI (Figure 21) presents again a

bathochromic shift of ligand absorption in the ultraviolet region. The












(" )
(-)


Solution Solid


-


800


Wavelength (nm)


Fig; 17 Electronic spectrum of Cu(D2A ) (N03)2-4H20 (I)


900


1000


1100
























(1)


O






I4














300 400 500 600 700 800 900 1000 Wavelength (nm) Fig. 18 Diffuse reflectance spectrum of Cu2(D2A2)Cl4"2H20 (II)













( )
(-)


Solution Solid


I I I


500


Fig. 19 Electronic spectrum


600 700 Wavelength (nm) of Cu(D2A2)(C104)


800


2"4H20 (III)


200


900


1000


1100





















W
o
0


Cl)















300 400 500 600 700 800
Wavelength (nm) Fig. 20 Diffuse reflectance spectrum of Cu2('P2A2 )C144H20 (V)


900. 1000











( ,) (----)


Solution Solid


-


N


- - -s


400


Fig. 21 Electronic


500 600 700 800 Wavelength (nm)
spectrum of Cu(P2A2')(Cl04)2'2H20 (VI)


200


900


1000


1100








bands lying at 50.5 and 44.4 kK in the macrocycle are shifted to 46.5 and 43.5 kK, respectively, while the band lying at 42.5 kK is now found at 35.4 kK. In the reflectance spectrum the shoulder at 21.1 kK may be attributed either to charge transfer absorption or to a metal d-d transition. The broad envelope centered at 16.7 kK can be assigned to d-d transitions on the metal ion.

As a reference compound, complex VII was synthesized as a model for determining the stereochemistry of the above complexes. It was observed that VII forms suspensions with solvents commonly used for study in the ultraviolet region. The reflectance spectrum of VII (Figure 22) presents several more distinct absorptions than the macrocyclic complexes. The high energy absorptions at 20.4 and 31.3 kK can be assigned to chargetransfer absorptions. In the spectral study of [Cu(en)2](N03)2, Hathaway et al.41 have reported absorption at 19.7, 17.9, and 14.1 kK, the latter being a shoulder. The absorptions of VII at 19.0, 17.5, and 16.4 kK are very similar, realizing crystal field splitting parameters of the two 43
diamines, no doubt, differ slightly. Yamada and Tsuchida. in a previous work reported a band at 16.7 kK for the [Cu(en)2] 2+ ion. The three absorptions can be assigned to the transions 2B 4 2 E_, 2B 2 B lg g lg 2g
and 2 ABg + 2Alg, in decreasing order of energies. The model cation is known to have the four nitrogens in a plane about the metal ion with two other groups possibly coordinated along the tetragonal axis.43,44

Although, visible spectral characteristics have been discussed in

terms of d-d transitions, there is also the possibility of superposition of charge-transfer absorptions on d-d absorptions. The types of absorptions cannot be separated in the absence of solution spectra in the visible and far infrared regions. All solid spectra were extended





























w
(31

0 Ca



4j







CaC






SI I . ,1
1

300 400 500 600 700 800 900 1000
Wavelength.(nm) Fig. 22 Diffuse reflectance spectrum of Cu(DAN)2(N03)2 (VII)








to 1900 nm (5.3 kK); no absorptions were observed in the far infrared region. "Solution" spectra also are actually composites of solution and solid state spectra because of the formation of colloidal dispersions. Relative intensities of solution and diffuse reflectance spectra are only approximate.

The visible spectrum of Complex IV is presented in Figure 23. 'Only two broad absorptions appear; that at 27.8 kK can be assigned to chargetransfer absorption and the very broad band centered at 20.8 kK most likely arises from a d-d transition(s) of the cobalt(II). One would wish to relate the d-d transitions centered upon the metal ion to the symmetry of the complex and to the crystal field splitting parameter, Dq. But, the stereochemistry of the metal ion must be approached somewhat indirectly. A tetrahedral environment about cobalt(II) should give bands in the near infrared region, i.e., between 5 and 6 kK, which might be assigned to the 4A2 4T2(F) transition,45 and an intense multicomponent band associated with the 4A2 * 4T1(P) transition lying between 12.5 and 16.6 kK.46 In the spectrum of IV neither of these absorptions is seen.

In the last decade numerous high-spin, pentacoordinate complexes of cobalt(II) have been reported.33 These compounds generally exhibit four principal bands lying at 5.6 to 6.0 kK, 12 to 12.6 kK, 15.5 to 16.5 kK, and 19 to 20.5 kK.47 But, this pattern is not observed for IV. Because the complex must be considered spin-free (vide infra) and because only spin-paired, square planar cobalt(II) species containing four Co-N bonds have been isolated, this stereochemistry is ruled out. Moreover, the low energy absorption characteristics of square planar complexes are not observed in this case.48 Rather, the broad absorption of IV centered at 20 kK is characteristic of distorted octahedrally coordinated cobalt(II).49-51























(1)
U

0






-1












l I I I I 300 400 500 600 700 800 900 1000 Wavelength (nm) Fig. 23 Diffuse reflectance spectrum of Co(P2A2)C12.4H20 (IV) un









This envelope is generally considered to contain two spin-allowed d-d transitions, viz., those from the ground term, 4T1g(F), and the terms 4A2g(F) and 4T g(F).49 A small shoulder sometimes is observed on the high frequency side as a consequence of spin-orbit coupling serving to lift the degeneracy of the 4Tg (P) term. The weak 4Tg (F) -* 4T2g(F) transition is frequently found between 8 and 9 kK. This absorption is not observed in this case. The appearance of this band leads to some confusion in the literature, which is most probably associated with the fact that it is invariably weak (6<10).52 The complexes [Co(NH 3)62+
2+ 2+
and [Co(en)31 , as compared to [Co(H20)6]2 , do not exhibit this absorption.53 However, Ballhausen and J~rgensen54 have reported the transition for the amine complexes to lie at 9.0 and 9.8 kK, respectively.

Assuming the transition occurring at 20.8 kK is from the ground term, 4T g(F),. to the 4A2g term, the crystal field splitting parameter, Dq, for [Co(DDnTC)]2+ is 1156 cm-1 The Dq values of [Co(en)] 2+ and [Co(NH3)6]2+ for the same transition should be slightly smaller in the absence of conjugation in the ligands. The [Co(en)3] 2+ ion exhibits a Dq value of 1130 cm with the value for [Co(NH3)6]2+ only slightly less.45

Strictly speaking, Jahn-Teller distortions would not be expected to occur because of the fact that spin-orbit coupling has removed the degeneracy of the low-lying 4Tlg term, a Kramers doublet.42 Not realizing the amount of tetragonal distortion, if it occurs in this case, one cannot definitively account for all spectral differences. This and similar systems are complicated and, in general, poorly understood.49 Magnetic Properties

Most transition metal compounds are relatively magnetically dilute, i.e., their paramagnetic centers are isolated from each other by inert









ligand molecules. In such compounds the paramagnetic ions act independently of each other. In these cases the idealized behavior for the variation of magnetic susceptibility with temperature is the Curie law, xA = C/T, where x is the susceptibility per mole of paramagnetic material corrected for the diamagnetic effect of the ligand molecules, C is the Curie constant and T the absolute temperature. However, the majority of paramagnetic substances obey a modified version called the Curie-Weiss law, Xt = C/(T-e), in which the e is an empirical quantity and is a measure of the deviation of the paramagnetism from the idealized Curie law description. This parameter is determined by plotting l/X' versus T and determining the intercept on the T axis. For magnetically dilute paramagnetics 0 is usually a small quantity.

Spin-free octahedral cobalt(Il) complexes customarily have magnetic moments between 3.89 and 5.2 B. M. and values of e between 150 and 300.55 In octahedral complexes one rarely observes magnetic moments as low as the Ispin-only" moment of 3.89 B. M. High-spin octahedral cobalt(II) compounds give very high orbital contributions to the magnetic moments. This high orbital contribution is attributable to the threefold degeneracy of the 4Tlg ground state. Considering the interaction between the total spin quantum number, S, and the orbital angular momentum quantum number, L, the magnetic moment may be calculated from [4S(S+l) + L(L+)]12 where for cobalt(II) S = 3/2 and L = 1. The moment is now found to be 4.12 B. M.; therefore, in cobalt(II) there is additional orbital angular momentum contributed in some other manner, i.e., from excited states or higher lying terms. Both the spinning of an electron and the movement of an electron in a closed path about the nucleus will produce a magnetic moment. The magnetic properties of any individual atom or ion will result,









therefore, from some combination of these two factors. Orbital contribution may be quenched wholly or partially by the lowering of symmetry. Electric fields of other atoms, ions and molecules surrounding the metal ion in its compound restrict the orbital motion of the electron.

The orbital contribution in the cobalt(II) ion is considered to

arise from the mixing of some of the next higher orbital triplet level into the singlet by the operation of spin-orbit coupling. The F ground term of cobalt (II) is split into two orbital triplet sets and a singleone.55 Considerable orbital contribution is seen in Co(P2A2)Cl2"4H20 as evidenced by a magnetic moment of 4.57 B. M. at room temperature. This moment is on the low end of the range expected for octahedral or distorted octahedral cobalt(II) complexes.55 Table 2 lists the room temperature magnetic moments of the compounds herein investigated.

There are instances in which the paramagnetic ions influence each other; these spin-spin interaction phenomena are referred to as magnetic exchange interactions. They may arise because the distance between the paramagnetic constituents is small, i.e., there is a direct overlap of atomic orbitals or a direct exchange, or because the intervening diamagnetic atoms are capable of transmitting the magnetic interaction, i.e., super-exchange.56 In cases where there is a spin-spin interaction the Weiss constant, 0, is large, e.g., the copper(II) acetate dimer which exhibits a Weiss constant of 1080.57 Ordinary mononuclear copper(II) complexes present 0 values of 9' or less.58

In this study those compounds possessing one copper ion per mole of complex were expected to exhibit magnetic moments indicative of the mononuclear species, i.e., 1.70-1.90 B. M. (see Table 2). The exceptions to this rule were the perchlorate salts of each macrocyclic complex.

In attempting to completely characterize magnetic properties over an











Compound Cu(D2A2) (NO3)2' 4H20

Cu(D2A2) (NO3)2 Cu2 (D2A2 )C14" 3H20 Cu2 (D2A2) Cl4 Cu(D2A2) (CI04)2" 4H20 Cu(D2A2) (ClO4)2 Co(P2A2)C2� 4H20 Cu2 (P2A2) CI4" 4H20 Cu2 (P2A2)C14 Cu(P2A2) (C104)2' 3H20 Cu(P2A2) (C104)2 Cu (DAN) 2 (NO3)2


Table 2

Average Magnetic Susceptibilities and Moments

Temperature, �K XA,* x 103 eff' B. M.*

292 1.352 1.78 299 1.256 1.74 293 0.930 1.50 299 0.539 1.14 296 1.176 1.67


0.906 8.859

1.359 0.995

1.005 0.569 1.492


1.47 4.57 1.78 1.54 1.56 1.16 1.88


Expected p eff


1.70-1.90 1.70-1.90 1.70-1.90

1.70-1.90 1.70-1.90 1.70-1.90 4.30-5.20 1.70-1.90

1.70-1.90 1.70-1.90 1.70-1.90 1.70-1.90


*Veff 2.83/'X x T per metal atom









extended range of temperature, it was found that both the hydrated and anhydrous forms of the perchlorate salts developed an abnormally large electrostatic charge while measurements were performed. A small amount of beta-emitter was placed in the cryostat in the form of a cesium-137 salt solution impregnating a piece of filter paper. This improved the ease of measurements at room temperature but data at low temperatures were impossible to obtain. The eleclxon spin resonance measurements indicate some form of spin-spin interaction (vide infra).

The data obtained from a temperature dependent magnetic susceptibility study of Cu2(P2A2)Cl4 and Cu2(P2A2)C4"4H20 are given in Table 3 and reproduced graphically in Figures 24 and 25. Each shows a slight increase of magnetic moment with decreasing temperature. The 8 values of 240 and 34% respectively, which are consistent with a weak spin-spin interaction, were determined from a plot of /Xk versus T. The results of esr measurements are also consistent with the observed interaction (vide infra).

Measurements from the temperature-dependent magnetic susceptibility study of Cu2(D2A2)C14 are included also in Table 3 and graphically presented in Figure 26. The hydrated form of the complex developed too great an electrostatic charge at low temperature for data to be obtained. The anhydrous form did not present a straight line upon plotting 1/Xk versus T but does show an increase in magnetic moment and susceptibility with decreasing temperature. The "knee" observed in this graph was reproduced several times with a rise and fall of temperature. This unusual phenomenon has been noted several times in recent studies,59-61 particularly with chloride salts of copper(II) complexes. It would appear that above 2100 K and below 1950 K the complex obeys the Curie-Weiss law. In








Table 3


Temperature Dependence


Compound Cu2 (D2A2)Cl4

6 = 580 PTIM = 1.24 B. M.**























Cu2 (P2A2)Cl4 o = 24� PTIM = 1.60 B.M.







Cu2 (P2A2)CI4" 4H20 o = 340 ]ITIM = 1.87 B. M.


T, OK 135.2 167.3 180.3 190.3 197.2 200.4 208.7 211.7 220.4 221.2 248.4 259.9 297.8 136.9 167.3 197.3 229.8 259.2 299.0 135.7 166.5 196.8 227.6


of Average Magnetic and Moments XA,* x 103


1.782 1.252 1.146 1.057 0.787 0.785 0.787 0.822 0.764 0.769 0.710 0.660 0.539 2.452 1.909 1.439 1.233 1.129 0.995 3.021 2.294 1.766 1.521


Susceptibilities


Peff' B. M.*

1.39 1.30 1.29 1.27 1.12 1.12 1.15 1.18 1.16 1.17 1.19 1.18 1.14 1.64 1.60 1.51 1.51 1.53 1.54 1.82 1.75 1.67 1.67








Table 3 (continued)


Xk* x 10 3
x


1.364 1.273 1.328 1.309


Peff' B. M.*

1.66

1.69 1.78 1.77


*e = 2.83VYX x T per metal atom


**PTIM = 2.83VX?(T+0) per metal atom


Compound


T, OK


252.7

279.6 298.5 299.5








10.0 +






8.0 +

+




C 6.0




C

N 4.0

X/
, /

o/





2.0
// x / // // // //





50 100 150 200 250 300 Temperature (0K) Fig. 24 Temperature dependence of inverse susceptibility per copper atom in Cu2 (+P2A2 )C14.













8.0,

� +



+
6.0




C

'-.4
S4.0


'-o4
i4




2.0






50 100 150 200 250 300 Temperature (OK)
,II





Fig. 25 Temperature dependence of inverse susceptibility per copper atom in
Cu2 ( P2 A2 )C14'4H20.









































100


200


250


Temperature ('K)


Temperature dependence of inverse susceptibility per copper atom in uJ Cu2 (D2A2) Cl4.


20.0.


U




x
H-


15.0r


io.Or


Fig. 26









other words, with a decrease in temperature from room temperature, X. increases and passes through a broad maximum and again increases in accordance with the Curie-Weiss behavior. Assuming the copper(II) ions are not interacting directly in the "hole" of the macrocycle, several rationalizations have been proposed to explain this behavior. 62,63 These investigators account for the behavior by (1) the presence of magnetically isolated copper(II) ions causing extraneous paramagnetism obeying the Curie-Weiss law at low temperatures or (2) the magnetic interaction between one-dimensional lattices. For an infinite one-dimensional Heisenberg spin lattice, the magnetic susceptibility has been calculated as a function of temperature. It shows a broad maximum and tends to a 64
finite value at very low temperatures. Since this "knee" would appear to be a Nel temperature, the phenomenon has been called linear 65
antiferromagnetism. A 0 value of 580 is consistent with a weak to moderate spin-spin interaction as determined from the plot of 1/X versus
M
T. Barraclough and Ng65 have shown that the one-dimensional Heisenberg model can be used to explain the observed results of anhydrous copper(II) chloride, which possesses a linear structure of copper atoms linked by bridging chlorides and one unpaired electron on each copper atom. In this compound an ordering process takes place below the N6el temperature. In the Heisenberg model one allows for interaction between nearest neighbors and assumes that if a pair of adjacent atoms have their electron spins parallel there is an interaction energy of J. If electron spins on adjacent pairs of atoms are parallel, the interaction energy is negative and the system shows antiferromagnetism. For the critical or N6el temperature 2J, the energy level separation between the singlet and triplet states, can be estimated from 2J=1.6kT where k is the Boltzmann constant, The compound Cu2(D2A2)CI4 presents a -2J value of 226 cm-. The copper(II)









acetate dimer and one of its derivatives exhibit a-2J value of 300 cm-I 57,66 and salts of the planar, bridged Cu2C162- ion show a-2J value of
-l 67
158 cm . The dimeric complex [Cu2(TMCD)](N03)4 possesses a-2J value of 56 cml.6

The break or "knee" in the 1I/X versus T plot (Figure 26) also may be attributed to a phase transformation in the solid, which, for a small temperature range, causes ordering of the magnetic domains.

Indeed, if the subnormal magnetic moments are not due to a superexchange mechanism, it is informative to have an understanding of the type of direct metal-metal bonding involved in dimeric complexes. If the copper ions are joined along the z-axis, both the unpaired electrons of the dimer are in the 3d 2 2 type orbital.68 Figgis and Martin,57 using
--x -y

arguments based on a theoretical treatment of Craig et al.,69 have proposed that the physical origin of the exchange interaction arises from the lateral overlap of these 3d 2 2 orbital functions, i.e., a 6 bond.
-x -y

More recent calculations by Ross70 and by Boudreaux71 and an nmr study
72 73
by Royer have supported this model. Hare et al. have considered the bond length in the Cu2 molecule. They concluded that the only slightly


longer copper-copper distances in dimers would seem to rule out existence 74 69
of a strong direct metal-metal interaction. Craig et al. have calculated that a decrease in the overlap integral of only 10% can halve the interaction energy. Thus, if the copper-copper distance is only slightly greater than that in the acetate complex, the amount of interaction would be decreased markedly.

