UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Transition-metal imidazolate polymers : a new family of molecule-based magnets Sánchez, Víctor 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-715337.pdf [ 31.9MB ]
Metadata
JSON: 831-1.0061270.json
JSON-LD: 831-1.0061270-ld.json
RDF/XML (Pretty): 831-1.0061270-rdf.xml
RDF/JSON: 831-1.0061270-rdf.json
Turtle: 831-1.0061270-turtle.txt
N-Triples: 831-1.0061270-rdf-ntriples.txt
Original Record: 831-1.0061270-source.json
Full Text
831-1.0061270-fulltext.txt
Citation
831-1.0061270.ris

Full Text

TRANSITION-METAL IMIDAZOLATE POLYMERS: A NEW FAMILY OF MOLECULE-BASED MAGNETS by VICTOR SANCHEZ B.Sc, Universidad Autonoma del Estado de Mexico, 1986 M.Sc., Universidad Autonoma del Estado de Morelos, 1992 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Chemistry) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA July 2001 © Victor Sanchez, 2001 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada Date DE-6 (2/88) ABSTRACT One-, two- and three-dimensional transition-metal coordination polymers involving imidazolate-based ligands have been prepared and characterized structurally and magnetically. A 1-D material, [Fe(pz)2]x (pz = pyrazolate), which exhibits weak antiferromagnetic exchange (short-range), was found to possess a chain type structure in which metal ions are doubly bridged by pyrazolate ligands. In contrast, when imidazolate-type ligands were utilized in the synthesis of binary metal-azolate complexes, 3-D extended systems were produced as a consequence of the single-bridging of metal ions characteristic of imidazolate ligands. Hence, [Fe(4-abimid)2]x (4-abimid = 4-azabenzimidazolate), and its cobalt analogue, both of which have a novel 3-D single diamondoid structure, were prepared. Both of these materials exhibit long-range ferromagnetic ordering at low temperatures. [Co(imid)2]x, (imid = imidazolate); [Cu(2-meimid)2]x (2-meimid = 2-methylimidazolate); [Co(benzimid)2]x, [Ni(benzimid)2]x and [Cu(benzimid)2]x (benzimid = benzimidazolate); [Cu(4,5-dichloroimid)2]x (4,5-dichloroimid = 4i5-dichloroimidazolate); and [Co3(imid)6(imidH)2]x (imidH = imidazole), all exhibit magnetic behaviour that classifies them as molecule-based magnets. Indirect evidence suggests that these materials also have extended 3-D lattices. [Fe2(imid)4(bipy)]x (bipy = 2,2'-bipyridine), [Co2(imid)4(bipy)]x and [Fe4(imid)8(terpy)]x (terpy = 2,2':6',2"-terpyridine), have 2-D structures, a structural motif never before seen in polymetallic imidazolates. The 'pyridine' molecules act as chelating, ii capping, ligands which separate the extended sheets of imidazolate-bridged metal ions in these materials. [Fe2(imid)4(bipy)]x is unique in exhibiting two structural phase transitions. Both [Fe2(imid)4(bipy)]x and [Co2(imid)4(bipy)]x exhibit long-range ferromagnetic ordering at low temperatures while [Fe4(imid)8(terpy)]x shows more complex magnetization behaviour. Al l three of these materials can be considered molecule-based magnets. [Fe(l-Me-2-S-imid)20.5Cp2Fe]x (l-Me-2-S-imid = l-methyl-2-thioirnidazolate; Cp 2Fe = ferrocene), was obtained as a rare example of a 1-D chain polymer that exhibits long-range magnetic ordering. Alternating FeN 4 and FeS 4 chromophores along the chains is a unique structural feature of this material. The single-bridging imidazolate ligands involved in most of the compounds studied here are efficient mediators of magnetic exchange interaction between metal centres. The observation of antiferromagnetic behaviour above a critical temperature, T c , and long-range ferromagnetic ordering below T c suggests canted spin structures for many of these compounds. Importantly, long-range tJhree-dimensional ordering of the residual spins, arising from the canting, leads to net magnetization at zero applied field. These magnetic properties classify these novel materials as molecule-based magnets. iii T A B L E OF CONTENTS ABSTRACT ii LIST OF TABLES xii LIST OF FIGURES xiv LIST OF ABBREVIATIONS A N D SYMBOLS xxvi A C K N O W L E D G M E N T S xxx Chapter 1 INTRODUCTION 1 1.1 MAGNETISM 2 1.1.1 INTRODUCTION 2 1.1.2 MAGNETIC E X C H A N G E 14 1.1.3 M O L E C U L E - B A S E D MAGNETS 17 1.2 DIMENSIONALITY A N D CONNECTIVITY 19 1.3 COORDINATION POLYMERS 22 1.4 DIAZOLES A N D DIAZOLATES 24 1.5 PHYSICAL METHODS OF CHARACTERIZATION 27 1.5.1 MAGNETIC SUSCEPTIBILITY DETERMINATION. 27 1.5.2 X - R A Y DIFFRACTION 30 1.5.3 SPECTROSCOPIC METHODS .31 1.5.3.1 INFRARED 31 1.5.3.2 UV-VIS-NTR 32 1.5.3.3 NMR 32 iv 1.5.3.4 MOSSBAUER 32 1.5.4 T H E R M A L GRAVIMETRIC ANALYSIS (TGA) 34 1.5.5 E L E M E N T A L ANALYSIS 34 1.6 OBJECTIVES A N D ORGANIZATION OF THIS THESIS 35 REFERENCES 38 Chapter 2 POLYBIS(PYRAZOLATO)IRON(Il). A ONE-DLMENSIONAL MATERIAL SHOWING W E A K ANTIFERROMAGNETIC E X C H A N G E 44 2.1 INTRODUCTION 44 2.2 RESULTS A N D DISCUSSION 45 2.2.1 SYNTHESIS A N D PHYSICAL PROPERTIES 45 2.2.2 SINGLE-CRYSTAL X - R A Y DIFFRACTION CHARACTERIZATION 46 2.2.3 INFRARED SPECTROSCOPY 49 2.2.4 MAGNETIC BEHAVIOR 49 2.3 S U M M A R Y A N D CONCLUSIONS 52 REFERENCES 54 Chapter 3 POLYBIS(4-AZABENZIMIDAZOLATO) LRON(II) A N D COBALT(n). 3-D SINGLE DIAMONDOID MATERIALS EXHIBITING W E A K FERROMAGNETIC ORDERING 56 3.1 INTRODUCTION 56 3.2 RESULTS A N D DISCUSSION 59 V 3.2.1 SYNTHESES, STRUCTURES A N D PHYSICAL MEASUREMENTS 59 3.2.2 MAGNETIC PROPERTIES 70 3.2.3 MOSSBAUER SPECTROSCOPY 83 3.3 S U M M A R Y A N D CONCLUSIONS 86 REFERENCES 88 Chapter 4 BINARY IMLDAZOLATES OF COBALT(II), NICKEL(H), AND COPPER(II) 91 4.1 INTRODUCTION 91 4.2 COBALT(U) IMIDAZOLATE POLYMERS 91 4.2.1 INTRODUCTION 91 4.2.2 RESULTS A N D DISCUSSION 93 4.2.2.1 SYNTHESES, PHYSICAL, T H E R M A L A N D STRUCTURAL CHARACTERIZATION 93 4.2.2.2 MAGNETIC PROPERTIES 104 4.3 A NICKEL(II) BENZIMIDAZOLATE POLYMER 116 4.3.1 INTRODUCTION 116 4.3.2 RESULTS A N D DISCUSSION 117 4.3.2.1 SYNTHESIS, STRUCTURAL, T H E R M A L AND PHYSICAL CHARACTERIZATION 117 4.3.2.2 MAGNETIC PROPERTIES 120 4.4 COPPER(II) IMIDAZOLATE POLYMERS 126 vi 4.4.1 INTRODUCTION 126 4.4.2 RESULTS AND DISCUSSION 127 4.4.2.1 SYNTHESES, STRUCTURAL, THERMAL AND PHYSICAL CHARACTERIZATION 127 4.4.2.2 MAGNETIC PROPERTIES 137 4.5 SUMMARY AND CONCLUSIONS 150 REFERENCES 152 Chapter 5 TWO-DIMENSIONAL IRON(n) A N D COBALT(II) IMIDAZOLATE POLYMERS EXHIBITING LONG-RANGE FERROMAGNETIC ORDERING 155 5.1 INTRODUCTION 155 5.2 POLY-2,2'-BIPYRIDrNETETRAKIS(IMIDAZOLATO) DHRON(n) 156 5.2.1 RESULTS A N D DISCUSSION 156 5.2.1.1 SYNTHESIS, PHYSYCAL AND THERMAL CHARACTERIZATION 157 5.2.1.2 X-RAY DIFFRACTION STUDIES 158 5.2.1.3 MAGNETIC PROPERTIES 169 5.2.1.4 MOSSBAUER SPECTROSCOPY 188 vii 5.3 P O L Y - 2 , 2 ' - B I P Y P J D r N E T E T R A K I S ( I M I D A Z O L A T O ) DICOBALT(H) 194 5.3.1 RESULTS A N D DISCUSSION 194 5.3.1.1 SYNTHESIS, T H E R M A L A N D STRUCTURAL CHARACTERIZATION 194 5.3.1.2 MAGNETIC PROPERTIES 197 5.4 S U M M A R Y A N D CONCLUSIONS 204 REFERENCES 208 Chapter 6 POLY-2,2':6\2''-TEPvPYRIDrNEOCTAKIS(rMIDAZOLATO)-TETRArRON(II). A V E R Y SOFT 2-D M O L E C U L E - B A S E D M A G N E T 210 6.1 INTRODUCTION 210 6.2 RESULTS A N D DISCUSSION 211 6.2.1 SYNTHESIS A N D PHYSICAL PROPERTIES 211 6.2.2 X - R A Y DIFFRACTION STUDIES 213 6.2.3 MOSSBAUER SPECTROSCOPY 218 6.2.4 MAGNETIC PROPERTIES 219 6.3 S U M M A R Y A N D CONCLUSIONS 229 REFERENCES , 231 Chapter 7 POLYBIS( 1 -METHYL-2-THIOIMIDAZOLATO)jRON(n). A ONE DIMENSIONAL MATERIAL EXHIBITING LONG-RANGE MAGNETIC ORDERING 232 viii 7.1 INTRODUCTION 232 7.2 RESULTS A N D DISCUSSION 233 7.2.1 SYNTHESIS, PHYSICAL A N D T H E R M A L PROPERTIES 233 7.2.2 X - R A Y C R Y S T A L L O G R A P H Y 235 7.2.3 MAGNETIC PROPERTIES 238 7.2.4 MOSSBAUER SPECTROSCOPY 247 7.3 S U M M A R Y A N D CONCLUSIONS 252 REFERENCES 254 Chapter 8 G E N E R A L S U M M A R Y A N D SUGGESTIONS FOR FUTURE WORK 256 8.1 G E N E R A L S U M M A R Y 256 8.2 SUGGESTIONS FOR FUTURE WORK 262 Chapter 9 EXPERIMENTAL 265 9.1 INTRODUCTION 265 9.2 SYNTHESES 265 9.2.1 IRON(II) A Z O L A T E POLYMERS 267 9.2.1.1 Polybis(pyrazolato)iron(II), [Fe(pz)2]x 267 9.2.1.2 Poly-2,2'-bipyridinete1rakis(imidazolato)diiron(Il), [Fe2(imid)4(bipy)] x 268 9.2.1.3 Polybis(4-azabenzimidazolato)iron(II), [Fe(4-abimid)2]x 268 ix 9.2.1.4 Poly-2,2':6', 2"-terpyridine octakis(imidazolato) tetTairon(n),[Fe4(imid)8(terpy)]x 269 9.2.1.5 Polybis( 1 -memyl-2-tJiioirnidazolato) iron(II)hemidicyclopentadienyliron (II), [Fe(l-Me-2-S-imid)2-0.5 Cp 2Fe] x 270 9.2.2 COBALT(II) IMIDAZOLATE POLYMERS 271 9.2.2.1 Polybis(irnidazolato)cobalt(II), [Co(imid)2]x 271 9.2.2.2 Polybis(2-methylimidazolato)cobalt(II), [Co(2-meimid)2]x 271 9.2.2.3 Polybis(4-methylimidazolato)cobalt(II), [Co(4-meimid)2]x 272 9.2.2.4 Polybis(benzimidazolato)cobalt(II), [Co(benzimid)2]x 272 9.2.2.5 Polybis(irmdazole)hexa(irmdazolato)tricobalt(II), [Co3(imid)6(imidH)2]x 273 9.2.2.6 Polybis(4-azabenzimidazolato)cobalt(II), [Co(4-abimid)2]x ....273 9.2.2.7 Poly-2,2 '-bipyridinetetjakis(imidazolato)dicobalt(n), [Co2(irnid)4(bipy)]x 274 9.2.3 NICKEL(II) IMIDAZOLATE POLYMER 274 9.2.3.1 Polybis(benzimidazolato)nickel(ir), [Ni(benzimid)2]x 274 9.2.4 COPPER IMIDAZOLATE POLYMERS 275 9.2.4.1 Polybis(imidazolato)copper(II), [Cu(imid)2]x 275 X 9.2.4.2 Polybis(2-methylimidazolato)copper(II), [Cu(2-meimid)2]x 275 9.2.4.3 Polybis(4-memylimidazolato)copper(LT), [Cu(4-meimid)2]x • 276 9.2.4.4 Polybis(benzimidazolato)copper(Il), [Cu(benzimid)2]x 276 9.2.4.5 Polybis(4,5-dichloroimidazolato)copper(Il), [Cu(4,5-dichloroimid)2]x 277 9.3 PHYSICAL METHODS 277 9.3.1 MAGNETIC SUSCEPTIBILITY MEASUREMENTS 277 9.3.2 SINGLE C R Y S T A L X - R A Y DIFFRACTION 279 9.3.3 POWDER X - R A Y DIFFRACTION 279 9.3.4 E L E M E N T A L ANALYSIS 280 9.3.5 MOSSBAUER SPECTROSCOPY 280 9.3.6 ELECTRONIC SPECTROSCOPY 281 9.3.7 T G A 281 9.3.8 INFRARED SPECTROSCOPY 281 9.3.9 NMR SPECTROSCOPY 282 REFERENCES. . . 283 APPENDIX I SINGLE C R Y S T A L X - R A Y DIFFRACTION D A T A 284 xi LIST OF TABLES Number Page Table 1.1 Mechanisms for achieving ferro- or antiferromagnetic spin coupling 15 Table 4.1 Magnetic parameters for some cobalt(II) weak ferromagnets 112 Table 4.2 UV-Vis-NIR spectra of copper(II) imidazolates. Approximate wavelength values or regions (nm) 131 Table 5.1 Magnetic parameters for three pairs of analogous iron(II) and cobalt(II) weak ferromagnets 206 Table 9.1 Commercial source of most chemical reagents employed in this thesis 266 Table I-1 Crystallographic data for [Fe(pz)2]x 284 Table 1-2 Selected bond lengths (A) and angles (°) for [Fe(pz)2]x with estimated standard deviations in parentheses 285 Table 1-3 Crystallographic data for [Fe(4-abimid)2]x 285 Table 1-4 Selected bond lengths (A) and angles (°) for [Fe(4-abimid)2]x with estimated standard deviations in parentheses 287 Table 1-5 Crystallographic data for a- and y-[Fe2(imid)4(bipy)]x 288 Table 1-6 Selected bond lengths (A) for a- and y-[Fe2(imid)4(bipy)]x, with estimated standard deviations in parentheses 289 xii Table 1-7 Selected bond angles (°) for a- and Y-[Fe2(imid)4(bipy)]x, with estimated standard deviations in parentheses 291 Table 1-8 Crystallographic data for P-[Fe2(imid)4(bipy)]x 294 Table 1-9 Selected bond angles (°) for (3-[Fe2(imid)4(bipy)]x, with estimated standard deviations in parentheses 295 Table I-10 Crystallographic data for [Fe4(imid)g(terpy)]x 300 Table 1-11 Selected bond lengths (A) and angles (°) for [Fe4(imid)8(terpy)]x 301 Table 1-12 Crystallographic data for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x 303 Table I-13 Selected bond lengths (A) and angles (°) for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x 304 xiii LIST OF FIGURES Number Page Figure 1.1 Schematic illustration of magnetic behaviors 4 Figure 1.2 The reciprocal susceptibility % x extrapolated from the high-temperature region as a function of temperature for independent g = 2,S=Vi spins as well as ferromagnetically coupled (0 = 10 K) and antiferromagnetically coupled (9 =-10 K) spins Figure 1.3 Schematic illustration of the magnetization, M, as a function of applied magnetic field, H, for several types of commonly observed magnetic behavior 10 Figure 1.4 A typical magnetic hysteresis loop 12 Figure 1.5 Illustration of the two types of superexchange. (a) kinetic exchange, and (b) potential exchange 16 Figure 1.6 The three regular planar nets, (a) 3-, (b) 4- and (c) 6-connected 21 Figure 1.7 Structures of pyrazole (a), and imidazole (b) 24 Figure 1.8 Schematic representation of different bridging modes for pyrazolate (1,2-diazolate) (top) and imidazolate (1,3-diazolate) (bottom) metal complexes. Spin orientations (arrows) are also illustrated as expected for the two different structural motifs 26 xiv Figure 2.1. Section of the polymer chain of [Fe(pz)2]x showing the atom numbering scheme. Hydrogen atoms are omitted. (50 % probability thermal ellipsoids shown) 47 Figure 2.2. View looking almost down the c axis in the structure of [Fe(pz)2]x. (50 % probability thermal ellipsoids shown) 48 Figure 2.3 x Heff versus temperature plot at 10 000 G for [Fe(pz)2]x. Lines are from theory as described in the text 50 Figure 3.1 View of the repeat unit of [Fe(4-abimid)2]x and atom numbering scheme (33% probability thermal ellipsoids). Hydrogen atoms are omitted 60 Figure 3.2 Stereoscopic view of a section of the diamond-like framework of [Fe(4-abimid)2]x. For clarity only the iron ions and the bridging N - C - N atoms of the imidazolate rings are shown 61 Figure 3.3 Iron ion connectivity diagram for a section of [Fe(4-abimid)2]x 62 Figure 3.4 View of [Fe(4-abimid)2]x looking down the b axis. Notice the voids being occupied by the 4-azabenzene part of the ligand 63 Figure 3.5 View of [Fe(4-abimid)2]x looking down the b axis. For clarity only the iron ions and the bridging N-C-N are shown 64 Figure 3.6 X-ray powder diffractograms of [Co(4-abimid)2]x (top, experimental) and [Fe(4-abimid)2]x (bottom, calculated) 67 Figure 3.7 Electronic spectra of [Co(4-abimid)2]x at two different mull concentrations 68 \ XV Figure 3.8 T G A plots for [Fe(4-abimid)2]x and [Co(4-abimid)2]x 69 Figure 3.9 % and %T versus T plots at 500 G for [Fe(4-abimid)2]x (top) and [Co(4-abimid)2]x (bottom) 71 Figure 3.10 Magnetization versus applied field plots at different temperatures for [Fe(4-abimid)2]x (top) and [Co(4-abimid)2]x (bottom) 73 Figure 3.11 Magnetic hysteresis plots at 4.8 K for [Fe(4-abimid)2]x (top) and at 10 K for [Co(4 abimid)2]x (bottom) 74 Figure 3.12 %sndyj versus T plots at 10 000 G for [Fe(4-abimid)2]x (top) and [Co(4-abimid)2]x (bottom) 76 Figure 3.13 A C susceptibility of [Fe(4-abimid)2]x; Hac = 1 G, f = 125 Hz 78 Figure 3.14 % versus T plots for[Co(4-abimid)2]x at 50, 100, 500 and 10 000 G 80 Figure 3.15 A C susceptibility for [Co(4-abimid)2]x; HAC = 1 G, f = 125 Hz (top) and HAC = 1 G, #DC = 20 G, f - 125 Hz (bottom) 82 Figure 3.16 Mossbauer spectrum of [Fe(4-abimid)2]x at 77.3 K 84 Figure 3.17 Selected Mossbauer spectra for [Fe(4-abimid)2]x at various temperatures...85 Figure 4.1 Asymmetric unit of [Co(imid)2]x. View looking down the c axis. Hydrogen atoms are omitted 94 Figure 4.2 T G A plots for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(imid)6(imidH)2]x 96 Figure 4.3 UV-Vis-NIR spectra for compounds [Co(imid)2]x, (a); [Co(2-meimid)2]x, (b); [Co(4-meimid)2]x, (c); [Co(benzimid)2]x, (d); and [Co3(imid)6(imidH)2]x, (e) 97 xvi Figure 4.4 X-ray powder diffractograms of [Co(imid)2]x (top, experimental; bottom, calculated) Figure 4.5 X-ray powder diffractograms of [Co(2-meimid)2]x (a) and [Co(benzimid)2]x (b) Figure 4.6 X-Ray powder diffractograms of [Co3(imid)6(imidH)2Jx (top, experimental) and Fe3(imid)6(imidH)2 (bottom, calculated). Figure 4.7 % versus T plots at 10 000 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(imid)6(imidH)2]x Figure 4.8 Lieff versus T plots at 10 000 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(imid)6(imidH)2]x Figure 4.9 Lieff versus T plots at 500 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(imid)6(imidH)2]x Figure 4.10 x versus T plots at 500 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(imid)6(imidH)2]x Figure 4.11 Magnetization versus applied field plots at different temperatures for compounds [Co(imid)2]x, (a); [Co(2-meimid)2]x, (b); [Co(4-meimid)2]x, (c); [Co(benzimid)2]x, (d); and [Co3(imid)6(imidH)2]x, (e) xvii Figure 4.12 Magnetic hysteresis plots at 4.8 K for compounds [Co(imid)2]x, (top); [Co(benzimid)2]x, (middle); [Co3(imid)6(imidH)2]x, (bottom) I l l Figure 4.13 Magnetic hysteresis plots at 4.8 K for compounds [Co(2-meimid)2]x, (top); and [Co(4-meimid)2]x (bottom) 115 Figure 4.14 UV-visible-near-IR spectrum for [Ni(benzimid)2]x. Insert plot shows the two highest energy d-d transition bands 118 Figure 4.15 T G A plot for [Ni(benzimid)2]x 119 Figure 4.16 Plots of % and Ueff versus T for [Ni(benzimid)2]x 121 Figure 4.17 Magnetization versus applied field plots at different temperatures for [Ni(benzimid)2]x 122 Figure 4.18 Magnetic hysteresis plot at 2 K for [Ni(benzimid)2]x. The insert plot shows a magnification of the central part of the hysteresis curve 123 Figure 4.19 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Ni(benzimid)2]x at 50 G 125 Figure 4.20 T G A plots for compounds [Cu(imid)2]x, [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x 129 Figure 4.21 UV-Vis-NIR spectra for [Cu(imid)2]x, (a); [Cu(2-meimid)2]x, (b); [Cu(4-meimid)2]x, (c); [Cu(benzimid)2]x, (d); and [Cu(4,5-dichloroimid)2]x, (e) 130 Figure 4.22 Repeat unit of blue-[Cu(imid)2]x. Hydrogen atoms are omitted 133 Figure 4.23 Stereoview of a section of blue-[Cu(imid)2]x including the unit cell. Projection (001). No hydrogen atoms shown 134 xviii Figure 4.24 X-ray powder diffractograms of blue-[Cu(imid)2]x (top, calculated) and [Cu(imid)2]x prepared here (bottom, experimental) 135 Figure 4.25 X-ray powder diffraction patterns of [Cu(2-meimid)2]x, (a); [Cu(4-meimid)2]x,(b); [Cu(benzimid)2]x, (c); and [Cu(4,5-dichloroimid)2]x, (d) 136 Figure 4.26 % versus T plots at 10 000 G for [Cu(imid)2]x, [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x 138 Figure 4.27 LL F^ versus T plots at 10 000 G for [Cu(imid)2]x, [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x 139 Figure 4.28 Plot of x versus T for [Cu(4-meimid)2]x at 500 G 140 Figure 4.29 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(benzimid)2]x at 50 G 142 Figure 4.30 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(2-meimid)2]x at 50 G 143 Figure 4.31 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(4,5-dichloroimid)2]x at 50 G 144 Figure 4.32 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(imid)2]x at 50 G 145 Figure 4.33 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(4-meimid)2]x at 50 G 146 xix Figure 4.34 Magnetic hysteresis plots at 4.8 K for [Cu(4,5 dicloroimid)2]x, (top); [Cu(benzimid)2]x, (middle); and [Cu(2-meimid)2]x, (bottom). The insert plots show magnifications of the central part of the hysteresis curves 148 Figure 4.35 Magnetic hysteresis plots at 4.8 K for [Cu(imid)2]x, (top), and [Cu(4-meimid)2]x, (bottom). The insert plots show magnifications of the central part of the hysteresis curves 149 Figure 5.1 T G A plot for [Fe2(imid)4(bipy)]x 158 Figure 5.2 View of the repeat unit of [Fe2(imid)4(bipy)]x (a-phase, 294 K) and atom numbering scheme (33% probability thermal ellipsoids) 159 Figure 5.3 ORTEP diagrams of [Fe2(imid)4(bipy)]x (a-phase) looking down the c axis. In the bottom view, bipyridine ligands have been removed to reveal the double-layer sheet extended framework. (50 % probability thermal ellipsoids) 161 Figure 5.4 Iron ion connectivity diagram of a section of two double-layer sheets for the a-phase of [Fe2(imid)4(bipy)]x. Octahedral iron (red), tetrahedral iron (green). View looking approximately down the c axis 162 Figure 5.5 View of the asymmetric unit of [Fe2(imid)4(bipy)]x (y-phase, 113 K) and atom numbering scheme (33% probability thermal ellipsoids) 164 XX Figure 5.6 Iron ion connectivity diagram of a section of two double-layer sheets for the y-phase of [Fe2(imid)4(bipy)]x. Octahedral iron v (red or semi-filled), tetrahedral iron (green or non-filled). View looking approximately down the c axis 166 Figure 5.7 View of the asymmetric unit of [Fe2(imid)4(bipy)]x (P-phase, 143 K) and atom numbering scheme 167 Figure 5.8 Comparison of coordination sphere geometries by overlapping octahedral irons (red circle) in the a - (black bonds) and y- (green bonds) phases of [Fe2(imid)4bipy]x 168 Figure 5.9 % and Lieff versus T plots at 10 000 G for [Fe2(imid)2(bipy)]x 170 Figure 5.10 % ^  Meff versus T plots at 500 G for Y-[Fe2(imid)2(bipy)]x 171 Figure 5.11 Magnetization versus applied field plots at different temperatures for y-[Fe2(imid)2(bipy)]x 172 Figure 5.12 Magnetic hysteresis plot at 4.8 K for y- [Fe2(imid)4(bipy)]x 173 Figure 5.13 Plot of %_1 versus temperature at 10 000 G for [Fe2(imid)4(bipy)]x 174 Figure 5.14 Plots of Z F C M , F C M and R E M for y-[Fe2(imid)4(bipy)]x at a DC field o f50G 177 Figure 5.15 Temperature dependences of the in-phase, and out-of-phase, X " , A C magnetic susceptibilities for y-[Fe2(imid)4(bipy)]x at f=125 Hzandi /= 1 G 178 Figure 5.16 %T versus T for [Fe2(imid)4(bipy)]x. Hoc = 10 000 G. Cooling mode 179 xxi Figure 5.17 Temperature dependence of the derivatives d(%T)/dT (DC) in the cooling mode and determination of the transition temperatures for [Fe2(imid)4(bipy)]x. Hue = 10 000 G 180 Figure 5.18 Cooling and warming modes yj versus temperature plots for [Fe2(imid)4(bipy)]x. / / D C = 1 000 G. Insert plot shows an augmentation of the a<-»(3 transition region 181 Figure 5.19 Temperature dependence of the derivatives d(%T)/dT (DC) in the cooling and warming modes and determination of the transition temperatures for [Fe2(imid)4(bipy)]x. Hoc = 1 000 G 183 Figure 5.20 A C % versus T plot for [Fe2(imid)4(bipy)]x. Cooling mode. f=500Hz, i / A c = 2.5G 184 Figure 5.21 A C % versus T plot for [Fe2(imid)4(bipy)]x. Warming mode. f=500 H z , / / A c = 2.5 G 185 Figure 5.22 Temperature dependence of A C %T in the cooling and warming modes for [Fe2(imid)4(bipy)]x. f = 500 Hz, HAc - 2.5 G. Arrow down refers to cooling mode, arrow up refers to warming mode 186 Figure 5.23 Temperature dependence of the derivatives d(%T)/dT (AC) in the cooling and warming modes and determination of the transition temperatures for the Y-phase of [Fe2(imid)4(bipy)]x. f = 500 Hz, HAC = 2.5 G 187 Figure 5.24 Cooling and warming mode A C % versus temperature plots for [Fe2(imid)4(bipy)]x. f = 500 Hz, HAC = 2.5 G 188 xxii Figure 5.25 Mossbauer spectrum of [Fe2(imid)4(bipy)]x at 293 K 189 Figure 5.26 Mossbauer spectra in the warming mode for [Fe2(imid)4(bipy)] 191 Figure 5.27 Mossbauer spectra in the cooling mode for [Fe2(imid)4(bipy)]x 193 Figure 5.28 X-Ray powder diffractograms of [Co2(imid)4(bipy)]x (top, experimental) and [Fe2(imid)4(bipy)]x (bottom, calculated) 195 Figure 5.29 T G A plot for [Co2(imid)4(bipy)]x 196 Figure 5.30 % and ^  versus T plots at 10 000 G for [Co2(imid)4(bipy)]x 198 Figure 5.31 % and p^ff versus T plots at 500 G for [Co2(imid)4(bipy)]x 199 Figure 5.32 Magnetization versus applied field plots at different temperatures for [Co2(imid)4(bipy)]x 200 Figure 5.33 Magnetic hysteresis plots at 4.8 K for [Co2(imid)4(bipy)]x 201 Figure 5.34 Plot of y£x versus temperature at 10 000 G for [Co2(imid)4(bipy)]x 202 Figure 5.35 Plots of Z F C M , F C M and R E M for [Co2(imid)4(bipy)]x using a DC field of 50 G 204 Figure 6.1 T G A plot of [Fe4(imid)8(terpy)]x 212 Figure 6.2 Repeat unit of [Fe4(imid)8(terpy)]x showing the atom numbering scheme; 33 % probability thermal ellipsoids are shown 214 Figure 6.3 View of a section of [Fe4(imid)8(terpy)]x looking down the a axis. Terpy ligands and C-4 and C-5 of imidazolate ligands have been omitted in the bottom view for clarity. Hydrogen atoms are not shown 215 xxiii Figure 6.4 View of a section of [Fe4(imid)8(terpy)]x looking down the b axis. For clarity, terpy ligands and C-4 and C-5 of imidazolate ligands have been omitted in the bottom view. Hydrogen atoms are not shown 216 Figure 6.5 Iron ion connectivity diagram for a section of [Fe4(imid)g(terpy)]x. Four-coordinate ions (green and pink/black ellipsoids), six-coordinate ions (blue ellipsoids) and five-coordinate ions (red ellipsoids) 217 Figure 6.6 Mossbauer spectrum of [Fe4(imid)g(terpy)]x at 293 K 218 Figure 6.7 DC % and L^ff versus T at 10 000 G for [Fe4(imid)8(terpy)]x 220 Figure 6.8 DC % and LL-ff versus T at 500 G for [Fe4(imid)8(terpy)]x 221 Figure 6.9 A C magnetic susceptibility for [Fe4(imid)g(terpy)]x, H\c = 1G, f = 125 Hz 222 Figure 6.10 Plots of Z F C M , F C M and R E M for [Fe4(imid)8(terpy)]x. Hoc = 50 G 223 Figure 6.11 Plot of magnetization versus applied field at different temperatures for [Fe4(imid)g(terpy)]x 224 Figure 6.12 Field dependence of magnetization at 4.8 K for [Fe4(imid)8(terpy)]x. Central portion of hysteresis loop shown. The data obtained on decreasing the applied field are shown as • while the data obtained on increasing the applied filed are shown as A 225 Figure 6.13 Field dependence of magnetization at 4.8 K for [Fe4(imid)g(terpy)]x. Central portion of hysteresis loop shown .227 Figure 7.1 T G A plot for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x 234 xxiv Figure 7.2 Molecular structure of the polymer chain of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x showing the atom numbering scheme; 33 % probability thermal ellipsoids are shown. (Hydrogen atoms are omitted) 236 Figure 7.3. View of the crystal structure of [Fe(l -Me-2-S-imid)2-0.5Cp2Fe]x down the c axis. 50 % thermal ellipsoids are shown 237 Figure 7.4 DC % and %T versus temperature plots at 500 G for [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x The line is from theory as described in the text. ; 239 Figure 7.5 A C magnetic susceptibility for [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x, HAC=l G , f = 125 Hz 240 Figure 7.6 Plot of magnetization versus applied field at three temperatures for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x 241 Figure 7.7 Field dependence of magnetization at 4.8 K for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x. Central portion of hysteresis loop shown ...242 Figure 7.8 Plot of %T versus temperature at three values of applied field for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x 243 Figure 7.9 Mossbauer spectrum of [Fe( 1 -Me-2-S-imid) 20.5Cp 2Fe] x at 293 K 248 Figure 7.10 Mossbauer spectrum of [Fe(l -Me-2-S-imid) 20.5Cp 2Fe] x at 77 K 249 Figure 7.11 Mossbauer spectrum of [Fe( 1 -Me-2-S-imid) 20.5Cp 2Fe] x at 4.2 K .251 XXV LIST OF ABBREVIATIONS A N D SYMBOLS ~ approximately % percent X magnetic susceptibility %' in-phase magnetic susceptibility X" out-of-phase magnetic susceptibility ° or deg degree(s) r| hapticity 8 isomer shift ° C degree(s) Celsius U.B Bohr magneton Lieff effective magnetic moment 0 X-ray diffraction angle or Curie-Weiss law correction term AE quadrupole splitting H applied magnetic field H0 applied magnetic field static component H\ applied magnetic field oscillating component 1- D one-dimensional 2- D two-dimensional xx vi 3-D three-dimensional F goodness-of-fit parameter f frequency cm"1 wavenumber(s) S length of the coil V strength of the induced voltage permeability of a material CO period of oscillation <P phase angle Cp2Fe or FeCp2 ferrocene o A Angstrom(s) bipy 2,2'-bipyridine C Curie constant cm centimeter(s) cos cosine coth hyperbolic cotangent F C M field-cooled magnetization fw formula weight g Lande splitting factor G Gauss g gram(s) Hz hertz xxvii imidH imidazole imid imidazolate IR infrared J magnetic exchange coupling constant k Boltzmann constant K Kelvin Me or me methyl ml mililiter(s) mmol milimole(s) mol mole(s) M magnetization Ms saturation magnetization N Avogadro's number NIR near infrared nm nanometer(s) NMR nuclear magnetic resonance ORTEP Oakridge thermal ellipsoid plot P relative proportion of paramagnetism impurity pz pyrazolate Ref. reference R E M or A/rem remnant magnetization S total spin xx viii sin sine SQUID superconducting quantum interference device t time T or T temperature tan tangent Tc Curie temperature T c critical temperature terpy 2,2':6',2"-terpyiridine T N Neel temperature T G A thermal gravimetric analysis U V ultraviolet Vis or VIS visible V S M vibrating sample magnetometer Z F C M zero-field-cooled magnetization xx ix A C K N O W L E D G M E N T S I would like to express my sincere gratitude to my research supervisors, Drs. Robert C. Thompson and Alan Storr, for their remarkable support, guidance and great professional commitment during the development of this thesis. I am grateful to Dr. William M . Reiff at Northeastern University for his valuable contribution to this work and for fruitful discussions on Mossbauer spectroscopy. I would like also to thank the members of my guidance committee, Drs. C. Orvig, M . Wolf and D. Douglas for their knowledgeable suggestions during the final preparation of this dissertation. My gratitude is also extended to Drs. S. J. Rettig and B. O. Patrick of this Department for their expert crystal structure determinations, to Mr. P. Borda of this Department for microanalytical services, and to Mr. Pinder Dosanjh of the Advanced Materials and Process Engineering Laboratory at U.B.C. for assistance on the SQUID magnetometer. Many thanks are also extended to the experts in the glassblowing, electronics and mechanical shops for their aid. I also thank my colleagues Dr. D. A. Summers, Dr. Z. Xu, Ms. I. Sham and Ms. C. L. Stevens for their assistance and for making my time in this Department more enjoyable. I want to give special thanks to my parents, and dedicate this work in commemoration of their fiftieth wedding anniversary. Finally, I wish to offer my deepest gratitude to my wife Emma Veronica, for her love and endless encouragement throughout this endeavour. xxx Chapter 1 INTRODUCTION The search for and study of new molecule-based magnets, molecular materials that show spontaneous magnetization below some critical temperature, has become an area of considerable interest and activity in recent years [1]. Coordination polymers of the transition metals exhibiting such properties constitute a sub-classification of molecule-based magnets. The structural diversity of these materials and the fact that many contain ligands that are amenable to systematic derivatization, raises the possibility of tailoring their properties to specific desired applications [2]. Molecular magnetism is a multidisciplinary area of investigation that requires the combined efforts of chemists, physicists and materials scientists to establish the necessary fundamentals. In magnetochemistry, a chemist must be able to explain the magneto-structural correlations in a material. This work requires the use of a variety of techniques for both structural and magnetic characterization. Also important is the ability to synthesize multidimensional extended molecular systems, such as coordination polymers, with potentially interesting magnetic properties. One of the aims in the current work was to synthesize coordination polymers of transition metals, with different dimensionality, capable of exhibiting long-range magnetic ordering. The ligands chosen to connect the transition metals in this study l were diazolates, mainly 1,3-diazolates. It was anticipated that, due to the positioning of the nitrogen coordinating atoms in 1,3-diazolates, the formation of 2-D or 3-D structures (involving single azolate bridges) could be achieved. Fundamental concepts in (i) magnetism and molecule-based magnets, (ii) dimensionality and connectivity in extended systems (coordination polymers) and (iii) the azolates as bridging ligands of transition metals are outlined in this Chapter. The techniques employed in this work in the physical characterization of materials are also briefly described. The general objectives of the research and a description of the organization of this dissertation are also included at the end of this Chapter. 1.1 MAGNETISM 1.1.1 INTRODUCTION Magnetism can be measured using two different techniques, by the response (attraction or repulsion) of a material to a magnetic field, which is the basis for the outdated force methods [3], or by state-of-the-art induction methods that determine directly the change in magnetic flux density resulting from placement of the material in a magnetic field [4]. Magnetism is a collective effect based on the coupling of the spin or internal angular momentum of unpaired electrons throughout an entire material [5]. To simplify the following discussion, the orbital angular momentum contribution to the bulk magnetization of materials will be ignored. In other words, a spin-only model will 2 be employed. Paramagnetism is typically a consequence of the spin associated with an unpaired electron (ms = lA T or m s = -Vi I) [5]. If a molecule or ion has only paired electrons, it is diamagnetic and will be repelled slightly in an applied magnetic field. If a molecule has an odd number of electrons, there will be at least one unpaired electron, and the molecule would have a net spin. Such molecules or ions are usually sufficiently isolated so that their spin-spin coupling energy, / (as deduced from the Hamiltonian H= -2 /SA - SR) is small compared to (coupling-breaking) thermal energy. Their spins do not couple in the absence of an applied field; these are called simple paramagnets (Figure 1.1). When two paramagnetic metal ions interact directly such that their magnetic orbitals (those that contain the unpaired electrons) overlap, direct spin coupling operates [6]. In contrast, when paramagnetic metal ions are bridged by ligands, as in many molecule-based magnets (vide infra), superexchange coupling [7] takes place as the main mechanism responsible for spin interactions. Hence, spin-spin interactions may be large enough to enable an effective parallel -called ferromagnetic (TT)- or antiparallel -called antiferromagnetic (-IT)- coupling (Figure 1.1). It is important to describe at this point a distinction between short-range and long range magnetic ordering. Short-range order may be described as the tendency of the paramagnetic spins to orient themselves locally relative to one another when a 3 A A A A A A A A A A A A Ferromagnetic A A A A A V A Antiferromagnetic A X A X A A X A Y Y A Ferrimagnetic Canted Antiferromagnetic Figure 1.1 Schematic illustration of magnetic behaviors. material is cooled [8]. Magnetic interactions in clusters are an example of short-range order [9]. Long-range magnetic ordering, on the other hand, is the result of an extended and cooperative ordering of the spins throughout the lattice. Long-range order accompanies a change in the spin phase of a material [10]. Therefore, bulk ferromagnetic behaviour occurs when the spins in a material undergo long-range alignment in the same direction, resulting in a net magnetic moment. Hence, ferromagnetism requires that the individual unpaired spins interact collectively with each other aligning themselves parallel and in the same direction. Ferrimagnetism occurs when, due to the presence of magnetic dipoles of different size, antiferromagnetic coupling does not lead to complete cancellation of moments and a net moment remains (Figure 1.1). It is important to note that potentially commercially useful ferro- or ferrimagnetic behavior is not a property of a single molecule or ion; it, like superconductivity, is a cooperative solid-state bulk property [11]. As mentioned previously, paramagnets are characterized by their response to an applied magnetic field, H. For ideal, non-interacting spins a net magnetic moment or magnetization, M, is induced in the material when exposed to an applied field, H; where M i s proportional to H, M=%H The proportionality constant is termed the molar magnetic susceptibility, %. A material is magnetically isotropic when both the magnitude and the direction of M do not depend on orientation. In this case the direction of M is coincident with that of H, regardless of 5 the specimen orientation. If, on the contrary, a material is magnetically anisotropic, the direction and magnitude of M depend on orientation [12]. The magnetic susceptibility of a simple paramagnet has a temperature dependence that is characterized by the Curie expression [13], _ C The Curie constant C in cm 3 K mol"1, is defined according to the following equation, c^Ng2n2BS(S + l) 3k in which S is the spin quantum number, N is Avogadro's number, g is the Lande factor, LIB is the Bohr magneton, and k is the Boltzmann constant. If the spins experience an effective parallel (or antiparallel) exchange field due to cooperative interactions with neighboring spins this will increase (or decrease) the measured susceptibility from that predicted for independent spins by the Curie law. In these instances, the high-temperature susceptibility data often can be fit to the Curie- Weiss law [14], C X = T-Q 6 where for parallel (ferromagnetic) or antiparallel (antiferromagnetic) interactions, 9 is greater or less than zero, respectively. The value of 0 can be determined from the intercept obtained in the linear extrapolation of the plot of %' versus T at high temperature. The temperature dependencies of %~l are illustrated for independent spins (Curie law) and spins with ferro-, and antiferromagnetic interactions in Figure 1.2. T ( K ) Figure 1.2 The reciprocal susceptibility %A extrapolated from the high-temperature region as a function of temperature for independent g = 2, S = lA spins as well as ferromagnetically coupled (9 = 10 K) and antiferromagnetically coupled (9 = -10 K) spins. 7 Very frequently in magnetic studies of materials the effective moment, 10^ , or simply %T, is reported [15]. The effective moment, in units of the Bohr magneton (u.B), is defined by the following equation, For a simple system comprising one mole of non-interacting spins (i.e., 0 = 0), and where S is a valid spin quantum number, |ieff is temperature-independent [15], Heff=Jg2S(S + l) At low temperatures and high magnetic fields, conditions under which the magnetic energy (gS[ief0) is comparable in magnitude to the thermal energy (kT), the magnetization no longer obeys the equation M = %H, but approaches the limiting value or saturation magnetization, Ms, Ms=[iBNSg For a system with non-interacting spins (i.e., 0 = 0) the temperature dependence of M can be calculated from the Brillouin function [16], M = \LBNSgB 8 where B = 25 + 1 25 coth '25 + 1 25 1 ( x \ coth — 25 \2S j and gSfiBB kT Because the response to an applied magnetic field varies depending on the type of magnetic coupling within the material, a plot of magnetization as a function of applied field produces a curve with a shape characteristic of the type of magnetic coupling occurring in the material [5] (Figure 1.3). In Figure 1.3 the line labelled paramagnetic represents a simple paramagnetic system where there is no spin coupling. In this case the initial slope of the observed M versus H data is as expected for the equation M = [isNSgB. Antiferromagnetic coupling is evident if the initial slope is less than this and ferromagnetic coupling is evident if the initial slope exceeds the value expected by this equation. The M versus H plot of a diamagnetic sample will contrast with that of a paramagnetic one by having a negative slope (Figure 1.3). Metamagnetism is a transformation from an antiferromagnetic state to a high-moment ferromagnetic state; that is, the spin alignment changes from antiparallel to parallel by the occurrence of an applied magnetic field [17]. This generates an M versus H curve of the type shown in Figure 1.3. 