UNIVERSITY of CALGARY Heterocyclic Cyclopentadienyl

Home , Bismole, Borole, Dimethoxyethane, Ruthenocene, Sandwich compound, Transmetalation

UNIVERSITY OF CALGARY

Heterocyclic Cyclopentadienyl Analogs with BN Frameworks

Hanh Vien Ly

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEMISTRY

CALGARY, ALBERTA

DECEMBER, 2007

ISBN: 978-0-494-38226-4

Abstract

The heterocyclic cyclopentadienyl analogs have provided a fertile area of

research, owing to the interest in tuning the electronic properties of cyclopentadienyl

(Cp) ligands for the development of more efficient catalysts and new materials. A variety of five-membered heterocyclic Cp analogs have been synthesized by the formal replacement of ring carbon atoms with isoelectronic main group fragments. Metal complexes containing these heterocyclic ligands have proven to be feasible ancillary ligands in catalysis.

The primary objective of this thesis is the development of a novel class of heterocyclic Cp ligands having ring carbon fragments (CR) formally substituted with isolobal boron (BR−) and nitrogen (NR’+) fragments. The synthesis and characterization of the 1,2-diaza-3,5-diborolyl ligands with cyclic CB2N2 frameworks is described and

their coordination to a variety of metal fragments is reported. Just like cyclopentadienyl,

the 1,2-diaza-3,5-diborolyl ligands are excellent π ligands that exhibit interesting

coordination properties. Alkali metal complexes of 1,2-diaza-3,5-diborolyls were

synthesized and their investigation by single-crystal X-ray diffraction analysis revealed

substantial coordinative similarities, but also remarkable differences in comparison to

cyclopentadienyl ligands. Sandwich complexes containing these new heterocyclic π

ligands were synthesized by metathesis reaction of the corresponding alkali metal salts

with various metal halides. The structural characterization of a series of electron-rich

group 12 and 14 metallocenes, as well as metallocenes of the early and late transition metals, revealed some unique coordinative properties of the ligands: η1, η3, η4 and η5- coordination modes of the 1,2-diaza-3,5-diborolyl ligands were observed, depending on

ii the electron properties of the coordinated metals. In addition, the synthesis of the

heterobicyclic 1,5-diaza-2,4,6,8-tetraborolyl ligand featuring a C2B4N2 framework is

described and the structural investigation of its dipotassium salt showed that this

heterobicyclic ligand is a promising π bridging linker for the construction of polydecker

sandwich complexes. A triple-decker ruthenium sandwich complex featuring an unusual

eight-membered C2B4N2 ring as the middle deck was synthesized through the insertion of

two RuCp* fragments into the N-N bond of the heterobicyclic ligand. Electrochemical

studies of the triple-decker ruthenocene are presented.

The coordination chemistry of the 1,2,4-triaza-3,5-diborolyl ligand, a carbon-free heterocyclic Cp analog, was also investigated. The alkali metal salts of this ligand were synthesized via selective deprotonation of the ring nitrogen utilizing appropriate metalating agents. In contrast to the former ligands, the solid-state structures of the alkali metal complexes featuring these ligands are dominated by σ interactions of the ligand to the metal ions. A rhodium dimer containing 1,2,4-triaza-3,5-diborolyl ligands σ-bridging two Rh(cod) fragments was synthesized and characterized. The synthesis and characterization of a tricyclic compound with a B8N4 framework consisting of fused five- and six-membered rings was discussed, along with a study of its electrochemical properties.

iii Acknowledgements

I would like to thank my supervisor, Dr. Roland Roesler for his encouragement,

enthusiasm and suggestions with respect to my graduate research. His instruction and

trust in me have made me the chemist I am today. Special thanks to my PhD committee

members, Dr. Tristram Chivers, Dr. Warren Piers, Dr. Richard Oakley, and Dr. Ray

Turner.

To all the past and present group members, Kelly, Matt, Taryn, Javier, Hongsui,

Doaa and Andrea; I thank them for all the support and wonderful times that they have

shared with me in and out of the lab. To my friends Bonnie and Vicky, thank you for the routine lunch calls and entertaining talks. All of these wonderful friendships are greatly cherished and I wish them all the best of luck in their future accomplishments.

I would like to thank Dr. Masood Parvez, Dr. Dana Eisler, Dr. Jari Konu and Dr.

Robert McDonald (University of Alberta) for their contribution and expertise with regard to X-ray crystallography. I would like to thank Dr. Heikki Tuononen for all the computational calculations, as well as Tracey Roemmele and Dr. René Boeré for their help with the EPR studies. I would also like to thank Dorothy Fox, Qiau Wu, Jian Jun Li,

Roxanna Simank, Olivera Blagojevic and Dr. Raghav Yamdagni for their help and technical expertise. Special thank to Bonnie King for all of the administration work that she kept me (and Roland) on track and to Mark Toonen for his help with glassware.

Most importantly, I would like to express my deepest gratitude to my parents, my grandma and my two sisters, Vien and Cindy, for their unconditional love, encouragement and endless support throughout this journey. I could not have done it without them. Thank you all.

iv Table of Contents

Abstract ...... ii Acknowledgements ...... iiv List of Tables ...... iix List of Figures ...... xii List of Abbreviations ...... xvii List of Publications ...... xix

CHAPTER ONE: Introduction ...... 1

1.1 Cyclopentadienyl ...... 1 1.2. Heterocyclic Analogs of Cyclopentadienyl ...... 4 1.2.1. General Considerations ...... 4

1.2.2. Cyclopentadienyl Analogs with C4E Frameworks ...... 5

1.2.3. Cyclopentadienyl Analogs with C5-nEn (n = 2 – 4) Frameworks ...... 9

1.2.4. Carbon-Free Cyclopentadienyl Analogs with E5 Frameworks ...... 21 1.2.5. Multidecker Sandwich Complexes of Boron-Containing Heterocycles ....24 1.3. Objectives and Outline of Thesis ...... 25

CHAPTER TWO: 1,2-Diaza-3,5-diborolyl Ligands and their Alkali Metal Complexes ..28

2.1. Introduction ...... 28 2.2. Methodology for the Synthesis of 1,2-Diaza-3,5-diborolidines, their Spectroscopic Characterization and the X-ray Structures of 2.3c and 2.3d...... 30 2.3. Alkali Metal Salts Containing 1,2-Diaza-3,5-diborolyl Ligands (2.4 – 2.6)...... 39 2.3.1. Synthesis and Spectroscopic Characterization of Alkali Metal Salts Incorporating the 1,2-Diaza-3,5-diborolyl Ligands (2.4a, 2.4b – 2.6b and 2.4c – 2.6c)...... 39

2.3.2. The X-ray Structures of Lithium (2.4a and 2.4c(thf)3), Sodium (2.5b,

2.5b(thf)3 and 2.5c(thf)3) and Potassium (2.6b(thf), 2.6b(thf)2 and

v 2.6c(thf)) Complexes Incorporating the 1,2-Diaza-3,5-diborolyl Ligands...... 44 2.4. Synthesis and Spectroscopic Characterization of the Bicyclic 1,2-Diaza-3,5- diborolidine (2.3e) and their Alkali Metal Complexes (2.4e – 2.6e)...... 58 2.5. Conclusions ...... 61

CHAPTER THREE: Group 14 Metallocenes Incorporating 1,2-Diaza-3,5-diborolyl Ligands: Silicon, Germanium and Tin Complexes ...... 63

3.1. Introduction ...... 63 3.2. Synthesis and Spectroscopic Characterization of the Trichlorosilyl-1,2-diaza- 3,5-diborolyl Complexes (3.1) and the X-ray Structure of 3.1b...... 66 3.3. Synthesis, Spectroscopic Characterization and X-ray Structure of Bis(1,2- diaza-3,5-diborolyl)germanium (3.2) and Bis(1,2-diaza-3,5-diborolyl)tin (3.3) Complexes...... 70 3.4. Synthesis, Spectroscopic Characterization of Bis(1,2-diaza-3,5-diborolyl) chloromethyl tin chloride (3.4) and 1,2-Diaza-3,5-diborolyltin chloride (3.5) and the X-ray Structure of 3.4...... 79 3.5. Synthesis and Spectroscopic Characterization of Cationic 1,2-Diaza-3,5- diborolyltin Borate Complexes (3.6)...... 84 3.6. Conclusion ...... 86

CHAPTER FOUR: Group 12 Metallocenes Incorporating 1,2-Diaza-3,5-diborolyl Ligands: Zinc, Cadmium and Mercury Complexes ...... 88

4.1. Introduction ...... 88 4.2. Synthesis and Spectroscopic Characterization of Bis(1,2-diaza-3,5-diborolyl) Zinc (4.1), Cadmium (4.2) and Mercury (4.3) Sandwich Complexes...... 90 4.3. X-ray Structures of the Bis(1,2-diaza-3,5-diborolyl) Zinc (4.1), Cadmium (4.2) and Mercury (4.3) Sandwich Complexes...... 95 4.4. Conclusion ...... 104

vi CHAPTER FIVE: Group 4, 8 and 9 Metallocenes Incorporating 1,2-Diaza-3,5- diborolyl Ligands, and a Ruthenium Triple-Decker Sandwich

Containing an Unusual Eight-Membered C2B4N2 Ring ...... 106

5.1. Introduction ...... 106 5.2. Synthesis, Spectroscopic Characterization and X-ray Structures of the 1,2- Diaza-3,5-diborolyl Rhodium Cyclooctadiene Complexes (5.1) ...... 108 5.3. Synthesis, Spectroscopic Characterization, Electrochemical Studies and X-ray Structures of Bis(1,2-diaza-3,5-diborolyl)iron (5.2) and 1,2-Diaza-3,5- diborolyl Ruthenium Pentamethylcyclopentadienyl (5.3) Complexes...... 115 5.4. Synthesis, Spectroscopic Characterization and the X-ray Structure of the 1,2- Diaza-3,5-diborolylzirconium dichloropentamethylcyclopentadienyl Complex (5.4)...... 125 5.5. Synthesis and Spectroscopic Characterization of the 1,2-diaza-4-oxa-3,5- diborolidines (5.6) and the X-ray Structure of 5.6c...... 130 5.6. Synthesis, Spectroscopic Characterization and X-ray Structures of 1,5-Diaza- 2,4,6,8-tetraborolidine (5.7), its Mono and Dipotassium Salts (5.8 and 5.9) and

the Triple-decker Ruthenocene (6.10) with C2B4N2 Ring...... 136 5.7. Conclusion ...... 149

CHAPTER SIX: Alkali Metal and Rhodium Complexes Incorporating 1,2,4-Triaza-

3,5-diborolyl Ligand and a Tricyclic BN Compound with B4N8 Framework ...... 152

6.1. Introduction ...... 152 6.2. Synthesis, Spectroscopic Characterization and X-ray Structures of Lithium (6.1a), Sodium (6.1b) and Potassium (6.1c) Complexes Incorporating the 1,2,4-Triaza-3,5-diborolyl Ligand...... 154 6.3. Synthesis, Spectroscopic Characterization and the X-ray Structure of a Dimeric Rhodium (6.2) Complex Incorporating the 1,2,4-Triaza-3,5-diborolyl Ligand ...... 163

vii 6.4. Synthesis and Spectroscopic Characterization of the Alkali Metal Triphenylborane Adducts (6.3) and the X-ray Structures of the Lithium (6.3a) and Potassium Adducts (6.3c)...... 167 6.5. Synthesis, Spectroscopic Characterization and the X-ray Structures of the Tricyclic Dilithiodiborate (6.5) and Tetrahydrazidotetraborane (6.6)

Complexes with B4N8 Framework...... 171 6.6. Conclusion ...... 181

CHAPTER SEVEN: Conclusions and Future Research ...... 183

7.1 Coordination Chemistry of 1,2-Diaza-3,5-diborolyl Ligands ...... 183 7.2. Coordination Chemistry of 1,2,4-Triaza-3,5-diborolyl Ligand ...... 189 7.3. Suggested Future Research Directions ...... 191

CHAPTER EIGHT: Experimental Methods ...... 195

8.1. General Considerations ...... 195 8.1.1. Solvents ...... 195 8.1.2. Reagents ...... 196 8.1.3. Instrumentations ...... 197 8.1.4. X-ray Crystallography ...... 198 8.1.5. Computational Calculations ...... 199 8.2. Experimental Details for Chapter 2 ...... 200 8.3. Experimental Details for Chapter 3 ...... 213 8.4. Experimental Details for Chapter 4 ...... 220 8.5. Experimental Details for Chapter 5 ...... 224 8.6. Experimental Details for Chapter 6 ...... 234

REFERENCES ...... 243

APPENDIX ONE: Tabulated Selected Crystal Data Collection Parameters and Selected Bond Lengths and Bond Angles ...... 263

viii List of Tables

Table 1.1. Comparison of the catalyst efficiency in ethylene/1-octene polymerization for various zirconium derivatives featuring boron heterocycles...... 13

Table 5.1. Reversible oxidation potentials of 5.2 and 5.3 and cyclopentadienyl analogs of iron and ruthenium complexes (vs. SCE).a...... 124

Table 2.1. Selected Data and Structure Refinement Details for 2.3c, 2.3d and 2.4a. .... 264

Table 2.2. Selected Data and Structure Refinement Parameters for 2.5b, 2.5b(thf)3, 2.6b(thf) and 2.6b(thf)2...... 265

Table 2.3. Selected Data and Structure Refinement Parameters for 2.4c(thf)3, 2.5c(thf)3 and 2.6c(thf)...... 266

Table 2.4. Selected Bond Lengths (Å) and Bond Angles (°) for 2.3c...... 267

Table 2.5. Selected Bond Lengths (Å) and Bond Angles (°) for 2.3d...... 267

Table 2.6. Selected Bond Lengths (Å) and Bond Angles (°) for 2.4a...... 268

Table 2.7. Selected Bond Lengths (Å) and Bond Angles (°) for 2.5b, 2.5b(thf)3, 2.6b(thf) and 2.6b(thf)2...... 269

Table 2.8. Selected Bond Lengths (Å) and Bond Angles (°) for 2.4c(thf)3, 2.5c(thf)3, and 2.6c(thf)...... 271

Table 3.1. Selected Data and Structure Refinement Parameters for 3.1b · 0.25 C5H12, 3.2, 3.3a, 3.3b and 3.4...... 273

Table 3.2. Selected Bond Lengths (Å) and Bond Angles (°) for 3.1b · 0.25 C5H12...... 274

Table 3.3. Selected Bond Lengths (Å) and Bond Angles (°) for 3.2...... 275

Table 3.4. Selected Bond Lengths (Å) and Bond Angles (°) for 3.3b...... 276

Table 3.5. Selected Bond Lengths (Å) and Bond Angles (°) for 3.3a...... 277

Table 3.6. Selected Bond Lengths (Å) and Bond Angles (°) for 3.4...... 278

Table 4.1. Selected Data and Structure Refinement Details for 4.1a, 4.1b, 4.2a · BrLi(thf)3, 4.2b and 4.3...... 279

Table 4.2. Selected Bond Lengths (Å) and Bond Angles (°) for 4.1a...... 280

Table 4.3. Selected Bond Lengths (Å) and Bond Angles (°) for 4.1b...... 282

ix Table 4.4. Selected Bond Lengths (Å) and Bond Angles (°) for 4.2a · LiBr(thf)3...... 283

Table 4.5. Selected Bond Lengths (Å) and Bond Angles (°) for 4.2b...... 284

Table 4.6. Selected Bond Lengths (Å) and Bond Angles (°) for 4.3...... 285

Table 5.2. Selected Data and Structure Refinement Parameters for 5.1b, 5.1e, 5.2, 5.3 and 5.4...... 286

Table 5.3. Selected Data and Structure Refinement Parameters for 5.6c, 5.7, 5.9(tmeda)2 and 5.10...... 287

Table 5.4. Selected Bond Lengths (Å) and Bond Angles (°) for 5.1b...... 288

Table 5.5. Selected Bond Lengths (Å) and Bond Angles (°) for 5.1e...... 289

Table 5.6. Selected Bond Lengths (Å) and Bond Angles (°) for 5.2...... 290

Table 5.7. Selected Bond Lengths (Å) and Bond Angles (°) for 5.3...... 291

Table 5.8. Selected Bond Lengths (Å) and Bond Angles (°) for 5.4...... 292

Table 5.9. Selected Bond Lengths (Å) and Bond Angles (°) for 5.6c...... 293

Table 5.10. Selected Bond Lengths (Å) and Bond Angles (°) for trans-5.7...... 293

Table 5.11. Selected Bond Lengths (Å) and Bond Angles (°) for 5.9(tmeda)2...... 294

Table 5.12. Selected Bond Lengths (Å) and Bond Angles (°) for 5.10...... 295

Table 6.1. Selected Data and Structure Refinement Parameters for 6.1a, 6.1b, 6.1c, and 6.2 · thf...... 296

Table 6.2. Selected Data and Structure Refinement Parameters for 6.3a(CH3CN)3, 6.3c, 6.5(thf) and 6.6...... 297

Table 6.3. Selected Bond Lengths (Å) and Bond Angles (°) for 6.1a...... 298

Table 6.4. Selected Bond Lengths (Å) and Bond Angles (°) for 6.1b...... 299

Table 6.5. Selected Bond Lengths (Å) and Bond Angles (°) 6.1c...... 300

Table 6.6. Selected Bond Lengths (Å) and Bond Angles (°) for 6.2 · thf...... 301

Table 6.7. Selected Bond Lengths (Å) and Bond Angles (°) for 6.3a(CH3CN)3...... 302

Table 6.8. Selected Bond Lengths (Å) and Bond Angles (°) for 6.3c...... 303

x Table 6.9. Selected Bond Lengths (Å) and Bond Angles (°) for 6.5(thf)...... 304

Table 6.10. Selected Bond Lengths (Å) and Bond Angles (°) for 6.6...... 305

xi List of Figures

Figure 1.1. Most common coordination modes of the Cp ligands in metal complexes: η5 (pentahapto), η3 (trihapto) and η1 (monohapto)...... 2

Figure 1.2. Examples of representative metallocenes utilized in various applications...... 2

Figure 1.3. Side and top views of the molecular structure of 1.30a and 1.30b, illustrating the two diastereoisomers BN/BN (a) and BN/NB (b)...... 11

Figure 1.4. Ruthenocenes with boron-containing heterocyclic ligands...... 12

t − t − Figure 1.5. The coordination modes of the [P2C3 Bu3] and [P3C2 Bu2] rings...... 17

Figure 1.6. Different coordination modes of the heterocyclic ligands in metal complexes...... 26

1 Figure 2.1. The H NMR spectra of 2.3a in THF-d8 (a), 2.3b in C6D6 (b), 2.3c in THF-d8 (c) and 2.3d in toluene-d8 (d) at 25 ºC...... 35

Figure 2.2. Molecular structures of 2.3c (left) and 2.3d (right) with 50% probability thermal ellipsoids; all hydrogen atoms of the organic substituents omitted for clarity...... 37

1 Figure 2.3. The H NMR spectrum of the lithium salt 2.4a in THF-d8 at 25 ºC...... 40

Figure 2.4. Fragment of the polymeric structure of the lithium salt 2.4a with thermal ellipsoids drawn at the 50% probability level and all hydrogen atoms omitted for clarity...... 45

Figure 2.5. Side view of the sodium salts 2.5b (left), 2.5b(thf)3 (center) and 2.5c(thf)3 (right) illustrating the slight folding of the CB2N2 rings along the B(1)⋅⋅⋅N(1) axis. For clarity, only the metal and ring atoms are represented...... 47

Figure 2.6. Fragment of the polymeric structure of the sodocene 2.5b with 50% probability thermal ellipsoids. For clarity, only the α-carbon atoms of the isopropyl substituents are represented. The disorder involving the Na(2) atom and all hydrogen atoms have been omitted...... 49

Figure 2.7. Two views of the molecular structure of 2.5b(thf)3, with thermal ellipsoids drawn at 50% probability level. Left: only the ipso carbon of the phenyl group on B1 is represented. Right: perpendicular projection onto the CB2N2 plane with all THF molecules and all hydrogen atoms omitted for clarity ...... 51

xii Figure 2.8. Molecular structures of 2.4c(thf)3 (left) and 2.5c(thf)3 (right) with 50% probability thermal ellipsoids. For clarity, only the ipso carbon atom of the phenyl substituent on B1 and B21 atoms are represented, and all hydrogen atoms omitted. Bottom: perpendicular projection of the metal ions onto the CB2N2 planes, with the three tetrahydrofuran molecules omitted for clarity...... 53

Figure 2.9. Fragment of the polymeric chain of 2.6b(thf) and 2.6b(thf)2 in the solid- state with 50% probability level thermal ellipsoids. For clarity, only the α- carbon atoms of the organic substituents are represented and all hydrogen atoms omitted...... 54

Figure 2.10. Polymeric structure of 2.6c(thf) in the solid state, with thermal ellipsoids drawn at 50% probability level. For clarity, only the ipso carbon atoms of the phenyl substituents on the nitrogen atoms are represented and all hydrogen atoms omitted...... 57

1 Figure 2.11. H NMR spectra of 2.3e in C6D6 (a) and the lithium salt 2.4e in THF-d8 (b) at 25 ºC...... 60

Figure 3.1. Molecular structure of 3.1b with 50% probability level thermal ellipsoids and all hydrogen atoms and a co-crystallized pentane molecule omitted for clarity...... 68

1 Figure 3.2. H NMR spectra of 3.3b (a) and 3.3a (b) recorded in C6D6 at 25 °C...... 71

Figure 3.3. Molecular structure of 3.2 with 50% probability level thermal ellipsoids. For clarity, only the ipso carbon atom of the phenyl substituent on B1 is shown and all hydrogen atoms omitted...... 73

Figure 3.4. View of the perpendicular projection onto the CB2N2 planes, illustrating the hapticity of the ligands. From left to right: 3.2 (η3,η3), 3.3b (η3,η3) and 3.3a (η4,η4)...... 74

Figure 3.5. Molecular structure of 3.3b (top) and 3.3a (bottom) with 50% probability level thermal ellipsoids. For clarity, only the α-carbon of the isopropyl substituent on N4 atom is represented (bottom) and all hydrogen atoms omitted...... 75

Figure 3.6. Side view of one of the CB2N2 rings in 3.3a (left) and 3.3b (right) illustrating the envelope conformation of the ring structure with folding along the C⋅⋅⋅N axis. For clarity, only the metal and ring atoms are represented...... 77

1 Figure 3.7. H NMR spectrum of complex 3.4 recorded in C6D6 at 25 °C...... 81

xiii Figure 3.8. Molecular structure of 3.4 with thermal ellipsoids drawn at the 50% probability level and all hydrogen atoms omitted for clarity...... 81

1 Figure 3.9. H NMR spectrum of the NMR reaction of 3.3a with [H(Et2O)2]B(C6F5)4 in THF-d8 at 25 °C, revealing a 1:1 molar formation of the 1,2-diaza-3,5- diborolyl tin borate 3.6a, and the neutral 1,2-diaza-3,5-diborolidine 2.3a. ... 85

Figure 4.1. Representative 1H NMR spectra of zincocene 4.1a (a), cadmocene 4.2a (b) and mercurocene 4.3 (c) in C6D6 at 25 °C ...... 92

1 Figure 4.2. (a) The H NMR spectrum of 4.2b in THF-d8, (b) the proton coupled 113 13 Cd NMR spectrum of 4.2b in C6D6 and (c) a portion of the C NMR spectrum of 4.2b in THF-d8 at 25 °C...... 94

Figure 4.3. Side view illustrating the envelope conformation of one of the CB2N2 rings in 4.2a · BrLi(thf)3 with folding along the B(3)···B(4) (left) and C(11)···N(3) axes (right). For clarity, only the metal and the ring atoms are represented...... 95

Figure 4.4. Molecular structure of one of the two independent molecules in 4.1a with 50% probability level thermal ellipsoids. All hydrogen atoms omitted for clarity. The disorder involving the isopropyl groups (C15 and C18) is not depicted...... 98

Figure 4.5. Molecular structure of 4.1b with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity...... 98

Figure 4.6. Molecular structure of 4.2a · BrLi(thf)3 with 50 % probability level thermal ellipsoids and all hydrogen atoms omitted for clarity. The disorder involving the bromine atom and the isopropyl groups (C5 and C8) is not depicted...... 100

Figure 4.7. Molecular structure of 4.2b with 50 % probability level thermal ellipsoids and all hydrogen atoms omitted for clarity...... 100

Figure 4.8. Molecular structure of 4.3 with 50 % probability level thermal ellipsoids. For clarity, only two of the disordered ligands around the mercury center are represented and all hydrogen atoms omitted. Bottom: a view of the disorder in the crystal structure of 4.3. The two molecules are equally abundant and only the metal and ring atoms are represented...... 103

1 Figure 5.1. H NMR spectra of rhodium compounds 5.1a recorded in THF-d8 (a), 5.1b (b) and 5.1e recorded in C6D6 (c) at 25 °C...... 109

Figure 5.2. A side view of 5.1b (left) and 5.1e (right) illustrating the envelope conformation of the CB2N2 ring. For clarity, only the metal and the ring atoms are represented...... 111

xiv Figure 5.3. Molecular structures of 5.1b (top) and 5.1e (bottom) with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity...... 113

Figure 5.4. Perpendicular projection onto the CB2N2 planes, illustrating the coordination mode of the ligands. From left to right: 5.1b, 5.1e, 5.2 (two CB2N2 rings) and 5.3...... 114

1 Figure 5.5. H NMR spectra of 5.2 (a) in CD2Cl2 and 5.3 (b) in C6D6 at 25 °C...... 117

Figure 5.6. Views of the molecular structure of 5.2 with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity. Bottom: top view of the structure, illustrating the orientation of the heterocyclic rings. . 118

Figure 5.7. Views of the molecular structure of 5.3 with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity. Bottom: top view of the structure, illustrating the relative position of the CB2N2 and Cp* rings...... 120

Figure 5.8. Cyclic voltammograms of 5.2 (a) and 5.3 (b) recorded in 0.1 M nBu4PF6 / DCM at 25 °C with a platinum working electrode and potential reported versus SCE at a scan rate = 50 mV/s...... 123

Figure 5.9. Molecular structure of 5.4 with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity...... 127

Figure 5.10. Side view of 5.4 displaying the folding of the CB2N2 ring along the B(1)⋅⋅⋅B(2) axis and the perpendicular projection onto the best plane of the CB2N2 ring...... 128

Figure 5.11. Molecular structure of 5.6c with 50 % probability level thermal ellipsoids and all hydrogen atoms omitted for clarity...... 132

Figure 5.12. Top: molecular structure of trans-isomer of 5.7 with 50% probability thermal ellipsoids and hydrogen atoms of the organic substituents omitted for clarity. Bottom: side view of 5.7 illustrating the folding of the bicyclic framework along the B(1)⋅⋅⋅B(2) axis. For clarity, only the ipso carbons of the phenyl groups are represented...... 138

Figure 5.13. Representative 1H NMR studies of the stepwise deprotonation of the 1 heterobicyclic C2B4N2 compound. From top to bottom: H NMR spectra of 5.7, 5.8 and 5.9 recorded in THF-d8 at 25 °C...... 139

Figure 5.14. Top: molecular structure of 5.9(tmeda)2 with 50% probability thermal ellipsoids and all hydrogen atoms omitted for clarity. Bottom: side view of 5.9(tmeda)2 displaying the orientation of the ring structure; substituents and the TMEDA molecules omitted for clarity...... 141

xv Figure 5.15. Top: molecular structure of 5.10 with 50% probability level thermal ellipsoids. For clarity, only the ipso carbons of the phenyl groups are represented and all hydrogen atoms omitted. Bottom: side and top views of 5.10. For clarity, only the metal and the ring atoms of the middle deck are represented...... 144

Figure 5.16. Cyclic voltammogram of 5.10 in 0.1 M nBu4PF6 / THF at 25 °C recorded at platinum working electrode versus SCE, scan rate = 500 mV/s...... 148

Figure 6.1. (a) The 1H NMR spectrum of the 1,2,4-triaza-3,5-diborolidine 1.63b and (b) 1,2,4-triaza-3,5-diborolyl sodium salt 6.1b in THF-d8 at 25 °C...... 156

Figure 6.2. Fragments of the 2D polymeric structures of 6.1a (top) and 6.1b (bottom), illustrating the environment of the ligand, with 50% probability level thermal ellipsoids. For clarity, only the ipso carbon atoms of the phenyl groups are represented and all hydrogen atoms on the organic substituents omitted...... 158

Figure 6.3. Fragment from the 1D polymeric structure of 6.1c with 50% probability level thermal ellipsoids. For clarity, only the ipso carbon atom of the phenyl group on B2 is represented and all hydrogen atoms on the organic substituents omitted...... 160

1 13 Figure 6.4. H and C NMR spectra of 6.2 recorded in CD2Cl2 at 25 °C...... 164

Figure 6.5. Molecular structure of 6.2 with 50% probability level thermal ellipsoids. For clarity, only the ipso carbon atoms of the phenyl groups are represented and all hydrogen atoms on the organic substituents omitted. ... 165

Figure 6.6. Molecular structure of 6.3a(CH3CN)3 with 50% probability level thermal ellipsoids. Hydrogen atoms of the organic substituents omitted for clarity. 169

Figure 6.7. Fragment of the 1D polymeric structure of 6.3c with 50% probability level thermal ellipsoids. Hydrogen atoms of the organic substituents omitted for clarity ...... 169

1 1 11 Figure 6.8. (a) H NMR spectrum of 6.4 recorded in C6D6 at 25 °C. (b) H and B 1 NMR spectra of 6.5 recorded in THF-d8 at 25 °C. (c) H NMR spectrum of 6.6 recorded in THF-d8 at 25 °C...... 173

Figure 6.9. Molecular structure of 6.5 (top) and 6.6 (bottom) with 50% probability ellipsoids and all hydrogen atoms omitted for clarity...... 176

Figure 6.10. Cyclovoltammograms of 6.6 at 50 mV/s scan rate in the absence of

internal standard [Cp2Co]PF6 (top) and at 200 mV/s scan rate in the 0/+1 0 presence of [Cp2Co]PF6 ([Cp2Co] with E = –1.36 V vs. ferrocene and -0.82 V vs. the SCE)...... 179

xvi List of Abbreviations

∠ angle acac acetylacetone

AgOTf silver triflate cod 1,5-cyclooctadiene

Cp cyclopentadienyl

Cp* pentamethylcyclopentadienyl d doublet

DCM dichloromethane

DME dimethoxyethane

DMF dimethylformamide

EA elemental analysis

ESI electrospray ionization

Fc ferrocene

HRMS high resolution mass spectroscopy

KHMDS potassium hexamethyldisilazane

LDA lithium diisopropylamide

LiTMP lithium 2,2,6,6-tetramethylpiperidide m multipet

MAO methylaluminoxane

NaHMDS sodium hexamethyldisilazane nBuLi n-butyllithium

PMDETA pentamethyldiethylenetriamine

Py pyridine

xvii Pz-boc bis-tert-butoxycarbonylpyrazolidine q quartet

RCM Ring-closing metathesis

SCE standard calomel electrode sep septet t triplet

THF tetrahydrofuran

TMEDA tetramethylethylenediamine

TMP 2,2,6,6-tetramethylpiperidine

xviii List of Publications

Published works from Chapter Two:

1. Ly, H. V.; Forster, T. D.; Maley, D.; Parvez, M.; Roesler, R. Chem. Commun. 2005, 4468 – 4470.

2. Ly, H. V.; Forster, T. D.; Corrente, A. M.; Eisler, D. J.; Konu, J.; Parvez, M.; Roesler, R. Organometallics 2007, 26, 1750 – 1756.

The experimental works were done by Hanh Ly, Taryn Forster and Darren Maley. The X-ray crystallographic analysis was done by Andrea Corrente, Dr. D. J. Eisler, Dr. J. Konu and Dr. M. Parvez.

Published works from Chapter Four:

3. Ly, H. V.; Forster, T. D.; Parvez, M.; McDonald, R.; Roesler, R. Organometallics 2007, 26, 3516 – 3523.

The experimental works were carried out by Hanh Ly. The X-ray crystallographic analysis was performed by Dr. M. Parvez and Dr. R. McDonald.

Published works from Chapter Five:

4. Ly, H. V.; Tuononen, H. M.; Parvez, M.; Roesler, R. Angew. Chem. Int. Ed. 2007, 46, ASAP.

The experimental works were carried out by Hanh Ly. The theoretical calculation was performed by Dr. H. M. Tuononen and the X-ray crystallographic analysis was done by Dr. M. Parvez.

Published works from Chapter Six:

5. Ly, H. V.; Tuononen, H. M.; Parvez, M.; Roesler, R. Chem. Commun. 2007, 4522 – 4524.

6. Ly, H. V.; Chow, J. H.; Parvez, M.; McDonald, R.; Roesler, R. Inorg. Chem. 2007, 46 9303 – 9311.

The experimental works were carried out by Hanh Ly. The theoretical calculation was performed by Dr. H. M. Tuononen and the X-ray crystallographic analysis was done by Dr. M. Parvez and Dr. R. McDonald.

xix 1

CHAPTER ONE

Introduction

1.1 Cyclopentadienyl

Cyclopentadienyl (Cp) and its derivatives are the most widely used organic

ligands in organometallic chemistry. The discovery of ferrocene (1.1) in 19511 with its fascinating sandwich structure has caught the attention of numerous chemists. Shortly thereafter a variety of metal cyclopentadienyl complexes (metallocenes) were synthesized and characterized. For more than half a century, the chemistry of metallocenes has been exhaustively investigated experimentally and theoretically. Cyclopentadienyl complexes can be classified into three categories based on the type of bonding between the metals and the Cp ligands: π-complexes, σ-complexes and ionic complexes. The term

“hapticity” is used to describe the coordination modes of the Cp ligand to the central metal atom. It is denoted by the Greek letter η (eta) followed by a superscripted number

(e.g. ηx, x = 1 – 5) representing the number of carbon atoms of the ligand that are coordinated to the metal center (Figure 1.1). The most common bonding modes in cyclopentadienyl complexes is η5, with all five carbon atoms of the ligand coordinated to

the metal center. In complexes of metal centers having a complete valence shell (18

electrons), η3 and η1 coordination modes can be preferred.

M MLn M MLn MLn Ln Ln

5 - η3-Cp (4π-e- donor) η1-Cp (2π-e- donor) η -Cp (6π-e donor)

Figure 1.1. Most common coordination modes of the Cp ligands in metal complexes: η5

(pentahapto), η3 (trihapto) and η1 (monohapto).

Aside from their intriguing bonding and structural motifs, the numerous

applications of various metallocenes are of great importance in materials science and

especially in catalysis. One of the best known applications is provided by the group 4

metallocene derivatives (1.2), which have proven to be efficient catalysts for olefin polymerization.2 Group 8 metallocenes have been used as building blocks in

organometallic polymers (i.e. poly(vinylferrocene) 1.3) possessing interesting electronic

properties,3 and ferrocenes containing chiral ligands (1.4) have been utilized in

asymmetric catalysis.4

H2 H C C P R n Ar Me Fe M Fe Fe Me

1.1 1.2 1.3 1.4 M = Ti, Zr R = alkyl Ar = aryl

Figure 1.2. Examples of representative metallocenes utilized in various applications.

o o o B N H H 1 Cp [ H]-boratabenzene benzene pyridinium

The six-membered heterocyclic [1H]-boratabenzene, isoelectronic with benzene and pyridinium, and its derivatives have also been subject to intense studies, primarily due to the promising catalytic properties of the zirconium derivatives in olefin polymerization.5 Given that boratabenzenes are 6π-electron anions, they closely resemble the Cp anions and can be considered heterocyclic Cp analogs. The first boratabenzene derivative was reported in 1970 by Herberich and coworkers.6a Reaction of cobaltocene with boron halides yielded complexes 1.5 and 1.6 through insertion of the boranediyl group into the Cp rings, and the alkali metal salts of boratabenzenes (1.7) were isolated by degradation of 1.6 with alkali metal cyanides (Scheme 1.1a).6b A more general synthetic approach for the synthesis of lithium boratabenzenes that serve as useful precursors to other metal complexes was established by Ashe and coworkers, as shown in

Scheme 1.1b.6c

Scheme 1.1.

BR BR

RBX2 MCN (a) Co Co + Co B R = Me, Ph M = Na, K M BR R

1.5 1.6 1.7

Bu SnH RBX LDA (b) 2 2 2 Sn B B Bu2 R Li R

The remarkable success of cyclopentadienyl and boratabenzene as ligands in organometallic chemistry has prompted the search for other heterocyclic analogs containing heteroatomic skeletons. The primary focus of this thesis is the study of novel heterocyclic Cp analogs containing boron (BR−) and nitrogen (NR’+) fragments, and their coordination chemistry. This chapter will provide a brief summary of the coordination chemistry of five-membered heterocyclic Cp analogs. The objectives and outline of the thesis will be also addressed.