Electron Spin Resonance Studies

For an electron of spin s = , the spin angular momentum quantum number can have values of m = �,, which in the absence of a magnetic









field leads to a doubly degenerate spin state. When a magnetic field is applied, the degeneracy is resolved as represented in the diagram. In an electron spin resonance experiment a transition from the MS = - to the




2Blg S'= MS' =

No field MS' = -S

H



M = + state occurs upon absorption of a quantum of radiation. The spectroscopic splitting factor, _, is inversely proportional to the field strength at which the resonance is observed, i.e., E = g H, where H is the field strength and is the Bohr magneton. For a free electron - has the value of 2.0023. In general, the magnitude of _ depends upon the orientation of the molecule containing the unpaired electron with respect to the magnetic field75 and the spin and orbital angular momenta.

The two isotopes of copper possess a nuclear spin, I, of 3/2.

Therefore, the number of allowed transitions for equivalent interacting copper ions is (2V1 + 1) or 7, where I' = I(i) + 1(2), as compared to four transitions in the mononuclear case. The situation is now further complicated by the two isotopes of copper which possess nearly identical nuclear
76
moments. In antiferromagnetically coupled copper(II) systems, either four or seven lines may be seen; never have sixteen lines been seen, based on 2 nonequivalent copper(II) ions. In solid polycrystalline samples hyperfine splitting is often not seen at low temperatures. The shape and line width of an absorption for a solid is frequently broader than its solution spectrum. Line widths are also altered considerably by exchange processes, i.e., electrons on neighboring lattice sites exchange









76
spin states rapidly. If the exchange occurs between equivalent ions, the lines broaden at the base and become narrower at the center. When exchange involves dissimilar ions, the resonances of the separate lines may merge to produce a single broad line. For the former situation van Vleck77 has shown that the isotropic exchange interaction contributes to the fourth moment and not to the second moment of the wave function describing the system. This "exchange narrowing" explains why microwave paramagnetic absorption lines are much narrower than one first conjectures from the amount of dipolar coupling. From esr data in Table 4 it is seen that the resonances of those copper(II) complexes presenting subnormal magnetic moments generally show some narrowing as expected for exchange between equivalent ions. The magnitude of narrowing varies from 5 to 30 gauss, in closq agreement with that found in the [Cu2(TMCD)](N03)4
6
system.

Three complexes exhibited unexpected, slight broadening at low temperature, i.e., Cu2(P2A2)C4, Cu(D2A2) (C104)2"4H20, and Cu(P2A2)(CO4)2.
78
Bleaney and Bowers also noted anomalous broadening at low temperature in their study of the copper(II) acetate dimer. They attributed the broadening to unresolved components of hyperfine structure and possibly thermal vibrations. Owing to the different nuclear magnetic moments of the two abundant isotopes of copper, each of the hyperfine lines may split into a number of components of unequal intensity (vide supra). At low temperature lattice vibrations should have little effect on esr absorptions. However, thermal vibrations may cause small fluctuation in the distance between two interacting ions. If the exchange integral, which is a measure of overlap of magnetic wave functions, is sensitive to distance, it will be sensitive to such fluctuations. Hare et al.73 and









Table 4

Temperature Dependence of R Values and Line Widths*
and Singlet-Triplet Separations


T, -K


Cu2 (D2A2)Cl4 Cu2 (D2A2) C14 3H20 Cu2 (P2 A 2 )Cl 4H 20 Cu2 (P2A2)Cl4 Cu (D2A2) (C104)2" 4H20 Cu(D2A2) (C104)2 Cu(P2A2) (C104)2" 3H O Cu(P2A2) (C104)2 Cu (DAN) 2 (NO3)2


298
77 298
77 298 77 298 77 298

77 298

77 298

77 298

77


[Cu2 (TMCD)] (NO3)4 Co (P2A2) C12� 4H20


g�Value


2.12
2.12

2.14 2.13

2.12 2.14

2.10 2.10

2.09 2.10

2.10
2.37(,)
2.09

2.08
2. 36(gll)
2.08

2.06
2. 11(g)
2.05
2.13(g)

2.11 2.11


W, gauss -2J, cm1


175 155

155 140 180 190 145 165 140 135 190 160 140


3.72


* Full width between extrema of first-derivative curves


Compound









Craig et al.69 have considered the distance factor to be important (vide supra).

In polycrystalline samples, one usually sees those electronic transitions for which AMS, = �1. In an axially symmetric field one would




M St =+1

S' = 1 MS= 0

MST= 1
S' =0 MS, =-i
No field H




expect two fundamental transitions and two resonances for each. In practice two resonances are seen from just one transition, viz. at approximately 2500 and 3200 gauss. The remaining resonances are not seen because of large zero-field splitting relative to the microwave energy employed in the measurement. If dipole-dipole interaction is sufficiently large, the low field resonance is of such low intensity it may not be observed in certain orientations.79 Therefore, in a sample of randomly oriented axially symmetric crystallites, the probability that the axially symmetric axis is perpendicular to the direction of the magnetic field is highest. The perpendicular postion of the spectrum is thus considerably stronger than the parallel portion.79 This much less intense parallel portion manifests itself by causing a dissymmetry in the first derivative curve. A very low intensity absorption at low field was seen only for Cu(P2A2) (CG04)2 and Cu(P2A2) (Cl04)2" 2H20.

The values of __ and yi, for the compounds in this study agree well with those compounds for which crystal structures have been determined,









i.e., known dimeric axially symmetric copper(ll) complexes. For various dimeric copper(II) complexes Y1varies from 2.05 to 2.10 and R, 2.20 to 24061,66,78,8083 Mononuclear copper(II) species customarily present_& values of 2.00 to 2.08 and.&, values of 2.1 to 2.2.41,44

Also listed in Table 4 are the estimated values of-2J for those

copper complexes with subnormal magnetic moments. The energy level separations have been calculated from the modified form of the van Vleck equation, viz.,

l3kT exp (k + Na


where _ is the spectroscopic splitting factor (determined from the esr spectrum), N is Avogadro's number, the Bohr magneton, k the Boltzmann constant, and Na the temperature independent paramagnetictcontribution to the molar susceptibility. Na is approximately 60 x 10.6 erg/gauss for copper complexes58 and arises from the second order interaction through the magnetic field between the T2g and E levels, i.e., it is a second order Zeeman effect.

The possibility exists that the observed interaction is intermolecular rather than intramolecular. To eliminate this possibility, esr measurements may be made on solutions of different concentrations. Because an exchange interaction requires overlap of wave functions, an intermolecular electron spin exchange is proportional to concentration. Thus, in a dilute system paramagnetic ions having like charge, intermolecular exchange is expected to be small84, 85 and the first effect of dilution upon exchange narrowed resonance is a broadening. Because of the very low solubility of all compounds investigated, this technique was not feasible. Stotz6 observed that solutions of [Cu2(TMCD)](N03)4 showed random variation









in line half widths of � 10 gauss. This is indicative of intra- rather than intermolecular exchange. Unfortunately, this technique may not be used to explain the behavior of the mononuclear copper complexes studied.

Electron spin resonance spectra for Co(P2A2)Cl2"4H20 were attempted at room temperature and at liquid nitrogen temperature (770 K). At room temperature the esr absorption becomes so broad as to escape detection by the equipment. Presumably the broadening arises from a decrease in the spin-lattice relaxation time which accompanies either an increase in temperature or an increase in orbital contribution to the magnetic susceptibility.86 The spectrum of microcrystalline solid [Co,(o-phen)3]E 55
(CG04)2 yields a g value of 4.2 � 0.1, considered to be characteristic of octahedral high-spin cobalt(II). An isotropic value of 4 is expected in the strong field limit and of 4.333 in the weak field limit for the S' = J level of an octahedral high-spin d7 ion. The absorption observed at = 3.72 for Co(DDnTC)Cl2"4H20 is seen only at high signal amplification and also presents a very broad envelope. Low-spin octahedral cobalt(II) complexes are expected to give an isotropic _ value of 2, observed for the 2E( 2G) resonance level.86 The esr absorption for a lowspin complex is generally narrower than that of high-spin and broadens but little as the temperature is increased.86 Because room temperature measurements were not feasible the author cannot estimate any possible low-spin contribution in the complex. In light of the magnetic moment, the large Z value at 77' K and the absence of an absorption of g = 2.0, it can be reasonably concluded that the cobalt(II) ion in the complex is essentially high-spin with very little, if any, low-spin contribution present.

In view of magnetic and esr data, the task of reasonably concluding









the mechanism of interaction in the copper(II) complexes is not as simple. Lacking structural data one can postulate several possibilities. Indeed, the "hole" of the proposed macrocycles, TMTC and DDnTC, may accommodate two copper(II) ions as shown previously in TMCD.6 Two metal ions may feasibly "fit" in the plane presented by the six nitrogen atoms, thus producing weak metal-metal interaction and the observed magnetic properties and electron spin resonances. The question arises, at this point, as to why the chloride salt permits such enclosure of two metal ions while the perchlorate and nitrate salts give a metal:ligand ration of 1:1. Stotz6 also observed that TMCD will accommodate two metal ions in the nitrate salt but only one in the perchlorate. It is not altogether unlikely that the perchlorate salts of this study may possess a crystal structure such that the cations are layered or are in such orientation in the solid that metal ions may interact intermolecularly.

Goedkan and Christoph87 and Mangia et al.88 have observed an unusual helical coordination about one and two metal ions. These studies have involved ligands containing diimines formed with 2,6-diacetylpyridine. The ligands do not present all donor atoms about the metal ion in a planar array. However, the methyl groups are undoubtedly causing steric interactions and force the ligand to "wrap" itself about the two metal ions in a spiral fashion. There is the possibility, also, that the macrocycles TMTC and DDnTC are not formed but that a polymer is produced which coordinates to the metal ion(s) in an helical manner, leaving a coordinated amine group at one end and a free carbonyl at the other. The few end groups should not give strong absorptions in the infrared spectra. The helical form of the polymer would certainly lessen steric interactions of the methyl groups with the naphthalene rings, and would not destroy the









planarity of the imine nitrogens. Such a theory could only be conclusively accepted upon the determination of the crystal structure. Unfortunately, attempts to grow crystals from various solvents have resulted in failure.

Further Characterization Attempts

Mass spectra of compounds I and II were taken at elevated temperatures. The highest weight fragment in both instances was m/e 158, the molecular weight of DAN. Because no macrocyclic parent ion was obtained, attempts to obtain mass spectral data were abandoned. Indeed, neither melting nor sublimation appeared to have occurred during these runs.

Conductance measurements were made on those compounds appearing to be soluble in methanol (I, II, IV, and VI), but, the resulting data were very erratic. The solutions employed in the conductivity measurements were saturated solutions obtained by stirring a mixture of complex and methanol for several hours. Subsequently, these mixtures were permitted to stand overnight. A portion of the supernatant mixture was removed after which concentrations were obtained by pipetting aliquots from the supernatant, and evaporating the methanol with a gentle stream of nitrogen and weighing the dry residue in the tared flask. Concentrations were in the range of 10-4 to 10-6 M. Because the resulting solutions gave scattered concentrations, these data suggested the possibility that true solutions were not formed, but, rather, colloidal dispersions. This suspicion was confirmed subsequently by simple light scattering experiments, i.e., the Tyndall effect. Thus, further measurements were considered to be of no value. This phenomenon has been observed in other complex "solutions."14 However, the mixtures do contain some solvated ions as observed by their low conductances. Other solvents were considered, but again solubilities









are very small.

Complexes II and V have much larger solubilities in DMF than do the other compounds. To increase solubility in DMF the solutions were heated. The unexpected observation was made that copper(II) chloride was removed as a consequence of refluxing. DMF is commonly known as a good solvent for copper salts. On this basis it was initially believed that the copper removed was present in II and V as a counterion, viz. tetrachlorocuprate(II) ion. This belief was discarded, subsequently on the basis of electronic spectral measurements (vide supra). Upon addition of a few milliliters of silver nitrate solution to the hot DMF solutions, the formation of silver chloride was noted. Precipitation is not observed at room temperature. This further substantiates the exclusion of copper(II) chloride from the pomplex.














SUMMARY

The preparation of six (l-VI) new complexes, believed to contain macrocyclic ligand molecules, was accomplished by the Schiff base condensation of 1,8-diaminonaphthalene and either 2,6-diacetylpyridine or 2,6-pyridinedicarboxaldehyde in the presence of copper(II) and cobalt(II) salts. Attempts to isolate free base, prepared via the template method, from the complexes by precipitation of the metal ions with sulfide ion were unsuccessful. Attempts to prepare the free ligands resulted in either incomplete cyclization or oligomerization or both as evidenced by elemental analyses and mass and infrared spectral data.

The complexes were characterized by elemental analysis, infrared, ultraviolet, visible and electron spin resonance spectra, and magnetic susceptibility determinations. The results of these studies support the formulation of each of the complexes as metal ion(s) surrounded by a planar, quadra- or hexadentate ligand with counterions or water either very loosely held in the axial positions or present as part of the crystal lattice.

Subnormal, room temperature magnetic moments were observed for Cu2(D2A2)C14"2H20, Cu2(D2A2)C14, Cu(D2A2)(C104)2, Cu2(P2A2)C14, Cu(P2A2)(C104)2 and Cu(P2A2)C104)2"2H20. Because of a large electrostatic charge accumulation upon some of the samples, particularly at lower temperatures, reliable temperature-dependent magnetic susceptibility data could be obtained only for Cu2(D2A2)C14, Cu2(P2A2)Cl4 and Cu2(P2A2)CI4"4H20. The subnormal moments of these complexes range from 1.5 to 1.1 B. M., less








than the expected 1.7 to 1.9 B. M. characteristic of the d9 system. The complex Cu2(D2A2 )C4 presented a transition temperature in the range of 1970 to 2090 K, following the Curie-Weiss law above and below this plateau.

Electron spin resonance spectra of those compounds possessing subnormal magnetic moments give y values and general peak narrowing indicative of spin-spin exchange of either intra- or intermolecular character. Extensive structural analysis would be required to aid in the elucidation of the mechanism of this exchange.








































APPENDIX















Table 1-A

Mass Spectral Cracking Pattern of Impure Trimer of TMTC


m/e (70 eV, 1000 C)


Relative Intensity


443 304
299
262 246 205 204 184 183


Table 2-A

Mass Spectral Cracking Pattern of Impure DDnTC


m/e (70 eV, 340� C)


Relative Intensity


31.1 100.0 24.0 29.8 10.2 23.6 7.4















Table 3-A

Octahedral Ionic Radii of Metal Ions in Oxide Salts


Radius, X *

0.73


Metal Ion

Cu(II) Co(II) (high spin)

Ni(II) Mn(II) (high spin)


0.74 0.70 0.82 0.75


Zn(II)


* Taken from R.D. Shannon and C.T. Prewitt, Acta
Crystallogr., B26, 1046 (1970).














BIBLIOGRAPHY


1. L.F. Lindoy and D.H. Busch in "Preparative Inorganic Reactions," W.L.
Jolly, Ed., John Wiley & Sons, Inc., New York, N.Y., 1971, vol. 6.

2. D. St.C. Black and A.J. Hartshorn, Coord. Chem. Rev., 9, 219 (1973). 3. D.W. Wester, Ph.D. Thesis, University of Florida, Gainesville, Fla., 1975.

4. D. St.C. Black and I.A. McLean, Aust. J. Chem., 24, 1401 (1971).

5. K. Travis and D.H. Busch. Chem. Commun., 1041 (1970).

6. R.W. Stotz, Ph.D. Thesis,University of Florida, Gainesville, Fla., 1970.

7. F.A. Cotton and E. Pedersen, Inorg. Chem., 14, 383 (1975).

8. F.A. Cotton and E. Pedersen, ibid., 14, 388 (1975). 9. F.A. Cotton and E. Pedersen, ibid., 14, 399 (1975). 10. F.A. Cotton, T.G. Dunne, and J.S. Wood, ibid., 3, 1495 (1964). 11. J.N. van Niekerk and F.R.L. Schoening, Acta Crystallogr., 6, 227 (1953). 12. D.H. Busch, K. Farmery, V. Goedken, V. Katovic, A.C. Melnyk, C.R.
Sperati, and N. Tokel, Advan. Chem. Ser., 100, 44 (1971).

13. J.L. Love and H.K.J. Powell, Inorg. Nucl. Chem. Lett., 3, 113, (1967).

14. S.R. Weller, Ph.D. Thesis, University of Florida, Gainesville, Fla.,
1973.

15. F.F. Myers,, Jr., personal communication. 16. R.W. Layer, Chem. Rev., 63, 489 (1963). 17. N.F. Curtis, Coord. Chem. Rev., 3, 3 (1968). 18. D.H. Busch, Record Chem. Progr., 25, 107 (1964). 19. D.H. Busch, Helv. Chim. Acta (spec. ed.), 174 (1967). 20. D.H. Busch, Advan. Chem. Ser., 37, 125 (1963). 21. D.H. Busch, D.C. Jicha, M.C. Thompson, J.W. Wrathall, and E. Blinn,
J. Amer. Chem. Soc., 86, 3642 (1964).
82








22. M.C. Thompson and D.H. Busch, ibid., 86, 3651 (1964). 23. F.P. Papadopoulos, A. Jarrar, and C.H. Issidorides, J. Org. Chem.,
31, 615 (1966).

24. W. Mathes, W. Sauerlich, and T. Klein, Chem. Ber., 86, 585 (1953). 25. F. Sondheimer, '. Mancera, M. Urquiza, and G. Rosenkranz, J. Amer.
Chem. Soc., 77,"4145 (1955).