9 M F e r r o m a g n e t / / J j / P a r a m a g n e t i c j /F e r r o m a g n e t i c / M e t a m a g n e t i c —-** " ~* A n t i f e r r o m a g n e t i c D i a m a g n e t i c ~ ' . _ J H —> Figure 1.3 Schematic illustration of the magnetization, M, as a function of applied magnetic field, H, for several types of commonly observed magnetic behavior. When long-range magnetic ordering is present in a material, the temperature at which the spins order is termed the critical temperature, Tc. If the spins align ferromagnetically, a spontaneous magnetization in zero applied field is present and below T c the material is a ferromagnet and the M versus H plot (Figure 1.3) does not extrapolate to zero magnetization at zero applied field. The critical temperature in this case is sometimes referred to as the Curie temperature, Tc- If a long-range antiparallel 10 alignment of spins occurs, there is no net moment below T c and the susceptibility is zero. In this situation the critical temperature is sometimes referred to as the Neel temperature, T^. When ferrimagnetism is present in the system, the critical temperature is also referred to as the Neel temperature, 7N, and below the material is a ferrimagnet and under these conditions the M versus H plot (Figure 1.3), like that of a ferromagnet, does not extrapolate to zero magnetization at zero applied field. Below T c the magnetic moments for ferro- and ferrimagnets align in small regions (domains). The direction of the magnetic moment of adjacent domains differs, but can be aligned by application of a minimal magnetic field. This leads to history dependent magnetic behavior (hysteresis) characteristic of ferri- and ferromagnets. Thus, applying an external magnetic field will cause the domains to coalesce and form a single domain aligned with the external field. At low temperatures and high applied magnetic field, the magnet can rapidly reach a maximum magnetization, which is limited by its saturation magnetization, Ms. When the applied field is decreased, the spin alignment of the domains relaxes, but more slowly than the original alignment occurred, so that when the external field reaches zero, some remnant magnetization, Mrem, remains. Reversing the direction of the external magnetic field will cause the spins within the magnet to reverse. At a large enough applied magnetic field, the magnetization reaches saturation again, but in the opposite direction. Increasing, then, the applied magnetic field to positive values results in an approximately syrnmetric closed loop termed a hysteresis loop. A typical hysteresis loop is shown in Figure 1.4. 11 The coercive field, Hc, is the reverse magnetic field required to reduce the magnetization of a sample to zero starting from a saturation condition magnetization. "Hard" magnets have values of Hc> 100 G, whereas "soft" magnets have values of < 10 G. Medium to large values (hundreds of G) of Hc are necessary for permanent magnetic storage of data, while low values (mG) are required for ac motors, magnetic shielding, and in the recording heads necessary to write, read and erase the recording information [18]. Hence, the T c , Ms, and Hc are key parameters in ascertaining the commercial utility of a magnet. -400 -200 0 200 400 Figure 1.4 A typical magnetic hysteresis loop. 12 Canted antiferromagnetism, canted ferromagnetism, metamagnetism, and spin-glass behaviors are examples of other possible magnetic phenomena. The canted antiferromagnetic behavior (Figure 1.1, page 4) results from the action of antisymmetric exchange in anisotropic materials in which the coupled magnetic dipoles are not related by an inversion center [19]. The weak ferromagnetism, produced in this type of spin coupling, is due to the fact the antiparallel alignment of spins on the two sublattices have orientations slightly canted to each other. This canting leads to a non-zero magnetization at zero-applied field [20]. A material exhibiting canted antiferromagnetic behavior is referred to as a weak ferromagnet [20]. It should be noted that canted antiferromagnetism is considered to be the primary mechanism accounting for the residual magnetization observed in the molecule-based magnets studied in this dissertation. A canted ferromagnet, on the other hand, results from the relative tilting of ferromagnetically coupled spins (Figure 1.1, page 4) such that, though the material is ferromagnetic, there is a reduction of the residual moment. Finally, a spin glass occurs when there are local spatial correlations in the directions of neighboring spins, but no long-range order. For a spin glass the spin alignment is as described for a paramagnet (Figure 1.1, page 4). However, unlike a paramagnet where the directions of the spins vary with time, the spins remain fixed in their orientations for a spin glass. Hence, a spin glass has spins pointing in similar directions for short distances, but no long-range order [17]. 13 1.1.2 MAGNETIC E X C H A N G E The type of spin coupling, whether ferromagnetic or antiferromagnetic, and the magnitude, can be described by the coupling interaction energy, J. When J is positive, the coupling is ferromagnetic; when it is negative, the coupling is antiferromagnetic. The value of J cannot be measured directly, but can be deduced from a mathematical model fitted to the magnetic data. Different models can give different values of J even in systems with a well defined T c . Three distinct mechanisms for spin coupling have been proposed to lead to ferro- or antiferromagnetism. (Table 1.1). It is difficult to predict in advance which mechanism may dominate in a particular system, and more than one mechanism may be operational. The most straightforward type of spin coupling occurs between unpaired electrons in orthogonal orbitals in the same spatial region. These spins couple o ferromagnetically, and the closer the orbitals are in space (less than 3 A), the stronger the coupling [21]. This mechanism can lead only to ferromagnetic coupling. Ferromagnetic coupling can also occur between unpaired electrons that are nominally not in the same spatial region. The most important and powerful exchange mechanism of this type is the termed superexchange. In superexchange [7], non-magnetic moieties (atoms or molecules) can function as mediators for spin coupling between spin carriers. Anderson's theory [22] established that superexchange occurs 14 Table 1.1 Mechanisms for achieving ferro- or antiferromagnetic spin coupling. Mechanism Spin interaction Spin coupling 1. Spins in Orthogonal Orbitals Intramolecular Ferromagnetic 2. Spin coupling via Superexchange Intramolecular Ferro- or antiferromagnetic 3. Dipole-dipole (through-space interactions) Intra- or intermolecular Ferro- or antiferromagnetic because the metal d-orbitals, where the unpaired spins originate, overlap with filled s or p orbitals of the mediator atom or atoms. Consequently, delocalized magnetic antibonding orbitals, which include the metal ion and the intermediary atom or atoms, are formed. The spins in two such delocalized magnetic orbitals can interact in two ways: kinetic and potential exchange. Kinetic exchange arises when there is a non-orthogonal orbital interaction pathway between the bridging ligand and two magnetic centres. This process, which is illustrated in Figure 1.5, yields antiferrromagnetic coupling. On the contrary, potential exchange (Figure 1.5) happens when there is orthogonality in the orbital interaction pathway, and ferromagnetic coupling results. 15 (a) (b) Figure 1.5 Illustration o f the two types o f superexchange. (a) kinetic exchange, and (b) potential exchange. The majority o f the compounds studied in this work are paramagnetic transition metal ions separated by imidazolate or pyrazolate ligands, which yields distances between metal ions larger than 3 A . Hence, superexchange is the main mechanism involved in the magnetic coupling exhibited by the compounds described in this work. A third type o f spin coupling can occur through space between spins in orbitals that do not overlap. These interactions, which occur via the magnetic fields generated with each spin, are very weak and only produce magnetic ordering at temperatures below a few degrees Ke lv in [21]. 16 1.1.3 M O L E C U L E - B A S E D MAGNETS Traditional magnetic materials are atom-based, which means that they have d- or f- orbital spin sites, and posses extended network magnetic "bonding" in at least two dimensions. Furthermore, they are prepared by high temperature methodologies. In contrast, molecule-based magnets [23] are materials prepared utilizing the low-temperature synthetic procedures of organic, organometallic, or coordination metal chemistry. As a result, the organic fragment may (i) be an active component with spin sites (free radicals) contributing to both the high magnetic moment and the spin coupling, or (ii) can mediate the spin coupling interaction by derealization of the spins on the metal ions throughout the ligand molecules (superexchange) [24]. The first molecular ferromagnetic compound was reported by Wickman et al. in 1967 [25]. This five-coordinate, square-pyramidal complex, cMorobis(diemyldithiocarbamato)iron(III), orders ferromagnetically at 2.46 K and is rather unusual because it represents a true molecular solid. By definition, a true molecular solid consists of neutral species bonded intermolecularly only by van der Waals interactions and/or hydrogen bonds [26]. Other examples of true molecular ferromagnets that have been described include the purely organic ferromagnets such as P-(p-nitrophenyl)nitronylnitroxide, which has a Curie temperature of 0.6 K [27], and Rassat's dinitroxide [28], which has the highest reported ordering temperature for a purely organic compound with T c = 1.48 K. In contrast, most of the recently reported 17 ferromagnetic molecular solids are more accurately described as "molecule-based". Examples include the metallocene charge-transfer salts which consist of organic and organometallic anions and cations bound by Coulomb interactions [29] and the Prussian Blue analog magnets which are three-dimensional coordination polymers involving bridging of metal ions by cyanide anions [30]. The field of molecule-based magnets is a relatively new branch of chemistry [31]. The first compounds of this kind were reported in 1986 [32], and in the last few years an increasing number of research groups have started some activity along this line. The typical synthetic approach to design molecule-based magnets consists of starting from precursors bearing a spin, then assembling them in such a way that there is no compensation of the spins at the scale of the crystal lattice. Several chemical features can stabilize long-range ferromagnetic coupling. These include having as many spins as possible in orbitals oriented so that the spins can couple strongly to form a magnet. As described, the interactions between spin carriers may occur through space or through bonds. In the former case, a genuine molecular lattice with molecules or molecular ions at the lattice points is involved. In the latter case, a polymeric or extended structure is involved and in this case the magnets are termed "molecule-based". Most often the interactions are much stronger when they occur through bonds (superexchange). This is particularly true when the bridging linkages (ligands) are conjugated [7]. Therefore, an efficient strategy to obtain a molecule-based magnet is to assemble the spin bearing precursors using conjugated 18 bridging ligands with the structural capability of forming extended structures. This fundamental strategy was followed in the preparation of the molecule-based magnets investigated in the present thesis Many other important aspects of magnetism may have been left out of the brief account presented here. The interested reader is encouraged to explore further some of the many interesting books and reviews published about this fascinating subject [1,2, 33-42]. 1.2 DIMENSIONALITY A N D CONNECTIVITY Dimensionality plays a critical role in determining the properties of materials due to, for example, the different ways that electrons interact in three-dimensional (3-D), two-dimensional (2-D), and one-dimensional (1-D) structures [43]. The study of dimensionality has a long history in chemistry and physics, although this has been primarily with the prefix "quasi" added to the description of materials. That is, quasi-ID solids, such as square-planar platinum chain compounds [44], and quasi-2D layered solids, such as copper oxide superconductors [45]. The control of dimensionality is a major challenge within the metal coordination polymer field [46]. Even when polyfunctional ligands are used to obtain high dimensional polymers, ancillary ligation by water or other solvent ligands may result in 19 low dimensionality [47]. Hence, the relatively new field of metallo-organic polymers [48], although offering great potential for chemical and structural diversity, suffers from general difficulties in the control of polymer dimensionality or framework stability. Sometimes low-dimensional coordination polymers can lack framework integrity [49], in other cases the resulting coordination polymers are frequently plagued by lattice interpenetration [50] or a framework breakdown upon removal of absorbates [51]. Synthesis at relatively higher temperatures can promote the formation of polymer frameworks of higher dimensionality through the loss of terminal ancillary ligands. Thus, for example, Wood et al. [52] have found a condensed 3D structure for [Mn(TMA)] [TMA = trimesate (benzene-1,3,5-tricarboxylate)], whereas a discrete molecular complex [Mn(TMA-H2)2(H20)4] is formed at room temperature [53]. Hence, the use of synthetic methods, such as the reaction of metallocenes with molten-ligands [54], and those involving solvothermal chemistry [55], have been used in attempts to form higher dimensionality frameworks. Connectivity is an influential concept that describes the way a set of points connects to build a lattice that is infinite in one to three dimensions, like a crystal [56]. In two dimensions, there are only three regularly connected nets [57]. In this case "regular" means a network that not only has the same number of neighbours at each site, but where there is only one type of polyhedron in the net. For 3-connected planar nets, there is a hexagonal arrangement, while the 4-connected one corresponds to a net 20 made up of quadrilaterals (not necessarily squares) and regular 6-connected nets are an array of triangles (not necessarily equilateral). The most symmetrical forms of these three regular networks are shown in Figure 1.6. (a) (b) (c) Figure 1.6 The three regular planar nets, (a) 3-, (b) 4- and (c) 6-connected. It has become more evident that connectivity (i.e. the number and arrangement of interaction pathways between neighbouring centres with localized spins) is critical in generating long-range ordered systems. Since the interaction mechanism between spins 21 connected through a chemical bonding pathway is superexchange, it is not only the connectivity between atoms that should be considered, but also connectivity between orbitals. Throughout this thesis, the concept of connectivity has been utilized for a better understanding of the relationship between extended structures and magnetic properties in several compounds studied. Ferromagnetic interactions do not necessarily lead to long-range ferromagnetic ordering. The central point is that such long-range magnetic ordering is rigorously impossible for a system consisting of isolated molecules (zero-dimensional), or of isolated chains (one-dimensional). It may occur for a system consisting of isolated layers (two-dimensional), provided that the spins are not strictly isotropic. On the contrary, long-range magnetic ordering is the normal behaviour of a three-dimensional spin network. Therefore, the design of a molecule-based magnet requires one to create spin interactions along the three directions of space. Furthermore, these interactions must be either ferromagnetic, antiferromagnetic between non-equivalent spin lattices (ferrimagnetism), or as determined in this work, canted antiferromagnetic leading to weak ferromagnetism. 1.3 COORDINATION POLYMERS A coordination polymer may be defined as a material consisting of metal ions linked by coordinate bonds by mono-atomic or poly-atomic species forming an 22 extended structure. About 30 years ago, the main interest in coordination polymers depended mainly on the expectation of increased thermal stability for the materials. Over the last 10 years or so, the properties of ordered infinite aggregates of metal ions, connected by bridging ligands, have come to the fore as a subject for synthetic study. As in classical coordination chemistry the local electronic structure at the metal ion remains important but the connectivity of the lattice and the nature of the bridging groups tuning the inter-metallic interactions are essential in determining the properties of the bulk material. The increasing interest in the area of coordination polymers has been motivated by the ability of the metal-ligand coordination to provide a facile approach to the controlled assembly of one-, two- and three-dimensional extended solids. This strategy presents an excellent opportunity for the construction of functional materials with interesting properties such as, second-order non-linear optical [58], electronic [59], magnetic [60], inclusion [61], and catalytic properties [62]. A general problem in the characterization of coordination polymers arises because of the extreme intractability of many of these of materials, and the consequent lack of available structural information. In addition, owing to the difficulty of obtaining single crystals of these materials, relatively few X-ray structure determinations have been carried out on coordination polymers, necessitating the use of indirect methods of structural characterization. 23 1.4 DIAZOLES A N D DIAZOLATES Diazoles are five-membered, aromatic, two nitrogen-containing heterocyclic molecules. Among the best known of these compounds are the 1,2-diazole, pyrazole, and the 1,3-diazole, imidazole, which are shown in Figure 1.7. The role of diazoles and their anions (diazolates) as ligands in coordination chemistry is well documented by several reviews published on this topic [63 -65]. (a) (b) Figure 1.7 Structures of pyrazole (a), and imidazole (b). From the point of view of molecular architecture, the most important characteristic of the diazolates is the way they bridge metal centres. Both, 1,2-diazolates and 1,3-diazolates act as exo-bidentate ligands; however, while pyrazolates generally form double ligand bridges between metals, imidazolates form only single-ligand 24 bridges [66] (Figure 1.8). This difference in the coordination modes of these two diazolates can be explained by the steric hindrance imposed by the C H group in the 2-position between the two nitrogens in the 1,3-diazolate (i.e., imidazolate). As a consequence, binary transition-metal imidazolate compounds are typically 3-D coordination polymers, whereas binary transition-metal pyrazolates have linear chain 1-D extended structures [67, 68] (Figure 1.8). Another important property of diazolate polymers is the conjugation of electronic density throughout their structures, This is an important characteristic of these ligands in terms of their ability to mediate magnetic exchange between paramagnetic centers [68]. As a consequence of their structural characteristics, paramagnetic transition metal pyrazolate and imidazolate polymers exhibit different spin coupling behaviours. Thus, while metallic pyrazolate-bridged polymers generally exhibit antiferromagnetic coupling due to antiparallel alignment of their spins (Figure 1.8 (top)), metallic imidazolate-bridged polymers possess a canted-spin structure resulting from an imperfect antiparallel alignment of the spins (Figure 1.8 (bottom)). As previously mentioned, canted-spin antiferromagnetism leads to weak ferromagnetism at low temperatures [19]. Hence, the 3-D long-range magnetic ordering exhibited by several paramagnetic transition-metal imidazolate polymers studied in this thesis, classify them as molecule-based magnets. 25 Figure 1.8 Schematic representation of different bridging modes for pyrazolate (1,2-diazolate) (top) and imidazolate (1,3-diazolate) (bottom) metal complexes. Spin orientations (arrows) are also illustrated as expected for the two different structural motifs. 26 1.5 PHYSICAL METHODS OF CHARACTERIZATION This section is intended to give a brief description of the principles and general usefulness of the main methods of characterization utilized in this thesis. References to detailed reviews on these techniques are given in the appropriate sections. DC magnetic susceptibility, thermal gravimetric analysis, X-ray powder diffraction, UV-Vis-NIR spectroscopy and IR spectroscopy studies were performed by the author. A C magnetic susceptibility and Mossbauer spectroscopy studies were performed by Prof. William M . Reiff at the Department of Chemistry of Northeastern University, Boston, Mass., USA. X-ray single crystal diffraction studies were carried out by Dr. Steven J. Rettig and Dr. Brian O. Patrick of this Department. Mr. Peter Borda of this Department performed the elemental analyses. Experimental details for most of the physical methods of characterization employed in this work are given in Chapter 9, section 9.3. 1.5.1 MAGNETIC SUSCEPTIBILITY DETERMINATION The magnetic properties of the materials investigated here were mainly studied by DC and A C magnetization measurements. Classical methods to measure magnetization are based on the force acting on a sample when it is placed into an 27 inhomogeneous magnetic field or on the total electromotive force induced in a pick-up coil when the sample is moved in a constant magnetic field. When an isotropic paramagnetic material is placed in an inhomogeneous magnetic field, a displacement force is applied on the sample drawing it into a region of higher field. Since the displacement force depends on the both magnetization and field gradient, measurement of the force gives direct information on the magnetic susceptibility of a material. These are the fundamentals of the force methods, such as the Gouy method, [69] the Faraday method [70, 71], and the alternating force magnetometer [72, 73]. Alternatively, the change in magnetic flux density that results from placement of the material in a magnetic field may be examined by inductive methods. If a sample is inserted into an induction detection coil, then a change in the voltage is induced in this coil associated with the insertion of the sample into the detection coil. The strength of the induced voltage is given by the equation u N 2 A d i v = — S dt where N is the number of turns of wire, A is the cross-sectional area, S is the length of the coil, and di/dt is the frequency of current oscillation. The quantity p: is the 28 permeability of the material within the coil and is related to the magnetic susceptibility by the equation Lt = l + 47tx The general inductive response described in the first equation is the basis of several techniques that measure magnetic susceptibility, such as the vibrating sample magnetometer [74], A C susceptometer [75, 76], and superconducting susceptometer (SQUID) [77 - 82]. In DC (direct current) measurements, a static field, Ho, is applied, which induces a magnetization, M. In A C (alternating current) measurements, an oscillating field, Hi, is applied. Thus, the applied magnetic field can be defined as consisting of a static component HQ and an oscillating component Hi, then the magnetic field at any time, t, can be written as H(t) = H0+Hlcos(ox) where co is the period of oscillation. The resulting magnetization of a sample in the oscillating magnetic field may be written as M(t) = M0+Mt cos(aX - <p) 29 where cp is the phase angle by which the magnetization lags the oscillating component of the magnetic field. It can be then written M(t) = Xo#o +X'#> cos(a>t)+x,,^1 sin(cot) where Xo — Mx cos((p) and x' and X" depend on the frequency and magnitude of the oscillating field. Hence, the A C magnetic susceptibility is determined from its two components, the in-phase (or real) component, x', and the out-of-phase susceptibility (or imaginary) component, The in-phase susceptibility is an initial susceptibility with the same phase as the oscillating field. The out-of-phase susceptibility is related to the phase delay with respect to the oscillating field in the magnetically ordered phase. The presence of a non-zero x" response is characteristic of a magnet [83]. 1.5.2 X - R A Y DIFFRACTION Currently the most important method for determining molecular structure is single crystal X-ray diffraction. An X-ray diffraction experiment consists of placing a crystal in a monochromatic X-ray beam and then measuring the position and intensity 30 of the diffracted (scattered) rays [84]. This is also the most widely used method to obtain accurate structural information about bond lengths and angles. The necessity of a single crystal to apply this technique precluded the in-depth structural characterization of several coordination polymers obtained as microcrystalline powders in this study. Nonetheless, several key structures were, in fact, obtained here by this method. X-ray powder diffraction [85] is used to get structural information on microcrystalline powders; however, the information that can be obtained from a powder diffractogram is much more limited than that from single crystal X-ray diffraction. In this work, the powder diffraction patterns obtained are used to determine isomorphism between analogous compounds, and this is particularly useful when the molecular structure of one of the compounds is known. 1.5.3 SPECTROSCOPIC METHODS 1.5.3.1 INFRARED Infrared (IR) spectroscopy involves direct absorption of radiation that can only occur when the vibrational motion in a molecule involves some change in its dipole moment. If this happens, a vibration is said to be infrared active. This very useful technique [86] provides, among other structural characteristics in a material, immediate information on the different chemical moieties present. IR spectroscopy was useful for 31 determining the presence or absence of neutral azoles in the compounds prepared in this work. 1.5.3.2 UV-VIS-NIR Electronic spectroscopy of transition metal complexes can provide useful information about the metal chromophore geometry [87]. This was helpful in the current work primarily when microcrystalline samples of the compounds were obtained. Hence, electronic spectroscopy was used as an indirect method for structural characterization of some of the compounds obtained here. 1.5.3.3 NMR Due to the paramagnetic nature and solubility properties of the main compounds synthesized here, this very powerful structural characterization tool [88, 89] had limited use in the present research. NMR was utilized here primarily to confirm the structure, and purity, of some ligand precursors. 1.5.3.4 MOSSBAUER The discovery of the Mossbauer effect (Nuclear Gamma Resonance (ngr) spectroscopy) in 1958 [90] has led to an elegant and welcome spectroscopic technique 32 for direct microscopic observation of single ion effects and cooperative magnetic behavior in solids. Information on the type of magnetism and the onset of magnetic ordering (Tc or 7N), can be obtained with this technique. The Mossbauer effect experiment is based on the recoil free emission and subsequent recoil free, resonant absorption of low energy gamma rays (Ey generally < 100 kev) in the solid state for identical isotopes of a given element. The Mossbauer spectra of magnetic materials frequently show interesting features in the critical region near magnetic ordering transitions [91]. Hence, the onset of nuclear Zeeman splitting resulting from the growth of internal hyperfine fields owing to exchange fields and long-range magnetic ordering can be observed directly in Mossbauer effect spectra of appropriate metal nuclides, e.g. Fe-57 [92]. All the iron(H) imidazolate polymers obtained in this thesis were evaluated by Mossbauer spectroscopy studies. These studies were performed in order to confirm the presence of a magnetic transition in the compounds as well as the determination of the critical temperature, T c , at which their long-range magnetic ordering occurs. Also this spectroscopic technique was utilized to further identify the different coordination geometries found in the iron(II) imidazolate type compounds. In addition, evidence of structural phase transitions occurring in one of the compounds studied was obtained from Mossbauer spectroscopy studies. 33 1.5.4 T H E R M A L GRAVIMETRIC ANALYSIS (TGA) By monitoring the thermal decomposition of a material, additional information about its composition, and the thermal stability of a particular component of the material can be determined. In a typical experiment, the sample is placed in a very sensitive microbalance and the weight of the sample is monitored as the temperature is increased [93]. The programming capabilities of modern instrumentation involving this technique were also used here for the preparation of a coordination polymer, by thermal elimination of neutral molecules from precursor (Chapter 4, section 4.2.2.1). 1.5.5 E L E M E N T A L ANALYSIS As an essential tool this technique provides the relative percentages of analyzed elements from which an empirical formula can be determined [94]. Purity and the composition of all the compounds described in this thesis were assessed by this technique. This technique was particularly important in the initial characterization of microcrystalline coordination polymers since spectroscopic techniques, such as IR, did not provide enough composition/structural information. 34 1.6 OBJECTIVES A N D ORGANIZATION OF THIS THESIS The synthesis and structural and magnetic characterization of several paramagnetic transition metal diazolate polymers was carried out in the present thesis with the main purpose of contributing to fundamental understanding in the growing field of molecule-based magnets. Previous studies of compounds having mainly pyrazolate as the bridging ligand, established the ability of diazolate bridging ligands to promote magnetic interaction between paramagnetic transition metal ions [68, 95 - 98]. In the current work, new transition metal diazolate polymers, mainly with imidazolate type bridging ligands, were synthesized and found to have structures with one-, two- and three-dimensional frameworks. Hence, one aim of this work was achieved by correlating the magnetic properties of the compounds studied with their different structural dimensionalities. In this regard, the iron(II) diazolate polymers prepared here were the most successfully studied of the different metal systems because they were formed as macroscopic crystalline materials in a form suitable for single crystal X-ray diffraction characterization. Nevertheless, related cobalt(LT), nickel(II) and copper(H) imidazolate systems, which were obtainable only as microcrystalline powders, were also studied extensively since they were found to have magnetic properties that also characterize them as molecule-based magnets. Another pursued objective, the comparison of the magnetic properties of isostructural systems possessing different transition metal ions was also possible. Thus, 35 2-D and 3-D coordination polymers of iron(II) and cobalt(II) isostructural systems provided a unique opporrxrnity to study the influence of the d n electronic configuration on the magnetic properties of molecule-based magnets. This thesis is structured in nine Chapters and an Appendix. Chapter 1 has introduced concepts and provided general information about the different topics involved in this work. In addition, a brief description of the physical methods of characterization has been presented. Chapter 2 concerns the characterization of polybis(pyrazolato)iron(II), a one-dimensional material showing weak antiferromagnetic exchange and no long-range magnetic order. Chapters 3 to 7 describe the synthesis and structural and magnetic characterization of several new imidazolate-based magnetic materials most of which are shown to exhibit properties of molecule-based magnets at low temperature. A binary iron(II) imidazolate (and its cobalt(H) analogue) with a novel 3-D single diamondoid structure is reported in Chapter 3. Binary imidazolates of Co(LT), Ni(II) and Cu(II) are described in Chapter 4. Chapters 5 and 6 describe compounds which incorporate neutral "capping ligands" and which have 2-D extended structures, a motif never before seen for metal imidazolate polymers. A rare example of a 1-D chain polymer exhibiting long-range magnetic ordering is described in Chapter 7. In Chapter 8, a general summary of this work and suggestions for future work are provided. Lastly, Chapter 9 provides experimental details of the syntheses and the methods utilized in the physical characterizations of the compounds studied in this 36 dissertation. Crystallographic data, and selected bond distances and bond angles for all the macro-crystalline compounds studied here, are provided in the Appendix. 37 References 1. O. Kahn. Molecular Magnetism. V C H . New York. (1993). 2. D. Gatteschi. Adv. Mat. 6, 635 (1994). 3. P. W. Selwood. Magnetochemistry. Interscience. New York. 1956. 4. A. P. Long, T. D. Clark, and R. J. Prance. Rev. Sci. Instrum. 51, 8 (1980). 5. J. S. Miller, and A. J. Epstein. Angew. Chem. Int. Ed. Engl. 33, 385 (1994). 6. P. W. Anderson, in Magnetism. G. T. Rado, and H. Suhl. Eds. Vol. 1. Academic Press. New York. 1963. 7. A. P. Ginsberg. Inorg. Chim. Acta, Rev. 5, 45 (1971). 8. C. J. O'Connor. Prog. Inorg. Chem. 29, 203 (1982). p. 238. 9. L. K. Thompson, S. S. Tandon, M . E. Manuel, M . K. Park, and M . Handa in Molecule-based Magnetic Materials. M . M . Turnbull, T. Sugimoto and L. K. Thompson. Eds. ACS symposium series. Washington. 1996. 10. F. Palacio, F. J. Lazaro, and A. J. van Duyneveldt. Mol. Cryst. Liq. Cryst. 176, 289 (1989). 11. O. Kahn. Acc. Chem. Res. 33 (10) 647 (2000). 12. C. J. O'Connor. Prog. Inorg. Chem. 29, 203 (1982). p. 205. 13. O. Kahn. Molecular Magnetism. V C H . New York. (1993). p. 1. 14. L. Carlin. Magnetochemistry. Springer. Berlin. 1986. pp. 5, 11, 25. 15. J. S. Miller, and A. J. Epstein. Angew. Chem. Int. Ed. Engl. 33, 385 (1994). p. 389. 16. O. Kahn. Molecular Magnetism. V C H . New York. (1993). p. 11. 17. J. S. Miller, and A. J. Epstein. Chem. & Eng. News, 73, (40), 30 (1995). 38 18. M . Cohen. Advancing Materials Reasearch. P. A. Psaras and H. D. Langford. National Academy Press. Washington, DC. USA. 1987. p. 91. 19. T. Moriya, in Magnetism. G. T. Rado, and H. Suhl. Eds. Vol. 1. Academic Press. New York. 1963. 20. O. Kahn. Molecular Magnetism. V C H . New York. (1993). p.321. 21. F. Palacio, M . C. Moron, J. Pons, J. Casabo, K. E. Merabet, and R. L. Carlin. Phys. Lett. A, 135, 231 (1989). 22. P. W. Anderson. Phys. Rev. 115, 2 (1959). 23. Conference Proceedings in the Conference on Molecule-Based Magnets, J. S. Miller and A. J. Epstein. Mol. Cryst. Liq. Cryst., 271 (1995). 24. O. Kahn. Adv. Inorg. Chem. 43, 179 (1995). 25. H. H. Wickman, A. M . Trazzolo, H. J. Williams, G. W. Hull, and F. R. Merrit. Phys. Rev., 155, 563 (1967). 26. B. Mahan, and R. Myers, University Chemistry. Benjamin/Cummins. Menlo Park, C A , 1987; p 991. 27. K. Agawa, T. Inabe, U. Nagashima, and Y. Muruyama. J. Chem. Soc, Chem. Commun., 1617(1989). 28. R. Chiarelli, M . A . Novak, A . Rassat, and J. L. Tholance. Nature 363, 147 (1993). 29. D. A. Eichhorn, D. C. Skee, W. E. Broderick, and B. M . Hoffman. Inorg. Chem. 32, 491 (1993). 30. T. Malah, S. Thiebaut, M . Verdaguer, and P. Veillet. Science, 262, 1554 (1993). 31. Proceedings of the Fifth International Conference on Molecule-Based Magnets, Mol. Cryst. Liq. Cryst. (Eds.: K. Itoh, J. S. Miller, K. Takui), 305-306 (1997). 32. J. S. Miller, J. C. Calabrese, A. J. Epstein, R. W. Bigelow, J. H. Zang, and W. M . Reiff. J. Chem. Soc, Chem. Commun., 1026 (1986). 33. L. Carlin. Magnetochemistry. Springer. Berlin. 1986. 39 34. P. Day and I. D. Hall. J. Chem. Soc. (A), 2679 (1970). 35. F. E . Mabbs and D. J. Machin. Magnetism and Transition Metal Chemistry. Chapman and Hall. London. 1973. 36. A. Earnshaw. Introduction to Magnetochemistry. Academic Press. New York. 1968. 37. M . Gerloch. Magnetism and Ligand Field Analysis. Cambridge University Press. Cambridge. 1983. 38. J. S. Miller and A. J. Epstein. Angew. Chem. Int. Ed. Engl., 106, 399 (1994). 39. A. L. Buchachenko. Russ. Chem. Rev. 59, 307 (1990). 40. R. L. Carlin and A. J. van Duyneveldt. Magnetic Properties of Transition Metal Compounds. Springer. Berlin. 1986. 41. C. J. O'Connor. E d Research Frontiers in Magnetochemistry. World Scientific. Songapore. 1993. 42. D. H. Martin. Magnetism in Solids. MIT Press. Cambridge. 1967. 43. J. Hu, T. W. Odom, and C. M . Lieber. Acc. Chem. Res. 32, 435 (1999). 44. J. Rouxel. Ed. Crystal Chemistry and Properties of Materials with Quasi-One-Dimensional Structures. Reidel. Boston (1986). 45. C. M . Lieber, J. Liu, and P. E. Sheehan. Angew. Chem., Int. Ed. Engl. 35, 686 (1996). 46. M . Fujita, Y . J. Kwon, S. Washizu, and K. Ogura. J. Am. Chem. Soc. 119, 2861 (1994). 47. R. Pech, and J. Pickardt. Acta Crystallogr. C44, 992 (1988). 48. B. F. Abrahams, B. F. Hoskins, D. M . Michail, and R. Robson. Nature. 369, 727 (1994). 49. O. M . Yaghi, G. Li , and H. Li . Nature. 378, 703 (1995). 50. R. W. Gable, B. F. Hoskins, and R. Robson. / . Chem. Soc. Chem. Commun. 1697 (1990). 40 51. S. Subramanian, and M . Zaworotko. Angew. Chem. Int. Ed. Engl. 34, 2127 (1995). 52. S. O. H. Gutschke, A. M . Slawin, and P. T. Wood. J. Chem. Soc. Chem. Commun. 823 (1996). 53. S. S. -Y. Chui, S. M . - F , Lo, J. P. H. Charmant, A. G. Orpen, and I. D. Williams. Science. 283, 1148 (1999). 54. Rettig, S. J.; Storr, A.; Summers, D. A.; Thompson, R. C ; Trotter, J. J. Amer. Chem. Soc. 119, 8675 (1997). 55. D. J. Price, A. K. Powell and P. T. Wood. J. Chem. Soc, Dalton Trans., 3566, 2000. 56. P. Day. Coord. Chem. Rev. 190-192, 827 (1999). 57. A. F. Wells, Structural Inorganic Chemistry, 4 t h ed., Oxford, Clarendon Press. 1975, pp. 70-74. 58. W. Lin, O. R. Evans, R.-G. Xiong, and Z. Wang. J. Am. Chem. Soc. 120, 13272 (1998). 59. K. Sinzger, et. al. J. Am. Chem. Soc. 115, 7696 (1993). 60. W. R. Entley, and G. S. Girolami. Inorg. Chem. 33, 5165 (1994). 61. C. Janiak. Angew. Chem., Int. Ed. Engl. 36, 1431 (1997). 62. M . Fujita, Y . J. Kwon, S. Washizu, and K. Ogura. J. Am. Chem. Soc. 116, 1151 (1994). 63. H. C. Freeman. Adv. Protein Chem. 22, 257 (1967). 64. S. Trofimenko. Prog. Inorg. Chem. 34, 115 (1986). 65. G. La-Monica and G. A. Ardizzoia. Prog. Inorg. Chem. 46, 151 (1997). 66. S. J. Rettig, A. Storr, D. A. Summers, R. C. Thompson, and J. Trotter. Can. J. Chem. 77, 425 (1999). 67. M . K. Ehlert, S. J. Rettig, A . Storr, R. C. Thompson, and J. Trotter. Can. J. Chem. 67, 1970(1989). 41 68. M . K. Ehlert, A. Storr, D. A. Summers, and R. C. Thompson. Can. J. Chem. 75, 491 (1997). 69. C. J. O'Connor, in Molecule-based Magnetic Materials. M . M . Turnbull, T. Sugimoto and L. K. Thompson. Eds. ACS symposium series. Washington. 1996. p. 45. 70. D. A. Cruse and M . J. Gerloch. J. Chem. Soc. Dalton Trans. 152, 1977. 71. R. T. Lewis. Rev. Sci. Instrum. 42, 31 (1971). 72. R. Reeves. J. Phys. E. 5, 547 (1972). 73. C. J. O'Connor, in Molecule-based Magnetic Materials. M . M . Turnbull, T. Sugimoto and L. K. Thompson. Eds. ACS symposium series. Washington. 1996. p. 47. 74. S. Foner. Rev. Sci. Instrum. 30, 548 (1959). 75. R. L. Carlin, K. O. Joung, A. Paduan-Filho, C. J. O'Connor, and E. J. Sinn. J. Phys. C. 12, 293(1979). 76. H. A. Groenendljk, A. J. van Deyneveldt, and R. D. Willet. Physica. 101B, 320 (1980). 77. R. C. Jacklevic, J. Lambe, A. H. Silver, and J. E. Marcereau. Phys. Rev. Lett. 12, 159(1964). 