1.2. Heterocyclic Analogs of Cyclopentadienyl

1.2.1. General Considerations

The coordination studies of heterocyclic analogs of cyclopentadienyl have attracted attention primarily due to the possibility to tune the electronic properties of the

Cp ligands by modifying the skeletal configuration of the five-membered ring. Formally, the preparation of heterocyclic Cp analogs involves the substitution of skeletal carbon fragments, CH, in the ring by isoelectronic main group fragments:

BH o CH o NH o O

To date, the synthesis and characterization of various five-membered heterocyclic analogs having up to four ring carbons substituted by isolobal heteroatoms has been

described. Carbon-free cyclopentadienyl analogs containing a homoatomic E5 framework

(E = P, As and Sb) have also been reported and the coordination chemistry of the lighter analogs has been thoroughly investigated.7

1.2.2. Cyclopentadienyl Analogs with C4E Frameworks

Metal complexes containing monoanionic ligands with C4E skeletons, where E

belongs to group 14 and 15 elements, have been isolated. In particular, a series of

diheteroferrocenes (1.10) featuring group 15 heterocyclopentadienyls (1.9), including pyrrolyl,8 phospholyl,9 arsolyl,10 stibolyl11 and bismolyl12 has been synthesized and

characterized. The preparation of the heavier group 15 heteroles (1.8b – e) was achieved

utilizing the synthetic route developed by Fagan and Nugent,13 involving the zirconium-

mediated alkyne coupling followed by electrophilic substitution with corresponding

group 15 metal halides (Scheme 1.2, route A). The alternative route B for the synthesis

of stibole and bismole derivatives was reported by Ashe.14 The 1,4-diiodobutadiene and

1,4-dilithiobutadiene were prepared in this fashion through halogen/metal exchange and

were subsequently subjected to ring closing with metal halides.

Scheme 1.2.

2 R2 R R2 I R1 R2 I Mg 2 R1 + 1 1 1 R - MgCl2 R Zr R route B Cp2ZrCl2 I R2 Cp Cp

PhECl 2 nBuLi route A 2 - Cp2ZrCl2

PhSbX2 2 R2 R2 or R Li 1 PhBiX2 R E R1 1 1 - Cp2ZrX2 2 P 1.8b As 1.8c R E R Li R Sb 1.8d Bi 1.8e Ph

The 1,1-diheteroferrocenes (1.10) were synthesized by reaction of the corresponding 1-phenylheteroles (1.8) with lithium metal in tetrahydrofuran (THF)

followed by reaction with FeCl2 (Scheme 1.3). The structural analysis of 1,1- heteroferrocenes confirmed the η5-coordination of the planar and parallel heterocyclic ligands to the iron center.8-12 The alkali metal complexes of the group 15 heterocycles can

be generated in solution, however, their structural characterization was rather limited. In

fact, only two crystal structures of the lithium derivatives of the phospholide

15 16 Li(C4Me4P) and arsolide Li(C4Me4As) have been reported. These derivatives feature

monomeric structures with the heterocyclic ring η5-bonded to the lithium ion and

additionally chelated by a TMEDA ligand. The C-C bond lengths within the heterocycles

(1.37 – 1.43 Å) are similar, indicating a high degree of delocalization. In addition, the P-

C and As-C bonds are shorter than the P-C and As-C single bonds, consistent with partial

double-bond character.

Scheme 1.3.

1 R2 R

2 2 2 2 E R R R R R2 E 1) NH or AlCl 1 Li / THF 3 3 R a = N Fe 2 R1 b = P 1 1 - LiPh 1 1 2) FeCl R R E R R E R 2 c = As R2 E d = Sb Ph Li R1 e = Bi 1.8 1.9 1.10

Symmetric ruthenium metallocenes containing the pyrrolyl (1.11)17 and phospholyl

(1.12)18 ligands were prepared by metathesis reaction of the corresponding lithium salts

with LnRu(II)Cl2 reagents (Ln = cod, C6H6, PPh3). The structural characterization of

diheteroruthenocenes revealed classical sandwich arrangements with ligands binding η5 to the ruthenium atom.

Scheme 1.4.

Me Me Li nBuLi, -78 oC 1/4 [Cp*RuCl]4 Me Ge Me E Si(SiMe ) Me Me Si(SiMe3)3 3 3 1.14 Ru Me E Me

(Me3Si)3Si H (thf)3LiSi(SiMe3)3 E = Si (1.13a) Ru H E = Ge (1.15) 1) [Cp*Ru(μ-OMe)]2 Ge (1.13b) 2) Me SiOTf Si (1.17) 3 Si 3) NaBPh4 Si(SiMe3)3 1.16

Similar metal complexes incorporating the heavier group 14 elements, silicon and

germanium in the ring backbone have been reported.19-22 The first isolated η5- germolylruthenium complex (1.15) was prepared by reaction of [Cp*RuCl]4 with the

germolyl anion (1.14) that was generated in situ by the deprotonation of 1.13b with n- butyllithium (nBuLi) at -78 °C (Scheme 1.4).19 The spectroscopic characterization of

1.15 revealed two resonances for the ring carbon atoms with similar chemical shifts at δ

80.23 and 87.82 ppm, reflecting the aromatic character of the C4Ge ring. The crystal

structure of 1.15 showed that the planar C4Ge and Cp* rings were bonded to the

ruthenium atom in an η5 fashion. The germanium atom deviates only slightly (ca. 0.02

Å) from the best plane of the C4Ge ring and the sum of the bond angles at the Ge atom

(358.1°) is consistent with the sp2 hybridization. The synthesis of the analogous silolylruthenium complex (1.17), and its characterization in solution, was also published.20 In contrast to the germanium analog, compound 1.17 was obtained by

deprotonation of the cationic ruthenium complex (1.16) that was prepared directly via

oxidative addition of 1.13a to the Cp*Ru+ fragment. The crystal structure of 1.16

revealed a slightly bent sandwich structure with a centroid-Ru-centroid angle of

164.2(3)°. The five-membered C4Si ring is nearly planar with the bulky Si(SiMe3)3 substituent bent away from the Ru center.

Scheme 1.5.

Me Me Me Me M MCH2Ph Cp*M'Cl3 Cl Hf Cl Me Me E Me E Me M = Li, K M' = Zr, Hf E SiMe3 Me3Si SiMe3 SiMe3

E Li K E Zr Hf E = Si (1.18a) Ge (1.18b) Si 1.19a 1.19c Si 1.21 1.22a Ge 1.19b Ge 1.22b

The hafnium metallocenes (1.22) containing the silolyl and germolyl ligands were

prepared by the desilylation of the silole (1.18a) and germole (1.18b) with lithium and

potassium benzyl (Scheme 1.5), followed by reaction of the lithium salts 1.19 with

21a Cp*HfCl3, to yield the silolyl and germolyl complexes (1.22) in low yields. An optimization of the synthesis was later reported, utilizing the magnesium derivative,

1 Mg[η -C4Me4SiSiMe3]2 (1.20), prepared by treatment of MgBr2(Et2O) with the

21b potassium salt 1.19c in THF. Reaction of 1.20 with Cp*M’Cl3 (M’ = Zr and Hf)

resulted in the quantitative formation of the zirconium and hafnium complexes 1.21 and

1.22a. The molecular structures of these group 4 metallocenes feature the expected bent

sandwich geometry with two planar ligands bonded to the metal atom in an η5 fashion.

1.2.3. Cyclopentadienyl Analogs with C5-nEn (n = 2 – 4) Frameworks

The chemistry of heterocyclic Cp analogs continues to be a fertile area of research

and a variety of anionic ligands containing isoelectronic main-group elements, such as B,

N, P, Si, Ge, O and S, have been isolated. The synthesis of the 1,2-heteroborolyls with

− 22 − 23 −24 anionic C3BN , C3BO , and C3BS frameworks was of particular interest primarily

due to their isoelectronic relationship with cyclopentadienyls and boratabenzenes, as well

as to the facile tunability of their electronic properties via their substituent groups.25

Studies had shown that these ligands possess similar coordinative properties to those of

Cp and they proved to be feasible ancillary ligands in transition metal catalysis.24, 26 The synthesis of 1,2-heteroborolyls (1.26) was mainly approached by the two synthetic routes outlined in Scheme 1.6.

Scheme 1.6.

Route A Route B R Grubbs R C3H5EH B catalysts RBCl2 R'' B B Sn E base E R E R'' Cl PCy3 Cl Ph E = NR', O, S Ru 1.23 Cl 1.24 1.25 PCy3 LDA or tBuLi

E Li a = NR' B R b = S E c = O 1.26

R S R S t R S 1) BCl3 BuLi Li

2) HN(iPr)2 R Sn R B R B Bu2 N(iPr)2 N(iPr)2 1.28 1.27

R2 = H2 ,

The key step in route A involves a ring-closing metathesis (RCM)27 of heteroallyl- vinylboranes (1.23) using Grubbs catalyst to give the 1,2-heteroborolines (1.24). The alternative route involves a transmetalation of the stannacycle (1.25) with boron halides affording the corresponding 1,2-heteroborolines.28 The deprotonation of 1.24 using strong

bases led to the formation of 1.26. The 1,3-thiaborolyls (1.27) were isolated using a

similar synthetic approach (route B) via reaction of the monocyclic 1,3-thiastannolenes with borane halides, yielding the corresponding 1,3-thiaborolines 1.28. Subsequently, deprotonation using strong bases yielded the target compounds 1.27.29

Scheme 1.7.

SiMe3 N B Me o -4 N -130 C/10 torr N 3 + M + SiMe3 M SiMe3 B SiMe3 B CH3C6H11 N Me B Me Me

1.29 1.30a: M = Fe 1.31 1.30b: M = Co

The synthesis of the 1,2-azaborolyls (1.26a) and their transition metal chemistry was studied extensively by Schmid and coworkers.30 The ferrocene and cobaltocene

analogues 1.30a and 1.30b, respectively, were first prepared by reaction of the neutral

1,2-azaboroline 1.29 with transition metal atoms to form the corresponding metal

complexes and the saturated 1,2-azaborolidine 1.31 (Scheme 1.7).30b The structural

analysis of both complexes revealed two diastereoisomers, featuring the η5-coordinated

1,2-azaborolyl rings in staggered (BN/BN isomer) and eclipsed (BN/NB isomer)

arrangements (Figure 1.3).30d

SiMe SiMe N 3 N 3 B Me B Me

M M Me B N N Me3Si B Me3Si Me

SiMe 3 SiMe3 N Me N B M M B N B Me N Me3Si B Me Me3Si Me BN/BN BN/NB (a) (b) M = Fe (1.30a), Co (1.30b)

Figure 1.3. Side and top views of the molecular structure of 1.30a and 1.30b, illustrating

the two diastereoisomers BN/BN (a) and BN/NB (b).

The ruthenocene and nickelocene analogues 1.30c and 1.30d, respectively, were

prepared by treatment of the 1,2-azaborolyllithium salt that was generated in situ with the

corresponding transition metal halides.31, 32 Analogous to the iron and cobalt complexes,

the molecular structure of 1.30c exhibited the typical sandwich geometry with the Ru

atom positioned between the 1,2-azaborolyl ring centers in an η5 fashion, and nearly

parallel 1,2-azaborolyl rings (dihedral angle 8.6°).31 In contrast, the molecular structure

of 1.30d showed a considerable slip distortion from the η5 arrangement of the rings with

the Ni atom displaced from the ring center by 0.45 Å, and the structural motif is best described as an η3-complex.32 The electron-rich Ni atom shows no contact with the N

atom, resulting in Ni-N distances longer than the Ni-B distances. The 11B NMR data also

supported the structural results. Due to the lack of notable Ni-B interaction, the shift of

the 11B signal is not influenced by the coordination to the metal.

S S S B Ph B N(iPr)2 B B O N(iPr)2 N(iPr)2 Ru Ru Ru Ru

1.32a 1.32b 1.32c 1.32d

Figure 1.4. Ruthenocenes with boron-containing heterocyclic ligands.

The transition metal chemistry of thia and oxaborolyl ligands was investigated mainly by Ashe and coworkers.23, 24, 29 Similar to 1,2-azaborolyls, these ligands readily

form transition metal complexes. The reaction of [Cp*RuCl]4 with heteroborolides

resulted in the formation of the diheteroruthenocenes (1.32) containing the thia and

oxaborolyl ligands (Figure 1.4). The molecular structures of 1.32 showed that the

heteroborolyl rings are η5 bonded to the ruthenium atom and are virtually coplanar with

the Cp* ring. In the structure of 1.32a, the Ru atom is shifted away from B toward the S

atom. Similar distortions away from the boron atom are common features of the π-

coordinated boron heterocycles.33 Although disorder within the ring frameworks of the other three ruthenium structures limits the accuracy of the bond distances, the structural data proved the close similarity of the π-coordinated heteroborolyl ligands to cyclopentadienyls.

Metallocene derivatives of the group 4 metals have been extensively used as homogeneous catalysts for olefin polymerization. A series of novel zirconium complexes

(1.34) incorporating boron-containing heterocycles were prepared by reaction of 1,2- heteroborolides 1.33 with Cp*ZrCl3. These complexes were shown to catalyze the

polymerization of ethylene upon activation with excess methylaluminoxane (MAO).

Scheme 1.8.

Li Cp*ZrCl3 Cl B R Zr Cl E 1.33a: E = NEt, R = Ph B R E 1.33b: E = S, R = NiPr2 1.33c: E = O, R = NiPr 2 1.34a: E = NEt, R = Ph Me SiCl 2 2 1.34b: E = S, R = NiPr2 1.34c: E = O, R = NiPr2

Ph SiMe2Cp SiMe2Cl Me2 Si B LiCp 1.33a B R B R N Et B E E N Ph Et 1.35a: E = NEt, R = Ph 1.35b: E = S, R = NiPr 1) LDA 2 1) LDA 2) ZrCl 2) ZrCl4 4

Et N Ph B

Me Si Cl Cl 2 Zr Me2Si Zr Cl Cl R B B E Ph N Et 1.37a: E = NEt, R = Ph 1.37b: E = S, R = NiPr2 1.36

Table 1.1. Comparison of the catalyst efficiency in ethylene/1-octene polymerization for

various zirconium derivatives featuring boron heterocycles

Complex Efficiency (g of polymer / atom-gram of Zr) 1.34a 234 x 104 BN(iPr)2 BN(iPr)2 1.34b 6.0 x 104 Cl Cl 1.36 126 x 104 Zr Me Si Zr Cl 2 Cl 1.37a 66 x 104 BN(iPr) BN(iPr) 1.38 20.4 x 104 2 2 4 1.38 1.39 1.39 2.4 x 10

Under identical conditions the relative activities of 1.34a and 1.34b were measured to be

2.34 x 106 and 6.0 x 104 (g of polymer)/(atom-gram of Zr), respectively.24, 34 According

to these results, the 1,2-heteroborolyls are promising replacement ligands for Cp in

metallocene-based polymerization catalysts. Studies were extended to the synthesis of

bridging 1,2-heteroborolyl zirconium complexes, as outlined in Scheme 1.8.

Silylation of 1.33a and 1.33b with excess dichlorodimethylsilane afforded

complexes 1.35 in high yields. Sequential reaction of 1.35a with 1.33a followed by

lithium diisopropylamine (LDA) and ZrCl4 yielded the desired ansa-1,2-azaborolyl

zirconocene 1.36. The mixed ansa 1,2-heteroborolyl zirconium derivatives 1.37 were

also synthesized using the same method. The polymerization using complexes 1.36 and

1.37a with mixtures of ethylene and 1-octene was examined and the results are

summarized in Table 1.1.35 The results showed that the bridging 1,2-azaborolyl

zirconium complexes are significantly more active catalysts than the 1-boratabenzene

zirconium complexes 1.3836 and 1.39,37 and than the 1,2-thiaborolyl zirconium complex

1.34b.24 The dimethylsilyl bridging in 1.36 and 1.37a led to reduction of the polymerization activity relative to the unbridged 1,2-azaborolyl zirconium derivative

1.34a. This was explained by the lack of conformational mobility of the 1-ethyl and 2- phenyl substituents that were held in position by the bridging unit, interfering with the propagation steps. A similar but larger drop in activity was observed between bridged and unbridged 1-boratabenzene derivatives.37

The X-ray structures of the zirconium derivatives 1.34, 1.36 and 1.37 exhibited

the expected bent sandwich arrangements with the 1,2-heteroborolyl rings η5-coordinated

to Zr, with the exception of complex 1.34c where the 1,2-oxaborolyl ring is approaching

η4-ligation to the Zr center.34, 35 In all structures, the Zr atom is slipped away from the

boron atom (Zr-B distances ranging from 2.68 to 2.80 Å). This is most likely due to the

high electron demand of the electron-deficient Zr (d0 configuration), which prefers to

coordinate to the more electron-rich ring atoms.

In addition to the 1,2-heteroborolyls, the transition metal chemistry of the anionic

38 pyrazolyl ligands with C3N2 ring skeletons, has been explored. The structural

investigation of various reported pyrazolyl metal complexes revealed predominately η1 and η2-coordination modes of the ligand.39 Metal complexes containing η5-pyrazolyl

binding are extremely rare, in fact to date only the crystal structure of the ruthenium

complex 1.41a bearing an η5-pyrazolyl ligand has been reported and it is the first

documentation of such coordination mode in any metal.40 The synthesis of complexes

1.41 was accomplished by treatment of [Cp*RuCl]4 with the potassium salts 1.40 in THF

(Scheme 1.9). The pyrazolyl potassium derivatives were prepared by deprotonation of pyrazoles with KH. The molecular structure of 1.41a featured a sandwich structure with the pyrazolyl and Cp* ligands η5 bonded to the Ru center. The structural data was

5 5 supported by the theoretical studies conducted for the [(η -C5H5)Ru(η -C3H3N2)] model.

Scheme 1.9.

N N K THF R R R R KH R R reflux + 4 [Cp*RuCl]4 Ru N N - H2 NN -4 KCl H 1.40

1.41a: R = Me 1.41b: R = tBu 1.41c: R = Ph

Scheme 1.10.

tBu t P Bu Li(dme)3 P PP P t P Bu CP P t DME FeCl2 Bu Fe Fe + t + - LiCl Bu tBu LiP(SiMe3)2 PP P P t P t Bu P tBu P Bu DME = dimethyoxyethane 1.42 1.43 1.44

NiBr2 CoCl2 CrCl2(thf)2 [Ru3(CO)12]

t t Bu t Bu Bu t t t t P Bu P Bu Bu H P Bu (CO) P PP 3 P P P Ru tBu PP tBu Ni Co Cr Fe Ru(CO)4 tBu P P P P P P P Ru t t P t t (CO)3 tBu P Bu tBu P Bu tBu P Bu P Bu 1.47 1.46 1.45 1.48

The heterocyclic anionic ligands with P5-n(CR)n skeletons (n = 1 – 3) exhibit a rich transition metal coordination chemistry and have been extensively studied. The

t − 41 anionic [P3C2 Bu2] ligand (1.42) was initially reported by Becker, as a product of the

t reaction of the phosphaalkyne BuC≡P with LiP(SiMe3)2. The green air-stable pentaphospha (1.44) and hexaphosphaferrocenes (1.43) were synthesized by reaction of

1.42 with FeCl2 (Scheme 1.10), and the compounds were separated by fractional crystallization.42 The characterization of 1.43 and 1.44 by X-ray diffraction revealed typical sandwich structures with the two η5 rings binding to the iron center in an eclipsed arrangement. The chromium,43 cobalt44 and nickel45 derivatives (1.45 – 1.47) containing

t − t − the anionic [P2C3 Bu3] and [P3C2 Bu2] ligands were obtained by reaction of 1.42 with the corresponding metal halides. In the solid state, 1.45 is isostructural with the iron

analogue, whereas the diamagnetic cobalt and nickel metallocenes featured η3 and η4-

t t coordination modes of the P2C3 Bu3 and P3C2 Bu2 ligands.

Several types of ligation have been observed for the di and triphosphacyclopentadienyl ring systems. The P2C3R3 ring generally binds to metals in

either an η5 fashion as shown in complexes 1.43 – 1.45 or in an η3 mode as in the nickel

5 derivative 1.47. On the other hand, the P3C2R2 ring exhibits η -ligation in a variety of

complexes, as well as coordination through the lone pair of electrons of the P atoms in

the ring (Figure 1.5). A particularly interesting example for the utilization of the

phosphorus lone pair of electrons is the preparation of the tetrametallic complex 1.48, where the two P3C2R2 rings of the hexaphosphaferrocene were linked by the [Ru3(CO)10]

unit.46

P P

P P M M η5 η3

M M P P P M P P P

P P P P P M'' P P P P P P P M M M' M M' M'

1 1 1 4 5 1 5 1 1 5 η η :η η η η :η η :η :η

t − t − Figure 1.5. The coordination modes of the [P2C3 Bu3] and [P3C2 Bu2] rings.

Recently, Scheer and coworkers have isolated a mixture of phosphorus-based

t − t ferrocene analogs with [( BuC)nP5-n] ligands (n = 0 – 2) by reaction of 1.49 with BuC≡P

(Scheme 1.11).47 These phosphaferrocenes were separated by chromatography. The

1,2,3,4-tetraphospha (1.50) and 1,2,4-triphosphaferrocenes (1.51) were isolated as major

products and the pentaphosphaferrocene (1.52) in only 2% yield. The structural

characterization of 1.50 and 1.51 revealed the classic structure with the Fe atom

sandwiched between the two virtually parallel five-membered rings in an η5 arrangement.

Notably, the tetraphosphaferrocene is to date the only heterometallocene incorporating a

heterocyclic Cp ligand containing only one carbon atom in the ring skeleton.

Scheme 1.11.

P C P C P P P P C P P P ButCP P P P + Fe + Fe + Fe Cp'''(CO) Fe P P Fe(CO) Cp''' 2 P 2 P 1.501.51 1.52 1.49 P P Cp''' = η5-C H tBu 5 2 3 + Cp'''Fe FeCp''' P

− Besides the triphosphacyclopentadienyl ligands (RC)2P3 , there are limited studies

on the coordination chemistry of other heterocyclic Cp ligands containing only two

carbon atoms in the ring framework. Only the structures of the lithium and tungsten

complexes featuring η1,η1 and η1-1,2,4-diazaphospholide ligands, respectively, have been reported by Gudat and coworkers.48 The ruthenium complexes (1.55) containing the

5 − η -1,2,4-diazaphospholyl ligand [(RC)2N2P] were synthesized and characterized very

49 recently. The synthesis of 1.55 was carried out by reaction of [Cp*RuCl]4 with the potassium salt generated in situ via the deprotonation of 1.54 with potassium metal in

THF (Scheme 1.12). A synthetic approach for the preparation of the substituted 1,2,4-

diazaphospholes (1.54) was provided by the condensation of 2-phosphaallyl chlorides

(1.53) with monosubstituted hydrazines.50 The sandwich structure of 1.55a consisted of coplanar η5-diazaphospholyl and Cp* ligands bonded to the Ru atom with nearly equal

metal-to-ring distances, demonstrating the good coordinating ability of the heterocyclic

ligand. In addition, theoretical investigations suggested that the heterocyclic ligand in

1.55a is a poorer electron donor relative to the Cp* ligand and that the lone pair electrons

on the heteroatoms hardly participate in the bonding to the ruthenium center.

Scheme 1.12.

R R R Cl P Cl + P(SiMe3)3 + R'HN-NH2 R R 2 - 3 Me SiCl Me2N Cl 3 Me2N P NMe2 - Me2NH2Cl N N - Me NH 2 R' 1.53 1.54

N N K R P P R P R R K / THF R R + 1/4 [Cp*RuCl]4 Ru NNH - H2 NN - KCl

1.54a: R = tBu 1.54b: R = Ph 1.55a: R = tBu 1.55b: R = Ph

A heavy analogue of the Cp ligand, the lithium 1,2-disila-3-

germacyclopentadienyl complex (1.57) with C2GeSi2 framework, was synthesized and

51 structurally characterized by Sekiguchi and coworkers. Reduction of 1.56 with KC8 in

THF afforded the potassium complex that was subsequently treated with an excess of dry

LiBr, yielding the desired lithium derivative 1.57 (Scheme 1.13). The crystal structure of

1.57 revealed a planar heterocycle that is η5-coordinated to the lithium ion. The changes

in the skeletal bond lengths upon reduction of 1.56 included the elongation of the double

bonds (Si=Ge and C=C) and a shortening of the single bonds (Si-Si, Si-C and Ge-C).

This is indicative of the delocalization of the electron density over the ring framework.

Scheme 1.13.

R R R R R R K Si Si Si Ph H Ge R 2 KC8 / THF Ge Si Si R Ge Si C6D6, rt R - RK R R Ph Ph H H t R = SiMe Bu2 1.56 LiBr / THF - KBr R R Si R Ge Si Li+(thf) R Ph H Cp*Fe(acac) Si R Fe R Ge Si - Li(acac) Ph H O O acac = 1.57 1.58 C H

A heavy ferrocene analog incorporating the 1,2-disila-3-germacyclopentadienyl

− 52 [C2GeSi2] ligand, was recently reported. The synthesis of 1.58 involved the utilization

of the Cp*Fe(acac) complex (acac = acetylacetone) as a convenient source for the Cp*Fe

fragment, since the classical coupling reactions of either FeCl2 or FeCl2(thf)n with 1.57

were unsuccessful. The structural characterization of 1.58 showed that the nearly parallel ligands were η5 bonded to the Fe center. The Fe atom is situated slightly closer to the

C2GeSi2 ring (1.686 Å vs. 1.705 Å for Cp* ring). The skeletal bond distances in 1.58 are

comparable to those of the lithium derivative 1.57, and markedly distinct from those of

the neutral 1,2-disila-3-germacyclopentadiene 1.56.53 Consequently, all skeletal bonds are

intermediate in length between the typical single and double bonds, indicating the

presence of π-electron delocalization over the heterocycle ring. Computational studies on

1.58 and other model compounds containing Me3Si-, H3Si-, and H-substituents showed

that upon the decrease in bulkiness and increase in electronegativity of the substituents,

the heterocyclic ring becomes gradually more distorted and the conformation of the

ligands changes from a staggered to an eclipsed arrangement. Moreover, the cyclic

voltammetry measurement of 1.58 displayed two irreversible oxidation waves at peak

+ potentials of -0.53 and -0.24 V (vs. Ag/Ag in CH2Cl2). The first oxidation step that was

assigned to the Fe2+/Fe3+ process is significantly more negative in potential than the

corresponding one-electron reversible oxidation of Cp*2Fe (-0.32 V) measured under the

same conditions. This suggests that the C2GeSi2 ligand in 1.58 is a better electron donor

than the Cp* ligand.

1.2.4. Carbon-Free Cyclopentadienyl Analogs with E5 Frameworks

The heterocyclic 1,2,4-triaza-3,5-diborolidines B, related to borazines A, have

been studied extensively due to their applications as precursors in boron nitride fibres and

ceramics,54 and in molecular materials.55 These heterocycles are also suitable precursors for the synthesis of heterocyclic anions C and D featuring a carbon-free skeletal framework.

R R'' R'' R' B R' N N N N N R B B R R B B R R B B R o o B B N N N N N N R N R R' R' R' R' R' R' ABCD

Scheme 1.14.

R'' route B R' R' N H H N N R R + 2 RB(SMe)2 + H2NR'' B B N N R B B R - 2 HSMe - HSMe N N R' R' SMe SMe R' R' 1.60 H R N R 1.59 R R' R'' route A B B a Ph Me H NMe NMe 2 2 b Me Me H R' R' c Me Ph H H N N N d Me Me NMe2 R B B R e Me Me Me + 2 HNMe R B B R + RB(SMe)2 2 f Me Ph Me NN NN R' R' R' R'

1.59a: R = Ph, R' = Me 1.61

The 1,2,4-triaza-3,5-diborolidine 1.59a was first reported in 1963 by Nöth and

coworkers and was obtained through the condensation reaction of bis(aminoboryl)amines

with substituted hydrazines (Scheme 1.14 route A).56a A more convenient method for the

synthesis of various substituted B2N3 heterocycles was also described (route B). The treatment of di(methylthio)borane with hydrazines afforded preferentially compounds

1.60 and the tertaazadiborolidines 1.61 as the minor products. Compound 1.60 readily

56b, 56c reacted with NH3 and primary amines yielding derivatives of 1.59. The lithium

− complexes 1.62 containing the anionic B2N3 ligands were prepared by deprotonation of

1.59a – 1.59c with methyllithium.57a The synthesis and reactivity of the protonated

− precursors to the anionic B2N3 rings were further investigated but no references to metal

complexes incorporating these ligands were made.57b The 1,2,4-triaza-3,5-diborolidines

1.63, precursors to less symmetric anionic rings D with a ring proton in the 2-position

instead of the 4-position in ligands of type C, have been reported as well. The reaction of

methylhydrazine with MeB(SMe)2 and PhB(NMe2)2 yielded the 1,2,4-triaza-3,5-

diborolidines 1.63a and 1.63b, respectively (Scheme 1.15).58 Structural characterization

of 1.63b revealed a planar five-membered ring with the skeletal B-N and N-N bond

distances ranging between 1.40 and 1.44 Å, respectively.58b

Scheme 1.15.

Me HN R N R R 22B + Me(H)N NH2 B B + 4 HER' R'E ER' N NH Me ER' = SCH3, N(CH3)2 1.63a: R = Me 1.63b: R = Ph

A ferrocene analog containing ligands of type C was prepared and extensively

investigated, however no structural characterization confirming its identity was reported to date.57a Nevertheless, based on the 1H and 11B NMR, and IR spectroscopy, molecular

weight determination and elemental analysis, a symmetric structure containing π-

coordinating ligands was postulated for the diamagnetic, crystalline compound. In

addition, theoretical studies also supported the sandwich structure proposed by Nöth, predicting that this compound would be “a fully inorganic analog of ferrocene”.59 A

“fully inorganic” analog of ferrocene, a sandwich complex containing no carbon atoms in the ligand skeletons, remains elusive to date. No such complex has been characterized

using single-crystal X-ray diffractometry, and the closest confirmed relative is that of the

5 2- inorganic titanocene [Ti(η -P5)2] that was prepared by heating white phosphorus with

60 TiCl4(thf)2 and potassium naphthalenide in the presence of 18-crown-6.

Phosphorus, arsenic, and antimony are to date the only elements other than carbon

that were shown to generate sandwich structures containing pentaatomic homocyclic

5 ligands. Triple-decker sandwich complexes containing cyclo-(μ,η -E5) ligands have been

isolated and structurally characterized for E = P,61 As,62 and Sb.63 These complexes were

synthesized by thermolysis reaction of elemental pnictogens with appropriate

cyclopentadienyl metal carbonyl complexes, as depicted in Scheme 1.16.

Scheme 1.16.

Mo Cr

P (CH3As)5 As 4 P P As As [CpMo(CO)3]2 P [Cp*(CO)2Cr]2 (Cr Cr) 140 oC PP 190 oC As As Mo Cr

1.64 1.65

cyclo-tBu Sb Sb 4 4 Sb Sb Cp'''(CO)3MoCH3 200 oC Sb Sb Mo

1.66

1.2.5. Multidecker Sandwich Complexes of Boron-Containing Heterocycles

Boron-containing heterocyclic ligands played an important role in the chemistry of multidecker sandwich compounds.64 Experimental evidence supported by theoretical

calculations indicates that heteroatomic boron rings have increased stacking tendency

with respect to their organic analogs and form multidecker complexes. The only

polydecker sandwich compound 1.67 containing 1,3-diborolyl ligands was reported by

Siebert and coworkers in 1986, and displayed semiconducting properties (Scheme

1.17).65

Scheme 1.17.

B B

R' R' Ni Ni

Ni(C3H5)2 / hv B B B B B B Me Me - C3H6 - Ni(C H ) 3 5 2 n R H Ni Ni - C6H10

B B

1.67

The synthetic approach towards multidecker sandwich complexes commonly involves the

successive stacking of smaller sandwich compounds to produce larger and larger

complexes. The method of stacking allows for an efficient control of the size of the

oligomers, but requires multiple synthetic steps.64

1.3. Objectives and Outline of Thesis

The objective of this thesis primarily focuses on the fundamental investigation of

heterocyclic Cp analogs with skeletal frameworks containing boron and nitrogen. The

initial goals involved the development of synthetic methods, the spectroscopic

characterization, and the structural investigation of metal complexes of the novel

heterocyclic Cp ligands. The synthesis and structural characterization of metal complexes

containing the heterocyclic cyclopentadienyl ligands will be examined in detail, seeking

to expand our knowledge on the coordination chemistry of heterocyclic Cp ligands and to

provide an understanding of their reactivity in comparison to their carbon-based analogs.

The replacement of the carbon atoms by isoelectronic main-group atoms

unquestionably disturbs the electronic standards of the Cp ligands. The differences in

electronegativity and atom size between the skeletal boron, nitrogen and carbon atoms

prevent the complete delocalization of the π electrons along the CB2N2 framework.

Consequently, metal complexes incorporating the 1,2-diaza-3,5-diborolyl ligands display

more diverse coordination properties in comparison to the Cp ligands. The description of

the coordination modes of the diazadiborolyl ligands in metal complexes was obtained by

assessing the bond distances of the metal-to-ring atoms of the ligand and by examining

the position of the metal atoms with respect to the ring center, using the perpendicular

projection on the best plane of the ligands (Figure 1.6).

R' R' R' R' R' R' R' R' R' N N N N N N N N N R' N B R B B B B BB B B R R R R R R R R B Me Me Me Me R Me M M M M M

5 1 2 3 4 η η η η η

Figure 1.6. Different coordination modes of the heterocyclic ligands in metal complexes.

Chapter Two of this thesis presents a convenient synthetic approach for the synthesis of the 1,2-diaza-3,5-diborolidines. The coordination properties of the alkali metal complexes of 1,2-diaza-3,5-diborolyls, E, were investigated by single-crystal X-ray diffraction analysis. These alkali metal salts are most suitable synthons for the

preparation of transition metal derivatives. Chapter Three and Four of this thesis address

the synthesis and characterization of novel heterometallocenes of group 12 and 14 metals.

Finally, Chapter Five is dedicated to the synthesis of metallocenes of early and late

transition metals and the study of their structures and properties. In addition, a

heterobicyclic compound with C2B4N2 framework, F, was prepared and characterized,

and proved to be a promising precursor for the synthesis of polydecker sandwich

compounds. The synthesis and characterization of a triple-decker ruthenocene containing

an unprecedented eight-membered C2B4N2 ring as the middle deck will be described in

this chapter as well. Chapter Six of this thesis revisits the studies of the coordination

chemistry of 1,2,4-triaza-3,5-diborolyl ligands, investigating the synthesis and structural

characterization of alkali metal and rhodium complexes of pentaatomic heterocyclic B2N3 ligands, G. The self assembly of the anionic ligands led to the formation of a novel tricyclic BN compound containing an inorganic B4N8 backbone, H.

Me Me Me Me Ph Ph B B Ph N N N R' B B B R' R' B B R' B N N N Me N N Me N B N N N N B Ph BBPh N N R R R HN Me Ph Me Me Me EGHF

This thesis concludes with a summary of the results generated by the metal complexes containing heterocyclic 1,2-diaza-3,5-diborolyl and 1,2,4-triaza-3,5-diborolyl ligands, as well as a few proposals for future research directions in this area. Lastly,

Chapter Eight of the thesis summarizes all the experimental procedures pertaining to

Chapters Two through Six. In addition, the tabulated crystal data collection parameters and the selected bond lengths and bond angles of the complexes characterized in the solid state by X-ray analyses are presented in Appendix One of the thesis.

CHAPTER TWO

1,2-Diaza-3,5-diborolyl Ligands and their Alkali Metal Complexes

2.1. Introduction

Alkali metal complexes of cyclopentadienyls (M−Cp) are essential precursors used in organometallic chemistry.66 A diverse array of structural types has been described for M−Cp derivatives, including the monomeric (A), the oligomeric (B), and the polymeric (C) structures, which are dependent on the substitution pattern on Cp and the presence of coordinating Lewis bases (L) in the structural arrangement.66b In an attempt to tune the electronic properties of the cyclopentadienyl ligand, various five-membered heterocyclic analogs have been synthesized through the formal substitution of ring carbon atoms in Cp with isolobal main-group fragments.

This chapter describes the synthesis of a novel class of heterocyclic Cp analogs with a CB2N2 ring framework, and the structural investigation of their alkali metal complexes. The synthetic route for the synthesis of the heterocyclic 1,2-diaza-3,5- diborolidines (2.3a – 2.3e) will be presented. The study of the electronic and steric influence of the substituents attached to the skeletal boron and nitrogen atoms will be

examined, along with their spectroscopic characterization. The crystal structures of

compounds 2.3c and 2.3d are also described.

M M M M R2E M L L L L M M M L M ER M 2 M M L L L L M L Type A Type B Type C

Me H Me H Me H Me H Me H

Me Me Ph Ph Ph Ph Me2N NMe2 Ph Ph B B B B B B B B B B N N N N N N N N N N Ph Ph Ph Ph

2.3a 2.3b 2.3c 2.3d 2.3e

Alkali metal cyclopentadienyls are among the most commonly used

cyclopentadienyl transfer agents, however, to date the structural characterization of the

alkali metal complexes containing heterocyclic Cp analogs is surprisingly scarce: only

15 16 30c the structures of the lithium derivatives containing ligands with C4P, C4As, C3BN,

26 and C2GeSi2 skeletons have been reported. Seeking to develop useful precursors for the

synthesis of novel heterometallocenes, a series of alkali metal complexes containing the

1,2-diaza-3,5-diborolyl ligands (2.4 – 2.6) were prepared and characterized. The

structural characterization of lithium (2.4a, 2.4c), sodium (2.5b, 2.5c) and potassium

(2.6b, 2.6c) salts, displaying many similarities but also differences in comparison to the

alkali metal cyclopentadienyl analogs will also be described in this chapter.

2.2. Methodology for the Synthesis of 1,2-Diaza-3,5-diborolidines, their

Spectroscopic Characterization and the X-ray Structures of 2.3c and 2.3d.