26. E.A. Clevenger, M.S. Thesis, University of Florida, Gainesville,
Fla., 1961.

27. B.N. Figgis and R.S. Nyholm, J. Chem. Soc., 4190 (1958). 28. B.N. Figgis and R.S. Nyholm, ibid., 12 (1954). 29. B.N. Figgis and R.S. Nyholm, ibid., 331 (1959). 30. R.T. Morrison and R.N. Boyd, "Organic Chemistry," 2nd ed., Allyn and
Bacon, Inc., Boston, Mass., 1966, Chap. 19.

31. D.H. Busch, L.Y. Martin, L.J. DeHayes, and L.J. Zompa, J. Amer. Chem.
Soc., 96, 4046 (1974).

32. D.H. Busch and' G.A. Melson, Proc. Chem. Soc., London, 223 (1963). 33. J.P. Lafornara, Ph.D. Thesis, University of Florida, Gainesville,
Fla., 1970.

34. K. Nakamoto, "Infrared Spectra of Inorganic Complex Molecules," John
Wiley & Sons, Inc., New York, N.Y., 1970.

35. B. Witkop, J. Amer. Chem. Soc., 76, 5597 (1954). 36. L.J. Bellamy, "The Infrared Spectra of Complex Molecules," John Wiley
& Sons, Inc., New York, N.Y., 1958, Chap. 16.

37. B.M. Catehouse, S.E.Livingstone, and R.S. Nyholm, J. Chem. Soc., 4222
(1957).

38. B.J. Hathaway and A.E. Underhill, ibid., 3091 (1961). 39. D.J. Pasto and C.R. Johnson, "Organic Structure Determination,"
Prentice-Hall, Inc., Englewood Cliffs, N.J., 1969, Chap. 3. 40. B.J. Hathaway, J. Chem. Soc., Dalton Trans., 1196 (1972). 41. I.M. Proctor, B.J. Hathaway, and P. Nicholls, J. Chem. Soc. A, 1678
(1968).

42. C.J. Ballhausen, "Introduction to Ligand Field Theory," McGraw-Hill
Co., New York, N.Y., 1962, Chap. 10.








43. S. Yamada and R. Tsuchida, Bull. Chem. Soc. Jap., 29, 289 (1956). 44. B.J. Hathaway, D.E. Billings, P. Nicholls, and I.M. Proctor, J.
Chem. Soc. A, 319 (1969).

45. L.E. Orgel, J. Chem. Phys., 23, 1004 (1955). 46. F.A. Cotton and R.H. Holm, J. Amer. Chem. Soc., 82, 2979, 2983 (1960). 47. J.S. Wood, Inorg. Chem., 7, 852 (1968). 48. F.R. Urbach, R.D. Bereman, J.A. Topick, M. Hariharan, and B.J. Kalbacher, J. Amer. Chem. Soc., 96, 5063 (1974).

49. F.A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry," John
Wiley & Sons, Inc., New York, N.Y., 2nd ed., 1966, Chap. 29. 50. J. Ferguson, J. Chem. Phys., 32, 533 (1960). 51. W. Low, Phys. Rev., 109, 256 (1958). 52. B.N. Figgis, "Introduction to Ligand Fields," John Wiley & Sons, Inc.,
New York, N.Y., 1966, Chap. 9.

53. G.L. Roberts and F.H. Field, J. Amer. Chem. Soc., 72, 4232 (1950). 54. C.J. Ballhausen and C.K. Jfrgensen, Acta Chem. Scand., 9, 397 (1955). 55. B.N. Figgis and R.S. Nyholm, J. Chem. Soc., 338 (1959). 56. M. Kato, H.B. Jonassen, and J.C. Fanning, Chem. Rev., 64, 99 (1964). 57. B.N. Figgis and R.L. Martin, J. Chem. Soc., 3837 (1956). 58. B.N. Figgis and C.M. Harris, ibid., 855 (1959). 59. M. Inoue, S. Emori, and M. Kubo, Inorg. Chem., 7, 1427 (1968). 60. F.J. Rioux and B.C. Gerstein, J. Chem. Phys., 50, 758 (1969). 61. G.F. Kokoszka, K. Hyde, and G. Gordon, J. Inorg. Nucl. Chem., 31,
1993 (1969).

62. H. Kobayashi, T. Haseda, E. Kanda, and S. Kanda, J. Phys. Soc. Ja.,
18, 349 (1963).

63. S. Kadota, I. Yamada, S. Yomayama, and K. Hiradawa, ibid., 23, 75
(1967).

64. J.C. Bonner and M.E. Fisher, Phys. Rev., 135, A640 (1964). 65. C.G. Barraclough and C.F. Ng, Trans. Faraday Soc., 60, 836 (1964). 66. J.R. Wasson, C. Shyr, and C. Trapp, Inor. Chem., 7, 469 (1968).








67. G.J. Maass, B.C. Gerstein, and R.D. Willet, J. Chem. Phys., 46, 410
(1967).

68. R.W. Jotham and S.F.A. Kettle, Inorg. Chem., 9, 1390 (1970). 69. D.P. Craig, A. Maccoll, R.S. Nyholm, L.E. Orgel, and L.E. Sutton,
J. Chem. Soc., 322 (1954).

70. I.G. Ross, Trans. Faraday Soc., 55, 1057 (1959). 71. E.A. Boudreaux, Inorg. Chem., 3, 506 (1964). 72. D.J. Royer, ibid., 4, 1830 (1965). 73. C.R. Hare, T.P. Sleight, W. Cooper, and G.A. Clarke, ibid., 7, 669
(1968).

74. G.F. Kokoszka, M. Linzer, and G. Gordon, ibid., 7, 1730 (1968). 75. R.S. Drago, "Physical Methods in Inorganic Chemistry," Reinhold
Publishing Co., New York, N.Y., 1965, Chap. 10.

76. A. Carrington and A.D. McLachlan, "Introduction to Magnetic Resonance,"
Harper & Row, Publishers, New York, N.Y., 1967. 77. J.H. van Vleck, Phys. Rev., 74, 1168 (1948). 78. B. Bleaney and K.D. Bowers, Proc. Roy. Soc., Ser. A, 214, 451 (1952). 79. G.F. Kokoszka and R.W. Duerst, Coord. Chem. Rev., 5, 209 (1970). 80. R.W. Duerst, S.J. Baum, and G.F. Kokoszka, Nature (London), 222,
665 (1969).

81. D.J. Hodgson, P.K. Hale, J.A. Barnes, and W.E. Hatfield, Chem. Commun.,
786 (1970).

82. J. Lewis, F.E. Mabbs, L.K. Royston, and W.R. Smail, J. Chem. Soc. A,
291 (1969).

83. S.J. Grubber, C.M. Harris, and E. Sinn, Inorg. Chem., 7, 268 (1968). 84. D.M.S. Bagguley and J.H.E. Griffths, Nature: (London), 162, 538 (1948). 85. D. Kivelson and G. Collins in "Paramagnetic Resonance," Vol. II, W.
Low, Ed., Academic Press, Inc., New York, N.Y., 1963, p. 496.

86. J.G. Schmidt, W.S. Brey, and R.C. Stoufer, Inorg. Chem., 6, 268 (1967). 87. V.L. Goedken and G.G. Christoph, ibid., 12, 2316 (1973). 88. A. Mangia, C. Pelizzi, and G. Pelizzi, Acta Crystallogr., B30, 2146
(1974).
















BIOGRAPHICAL SKETCH

Barbara Judith Romanik was born May 21, 1942, in Millville, New

Jersey. She attended Bridgeton (New Jersey) Senior High School and was graduated in June, 1960. In June, 1964, she received the degree of Bachelor of Science in Chemistry from Washington College, Chestertown, Maryland.

Ms. Romanik taught in the public school system of Polk County,

Florida, between January, 1966, and June, 1971. Graduate study was begun in June, 1968, at Purdue University. Under a National Science Foundation Fellowship and a four-summer sequential program she received the Master of Science degree in Chemistry in August, 1971. In September, 1971, Ms. Romanik enrolled in graduate school at the University of Florida. From that time until June, 1974, she held a graduate teaching assistantship and was an interim instructor for the 1972-73 year. A DuPont Teaching Award was received in June, 1973. From June, 1974, until June, 1975, she supported her own graduate education.

Ms. Romanik is a member of the American Chemical Society and its Division of Chemical Education.










I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
















I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.





Gus. J. nik
Professor of Chemistry








I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.


Richard D. Diesdner Professor of Chemistry











I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.





Joh A. Zoltewicz Proessor of Chemistry








I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.





Larry L. Hench
Professor of Materials Science and Engineering








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

August, 1975


Dean, Graduate School




Full Text

PAGE 1

PREPARATION AND CHARACTERIZATION OF COPPER (II) COMPLEXES OF THE CONDENSATION PRODUCTS OF 2 , 6-DIACETYLPYRIDINE AND 2 , 6-PYRIDINEDICARBOXALDEHYDE WITH 1 , 8-DIAMINONAPHTHALENE By BARBARA JUDITH ROMANIK i A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA

PAGE 2

To my parents and grandparents for love and encouragement

PAGE 3

ACKNOWLEDGEMENTS The author wishes to express her sincere appreciation to Dr. R.C. Stoufer, Chairman of the author's Supervisory Committee, and to the other members of her Supervisory Committee. Special thanks to Dr. R.W. King for his aid in obtaining the mass spectral data given in the manuscript. Much appreciation is expressed to the typist. Miss Oonagh Kater. iii

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii LIST OF TABLES v LIST OF FIGURES vi KEY TO SYMBOLS USED IN TEXT viii ABSTRACT ix INTRODUCTION 1 EXPERIMENTAL 8 Reagents 8 Preparation of Starting Materials 8 Procedures 9 Apparatus 15 RESULT AND DISCUSSION 18 Ligands 18 Complexes 22 Infrared Spectra 25 Electronic Spectra 38 Magnetic Properties 56 Electron Spin Resonance Studies 67 Further Characterization Attempts 75 SUMMARY 77 APPENDIX 79 BIBLIOGRAPHY 82 BIOGRAPHICAL SKETCH 86 iv

PAGE 5

LIST OF TABLES Table Page 1. Near Infrared, Visible and Ultraviolet Spectral Data 41 2. Average Magnetic Susceptibilities and Moments 59 3. Temperature Dependence of Average Magnetic Susceptibilities and Moments 61 4. Temperature Dependence of Values and Line Widths and Singlet-Triplet Separations 70 1A. Mass Spectral Cracking Pattern of Impure Trimer of TMTC 80 2A. Mass Spectral Cracking Pattern of Impure DDnTC 80 i 3A. Octahedral Ionic Radii of Metal Ions in Oxide Salts 81 v

PAGE 6

LIST OF FIGURES Figure 1. Structural formula of TMTC 2. Structural formula of DDnTC 3. Structural formula of TMCD 4. Structural formula of CuCDAN)^ 5. Infrared spectra of DDnTC and its precursors 6. Infrared spectra of impure trimer of TMTC and its precursors 7. Infrared spectrum of CuCD^A^^O^)^ 4H 0 .8. Infrared spectrum of Cu„ (D_A„)C1. • 2H.0 f 2 2 2 4 2 9. Infrared spectrum of Co (P 2 A 2 )Cl 2 411^0 10. Infrared spectrum of Cu 2( P 2 A 2) C1 4* ^ H 2^ 11. Infrared spectrum of Cu(D 2 A 2 ) (CIO^^^H^O 12. Infrared spectrum of Cu(P 2 A 2 ) (C10^) ^ • 2 H 2 O 13. Infrared spectrum of Cu (DAN ) ^ (NO ^) ^ 14. Electronic spectrum of impure trimer of TMTC 15. Electronic spectrum of DDnTC 2 16. The splitting of the D term of the copper (II) ion in ligand fields of different symmetries 17. Electronic spectrum of Cu(D 2 A 2 ) (NO^^'^I^O 18. Diffuse reflectance spectrum of Cu2(D2A2)C1^*2H.20 ... 19. Electronic spectrum of Cu (D 2 A 2 ) (C10^) ^ ' 41^0 20. Diffuse reflectance spectrum of C^ 41^0 ••• 21. Electronic spectrum of Cu(P 2 A 2 ) (CIO^^* ZI^O vi Page 5 6 7 26 27 29 31 32 33 34 35 36 37 39 40 45 47 48 49 50 51

PAGE 7

LIST OF FIGURES (Continued) Figure 22. Diffuse reflectance spectrum of Cu (DAN) ^ (NO^) ^ 23. Diffuse reflectance spectrum of Co (P 2 A 2 )C1 2 • 4^0 24. Temperature dependence of inverse susceptibility per copper atom in Cu„ (P 0 A„)C1, 2 2 2 4 25. Temperature dependence of inverse susceptibility per copper atom in Cu„ (P„A„)C1 • 4H„0 2 2 2 4 2 Temperature dependence of inverse susceptibility per copper atom in Cu 2 (D 2 A 2 )C 1 ^ 26 .

PAGE 8

KEY TO THE SYMBOLS USED IN TEXT DDnTC 13,9: 28-24-dinitrilo-9H, 24H-dinaphtho [ 1 , 8-bc : 1 ' , 8 ' -no] 1 [1,5,13,17] tetraazacyclotetracosine TMTC 8,14,23 , 29-tetramethyl-13 , 9 : 28-24-dinitrilo-9H , 24H-dinaphthoE [l,8-bc:l' , 8 ' -no ] [l,5,13,17]tetraazacyclotetracosine TMCD 6,12,19, 25-tetramethyl-7 , 11:20, 24-dinitrilodibenzo [b ,m] 5 [1,4,12,15] tetraazacyclodocosine DAP 2,6-diacetylpyridine PDC 2,6-pyridinedicarboxaldehyde DAN 1,8-diaminonaphthalene D> 2^2 condensate of two moles DAP and two moles DAN ^2^2 condensate of two moles PDC and two moles DAN DMSO dimethylsulf oxide DMF N,N' -dimethylf ormamide NH^ ammonia en ethylenediamine CDCl^ chloroform, deuterated j)-phen 1,10-phenanthroline viii

PAGE 9

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PREPARATION AND CHARACTERIZATION OF COPPER(II) COMPLEXES OF THE CONDENSATION PRODUCTS OF 2 , 6-DIACETYLPYRIDINE AND 2 , 6-PYRIDINEDICARBOXALDEHYDE WITH 1, 8-DIAMINONAPHTHALENE By Barbara Judith Romanik August, 1975 Chairman: R. Carl Stoufer Major Department: Chemistry i Six new complexes proposed to contain the macrocyclic ligands, 13,9:28, 24-dinitr ilo-9H, 24H-dinaphtho [ 1 , 8-bc : 1 ' , 8 ' -no ] [ 1 , 5 , 13 , 17 ] S tetraazacyclotetracosine (DDnTC) and 8,14,23,29-tetramethyl-13,9:28,24dinitrilo-9H , 24H-dinaphtho [ 1 , 8-bc : 1 ' , 8 ' -no] [1,5,13,17] tetraazacyclotetracosine (TMTC) , have been prepared via the template method. The former ligand was produced by the Schiff base condensation of 1,8diaminonaphthalene and 2 , 6-pyridinedicarboxaldehyde with the salts CuCClO^)^ • 6^0, CuCl^, and CoCl^. The latter was prepared by the Schiff base condensation of 1, 8-diaminonaphthalene and 2 , 6-diacetylpyridine with the salts CuCNO^^’xH^O, CuCl^, and Cu(ClO^) ^ * 6^0. Isolation of the free macrocycle base could not be accomplished by precipitation of the metals with sulfide ion. The complexes were characterized by elemental analysis, infrared, ultraviolet, visible and electron spin resonance spectra, and magnetic susceptibility determinations. The results of these studies support the ix

PAGE 10

formulation of each of the complexes as metal ion(s) surrounded by a planar, quadraor hexadentate ligand with counterions or water either very loosely held in the axial positions or present as part of the crystal lattice. The copper (II) chloride and perchlorate complexes presented subnormal room temperature magnetic moments. Temperature-dependent magnetic susceptibility data were obtained for Cu 2 (TMTCOCl^, Cu 2 (DDnTC)Cl , and Cu 2 (DDnTC) Cl^ ' 4H 2 0 . The complex Cu 2 (TMTC)Cl^ presented a transition temperature in the range of 197° to 209° K, following the Curie-Weiss law above and below this plateau. Electron spin resonance data of those compounds possessing anomalous magnetic moments gave evidence of spin-spin exchange of either intraor intermolecular character. A structure determination would be desired to aid in the elucidation of the mechanism of this exchange, i.e., direct metal-metal interaction or super-exchange. Unfortunately, extreme insolubilities of the complexes have prevented growth of single crystals. x

PAGE 11

INTRODUCTION In the past fifteen years the possibility of using synthetic macrocycles as simple models for complicated biological systems has been actively pursued.-*Consequently the area of coordination chemistry involving the synthesis of transition metal complexes incorporating these macrocyclic ligands has grown significantly. A macrocyclic ligand is one in which the donor atoms are contained as integral parts of a closed ring. The majority of the macrocyclic ligands reported to date are quadridentate ligands. Examples of penta-, hexa-, and octadentate macrocycles are rare2 5 in most instances, the i donor atoms are located approximately in a planar array about the central metal ion. The complete characterization of a variety of model macrocyclic complexes should lead to a better understanding of naturally occurring analogues, e.g., porphyrins, corrins, which themselves are highly conjugated and which may contain metal ions confined within a cyclic, approximately planar array of four donor atoms. Initially, the study of these new synthetic systems described herein was undertaken in an attempt to prepare and to characterize complexes containing metal-metal bonds.^ ^ There are but a very few compounds of this kind which have been prepared and these are incompletely characterized. Macrocyclic ligands with significantly large "holes" lend themselves to incorporating more than one metal ion within the ring, thus increasing the tendency for metal-metal interaction. 1