78. B. S. Jr. Deaver, T. J. Bucelot, and J. J. Finney, in Future Trends in Superconducting Electronics. ALP Conference Proceedings No. 44. Charlotesville, V A . 1978. B. Deaver, C. Falco, J. Harris, and S. Wolf. Eds. ALP. New York. 1978. 79. Quantum Design Corporation. San Diego, C A 92121. 80. P. A. Salyer and L. W. ter Haar, in Molecule-based Magnetic Materials.. M . M . Turnbull, T. Sugimoto and L. K. Thompson. Eds. ACS symposium series. Washington. 1996. 81. J. A. Bertrand, E. Fujita, and D. G. Derveer. Inorg. Chem. 19, 2022 (1980). 82. W. G. Jenks, S. S. H. Sadeghi, and J. P. Jr. Wikswo. J. Phys. D: Appl. Phys. 30 (3), 293 (1997). 42 83. O. Kahn, J. Larinova, and J. V. Yakhmi. Chem. Eur. J. 5 (12) 3443 (1999). 84. M . F. Ladd and R. A. Palmer. Structure Determination by X-ray Crystallography. Plenum Press. New York. 1988. 85. L. V. Azaroff and M . J. Buerger. The Powder Method in X-ray Crystallography. McGraw-Hill. New York. 1958. 86. K. Nakamoto. Infrared Spectra of Inorganic and Coordination Compounds. Wiley-Interscience. New York. 1969. 87. A. P. B. Lever. Inorganic Electronic Spectroscopy. Elsevier. Amsterdam. 1984. 88. J. W. Akitt. NMR and Chemistry. Chapman and Hall. London. 1983. 89. J. K. M . Sanders, E. C. Constable, and B. K. Hunter. Modern nmr Spectroscopy: A Workbook of Chemical Problems. Oxford. Toronto. 1989. 90. D. P. E. Dickson and F. J. Berry. Eds. Mossbauer Spectroscopy. Cambridge University Press. Cambridge. 1986. 91. N. N. Greenwood and T. C. Gibb. Mossbauer Spectroscopy. Chapman and Hall. London. 1971. 92. W. M . Reiff. Mol. Cryst. Liq. Cryst. 176, 391 (1989). 93. A. R. West. Solid-state Chemistry and its Applications. John Wiley & Sons. New York. 1984. 94. T. S. Ma and R. C. Rittner. Modern Organic Elemental Analysis. Marcel Decker Inc. New York. 1979. 95. M . K. Ehlert, S. J. Rettig, A. Storr, R. C. Thompson, and J. Trotter. Can. J. Chem. 68, 1494 (1990). 96. M . K. Ehlert, S. J. Rettig, A. Storr, R. C. Thompson, and J. Trotter. Can. J. Chem. 69, 432(1991). 97. M . K. Ehlert, S. J. Rettig, A. Storr, and R. C. Thompson. Can. J. Chem. 70, 1121 (1992). 98. M . K. Ehlert, S. J. Rettig, A. Storr, and R. C. Thompson. Can. J. Chem. 71, 1412(1993). 43 Chapter 2 POLYBIS(PYRAZOLATO)IRON(II). A ONE-DIMENSIONAL MATERIAL SHOWING WEAK ANTIFERROMAGNETIC EXCHANGE 2.1 INTRODUCTION Metal coordination complexes with one-dimensional structures have long been investigated as materials with unusual properties. Molecule-based ferromagnets, electrical conductors, and nonlinear optical materials represent several potential applications of one-dimensional coordination polymers. Several studies have demonstrated that it is possible to modify the bulk magnetic, electronic, and optical properties of such materials [1-3]. Compared with high-dimensional materials, one-dimensional molecule-based magnetic materials sometimes have larger anisotropy favoring hard magnetic materials (stronger coercive field and larger hysteresis loop) [4]. Coordination compounds containing pyrazoles or pyrazolates as terminal or bridging ligands to transition metals have been extensively studied. Reviews such as those authored by Trofimenko [5, 6] and La Monica and Ardizzoia [7] give a good account of the development and importance of this field in coordination chemistry. The ability of pyrazolates to form double bridges between transition metal ions and 1-D extended chain structures in which the metal ions are antiferromagnetically coupled via 44 the ligands has been well documented in previous research on binary M(II) pyrazolates (M = Mn, Co, Ni, or Cu) [8-11]. While iron(n) pyrazolate complexes or polymers are relatively rare, the synthesis of iron pyrazolate polymers, [Fe(pz)2]x and [Fe(pz)3]x, appeared in 1968 [12]. The work involved reacting pyrazole with Fe(CO)s or [CpFe(CO)2]2 in benzene or toluene solvent. Only the elemental analysis of the polymeric compounds was reported without further characterization. The extensive characterization of the 1-D polymer, [Fe(pz)2]x, described here, is important for further discussion of other azolate transition metal polymers studied in this thesis. This model compound, [Fe(pz)2]x, has permitted the determination of the main structural differences between the 1,2 and 1,3-diazolate bridging motifs as well as permitted a better understanding of the influence of structure on the different magnetic behaviors observed. 2.2 RESULTS A N D DISCUSSION 2.2.1 SYNTHESIS A N D PHYSICAL PROPERTIES Details of the synthesis employed for [Fe(pz)2]x are given in Chapter 9, section 9.2.1.1. In this approach, which was used previously to obtain other iron(U) azolate polymers [13], ferrocene was reacted with an excess of pyrazole under inert atmosphere conditions. The product was isolated as red brown air-sensitive crystals. In this synthetic "metallocene" approach, the formation of the polymer involves an acid-base reaction in 45 which the pyrazole N H proton is transferred to a Cp ring of ferrocene. The resulting pyrazolate binds to the iron ion liberating cyclopentadiene, in the process: x(FeCp2) + xs(pzH) • [Fe(pz)2]x + 2x(CpH) The reaction of metallocenes with molten imidazoles has been a successful method for the preparation of divalent transition metal imidazolate coordination polymers. Often, when ferrocene is involved in this type of reaction, single crystals of the coordination polymers are obtained [13, 14]. Here, the reaction of ferrocene with pyrazole produced the polymeric Fe(II) material, [Fe(pz)2]x, in needle-like crystalline form suitable for single crystal X-ray diffraction studies. 2.2.2 SINGLE-CRYSTAL X - R A Y DIFFRACTION CHARACTERIZATION A section of the extended structure of [Fe(pz)2]x is shown in Figure 2.1. Selected crystallographic data, atom coordinates, bond lengths and bond angles appear in Appendix I, Tables 1-1 and 1-2. In this coordination polymer iron ions are linked by double pyrazolate bridges in the structure generating tetrahedral FeNj metal chromophores and extended 1-D chains. The double pyrazolate bridge between the Fe ions generates a linear chain structure (Figure 2.1). A projection of the structure down the c axis is shown in Figure 2.2. This type of structure has been determined by X-ray crystallography previously for [Cu(pz)2]x [8]. 46 C 2 Figure 2.1. Section of the polymer chain of [Fe(pz)2]x showing the atom numbering scheme. Hydrogen atoms are omitted. (50 % probability thermal ellipsoids shown). The Fe(l)—N(l) bond distance of 2.07(1) A is normal for an iron(II) azolate system [13, 14]. The N(l)—Fe(l>—N(l) bond angles, which range from 108.93(9) to 110.27(9)° are remarkably close to the value for a regular tetrahedron. The corresponding angles for the copper(U) analogue, for example, lie in the range 94.3(1) to 139.5(1)° [8]. 47 Figure 2.2. View looking almost down the c axis in the structure of [Fe(pz)2]x. (50 % probability thermal ellipsoids shown). 48 2.2.3 INFRARED SPECTROSCOPY Infrared spectroscopy was utilized to ensure that no neutral pyrazole molecules were present in the polymeric structure of [Fe(pz)2]x- The sharp and intense band at ~ 3380 cm"1, expected for neutral pyrazole (N-H stretch), was absent in the LR spectrum of this compound. Assignments of the vibrational spectrum of the pyrazolate ion were made previously [15]. 2.2.4 MAGNETIC BEHAVIOR DC magnetic susceptibility measurements were made on powdered samples of [Fe(pz)2]x at an applied field of 10 000 G over the temperature range 2 to 300 K. Plots of X and p^ ff against T are given in Figure 2.3. Although no maximum is observed in the susceptibility plot, the decrease in LLsff with decreasing temperature (4.8 U , B at 300 K; 1.75 U - B at 2 K), suggests antiferromagnetic coupling, albeit very weak. In an attempt to quantify the magnitude of the antiferromagnetic exchange in this system the magnetic data were fitted to a model appropriate for a 1-D compound. Accordingly, fits of the susceptibility data to the isotropic Heisenberg model for antiferromagnetically coupled linear chains developed by Weng [16] and Hiller et al. [17] were examined. To obtain a satisfactory fit it was necessary to include a term in the 49 expression to accommodate a fraction P of the iron ions present as a structural paramagnetic impurity. The necessity of allowing for paramagnetic impurity in analyzing magnetic data of antiferromagnetically coupled systems is not uncommon, as was found 50 to be the case in analyzing the magnetic properties of the copper(II) analogue [8]. The modified expression used was: + P(2Ng2p2 IkT) where x = \J\fkT, J is the exchange coupling constant, and g is the Lande factor. In the least squares fitting procedure used, the function minimized, F, was: 2-il/2 X calc ~ 5C o b s % obs where n = the number of data points. The value of F provides a measure of the goodness of fit. Employing g, J (the exchange coupling parameter) and P as variable parameters the best fit of experiment to theory was obtained for g = 2.01, -J= 0.591(5) cm"1 and P = 0.033(2) with F= 0.011. The solid lines shown in Figure 2.3 were calculated from theory employing these best fit parameter values and provide a visual indication of the goodness of fit. Clearly the magnetic properties of [Fe(pz)2]x are well described by the model used. X = (l-P){Ng2p2/kT 2 + 71.938*2 l + 10.482x + 955.56x3 51 The antiferromagnetic coupling in [Fe(pz)2]x is judged to be weak, based on the absence of a detectable maximum in the % versus T plot above 2 K and on the small value of the exchange coupling constant. To put the magnitude of the exchange observed here in some perspective it is necessary to compare it with that recorded for linear chain pyrazolates of other metals. To compare the magnitude of coupling in systems of different spin, rather than comparing J values it is appropriate to compare hJS21 values [18]. For [Fe(pz)2]x this is 9.4 cm"1, a value significantly smaller than the 78 cm'1 of the copper(II) analogue [8]. In an earlier summary of [US2 I values for Ni(H), Mn(II), Cu(LI) and Co(H) linear chain pyrazolates it was observed that the strength of coupling appears related to the total number of d electrons in the system, suggesting that covalency of the metal - ligand bonds is an important factor [19]. Consistent with this, the HJS2 I value of [Fe(pz)2]x is close to the range 10 to 50 cm"1 observed in Mn(II) systems and clearly outside the range 58 to 105 cm"1 observed in Cu(II) systems [19]. 2.3 S U M M A R Y A N D CONCLUSIONS Polybis(pyrazolato)iron(II), [Fe(pz)2]x, has an extended chain 1-D structure in which iron ions are doubly bridged by pyrazolate ligands. The compound exhibits weak antiferromagnetic interactions. The magnetic susceptibility data were fit to a Heisenberg 52 model for chains of antiferromagnetically coupled S = 2 metal centers, yielding the magnetic parameters -J= 0.591(5) cm"1 andg= 2.01. In characterizing polybis(pyrazolato)iron(II), [Fe(pz)2]x, as a 1-D linear chain polymer which incorporates double azolate bridges and which exhibits weak antiferromagnetic exchange mediated by the bridging ligands, the work confirms that the structural motif and consequent magnetic properties, that characterize binary metal pyrazolates of other metals, extend to iron(II) systems. As mentioned previously, the analysis made here for [Fe(pz)2]x will be useful in establishing a better understanding of the different structures, structural dimensionalities, and magnetic properties of the other transition metal azolate polymers presented in the following Chapters of this thesis. 53 References 1. C. T. Chen and K. S Suslick. Coord. Chem. Rev. 128, 293 (1993). 2. O. Kahn. Molecular Magnetism. V C H . New York. 1993. 3. J. S. Miller. Ed. Extended Linear Chain Compounds. Vol. 3.Plenum. New York, 1982. 4. P. Delahes and M . Drillon. Eds. Organic and Inorganic Linear Dimensional Crystalline Materials. NATO: New York, 1989. 5. S. Trofimenko. Chem. Rev. 72, 497 (1972). 6. S. Trofimenko. Prog. Inorg. Chem. 34, 115 (1986). 7. G. La-Monica and G. A. Ardizzoia, Prog. Inorg. Chem. 46, 151 (1997). 8. M . K. Ehlert, S. J.Rettig, A. Storr, R. C. Thompson, and J. Trotter. Can. J. Chem. 67, 1970(1989). 9. M . K. Ehlert, S. J.Rettig, A. Storr, R. C. Thompson, and J. Trotter. Can. J. Chem. 69,432(1991). 10. M . K. Ehlert, S. J.Rettig, A. Storr, R. C. Thompson, F. W. B. Einstein, and R. J. and Batchelor. Can. J. Chem. 71, 331 (1993). 11. M . K. Ehlert, A. Storr, and R. C. Thompson. Can. J. Chem. 71, 331 (1993). 12. F. Seel, and V. Sperber. Angew. Chem. Int. Ed. 7, 70 (1968). 13. S. J. Rettig, A. Storr, D. A. Summers, R. C. Thompson, and J. Trotter. J. Amer. Chem. Soc. 119, 8675 (1997). 14. S. J. Rettig, A. Storr, D. A. Summers, R. C. Thompson, and J. Trotter. Can. J. Chem. 77, 425(1999). 15. J. G. Voss; and W. L. Groeneveld. Inorg. Chim. Acta. 24,173 (1978). 16 C. H. Weng, Doctor of Philosophy Thesis, Carnegie-Mellon University, 1968. 54 17. W. Hiller, J. Strahle, A. Datz, M . Hanack, W. F. Hatfield, and P. Gutlich. J. Am. Chem. Soc. 106, 329. (1984). 18. S. L. Lambert, and D. N. Hendrickson. Inorg. Chem. 18, 2683 (1979). 19. A. Storr, D. A. Summers, and R. C Thompson. Can. J. Chem. 76,1130 (1998). 55 Chapter 3 POLYBIS(4-AZABENZIMIDAZOLATO) IRON(II) AND COBALT(II). 3-D SINGLE DIAMONDOID MATERIALS EXHIBITING WEAK FERROMAGNETIC ORDERING 3.1 INTRODUCTION As described in Chapter 1, the ability of diazolate ligands to bridge metal ions and to mediate magnetic interactions between paramagnetic centers has been well documented [1-3]. Studies on binary copper(II), cobalt(II) nickel(II) and manganese(ir) [4-9] and iron (IT) (Chapter 2) pyrazolates show that the 1,2 positioning of the nitrogen atoms in the ligands leads to polymeric 1-D linear chain structures with double azolate bridges and antiferromagnetically coupled metal centers. In contrast, it has been suggested that the 1,3 positioning of the donor nitrogens in imidazolate ligands results in steric constraints which prevent the formation of double azolate bridges between metal ions [10]. These imidazolate ligands characteristically singly bridge metals leading to higher dimensional, 2-D or 3-D, structures often possessing interesting long-range magnetic interactions. This structural motif is exhibited by [Fe3(imid)6(imidH)2]x (where imidH = imidazole), a compound in which iron(II) ions are singly bridged by imidazolate ligands in an extended 3-D lattice [10]. 56 Magnetic studies on [Fe3(inud)6(inudH)2]x revealed antiferromagnetic coupling at higher temperatures but long-range ordering and weak ferromagnetism at lower temperatures. The use of 2-methylimidazolate as the bridging ligand in an analogous reaction led to an entirely different 3-D material, [Fe(2-meimid)2* 0.13Cp2Fe]x (2-meimid = 2-methylimidazolate and Cp = cyclopentadienyl) [11], which again was found to exhibit properties characteristic of molecule-based magnets at low temperatures. In expanding these studies we decided to try to prepare a 3-D Fe(U) polymer using 4-azabenzimidazole as a precursor of the 4-azabenzimidazolate ion. Compared to imidazolate or 2-methylimidazolate, this ligand has an extra nitrogen, the 4-aza nitrogen, with potential to be coordinated to a metal ion. We were curious to know if this nitrogen would get involved in the coordination of the metal ion. Another interesting difference, compared to imidazolate or 2-methylimidazolate, is the additional bulkiness that the 4-azabenzimidazolate ligand provides. Hence we were also interested in what effect the steric hindrance in the ligand would have on the structural dimensionality of the polymeric product. The direct reaction between ferrocene and excess molten 4-azabenzimidazole yields amber-green crystals of polybis(4-azabenzimidazolato)iron(II), [Fe(4-abimid)2]x. Single crystal X-ray diffraction studies reveal that in [Fe(4-abimid)2]x the 4-aza nitrogen is not involved in coordination to the metal, the 4-azabenz-substituent serving instead to create sufficient steric bulk in the imidazolate moiety to generate a unique 3-D diamond-like (diamondoid) extended lattice. The diamond-like structure in coordination 57 polymers has attracted the attention of synthetic and materials chemists for some time now. They provide examples of 3-D "scaffolding-like materials" of potential practical importance [12, 13] and they are part of supramolecular chemistry and the emerging cross-disciplinary field of crystal engineering [14, 15]. Since the work of Kinoshita et al., [16] on bis(adiponitrilo)copper(I) nitrate, there have been a number of papers devoted to this particular molecular motif [17-25]. The title compound, [Fe(4-abimid)2]x, is the first to be reported with a totally covalent, non-interpenetrating (single) diamond-like network, exhibiting spontaneous magnetization at low temperatures. The need to explore factors affecting characteristic properties of molecule-based magnets, such as coercivity, has been recently addressed [26]. To explore the effects of altering the d n configuration, polybis(4-azabenzimidazolato)cobalt(n), [Co(4-abimid)2]x, has also been synthesized and characterized here. Evidence indicates that the polymer [Co(4-abimid)2]x is isomorphous, and probably isostructural, with the polymer [Fe(4-abimid)2]x. Although [Co(4-abimid)2]x also exhibits the properties of a molecule-based magnet its critical temperature, as well as coercive field and remnant magnetization at 4.8 K are all distinctly different from those of [Fe(4-abimid)2]x. An article regarding the structural and magnetic properties of the compounds described in this chapter has been published recently [27]. 58 3.2 RESULTS A N D DISCUSSION 3.2.1 SYNTHESES, STRUCTURES A N D PHYSICAL MEASUREMENTS The reaction of metallocenes with molten azoles has been utilized as an effective method for preparing divalent metal azolate polymers, often in macroscopic crystalline form [10, 11]. In the present work the reaction between ferrocene and excess molten 4-azabenzimidazole has afforded a polymeric Fe(II) material, [Fe(4-abimid)2]x, in macroscopic crystalline form. While, the reaction of cobaltocene with the same molten ligand generated the Co(LT) material, [Co(4-abimid)2]x, as a microcrystalline powder. Details about these syntheses are in Chapter 9, sections 9.2.1.3 and 9.2.2.6, respectively. [Fe(4-abimid)2]x, was obtained in crystalline form suitable for single crystal X-ray diffraction studies. Crystallographic data for [Fe(4-abimid)2]x, appear in Appendix I, Table 1-3. Selected bond lengths and bond angles are shown in Appendix I, Table 1-4. The repeat unit of [Fe(4-abimid)2]x is shown in Figure 3.1 and a stereoscopic view of a segment of the structure is shown in Figure 3.2. The structure consists of iron(n) ions linked by single 4-azabenzimidazolate bridges bridging through the 1,3 nitrogens giving a 3-D extended array. Coordination of the 4-aza nitrogen is not observed. The bond distances between Fe and the four nitrogen atoms in the tetrahedral chromophore are: Fe(l)—N(l) = 2.030 A, Fe(l)—N(2) = 2.046 A, Fe(l)—N(4) = 2.044 A, and Fe(l)—N(5) = 2.034 A. These values are within the expected Fe—N bond distances for 59 tetrahedral iron(IT) imidazolate complexes [10, 11]. The N—Fe(l)—N bond angles range from 102.10° to 118.24°, which correspond to angles typically obtained for FeN 4 chromophores having a distorted tetrahedral geometry [10, 11]. Fused rings of six Figure 3.1 View of the repeat unit of [Fe(4-abimid)2]x and atom numbering scheme (33% probability thermal ellipsoids). Hydrogen atoms are omitted. 60 Figure 3.2 Stereoscopic view of a section of the diamond-like framework of [Fe(4-abimid)2]x- For clarity only the iron ions and the bridging N - C - N atoms of the imidazolate rings are shown. distorted-tetrahedral iron centers form a unique covalently bonded diamond-like framework. This framework can be viewed easily in the iron ion connectivity diagram shown in Figure 3.3. Further views of this structure looking down the b axis of the unit cell are shown in Figures 3.4 and 3.5. 61 Figure 3.3 Iron ion connectivity diagram for a section of [Fe(4-abimid)2]x-62 Figure 3.4 View of [Fe(4-abixnid)2]x looking down the b axis. Notice the voids being occupied by the 4-azabenzene part of the ligand. 63 Figure 3.5 View of [Fe(4-abimid)2]x looking down the b axis. For clarity only the iron ions and the bridging N - C - N are shown. Although coordination polymers with diamond-like structures have been reported before [17-25, 28], none of them involve iron(LT). Moreover, most of these materials present different degrees of interpenetration in their diamond-like arrays. In contrast, [Fe(4-abimid)2]x consists of a single, non-mterpenetjating, diamond-like framework. 64 There are few examples of molecular compounds showing a single diamond-like motif. One of the most interesting studies of such structures was done by Hoskins and Robson [12], who employed large counter ions, such as N(CHs)4 + in [N(CH 3)4][Cu IZn I I(CN)4], and BF 4" in C u 1 ^ ^ ' ^ " ^ ' " -tetracyanotetraphenylmethane]BF4JcC6H5N02, to block the adamantane-like cavities and prevent interpenetration. The latter compound also contains molecules of nitrobenzene occupying the cavities. Another example of a single diamond-like framework was obtained in this laboratory [29]. The compound poly-bis(/j-2,5-dimethylpyrazine)copper(I) hexafluorophosphate has a cationic diamond-like lattice where the PF6 ions occupy positions in the lattice cavities. It should be noted that these reported single diamond-like structures have ionic lattices with counter ions occupying positions within the extended lattice, thus preventing interpenetration. Therefore, [Fe(4-abimid)2]x is the first example of a totally covalent coordination complex having a single diamond-like framework. Extended non-diamondoid 3-D networks in compounds incorporating iron(II) and bridging imidazolate ligands have been observed before [10, 11]. The 2-methylimidazolate iron(LI) polymer has a complex 3-D network of linear channels in which ferrocene molecules, utilized in the synthesis of the material, are trapped [11]. This contrasts with the situation seen for [Fe(4-abimid)2]x in which there is a single non-interpenetrating diamond-like lattice with nothing trapped in the lattice cavities, a consequence, presumably, of the steric bulk of the 4-azabenz-substituent (Figure 3.5). 65 By employing cobaltocene instead of ferrocene, and following the same synthetic procedure as described for [Fe(4-abimid)2]x, an analogous cobalt(U) compound, [Co(4-abimid)2]x, was obtained in microcrystalline form, and its X-ray diffraction powder pattern was determined. The powder pattern for [Fe(4-abimid)2]x was calculated, for comparison with the experimentally determined one for [Co(4-abimid)2]x, employing the program Powdercell [30]. The X-ray powder diffractogram of [Co(4-abimid)2]x, corresponds well with that calculated from the single-crystal data of [Fe(4-abimid)2]x (Figure 3.6) indicating the two materials are isomorphous. Indexing the powder data [31] for [Co(4-abimid)2]x gave an orthorhombic unit cell with lattice o parameters a = 9.72, b = 10.37 and c = 12.45 A, in close agreement with those of [Fe(4-abimid)2]x (a = 9.65, b = 10.34 and c = 12.46 A) (Appendix I, Table 1-3.). Evidence in support of the fact the iron and cobalt compounds are isostructural in addition to being isomorphous comes from spectroscopic studies. The electronic spectrum of [Co(4-abimid)2]x shows two principal absorption regions at around 1125 (broad), and between 580 nm to 540 nm (Figure 3.7). These bands can be assigned to the 4A2 -> 4Ti(F) and - » 4Ti(P) transitions, respectively, for tetrahedral cobalt(H) [32, 33]. The latter band seems to consist of two bands, one of them split. This complexity may result from transitions to doublet excited states occurring in this region [34]. Hence, complex envelopes in the visible region are generally observed for tetrahedral Co(II) chromophores [34]. It is also important to notice that, due to the strong bands 66 u —I 50 10 20 30 40 29 (deg) Figure 3.6 X-ray powder diffractograms of [Co(4-abimid)2]x (top, experimental) and [Fe(4-abimid)2]x (bottom, calculated). arising presumably from charge transfer in the lowest wavelength region of the spectrum, the bands in the 540 - 580 nm region are not as well defined as usual for tetrahedral Co(II) compounds (See Chapter 4, section 4.2.2.1). Interesting also is the presence of another band at around 1750 nm, which appears very weakly in the higher-concentration mull spectrum (Figure 3.7). This low energy band corresponds to the 4 A 2 - » 4 T 2 transition. This band typically appears in the 1000-2000 nm region in the 67 " - • 1 ' 1 1 1 1 1 500 1000 1500 2000 Wavelength (nm) Figure 3.7 Electronic spectra of [Co(4-abimid)2]x at two different mull concentrations. tetrahedral Co(II) compounds, but is often too weak to be observed [33]. Comparison of the spectral data with the corresponding Tanabe-Sugano diagram for d 7 tetrahedral systems [35], shows that these transitions are consistent with Dq and B values of 541 cm"1 and 731 cm"1 respectively. In summary, the electronic spectra of [Co(4-abimid)2]x is consistent with tetrahedrally coordinated cobalt centers, and a structure akin to that of the iron compound. Further support that the compounds are isomorphous comes from infrared spectroscopy. The infrared spectra of [Fe(4-abimid)2]x and [Co(4-abimid)2]x show very similar vibrational bands at almost identical frequencies. 68 The 3-D diamond-like structure seems to confer high thermal stability on both [Fe(4-abiniid)2]x and [Co(4-abimid)2]x, as shown by thermal gravimetric analysis (TGA) (Figure 3.8). [Fe(4-abimid)2]x is thermally stable to 402 °C. Decomposition with continuous weight loss occurs from 402 to 476 °C with a total weight loss of 70% of the initial mass. No additional loss of mass was observed up to the maximum temperature reached of 800 °C. [Co(4-abimid)2]x is thermally more robust than [Fe(4-abimid)2]x, and it does not show significant weight loss until the temperature exceeds ~ 600 °C. This material shows a gradual weight loss amounting to 50 % of the initial mass over the temperature range of - 600 °C to 800 °C. 110-100-90- V V \ \ s5 80-cent 70- [Fe(4-abimid)2] i \ \ • a fib 60- [Co(4-abimid)2] i j \ 1 \_ -= 3 50-"s. 40- \ 30-20- — i 1 1 1 1 1 1 1 1 1 ' 1 • 1 1 100 200 300 400 500 600 700 800 Temperature (°C) Figure 3.8 T G A plots for [Fe(4-abimid)2]x and [Co(4-abimid)2]x. 69 3.2.2 MAGNETIC PROPERTIES Magnetic susceptibilities of a powdered sample of [Fe(4-abimid)2]x were measured at a field of 500 G from 2 to 300 K. Figure 3.9 presents the % versus T and %T versus T data obtained below 150 K. The value of yT decreases smoothly with temperature from the value 3.38 cm3Kmol"1 at 300 K (corresponding to p^ff = 5.20 LIB) to a low of 1.60 cn^Kmol"1 just above a critical temperature, T c , of 21 K. Below T c , %T increases abruptly to a maximum value of 32.8 cm3KmoT1 at 14 K before decreasing again with temperature to 6.10 cm 3KmoT 1 at 2 K. The magnetic transition at T c is also observed in the % versus T plot (Figure 3.9). The magnetic susceptibility, which decreases smoothly with decreasing temperature below 300 K, rises abruptly (below T c), as the temperature decreases before leveling off and approaching a saturation value. This magnetic behavior exhibited by [Fe(4-abimid)2]x is similar to that reported for [Fe3(imid)6(imidH)2]x [10] and [Fe(2-meimid)20.13Cp2Fe]x [11]. It suggests antiferromagnetic coupling in which perfect antiparallel alignment of spins on neighboring metal ions does not occur due to canting of spins. This leads to a residual spin on the metal centers as the temperature is lowered. Long-range ferromagnetic ordering of these spins below T c generates a ferromagnetic transition. 70 s u J(K) £ J(K) Figure 3.9 % and %T versus T plots at 500 G for [Fe(4-abimid)2]x (top) and [Co(4-abimid)2]x (bottom). 71 The magnetization versus field plots at three temperatures, shown in Figure 3.10, reflect this anomalous magnetic behavior. The plot is linear at 35 K and extrapolates to zero magnetization at zero applied field while at 10 and 4.8 K (below T c ) the plots extrapolate to give a net magnetization at zero applied field. These results confirm that [Fe(4-abimid)2]x exhibits long-range ferromagnetic order below T c . Cycling the applied field between +55 000 and -55 000 G at 4.8 K generates a hysteresis loop, the central portion of which is shown in Figure 3.11. From this is obtained a remnant magnetization of 2100 cm3Gmol"1 and a coercive field of 80 G. This hysteresis magnetization result provides conclusive evidence that [Fe(4-abimid)2]x behaves as a magnet at low temperatures. A spin-canted structure for [Fe(4-abimid)2]x is supported by the fact that the highest magnetization reached, 6690 cm3Gmol"1, (at 4.8 K and 55 000 G) is significantly smaller than the theoretical saturation value of 22 300 cm 3Gmor 1 [36]. Further evidence for a canted spin structure comes from the structural data of [Fe(4-abimid)2]x which show a feature observed before in this type of system, that is a systematic alternation of the relative orientation of neighboring metal chromophores [10, 11]. As a measure of this the dihedral angles between the N(l)-Fe(l)-N(5) planes on adjacent, symmetry related, iron centers was examined. On every iron center, the N(l)-Fe(l)-N(5) plane forms dihedral angles of 75.4° with the corresponding planes on two of its nearest neighbors and angles of 172.5° with the corresponding planes on the other two neighbors. 72 8000 ^6000 o S s 4000 2000 • 4.8 K A 10 K • 35 K 30000 H (G) 60000 4000 3000 © S o rn S u 2000 i 1000 • 4.8 K • 10 K • 35 K A § A t i 30000 # ( G ) 60000 Figure 3.10 Magnetization versus applied field plots at different temperatures for [Fe(4-abimid)2]x (top) and [Co(4-abimid)2]x (bottom). 73 -1000 -750 -500 -250 0 250 500 750 1000 H (G) -10000 -5000 0 5000 10000 H (G) Figure 3.11 Magnetic hysteresis plots at 4.8 K for [Fe(4-abimid)2]x (top) and at 10 K for [Co(4-abimid)2]x (bottom). 74 The extent of the spin canting can be estimated by extrapolating the plot of magnetization (M) versus applied field (H) obtained at 4.8 K (Figure 3.10) to H = 0. This extrapolation gives a saturated moment (Ms(0)) of ~ 2650 cm3GmoT1 for [Fe(4-abimid)2]x- From this, an estimation of the spin canting angle, y, can be obtained using the following equation [37] Hence, the canting angle for [Fe(4-abimid)2]x is calculated to be y ~ 7°. In the earlier study on the related [Fe(2-meimid)2-0.13Cp2Fe]x, it was observed that the ferromagnetic ordering appears to be repressed by the applied field [11]. The same situation, arising presumably through saturation effects, pertains to [Fe(4-abimid)2]x. Plots of %T and % versus T (2 to 150 K range) obtained at 10 000 G are shown in Figure 3.12. Although the ferromagnetic transition is still observed at this applied field, the maximum in yT and the saturation value of % are both smaller than observed at an applied field of 500 G. 7 = tan 75 Figure 3.12 % and %T versus T plots at 10 000 G for [Fe(4-abimid)2]x (top) and [Co(4-abimid)2]x (bottom). 76 In addition to DC measurements, the magnetic susceptibility was deterauhed in an applied A C field of 1 G at 125 Hz for [Fe(4-abimid)2]x and the resulting data are consistent with long-range ferromagnetic order. An extremely sharp peak in the real part of the A C susceptibility, %', at 17.27 K is further confirmation of the spontaneous magnetization exhibited by this compound (Figure 3.13). An out-of-phase component (imaginary), %", characteristic of a non-compensated moment is present also with a peak at 17.23 K (Figure 3.13). These peak maxima in x' and %" provide more accurate measures of T c [38] than DC magnetic susceptibility studies (vide supra) which indicate the critical temperature to be 21 K. The magnetic properties of [Co(4-abimid)2]x suggest that it too can be classified as a spin canted low temperature molecule-based magnet. As for [Fe(4-abimid)2]x, %T measured at an applied field of 500 G decreases with decreasing temperature from 2.30 cn^Kmol"1 at 300 K (corresponding to p^ ff = 4.29 LIb) to a critical temperature T c = 11 K. Below 11 K it increases abruptly, signaling the onset of long-range ferromagnetic ordering (Figure 3.9). Because of the very small remnant magnetization present in [Co(4-abimid)2]x the magnetization versus field plots shown in Figure 3.10 do not clearly display the net magnetization at zero applied field for the data obtained below T c . The non-linearity of the plots obtained at 4.8 and 10 K is, however, evident. The highest magnetization measured for [Co(4-abimid)2]x was 3469 cm 3Gmor 1 at 10 K and 55 000 G. This is significantly lower than the theoretical saturation value of 16 766 77 8 6H i 4 s w 2 0 + 10 -1 1 1 1 1-• X' T 1 r 15 T(K) 20 25 Figure 3.13 A C susceptibility of [Fe(4-abirnid)2]x; = 1 G, f = 125 Hz crn3Gmor1 for a S = 3/2 system [36], again consistent with spin canting providing the source of the residual spin at low temperatures. The canting angle for [Co(4-abimid)2]x was estimated to be very small (~ 0.1°) employing the method described above for the iron analogue. The x versus T plot for [Co(4-abimid)2]x is somewhat different from that observed for [Fe(4-abimid)2]x (Figure 3.9). For [Co(4-abimid)2]x, the susceptibility shows an incipient maximum just above T c and the expected abrupt rise below T c ; 78 however, as the temperature is lowered further, instead of showing saturation, as in [Fe(4-abimid)2]x, % passes through a maximum at 9 K and then decreases in value as the temperature is lowered further to 2 K (Figure 3.9). This type of behavior, which suggests the loss of long-range ferromagnetic order at the lowest temperatures, was observed previously for another cobalt(II) spin-canted molecule-based magnet, polybis(formamide)bis(jU-formato)cobalt(II) [39]. In this earlier study it was observed that for the formate compound this disruption of long-range order is not seen at lower applied fields. This prompted us to examine the susceptibility versus temperature behavior of [Co(4-abimid)2]x at applied fields below and above 500 G. Plots of % versus T obtained at applied fields ranging from 50 to 10 000 G in the temperature range 2 -25 K are shown in Figure 3.14. Above T c the susceptibilities are essentially field independent. Below T c , at fields of 50, 100 and 500 G there is an abrupt rise in % signaling ferromagnetic ordering. At all three of these fields the susceptibility on further cooling passes through a maximum, the magnitude of which increases with decreasing applied field strength, consistent with earlier observations that ferromagnetic ordering in such systems appears to be repressed by applied fields. At the largest field studied, 10 000 G, there is no evidence of long-range ferromagnetic ordering as the susceptibility simply passes through a single maximum at about 10 K, indicative of the antiferromagnetic coupling. This is more clearly seen in Figure 3.12 which also shows that there is no magnetic anomaly in the %T plot for [Co(4-abimid)2]x at 10 000 G. In Figure 3.12, the susceptibility for [Co(4-abimid)2]x is seen to approach a constant value of ~ 0^ 043 cn^mol"1 at the lowest temperatures, a consequence, presumably, of the spin 79 0.5 • • 50 G # 100 G 0.4 - • • o 500 G A 10 000 G ' - . 0 3 O 2 « S e 0 . 2 0.1 { 0.0 • • • _ • o 0 ° 0 I 10 15 J(K) ft ft 20 25 Figure 3.14 % v e r s u s T P l o t s f o r [Co(4-abimid) 2] x at 50, 100, 500 and 10 000 G . canting. A t 2 K this % value corresponds to an effective magnetic moment o f 0.83 \\%. A t the three lowest fields studied, where ferromagnetic ordering is seen, the susceptibilities decrease below 9 K and, in all cases, approach at the lowest temperature studied the same value as that recorded at 10 000 G (Figure 3.14). There is no simple explanation for this apparent loss in ferromagnetic order at low temperatures. The cause could be at the single-ion level. The 4 A 2 electronic ground state o f tetrahedral cobalt(D) is subject to zero-field splitting and i f this is large enough significant changes in the 80 population of zero-field split levels at low temperatures could affect the exchange. The significance of this factor will depend of course on the magnitude of the zero-field splitting. This may account for the fact that, although the 5 E ground state of tetrahedral iron(II) is also subject to zero-field splitting, loss in ferromagnetic ordering at low temperatures is not seen for the iron analogue, [Fe(4-abimid)2]x. A consequence of the phenomenon just discussed for the cobalt compound, [Co(4-abimid)2]x, is that its hysteresis properties measured below T c depend significantly on temperature. Measured at 10 K the hysteresis plot (central portion shown in Figure 3.11) yields = 22 cm3Gmol"1 and H c o e r = 400 G while at 4.8 K (the temperature at which the hysteresis behaviour was measured for [Fe(4-abimid)2]x) Mrem = 6 cm^Gmol"1 and H c o e r = 100 G. A C magnetic susceptibility measurements determined for [Co(4-abimid)2]x show this compound has a non magnetic ground state (Figure 3.15). In contrast to the A C susceptibility behaviour of [Fe(4-abimid)2]x (Figure 3.13), the cobalt analogue displays just a discontinuity on the real component of the susceptibility, x', at 11 K (Figure 3.15). The imaginary component, does not show a maximum. This result is not totally inconsistent with the DC susceptibility study which clearly indicated a loss in ferromagnetic order at low temperatures for [Co(4-abimid)2]x- In order to further investigate this behavior, the A C susceptibility measurements were done with application of a small DC field. It was expected that the presence of a low DC magnetic 81 0.060 0.055 -0.050 -0.045 -0.040 -| 0.035 H 'g 0.030 H ^ 0.025-0.020-0.015 -0.010 -0.005 -0.000 • i 1 i • i 1 i • i • i 1 i • i • i 1 i 1 i 1 i x' A • i 1 i • i 1 i • i—1—r i 1 i 3 4 5 6 7 8 9 10 11 12 13 14 15 16 T ( K ) 0.060 -0.055-0.050-0.045-0.040-— r • i 1 i 1 i x' . * : •• © 0.035- -s 0.030- -s w 0.025-0.020 -0.015--0.010- x" : 0.005- -0.000 -1 — i 1 1 1 1 1 1 T ( K ) Figure 3.15 A C susceptibility for [Co(4-abirnid)2]x; HAC = 1 G, f = 125 Hz (top) and HAC = 1 G, Hoc = 20 G, f = 125 Hz (bottom). 82 field (20 G) would help to reveal the magnetic transition in the A C susceptibility study of [Co(4-abimid)2]x- As can be seen in Figure 3.15 the application of the small DC field had basically no effect on the A C susceptibility behaviour. Hence, in concordance with the magnetic studies discussed above it was found that, although [Co(4-abimid)2]x also exhibits the properties of a molecule-based magnet its critical temperature, as well as coercive field and remnant magnetization at 4.8 K are all markedly different from those exhibited by [Fe(4-abimid)2]x. 3.2.3 MOSSBAUER SPECTROSCOPY The Mossbauer spectrum of [Fe(4-abimid)2]x shows a single quadrupole split doublet at 77.3 K (Figure 3.16) corresponding to a single tetrahedral site, and consistent with the structure determined single crystal X-ray diffraction studies (Figure 3.1). The isomer shift of about 0.83 mm s"1 at 77.3 K is typical of a tetrahedral ferrous chromophore [40] whereas the quadrupole splitting of 3.01 mm s"1 indicates a large low symmetry ligand field component lifting the degeneracy of the nominal 5 E ground state of regular Tj symmetry [41]. This single, sharp, narrow line-width doublet is maintained down to 18.5 K (Figure 3.17). At a temperature of about 18 K magnetic hyperfine splitting occurs signaling the onset of long-range magnetic order. This is consistent with the findings of the magnetization experiments discussed above and the 83 10 -2 V e l o c i t y R e l a t i v e to F e ( m m s ) Figure 3.16 Mossbauer spectrum of [Fe(4-abimid)2]x at 77.3 K. temperature is in good agreement with the ones determined by the DC and A C susceptibility measurements (21 K and 17.27 K respectively). For a unique Fe(II) site the number of hyperfme lines should be six. Careful examination of the spectra, particularly the one at 4.3 K (Figure 3.17), reveals more than six lines suggesting the presence of at least two unique iron sites. This is not in agreement with the higher temperature X-ray diffraction structure and suggest a structural phase change at low temperatures. A low temperature X-ray diffraction structure determination for [Fe(4-abimid)2]x (below 18 K) is needed to confirm this finding. 