The strategy for the synthesis of the five-membered heterocyclic 1,2-diaza-3,5- diborolidines involves the coupling of a symmetrically substituted 1,1-diborylethane,

MeCH(BRCl)2, with a disubstituted hydrazine, R’HN-NHR’, in the presence of a base.

The 1,1-bis(dichloroboryl)ethane, MeCH(BCl2)2, was prepared through the hydroboration of acetylene with dichloroborane that was generated in situ by the reaction of BCl3 and

67 Me3SiH in hexane at -78 °C (Eq. 1). The product was isolated as an air and moisture-

sensitive colorless liquid in low but reproducible yield (30%) upon reduced pressure

distillation (b.p. 55 – 65 °C / 50 torr). The purity of the 1,1-bis(dichloroboryl)ethane was

confirmed by the 11B NMR spectrum that revealed a broad resonance at δ 59.9 ppm and

the 1H NMR spectrum exhibiting two broad signals at δ 0.93 and 1.89 ppm corresponding to the methyl and methine protons, respectively. The expected coupling pattern for the methyl and methine protons was not observed due to the large signal width caused by the interaction with the quadrupolar boron nuclei.

Me H hexane, -78 ° C Cl Cl HC CH Cl Cl B (1) BCl3 + Me3SiH B B -Me SiCl 3 H Cl Cl

The symmetrically disubstituted compounds, MeCH(BRCl)2 (R = Me (2.1a), Ph

(2.1b) and NMe2 (2.1c)), were prepared by heating a neat mixture of MeCH(BCl2)2 with

68 SnMe4 at 40 °C for 2 h (2.1a), SnPh4 at 80 °C for 18 h (2.1b) and Me2N-SiMe3 under

69 reflux for 2 h (2.1c), as shown in Scheme 2.1. Compound 2.1a and the Me3SnCl by-

product have similar boiling points and the purification of 2.1a requires a number of

distillations at reduced pressure, resulting in significant product loss. It was, therefore

preferable to generate the compound in situ and use it in further reactions without purification. The 11B NMR spectrum of 2.1a displayed a broad signal at δ 74.3 ppm in

C6D6 solution, which is ca. 14 ppm downfield shifted from the 1,1- bis(dichloroboryl)ethane (δ 59.9 ppm). A broad singlet and a doublet corresponding to the methine and methyl protons were observed at δ 2.11 and 1.88 ppm, respectively, and a singlet associated to the methyl groups on the boron atom was observed at δ 0.79 ppm in the 1H NMR spectrum.

Scheme 2.1.

2 SnMe4 or Me H 2 Me3SnCl 2/3 SnPh4 R R B B + or neat Cl Cl 2/3 PhSnCl3 Me H R = Me (2.1a) Cl Cl B B Ph (2.1b)

Cl Cl Me H 2 Me NSiMe 2 3 Me2N NMe2 B B + 2 Me3SiCl neat Cl Cl 2.1c

The 1,1-diphenyl-diborylethane (2.1b) was isolated in 85% yield from the

PhSnCl3 by-product through reduced pressure distillation as a colorless viscous liquid

(b.p. 100 – 120 ºC / 0.1 torr). It is an air and moisture-sensitive compound that is highly

soluble in organic solvents. The 1H NMR spectrum of 2.1b displayed the characteristic

quartet (δ 3.22 ppm) and doublet (δ 1.51 ppm) signals for the methine and methyl protons

and multiplets (δ 7.10 – 7.98 ppm) corresponding to the phenyl protons on the boron

atoms. As expected, a broad boron resonance was also detected in the 11B NMR spectrum at δ 66.9 ppm.

The 1,1-dimethylamino-diborylethane (2.1c) was obtained as a colorless liquid by distillation (78% yield, b.p. 32 °C / 0.01 torr). The 11B NMR spectrum of 2.1c exhibited

a broad boron resonance at δ 38.7 ppm, which is significantly upfield shifted in

comparison to compounds 2.1a and 2.1b. The 1H NMR spectrum provided evidence for the π-bonding between B and N atoms. Two singlets attributed to the methyl protons of the amine group at δ 2.36 and 2.60 ppm were observed and their inequivalence is probably due to restricted free rotation around the B-N bond. It is worth noting that the distinctive quartet and doublet of the methine and methyl protons on the carbon atom were not observed, instead the signals overlapped into a broad singlet at δ 1.32 ppm with the intensity ratio of 4:6:6 relative to the two inequivalent singlets of the NMe2 group.

2 Me C=O H 2 3 Å mol. H2 + N N NN (2) sieves Pt/C H2N-NH2 H

2.2a

For the synthesis of the symmetric diisopropylhydrazine (2.2a), the diketazine

obtained through the condensation of hydrazine with acetone was hydrogenated with

molecular hydrogen in the presence of platinum deposited on carbon (Eq. 2).70 The air- sensitive product was dried on amalgamated aluminum foil and isolated as a colorless liquid in 59% yield via atmospheric distillation at 124 °C. The 1H NMR spectrum

featured a doublet and a septet at δ 0.97 and 2.74 ppm corresponding to the methyl and

methine protons of the isopropyl group, and a broad signal for the amino proton at δ 2.34

ppm. The diisopropylhydrazine is easily oxidized upon exposure to oxygen to form the

corresponding diazo derivative.

Successful ring closure between MeCH(BRCl)2 (R = Me, Ph, NMe2) and R’HN-

NHR’ (R’ = iPr, Ph) occurred easily in the presence of triethylamine, yielding the ring

system of 1,2-diaza-3,5-diborolidines (2.3a – 2.3d), as depicted in Scheme 2.2. Ligand

2.3a was isolated as a colorless liquid by reduced pressure distillation (b.p. 75 °C / 5 torr)

in 76% yield. This air and moisture-sensitive compound hydrolyzes rapidly upon

exposure to the atmosphere and has good solubility in organic solvents.

Scheme 2.2.

H H Me H + N N R R B B iPr iPr N N 2 NEt / hexane 3 iPr iPr Me H R R R = Me (2.3a); Ph (2.3b) B B H H Cl Cl Me H + N N R R 2.1a-c Ph Ph B B

2 NEt3 N N benzene/hexane Ph Ph

R = Ph (2.3c); NMe2 (2.3d)

The compound was characterized by multinuclear (1H, 11B and 13C) NMR spectroscopy and mass spectrometry, confirming its identity and purity. The 1H NMR spectrum of 2.3a in THF-d8, as illustrated in Figure 2.1a, revealed a broad quartet and a

doublet (δ -0.02 and 0.92 ppm) corresponding to the coupled methine and methyl protons

on the ring carbon atom. A singlet resonance for the methyl substituent (δ 0.45 ppm) on

the boron atoms and a doublet and septet for the methyl and methine protons (δ 1.29 and

3.93 ppm) of the isopropyl group on the nitrogen atoms were also observed in the 1H

NMR spectrum. The 11B NMR spectrum showed a broad resonance at δ 46.9 ppm, which

is comparable to the reported 11B NMR resonance in the protonated boron-containing

23, 24, 29, 30 11 heterocyclic C3BE rings (E = N, S, O) ranging from δ 41 to 47 ppm. The B

NMR resonance is ca. 28 ppm upfield shifted with respect to the resonance in the 1,1-

bis(chloromethylboryl)ethane 2.1a.

The 1,2-diisopropyl-diaza-3,5-diphenyl-diborolidine, 2.3b, was prepared using

the same procedure as for its methyl derivative 2.3a, by reaction of diphenyl

diborylethane with diisopropylhydrazine (Scheme 2.2). Ligand 2.3b was obtained as a

colorless, viscous liquid upon removal of the volatiles under vacuum in a reasonable

yield (71%), and could not be distilled but had excellent purity according to the NMR data. A broad 11B NMR signal at δ 46.0 ppm was observed, which is comparable to the

11B resonance of 2.3a (δ 46.9 ppm) suggesting only a minor electronic influence through

the phenyl substituents. The 1H NMR spectrum exhibited two doublets (δ 1.13 and 1.14 ppm) and a septet (δ 3.91 ppm) representing the methyl and methine protons of the isopropyl group on the nitrogen atoms (Figure 2.1b). The characteristic quartet and doublet of the methine and methyl protons on the ring carbon atom were not clearly observed because of the overlap with the set of doublets of the isopropyl groups. A set of multiplets ranging from δ 7.20 – 7.46 ppm were observed with the intensity ratio of 2:4:4 representing the para, meta and ortho-protons of the phenyl substituent on the boron atoms. It is worth noting that a broad 13C NMR signal corresponding to the ring carbon was detected at δ 26.8 ppm and all carbon signals of the ligand were identified in the expected region of the 13C NMR spectrum. In addition, the electron impact (EI) mass

spectra of 2.3a and 2.3b revealed the molecular ion peak [M]+ (m/z = 194 and 318,

respectively) as well as the fragment ion [M - Me]+ (m/z = 179 and 303, respectively),

resulting from the loss of a methyl substituent from the ligand.

(a) Me H Me Me B B N N iPr iPr

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 -0.5 ppm

(b) Me H Ph Ph B B N N iPr iPr

1.20 1.10 1.00

7.5 6.5 5.5 4.5 3.5 2.5 1.5 0.5 ppm (c) Me H Ph Ph B B N N Ph Ph

7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm

(d) Me H

Me2N NMe2 B B N N Ph Ph

7.5 6.5 5.5 4.5 3.5 2.5 1.5 0.5 ppm

1 Figure 2.1. The H NMR spectra of 2.3a in THF-d8 (a), 2.3b in C6D6 (b), 2.3c in THF-d8

To gain further insight into how the boron and nitrogen substituents affect the

electronic nature of the heterocyclic ligands, the 1,2,3,5-tetraphenyl-1,2-diaza-3,5-

diborolidine (2.3c) and 1,2-dimethylamino-diaza-3,5-diphenyl-diborolidine (2.3d) were

synthesized by similar ring-closing reactions of 2.1b and 2.1c, respectively, with

diphenylhydrazine in the presence of NEt3. Diphenylhydrazine is a commercial reagent that was purified by washing with hexane to remove the diazobenzene impurity prior to usage. The protonated ligands 2.3c and 2.3d were isolated as colorless solids (63% yield for 2.3c and 51% yield for 2.3d) and had good solubility in organic solvents, but were only sparingly soluble in pentane. Colorless thin plate crystals of 2.3c and 2.3d were obtained by cooling the concentrated solutions of 2.3c and 2.3d in a mixture of toluene and hexane at -35 ºC. The compounds were characterized by multinuclear (1H, 11B, and

13C) NMR spectroscopy, HRMS and EI-MS spectrometry and single crystal X-ray

diffraction. The distinctive sets of quartet and doublet resonances corresponding to the

methine and methyl protons (δ 1.45 and 1.65 ppm for 2.3c and δ 0.65 and 1.01 ppm for

1 2.3d) of the HMeCB2 fragment were observed in the H NMR spectra of 2.3c and 2.3d

(Figure 2.1c and 2.1d). The set of multiplets corresponding to the phenyl substituents on

the boron and nitrogen atoms ranging from δ 6.65 – 6.87 ppm (NPh) and δ 7.11 – 7.44 ppm (BPh) were also detected in the 1H NMR spectrum of 2.3c. The 1H NMR spectrum

of 2.3d exhibited only a single resonance at δ 2.57 ppm for the NMe2 group on the boron

atoms, indicating that the methyl protons are equivalent due to rapid exchange on the

NMR time scale, and a set of multiplet (δ 6.82 – 7.19 ppm) for the phenyl substituents on

the nitrogen atoms. The 11B NMR spectra of 2.3c and 2.3d also features a broad boron resonance at δ 44.0 and 37.4 ppm, respectively, which is slightly upfield shifted in

comparison to the previously discussed ligands (δ 46.9 ppm for 2.3a and δ 46.0 ppm for

2.3b). The electron-rich NMe2 substituent on the boron atoms leads to an upfield shift of

the 11B NMR signal. In the 13C NMR spectra of 2.3c and 2.3d, broad resonances for the ring carbon were observed at δ 23.0 and 8.9 ppm, respectively, higher in frequency than

the value observed in 2.3b. The mass spectra of 2.3c and 2.3d showed the molecular ion

peak [M]+ at m/z = 386 and 320, respectively, confirming the identity of the compounds.

Figure 2.2. Molecular structures of 2.3c (left) and 2.3d (right) with 50% probability thermal ellipsoids; all hydrogen atoms of the organic substituents omitted for clarity.

The molecular structures of 2.3c and 2.3d are shown in Figure 2.2. The selected structural parameters and the relevant bond lengths and bond angles are summarized in

Table 2.1, 2.4 and 2.5 (Appendix One). In the crystal structure of 2.3c, the CB2N2 ring is virtually planar with the sum of the intraannular ring angles of 539.9º. The B-N bonds

(B(1)-N(1) = 1.412(3) Å and B(2)-N(2) = 1.416(3) Å) are found to be marginally shorter than those observed in borazines (1.42 – 1.44 Å),71 indicative of a multiple bond character. The extraannular B-C bonds (B(1)-C(15) = 1.568(4) Å and B(2)-C(21) =

1.565(4) Å) are slightly shorter than those within the ring structure (B(1)-C(1) = 1.572(4)

Å and B(2)-C(1) = 1.581(4) Å), supporting the single-bond character of the intraannular

B-C bonds. The N-N bond in 2.3c (N(1)-N(2) = 1.452(3) Å) is similar in length to the

single N-N bond in hydrazine (1.45 Å).72

The ring carbon, C(1), adopted a distorted tetrahedral geometry with bond angles

B(1)-C(1)-B(2) = 102.3(2)º, B(1)-C(1)-C(2) = 121.0(2)º and B(2)-C(1)-C(2) = 125.3(2)º,

indicating an sp3 character. The methyl substituent on the ring carbon lies noticeably out

of the plane of the ring, with the C(1)-C(2) bond forming an angle of 27.4º with the ring plane. The extraannular B-C bonds of the two phenyl substituents on the boron atoms are positioned in the plane of the CB2N2 ring and the two phenyl rings are twisted by 44º

from the ligand plane. For the phenyl substituents on the nitrogen atoms, the

extraannular N-C bonds are clearly situated above and below the plane with the angles

formed between the N-C bonds and ligand plane of 10 and 15º (torsion angle C(3)-N(1)-

N(2)-C(9) of 30.1(3)º).

In contrast to 2.3c, the planarity of the CB2N2 ring in 2.3d is relatively poor with

the sum of the intraannular angles of 531.0º. The ring skeleton revealed a noticeable

folding along the transannular B(1)···N(2) axis (∠ CB2N / NBN = 29º), resulting in an

envelope conformation. The intraannular B-C and N-N bonds (B(1)-C(1) = 1.578(2) Å,

B(2)-C(1) = 1.592(2) Å and N(1)-N(2) = 1.457(2) Å) are similar in length to those

observed in 2.3c, while the intraannular B-N bonds (B(1)-N(1) = 1.497(2) Å and B(2)-

N(2) = 1.467(2) Å) are considerably longer than those in 2.3c (1.412(3) and 1.416(3) Å).

The lengthening of the ring skeletal B-N bonds is a result of the loss of multiple bond

character between the boron and nitrogen atoms due to the competition with the

extraannular B-N bonds. A similar observation was reported for the stable N- heterocyclic carbene, :C(NDipp)2(BNMe2)2 (endocyclic B-N bonds of 1.460(2) and

1.470(2) Å vs. exocyclic B-N bonds of 1.374(2) and 1.380(2) Å).73 As expected, the

extraannular B-N bonds (B(1)-N(3) = 1.389(2) Å and B(2)-N(4) = 1.399(2) Å) are ca. 0.1

Å shorter than the intraannular B-N bonds, indicating the existence of multiple bonding

between the ring boron and the dimethylamino substituent. In comparison to 2.3c, the

methyl substituent on the ring carbon is more tilted away from the ring plane forming an

angle of 59.1º between the C(1)-C(2) bond and B(1)-C(1)-B(2)-N(2) plane. One of the

NMe2 substituents on the boron atom is bent out of plane by an angle of 18.0º formed

between the B(1)-N(3) bond and the CB2N plane, while the other NMe2 substituent lies

nearly in the plane of the ring (B(2)-N(4) bond and CB2N plane forming an angle of

0.8º). The skeletal nitrogen, N(1) and N(2), have a distorted pyramidal geometry with the

sum of the angles around nitrogen atoms of 341° and 353º, respectively.

2.3. Alkali Metal Salts Containing 1,2-Diaza-3,5-diborolyl Ligands (2.4 – 2.6).

2.3.1. Synthesis and Spectroscopic Characterization of Alkali Metal Salts

Incorporating the 1,2-Diaza-3,5-diborolyl Ligands (2.4a, 2.4b – 2.6b and 2.4c – 2.6c).

The synthesis of the alkali metal salts of the 1,2-diisopropyl-3,5-dimethyl-1,2-

diaza-3,5-diborolyl ligand was accomplished by deprotonation of 2.3a using the

appropriate alkali metal reagents (Scheme 2.3). The deprotonation attempts of 2.3a using

lithium diisopropylamide (LDA) and potassium hexamethyldisilazane (KHMDS) were

unsuccessful. The reaction of 2.3a with nBuLi, which is a strong deprotonating reagent,

also failed. The NMR analysis of the reactions indicated the decomposition of the ligand.

Deprotonation of 2.3a with lithium 2,2,6,6-tetramethylpiperidide (LiTMP), which was

prepared in situ by treatment of nBuLi with 2,2,6,6-tetramethylpiperidine (TMP) in THF,

proceeded cleanly and quantitatively to form the lithium salt 2.4a. The lithium salt was

obtained after work-up as a solvent-free colorless powder (91% yield) that hydrolyzed

upon exposure to the atmosphere. The compound was highly soluble in THF but

insoluble in hydrocarbons.

Scheme 2.3.

Me M Base Me Me B B X Me H THF N N Me Me Base = KHMDS, iPr iPr B B LDA, N N nBuLi Me Li iPr iPr Me Me LiTMP B B 2.3a THF N N iPr iPr 2.4a

Me Li+ Me Me B B N N iPr iPr

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 ppm

1 Figure 2.3. The H NMR spectrum of the lithium salt 2.4a in THF-d8 at 25 ºC.

The lithium salt was examined using multinuclear (1H, 11B, 7Li and 13C) NMR

spectroscopy. In the 1H NMR spectrum, the signal corresponding to the methine proton

disappeared and the doublet corresponding to the methyl substituent on the ring carbon

collapsed into a singlet, clearly supporting the successful deprotonation of 2.3a (Figure

2.3). An upfield shift of the resonances for the isopropyl and methyl substituents on

nitrogen and boron atoms, respectively, was observed in the 1H NMR spectrum. The

broad boron signal at δ 38.3 ppm in the 11B NMR spectrum is ca. 9 ppm upfield shifted

relative to the protonated ring (δ 46.9 ppm). A similar observation was reported for the

23, boron-containing C3BE rings (E = N, S, O) upon the formation of the anionic ligands.

24, 29, 30 In addition, a broad resonance corresponding to the ring carbon was detected at δ

13 86.7 ppm in the C NMR spectrum run at -50 ºC in THF-d8. This is characteristic for metal cyclopentadienyl complexes with pronounced ionic character. For example, the deprotonation of pentamethylcyclopentadiene results in a shift of the analogous carbon

* 7 signal from δ 52.2 ppm (in C6D6) to δ 105.2 ppm (Cp Na in THF-d8). The Li NMR

spectrum recorded in THF-d8 showed a resonance at δ -2.34 ppm, indicating that the anionic ligand is separated from the THF-coordinated lithium cation in solution. A doublet representing all four equivalent methyl groups provided further evidence for the separation of anionic ligand from the solvated lithium cation.

In contrast to the methyl derivative 2.3a, the deprotonation of 2.3b was achieved

more easily by using various metalating reagents, as illustrated in Scheme 2.4. The

reactions of 2.3b with LiTMP, Na[N(SiMe3)2] and K[N(SiMe3)2] in THF quantitatively

yielded the lithium, sodium and potassium salts 2.4b, 2.5b and 2.6b, respectively. These

salts were isolated as colorless solids that were soluble in THF and sparingly soluble in

diethyl ether, but insoluble in hydrocarbons. Purification of the lithium salt, 2.4b, was a tedious procedure as it required a number of washings of the solid with hexane, resulting in loss of product and ultimately a 41% yield obtained. The sodium and potassium salts were more easily purified (75 % yield in 2.5b and 93 % yield in 2.6b) by washing with

hexane.

Scheme 2.4.

Me H Me M + LM Ph B B Ph Ph B B Ph THF N N N N - LH R R R R

R = iPr (2.3b) LM = LiTMP M = Li; R = iPr (2.4b) , Ph (2.4c) Ph (2.3c) NaHMDS M = Na; R = iPr (2.5b), Ph (2.5c) KHMDS M = K; R = iPr (2.6b), Ph (2.6c)

Deprotonation resulted in a shift of the 11B NMR signal from δ 46 ppm in 2.3b to

δ 38 – 39 ppm in the alkali metal salts. Additionally, a dramatic low field shift was observed for the broad 13C NMR signal corresponding to the ring carbon, from δ 26.8

ppm in the protonated ligand, 2.3b, to δ 91.5 – 96.8 ppm in 2.4b – 2.6b. The 1H NMR

spectra of the salts described here featured only one doublet for the methyl groups of the isopropyl substituents on nitrogen, indicating that the coordination environment on both faces of the ligand is identical on the NMR time scale. This provides evidence of a solvent-separated ion-pair structure in solution, consisting of free anionic ligands and

7 solvated alkali metal ions. The Li NMR spectrum of 2.4b in THF-d8 revealed a singlet

resonance at δ -2.8 ppm, further supporting this hypothesis: the chemical shift typical for

+ 74 [Li(thf)n] ranges from δ 0.35 to 1.65 ppm.

The alkali metal salts of the tetraphenyl-substituted ligand, 2.4c – 2.6c, were

prepared using the same bases presented in Scheme 2.4. These salts were isolated as air-

and moisture-sensitive solids in very good yields (89 – 97%), and they displayed very

similar solubility to the other alkali metal derivatives (2.4a, 2.4b – 2.6b). The

spectroscopic characterization of 2.4c – 2.6c revealed a small chemical shift of the broad

11B signal from δ 44.0 ppm for the protonated ring 2.3c to ca. δ 40.0 ppm for the alkali metal salts. Analogous to the other alkali metal salts, a singlet (δ 1.73 – 1.82 ppm)

corresponding to the methyl protons on the ring carbon and the resonances for the phenyl

substituents on the boron (δ 6.94 – 7.34 ppm) and nitrogen atoms (δ 6.35 – 6.84 ppm)

were observed in the 1H NMR spectra of 2.4c – 2.6c. The broad signal corresponding to the ring carbon of the alkali metal salts was observed as expected at low field (δ 93.6 –

98.1 ppm) in the 13C NMR spectrum, similar to those in 2.4b – 2.6b.

Attempts to synthesize the anionic 1,2-dimethylamino-3,5-diphenyl-1,2-diaza-

3,5-diborolyl ligand by reaction of 2.3d with strong bases such as LiTMP and KHMDS failed; no changes in the 1H and 11B NMR resonances were observed. The use of

potassium benzyl, which is a stronger base, to deprotonate 2.3d also failed and the 1H

NMR spectrum was consistent with ligand decomposition. The failure to synthesize the anionic ligand is likely due to the increase in the basicity of the ring carbon as an effect of the electron-rich amine substituents attached to the boron atoms.

2.3.2. The X-ray Structures of Lithium (2.4a and 2.4c(thf)3), Sodium (2.5b,

2.5b(thf)3 and 2.5c(thf)3) and Potassium (2.6b(thf), 2.6b(thf)2 and 2.6c(thf)) Complexes

Incorporating the 1,2-Diaza-3,5-diborolyl Ligands.

Colorless crystals of 2.4a suitable for X-ray diffraction analysis were obtained by

slow evaporation of a solution of 2.4a in a mixture of THF and benzene at ambient temperature. The selected bond lengths and bond angles of 2.4a are provided in Table

2.6 (Appendix One). The structure of 2.4a (Figure 2.4) revealed a polymeric sandwich

structure with a bifacially coordinating π ligand, typical for base-free alkali metal

cyclopentadienyls.66 The bridging ligand coordinates to lithium ion in a η1 fashion with one face and η4 with the other face. The η1-coordination of the lithium ion involves a

short distance to carbon (Li(1’)-C(1) = 2.225(6) Å) while the η4-coordination is comprised of two short Li-N distances (Li(1)-N(1) = 2.075(5) Å and Li(1)-N(2) =

2.112(5) Å) and two Li-B distances (Li(1)-B(2) = 2.354(6) Å and Li(1)-B(1) = 2.370(6)

Å). The Li-N distances are shorter than the ones observed in lithium complexes

containing η5-coordinating 1,2-azaborolyl (2.142(7) Å)30c and pyrrolyl type ligands

(2.161(4) and 2.185(1) Å).75 The Li-C and Li-B distances fall in the range observed for

polymeric lithium cyclopentadienyls (Li-C = 2.22 – 2.37 Å)76 and sandwich complexes

containing η5-coordinated five-membered boron heterocycles (Li-C = 2.05 – 2.41 Å and

Li-B = 2.18 – 2.42 Å).77

In lithium boratabenzene sandwiches, the Li-C and Li-B distances are somewhat longer, usually ranging from 2.30 to 2.56 Å and 2.45 to 2.58 Å, respectively.78 The Li(1)-

C(1) distance of 2.480(6) Å is found to be longer than all the Li-C distances observed in

polymeric lithiocenes containing cyclopentadienyl ligands. The longer Li-B distances

(Li(1’)-B(1) = 2.543(6) Å and Li(1’)-B(2) = 2.751(6) Å) are situated outside the typical

range observed for Li-B interactions in related compounds, therefore the binding mode of

the ligand is best described as μ,η1,η4. The planes of the alternating ligands form angles of 21.4° with each other and the consecutive rings are staggered at an angle of 180°. The distance between the neighboring lithium ions (Li···Li = 4.381(8) Å) is larger than the distances observed in polymeric lithiocenes incorporating carbon-based ligands (Li···Li =

3.9 – 4.0 Å).75 The lithium ions are positioned closer to the plane of the η4 face (1.861(5)

Å) than to the plane of the η1 face (2.196(5) Å).

Figure 2.4. Fragment of the polymeric structure of the lithium salt 2.4a with thermal

ellipsoids drawn at the 50% probability level and all hydrogen atoms omitted for clarity.

The CB2N2 ring of 2.4a is practically planar, with a sum of the bond angles within the ring of 539.9°. Consistent with the expected electron delocalization, the intraannular

B-C and B-N bond distances (B(2)-C(1) = 1.498(4) Å and B(1)-C(1) = 1.513(4) Å; B(1)-

N(1) = 1.457(4) Å and B(2)-N(2) = 1.479(4) Å) are comparable to those found in the

base-stabilized 1,2-azaborolyl lithium complex (B-C = 1.496(6) Å and B-N = 1.503(6)

Å).30c The N(1)-N(2) bond of 1.466(3) Å is similar in length to the N-N single bond in hydrazine (1.45 Å)72 and much longer than a double bond in pyridazine (1.34 Å).79 The extraannular B-C bonds in 2.4a are ca. 8 pm longer than the intraannular B-C bonds, indicating electron delocalization within the ring structure. Most substituents of the ring skeleton are situated in the ligand plane, with only one of the isopropyl groups lying significantly outside the plane (the N-C bond forming an angle of 23.2° with the CB2N2 plane).

As predicted, the replacement of the alkyl groups with phenyl substituents on the ligand improved the crystallization properties of the alkali metal salts and crystals of these alkali metal salts were grown using various crystallization methods. Thin plate crystals of 2.5b were obtained by slow evaporation of a benzene / THF solution of 2.5b at ambient temperature. Colorless block crystals of 2.5b(thf)3, 2.5c(thf)3 and 2.6b(thf)2 were obtained by cooling the concentrated solutions in a mixture of hexane and THF to -

35 ºC. The colorless prismatic crystals of 2.4c(thf)3, 2.6b(thf) and 2.6c(thf) were obtained by slow diffusion of hexane into the respective THF solutions at ambient temperature. Single-crystal X-ray diffraction experiments were performed for compounds

2.5b, 2.5b(thf)3, 2.6b(thf), 2.6b(thf)2, 2.4c(thf)3, 2.5c(thf)3 and 2.6c(thf) and the

selected bond lengths and bond angles are given in Table 2.7 and 2.8 (Appendix One).

The metric parameters of the π ligand display little variation between all of the structures.

The planarity of the CB2N2 ring is relatively good in all structures, with the sum of

intraannular bond angles ranging between 539.4 – 539.6°. Careful examination reveals a

slight folding of 6 – 7° along the B(1)···N(2) axis, resulting in an envelope conformation

of the ring skeleton, as illustrated in Figure 2.5. The ligand geometry in 2.5b deviates

slightly from this generalization, with the sum of the ring bond angles being 539.2° and

the envelope conformation is folded by an angle of 8.5° along the transannular

B(1)···N(2) axis.

Figure 2.5. Side view of the sodium salts 2.5b (left), 2.5b(thf)3 (center) and 2.5c(thf)3

(right) illustrating the slight folding of the CB2N2 rings along the B(1)⋅⋅⋅N(1) axis. For

clarity, only the metal and ring atoms are represented.

In structures 2.5b, 2.5b(thf)3, 2.6b(thf) and 2.6b(thf)2, the intraannular B-C

bonds ranging from 1.486(6) to 1.514(3) Å are ca. 10 pm shorter than the corresponding

extraannular bonds (1.579(7) – 1.595(4) Å) and similar in length to those observed in the

1,2-azaborolyl (1.48 – 1.53 Å)30 and boratabenzene (1.50 – 1.55 Å)80 complexes. The B-

N bonds measure between 1.452(3) and 1.481(3) Å, intermediate in length between the

B-N bonds observed in borazines (1.42 – 1.44 Å)71 and 1,2-azaborolyl complexes (1.46 -

1.50 Å),30 indicating partial multiple bond character. The N-N bonds of 1.444(3) –

1.460(3) Å are much closer in length to the single bond in hydrazine (1.45 Å)72 than to a double bond in pyridazine (1.34 Å).79 For comparison, the N-N bonds in η5-pyrazolyl

ligands measure 1.38 – 1.40 Å,40, 81 indicating that the electron delocalization over the N-

B-C-B-N skeleton does not extend over the N-N bond. The substituents on the ring

carbon and boron atoms lie nearly in the plane of the ring, with the extraannular C-C and

B-C bonds forming angles of less than 4° with the ring plane. The isopropyl groups in

structures 2.5b(thf)3, 2.6b(thf) and 2.6b(thf)2 lie noticeably above and below the ligand

plane, with the extraannular N-C bonds forming angles of 26 – 33° with the plane of the

ring.

The intraannular B-C and B-N bonds of 2.4c(thf)3, 2.5c(thf)3 and 2.6c(thf) range

from 1.479(3) to 1.501(3) Å and 1.458(3) to 1.494(3) Å, respectively, and are comparable

to the corresponding bonds in the other alkali metal complexes of 1,2-diaza-3,5-diborolyl

ligands described herein. The N-N bonds of 1.437(2) – 1.443(2) Å are marginally shorter

than those observed in 2.4a and 2.5b – 2.6b. Consistent with the expected multiple bonding character of the anionic ligands, the shortening of the B-C bonds (ca. 8 – 10 pm)

accompanied by the lengthening of the B-N bonds (ca. 3 – 7 pm) is observed in comparison to the protonated ligand 2.3c. The extraannular B-C bonds of 1.579(3) –

1.589(3) Å are markedly longer than the intraannular bonds. Similar to alkali metal complexes 2.5b and 2.6b, the ring carbon and boron substituents are situated in the ligand plane, with the extraannular B-C and C-C bonds slightly tilted out of the plane by 7 – 9°.

The phenyl substituents on the nitrogen atoms are positioned above and below the plane of the ring, with the N-C bonds forming angles of 13º and 37º with the ligand plane. The orientation of the phenyl groups is perhaps due to the steric interaction between the substituents.

A fragment of the polydecker sandwich structure of the solvent-free sodocene,

2.5b, is shown in Figure 2.6. The structure is slightly disordered, with one of the two independent sodium ions, Na(2), occupying several different positions in relation to the anion. While Na(1) is symmetrically coordinated between the η3 faces of the perfectly

parallel ligands, Na(2) is disordered around an η2 position with respect to the other face of the ligands, resulting in an μ,η2,η3-coordination mode. The separation between

neighboring sodium ions (Na(1)···Na(2) = 5.01(1) – 5.21(1) Å) is larger than the Na···Na distance measured in reported polymeric solvent-free sodium cyclopentadienyls (4.61 -

4.86 Å),76b, 82f but comparable to the solvated sodium structures (4.89 – 5.21 Å).82 The

disorder of the metal ion prevents a detailed discussion of the η2-coordination of Na(2) in

2.5b. The η3-coordination of Na(1) consists of short distances to nitrogen (Na(1)-N(2) =

2.612(2) Å) and boron (Na(1)-B(2) = 2.631(2) Å) and a longer distances to carbon

(Na(1)-C(1) = 2.801(2) Å). The Na(1)-C(1) distance falls in the typical range for sodium

cyclopentadienyl salts (2.60 –3.20 Å).76b, 82 To date, sodocenes featuring five-membered,

boron-containing π ligands have not been structurally characterized, however a couple of

sodium boratabenzene structures have been reported.83 The Na-B and Na-C distances in

2.5b are found to be much longer than those observed in the sodium boratabenzene

complexes (Na-B = 2.910(6) – 2.976(2) Å; Na-C = 2.732(2) – 3.000(2) Å). The Na-N

distances measured in sodium salts containing π-coordinating pyrazolyl (2.494(2) –

2.638(2) Å)84 and pyrrolyl (2.694(1) Å)85 ligands, are comparable to those measured in

2.5b. The Na(1)-B(1) and Na(1)-N(1) distances in 2.884(2) and 2.872(2) Å, respectively,

are substantially longer than the bonding interactions described above, and therefore the

bonding mode of the ligand is best described as η3.

In contrast to the polymeric structure of 2.4a and 2.5b, the monomeric structure of

2.5b(thf)3, featured in the solid state a piano stool configuration with the sodium cation being coordinated by the π ligand and three tetrahydrofuran molecules (Figure 2.7). The

Na-O distances (2.333(2) - 2.344(2) Å) and O-Na-O angles (87 – 100°) fall within the typical ranges for solvated sodium cations.82b, 82c, 82d At 2.616(2) Å, the Na(1)-C(1)

distance is comparatively short, and in fact shorter Na-C distances (2.544(3) – 2.575(6)

Å) have been observed only for dicyclopentadienyl sodium anions,86 a solvent-free

cyclopentadienyl sodium derivative,76b and a strained-geometry cyclopentadienyl sodium

87 featuring a pendant azamacrocycle. The sodium atom in 2.5b(thf)3 is positioned above the B(1)-C(1) bond and as a result the Na(1)-B(2) distance is considerably longer than the

Na(1)-B(1) distance (3.101(3) vs. 2.874(2) Å). Both distances, however, are more than

0.3 Å longer than the shortest Na(1)-B(2) distance observed in 2.5b, and therefore the binding mode of the heterocyclic ligand is best described as η1. The shortest Na-N

distance measures more than 3.35 Å and the distance of the metal to the ligand plane is

2.609(2) Å.

Figure 2.7. Two views of the molecular structure of 2.5b(thf)3, with thermal ellipsoids

drawn at 50% probability level. Left: only the ipso carbon of the phenyl group on B1 is

represented. Right: perpendicular projection onto the CB2N2 plane with all THF

molecules and all hydrogen atoms omitted for clarity

The salts 2.4c(thf)3 and 2.5c(thf)3 (Figure 2.8) exhibited monomeric structures

featuring a piano stool arrangement similar to 2.5b(thf)3. In both structures, the lithium

and sodium cations are η1-coordinated by the π ligand and three tetrahydrofuran

molecules completing the coordination sphere. The Li-O distances (1.947(4) – 2.009(3)

74 Å) fall within the typical range. The Li(1)-C(1) distance of 2.351(4) Å in 2.4c(thf)3 is

somewhat longer than the Li-C distances observed in the polymeric 2.4a and lithium

cyclopentadienyls.76 The lithium ion is positioned above the C(1)-C(2) bond as depicted

in Figure 2.8 (bottom), and the Li(1)-C(2) distance of 2.635(4) Å is substantially longer

than the value for Li(1)-C(1). The Li-B distances (Li(1)-B(1) = 2.908(4) Å and Li(1)-

B(2) = 2.983(4) Å) are clearly significantly longer than the Li-B distances in 2.4a

(2.354(6) and 2.370(6) Å). The distance of the lithium ion to the CB2N2 plane of

2.255(4) Å is comparable to those observed in 2.4a. The angle formed between the

Li(1)-C(1) axis and the ligand plane measures 97.3°, thus clearly supporting an η1-

coordination mode of the ligand.

The Na(1)-C(1) distance of 2.680(4) Å measured in 2.5c(thf)3 is slightly longer

than those observed in 2.5b(thf)3, but considerably shorter than observed in the solvent-

free polymeric sodocene, 2.5b. The sodium ion is located closer to the B(2)-C(1) bond

(Figure 2.8), and consequently the Na(1)-B(2) distance of 2.939(5) Å is much shorter

than Na(1)-B(1) (3.189(5) Å). Nevertheless, these Na-B distances are noticeably longer

than those observed in 2.5b, hence the hapticity of the ligand is best represented as η1 bonding. The distance of the sodium ion to the ring plane measures 2.678(2) Å with an angle of 84.7° measured between the ring plane and Na(1)-C(1) axis. The Na-O

distances (2.325(3) – 2.360(4) Å) and O-Na-O angles (89.1(1) – 110.3(1)º) are

comparable to those found in the solvated sodium complex 2.5b(thf)3.