PAGE 12

2 Metal-metal bonding is now noted as a widespread phenomenon which is both interesting and important. Of the numerous compounds now known in which metal— metal bonding is thought to occur, most examples contain metals of the second and third transition series. Too, the majority of these compounds contain metal clusters in which the metal ions are in a low oxidation state, bridged by oxo-, halo-, or carbonyl groups. Although direct metal to metal bonding is rare for first transition series metal ions, copper complexes have been reported to involve this type of interaction.^®’ Macrocyclic ligands and complexes are generally uncommonly stable both in the thermodynamic and the kinetic sense; they are inert toward dissociation, even in strong mineral acid. Because of the ring structure, a simple dissociative step cannot occur; the ring has no "end." It is not i possible to extend metal-donor distance sufficiently to constitute bond rupture without additional bond rupture involving the ligand or extensive rearrangement within the coordination sphere. Thus, activation energies are considered to be very large. It is suspected that the more flexible unconjugated rings fold before the first bond between the donor atom and metal ion is broken. It is relatively simple to move a donor atom away from the metal ion when the ring is in a folded configuration. The in-plane ligand field strengths characteristic of macrocyclic ligands have been shown to be somewhat greater than expected on the basis of kinds of donor atoms,-^ leading to a more thermodynamically stable product. Accordingly, it is without surprise that one discovers that, although various methods have been attempted for the displacement of macrocyclic ligand from the metal ion, e.g., cyanide ion,^^ sulfide ion^ and even solvated electrons,-*-^ rarely is the free macrocycle base released; either

PAGE 13

3 the macrocyclic complex remains intact or the ligand is destroyed completely. Many synthetic macrocyclic complexes have been prepared containing nitrogen as the donor atoms. Many of these are derived via a Schiff base condensation of a carbonyl.— containing compound with an amine moiety. Such condensations are known to proceed by way of nucleophilic attack of the amine nitrogen on the carbon atom of the carbonyl group to yield a carbinolamine intermediate?6 Normally, this reaction is acid-catalyzed. Thus coordination of the carbonyl oxygen to a positive center favors condensation by making the carbonyl carbon atom more susceptible to nucleophilic attack. Schiff base-containing macrocycles have been widely studied by Curtis 1 ^ and Busch ^ > ^ and their co-workers. Another type of macrocyclic complex is that obtained by the reaction of coordinated mercaptides with alkyl and aryl dihalides. Numerous examples of these complexes have been prepared by Busch and coworkers.^® 22 Although a large number of macrocyclic complexes have been prepared, free macrocyclic ligands are rarely isolated because of contamination by acyclic impurities. Accordingly macrocyclic ligands are prepared in the presence of a metal ion, such as Ni(II), Co(II)„, or Cu(II);. The metal ion serves to orient the reacting species in such a way that cyclization occurs with minimal side reactions. This influence, exhibited by the metal ion, is known as the "coordinate template effect." The particular systems chosen for these investigations result from the Schiff base condensation of 2, 6-pyridinedicarboxaldehyde (PDC) with 1,8-diaminonaphthalene (DAN) and 2 ,6-diacetylpyridine (DAP) with the same diamine. In each instance a 20-member ed macrocyclic ring is possibly formed, i.e., a potential hexadentate ligand. The macrocycles do not

PAGE 14

4 satisfy the Hiickel (4n + 2) criterion but do exhibit a large degree of conjugation (Figures 1 and 2) . Often, in the case of large annulenes and their metal complexes, the Hiickel rule is not stringent. In view of this fact and Framework Molecular Model representations, this author has concluded that the donor atoms can be planar and the Tr-system of the ligand delocalized. The ligands of this study were conceived as modifications of 6,12,19 , 25-tetramethyl-7 ,11:20, 24-dinitr ilodibenzo [b ,m] [1,4,12,15] tetraazacyclodocosine , hereafter TMCD 6 (Figure 3) in the attempt to promote metal-metal bonding in a slightly larger ring, 20-member ed rather than 18-membered. Also, most synthetic macrocycles have been formulated from alkyl amines rather than aromatic amines. For broader scope and application to natural systems, continued synthesis and study should envelop the incorporation of aromatic amines in the model systems.

PAGE 15

8,1 4, 2 3, 2 9 -T E T R A ME T HY L-1 3,9 : 28,24 DlNITRILO-9P),24HD INA PH THO [l,8-bc:T,8',-ni]c [1,5,13,17] T ETRA A ZACYCLOTETRACOSI NE Fig. 1 Structural formula of TMTC

PAGE 16

1 3, 9:20, 24 DIN I TRILO-9H,24HD INAPHT HO [l,8be :1',8'-no] [l,5,13,17]c I ETR A A ZAC YC LOTETRACOS1NE Fig. 2 Structural formula of DDnTC

PAGE 17

H 3 qr^N^iCH 3 C-N N o N N h 3 c CH 3 6,12,19,2 5 -TE TRA METHYL-^II :20,24-Dl NITR1LOD I B E N ZO [b, mj |i,4 ,1 2,1 5] TETRAAZAOC LODOCOSI NE Fig. 3 Structural formula of TMCD

PAGE 18

EXPERIMENTAL Reagents Unless otherwise specified all chemicals were commercially available as reagent grade and were used without further purification. 2.6Pyridinedimethanol Aldrich Chemical Co., Inc.; used as received; mp, 111-114° C. 2 . 6Diace tylpyridine Aldrich Chemical Co., Inc.; used as received (97%); mp, 78-79° C. 2,2'-Dimethoxypropane Eastman Kodak Co,; used as received (practical); bp, 76-80° C. Preparation of Starting Materials Copper (II) perchlorate hexahydrate . The salt was prepared by reacting an aqueous slurry of reagent grade copper (II) carbonate with a slight excess of 70% perchloric acid, concentrating the resulting solution on a hot plate, cooling, filtering and drying over P^O^q in vacuo . The salt contained some residual perchloric acid reflected by the acidity of its aqueous solutions. 2 . 6Pyridinedicarboxaldehyde . The dialdehyde was prepared by em22 ploying a modification of the method of Papadopoulos j^t al. Sixty grams (0.69 mol) of freshly prepared manganese dioxide was suspended in 500 ml of chloroform containing 5.7 g (0.041 mol) of 2,6-pyridinedimethanol. The mixture was stirred at reflux for 5 hr., filtered, and the oxide washed with five 100-ml portions of ether. The filtrates were combined and evaporated under a stream of nitrogen gas. The residue was taken up in a minimum of solvent, 80% benzene-20% ethyl acetate. This solution 8

PAGE 19

9 was placed on a silica gel (60-200 mesh) column and eluted with 500 ml of the solvent mixture; the middle portion (350 ml) was collected after discarding the first 125 ml. The desired fraction was blown dry under a stream of nitrogen gas to yield a white crystalline product. Yield: 1.75 g (32%); mp 115-122° C; lit. mp, 124° a 24 Fresh manganese dioxide . The oxide was prepared by a modification of the procedure of Sondheimer et al . 25 A solution of 70 g (0.44 mol) potassium permanganate in 700 ml water was made acidic with 25 ml concentrated sulfuric acid. To this hot solution was added 100 g (0.59 mol) manganese sulfate monohydrate in 400 ml of water. The mixture was stirred for 2 hr while maintaining a temperature of 90° C. An excess of permanganate was also maintained throughout the reaction. The resulting solid was filtered by suction and washed five times i with water by decantation. The solid was dried subsequently in an oven at 135° C for 24 hrs, pulverized, and stored over calcium chloride or ^4°10’ 1 , 8-Diaminonaphthalene (Aldrich Chemical Co., Inc., 97%). The diamine was recrystallized from a saturated solution of 95% ethanol at room temperature. An excess of distilled water was added with stirring. The resulting mixture was an opaque flesh color. Formation of red— brown crystals was allowed overnight in the refrigerator. The product was collected and washed with water on a Btichner funnel. This procedure was repeated 2-3 times. The solid was dried over P.0,in vacuo at room 4 10 temperature, mp, 64-65° C; lit, mp, 66° C. Procedures Cu(D 2 A 2 ) (N0 3 ) 2 ‘4H 2 0.* In a 500-ml flask, equipped with a reflux *D 2 A 2 = condensate of two moles DAP and two moles DAN

PAGE 20

10 condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 2.0 g (0.012 mol) 2,6-diacetylpyridine. To this was added a filtered solution of 2.8 g (0.012 mol) hydrated copper (II) nitrate in 100 ml absolute methanol. Upon bringing the solution to reflux the color changed from light blue to light green. Added to the hot solution dropwise was 1.9 g (0.012 mol) 1,8-diaminonaphthalene in 100 ml absolute methanol. The mixture was refluxed for 48 hr then cooled and filtered. The product was collected on a sintered glass filter, washed with several small portions of solvent and ether and dried over P^O^q in vacuo . Yield: 2.5 g (30%) The complex is obtained as black powder containing some shiny black flakes. It is soluble in dimethylformamide, dime thylsulf oxide, and methanol to a limited degree. The complex was found to absorb atmospheric , i moisture reversibly, i.e., the water can be removed entirely in a drying pistol at 100° over P^O^q in. vacuo . After 24 hr in the drying pistol, the complex loses an average of 9.3% by weight. A comparable gain of weight in the atmosphere requires a 72 hr exposure. Anal. Calcd for Cu(C QQ H. n N,) (NO.) . : 4H 0 : C, 54. 94; H, 4.58; jo jU o i J. I N, 13.49; Cu, 7.71. Found: C, 54.93; H, 4.23; N, 12.43; Cu, 8.29. ^ U 2 ^2^2^'*"4 "^^2^” i n a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying rube, containing 100 ml absolute methanol, was dissolved 1.63 g (0.010 mol) 2,6-diacetylpyridine. To this was added a solution of 2.70 g (0.020 mol) anhydrous copper (II) chloride in 100 ml absolute methanol. Upon bringing this solution to reflux, 1.98 g (0.012 mol) 1,8-diaminonaphthalene in 100 ml absolute methanol was added dropwise. After 48 hr reflux the hot mixture was filtered. The product was collected on a sintered glass filter, washed with several small

PAGE 21

11 portions of solvent and ether and dried over P.0, „ in vacuo. Yield: 4.4 g (100%) The complex is obtained as a dark brown powder. It is slightly soluble in pyridine, dimethylf ormamide, and dimethylsulf oxide. The complex was found to absorb atmospheric moisture. After 24 hr in a drying pistol at 100° over P^O^g in vacuo the complex loses 6.3% by weight. A comparable gain of weight in the atmosphere requires the same amount of time. Anal . Calcd for ^(CjgH NjjCl^ 2^0: C, 52.11; H, 3.89; N, 9.60; Cu, 14.51. Found: C, 51.90; H, 3.82; N, 9.42; Cu, 13.01. Cu(D 2 A 2 ) (CIQ^) ^ ^H^O. In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 1.63 g (0.010 mol) 2,6-diacetylpyridine. To this ( was added a solution of 7.41 g (0.020 mol) hydrated copper (II) perchlorate in 100 ml solvent. Upon bringing this solution to reflux, 1.58 g (0.010 mol) 1,8-diaminonaphthalene in 100 ml solvent was added dropwise. A dark solid formed within 24 hr. After 48 hr of reflux 150 ml solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of methanol and ether and dried over P^O^q in vacuo . Yield: 1.9 g (21%) The complex is obtained as a black powder. It is slightly soluble in dimethylf ormamide, pyridine, dimethylsulf oxide, and acetonitrile. Over a period of several days the complex gains 7.66% in weight on exposure to the atmosphere. This moisture may be removed at 100° over P^O^q in vacuo . Anal. Calcd for Cu(C 0 H N,) (C10,)„-4H 0: C, 50.39; H, 4.20; N, 9.28 JO JU b 422 Cu, 7.07. Found: C, 50.58; H, 3.68; N, 9.42; Cu, 7.12.

PAGE 22

12 Co(P2A2)Cl2'4H20.* In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 1.35 g (0.010 mol) 2 , 6-pyridinedicarboxaldehyde. To this was added a solution of 2.60 g (0.020 mol) anhydrous cobalt (II) chloride in 100 ml of solvent. Upon bringing this solution to reflux, 1.58 g (0.010 mol) 1, 8-diaminonaphthalene in 100 ml solvent was added dropwise. The mixture changed from a deep blue-violet color to black during the reaction. After 48 hr of reflux 150 ml of solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of methanol and ether and dried over P^O^ in vacuo . Yield: 2.0 g (56%) The complex is obtained as a black powder. It is slightly soluble in dimethylformamide, pyridine, and dimethylsulf oxide. The complex is i considered not to be hygroscopic. Anal . Calcd for Co(C 34 H 22 N 6 )C1 2 *4H 0: C, 56.98; H, 4.19; N, 11.73; Co, 8.24. Found: C, 57.30; H, 3.55; N, 11.50; Co, 8.33. ^ U 2^2^2^^4 *^2^‘ a ^00-ml f3as k, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 absolute methanol, was dissolved 1.21 g (0.009 mol) 2 , 6-pyridinedicarboxaldehyde. To this was added a solution of 5.40 g (0.040 mol) anhydrous copper (II) chloride in 100 ml solvent. To the refluxing solution was added dropwise 1.58 g (0.010 mol) 1, 8-diaminonaphthalene in 100 ml solvent. The mixture immediately changed from bright green to dark brown in color. After 48 hr of reflux, 150 ml solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small * P 2 A 2 = condensate of two moles PDC and two moles DAN

PAGE 23

13 portions of methanol and ether and dried over P,0„„ in vacuo. Yield: 4 10 2.8 g (73%) The complex is obtained as a black powder. It is slightly soluble in dimethylformamide and pyridine. The complex absorbs atmospheric moisture. After 24 hr in a drying pistol at 100° over P,0^ in vacuo the 4 10 complex loses 13.4% by weight. Anal . Calcd for CuCC^H^N^Cl^I^O: C, 47.72; H, 3.51; N, 9.82; Cu, 14.85. Found: C, 48.36, H, 3.17; N, 9.57; Cu, 14.85. Cu (P^A^) (C10^) ^ * 2^0). In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 100 ml absolute methanol, was dissolved 1.35 g (0.010 mol) 2,6-pyridinedicarboxaldehyde. To this was added a solution of 7.41 g (0.020 mol) hydrated copper (II) perchlorate in 100 ml of solvent. To the refluxing, light blue solution ( was added dropwise 1.58 g (0.010 mol) 1,8-diaminonaphthalene in 100 ml of solvent. The mixture slowly turned very dark brown in color. After 24 hr of reflux a dark solid was observed. After 48 hr 150 ml of solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of methanol and ether and dried over P^O^ in vacuo . Yield: 4.2 g (100%) The complex is obtained as a black powder. It is slightly soluble in dimethylformamide, pyridine, and dimethylsulf oxide. Over a period of several days the complex gains 5.28% by weight while exposed to the atmosphere. Anal. Calcd for Cu(C 0/ H 00 N,) (CIO.) *2H 0: C, 50.18; H, 3.20; 34 ZZ 0 4 z z N, 10.33; Cu, 7.87. Found: C, 51.02; H, 3.10; N, 9.53, Cu, 8.04. CutDAN^CNO^^In a 250-ml flask, equipped with a reflux condenser. magnetic stirrer and drying tube, containing 75 ml absolute methanol, was

PAGE 24

14 dissolved 1.58 g (0.010 mol) 1,8-diaminonaphthalene. To this was added a filtered solution of 1.17 g (0.005 mol) hydrated copper (II) nitrate in 75 ml solvent. The mixture refluxed for 2 hr and was filtered hot. The filtrate was deep purple in color. The product was collected on a sintered glass filter, washed with several small portions of solvent and ether and dried over P^O^ in vacuo . Yield: 2.3 g (46%) The complex is obtained as a black-green powder. It is slightly soluble in acetone, dimethylformamide, pyridine and dimethylsulf oxide . The complex is not considered to be hygroscopic; over a period of a week, one sample gained 0.59% by weight. Anal . Calcd for Cu(C Q H N ) (NO.^: C, 47.62; H, 3.97; N, 16.67; Cu, 12.70. Found: C, 47.70; H, 4.09; N, 16.73; Cu, 12,68, Attempted preparation of DDnTC . In a 500-ml flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 200 ml absolute methanol, was dissolved 1.35 g (0.010 mol) 2 , 6-pyridinedicarboxaldehyde. The solution was brought to reflux and 1.98 g (0.012 mol) 1,8-diaminonaphthalene in 100 ml solvent was added dropwise. After 6 hr of reflux, a brick-red solid was observed. The mixture was refluxed for 48 hr, after which 100 ml of solvent was distilled off and the hot mixture filtered. The product was collected on a sintered glass filter, washed with several small portions of solvent and ether,' and dried over P^O^q in vacuo . Yield: 1.0 g (36%) The ligand is obtained as a red-orange powder which darkens to winered on standing. It is slightly soluble in ethyl acetate, acetone, chloroform, methylene chloride, dimethylformamide, pyridine, and dimethylsulf oxide , Anal . Calcd for • 2H 2 0: C, 74.18; H, 4.73; N, 15,27. Found: C, 75.12; H, 5.00; N, 15.18

PAGE 25

15 Attempted preparation of TMTC . In a 1-liter flask, equipped with a reflux condenser, magnetic stirrer and drying tube, containing 400 ml n-butanol, was dissolved 6.52 g (0.040 mol) 2,6-diacetylpyridine. This solution was brought to reflux and 2 drops of concentrated sulfuric acid added. A solution of 7.90 g (0.050 mol) 1 , 8-diaminonaphthalene in 400 ml n-butanol was added dropwise to the hot mixture. After 48 hr of reflux 300-400 ml of solvent was distilled off. No solid was observed in the dark brown reaction mixture. A small portion (10-15 ml) of the mixture was combined with 75 ml hexane and allowed to stand for 30 min. A light tan-grey solid was collected on a Buchner funnel, washed with a small portion of hexane, and sucked dry. mp, 105-6° C with decomposition. The solid is slightly soluble in chloroform, dimethylf ormamide, methanol, methylene chloride, and ethanol. ( Anal . Calcd for C 3g H 3() N 6 : C, 80.00; H, 5.26; N, 17.74. Found: C, 76.57; H, 6.10; N, 14.41. Elemental analysis . All CHN microanalyses were performed by Atlantic Microlab, Inc., Atlanta, Ga. Metal analysis . Metal analyses were obtained using a Perkin-Elmer Model 290-B atomic absorption spectrometer. All samples were analyzed in aqueous solution after digesting almost to dryness in 20 ml 1:1 mixture of concentrated nitric and perchloric acids. Calibration curves were obtained by employing aqueous solutions prepared from certified atomic absorption reference solutions (Fisher Scientific Co.). Apparatus Magnet . The magnetic susceptibilities were determined by the Guoy 26 method. The magnet used was a Varian Associates Model V-4004 equipped with four-inch cylindrical pole pieces, separated by an air gap of 2-1/4 inches. A Varian Associates Model V-2501-A current regulator was used to