84 s e '•G a e = S o Oj. 3-6 9 -6 0 3 6 9^  12 » i . mm i 77.3 K 0 2 4 19 K - 6 ^ - 2 0 2 18.5 K "". f 6 -2 18 K V -2 0 15 K "V -2 0 4.3 K \ /\ /V^ w -2 V e l o c i t y re la t ive to F e / m m s" Figure 3.17 Selected Mossbauer spectra for [Fe(4-abimid)2]x at various temperatures 85 3.3 S U M M A R Y A N D CONCLUSIONS [Fe(4-abimid)2]x and [Co(4-abimid)2]x provide the first examples of isomorphous and presumably isostructural molecule-based magnets of two different metals. The objective in comparing two such materials was to examine the effect on the magnetic properties of changing the d n configuration of the metal. In both [Fe(4-abimid)2]x and [Co(4-abimid)2]x the metal is in a pseudo-tetrahedral geometry. In terms of single ion effects there is no first-order orbital contribution to the magnetic moment in either case [42] and the primary difference lies in the spin contributions, S = 2 for [Fe(4-abimid)2]x and S = 3/2 for [Co(4-abimid)2]x- This may, partly at least, contribute to the smaller remnant magnetization observed for [Co(4-abimid)2]x (6 cn^Gmol"1 at 4.8 K and 22 cn^Gmol"1 at 10 K) compared to that for [Fe(4-abimid)2]x (2100 cm 3 Gmor 1 at 4.8 K). It seems likely, however, that the degree of spin canting, calculated and described above to be very small for the cobalt compound, would be much more important in this regard. The coercive fields are not as remarkably different in the two materials. This quantity, defined as the applied field required to return the magnetization of the sample to zero, is 80 G (at 4.8 K) for [Fe(4-abimid)2]x and 100 G at 4.8 K (400 G at 10 K) for [Co(4-abimid)2]x. In terms of hysteresis behavior another characteristic property is the range of applied fields over which the magnetization of the sample is dependent on the history of the field sweep (increasing or decreasing). In this regard the samples are quite different. For [Fe(4-abimid)2]x this field range is approximately ±1 000G at 4.8 K while for [Co(4-abimid)2]x it is about ±10 000 Gat 86 10 K. At 4.8 K the magnetization of [Co(4-abimid)2]x is dependent on the history of the field sweep over the entire range of fields studied. Finally it is noticeable that below T c ferromagnetic ordering persists at higher applied fields for [Fe(4-abimid)2]x than for [Co(4-abimid)2]x and that the apparent loss in order at low temperatures and all fields exhibited by [Co(4-abimid)2]x is not observed for [Fe(4-abimid)2]x. On the basis of this single study it would be dangerous to draw general conclusions. To establish the generality of such findings investigation of other isomorphous pairs of iron and cobalt compounds is required. Also to better understand the detailed aspects of the magnetic properties of both compounds single crystal and powder X-ray diffraction studies at He temperatures are needed. Ideally, neutron diffraction experiments would also be very useful. The difficulty here is that larger single crystals of the iron(II) compound are required and, since presence of hydrogen atoms interfere in the neutron diffraction results, a deuterated [Fe(4-abimid)2]x would need to be prepared [43,44]. 87 References 1. M. K. Ehlert, S. J. Retig, A. Storr, R. C. Thompson, and J. Troter. Can. J. Chem. 67, 1970(1989). 2. M. A. Martinez-Lorente, V. Petrouleas, R. Poinsot, J. P. Tuchagues, J. M. Savariault, and M. Drilon. Inorg. Chem. 30, 3587 (1991). 3. M. K. Ehlert, S. J. Retig, A. Storr, R. C. Thompson, and J. Troter. Can. J. Chem. 68, 1444(1990). 4. M. K. Ehlert, S. J. Retig, A. Storr, R. C. Thompson, and J. Troter. Can. J. Chem. 68, 1494(1990). 5. M. K. Ehlert, S. J. Retig, A. Storr, R. C. Thompson, and J. Troter. Can. J. Chem. 69,432(1991). 6. M. K. Ehlert, S. J. Retig, A. Storr, and R. C. Thompson. Can. J. Chem. 70, 1121 (1992) . 7. M. K. Ehlert, S. J. Retig, A. Storr, and R. C. Thompson. Can. J. Chem. 71, 1412 (1993) . 8. Ehlert, M. K.; Retig, S. J.; Storr, A.; Thompson, R. C; Einstein, F. W. B.; Batchelor, R. J. Can. J. Chem. 71, 331 (1993). 9. M. K. Ehlert, A. Storr, D. A. Summers, and R. C. Thompson. Can. J. Chem. 75, 491 (1997). 10. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter. J Amer. Chem. Soc. 119, 8675 (1997). 11. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter. Can. J. Chem. 77, 425 (1999). 12. B. F. Hoskins and R. Robson. J. Amer. Chem. Soc. Ill, 1546 (1990). 13. L. R. MacGilivray, S. Subramanian, and M. J. Zaworotko. J. Chem. Soc. Chem. Commun. 1325 (1994). 14. M. Ward. Chemistry in Britain. September, 52 (1998). 15. M. Munakata, L. P. Wu, and T. Kuroda-Sowa. Adv. Inorg. Chem. 46, 173 (1999). 88 16. Y. Kinoshita, I. Matsubara, T. Higuchi and Y. Saito. Bull. Chem. Soc. Jpn. 34, 1221 (1959). 17. 0. Ermer. Adv. Mater. 3, 608 (1991). 18. B. F. Abrahams, M. J. Hardie, B. F. Hoskins, R. Robson, and G. A. Wilams. J. Am. Chem. Soc. 114, 10641 (1992). 19. O. M. Yaghi, D. A. Richardson, G. Li, E. Davis, and T. L. Groy, Mater. Res. Soc. Symp. Proc, 371 (Advances in Porous Materials), 15, 1995. 20. T. Vossmeyer, G. Reck, L. Katsikas, E. T. K. Haupt, B. Schulz, and H. Weler. Science. 267, 1476(1995). 21. T. Kuroda-Sowa, M. Yamamoto, M. Munakata, M. Seto, and M. Maekawa. Chem. Lett., 349 (1996). 22. T. Kuroda-Sowa, T. Horino, M. Yamamoto, Y. Ohno, M. Maekawa, and M. Munakata. Inorg. Chem. 36, 6382 (1997). 23. S. Lopez, M. Kahraman, M. Harmata, and S. W. Keler. Inorg. Chem. 36, 6138 (1997). 24. S. M. Stalder and A. P. Wilkinson. Chem. Mater. 9,2168(1997). 25. E. Siebel, A. M. A. Ibrahim, and R. D. Fischer. Inorg. Chem. 38, 2530 (1999). 26. M. G. F. Vaz, L. M. M. Pinheiro, H. O. Stumpf, A. F. C. Alcantara, S. Golhen, L. Ouahab, O. Cador, C. Mathoniere, and O. Kahn. Chem. Eur. J. 5, 1486 (1999). 27. S. J. Retig, V. Sanchez, A. Storr, R. C. Thompson, and J. Troter. J. Chem. Soc, Dalton Trans. 3931 (2000). 28. W. Choe, Y.-H. Kiang, Z. Xu, and S. Lee. Chem. Mater. 11, 1776 (1999). 29. T. Otieno, S. J. Retig, R. C. Thompson, and J. Troter. Inorg. Chem. 32, 1607 (1993). 30. PowderCel, version 2.3, W. Krauss and G. Nolze, Federal Instiute for Materials Research and Testing (BAM). Berlin, 1997. 31. R. Shirley, The CRYSFIRE System for Automatic Powder Indexing: User's Manual, The Latice Press, Guildford, 1999; J. W. Visser, /. Appl. Crystallogr. 2, 89(1969). 32. A. M. Vecchio-Sadus. Transition Met. Chem. 20,46 (1995). 89 33. M. K. Ehlert, A. Stor  and R. C. Thompson. Can. J. Chem. 71, 1412 (1993). 34. F. A. Coton and G. Wilkinson. Advanced Inorganic Chemistry. A Comprehensive Text. 4th Edition. Wiley-Interscience. New York. 1980. pp. 770-772. 35. Y. Tanabe and S. Sugano. J. Phys. Soc. Jpn. 9, 753 (1954). 36. R. L. Carlin, Magnetochemistry. Springer-Verlag. Berlin. 1986. pp. 7-9. 37. F. Palacio, M. Andres, R. Home, and A. J. van Duyeneveldt. J. Magn. Magn. Mater. 54-57, 1487 (1986). 38. O. Kahn. Acc. Chem. Res. 33 (10) 647 (2000). 39. S. J. Retig, R. C. Thompson, J. Troter and S. Xia. Inorg. Chem. 38, 1360 (1999). 40. B. W. Dockum and W. M. Reiff. Inorg. Chim. Acta. 120, 61 (1986). 41. N. N. Greenwood and T. C. Gibb. Mossbauer Spectroscopy. Chapman and Hal Ltd. London. 1971. 42. B. N. Figgis and M. A. Hitchman, Ligand Field Theory and Its Applications. Wiley-VCH. New York. 2000. 43. J. L. Manson, Q. Huang, J. W. Lynn, H-J. Koo, M-H. Wahngbo, R. Bateman, T. Otsuka, N. Wada, D. N. Argyriou, and J. S. Miler. J. Am. Chem. Soc. 123, 162 (2001). 44. C. Hardacre, J. D. Holbrey, and S. E. J. McMath. J. Chem. Soc. Chem. Commun. 367 (2001). 90 Chapter 4 BINARY IMIDAZOLATES OF COBALT(II), NICKEL(II), AND COPPER(II) 4.1 INTRODUCTION As previously stated, the non-centrosyrnmetric M-L-M exchange pathway provided by single-bridging 1,3-diazolate ligands (Chapter 1, section 1.4) appears to be a key factor in generating spin canting and weak feromagnetism in [Fe3(imid)6(imidH)2]x [1], [Fe(2-meimid)2- 0.13Cp2Fe]x [2], and [Fe(4-abimid)2]x (Chapter 3). The phenomenon should not be restricted to coordination polymers of iron(Fl). This fact, coupled with the observation that the iron centers involved in the primary exchange pathways in the above systems are tetrahedraly coordinated, a common geometry for cobalt(H), prompted interest in broadening the investigation to include the magnetic properties of related cobalt systems. This work was later extended to the nickel and copper imidazolate systems. 4.2 COBALT(Il) IMIDAZOLATE POLYMERS 4.2.1 INTRODUCTION Reported here are the synthesis, structural studies and magnetic properties of five cobalt systems: [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x 91 (benzimid = benzimidazolate), and [Co3(imid)6(imidH)2]x. [Co(imid)2]x has been reported previously and its structure determined by single crystal X-ray difraction studies [3]. We had hoped to determine the molecular structures of the other cobalt(I) imidazolates studied here; unfortunately, in spite of utilizing a synthetic route which has been very successful in obtaining single crystals of related iron(I) imidazolate polymers [1, 2], it was not possible to obtain macroscopic single crystals of these Co(H) compounds. In the present work we were, however, able to show by X-ray powder difraction studies that [Co3(imid)6(imidH)2]x, is isomorphous, and presumably isostructural, with the iron analogue. The later has an extended 3-D latice structure [1]. [Co(2-meimid)2]x, [Co(4-meimid)2]x and [Co(benzimid)2]x, have been reported previously [4-6] and although definitive structures of these compounds remain elusive, spectroscopic and thermal analysis data described below atest to their polymeric nature. [Co(imid)2]x was chosen to be investigated, in particular, because its structure is known and because previous magnetic measurements were conducted at high temperatures only [3, 6]. Of the other four compounds only [Co(benzimid)2]x has been subjected to magnetic studies previously, again only at high temperatures [7]. New magnetization studies to cryogenic temperatures on the five cobalt systems are reported here. All five compounds show the presence of antiferomagnetic exchange, and three of them, [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x show clear transitons below critical temperatures, Tc, to low-temperature long-range feromagneticaly ordered states. 92 4.2.2 RESULTS AND DISCUSSION 4.2.2.1 SYNTHESES, PHYSICAL, THERMAL AND STRUCTURAL CHARACTERIZATION [Co(imid)2]x, was prepared by reacting cobalt(IT) nitrate hexahydrate with an excess of imidazole in water (Chapter 9, section 9.2.2.1). This method yields a purple microcrystaline powder. A method involving basic conditons and another, an electrochemical procedure, have been reported in the literature [3, 4-6]. [Co(imid)2]x was also produced in the present study by thermal decompositon of [Co3(imid)6(imidH)2]x, at ~ 325 °C, at which temperature the loss of the neutral imidazole molecules occurs (vide infra). The methods described for the synthesis of [Co(2-meimid)2]x, [Co(4-meimid)2]x, and [Co(benzimid)2]x (Chapter 9, sections 9.2.2.2 through 9.2.2.4), also led to purple microcrystaline materials and eforts to produce samples suitable for single crystal X-ray studies were unsuccessful. The structure of [Co(imid)2]x, was determined previously by X-ray crystalography [3]. This crystaline form of the compound displays tetrahedraly coordinated cobalt(I) ions. It is a 3-D polymer consisting of fused puckered rings of four tetrahedraly coordinated cobalt(l) ions linked by single-bridging imidazolates. A representation of the asymmetric unit of [Co(imid)2]x is shown in Figure 4.1. This ilustration was obtained employing the software Powdercel [8] using the crystalographic data previously reported [3]. 93 Figure 4.1 Asymmetric unit of [Co(imid)2]x. View looking down the c axis. Hydrogen atoms are omited. Although the exact structures of [Co(2-meimid)2]x, [Co(4- meimid)2]x, and [Co(benzimid)2]x remain unknown the polymeric nature of the materials is indicated by their physical properties, which to some extent resemble those determined for [Co(imid)2]x. All four of these compounds are stable both in air and in contact with moisture. They are insoluble in water and common organic solvents; they are nonvolatile, and thermaly robust, and they decompose when treated with concentrated mineral acids. Thermal gravimetric analyses of these materials show no mass loss due to thermal 94 decompositon or sublimation below 200 °C and at higher temperatures the compounds present similar thermal decompositon behaviors with no significant weight loss till temperatures above 400 °C are reached (Figure 4.2). [Co(benzimid)2]x is the most thermaly robust of the four compounds in that there is no indication of thermal decompositon below 600 °C for this material. The electronic spectra of [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, and [Co(benzimid)2]x are very similar (Figure 4.3). They show two intense d-d transition bands, the first one centered (determined visualy) around 1110 - 1130 nm (broad and structured), and the second one observed between 570 - 590 nm, the later with a shoulder at about 520 -540 nm. These can be tentatively assigned to the 4A2 -» 4Ti(F) and 4A2 -» 4Ti(P) transitions, respectively, of (distorted) tetrahedral cobalt(H) [9]. The absorptions seen at the lowest wavelengths (Figure 4.3) arise from charge transfer transitions. An expected third d-d band due to the 4A2 —»4T2 transition, expected in the 1500 - 2500 nm region, is usualy too weak and broad, to be clearly identified in spectra of muls of the type studied here. Nevertheless, this third band is just apparent in the spectra of [Co(imid)2]x, [Co(4-meimid)2]x and [Co(benzimid)2]x between 1750 to 2020 nm. Using average band wavelengths of 580nm and 1120 nm, Dq and B parameters of 525 cm"1 and 700 cm"1, respectively, were calculated using the appropriate Tanabe-Sugano corelation diagram [10]. These values of Dq and B are in close agreement with those reported previously for other tetrahedral Co(I) imidazolate and pyrazolate complexes [6,11]. 95 Ok - 60 — o 40 20 [Cotimid),], [Co(2-meimid)2]x [Co(4-meimid)2]I [Co(benzimid)2]x [COjCimid^imidH),], 200 400 Temperature (°C) 600 800 Figure 4.2 TGA plots for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(innd)6(imidH)2]x . Using these Dq and B values the 4A2 —> ^2 transition is predicted to appear at ~ 5250 cm"1 (~ 1904 nm), which is within the wavelength range where the very weak, broad, absorption was observed in some of the spectra (vide supra). Electronic spectral data for al four of these compounds have been reported previously [6, 7, 12]. Previous studies [6] on the difuse reflectance spectra of [Co(imid)2]x and [Co(4-meimid)2]x reported the 96 Figure 4.3 UV-Vis-NER spectra for compounds [Co(imid)2]x, (a); [Co(2-meimid)2]x, (b); [Co(4-meimid)2]x, (c); [Co(benzimid)2]x, (d); and [Co3(imid)6(imidH)2]x, (e)-97 presence of bands at 1840 and 2290 nm for [Co(imid)2]x, and 1860, 1990 and 2280 nm for [Co(4-meimid)2]x. These are likely components of the 4A2 —» 4T2 transition seen only very weakly in the spectra studied here. An earlier electronic spectroscopy study [7] on [Co(benzimid)2]x, using both reflectance and mul methods, reported the observation of only the two major d-d transitions, with the mul spectrum bands at 1150 and 595 nm (a shoulder at -540 nm). Hence, the spectra shown in Figure 4.3 are in general agreement with the earlier work which also concluded that these compounds have structures incorporating tetrahedraly coordinated cobalt(Il) centers. The absence of a V N - H stretching vibration in these electronic spectra and in the infrared spectra of these complexes (see the discussion below concerning the observation of this band in [Co3(imid)6(imidH)2]x) is indicative of the absence of neutral imidazole in these materials [1]. The X-ray powder diffactogram of a sample of [Co(imid)2]x synthesized in this work agrees wel with that calculated [8] from single crystal X-ray difraction data reported for the same compound [3] (Figure 4.4). Indexing the X-ray powder difractogram using the program Celref [13] yields calculated latice parameters of a = 22.834, b = 22.834 and c = 12.983 A. These values are very similar to those reported for the same compound in the previous study (a = 22.872, b = 22.872, c = 12.981 A) [3]. We conclude that [Co(imid)2]x synthesized in the present work is the same as that reported earlier [3]. 98 5 10 15 20 25 30 35 40 45 50 26 (deg) Figure 4.4 X-ray powder difractograms of [Co(imid)2]x (top, experimental; botom, calculated). Unfortunately, relatively little structural information was obtained from the powder X-ray difraction studies of the other three cobalt compounds. The X-ray powder difraction paterns for [Co(2-meimid)2]x and [Co(benzimid)2]x are shown in Figure 4.5. From these difractograms it can be seen that these compounds are not isomorphous with each other, nor with [Co(imid)2]x. A very poor difraction patern with no detectable peaks, characteristic of an amorphous solid, was obtained for [Co(4-meimid)4]x. 99 20 (deg) 20 (deg) Figure 4.5 X-ray powder difractograms of [Co(2-meimid)2]x (a) and [Co(benzimid)2]x (b). In summary, concerning structures, the similarities in stoichiometry and physical and spectroscopic properties of [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, and [Co(benzimid)2]x strongly support the conclusion that, as known definitively for [Co(imid)2]x, al have extended structures with tetrahedraly coordinated metal centers and singly bridging azolate ligands. 100 In an earlier study [1] the synthesis of a new molecule-based magnet, [Fe3(imid)6(imidH)2]x, by the reaction of ferocene with excess molten imidazole, was reported. The same reaction, employing cobaltocene in place of ferocene, yields the analogous cobalt compound, [Co3(imid)6(imidH)2]x, as a microcrystaline powder (Chapter 9, section 9.2.2.5). The X-ray powder difractogram of [Co3(imid)6(imidH)2]x coincides wel with that calculated (employing single crystal X-ray difraction data [1] and the program PowderCel, [8]), for [Fe3(imid)6(imidH)2]x (Figure 4.6). Indexing the X-ray powder difractogram of [Co3(imid)6(imidH)2]x using the program Celref [13] yields calculated latice parameters of a = 10.568, b = 12.964 and c = 10.634 A. These are very similar to those of [Fe3(imid)6(imidH)2]x (a = 10.591, b = 12.958, c = 10.617 A) [1]. Additonal evidence supporting the conclusion that the iron and cobalt compounds are isostructural in additon to isomorphous comes from spectroscopic studies on [Co3(imid)6(imidH)2]x- The X-ray determined structure of the iron compound shows the presence of both tetrahedral and octahedral metal centers, the additonal coordination sites on the later being filed by neutral imidazole molecules. The electronic spectrum of [Co3(imid)6(imidH)2]x (Figure 4.3) shows the absorption bands assignable to the 4A2 -> 4Ti(F) and 4A2 -> 4Ti(P) transitons of tetrahedraly coordinated Co ions (see previous discussion). Transitons assignable to octahedral cobalt(I) centers are not observable; however, this is not too surprising in view of the fact they would be expected to be an order of magnitude or more weaker than the bands arising from the tetrahedral centers [14]. Evidence for neutral imidazole molecules (coordinated, presumably, to octahedral cobalt centers) in [Co3(imid)6(imidH)2]x comes from the observation of a sharp peak at 101 2950 nm in the electronic spectrum (Figure 4.3) coresponding to the V N - H stretching vibration of neutral imidazole [1]. Finaly we note that the infrared spectrum of LU —I— 10 ^0 To" ~50 40 2 0 ( d e g ) Figure 4.6 X-Ray powder diffractograms of [Co3(imid)6(imidH)2]x (top, experimental) and Fe3(imid)6(imidH)2 (bottom, calculated). 102 [Co3(imid)6(imidH)2]x and the iron analogue are virtualy identical exhibiting the same vibrational bands at almost the same frequencies. It can be concluded that [Co3(imid)6(imidH)2]x has the same structure as the iron analogue [1]. In this structure, tetrahedraly coordinated metal ions are connected in chains by singly bridging imidazolate ions. The chains are cross-linked by octahedraly coordinated metal ions. Each tetrahedral center is linked to two others in the same chain and via octahedral centers to two additonal chains. Each chain is linked to four diferent chains via the octahedral centers. [Co3(imid)6(imidH)2]x is nonvolatile and does not melt up to the temperature at which its thermal decompositon begins (208 °C). Thermal gravimetric analysis of the compound shows that it decomposes in several steps (Figure 4.2). Approximately 15% of the initial weight is lost between 208 and 255 °C. The next 4% is lost between 255 and 320 °C. This total weight loss of 19% coresponds to that expected for the loss of the neutral imidazole ligands. The remaining material has the same compositon as [Co(imid)2]x- This is evidenced by the thermogravimetric data and by the experiment described in the experimental section in which a sample was retrieved from the TA instrument, after heating for 30 minutes at 325 °C, and subjected to elemental analysis. The rest of the thermolysis curve of [Co3(imid)6(imidH)2]x above 320 °C paralels closely that of the authentic sample of [Co(imid)2]x, as expected. 103 4.2.2.2 MAGNETIC PROPERTIES Magnetic susceptibilty, %, and magnetic moment, ji^ f, versus T data over the temperature range 2-300 K on powdered samples of the five cobalt compounds in an applied field of 10 000 G are shown in Figures 4.7 and 4.8, respectively. For al five [Co(imid)2]x [Co(2-meimid)2]I [Co(4-meimid)2]x [Co(benzimid)2]x [Co3(iimd)6(iimdH)2]x 50 100 150 200 250 300 T ( K ) Figure 4.7 % versus T plots at 10 000 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(imid)6(hnidH)2]x. 104 3 i • • • • • • • 0 0 0 0 0 6 ne ° 0 X 4 • [Co(imid)2]x [C o(2-meimid)2] B • [Co(4-meimid)2]x • |Co(benzimid),]x 0 [COjCimid^CimidH),] 50 100 150 200 250 300 T(K) Figure 4.8 [i^s versus T plots at 10 000 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(innd)6(imidH)2]x. compounds the magnetic moment decreases with decreasing temperature from a value above 4 p.R at 300 K to a value below 1 u.B at 2 K (Figure 4.8). This suggests antiferomagnetic coupling between magnetic centers, a conclusion further supported by the susceptibilty data for [Co(2-meimid)2]x and [Co(4-meimid)2]x which show broad maxima at low temperatures. In contrast to the results for [Co(2-meimid)2]x and [Co(4-meimid)2]x, the data for [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x show 105 clear evidence of magnetic anomalies in the region below 20 K. These are noticeable in both the Lieff and % plots, particularly in the former. The abrupt increase in % suggests a transition to a feromagnetic state for these compounds and since saturation efects caused by large applied fields can mask such behavior (See Chapter 3, section 3.2.2) the magnetic properties of al five compounds were examined at the lower applied field of 500 G. Lieff and % data at 500 G over the low temperature region (2-50 K) are shown in Figures 4.9 and 4.10, respectively. While the magnetic properties of [Co(2-meimid)2]x and [Co(4-meimid)2]x are virtualy unchanged at this lower applied field, as expected where only short range antiferomagnetic interactions are involved, % and Lieff data of [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x are dramaticaly altered below the temperature of the anomaly. Below a critical temperature, Tc, (16 K, 13 K and 15 K for [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x respectively) LL-ff increases abruptly to a maximum value (7.84 LIB at 9 K for [Co(imid)2]x, 5.59 Li B at 10 K for [Co(benzimid)2]x, and 6.02 |iB at 13 K for [Co3(imid)6(imidH)2]x) before decreasing again as the temperature decreases to 2 K (Figure 4.9). The magnetic transition at Tc is also seen clearly in the % versus T plots (Figure 4.10). The susceptibilty increases with decreasing temperature below 300 K and above Tc for [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x. Below Tc the susceptibilty rises abruptly as the temperature decreases before leveling of and approaching a saturation value for 106 ea is 5 zL 10 8 6 8 [Co(imid)2]x [Co(2-meimid)2]x [Co(4-meimid)2]x [Co(benziinid)2]x [Co3(iiiud)6(iinidH)2]J I 1*1 8 8 8 10 20 30 T ( K ) 40 50 Figure 4.9 Lieff versus T plots at 500 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(iniid)6(imidH)2]x. [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x. The susceptibilty of [Co(imid)2]x maximizes at 11 K then decreases on cooling further before leveling of  at the lowest temperatures studied (Figure 4.10). The magnetic behaviors of [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x indicate in al three compounds the presence 107 1.0 0.8 J - 0.6 o £ & 0.4 0.2 i 0.0 OOOOOS6OQ0 o * o o o [Co(imid)2]x [Co(2-meimid)2]x [Co(4-meimid)2]x [Co(benzimid)2]x [Co3(imid)6(imidH)2]x 0 10 20 30 T ( K ) 40 50 Figure 4.10 % versus T plots at 500 G for compounds [Co(imid)2]x, [Co(2-meimid)2]x, [Co(4-meimid)2]x, [Co(benzimid)2]x, and [Co3(inn^ )6(imidH)2]x. of antiferomagnetic coupling between paramagnetic centers as the primary exchange mechanism combined with a magnetic phase transition to a feromagneticaly ordered state at low temperatures. This magnetic behavior is very similar to that reported for [Fe3(imid)6(imidH)2]x [1] [FeC2-meimid)2. 0.13(FeCp2)]x [2] and [Fe(4-abimid)2]x (Chapter 3), which suggests a form of canted-spin antiferomagnetic coupling leading to weak feromagnetism at low temperatures, as discussed in Chapter 3, section 3.2.2. Weak 108 Figure 4.11 Magnetization versus applied field plots at diferent temperatures for compounds [Co(imid)2]x, (a); [Co(2-meimid)2]x, (b); [Co(4-meimid)2]x, (c); [Co(benzimid)2]x, (d); and [Co3(imid)6(imidH)2]x, (e). feromagnetism is also evident in the magnetization versus field plots at several temperatures for [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x (Figure 4.11). Plots of magnetization versus applied field for these compounds are linear above Tc and extrapolate to zero magnetization at zero applied field, while below Tc they show extrapolated net magnetization at zero field. In contrast, the data obtained at coresponding temperatures for [Co(2-meimid)2]x and [Co(4-meimid)2]x extrapolate to 109 zero magnetization at al temperatures studied (Figure 4.11). That [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x exhibit long-range feromagnetic order and spontaneous magnetization below Tc is further ilustrated by hysteresis studies. In these studies, the magnetization was measured as the applied field was cycled between +50 000 G and -50 000 G at 4.8 K. Preliminary hysteresis studies for these compounds showed evidence that the micro-crystals were aligning with the applied field, resulting in abnormal shapes of the hysteresis plots, as shown in Figure 4.12 (middle plot) for [Co(benzimid)2]x. These preliminary results prompted us to mul the sample in nujol to prevent the alignment of the micro-crystals with the applied field. The resulting hysteresis loops using nujol are shown in Figure 4.13. These loops give remnant magnetizations of 350, 280 and 200 cm3 G mol"1 and coercive fields of 5500, 2500 and 2000 G for [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x, respectively. A spin-canted structure, for [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(inidH)2]x is also supported by the fact that their highest magnetizations measured were 2924, 2253 and 2267 cm3Gmol"1, respectively, at 4.8 K and 55 000 G. These values are considerably lower than the theoretical saturation magnetization value of 16 766 cm 3Gmol"1 for an b = 3/2 system [15]. Magnetic parameters for five cobalt(I) 1,3-diazolate compounds, including [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x, which exhibit weak feromagnetism are given in Table 4.1. The only parameter which is relatively constant in this group of compounds is the critical temperature which lies in the narow range of 11 110 -60000 -40000 -20000 0 20000 40000 60000 2000 1000 J 0-s -1000 -60000 -40000-20000 0 20000 40000 60000 H(G) Figure 4.12 Magnetic hysteresis plots at 4.8 K for compounds [Co(imid)2]x, (top); [Co(benzimid)2]X5 (middle); [Co3(imid)6(imidH)2]x, (botom). Ill Table 4.1 Magnetic parameters for some cobalt(H) weak feromagnets Compound TC(K) Hcoer Ref. (G) (cnr'Gmor1) Co(4-abimid)2 11 400 22 Chapter 3 Co2(imid)4(bipy) 13 125 1900 Chapter 5 Co(imid)2 16 6 620 334 This Chapter Co(benzimid)2 13 5 280 257 This Chapter Co3(imid)6(imidH)2 15 4 140 175 This Chapter Abbreviations: imid = imidazolate, benzimid - benzimidazolate, 4-abimid = 4-azabenzimidazolate, bipy = 2,2'-bipyridine. to 16 K. Single crystal X-ray difraction studies on [Co(imid)2]x [3], and powder difraction studies on [Co(4-abimid)2]x (Chapter 3) show these compounds to have extended 3-D latices with tetrahedral cobalt centers linked via singly-bridging azolates to four nearest neighbors. In spite of the structural similarities, they have significantly diferent magnetic properties, [Co(imid)2]x being both a stronger magnet (larger remnant 112 magnetization) and a harder magnet (greater coercive field). The magnetic parameters of Co(4-abimid)2 have been discussed in some detail in Chapter 3. It is suficient to point out here that in Chapter 3 it was concluded that the properties of cobalt(I) molecule-based magnets may be influenced significantly by single ion efects such as zero-field spliting. Diferences in the magnitude of the zero-field spliting, in turn brought on by factors such as diferences in the nature and degree of distortion of the C0N4 chromophore, wil cause significant diferences in the populations of zero-field split levels at low temperatures with concomitant efects on the exchange and the magnetic properties. Unfortunately the number of compounds of this class with known structures is too smal at this time to try to corelate magnetic parameters with detailed structural features. In contrast to the behavior discussed above, [Co(2-meimid)2]x and [Co(4-meimid)2]x exhibit only short-range antiferomagnetic coupling. Plots of magnetic susceptibilty, %, and magnetic moment, p^ ff, versus T data in an applied field of 10 000 G are shown in Figure 4.7 and 4.8. A maximum at approximately 21 K is observed clearly in the magnetic susceptibilty plot for [Co(2-meimid)2]x. A paramagnetic impurity in this compound might be the cause of the smal increase in magnetic susceptibilty at the lowest temperatures studied (Figure 4.7). The maximum in the magnetic susceptibilty plot of [Co(4-meimid)2]x is less noticeable since it appears at a very low temperature of 4 K (Figure 4.7). Antiferomagnetic behavior for these compounds is also apparent in the Lieff versus T plots determined at 500 G (Figure 4.10). The magnetic moment decreases as 113 temperature is lowered. That these two polymers do not present a net magnetization at zero applied field is shown in their magnetization plots at two diferent temperatures which extrapolate to zero at zero applied field (Figure 4.11). In addition, magnetization studies at 4.8 K on these compounds, in which the applied field is cycled between +55 000 and - 55 000 G, reveal no evidence of significant hysteresis behavior (Figure 4.13). It is not possible to quantify the magnitude of the antiferomagnetic coupling in [Co(2-meimid)2]x and [Co(4-meimid)2]x due to the lack of a suitable model for S = 3/2 extended 3-D latice systems. 114 2000 H 1 0 0 0 1 O E P 0 S u -looo H i -60000 -40000 -20000 0 20000 40000 60000 Applied Field (G) o £ rj o E -2000 -3000 1 1 1 1 1 I 1 / . . . . -60000 -40000 -20000 0 20000 40000 60000 Applied Field (G) Figure 4.13 Magnetic hysteresis plots at 4.8 K for compounds [Co(2-meimid)2]x, (top); and [Co(4-meimid)2]x (botom). 115 4.3 4.3.1 A NICKEL(I) BENZLMLDAZOLATE POLYMER INTRODUCTION The magnetic properties of several 1-D nickel(I) pyrazolate polymers have been studied previously [16]. In those studies it was found that nickel(I) pyrazolates with a square planar chromophore geometry are diamagnetic, while those with tetrahedral geometry are paramagnetic and exhibit antiferomagnetic exchange coupling [17]. In contrast, few nickel(I) imidazolate polymers have been synthesized previously [6]. Those that have been made are square planar and diamagnetic [6]. None of the nickel(H) pyrazolate or imidazolate polymers reported has been obtained as macroscopic crystals suitable for single crystal X-ray difraction studies. In this section, a benzimidazolate complex of nickel(H) is examined. This microcrystaline material, [Ni(benzimid)2]x, was reported earlier [18] in a study which included some spectroscopic characterization. The magnetic properties of the compound were not studied in the earlier work. The magnetic studies presented here for [Ni(benzimid)2]x, show that this material behaves as a very weak feromagnet at low temperatures. Therefore, [Ni(benzimid)2]x is the first reported Ni(LI) imidazolate-based compound that behaves as a molecule-based magnet. 116 Atempts to prepare other magneticaly interesting Ni(I) imidazolate polymers, such as [Ni(imid)2]x and [Ni(4-meimid)2]x, were made, however, the materials obtained were found to be diamagnetic and for this reason were not investigated further. 4.3.2 RESULTS AND DISCUSSION 4.3.2.1 SYNTHESIS, STRUCTURAL, THERMAL AND PHYSICAL CHARACTERIZATION [Ni(benzimid)2]x was prepared, with some modifcations, folowing the method outlined by Goodgame and Coton [7]. In this method, a solution of benzimidazole in hot water was added to an aqueous solution of Ni(N03)2* 6H2O. The mixture was heated to boilng, and on cooling, a lavender precipitate formed. Details on this synthesis are given in Chapter 9, section 9.2.3.1. This complex is stable both in air and in contact with moisture. It is insoluble in water and most organic solvents, nonvolatile and thermaly robust. The UV-Vis-NLR spectrum of [Ni(benzimid)2]x is shown in Figure 4.14. The spectrum, which has not been previously analyzed, consists of three main absorptions in the ranges of 1680-1700 nm (barely observable), 755-780 nm and 500-550 nm, which can be tentatively assigned to the folowing transitions, 3Ti(F) -> 3T2,3Ti(F) -> 3A2, and 3Ti(F) -> 3Ti(P), respectively. There is another absorption in the range 835-860 nm, which has been assigned previously to the spin-forbidden transition 3Ti(F) -» !T2(D) [19]. These observable d-d 117 I—' 1 ' 1 1 1 ' — I 1 1 — ' 1 — 1 — I — 1 — I — ' — I 200 400 600 800 1000 1200 1400 1600 1800 2000 Wavelength (nm) Figure 4.14 UV-visible-near-LR spectrum for [Ni(benzimid)2]x- Insert plot shows the two highest energy d-d transition bands. transitions, three spin alowed and one spin forbidden, are expected for tetrahedral or distorted tetrahedral Ni(LT) [20]. Using the two highest energy bands (770 nm and 525 nm), and the coresponding Tanabe-Sugano corelation diagram [10] for a d8 tetrahedral system, Dq and B were calculated to be 711 cm"1 and 960 cm"1 respectively. In addition, strong charge transfer bands are seen in the region 200-400 nm. 118 The TGA plot for [Ni(benzirnid)2]x is shown in Figure 4.15. As can be seen there is no evidence of loss in mass due to thermal decompositon or sublimation at any temperature below ~ 450 °C. The TGA data suggest that two steps occur in the thermal decompositon of fNi(benzimid)2]x. The first one involves a rapid lost of nearly 50 % of the initial mass of the sample between 450 °C and 612 °C. The second event is more gradual with an additonal approximately 25 % of the initial mass being lost between 612 °C and 800 °C. 110 100 100 200 300 400 500 600 700 800 T(°C) Figure 4.15 TGA plot for [Ni(benzimid)2]x. 119 Hence, according to the electronic spectroscopic and thermal characterization of [Ni(benzimid)2]x, this compound shows the physical properties of a coordination polymer and it, most likely, possesses metal ions with a tetrahedral coordination geometry, similarly to those found in [Co(4-abimid)2]x (Chapter 3) or [Co(imid)2]x (vide supra). 4.3.2.2 MAGNETIC PROPERTIES Magnetic susceptibilty and magnetic moment versus temperature data on powdered samples of [Ni(benzimid)2]x in an applied field of 500 G are shown in Figure 4.16. The magnetic moment of [Ni(benzimid)2]x decreases from 2.53 ii.B at 300 K decreases to 1.35 [LB as the temperature is lowered to 2 K. This suggests antiferomagnetic coupling although confirmation of this, in the form of a maximum in the % versus T plot, is not seen (Figure 4.16). Furthermore, no evidence for long-range feromagnetic interaction as seen for the cobalt analogue, [Co(benzimid)2]x, is observed in this 500 G data. Nonetheless, the possibilty that this nickel compound has a structure similar to that of the Co analogue (which exhibits magnetic properties of a molecule-based magnet) prompted us to investigate the magnetic properties of the nickel compound ftrrther. Figure 4.17 shows the magnetization plots for [Ni(benzimid)2]x at diferent 120 as Figure 4.16 Plots of % and p^r versus T for ^ i(benzimid)2]x. temperatures. All the magnetization plots extrapolate to zero magnetization at zero field, except for the one determined at 4.8 K. In addition, except for the one determined at 4.8 K, al other magnetization plots in Figure 4.17 are linear. This result supports a possible long-range feromagnetic ordering at temperatures ~ 4.8 K and below for [Ni(benzimid)2]x-121 3500 3000 2500 H | 2000 -| o .5 1500 H IOOO H 500 0 • 4.8 K • 15 K • 25 K o 55 K J i - A o A o • A o 0 10000 20000 30000 40000 Applied Field (G) 50000 60000 Figure 4.17 Magnetization versus applied field plots at diferent temperatures for [Ni(benzimid)2]x-That [Ni(benzimid)2]x exhibits long-range feromagnetic order and spontaneous magnetization at ~ 4.8 K is further ilustrated by a hysteresis study. Magnetization was measured as the applied field was cycled between +55 000 G and -55 000 G at 2 and 4.8 K. The resulting hysteresis loop at 2 K is shown in Figure 4.18. This loop gives a remnant magnetization of ~ 7 cm3 G mol"1 and a coercive field of ~ 60 G for [Ni(benzimid)2]x which characterize this compound as a very weak and soft molecule-based magnet. 122 Figure 4.18 Magnetic hysteresis plot at 2 K for |>Ji(penzinid)2]x. The insert plot shows a magnifcation of the central part of the hysteresis curve. As was found for the cobalt analogue, [Co(benzimid)2]x, a spin-canted structure, for [Ni(benzimid)2]x, is also supported by the fact that the highest magnetization measured was 3089 cm3Gmor' at 4.8 K and 55 000 G. This value is considerably lower 123 than the theoretical saturation magnetization value o f 11 177 cm3Gmor1 expected for an S= 1 system [15]. To investigate further the possibilty of a feromagnetic ground state for [Ni(benzimid)2]x, DC magnetic susceptibilty measurements at an applied field of 50 G were caried out on a sample of this polymer as folows: the sample was cooled in zero field to 2 K, a magnetic field of 50 G was applied and data were colected while warming the sample (zero-field-cooled magnetization - ZFCM); then, the sample was cooled in the same field (50 G) to 2 K, and data were colected in the warming mode (field-cooled magnetization - FCM); finaly, the sample was cooled again to 2 K, in a field of 50 G, then the field is removed and data are colected while warming the sample (remnant magnetization - REM). The results of this data colection scheme are shown in Figure 4.19. The ZFCM data increases gradualy to a maximum at ~ 6.5 K, then decreases slightly to increase again to a second maximum at ~ 2.5 K. The Tc value, determined as the first maximum on the ZFCM plot, is ~ 6.5 K,. When the applied field was switched of at 2 K a smal remnant magnetization of ~ 0.95 cm3Gmol"1 was found (in good agreement with the value of Mrem ~ 1 cm3Gmol"1 obtained in the hysteresis study shown above). This Mrem decreased significantly on warming to 2.5 K then decreased further upon warming and vanished at ~ 5.5 K. [Ni(benzimid)2]x exhibits another transition at ~ 2.5 K. The origin of this lower-temperature transition is not clear. A similar double transition has been recently observed in the ZFCM-FCM-REM studies in a 1-D 124 molecule-based ferimagnet having Cu(I) and Mn(LI) ions [20]. A possible source of this phenomenon was not discussed in this earlier report. From these results it is evident that [Ni(penzimid)2]x exhibits long-range feromagnetic order at low temperatures, and that this material can be considered a molecule-based magnet. 