Figure 2.8. Molecular structures of 2.4c(thf)3 (left) and 2.5c(thf)3 (right) with 50%

probability thermal ellipsoids. For clarity, only the ipso carbon atom of the phenyl substituent on B1 and B21 atoms are represented, and all hydrogen atoms omitted.

Bottom: perpendicular projection of the metal ions onto the CB2N2 planes, with the three

tetrahydrofuran molecules omitted for clarity.

Figure 2.9. Fragment of the polymeric chain of 2.6b(thf) and 2.6b(thf)2 in the solid-state with 50% probability level thermal ellipsoids. For clarity, only the α-carbon atoms of the organic substituents are represented and all hydrogen atoms omitted.

Compounds 2.6b(thf) and 2.6b(thf)2 are polymeric in the solid state, featuring

polydecker sandwich structures containing potassium ions intercalated with parallel, π- coordinating diazadiborolyl ligands as shown in Figures 2.9. This is not typical for polymeric potassium cyclopentadienyls, which usually display bent geometries with non- parallel ligands.66 The main difference between the two structures is the number of

tetrahydrofuran molecules completing the coordination sphere of each of the potassium

ions: one in 2.6b(thf) and two in 2.6b(thf)2. The K-O distances measuring between

2.723(2) and 2.816(4) Å fall in the expected range for solvated potassium salts. The

multidecker sandwich structure of 2.6b(thf), where the potassium ions have a lower

coordination number, is more compact, with a K(1)···K(1’) separation of 6.055(3) Å

versus 6.165(4) Å in 2.6b(thf)2. In comparison, shorter K···K distances ranging from

5.52 – 5.85 Å are observed in the bent, polymeric potassium cyclopentadienyl salts.76b, 82c,

82g, 88 Uncharacteristic for alkali metal cyclopentadienyls, the potassium ions in 2.6b(thf)

and 2.6b(thf)2 are not contained between the ligands but are instead situated outside of

the columns formed by the parallel CB2N2 rings, alternating on both sides in the vicinity

of the two endocyclic C-B bonds. The K-C distances in 2.6b(thf) (3.033(2) and 3.058(2)

Å) and 2.6b(thf)2 (3.127(4) and 3.156(4) Å) fall in the range observed in the polymeric

potassium cyclopentadienyls (2.95 to 3.30 Å), with the shortest K-C distances in the

complexes measuring 2.952(2) – 3.074(7) Å.76b, 82c, 82g, 88

Potassium borole salts have not been structurally characterized, but a few

potassium boratabenzene and diboratabenzene complexes displaying monomeric and

polymeric sandwich structures have been reported.89 The K-B distances in these

complexes range between 3.06(1) – 3.686(6) Å, and the K-C distances are situated

89 between 2.952(3) – 3.572(5) Å. In the two structures 2.6b(thf) and 2.6b(thf)2, the

potassium ions are located closer to one of the boron atoms of each of the coordinating

ligands. The two shorter K-B distances of 3.211(3) and 3.315(3) Å in 2.6b(thf) and

3.364(4) and 3.425(4) Å in 2.6b(thf)2 are within the corresponding range for the reported

potassium boratabenzenes. The two longer K-B distances of 3.477(3) and 3.752(3) Å in

2.6b(thf) and 3.802(4) and 3.895(4) Å in 2.6b(thf)2 are outside the usual range. Taking

into account the geometry and metric parameters, it appears that the best description for

the coordination mode of the 1,2-diaza-3,5-diborolyl ligands in 2.6b(thf) and 2.6b(thf)2 is μ,η2,η2. The distances between the potassium atoms and the methyl substituents of the

ring carbon are quite short as a result of the unusual position of the potassium ions, on the

outside of the columns formed by the CB2N2 rings. The K-C(2) distances range between

3.203(4) and 3.331(3) Å, are only 0.05 to 0.30 Å longer than the K-C(1) distances, and

shorter than some of the distances between potassium and the ring carbon in polymeric

potassium cyclopentadienyls (2.95 to over 3.30 Å).76b, 82c, 82g, 88

In contrast to the polymeric potassium salts of 2.6b(thf) and 2.6b(thf)2, the crystal structure of 2.6c(thf) (Figure 2.10) revealed a polymeric structure where the potassium cation is η2-coordinated to the π ligand with one THF molecule and the phenyl substituent of a successive ligand completing the coordination sphere of the metal center.

Due to the presence of the weak interaction of the potassium cation with the phenyl substituent of the ligand, the polymeric structure of 2.6c(thf) features an unusual zig-zag arrangement. The alternating CB2N2 ligand planes form dihedral angles of 21.4(9)º with

each other and are staggered by an angle of 180º. The distance of the potassium to the

ring carbon (K(1)-C(1) = 3.021(2) Å) is slightly shorter than to the phenyl substituent

(K(1)-C(21’) = 3.232(2) Å, K(1)-C(25’) = 3.222(2) Å and K(1)-C(26’) = 3.052(2) Å).

These K-C distances fall within the range observed for the polymeric potassium

cyclopentadienyls and comparable to those observed in 2.6b(thf), but moderately shorter

than those in 2.6b(thf)2. The short and long K-B distances of 3.249(3) and 3.495(3) Å,

respectively, fall within the measured range in the potassium borata and diboratabenzene

complexes (3.06 – 3.69 Å).89

A shorter K(1)-O(1) distance of 2.644(2) Å is observed for 2.6c(thf), in comparison to

the polymeric potassium salts in 2.6b(thf) (2.733(2) Å) and 2.6b(thf)2 (2.726(4) and

2.816(4) Å). The geometry of these potassium complexes incorporating the 1,2-diaza-

3,5-diborolyl ligands was not observed in any other potassium sandwich compounds.

2.4. Synthesis and Spectroscopic Characterization of the Bicyclic 1,2-Diaza-3,5- diborolidine (2.3e) and their Alkali Metal Complexes (2.4e – 2.6e).

The previously described 1,2-diaza-3,5-diborolyl complexes showed that in all cases the organic substituents on the nitrogen atoms are positioned on either side of the ligand plane. To enhance the delocalization of the π electrons, the bicyclic 1,2-diaza-3,5- diphenyl-diborolidine (2.3e), was designed with the intention to force the nitrogen substituents in the plane of the ring. The preparation of the pyrazolidine hydrochloride salt (2.2c·HCl) was carried out according to a slightly modified reported procedure, as illustrated in Scheme 2.5.90 The bis-tert-butoxycarbonylpyrazolidine (Pz-boc) was

synthesized by the deprotonation of the di-tert-butyl hydrazoformate with NaH in

dimethylformamide (DMF) followed by the addition of 1,3-dibromopropane. The

product was obtained as colorless oil in quantitative yield. Addition of an excess

hydrogen chloride solution to a hexane solution of Pz-boc followed by stirring at ambient

temperature resulted in the formation of a colorless suspension of 2.2c·HCl. The product

was collected as a colorless hygroscopic solid (> 95% yield) and proved to be insoluble

in THF and other hydrocarbons.

Pyrazolidine (2.2c) was prepared in situ by the reaction of the 2.2c·HCl with 2

equivalents of KHMDS in THF. Addition of the stoichiometric amount of the 1,1-

bis(chlorophenylboryl)ethane and triethylamine to the mixture resulted in the successful

synthesis of the desired cyclic substituted 1,2-diaza-3,5-diborolidine 2.3e (Scheme 2.5).

Compound 2.3e was obtained as a colorless solid in 71% yield and had good solubility in organic solvents, with the exception of pentane. Colorless thin needles of 2.3e were grown by cooling a concentrated hexane solution of 2.3e to -35 °C, but the crystals were too thin for X-ray diffraction analysis.

Scheme 2.5.

O O Cl 2 NaH, DMF 4M HCl O NH O N H2N

O NH 1,3-dibromo- O N dioxane H2N propane O O Cl Pz-boc 2.2c.HCl

2 KHMDS Me H THF H Me M Me Ph Ph B B Ph Ph Ph Ph B B LM B B Cl Cl HN N N N N HN THF 2 NEt3 benzene LM = LiTMP 2.2c M = Li (2.4e) NaHMDS 2.3e Na (2.5e) KHMDS K (2.6e)

The characterization of 2.3e using multinuclear (1H, 11B and 13C) NMR, HRMS

and EI-MS analysis unambiguously confirmed the identity and purity of the compound.

As anticipated, the 11B NMR spectrum revealed a broad signal at δ 39.3 ppm for the

tricoordinated boron atom of the ring. The signature doublet and quartet signals of the

methyl and methine proton for the MeHCB2 fragment were observed at δ 1.44 and 1.73

ppm, respectively, in the 1H NMR spectrum (Figure 2.11). Similar to the described 1,2-

diaza-3,5-diborolidines, the broad 13C NMR signal of the ring carbon was observed at δ

22.5 ppm.

(a) Ph B H N Me N B Ph

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm (b) Ph Li+ B N Me N B Ph

7.5 6.5 5.5 4.5 3.5 2.5 1.5 ppm

1 Figure 2.11. H NMR spectra of 2.3e in C6D6 (a) and the lithium salt 2.4e in THF-d8 (b) at 25 ºC.

The deprotonation of 2.3e using the appropriate alkali metal reagents (Scheme

2.5) resulted in the formation of the corresponding salts (2.4e – 2.6e), which were isolated as colorless powders. The lithium and sodium salts are readily soluble in THF, but insoluble in other solvents. Meanwhile, the potassium salt is only sparingly soluble in THF and colorless thin long needles of 2.6e precipitated out of a THF solution at ambient temperature. However, because the crystals were too thin, they were not suitable for X-ray crystallographic analysis. The broad 11B NMR signal of the alkali metal salts

observed at δ 31.5 – 32.1 ppm was ca. 8 ppm upfield shifted with respect to the neutral

ligand 2.3e, in good agreement with the formation of an anionic ligand. In the 1H NMR

spectrum of these alkali metal salts, the single resonance for the methyl protons of the

MeCB2 fragment is shifted to a lower frequency at δ ranges from 2.23 – 2.29 ppm in 2.4e

– 2.6e, in comparison to previously described 1,2-diaza-3,5-diborolyl alkali metal salts.

The multiplet resonances corresponding to the (CH2)3N2 fragment in 2.3e was separated

into a quintet (-CH2) and a triplet (-NCH2) with the intensity ratio of 2:4, indicating the

equivalence of the methylene protons. The 13C NMR signal for the ring carbon of the

alkali metal salts was detected in the expected low field range of δ 91.1 – 92.4 ppm,

similar to the other 1,2-diaza-3,5-diborolyl alkali metal complexes.

2.5. Conclusions

The cyclopentadienyl analogs with a CB2N2 framework, 1,2-diaza-3,5-diborolyl,

have been synthesized using the ring closure method described in this chapter. The

coordination chemistry of the 1,2-diaza-3,5-diborolyl ligand towards alkali metals was

examined in detail and the results reveal substantial similarities, but also important differences with respect to the all-carbon analogs. The spectroscopic analyses provided conclusive information for the identity of the complexes in solution.

In agreement with the 13C NMR data, the structural analyses of the 1,2-diaza-3,5-

diborolyl ligands indicate electron delocalization over the NBCBN skeleton. The

intraannular B-C and B-N distances are intermediate in length between those

corresponding to single and double bonds, however the N-N bond lengths are typical for

single bonds. The latter observation, together with the fact that the substituents on

nitrogen feature significant deviations from the ring plane, implies that the electron

delocalization over the ring framework does not extend over the N-N bond. Similar to

cyclopentadienyl, these new π ligands coordinate face-on to metal ions, forming

monomeric (2.4c(thf)3, 2.5b(thf)3 and 2.5c(thf)3) and polymeric (2.4a, 2.5b, 2.6b(thf),

6b(thf)2 and 2.6c(thf)) sandwich complexes. The distances separating the metal ions are

systematically larger in the complexes described in this chapter than in the polymeric

cyclopentadienyl analogs, suggesting the lower coordination ability of the new ligands.

These ligands exhibit nevertheless a remarkable variability in their coordination manner

and, due to the difference in the radii of the ring atoms, the assignment of the

coordination mode is not explicitly apparent. Taking into consideration the geometry and

bond distances reported in the literature, η1 (C), η2 (BC), η3 (NBC), and η4 (BNNB) coordination modes have been assigned to the alkali metal complexes incorporating these novel π ligands. Relatively short distances between the metal and the ring carbon are present in all compounds, while only some of the complexes feature short distances between the metal and boron or nitrogen atoms. By comparison, the cyclopentadienyl ligand displays almost exclusively the η5-coordination in alkali metal complexes. No

trends regarding the preference of the heterocyclic ligand for a specific coordination

mode can be formulated at this point, but it appears obvious that the energetic differences

between the various coordination modes are small. This is not surprising, given the

predominantly ionic character of the bonding in the complexes. The coordination

chemistry of the 1,2-diaza-3,5-diborolyl ligands towards main-group and transition

metals will be investigated thoroughly in the following chapters.

CHAPTER THREE

Group 14 Metallocenes Incorporating 1,2-Diaza-3,5-diborolyl Ligands: Silicon,

Germanium and Tin Complexes

3.1. Introduction

Cyclopentadienyl metal complexes of the main group elements have been extensively investigated due to their diverse array of structural features.66b, 91 A variety of the group 14 sandwich complexes have been described and their structural characterization, reactivity and more recently their catalytic properties92 have been the subject of great interest. The first structurally characterized divalent silicon sandwich complex, Cp*2Si, revealed two independent molecules with surprisingly different structures: one with a centrosymmetric sandwich structure and the other featuring a bent sandwich structure, both with η5-coordinated cyclopentadienyl ligands.93 A stable η5-

+ - pentamethylcyclopentadienyl silicon cation, Cp*Si B(C6F5)4 , was recently isolated and its reactivity was investigated.94 In comparison to the silicon analog, the cyclopentadienyl derivatives of the heavier Group 14 metals such as germanium, tin and lead, have been investigated much more extensively. The synthesis and crystal structures of the divalent

95 96 97 germanium sandwich complexes, Cp2Ge, [(PhCH2)5C5]2Ge, [(Me3Si)3H2C5]2Ge, and

98 more recently Cp*2Ge, have been reported. The half-sandwich complex of germylenes

99 Cp*GeCl, [Cp*GeBr]2 and [Cp*Ge][BF4], were also structurally characterized.

β α M

M = group 14 metals

The structures of divalent tin sandwich compounds typically exhibit a bent

geometry and the increase in the centroid-metal-centroid angle (α) is dependent on the

100 degree of steric substitution of the cyclopentadienyl ring: Cp2Sn (143.7º), Cp*2Sn

101a i 101b 101c (144.1º), [( Pr2N)2PC5H4]2Sn (150.2º), [(PhCH2)5C5]2Sn (155.9º),

101d i 101e 101f [(Me3Si)3C5H2]2Sn (162º), , [ Pr4C5H]2Sn (165.0º) , [Ph5C5]2Sn (180º) and

i 101g [ Pr5C5]2Sn (180º). Decaphenylstannocene was the first symmetrical Group 14

sandwich compound with the Sn(II) atom situated on an inversion center between exactly

parallel and planar cyclopentadienyl rings.101f A series of centrosymmetric metallocenes

t of germanium, tin and lead, [( BuSiMe2)C5Me4]2M, were also synthesized and

structurally characterized.102 The crystal structure of a tin sandwich complex containing a

5 103 Lewis base weakly bonded to the tin atom, [(η -Cp)2Sn · TMEDA] (TMEDA =

5 5 [Me2NCH2]2) and the related adducts [(η -Cp)2Sn(μ-η -Cp)Na · PMDETA] (PMDETA =

104a 3 104b (Me2NCH2CH2)2-NMe) and [(η -Cp)3Sn]2[Mg(THF)6], were reported. A number of neutral and cationic half-sandwiches have also been reported: CpSnCl,105a

105b 105c 105d t 105e Cp*SnCl, [Cp*Sn][BF4], [Cp*Sn(Py)][CF3SO3] and [( BuC5H4)Sn][BF4].

The Group 14 metallocenes incorporating heterocyclic cyclopentadienyl rings are

reasonably rare: to date, only three structures of tin sandwich complexes containing 1,2-

106 107 108 azaborolyl (C3NB⎯), 1,3-diborolenyl (C3B2⎯) and pyrrolyl (C4N⎯) ligands, were

structurally characterized.

This chapter describes investigations of the coordination chemistry of the

heterocyclic 1,2-diaza-3,5-diborolyl ligands towards the Group 14 elements. The synthesis and spectroscopic characterization of the silicon, germanium and tin sandwich complexes (3.1 – 3.3) incorporating these ligands will be discussed. The X-ray structures of the trichloro(1,2-diaza-3,5-diborolyl)silyl complex (3.1b), the heterocyclic germanocene (3.2) and stannocenes (3.3a and 3.3b), will be described. The oxidative

addition reaction of the tin sandwich complexes with dichloromethane, yielding the

tetravalent tin complex (3.4) and its X-ray structure will also be presented. In addition,

the syntheses of the divalent 1,2-diisopropyl-3,5-diphenyl-1,2-diaza-1,3-diborolyltin(II)

chloride (3.5) and cationic 1,2-diaza-3,5diborolyltin(II) borate complexes (3.6a and 3.6b)

will be examined in this chapter.

3.2. Synthesis and Spectroscopic Characterization of the Trichlorosilyl-1,2-diaza-

3,5-diborolyl Complexes (3.1) and the X-ray Structure of 3.1b.

The synthesis of the stable divalent decamethylsilicocene by reaction of

cyclopentadienyl alkali metal salts with silicon tetrachloride, followed by the reduction of

the Cp*2SiCl2 with Li[C10H8], has attracted much attention. The preparation of silicon

sandwich complexes incorporating 1,2-diaza-3,5-diborolyl ligands utilizing the described

procedure was investigated, however the results were unsatisfying. Treatment of the 1,2-

diisopropyl-diaza-3,5-dimethyl-diborolyl lithium (2.4a) and 1,2-diisopropyl-diaza-3,5- diphenyl-diborolyl potassium (2.6b), respectively, with half an equivalent of silicon

tetrachloride resulted in the formation of trichloro(diazadiborolyl)silanes (3.1a and 3.1b)

and excess alkali metal salts, as illustrated in Scheme 3.1. The reaction resulted in a

mono-substitution at silicon and no further reaction was observed even upon heating the

reaction mixture at 70 °C for 24 hours. The reaction of a 1:1 molar ratio of 2.4a or 2.6b with SiCl4 in THF at ambient temperature afforded colorless solutions of the compounds,

3.1a and 3.1b, respectively. The products were isolated as colorless solids and after recrystallization from hexane colorless crystalline solids were obtained in reasonable yields (75% for 3.1a and 79% for 3.1b).

Scheme 3.1.

Me M Me M N R R R + 1/2 SiCl N B R R BB 4 Me BB 1/2 B + 1/2 NN THF R Si NN Cl - 1/2 MCl Cl Cl

M = Li, R = Me (2.4a) R = Me (3.1a), Ph (3.1b) M = K, R = Ph (2.6b)

These silicon compounds were characterized by multinuclear (1H, 11B, 13C and

29Si) NMR spectroscopy and mass spectrometry. The 1H and 13C NMR spectra of these

1 derivatives featured the expected signals corresponding to the Cs symmetry. The H

NMR spectra of 3.1a and 3.1b each revealed a singlet resonance for the methyl substituent on the ring carbon atom (δ 1.26 ppm for 3.1a and 1.43 ppm for 3.1b) and 29Si

3 satellites were observed with coupling constants JSi-H = 13.0 Hz. Two doublets (δ 1.34

and 1.37 ppm for 3.1a and 1.20 and 1.31 ppm for 3.1b) and a septet (δ 4.07 ppm for 3.1a and 4.19 ppm for 3.1b) corresponding to the methyl and methine protons of the isopropyl group at the nitrogen atoms were observed, indicating the diastereotopic inequivalence of the methyl groups. The methyl substituents on the boron atoms for 3.1a give rise to a singlet at δ 0.55 ppm and the set of multiplet signals for the phenyl substituents on the boron atoms of 3.1b were observed between δ 7.25 – 7.37 ppm. A broad 13C NMR resonance corresponding to the ring carbon appeared at δ 35.4 and 36.2 ppm in the two silicon derivatives, only ca. 13 ppm downfield from the signals displayed by the protonated ligands (δ 23.0 ppm for 2.3b) and significantly upfield shifted in comparison to the signals displayed by their alkali metal salts (δ 86.7 ppm for 2.4a and 96.7 ppm for

2.6b). This is a clear indication for the pronounced sp3 character of the ring carbon in these silicon compounds, which is typical for the σ complexes. Because of the electronic influence of the neighboring boron atoms, a direct comparison to cyclopentadienyl complexes is not feasible. As expected, the 11B and 29Si NMR spectra of 3.1a (δ 43.9 and

5.7 ppm, respectively) and 3.1b (δ 41.8 and 7.0 ppm, respectively) displayed singlet

resonances. The identity of compounds 3.1a and 3.1b was confirmed by low and high resolution mass spectrometry.

Figure 3.1. Molecular structure of 3.1b with 50% probability level thermal ellipsoids and all hydrogen atoms and a co-crystallized pentane molecule omitted for clarity.

Colorless thin plate crystals of 3.1b were obtained by cooling a concentrated pentane solution to -35 ºC. The crystallographic determination was performed and the selected structural and metric parameters are summarized in Tables 3.1 and 3.2

(Appendix One). The X-ray structure of 3.1b features the proposed structure containing

a pyramidal SiCl3 moiety that is σ-coordinated by the diazadiborolyl ligand through a Si-

C bond, as shown in Figure 3.1. The lengths of the Si(1)-C(1) (1.829(5) Å) and Si-Cl

bonds (2.038(2) – 2.040(2) Å) are comparable to those found in other

alkyltrichlorosilanes such as 1,2-bis(trichlorosilyl)ethane (Si-C 1.847(2) Å and Si-Cl

2.0225(6) – 2.0283(6) Å).109 Typical for a distorted tetrahedral geometry, the C-Si-Cl and

Cl-Si-Cl angles measured between 111.4(2) - 113.3(2)° and 104.02(8) - 107.29(9)°,

respectively. The geometry of the ligand skeleton in 3.1b is similar to that of the

protonated 1,2-diaza-3,5-diborolidine 2.3c. The CB2N2 ring is practically planar with the

sum of the intraannular ring angles of 539.8º. The CB2N2 ring has a folding angle of only

4.1° along the transannular C(1)···N(1) axis and the organic substituents on boron and

nitrogen atoms deviate only slightly out of the ring plane (less than 4.3°). This is

significantly less pronounced in comparison to 2.3c where the extraannular N-C bonds are tilted out of the CB2N2 plane by ca. 10 – 15º. Distinctive for a σ-coordination mode, the ring carbon adopted a distorted tetrahedral geometry, with the Si(1)-C(1) and C(2)-

C(1) bonds forming angles of 64.1 and 48.1° with the CB2N2 plane, respectively. In

agreement with the more pronounced sp3 character of the ring carbon in 3.1b, the C(1)-

C(2) (1.568(6) Å) and C(1)-B bond lengths (1.585(7) and 1.592(7) Å) are slightly but

systematically longer than those in 2.3c (1.462(4) Å, 1.572(4) and 1.581(4) Å,

respectively). The B-N bonds in 3.1b (1.382(6) and 1.404(6) Å) are marginally shorter in

comparison to those described in 2.3c. A similar structure was reported for Cp*SiCl3, with a planar ring skeleton and a Si-C bond length of 1.867(3) Å.110 In this derivative, the

angles formed by the Si-C and C-C(Si) bonds with the ligand plane measure 64.9 and

45.8°, respectively, nearly identical to the values observed in 3.1b.

3.3. Synthesis, Spectroscopic Characterization and X-ray Structure of Bis(1,2-

diaza-3,5-diborolyl)germanium (3.2) and Bis(1,2-diaza-3,5-diborolyl)tin (3.3)

Complexes.

The germanium and tin sandwich complexes incorporating 1,2-diaza-3,5-

diborolyl ligands were prepared utilizing the metathesis reaction of the corresponding

alkali metal salts with the appropriate Group 14 metal dihalides, as illustrated in Scheme

3.2. Treatment of the potassium salts, 2.6b, with germanium dichloride dioxane-adduct

in diethyl ether at ambient temperature immediately yielded a bright red solution.

Solvent removal and extraction of the reaction mixture with hexane afforded the

germanium sandwich complex (3.2) in 91% yield. Compound 3.2 was isolated as a

highly air and moisture-sensitive orange-red solid, with excellent solubility in

hydrocarbons. Red crystals of 3.2 were grown by cooling a concentrated pentane

solution to -35 ºC. Surprisingly, the treatment of the lithium salt, 2.4a, with GeCl2 ·

dioxane failed to produce the desired product based on the NMR analysis, which showed

the decomposition of the ligand. Attempted synthesis of the germanium complex at low

temperature also failed, resulting in the formation of a mixture of inseparable products.

Scheme 3.2.

R N B Me Me M N B R R + 1/2 M'X2 BB R Me M' R NN THF or Et2O B R B . N M'X2 = GeCl2 dioxane N SnCl2

M = Li, R = Me (2.4a) M' = Ge; R = Ph (3.2) M = K, R = Ph (2.6b) M' = Sn; R = Me (3.3a), Ph (3.3b)

The synthesis of the divalent stannocenes, 3.3a and 3.3b, was carried out by reaction of the lithium and potassium salts, 2.4a and 2.6b respectively, with tin dichloride in THF at ambient temperature. The tin sandwich complexes were obtained as red crystalline solids in high yields (82% in 3.3a and 95% in 3.3b). The complexes are soluble in organic solvents and decompose rapidly upon exposure to the atmosphere, immediately forming colorless solids. Red crystals of 3.3a and 3.3b were obtained by cooling the concentrated pentane solutions to -35 °C. The crystals of 3.3b were easier to

obtain than those of 3.3a due to the better crystallization properties of the diphenyl

substituted diazadiborolyl ligand.

(a) Ph N B N B Ph Sn THF Ph THF B impurity B Ph impurity N N

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm (b)

N B N B Sn B B N 119Sn satellites N

7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm

1 Figure 3.2. H NMR spectra of 3.3b (a) and 3.3a (b) recorded in C6D6 at 25 °C.

Characterization of the germanium and tin sandwich complexes by multinuclear

(1H, 11B and 13C) NMR spectroscopy and mass spectrometry was performed. The 1H

NMR spectra of 3.2 and 3.3b were very similar (Figure 3.2a). The signals corresponding to the methine and methyl protons of the isopropyl group on the nitrogen atoms were observed as a broad singlet and doublet. A single resonance for the methyl substituent on the ring carbon atom and a set of multiplets for the phenyl group on the boron atoms were also observed. The distinctive broad 13C NMR signals corresponding to the ring carbon

atom of 3.2 and 3.3b were observed at a significant lower frequency (δ 110.9 and 112.0

ppm, respectively) than for 3.1b (δ 36.2 ppm), supporting the sp2 character of the ring

carbon. Broad 11B NMR resonances at δ 36.4 ppm for 3.2 and 30.0 ppm for 3.3b were

observed in the 11B NMR spectra.

The 1H NMR spectrum of 3.3a shown in Figure 3.2b displayed two sets of

doublets and a broad singlet associated to the isopropyl groups on the nitrogen atoms.

119Sn satellites for the methyl substituents on the ring carbon atom were detected, with

3 13 coupling constants J119Sn-H = 11.3 Hz. The broad C signal of the ring carbon (δ 104.4 ppm) was found at a slightly higher field than for complexes 3.2 and 3.3b. The 11B NMR spectrum of 3.3a displays a broad resonance at δ 33.8 ppm. A sharp 119Sn NMR signal at

δ -1975.9 ppm was also observed, slightly downfield shifted in comparison to cyclopentadienyl tin sandwich complexes (δ -2129 to -2204 ppm).102 No signal was

detected in the 119Sn NMR spectrum of 3.3b, even at low temperature.

Figure 3.3. Molecular structure of 3.2 with 50% probability level thermal ellipsoids. For

clarity, only the ipso carbon atom of the phenyl substituent on B1 is shown and all hydrogen atoms omitted.

Single-crystal X-ray diffraction experiments were performed for compounds 3.2,

3.3a and 3.3b. The relevant structural parameters are summarized in Table 3.1 (Appendix

One). The structures of these complexes exhibited bent-sandwich geometries, typical for

the Group 14 metallocenes. In the molecular structures of 3.2 (Figure 3.3), the 1,2-diaza-

3,5-diborolyl ligands are η3-coordinated to the germanium atom through the skeletal

carbon and boron atoms. The Ge-C bonds (2.186(4) and 2.199(3) Å) are considerably

shorter in comparison to those reported for the cyclopentadienyl germanocenes (2.24 –

2.73 Å).95b, 96, 97, 98b To date, no structural characterization of germanocenes containing

heterocyclic Cp ligands has been reported, consequently no direct comparison of the Ge-

B bond lengths can be made. The Ge-B bond distances in 3.2, ranging from 2.520(4) –

2.600(4) Å, are noticeably longer than those in the published germacarborane complexes:

111a Ge[(Me3Si)2C2B4H4]2 (2.08(3) - 2.15(2) Å), [C10H8N2]Ge[(Me3Si)2C2B4H4] (2.208(7)

111b 111c – 2.371(7) Å) and Ge[R2C2B4(GeCl3)H3] (2.234(6) – 2.272(5) Å). The Ge···N

distances measure between 2.794(3) and 2.946(3) Å and are too long for bonding interactions. A view of the perpendicular projection of the Ge atom onto the planes of

the rings clearly illustrates the η3 binding mode of the ligand to the metal center (Figure

3.4). The displacement of the germanium atom away from the skeletal N atoms can be explained by the electron-rich character of the metal which avoids the contact with the electron-rich N atoms; similar observations had been made for 1,2-azaborolyltin complexes. The distances from the germanium atom to the planes of the CB2N2 rings measure 2.173(4) and 2.177(4) Å, slightly shorter in comparison the cyclopentadienyl

97, 98b analogs (2.20 – 2.26 Å). The dihedral angle that forms between the two CB2N2 planes (53.1(1)º) is considerably wider than the angles measured in the cyclopentadienyl germanocenes ([(Me3Si)3H2C5]2Ge = 8.2 and 10.5º, [(PhCH2)5C5]2Ge = 31º, and Cp*2Ge

= 31.3 and 31.6º).96, 97, 98b

Figure 3.4. View of the perpendicular projection onto the CB2N2 planes, illustrating the hapticity of the ligands. From left to right: 3.2 (η3,η3), 3.3b (η3,η3) and 3.3a (η4,η4).

The molecular structure of 3.3b (Figure 3.5) featured a η3,η3-coordinated, bent sandwich structure similar to 3.2. The dihedral angle between the planes of the two

CB2N2 rings of 50.2(1)º is larger than the corresponding angle observed in the 1,2- azaborolyl stannocene (46.5º)106 and intermediate between the corresponding angles in

100 101a Cp2Sn (55º) and Cp*2Sn (36º). The Sn-C bond lengths of 2.435(3) and 2.451(3) Å are shorter than those reported for the η5-cyclopentadienyl stannocenes (2.53 – 2.82

100, 101 5 t Å). For the boron-containing heterostannocenes, Sn[η - BuMe2H2C3NB]2 and

106 107 Sn[CpCo(C3B2)]2, similar short Sn-C bonds were observed (2.48 Å and 2.43 Å , respectively). The Sn-B bonds measuring between 2.666(4) and 2.749(4) Å in 3.3b are comparable to those observed in stannocenes containing boron heterocycles, Sn[η5-

t 106 107 BuMe2H2C3NB]2 (2.74(1) Å) and Sn[CpCo(C3B2)]2 (2.59 – 2.71 Å). Analogous to

3.2, long Sn···N distances ranging from 2.846(3) to 2.981(3) Å were observed in 3.3b.

However, unusually long Sn-N bond distances for the heterocyclic stannocenes, Sn[η5-

t 106 108 BuMe2H2C3NB]2 (2.917(9) Å) and Sn[tBu2H2C4N]2 (2.943(5) Å), have been

reported. Although the shortest Sn···N distance of 2.846(3) Å in 3.3b is ca. 8 – 10 pm

shorter than in the reported heterostannocenes, based on the positioning of the tin atom

with respect to the ring planes, it is apparent that the coordination mode of the ligand to

the metal center is best described as η3,η3 (Figure 3.4). The distances from the tin atom

to the planes of the CB2N2 rings measure 2.330(3) and 2.355(3) Å and are slightly shorter

in comparison to substituted cyclopentadienyl analogs (parallel stannocene derivatives:

2.38 – 2.49 Å;101f, 101g, 102 bent stannocenes: ca. 2.42 Å).101e

The structure of 3.3a exhibited a monomeric bent sandwich geometry with the tin metal center η4-coordinated by the 1,2-diaza-3,5-diborolyl ligands through the skeletal

carbon, boron and one of the nitrogen atoms. The bent structure of 3.3a features short

Sn-C bonds (2.428(2) and 2.432(2) Å) that are comparable to those observed in 3.3b

(2.435(3) and 2.451(3) Å). The Sn-B bonds ranging from 2.663(2) to 2.774(3) Å are

comparable to those observed in 3.3b and the boron-containing heterocycle stannocenes.

The two Sn-N bond distances of 2.744(2) and 2.834(2) Å are clearly shorter than the

Sn···N distances of 2.936(2) and 2.998(2) Å in 3.3a and those measured in 3.3b (2.846(3)

- 2.981(3) Å). The hapticity of the ligands is best illustrated by the perpendicular projection of the tin atom onto the ring planes, shown in Figure 3.4, revealing a η4,η4- coordination moiety of the tin sandwich complex. The metal-to-ligand plane distances of

2.347(1) and 2.366(2) Å are slightly greater than those measured in 3.3b. The dihedral angle measured between the ring planes is 50.0(7)º, which is very similar to the one measured in 3.3b.

Figure 3.6. Side view of one of the CB2N2 rings in 3.3a (left) and 3.3b (right) illustrating

the envelope conformation of the ring structure with folding along the C⋅⋅⋅N axis. For

clarity, only the metal and ring atoms are represented.

The bond lengths and bond angles of the CB2N2 ring display little variation

between all three structures, showing a minor dependence on the organic substituents on

boron (methyl or phenyl) or on the coordinated metal. The planarity of the rings is

slightly better for the phenyl derivatives 3.2 and 3.3b than for the methyl derivative 3.3a,

with the sum of the intraannular angles ranging between 538.5 and 539.7° for the phenyl

derivatives and 537.5° for the methyl derivative. The slight deviation from planarity

results in an envelope conformation, which is best described by a folding along a C···N

axis, with one boron atom pointing away from the metal center, as shown in Figure 3.6.

The folding angles observed between the CBN2 and CBN planes for the phenyl derivatives are smaller (8 and 12º in 3.2; 6 and 9º in 3.3b) than those for the methyl derivatives (13 and 16º in 3.3a).

In these group 14 sandwich complexes, the intraannular B-C bonds (1.515(5) –

1.543(5) Å) are shorter than the extraannular B-C bonds (1.571(4) – 1.600(5) Å), as a

result of the expected multiple bond character of the former. The B-N bond lengths,

ranging between 1.429(5) and 1.453(4) Å, are comparable to the lengths of B-N bonds

observed in borazines (1.42 – 1.44 Å),71 and noticeably longer than those in 3.1b

(1.382(7) and 1.405(7) Å). The intraannular B-C bonds are slightly longer in the group

14 metallocenes presented here than in the alkali metal salts of the same ligands (1.486(6)

– 1.514(3) Å), while the B-N bonds are marginally shorter than the ones observed in the

alkali metal salts (1.452(3) – 1.481(3) Å). A similar behaviour was observed for the

corresponding B-C and B-N bonds in the bis(1,2-azaborolyl)tin complex in comparison

to the 1,2-azaborolyl lithium complex.106 The N-N bonds in complexes 3.2 – 3.3

(1.425(3) – 1.443(2) Å) are comparable to the single N-N bond in hydrazine (1.45 Å).72 It can therefore be concluded that the π electron delocalization over the N-B-C-B-N skeleton of the ligand is slightly less significant than in the alkali metal complexes of the same ligands.

3.4. Synthesis, Spectroscopic Characterization of Bis(1,2-diaza-3,5-diborolyl)

chloromethyl tin chloride (3.4) and 1,2-Diaza-3,5-diborolyltin chloride (3.5) and the

X-ray Structure of 3.4.

The reactivity properties of the novel heterostannocenes, 3.3a and 3.3b, was

investigated in comparison to the Group 14 metallocenes. Dissolution of the red crystals

of 3.3a in dichloromethane resulted in a slow color change from a dark red to a light

yellow solution at ambient temperature. This was a consequence of an oxidative addition of the tin(II) complex, yielding a tetravalent tin complex as shown in Scheme 3.3. The tin(IV) complex (3.4) was isolated as a colorless solid in 45% yield upon purification of the thick pale yellow residue by washing with a minimal amount of pentane and drying under vacuum. The complex had good solubility in organic solvents. Colorless crystals of 3.4 were grown by slow evaporation of the pentane and THF solution of 3.4 at ambient

temperature.

Scheme 3.3.