PAGE 26

16 _3 provide a constant current (+ 1x10 amp) . The maximum field strength 3 attained was 6.75 x 10 gauss. The magnetic field was calibrated by 27 using mercury(II) tetrathiocyanatocobaltate(II) . The current regulator was powered by a Varian Associates Model V-2300-A power supply. Temperatures between 95° and 350° K could be maintained within +0.1 degree as determined by a platinum resistance thermometer. Cryostat and temperature control . The cryostat and temperature con^ 28 29 trol apparatus used were of the basic design of Figgis and Nyholm. ’ Temperatures between 95° and 350° K could be maintained with less than +0.1 degree fluctuation. Sample tube . The sample tube was made of a cylindrical piece of quartz approximately 3.5 mm inside diameter and 17 cm in length and sealed at one end. Approximately 16 cm was used for containing the sample volume. It was suspended in the cryostat from a semi-micro balance by a gold chain attached to a tapered Teflon plug. The diamagnetic correction of the tube was measured as a function of temperature. Balance . A Mettler Model B-6 semi-micro balance of 0.01 mg sensitivity was used to measure the force exerted by the magnetic field upon the sample. Spectrometers . The solution visible and ultraviolet electronic spectra were obtained by using a Cary Model 15 recording spectrophotometer. The solid state diffuse reflectance spectra were obtained by using a Cary Model 1411 Diffuse Reflectance Accessory in conjunction with a Cary Model 14 recording spectrophotometer. Magnesium carbonate was employed as reference material. Infrared spectra were obtained using a Perkin-Elmer Corp. Model 137B NaCl prism and 237B grating spectrophotometers, Also employed was a Beckman Model IR-10 grating spectrophotometer. All spectra were calibrated

PAGE 27

17 with polystyrene. The pressed KBr pellet technique was used. "^H nmr spectra were measured on Varian Associates Models A60-A and XL-100 nuclear magnetic resonance spectrometers (TMS reference) . Electron spin resonance spectra were obtained on microcrystalline samples using the Varian Associates Model E-3 recording spectrometer. Mass spectra were obtained on AEI Scientific Apparatus Model MS-30 double-beam, double-focusing mass spectrometer equipped with a DS-30 data system. Each solid sample was run by direct introduction probe. Probe temperatures ranged from 200° to 340° C. Dr. R.W. King kindly performed these spectral analyses. Melting point apparatus . A Thomas Hoover "Uni-Melt" capillary melting point apparatus was used; the temperatures are uncorrected. Conductance apparatus . Conductances were measured using an Industrial Instruments, Inc., Model RC-18 Conductivity Bridge and a cell with a constant of 1.464 cm -*-. A constant temperature of 25° C ± 0.02 was maintained by the use of a water bath, regulated by a Sargent Thermonitor, Model SW. The units of conductance obtained were mho cm specific conductance.

PAGE 28

RESULTS AND DISCUSSION Ligands Attempts were made to prepare the macrocycles DDnTC and TMTC using procedures generally followed for such ligands. 1 ’ 18-20 Reaction mixtures were dilute in reactants in order to avoid the formation of oligomeric side-products. Several dehydrating and azeotroping solvents were used to ensure completion of reaction, i.e., the removal of product water. Variations in reactant mole ratios (other than 1:1) were employed also. The initial attempt to produce TMTC proved to be the most encouraging; however, cyclization did not occur. Following the procedure of Stotz, 8 a slight excess of diamine was added to a dilute refluxing solution of DAP in n-butyl alcohol. In view of the fact that Schiff base condensations are frequently acid catalyzed 10 and because of the lesser relative reactivity of a ketone as compared to an aldehyde, several drops of concentrated sulfuric acid were used as catalyst. Because of the insolubility of macrocycles in most organic solvents, solid product was expected to begin forming early in the course of reaction. However, after 48 hr of reflux no solid was observed. Upon addition of excess hexane to a small portion of reaction mixture, a product precipitated which proved to be the impure acyclic trimer consisting of two moles of DAN condensed with one mole of DAP. The mass spectrum of this impure product which decomposes at 105-106° C contained the parent ion at m/e 443, the molecular weight of the trimer. Other lighter fragments in the mass spectrum could not be identified 18

PAGE 29

19 (Table 1-A) . The electronic and infrared spectra support the conclusion that the product isolated contained some trimer ( vide infra) . Using 2,2'-dimethoxypropane as solvent and two drops of concentrated sulfuric acid, a small portion of greyish, light brown solid was obtained. This material decomposed above 190° C. Elemental analysis was inconsistent with either the formation of the trimer or the macrocycle. The reactants were also refluxed in n-butyl alcohol using ammonium chloride as catalyst. The ammonium ion is a much weaker acid than the hydronium ion and should not react readily with the diamine. A small amount of dark brown solid was formed but elemental analysis data do not agree with the calculated values for either the trimer or the macrocycle. The color of the reaction mixture was identical to that of the n-butyl alcohol preparation. ( Thin-layer chromatography on alumina and silica gel did not show any separation of components in the above solids. Various polar to nonpolar organic solvents were employed as eluants. A small amount of the impure trimer was dissolved and refluxed in n-butyl alcohol. Several drops of concentrated sulfuric acid were added with slow addition of a dilute DAP solution in the hope that cyclization would occur. But, after 23 hr of reflux, no solid had formed. An attempt was made to deuterate the impure trimer to show evidence of a primary amine. A sample of the trimer was dissolved in a small portion of methanol and several drops of D 2 0 added. This solution was evaporated in a vacuum desiccator and a KBr pellet of the product was prepared. The infrared spectrum showed no observable N-D or 0-D absorptions. The -4i nmr spectrum of a saturated solution of the impure trimer in CDCI 3 was attempted; however, solute concentration is so low that no

PAGE 30

20 resonances could be observed. The spectrum of saturated solution of the trimer in d^-DMSO gave resonances at 66.55-7.15 (broad multiplet) and 61.55 (singlet). The former resonance is indicative of the overlap of pyridine and naphthalene protons and the latter of the methyl groups. Integration revealed a ratio of 2:1, respectively, as compared to the 15:6 ratio expected for the pure trimer. In the attempt to produce the free ligand, DDnTC, methanol proved to be the best reaction medium. No acid catalyst was used in the reactions. N-butyl alcohol and 2,2'-dimethoxypropane were also employed as reaction media. In all cases the solid formed was red-brown to brick-red in color. Elemental analyses showed consistently low carbon and nitrogen percentages except for the product separated from methanolic mixtures. Similarities of the compounds include not only limited solubilities t but also relative high decomposition temperature. At atmospheric pressure the melting point is above 300° C. Upon examination of the sample after the mass spectrum was taken, it was noted that some melting and sublimation had occurred (340° C and 10 ^ torr) . A small portion of the impure ligand dissolved in hot chloroform and was recrystallized by dropwise addition of cold cyclohexane. This procedure did not improve the purity of the macrocycle. An attempt was made to prepare the hydrogen chloride adduct in the hope of obtaining a single product. A small amount of the impure ligand was dissolved in hot chloroform and a moderate stream of dry hydrogen chloride was passed through the solution for 15 min. After 1 min the deep red solution turned brown and solid formed. The dried product was a tan powder which had an infrared spectrum only very slightly different than that of the free ligand, i.e., a broader absorption between 2600 and 3300 cm -*-.

PAGE 31

21 However, again elemental analysis showed the presence of a mixture of products. The adduct placed in distilled water changed the pH gradually from 6 to 2 over a period of five days; a similar change of pH was noted for the TMCD adduct. ^ The excess solid in contact with the saturated solution did not change color during this time. The mass spectrum of the original reaction product is characteristic of a mixture rather than the free ligand sought (DDnTC) . A small peak was observed at m/e 514, which would represent the parent ion of the free macrocycle; but, a more intense peak at higher m/e was also noted. The only other identifiable peak was at m/e 410. This would be the parent ion minus a (C^lOCHN fragment (Table 2-A) . The 1 H nmr of the impure ligand was attempted in CDC1_ and d,-DMS0. -3 6 The saturated d^-DMSO solution gave absorptions at 66.6-7.05 (broad f mu ltiplet) and at 65.55 (broad singlet). The former is indicative of the overlap of pyridine and naphthalene protons. Integration revealed a ratio of 26:3, respectively, a disproportionate ratio in view of the possible overlap and presence of impurities. The ligand was not sufficiently soluble in CDCl^ to give any observable absorptions. Electronic and infrared spectral data support the inclusion of some macrocycle in the impure product ( vide infra ) . Thin— layer chromatography shows no separation into components on either alumina or silica gel even though various polar and nonpolar organic solvents were used as eluants. The unsuccessful attempts to isolate the free macrocycles are not 1 14 18 19 considered unusual. ’ ’ * Because the initial intent was to investigate the complexes of these macrocycles, the impure "free" ligands were not further characterized. Because the attempts were unsuccessful in isolating the free macrocycle

PAGE 32

22 and because of the lack of proof ( vide infra ) for macrocyclic structure in the complexes, ^ 2^2 "*" S usec ^ as general notation for the condensation product of two moles of DAP with two moles of DAN and P 2 A 2 for two moles of PDC condensed with two moles of DAN. Complexes For the purpose of brevity the seven complexes will be hereafter denoted as follows (see also Figures 1 and 2) : Compound Number Cu(D 2 A 2 )(N0 3 ) 2 *4H 2 0 I Cu 2 (D 2 A 2 )C1 4 Â’2H 2 0 II Cu(D 2 A 2 )(C10 4 ) 2 -4H 2 0 III Co(P 2 A 2 )C1 2 -4H 2 0 IV Cu 2 (P 2 A 2 )C1 4 '4H 2 0 V Cu(P 2 A 2 ) (C10 4 ) 2 '2H 2 0 VI cu(dan) 2 (no 3 ) 2 VII Synthesis of each complex followed the straightforward template method."*' All attempts made to produce each of the complexes via the direct method were unsuccessful. First row transition metals, in general, form very stable amine complexes, hence the order of addition of reactants to the reaction mixture is important. Accordingly, amine solutions were added to either ketoneor aldehyde-metal salt solutions. Anhydrous methanol proved to be a satisfactory solvent in that all starting materials possessed high solubility in it and the products were of low solubility. The strategy of the syntheses was to choose counterions which generally do not coordinate with the central ion and yet give reacting salts which are soluble in organic solvents. For these reasons the

PAGE 33

23 nitrate, chloride and perchlorate salts of several metals were used. The only isolable compounds which appeared to be macrocyclic complexes were those of Cu(II) and one of Co (II), although salts of nickel (II), manganese (II) and zinc (II) also were employed in the attempts. The analytical results for the nickel (II) and manganese (II) products did not fit the formation of complexes. In the zinc-containing system, the only compound isolated appeared to be the bis (diamine) complex. These results are not surprising in view of the selectivity of the "hole" of the macrocycle in relation to metal ion size^ (Table 3-A) . Also many first row transition metals form thermodynamically stable amine complexes. An unusual phenomenon for some macrocycles is the difficulty encountered in obtaining elemental analyses which agree precisely with reasonable formulation of the complexes. ^ Because of the unusual stai bility of these complexes, accurate elemental percentages may be impossible to obtain. Complex I repeatedly gave low nitrogen analyses ( ca . 1%) . Stotz^ also noted this in his related complex prepared by the in situ method (ca . 1.5% low). But these complexes are nitrates. Therefore, it may be that, since both of these compounds contain nitrogen in the counterion as well as in the ligand, analytical procedures presently used may be inadequate for accurate nitrogen determinations. Complex II presents consistently low copper analyses ( ca . 1.5%). Copper standards were checked several times; furthermore, numerous copper determinations, of good precision, were made. This difficulty was considered to be a characteristic of the complex because analyses for carbon, hydrogen, and nitrogen are in good agreement with the theoretical amounts. At this point the author cannot suggest a reasonable solution to the problem.

PAGE 34

24 Several other complexes show small inconsistencies in carbon and hydrogen analyses. But, these problems appear to go hand-in-hand with compounds of very high stability. 2 Because stabilities of macrocycle complexes are known to be abnormally high, I more than 8 hr was required for the complete digestion of samples in preparation for metal analyses. The extreme stability of all seven complexes is further noted in their inactivity towards hydrogen sulfide in dilute pyridine solutions. Metal sulfide did not precipitate although mixtures were allowed to stand for a period of several days. Another factor which may lead to the troublesome results in elemental analyses is that all the complexes, with the exception of IV, will dehydrate and hydrate reversibly. During microanalytical procedures rapid weight gain was reported by Atlantic Microlab, Inc.; this was substantiated by the author after water was reversibly removed at elevated temperatures over in vacuo and gained under atmospheric conditions. Small amounts of samples were placed in tared vessels and subsequently dried, weighed, exposed to atmospheric moisture and weighed again. Anhydrous products appear to be formed in absolute methanol which hydrate. Water in analyzed Water gained by Compound product anhydrous form 1 4H 2 0 4H 2 0 II 2H 2 0 3H 2 0 HI 4H 2 0 4H 2 0 IV 4H 2 0 V 4H 2 0 4H 2 0 VI 2H 2 0 3H 2 0 with ease, on work-up in the atmosphere. One cannot definitively differentiate between water in lattice sites, water coordinated to the metal

PAGE 35

25 ion, or water included as carbinolamine linkage(s) . 33 The 1 H nmr spectra are not obtainable because of low solubilities and the presence of paramagnetic ions. Infrared spectra are inconclusive because 0-H absorptions of the three possibilities occur in the same high energy infrared region and because metal-oxygen absorptions are generally present in broad regions where aromatic peaks also occur. 3 ^ All measurements were taken on either completely hydrated or dehydrated samples . Since it is highly probable that the metal ions in these complexes are coordinated in an approximately planar arrangement of nitrogen atoms, it was desired to prepare a simpler model complex for comparison (Figure 4). Compound VII was selected for this purpose. The preparation of VII was completely straightforward and rapid; it is not hygroscopic. Spectral comparisons will be noted in the next sections. The preparation of the analogous cobalt (II) complex was attempted using chloride, nitrate, and perchlorate as counterions. The bis (diamine) complex of Co (II) was not successfully prepared in these trials, which were performed in absolute methanol. Infrared Spectra The infrared spectra of 2,6-pyridinedicarboxaldehyde, 1,8-diaminonaphthalene, and impure DDnTC are presented in Figure 5. The noteworthy differences between the infrared spectra of the ligand and its precursors are the disappearance of the characteristic carbonyl band of the dialdehyde at 1700 cm -*and of the primary amine bands of the diamine lying near 3300 cm -*and 900 cm -*-. With the exception of the characteristic carbonyl and primary amine absorptions, the spectrum of the macrocycle is essentially a composite of the spectra of the parent compounds.