4.5 4.0 3.5-3.0 H O 2.5 • s P 2.0 -\ B & 1.5 1.0-0.5-0.0 • ZFCM o F C M £ REM * * JL. I 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -2 4 6 8 10 12 14 16 18 20 T(K) Figure 4.19 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Ni(penzimid)2]x at 50 G. 125 4.4 COPPER(I) IMIDAZOLATE POLYMERS 4.4.1 INTRODUCTION Imidazolate-bridged copper(I) dimeric and oligomeric complexes have been actively studied [22-23] mainly to understand the factors determining the extent of coupling between two metal ions and to use these simple compounds as models for metaloenzymes that contain the same structural units [24, 25]. Several studies have already been reported in which structural data have been used to find useful corelations between structure and magnetic coupling [26-30]. On the contrary, structural and magnetic properties of imidazolate-bridged copper(I) polymers have not been extensively studied. A blue form of [Cu(imid)2]x is the only compound of this family with a known molecular structure, and its magnetic properties have also been determined but only at temperatures between 80 - 300 K [31, 32]. In the present section, the synthesis, characterization, and low-temperature magnetic studies of [Cu(imid)2]x and four other Cu(I) systems incorporating substiuted imidazolate ligands: [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(benzimid)2]x, and [Cu(4,5-dichloroimid)2]x, are discussed. None of the materials studied here was isolated in a form suitable for single crystal X-ray difraction studies. Moreover, no definitive details on the structures of these compounds were obtained as neither the electronic spectra nor the X-ray powder difractograms were particularly informative for these systems. Nonetheless, 126 as with most of the other compounds studied in this work, the compounds are considered to be polymeric based on solubilty and thermal gravimetric studies. With the exclusion of [Cu(imid)2]x and [Cu(4-meimid)2]x, which showed only weak antiferomagnetic coupling, the other Cu(LI) imidazolate polymers exhibited magnetic properties that classify them as weak low-temperature molecule-based magnets. 4.4.2 RESULTS AND DISCUSSION 4.4.2.1 SYNTHESES, STRUCTURAL, THERMAL AND PHYSICAL CHARACTERIZATION Detailed descriptions of the syntheses of the copper(n) imidazolate complexes studied here, can be found in Chapter 9, sections 9.2.4.1 through 9.2.4.5. The molten ligand-copper shot method, which has been successfuly used to obtain single crystals of binary copper(I) pyrazolates [33], was first tried in an atempt to obtain single crystals of the compounds; however, al atempts to produce macroscopic crystals were unsuccessful. Ultimately, the only compound prepared using this method was [Cu(imid)2]x, which was obtained as a dark blue powder. The other Cu(LI) imidazolate compounds were prepared by wet methods which involved the use of copper shot with an ethanolic solution of the appropriate ligand or, the reaction of an appropriate salt of Cu(I) with the appropriate imidazolate in water. These synthetic procedures are modifcations of previously reported methods [7, 34]. 127 The five copper(I) compounds, [Cu(imid)2]x, [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x, are stable both in air and in contact with moisture. They are insoluble in water and common organic solvents, nonvolatile and thermaly robust. They decompose when treated with concentrated mineral acids. Thermal gravimetric analyses of the five copper compounds show no mass loss due to thermal decompositon or sublimation below 180 °C. At higher temperatures the compounds present similar thermal decompositon behavior, showing a weight loss in three stages over the temperature range studied (Figure 4.20). [Cu(benzimid)2]x is the most thermaly robust of the five in that there is no indication of thermal decompositon below 320 °C. In general, the Cu(I) compounds are less thermaly stable than the Co(I) imidazolates reported on above. The absence of a V N - H stretching vibration in the electronic spectra (vide infra) and in the infrared spectra of these complexes (See the discussion above concerning the observation of this band in [Co3(imid)6(imidH)2]x) is indicative of the absence of neutral imidazole in these materials. 128 it 5<H £ 40H 30-20-10-i 100 200 [CuCimid)^  [Cu(2-meimid)z]i [Cu(4-meimid)2]i [Cu(benzimid)2]i [Cu(4,5-dichloroimid)2]i 300 400 — I — 500 I 1 1— 600 700 800 Temperature (°C) Figure 4.20 TGA plots for compounds [Cu(imid)2]x, [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dicMoroimid)2]x. The electronic spectra of the copper derivatives are shown in Figure 4.21 and a summary is presented in Table 4.2. In the NIR region, a band maximum is observed in the range 800 to 950 nm for [Cu(imid)2]x, [Cu(4,5-dichloroimid)2]x, [Cu(benzimid)2]x and, less clear, for [Cu(4-meimid)2]x. All five compounds show bands around 550 - 600 129 200 400 600 800 1000 1200 1400 W a v e l e n g t h (nm) Figure 4.21 UV-Vis-NLR spectra for [Cu(imid)2]x, (a); [Cu(2-meimid)2]x, (b); [Cu(4-meimid)2]x, (c); [Cu(benzimid)2]x, (d); and [Cu(4,5-dichloroimid)2]x, (e). 130 nm and 400 - 500 nm in the visible region although for [Cu(4,5-dichloroimid)2]x and [Cu(benzimid)2]x the later bands are somewhat obscured by intense bands in the UV region. All five compounds show bands assignable to charge-transfer in the 200 - 400 nm region. Table 4.2 UV-Vis-NLR spectra of copper(I) imidazolates. Approximate wavelength values or regions (nm) Compound 200-400 nm 400-500 nm 1 550-600 nm 1 800-950 nm (a) imid - 200 - 300 -400 - 600 (split) -910 (b) 2-meimid ~ 200 - 380 ~ 400 and 490 -600 None (c) 4-meimid ~ 200 - 250 -400 600 (split) - 850 - 950 (d) benzimid - 200 - 400 ? -570 -890 (e) 4,5-diclimid ~ 200 - 290 ? -560 -800 According to Hathaway [35], complexes with CuN4 chromophores that exhibit d-d bands in the 500 to 550 nm region are likely to have square-planar stereochemistries, 131 while bands in the 625 to 850 nm region are characteristic of compressed tetrahedral Q1N4 chromophores. These criteria for C11N4 geometry have been used for related copper(I) pyrazolates [36] and pyrazolyl galate systems [37]. As can be seen in Figure 4.21 and Table 4.2, the compounds studied here exhibit bands in wavelength regions coresponding to both chromophores. A mixture of two or more chromophore geometries is possible for these systems. As described above, one form of [Cu(imid)2]x has been shown by single crystal X-ray difraction studies to have both square-planar and distorted tetrahedral C U N 4 chromophores. As previously mentioned, the crystal structure of a blue form of [Cu(imid)2]x is known. Three magneticaly diferent crystal modifcations have been found for [Cu(imid)2]x [31, 32], a blue modifcation (p^f  = 1.57 u,B at 293 K ) , a green modifcation (Meff = 1-62 p.B at 303 K ) and a brown modifcation ([igs = 1.46 jxB at 303 K ) . The [Cu(imid)2]x synthesized in the present work has a dark blue color, with a |ieff = 1.54 p:B at 300 K . The color of our compound and its high temperature magnetic properties suggested to us initialy that we had prepared the blue form of the compound. The blue modifcation is the only one that has been studied by single crystal X-ray difraction [32]. In this structure, imidazolato groups bridge Cu(I) ions to form chains, in which there is a systematic alternation of a Cu(l) ion with square planar coordination, folowed by another Cu(I) ion in a flatened tetrahedral coordination. The chains are linked at the flatened tetrahedral Cu(I) ion so that the whole assembly forms a three-dimensional network [32]. A representation of the repeat unit of blue-[Cu(imid)2]x, and a stereoview 132 showing the 3-D structure of this polymer are shown in Figures 4.22 and 4.23, respectively. These views were achieved employing the software Powdercel [8] and using the crystalographic data previously reported [32]. Figure 4.22 Repeat unit of blue-[Cu(imid)2]x. Hydrogen atoms are omited. 133 Figure 4.23 Stereoview of a section of blue-[Cu(imid)2]x including the unit cell. Projection (001). No hydrogen atoms shown. As shown in Figure 4.24, the experimental X-ray powder diffraction pattern of the sample of [Cu(imid)2]x prepared in this thesis work does not coincide well with that calculated [8], employing single crystal X-ray diffraction data reported for the blue-[Cu(imid)2]x [32]. This implies that the [Cu(imid)2]x synthesized in this work is not isomorphous with blue-[Cu(imid)2]x [32]. The detailed structure of our compound remains unknown. 134 i 5 10 15 20 25 30 35 40 45 50 u ^ ^ ^ ^ 5 10 15 20 25 30 35 40 45 50 29 (deg) Figure 4.24 X-ray powder difractograms of blue-[Cu(imid)2]x (top, calculated) and [Cu(imid)2]x prepared here (botom, experimental). The X-ray powder difraction paterns for [Cu(2-meimid)2]x, [Cu(4-meimid)2]x, [Cu(4,5-dichloroimid)2]x and [Cu(benzimid)2]x, are shown in Figure 4.25. Each patern is unique showing there is no isomorphism associated with the diferent copper systems. Interestingly, the paterns for the imid, 2-meimid and benzimid copper compounds are also diferent from those of the coresponding cobalt compounds (see Figures 4,4 and 4.5). In these studies we have found no examples of isomorphous pairs of copper(n) and 135 (a) (b) u - J 10 20 30 40 50 10 20 30 40 50 26 26 (c) | 10 i • i 20 30 40 50 10 20 30 40 50 26 26 Figure 4.25 X-ray powder difraction paterns of [Cu(2-meimid)2]x, (a); [Cu(4-meimid)2]x, (b); [Cu(benzimid)2]x, (c); and [Cu(4,5-dicWoroimid)2]x, (d). cobalt(I) imidazolates. This contrasts sharply with the situation for iron(I) and cobalt(I) where we have discovered three examples of isomorphous pairs. 136 The structures of the copper(I) imidazolates studied here while likely involving 3-D connectivites of the type seen in blue-[Cu(imid)2]x [32], appear to be somewhat unique. This may be caused by the presence of two or more diferent chromophores in the latice as indicated by the electronic spectroscopy studies. At this point it is not possible to make a more conclusive statement regarding structures. 4.4.3 MAGNETIC PROPERTIES Magnetic susceptibilties were measured at 10 000 G from 2 to 300 K for al five Cu(I) imidazolate compounds. Magnetic susceptibilty, %, and magnetic moment, \i^s, versus T (2 to 100 K) data on powdered samples are shown in Figure 4.26 and 4.27, respectively. All five compounds show broad maxima in their % versus T plots over the 25 -150 degree range indicating the presence of antiferomagnetic exchange. This is confirmed by the fact al five compounds show decreasing magnetic moments with decreasing temperature over this range. All five of the compounds exhibit an increase in X at lower temperatures. The p^f versus T plot, in particular at low temperatures, clearly shows distinctive behavior for the five compounds. [Cu(4-meimid)2]x shows a decrease in 137 0.020 0.015 o •5 o.oio s 61 0.005 0.000 • V V o [Cu(imid)2]x V • [Cu(2-meimjd)2]x V • [Cu(4-meimid)2]x V v [Cu(benzimid)2]x V O [Cu(4,5-dichloroimid)2]x V v V V ^ V V V v » 6 6 v o o 25 50 T ( K ) 75 I I 100 Figure 4.26 x v e r s u s T Plots a t 10 000 G for [Cu(imid)2]x, [Cu(2-meimid)2]x [Cu(4-meimid)2]x, [Cu(berizimid)2]x and [Cu(4,5-dicMoroimid)2]x. LL-f with decreasing temperature down to 2 K, consisted with antiferomagnetic exchange over the entire range studied. This is confirmed by the % data obtained for this compound employing an applied field of 500 G. % shows a broad maximum at ~ 75 K (Figure 4.28). The increase in at the lowest temperatures studied seen in both the 10 000 and 500 G 138 3 1.75 1.50 1.25 1.00 is 6 * 0.75 0.50 0.25 0.00 V s o o o v o • * o • [Cu(imid)2]x [Cu(2-meimid)2]x [Cu(4-meimid)2]x [Cu(benzimid)2]x [Cu(4,5-dicWorimid)2]x 25 50 75 T(K) 100 125 150 Figure 4.27 LUff versus T plots at 10 000 G for [Cu(imid)2]x, [Cu(2-meirnid)2]x, [Cu(4-meirnid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x. data is not large and likely arises from the presence of paramagnetic impurities. The magnetic behavior of [Cu(imid)2]x is similar to that of [Cu(4-meimid)2]x with the exception that the "paramagnetic tail" in the % plot is much more pronounced for this compound (Figure 4.26). As a result the p<.ff versus T plot shows a clear tendency to level 139 0.0045 0.0020 100 150 200 250 300 T (K) Figure 4.28 Plot of % versus T for [Cu(4-meimid)2]x at 500 G. of (smaler slope) below ~ 25 K. This suggests: (i) either a higher level of paramagnetic impurity in this material or, (ii) more complex behavior with a decrease in the strength of the antiferomagnetic coupling as the temperature is lowered or, (iii) a change in the magnetic exchange mechanism from primarily antiferomagnetic to primarily feromagnetic coupling. 140 The other three compounds give clear evidence of a transition from antiferomagnetic to feromagnetic exchange at low temperatures. This is clearly the case for [Cu(benzimid)2]x, which shows a magnetic anomaly at ~ 15 K below which Lieff increases on decreasing the temperature before decreasing again below 8 K. Similar anomalies are less pronounced for [Cu(4,5-dichloroimid)2]x and [Cu(2-meimid)2]x where the LL-ff values are seen just to level of  on decreasing the temperature before decreasing again as the temperature is lowered (Figure 4.27). To test for the presence (or absence) of long-range feromagnetic order in these five copper systems we decided to undertake field-cooled (FC) and zero-field-cooled (ZFC) DC susceptibilty measurements at a relatively low applied field of 50 G. The protocol for this consists of cooling the sample at zero field to 2.0 K, then, applying a magnetic field (50 G) the data are colected while the sample warms (ZFC data); the sample is then cooled in the field of 50 G to 2.0 K, and data are colected while warming the sample (FC data). The results of this data colection scheme are shown in Figures 4.29, 4.30, 4.31, 4.32, and 4.33, for [Cu(benzimid)2]x, [Cu(2-meimid)2]x, [Cu(4,5-dichloroimid)2]x, [Cu(imid)2]x, and [Cu(4-meimid)2]x, respectively. In a compound exhibitng long-range feromagnetic ordering, the temperature-dependence of magnetization curves wil show, generaly, a break in the field-cooled magnetization (FCM) curve and a peak for the zero-field-cooled magnetization (ZFCM) curve at the onset of the magnetic transition (Tc). Also, the values of the ZFCM are 141 always lower than the FCM values at temperatures below the magnetic transition. Hence, these plots can provide one of the most accurate ways to determine the critical temperature (Tc) [38]. 16-• 14-ZFCM a FCM 12-10- o 8 8 - O 0 6- o 4- o ° a o 2- a ^tato^r, 0-1 1 1 1 1 1 1 1 1 1 0 5 10 15 20 25 T(K) Figure 4.29 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(benzimid)2]x at 50 G. 142 0.44 H 0 . 0 8 4 — | — i — | — i — | — i — | — i — j — i — | — i — i — i — i — i — | 0 5 10 15 20 25 30 35 40 T(K) Figure 4.30 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(2-meimid)2]x at 50 G. As can be seen in Figure 4.29, [Cu(benzimid)2]x exhibits the FCM and ZFCM characteristics of a material showing long-range feromagnetic order. The critical temperature (Tc) for this compound is determined to be ~ 8 K. Comparing the behavior of [Cu(2-meimid)2]x (Figure 4.30) and [Cu(4,5-dicMoroimid)2]x (Figure 4.31) to that of [Cu(benzimid)2]x, it appears that these two polymers also exhibit long-range feromagnetic order at low temperatures. Employing the maximum in the ZFCM plot as the measure of the critical temperature generates Tc values of~ 15Kand- 14Kfor 143 0.23 -I 0.22- • o ZFCM 0.21-• • FCM ° • "o 0.20- o • £ O r| 0.19- O °P i S 0 S 0.18- • • g 0.17-0.16-0.15- 1 1 ' — 1 ' 1 ' I 1 - 1 " i 0 5 10 15 20 25 T(K) Figure 4.31 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(4,5-dichloroimid)2]x at 50 G. [Cu(2-meimid)2]x and [Cu(4,5-dichloroimid)2]x, respectively. [Cu(2-meimid)2]x shows a second maximum in its ZFCM plot at ~ 9 K. It is not at al clear what the origin is of this lower-temperature transition. A similar double transition appears in the ZFCM-FCM-REM studies caried out on [Ni(benzimid)2]x and described earlier in this Chapter (section 4.3.2.2). Also, as mentioned earlier, a similar double transition has been recently reported in a 1-D molecule-based ferimagnet involving Cu(H) and Mn(I) ions [20]. 144 In contrast to the behavior just described, the studies on [Cu(irnid)2]x and [Cu(4-meimid)2]x do not reveal peaks or large discontinuites in their ZFCM or FCM plots (Figures 4.32 and 4.33). For [Cu(imid)2]x the ZFCM and FCM plots show only smal discontinuites at around 15 K, the temperature at which the values of the FCM and ZFCM are the same folowing a warming mode (Figure 4.32). Interestingly, this compound shows a second minor anomaly at around 10 K. Similarly, the magnetization T (K) Figure 4.32 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(imid)2]x at 50 G. 145 curves of [Cu(4-meimid)2]x have smal breaks at around 15 K (Figure 4.33). In conclusion, [Cu(imid)2]x and [Cu(4-meimid)2]x do not give evidence for long-range feromagnetic order even in applied fields as low as 50 G. This is consistent with the conventional DC magnetization studies done at 500 and 10 000 G. o S o in S 0.34-0.32-J • 0.30-J o 0.28-• 0.26; 0.24- • 0.22-0.20; • 0.18- ° o c 0.16- o - 0 0.14- o 0.12-0.10-0.08-1 1 1 0 5 o Z F C M • F C M i 10 15 20 i 25 -r— 30 35 40 T ( K ) Figure 4.33 Zero-field cooling (ZFC) and field-cooling (FC) magnetization plots for [Cu(4-meimid)2]x at 50 G. 146 As described in earlier sections of this Chapter, hysteresis studies may be used to confirm the presence of long-range feromagnetic order. Accordingly we examined the magnetization of al five copper compounds as the applied field was cycled between +55 000 G and -55 000 G at 4.8 K. The resulting hysteresis loops are shown in Figure 4.34. These loops give remnant magnetizations of 0.25, 6.0 and 0.02 cm3 G mol"1 and coercive fields of 65, 45 and 4 G for [Cu(2-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x, respectively, which characterize these compounds as very soft molecule-based magnets. In contrast, no wel defined magnetic hysteresis was found, at 4.8 K, for [Cu(imid)2]x and [Cu(4-meimid)2]x, as shown in Figure 4.35. Thus, these two copper compounds cannot be regarded as molecule-based magnets. A spin-canted structure, leading to residual spin at low temperatures, for [Cu(2-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x, is also supported by the fact that their highest magnetizations measured were 80, 347 and 251 cm3Gmol"1, respectively, at 4.8 K and 55 000 G. These values are considerably lower than the theoretical saturation magnetization value of 5588 cm3Gmor1 for an S = 1/2 system [15]. 147 300 0 -100 -200 -300 400 C 2 0 0 "3 £ P o E CD *s -200 J v i 1 o 1 f •40000 -20000 0 20000 40000 o -e-° o O ° ' ° ° -40000 -20000 0 20000 40000 -40000 -20000 0 20000 40000 Applied Field (G) Figure 4.34 Magnetic hysteresis plots at 4.8 K for [Cu(4,5 dicloroirnid)2]x, (top); [Cu(benzimid)2]x, (middle); and [Cu(2-meimid)2]x, (botom). The insert plots show magnifcations of the central part of the hysteresis curves. 148 o s o m E -60000 -40000 -20000 0 20000 40000 60000 Applied Field (G) 200 -I 150 -100 J". "3 50 E 0 w 0 -E -50 --100 -150 --200 -ii © © @ 8 0 @ r- 1 0 © • © -60000 -40000 -20000 0 20000 40000 60000 Applied Field (G) Figure 4.35 Magnetic hysteresis plots at 4.8 K for [Cu(imid)2]x, (top), and [Cu(4-meimid)2]x, (botom). The insert plots show magnifcations of the central part of the hysteresis curves. 149 The magnetic behaviors of [Cu(2-meimid)2]x, [Cu(benzimid)2]x and [Cu(4,5-dichloroimid)2]x, indicate in al three compounds the presence of antiferomagnetic coupling between paramagnetic centers as the primary exchange mechanism combined with a magnetic phase transition to a feromagneticaly ordered state at low temperatures. Therefore, these three copper compounds can be considered as molecule-based magnets. Again, this magnetic behavior is very comparable to that reported for [Fe3(imid)6(imidH)2]x [1], [Fe(2-meimid)2.0.13(FeCp2)]x [2] and [Fe(4-abimid)2]x (Chapter 3), and the one observed previously in this Chapter for two Co(I) imidazolates and for the Co(I) and Ni(I) benzimidazolates. 4.5 SUMMARY AND CONCLUSIONS Earlier studies, mainly on compounds of iron(I), led to the suggestion [1,2] that azolate ligands which bridge through nitrogen atoms separated by one carbon in the heterocyclic ring wil generate structures with single azolate bridges and extended arays. The work described in this Chapter indicates that, even though no new structures were determined by single crystal X-ray difraction, this structural motif extends beyond compounds of iron(I) to include those of cobalt(I), nickel(LI) and copperfl). Moreover, al of the systems studied in this Chapter show antiferomagnetic exchange coupling, mediated by the bridging imidazolate ligands, as the primary exchange process, and many, although not all, of the systems give evidence for a magnetic transition to long-150 range feromagnetic order at low temperatures. It appears that the phenomenon of spin-canting, a consequence of the non-cenlxosymmetric M-L-M exchange pathway provided by the single-bridging imidazolate ligands does indeed extend to metal systems other than iron(LT). It is not clear at this point why some, but not all, of the binary imidazolates of coppertjl) and cobalt(I) show weak feromagnetism. There appears to be no obvious ligand characteristic determining this. For example, while there is evidence for feromagnetic order in the binary benzimidazolates of both cobalt and copper (nickel too, albeit very weak), the binary imidazolate of cobalt(I) shows low temperature order while that of copper(I) does not. In contrast, the 2-methylimidazolate of copper shows order and that of cobalt does not, while neither the cobalt nor the copper 4-methylimidazolate shows order. It seems that the presence or absence of measurable long-range feromagnetic order in imidazolate systems depends on structural details which in turn afect factors such as the degree of spin-canting. Less important is the dn configuration of the metal center (other than that it be a paramagnetic configuration). 151 References 1. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter. /. Am. Chem. Soc. 119, 8675 (1997). 2. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter. Can. J. Chem. 77, 425 (1999). 3. V. M. Sturm, F. Brandl, D. Engel, W. Hoppe. Acta Cryst., B31, 2369 (1975). 4. G. P. Brown, and S. Aftergut. J. Polymer Sci., A2, 1839 (1964). 5. F. Seel, and J. Rodrian. J. Organomet. Chem., 16, 479 (1969). 6. A. M. Vecchio-Sadus. Trans. Met. Chem., 20,46 (1995). 7. M. Goodgame, and F. A. Coton. J. Am. Chem. Soc. 84, 1543 (1962). 8. PowderCell, version 2.3, W. Krauss and G. Nolze, Federal Instiute for Materials Research and Testing (BAM), Berlin, 1997. 9. F. A. Coton, and G. Wilkinson. Advanced Inorganic Chemistry. Fourth Edition. John Wiley & Sons. New York. 1980. p. 770 10. Y. Tanabe and S. Sugano. J. Phys. Soc. Jpn. 9, 753 (1954). 11. M. K. Ehlert, A. Storr, and R. C. Thompson. Can. J. Chem. 71,1412 (1993). 12. W. J. Eilbeck, F. Holmes, C. E. Taylor, and A. E. Underhil. J. Chem. Soc. (A), 128 (1968). 13. MGP-Suite of Programs for Interpretation ofX-Ray Experiments, by Jean Laugiert and Bernard Bochu, ENSP/Laboratoire des Materiaux et du Genie Physique. BP 46. 38042 Saint Martin d'Heres, France. htp:/www.inpg.fr/LMGP and htpp:/www.ccpl4.ac.uk/tutorial/mgp/. 14. F. A. Coton, and G. Wilkinson. Advanced Inorganic Chemistry. Fourth Edition. John Wiley & Sons. New York. 1980. p. 771. 15. R. L. Carlin. Magnetochemistry. Springer-Verlag. Berlin. 1986. pp. 7-9. 16. A. Storr, D. A. Summers, and R. C. Thompson. Can. J. Chem. 76, 1130 (1998). 152 17. D. A. Summers. Ph. D. Thesis. The University of British Columbia. 1997. 18. M. M. Cordes and J. L. Walter. Spectrochim. Acta. 24, 1421 (1968). 19. F. A. Coton, and G. Wilkinson. Advanced Inorganic Chemistry. Fourth Edition. John Wiley & Sons. New York. 1980. p. 789. 20. N. Fukita, M. Ohba, T. Shiga, H. Okawa, and Y. Ajiro. J. Chem. Soc, Dalton. Trans. 64 (2001). 21. J. C. Dewan and S. J. Lippard. Inorg. Chem. 19, 2079 (1980). 22. H. M. J. Hendricks, P. J. M. W. L. Birker, G. C. Vershoor, and J. Reedijk. J. Chem. Soc, Dalton Trans. 623 (1982). 23. C. Beneli, R. K. Bunting, D. Gateschi, and C. Zanchini. Inorg. Chem. 23, 3074 (1984). 24. J. A. Ibers and R. H. Holmes. Science. 209, 223 (1980). 25. E. Colacio, J. M. Dominguez-Vera, M. Ghazi, R. Kivekas, M. Klinga, and J. M. Moreno. Inorg. Chem. 37, 3040 (1998). 26. J. T. Landrum, C. A. Reed, K. Hatano, and W. R. Scheidt. J. Am. Chem. Soc. 100, 3232 (1978). 27. M. S. Haddad and D. N. Hendrickson. Inorg. Chem. 17, 2636 (1978). 28. C. L. O'Young, J. C. Dewan, H. R. Lilenthal, and S. J. Lippard. J. Am. Chem. Soc. 100, 7291 (1978). 29. G. Kolks, C. R. Frihart, P. K. Coughlin, and S. J. Lippard. Inorg. Chem. 20,2933 (1981). 30. G. Kolks, S. J. Lippard, J. V. Waszczak, and H. R. Lilienthal. J. Am. Chem. Soc. 104, 717 (1982). 31. M. Inoue, M. Kishita, and M. Kubo. Inorg. Chem. 4, 626 (1965). 32. H. C. Freeman. Advan. Protein Chem. 27, 257 (1967). 33. M. K. Ehlert, S. J. Retig, A. Storr, R. C. Thompson, and J. Troter. Can. J. Chem. 67, 1970(1989). 153 34. G. P. Brown, S. Aftergut. J. Polymer Sci. A2, 1839 (1964). 35. B. J. Hathaway. J. Chem. Soc. Dalton Trans. 1196 (1972). 36. M. K. Ehlert, A. Storr, and R. C. Thompson. Can. J. Chem. 70,1121 (1992). 37. F. G. Herring, D. J. Patmore, and A. Storr. J. Chem. Soc. Dalton Trans. 711 (1975). 38. M. G. F. Vaz, L. M. M. Pinheiro, H. O. Stumpf, A. F. C. Alcantara, S. Golhen, L. Ouahab, O. Cador, C. Mathoniere, and O. Kahn. Chem. Eur. J. 5, 1486 (1999). 154 Chapter 5 TWO-DIMENSIONAL IRON(II) AND COBALT(II) IMIDAZOLATE POL YMERS EXHIBITING LONG-RANGE FERROMAGNETIC ORDERING 5.1 INTRODUCTION It has already been suggested that steric constraints imposed by the 1,3 positoning of the nitrogens prevent double imidazolate bridging between metal centers [1]. One important consequence of the single azolate bridging in metal imidazolates has been the generation of extended structures with 3-D covalent connectivites (see Chapter 3). Another important characteristic property, not seen in coresponding pyrazolates (see Chapter 2), is that metal imidazolate polymers exhibit antiferomagnetic coupling above a critical temperature and long-range feromagnetic ordering below that temperature, behavior that characterizes them as low temperature molecule-based magnets [ 1 -3 ]. The work described in this Chapter achieves an important objective in the study of transition metal azolate polymers. It explores what efect significantly altering the extended structures of metal imidazolate systems would have on their magnetic properties. In the work described in this section, major structural modifcation has been achieved by incorporating 2,2'-bipyridine as a capping ligand in the compound 155 poly-2,2'-bipyridinetetrakis(imidazolato)diron(I), [Fe2(imid)4(bipy)]x. This material has a unique double layer 2-D extended latice. Like the iron(I) 4-azabenzimidazolate system, [Fe(4-abimid)2)]x, (Chapter 3) it incorporates single imidazolate bridges and, moreover, it also exhibits long range feromagnetic order and spontaneous magnetization at low temperatures. To see whether a similar structural modifcation is possible with other metals, the 2,2'-dipyridine complex of cobalt (I) imidazolate analogue was also investigated. Poly-2,2'-bipyridinetetrakis(imidazolato)dicobalt(I), [Co2(imid)4(bipy)]x, was synthesized and was found to be isomorphous with [Fe2(imid)4(bipy)]x. An investigation of the magnetic properties of the cobalt compound revealed that it too exhibits long range feromagnetic order at low temperatures. The polymer [Fe2(imid)4(bipy)]x is unique in that it exhibits two reversible structural phase transitons over the temperature range 2 to 300 K. These phase transitions, one of which exhibits thermal hysteresis, have been studied by both DC and AC susceptibilty measurements, in additon to Mossbauer spectroscopy and X-ray crystalography. Part of the material discussed in this chapter is curently in press. 5.2 POLY-2,2,-BIPYRIDINETETRAKIS(IMIDAZOLATO)DIRON(I) 5.2.1 RESULTS AND DISCUSSION 156 5.2.1.1 SYNTHESIS, PHYSICAL AND THERMAL CHARACTERIZATION The reaction of ferocene with imidazole and excess molten 2,2'-bipyridine produced the polymeric material, [Fe2(imid)4(bipy)]x, in macroscopic crystaline form, suitable for single crystal X-ray difraction studies. Details of the synthesis of this compound are given in Chapter 9, section 9.2.1.2. [Fe2(imid)4(bipy)]x is moderately air-sensitive, it does not dissolve in common organic solvents and it is non-volatile. This later characteristic is further supported by the thermal gravimetric analysis of [Fe2(imid)4(bipy)]x. As shown in Figure 5.1, [Fe2(imid)4(bipy)]x does not start to decompose until ~ 250 °C. The TGA plot folows a two stage weight loss. The first major weight loss (~ 40 %) occurs between 250 °C and 335 °C, and the second one, with a weight loss of ~ 25 % of the initial weight, happens between 335 °C and 580 °C. No further weight loss of any signifcance occurs up to the highest temperature studied of 800 °C. The calculated weight loss for the dissociation of 2,2'-bipyridine is 29 %. However, the TGA plot does not show the expected plateau, coresponding to such an event which would lead to the formation of polybis(imidazolate)iron(I), [Fe(imid)2]x. It seems that this iron compound is very thermaly unstable. This contradicts with the thermal properties of the analogue, [Co(imid)2]x, which can be obtained under the TGA analyser conditons from thermolysis of [Co3(imid)6(imidH)2]x (see Chapter 4, section 4.2). 157 100 200 300 400 500 600 700 800 T(°C) Figure 5.1 TGA plot for [Fe2(imid)4(bipy)]x. 5.2.1.2 X-RAY DIFFRACTION STUDIES Room temperature (294 K) X-ray difraction studies revealed an extended structure involving double layer sheets of iron ions linked by single imidazolate bridges. As mentioned in the introduction, we have identified three diferent structural phases in this material and have labeled this high temperature phase the a-phase. Crystalographic data for [Fe2(imid)4(bipy)]x are shown in Appendix I, Table 1-5. The repeat unit of the a-phase of [Fe2(imid)4(bipy)]x is shown in Figure 5.2. 158 (71 Figure 5.2 View of the repeat unit of [Fe2(imid)4(bipy)]x (a-phase, 294 K) and atom numbering scheme (33% probabilty thermal elipsoids). The capping of iron centres by 2,2'-bipyridine afects the dimensionality of this system, resulting in a 2-D polymer, as seen in Figures 5.3 and 5.4. Four- and six-coordinated Fe(I) ions alternate in the lattice, the later ions being coordinated by bipy 159 ligands in additon to bridging imidazolates. Each tetrahedraly coordinated iron is bonded via the ligand to four octahedral irons with each of these bonded to four tetrahedral ones. The capping bipy ligands prevent bridging of metal centres in the third dimension and occupy space between the sheets, isolating the sheets from each other. By looking at the iron ion connectivity diagram (Figure 5.4), it can be seen that the connectivites between the two layers of a sheet form four-membered fused rings, while, the connectivites within the layers, top and botom, form six-membered fused rings. When DC magnetic susceptibilty measurements were performed on [Fe2(imid)4(bipy)]x, a major discontinuity was detected (vide infra) in the magnetic moment of this compound at around 135 K. Another discontinuity was also barely apparent at ~ 150 K. The later was further confirmed by higher density DC susceptibilty measurements (vide infra). These findings prompted the determination of the single-crystal X-ray structure of [Fe2(imid)4(bipy)]x at low temperatures to determine whether crystalographic phase transitons were involved. Subsequently, single crystal X-ray difraction studies of [Fe2(imid)4(bipy)]x at ~ 143 K and ~ 113 K revealed two new structural phases. These structural studies were caried out on diferent crystals of [Fe2(imid)4(bipy)]x; therefore, in order to eliminate the possibilty that these three diferent polymorphs were isolated from the room temperature synthesis, the unit cel of 160 Figure 5.3 ORTEP diagrams of [Fe2(imid)4(bipy)]x (a-phase) looking down the c axis. In the botom view, bipyridine ligands have been removed to reveal the double-layer sheet extended framework. (50 % probabilty thermal elipsoids). 161 Figure 5.4 Iron ion connectivity diagram of a section of two double-layer sheets for the a-phase of [Fe2(imid)4(bipy)]x. Octahedral iron (red), tetrahedral iron (green). View approximately looking down the c axis. the crystal used for the lowest temperature (113 K) study was determined initialy at 173 K. The unit cel at 173 K was found to be the same as that obtained at 294 K [a = 10.507(4), b = 13.730(4), c = 9.188(3) A, a = 106.51(3), fj = 108.32(3), y= 80.84(3) deg, V= 1202.9(2), AJ]. When the same crystal was cooled down from 173 to 113 K the unit cel parameters changed [a = 10.414(5), b = 13.508(5), c = 26.060(1) A, a = 104.53(2), P = 93.892(2), y= 100.512(2) deg, V= 3646.0(2), A3]. The phase with these 162 cel parameters is labeled the y-phase. A similar procedure was utilzed in determining the structure of [Fe2(imid)4(bipy)]x at ~ 143 K. A crystal of [Fe2(imid)4(bipy)]x was examined above 170 K and found to have the same cel parameters as those determined earlier at 294 K. The same crystal was then cooled to ~ 143 K and the cel parameters o were determined and found to have changed [a = 17.1338, b = 18.5426, c = 23.6199 A, a = 80.424, P = 75.364, y = 80.826 deg, V = 7105.1(2) A3]. The phase with these cel parameters is labeled the p-phase. The Pi space group is retained in the three structures determined at 294 K (a-phase), 143 K (p-phase) and 113 K (y-phase); therefore, there are no crystalographic transitons involved, but structural phase transitons are evident in [Fe2(imid)4(bipy)]x. That the volume of the unit cel in the y-phase is three times larger than the one in the a-phase is due to the existence of six unique Fe(I) chromophores in the structure of the y-phase, as shown by its asymmetric unit depicted in Figure 5.5. The FeN6 cores involving the Fe(l), Fe(3) and Fe(5) sites in the y-phase have been slightly modifed compared to the octahedral Fe(l) site in the a-phase (Figure 5.2). The Fe(l)-N(9) and Fe(l)-N(10) bond lengths are 2.217(2) and 2.314(2) A, respectively, compared to 2.262(2) and 2.299(2) A, respectively, at 294 K. Selected bond lengths for the a- and y-phases, are shown in Appendix I, Table 1-6. The bond angles are also diferent at the two temperatures; for instance, the N(10)-Fe(l)-N(l) angle of 81.33(9)° at 113 K 163 coresponds to the N(2)-Fe(l)-N(3) angle of 88.99° at 294 K. As a consequence of the reduction in this angle the adjacent N(10)-Fe(l)-N(7) angle of 91.00(9)° at 113 K difers significantly from the coresponding N(2)-Fe(l)-N(8) angle of 83.61(9)° at 294 K. Figure 5.5 View of the asymmetric unit of [Fe2(imid)4(bipy)]x (y-phase, 113 K) and atom numbering scheme (33% probabilty thermal elipsoids). 164 Significant diferences between the a- and y- phases in bond lengths and angles involving the FeN4 core are also seen. Fe(2)-N(6) is 2.019(2) A at 113 K against 2.036(2) A at 294 K. Furthermore, the N(2)-Fe(2)-N(4) bond angle is 107.4° at 113 K compared to the coresponding N(7)-Fe(2)-N(9) angle of 117.6° at 294 K. Selected bond angles for the a- and y-phases of [Fe2(imid)4(bipy)]x, are shown in Appendix I, Table 1-7. The similtude of the iron ion connectivity diagrams of the a- and y-phases of [Fe2(imid)4(bipy)]x (Figures 5.4 and 5.6) show that the structural changes are relatively subtle. Nonetheless, these structural phase transitons were properly identified by DC and AC magnetic susceptibilty measurements as wel as Mossbauer spectroscopy as wil be shown in subsequent sections. As previously mentioned, the c axis of the unit cel of the y-phase is almost three times longer than that of the a-phase. This situation can be understood beter by examining the octahedral iron centers in the connectivity diagrams of these phases (Figures 5.4 and 5.6, respectively). Hence, for the a-phase, looking approximately along the c axis, octahedral iron chromophores equivalent to that on the first row, can be found on the second and third rows shown (Figure 5.4). In contrast, in the y-phase there are three diferent, and unique, octahedral iron chromophores in the first, second and third rows. The fourth row (not shown) is equivalent to the first. (Figure 5.6). As a consequence, the unit cel increases by approximately three times in volume compared to that of the a-phase (vide supra). 165 Figure 5.6 Iron ion connectivity diagram of a section of two double-layer sheets for the y-phase of [Fe2(imid)4(bipy)]x. Octahedral iron (red or semi-filed), tetrahedral iron (green or non-filed). View looking approximately down the c axis. Interestingly, the P-phase has a unit cel volume six times larger than the a-phase, and two times larger than that of the y-phase discussed above. This is atributed to the existence of six unique octahedral iron(I) chromophores and six unique tetrahedral iron(I) chromophores in the P-phase, as shown in Figure 5.7. Changes in bond distances and angles occuring in the P phase, in comparison to the other structural phases, are similar to those discussed for the a and y-phases above. Crystalographic data and selected bond lengths and angles for the P-phase, are listed in Appendix I, 166 Figure 5.7 View of the asymmetric unit of [Fe2(imid)4(bipy)]x (P-phase, 143 K) and atom numbering scheme. Tables 1-8 and 1-9, respectively. In the connectivity diagram for the y-phase (Figure 5.6) going from left to right along a "row" (along the a axis) the octahedral centers are identical as are the tetrahedral centers. In contrast, in the (3-phase there are two diferent and unique octahedral chromophores which alternate along the a axis. The same applies to the tetrahedral chromophores. The situation regarding the c axis is the same as for the y-phase. As a result, the unit cel volume of the (3-phase increases approximately two times compared to that of the y-phase {vide supra). 167 The change in the asymmetric units determined for the a and y phases (Figures 5. 