N B Me N Me B Cl CH2Cl2 Me Cl Me Sn R Me C B B H2 N Me N B Me N B N

RMe Sn 3.4 R B B R Ph N N B N SnCl2 N Me 2 B CH Cl 2 2 Ph Sn Cl R = Me (3.3a), Ph (3.3b) 3.5

No reaction was observed upon dissolution of 3.3b in dichloromethane, even over several

days, according to the NMR analysis. The reduced reactivity of 3.3b towards

dichloromethane may be due to the reduction of the nucleophilicity of the tin atom as a

result of the replacement of the more electron-donating methylated ligands with the

phenylated ligands. This is consistent with the study of the B-heteroatom-substituted 1,2-

azaborolyl complexes by Fu, showing that electron-rich substituents on the skeletal boron

atoms will increase the electron-donating ability of the 1,2-azaborolyl ligands.25

The tetravalent tin complex, 3.4, was characterized by multinuclear (1H, 11B, 13C

and 119Sn) NMR and mass analysis. The 1H and 13C NMR spectra of 3.4 exhibited the

expected resonance signals for two equivalent 1,2-diaza-3,5-diborolyl ligands with four

sets of doublets (δ 1.15 – 1.30 ppm) corresponding to the methyl protons of the isopropyl groups. The methine protons of the isopropyl groups overlapped into a sextet at δ 3.78 ppm and two singlet resonances were observed, corresponding to the methyl substituents on the boron atoms (Figure 3.7). Tin satellites associated with the protons of the methyl

3 3 group on the ring carbon (δ 1.68 ppm, J117Sn-H = 111 Hz and J119Sn-H = 116 Hz) and the

2 methylene protons of the CH2Cl fragment (δ 3.25 ppm, JSn-H = 11 Hz) were observed. A

broad 11B signal (δ 42.6 ppm) was found at a lower field than the one detected in 3.3a (δ

33.8 ppm). As expected, the 119Sn NMR spectrum of 3.4 revealed a singlet resonance at

δ 45.0 ppm, dramatically downfield shifted from the divalent stannocene 3.3a (δ -1975.9 ppm), which is typical of the triorganotin(IV) halides (δ -89.4 to 69.2 ppm; 164 ppm in

112 + Me3SnCl). The mass spectrum of 3.4 featured the molecular ion peak [M] at m/z =

+ + + 590, as well as the ion fragments [M-LCH2Cl] , [M-LCH2Cl2] and [L-H] .

N B N B Cl Sn Cl C 117 119 B H2 Sn and Sn 117 119 Sn and Sn N B satellites satellites N

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm

1 Figure 3.7. H NMR spectrum of complex 3.4 recorded in C6D6 at 25 °C.

Figure 3.8. Molecular structure of 3.4 with thermal ellipsoids drawn at the 50%

probability level and all hydrogen atoms omitted for clarity.

The molecular structure of 3.4 was determined by single-crystal X-ray diffraction

(Table 3.1 and 3.5, Appendix One). Consistent with the NMR data, the structure revealed a tetracoordinated tin with a distorted tetrahedral arrangement (Figure 3.8). The tin atom is bonded to two 1,2-diaza-3,5-diborolyl ligands in a σ fashion, as well as to the

fragments of CH2Cl and Cl. The three Sn-C bonds (2.167(4) – 2.195(3) Å) are noticeably shorter than those in 3.3a and 3.3b, in agreement with the σ bonding mode. The Sn-Cl

bond distance of 2.385(1) Å is shorter in comparison to corresponding distances found in

the cyclopentadienyl tin chloride derivatives, CpSnCl (2.679(5) Å)105a and Cp*SnCl

(2.657(1) and 2.693(1) Å),105b but similar to those observed in diorganotin(IV) dichloride,

113 Ph(Et)SnCl2 (2.359(2) and 2.367(3) Å). Characteristic of a distorted tetrahedral

geometry around the metal center, the C-Sn-C and C-Si-Cl angles measure from 109.7(1)

to 120.8(1)° and 101.2(1) to 107.43(9)°, respectively.

The geometry of the ligand skeletons differ slightly from each other. The C(12)-

B(3)-N(3)-B(4)-N(4) ring is nearly planar with the sum of the intraannular ring angles of

539.9º, while the C(2)-B(1)-N(1)-B(2)-N(2) ring deviates slightly from planarity with the

sum of the intraannular ring angles of 538.5º. The two CB2N2 rings are slightly folded

along the transannular B(1)···B(2) axis by angles of 3.1º and l1.5º. The boron and

nitrogen substituents are bent out of the ring planes by 3 – 12º. The C(12)-C(13) and

Sn(1)-C(12) bonds forming angles of 42.7 and 61.7º with the ring plane are comparable

to those observed in silicon complex 3.1b (48.1 and 64.1º, respectively). In comparison,

the angles created between the other ring plane and the C(2)-C(3) and Sn(1)-C(2) bonds

(26.4 and 81.2º) are indicative of a less pronounced sp3 character of the ring carbon atom

C(2). The intraannular B-C bonds (1.554(5) – 1.572(6) Å) are longer than those observed

in 3.3a; this is consistent with the increased sp3 character of the ring carbon atoms.

Moreover, the skeletal B-C bond distances are closer in lengths to the exocyclic B-C bonds (1.573(6) – 1.582(6) Å), further supporting the single B-C bond character. In

agreement with the insufficient π-electron delocalization along the CB2N2 ring, the B-N

bonds of 1.404(4) – 1.414(5) Å are notably shorter in comparison to those measured in

3.3a (1.433(3) – 1.453(4) Å), and comparable to those observed in 3.1b (1.382(7) and

1.405(7) Å) and the protonated ring of 2.3c (1.412(3) and 1.416(3) Å). The N-N bond lengths of 1.447(4) and 1.457(4) Å are similar to those in 3.1b and the protonated rings

2.3c and 2.3d, characteristic for a single N-N bond.

Addition of SnCl2 to the dichloromethane solution of 3.3b at ambient temperature

immediately resulted in a color change from dark red to orange (Scheme 3.3). The

comproportionation reaction yielded the 1,2-diaza-3,5-diborolyltin chloride (3.5) in a

64% yield. The product was obtained as a yellow solid, stable under inert atmosphere

and with good solubility in organic solvents. The characterization of 3.5 using MS and

NMR analysis confirmed the identity and purity of the compound. The 1H NMR spectrum of 3.5 featured two sharp doublets (δ 1.05 and 1.14 ppm) and a broad singlet (δ

4.06 ppm) corresponding to the isopropyl substituents on the nitrogen atoms. A downfield shift for the single resonance (δ 2.87 ppm) of the methyl substituent on the ring carbon atom was observed in comparison to 3.3b. A doublet and two triplets with intensity ratio of 4:4:2 corresponding to the ortho, meta, and para-protons of the phenyl ring could be distinguished. An upfield shift of the broad 13C signal of the ring carbon

was observed at δ 95.1 ppm in comparison to the stannocene 3.3b (δ 112 ppm),

suggesting a reduction of the sp2 character of the ring carbon. The 11B NMR spectrum

also showed a broad resonance at 36.0 ppm, only ca. 4 ppm downfield shifted from

complex 3.3b. No signal was detected in the 119Sn NMR spectrum of 3.5. The molecular

ion peak [M]+ (m/z = 472) was detected in the mass spectrum of 3.5, further confirming

its identity.

3.5. Synthesis and Spectroscopic Characterization of Cationic 1,2-Diaza-3,5-

diborolyltin Borate Complexes (3.6).

The synthesis of Group 14 triple-decker cationic sandwich complexes by reaction

114 of [Cp*M][B(C6F5)4] with Cp*2M (M = Sn, Pb) has recently been reported. The

synthesis of a heterocyclic analog of the cationic tin half -sandwich complexes containing

the 1,2-diaza-3,5-diborolyl ligands was attempted by the treatment of the stannocenes

(3.3a and 3.3b) with [H(Et2O)2]B(C6F5)4, as illustrated in Scheme 3.4. The NMR scale

reaction of 3.3a with [H(Et2O)2]B(C6F5)4 in THF-d8 at ambient temperature immediately

afforded an orange solution. The 1H NMR spectrum of the reaction mixture is shown in

Figure 3.9, and it reveals the formation of the 1,2-diaza-3,5-diborolyltin borate complex

(3.6a) and the neutral 1,2-diaza-3,5-diborolidine ligand (2.3a) in a 1:1 ratio. Complex

3.6a was isolated as a yellow solid in 94% yield after washing the yellow residue with

pentane. The product is an air- and moisture-sensitive compound, which is stable for

months in an argon atmosphere and is readily soluble in organic solvents except for

pentane.

Scheme 3.4.

R B N Me Me H N N B N R [H(Et2O)2]B(C6F5)4 B R R R B BB Me Sn R Me + R THF NN B Sn B R - 2 Et O N 2 N B(C6F5)4

R = Me (3.3a), Ph (3.3b) R = Me (3.6a), Ph (3.6b) R = Me (2.3a), Ph (2.3b)

NiPr Et O N Me H Me 2 N B Me B B Me B N Me Me N Sn BMe B(C6F5)4 CMe NiPr Et2O BMe

HCMe NiPr NiPr HCMe

4.5 3.5 2.5 1.5 0.5 -0.5 ppm

1 Figure 3.9. H NMR spectrum of the NMR reaction of 3.3a with [H(Et2O)2]B(C6F5)4 in

THF-d8 at 25 °C, revealing a 1:1 molar formation of the 1,2-diaza-3,5-diborolyl tin borate

3.6a, and the neutral 1,2-diaza-3,5-diborolidine 2.3a.

The 1H NMR spectrum of 3.6a revealed distinctive chemical shifts for the signals

of the methyl and isopropyl substituents on the skeletal ring atoms ca. 0.1 – 0.2 ppm

downfield shifted in comparison to the resonances in 3.3a. Although the broad 13C NMR

signal of the ring carbon atom was not observed due to quadrupolar relaxation over the

boron atoms, the other 13C signals corresponding to the ligand and the borate ion were detected. As expected, a broad resonance corresponding to the tricoordinated boron atoms (δ 36.9 ppm) and a sharp resonance associated with the tetracoordinated borate ion

(δ -17.4 ppm) were observed in the 11B NMR spectrum. The positive and negative mode

electrospray mass spectra of 3.6a featured signals for the cationic

+ [Sn(MeC(BMe)2(NiPr)2)] (m/z = 312) and anionic [B(C6F5)4]⎯ (m/z = 678) fragments of the complex. Various crystallization methods such as slow evaporation of the dichloromethane solution, cooling of a concentrated solution to -35 °C, liquid-liquid

diffusion of THF and hexane solution, and gas-liquid diffusion of Et2O and C6H5Br solution, failed to produce crystals of 3.6a suitable for structural determination.

Owing to the better crystallization properties of the diphenyl 1,2-diaza-3,5- diborolyl ligand, the synthesis of the phenyl derivative of the tin borate complex (3.6b) utilizing the same procedure was carried out, in an attempt to obtain a crystal structure of the tin borate complex. Complex 3.6b was obtained as a yellow solid after washing with hexane and drying under vacuum in 86% yield. Similar to 3.6a, the multinuclear NMR and MS analysis of 3.6b showed very similar results, confirming the purity and identity of the complex. Unfortunately, the attempted crystal growth was also unsuccessful.

3.6. Conclusion

Group 14 metallocenes containing the 1,2-diaza-3,5-diborolyl ligands (3.2 and

3.3) were synthesized and characterized by multinuclear NMR spectroscopy and X-ray crystallography. The NMR analysis of the novel germanocene (3.2) and stannocenes

(3.3a and 3.3b) revealed that in solution the molecular structure of the group 14 metallocenes is retained. In the solid state, the molecular structures of these metallocenes exhibited similarities, but also differences with respect to the cyclopentadienyl analogs.

Typical for cyclopentadienyl complexes, bent sandwich structures were obtained for 3.2,

3.3a and 3.3b with dihedral angles formed between the two CB2N2 rings of 50.1(1) –

53.1(1)º, wider than those measured for cyclopentadienyl analogs. In contrast to the η5-

coordinated cyclopentadienyl derivatives, the 1,2-diaza-3,5-diborolyl ligands display

more diverse coordination properties. The phenyl derivative of 3.2 and 3.3b revealed a

η3,η3-coordination of the 1,2-diaza-3,5-diborolyl ligands to the metals, while the methyl

derivative 3.3a displayed a η4,η4-coordination mode. In the structures, the metal ions are

preferentially bonded to the skeletal carbon and boron atoms of the 1,2-diaza-3,5-

diborolyl ligands. The lack of coordination to ring N atoms is likely due to the electron

richness of the group 14 metals possess. The attempt to synthesize a silicocene

containing the 1,2-diaza-3,5-diborolyl ligands led instead to the isolation of the

trichloro(diazadiborolylsilyl) complexes (3.1a and 3.1b). The solid-state structure of the trichlorosilyl derivative 3.1b was determined, and featured a pyramidal SiCl3 moiety that

is σ-coordinated by the 1,2-diisopropyl-diaza-3,5-diphenyl-diborolyl ligand. The multinuclear NMR and MS analysis of the silicon complexes also provided conclusive evidence for the identity of the isolated trichlorosilyl complexes.

The oxidative addition of 3.3a with dichloromethane led to the tetravalent tin complex 3.4, through the insertion of the tin center into the C-Cl bond. Characterization of 3.4 showed that in solid state the tetracoordinated tin complex contains two σ- coordinated 1,2-diaza-3,5-diborolyl ligands. The oxidative addition reaction of 3.3a is quite unusual in comparison to the cyclopentadienyl tin complexes and in fact no such reaction had been reported for organic derivatives of stannocenes. These heterostannocenes (3.3a and 3.3b) displayed similar chemical properties to the carbon- base stannocenes. The synthesis of the neutral 1,2-diaza-3,5-diborolyltin chloride 3.5 and

the cationic tin borate complexes 3.6a and 3.6b was achieved by the reactions of the

stannocenes 3.3a and 3.3b with tin chloride and Jutzi’s acid ([H(Et2O)2]B(C6F5)4), respectively. Characterization of these complexes by multinuclear NMR and mass spectrometry provided conclusive evidence for the identity of the complexes.

CHAPTER FOUR

Group 12 Metallocenes Incorporating 1,2-Diaza-3,5-diborolyl Ligands: Zinc,

Cadmium and Mercury Complexes

4.1. Introduction

In comparison to the rich cyclopentadienyl coordination chemistry of main-group

and transition metals, only a handful of metallocenes of the electron-rich Group 12

elements have been reported. The first structurally characterized sandwich complex of

1 2 115 zinc, Cp2Zn, displays a polymeric structure with η and η -coordinated ligands. The other base-free zincocenes feature asymmetric structures with η1 and η5 bonded

116 5 cyclopentadienyls. Recently, the synthesis of dizincocenes, [(η -C5Me4R)2Zn2] has attracted a lot of interest and their structures were subsequently the subject of theoretical calculations.117 Cyclopentadienyl derivatives of the heavier Group 12 metals, cadmium

and mercury, have been even less studied. The crystal structures of cadmium sandwich complexes containing Lewis bases coordinated to the metal have been published:

118a, 118b 118c Cp2CdPy2, Cp2Cd(TMEDA) and Cp2Cd(PMDETA). The structures of two

t i 119 base-free cadmocenes, ( Bu3C5H2)2Cd and ( Pr4C5H)2Cd were recently reported. A

dimeric half-sandwich complex of cadmium, [Cp*CdN(SiMe3)2]2, was also structurally

characterized.120 Three sandwich compounds of mercury have been crystallographically

1 121a t 121b characterized and they all exhibit η -coordination modes: Cp2Hg, (Bu3C5H2)2Hg,

t 121c 1 and [( BuMe2Si)C5Me4]2Hg. The structures of three η bonded mercury half-

122a 122b sandwiches have also been determined: (Cl5C5)HgC6H5, Cp*HgCl and

t 121c [( BuMe2Si)C5Me4]HgCl. Group 12 metallocenes incorporating heterocyclic

cyclopentadienyl analogs have not been reported.

This chapter describes the coordination chemistry of the heterocyclic 1,2-diaza-

3,5-diborolyl ligands with electron-rich Group 12 elements. The synthetic methods and

the characterization of a series of zinc, cadmium and mercury complexes (4.1 – 4.3) containing the 1,2-diaza-3,5-diborolyl ligands introduced in Chapter 2 will be discussed.

The X-ray structural analysis of the zincocenes (4.1a, 4.1b), cadmocenes (4.2a, 4.2b) and

a mercurocene (4.3) featuring sandwich structures with η1 and η3-coordinating ligands

will be presented.

4.2. Synthesis and Spectroscopic Characterization of Bis(1,2-diaza-3,5-diborolyl)

Zinc (4.1), Cadmium (4.2) and Mercury (4.3) Sandwich Complexes.

A series of group 12 metallocenes incorporating the 1,2-diaza-3,5-diborolyl ligands were synthesized by reaction of the corresponding alkali metal salts with the appropriated group 12 metal dihalides, as depicted in Scheme 4.1. Treatment of the 1,2- diaza-3,5-diborolyllithium (2.4a) and 1,2-diaza-3,5-diborolylpotassium (2.6b) with zinc dichloride, cadmium dibromide and mercury dichloride in THF cleanly produced the zincocenes (4.1a, 4.1b), cadmocenes (4.2a, 4.2b) and mercurocene (4.3), respectively.

These group 12 complexes were isolated as air and moisture-sensitive colorless solids in reasonable yields (65% for 4.1a, 58% for 4.1b, 94% for 4.2a, 80% for 4.2b and 94% for

4.3) and had excellent solubility in organic solvents.

Scheme 4.1.

N R Me M N B B + 1/2 M'X Me R R 2 R BB M' THF R NN Me B R B N M'X2 = ZnCl2 N CdBr2 HgCl2

M = Li, R = Me (2.4a) M' = Zn; R = Me (4.1a), Ph (4.1b) M = K, R = Ph (2.6b) M' = Cd; R = Me (4.2a), Ph (4.2b) M' = Hg; R = Me (4.3)

Colorless crystals of the zinc complexes were grown by cooling the concentrated solutions 4.1a and 4.1b in pentane at -35 °C. It is worth noting that the crystals of 4.1a are very soluble at ambient temperature and slowly dissolved in Paratone oil. Pale yellow prismatic crystals of cadmocene 4.2b were obtained by cooling a concentrated

pentane solution of the material at -35 °C. Colorless needle-like crystals of 4.2a co-

crystallized with LiBr were obtained by cooling a concentrated solution of the mixture at

-35 ºC. Colorless crystals of 4.3 were obtained by slow evaporation of a concentrated

pentane solution at -35 °C.

These group 12 complexes were characterized using multinuclear (1H, 11B, 13C,

113Cd and 199Hg) NMR spectroscopy and mass spectrometry, which provided conclusive

results for the identity and purity of the complexes. The 1H NMR spectra of the B,B’-

dimethyl derivatives (Figure 4.1) display similar resonance patterns and the chemical

shifts of the signals are only slightly affected by the metal. Unlike the lithium salt 2.4a,

two resonances corresponding to diastereotopic methyl protons of the isopropyl

substituents were observed in the 1H and 13C NMR spectra, thus implying that the two faces of the ligand are inequivalent and finally that the expected sandwich structures are

retained in solution. In addition, the 1H and 13C NMR spectra of 4.2a revealed the

presence of 113Cd satellites for the singlet resonances corresponding to the methyl

3 2 4 substituents on the ring carbon ( JCd-H = 59 Hz and JCd-C = 42 Hz) and ring boron ( JCd-H

= 8 Hz), confirming that no ligand dissociation occurred in solution. In the 1H and 13C

NMR spectra of 4.3, 199Hg satellites for the singlet resonances associated with the methyl

3 4 2 substituents on the ring carbon and boron ( JHg-H = 143 Hz, JHg-H = 20 Hz and JHg-C = 65

Hz) were also detected. The low temperature 13C NMR spectra of these complexes

13 recorded in THF-d8 revealed a broad C signal for the ring carbon at δ 63.7 ppm in 4.1a

(-50 °C), 70.6 ppm in 4.2a (-20 °C) and 77.3 ppm in 4.3a (-20 °C), which is ca. 10 – 13

ppm upfield shifted with respect to the signal observed in the lithium salt 2.4a (δ 86.7

ppm at -50 °C). This is a consequence of an increase in the covalent character of the

metal-ligand interaction in the group 12 sandwich complexes. The 11B NMR spectra

display only one broad signal with a chemical shift situated in a narrow range of δ 38 –

41 ppm, which is comparable to the signals observed for the lithium salt of the ligand (δ

38 ppm in 2.4a). There were minor differences between the 1H, 13C and 11B NMR spectra recorded for the cadmium and mercury complexes in C6D6 and in THF-d8.

(a) N N B B Zn B B N N

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm (b)

N N B B Cd 113Cd 113Cd B satellites B N satellites N

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm (c)

N N B B Hg 199Hg 199Hg B satellites satellites B N N

4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm

Figure 4.1. Representative 1H NMR spectra of zincocene 4.1a (a), cadmocene 4.2a (b) and mercurocene 4.3 (c) in C6D6 at 25 °C

However, the chemical shift of the 113Cd resonance observed in the 113Cd{1H} NMR spectrum of 4.2a recorded in different solvents (δ 256.2 ppm in C6D6 and δ 397.9 ppm in

THF-d8) was significantly different. This is likely due to the formation of 4.2a(thf)n adducts in the coordinating THF solvent. No 113Cd NMR data for cyclopentadienyl

cadmium compounds are available in the literature for comparison. The 199Hg{1H} NMR spectrum of 4.3 in C6D6 displays a single resonance at δ -2063.3 ppm, significantly

upfield shifted in comparison to the reported 199Hg resonance of Hg[η1-

t 121c C5Me4(SiMe2 Bu)]2 (δ = -1112 ppm). The EI mass spectra of all complexes described

+ here have shown peaks that were assigned to the molecular ions [ML2] , as well as to the

fragments [ML]+ and [L]+.

The multinuclear NMR analysis of the zinc (4.1b) and cadmium (4.2b) complexes

containing the 1,2-diisopropyl-diaza-3,5-diphenyl-diborolyl ligand featured very similar

spectra with only slight differences in the chemical shift of the signals. The 1H and 13C

NMR spectra exhibited two resonances that were assigned to the methyl protons of the

isopropyl substituents, indicating inequivalent faces of the ligand (Figure 4.2). 113Cd

3 satellites for the singlet resonance of the methyl substituent on the ring carbon ( JCd-H =

2 1 13 59 Hz and JCd-C = 42 Hz) were observed in the H and C NMR spectra of 4.2b. The

proton-coupled 113Cd NMR spectrum of 4.2b revealed a septet resonance at δ 286.1 ppm,

as a result of the coupling to the six equivalent protons of the methyl substituents on the

3 13 ring carbon ( JCd-H = 60 Hz). The broad C signals corresponding to the ring carbon

were observed at δ 69 and 77 ppm in the 13C NMR spectra of 4.1b and 4.2b, respectively.

This upfield shift of ca. 18 – 20 ppm with respect to the corresponding potassium salt (δ

97 ppm in 2.6b) is indicative of the increase covalent character of the metal-ligand

interaction in the former compounds. The 11B NMR spectra of 4.1b and 4.2b showed the

expected broad boron signal at δ 39 and 37 ppm, respectively. These values are

comparable to those detected for the B,B’-dimethyl derivatives 4.1a and 4.2a. The

electrospray ionization (ESI) mass spectra of 4.1b and 4.2b recorded using acetonitrile

+ solutions displayed signals that were assigned to the molecular ions [ML2] , and the fragments [ML]+ and [L]+.

(a) (b) N N B Ph B Ph Cd Ph B B N Ph N 113Cd satellites 290 285 280 ppm

8.0 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 ppm (c)

113Cd satellites

ring carbon 1314 12

80 70 60 50 40 30 20 10 ppm

1 113 Figure 4.2. (a) The H NMR spectrum of 4.2b in THF-d8, (b) the proton coupled Cd

13 NMR spectrum of 4.2b in C6D6 and (c) a portion of the C NMR spectrum of 4.2b in

THF-d8 at 25 °C.

4.3. X-ray Structures of the Bis(1,2-diaza-3,5-diborolyl) Zinc (4.1), Cadmium

(4.2) and Mercury (4.3) Sandwich Complexes.

The crystal structures of 4.1a, 4.1b, 4.2b, 4.2a · BrLi(thf)3 and 4.3 have been

determined using single-crystal X-ray diffractometry. Selected structure refinement

details are summarized in Table 4.1 and selected bond lengths and angles are provided in

Tables 4.2 – 4.6 (Appendix One). The metric parameters of the CB2N2 ring display little

variation between all structures, showing a minor dependence on the organic substituent

on boron (methyl or phenyl) or the coordinated metal. The planarity of the rings is

slightly better for the phenyl derivatives 4.1b and 4.2b than for the methyl derivatives

4.1a, 4.2a · BrLi(thf)3 and 4.3, with the sum of the intraannular angles ranging between

539.5 and 539.7° for the former and between 538.6 and 539.1° for the latter. The slight deviation from planarity results in an envelope conformation, which can be described as either a folding along the B···B axis, with the ring carbon pointing towards the metal, or a folding along a C···N axis, with one boron atom pointing away from the metal (Figure

4.3).

Figure 4.3. Side view illustrating the envelope conformation of one of the CB2N2 rings in

4.2a · BrLi(thf)3 with folding along the B(3)···B(4) (left) and C(11)···N(3) axes (right).

For clarity, only the metal and the ring atoms are represented.

In all cases, the latter model offers a slightly better description of the structure, with the

CBN2 / CBN folding angle measuring 9.6 to 11.3° for the methyl derivatives and 5.1 to

6.5° for the phenyl derivatives.

In all complexes, the intraannular B-C bonds (1.507(6) - 1.557(9) Å) tend to be shorter than the extraannular B-C bonds (1.570(7) - 1.597(11) Å), in agreement to the expected multiple bond character. The B-N bond lengths, ranging between 1.390(12) and

1.444(7) Å, are comparable to those observed in borazines (1.42 – 1.44 Å).71 The intraannular B-C and B-N bonds in these group 12 complexes are slightly longer and shorter, respectively, than in the alkali metal salts of the same ligands (B-C = 1.486(6) –

1.514(3) Å and B-N = 1.452(3) – 1.481(3) Å).123 A similar behavior was observed for the

corresponding C-C bonds in cyclopentadienyl complexes of the group 12 metals by

comparison to alkali metal cyclopentadienyls. Such behavior was attributed to the strong

involvement of the ring carbon atom in the binding to the metal and as a result the

decreased aromaticity of the ligand in these group 12 metallocenes. The N-N bonds in

complexes 4.1 – 4.3 (1.424(4) - 1.486(12) Å) are much closer in length to the single N-N

bond in hydrazine (1.45 Å)72 than to the double N=N bond in pyridazine (1.34 Å).79

Hence, the π-electron delocalization over the N-B-C-B-N skeleton of the ligand in these group 12 complexes can best be described by the resonance structures shown below:

Me Me Me

R R BB R R BB R BB R

NN NN NN

The orientation of the ring substituents with respect to the ligand plane is similar

in all complexes. The methyl substituent on the ring carbon is bent noticeably away from

the metal, with the C-C bond forming angles of 12.2 to 21.7° with the CB2N2 plane. This

is common in metallocenes of group 12 metals and was correlated to the partial sp3 hybridization of the ring carbon bonded to the metal and to the corresponding σ character of the η1-coordination mode. In the methyl derivatives, one of the exocyclic B-C bonds

lies closer to the ligand plane (∠ B-C / CB2N2 = 1.0 to 5.3°) while the other is more

pronouncedly bent away from the metal (∠ B-C / CB2N2 = 7.9 to 13.8°). In both 4.1b

and 4.2b the phenyl substituents on boron are only slightly bent away from the metal,

with angles between the B-C bonds and the CB2N2 plane of 0.1 to 4.9°. As observed in

the alkali metal salts of these ligands, the isopropyl substituents on nitrogen alternate above and below the ligand plane, with one of the N-C bonds tilted towards the metal

(∠N-C / CB2N2 = 10.5 to 25.8°) and the other away from the metal (∠ N-C / CB2N2 =

5.8 to 18.4°). This orientation is in disagreement with the partial multiple bond character of the B-N bonds and does not appear to be sterically induced, since it is not visibly affected by the replacement of methyl substituents on boron with phenyl.

The crystal structure of 4.1a contains two independent molecules that are very similar and will be discussed together (Figure 4.4).123a The zinc atom is nearly

symmetrically coordinated by the two parallel ligands that form angles of 8.9 (molecule

1) and 0° (molecule 2) with each other. The zincocenes that have reported previously are

all asymmetric, featuring η1,η2 or η1,η5-coordination modes of the cyclopentadienyl ligands.115, 116 The only symmetric η5-coordination mode has been observed only in

dizincocenes.117

Figure 4.4. Molecular structure of one of the two independent molecules in 4.1a with

50% probability level thermal ellipsoids. All hydrogen atoms omitted for clarity. The disorder involving the isopropyl groups (C15 and C18) is not depicted.

Figure 4.5. Molecular structure of 4.1b with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity.

The zinc center in 4.1a lies almost above the ring carbon, with its projection onto the

ligand plane falling only slightly inside the ligand pentagon. Its coordination can be

described as either η1,η1 or η3,η3-modes. Although, the Zn···B distances (2.443(4) -

2.546(5) Å) are longer than the shortest Zn-B bonds observed in sandwich compounds

2- 124a, 124b containing borane as ligands ([Zn(B10H12)2] (2.191(7) – 2.209(10) Å) and [(nido-

124c C2B9H11)ZnNMe3]2 (2.160(3) – 2.177(3) Å) ), the latter description was chosen given

the small yet distinct difference with respect to the other structures reported herein.

The structure of 4.1b (Figure 4.5) is similar to its methyl derivative. The main difference is that the Zn-C bond is nearly orthogonal to the CB2N2 plane forming angles

of 86.9 and 87.8° in 4.1b vs. 81.1, 82.6 and 80.3° in 4.1a. Therefore, the coordination

mode of the ligand is typically η1(π). The Zn-C bond distances are very similar in the two complexes (1.999(2) - 2.004(4) Å), which fall in the range expected for cyclopentadienyl complexes (1.99 – 2.37 Å for η5-coordination and 1.95 – 2.22 Å for η1-

coordination).115, 116 From all five crystal structures of zincocenes with η1,η5-coordination

that have been determined, only Zn[C5Me4(SiMe3)]2 is not disordered and provides a

reliable comparison for the ligand geometry. As opposed to 4.1a and 4.1b, this

compound features a structure with a σ-coordinating cyclopentadienyl and a distorted

tetrahedral ring carbon. The Zn-C and C-C bonds involving the ring carbon form angles

of 69.0 and 43.0°, respectively, with the ring plane. With C-C / CB2N2 angles of 12.2 –

20.2°, the methyl substituent on the ring carbon is bent much less out of the ring plane in

the two zincocenes incorporating diazadiborolyl ligands. The Zn···B distances in 4.1b

(2.564(2) - 2.626(2) Å) are even larger than in 4.1a.

100

Figure 4.7. Molecular structure of 4.2b with 50 % probability level thermal ellipsoids and all hydrogen atoms omitted for clarity.

101

The molecular structures of the two cadmium complexes 4.2a · BrLi(thf)3 (Figure

4.6) and 4.2b (Figure 4.7) feature very similar η1(π)-coordination mode of the ligands,

with the Cd-C bonds forming angles of 85.0 – 87.9° with the ring planes. In both

compounds, the angles formed by the C-C bonds involving the ring carbons with the ring

plane fall in a narrow range between 15.6 and 19.1°. Derivative 4.2b is isostructural with

4.1b, containing parallel ligands (dihedral angle 7.5°) and displaying a nearly linear

geometry at the cadmium center (C-Cd-C 175.3°). Only two base-free cadmocenes have

been structurally characterized, featuring η1,η1 and η1,η2-coordinating cyclopentadienyl

ligands.119 The latter structure is disordered and therefore only the former,

t ( Bu3C5H2)2Cd, provides a basis for a rigorous comparison of the metric parameters.

Comparable to 4.2b, the Cd-C and C-C bonds forming angles of 87.2 and 28.6° with the plane of the ring skeleton were observed. The distances of the Cd-C bonds are identical

t 119 in 4.2b (2.201(5) and 2.206(5) Å) and ( Bu3C5H2)2Cd (2.201(2) Å).

In complex 4.2a · BrLi(thf)3, coordination of the BrLi(thf)3 unit through a long

2- 125 Br-Cd bond (2.731(8) and 2.7916(14) Å vs. 2.55 – 2.59 Å in CdBr4 ) imposes a narrower C-Cd-C angle of 159.3(1)°, and this prevents the ligands from adopting a parallel arrangement (dihedral angle 19.6°). The Br atom is disordered between two positions and the lithium atom has the expected distorted tetrahedral geometry. All three reported structures of cadmocenes containing Lewis bases coordinated to the metal feature cyclopentadienyl as a ligand and have two or three nitrogen atoms completing the coordination sphere of cadmium.118 The cyclopentadienyl ligand displays η1(σ), η1(π) and

η2-coordination in these compounds, and the Cd-C bonds are longer in the complexes

1 featuring η -coordination (2.307(5) – 2.410(6) Å) than in 4.2a · BrLi(thf)3 (2.248(3) and

102

2.255(3) Å). This is likely due to the better donor ability of the nitrogen bases and their

capacity to reduce the electrophilicity of cadmium and hence its affinity for the

cyclopentadienyl ligands. The angles formed by the Cd-C bonds with the planes of the

cyclopentadienyl rings measure ca. 77 - 81° for the η1(π) binding mode and ca. 59° for the η1(σ) binding mode, being noticeably smaller than the corresponding angles in 4.2a ·

BrLi(thf)3 (85.0 and 87.7°). In addition, the C-Cd-C angles are more acute in the

cyclopentadienyl derivatives (111.4(2) and 129.1(2)°). The Cd···B distances in 4.2a ·

BrLi(thf)3 and 4.2b range from 2.706(6) Å to 2.872(4) Å and are substantially longer

126a than the shortest Cd-B bonds observed in [(Et2O)2Cd(B10H12)]2 (2.37(3)-2.46(3) Å),

2- 126b 2- 126c [Cd(B9H13)2] (2.272(4)-2.302(4) Å), and [Cd(B6H6)2] (2.388(10)-2.451(14) Å).

The structure of the mercury metallocene 4.3 is heavily disordered, with two equally abundant molecules occupying the same position in the crystal lattice. The Hg center in the two molecules having nearly perpendicular C-Hg-C axes occupies the same crystallographic position (Figure 4.8 bottom). The isopropyl substituents in each of these molecules are also disordered between two positions. Careful modeling allowed for a good refinement of the structure. Similar to the previously examined structures, the

1 mercury complex also possess the η (π)-coordinating ligands (Hg-C / CB2N2 angles of

87.6 and 88.9°) and a nearly linear geometry at the mercury center(C-Hg-C angles of

175.6(3) and 176.1(3)°). The angles formed by the C-C bonds involving the ring carbons

with the ring planes are relatively acute (20.2 and 21.7°), which further supporting the

η1(π) binding mode of the ligands. The Hg-C bonds measure 2.168(6) and 2.177(6) Å are

comparable in length to the bonds observed in the three mercurocenes that have been

crystallographically characterized, all featuring η1 bonding modes (2.09(2) - 2.163(3)

103

Å).121 The bonding in these latter compounds can also be described as η1(π), however, the

geometries of these derivatives indicate a significant σ-bonding component. The Hg-C

and C-CHg bonds in these complexes form angles of 68.1 – 77.3° and 31.9- 37.7°, respectively, with the planes of the cyclopentadienyl rings.

Figure 4.8. Molecular structure of 4.3 with 50 % probability level thermal ellipsoids. For

clarity, only two of the disordered ligands around the mercury center are represented and

all hydrogen atoms omitted. Bottom: a view of the disorder in the crystal structure of 4.3.

The two molecules are equally abundant and only the metal and ring atoms are

represented.

104

The Hg···B distances in 4.3 (2.729(7) - 2.780(6) Å) are much longer than the shortest Hg-

B bonds observed in mercury half-sandwich compounds featuring 7,8-dicarba-nido-

2- 127 undecaboranes, [B9C2R2H9] , as ligands (2.178(8) – 2.208(6) Å).

4.4. Conclusion

A series of the new heterocyclic zinc, cadmium, and mercury metallocenes

containing 1,2-diaza-3,5-diborolyl ligands were synthesized and characterized. The

multinuclear NMR analysis shows that in contrast to the related alkali metal derivatives,

no ligand dissociation takes place and the sandwich structure of the group 12 metallocenes is retained in solution. This is in agreement with the more pronounced covalent character of the metal-ligand interaction in the latter compounds and is

consistent with the behavior of the corresponding cyclopentadienyl analogs. In all complexes, the ring carbon plays the essential role in the binding of the ligand to the metal and the geometry of the complexes is fairly symmetric. Four base-free complexes were isolated, featuring a linear environment around the metal, with C-M-C angles situated between 174 and 180°. Compound 4.2a crystallized with one unit of BrLi(thf)3 coordinated to the cadmium center and as a result a more acute C-Cd-C angle of 159° was observed. Except for the zinc complex 4.1a, where the Zn-C bond is noticeably tilted towards the ligand, and the coordination is better described as η3,η3, the coordination mode of the ligand is typically η1(π),η1(π) in all other group 12

metallocenes. The M-C bond is nearly perpendicular onto the CB2N2 plane in

compounds 4.1b, 4.2a · BrLi(thf)3, 4.2b and 4.3 (85.0 – 88.9°), while the C-C bond

involving the ring carbon forming angles of 12.2 to 21.7° with the same ring plane. In

105

the group 12 cyclopentadienyl derivatives available for comparison, the σ character of the

η1 bonding is more pronounced than in these heterocyclic group 12 complexes. This suggests that the stabilization of a planar configuration for the negatively charged carbon, through π donation to the electron-deficient boron atoms in 1,2-diaza-3,5-diborolyl is more effective than the stabilization through delocalization over the ring in cyclopentadienyl. The metal-carbon bond lengths in all complexes are comparable to those observed in the cyclopentadienyl analogs. The coordination properties of the 1,2- diaza-3,5-diborolyl ligands with transition metal complexes will be examined in the following Chapter.