PAGE 36

26 B IS (1 ,8 -Dl AMINONAPHTHALENE) COPPER(H) NITRATE Fig. 4 Structural formula of Cu (DAN) 2 (NO3) 2

PAGE 37

Transmittance (%) 27 PDC DDnTC DAN * 1 1 * * » 1 I 4000 3000 2000 1500 1200 I000 POO 800 700 Frequency (cm -*-) Fig. 5 Infrared spectra of DDnTC and its precursors

PAGE 38

28 The broad band at 1600 cm has been assigned to the imine stretching mode, and 0-H deformation, and higher frequency ring vibrations. An imine stretching mode in this region is characteristic of an imine conjugated with an aromatic system. Isolated imines absorb at higher energies, ^a. 1670 cm -*-. Unsubstituted, uncoordinated pyridine has characteristic absorptions lying at 1580, 1570 and 1485 cm ^.^ The two bands at higher energy arise from the interaction between the C=C and C=N vibrations of the ring. The infrared spectra of 2,6-diacetylpyridine, 1,8-diaminonaphthalene, and the impure trimer of TMTC are presented in Figure 6. It is apparent that primary amine groups are contained in the trimer because of the presences of the band at 3250 cm -*-. The mediumintense absorption at i 2900 cm”*-, which also appears in the Sadtler spectrum, may be due to an impurity in the DAN. The relative intensity of this absorption decreases after recrystallization of DAN; the freshly recrystallized diamine oxides very rapidly even in the absence of light. The carbonyl band of the diketone at 1700 cm *has all but disappeared. The strong absorption at 1600 cm *contains the imine stretching mode. With these exceptions, the spectrum of the trimer is a composite of the absorption bands of its precursors. The infrared spectra of all the complexes investigated contain bands which are much broader than those of the respective precursors. This is typical of metal complexes prepared as wafers. All samples exhibit a disappearance of the primary amine bands at 3300 and 900 cm *and of the carbonyl band at 1700 cm * . Water vibration, be it lattice or coordinated, appears as a broad band centered at 3100-3400 cm *-.-*^ Each

PAGE 39

Transmittance (%) 29 DAP Trimer of TMTC DAN i i i i 1 1 1 — 4000 5000 2000 1500 1200 \000 £00 800 Frequency (cm 1) Fig. 6 Infrared spectra of impure trimer of TMTC and its precursors

PAGE 40

30 displays a strong, broad absorption centered at approximately 1600 cm -1 , containing the inline stretching mode, 0 — H deformation, and high frequency ring vibrations. Complex I (Figure 7) displays bands at 1390 and 825 cm -1 , both typical of ionic nitrate . ^ The former is attributed to the asymmetric stretching mode of NO 3 and the latter to the deformation mode of the ion. Complexes II, IV, and V (Figure 8-10), metal chloride salts, have infrared spectra which are virtually identical. The imine and aromatic absorption, which are broad, lie at the expected frequencies. Complexes III and VI (Figure 11 and 12) are both perchlorate salts. The intense, broad, semistructured band centered at 1100 cm -1 and a peak at approximately 625 cm ^ are typical of ionic perchlorate."^ The former absorption represents the asymmetric stretching modes, and the latter an asymmetric bending mode. Also, an infrared-forbidden band at 930 cm l (symmetric stretching mode) is often observed as a weak absorption because of the reduction of the Tj symmetry of the anion within the lattice; it is not observed in these compounds. The broad and composite imine and aromatic absorption envelope is the same as that noted in the spectra discussed earlier. Complex VII (Figure 13) was synthesized as a comparison to the above complexes. Primary amine bands are evident at 3300, 3100 and 1635 cm -1 . A strong absorption at 1390 cm is characteristic of ionic nitrate. Other bands are similar to the absorptions of the precursor. A Framework Molecular Model representation of each macrocycle presents the pyridine nitrogens at significantly greater distance from the metal ion in comparison to the imine nitrogens. Upon complexation for a

PAGE 41

31 (%) aouB^^TmsuBax Fig. 7 Infrared spectrum of CuCD^Ap (N 03 ) 2 * 4 H 20 (I)

PAGE 42

32 (%) aouBruyuisuBJx Fig. 8 Infrared spectrum of Cu 2 (D^pCl^* 2H 2 0 (II)

PAGE 43

33

PAGE 44

34 Frequency (cm !) Fig. 10 Infrared spectrum of Cu 2 ( P 2 A 2 ) C1 4 ’ ^ H 2° (V)

PAGE 45

35 (%) aouB^^TtnsuBJX 4000 3000 2000 1500 1200 1000 Frequency (cm -*-) Fig. 11 Infrared spectrum of Cu^Ap (C 104 ) 2 ' 4 H 20 (HI)

PAGE 46

36 o o " o CM & u a c Q) 2 cr Q) u (%) sonB^TtnsuHJX Fig. 12 Infrared spectrum of Cu ( P,^ ) (CIO 4 ) 2 ' 2H 2 0 (VI)

PAGE 47

37 (%) aoireinxmsuBJi Fig. 13 Infrared spectrum of Cu (DAN) 2 (NO3) 2 (VII)

PAGE 48

38 given metal ion, the pyridine absorptions at 604 and 405 cm ^ are sensii tive to the stereochemistry of a complex. These absorptions are shifted to higher frequencies ( ca . 50 cm upon coordination. ^ Both bands are weak in uncoordinated pyridine but have proved to be of diagnostic value 33 in determining coordination sites. Spectra of each complex revealed no absorptions at the expected frequencies. If the nitrogens of the pyridine rings are uncoordinated, a strong Lewis acid such as hydrogen chloride should form an adduct at these sites. A stream of dry hydrogen chloride was passed through a weighed sample of Complex I, followed by a gentle flow of nitrogen gas. Upon weighing it was found that the sample had gained a small amount of weight but elemental analysis revealed a nonintegral number of moles of hydrogen chloride added, viz . , between three and four. Although the number of moles of hydrogen chloride added is puzzling, the fact that the hydrogen chloride is absorbed is certainly taken to mean that there are free base sites available. The identification of these base sites could not be accomplished because the infrared spectrum of the adduct was essentially identical to that of the precursor. Electronic Spectra The electronic spectra of the proposed macrocycles are presented in Figures 14 and 15 with absorption maxima listed in Table 1. The spectrum of impure DDnTC in acetonitrile is a composite of its precursors. Absorptions between 50.0 and 40.0 kK encompass the range of the E and K 39 bands observed in benzenoid compounds. These absorptions are assigned to n to tt* transitions. The band at 28.6 kK, attributed to naphthalene, has shifted to slightly lower energy as compared to pure DAN; this would be expected with an electron-donating substituent and the fusion of an additional ring.

PAGE 49

39 aouBqaosqv OATqeqe-g 200 300 400 Wavelength (nm) Fig. 14 Electronic spectrum of impure trimer of TMTC

PAGE 50

40 oounqaosqv aAxqBqa-a 200 300 400 Wavelength (nm) Fig* 15 Electronic spectrum of DDnTC

PAGE 51

Table 1 Near Infrared, Visible and Ultraviolet Spectral Data Compound DDnTC Trimer of TMTC PDC DAP DAN I II III Solvent Wavelength , nm Frequency, kK CH 3 CN 200 50.0 225 44.4 235 42.5 350 28.6 MeOH 203 (sh) 49.3 237 42.2 350 28.6 MeOH Strong absorption below 220 45.4 t 260 38.5 267 37.5 MeOH 211 47.4 237 42.2 275 36.4 MeOH 231 43.3 334 29.9 MeOH 270 37.0 Solid 500 20.0 Solid 363 27.5 480 20.8 ch 3 cn 225 (sb) 44.4 270 37.0 Solid 600 16.7 Solid 360 27.8 IV

PAGE 52

Compound Solvent Table 1 (continued) Wavelength, nm Frequency, kK V Solid 500 20.0 VI ch 3 cn 215 (sh) 46.5 230 43.5 273 35.4 Solid 475 21.1 600 16.7 VII Solid 320 31.3 490 20.4 525 (sh) 19.0 570 17.5 610 16.4

PAGE 53

43 The impure trimer of TMTC in methanol also shows similarities in its spectrum as compared with the precursors. Intense bands lying at 49.3 and 42.2 kK are typical of E and K bands in aromatic moieties. The naphthalene absorption is again observed at 28.6 kK. Bathochromic shifts are expected when a substituent is electron-donating or capable of con39 jugation. Both of the proposed macrocycles as well as the impure trimer have increased conjugation in comparison to the reactants. The electronic spectra of the copper (II) complexes in this study, in general, present over a range of energy from near infrared to the ultraviolet, only one band; in this region diffuse reflectance spectra were required because of the extremely low solubilities of the complexes. Each of the bands has much the same appearance, indicating some common factor in origin. ( Copper (II) complexes are known to exhibit a wide range of possible stereochemistries. Considering the ability of the two macrocycles to coordinate four to six nitrogens, it would be likely to limit the possible stereochemistries to octahedral, square coplanar, tetrahedral, or square pyramidal, including anion coordination. Because macrocycles similar to DDnTC and TMTC exhibit the tendency to remain planar upon complexation, the author will eliminate the tetrahedral arrangement ( vide infra ) . Hathaway^® has summarized the correlation of electronic properties and stereochemistry of numerous {CuN^_^} chromophores. He has shown that square coplanar {CuN^} species generally absorb between 18 and 20 kK. The range of absorption of species studied is between 17 and 21 kK. The [Cufen^]^ cation is often used as a model complex. ^ The approximately tetragonal ligand environment of [Cu(en)~]^ + is D., . 2 4h Its complexes possess a magnetic moment between 1.80 and 1.90 B. M. This is typical of an orbitally non-degenerate B-type ground state.

PAGE 54

44 It is rare that copper (II) complexes are found in pure octahedral 9 2 symmetry. In a cubic field a d ion possesses an E ground term which ~ g is highly susceptible to Jahn-Teller configurational instability.^ Upon tetragonal distortion to an approximate symmetry the 2 E term 2 2 2 9 9 splits into B. and A terms, and the T_ term into B„ and E„ -*-§ 2g 2g g terms. The broad absorption envelopes seen in the copper (II) complexes following can, indeed, be related to symmetry and the d-d transitions centered on the metal ion. In principle the maximum number of d.-d transitons in the electronic spectrum of the copper (II) complex, (CuN^ ^}, can be as many as four (Figure 16) ; in practice, very few complexes give any indication of more than two bands. Many spectra show a main band with a low frequency shoulder only partially resolved. The effect of coordination number upon energies of d-d_ transitions of copper (II) complexes is complicated. In practice, it is the degree of tetragonal distortion which has the major effect in determining the energy of bands in the d-d spectrum of a copper (II) complex. As tetragonal distortion increases from regular octahedral to tetrahedral stereochemistry towards the square coplanar one, the center of gravity of d-d transitions moves to higher energy. ^ In such cases of tetragonal distortion the precise energy levels in synHEetry are uncertain. Depending upon the degree of tetragonal distortion present, three possible energy level sequences may arise: (octahedral) 2 2 2 9 a) Z B < \ < V < Z E lg lg 2g g b) 2 B < 2 B < 2 A < 2 E !g 2g lg g c) 2 B lg < 2 B 2g < ^E g < 2 A lg (square planar) 2 2 The B-^ g -y A^ g transition is believed to be the lowest energy transition and possibly reflects the degree of distortion as the shoulder on the low

PAGE 55

45 CO M CM 60 60 CN] rH 60 60 PQ PQ <2 <2 CN] CN] CN CN k > 1 \ / \ 00 w CM \ N bO CM H CM Y b0 60 b0 CM rH T— ( pq <2 pq CM CN CM k y y \ y y bO W CM O X -H cm ,n q a o x u x n j-i i— i ni ca d d nj O r1 bo d, cfl d
PAGE 56

46 frequency side of the absorption envelope. Polarization spectra of 41 single crystals have tentatively verified this. Thus, the broad absorptions presented by the copper (II) complexes herein are actually envelopes including three absorptions of similar energies, indicative of the {CuN^} chromophore in the presence of a tetragonal distortion. Complex I presents a broad maximum at 20.0 kK (Figure 17). The higher energy absorptions of the macrocycle have shifted to a less intense band lying at 37.0 kK. Bathochromic shifts of ligand absorptions are commonly observed as metal ions complex and participate in the TT-system. The DAN absorption, shifted to lower energy, is probably hidden by the broad envelope centered at 20.0 kK. Complex II has extremely low solubility in solvents transparent in the ultraviolet region. The reflectance spectrum of II (Figure 18) presents two broad absorptions lying at 27.5 kK and 20.8 kK; the former is attributed to charge-transfer absorption and the latter to metal d-cl transitions . The solution spectrum of complex III (Figure 19) shares some similarity with the impure trimer in the ultraviolet region in that bands lying at 42.2 and 28.6 kK have been shifted to longer wavelength upon complexation. The broad envelope centered at 16.7 kK can be assigned to metal d-c[ transitions. Like its analogue, II, complex V (Figure 20) has low solubility in solvents transparent in the ultraviolet region. The only feature in the visible region is the broad absorption attributed to d-d transitions on the metal ion. The solution spectrum of complex VI (Figure 21) presents again a bathochromic shift of ligand absorption in the ultraviolet region. The

PAGE 57

Solution Solid 47 aou-eqnosqv aAxqBja'a Wavelength (nm)

PAGE 58

48 aoueqaosqy aATqBqa-g 300 400 500 600 700 800 900 1000 Wavelength (nm) Fig. 18 Diffuse reflectance spectrum of Cu 2 (D^^Cl^* 2H 2 0 (II)

PAGE 59

Solution 49 aoueqnosqy aATqnqo^ 200 300 400 500 600 700 800 Wavelength, (nm) Fig. 19 Electronic spectrum of CuCD^A^) (ClO^^^l^O (III)

PAGE 60

50 soireqjosqv aATqnqa'H Wavelength (nm)

PAGE 61

Solution Solid 51 aouBquosqy aAT^Bja'a 200 300 400 500 600 700 800 900 Wavelength (nm) Fig. 21 Electronic spectrum of Cu( ^2 A 2'^ (CIO^) 2 ' 2 H 2 O (VI)

PAGE 62

52 bands lying at 50.5 and 44.4 kK in the macrocycle are shifted to 46.5 and 43.5 kK, respectively, while the band lying at 42.5 kK is now found at 35.4 kK. In the reflectance spectrum the shoulder at 21.1 kK may be attributed either to charge transfer absorption or to a metal d-d transition. The broad envelope centered at 16.7 kK can be assigned to d-d transitions on the metal ion. As a reference compound, complex VII was synthesized as a model for determining the stereochemistry of the above complexes. It was observed that VII forms suspensions with solvents commonly used for study in the ultraviolet region. The reflectance spectrum of VII (Figure 22) presents several more distinct absorptions than the macrocyclic complexes. The high energy absorptions at 20.4 and 31.3 kK can be assigned to chargetransfer absorptions. In the spectral study of [CuCen)^] (NO^)^ , Hathaway et_ al.^ have reported absorption at 19.7, 17.9, and 14.1 kK, the latter being a shoulder. The absorptions of VII at 19.0, 17.5, and 16.4 kK are very similar, realizing crystal field splitting parameters of the two 43 diamines, no doubt, differ slightly. Yamada and Tsuchida : in a pre2+ vious work reported a band at 16.7 kK for the [CuCen)^] ion. The three 2 2 2 2 absorptions can be assigned to the transions B^ -* , B^ ®2g’ 2 2 and B^ -> A^, in decreasing order of energies. The model cation is known to have the four nitrogens in a plane about the metal ion with two other groups possibly coordinated along the tetragonal axis.^,^4 Although, visible spectral characteristics have been discussed in terms of d-d transitions, there is also the possibility of superposition of chargetransfer absorptions on d-d_ absorptions. The types of absorptions cannot be separated in the absence of solution spectra in the visible and far infrared regions. All solid spectra were extended

PAGE 63

53 eouFqaosqy FAxaEqa^ 300 400 500 600 700 800 900 1000 Wavelength (nm) Fig22 Diffuse reflectance spectrum of Cu (DAN) 2 (NO 3) 2 (VII)

PAGE 64

54 to 1900 nm (5.3 kK) ; no absorptions were observed in the far infrared region. "Solution" spectra also are actually composites of solution and solid state spectra because of the formation of colloidal dispersions. Relative intensities of solution and diffuse reflectance spectra are only approximate. The visible spectrum of Complex IV is presented in Figure 23. Only two broad absorptions appear; that at 27.8 kK can be assigned to chargetransfer absorption and the very broad band centered at 20.8 kK most likely arises from a d-d transition(s) of the cobalt (II). One would wish to relate the d-d_ transitions centered upon the metal ion to the symmetry of the complex and to the crystal field splitting parameter, Dq. But, the stereochemistry of the metal ion must be approached somewhat indirectly. A tetrahedral environment about cobalt (II) should give bands i in the near infrared region, i.e., between 5 and 6 kK, which might be assigned to the -> ^T (F) transition, ^ and an intense multicomponent 4 4 band associated with the A^ -* T^(P) transition lying between 12.5 and 16.6 kK.^ j n the spectrum of IV neither of these absorptions is seen. In the last decade numerous high-spin, pentacoordinate complexes of 33 cobalt (II) have been reported. These compounds generally exhibit four principal bands lying at 5.6 to 6.0 kK, 12 to 12.6 kK, 15.5 to 16.5 kK, and 19 to 20.5 kK.^ But, this pattern is not observed for IV. Because the complex must be considered spin-free ( vide infra ) and because only spin-paired, square planar cobalt (II) species containing four Co-N bonds have been isolated, this stereochemistry is ruled out. Moreover, the low energy absorption characteristics of square planar complexes are not observed in this case. ^ Rather, the broad absorption of IV centered at 20 kK is characteristic of distorted octahedrally coordinated cobalt (II) . ^9-51

PAGE 65

55 ooueqaosqv OATaeqa^ Wavelength (nm) Fig. 23 Diffuse reflectance spectrum of Co ( ? 2^2 ) C ^2" ^ H 2° (IV)

PAGE 66

56 This envelope is generally considered to contain two spin-allowed d-d 4 transitions, viz . , those from the ground term, T (F) , and the terms *~o 4 4 49 A„ (F) and T, (F) . A small shoulder sometimes is observed on the 2g lg high frequency side as a consequence of spin-orbit coupling serving to 4 4 4 lift the degeneracy of the T (P) term. The weak T (F) -> T (F) -*-o *** o transition is frequently found between 8 and 9 kK. This absorption is not observed in this case. The appearance of this band leads to some confusion in the literature, which is most probably associated with the fact that it is invariably weak (e<10).^^ The complexes [Co (NH_) , ]^ + 3 b 2 + 2 + and [Co(en) ] , as compared to [Co (HO),] , do not exhibit this absorp3 Zb • 54 tion. J However, Ballhausen and Jorgensen have reported the transition for the amine complexes to lie at 9.0 and 9.8 kK, respectively. Assuming the transition occurring at 20.8 kK is from the ground term, 4 4 T^(F), to the A^g term, the crystal field splitting parameter, Dq, for [Co (DDnTC) ] is 1156 cm The Dq values of [Co(en)^]^ + and [Co(NH^)g]^ + for the same transition should be slightly smaller in the absence of con2+ jugation in the ligands. The [CoCen)^] ion exhibits a Dq value of 1130 cm ^ with the value for [Co(NH,),]^ + only slightly less.^ 3 b Strictly speaking, Jahn-Teller distortions would not be expected to occur because of the fact that spin-orbit coupling has removed the de4 42 , . . generacy of the low-lying T^ term, a Kramers doublet. Not realizing the amount of tetragonal distortion, if it occurs in this case, one cannot definitively account for all spectral differences. This and similar systems are complicated and, in general, poorly understood. 1 Magnetic Properties Most transition metal compounds are relatively magnetically dilute, i.e., their paramagnetic centers are isolated from each other by inert 49

PAGE 67

57 ligand molecules. In such compounds the paramagnetic ions act independently of each other. In these cases the idealized behavior for the variation of magnetic susceptibility with temperature is the Curie law, = C/T, where is the susceptibility per mole of paramagnetic material corrected for the diamagnetic effect of the ligand molecules, iC is the Curie constant and T_ the absolute temperature. However, the majority of paramagnetic substances obey a modified version called the Curie-Weiss law, Xj!j = C/(T-0), in which the 6^ is an empirical quantity and is a measure of the deviation of the paramagnetism from the idealized Curie law description. This parameter is determined by plotting 1/x^ versus T and determining the intercept on the T axis. For magnetically dilute paramagnetics 0 is usually a small quantity. Spin-free octahedral cobalt (II) complexes customarily have magnetic moments between 3.89 and 5.2 B. M. and values of 0 between 15° and 30°. In octahedral complexes one rarely observes magnetic moments as low as the "spin-only" moment of 3.89 B. M. High-spin octahedral cobalt (II) compounds give very high orbital contributions to the magnetic moments. This high orbital contribution is attributable to the threefold degeneracy of the 4 T^ ground state. Considering the interaction between the total spin quantum number, S_, and the orbital angular momentum quantum number, L, the magnetic moment may be calculated from [4S(S+1) + L(L+1)]^ where for cobalt(II) S = 3/2 and L = 1. The moment is now found to be 4.12 B. M. ; therefore, in cobalt (II) there is additional orbital angular momentum contributed in some other manner, i.e., from excited states or higher lying terms. Both the spinning of an electron and the movement of an electron in a closed path about the nucleus will produce a magnetic moment. The magnetic properties of any individual atom or ion will result.