2 and 5.5, respectively) is revealed on atempting to overlap the two structures around an octahedral iron centre. Thus, by matching the bipy ligands and one of the imidazolate bridging ligands from the a- and y-phases, as shown in Figure 5.8, another imidazolate ligand (vertical positon in Figure 5.8) in the y-phase is oriented diferently from the coresponding one in the a-phase. Figure 5.8 Comparison of coordination sphere geometries by overlapping octahedral irons (red circle) in the a- (black bonds) and y- (green bonds) phases of [Fe2(imid)4bipy]x. In general, accordingly to X-ray difraction, rotation of imidazolate moieties about the bridge axis seems to be the major structural diference between the three phases in [Fe2(imid)4(bipy)]x. 168 5.2.2.3 MAGNETIC PROPERTIES Variable temperature DC magnetic susceptibilties of a powdered sample of [Fe2(imid)4(bipy)]x were measured at fields of 500 and 10 000 G from 2 to 300 K. The Peff versus T plot (10 000 G, 2 to 300 K, Figure 5.9) shows a clear discontinuity near 135 K. Another irregularity in the same plot is barely apparent at slightly above 150 K. These two anomalies are atributed to structural phase transitions, as confirmed by low temperature X-ray difraction studies (vide supra). The structural phase transitons have been studied in detail by AC and DC susceptibilty measurements. These are presented folowing the discussion of the magnetic phase transition found in the y-phase of this compound. Figure 5.10 presents the % versus T and u^r versus T data obtained for y-[Fe2(imid)4(bipy)]x below 100 K at 500 G. The value of peff decreases smoothly with temperature from 5.3 U\B at 300 K to a low of 3.96 (IB just above 11 K. Below this temperature, ji^ f  increases abruptly to a maximum value of 12.6 (XB at 3 K before decreasing again with temperature to 11.2 \i& at 2 K. The onset of the magnetic transition at around 11 K is also observed in the % versus T plot (Figure 5.10). The magnetic susceptibilty, which increases smoothly with decreasing temperature below 169 50 100 150 T ( K ) 200 250 300 Figure 5.9 % and |i«ff versus T plots at 10 000 G for [Fe2(imid)2(bipy)]x. 300 K, rises abruptly as the temperature decreases below 11 K. The magnetization versus field plots at three temperatures shown in Figure 5.11, reflect this unusual magnetic behaviour. The plots are linear at 30 and 20 K and extrapolate to zero magnetization at zero applied field while distinct curvature is seen at 10 and 4.8 K. At 4.8 K the plot extrapolates to give a net magnetization at zero applied field. Cycling the applied field between +55 000 and -55 000 G at 4.8 K generates a hysteresis loop, the 170 0 10 20 30 40 50 60 70 80 90 100 T(K) Figure 5.10 x M«f versus T plots at 500 G for y- [Fe2(imid)2(bipy)]x. central portion of which is shown in Figure 5.12. From this is obtained a remnant magnetization of 200 cm3Gmor' and a coercive field of 15 G. Further evidence for structural phase transitons in [Fe2(imid)4(bipy)]x can be seen in the traditional Curie-Weiss analysis of the magnetic data. This analysis can provide evidence for the primary exchange process present in the system. A plot of %" 171 c £ £ s o a N c OX) 12000 10500 9000 7500 6000 4500 3000 1500 0 t D A | 6 ° o A o A o 10000 20000 30000 40000 Applied Field (G) 50000 60000 Figure 5.11 Magnetization versus applied field plots at diferent temperatures for Y-[Fe2(imid)2(bipy)]x. versus T (10 000 G ; 2 - 300 K) for [Fe2(imid)4(bipy)]x (Figure 5.13) reveals two linear regions, coresponding to two of the three structural phases present. Detection of the |3-phase in this plot is not possible, consistent with the fact that its magnetic properties are barely distinguishable from those of the a-phase. Fiting the data in Figure 5.13 to the Curie-Weiss equation yields: (i) employing data in the temperature range 300 - 170 K, C (Curie constant) = 3.60 cm3Gmor', 6 (Weiss constant) = -6.7 K and (ii) for the range 172 -200 -150 -100 -50 0 50 100 150 200 A p p l i e d F i e l d ( G ) Figure 5.12 Magnetic hysteresis plot at 4.8 K for y- [Fe2(imid)4(bipy)]x. 120 - 30 K, C = 3.04 cn^Gmol"1, 0 = -9.9 K. The observed negative Weiss constants are consistent with antiferomagnetic coupling as the primary exchange process operating here. However, the interpretation of these parameters is complicated by the fact that the three phases of [Fe2(imid)4(bipy)]x contain octahedral centres, which for high spin d6 means first order orbital contribution and spin-orbit coupling. As a result these centres would show decreasing moments with temperature and negative Weiss 173 0 50 100 150 T(K) 200 250 300 Figure 5.13 Plot of % versus temperature at 10 000 G for [Fe2(imid)4(bipy)]x. constants even in the absence of antiferomagnetic exchange. It seems reasonable to assume that the 9 values here are a composite of single ion efects and exchange interactions but to determine the relative contributions would be difficult. Overal the results are consistent with [Fe2(imid)4(bipy)]x exhibitng antiferomagnetic exchange 174 interactions at high temperatures and, coupled with the observed net moment ground state, this classifies it as a weak feromagnet. The magnetic behavior exhibited by y-[Fe2(imid)4(bipy)]x is similar to that reported for [Fe3(imid)6(imidH)2]x [1], [Fe(2-meimid)2-0.13Cp2Fe]x [2], and [Fe(4-abimid)2]x (Chapter 3) and it is possible that y-[Fe2(imid)4(bipy)]x like the other systems listed, also exhibits canted-spin antiferomagnetism. This conclusion is supported by the fact that the highest magnetization reached, 10 940 cm3Gmor' (at 4.8 K and 55 000 G), is significantly smaler than the theoretical saturation value of 22 300 cm ^mol"1 [5]. As described in Chapter 3, the spin canting angle, y, can be estimated by extrapolating the plot of M versus H to H = 0 at a temperature below Tc. Doing this for y-[Fe2(imid)4(bipy)]x gives a saturation moment (Ms(0)) of 6980 cm3Gmof1, and from this, an estimation of the spin canting angle, y, of ~ 17°. It should be noted that unlike [Fe(2-meimid)2-0.13Cp2Fe]x [2] and [Fe(4-abimid)2]x (Chapter 3), y-[Fe2(imid)4(bipy)]x has a structure in which the nearest neighbor interactions are between iron ions that difer significantly in their ligand chromophores. Because of the diference in g values, the size of the interacting magnetic dipoles wil difer and, hence, even perfect antiparalel alignment of spins between neighbours would lead to a residual moment on the sheets. This form of ferimagnetism, which was suggested to possibly occur in the 1-D polymer, polybis(l-methyl-2-thioimidazolate)iron(I) [7] (See Chapter 7), could be the cause of the 175 anomalous magnetic behaviour observed for y-[Fe2(imid)4(bipy)]x or, at least, contribute significantly to it. In an efort to determine the onset of the magnetic transition in y-[Fe2(imid)4(bipy)]x the temperature dependence of the field-cooled magnetization (FCM), zero-field-cooled magnetization (ZFCM) and remnant magnetization (REM) were examined (Figure 5.14). The FCM curve, measured by cooling the sample under a DC field of 50 G, shows a clear increase in M in the 6 - 8 K region. On further cooling, this curve exhibits a maximum in the 2 - 3 K region. The derivative curve, d(FCM)/dT, has an extremum at 3.5 K, which can be considered as the critical temperature, Tc [8]. The ZFCM curve, measured by cooling the sample in zero field, then warming in a DC field of 50 G, exhibits a maximum at ~ 3 K and is significantly lower than the FCM curve at this temperature and at lower temperatures. The derivative curve, d(ZFCM)/dT, shows an extremum at 4.0 K. Thus, Tc for this system, as measured by the FCM-ZFCM experiment, can be considered as Tc = 3.75 K, which is the average temperature of those obtained from the FCM and ZFCM plots. Finaly, REM, obtained by cooling the sample in a 50 G DC field, then colecting the data while warming it in zero field vanishes at a slightly higher temperature (at 5 K, as confirmed by looking at the data obtained) than Tc, which agrees beter with the Tc determined by AC susceptibilty measurements of y-[Fe2(imid)4(bipy)]x, shown below. 176 800-700-600-~_ 500H o g rj 400-1 £ S 300 200-100-0-d ZFCM / dT • FCM • ZFCM * REM 2 10 12 T ( K ) Figure 5.14 Plots of ZFCM, FCM and REM for Y-[Fe2(imid)4(bipy)]x at a DC field of 50 G. Further information on the magnetic phase transition of Y-[Fe2(imid)4(bipy)]x can be deduced from AC magnetic susceptibilty measurements. The in-phase, x\ and out-of-phase, AC molar magnetic susceptibilties, with a zero static field and a 125 Hz oscilating field of 1 G, for y-[Fe2(imid)4(bipy)]x are displayed in Figure 5.15. x" is non-zero below 6.1 K, which is in agreement with the onset of a magnetic transition at that temperature. %'X" Pea^ a t 5-8 and 5.7 K, respectively, proving the material y-[Fe2(imid)4(bipy)]x is a genuine magnet at these temperatures. 177 20 15 * X' s 5 H OH : : : : : : t t t i n 5.5 —I ' 1 1 1 1 1 1 6.0 6.5 7.0 7.5 8.0 T(K) Figure 5.15 Temperature dependences of the in-phase, %\ and out-of-phase, AC magnetic susceptibilties for y-[Fe2(imid)4(bipy)]x at f=125 Hz and H= 1 G. In an efort to beter characterize the structural phase transitons that occur in [Fe2(imid)4(bipy)]x detailed AC and DC magnetic susceptibilties measurements were made over the appropriate temperature ranges. It is the first time such studies have been made on a molecule-based magnet. The yj (DC) versus T plot, shown in Figure 5.16, reveals two inflections in the 130 - 160 K region folowing a cooling mode, the higher 178 3.5-3 .4 -3.3-• "o 3.2-£ • 3.1-• 3.0-H 2.9-2.8-2.7-100 120 T(K) Figure 5.16 yj versus T for [Fe2(imid)4(bipy)]x. / / D C = 10 000 G. Cooling mode. temperature anomaly being less evident than the lower one. In order to determine the transition temperatures accurately, the temperature dependence of the derivative d(xT)/dT, in the cooling mode, was calculated and ploted as shown in Figure 5.17. This gave transition temperatures of 150 K for the a —> p transition and 135 K for the p —> y transition. 179 Figure 5.17 Temperature dependence of the derivatives d(^T)/dT (DC) in the cooling mode and determination of the transition temperatures for [Fe2(imid)4(bipy)]x. HDC= 10 000 G. Further details on the two phase transitons occuring in [Fe2(imid)4(bipy)]x, were obtained from DC magnetic susceptibilty measurements in both, cooling and warming modes. Also, for higher accuracy, a higher-density data acquisiton was caried out close to the phase transition temperature ranges. A DC field of 1 000 G was utilzed in these studies with the hope of obtaining a beter resolution of the high temperature phase transition. In Figure 5.18 the temperature-dependence of yl (100 -190 K) is shown. As the temperature is lowered (cooling mode), yl is decreases 180 Figure 5.18 Cooling and warming modes yT versus temperature plots for [Fe2(imid)4(bipy)]x. //DC= 1 000 G. Insert plot shows an augmentation of the OK-»|3 transition region. slightly until it reaches a temperature of ~ 152 K, it then decreases abruptly until ~ 150 K (a —> (3) ataining a "plateau" between ~ 150 138 K. Finaly, yT diminishes abruptly between the temperatures of ~ 138 K and ~ 132 K (3 -» y) before "leveling of" again. As the temperature is increased (warming mode), the temperature where the yT rises sharply is shifted by ~ 5 K to ~ 137 K, giving rise to the hysteretic behaviour 181 shown in Figure 5.18. Hence, thermal hysteresis is evident for the low-temperature transition (y P). This low temperature phase transition, occuring with thermal hysteresis, may be considered as being a first order transition [9]. Above 137 K, and folowing a warming mode, %T increases gradualy, folowing the cooling curve in this region reasonably closely. The %T then increases abruptly over the 150 to 152 K region, again folowing the cooling mode closely, the abrupt change coresponding to the P —> a transition. In contrast to the P <-> y, there appears to be no significant hysteresis associated with the P <-> a transition. We note that above and below each of these transitons the data obtained in the warming mode are slightly below those obtained in the cooling mode. This could be because the a —> P and p —> y transitons involve cooperative interaction between chromophores which cause the final stages of the conversion to occur very slowly. Hence at each temperature studied on cooling a smal amount of the higher temperature phase remains. On warming, at each temperature there is less of the high temperature phase "impurity" present and hence the %T value is slightly lower. This could of course be tested by alowing long periods (longer than 15 minutes alowed in the standard run) between colecting data points on cooling. In view of the relative minor efect observed, and the cost of doing the extended time runs, these experiments were not performed. In order to determine the transition temperatures more accurately, the temperature dependence of the derivative d(%T)/dT, in the cooling and warming modes 182 [10], was calculated and ploted as shown in Figure 5.19. This gave transition temperatures of 151 K for the a <-»• P transition, and 135 K for the p <-» y transition, the later with a thermal hysteresis width of 4 K (133 - 137 K). 0.30 0.25-_~ 0.20-i "o £ "a 0.15 H o H 3| 0.10-1 H >< "° 0.05-1 0.00 137 K 133 K warming cooling - i — | — i — | — i — | — i — | — i — | — i — | — i — | — i — | — i -100 110 120 130 140 150 160 170 180 190 T(K) Figure 5.19 Temperature dependence of the derivatives d(%T)/dT (DC) in the cooling and warming modes and determination of the transition temperatures for [Fe2(imid)4(bipy)]x. = 1 000 G. 183 A high-density AC magnetic susceptibilty measurement study was also performed for the (3 <-» y structural phase transition. % versus temperature data, in the cooling mode, are shown in Figure 5.20. As in the DC magnetic susceptibilty 0.0250 0.0150 Figure 5.20 AC % versus T plot for [Fe2(imid)4(bipy)]x. Cooling mode, f = 500 Hz, HAC = 2.5 G. 184 measurements, a break in the plot appears between 135 K and 132 K. Meanwhile, in the warming mode plot (Figure 5.21) the anomaly is shifted to higher temperatures, appearing between 135 K and 139 K. The thermal hysteresis of the low-temperature phase transition of [Fe2(imid)4(bipy)]x is also confirmed by the plot of yT (AC) versus T as shown in Figure 5.22. As before, transition temperatures may be determined more o £ £ 0.0250 0.0225 4 0.0200 H 0.0175 H 0.0150 fi-phase J . Y-phase * V * ^ ^ ^ 110 120 130 140 150 160 T(K) Figure 5.21 AC x versus T plot for [Fe2(imid)4(bipy)]x. Warming mode, f: 500 Hz, HAC = 2.5 G. 185 3.1 © E E H 2.7 H 2.6 4 2.5 120 125 130 135 T(K) 140 145 150 Figure 5.22 Temperature dependence of AC yj in the cooling and warming modes for [Fe2(imid)4(bipy)]x. f = 500 Hz, HAc = 2.5 G. Arow down refers to cooling mode, arow up refers to warming mode. accurately from the extremes of the temperature dependences of the derivatives d(xT)/dT in both the cooling and the warming mode [10]. This is shown in Figure 5.23. These transition temperatures are found to be 133 K in the cooling mode and 137 K in the warming mode in total agreement with the temperatures determined by the DC susceptibilty measurements (vide supra). Hence, the width of the thermal hysteresis in the (3 <-> y transition is 4 K. 186 0.35-0.30-0.25-o £ • E 0.20-0.15-H 0.10-H 0.05-0.001 Figure 5.23 Temperature dependence of the derivatives d(^T)/dT (AC) in the cooling and warming modes and determination of the transition temperatures for the y-phase of [Fe2(imid)4(bipy)]x. f = 500 Hz, HAc = 2.5 G. In an atempt to study the a <-» p transition in more detail, AC magnetic susceptibilty measurements were caried out in the temperature range from 131 K to 171 K in both cooling and warming modes. Unfortunately, as seen in Figure 5.24, only the p y transition (showing hysteresis) was detected but no other anomaly was found in the cooling or warming mode coresponding to the a <-> P transition. Apparently the diference in the magnetic properties between the a and p forms is too smal to be observed in the AC measurements. 187 0.034 4 0.032 4 o g 0.030 -| ^ 0.028 4 0.026 4 • *•. 130 140 150 T(K) 160 -< 1 — 170 Figure 5.24 Cooling and warming mode AC % versus temperature plots for [Fe2(imid)4(bipy)]x. f = 500 Hz, HAC = 2.5 G. 5.2.1.4 MOSSBAUER SPECTROSCOPY The room temperature Mossbauer spectrum of [Fe2(imid)4(bipy)]x (Figure 5.25) coresponds to two overlapping quadrupole doublets consistent with the X-ray 188 Velocity Relative to Fe (mm s*) Figure 5.25 Mossbauer spectrum of [Fe2(imid)4(bipy)]x at 293 K. crystalography results at that temperature, that indicate equal population of distorted six-and four-coordinate sites. The Mossbauer parameters obtained herein for the nominal Td and Oh sites at 293K are: isomer shifts (5) of 0.74 mm s"1 and 1.01 mm sf1, and quadrupole splitings (AE) of 2.01 mm s"1 and 1.06 mm s"1, respectively, values fairly characteristic of high-spin iron(I) with "al" nitrogen ligation [11]. AE values are not reliably diagnostic of coordination number for high-spin ferous complexes. The 189 spliting value ranges can often overlap for four, five and six coordination environments (even with similar ligands) depending on the degree of distortion of the local coordination environment. On the other hand, isomer shifts have been found to be quite sensitve to coordination number for fixed spin state and similar ligation. Specificaly, for spin quintet iron(I) with nitrogen or mixed nitrogen - halogen ligation, 8 values for six coordination generaly range from ~ +1.0 to 1.2 mm s"1 [11-13]; five-coordination from -+0.85 to 0.95 mm s"1 [13-15] and four-coordination from ~ +0.70 to 0.85 mm s"1 [16] at room temperature. Hence, the present 8 values in combination with the available literature results appear to unequivocaly confirm the six- and four-coordinate nature of the iron of [Fe2(imid)4(bipy)]x. The structural phase transitons occuring in [Fe2(imid)4(bipy)]x were also examined meticulously with Mossbauer spectroscopy. In fact, the high-temperature transition (a <-> (3) of [Fe2(imid)4(bipy)]x was first found using this spectroscopic technique, and later confirmed with the X-ray difraction and DC susceptibilty studies as already described. A set of Mossbauer spectra for [Fe2(imid)4(bipy)]x, determined in the warming mode from 145.5 K to 175.1 K, is shown in Figure 5.26. According to this study the low-temperature (P «-» y) phase transition occurs between about 146.9 and 148.3 K, while the high-temperature phase transition occurs between about 160.2 and 167.7 K (Figure 5.26). 190 —1 1 1 1 1 1 • T = 145.5K ^ — i 1 1 1 1 1 1 T = 157.3K ~w T = 146.9K y 1 1 1 i i v T = 160.2K V —« 1 , , T = 148.3K V A. j 1 ! "•: • : * il *' i : •i l ' I • - i w T = 167.7K ~ ^ A f T =150.0K V ~ ~ ^ A r\ r ; f : 1 j • * I i } i • • • • y M • •* T = 175.1K JJY Velocity Relative to Fe (mm s ) Figure 5.26 Mossbauer spectra in the warming mode for [Fe2(imid)4(bipy)]. 191 After heating a sample of [Fe2(imid)4(bipy)]x to 169.7 K, Mossbauer spectra were measured from 165.1 to 129.9 K (cooling mode). As shown in Figure 5.27, the data indicate that the high-temperature phase transition occurs between 165.1 and 162.7 K, while, the low-temperature phase transition occurs between 144.9 and 139.9 K. In the Mossbauer study, the low-temperature transition exhibits hysteresis of around 3 to 5 K and the high-temperature transition exhibits no significant hysteresis, in qualitative agreement with the AC and DC magnetic susceptibilty studies. However, both structural phase transitons appear to occur at ~ 10 K higher than indicated by the DC and AC susceptibilty studies. There is not a straightforward explanation for this, although, a possible explanation may involve the diferent time-scale of Mossbauer spectroscopy. It seems the microscopic Mossbauer efect detects the two phase transitons "earlier" than the bulk AC and DC magnetic susceptibilty measurements in the cooling mode. However, in the warming mode, AC and DC susceptibilty measurements detect the transitons "earlier" (vide supra). This situation may also be an experimental efect regarding these techniques. Mossbauer measurements were done on muls while the AC and DC susceptibilty measurements were performed on crystaline solids, which could be the source of the diferent phase transition detection temperatures. 192 S o -*-»• a i -o X ! 13 I* t • i * * i I' i 5 1 V V \\ T = 165.1K | V W M V if T = 139.9K j V i • i i \h\ i / V I T = 163.6K j \ i 1 i 1 i I i — * \ A r 1 • l i H 1 V • -1 T = 135.0K J V T = 162.7K | V v y T = 132.7K | y ~~' 1 ' I ' 1 ' W \ V \\ T = 144.9K | y i i i \ A * V / i * V t < i» »• T = 129.9K | V Velocity Relative to Fe (mm s ) Figure 5.27 Mossbauer spectra in the cooling mode for [Fe2(imid)4(bipy)]x. 193 5.3 P O L Y - 2 , 2 ' - B J P Y R I D I N E T E T R A K I S ( I M I D A Z O L A T O ) D I C O B A L T ( I I ) 5.3.1 RESULTS AND DISCUSSION 5.3.1.1 SYNTHESIS, THERMAL AND STRUCTURAL CHARACTERIZATION The reaction of cobaltocene with imidazole and excess molten 2,2'-bipyridine generated, [Co2(imid)4(bipy)]x, as a microcrystaline powder. Details of this reaction are given in Chapter 9, section 9.2.2.7. Although this material could not be obtained in a form suitable for single crystal X-ray difraction studies, it was possible to show by powder difraction studies that it is isomorphous and probably isostructural with the room temperature form of [Fe2(imid)4(bipy)]x. The X-ray powder difractogram of [Co2(imid)4(bipy)]x coincides wel with that calculated from the single-crystal data of [Fe2(imid)4(bipy)]x obtained at room temperature, employing the program PowderCel [17] (Figure 5.28). Indexing the X-ray powder difractogram of [Co2(imid)4(bipy)]x using the program Celref [18], generated the latice parameters: a = 10.521, b = 13.729 and c = 9.181 A. These are very similar to the room temperature latice parameters obtained for a-[Fe2(imid)4(bipy)]x (a = 10.507, b = 13.730 and c = 9.188 A). The room temperature infrared spectra of a-[Fe2(imid)4(bipy)]x and [Co2(imid)4(bipy)]x were also found to exhibit the same vibrational bands at almost the same frequencies. 194 •—w 10 20 30 40 50 20" Figure 5.28 X-Ray powder difractograms of [Co2(imid)4(bipy)]x (top, experimental) and [Fe2(imid)4(bipy)]x (botom, calculated). In contrast to the iron(I) analogue, [Co2(imid)4(bipy)]x is air-stable, but both compounds have properties consistent with polymeric structures, for example [Co2(imid)4(bipy)]x does not dissolve in common organic solvents and it is non-volatile. A thermal gravimetric study performed on the Co(I) material gave no evidence for decompositon until ~ 250 °C (Figure 5.29). The TGA plot folows a two step weight 195 0 1 1 1 1 1 1 1 1 200 400 600 800 T(°C) Figure 5.29 TGA plot for [Co2(imid)4(bipy)]x. loss profile and resembles that of [Fe2(imid)4(bipy)]x. The first major weight loss (~ 30 %) occurs between 225 °C and 300 °C, and the second one, with a weight loss of ~ 47 % of the initial weight, happens between 390 °C and 660 °C. A weight percentage of ~ 25 % remains at the highest temperature studied, 800 °C. In contrast to the TGA plot of [Fe2(imid)4(bipy)]x, there is a plateau between 300 °C and 390 °C in the plot for [Co2(imid)4(bipy)]x. This plateau seems to corespond to the formation of [Co(imid)2]x by loss of the bipyridine ligand (~ 29 %) in the first step of the thermal decompositon. A similar situation occured when the thermal gravimetric study of 196 [Co3(imid)6(imidH)2]x was caried out and [Co(imid)2]x was obtained by the loss of the neutral imidazole molecules (Chapter 4, section 4.2). 5.3.1.2 MAGNETIC PROPERTIES A noticeable diference in the magneto-structural behaviour of [Co2(imid)4(bipy)]x versus [Fe2(imid)4(bipy)]x is that whereas the later shows a discontinuity in the (leff versus temperature plot at 135 K (Figure 5.9), no equivalent discontinuity is observed for [Co2(imid)4(bipy)]x (Figure 5.30). The cobalt derivative does not appear to undergo the structural phase transitons exhibited by [Fe2(imid)4(bipy)]x. Otherwise the magnetic properties of [Co2(imid)4(bipy)]x paralel those of [Fe2(imid)4(bipy)]x. For [Co2(imid)4(bipy)]x, the value of jieff, measured at an applied field of 500 G, decreases with decreasing temperature from 4.35 U , B at 300 K to 2.89 p.B at 13 K. Below 13 K it increases abruptly, signaling the onset of long-range feromagnetic ordering (Figure 5.31). This magnetic behaviour is also seen in the % versus T plot for the compound which shows an incipient maximum in % just above 13 K and an abrupt rise in % below this temperature. As the temperature is lowered further to 2 K, % tends to saturation (Figure 5.31). Studies at an applied field of 10 000 G (Figure 5.30) show a less prominent increase in % and u«f below Tc than observed in the 500 G data. This 197 0.30 5.0 0.25 !_ 0.20 o £ I 0.15 0.10 0.05 0.00 A A A A A A A A A A A A A A A 50 100 150 T ( K ) 200 250 4.5 4.0 3.5 m 3.0 =L 2.5 2.0 1.5 1.0 300 Figure 5.30 % and jieff versus T plots at 10 000 G for [Co2(imid)4(bipy)]x resembles the behaviour observed for [Fe2(imid)4(bipy)]x. Nonetheless, for both [Fe2(imid)4(bipy)]x and [Co2(imid)4(bipy)]x the magnetic transition is clearly observed at this higher field (Figures 5.9 and 5.30, respectively). 198 4.0 3.5 i -T 3.0 } | 2.5 | 2.0 1.5 { 1.0 0.5 H 0.0 A A A A A A A A A A A A A 16 14 12 as h 10 f =L 8 6 t 4 A 2 0 10 20 30 40 50 60 70 80 90 100 T(K) Figure 5.31 x m^ M-ef versus T plots at 500 G for [Co2(imid)4(bipy)]x. The efect of the low temperature magnetic transition in [Co2(imid)4(bipy)]x is seen in the magnetization versus applied field plots at diferent temperatures (Figure 5.32). At 4.8 K the plot is non-linear and, in contrast to the higher temperature plots, it does not extrapolate to zero at zero applied field. Cycling the applied field between +55 000 and -55 000 G at 4.8 K generates a hysteresis loop (Figure 5.33) from which a 199 I *© s o £ a © 5000 4000 j 3000 i '5 2000 US N *-C S §f 1000 2 • A O o • • • A • 10000 20000 30000 40000 Applied Field (G) 50000 60000 Figure 5.32 Magnetization versus applied field plots at diferent temperatures for [Co2(imid)4(bipy)]x. coercive field of 125 G and a remnant magnetization of 1900 cm3Gmol"1 are obtained. A plot of x 1 versus T (10 000 G data) is linear over the temperature range 300 - 25 K and a Curie-Weiss analysis of this data yields C = 2.64 cm 3Gmor1 and 0 = -25 K (Figure 5.34). As discussed above for [Fe2(imid)4(bipy)]x, overal these magnetic 200 Applied Field (G) Figure 5.33 Magnetic hysteresis plots at 4.8 K for [Co2(imid)4(bipy)]x. properties classify [Co2(imid)4(bipy)]x as a weak feromagnet. However, it should be noted that, since [Co2(imid)4(bipy)]x contains high spin octahedral d7 metal centers and since such centres have first order orbital efects contributing to their magnetism, the comments made above concerning the interpretation of 6 values for [Fe2(imid)4(bipy)]x apply here also. 201 T(K) Figure 5.34 Plot of % versus temperature at 10 000 G for [Co 2(imid) 4(bipy)]x. The highest magnetization measured for [Co2(imid)4(bipy)]x (4 515 cn^GmoF 1 at 4.8 K and 55 000 G) is significantly lower than the theoretical saturation value (16 766 c m 3 G m o r ' for a S = 3/2 system [5]). As discussed above for [Fe2(imid)4(bipy)]x, this result is not inconsistent with spin canting or ferrimagnetic exchange providing the source o f the residual spin at low temperatures. 202 Extrapolation of the M versus H plot to H = 0 on the magnetization curve obtained at 4.8 K (Figure 5.32), yields the spontaneous magnetization, Ms(0) = 2640 cm3Gmol"1. From this, the canting angle is calculated as described previously for [Fe2(imid)4(bipy)]x, to be y~ 9°. That this polymer behaves as a magnet below Tc is confirmed by the shape of the ZFCM, FCM and REM plots in Figure 5.35. The zero-field-cooled magnetization (ZFCM) under an applied field of 50 G showed a maximum at 9.0 K. The field-cooled magnetization (FCM) under an applied field of 50 G increased rapidly below 10 K to a maximum value of - 1370 cn^Gmol"1 at 2 K (Figure 5.35). When the applied field was switched of at 2 K a remnant magnetization of - 1190 cm3Gmor' remained that decreased upon warming and vanished at - 10 K. No ambiguites are found in these plots, in contrast to the ones for [Fe2(imid)4(bipy)]x where the derivative of the plots had to be calculated in order to estimate Tc. Thus, the temperatures for the highest slope of the FCM plot, the peak on the ZFCM plot, and the temperature at which REM vanishes, are al within the range between 9.0 and 10 K. Hence, the critical temperature of [Co2(imid)4(bipy)]x is determined as Tc = 9.5 K [8]. This temperature is in reasonably agreement with the onset temperature (~ 13 K) of the long-range magnetic ordering determined by DC susceptibilty. 203 1400 H _ ^ 1200 "o E 1000 u | 800 | 600-| N « 400 -\ C I 200-0-0 • * - T 5 4i 10 15 • ZFCM • F C M REM 20 25 T ( K ) Figure 5.35 Plots of ZFCM, FCM and REM for [Co2(imid)4(bipy)]x using a DC field of 50 G. 5.4 SUMMARY AND CONCLUSIONS [Fe2(imid)4(bipy)]x, and the isomorphous cobalt compound, [Co2(imid)4(bipy)]x, have 2-D extended structures in which double layered sheets of alternating tetrahedraly and octahedraly coordinated metal ions are linked by single bridging imidazolates. The octahedral metal centers are additionaly coordinated by 2,2'-bipyridine ligands which occupy positons between the sheets, isolating the sheets from each other. The presence 204 of the two diferent iron centers in [Fe2(imid)4(bipy)]x is confirmed by ambient temperature Mossbauer studies. The magnetic properties of these materials reveal a transition to long-range feromagnetic order below 6 K for [Fe2(imid)4(bipy)]x and below 9.5 K for [Co2(imid)4(bipy)]x. Both materials reveal magnetic hysteresis at 4.8 K. Analysis of the data yield, for [Fe2(imid)4(bipy)]x and [Co2(imid)4(bipy)]x respectively, coercive fields of 15 and 125 G and remnant magnetizations of 200 and 1900 cm3Gmol"1. Magnetic parameters for three pairs of analogous iron(I) and cobalt(I) imidazolate compounds, including y-[Fe2(imid)4(bipy)]x and [Co2(imid)4(bipy)]x, which exhibit weak feromagnetism are shown in Table 5.1. For each pair, Hcoet is always greater for the cobalt system suggesting cobalt magnets are generaly harder than analogous iron ones. In contrast to what might have been expected it seems that the magnet strength, as measured by MKm, is not a simple function of the number of unpaired electrons or canting angle. While the cobalt material has the highest Miem in one pair it has the lower Mrem in the other two. Moreover, while the low MKm for [Co(4-abimid)2]x seems consistent with a low canting angle (Chapter 3), the canting angle for [Co2(imid)4(bipy)]x is lower than that of its iron analogue yet the MKm is greater for the cobalt system. [Fe2(imid)4(bipy)]x is unique in presenting two structural phase transitions, at ~ 151 K and ~ 135 K, as determined by DC and AC magnetic susceptibilty 205 Table 5.1 Magnetic parameters for three pairs of analogous iron(I) and cobalt(I) weak feromagnets. Compound Hcoet Ref. (G) (cn^Gmof1) Y-Fe2(imid)4(bipy) 15 200 This Chapter Co2(imid)4(bipy) 125 1900 ThisChapter Fe(4-abimid)2 80 2100 Chapter 3 Co(4-abimid)2 400 22 Chapter 3 Fe3(imid)6(imidH)2 200 2000 [1] Co3(imid)6(imidH)2 4 140 175 Chapter 4 Abbreviations: imid = imidazolate, 4-abimid = 4-azabenzimidazolate, bipy = 2,2'-bipyridine. measurements. These structural phase transitons have also been studied in detail by Mossbauer spectroscopy. These studies yielded transition temperatures of ~ 10 K higher than those determined by magnetic studies. Thermal hysteresis behavior is evident for 206 the low-temperature phase transition (p <-> y) a s determined by DC and AC susceptibilty as wel as Mossbauer spectroscopy. Single crystal X-ray difraction studies at 113 K and 143 K revealed that while the crystal integrity was retained, as wel as the basic double layer structural motif found in the room temperature a-phase, there were significant changes in cel and bond parameters for the P- and y-phases of [Fe2(imid)4(bipy)]x. To the author's knowledge this is the first time that two phase transitons have been detected in a molecule-based magnetic material ([Fe2(imid)4(bipy)]x) and studied in detail by a combination of X-ray crystalography, AC and DC magnetic susceptibilty measurements and Mossbauer spectroscopy. 207 References 1. S.J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter, J Amer. Chem. Soc. 119, 8675 (1997). 2. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter, Can J. Chem. 77, 425 (1999). 3. S. J. Retig, V. Sanchez, A. Storr, R. C. Thompson, and J. Troter, J. Chem. Soc, Dalton Trans. 3931 (2000). 4. O. Kahn. Molecular Magnetism. VCH. New York. 1993. p. 321. 5. R. L. Carlin. Magnetochemistry. Springer-Verlag. Berlin. 1986. pp. 7-9. 6. F. Palacio, M. Andres, R. Home, and A. J. van Duyeneveldt. J. Magn. Magn. Mater. 54-57, 1487(1986). 7. S.J. Retig, V. Sanchez, A. Storr, R. C. Thompson, and J. Troter. Inorg. Chem. 38, 5920 (1999). 8. O. Kahn. Acc Chem. Res. 33 (10) 647 (2000). 9. C. N. R. Rao, and K. J. Rao. Phase Transitions in Solids. An Approach to the Study of the Chemistry and Physics of Solids. McGraw-Hil. London. 1978. pp. 17-24. 10. Y. Garcia, O. Kahn, L. Rabardel, B. Chansou, L. Salmon, and J. P. Tuchagues. Inorg. Chem. 38, 4663 (1999). 11. N. N. Greenwood, and T. C. Gibb. Mossbauer Spectroscopy. Chapman and Hal Ltd. London. 1971. 12. B. W. Dockum, and W. M. Reiff. Inorg. Chim. Acta. 35, 285 (1979). 13. F. F. Charon, and W. M. Reiff. Inorg. Chem. 25, 2786 (1986). 14. W. M. Reiff, N. E. Erickson, and W. A. Baker. Inorg. Chem. 8, 2119 (1969). 15. D. Sedney, M. Kahjehnassiri, and W. M. Reiff. Inorg. Chem.20, (1981) 3476. 16. B. W. Dockum, and W. M. Reiff. Inorg. Chim. Acta. 120, 61 (1986). 208 17. PowderCel, version 2.3, W. Krauss and G. Nolze, Federal Instiute for Materials Research and Testing (BAM), Berlin, 1997. 18. LMGP-Suite of Programs for Interpretation of X-Ray Experiments, by Jean Laugiert and Bernard Bochu, ENSP/Laboratoire des Materiaux et du Genie Physique. BP 46. 38042 Saint Martin d'Heres, France. htp:/www.inpg.fr/LMGP and htpp:/www.ccp 14.ac.uk/tutorial/mgp/. 209 Chapter 6 POLY-2,2 ':6',2"-TERPYRIDINEOCTAKIS(IMIDAZOLATO)-TETRAIRON(II). A VERY SOFT 2-D MOLECULE-BASED MAGNET 6.1 INTRODUCTION The remarkable influence of 2,2'-bipyridine on the structures of the iron(I) and cobalt(I) imidazolate polymers was described in the previous chapter. The efect of their structural dimensionality on the physical and magnetic properties of these 2-D polymers, encouraged the investigation of similar systems utilizing a diferent neutral multi-dentate ligand, namely 2,2':6',2"-terpyridine (terpy). It was expected that, by virtue of the tridentate meridional coordination mode of terpy, the considerable geometric demands of this ligand might influence the structure of transition metal imidazolate polymers, in a similar way to the bipy ligand, yielding further examples of interesting two-dimensional extended systems. The title compound, [Fe4(imid)g(terpy)]x, was synthesized by the reaction of ferocene with imidazole in the presence of terpyridine in a sealed tube at elevated temperatures. This polymer has a novel 2-D extended structure, as determined by an X-ray crystalography study, in which wrinkled sheets of four-coordinate (two unique centers of this type), five-coordinate and six-coordinate iron(I) ions are linked by 210 single bridging imidazolates. The presence of the four unique iron sites is confirmed by Mossbauer spectroscopy. DC magnetic susceptibilty measurements reveal antiferomagnetic interactions between metal centers above about 12 K, and the onset of a transition to a feromagnetic state below this temperature. Long-range feromagnetic ordering below Tc = 6.5 K is confirmed by AC magnetic susceptibilty, zero-field-cooled magnetization and Mossbauer studies. A very unique hysteresis loop, determined by field-dependant magnetization studies, show the material to be a very soft molecule-based magnet. 6.2 RESULTS AND DISCUSSION 6.2.1 SYNTHESIS AND PHYSICAL PROPERTIES Details of the synthesis of [Fe4(imid)g(terpy)]x are described in Chapter 9, section 9.2.1.4. Ferocene, imidazole and terpyridine were sealed under vacuum in a Carius tube and the mixture was heated. The desired compound was isolated as green crystals suitable for X-ray crystalography. [Fe4(imid)8(terpy)]x appeared to be moisture sensitve but could be handled briefly under normal atmospheric conditons. Powdered samples of this compound would turn red upon exposure to air for around a week. The compound is insoluble in 211 common organic solvents. The TGA plot obtained for [Fe4(imid)8(terpy)]x (Figure 6.1) shows the thermal robustness of this compound. There is no apparent decompositon of 100 200 300 400 500 600 700 800 T(°C) Figure 6.1 TGA plot of [Fe4(imid)g(terpy)]x. this polymer until a temperature of 270 °C is reached. A decrease in weight is observed between 270 °C and 630 °C, with ~ 69 % of the initial weight lost at the highest temperature. The thermal decompositon behavior shown by [Fe4(imid)8(terpy)]x is very similar to that exhibited by [Fe2(imid)6(bipy)]x (Chapter 5, section 5.2.1.1.) 212 6.2.2 X-RAY DIFFRACTION STUDIES Single crystal X-ray difraction studies reveal [Fe4(imid)8(terpy)]x to be comprised of distorted tetrahedral (two unique chromophores of this type), distorted trigonal bipyramidal and distorted octahedral iron(I) ions linked by single bridging imidazolates (Figure 6.2), forming a novel 2-D extended structure resembling wrinkled sheets (Figures 6.3, 6.4 and 6.5). Crystalographic data, atom coordinates, selected bond lengths and angles appear in Appendix I, Tables I-10 and 1-11. The six-coordinated metal centers are coordinated by bridging imidazolate ligands and terpy ligands (Figure 6.2) that occupy positons between the sheets, isolating the sheets from each other, as can be seen in Figure 6.3. This capping of iron centres by terpy controls the dimensionality, resulting in a 2-D polymer (Figure 6.3), in a manner similar to that observed in [Fe2(imid)6(bipy)]x. An iron ion connectivity picture of a section of the structure of [Fe4(imid)g(terpy)]x is given in Figure 6.5. The use of terpy to influence the structural dimensionality has been reported recently for hybrid metal oxides extended systems [1]. However, no reports on the presence of three diferent iron(I) chromophore geometries in the same coordination compound, were found in the literature. 213 Figure 6.2 Repeat unit of [Fe4(imid)g(terpy)]x showing the atom numbering scheme; 33 % probabilty thermal elipsoids are shown. 214 Figure 6.3 View of a section of [Fe4(imid)8(terpy)]x looking down the a axis. Terpy ligands and C-4 and C-5 of imidazolate ligands have been omited in the botom view for clarity. Hydrogen atoms are not shown. 215 Figure 6.4 View of a section of [Fe4(imid)g(terpy)]x looking down the b axis. For clarity, terpy ligands and C-4 and C-5 of imidazolate ligands have been omited in the botom view. Hydrogen atoms are not shown. 216 Figure 6 .5 Iron ion connectivity diagram for a section of [Fe4(imid)g(terpy)]x. Four-coordinate ions (green and pink/black elipsoids), six-coordinate ions (blue elipsoids) and five-coordinate ions (red elipsoids). 217 6.2.3 MOSSBAUER SPECTROSCOPY The room temperature (293 K) Mossbauer spectrum of [Fe4(imid)8(terpy)]x shows the expected number of lines (eight) coresponding to the overlap of four quadrupole doublets [2] (Figure 6.6). This is consistent with the X-ray single crystal difraction studies that show there are two unique tetrahedral, one unique trigonal pyramidal and one unique octahedral iron site, al of them distorted, in the framework of this compound (Figure 6.2). 