106

CHAPTER FIVE

Group 4, 8 and 9 Metallocenes Incorporating 1,2-Diaza-3,5-diborolyl Ligands, and a

Ruthenium Triple-Decker Sandwich Containing an Unusual Eight-Membered

C2B4N2 Ring

5.1. Introduction

The coordination chemistry of the novel 1,2-diaza-3,5-diborolyl ligands toward the group 12 and 14 metals displayed similar properties to the properties of their cyclopentadienyl analogs. As presented in the beginning of this thesis, numerous transition metal metallocenes containing heterocyclic cyclopentadienyl ligands have been characterized and their structural characterization revealed η5-coordination modes of the ligands.

This chapter extended the investigations of the coordination chemistry of the 1,2- diaza-3,5-diborolyl ligands to transition metals, targeting the synthesis of sandwich complexes containing η5-coordinating π ligands. These complexes will provide a better comparison of the electronic properties of the heterocyclic ligands to the well known cyclopentadienyl derivatives. The synthesis and spectroscopic characterization of

107

rhodocene (5.1b and 5.1e), ferrocene (5.2), ruthenocene (5.3) and zirconocene (5.4)

complexes containing the 1,2-diaza-3,5-diborolyl ligands, along with the structural

characterization of these sandwich complexes will be discussed. The electrochemical

properties of the ferrocene and ruthenocene analogs were examined using cyclic

voltammetry. The alkali metal complexes of the examined 1,2-diaza-3,5-diborolyl

ligands acted as oxygen scavengers towards metal carbonyl compounds, producing 1,2-

diaza-4-oxa-3,5-diborolidines (5.6) and alkynyl metal carbonyl complexes

− [(CO)5M(C≡CMe)] (5.5).

In addition, the investigation of the 1,2-diaza-3,5-diborolyl ligands has been

extended to the study of a heterobicycle with C2B4N2 framework. The synthesis and

spectroscopic characterization of the 1,5-diaza-2,4,6,8-tetraborolidine (5.7), as well as the stepwise deprotonations of 5.7 to its mono and dipotassium salts (5.8 and 5.9, respectively), and the X-ray structures of 5.7 and 5.9 will be presented. The dianionic

1,5-diaza-2,4,6,8-tetraborolyl ligand has the potential to act as bridging π ligand and form multidecker sandwich compounds. A triple-decker ruthenium sandwich complex (5.10)

featuring the novel eight-membered C2B4N2 ring as the middle deck was synthesized and

characterized. The crystal structure of the triple-decker ruthenocene along with its

electrochemical and theoretical investigation will be described.

108

5.2. Synthesis, Spectroscopic Characterization and X-ray Structures of the 1,2-

Diaza-3,5-diborolyl Rhodium Cyclooctadiene Complexes (5.1)

The synthesis of 1,2-diaza-3,5-diborolyl rhodium cyclooctadienes (5.1) was

achieved utilizing the salt metathesis reaction of the corresponding alkali metal salts with

[Rh(cod)Cl]2, as shown in Scheme 5.1. Treatment of the lithium (2.4a) and potassium salts (2.6b) with half an equivalent of [Rh(cod)Cl]2 in THF immediately produced dark

brown solutions that yielded 5.1a and 5.1b, respectively, as brown solids upon removal of the solvent. The crude solids were re-crystallized from hexane and the orange-brown rhodium complexes were obtained in moderate yields (57% in 5.1a and 62% in 5.1b).

Mixing of the lithium salt 2.4e with the stoichiometric amount of [Rh(cod)Cl]2 in THF afforded a dark red solution at ambient temperature. Recrystallization of the product from hexane at -35 °C yielded 5.1e in 45% yield as a yellow-orange solid. These air and moisture-sensitive solids are readily soluble in organic solvents.

Scheme 5.1.

Me N M N B R R R + 1/2 [Rh(cod)Cl] B BB 2 R Me Rh NN THF - MCl

M = Li, R = Me (2.4a) R = Me (5.1a) M = K, R = Ph (2.6b) Ph (5.1b)

Me Li N N B Ph Ph Ph B B B + 1/2 [Rh(cod)Cl]2 Ph Me Rh N N THF - LiCl

2.4e 5.1e

109

(a)

N N B B Rh

cod-CH cod-CH 2

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 ppm (b)

N BPh N B Ph B Ph Rh cod-CH2 cod-CH

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm (c) H H H c b c H H a BPh b N N B Ph Ha B Ph cod-CH cod-CH Rh 2 Ha + Hb Hc

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm

1 Figure 5.1. H NMR spectra of rhodium compounds 5.1a recorded in THF-d8 (a), 5.1b

(b) and 5.1e recorded in C6D6 (c) at 25 °C.

The spectroscopic characterization of rhodium complexes (5.1) provided

conclusive evidence regarding their identity. The 1H and 13C NMR spectra of 5.1a, 5.1b

and 5.1e featured all the anticipated resonances for the organic substituents of the

heterocyclic ligand and the cyclooctadiene group (Figure 5.1). Similar to the group 12

and 14 metallocenes, upon the coordination of the diazadiborolyl ligand to the rhodium

center two resonances associated to the diastereotopic methyl protons of the isopropyl

group were observed in the 1H and 13C NMR spectra of 5.1a and 5.1b. The multiplet

110

resonance patterns corresponding to the (CH2)3N2 fragment were observed at δ 1.36 (-

CH2), 2.85 (-NCH2) and 3.19 ppm (-NCH2) with 2:2:2 intensity ratio, respectively, in the

1H NMR spectrum of 5.1e, clearly indicating the inequivalence of the methylene protons.

These results confirmed that in solution the heterocyclic ligand is coordinated to the

Rh(cod) fragment forming the expected half sandwich compounds. Moreover, the skeletal carbon and boron atoms of the 1,2-diaza-3,5-diborolyl in complexes 5.1 (δ 13C =

81.3 ppm in 5.1a, 82.6 ppm in 5.1b and 82.6 ppm in 5.1e; δ 11B = 26.5 ppm in 5.1a, 28.9 ppm in 5.1b and 23.8 ppm in 5.1e) were evidently upfield shifted upon complexation relative to their corresponding alkali metal salts (δ 13C = 86.7 ppm in 2.4a, 96.8 ppm in

2.6b and 91.6 ppm in 2.4e; δ 11B = 38.3 ppm in 2.4a, 38.6 ppm in 2.6b and 31.5 ppm in

2.4e). The coupling of the rhodium-carbon nuclei for the vinylic proton of the

13 1 cyclooctadiene was observed in the C NMR spectrum of 5.1e ( JRhC = 13.6 Hz) and it was comparable to the value observed for the cyclooctadiene rhodium derivatives.128 The electron impact mass spectrum of 5.1b and 5.1e features the molecular ion peak

[Rh(cod)L]+ (m/z = 528 and 484, respectively), as well as ion fragments [RhL]+ (m/z =

418 and 376) and [L]+ (m/z = 317 and 274) resulting from the successive loss of

cyclooctadiene and rhodium. In addition, the results obtained from the HRMS and

elemental analysis of 5.1b and 5.1e also confirmed the identity and purity of the rhodium

compounds.

The crystal structures of the heterorhodocenes 5.1b and 5.1e were determined by

X-ray diffraction and the relevant bond lengths and angles are summarized in Table 5.4

and 5.5 (Appendix One). Despite the difference in the organic substituents on the ring

nitrogen atoms, the CB2N2 ring structures of both rhodium compounds are very similar.

111

The planarity of the heterocyclic rings is nearly perfect with the sum of the ring angles of

538.5° in both compounds. The geometry of the ring skeleton is best characterized as an

envelope conformation with a folding along the B(1)···B(2) vector, and the ring carbon

pointing towards the rhodium center (Figure 5.2). The folding angle formed between the

B2N2 / BCB planes measured 12.0° for 5.1b and 11.3° for 5.1e. A similar folding of the

6 5 129, 130 diborolyl ring was observed in (η -arene)Rh(η -C3B2) with an angle of 7 – 15°.

Figure 5.2. A side view of 5.1b (left) and 5.1e (right) illustrating the envelope conformation of the CB2N2 ring. For clarity, only the metal and the ring atoms are

represented.

Although the conformation of the ring skeletons in these rhodocenes is very

similar, the bond distances of the CB2N2 rings reveal slight differences between the mono

and bicyclic ligands 5.1b and 5.1e, respectively. The most noticeable difference between

the two rings is the shorter N-N bond observed for 5.1e (1.449(4) Å) relative to 5.1b

(1.481(3) Å). The endocyclic B-C bonds of 5.1b (1. 512(4) and 1.516(4) Å) are

comparable to those in 5.1e (1.528(6) and 1.540(6) Å). The B-N bond distances of the

ligands (1.457(4) and 1.472(4) Å in 5.1b; 1.452(5) and 1.454(5) Å in 5.1e) fall

marginally outside the typical range in borazines (1.42 – 1.44 Å)71 and are shorter in

comparison to the value observed in η5-azaborolyl rhodium cyclooctadiene (1.512(3)

112

Å).22 These B-C and B-N bonds are comparable in length to the corresponding bonds measured in the group 1, 12 and 14 sandwich complexes featuring the 1,2-diaza-3,5- diborolyl ligands, which were described in the previous chapter. Consistent with the expected multiple bond character of the intraannular B-C bonds, longer extraannular B-C bonds were observed (1.579(4) and 1.585(4) Å in 5.1b; 1.575(6) and 1.582(6) Å in 5.1e).

- 131 - 129, In the rhodium sandwich complexes containing borolyl (C4B ), 1,3-diborolyl (C3B2 )

130 - 22 and 1,2-azaborolyl (C3BN ) ligands, the B-C bonds ranging from 1.53 to 1.56 Å are slightly longer than those in 5.1b and 5.1e.

The structures of 5.1b and 5.1e are relatively similar (Figure 5.3) and typical for cyclopentadienyl rhodium cyclooctadiene half sandwich complexes. Only one such structure containing a heterocyclic η5-azaborolyl analog having a very similar geometry

has been reported.22 The rhodium atom is η5 and η4-coordinated by the 1,2-diaza-3,5- diborolyl ligand and the cyclooctadiene molecule, respectively. The Rh-C bonds involving the cyclooctadiene moiety range from 2.076(4) to 2.185(3) Å and fall in the

5 4 132 5 4 typical range for (η -Cp)Rh(η -cod) (2.107(7) – 2.128(8) Å) and (η -C5Br5)Rh(η -cod)

(2.11(2) – 2.12(2) Å).133 The distance measured from the rhodium atom to the plane of

the heterocyclic ring is only slightly larger in 5.1b (1.921(3) Å) than in 5.1e (1.908(3) Å),

and comparable to the distance observed in (η5-azaborolyl)Rh(cod) (1.934(2) Å).22 This

6 5 distance is noticeably smaller in boron-containing rhodocenes: (η -arene)Rh(η -C3B2)

129, 130 5 5 131 (1.79 – 1.82 Å) and (η -Cp*)Rh(η -C4B) (1.80 Å). Consequently, the Rh-C and

Rh-B bonds (Rh(1)-C(1) = 2.218(3) Å, Rh-B = 2.335(3) and 2.377(3) Å for 5.1b; Rh(1)-

C(1) = 2.217(3) Å, Rh-B = 2.324(4) and 2.385(4) Å for 5.1e) are marginally longer than

113

those in the rhodium sandwich complexes containing borolyl (Rh-C = 2.145(4) –

2.199(5) Å; Rh-B = 2.275(6) Å)131 and diborolyl (Rh-C = 2.159(2) – 2.206(2) Å; Rh-B =

2.253(6) – 2.30(3) Å) 129, 130 ligands. The Rh-N bond distances (2.284(2) and 2.334(2) Å

for 5.1b; 2.249(3) and 2.309(3) Å for 5.1e) are shorter than the Rh-B bonds as a

consequence of the small folding along the B⋅⋅⋅B axis of the ring structure (Figure 5.2).

Figure 5.3. Molecular structures of 5.1b (top) and 5.1e (bottom) with 50% probability

level thermal ellipsoids and all hydrogen atoms omitted for clarity.

114

Longer Rh-B and Rh-N bonds of 2.544(2) and 2.408(2) Å respectively, were determined

for the asymmetrically coordinated η5-azaborolylrhodium complex featuring shorter Rh-

C bond distances (2.141(2) – 2.266(2) Å).22 In contrast, the 1,2-diaza-3,5-diborolyl

ligands are bonded to the rhodium atom in an arrangement much closer to the ideal η5 geometry, as illustrated by the orthogonal projection of the Rh atom onto the plane of the

CB2N2 ring (Figure 5.4).

Figure 5.4. Perpendicular projection onto the CB2N2 planes, illustrating the coordination

mode of the ligands. From left to right: 5.1b, 5.1e, 5.2 (two CB2N2 rings) and 5.3.

In complex 5.1e, the cyclic propylene group deviates less from the ring plane (N-

C bonds are 4 and 7° tilted above the ligand plane) with a folding angle of 5.7° between

the C(3)-N(1)-N(2)-C(5) and C(1)-B(1)-B(2)-N(1)-N(2) planes. In complex 5.1b, one of

the isopropyl substituents lies nearly in the ligand plane (less than 2° out of plane), and

the other isopropyl group is clearly situated above the CB2N2 ring, forming an angle of

16° between this plane and N(2)-C(6) bond. As a result of the steric restriction in the

cyclic substituent in 5.1e, a much smaller C(3)-N(1)-N(1)-C(5) torsion angle of 3.6(4)°

was observed relative to 5.1b (torsion angle of C(3)-N(1)-N(2)-C(6) = 20.7(3)°). Our

target was to improve the π delocalization ability in the heterocyclic ligand by enforcing planarity of the nitrogen substituents. In both structures, the methyl substituent on the

115

ring carbon is situated in the ring plane, with a deviation angle of only 1°, and the phenyl

groups deviate only slightly out of plane with angles between the extraannular B-C bonds

and the CB2N2 plane of 6 – 8°. In addition, the orientation of the phenyl groups with

respect to the ring plane is also quite different in both structures. For compound 5.1b, the

two phenyl rings are orientated nearly perpendicular to the CB2N2 plane (83.2(1) and

83.7(1)°). In compound 5.1e, the phenyl rings are twisted by 35.1(1) and 39.2(1)° with

respect to the ligand plane.

5.3. Synthesis, Spectroscopic Characterization, Electrochemical Studies and X-

ray Structures of Bis(1,2-diaza-3,5-diborolyl)iron (5.2) and 1,2-Diaza-3,5-diborolyl

Ruthenium Pentamethylcyclopentadienyl (5.3) Complexes.

The group 8 metallocenes incorporating the heterocyclic 1,2-diaza-3,5-diborolyl

ligand were prepared by reaction of the lithium salt 2.4e, with the stoichiometric amount

of FeCl2(thf)2 and [Cp*RuCl]4, respectively, in THF, producing dark brown solutions of

the iron and ruthenium complexes (Scheme 5.2). The products were isolated as a reddish

brown solid for the iron derivative 5.2 (89% yield) and a light yellow solid for the

ruthenium derivative 5.3 (80% yield). The iron and ruthenium compounds are readily

soluble in organic solvents and X-ray quality crystals of 5.2 (red thin plates) and 5.3

(yellow blocks) were obtained by slow evaporation of the hexane solutions at ambient

temperature. The reactions of the lithium 2.4a and potassium 2.6b salts with FeCl2(thf)2 and [Cp*RuCl]4 also generated dark brown solutions, however the isolation of the desired products was not possible and the NMR spectra revealed a complex mixture of intractable products. The attempted crystallization failed to yield crystals suitable for

116

structural analysis. The failure to synthesize these group 8 sandwich complexes

containing the diisopropyl substituted 1,2-diaza-3,5-diborolyl ligands may be a result of

the inferior π-coordinating ability of these ligands or, most likely, due to their poor

crystallization properties.

Scheme 5.2.

N N B Ph B Me + 1/2 FeCl2(thf)2 Ph Fe Ph THF Me B - LiCl B N Ph Ph N B Li N 5.2 Me N B N Ph Ph N B 2.4e B Me + 1/4 [Cp*RuCl]4 Ph Ru THF - LiCl

5.3

Compounds 5.2 and 5.3 were characterized using multinuclear (1H, 11B and 13C)

NMR spectroscopy and MS spectrometry, and good elemental analysis results were also obtained. The NMR spectral data of 5.2 and 5.3 are compatible with the proposed η5- coordination of the 1,2-diaza-3,5-diborolyl ligand to the iron and ruthenium centers. The

1H and 13C NMR spectra exhibited all resonances expected for the organic substituents of

the ligand, as well as the pentamethylcyclopentadienyl (Cp*) group, with the appropriate

integration ratio. Analogous to rhodocene 5.1e, the inequivalent methylene protons of

the cyclic propylene substituent (CH2)3N2 were observed as four multiplet resonances

1 with intensity ratio of 1:1(–CH2):2:2(–NCH2) in the H NMR spectra of 5.2 and 5.3

117

(Figure 5.5). The broad 13C NMR signal of the ring carbon was observed at δ 65.5 ppm

for 5.2 and 80.0 ppm for 5.3, upfield shifted with respect to the alkali metal salts of the

same ligands (δ 91.1 – 92.4 ppm, 2.4e – 2.6e). The 11B NMR signals corresponding to

the skeletal boron atoms in both the iron and ruthenium complexes were observed at δ

13.6 and 14.7 ppm, respectively, are more upfield shifted relative to analogous iron and

5 30b ruthenium complexes containing heterocyclic ligands: (η -C3BN)2Fe (δ = 24 ppm),

5 134 5 135 5 (η -C3B2)FeCp* (δ = 19 ppm), (η -C3B2)RuCp* (δ = 19 – 22 ppm) and (η -

24, 29 C3BS)RuCp* (δ = 21 – 22 ppm). The electron impact mass spectrum of 5.2 exhibited

+ + the molecular ion peaks [FeL2] (m/z = 602) and the ion fragments [FeL] (m/z = 328) and [L]+ (m/z = 274) resulted from the successive loss of the ligand and iron.

(a) Hd Hb Hc Ha Hb N Ph H N B BPh a B Ph Me Fe Ph Me B 2:2 ratio (H + H ) 1:1 ratio B N a b (H + H ) Ph N c d

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm

(b) Hd Hb Hc Ha Hb 2:2 ratio (H + H ) N Ph a b 1:1 ratio (H + H ) H N B c d a B Ph Me Ru

BPh 3.0 2.5 2.0 1.5

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm

1 Figure 5.5. H NMR spectra of 5.2 (a) in CD2Cl2 and 5.3 (b) in C6D6 at 25 °C.

118

Only the molecular ion peak of the ruthenocene [Cp*RuL]+ (m/z = 509) was observed in

the mass spectrum of 5.3. The identity and purity of these group 8 metallocenes was also

supported by the HRMS analysis.

119

X-ray crystal structure determinations were performed for 5.2 and 5.3 and the

selected bond lengths and bond angles are given in Table 5.6 and 5.7 (Appendix One).

The molecular structure of 5.2 revealed a classic sandwich structure with the two π

ligands η5-coordinated to the iron center (Figure 5.6), very similar to the reported structure of the bis(1,2-azaborolyl)iron complex.30b The 1,2-diaza-3,5-diborolyl ligands

appear to be essentially coplanar with dihedral angle of 1.3(1)° between the two CB2N2 ring planes. The two five-membered rings in 5.2 are twisted by ca. 22° from the eclipsed conformation (Figure 5.6 bottom). Similar orientation of the ligands was observed in related iron sandwich complexes.8 The metal-to-ring plane distances measured between

the iron and CB2N2 planes of 1.674(3) and 1.676(3) Å showed that the ligands are

equidistant from the Fe atom. These distances are in the expected range of the reported

heteroferrocenes containing pyrrolyl ligands (1.651 – 1.683 Å).8 Within the

diazadiborolyl ligand, the Fe-N bonds ranging from 1.971(2) to 1.997(2) Å are shorter than the Fe-C and Fe-B bonds, measuring 2.177(2) – 2.192(2) Å and 2.198(3) – 2.217(3)

Å, respectively. This reflects the asymmetric coordination of the iron atom relative to the ring center, which is displaced slightly towards the nitrogen atoms as illustrated in Figure

5.4. The Fe-N, Fe-C and Fe-B distances fall in the expected range associated with iron sandwich complexes containing the 1,2-azaborolyl (Fe-N = 2.01 – 2.09 Å; Fe-C = 2.01 –

2.14 Å; Fe-B = 2.14 – 2.24 Å),136 pyrrolyl (Fe-N = 2.00 – 2.02 Å; Fe-C = 2.02 – 2.08 Å)8 and 1,3-diborolyl (Fe-C = 1.899(6) and 2.121(4) Å; Fe-B = 2.249(5) Å)134 ligands.

120

The crystal structure of 5.3 features a typical sandwich complex with the

ruthenium ion η5 ligated by the 1,2-diaza-3,5-diborolyl and pentamethylcyclopentadienyl

ligands, as depicted in Figure 5.7. A number of ruthenium sandwich complexes

containing π-heterocyclic ligands and pentamethylcyclopentadienyl and displaying very

121

similar sandwich structures have been reported. Analogous to 5.2, the two π-bonded

ligands in 5.3 are practically coplanar, with a dihedral angle of 1.1(1)° between the planes

of the CB2N2 and Cp* rings. This angle is comparable to those observed in the related

5 40 5 heterocyclic ruthenocenes: 2.6(4)° in [(η -Me2C3HN2)RuCp*], 5.7(1)° in [(η -

t 49 5 23 5 Bu2C2N2P)RuCp*], 2.5(1)° in [(η -H3C3OBPh)RuCp*] and 6.6(1)° in [(η -

29b (iPr2N)BSC3H3)RuCp*]. The orientation of the CB2N2 ring with respect to the Cp*

ring deviates from the eclipsed conformation in the structure of 5.3 as shown in Figure

5.7 (bottom). The ruthenium atom is situated closer to the Cp* ring (Ru-Cp* plane

distance of 1.797(2) Å vs. Ru-CB2N2 plane distance of 1.836(2) Å). Similar metal-to-

5 ligand distances were observed for other heteroruthenocenes: [(η -Me2C3HN2)RuCp*]

40 5 t 49 5 (1.785(3) and 1.837(3) Å), [(η - Bu2C2N2P)RuCp*] (1.817(3) and 1.823(3) Å), [(η -

23 5 H3C3OBPh)RuCp*] (1.789(2) and 1.853(2) Å) and [(η -(iPr2N)BSC3H3)RuCp*]

(1.797(2) and 1.851(2) Å).29b The two Ru-N bond distances of 2.126(2) and 2.127(2) Å in

5.3 are shorter than the Ru-C (2.296(2) Å) and Ru-B (2.322(3) and 2.323(3) Å) bonds,

even though the ruthenium atom is more evenly positioned relative to the center of the

diazadiborolyl ring (Figure 5.4). The Ru-N and Ru-B bond lengths observed in reported

heteroruthenocenes measured between 2.165(2) – 2.195(5) Å and 2.278(2) – 2.449(2) Å,

respectively, and are comparable to those in 5.3. The Ru-C bond distances of the Cp*

ligand ranging from 2.154(2) to 2.194(2) Å are within the expected range for

cyclopentadienyl-based ruthenium complexes137 and somewhat shorter than the Ru-C

bond of the diazadiborolyl ligand.

122

As expected, the structures of the CB2N2 rings in 5.2 and 5.3 resemble closely

those observed in the rhodocenes 5.1 (Figure 5.2): the rings are nearly planar with the

sum of the ring angles of 538.8 and 539.4° for the iron and 539.4° for the ruthenium

derivatives. The same envelope conformation of the ring skeleton was observed in 5.2

and 5.3, and the folding angles formed between the B2N2 and BCB planes measured

7.9(2) and 11.1(2)° for 5.2 and 8.4(2)° for 5.3. The intraannular bond lengths and angles

of the CB2N2 ring are practically equivalent in 5.2 and 5.3. The ring B-C bonds ranging

from 1.511(4) to 1.522(4) Å are shorter than the exocyclic B-C bonds (1.568(4) –

1.578(4) Å), in agreement with the multiple bonding character of the former. The B-N

bond distances of 5.2 (1.472(3) – 1.489(4) Å) and 5.3 (1.479(3) and 1.487(3) Å) are

comparable to the B-N bond distances in 5.1e, and so are the N-N bonds (1.436(3) and

1.439(3) Å for 5.2; 1.431(3) Å for 5.3). The B-N bonds are slightly longer than those in

borazines (1.42 – 1.44 Å),71 and similar to those measured in bis(1,2-azaborolyl)iron

complexes (1.45 – 1.50 Å).136 The N-N bonds are closer in value to the single N-N bond

in hydrazine (1.45 Å),72 and somewhat longer than the value observed in the η5- pyrazolato (1.395(5) Å)40 and η5-diazaphospholide (1.395(3) Å)49 ruthenium complexes.

The intraannular B-C bonds are comparable to the shortest B-C bonds observed in the

− 134, ferrocenes and ruthenocenes containing the diborolyl C3B2 (1.538(3) – 1.590(3) Å),

135 − 136 − 24, 29 azaborolyl C3BN (1.53(1) Å) and thiaborolyl C3BS (1.523(4) – 1.579(3) Å) ligands.

The orientation of the ring substituents with respect to the ligand plane is very similar in 5.2 and 5.3. The cyclic propylene group on the nitrogen atoms lies above the

123

ring plane forming angles of 7.0 – 10.6° between the N-C bonds and CB2N2 plane. In both structures, the methyl substituent is slightly more tilted out of plane (∠ C-C / CB2N2 of 2 – 6°) relative to 5.1e. The phenyl substituents deviate by 2 – 8° from the CB2N2 plane and are twisted by 50.3(1) and 37.6(8)° with respect to the ligand plane.

(a)

E (V) vs SCE

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

(b)

E (V) vs SCE

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

Figure 5.8. Cyclic voltammograms of 5.2 (a) and 5.3 (b) recorded in 0.1 M nBu4PF6 /

DCM at 25 °C with a platinum working electrode and potential reported versus SCE at a scan rate = 50 mV/s.

124

The electrochemical behavior of the 1,2-diaza-3,5-diborolyl iron and ruthenium

complexes was examined by cyclic voltammetry. The cyclic voltammograms of 5.2 and

5.3 recorded in CH2Cl2 (Figure 5.8) exhibited one reversible oxidation to their radical cations at E1/2 = -0.030 and 0.54 V vs. SCE (-0.49 and 0.080 V vs. Fc), respectively. The expansion of the electrochemical window in THF displayed the same reversible oxidation step (-0.050 V in 5.2 and 0.60 V in 5.3 vs. SCE) and no further redox process were observed. The reversible process is attributed to the oxidation of the metal atoms. The oxidation potential of 5.2 and 5.3 is comparable to those in Cp*2Fe (-0.13 V vs. SCE in

138 139 CH2Cl2) and Cp*2Ru (0.55 V vs. SCE in CH2Cl2) complexes (Table 5.1). The results of the electrochemical studies showed that the electron-donating ability of the 1,2- diaza-3,5-diborolyl ligands to the Fe and Ru atoms is similar to the Cp* ligand.

However, the oxidation potential of 5.2 and 5.3 to the cyclopentadienyl analogs containing the same substituent pattern is not available for a direct comparison.

Table 5.1. Reversible oxidation potentials of 5.2 and 5.3 and cyclopentadienyl analogs of

iron and ruthenium complexes (vs. SCE).a

Compound E1/2 (V) in CH2Cl2 E1/2 (V) in THF

Cp*2Fe -0.13 - 5.2 -0.03 -0.05

Cp2Fe +0.46 +0.56

Cp*2Ru +0.55 - +0.42b

- Cp*RuCp +0.54b

5.3 +0.54 +0.60

a All data measured at room temperature, supporting electrolyte [nBu4N][PF6]. b Reference 139c; supporting electrolyte [nBu4N][ClO4].

125

Not taking into account the electronic influence of the ligand substituents on the metal,

the results indicated that by replacing the carbon atoms of the Cp ligand with two

isoelectronic B-N fragments, the heterocyclic CB2N2 ligand definitely has an electronic

impact on the central metal atoms. This is supported by the electrochemical studies of

the 1,2-azaborolyliron complexes reported by Fu and coworkers, which showed that the

1,2-azaborolyl is more electron-rich than the isostructural Cp ligands.25 In contrast to the

5 diheteroferrocenes (η -R4C4E)2Fe (E = group 15 elements) that display a one-electron

reduction of the heterocyclic π ligands (E1/2 ranging from -2.004 to -2.352 V vs. SCE in

DME),140 no reduction of the 1,2-diaza-3,5-diborolyl ligand was detected for 5.2 and 5.3.

5.4. Synthesis, Spectroscopic Characterization and the X-ray Structure of the 1,2-

Diaza-3,5-diborolylzirconium dichloropentamethylcyclopentadienyl Complex (5.4).

Metallocene of the group 4 metals have drawn a tremendous interest to the field

of organometallic chemistry owing to their catalytic applications in olefin

polymerization. The synthesis of zirconocenes containing heterocyclic Cp ligands has

been examined extensively and studies have shown that they possesses comparable

catalytic activity to their analogs featuring carbon-based ligands.35, 141 The synthesis of a

novel zirconium sandwich complex containing a substituted 1,2-diaza-3,5-diborolyl

ligand was investigated. Utilizing the same synthetic approach as described in Scheme

5.3, the lithium salt 2.4e was treated with one equivalent of ZrCp*Cl3 in THF at -78 °C,

affording a thick yellow-orange residue. The NMR analysis of the crude sample revealed

a product mixture containing the desired product 5.4 and the protonated ligand 2.3e,

along with a small quantity of unidentified products. The separation of the protonated

126

ligand from the product mixture was done by hexane extraction of the crude sample,

however, the desired zirconium complex 5.4 was inseparable from the remaining

impurities. As a result of the tedious purification process, a low yield of isolated

zirconium complex was obtained (15 – 20%). No catalytic studies using 5.4 could be

performed and efforts towards optimizing the synthesis are ongoing. Fortunately,

colorless crystals of 5.4 could be isolated by slow evaporation of a solution of the impure

product in a hexane and benzene mixture at ambient temperature.

Scheme 5.3.

N Ph N B Ph Li B B + Cp*ZrCl Ph Me N 3 Cl Me Zr N THF Cl B - LiCl Ph

2.4e 5.4

The characterization of 5.4 by multinuclear (1H, 11B and 13C) NMR and MS

analysis provided conclusive evidence for the identity of the compound. The resonances

associated with the organic substituents of the diazadiborolyl ligand and

pentamethylcyclopentadienyl group were apparent in the expected region in the 1H and

13C NMR spectra of 5.4. The set of distinctive multiplet signals corresponding to the inequivalent methylene protons of (CH2)3N2 fragments were observed at δ 1.42 (-CH2),

3.22 and 3.76 ppm (-NCH2) with the intensity ratio of 2:2:2, respectively. In contrast to

the iron, ruthenium and rhodium analogs, the broad 13C and 11B NMR signals of the ring

carbon (δ 108.5 ppm) and boron (δ 42.6 ppm) atoms were downfield-shifted in the 13C

and 11B NMR spectra of 5.4 in comparison to the former metal complexes. A signal

127

+ associated with the molecular ion [Cp*ZrLCl2] at m/z = 568 was detected in the mass spectrum of 5.4. Due to the difficulties related to the isolation of a pure sample of 5.4 no acceptable elemental analysis was obtained, nevertheless HRMS analysis confirmed the identity of the compound.

Figure 5.9. Molecular structure of 5.4 with 50% probability level thermal ellipsoids and all hydrogen atoms omitted for clarity.

The structure of 5.4 determined by X-ray crystallography unambiguously confirms the identity of the zirconium complex and the relevant structural bond lengths and angles are summarized in Table 5.8 (Appendix One). The molecular structure of 5.4 exhibited the archetypal bent-sandwich structure with the zirconium atom η5-coordinated by the 1,2-diaza-3,5-diborolyl and pentamethylcyclopentadienyl ligands, together with two chloride ions (Figure 5.9). The dihedral angle of 41.5(1)° created between the mean planes of the CB2N2 and the Cp* rings of 5.4 is smaller than those determined in the

128

5 141, analogous zirconocene dichlorides: 52° in (η -H3C3BPhNEt)ZrCp*Cl2 and Cp2ZrCl2,

142 5 5 5 143 and 57° in Me2Si(η -H3C3BPhNEt)(η -C5H4)ZrCl2 and Me2Si(η -C5H4)2ZrCl2. The

Zr atom is positioned closer to the diazadiborolyl ring than to the Cp* ring (Zr−CB2N2 plane of 2.156(3) Å vs. Zr−Cp* plane of 2.231(3) Å). As illustrated in Figure 5.10, the

CB2N2 ring is severely folded with the boron atoms situated further away from the Zr

atom. Consequently, the Zr-B bond distances of 2.715(3) and 2.757(3) Å are considerably

longer than the Zr-C (2.597(3) Å) and Zr-N bonds (2.357(2) and 2.372(2) Å). The two

short Zr-N bond distances in 5.4 are most likely associated to the stronger bonding

interaction between the electron-rich nitrogen and the electron-deficient zirconium atoms.

In other heterozirconocenes, the Zr-C and Zr-B bonds range from 2.402(2) to 2.525(2) Å

and from 2.681(2) to 2.802(3) Å, respectively, being somewhat shorter than those

measured in 5.4. However, longer Zr-N bonds of 2.525(1) – 2.531(1) Å were observed in

these compounds.34, 141 The Zr-C bond distances corresponding to the Cp* ligand

measure between 2.524(3) and 2.561(3) Å and along with the Zr-Cl bonds (2.423(1) and

2.430(1) Å), they fall in the expected range for zirconocene dichloride complexes.142, 143

Figure 5.10. Side view of 5.4 displaying the folding of the CB2N2 ring along the

B(1)⋅⋅⋅B(2) axis and the perpendicular projection onto the best plane of the CB2N2 ring.

129

In contrast to the ligand conformations in 5.1 – 5.3, the diazadiborolyl ligand

exhibited a more severe folding along the transannular B(1)···B(2) axis with an angle of

27.4° and hence the planarity of the ring was poor with the sum of the pentagon angles of

532.3°. Regardless of the large folding angle of the ring skeleton, the intraannular bond

distances and angles in 5.4 do not differ by much from those observed in complexes 5.1 –

5.3. The B-C (1.509(4) and 1.517(4) Å) and B-N bonds (1.467(4) and 1.489(4) Å) are in

the range of the bond distances of previously described complexes, while N-N bond of

1.467(3) Å is slightly longer than those in 5.1e, 5.2, and 5.3. The extraannular B-C bonds

(1.569(4) and 1.572(4) Å) are markedly longer than the intraannular ones supporting the

multiple bonding character and the delocalization of the π-electron over the ligand

framework. The endocyclic B-C and B-N bonds observed in the 1,2-azaborolylzirconium

complex with 1.510(2) and 1.482(2) Å, respectively, are comparable to the corresponding

bond distances in 5.4.

In complex 5.4, the cyclic propylene substituent of the nitrogen atoms is pushed

away from the metal center, resulting in a dihedral angle of 30.8(1)° between the B(1)-

N(1)-N(2)-B(2) and C(3)-N(1)-N(2)-C(5) planes. The drastic tilting of the ring nitrogen

substituent away from the Zr center is probably due to the steric interaction with the

chloride substituent. The torsion angle C(3)-N(1)-N(1)-C(5) is 0°, indicating that the two

N-C bonds are evenly situated above the ring plane at an angle of 22°. Similar to 5.1e –

5.3, the methyl substituent deviates slightly from the ring plane (∠ C-C / CB2N2 =

2.8(1)°) whereas the phenyl groups are 15.3(1) and 18.4(1)° above the mean plane of the

CB2N2 ring.

130

5.5. Synthesis and Spectroscopic Characterization of the 1,2-diaza-4-oxa-3,5-

diborolidines (5.6) and the X-ray Structure of 5.6c.

Transition metal carbonyl complexes containing heterocyclic analogs of the

cyclopentadienyl ligand have been prepared and characterized.144 Pursuing the

coordination chemistry of transition metal carbonyl complexes containing the 1,2-diaza-

3,5-diborolyl ligands, the lithium salt of 2.4e was treated with an equivalent amount of

tungsten hexacarbonyl in THF. This immediately produced an orange mixture of the

complexes 5.5 and 5.6e, instead of the desired 1,2-diaza-3,5-diborolyltungsten complex illustrated in Scheme 5.4.

Scheme 5.4.

R N Me + W(CO)6 R N B X B Me Me Li THF-d8 W Me - CO CO Li OC CO Ph BB Ph NN R R Me O CO C Ph Ph R = Ph (2.4c) B B + W(CO)6 OC C cyclo-(CH ) (2.4e) W + 2 3 Li(thf) NN THF OC CO n CO R R

5.5 R = Ph (5.6c) R = cyclo-(CH2)3 (5.6e)

Compounds 5.5 and 5.6e can be easily separated by washing the orange residue with hexane. The lithium tungsten complex 5.5 was isolated as a yellow solid after filtration and dried under vacuum (75% yield), while the 1,2-diaza-4-oxa-3,5-diborolidine 5.6e contained in the orange filtrate was obtained as a creamy tan solid (56% yield). 5.5 is an

131

air stable compound that is soluble in THF. Yellow crystals of 5.5 were obtained from a

THF solution at -35 °C in the presence of crown ether to aid the crystallization process.