PAGE 68

58 therefore, from some combination of these two factors. Orbital contribution may be quenched wholly or partially by the lowering of symmetry. Electric fields of other atoms, ions and molecules surrounding the metal ion in its compound restrict the orbital motion of the electron. The orbital contribution in the cobalt (II) ion is considered to arise from the mixing of some of the next higher orbital triplet level into the singlet by the operation of spin-orbit coupling. The F ground term of cobalt (II) is split into two orbital triplet sets and a singleone.-^ Considerable orbital contribution is seen in Co (P^A^C^* 4H^0 as evidenced by a magnetic moment of 4.57 B. M. at room temperature. This moment is on the low end of the range expected for octahedral or distorted octahedral cobalt (II) complexes. 55 Table 2 lists the room temperature magnetic moments of the compounds herein investigated. There are instances in which the paramagnetic ions influence each other; these spin-spin interaction phenomena are referred to as magnetic exchange interactions. They may arise because the distance between the paramagnetic constituents is small, i.e., there is a direct overlap of atomic orbitals or a direct exchange, or because the intervening diamagnetic atoms are capable of transmitting the magnetic interaction, i.e., superexchange. 55 In cases where there is a spin-spin interaction the Weiss constant, 0, is large, e.g., the copper(II) acetate dimer which exhibits a Weiss constant of 108°. 57 Ordinary mononuclear copper (II) CO complexes present 0 values of 9° or less. In this study those compounds possessing one copper ion per mole of complex were expected to exhibit magnetic moments indicative of the mononuclear species, i.e., 1.70-1.90 B. M. (see Table 2). The exceptions to this rule were the perchlorate salts of each macrocyclic complex. In attempting to completely characterize magnetic properties over an

PAGE 69

Magnetic Susceptibilities and Moments 59 4-1 4-1
PAGE 70

60 extended range of temperature, it was found that both the hydrated and anhydrous forms of the perchlorate salts developed an abnormally large electrostatic charge while measurements were performed. A small amount of beta-emitter was placed in the cryostat in the form of a cesium-137 salt solution impregnating a piece of filter paper. This improved the ease of measurements at room temperature but data at low temperatures were impossible to obtain. The electron spin resonance measurements indicate some form of spin-spin interaction ( vide infra ) . The data obtained from a temperature dependent magnetic susceptibility study of Cu^CP ^ 2 ) 01 ^ and Cu 2 ( P 2 A 2^ are S^ven in Ta ble 3 and reproduced graphically in figures 24 and 25. Each shows a slight increase of magnetic moment with decreasing temperature. The 6 values of 24° and 34°, respectively, which are consistent with a weak spin-spin interaction, were determined from a plot of 1 /yi versus T. The results M of esr measurements are also consistent with the observed interaction (vide infra ) . Measurements from the temperature-dependent magnetic susceptibility study of Cu 2(D2A2)C1^ are included also in Table 3 and graphically presented in Figure 26. The hydrated form of the complex developed too great an electrostatic charge at low temperature for data to be obtained. The anhydrous form did not present a straight line upon plotting 1/ XjJ^ versus Tbut does show an increase in magnetic moment and susceptibility with decreasing temperature. The "knee" observed in this graph was reproduced several times with a rise and fall of temperature. This unusual phenomenon has been noted several times in recent studies , 59-61 particularly with chloride salts of copper (II) complexes. It would appear that above 210° K and below 195° K the complex obeys the Curie-Weiss law. In

PAGE 71

61 Table 3 Temperature Dependence of Average Magnetic and Moments Susceptibilities Compound H o X£>* x !0 3 P e ff’ cu 2 (d 2 a 2 )ci 4 135.2 1.782 1.39 O 00 II CD 167.3 1.252 1.30 P TIM = 1,24 B M '** 180.3 1.146 1.29 190.3 1.057 1.27 197.2 0.787 1.12 200.4 0.785 1.12 208.7 0.787 1.15 211.7 0.822 1.18 220.4 0.764 1.16 ( 221.2 0.769 1.17 248.4 0.710 1.19 259.9 0.660 1.18 297.8 0.539 1.14 Cu 2 (p 2 a 2 )ci 4 136.9 2.452 1.64 0 = 24° 167.3 1.909 1.60 P TIM = 1-60 B ' M ’ 197.3 1.439 1.51 229.8 1.233 1.51 259.2 1.129 1.53 299.0 0.995 1.54 Cu 2 (P 2 A 2 )C1 4 -4H 2 0 135.7 3.021 1.82 e = 34° 166.5 2.294 1.75 P TIM = 1 ' 87 B M 196.8 1.766 1.67 227.6 1.521 1.67

PAGE 72

62 Table 3 (continued) Compound T, °K x A > * x 10^ p B. M.* M eff 252.7 1.364 1.66 279.6 1.273 1.69 298.5 1.328 1.78 299.5 1.309 1.77 * p j-r= 2.83/x' x T per metal atom ett M ** p^-^ = 2 . 83 /XjJj(T+ 0) per metal atom

PAGE 73

10.0 63 Fig. 24 Temperature dependence of inverse susceptibility per copper atom in

PAGE 74

64 Fig. 25 Temperature dependence of inverse susceptibility per copper atom in Cu 2 ( P 2 A 2 )C1 4 -4H 2 0.

PAGE 75

20.0 65 Fig. 26 Temperature dependence of inverse susceptibility per copper atom in

PAGE 76

66 other words, with a decrease in temperature from room temperature, X^ increases and passes through a broad maximum arid again increases in accordance with the Curie-Weiss behavior. Assuming the copper (II) ions are not interacting directly in the "hole" of the macrocycle, several 62 63 rationalizations have been proposed to explain this behavior. ’ These investigators account for the behavior by (1) the presence of magnetically isolated copper (II) ions causing extraneous paramagnetism obeying the Curie-Weiss law at low temperatures or (2) the magnetic interaction between one-dimensional lattices. For an infinite one-dimensional Heisenberg spin lattice, the magnetic susceptibility has been calculated as a function of temperature. It shows a broad maximum and tends to a 64 finite value at very low temperatures. Since this "knee" would appear to be a Neel temperature, the phenomenon has been called linear antif erromagnetism. ^ A 0 value of 58° is consistent with a weak to moderate spin-spin interaction as determined from the plot of 1/X^ versus 65 T. Barraclough and Ng have shown that the one-dimensional Heisenberg model can be used to explain the observed results of anhydrous copper (II) chloride, which possesses a linear structure of copper atoms linked by bridging chlorides and one unpaired electron on each copper atom. In this compound an ordering process takes place below the Ndel temperature. In the Heisenberg model one allows for interaction between nearest neighbors and assumes that if a pair of adjacent atoms have their electron spins parallel there is an interaction energy of ^J. If electron spins on adjacent pairs of atoms are parallel, the interaction energy is negative and the system shows antiferromagnetism. For the critical or Neel temperature 2.J, the energy level separation between the singlet and triplet states, can be estimated from 2J=1.6kT , where k is the Boltzmann constant, The compound ^ 2 ( 02 X 2 ) 0 !^ presents a -2^1 value of 226 cm ^ . The copper (II)

PAGE 77

67 acetate dimer and one of its derivatives exhibit a-2J_ value of 300 cm J ^’^6 and salts of the planar, bridged Cu Cl ^ ion show a -2 J value of zb — _ -j r-j 158 cm . The dimeric complex [Cu^ (TMCD) ] (NO^)^ possesses a-2J value of 56 cm' -1 6 The break or "knee" in the 1/y^ versus T plot (Figure 26) also may be attributed to a phase transformation in the solid, which, for a small temperature range, causes ordering of the magnetic domains. Indeed, if the subnormal magnetic moments are not due to a superexchange mechanism, it is informative to have an understanding of the type of direct metal-metal bonding involved in dimeric complexes. If the copper ions are joined along the z-axis, both the unpaired electrons of the dimer are in the 3d_ 2_ 2 type orbital. Figgis and Mar tin , D ' using x y 69 arguments based on a theoretical treatment of Craig et al . , have pro( posed that the physical origin of the exchange interaction arises from the lateral overlap of these 3d 2 2 orbital functions, i.e., a 6 bond. —x -y — More recent calculations by Ross^ and by Boudreaux^ 1 and an nmr study by Royer have supported this model. Hare et al. have considered the bond length in the Cu^ molecule. They concluded that the only slightly longer copper-copper distances in dimers would seem to rule out existence of a strong direct metal-metal interaction.^ Craig ad.. ^ have calculated that a decrease in the overlap integral of only 10% can halve the interaction energy. Thus, if the copper-copper distance is only slightly greater than that in the acetate complex, the amount of interaction would be decreased markedly. Electron Spin Resonance Studies For an electron of spin s = the spin angular momentum quantum number can have values of m g = ±%, which in the absence of a magnetic

PAGE 78

68 field leads to a doubly degenerate spin state. When a magnetic field is applied, the degeneracy is resolved as represented in the diagram. In an electron spin resonance experiment a transition from the M , = -% to the Mg, = +% state occurs upon absorption of a quantum of radiation. The spectroscopic splitting factor, g, is inversely proportional to the field strength at which the resonance is observed, i.e., E = gBH, where _H is the field strength and is the Bohr magneton. For a free electron £ has the value of 2.0023. In general, the magnitude of _g_ depends upon the orientation of the molecule containing the unpaired electron with respect to the magnetic field^ and the spin and orbital angular momenta. The two isotopes of copper possess a nuclear spin, I_, of 3/2. Therefore, the number of allowed transitions for equivalent interacting copper ions is ( 21 ' +1) or 7, where 1/ = ^(1) + 1/2), as compared to four transitions in the mononuclear case. The situation is now further complicated by the two isotopes of copper which possess nearly identical nuclear moments.^ In antif erromagnetically coupled copper (II) systems, either four or seven lines may be seen; never have sixteen lines been seen, based on 2 nonequivalent copper (II) ions. In solid polycrystalline samples hyperfine splitting is often not seen at low temperatures. The shape and line width of an absorption for a solid is frequently broader than its solution spectrum. Line widths are also altered considerably by exchange processes, i.e., electrons on neighboring lattice sites exchange

PAGE 79

69 spin states rapidly. If the exchange occurs between equivalent ions, the lines broaden at the base and become narrower at the center. When exchange involves dissimilar ions, the resonances of the separate lines may merge to produce a single broad line. For the former situation van Vleck^ has shown that the isotropic exchange interaction contributes to the fourth moment and not to the second moment of the wave function describing the system. This "exchange narrowing" explains why microwave paramagnetic absorption lines are much narrower than one first conjectures from the amount of dipolar coupling. From esr data in Table 4 it is seen that the resonances of those copper (II) complexes presenting subnormal magnetic moments generally show some narrowing as expected for exchange between equivalent ions. The magnitude of narrowing varies from 5 to 30 gauss, in closq agreement with that found in the [C^ (TMCD) ] (NO^)^ 6 system. Three complexes exhibited unexpected, slight broadening at low temperature, i.e., Cu^ (P2^2^'*'4 ’ CuCD^A^) (C10^) ^ * 4^0 , and CuCP^A^) (CIO^^' 7 8 Bleaney and Bowers also noted anomalous broadening at low temperature in their study of the copper (II) acetate dimer. They attributed the broadening to unresolved components of hyperfine structure and possibly thermal vibrations. Owing to the different nuclear magnetic moments of the two abundant isotopes of copper, each of the hyperfine lines may split into a number of components of unequal intensity ( vide supra ) . At low temperature lattice vibrations should have little effect on esr absorptions. However, thermal vibrations may cause small fluctuation in the distance between two interacting ions. If the exchange integral, which is a measure of overlap of magnetic wave functions, is sensitive to 73 distance, it will be sensitive to such fluctuations. Hare £t al. and

PAGE 80

70 Table 4 Temperature Dependence of _g_ Values and Line Widths* and Singlet-Triplet Separations Compound T, °K j^Value W, gauss -2J, cm Cu 2 (d 2 a 2 )ci 4 226 Cu ? (D 9 A 9 )C1 a *3H 9 0 298 2.12 175 273 77 2.12 155 Cu 9 (P 9 A 9 )C1 a -4H 9 0 298 2.14 155 77 2.13 140 Cu (P A )C1 298 2.12 180 231 77 2.14 190 Cu(D ? A 9 ) (C10 /i ) 9 *4H 9 0 298 2.10 145 144 77 2.10 165 Cu (DA) (CIO. ) 9 298 2.09 140 260 77 2.10 135 Cu(P ? A 9 ) (C10 /| ) 9 -3H cS 0 298 2.10 190 220 2 • 37 (g ) 77 2.09 160 cu(p 9 a 9 ) (cio a ) 9 298 2.08 140 422 2-36(6,) 77 2.08 155 Cu(DAN) 9 (N0 ) 298 2.06 45 2-lHiL,,) 77 2.05 40 2.13(jL u ) [Cu 9 (TMCD) ] (N0_) . 298 2.11 140 56 77 2.11 125 Co(P 2 A 2 )C1 2 -4H 2 0 77 3.72 * Full width between extrema of first-derivative curves

PAGE 81

71 £ Q Craig et al. have considered the distance factor to be important (vide supra ) . In polycrystalline samples, one usually sees those electronic transitions for which AM^, = ±1. In an axially symmetric field one would expect two fundamental transitions and two resonances for each. In practice two resonances are seen from just one transition, viz , at approximately 2500 and 3200 gauss. The remaining resonances are not seen because of large zero-field splitting relative to the microwave energy employed in the measurement. If dipole-dipole interaction is sufficiently large, the low field resonance is of such low intensity it may not be observed in certain orientations.^ Therefore, in a sample of randomly oriented axially symmetric crystallites, the probability that the axially symmetric axis is perpendicular to the direction of the magnetic field is highest. The perpendicular postion of the spectrum is thus considerably stronger than the parallel portion. This much less intense parallel portion manifests itself by causing a dissymmetry in the first derivative curve. A very low intensity absorption at low field was seen only for Cu(P 2 A 2 ) (C10 4 ) 2 and CuCP^) (C10 4 ) 2 * 2^0. The values of and for the compounds in this study agree well with those compounds for which crystal structures have been determined.

PAGE 82

72 i.e., known dimeric axially symmetric copper (II) complexes. For various dimeric copper (II) complexes varies from 2.05 to 2.10 and 2.20 to . 61,66,78,80-83 , , . 2.40. Mononuclear copper (II) species customarily present j* 41,44 Jvalues of 2.00 to 2.08 and j* u values of 2.1 to 2.2. Also listed in Table 4 are the estimated values of -2^J for those copper complexes with subnormal magnetic moments. The energy level separations have been calculated from the modified form of the van Vleck equation, viz . , '
PAGE 83

73 in line half widths of ± 10 gauss. This is indicative of intrarather than intermolecular exchange. Unfortunately, this technique may not be used to explain the behavior of the mononuclear copper complexes studied. Electron spin resonance spectra for Co(P2A2)Cl2'4H20 were attempted at room temperature and at liquid nitrogen temperature (77° K) . At room temperature the esr absorption becomes so broad as to escape detection by the equipment. Presumably the broadening arises from a decrease in the spin-lattice relaxation time which accompanies either an increase in temperature or an increase in orbital contribution to the magnetic 86 susceptibility. The spectrum of microcrystalline solid [Co(o-phen)^] (C10 ^) 2 yields a Rvalue of 4.2 ± 0.1, 55 considered to be characteristic of octahedral high-spin cobalt (II). An isotropic value of 4 is expected in the strong field limit and of 4.333 in the weak field limit for the ( S' = % level of an octahedral high-spin d 7 ion. The absorption observed at £ = 3.72 for Co (DDnTC) Cl^ * 41^0 is seen only at high signal amplification and also presents a very broad envelope. Low-spin octahedral cobalt (II) complexes are expected to give an isotropic value of 2, observed for the E( G) resonance level. The esr absorption for a lowspin complex is generally narrower than that of high-spin and broadens 8 6 but little as the temperature is increased. Because room temperature measurements were not feasible the author cannot estimate any possible low-spin contribution in the complex. In light of the magnetic moment, the large £ value at 77° K and the absence of an absorption of £ = 2.0, it can be reasonably concluded that the cobalt (II) ion in the complex is essentially high-spin with very little, if any, low-spin contribution present . In view of magnetic and esr data, the task of reasonably concluding