218 6.2.4 MAGNETIC PROPERTIES DC magnetic susceptibilty and p^ f versus temperature (2 - 300 K) data on powdered samples of [Fe4(imid)g(terpy)]x in an applied magnetic field of 10 000 G are shown in Figure 6.7. As the temperature is lowered, the u«ff value decreases from 4.96 U\B at 300 K to 4.27 | I B at ~ 17 K. At about 16 K, u«ff increases abruptly, reaching a maximum at about 7 K before decreasing with decreasing temperature in the lowest temperature region. The behavior exhibited by % and u«ff (Figure 6.7) suggests antiferomagnetic exchange between the metal centers in [Fe4(imid)8(terpy)]x above a temperature of ~ 16 K, and a feromagnetic transition below this temperature. DC susceptibilty studies were also performed at 500 G for [Fe4(imid)g(terpy)]x. The trend of the % and u^f data versus temperature (2 - 50 K) are shown in Figure 6.8. The results show field dependence at low temperatures, particularly below ~ 12 K. The magnetic transition which is clearly seen in the 500 G data is comparable to that obtained from the 10 000 G data (Figure 6.7). A noticeable diference is that the onset temperature for the magnetic transition (temperature at which % and ji^ f  values start to increase) appears to be lower, ~ 12 K, at the lower field studied. This situation arises, perhaps, from the fact that the magnetic transition observed in [Fe4(imid)g(terpy)]x is not an abrupt one, but rather a gradual one. Therefore, it is not easily determined by DC susceptibilty studies. 219 0.8 6.0 ^ 0.6 i I O E E ® 0.2 t 0.0 V A A A A A A A A A A A A A A A A A , © © V - | 1 -5.5 5.0 1 4.5 =L 4.0 3.5 3.0 0 50 100 150 200 250 300 T ( K ) Figure 6.7 DC % and peff versus T at 10 000 G for [Fe4(imid)8(terpy)]„ The actual critical temperature of [Fe4(imid)g(terpy)]x was later confirmed by AC susceptibilty measurements, Mossbauer spectroscopy and zero-field-cooled magnetization studies (vide infra). 220 0 10 15 20 25 30 35 40 45 T(K) j 22 - 20 18 16 pp. 14 u 12 10 < - 8 6 % i 2 50 Figure 6.8 DC x and u«ff versus T at 500 G for [Fe4(imid)8(terpy)]x. AC magnetic susceptibilty measurements on [Fe4(imid)8(terpy)]x revealed a feromagnetic ordering at Tc = 6.35 K, which is determined as the x' maximum (Figure 6.9). The fact that x" (out of phase component) is diferent from zero at Tc confirms definitively that there is a non-zero net moment ground state, as was also determined from the medium and high field (static) DC susceptibilty studies mentioned above. 221 4 6 8 10 12 14 16 T(K) Figure 6.9 AC magnetic susceptibilty for [Fe4(imid)g(terpy)]x, HAC = 1 G, f = 125 Hz. The temperature of the onset of the magnetic transition in [Fe4(imid)g(terpy)]x was confirmed by the temperature dependence of the field-cooled magnetization (FCM), zero-field-cooled magnetization (ZFCM) and remnant magnetization (REM) studies (Figure 6.10). The FCM, measured by cooling the sample under a DC field of 50 G, shows an inflection around 7.5 K. According to the d(FCM)/dT plot the temperature of ~ 7 K can be considered asTc (Figure 6.10). As is typicaly observed, the ZFCM, 222 Figure 6.10 Plots of ZFCM, FCM and REM for [Fe4(imid)8(terpy)]x. #DC = 50 G. measured by cooling the sample in zero field, then warming up under a DC field of 50 G , is lower than the FCM at al temperatures below Tc [3]. The REM, obtained by cooling the sample in a 50 G field, then warming up in zero field, vanishes at ~ Tc again as is expected These results confirm that [Fe4(imid)8(terpy)]x behaves as a weak feromagnet below Tc = 7 K. This Tc value is in close agreement with that determined (Tc = 6.35 K) employing AC susceptibilty measurements (vide supra). 223 Additonal evidence for a feromagnetic transition in [Fe4(imid)8(terpy)]x comes from magnetization versus DC applied field plots performed at diferent temperatures (Figure 6.11). At 20 K (above Tc), the plot is linear from 20 000 G to zero field and extrapolates to zero magnetization at zero applied field. The plot shows much more curvature at a temperature near Tc (10 K) and at temperatures below Tc (2 K and 4.8 K), the plots are not linear and extrapolate to a negative magnetization (not shown in Figure 6.11) at zero applied field (see folowing discussion). o S u s s .2 « N • a * CJ S ex 14000 12000 i 10000 10000 20000 30000 40000 50000 60000 Applied Field (G) Figure 6.11 Plot of magnetization versus applied field at diferent temperatures for [Fe4(imid)g(terpy)]x. 224 A hysteresis loop is produced by cycling the applied field between +55 000 and -55 000 G at 4.8 K (Figure 6.12), as expected for a material exhibitng long-range feromagnetic ordering. From this hysteresis loop, a coercive field of ~ 5 G and a -300 -200 -100 0 100 200 300 Applied Field (G) Figure 6.12 Field dependence of magnetization at 4.8 K for [Fe4(imid)8(terpy)]x. Central portion of hysteresis loop shown. The data obtained on decreasing the applied field are shown as T while the data obtained on increasing the applied filed are shown as A. 225 remnant magnetization of 40 cm3Gmol"1, are obtained. A magnet is said to be soft or hard according to whether the coercive field is smal or large [3]. The very low value of the coercive field classifies [Fe4(imid)8(terpy)]x as a very soft molecule-based magnet. In a normal hysteresis loop (i.e., see Chapter 5, section 5.2.2.3) the down-field magnetization data approaching zero applied field are positive, while the up-field magnetization values are negative close to zero applied field. For [Fe4(imid)g(terpy)]x, however, quite diferent behavior is observed. The down-field magnetization data actualy cross the zero field line at negative magnetization values while the up-field data cross the zero field line at positve magnetization values. This results in a "crossing" of the down-field and up-field magnetization lines, generating a rare magnetization loop in the central part of the plot (Figure 6.12). Confirmation of this distinctive behavior was obtained in a separate experiment in which the applied field was cycled (-55 000 to +55 000) three times at 4.8 K. The results of this study, shown in Figure 6.13, confirmed that this unique magnetization loop is real and reproducible, since it appears in the same region in every one of the cycles caried out, and cannot be "removed' by repeatedly oscilating the applied field. A possible explanation for the distinctive magnetization loop of [Fe4(imid)8(terpy)]x is the folowing. There are four unique iron chromophores in this compound, which can be identified accordingly to their coordination number as 4, 4', 5 226 Applied Field (G) Figure 6.13 Field dependence of magnetization at 4.8 K for [Fe4(imid)8(terpy)]x. Central portion of hysteresis loop shown. and 6. The size of the magnetic dipole on each site wil  be diferent. For the sake of discussion we assume the dipole on site 5 is significantly smaler than those on the other three sites which, in turn are approximately the same. Assuming antiferomagnetic coupling, the connectivites require the spins in 4 and 4' to be paralel to each other, with the 5 and 6 site spins antiparalel to these as folows: 227 This arangement would give a net spin T. We now speculate that as the positve H (T) is reduced to zero al of the spins randomize (as expected in a material which is not a "magnet") except for the spin on 5, which retains a X preference (caused by slow relaxation). This would give a negative M at H = 0. As H is increased in the negative direction H (-1) the spins on al sites orient with the reverse direction to that shown above, giving a net spin i. Then, reducing the negative H (1) to zero al spins randomize except for the spin on 5, which retains the T preference (slow relaxing). This situation gives a positve M at H = 0. Support for the speculation above comes from Mossbauer studies. Preliminary results suggest that one of the sites undergoes magnetic hyperfine spliting owing to slow paramagnetic relaxation (not ordering) in the decreasing temperature range 77 to 7 K. Below the later temperature al four sites are hyperfine split. The magnetic properties of [Fe4(imid)g(terpy)]x are quite diferent from those of the other molecule-based magnets described in this thesis. While the primary 228 mechanism of exchange is antiferomagnetic, as in the other systems, in this case the anomalous magnetic behavior leading to a phase transition below 7 K appears to arise from the coupling of dipoles of diferent strengths, a form of ferimagnetism. Moreover, rather than long-range ordering of al spins sites below Tc, three of the sites undergo rapid relaxation with the fourth retaining orientation in the field thus generating the net magnetization at zero applied field. It is hoped that more detailed study and analysis of the Mossbauer spectra of this material (curently ongoing) wil permit us to identify the slow relaxing site and thereby remove some of the speculation surounding the unique magnetic behavior of this material. 6.3 SUMMARY AND CONCLUSIONS The reaction of ferocene with molten imidazole and terpyridine yields green crystals of compositon [Fe4(imid)8(terpy)]x. The structure of this material has been determined by single crystal X-ray difraction studies and is shown to consist of novel 2-D wrinkled sheets formed by four-coordinate (two unique ions of this type), five-coordinate and six-coordinate iron(I) ions linked by single bridging imidazolates. The six-coordinate metal ions are additionaly coordinated by terpyridine ligands that occupy positons between the sheets, isolating the sheets from each other. The presence of four unique iron sites in [Fe4(imid)g(terpy)]x was confirmed by Mossbauer spectroscopy. 229 In the formation of this unique 2-D iron(I) imidazolate polymer, some degree of structural control has been introduced by the tridentate ligand, terpy, which occupies coordination sites on the metal ions and provides steric constrains, thus preventing spatial extension of the polymeric structure to three-dimensions. DC magnetic susceptibilty measurements reveal antiferomagnetic interactions between metal centers above a temperature of ~ 12 K and the onset of long range feromagnetic ordering below this temperature. A transition to a feromagnetic state below these temperatures was seen in both DC and AC susceptibilties and zero-field-cooled magnetization studies. The critical temperature, Tc ~ 6.5 K, for the magnetic transition was determined from these later studies. [Fe4(imid)g(terpy)]x represents another rare example of a 2-D coordination polymer exhibitng weak feromagnetism at low temperatures. The compound shows a unique negative hysteresis loop and speculation to account for this points to the possibilty of a form of feromagnetic coupling combined with slow paramagnetic relaxation at low temperatures of one of the four unique metal centers. Additonal insights into this mater are expected from ongoing Mossbauer studies. Regardless of the mechanism, this material does exhibit net magnetization at zero applied field (MOTI = 40 cm3GmorI) classifying it as a weak molecule-based magnet. Its coercive field of ~ 5 G classifies it as a very soft magnet. 230 References 1. P.J. Hagrman and J. Zubieta. Inorg. Chem. 39, 5218 (2000). 2. N. Greenwood, T. C. Gibb. Mossbauer Spectroscopy, Chapman and Hal Ltd. London, 1971. 3 O. Kahn. Acc. Chem. Res. 33 (10) 647 (2000). 231 Chapter 7 POLYBIS(l-METHYL-2-THIOIMIDAZOLATO)IRON(II). A ONE-DIMENSIONAL MATERIAL EXHIBITING LONG-RANGE MAGNETIC ORDERING 7.1 INTRODUCTION As described before, metal polymers formed with bridging imidazolate ligands are usualy of high dimensionality (2-D or 3-D) as a result of the 1,3-positoning of the N-donor atoms of the imidazolate moiety, which alows the formation of single bridges between metal centers. However, with the N-l positon blocked with a methyl substiuent, as it is in the case for the ligand precursor l-memyl-2-thiolimidazole (1-Me-2-SH-imid), deprotonation of the thiol functionality would generate a bridging ligand, 1-Me-2-S-imid, capable of forming double bridges between Fe(U) centers in a rod like polymer [1]. Previous studies had revealed this type of bonding mode for this ligand in the dimeric molecules [Me2Ga(l-Me-2-S-imid)]2 and [Mo(Ti3-C3H5)(CO)2(l-Me-2-S-imid)]2 [2]. The rod-like 1-D structural motif was observed in the material characterized here, [Fe(l-Me-2-S-imid)20.5Cp2Fe]x. Interestingly, the phenomenon of long-range 232 feromagnetic ordering below a critical temperature, previously seen only in 1,3-diazolate complexes with 2-D and 3-D extended latices (Chapters 3, 4, 5 and 6), is observed in this unique Fe(I) coordination polymer. An article containing most of the results discussed in this chapter has been published [1]. 7.2 RESULTS AND DISCUSSION 7.2.1 SYNTHESIS, PHYSICAL AND THERMAL PROPERTIES Details of the synthesis of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x are described in Chapter 9, section 9.2.1.5. Folowing the success of the previously reported reactions of ferocene with imidazole and 2-methyl imidazole, which generate crystaline polymers [3, 4], the synthetic method involving ferocene and the molten ligand precursor was utilzed here. The desired compound was isolated as golden needle-like crystals. [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x appeared to be fairly air-stable and could be handled briefly under normal atmospheric conditons. The compound is insoluble in common organic solvents and water. The complex is non-volatile. Thermal gravimetric analysis (35 °C to 800 °C) measurements were made and the TGA plot is shown in Figure 7.1. These results show the complex to be thermaly stable to 197 °C. Decompositon with continuous weight loss occurs from 197 to 800 °C with a total 233 200 400 600 Temperature ( °C ), Figure 7.1 TGA plot for [Fe(l -Me-2-S-imid)20.5Cp2Fe]x 800 weight loss of 69% of the initial mass. This compares favorably to a 65% loss that would occur if only FeS remains at 800 °C. 234 7.2.2 X-RAY CRYSTALLOGRAPHY The structure of a section of the polymer chain is shown in Figure 7.2, and a view of the lattice, almost paralel to the polymer chain, is depicted in Figure 7.3. Crystalographic data, atom coordinates, selected bond lengths and angles appear in Appendix I, Tables 1-12 and 1-13. The structure consists of chains of distorted tetrahedral iron(I) ions double-bridged by the l-memyl-2-thioimidazolate ligands, giving rise to eight-membered rings linked by the Fe ions in a pseudospiro conformation (Figure 7.2). The ligands bind through the unsubstiuted nitrogen (N2 in Figure 7.2) and the sulfur atoms and orient along the chain in a manner that leads to distinctive FeN4 and FeS4 chromophores which alternate throughout the polymer chains. These structural characteristics produce a rod-like shape to the polymeric chain (Figure 7.3 ). The Fe(2)-N bond distances are significantly shorter at 2.054 A than the Fe(l)-S bonds at 2.368 A. The S-Fe(l)-S bond angles are close to tetrahedral, ranging from 108.32° to 110.05°, while the N-Fe(2)-N angles are farther from tetrahedral ranging from 104.91° to 119.05°. Molecules of ferocene, one for every two repeating units in the chain, are trapped between the polymer chains (Figure 7.3). These molecules cannot be removed by thermolysis without decompositon of the polymer (See Figure 7.1). 235 Figure 7.2 Molecular structure of the polymer chain of [Fe( 1 -Me-2-S-imid)2-0.5Cp2Fe]x showing the atom numbering scheme; 33 % probabilty thermal elipsoids are shown. (Hydrogen atoms are omited). 236 Figure 7.3. View of the crystal structure of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x down the c axis. 50 % thermal elipsoids are shown. 237 7.2.3 MAGNETIC PROPERTIES The DC magnetic susceptibilties, %, and %T versus temperature, on powdered samples of [Fe(l-Me-2-S-imid)20.5Cp2Fe]x in an applied magnetic field of 500 G, are shown in Figure 7.4. As the temperature is lowered from 300 K, the %T value decreases. At about 8 K, %T increases abruptly, reaching a maximum at about 5 K before decreasing with decreasing temperature in the lowest temperature region. The behavior suggests antiferomagnetic exchange between the metal centers above the critical temperature, Tc, of 5 K (this temperature confirmed by AC susceptibilty and Mossbauer measurements (vide infra)), and a feromagnetic transition below this temperature. AC susceptibilty measurements indicate the feromagnetic ordering at Tc = 5 K, which is determined as the average temperature between the %' plot maximum at 5.20 K and the %" plot maximum at 4.85 K [3] (Figure 7.5). The fact that / " (out of phase or imaginary component) is diferent from zero at Tc confirms unequivocaly that there is a non-zero net moment ground state. Therefore, the near zero field results reveal the presence of a long-range feromagnetic order below Tc in concordance with the DC susceptibilty studies discussed above. 238 ^ 1.25 0.00 50 100 150 200 Temperature (K) 250 300 Figure 7.4 DC % and yj versus temperature plots at 500 G for [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x The line is from theory as described in the text. Support for a feromagnetic transition comes from magnetization versus applied field plots at diferent temperatures (Figure 7.6). At 15 K (above Tc), the plot is linear to over 20 000 G and extrapolates to zero magnetization at zero applied field. At temperatures below Tc, the plots are not linear and extrapolate to yield net magnetization at zero applied field. 239 Figure 7.5 AC magnetic susceptibilty for [Fe(l-Me-2-S-imid)20.5Cp2Fe]x, HAC = 1 G, f = 125 Hz. Cycling the applied field between +55 000 and -55 000 G at 4.8 K generates a hysteresis loop (Figure 7.7), as expected for a material exhibitng long-range feromagnetic ordering. From this hysteresis loop, a coercive field of 40 G and a remnant magnetization of 190 cm3 G mol"1, are obtained. 240 4000 A 3500 A - r 3 0 0 0 o S o -I £ c o -I—I 1 o a 2500 10000 "i 1 r 20000 30000 40000 50000 60000 Applied Magnetic Field (G) Figure 7.6 Plot of magnetization versus applied field at three temperatures for [Fe( 1 -Me-2-S-imid)20.5Cp2Fe]x. 241 -400 -300 -200 -100 0 100 200 300 400 Applied Magnetic Field (G) Figure 7.7 Field dependence of magnetization at 4.8 K for [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x. Central portion of hysteresis loop shown. Magnetic susceptibilties were measured as a function of temperature at 10 000 and 50 000 G as wel as at 500 G (Figure 7.8). The results show field dependence at low temperatures, particularly below Tc. The magnetic transition which is clearly seen in the 242 500 G data is much less pronounced in the 10 000 G data and is absent at 50 000 G. The feromagnetic ordering is clearly repressed by large applied fields. Temperature (K) Figure 7.8 Plot of %T versus temperature at three values of applied field for [Fe( 1 -Me-2-S-imid)20.5Cp2Fe]x. 243 The magnetic properties of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x paralel closely those of [Fe3(imid)6(imidH)2]x [4], [Fe(4-imidazol-acetate)2]x [5], and [Fe(2-Me-imid)20.13Cp2Fe]x [6], compounds for which the primary exchange process is considered to involve antiferomagnetic coupling between iron centers with a canting of spins. A spin-canted structure for the compound studied here is also supported by the fact that the highest magnetization reached (3960 cm3 G mol"1 at 2 K and 55 000 G) is significantly smaler than the theoretical saturation value (22 300 cm3 g mol"1) [7]. Additonal support for the canted spin structure comes from structural data. These show a feature characteristic of such systems, that of a systematic alternation of the relative orientation of neighboring metal chromophores [4]. As a measure of this, the dihedral angle between the S(l)-Fe(l)-S(l)c and N(l)d-Fe(2)-N(l)e planes is 27.6°. Therefore, antiferomagnetic coupling between neighboring metal centers along the chain can occur with imperfect antiparalel alignment of spins, leading to residual spins on the chains. Feromagnetic ordering of the residual spins generates long-range three-dimensional magnetic ordering and spontaneous magnetization at low temperatures. It should be noted that the chains in [Fe(l-Me-2-S-imid)20.5Cp2Fe]x are isolated (Figure 7.3), and therefore, any interchain interaction cannot be mediated by bonding interactions. This contrasts with the situation for [Fe3(imid)6(imidH)2]x [4] and [Fe(2-Me-imid)2-0.13Cp2Fe]x [6] where covalent bonding interactions connect the paramagnetic centers in three dimensions and for [Fe(4-imidazol-acetate)2]x [5], where hydrogen-bonding interactions connect sheets of covalently linked metal centers. Such considerations may play an important role in determining the magnitude of the coercive 244 field in these systems, as [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x exhibits the smalest coercive field of the four compounds considered here. Further evidence in support of the primary antiferomagnetic intrachain coupling in [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x can be gathered by examining fits of the magnetic susceptibilty in the high temperature region to the expression derived employing an isotropic Hamiltonian of the form H= -US1S2 for a linear chain of antiferomagneticaly coupled S - 2 centers [8], as was previously done for [Fe(pz)2]x (Chapter 2). That equation is where JC = \J\ I kT, J is the exchange coupling constant and g is the Lande factor. By employing susceptibilty data obtained at the three diferent fields (500,10 000 and 50 000 G), no satisfactory fits were obtained when data below 50 K were included. Good fits were, however, obtained for data in the range 50-300 K as is ilustrated for the 500 G %T data in Figure 7.4. The lack of agreement between theory and experiment at low temperatures is not surprising since the model does not accommodate the efects of residual spin due to spin canting or interchain interactions, both of which are more pronounced at low temperatures. The theory line shown in Figure 7.5 was calculated with 245 -J= 3.92(6) cm"1 and g = 2.21(1) (F= 0.00033). The model used is limited by the fact it employs a single g value, while the structure of the compound requires diferent g values for the FeN4 and FeS4 chromophores. The best-fit g value of 2.21 presumably approximates the average g for the system. The strength of the antiferomagnetic coupling in this compound, as judged by the magnitude of - J, seems to be slightly greater than that seen in [Fe3(imid)6(inudH)2]x [4] and [Fe(2-meimid)20.13Cp2Fe]x [6] for which - / values of 2.3 and 2.75 cm"1, respectively, have been reported. The alternation in chromophore type and, therefore, g value along the chain should generate important magnetic consequences for this antiferomagneticaly coupled system. The size of the individual magnetic dipoles wil alternate along the chain, and even perfect antiparalel alignment between neighbors wil lead to a residual moment on the chain, which can be considered as an example of ferimagnetism. This same phenomenon was described in earlier chapters as a possible cause of the residual spin on the latices of [Fe2(imid)4(bipy)]x (Chapter 5) and [Fe4(imid)8(terpy)]x (Chapter 6). While this probably contributes to the observed magnetic properties of the system, a simple calculation indicates it cannot be the sole source of the residual chain magnetization. The saturation magnetization, Ms, for a S =2 center is [7] Ms = Ng$S 246 The net saturation magnetization, A/net, for a chain of perfect antiparalel coupled 5 = 2 metal centers with regularly alternation of g values is, per mole of metal center, M n e t = 1 / 2 A M s = i V p A g Inserting 3960 cm3 G mol"1 (the magnetization measured at 2 K and 55 000 G) for Mnet in the above equation yields a Ag value of 0.71. Hence, an unrealisticaly large diference in g values for the FeN4 and FeS4 chromophores would be required to invoke ferimagnetism to account for the largest magnetization (not even saturation magnetization) observed. Therefore, spin canting is likely the primary source of the residual magnetization on the chains, although ferimagnetism cannot be ruled out as a contributing factor. This later phenomenon is further discussed in the folowing section. 7.2.4 MOSSBAUER SPECTROSCOPY The room temperature Mossbauer spectrum of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x coresponds to three overlapping quadrupole doublets (Figure 7.9), which are consistent with the X-ray single crystal difraction studies that indicate equal population of distorted four coordinated (tetrahedral) FeN4 and FeS4 sites and half as many ferocene molecules 247 {FeSJ 2.0 H 1 1 1 1 1 1 1 1 1 1 - 2 - 1 0 1 2 Velocity Relative to Fe (mm s ) Figure 7.9 Mossbauer spectrum of [Fe( 1 -Me-2-S-imid)2-0.5Cp2Fe]x at 293 K. between the polymeric chains (Figures 7.2 and 7.3). The coresponding Mossbauer parameters (isomer shift (8), quadrupole spliting (AE) for FeN4, FeS4 and Fe(Cp)2 sites at 293 K are (0.8 mms"1, 1.42 mms"1), (0.64 mms"1, 3.03 mms"1) and (0.45 mms"1, 2.26 mms"1) respectively. These are fairly typical values for iron(JJ) FeN4 and FeS4 and for the S = 0 ferocene [9]. The results at 77 K are (0.94 mms"1, 2.52 mms"1), (0.76 mms"1, 3.30 mms"1) and (0.52 mms"1,2.41 mms"1) respectively (Figure 7.10). 248 Velocity relative to Fe (mm s"1) Figure 7.10 Mossbauer spectrum of [Fe( 1 -Me-2-S-imid)2-0.5Cp2Fe]x at 77 K. As mentioned in Chapter 3, section 3.2.3, the spliting value ranges can often overlap for four, five and six coordination environments (even with similar ligands) depending on the degree of distortion of the local coordination environment [9-11]. On the other hand, isomer shifts have been found to be quite sensitve to coordination number for fixed spin states and similar ligation [9-14]. For tetrahedral FeS4 chromophores, there is even further reduction of the isomer shift relative to FeN4 to 249 values ranging from ~ +0.60 mms" to 0.75 mms" owing to the increased covalence of the sulfur ligation environment [15, 16]. Thus, the present 8 values in conjuction with the available literature results appear to unequivocaly farther confirm the four coordinate, tetrahedral FeN4 and FeS4, nature of the high spin iron(I) of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x and the presence of ferocene. As expected, the Mossbauer spectrum of the magneticaly ordered phase of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x (Figure 7.11) coresponds to the overlap of two Zeeman paterns and the wel known (8 = 0.55 mm s"1, AE = 2.47 mm s"1) quadrupole doublet of ferocene. The extreme transition pairs of the FeN4 and FeS4 hyperfine paterns, and the doublet of ferocene, are designated with arows in Figure 7.11. The FeS4 site exhibits a smaler temperature dependence than FeN4 (< +0.3 mm s"1 vs +1.1 mm s"1) over the decreasing temperature range 293 K to 77 K, and has an absolute value of quadrupole spliting larger than that for FeN4 chromophore by ~ 0.8 mm s"1 at 77 K. These observations indicate a significantly larger low symmetry ligand; field component spliting (A) of the ground 5E term of the regular tetrahedron to the nondegenerate 5B and 5A states of distorted FeS4 and, in addition, a concomitant greater quenching of the orbital contributions to the moment for FeS4 than FeN4 via the 250 { [Fe(Cp)2] 1 1 1 • 1 -5 0 5 Velocity Relative to Fe (mm s"1) Figure 7.11 Mossbauer spectrum of [Fe( 1 -Me-2-S-imid)2-0.5Cp2Fe]x at 4.2 K. usual first order and spin-orbit coupling efects [9, 17]. These efects can lead to significant diferences in the efective g values for the FeN4 and FeS4 chromophores with g likely closer to 2 for the later. These arguments support the possibilty of the existence of a subtler intrinsic "intra-chain ferimagnetism", which arises from g factor modulation (alternation) along the polymeric chain contributing, thus, to the weak feromagnetism exhibited by [Fe(l-Me-2-S-imid)20.5Cp2Fe]x. 251 7.3 SUMMARY AND CONCLUSIONS [Fe(l-Me-2-S-irmd)2-0.5Cp2Fe]x has a rod-like polymer structure in which iron ions are double bridged by l-methyl-2-thioimidazolate ligands. The metal chromophores alternate along the chain between FeS4 and FeN4. DC magnetization studies on powdered samples reveal a net magnetization at zero field and temperatures below 8 K. AC magnetization studies confirm the long-range feromagnetic order in this system and permit an accurate evaluation of the critical temperature, Tc = 5 K. Magnetization versus applied field studies at 4.8 K generate a hysteresis loop with a remnant magnetization of 190 cm3GmoT1 and a coercive field of 40 G. Antiferomagnetic intrachain coupling with spin canting generating residual spin on the chains that undergo long-range feromagnetic ordering is believed to be a probable cause of the magnetic behaviour observed for this material. An alternative explanation for the weak feromagnetism exhibited by [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x, which is the alternation of the g values along the polymer chain, has been considered folowing the Mossbauer spectroscopy results, in particular the temperature dependence and limiting values of the quadrupole splitings of the paramagnetic ferous sites of this system. It is concluded that both mechanisms are probably operative here. [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x possesses very unique and complex magnetic properties which may require further study to make a definitive assessment of the dominant phenomenon, spin canting versus g factor diferences, that generates the 252 magnetic ground state of this system. Determination of the magnetic structure by neutron difraction studies could provide new insights to this mater; however, suficiently large, deuterated, single crystal samples of [Fe(l-Me-2-S-imid)2-0.5Cp2Fe]x would be required for this work. 253 References 1. S. J. Retig, V. Sanchez, A. Storr, R. C. Thompson, and J. Troter, Inorg. Chem. 38, 5920(1999). 2. D. A. Cooper, S. J. Retig; A. Storr, and J. Troter. Can. J. Chem. 64, 1643 (1986). 3. O. Kahn. Acc. Chem. Res. 33 (10) 647 (2000). 4. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter. J. Am. Chem. Soc. 119, 8675 (1997). 5. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson, and J. Troter. Can. J. Chem. 77,425 (1999). 6. M. A. Martinez-Lorente, V. Petrouleas, R. Poinsot, J. P. Tuchagues, J. M. Savariault, and M. Drilon, Inorg. Chem. 30, 3587 (1991). 7. R. L. Carlin. Magnetochemistry. Springer-Verlag. Berlin. 1986. 8. W. Hiler, J. Strahle, A. Datz, M. Hanack, W. F. Hatfield, and P. Gutlich, Am. Chem. Soc. 106, 329 (1984). 9. N.N. Greenwood and T. C. Gibb. Mossbauer Spectroscopy. Chapman and Hal Ltd. London. 1971. 10. B. W. Dockum , and W. M. Reiff. Inorganica Chimica Acta. 35, 285 (1979). 11. F. F. Charon, and W. M. Reiff. Inorg. Chem. 25,2786(1986). 12. W. M. Reiff, N. E. Erickson, and W. A. Baker. Inorg. Chem. 8,2119 (1969). 13. D. Sedney, M. Kahjehnassiri, and W. M. Reiff. Inorg. Chem. 20, 3476 (1981). 14. B. W. Dockum, and W. M. Reiff. Inorganica Chimica Acta. 120, 61 (1986). 15. W. M. Reiff, I.E. Grey, A. Fan, Z. Eliezer, and H. Steinfink. J. Solid State Chem., 13, 32 (1975). 16. K. K. Rao, M. C. W. Evans, R. Cammack, D. O. Hal, C. L. Thompson, P. J. Jackson, and C. E. Johnson. Biochem, J. 129, 1063 (1972). 254 R. M. Golding, K. F. Mok, and J. F. Duncan. Inorg. Chem. 5, 774 (1966). 255 Chapter 8 GENERAL SUMMARY AND SUGGESTIONS FOR FUTURE WORK 8.1 GENERAL SUMMARY This dissertation describes an investigation of the synthesis, structural and physical characterization, and in particular the magnetic properties, of one-, two- and three-dimensional transition metal polymers involving pyrazolate and, mainly, imidazolate and imidazolate derived ligands as bridging species. 1,2-Diazolates (pyrazolates) in combination with divalent paramagnetic transition metals typicaly form 1-D chain structures with double-bridging azolate ligands, materials which show short-range antiferomagnetic interactions between metal centers. A new compound in this series, [Fe(pz)2]x, has been synthesized and structuraly and magneticaly characterized. As expected, the structure of this 1-D polymer resembles a chain, due to the double-bridging characteristic of the pyrazolate ligands. Also as expected, this compound exhibited short-range antiferomagnetic exchange, with -J 0.591(5) cm"1, a value that represents the weak antiferomagnetism occuring between the tetrahedral metal centers in this compound. In contrast, when 1,3-diazolates ligands (imidazolates) were utilized, extended 2-D and 3-D latices with singly bridging azolates between the metal centers were obtained. 256 Many of these materials were shown to exhibit, at low temperatures, net magnetization in zero applied field, a property which classifies them as molecule-based magnets. The reaction of ferocene with molten 4-azabenzimidazole resulted in the isolation of a 3-D iron(n) coordination polymer. This compound was obtained as a crystaline material and its structure determined by single crystal X-ray difraction. The structure involves a unique single (non-interpenetrating), totaly covalent, diamondoid aray of Fe(I) centers, which are single bridged by the 1,3-diazolate ligands into a 3-D aray of fused rings each containing six distorted tetrahedral Fe(I) centers. Replacing ferocene by cobaltocene produced the cobalt analogue as a microcrystaline compound. X-ray powder difraction studies revealed the two compounds are isomorphous. Variable temperature DC magnetic susceptibilty studies on these compounds reveal antiferomagnetic coupling between neighboring metal ions above about 20 K for the iron compound and 11 K for the cobalt analogue. Below these temperatures both materials show long-range feromagnetic ordering. This behaviour suggests canted-spin structures are present in these compounds. By cycling the DC applied field at 4.8 K, hysteresis loops with Mrem of 2100 and 22 cm Gmol" and Hcoer of 80 and 100 G were obtained for [Fe(4-abimid)2]x and the cobalt analogue, respectively. AC magnetic susceptibilty studies confirmed the magnetic transition occuring in [Fe(4-abimid)2]x. Mossbauer spectroscopy studies in this compound revealed a feromagnetic transition at 18 K and the possibilty of a structural phase transition at low temperature. 257 The cobalt(Il) compounds, [Co(imid)2]x, [Co(benzimid)2]x and [Co3(imid)6(imidH)2]x, al exhibited a sudden increase in their magnetic moments below temperatures of ~ 16, 13 and 15 K, respectively, and magnetic hysteresis behavior at temperatures lower than these. X-ray powder difraction studies showed [Co3(imid)6(imidH)2]x to be isomorphous with the iron analogue which was reported previously [1], and which has a 3-D structure with single-bridging imidazolate ligands. The structures of the other two cobalt compounds are not known with certainty but are thought to be 3-D as wel. Thus, according to their magnetic properties, these three cobalt compounds are considered to belong to the class of materials known as molecule-based magnets. In addition, two other cobalt imidazolate compounds, [Co(2-meimid)2]x and [Co(4-meimid)2]x, where studied. However, these two compounds, which likely have 3-D structures also, show only weak antiferomagnetic coupling with no strong evidence of a transition to a feromagnetic state at low temperatures. The lack of definitive structural information for these cobalt polymers precluded further explanation of their magnetic behaviours. Other transition metal imidazolate compounds synthesized included nickel(H) benzimidazolate, the characterization of which showed it to be the first reported molecule-based magnet containing nickel(I) ions. This compound exhibits net magnetization at zero applied field, and ZFCM and FCM curve shapes that reveal long-range feromagnetic ordering below a Tc of ~ 6.5 K. In addition, three copper(I) imidazolate compounds, [Cu(2-meimid)2]x, [Cu(benzimid)2]x and 258 [Cu(4,5-dichloroimid)2]x, were found to be molecule-based magnets. This was revealed by DC magnetic susceptibilty studies, that show an increase in both % and M«f  a t temperatures of -25, 15 and 14 K, respectively. Also, the presence of a hysteresis loop in field-dependent magnetization studies at low temperature as wel as ZFCM-FCM experiments confirmed the existence of a magnetic transition to a long-range feromagnetic state for these three copper compounds. Two additonal coppertTI) imidazolates, [Cu(imid)2]x and [Cu(4-meimid)2]x, were prepared. These showed only weak antiferomagnetic behaviour and gave no evidence for long-range magnetic order. Eforts to modify the molecular dimensionality, and hence the magnetic properties, of transition metal imidazolates by incorporating neutral chelating ligands into the latice resulted in the formation of [Fe2(imid)4(bipy)]x and its cobalt analogue. Single crystal X-ray difraction studies on [Fe2(imid)4(bipy)]x show a structure involving double layer sheets of iron ions single-bridged by imidazolate ligands. Four- and six-coordinated ions alternate in the lattice, the later ions being coordinated by the bipy ligands in additon to the bridging imidazolates. This capping of iron centres by bipy controls the dimensionality, resulting in the 2-D polymer. [Fe2(imid)4(bipy)]x is unique in showing two structural phase transitons at ~ 151 K and ~ 135 K, which have been characterized by a combination of single crystal X-ray-difraction, DC and AC magnetic susceptibilty and Mossbauer spectroscopy studies. In addition, [Fe2(imid)4(bipy)]x shows a sharp increase in the DC magnetic susceptibilty below 11 K, and a non-zero AC out-of-phase magnetic susceptibilty below that temperature, indicative of a transition to a 259 feromagneticaly ordered state at low temperatures. Variable temperature DC magnetic susceptibilty studies on [Fe2(imid)4(bipy)]x and the cobalt analogue revealed antiferomagnetic coupling between neighbouring metal ions above Tc (11 and 13 K respectively). Below Tc both materials exhibit long-range feromagnetic ordering. The presence of canted-spin structures in both compounds is suggested by this behaviour; however, a novel type of ferimagnetism, due to the systematic alternation of four- and six-coordinate chromophores has also been invoked as a possible mechanism for the observed magnetic behaviour. [Fe4(imid)8(terpy)]x has a novel 2-D extended structure in which "wrinkled" sheets of four-coordinate (two unique centers of this type), five-coordinate and six-coordinate iron ions are linked by single-bridging imidazolate ligands. The six -coordinate metal centers are additionaly coordinated by terpy ligands that occupy positons between the sheets, isolating the sheets from each other. The presence of four unique iron sites was confirmed by Mossbauer spectroscopy. DC magnetic susceptibilty measurements revealed antiferomagnetic interactions between metal centers above Tc and a transition to a feromagnetic state below this temperature. A Tc of 6.5 K, was confirmed by AC magnetic susceptibilty and ZFCM-FCM-REM studies. Magnetic hysteresis studies revealed the magnetic properties of this system to be unique among the imidazolate systems studied here. Negative magnetizations at zero field were explained by invoking a mechanism in which three of the four unique sites act as fast relaxing 260 paramagnets with the fourth acting as a slow relaxing paramagnet and generating the observed magnetization in zero field. [Fe(l-Me-2-S-imid)2-0.5 Cp2Fe]x has a unique rod-like polymer structure in which iron ions are double bridged by l-methyl-2-thioimidazolate ligands. FeS4 and FeN4 metal chromophores alternate along the chains. Magnetic studies showed a net magnetization at zero field and temperatures below 8 K. Magnetization versus applied field studies at 4.8 K generate a hysteresis loop with a remnant magnetization of 190 cm3Gmol"1 and a coercive field of 40 G. Antiferomagnetic intrachain coupling with spin canting was proposed to generate residual spin on the chains that undergo long-range feromagnetic order below ~ 8 K. Due to the alternation on the FeS4 and FeN4 chromophores throughout the chains, a possible novel type of ferimagnetism has also been proposed to contribute to the magnetic ordering exhibited by this compound. Mossbauer spectroscopy provided unambiguous evidence for the presence of the two iron chromophores and ferocene as wel as the magnetic transition in this compound. This material provides a relatively rare example of a molecule-based magnet in which the covalent connectivites in the latice are 1-D. In summary, a relatively new family of molecule-based magnets incorporating imidazolate-based ligands as mediators of magnetic exchange between the metal ions has been established in this work. It has been demonstrated that heterocyclic azolate ligands with two donor nitrogens separated by a single carbon in the ring (as in the imidazolate 261 ion) wil form single ligand bridges and extended structures. Moreover these ligands wil create a bridge geometry that leads to a systematic alternation in the relative orientation of neighboring chromophores in the lattice, a situation that can produce significant spin canting and, as a consequence, long-range feromagnetic ordering at low temperatures. Examples now exist of iron(I) complexes incorporating imidazolate-based ligands which, as confirmed by single crystal X-ray difraction, have 1-D {[Fe(l-Me-2-S-imid)2- 0.5Cp2Fe]x}, 2-D {[Fe2(imid)6bipy)]x and [Fe4(imid)8(terpy)]x}, and 3-D {[Fe(4-abimid)2]x} extended covalent latices. All of these compounds exhibit magnetic properties that classify them as molecule-based magnets. 