Unfortunately, the crystals were not of sufficient quality to allow for complete refinement

of the structure nevertheless, the disordered X-ray structure of 5.5 supported the proposed tungsten complex. The 1H and 13C NMR spectra of 5.5 revealed a single resonance at δ

1.65 ppm corresponding to the methyl protons of the alkynyl group on tungsten and three

1 distinctive carbon signals at δ 7.08 (WC≡CCH3), 92.1 ( JWC = 93.6 Hz, WC≡CCH3) and

2 102.6 ppm ( JWC = 23.9 Hz, WC≡CCH3) for the alkynyl group. The results of the NMR

characterization and the good elemental analysis obtained for 5.5, along with inspection

of the disordered structure clearly confirmed the identity of the compound. The air and

moisture-sensitive 5.6e is readily soluble in organic solvents but only sparingly soluble in hexane and pentane. Colorless thin needles of 5.6e were obtained by various crystallization methods but they were too small for X-ray diffraction study. The identity of compound 5.6e was established using multinuclear (1H, 11B and 13C) NMR and MS analysis. The NMR spectra of 5.6e exhibited the distinctive signals corresponding to the phenyl and propylene substituents, as well as a broad 11B resonance at 26.3 ppm.

Additionally, the mass spectrum of 5.6e exhibited the molecular ion [M]+ signal at m/z =

262 with 100% intensity.

The reaction of 2.4c with W(CO)6 in THF quickly produced an orange-yellow solution of the expected lithium tungsten complex and the 1,2,3,5-tetraphenyl-1,2-diaza-

4-oxa-3,5-diborolidine (5.6c), which was isolated in 67% yield. As anticipated, the phenyl substituents improved the crystallization properties of the ligand and hence

132 colorless crystals of 5.6c suitable for structural analysis were obtained by recrystallization of 5.6c from a mixture of THF and hexane at -35 °C. The NMR analysis of 5.6c validates the proposed structure. The 1H and 13C NMR spectra of 5.6c confirmed the disappearance of the resonances corresponding to the CCH3 fragment and the resonances corresponding to the phenyl substituents appeared as a multiplet (δ 7.04 – 7.62 ppm). A

11 broad B signal (δB 29.2 ppm) was also observed at a higher field compared to the anionic ligand 2.4c. The molecular ion [M]+ was detected in the mass spectrum of 5.6c at m/z = 374 (100%) and satisfactory results were obtained for the HRMS analysis.

Figure 5.11. Molecular structure of 5.6c with 50 % probability level thermal ellipsoids and all hydrogen atoms omitted for clarity.

A single-crystal X-ray diffraction analysis was performed for 5.6c and the selected bond lengths and angles are summarized in Table 5.9 (Appendix One). The molecular structure of 5.6c (Figure 5.11) exhibits a C2 symmetry with the axis bisecting

133

the N(1’)-N(1) bond and the O(1) atom. The five-membered ring is practically planar with the sum of the intraannular ring angles of 539.7°. The B-O and B-N bond distances

of 1.396(1) and 1.414(2) Å, respectively, are shorter than the N-N bond of 1.447(2) Å.

Heterocycles with a OB2N2 framework have been synthesized by reaction of 1,2,3,4-

diazadiboretidine with amine oxides, through the insertion of an O atom into the B-B

bond, however no structural characterization of such ring system has been reported.145a

Nevertheless, a few heterodiborolanes consisting of the C2B2E frameworks (E = N, O, S)

were structurally characterized and revealed a nonplanar ring structure due to the steric

interaction between the substituents.145b The B-O bonds ranging from 1.393(4) to

1.400(2) Å in these oxadiborolanes are similar to those observed in 5.6c. The short B(1)-

N(1) bond falls in the expected bonding range observed in borazines (1.42 – 144 Å)71 and

in 2.3c (1.412(3) and 1.416(3) Å). The skeletal oxygen displays a bent geometry with the

B(1’)-O(1)-B(1) angle of 108.7(1)°, slightly smaller than those observed in

oxadiborolanes (111.2(2)°).145b Similar to 2.3c, the phenyl groups on the boron atom lie nearly in the ring plane (∠ B-C / OB2N2 = 1.8(1)°) and the planes of the phenyl rings are

twisted by 22.6(1)° from the ligand plane. The phenyl substituents on the nitrogen atoms

are situated above and below the ligand plane (∠ N-C / OB2N2 = ± 9.3(1)°) with the

torsion angle C(2’)-N(1’)-N(1)-C(1) measures 23.1(1)°.

In order to continue with the investigation of the unprecedented reactivity of the

1,2-diaza-3,5-diborolyl ligands towards metal carbonyl complexes, other derivatives of

the anionic ligand (2.4 – 2.6) were utilized in NMR studies with various group 6 – 8

metal carbonyls. In most cases, the NMR data were supportive of the formation of the

134

− + 1,2-diaza-4-oxa-3,5-diborolidines and [(CO)nM(C≡CMe)] Li complexes. To date, the

mechanism for the synthesis of complexes 5.5 and 5.6 is not well understood. The

carbonylation of organoboranes by reaction of trialkylboranes with carbon monoxide and

aldehydes resulting in the formation of boron heterocycles has been reported.146 A similar behaviour was observed in the synthesis of a nickelocene containing the 1-oxa-2,6- diborocyclohexenyl ligand through insertion of CO from Ni(CO)4 into the five-membered

147 1,3-diborolene forming a six-membered C3B2O ring.

Scheme 5.5.

Mechanism A

M(CO)5 M(CO) M(CO) C 5 5 Li O Li Me Li Me C C Li Me O O Me O R' R' R' R' CO C B B BB R' OC C BB R' B B R' + R' M NN NN NN NN OC CO CO R R R R R R R R 2.4a - c, e I II 5.5 5.6

Mechanism B

(OC)5M C O M(CO)5 M(CO)5 (OC)5M Li O Li Li Me Li Me C Me C C O O R' R' R' R' R' BB R' BB B B B B Me R' R' NN NN NN NN R R R R R R R R 2.4a - c, e III IV V

On the basis of the reported synthesis involving the CO insertion of the 1,3- diborolene, two possible mechanisms for the synthesis of the 1,2-diaza-4-oxa-3,5- diborolidines and the unusual pentacarbonyl tungsten propynyl lithium complex via the formation of a C≡C bond is outlined in Scheme 5.5. The proposed mechanism A involved an insertion of the CO from the metal carbonyl into the C-B bond of the anionic

135

ligand, forming a strained four-membered ring intermediate (I). This is followed by a

substitution of the C−CH3 fragment by the oxygen atom of the inserted CO molecule and

formation of the intermediate seven-membered ring system (II). The alternative

mechanism B involves the formation of a typical Fisher carbene intermediate (III), which

rearranges with formation of a borate intermediate (V) and subsequently yields the

reaction products 5.5 and 5.6. It is worth noting that the treatment of the protonated ligands (2.3a – 2.3e) with the transition metal carbonyls showed no signs of reactions even with the heating of the reaction mixtures above 100 °C for days. Hence, the presence of an anionic ligand appears to be a critical contributor for the initiation of the

CO activation process in the synthesis.

136

5.6. Synthesis, Spectroscopic Characterization and X-ray Structures of 1,5-Diaza-

2,4,6,8-tetraborolidine (5.7), its Mono and Dipotassium Salts (5.8 and 5.9) and the

Triple-decker Ruthenocene (6.10) with C2B4N2 Ring.

Within the scope of our investigation on heterocyclic cyclopentadienyl analogs, the synthesis of a novel heterobicyclic 1,5-diaza-2,4,6,8-tetraborolidine (5.7) containing two fused five-membered rings with a C2B4N2 framework was attempted. The synthesis of 5.7 involves the condensation of 1,1-bis(phenylchloroboryl)ethane (2.1b) with hydrazine in the presence of triethylamine, as depicted in Scheme 5.6. Hydrazine was prepared in situ by reaction of two equivalents of KHMDS with hydrazine dihydrochloride in THF and the mixture was refluxed until all the hydrazine salt dissolved. The desired product was obtained as a colorless solid containing a mixture of the cis and trans-isomers in 52% yield, which had good solubility in aprotic solvents but was only sparingly soluble in pentane.

Scheme 5.6.

Ph Ph K B B + KHMDS Me N Me THF H N B B Ph Ph Ph Ph + H N-NH 5.8 Me H 2 2 B B Me Ph Ph + 4 NEt3 N Me 2 B B KHMDS THF / hexane H N H Cl Cl B B THF Ph Ph 2.1b 5.7 Ph Ph K K cis- & trans-isomers B B + 2 KHMDS N Me Me THF N B B Ph Ph 5.9

137

The NMR characterization of 5.7 revealed two distinct sets of resonances

corresponding to the methine and methyl protons of the CHMe fragments for the two

isomers (Figure 5.13a). The signals attributed to the phenyl substituents are superimposed

into a multiplet in the 1H and 13C NMR spectra. However, only one boron resonance was

detected in the 11B NMR spectrum of the isomer mixture as a broad singlet at 53.9 ppm.

The trans-isomer of 5.7 could be separated by crystallization and colorless crystals were

obtained by cooling a solution of 5.7 in a mixture of pentane and THF to -35 °C.

The crystallographic determination of trans-5.7 was performed and the relevant

bond lengths and angles are given in Table 5.10 (Appendix One). The molecular

structure of trans-5.7 confirmed the fused bicyclic structure with C2B4N2 framework

(Figure 5.12). The bicycle is nearly planar with the sum of the pentagon angles of 539.1°

and the two fused rings are related by an inversion center. Similar to the monocyclic analogs, the ring skeleton is also slightly folded along the B(1)⋅⋅⋅B(2) axis by an angle of

10.1(1)° as illustrated in Figure 5.12 (bottom). The ring structure features a long N(1)-

N(1’) bond of 1.480(2) Å, slightly longer than the N-N bonds observed in hydrazine

(1.45 Å)72 and the protonated monocyclic analogs (1.452(3) Å in 2.3c and 1.457(3) Å in

2.3d). The intraannular B-C (1.571(2) and 1.574(2) Å) and B-N bonds (1.436(2) and

1.437(2) Å) are equal and slightly longer, respectively, than the corresponding bonds in

2.3c. The exocyclic B-C bonds with 1.560(2) and 1.565(2) Å are closer in value to the endocyclic B-C bond, as expected for single bond character. The methyl groups on the ring carbon atom are orientated in a trans-configuration. Like in the protonated ring 2.3c,

138 the phenyl substituents deviate slightly from the ligand plane with angles of 4.4 and 6.7° between the exocyclic B-C bonds and the mean plane of the C2B4N2 ligand.

Figure 5.12. Top: molecular structure of trans-isomer of 5.7 with 50% probability thermal ellipsoids and hydrogen atoms of the organic substituents omitted for clarity.

Bottom: side view of 5.7 illustrating the folding of the bicyclic framework along the

B(1)⋅⋅⋅B(2) axis. For clarity, only the ipso carbons of the phenyl groups are represented.

139

(a) BPh Ph Ph B B Me N Me HCMe

H N H B B HCMe Ph Ph

cis- & trans-isomers HCMe HCMe

7.5 6.5 5.5 4.5 3.5 2.5 1.5 0.5 BPh ppm (b) Ph Ph K(thf) B B n Me N Me H N thf B B thf Ph Ph

7.5 6.5 5.5 4.5 3.5 2.5 1.5 0.5 ppm (c) BPh Ph Ph K(thf) n B B K(thf)n N Me Me N thf B B thf Ph Ph

7.5 6.5 5.5 4.5 3.5 2.5 1.5 0.5 ppm

Figure 5.13. Representative 1H NMR studies of the stepwise deprotonation of the

1 heterobicyclic C2B4N2 compound. From top to bottom: H NMR spectra of 5.7, 5.8 and

5.9 recorded in THF-d8 at 25 °C.

The deprotonation of 5.7 could be conducted in two successive steps utilizing the

appropriate metalating reagents, as shown in Scheme 5.6. The treatment of 5.7 with one

equivalent of K[N(SiMe3)2] in THF produced a bright yellow solution of the

monoanionic ligand 5.8 and the addition of another equivalent of K[N(SiMe3)2] to the reaction mixture immediately yielded a color change from yellow to a bright orange for

140

the solution of the dipotassium salt 5.9. Compounds 5.8 and 5.9 were isolated as yellow and orange solids, respectively in high yields (97% and 82%) with excellent solubility in

THF. As expected, both isomers produced the same mono and dianion salts, but it is worth noting that the purity of the sample is crucial for the stepwise deprotonation of 5.7.

The characterization of the mono and dipotassium salts using multinuclear NMR

confirmed the successful stepwise deprotonation of 5.7. The disappearance of the

methine protons was evident in the 1H NMR spectra of 5.8 and 5.9 (Figure 5.13) and, as

expected, the deprotonation resulted in a dramatic downfield shift of the broad 13C signal of the skeletal carbon atom (δ 108.2 ppm in 5.8 and 101.5 ppm in 5.9) with respect to the protonated ligand (δ 31.8 ppm for trans-5.7 and 34.3 ppm for cis-5.7). An upfield shift of the broad 11B NMR signal (δ 41.3 ppm in 5.8 and 39.8 ppm in 5.9) with respect to the

protonated ligand was also observed in the 11B NMR spectra. Orange prismatic crystals

of 5.9(tmeda)2 were grown from a solution of 5.9 in TMEDA, but crystals of 5.8 suitable

for X-ray crystallography were not obtained.

The molecular structure of 5.9(tmeda)2 was determined and the selected bond

lengths and angles are given in Table 5.11 (Appendix One). The crystal structure of

5.9(tmeda)2 also displays a planar C2B4N2 ligand coordinated by two potassium ions with a TMEDA molecule completing the coordination environment of each metal center, in a centrosymmetric arrangement (Figure 5.14). The planarity of the bicyclic skeleton is better than that in 5.7 with a folding angle of 6.7(2)° between the B2C and B4N2 planes

(sum of the pentagon angles of 539.5°). The deprotonation of 5.7 to the dianion salt

5.9(tmeda)2 resulted in a shortening of the intraannular N-N (1.442(3) Å) and B-C bonds

141

142

(1.494(4) and 1.512(4) Å) and a slight lengthening of the B-N bonds (1.474(4) and

1.482(4) Å), resembling the transition from cyclopentadiene to cyclopentadienyl. The

two exocyclic B-C bonds (1.585(4) Å) are longer than the endocyclic ones. The

potassium ions are positioned above and below the center of the N-N bond of the bicyclic

structure with K-N distances of 2.820(2) and 2.889(2) Å. These distances are comparable

to those observed in the potassium pyrrolyl148 and indolyl149 complexes (2.80 – 3.06 Å) featuring π interactions between the potassium ion and ligand plane. The K-B distances

(3.268(3) and 3.357(3) Å) fall within the range observed for potassium boratabenzene and diboratabenzene complexes (3.06 – 3.69 Å)89 and the potassium salts of 2.6b and

2.6c (3.21 – 3.45 Å). The K-C distances of 3.514(3) and 3.691(3) Å are longer than those observed in the polymeric potassium cyclopentadienyls (2.95 to 3.30 Å)76b, 82c, 82g, 88 and the dimeric potassium pentalenediyl (2.86 – 3.23 Å)150 complexes. The separation

between the two potassium ions of 5.525(1) Å is comparable to the K···K separation

observed in polymeric cyclopentadienyls (5.52-5.85 Å)76b, 82c, 82g, 88 and the dimeric

potassium pentalenediyl derivative (5.43, 5.48 Å).150 In contrast to 5.7, the methyl

substituent on the ring carbon atom is very close to the mean plane of the bicyclic ring,

with a deviation of 1.9° out of this plane. The phenyl substituents are orientated above

and below the ligand plane with angles of 8.5 and 10.0° formed between the exocyclic B-

C bonds and mean plane.

Owing to the easily accessible mono and dianionic ligands, the potential for 5.8

and 5.9 to function as a bridging π ligand in the synthesis of multi-decker sandwich

compounds is a promising area of research.64a, 151 Reaction of 5.9 with a stoichiometric

143

amount of [Cp*RuCl]4 in THF produced a dark brown solution of the triple-decker

sandwich 5.10 (Scheme 5.7). The product was obtained as a brown solid and dark brown

prismatic crystals were obtained (25% yield) by slow evaporation of a hexane solution of

5.10 at ambient temperature.

Scheme 5.7.

Ph Ph Ru K B B K Ph Ph N + 1/2 [Cp*RuCl]4 B N B Me Me Me Me N THF B N B B B Ph -2 KCl Ph Ph Ph Ru 5.9

5.10

The triple-decker sandwich 5.10 was fully characterized by multinuclear NMR,

MS and elemental analysis. The 1H and 13C NMR spectra of 5.10 feature only one set of resonances for the methyl and phenyl substituents of the middle deck as well as the Cp* groups, which are equivalent on the NMR time scale. A downfield shift of ca. 0.5 ppm was detected for the single resonance of the methyl protons on the ring carbon atom with respect to the dipotassium salt 5.9. Likewise, only one broad signal was observed for the equivalent skeletal boron atoms at δ 24.1 ppm. The molecular ion peak of the triple- decker ruthenocene was detected as an intense signal at m/z = 907 in the mass spectrum, and its identity and purity were further confirmed by the HRMS and elemental analysis.

144

Figure 5.15. Top: molecular structure of 5.10 with 50% probability level thermal ellipsoids. For clarity, only the ipso carbons of the phenyl groups are represented and all hydrogen atoms omitted. Bottom: side and top views of 5.10. For clarity, only the metal and the ring atoms of the middle deck are represented.

An X-ray diffraction study on a singe crystal of 5.10 was conducted and the pertinent crystallographic and metric parameters are summarized in Table 5.3 and 5.12

(Appendix One). The crystal structure of 5.10 validated the triple-decker sandwich structure with the ruthenium centers coordinated by a π bridging middle deck and two

145

Cp* ligands, as illustrated in Figure 5.15. The formation of 5.10 was achieved through

the formal insertion of the ruthenium atoms into the N-N bond, resulting in the

conversion of the 8π-electron bicyclic ligand into an eight-membered ring. A related

transformation has been reported for the two-electron oxidation of the pseudo-triple-

2+ decker complex Cp2Ru2(μ-cyclo-C8H8) to the dicationic species Cp2Ru2(μ-cat-C8H8) through the insertion of the Ru atoms into a C-C bond of the cyclooctadiene ligand.152

The geometry of the C2B4N2 ring can be best described as a severely elongated hexagon

with nearly linear B-N-B moieties (B-N-B angles of 166.4(6) and 172.1(6)°) and it is

rather unusual for eight-membered rings. The framework of the eight-membered C2B4N2 ligand is relatively nonplanar because of the asymmetric ligation to the ruthenium atoms, as shown in Figure 5.15 (bottom left). The intraannular B-C (1.542(10) – 1.582(9) Å) and B-N bonds (1.400(9) – 1.462(9) Å) of the N2B4C2 ring are longer and shorter,

respectively, than the values observed in dipotassium salt 5.9. The cleavage of the N-N

bond of 5.9, resulted in a significant increase of the B-C-B angles from 104.7(2)° in 5.9

to 1.238(6) and 130.1(6)° in 5.10, as well as a long N(1)⋅⋅⋅N(2) distance of 2.548(7) Å.

The extraannular B-C bonds ranging from 1.578(9) to 1.608(9) Å are longer than the

intraannular B-C bonds.

The structure of the triple-decker sandwich is asymmetric, with Ru(1) situated at

comparable distances from all B atoms (Ru-B bonds ranging 2.488(7) - 2.546(7) Å). The

Ru(2) atom is clearly η5-coordinated to one side of the eight-membered ring featuring

shorter Ru-B bonds (2.351(7) and 2.378(6) Å). The ruthenium centers are situated much

closer to the nitrogen atoms, as shown in Figure 5.15 (bottom right), and the Ru-N bonds

146

associated to the Ru(1) and Ru(2) atoms are nearly equivalent (Ru(1)-N = 2.118(5) and

2.128(5) Å and Ru(2)-N = 2.129(5) and 2.132(6) Å). The coordination mode of Ru(1)

can be described as η7 or η8, depending on the involvement of C(2) in bonding. By

comparison the Ru(1)-C(2) bond of 2.548(6) Å is markedly longer than the Ru(1)-C(1)

(2.402(5) Å) and Ru(2)-C(2) (2.257(6) Å) bonds, as well as those observed in

+ 153 154 [(Cp*Ru)2Cp*] (2.19 – 2.23 Å) and (Cp*Ru)2H5C5BCH3 (2.23 – 2.29 Å). The Ru-

C bond distances involving the Cp* ligands measure between 2.160(8) and 2.243(6) Å and are comparable to those observed in other triple-decker ruthenocenes. The metal atoms are situated closer to the middle deck (Ru(1)-N2B4C plane = 1.578(4) Å and Ru(2)-

N2B2C plane = 1.592(4) Å) than to the Cp* rings (Ru(1)-Cp* = 1.862(4) Å and Ru(2)-

Cp* = 1.816(4) Å). Consequently, the Ru atoms are relatively close to each other with a

Ru⋅⋅⋅Ru distance of 3.249(1) Å. For comparison, the separation between the metals

+ 153 154 measures 3.35 Å in [(Cp*Ru)2P5] , 3.48 Å in (Cp*Ru)2H5C5BCH3, 3.68 Å in

+ 155 + 156 157 [(Cp*Ru)2Cp*] , 3.71 Å in [(Cp*Ru)2Cp] , and 3.90 Å in (Cp*Ru)2B8H14.

The computational analysis performed on the model structure 5.10a, which features a hydrogen-substituted C2B4N2H6 ligand and Cp groups, revealed the molecular

geometries that are consistent with the X-ray diffraction data. The frontier molecular

orbital (MO) analysis of 5.10a shows a substantial mixing of the ligand orbitals with the

MOs of the [Cp*Ru]+ fragments takes place, resulting in a formal insertion of the metal

into the N-N bond. The MO analysis conducted for 5.10a indicates that the ruthenium d

orbitals interact mainly with the two nitrogen p orbitals that are oriented parallel and

perpendicular to the C2B4N2 plane. The primary thermodynamic driving force for the

formation of 5.10 is the cleavage of the N-N bond, and the subsequent formation of the

147

Ru-N interactions. This is confirmed by the energy decomposition analysis conducted

for 5.10a and the total orbital interactions between the formally cationic and anionic fragments of the complex is strong, with a calculated value of ca. -1900 kJmol-1 (total calculated bonding energy is -2400 kJ mol-1). A similar oxidative addition of hydrazines

to transition metals has recently been reported.158 The η5 and η8-coordination

descriptions of Ru(2) and Ru(1) atoms, respectively, were also supported by the Mayer

bond orders obtained for 5.10a; the largest bond orders (around 0.45) are calculated for

Ru-N interactions. On the basis of the computational study, the atypical linear B-N-B moieties in 5.10 can be described as formally sp-hybridized nitrogen atoms with the unhybridized p orbitals participating mainly in the metal-ligand bonding.

The exploration of the electrochemistry of 5.10 was performed by cyclic voltammetry. In the cyclic voltammogram of 5.10 recorded in THF (Figure 5.16), a reversible one-electron oxidation at 0.42 V and two reversible one-electron reduction

steps at -1.30 and -1.99 V vs. SCE (-0.14, -1.86 and -2.55 V vs. ferrocene) were observed.

The expansion of the electrochemical window using DME confirmed this behaviour and

allowed for the observation of an additional, irreversible oxidation at 1.10 V. The

electrochemical range of CH2Cl2 enabled the observation of only one reduction and one

oxidation wave at -1.49 and 0.18 V. The one-electron transfer steps are well separated,

indicative of significant electron delocalization over the framework. The reversible

oxidation is most likely associated with the Ru2+/Ru3+ process while the irreversible

electron transfer could be attributed to the oxidation of the Cp* ligand, as observed for

139 Cp*2Ru (0.55 and 1.25 V in CH2Cl2), (Cp*Ru)2C8H6 (0.11 and 0.40 V in THF, -0.02

159 and 0.48 V in CH2Cl2), and (Cp*Ru)2B8H14 (0.17 and 0.93 V in 8.5:1.5

148

157 CH2Cl2/toluene). The reduction processes of 5.10 show no equivalent in the chemistry

of related species containing all-carbon and all-boron ligands. The two reversible

2- reduction waves are attributed to processes centered on the C2B4N2 ligand. A two-

electron reduction in essence could transform the 8π-electron ring into a 10-π-electron

4- B4N2C2 ligand, formally aromatic according to Hückel.

E (V) vs SCE

-3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

Figure 5.16. Cyclic voltammogram of 5.10 in 0.1 M nBu4PF6 / THF at 25 °C recorded at

platinum working electrode versus SCE, scan rate = 500 mV/s.

The chemical reduction and oxidation of 5.10 was examined by NMR studies.

The chemical oxidation and reduction of 5.10 with two equivalents of AgOTf and

K[Naphthalene], respectively, in THF quickly yielded dark red solutions of the dication

and dianion complexes. The NMR spectral data of the oxidation and reduction process

exhibited characteristic chemical shifts for the resonances corresponding to the methyl

and phenyl substituents of the middle deck ligand in 5.10. The NMR analysis gave

149

promising results for the formation of the stable oxidized and reduced products of 5.10.

However, due to the low yielding synthesis of 5.10, the preparation of the [5.10]2+ and

[5.10]2- complexes has not yet been accomplished.

5.7. Conclusion

The synthesis of rhodium (5.1b and 5.1e), iron (5.2), ruthenium (5.3) and

zirconium (5.4) sandwich complexes containing the 1,2-diaza-3,5-diborolyl ligands was

achieved by salt metathesis reactions. The NMR spectroscopic characterization of these

transition metal metallocenes provided conclusive information for the identity of the

proposed sandwich structures in solution. The structural investigation of these metal complexes revealed typical sandwich structures that possess structural features similar to those reported for heterocyclic analogs of cyclopentadienyl. In all complexes, the metal ions are η5-coordinated by the 1,2-diaza-3,5-diborolyl ligand and the geometry of the

complexes is fairly symmetric depending on the identity of the metals. The electron-

deficient zirconium atom binds closer to the electron-rich nitrogen atoms, while the

rhodium and ruthenium atoms bind relatively symmetrically over the center of the π

ligand. The heterocyclic ring is nearly planar with the exception of the zirconium

complex, where the ring structure is folded considerably with the skeletal boron atoms

pointing away from the Zr center, resulting in unusually long Zr-B bond distances. The

geometry of the zirconocene is bent with a dihedral angle between the CB2N2 and Cp*

rings of 41.5(1)º, much smaller than those measured for boron-containing

cyclopentadienyl analogs. The iron and ruthenium sandwiches featured a reversible

oxidation of the metal ions M2+/M3+ and the oxidation potentials are comparable to those

150

observed in analogous Cp*2Fe and Cp*2Ru. Thus indicating that the 1,2-diaza-3,5- diborolyl ligands are more electron-rich than the cyclopentadienyl ligand.

The reaction of the heterocyclic 1,2-diaza-3,5-diborolyl ligands towards metal carbonyl complexes resulted in the formation of mixtures of 1,2-diaza-4-oxa-3,5-

diborolidine (5.6) and lithium metal propynyl complexes (5.5b), by replacement of the

CCH3 fragment with an O atom of the carbonyl group. The spectroscopic

characterizations of the isolated products confirmed the identities of the compounds and a

crystal structure of the 1,2-diaza-4-oxa-3,5-diborolidine (5.6c) was determined. The

proposed mechanism for the unprecedented reaction involved an insertion of the CO from

the metal carbonyl into the C-B bond of the anionic ligand, followed by a substitution of

the C-CH3 group for the O atom of the inserted CO molecule.

The synthesis of a novel heterobicyclic 1,5-diaza-2,4,6,8-tetraborolidine (5.7) with C2B4N2 framework was described and the NMR analysis indicated the isolation of a

mixture of the cis and trans-isomeric products, which were easily separated by

crystallization. The structure of 5.7 featured a trans-conformation that resembled the bicyclic pentalene C8H6, with a relatively planar skeletal framework. The deprotonation of 5.7 could be accomplished cleanly in two successive steps using K[N(SiMe3)2] and

yielded the mono and dipotassium salts (5.8 and 5.9), which are excellent precursors for

the synthesis of sandwich complexes. The crystal structure of 5.9(tmeda)2 showed that

the dianionic C2B4N2 ligand is bifacially coordinated to the potassium ions, which are

coordinated by a TMEDA molecule. The reaction of 5.9 with [Cp*RuCl]4 prompted the

cleavage of the N-N bond and yielded the triple-decker ruthenium sandwich complex

5.10, featuring a highly unusual monocyclic middle deck with an elongated-hexagonal

151

B4N2C2 backbone. The ruthenium centers in this derivative are closer to each other than

in any other reported multidecker sandwich complex. The cyclic voltammogram of 5.10 displayed three reversible, well separated one-electron transfer steps indicating efficient electron delocalization over the framework. The properties of the ligand render it a very promising building block for the design of larger sandwich architectures.

152

CHAPTER SIX

Alkali Metal and Rhodium Complexes Incorporating 1,2,4-Triaza-3,5-diborolyl

Ligand and a Tricyclic BN Compound with B4N8 Framework

6.1. Introduction

In light of the rich coordination chemistry displayed by the novel 1,2-diaza-3,5-

diborolyl ligands discussed in the previous chapters, investigation of the coordination

properties of the carbon-free heterocyclic Cp analog, 1,2,4-triaza-3,5-diborolyl with B2N3 framework, is also of great interest. As mentioned in section 1.2.4 of this thesis, the substituted heterocyclic 1,2,4-triaza-3,5-diborolidines, 1.59a – 1.59f, had been reported and a ferrocene analog containing ligands of type C was prepared and characterized, although a crystal structure confirming its identity and structure was not yet obtained.56, 57

The synthesis of 1,2,4-triaza-3,5-diborolidine, 1.63b, precursors to less symmetric anionic rings D, through self-assembly of methylhydrazine with bis(dimethylamino)borane was later reported,58b but the coordination properties of

complexes containing ligand of type D was not well explored.

153

Me Me R'' Me NH NH N N N N R' B B R' R' B B R' R' BB R' Ph B B Ph Ph B B Ph o o NN NN NN NN HN N R R R R R R Me Me 1.59 CD2.4 - 2.6 1.63b

This chapter describes the investigation of the coordination chemistry of the

anionic 1,2,4-triaza-3,5-diborolyl ligand D, a carbon-free cyclopentadienyl analog, towards the alkali and transition metals. The synthesis and structural investigation of the alkali metal salts (lithium 6.1a, sodium 6.1b, and potassium 6.1c) and rhodium complex

(6.2) featuring the 1,2,4-triaza-3,5-diborolyl ligand are presented. Furthermore, the π- coordination properties of the borane-adducts (6.3a – 6.3c), which were synthesized by reaction of triphenylborane with the alkali metal salts 6.1a – 6.1c, were investigated. The spectroscopic characterization of the borane adducts and the solid-state structure of the lithium (6.3a) and potassium (6.3c) adducts will be discussed. The synthesis, spectroscopic characterization, and the X-ray structure of the novel tricyclic dilithiodiborate (6.5) and tetrahydrazidotetraborane (6.6) containing B4N8 frameworks are

also described. In addition, theoretical and electrochemical investigations of the tricyclic

B4N8 compound 6.6 are briefly discussed in this chapter.

154

6.2. Synthesis, Spectroscopic Characterization and X-ray Structures of Lithium

(6.1a), Sodium (6.1b) and Potassium (6.1c) Complexes Incorporating the 1,2,4-

Triaza-3,5-diborolyl Ligand.

The 1,2,4-triaza-3,5-diborolidine 1.63b was prepared according to a literature

58b procedure by treatment of methylhydrazine with PhB(NMe2)2 in hexane and refluxing

the mixture for 72 hours. Compound 1.63b was obtained as an air and moisture-sensitive

colorless solid (75% yield) with good solubility in polar organic solvents and sparingly

soluble in hexane and pentane. Colorless crystals of 1.63b could be easily obtained by

cooling the hexane solution of 1.63b to -35 °C. The 1H and 13C NMR spectra of 1.63b

revealed different signals for the chemically inequivalent methyl and phenyl substituents.

The broad singlet resonance corresponding to the ring proton observed at δ 7.49 ppm,

significantly downfield-shifted with respect to the quartet resonance corresponding to the

proton of the pendant methylamino group (δ 4.14 ppm). The deshielding of the ring

proton could be a result of a ring current effect, indicating a certain degree of aromaticity

of the six-π-electron B2N3 ring. A smaller difference in the chemical shift for the methyl

groups (δ 3.21 ppm at the ring nitrogen vs. δ 2.45 ppm at the pendant nitrogen) is also observed. In the 11B NMR spectrum, the signals corresponding to the two distinct boron

nuclei overlapped into a broad singlet at δ 27.5 ppm.

Deprotonation of 1.63b occurred easily in THF in the presence of various metalating agents, as depicted in Scheme 6.1. The reactions of 1.63b with equimolar amount of LiTMP and NaHMDS cleanly yielding the 1,2,4-triaza-3,5-diborolyl lithium and sodium salts and were isolated as colorless powders (6.1a in 92.1% yield and 6.1b in

92.4% yield, respectively). The treatment of 1.63b with KH quantitatively afforded the

155

potassium salt (6.1c in 98% yield) and hydrogen gas. These alkali metal salts are readily

soluble in THF and insoluble in hydrocarbons. Colorless crystals of 6.1a, 6.1b and 6.1c

suitable for X-ray diffractometry were obtained by slow diffusion of hexane into their

THF solutions. The double deprotonation attempted at the ring and the pendant

methylamino group using two equivalents of the same metalating reagents led to

decomposition of the B2N3 ring with formation of unidentified products.

Scheme 6.1.

Me H Me H N N Ph M Me H hexane N + LM N Ph Ph Ph Ph 2 Me B Me + 2 NN B B B B N N refluxed 3 days - LH H H NN NN Me Me -2 HNMe 2 H Me Me

1.63b M = Li, L = TMP M = Li (6.1a) Na, L = HMDS Na (6.1b) K, L = H K (6.1c)

The multinuclear (1H, 11B, 13C and 7Li) NMR spectra of 6.1a – 6.1c were very

similar, with only small chemical shift differences observed for the resonances of the

methyl and phenyl substituents. The disappearance of the broad signal corresponding to

the ring proton in the 1H NMR spectra was indicative of the successful deprotonation of

1.63b. A slight upfield shift was detected for the quartet resonance corresponding to the proton on the pendant methylamino group (δ 3.75 – 3.87 ppm in 6.1a – 6.1c vs. δ 4.14 ppm in 1.63b), as illustrated in Figure 6.1. Two distinct broad 11B NMR resonances were

observed for the inequivalent boron nuclei of the lithium and sodium salts, while only

one broad resonance was found for the potassium salt. The boron resonances for these

alkali metal salts ranging from δ 24.2 to 27.8 ppm are comparable to the resonance for the protonated ligand (δ 27.5 ppm). Hence, the formation of the anionic ligand did not

156

seem to have a significant impact on the electronic effect of the boron atoms. The 7Li

NMR spectrum of the lithium salt 6.1a recorded in THF-d8, featured a sharp resonance at

δ -1.24 ppm, indicative of the lithium ion-solvent pair Li(thf)n, similar signals were

detected for the diaza-diborolyl lithium salts (δ -2.34 ppm in 2.4a; -2.80 ppm in 2.4b; -

2.16 ppm in 2.4e). The mass spectrum of 6.1a displayed signals corresponding to the

protonated ligand [L]+ and the fragment [L – Me]+.

(a) Me H N

BC H N 6 5 Ph B B Ph NN H Me

8.0 7.0 6.0 5.0 4.0 3.0 2.0 ppm (b) Me H N Na N Ph Ph BC6H5 B B NN Me

8.0 7.0 6.0 5.0 4.0 3.0 2.0 ppm

Figure 6.1. (a) The 1H NMR spectrum of the 1,2,4-triaza-3,5-diborolidine 1.63b and (b)

1,2,4-triaza-3,5-diborolyl sodium salt 6.1b in THF-d8 at 25 °C.

Single-crystal X-ray diffraction analyses have been performed for compounds

6.1a, 6.1b and 6.1c. The relevant structural parameters are summarized in Table 6.1 and

selected bond lengths and bond angles are provided in Table 6.3 – 6.5 (Appendix One).

The solid-state structures of 6.1a and 6.1b are isomorphous, featuring a 2D polymeric

157

framework (Figure 6.2) and will be discussed together. In the crystal, each B2N3 ligand coordinates to three alkali metals in a σ fashion and each alkali metal is also coordinated by three nitrogen atoms of the different ligands. A dimeric fragment is formed by the bridging of two alkali metals through the substituent-free ring nitrogen N(2) atoms of two ligands. The resulting M2N2 rhombs feature relatively short M-N distances (2.013(5) and

2.068(5) Å for M = Li and 2.350(3), 2.416(3) Å for M = Na) and the M2N2 planes are

found nearly perpendicular to the planes of the B2N3 rings (82.8° for M = Li and 88.9° for

M = Na). The distorted pyramidal coordination environment of each metal is completed

by an exocyclic nitrogen atom N(4) from a different ligand, with the sum of the angles at

the metal center of 353.8° for M = Li and 336.6° for M = Na. The M-N(4) distances

involving the pendant methylamino group (2.100(5) Å for M = Li and 2.449(3) Å for M

= Na) are only slightly longer than the M-N(2) bonds. The 2D structures of 6.1a and 6.1b

were formed by the connection of the dimeric units through the M-N interactions

involving the pendant nitrogen, N(4).