PAGE 84

74 the mechanism of interaction in the copper (II) complexes is not as simple. Lacking structural data one can postulate several possibilities. Indeed, the "hole" of the proposed macrocycles, TMTC and DDnTC, may accommodate two copper (II) ions as shown previously in TMCD.^ Two metal ions may feasibly "fit" in the plane presented by the six nitrogen atoms, thus producing weak metal-metal interaction and the observed magnetic properties and electron spin resonances. The question arises, at this point, as to why the chloride salt permits such enclosure of two metal ions while the perchlorate and nitrate salts give a metal: ligand ration of 1:1. Stotz also observed that TMCD will accommodate two metal ions in the nitrate salt but only one in the perchlorate. It is not altogether unlikely that the perchlorate salts of this study may possess a crystal structure such that the cations are layered or are in such orientation in the solid that metal ions may interact intermolecularly . 87 88 Goedken and Christoph and Mangia nt a^. have observed an unusual helical coordination about one and two metal ions. These studies have involved ligands containing diimines formed with 2,6-diacetylpyridine. The ligands do not present all donor atoms about the metal ion in a planar array. However, the methyl groups are undoubtedly causing steric interactions and force the ligand to "wrap" itself about the two metal ions in a spiral fashion. There is the possibility, also, that the macrocycles TMTC and DDnTC are not formed but that a polymer is produced which coordinates to the metal ion(s) in an helical manner, leaving a coordinated amine group at one end and a free carbonyl at the other. The few end groups should not give strong absorptions in the infrared spectra. The helical form of the polymer would certainly lessen steric interactions of the methyl groups with the naphthalene rings, and would not destroy the

PAGE 85

75 planarity of the imine nitrogens. Such a theory could only be conclusively accepted upon the determination of the crystal structure. Unfortunately, attempts to grow crystals from various solvents have resulted in failure. Further Characterization Attempts Mass spectra of compounds I and II were taken at elevated temperatures. The highest weight fragment in both instances was m/e 158, the molecular weight of DAN. Because no macrocyclic parent ion was obtained, attempts to obtain mass spectral data were abandoned. Indeed, neither melting nor sublimation appeared to have occurred during these runs. Conductance measurements were made on those compounds appearing to be soluble in methanol (I, II, IV, and VI), but, the resulting data were very erratic. The solutions employed in the conductivity measurements were t saturated solutions obtained by stirring a mixture of complex and methanol for several hours. Subsequently, these mixtures were permitted to stand overnight. A portion of the supernatant mixture was removed after which concentrations were obtained by pipetting aliquots from the supernatant, and evaporating the methanol with a gentle stream of nitrogen and weighing the dry residue in the tared flask. Concentrations were in the -4 -6 range of 10 to 10 M. Because the resulting solutions gave scattered concentrations, these data suggested the possibility that true solutions were not formed, but, rather, colloidal dispersions. This suspicion was confirmed subsequently by simple light scattering experiments, i.e., the Tyndall effect. Thus, further measurements were considered to be of no 1,14 value. This phenomenon has been observed in other complex "solutions." However, the mixtures do contain some solvated ions as observed by their low conductances. Other solvents were considered, but again solubilities

PAGE 86

76 are very small. Complexes II and V have much larger solubilities in DMF than do the other compounds. To increase solubility in DMF the solutions were heated. The unexpected observation was made that copper (II) chloride was removed as a consequence of refluxing. DMF is commonly known as a good solvent for copper salts. On this basis it was initially believed that the copper removed was present in II and V as a counterion, viz , tetrachlorocuprate(II) ion. This belief was discarded, subsequently on the basis of electronic spectral measurements ( vide supra ) . Upon addition of a few milliliters of silver nitrate solution to the hot DMF solutions, the formation of silver chloride was noted. Precipitation is not observed at room temperature. This further substantiates the exclusion of copper (II) chloride from the complex.

PAGE 87

SUMMARY The preparation of six (I-VI) new complexes, believed to contain macrocyclic ligand molecules, was accomplished by the Schiff base condensation of 1,8-diaminonaphthalene and either 2 , 6-diacetylpyridine or 2, 6-pyridinedicarboxaldehyde in the presence of copper (II) and cobalt (II) salts. Attempts to isolate free base, prepared via the template method, from the complexes by precipitation of the metal ions with sulfide ion were unsuccessful. Attempts to prepare the free ligands resulted in either incomplete cyclization or oligomerization or both as evidenced by elemental analyses and mass and infrared spectral data. The complexes' were characterized by elemental analysis, infrared, ultraviolet, visible and electron spin resonance spectra, and magnetic susceptibility determinations. The results of these studies support the formulation of each of the complexes as metal ion(s) surrounded by a planar, quadraor hexadentate ligand with counterions or water either very loosely held in the axial positions or present as part of the crystal lattice. Subnormal, room temperature magnetic moments were observed for Cu 2 (D 2 A 2 )C V 2H 2°’ Cu 2 ( D 2 A 2 )C 1 4 ’ Cu ( D 2 V (CIO^, Cu^A^C^ Cu (P^A^) (CIO^) ^ and Cu (P^A^) CIO^) ^ ’ 211^0 . Because of a large electrostatic charge accumulation upon some of the samples, particularly at lower temperatures, reliable temperature-dependent magnetic susceptibility data could be obtained only for Cu^ (D 2 A 2 )C 1 ^ , Cu^ and . The subnormal moments of these complexes range from 1.5 to 1.1 B. M. , less 77

PAGE 88

78 o than the expected 1.7 to 1.9 B. M. characteristic of the d_ system. The complex ^ presented a transition temperature in the range of 197° to 209° K, following the Curie-Weiss law above and below this plateau. Electron spin resonance spectra of those compounds possessing subnormal magnetic moments give £ values and general peak narrowing indicative of spin-spin exchange of either intraor intermolecular character. Extensive structural analysis would be required to aid in the elucidation of the mechanism of this exchange.

PAGE 89

APPENDIX

PAGE 90

80 Table 1-A Mass Spectral Cracking Pattern of Impure Trimer of TMTC m/e (70 eV, 100° C) 443 304 299 262 246 205 204 184 183 Relative Intensity 10 5 7 7 21 12 12 100 73 Table 2-A Mass Spectral Cracking Pattern of Impure DDnTC m/e (70 eV, 340° C ) 516 514 412 411 410 259 258 205 168 Relative Intensity 8.8 1.3 31.1 100.0 24.0 29.8 10.2 23.6 7.4

PAGE 91

Table 3-A Octahedral Ionic Radii of Metal Ions in Oxide Salts Metal Ion Radius, A * Cu (II) 0.73 Co (II) (high spin) 0.74 Ni (II) 0.70 Mn(II) (high spin) 0.82 Zn(II) 0.75 * Taken from R.D . Shannon and C.T. Prewitt, Acta Crystallogr. , B26 , 1046 (1970) .

PAGE 92

BIBLIOGRAPHY 1. L.F. Lindoy and D.H. Busch in "Preparative Inorganic Reactions," W.L. Jolly, Ed., John Wiley & Sons, Inc., New York, N.Y. , 1971, vol. 6. 2. D. St.C. Black and A.J. Hartshorn, Coord . Chem . Rev . , 9_, 219 (1973). 3. D.W. Wester, Ph.D. Thesis, University of Florida, Gainesville, Fla., 1975. 4. D. St.C. Black and I. A. McLean, Aust . J^. Chem . , 24 , 1401 (1971). 5. K. Travis and D.H. Busch. Chem . Commun . , 1041 (1970). 6. R.W. Stotz, Ph.D. Thesis , University of Florida, Gainesville, Fla., 1970. 7. F.A. Cotton and E. Pedersen, Inorg . Chem . , 14 , 383 (1975). 8. F.A. Cotton and E. Pedersen, ibid . , 14 , 388 (1975). 9. F.A. Cotton and E. Pedersen, ibid., 3 A, 399 (1975). 10. F.A. Cotton, T.G. Dunne, and J.S. Wood, ibid . , _3, 1495 (1964). 11. J.N. van Niekerk and F.R.L. Schoening, Acta Crystallogr ., 6, 227 (1953). 12. D.H. Busch, K. Farmery, V. Goedken, V. Katovic, A.C. Melnyk, C.R. Sperati, and N. Tokel, Advan Chem . Ser . , 100 , 44 (1971). 13. J.L. Love and H.K.J. Powell, Inorg . Nucl. Chem . Lett., _3, 113, (1967). 14. S.R. Weller, Ph.D. Thesis, University of Florida, Gainesville, Fla., 1973. 15. F.F. Myers,, Jr., personal communication. 16. R.W. Layer, Chem . Rev . , 63 , 489 (1963). 17. N.F. Curtis, Coord . Chem . Rev., _3, 3 (1968). 18. D.H. Busch, Recor d Chem . Progr . , 25 , 107 (1964). 19. D.H. Busch, Helv. Chim . Acta (spec, ed.), 174 (1967). 20. D.H. Busch, Advan . Chem . Ser . , 37 , 125 (1963). 21. D.H. Busch, D.C. Jicha, M.C. Thompson, J.W. Wrathall, and E. Blinn, J. Amer . Chem . Soc . , 8j5, 3642 (1964) . 82

PAGE 93

83 22. M.C. Thompson and D.H. Busch, ibid., 86 _, 3651 (1964). 23. F.P. Papadopoulos, A. Jarrar, and C.H. Issidorides, J. Org. Chem. , 31, 615 (1966). 24. W. Mathes, W. Sauerlich, and T. Klein, Chem . Ber . , 86 , 585 (1953). O. 25. F. Sondheimer, C. Mancera, M. Urquiza, and G. Rosenkranz, J. Amer. Chem . Soc . , 77 , 4145 (1955). 26. E.A. Clevenger, M.S. Thesis, University of Florida, Gainesville, Fla., 1961. 27. B.N. Figgis and R.S. Nyholm, J. Chem . Soc . , 4190 (1958). 28. B.N. Figgis and R.S. Nyholm, ibid . , 12 (1954). 29. B.N. Figgis and R.S. Nyholm, ibid . , 331 (1959). 30. R.T. Morrison and R.N. Boyd, "Organic Chemistry," 2nd ed., Allyn and Bacon, Inc., Boston, Mass., 1966, Chap. 19. 31. D.H. Busch, L.Y. Martin, L.J. DeHayes, and L.J. Zompa, J. Amer. Chem. Soc . , 96, 4046 (1974). 32. D.H. Busch and 1 G.A. Melson, Proc. Chem . Soc . , London , 223 (1963). 33. J.P. Lafornara, Ph.D. Thesis, University of Florida, Gainesville, Fla., 1970. 34. K. Nakamoto, "Infrared Spectra of Inorganic Complex Molecules," John Wiley & Sons, Inc., New York, N.Y. , 1970. 35. B. Witkop, J_. Amer . Chem . Soc . , 76 , 5597 (1954). 36. L.J. Bellamy, "The Infrared Spectra of Complex Molecules," John Wiley & Sons, Inc., New York, N.Y. , 1958, Chap. 16. 37. B.M. Gatehouse, S.E. Livingstone, and R.S. Nyholm, J. Chem. Soc., 4222 (1957). 38. B.J. Hathaway and A.E. Underhill, ibid . , 3091 (1961). 39. D.J. Pasto and C.R. Johnson, "Organic Structure Determination," Prentice-Hall, Inc., Englewood Cliffs, N.J., 1969, Chap. 3. 40. B.J. Hathaway, J^. Chem . Soc . , Dalton Trans . , 1196 (1972). 41. I.M. Proctor, B.J. Hathaway, and P. Nicholls, J. Chem. Soc. A, 1678 (1968) . 42. C.J. Ballhausen, "Introduction to Ligand Field Theory," McGraw-Hill Co., New York, N.Y. , 1962, Chap. 10.

PAGE 94

84 43. S. Yamada and R. Tsuchida, Bull . Cham. Soc . Jap . , 29 , 289 (1956). 44. B.J. Hathaway, D.E. Billings, P. Nicholls, and I.M. Proctor, J_. Chem . Soc . A, 319 (1969) . 45. L.E. Orgel, J_. Chem . Phys . , 23 , 1004 (1955). 46. F.A. Cotton and R.H. Holm, J_. Amer. Chem . Soc . , 82 , 2979, 2983 (1960). 47. J.S. Wood, Inorg . Chem . , _7, 852 (1968). 48. F.R. Urbach, R.D. Bereman, J.A. Topick, M. Hariharan, and B.J. Kalbacher, J^. Amer . C hem . Soc . , 96 , 5063 (1974). 49. F.A. Cotton and G. Wilkinson, "Advanced Inorganic Chemistry," John Wiley & Sons, Inc., New York, N.Y. , 2nd ed., 1966, Chap. 29. 50. J. Ferguson, J_. Chem . Phys . , 32 , 533 (1960). 51. W. Low, Phys . Rev . , 109 , 256 (1958). 52. B.N. Figgis, "Introduction to Ligand Fields," John Wiley & Sons, Inc., New York, N.Y. , 1966, Chap. 9. 53. G.L. Roberts and F.H. Field, J. Amer. Chem. Soc., 72, 4232 (1950). t 54. C.J. Ballhausen and C.K. Jorgensen, Acta Chem . Scand . , _9, 397 (1955). 55. B.N. Figgis and R.S. Nyholm, J_. Chem . Soc . , 338 (1959). 56. M. Kato, H.B. Jonassen, and J.C. Fanning, Chem . Rev . , 64 , 99 (1964). 57. B.N. Figgis and R.L. Martin, J^. Chem . Hoc., 3837 (1956). 58. B.N. Figgis and C.M. Harris, ibid . , 855 (1959). 59. M. Inoue, S. Emori, and M. Kubo, Inorg . Chem . , 1427 (1968). 60. F.J. Rioux and B.C. Gerstein, J_. Chem . Phys . , 50 , 758 (1969). 61. G.F. Kokoszka, K. Hyde, and G. Gordon, _J. Inorg . Nucl. Chem . , 31 , 1993 (1969). 62. H. Kobayashi, T. Haseda, E. Kanda, and S. Kanda, J^. Phys . Soc . Jap , 18, 349 (1963). 63. S. Kadota, I. Yamada, S. Yomayama, and K. Hiradawa, ibid . , 23 , 75 (1967) . 64. J.C. Bonner and M.E. Fisher, Phys . Rev . , 135 , A640 (1964). 65. C.G. Barraclough and C.F. Ng, Trans . Faraday Soc ., 60 , 836 (1964). 66. J.R. Wasson, C. Shyr, and C. Trapp, Inorg . Chem . , 1 _, 469 (1968).

PAGE 95

85 67. G.J. Maass, B.C. Gerstein, and R.D. Willet, J_. Chem . Phys . , 46 , 410 (1967) . 68. R.W. Jotham and S.F.A. Kettle, Inorg . Chem . , 9_, 1390 (1970). 69. D.P. Craig, A. Maccoll, R.S. Nyholm, L.E. Orgel, and L.E. Sutton, J. Chem . Soc . , 322 (1954). 70. I.G. Ross, Trans . Faraday Soc ., 55 , 1057 (1959). 71. E.A. Boudreaux, Inorg . Chem . , _3, 506 (1964). 72. D.J. Royer, ibid., 4_, 1830 (1965). 73. C.R. Hare, T.P. Sleight, W. Cooper, and G.A. Clarke, ibid., ]_, 669 (1968) . 74. G.F. Kokoszka, M. Linzer, and G. Gordon, ibid-, _7, 1730 (1968). 75. R.S. Drago, "Physical Methods in Inorganic Chemistry," Reinhold Publishing Co., New York, N.Y., 1965, Chap. 10. 76. A. Carrington and A.D. McLachlan, "Introduction to Magnetic Resonance, Harper & Row, Publishers, New York, N.Y. , 1967. 77. J.H. van Vleck, Phys . Rev . , 74 , 1168 (1948). 78. B. Bleaney and K.D. Bowers, Proc. Roy . Soc . , Ser . A, 214 , 451 (1952). 79. G.F. Kokoszka and R.W. Duerst, Coord . Chem . Rev . , 5 _, 209 (1970). 80. R.W. Duerst, S.J. Baum, and G.F. Kokoszka, Nature ( London ) , 222 , 665 (1969). 81. D.J. Hodgson, P.K. Hale, J.A. Barnes, and W.E. Hatfield, Chem . Commun. 786 (1970). 82. J. Lewis, F.E. Mabbs, L.K. Royston, and W.R. Smail, J^. Chem . Soc . A, 291 (1969). 83. S.J. Grubber, C.M. Harris, and E. Sinn, Inorg . Chem . , 7 _, 268 (1968). 84. D.M.S. Bagguley and J.H.E. Griffths, Nature ( London ) , 162 , 538 (1948) 85. D. Kivelson and G. Collins in "Paramagnetic Resonance," Vol. II, W. Low, Ed., Academic Press, Inc., New York, N.Y. , 1963, p. 496. 86. J.G. Schmidt, W.S. Brey, and R.C. Stoufer, Inorg . Chem . , 6^, 268 (1967) 87. V.L. Goedken and G.G. Christoph, i bid . , 12 , 2316 (1973). 88. A. Mangia, C. Pelizzi, and G. Pelizzi, Acta Crystallogr . , B30 , 2146 (1974) .

PAGE 96

BIOGRAPHICAL SKETCH Barbara Judith Romanik was born May 21, 1942, in Millville, New Jersey. She attended Bridgeton (New Jersey) Senior High School and was graduated in June, 1960. In June, 1964, she received the degree of Bachelor of Science in Chemistry from Washington College, Chestertown, Maryland . Ms. Romanik taught in the public school system of Polk County, Florida, between January, 1966, and June, 1971. Graduate study was begun in June, 1968, at Purdue University. Under a National Science Foundation Fellowship and a four-summer sequential program she received the Master of Science degree in Chemistry in August, 1971. In September, 1971, Ms. Romanik enrolled in graduate school at the University of Florida. From that time until June, 1974, she held a graduate teaching assistantship and was an interim instructor for the 1972-73 year. A DuPont Teaching Award was received in June, 1973. From June, 1974, until June, 1975, she supported her own graduate education. Ms. Romanik is a member of the American Chemical Society and its Division of Chemical Education. 86

PAGE 97

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. -4L Gus. J. Pal 2nik Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Richard D. Dresdner Professor of Chemistry

PAGE 98

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. This dissertation was submitted to the Graduate Faculty of the Department of Chemistry in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1975 Larry L. Hench Professor of Materials Science and Engineering Dean, Graduate School