8.2 SUGGESTIONS FOR FUTURE WORK In order to learn more about the possible low-temperature structural phase transition occuring in [Fe(4-abimid)2]x a single crystal X-ray difraction study at He temperatures is required. This may alow a beter understanding of the magnetic properties of this diamondoid compound. In addition, neutron difraction studies would contribute to the complete understanding of the magnetic properties. In regard to the cobalt(I), nickel(I) an copper(I) imidazolate polymers, further synthetic atempts to produce single crystals suitable for X-ray difraction studies would be worthwhile. 262 Incorporation of other chelating ligands (such as 5,5'-dimethyl-2,2'-bipy) into metal imidazolate structures to obtain compounds similar to [Fe2(imid)4(bipy)]x and [Fe4(imid)g(terpy)]x is encouraged to try to generate extended 2-D systems with subtle structural modifcations. This would alow the investigation of the efect of such structural changes on magnetic properties. Moreover, especialy for materials similar to [Fe2(imid)4(bipy)]x, the study of potential structural phase transitons in these systems would be of significant interest. Due to the complex magnetic properties found in [Fe(l-Me-2-S-imid)2-0.5 Cp2Fe]x, and in order to get more insight about the predominant phenomenon (spin-canting versus g factor diferences) responsible for the long-range feromagnetic ordering exhibited by this material, neutron difraction studies could be atempted. This would require the synthesis of large, deuterated, single crystals of the compound. Finaly, the pyrazole complexes [Cu(pzH)4(N03)2], [Co(pzH)4(NC»3)2], and [Ni(pzH)4(N03)2] as wel as [Co(2-(2-py)benzimidH)3(C104)2] (py = pyridine) were synthesized in the course of this work. These materials have the potential to be used as building blocks for the formation of hetero-bimetalic ferimagnetic 1-D, 2-D or 3-D networks [2]. 263 References 1. S. J. Retig, A. Storr, D. A. Summers, R. C. Thompson and J. Troter, J Amer. Chem. Soc. 119, 8675 (1997). 2. F. Lambert, J-P. Renault, C. Policar, I. Morgenstern-Badarau, and M. Cesario. Chem. Commun. 35 (2000). 264 Chapter 9 EXPERIMENTAL 9.1 INTRODUCTION The experimental details of this work including synthetic aspects and materials employed are described here. The instruments and methods employed in the physical characterization of the compounds studied are also described. 9.2 SYNTHESES The chemicals used in this thesis were of reagent grade and were used without further purification. The commercial sources for most of chemical reagents utilzed in the syntheses caried out in the present work are given in Table 9.1. The majority of compounds prepared here are air stable, hence they were synthesized without special precautions. Those compounds that exhibited sensitivity to oxygen or moisture were prepared using Schlenk techniques and handled in a Vacuum Atmospheres Corporation Model HE 43-2 Dri-Lab glovebox under a dinitrogen atmosphere. All solvents used with these air-sensitive compounds were free of water and deoxygenated. The solvents were dried using the folowing methods: acetonitrile was refluxed with phosphorous pentoxide and distiled under dinitrogen; benzene was refluxed with potassium metal 265 and distiled under dinitrogen; xylenes were refluxed with sodium and benzophenone and distiled under dinitrogen. Table 9.1 Commercial source of most chemical reagents employed in this thesis (continued overleaf). Compound Commercial source Acetonitrile Aldrich 4-Azabenzimidazole Aldrich Benzene Aldrich Benzimidazole Eastman 4,5-dichloroimidazole Aldrich Cobaltocene Strem Cobalt chloride hexahydrate Fisher Cobalt nitrate hexahydrate Malinckrodt Copper sulphate pentahydrate J. T. Baker Ferocene Strem imidazole Aldrich 2-mercapto-1 -methylimidazole Lancaster 2-methylimidazole Lancaster 266 4-methylimidazole Aldrich ! 3,5-dimethylpyra2ole Aldrich | Nickel nitrate hexahydrate Malinckrodt • Phosphorous pentoxide BDH Potassium hydroxide AnalaR ; Pyrazole Aldrich | 2-(2-pyridyl)benzimidazole Aldrich 2,2'-l>ipyridine Matheson, Coleman & Bel 2r2':6',2"-terpyridine Matheson, Coleman & Bel Sodium hydroxide AnalaR Xylenes Aldrich 9.2.1 LRON(I) AZOLATE POLYMERS 9.2.1.1 Polybis(pyrazolato)iron(n), [Fe(pz)2]x Ferocene (0.3 g, 1.6 mmol), and pyrazole (0.2 g, 2.9 mmol) were mixed in a Carius tube, and then sealed under vacuum. The mixture was heated to 145 °C for 5 days. During that time, product in the form of needles crystalized from the orange solution. After cooling to room temperature, the Carius tube was opened under a dinitrogen atmosphere. The product was extracted by washing the solid obtained 267 thoroughly with acetonitrile. It was isolated as red-brown air-sensitive crystals suitable for X-ray crystalographic analysis. Yield 0.11 g (37 %). Analysis calculated for C6H6FeN4: C37.9;H3.2;N29.5; found: C37.7;H3.4;N29.4. 9.2.1.2 Poly-2,2'-bipyridinetetrakis(imidazolato)dikon(n), [Fe2(imid)4(bipy)]x. Ferocene (0.3 g, 1.6 mmol), imidazole (0.2 g, 2.9 mmol) and excess 2,2-bipyridine (0.5 g, 3.2 mmol) were placed in a Carius tube which was sealed under vacuum. The tube was heated to 130 °C for 2 days. Under those conditons, the original orange solution of ferocene in molten ligand became a mixture of dark crystals and brown solution. Upon cooling to room temperature, the product appeared to be a crystaline dark brown-green solid embedded in non-reacted ligand, ligand precursors and ferocene. The Carius tube was opened under a dinitrogen atmosphere. The excess of ligand and ligand precursors was extracted with acetonitrile and benzene solvents. The product was isolated as dark green moisture sensitve crystals suitable for X-ray crystalographic analysis. Yield 0.52 g (60 %). Analysis calculated for C22H20 Fe2Ni0: C 49.3; H 3.8; N 26.1; found: C 49.6; H 3.7; N 26.4. 9.2.1.3 Polybis(4-azabenzimidazolato)iron(n), [Fe(4-abimid)2]x. Ferocene (0.2 g, 1.07 mmol) and 4-azabenzimidazole (0.512g, 4.3 mmol) were placed in a Carius tube which was sealed under vacuum. The tube was heated at 145 °C 268 for 6 days. Under these conditons, the original orange solution of ferocene in molten 4-azabenzimidazole became a mixture of brown-red solid and an orange solution. Upon cooling to room temperature, the product appeared as a crystaline brown solid embedded in the excess, non-reacted imidazole. The Carius tube was opened under a dinitrogen atmosphere. The excess of 4-azabenzimidazole was extracted with dry and oxygen-free acetonitrile and xylenes. The product, which did not decompose in air for periods up to two months (longer periods of time were not examined), was isolated as amber-green crystals suitable for single crystal X-ray analysis. Yield 0.052 g (16 %). Analysis calculated for Ci2H8FeN6: C 49.3; H, 2.8; N 28.7; found: C 49.4; H 2.7; N 28.4. 9.2.1.4 Poly-2,2':6', 2"-terpyridine octakis(imidazolato)tetrairon(I), [Fe4(imid)8(terpy)]x. Ferocene (0.30 g, 1.63 mmol), imidazole (0.22 g, 3.26 mmol) and 2,2':6',2"-terpyridine (0.38 g, 1.63 mmol) were placed in a Carius tube which was sealed under vacuum. The tube was heated to 135 °C for 3 days. Under those conditons, the original orange solution of ferocene in molten ligand became a mixture of dark crystals and brown solution. Upon cooling to room temperature, the product appeared to be a crystaline dark green solid embedded in non-reacted ligand, ligand precursors, and ferocene. The Carius tube was opened under a dinitrogen atmosphere. The excess of ligand and ligand precursors was extracted with acetonitrile and toluene solvents. The 269 product was isolated as dark green moisture sensitve crystals suitable for X-ray crystalographic analysis. Yield 0.43 g (26 %). Analysis calculated for C39H35Fe4Ni9: C 47.2; H 3.5; N 26.8; found: C 47.5; H 3.6; N 26.9. 9.2.1.5 Polybis( 1 -memyl-2-thioimidazolato)iron(I)hemidicyclopentadienyliron (H), [Fe(l-Me-2-S-imid)20.5 Cp2Fe]x. Ferocene (0.5 g, 2.68 mmol) and an excess of 2-mercapto-lmethylimidazole (1.5 g, 13.14 mmol) were combined in a Carius tube which was sealed under vacuum. This mixture was heated at 145 °C for 6 days. During this period, a dark red solution of ferocene in molten ligand is obtained and light yelow crystals deposited from the solution. The reaction mixture was then alowed to cool down to room temperature, and the Carius tube was opened under a dinitrogen atmosphere. The excess of ligand precursor was extracted with dry and oxygen-free acetonitrile, and residual ferocene was removed with dry toluene. The compound was isolated as golden needle crystals, which were stable for at least a month of exposure to air (longer periods of time were not investigated). Yield 0.094 g (9 %). Analysis calculated for Ci3Hi5Fei.5N4S2: C 41.6; H 4.0; N 14.9; found: C 41.5; H 4.0; N 14.7. 270 9.2.2 COBALT(H) IMIDAZOLATE POLYMERS 9.2.2.1 Polybis(imidazolato)cobalt(I), [Co(imid)2]x. Cobalt(I) nitrate hexahydrate (7.5 g, 33 mmol) was dissolved in 10 ml of hot water and added dropwise to a solution of imidazole (3.4 g, 50 mmol) in 40 ml of hot water. The mixture was brought to boil and then immediately filtered. The solid obtained was washed with hot water first, then with acetone. The product was isolated as a microcrystaline purple powder and was dried in vacuo at 100 °C. Yield 0.25 g (5%). Analysis calculated forC6H6N4Co: C 37.3, H 3.1, N 29.0; found: C 37.6, H 3.0, N 28.7. This compound was also obtained by removal (thermolysis) of two molecules of neutral imidazole from compound [Co3(imid)6(imidH)2]x (see section 9.2.2.5 below). This thermal treatment of [Co3(imid)6(imidH)2]x was caried out under a flux of dinitrogen utilizing a TA Instruments TGA 51 unit. A thirty minute isotherm at 325 °C was programmed in the instrument in order to remove the neutral imidazole completely. Analysis found: C 37.1, H 3.0, N 28.8. 9.2.2.2 Polybis(2-methylimidazolato)cobalt(I), [Co(2-meimid)2]x. Cobalt(I) nitrate hexahydrate (1.0 g, 3.7 mmol) was dissolved in 10 ml of ethanol and added dropwise to a solution of 2-methylimidazole (3.0 g, 37 mmol) in 40 ml of ethanol. The purple precipitate which formed immediately was filtered off, 271 washed with ethanol, and dried for 1 hour at 115 °C. Yield 0.15 g (18 %). Analysis calculated for C8H10CoN4: C 43.4, H 4.6, N 25.3; found: C 43.0, H 4.3, N 24.9. 9.2.2.3 Polybis(4-memylimidazolato)cobalt(LT), [Co(4-meimid)2]x. Cobalt(I) chloride hexahydrate (1.0 g, 4.2 mmol) was dissolved in 15 ml of water and added dropwise to a solution of 4-methylimidazole (3.0 g, 37 mmol) in 30 ml of water. Sodium hydroxide (1.48 g, 37 mmol) was dissolved in 5 ml of water and added dropwise to the mixture to complete the precipitation of a fine purple powder. The precipitate was filtered off, washed with water and then ethanol, and air-dried at room temperature. Yield 0.36 g (39 %). Analysis calculated for CgHioCoN^  C 43.5, H 4.6, N 25.3; found: C 43.6, H 4.5, N 25.0. 9.2.2.4 Polybis(benzimidazolato)cobalt(I), [Co(benzimid)2]x. Benzimidazole (11.8 g, 100 mmol) was dissolved in 20 ml of hot water and added to a solution of cobalt(LT) nitrate hexahydrate (15.0 g, 51.5 mmol) dissolved in 80 ml of hot water. A purple precipitate formed immediately. The mixture was brought to boil. After the solution was cooled, the precipitate was filtered off, washed with ethanol, water and acetone, and dried under vacuum. Yield 5.1 g (34 %). Analysis calculated, for C14H10C0N4: C 57.3, H 3.4, N 19.1; found: C 57.5, H 3.3, N 19.2. 272 9.2.2.5 Polybis(imidazole)hex^  [Co3(imid)6(imidH)2]x. This compound was obtained as a purple powder by heating, to 150 °C for 5 days, a mixture of an excess of imidazole (2.0 g, 26 mmol) with cobaltocene (0.2 g, 1.06 mmol) in a sealed and evacuated Carius tube. Dry xylenes and acetonitrile were used to isolate the product from residual ligand precursor. Yield 0.34 g (45 %). Analysis calculated for C24H26Co3N16: C 40.3, H 3.6, N 31.3; found: C 40.8, H 3.6, N 31.2. 9.2.2.6 Polybis(4-azabenzimidazolate)cobalt(I), [Co(4-abimid)2]x. Cobaltocene(0.2 g, 1.06 mmol) and 4-azabenzimidazole (0.512g, 4.3 mmol) were placed in a Carius tube which was sealed under vacuum. The tube was heated at 145 °C for 4 days. Upon cooling to room temperature, the Carius tube was opened under a dinitrogen atmosphere. The excess of 4-azabenzimidazole was extracted with dry and oxygen-free acetonitrile and xylenes. The polymer was isolated as a dark purple microcrystaline solid. Yield 0.09 g (29 %). Analysis calculated for Ci2H8CoN6: C 48.8; H 2.7; N 28.5; found: C 50.1; H 2.8; N 28.4. Repeated analysis on two diferent samples of [Co(4-abimid)2]x consistently gave good H and N results and high C. There is no explanation for the high C except to suggest there may be smal amounts of solvent trapped in the sample. 273 9.2.2.7 Poly-2,2'-bipyridinetetralds(imidazolato)dicobalt(lT^  [Co2(imid)4(bipy)]x. Cobaltocene (0.3 g, 1.6 mmol), imidazole (0.2 g, 2.9 mmol) and excess 2,2'-bipyridine (0.5 g, 3.2 mmol) were placed in a Carius tube which was sealed under vacuum. The tube was heated to 130 °C for 2 days. The Carius tube was opened under a dinitrogen atmosphere. The excess of ligand and ligand precursors was extracted with acetonitrile and benzene solvents. The product was obtained as a purple microcrystaline solid. Yield 0.47 g (54 %). Analysis calculated for C22H20C02N10'. C 48.7; H 3.7; N 25.8. Found: C 49.0; H 3.7; N 25.6. 9.2.3 NICKEL(H) IMIDAZOLATE POLYMER 9.2.3.1 Polybis(benzimidazolato)nickel(I), [Ni(benzimid)2]x. Benzimidazole (11.8 g, 100 mmol) was dissolved in 20 ml of hot water and added to a solution of nickel(I) nitrate hexahydrate (15.0 g, 51.5 mmol) dissolved in 80 ml of hot water. A light violet precipitate was formed immediately. The mixture was brought to the boil. After the solution was cooled, the precipitate was filtered off, washed with ethanol, water and acetone, and dried under vacuum. Yield 0.82 g (5 %). Analysis calculated for Ci4Hi0NiN4: C 57.5, H 3.4, N 19.2; found: C 57.2, H 3.4, N 19.2. 274 9.2.4 COPPER IMIDAZOLATE POLYMERS In the folowing syntheses, wherever copper is used as a reactant it refers to copper metal beads (3-5 mm diameter) which had been cleaned by washing them with 12 M HC1, water, and acetone prior to use. 9.2.4.1 Polybis(imidazolato)copper(I), [Cu(imid)2]x. Copper (8.13g, 128 mmol) and imidazole (10 g, 147 mmol) were placed in a 100 ml round-botomed flask fited with a condenser. The reaction mixture was heated to 110 °C and air was bubbled into the mixture via a Pyrex tube for 48 h. A dark blue solid began to form. Sublimed imidazole was periodicaly scraped back into the reaction flask during the reaction. Upon cooling, the solidified mixture was extracted with ethanol, suction filtered, and further washed with acetone, and air-dried at room temperature. The dark blue powdery compound was isolated by physical separation from the copper shot. A 84 % yield was obtained based on amount of copper reacted. Analysis calculated forC6H6CuN4: C 36.4, H 3.1, N 28.3; found: C 36.5, H 3.0, N 27.9. 9.2.4.2 Polybis(2-methylimidazolato)copper(I), [Cu(2-meimid)2]x. Copper (1.9 g, 30 mmol) and an ethanolic solution (~ 75 ml) of 2-methylimidazole (3.0 g, 37 mmol) were placed in a round-botomed flask. The reaction 275 mixture was stirred at room temperature for 3 days. A dark brown precipitate, which formed, was suction filtered and washed thoroughly with ethanol and acetone. Air drying at room temperature yielded a light brown powder. A 77 % yield was obtained (based on amount of copper reacted). Analysis calculated for CgHioCuN^  C 42.6, H 4.5, N 24.8; found: C 42.4, H 4.3, N 24.5. 9.2.4.3 Polybis(4-methylimidazolato)copper(U), [Cu(4-meimid)2]x. Copper(I) sulphate hexahydrate (1.0 g, 3.7 mmol) was dissolved in 15 ml of water and added dropwise to a solution of 4-methylimidazole (3.0 g, 37 mmol) in 50 ml of water. 35 ml of a 1 M aqueous solution of NH3 was added dropwise to the mixture to complete the precipitation of a fine brown powder. The precipitate was filtered off, washed with water and then ethanol, and air-dried at room temperature. A brown powdery solid was obtained. Yield 0.57 g (68 %). Analysis calculated for C8HioCuN4: C 42.6, H 4.5, N 24.8; found: C 42.6, H 4.5, N 24.5. 9.2.4.4 Polybis(benzimidazolato)copper(n), [Cu(benzimid)2]x. Clean copper beads (1.0 g, 16 mmol) and an ethanolic solution of benzimidazole (3.6 g, 30 mmol) were placed in a round botom flask. After 5 days of vigorous stirring of the solution at room temperature, a red precipitate was formed. The solid was suction filtered and washed with ethanol. After air drying at room temperature the compound 276 was obtained as a red powder. 79 % yield was obtained based on amount of copper reacted. Analysis calculated for Ci4Hi0CuN4: C 56.5, H 3.4, N 18.8; found: C 56.2, H 3.4, N 18.6. 9.2.4.5 Polybis(4,5-dichloroimidazolato)copper(I), [Cu(4,5-dichloroimid)2]x-Copper(I) sulphate pentahydrate (1.0 g, 4.0 mmol) was dissolved in 15 ml of water and added dropwise to a solution of 4,5-dichloroimidazole (3.0 g, 22 mmol) in 60 ml of hot water. 20 ml of a 1 M aqueous solution of NH3 was added dropwise to the mixture to complete the precipitation of a fine gray-pink precipitate. The precipitate was filtered off, washed with water and then ethanol, and air dried at room temperature. A violet powdery product was obtained. Yield 0.37 g ( 28 %). Analysis calculated for C6H 2Cl4CuN 4: C 21.5, H 0.6, N 16.7; found: C 21.6, H 0.8, N 17.0. 9.3 PHYSICAL METHODS 9.3.1 MAGNETIC SUSCEPTIBILITY MEASUREMENTS DC magnetic susceptibilty measurements were performed using a Quantum Design (MPMS) SQUID magnetometer. These measurements were made at temperatures over the range 2-300 K and, unless otherwise stated, at applied fields of 500 and 10 000 G. Magnetization studies as a function of field strength (0-55 000 G) 277 were made at several temperatures and hysteresis magnetization data were obtained by oscilating the applied magnetic field between +55 000 G and -55 000 G usualy at 4.8 K. The sample holder is made of P VC plastic and is designed to go undetected by the magnetometer. The PVC sample holder and details regarding the use of the equipment have been described before [1]. Magnetic susceptibilties were corected for background and for the diamagnetism of al atoms using Pascal's constants [2]. All magnetic measurements were done on fine powdered samples and the data reported here are on a per mole of metal ion basis. Temperature calibrations on the SQUID magnetometer are performed using an external platinum resistance thermometer and a temperature accuracy within 0.1% is atained. Magnetic susceptibilty signals were calibrated using ultra-pure nickel standard and accuracy within 1% is obtained. When a material is suspected to behave as a molecule-based magnet below a Tc, such as the majority of the transition metal imidazolate compounds described in this thesis, certain precautions must be taken in the determination of its magnetic properties, to avoid misleading results. Problems are encountered if the remnant magnetization present after the colection of one set of magnetic data is not removed before another set of magnetic data is colected. If not removed, the remnant magnetization may influence or dominate any new magnetic signal detected by the magnetometer. A convenient way to remove this magnetization is to heat the sample above its Tc, then, set the applied field to zero by oscilating the field between positve and negative values (decreasing the magnitude of the field in each oscilation), and finaly cooling the sample to the 278 desired temperature in zero applied field. This procedure was applied to every molecule-based magnet investigated in this dissertation. AC magnetic susceptibilty measurements were made by W. M. Reif at Northeastern University using a Lake Shore Cryotronics Co. Model 7000 AC susceptometer, generaly over the temperature range 4.2 K to 30 K in an AC field of 1.0 G or 2.5 G at frequencies of 125 Hz or 500 Hz. 9.3.2 SINGLE CRYSTAL X-RAY DIFFRACTION Measurements were made by S. J. Retig and B. O. Patrick of this Department using either a Rigaku/ADSC CCD diffactometer or a Rigaku AFC6S difractometer, both with graphite monochromated Mo-Ka radiation. The low-temperature single crystal X-ray difraction study of [Fe2(imid)4(bipy)]x was performed by B. O. Patrick. 9.3.3 POWDER X-RAY DIFFRACTION Powder difractograms were recorded at room temperature on a Rigaku Rotaflex RU-200BH rotating anode powder X-ray difractometer (graphite monochromated Cu Ka radiation). Samples were prepared by applying a hexanes slury of the compound onto a glass plate and alowing the solvent to evaporate. 279 9.3.4 ELEMENTAL ANALYSIS Microanalysis were performed by P. Borda of this Department. A Carlo Erba Model 1106 or a Fisons (Erba) Instruments EA 1108 CHN-O Elemental Analyzer were utilized for determination of carbon, hydrogen and nitrogen percentages. Elemental analyses are considered to have an absolute accuracy within ±0.3%. 9.3.5 MOSSBAUER SPECTROSCOPY The Mossbauer spectra were obtained by W. M. Reif at Northeastern University using a conventional constant acceleration spectrometer operated in multichannel scaling mode. The gamma ray source (Du Pont- Merck Co.) consisted of 51.5 mCi of Co57 in a rhodium metal matrix that was maintained at ambient temperature. The spectrometer was calibrated using a 6-micron thick natural abundance iron foil. Isomer shifts are reported relative to the center of the magnetic hyperfine patern of the later foil taken as zero velocity. Apiezon-N grease mul samples (where the compounds studied were in a macro-crystaline form), were used in al measurements caried out. Sample temperature variation was achieved using a standard exchange gas liquid helium cryostat (Cryo Industries of America, Inc.) with temperature measurement and control based on silcon diode thermometry in conjunction with a 10 microampere excitation source (Lakeshore Cryotronics, Inc). Spectra were fit to unconstrained Lorentzians using the program ORIGIN (Microcal Software, Inc.). 280 9.3.6 ELECTRONIC SPECTROSCOPY Electronic spectra (200-3000 nm) were obtained at room temperature using a Varian Cary 5 UV-Vis-NIR spectrophotometer. Samples were prepared as Nujol muls pressed between quartz plates. 9.3.7 TGA Thermal gravimetric analysis (35 °C to 800 °C) was done using a TA Instruments TA 2000 system with a TGA 51 unit. Powdered samples (7 -12 mg) were heated in a dinitrogen atmosphere at a rate of 10 °C per minute to a maximum temperature of 800 °C. 9.3.8 INFRARED SPECTROSCOPY Infrared spectra (4000-400 cm"1) were recorded at room temperature on a Bomem FTIR spectrophotometer using KBr disk samples or Nujol muls pressed between KBr disks. Band frequencies are accurate to within ±4 cm"1. 281 9.3.9 NMR SPECTROSCOPY Nuclear magnetic resonance spectra were recorded in a Bruker AC-200 FT-NMR Spectrometer. NMR solvents were used as internal standards for calibration of the observed chemical shifts. 282 References 1. M. K. Ehlert. Ph. D. Thesis. University of British Columbia, 1992. 2. E. Konig. Landolt-Bdrstein Numerical Data and Functional Relationships Science and Technology. New Series, Vol. I/2. K. H. Helwege and A. M Helwege Eds. Springer-Verlag, Berlin, 1966. 283 APPENDIX I SINGLE CRYSTAL X-RAY DIFFRACTION DATA Table 1-1 Crystalographic data for [Fe(pz)2]x Molecular formula C6H6FeN4 fw 94.99 Crystal system orthorhombic Space group Ibam (No. 72) a/A 7.515(2) blk 14.604(4) elk 7.359(1) V/k3 807.7(2) Z 8 DJg cm"3 1.562 /I (MoKoO/cm"1 18.02 Crystal size/mm 0.50x0.10x0.10 R(Ff 0.024 Rw iff 0.032 lR{F) = S||Fo|-|Fc||/2|F0|, Rw (F2) - (2Zw\\F02\-\Fc2\\rLw\F02\2) 284 Table 1-2 Selected bond lengths (A) and angles (°) for [Fe(pz)2]x with estimated standard deviations in parentheses. Fe(l)—N(l) 2.027(1) N(l>—Fe(l)—N(la) 110.27(6) N(l)—Fe(l)—N(lc) 109.22(6) N(la>—Fe(l)—N(lc) 108.93(9) N(l)—C(3) 1.337(2) N(l)—Fe(l)—N(lb) 108.93(6) N(la)—Fe(l}—N(lb) 109.22(9) N(lb)—Fe(l)—N(lc) 110.27(9) Table 1-3 Crystalographic data8 for [Fe(4-abimid)2]x. Molecular formula fw Crystal system Space group a/A b/A elk U/A3 Z C12H6FeN6 292.08 orthorhombic P2i2!2i (No. 19) 9.655(2) 10.3403(6) 12.4671(7) 1244.6(2) 4 285 ZVg cm"3 1.559 F(000) 592.00 H (MoKaVcm"1 12.04 Crystal size/mm 0.15x0.20x0.20 26max/° 60.1 Total reflections 11118 Unique reflections 3233 No. with/>3o(i) 1685 No. of variables 172 R;Rw(i%/>3a(7) 0.034; 0.024 R; Rw (F\ al data) 0.084; 0.055 gof 1.22 a Temperature 180 K, Rigaku/ADSC CCD difractometer, Mo Ka (k = 0.71069), graphite monochromator, takeof angle 6.0°, aperture 94.0 x 94.0 mm at a distance of 39.22(7) mm from the crystal, o2 )^ = (C + Z?)/Lp2 (C = scan count, B = background count), function minimized l.w(\Fo2\-\Fc2\)2 where w = l/o^ fF2), R(F) = S|Fo|-[Fcl|/S|F0|, Rw (F2) = (SW||Fo2|-|Fc2||/Iw|Fo2|2)1/2, and gof = [Sw(|F02|-|Fc2|)2/(m-n)]1/2. 286 Table 1-4 Selected bond lengths (A) and angles (°) for [Fe(4-abimid)2]x with estimated standard deviations in parentheses*. Fe(l)—N(l) 2.030(3) Fe(l)—N(2)a 2.046(3) Fe(l)—N(4) 2.044(3) Fe(l>—N(5)b 2.034(3) N(l)—Fe(l)—N(2)a 110.39(11) N(l)—Fe(l)—N(4) 102.10(11) N(l>—Fe(l)—N(5)b 111.76(11) N(2)a—Fe(l)—N(4) 104.31(11) N(2)a—Fe(l)—N(5)b 109.55(11) N(4)—Fe(l>—N(5)b 118.24(11) Fe(l)_N(l)—C(l) 128.8(3) Fe(l)—N(l)-C(6) 127.8(2) Fe(l)c—N(2>—C(2) 132.6(2) Fe(l)c—N(2)—C(l) 123.5(2) Fe(l)—N(4>—C(12) 131.1(2) Fe(l)—N(4)—C(7) 124.1(3) Fe(l)a—N(5)—C(7) 127.6(2) Fe(l)d—N(5)—C(8) 129.1(2) * Superscripts refer to symmetry operations: (a) -1/2+x, 3/2-y, 1-z (b) 1-x, -1/2+y, 3/2-z (c) 1/2+x, 3/2-y, 1-z (d) 1-x, 1/2+y, 3/2-z 287 Table 1-5 Crystalographic data for a- and y-[Fe2(imid)4(bipy)]x. a-phase y-phase Molecular formula fw Crystal system Space group o a, A b, A c, A oc, deg Meg y,deg V, A3 Z peak, g/cm3 p. (MoKa), cm"1 Crystal size, mm C22H2oFe2Nio 536.16 triclinic PI (No. 2) 10.507(4) 13.730(4) 9.188(3) 106.51(3) 108.32(3) 80.84(3) 1202.9(2) 2 1.480 12.36 0.35x0.15x0.15 293 C22H2oFe2Nio 536.16 triclinic PI (No. 2) 10.4138(5) 13.5075(5) 26.060(1) 104.530(2) 93.892(2) 100.512(2) 3646.0(2) 2 1.542 12.88 0.45 x 0.35 x 0.20 113 288 R(F)a 0.034 0.030 0.035 0.049 a R(F) = ZHFol-IFcl/ZFol, RwiF2) = (Ew|Fo2|-|Fc2|/2:>v|Fo2|2)1/ Table 1-6 Selected bond lengths (A) for a- and y-[Fe2(imid)4(bipy)]x, with estimated standard deviations in parentheses. a-phase y-phase Fe(l)—N(l) 2.262(2) Fe(l)—N(10) 2.314(2) Fe(l)—N(2) 2.299(2) Fe(l)—N(9) 2.217(2) Fe(l)—N(3) 2.196(2) Fe(l>—N(7) 2.207(2) Fe(l)—N(5) 2.153(2) Fe(l)—N(3) 2.162(3) Fe(l)—N(8) 2.192(2) Fe(l)—N(l) 2.195(2) Fe(l>—N(10) 2.153(2) Fe(l>—N(5) 2.136(2) Fe(2)—N(4) 2.024(2) Fe(2)—N(6) 2.019(2) Fe(2)—N(6) 2.036(2) Fe(2)—N(4) 2.032(2) Fe(2)—N(7) 2.027(2) Fe(2)—N(2) 2.030(2) Fe(2)—N(9) 2.028(2) Fe(2)—N(l) 2.024(2) Fe(3>—N(12) Fe(3)—N(13) Fe(3)—N(15) Fe(3)—N(17) Fe(3>—N(18) Fe(3)—N(22) Fe(4)—N(16) Fe(4)—N(19) Fe(4)—N(21) Fe(4>—N(23) Fe(5)—N(20) Fe(5>—N(24) Fe(5)—N(25) Fe(5)—N(27) Fe(5)—N(29) Fe(5)—N(30) Fe(6)—N(8) Fe(6)—N(14) Fe(6)—N(26) Fe(6)—N(28) 2.186(2) 2.152(3) 2.155(2) 2.263(2) 2.302(2) 2.187(2) 2.035(2) 2.024(2) 2.039(2) 2.037(2) 2.011(2) 2.019(2) 2.016(2) 2.003(2) 1.957(2) 1.970(2) 2.059(2) 2.056(3) 2.034(2) 2.038(2) Table 1-7 Selected bond angles (°) for a- and y-[Fe2(imid)4(bipy)]x, with estimated standard deviations in parentheses. a-phase y-phase N(l)—Fe(l)—N(2) 71.78(9) N(10)^-Fe(l>—N(9) 72.54(9) N(l)—Fe(l)—N(3) 89.68(9) N(10)—Fe(l>—N(7) 90.02(9) N(l>—Fe(l)—N(5) 168.22(9) N(10>—Fe(l)—N(3) 169.59(9) N(l)—Fe(l)—-N(8) 88.91(8) N(10)—Fe(l)—N(l) 81.33(9) N(l>—Fe(l)—N(10) 97.12(9) N(10>—Fe(l)—N(5) 96.67(9) N(2>—Fe(l)—N(3) 88.99(9) N(9)—Fe(l)—N(7) 90.02(9) N(2)—Fe(l)—N(5) 96.51(9) N(9)—Fe(l>—N(3) 98.33(9) N(2)—Fe(l)—N(8) 83.61(9) N(9)—Fe(l)—N(l) 89.42(9) N(2)—Fe(l)—N(10) 168.33(9) N(9>—Fe(l)—N(5) 169.21(9) N(3)—Fe(l)—N(5) 88.75(9) N(7)—Fe(l)—N(3) 87.98(9) N(3>—Fe(l)—N(8) 172.53(9) N(7>—Fe(l)—N(l) 172.19(9) N(3)—Fe(l)—N(10) 94.65(9) N(7)—Fe(l>—N(5) 93.82(9) N(5)—Fe(l)—N(8) 91.14(9) N(3)—Fe(l)—N(l) 91.13(9) N(5)—Fe(l)—N(10) 94.65(9) N(3)—Fe(l)—-N(5) 92.3(1) N(8)—Fe(l)—N(10) 92.80(9) N(l)—Fe(l)—N(5) 93.97(9) N(4)—Fe(2>—N(6) 111.3(1) N(6)—Fe(2)—N(4) 108.7(1) 291 N(4)—Fe(2>—N(7) 110.15(9) N(4)—Fe(2)—N(9) 106.96(9) N(6)—Fe(2)—N(7) 105.33(9) N(6)—Fe(2)—N(9) 105.47(9) N(7>—Fe(2)—N(9) 117.6(1) N(6)—Fe(2)—N(2) 109.1(1) N(6)—Fe(2)—N(l) 117.8(1) N(4)—Fe(2)—N(l) 102.9(1) N(2)—Fe(2)—N(l) 107.4(1) N(2)—Fe(2)—N(l 1) 107.4(1) N(12)—Fe(3)—N(13) 86.85(9) N(12)—Fe(3)—N(15) 93.51(9) N(12)—Fe(3)—N(17) 87.41(9) N(12)—Fe(3)—N(18) 89.78(9) N(12>—Fe(3)—N(22) 174.26(9) N(13)—Fe(3)—N(15) 97.3(1) N(13)—Fe(3)—N(17) 163.39(9) N(13)—Fe(3)—N(18) 92.67(9) N(13)—Fe(3>—N(22) 92.21(9) N(15>—Fe(3)—N(17) 98.56(9) N(15)—Fe(3)—N(18) 169.6(1) N(15)—Fe(3)—N(22) 92.22(9) N(17)—Fe(3>—N(18) 71.73(9) N(17)—Fe(3>—N(22) 91.95(9) N(18)—Fe(3)—N(22) 84.61(8) N(16)—Fe(4)—N(19) 109.1(1) 292 N(16>—Fe(4)—N(21) N(16)—Fe(4)—N(23) N(19)—Fe(4)—N(21) N(19>—Fe(4)—N(23) N(21>—Fe(4)—N(23) N(20>—Fe(5)—N(24) N(20)—Fe(5)—N(25) N(20>—Fe(5)—N(27) N(20)—Fe(5)—-N(29) N(20)—Fe(5>—N(30) N(24)—Fe(5>—N(25) N(24>—Fe(5)—N(27) N(24)—Fe(5)—N(29) N(24>—Fe(5>—N(30) N(25)—Fe(5>—N(27) N(25)—Fe(5>—N(29) N(25)—Fe(5)—N(30) N(27>—Fe(5>—N(29) N(27>—Fe(5)—N(30) N(29)—Fe(5>—N(30) N(8>—Fe(6)—N(14) 114.8(1) 104.2(1) 112.7(1) 109.9(1) 105.7(1) 88.41(1) 92.31(9) 177.71(9) 90.9(1) 90.98(9) 89.2(1) 91.7(1) 176.8(1) 95.1(1) 89.98(9) 94.0(1) 174.7(1) 88.9(1) 86.73(1) 81.8(1) 114.8(1) N(8)—Fe(6>—N(26) 105.95(9) N(8)—Fe<6)—N(28) 104.74(9) N(14)—Fe(6)—N(26) 102.7(1) N(14)—Fe(6)—N(28) 100.8(9) N(26)—Fe(6>—N(28) 128.3(1) Table 1-8 Crystalographic data for P-[Fe2(imid)4(bipy)]x. Molecular formula C22H2oFe2Nio fw 536.16 Crystal system triclinic Space group PI (No. 2) o a, A 17.1338(4) b, A 18.5426(4) c, A 23.6199(3) a, deg 80.424(3) P,deg 75.364(3) y,deg 80.826(3) V, A3 7105.1(2) pcaic g/cm3 1.440 \i (MoKa), cm"1 12.36 Crystal size, mm 0.25 x 0.20 x 0.20 T, K 143 i?(i^ ,/>2c(/)) 0.145 Rw(F2f 0.339 3 R(F) = SIlFol-IFcll/SIFol, RM.F2) = (2ZW\\Fo2\-\Fc2\\/2Zw\F02\2)m Table 1-9 Selected bond angles (°) for (3-[Fe2(imid)4(bipy)]x, with estimated standard deviations in parentheses. Fe(la)—N(5a) 2.139(2) Fe(la)—N(3) 2.144(2) Fe(la)—N(8a) 2.193(2) Fe(la)—N(20) 2.197(2) Fe(la)—N(2a) 2.263(2) Fe(la)—N(la) 2.299(2) Fe(lb)—N(13) 2.155(2) Fe(lb)—N(38) 2.170(2) Fe(lb)—N(10a) 2.181(2) Fe(lb)—N(15) 2.188(2) Fe(lb)—N(2b) 2.215(2) Fe(lb)—N(lb) 2.280(2) Fe(lc)—N(18) 2.145(2) Fe(lc)—N(23) 2.174(2) Fe(lc)—N(25) 2.201(2) Fe(lc)—N(2c) 2.246(2) Fe(ld)—N(28) 2.130(3) Fe(ld)—N(33) 2.167(2) Fe(ld)—N(ld) 2.232(2) Fe(le)—N(45) 2.155(2) Fe(le)—N(40) 2.170(2) Fe(le)—N(2e) 2.220(2) Fe(lf>—N(59) 2.132(2) Fe(lf)—N(55) 2.187(2) Fe(lf)—N(52) 2.199(2) Fe(2a)—N(9a) 2.031(2) Fe(2a)—N(57) 2.046(2) Fe(2b)—N(19) 2.018(2) Fe(2b)—N(16) 2.039(2) Fe(2c>—N(27) 2.017(2) Fe(2c)—N(26) 2.040(2) Fe(2d>—N(36) 2.015(2) Fe(2d>—N(37) 2.036(2) Fe(2e)—N(49) 2.004(2) Fe(2e)—N(47) 2.017(3) Fe(lc)—N(48) 2.218(2) Fe(lc)—N(lc) 2.273(2) Fe(ld)—N(30) 2.159(2) Fe(ld)—N(35) 2.185(2) Fe(ld>—N(2d) 2.276(2) Fe(le)—N(50) 2.168(2) Fe(le)—N(43) 2.189(2) Fe(le>—N(le) 2.310(2) Fe(lf)—N(58) 2.140(2) Fe(lf>—N(lf) 2.196(2) Fe(lf)—N(2f) 2.254(2) Fe(2a)—N(7a) 2.039(2) Fe(2a)—N(6a) 2.046(2) Fe(2b>—N(17) 2.023(2) Fe(2b)—N(14) 2.047(2) Fe(2c)—N(56) 2.038(3) Fe(2c>—N(29) 2.041(2) Fe(2d)—N(39) 2.024(2) Fe(2d>—N(24) 2.044(2) Fe(2e>—N(46) 2.013(2) Fe(2e)—N(51) 2.027(2) Fe(2f>—N(44) 2.027(2) Fe(2f>—N(34) 2.028(2) Fe(2f)—N(4) 2.034(2) Fe(2f)—N(60) 2.044(2) N(5a)—Fe(la)—N(3) 94.02(9) N(5a)—Fe(la)—N(8a) 90.32(9) N(5a)—Fe(la)—N(20) 91.28(9) N(5a)—Fe(la)—N(2a) 168.13(9) N(5a>—Fe(la)—N(la) 96.43(9) N(3)—Fe(la>—N(8a) 95.07(9) N(3>—Fe(la)—N(20) 92.93(9) N(3)—Fe(la)—N(2a) 97.83(9) N(3)—Fe(la)—N(la) 168.34(9) N(8a>—Fe(la)—N(20) 171.71(9) N(8a>—Fe(la)—N(2a) 89.37(9) N(8a)—Fe(la)—N(la) 90.03(9) N(20)—Fe(la)—N(2a) 87.40(9) N(20>—Fe(la)—N(la) 81.71(9) N(2a)—Fe(la)—N(la) 71.70(9) N(13)—Fe(lb)—N(38) 96.83(9) N(13)—Fe(lb)—N(10a) 89.03(9) N(13)—Fe(lb)—N(15) 86.65(9) N(13)—Fe(lb)—N(2b) 167.04(9) N(13)—Fe(lb)—N(lb) 94.46(9) N(38)—Fe(lb)—N(10a) 92.06(9) N(38)—Fe(lb)—N(15) 93.52(9) N(38)—Fe(lb)—N(2b) 95.95(9) N(38>—Fe(lb)—N(lb) 168.23(1) N(10a)—Fe(lb)—N(15) 173.31(1) N(10a)—Fe(lb)—N(2b) 92.68(9) N(10a)—Fe(lb)—N(lb) 84.80(9) N(15)—Fe(lb)—N(2b) 90.41(9) N(15)—Fe(lb)—N(lb) 90.44(9) N(2b)—Fe(lb)—N(lb) 72.92(9) N(18)—Fe(lc)—N(23) 95.44(9) N(18)—Fe(lc)—N(25) 93.39(9) N(18)—Fe(lc)—N(48) 94.65(9) N(18>—Fe(lc)—N(2c) 167.72(9) N(18)—Fe(lc)—N(lc) 95.59(9) N(23>—Fe(lc)—N(25) 92.08(9) 297 N(23)—Fe(lc)—N(48) 87.94(9) N(23)—Fe(lc>—N(lc) 168.92(9) N(25>—Fe(lc)—N(2c) 83.45(9) N(48>—Fe(lc)—N(2c) 88.52(9) N(2c)—Fe(lc)—N(lc) 72.51(9) N(28>—Fe(ld)—N(33) 87.64(9) N(28>—Fe(ld)—N(ld) 167.05(9) N(30)—Fe(ld)—N(33) 94.57(9) N(30)—Fe(ld)—N(ld) 96.06(9) N(33>—Fe(ld)—N(35) 174.05(9) N(33)—Fe(ld)—N(2d) 90.04(8) N(35>—Fe(ld)—N(2d) 84.09(9) N(45>—Fe(le)—N(50) 94.20(9) N(45>—Fe(le)—N(43) 89.94(9) N(45>—Fe(le)—N(le) 95.37(9) N(50)—Fe(le)—N(43) 91.25(9) N(50)—Fe(le)—N(le) 169.10(1) N(40)—Fe(le)—N(2e) 89.47(9) N(43)—Fe(le)—N(2e) 90.79(9) N(2e)—Fe(le)—N(le) 72.31(9) N(59>—Fe(lf)—N(55) 88.76(9) N(23>—Fe(lc)—N(2c) 96.52(9) N(25)—Fe(lc)—N(48) 171.92(9) N(25>—Fe(lc)—N(lc) 88.27(9) N(48>—Fe(lc)—N(lc) 90.17(9) N(28)—Fe(ld)—N(30) 96.83(9) N(28>—Fe(ld)—N(35) 91.86(9) N(28)—Fe(ld)—N(2d) 94.71(9) N(30>—Fe(ld>—N(35) 91.38(9) N(30)—Fe(ld)—N(2d) 167.74(9) N(33>—Fe(ld)—N(ld) 90.04(9) N(35)—Fe(ld)—N(ld) 89.13(9) N(ld)—Fe(ld)—N(2d) 72.54(9) N(45>—Fe(le)—N(40) 88.28(9) N(45)—Fe(le>—N(2e) 167.49(9) N(50)—Fe(le)—N(40) 95.81(9) N(50)—Fe(le>—N(2e) 98.27(8) N(40>—Fe(le)—N(43) 172.83(1) N(40>—Fe(le>—N(le) 89.70(9) N(43>—Fe(le)—N(le) 83.56(9) N(59>—Fe(lf)—N(58) 95.54(9) N(59>—Fe(lf)—N(lf) 168.91(9) 298 N(59)—Fe(lf)—N(52) 90.51(9) N(58)—Fe(lf)—N(55) 93.11(9) N(58)—Fe(lf>—N(52) 91.41(9) N(55)—Fe(lf)—N(lf) 90.83(9) N(55)—Fe(lf)—N(2f) 92.19(9) N(lf)—Fe(lf)—N(2f) 73.43(9) N(9a)—Fe(2a)—N(7a) 107.79(9) N(9a>—Fe(2a)—N(6a) 104.89(9) N(7a)—Fe(2a)—N(6a) 110.75(1) N(19)—Fe(2b)—N(17) 121.11(1) N(19)—Fe(2b)—N(14) 103.99(9) N(17)—Fe(2b)—N(14) 105.39(9) N(27)—Fe(2c>—N(56) 114.60(9) N(27)—Fe(2c)—N(29) 105.06(9) N(56)—Fe(2c)—N(29) 105.71(9) N(36>—Fe(2d)—N(39) 110.82(9) N(36)—Fe(2d)—N(24) 104.15(9) N(39)—Fe(2d)—N(24) 109.56(9) N(49)—Fe(2e)—N(46) 102.63(9) N(49)—Fe(2e)—N(51) 116.70(9) N(46>—Fe(2e)—N(51) 106.94(9) N(59)—Fe(lf)—N(2f) 95.51(9) N(58)—Fe(lf)—N(lf) 95.54(9) N(58)—Fe(lf>—N(2f) 167.83(9) N(55>—Fe(lf)—N(52) 175.47(9) N(lf)—Fe(lf)—N(52) 89.03(9) N(52)—Fe(lf)—N(2f) 83.43(9) N(9a)—Fe(2a)—N(57) 121.74(9) N(7a>—Fe(2a>—N(57) 107.55(9) N(57>—Fe(2a>—N(6a) 103.87(1) N(19)—Fe(2b)—N(16) 110.79(1) N(17)—Fe(2b)—N(16) 103.86(9) N(16)—Fe(2b>—N(14) 111.63(9) N(27)—Fe(2c)—N(26) 101.65(9) N(56)—Fe(2c>—N(26) 110.12(9) N(26)—Fe(2c)—N(29) 119.96(9) N(36)—Fe(2d)—N(37) 117.48(9) N(39>—Fe(2d>—N(37) 108.69(9) N(37>—Fe(2d)—N(24) 105.76(9) N(49)—Fe(2e>—N(47) 109.23(9) N(46)—Fe(2e)—N(47) 112.01(9) N(47)—Fe(2e)—N(51) 109.21(9) 299 N(44)—Fe(2f)—N(34) 109.88(9) N(44>—Fe(2f)—N(4) 118.24(9) N(44)—Fe(2f>—N(60) 105.60(1) N(34>—Fe(2f)—N(4) 107.07(1) N(34)—Fe(2f)—N(60) 112.57(1) N(4>—Fe(2f>—N(60) 103.44(9) Table I-10 Crystalographic data for [Fe4(imid)8(terpy)]x. Molecular formula C39H3sFe4Ni9 fw 993.22 Crystal system triclinic Space group Pi (No. 2) a/A 11.351(2) blk 13.628(1) elk 15.748(4) a, deg 113.550(3) P, deg 103.696(3) y,deg 91.826(3) V,k3 2147.6(6) Z 2 A/gem"3 1.536 F(000) 1012.0 300 jj. (MoKoO/cm"1 13.77 Crystal size/mm 0.6x0.10x0.03 56.0 Total reflections 18208 Unique reflections 8100 No. with/>3o(7) 4793 No. of variables 559 R;Rw(F,/>3o(7) 0.039; 0.049 R; Rw (F2, al data) 0.079; 0.114 gof 1.07 a Temperature 173 K, Rigaku/ADSC CCD difractometer, Mo Ka (k = 0.71069), graphite monochromator, takeof angle 6.0°, aperture 94.0 x 94.0 mm at a distance of 40.48 mm from the crystal, c 2 ^ ) = (C + i?)/Lp2 (C = scan count, B = background count), function minimized Zw(|F02|-|Fc2|)2 where w = l/o^ F*), R(F) = 2||Fo|-|Fc||/I|Fo|, Rw (F2) = (Zw||F02|-|Fc2||/Iw|Fo2|2)1/2, and gof = [Iw(|F02|-|Fc2|)2/(/n-n)]1/2. Table 1-11 Selected bond lengths (A) and angles (°) for [Fe4(imid)8(terpy)]x. Fe(l)—N(l) 2.023(4) Fe(l)—N(2) 2.037(4) 301 Fe(l>—N(3) 2.025(4) Fe(l)—N(4) 2.040(4) Fe(2)—N(12) 2.017(4) Fe(2>—N(10) 2.024(4) Fe(2)—N(l) 2.030(4) Fe(2)—N(9) 2.041(4) Fe(3>—N(15) 2.072(4) Fe(3)—N(8) 2.141(3) Fe(3)—N(5) 2.148(4) Fe(3>—N(13) 2.151(4) Fe(3>—N(6) 2.156(3) Fe(4>—N(18) 1.858(4) Fe(4>—N(19) 1.961(4) Fe(4)—N(17) 1.976(4) Fe(4)—N(14) 1.992(3) Fe(4)—N(16) 1.994(3) Fe(4>—N(7) 2.002(4) N(l)—Fe(l)—N(3) 108.64(2) N(l)—Fe(l)—N(4) 115.61(2) N(3>—Fe(l>—N(4) 106.08(2) N(12>—Fe(2>—N(10) 113.58(2) N(12>—Fe(2>—N(9) 114.32(2) N(10>—Fe(2)—N(9) 108.15(1) N(15>—Fe(3>—N(8) 104.10(1) N(15)—Fe(3)—N(13) 98.28(2) N(8>—Fe(3)—N(5) 86.73(1) N(8>—Fe(3)—N(6) 154.30(2) N(5)—Fe(3)—N(6) 90.75(1) N(l)—Fe(l)—N(2) 106.85(1) N(3>—Fe( 1 >—N(2) 113.50(2) N(2>—Fe(l)—N(4) 106.35(2) N(12>—Fe(2)—N(l) 101.56(2) N( 10)—Fe(2)—N( 11) 117.24(2) N(l)—Fe(2)—N(9) 101.52(2) N(15>—Fe(3>—N(5) 99.03(2) N(15)—Fe(3)—N(6) 101.55(1) N(8)—Fe(3>—N(13) 87.50(1) N(5)—Fe(3)—N(13) 162.61(2) N(13)—Fe(3)—N(6) 87.36(1) N( 18)—Fe(4)—N( 19) 81.95(2) N(18)—Fe(4)—N(14) 93.25(1) N(18)—Fe(4)—N(7) 178.60(1) N(19)—Fe(4)—N(14) 90.97(1) N(19)—Fe(4)—N(7) 98.37(2) N(17)—Fe(4>—N(16) 92.35(1) N(14)—Fe(4>—N(16) 175.62(1) N(16>—Fe(4)—N(7) 87.93(1) N(18)—Fe(4)—N(17) 81.48(2) N(18)—Fe(4)—N(16) 90.73(1) N(19)—Fe(4)—N(17) 163.43(2) N(19)—Fe(4>—N(16) 87.77(1) N(17>—Fe(4)—N(14) 90.07(1) N(17)—Fe(4)—N(7) 98.19(2) N(14>—Fe(4)—N(7) 88.11(1) Table I-12 Crystalographic data for [Fe( 1 -Me-2-S-imid)2-0.5Cp2Fe]x.a Molecular formula fw Space group a, A c, A z peak, g/cm3 F(000) CisHjsFej^ Sj 375.18 P4/n (No. 85) 13.2862(7) 8.7665(4) 1547.49(11) 4 1.610 768 303 radiation Mo u, cm"1 16.88 X,A 0.71069 R 0.077 Rw 0.063 T, °C -93 a R = I||Fo2|-Fc2||/L| Fo2|, tfw = (Lw(\Fo2\-\ Fc2\frLw Fo4)1'2 Table I-13 Selected bond lengths (A) and angles (°) for [Fe( 1 -Me-2-S-imid)2-0.5Cp2Fe]x* Fe(l)—S(l) S(l)-C(l) N(l)-C(2) N(2)-C(3) C(2)-C(3) 2.3677(8) 1.732(3) 1.383(4) 1.375(4) 1.358(4) Fe(2)—N(l) N(l)-C(l) N(2)-C(l) N(2)-C(4) 2.054(2) 1.342(3) 1.364(3) 1.455(4) S(l)—Fe(l)—S(l)fl 110.05(2) S(l>—Fe(l)—S(l)c 108.32(4) S(l)—Fe(l)—S(l)* 110.05(2) N(l>—Fe(2)—N(l/ 104.91(6) 304 N(l)—Fe(2)—N(l)e 104.91(6) Fe(l>—S(l)—C(l) 95.67(10) Fe(2>—N(l)—C(2) 128.0(2) C(l)—N(2)—C(3) 108.0(3) C(3)—N(2)—C(4) 125.5(3) S(l>—C(ly—N(2) 124.7(2) N(l)—C(2)—€(3) 109.9(3) N(l>—Fe(2)—N(l)c 119.05(13) Fe(2>—N(l)—C(l) 125.9(2) C(l>—N(l>—C(2) 106.0(2) C(l)—N(2)—C(4) 126.5(3) S(l)—C(l)—N(l) 125.4(2) N(l)—C(l)—N(2) 109.9(3) N(2)—C(3)—C(2) 106.2(3) * Superscript numbers refer to symmetry operation (a) 'A+y, 1-x, -z (b) 1-y, -1/2+x, -z (c) 3/2-x,l/2-y, z (d) '/a+yj-xj-z (e) 1-y,-1/2+x, 1-z 305 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0061270/manifest

Comment

Related Items