The structures of the alkali metal salts of the isoelectronic pyrrolyl, indolyl and

carbazolyl provide a good structural comparison for complexes 6.1a and 6.1b. The solid-

state structures of pyrrolyl lithium, sodium and potassium crystallized with the aid of

crown ethers are very similar to each other, with the pyrrolyl ligand σ-coordinated to the

alkali metal that is coordinated by the crown ether.148 The M-N distances measure

1.950(4) and 2.319 Å for Li and Na respectively, being slightly shorter than the shortest

M-N distances in 6.1a and 6.1b. The lithium salt of the sterically more demanding 2,3- ditert-butylpyrrole exhibits a monomeric σ-bonded structure with a Li-N distance of

1.932(7) Å and two THF molecules coordinated to the lithium ion in a trigonal planar

158

Figure 6.2. Fragments of the 2D polymeric structures of 6.1a (top) and 6.1b (bottom), illustrating the environment of the ligand, with 50% probability level thermal ellipsoids.

For clarity, only the ipso carbon atoms of the phenyl groups are represented and all hydrogen atoms on the organic substituents omitted.

159

arrangement.160 The same type of dimeric structures were observed for the lithium and sodium indolyl with the M-N distances measuring 2.004(7) and 2.231(10) Å for M = Li and 2.356(5) – 2.481(5) Å for M = Na, and with tertiary amines completing the tetrahedral coordination environment of the metals.161 The solid-state structure of the

lithium carbazolyl features σ-bonded dimers very similar to those of 6.1a, and Li-N distances measuring 2.012(2) – 2.165(2) Å.162 An array of σ-bonded carbazolyl sodium

salts containing various ethers has been structurally characterized, with Na-N distances

163 situated between 2.30 and 2.53 Å. Some of these structures feature [Na(carbazolyl)]2 dimers similar to those observed in 6.1b, and the Na-N distances for these compounds were measured between 2.40 and 2.34 Å. It can therefore be concluded that the coordination behavior of the 1,2,4-triaza-3,5-diborolyl ligand towards alkali metals is very similar to that of its organic analogs pyrrolyl, indolyl, and carbazolyl. These ligands coordinate to lithium and sodium in a σ fashion as well, and dimeric structures containing

M2N2 rhombs with M-N distances very similar in length were observed in some of their

solid-state structures. An η5-coordinating ligand was observed in the 1D polymeric

structure of the electron-rich tetramethylpyrrolyl sodium salt with Na-N distances ranging

between 2.631(2) and 2.666(2) Å. The polymeric structure also featured shorter Na-N σ- interactions (2.351(1) and 2.411(2) Å).85

Lewis base adducts of the lithium salts of 2,4,6-triorganoborazines, which

contained σ-interactions between lithium and nitrogen, have been structurally

164 characterized. Although dimers formed through Li2N2 bridges, similar to those observed in 6.1a, were present only in three out of nine structures, in six of the examples

160

the deprotonated nitrogen in borazine is connected to two lithium ions, resulting in a

distorted tetrahedral environment of the metal. In all derivatives the Li-N distances

ranged between 1.96 and 2.16 Å. Upon deprotonation, an elongation of the borazine

skeleton was observed in these compounds, as a result of the reduction in the width of the

B-N--B and the opposite N-B-N angles. This was explained through the greater

electronic repulsions between the B-N bonds and the σ lone pair electrons on nitrogen. A

similar effect is observed upon deprotonation of the 1,2,4-triaza-3,5-diborolidine, with the N(1)-N(2)-B(2) angle contracting from 110° in the neutral ligand 1.63b to 105° in the

alkali metal salts 6.1a – 6.1c.

161

The potassium salt 6.1c features a 1D polymeric structure in the solid state, with

the ligand coordinating in both σ and π fashions (Figure 6.3). The potassium ion is σ-

coordinated by the amide nitrogen, N(2), and the nitrogen of the pendant methylamino

group, N(4), of the successive ligand, with the K-N distances measuring 2.681(3) and

2.889(3) Å, respectively, leading to the formation of linear chains. In addition, the potassium ion is also coordinated by the ipso carbon, C(3), of a phenyl group in the ligand, with a K-C contact distance of 3.127 Å. Two such linear chains are connected by

the η3 inter-chain interactions involving the metal and the B(2)-N(2)-N(1) fragment in the

B2N3 ring. The K-B(2) distance of 3.149(3) Å is noticeably shorter than those observed

in the η2-coordinated diazadiborolyl potassium salts (3.211(3) and 3.315(3) Å in

2.6b(thf) and 3.364(4) and 3.425(4) Å in 2.6b(thf)2). The π-coordinated K-N(2) and K-

N(1) distances measure 2.800(3) and 3.009(3) Å, respectively, being as expected slightly

longer than the σ-coordinated K-N distances. The coordination environment of the

potassium ion could be described as a three-legged piano-stool. The polymeric double-

stranded arrangement observed in 6.1c is similar to the above-mentioned structure of

tetramethylpyrrolyl sodium,85 and to the structure of 2,3-dimethylindolyl potassium.149

Structures featuring exclusively σ interactions between potassium and the pyrrolyl ring are rather rare, and have only been observed for the pyrrolyl and carbazolyl salts with a crown ether hosting the metal (K-N distance of 2.712(3) and 2.774(2) Å).148, 166a More commonly, the pyrrolyl,164, 165 indolyl149 and carbazolyl166 ligands coordinate to

potassium in a similar fashion to the triazadiborolyl ligand in 6.1c, through σ (K-N

distance of 2.676(3) – 2.826(2) Å) and π interactions (potassium ion to ring plane

162

distance ranging 2.80 – 3.06 Å). By comparison, the distance between the potassium

center and the plane of the B2N3 ring in 6.1c measures 2.79 Å, which is situated at the

lower end of the range observed for the organic analogs.

The results show that the bond lengths and angles of the B2N3 ring skeletons are

very similar in all structures of the alkali metal salts. The ring is practically planar, with

the sum of the pentagon angles of 540.0º for compounds 6.1a and 6.1b and 539.7° for compound 6.1c. The phenyl substituents on the boron atoms show little deviation from the ring plane, with the B-C bonds forming angles of 0.9 – 3.6º with the B2N3 plane. The

methyl and the pendant methylamino substituents are bent slightly out of ring plane, with

the extraannular N-C and N-N bonds forming angles of 6.8 – 7.8º and 1.3 – 6.3º with the plane, respectively. The intra and extraannular N-N bonds are equal in length, ranging between 1.430(3) to 1.440(3) Å, closer in value to the N-N single bond in hydrazine (1.45

Å),72 suggesting the absence of any significant multiple bond character. The two B-N

bonds involving the hydrazine unit that contributes with both nitrogen atoms to the ring

skeleton are equal and relatively short (B(1)-N(1) = 1.400(4) – 1.411(4) Å and B(2)-N(2)

= 1.400(4) – 1.407(4) Å) in comparison to the other two skeletal B-N bonds (B(1)-N(3) =

1.443(4) – 1.454(3) Å and B(2)-N(3) = 1.446(4) – 1.470(4) Å). The B-N bond distances

in borazines are ranging between 1.42 to 1.44 Å,71 which are comparable to the values

found in these 1,2,4-triaza-3,5-diborolyl complexes.

Me H Me H Me H N M N M N M N N N Ph B B Ph Ph B B Ph Ph B B Ph NN NN NN Me Me Me 1.63b E F

163

The alternating bond lengths in the ring framework suggest that the cyclopentadiene-like resonance structure E has a major contribution to the structure of the salts 6.1a – 6.1c, as opposed to the resonance structure F. The extraannular B-C and N-C bonds of 1.56 –

1.59 Å and 1.44 – 1.47 Å, respectively, are in the expected single-bond ranges.

6.3. Synthesis, Spectroscopic Characterization and the X-ray Structure of a

Dimeric Rhodium (6.2) Complex Incorporating the 1,2,4-Triaza-3,5-diborolyl

Ligand

The transition metal coordination chemistry of the triazadiborolyl ligand was investigated using the alkali metal derivatives described above as starting materials. The reactions between the alkali metal salts 6.1a – 6.1c and various transition metal reagents, such as FeCl2(thf)2, [Cp*RuCl]4, ZrCp*Cl3, and CuBr, were examined by NMR. The

results of the NMR analysis revealed that the major product of most reactions was the

protonated ligand 1.63b, along with a complex mixture of by-products that was not

further investigated. The most likely source of protons necessary for the regeneration of

1.63b is the pendant methylamino group of the ligand itself. It is possible that the

presence of transition metal reagents promotes a disproportionation of the monoanionic salts 6.1a – 6.1c to the protonated ring and the dianionic analogs as depicted in Scheme

6.2. As described earlier, the dianionic salt of 1.63b is not isolable and the driving force

for the process could be the formation of the transition metal methylhydrazides.

Nonetheless, the treatment of 6.1a with [Rh(cod)Cl]2 in THF immediately

afforded a bright red solution. Upon cooling the concentrated solution to -35 ºC for 48

hours, the yellow crystalline rhodium complex 6.2 precipitated out. Complex 6.2 was

164

isolated as an air and moisture-sensitive, yellow crystalline solid in 42% yield. It is only

sparingly soluble in dichloromethane and THF. Thin yellow crystals suitable for X-ray

diffraction were obtained by cooling a concentrated solution of 6.2 in a mixture of hexane

and THF to -35 ºC for two days.

Scheme 6.2.

Me H Me H Me N N N M Transition M' N metal N N Ph Ph Ph Ph Ph Ph B B complex B B B B Decomposition + 1.63b + products NN NN NN Me H Me M' Me

M = Li, Na, K (6.1a - 6.1c) 1.63b M' = transition metals

Me Ph N B N Rh H N N B Me Ph

BPh BPh CDHCl2

9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 ppm m-Ph 1J = 13.0 Hz o-Ph RhC p-Ph

85 84 77 76

129 128

135 125 115 105 95 85 75 65 55 45 35 25 ppm

1 13 Figure 6.4. H and C NMR spectra of 6.2 recorded in CD2Cl2 at 25 °C.

165

The characterization of 6.2 was performed using multinuclear (1H, 11B and 13C)

NMR, MS and elemental analysis. The 1H and 13C NMR spectra of 6.2 revealed

resonances associated to the Rh(cod) fragments and triazadiborolyl ligands (L) in a 1:1

ratio. Signals for two sets of inequivalent cyclooctatetraene vinyl and methylene groups,

and the phenyl substituents were observed in the 1H and 13C NMR spectra, as shown in

Figure 6.4. Although two distinct sets of resonance signals were present for the phenyl

groups, only one broad resonance was observed for the two inequivalent boron nuclei at δ

30.6 ppm in the 11B NMR spectrum. The electron impact mass spectrum of 6.2 features a

parent ion corresponding to the formulation [Rh(cod)L] (m/z = 474) and the fragments

[RhL – 2H]+ (m/z = 364) and [L + H]+ (m/z = 264).

Figure 6.5. Molecular structure of 6.2 with 50% probability level thermal ellipsoids. For

clarity, only the ipso carbon atoms of the phenyl groups are represented and all hydrogen

atoms on the organic substituents omitted.

166

The crystal structure of 6.2 (Figure 6.5) revealed a centrosymmetric dimeric

structure [Rh(cod)L]2, with two B2N3 ligands bridging two Rh(cod) fragments through

the amide nitrogen N(2), similar to the dimeric fragments in 6.1a and 6.1b. The plane of

the Rh2N2 rhomb is oriented nearly perpendicular onto the planes of the planar B2N3 rings

(Σpentagon angles of 539.9º) forming a dihedral angle of 88.4°. The only noticeable

changes in the metric parameters of the triazadiborolyl ligand are the slight lengthening

of the N(1)-N(2) and B(2)-N(2) bonds (1.447(3) and 1.460(4) Å, respectively) and the

shortening of the B(2)-N(3) bond (1.419(4) Å). This indicates a change in the electron

distribution along the ring backbone, with the ligand being better described by the

resonance structure F with the negative charge localized on the amide nitrogen N(2). The

Rh-N bond lengths of 2.141(2) and 2.168(2) Å fall in the range observed for rhodium

cyclooctadiene complexes containing 2-aminomethylpyrrolidines (2.112(3) – 2.178(3)

Å)167 and 2-aminomethylpyrrolyl (2.034(3) – 2.231(2) Å).168 The Rh-N-Rh and N-Rh-N

angles measure 98.33(9) and 81.67(9)°, respectively. The cyclooctadiene ligand is oriented with the C=C bonds coordinating to the metal in trans to the amide ring nitrogen in a pseudo-square planar geometry, and the Rh-C distances measure 2.115(3) – 2.168(2)

Å, typical for cyclooctadienerhodium complexes. This type of geometry is common for

167 molecules containing the Rh(cod)N2 moiety, such as 2-aminomethylpyrrolidines and 2-

aminomethylpyrrolyl.168 It is worth mentioning that, although transition metal complexes

featuring π-coordinating pyrazolyl-type ligands are not uncommon, no such derivatives

of rhodium have been reported and hence the σ-bonded complex 6.2 is not an exception.

167

6.4. Synthesis and Spectroscopic Characterization of the Alkali Metal

Triphenylborane Adducts (6.3) and the X-ray Structures of the Lithium (6.3a) and

Potassium Adducts (6.3c).

The above results illustrated that the coordination chemistry of the triazadiborolyl ligand towards alkali metals and rhodium is dominated by the σ interactions involving

the amide ring nitrogen, closely resembling the chemistry of the organic analogs of

pyrrole, indole and carbazole. Aiming to enforce the π-coordination activities of the

ligand, the alkali metal salts 6.1a – 6.1c were treated with triphenylborane to block the σ

bonding site at ring nitrogen as illustrated in Scheme 6.3. The η5-coordination mode of

pyrrolyl-borane adducts was accomplished in lithium8b and iron complexes.166, 169

Scheme 6.3.

Me H Me H N M N M N N + Ph B Ph B B Ph 3 Ph B B Ph NN THF NN

Me Ph3B Me

M = Li (6.1a) M = Li (6.3a) Na (6.1b) Na (6.3b) K (6.1c) K (6.3c)

The formation of the 1:1 adducts of triazadiborolyl-triphenylborane 6.3a – 6.3c was

immediately observed by a color change of the reaction mixture from a bright to pale

yellow solution. The desired compounds were isolated as colorless powders in high

yields (93% in 6.3a, 86% in 6.3b and 83% in 6.3c). These adducts are only sparingly

soluble in diethyl ether and acetonitrile, but readily soluble in THF. Colorless crystals of

6.3a and 6.3b were obtained by cooling a concentrated solution of 6.3a in a mixture of acetonitrile and hexane to -35 °C and by slow evaporation of a THF and Et2O solution of

168

6.3c at ambient temperature. The triazadiborolyl-borane adducts were characterized by multinuclear NMR spectroscopy. The 11B NMR spectra of 6.3a – 6.3c exhibited a singlet

resonance characteristic for a borate ion at δ -4 – -7 ppm and a singlet resonance corresponding to the two inequivalent ring boron atoms overlapping into a broad signal at

δ 25 – 27 ppm. The 1H and 13C NMR spectra of the adducts showed no significant

differences from the corresponding spectra of the alkali metal salts 6.1a – c, except for

the additional set of signals associated to the BPh3 moiety.

Structural determination for the alkali metal adducts 6.3a and 6.3c have been

carried out using X-ray crystallography and the selected structural parameters of these

adducts are summarized in Tables 6.7 and 6.8 (Appendix One). The metric parameters of

the planar B2N3 ring revealed only marginal variations upon the coordination of the

triphenylborane. The intraannular B(2)-N(3) bond of ca. 1.45 Å is marginally longer than the other B-N bonds (1.41 – 1.43 Å) and the N(1)-N(2) bond of 1.45 Å is comparable to the single N-N bond in hydrazine (1.45 Å).72 As expected for a

tetracoordinated boron, the extraannular B(3)-N(2) bond of 1.579(3) in 6.3a and 1.576(2)

Å in 6.3c, is significantly longer than the intraannular B-N bonds. The extraannular B-C

(1.57 – 1.58 Å) and N-C bonds (1.45 – 1.66 Å) are typical for the single bond character.

The ring substituents are oriented slightly out of plane, with the B-C, N-N and N-B bonds

forming angles of 3 – 6º, 1 - 3º and 2 – 6º with the B2N3 plane, respectively. In the

lithium and potassium complexes, the deviation of the methyl substituent out of the ring

plane (∠ N-C / B2N3 = 14.6º for 6.3a and 7.5º for 6.3c) is more pronounced.

169

Figure 6.6. Molecular structure of 6.3a(CH3CN)3 with 50% probability level thermal

ellipsoids. Hydrogen atoms of the organic substituents omitted for clarity.

Figure 6.7. Fragment of the 1D polymeric structure of 6.3c with 50% probability level

thermal ellipsoids. Hydrogen atoms of the organic substituents omitted for clarity

170

The crystal structure of 6.3a reveals a monomeric structure featuring the BPh3 unit coordinated to the amide ring nitrogen, N(2) of the 1,2,4-triaza-3,5-diborolyl ligand, and the lithium ion σ-coordinated by the amino nitrogen N(4) of the pendant methylamino group (Figure 6.6). The distorted tetrahedral coordination environment of the lithium ion is completed by three acetonitrile molecules and no π interactions between the metal and the cyclic ligands are present. The Li-N distances fall in a narrow range, between 1.979(6) and 2.072(6) Å, and are comparable to the corresponding distances observed in 6.1a and other isoelectronic pyrrolyl, indolyl and carbazolyl lithium complexes.

The crystal structure of 6.3c features a polymeric chain containing potassium ions coordinated by the triphenylboryltriazadiborolyl ligands, as shown in Figure 6.7. As observed in 6.3a, the triphenylborane molecule is coordinated by the amide nitrogen of the cyclic ligand and the resulting B(3)-N(2) bond is relatively long. The potassium ion is σ-coordinated by the pendant amino group (K-N distance 2.833(2) Å) and π- coordinated by a phenyl substituent of the ligand and two other phenyl groups of the triphenylborane moiety. The distances between potassium and the phenyl planes measure

2.84 Å for the phenyl of B2N3 ring and 2.91 and 2.99 Å for the phenyl groups of

triphenylborane, with the shortest K···C contacts involving the ipso (3.04 Å), ortho (2.99

Å) and ipso (3.11 Å) carbon atoms, respectively. Such π-coordination by aryl groups is

typical for potassium borate complexes, e.g. in K[BPh4], which is used for the selective

precipitation of this ion from aqueous solutions.170 The distances measured between

171 potassium ion and the aryl planes in potassium borate complexes, K[BPh4] (2.98 Å)

171

172 and K[B(C6H4OC6H5)4] (2.92 – 3.02 Å), are comparable to those observed in 6.3c.

The B-C bonds of the tetracoordinated borane fragment measure 1.640(3) – 1.659(3) Å and C-B-C angles of 104.3(1) – 113.6(1)º are comparable to those in 6.3a (1.634(4) –

1.652(3) Å and 104.9(2) – 113.2(2)º) and other borate complexes (1.637(2) – 1.645(2) Å and 102.1(3) – 112.9(4)º).171

Although the solid-state structures of 6.3a and 6.3c show no π interaction between the metal ions and the triazadiborolyl ligand, the coordination chemistry of these adducts with transition metal complexes was investigated by NMR. Based on the complex NMR spectra obtained, the treatment of the transition metal reagents such as FeCl2(thf)2,

[Cp*RuCl]4, ZrCp*Cl3, and CuBr with 6.3a, once again resulted in the formation of a

mixture of by-products and a small quantity of the neutral ligand 1.63b. The attempted

separation of the product mixtures by solvent extraction and crystallization was

unsuccessful.

6.5. Synthesis, Spectroscopic Characterization and the X-ray Structures of the

Tricyclic Dilithiodiborate (6.5) and Tetrahydrazidotetraborane (6.6) Complexes

with B4N8 Framework.

The deprotonation of 1.63b with KH followed by methylation with MeI resulted in a clean replacement of the more acidic ring proton with a methyl group, yielding the neutral 1,2,4-triaza-3,5-diborolidine ligand 6.4 as depicted in Scheme 6.4. Compound 6.4 was obtained as a colorless powder in 88% yield and had excellent solubility in organic solvents. The 1H NMR spectrum of 6.4 indicated the disappearance of the signal

corresponding to the ring proton in 1.63b at δ 7.48 ppm and the resonance for the

172

equivalent methyl protons on the ring nitrogen atoms was observed at δ 2.88 ppm. This

clearly indicated the regiochemistry of the reaction. The signals corresponding to the

pendant methylamino group were represented as a doublet and quartet at δ 2.32 and 3.76

ppm, respectively (Figure 6.8a). As a result of the methylation, only one set of resonance

signals attributed to the equivalent phenyl groups on ring boron were observed in the 1H and 13C NMR spectra. A broad boron signal at δ 28.6 ppm for the equivalent boron

centers was also detected in the 11B NMR spectrum.

Scheme 6.4.

Li(thf) Me H Me H Me N N + KH Me N Ph Me N N N B N Ph Ph Ph + MeI Ph Ph + LiTMP N B B B B B N B N B THF THF Ph N Me NN NN Me Ph N H Me Me Me Li(thf) Me 1.63b 6.4 FeCl2(thf)2 6.5 THF

Me Ph Me B N N Fe N B + FePh2(thf)2 + Me N N Me Ph-Ph B N N N B Me Ph Me 6.6

The lithium amide salt was produced through deprotonation of the exocyclic nitrogen in 6.4 with LiTMP, and subsequent spontaneous dimerization through N-B bond formation in THF, yielding the dilithium salt of a tricyclic borate dianion 6.5. The dilithium salt was isolated as a colorless powder in 60% yield and thin colorless needles were obtained by recrystallization of 6.5 from a THF solution at -35 °C.

173

(a) Me H N N Ph Ph B B NN BPh Me Me

7.5 6.5 5.5 4.5 3.5 2.5 1.5 ppm -B-Ph (b) Li(thf) Me Me N Ph Me N B N Ph N B N B N B BPh, -BPh Ph N Me Me Ph N B-Ph Li(thf) Me 2040 0 -20 ppm

7.5 6.5 5.5 4.5 3.5 2.5 1.5 ppm (c) Me Ph Me B N N BPh N B Me N N Me B N N N B Me Ph Me

7.5 6.5 5.5 4.5 3.5 2.5 1.5 ppm

1 1 11 Figure 6.8. (a) H NMR spectrum of 6.4 recorded in C6D6 at 25 °C. (b) H and B NMR

1 spectra of 6.5 recorded in THF-d8 at 25 °C. (c) H NMR spectrum of 6.6 recorded in

THF-d8 at 25 °C.

The spectroscopic characterization of the dilithium salt was performed using

multinuclear (1H, 11B, 13C and 7Li) NMR analysis. The 1H and 13C NMR spectra

confirmed the presence of two inequivalent phenyl groups and three inequivalent methyl

groups, while the 11B NMR spectrum featured signals corresponding to the borane and

the borate moieties at δ 28.5 and 2.7 ppm, respectively (Figure 6.8b). The NMR results

174

7 are in agreement with a tricyclic B4N8 structure in solution. The chemical shift of the Li

+ NMR signal observed at δ -2.18 ppm is characteristic for Li(thf)n , suggesting the

presence of solvent-separated ion pairs in the THF solution.74

The reaction of the tricyclic dilithiodiborate complex 6.5 with various transition

metal reagents afforded relatively complicated NMR spectra. The attempt to purify and isolate the products failed. The reaction of 6.5 with FeCl2(thf)2 in THF promptly yielded

a dark mixture that could not be characterized by NMR, most likely due to the formation

of paramagnetic by-products. However, cooling of the concentrated dark mixture to -35

ºC for several days resulted in the precipitation of colorless crystals. The crystals were

separated from the supernatant, washed with cold pentane and isolated in low but reproducible yield (22%).

The characterization of the colorless crystals using multinuclear (1H, 11B and 13C)

NMR, mass spectrometry and single-crystal X-ray diffraction revealed a neutral tricyclic

1 13 tetrahydrazidotetraborane (6.6) with a B4N8 skeletal framework. The H and C NMR

spectra of 6.6 exhibited the resonances expected for one phenyl group and three inequivalent methyl groups, downfield shifted with respect to the signals observed for the dilithiodiborate complex (Figure 6.8c). In the 11B NMR spectrum, the signals

corresponding to the two inequivalent boron centers merged into a broad singlet at δ 25.1

ppm. The electron-impact mass spectrum of 6.6 featured a signal that was assigned to the

molecular ion [M]+ (m/z = 400). In addition, the HRMS analysis provided a conclusive

proof for the identity of the compound.

The other probable product of the reaction is FePh2(thf)2 that subsequently

decomposed to the black precipitate of metallic iron and biphenyl. The latter product was

175

identified in the reaction mixture using GC-MS analysis (Scheme 6.4). A similar

173 decomposition pathway was reported for FePh2(PEt3)2 at temperatures above 0 °C. The

electrochemical oxidation of tetraphenylborate to boric acid and biphenyl in aqueous

conditions at a potential of 0.216 V vs. SCE has also been reported,174 however, the

treatment of Li[BPh4] with FeCl2(thf)2 failed to produce BPh3. Hence, the formation of

the stable polycycle appears to be a critical thermodynamic contributor to the oxidation of the tricyclic dilithiodiborate 6.5 to the tetrahydrazidotetraborane 6.6 and biphenyl. In an attempt to obtain a better yield for 6.6, the dilithiodiborate salt was reacted with a stoichiometric quantity of I2 in THF, affording a bright red solution. The NMR data of the isolated solid showed no formation of the neutral tricyclic compound 6.6.

Crystallographic determinations of compounds 6.5 and 6.6 were performed and

the relevant structural parameters are provided in Tables 6.9 and 6.10 (Appendix One).

The crystal structure of 6.5 revealed a centrosymmetric structure containing a fused

tricyclic B4N8 skeleton framework comprised of two five-membered B2N3 rings and a 6-

membered B2N4 ring, as illustrated in Figure 6.9 (top). As suggested by the NMR

spectra, the skeleton backbone contains two borane and two borate moieties. The

tetracoordinated boron, B(2) is situated 0.6 Å outside the plane formed by the other

skeletal atoms, resulting in an envelope conformation of the five-membered rings

(folding angle along the N(1)···N(3) axis of 30º) and a chair conformation of the six-

membered ring (folding angle of 55° along the N(3)···N(4’) axis). The N-N bond of the

six membered ring (N(3)-N(4) = 1.445(2) Å) is marginally shorter than those of the five

membered (N(1)-N(2) = 1.475(2) Å), and comparable to the single N-N bond in

hydrazine (1.45 Å).72 The B-N bonds involving the borane center (B(1)-N(3) = 1.408(2)

176

Å and B(1)-N(2) = 1.441(2) Å) are much shorter than those involving the borate center

(B(2)-N(3) = 1.556(2) Å, B(2)-N(1) = 1.562(2) Å and B(2)-N(4) = 1.566(2) Å).

Figure 6.9. Molecular structure of 6.5 (top) and 6.6 (bottom) with 50% probability ellipsoids and all hydrogen atoms omitted for clarity.

177

The lithium ions are positioned 1.7 Å outside the best plane of the ligand, and are

coordinated by the three nitrogen atoms that are bonded to the borate center. Two short

Li-N distances of 2.035(3) and 2.189(3) Å are similar to the distances observed for the

lithium complexes containing η5-coordinating 1,2-azaborolyl and pyrrolyl, and η3,η4 and

1 η (π)-coordinating 1,2-diaza-3,5-diborolyl 2.4a and 2.4c(thf)3. The coordination sphere

of the lithium ion is completed by an intramolecular contact involving the ipso carbon of

the phenylborate moiety (Li···C distance of 2.775(3) Å) and a THF molecule. The Li-O

distance of 1.901(3) Å in 6.5, is comparable to those found in solvated lithium

cyclopentadienyl derivatives and the diazadiborolyl lithium salt 2.4c(thf)3 (1.947(4) –

2009(3) Å).

The X-ray diffraction analysis of 6.6 revealed two independent molecules that are very similar and will be discussed together. The solid-state structure of 6.6 featured a centrosymmetric, planar polycyclic B4N8 framework with s-indacene-like structure. The

tricyclic skeleton backbone is nearly planar, with the sum of the pentagon and hexagon

angles of 539.5 and 720º for both molecules, and the dihedral angles measured between

the 5- and 6-membered rings of 6.6 is ca. 1º for both molecules. The B4N8 polycycle is

regular, with all B-N and N-N bonds being very similar in lengths (B-N bonds = 1.423(3)

– 1.440(3) Å and N-N bonds = 1.436(2) – 1.441(2) Å). These intraannular bond lengths

are characteristic of a partial multiple bond character for N-B, as observed in borazines

71 (1.42 – 1.44 Å), and a single N-N bond compared to hydrazines. The phenyl

substituents display little deviation from the ring planes (5 – 10°). The methyl

substituents are clearly situated above and below the B4N8 plane in an all-trans

178

arrangement with the C-N bonds forming angles of 22 – 40° with the plane. This results

in the distorted pyramidal geometry around all nitrogen atoms.

The all-carbon analog of 6.6, s-indacene, is an unstable molecule with

antiaromatic character according to Hückel (12-π-electron compound)175 and the

reduction of its more stable substituted derivatives176 to the formally aromatic, 14-π- electron dianion was reported.177 It could therefore be expected that a two-electron

oxidation or reduction of 6.6 would yield a stable dication or dianion, respectively,

satisfying the Hückel condition of aromaticity. In order to investigate this hypothesis,

theoretical calculations were executed for 6.6 as well as for its doubly reduced and

oxidized forms.178 The geometry-optimized structure of 6.6 is in excellent agreement

with the data from the structural determination. The MO analysis confirms that 6.6 is

only formally a 16-π-electron system. Although the tricyclic B4N8 framework is

essentially planar, the geometry around all methyl-substituted nitrogen atoms is

pyramidal, giving rise to MOs with only partial π-like character. The HOMO is π-

antibonding through all four N-N linkages, which readily explains the observed long

bond distances indicative of single bonds.

2- The B4N8 framework is significantly nonplanar in the dianion [6.6] due to occupation of a MO with B-N antibonding character. Hence, a two-electron reduction of

6.6 leads to a structure which does not fulfill the general conditions of aromaticity. On the other hand, the two-electron oxidation of 6.6 yields a dication with a nearly planar structure because two electrons are removed from the N-N antibonding HOMO. The minor geometric distortions present in the system arise from steric interactions, as evidenced by the geometry optimized structure of a hydrogen substituted tricycle

179

2+ [B4N8H8] , displaying a perfectly planar geometry. The calculations showed that the

two possible spin states, the open-shell singlet diradical and the lowest energy triplet

state, are very close in energy with the formally aromatic open-shell singlet state being

ca. 5 kJ mol-1 lower than the triplet state.

2nd oxidation = 0.34 V 0.90 V vs SCE

1st oxidation = -0.21 V 0.35 V vs SCE

E (V) vs Fc

-1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8

[Cp2Co]PF6 = -1.36 V -0.82 V vs SCE 1st oxidation = -0.21 V 0.35 V vs SCE

E (V) vs Fc

-2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4

Figure 6.10. Cyclovoltammograms of 6.6 at 50 mV/s scan rate in the absence of internal

standard [Cp2Co]PF6 (top) and at 200 mV/s scan rate in the presence of [Cp2Co]PF6

0/+1 0 ([Cp2Co] with E = –1.36 V vs. ferrocene and -0.82 V vs. the SCE).

180

The cyclovoltammetric measurements were performed on 6.6 in THF in the range

of –2.5 to 1.5 V, revealing two irreversible oxidation peaks at – 0.21 and 0.34 V vs. Fc

(0.35 and 0.90 V vs. SCE) and no reduction process. It is worth noting that the first oxidation step appears to be partially reversible. The NMR analysis for the chemical oxidation of 6.6 with [Cp2Fe]PF6, which should only be a strong enough oxidizing agent

1 for the first oxidation process of 6.6, in THF-d8 afforded a pale yellow solution. The H

NMR spectrum of the reaction mixture revealed the signal corresponding to Cp2Fe and a

number of broad resonances. A brown solid isolated from the mixture was insoluble in

organic solvents and thus no further characterization can be done. An EPR study on the

chemical oxidation of 6.6 was conducted, but no signal was observed. No radical species

were detected by EPR upon in situ electrochemical oxidation.

181

6.6. Conclusion

The selective deprotonation of the 1-methyl-3,5-diphenyl-4-methylamino-1,2,4-

triaza-3,5-diborole 1.63b at the ring nitrogen, cleanly produced the alkali metal salts

featuring the two-dimensional (lithium, 6.1a, and sodium, 6.1b) and one-dimensional

(potassium, 6.1c) polymeric structures in the solid state. In all complexes, the structures

are dominated by the M-N σ interactions between the 1,2,4-triaza-3,5-diborolyl ligands

and the alkali metals. Interestingly, the polymeric structure of the potassium derivative

also features K-N and K-B π interactions involving the metal and the B2N3 ring.

Reaction of the alkali metal salts 6.1a – 6.1c with most transition metal reagents led to

reprotonation of the ligand in a potential disproportionation process that also produced an

intractable mixture of by-products. The treatment of 6.1a with [Rh(cod)Cl]2 successfully

yielded the dimeric rhodium complex 6.2 consisting of Rh(cod) fragments bridged by triazadiborolyl rings σ-coordinating through the amido ring nitrogen atoms.

The σ-coordination site at the amido ring nitrogen can be conveniently blocked

using triphenylborane. The synthesis of the borane adducts 6.3a – 6.3c did however not lead to the induction of any π-coordination between the B2N3 ring and alkali metal ions,

as illustrated by the crystal structure of the lithium and potassium derivatives. Instead,

the lithium ion is coordinated by the pendant methylamino group of the ligand and three

acetonitrile molecules in 6.3a(CH3CN)3, while the potassium ion is coordinated by the

pendant methylamino group of the ligand and three phenyl groups in 6.3c. The B2N3 ring is planar in all compounds described herein, and its metric parameters indicate that the

182

bonding arrangements in the ligand are most accurately described by the resonance

structures E and F.

The substitution of the amido ring proton with a methyl group produced a more

symmetrical 1,2,4-triaza-3,5-diborolidine 6.4. Following the deprotonation of the

pendant methylamino group in 6.4, a tricyclic dilithioborate complex 6.5, with a B4N8 framework was generated through the assembly of the two lithium amide complexes.

The oxidation of 6.5 with FeCl2(thf)2 resulted in the formation of the neutral tricycle tetrahydrazidotetraborane 6.6. The crystal structure of 6.6 revealed a planar structure with a tricyclic B4N8 ring backbone consisting of one 6- and two 5-membered rings,

similar to the organic analog s-indacene. Although the tricyclic B4N8 framework is

essentially planar, the geometry around all methyl-substituted nitrogen atoms is

pyramidal. This is indicative of the localization of the π-electrons on the nitrogen atoms,

which is supported by the theoretical results. The electrochemical study of 6.6 also

showed two irreversible oxidation peaks and no reduction process.

183

CHAPTER SEVEN

Conclusions and Future Research

7.1 Coordination Chemistry of 1,2-Diaza-3,5-diborolyl Ligands

This thesis focuses primarily on the fundamental investigation of heterocyclic cyclopentadienyl analogs containing CB2N2 and B2N3 frameworks and the chemistry of their metal complexes. The synthesis of a novel class of heterocyclic 1,2-diaza-3,5- diborolyl ligands was accomplished through the ring closing reactions between geminal bis(chloroboranes), MeCH(BRCl)2, and symmetrically substituted hydrazines,

R’HN−NHR’, in the presence of NEt3, followed by deprotonation with appropriate bases.

The coordination properties of these heterocyclic ligands towards alkali, main-group, early and late transition metals were examined. The results of the investigation showed that the variation of the skeletal boron and nitrogen substituents affects the coordination properties of the ligands. The coordinative behaviour of these 1,2-diaza-3,5-diborolyl ligands exhibits many similarities but also remarkable differences in comparison to the chemistry of the all-carbon cyclopentadienyl analogs.

184

A number of alkali metal complexes (2.4 – 2.6) containing 1,2-diaza-3,5-diborolyl

ligands were synthesized and characterized. The NMR analysis of these alkali metal

complexes showed that in solution these compounds exist as solvent-separated ion pairs,

− + [MeC(BR)2(NR’)2] [M(thf)n] . The structural analysis of the 1,2-diaza-3,5-diborolyl

ligands indicates electron delocalization over the NBCBN skeleton. The intraannular B-

C and B-N bonds in these alkali metal salts are intermediate in length between those

corresponding to single and double bonds while the N-N bonds are closer in lengths to

single bonds. The latter observation, in conjunction with the notable deviations of the

ring nitrogen substituents from the ring plane, indicate that the delocalization of the π-

electron over the ring framework does not extend over to the N-N bond. Analogous to cyclopentadienyl, these novel heterocyclic ligands bind face-on to the alkali metal ions, forming monomeric (2.4c(thf)3, 2.5b(thf)3 and 2.5c(thf)3) and polymeric (2.4a, 2.5b,

2.6b(thf), 2.6b(thf)2 and 2.6c(thf)) sandwich complexes. The metal-to-ligand plane distances in these alkali metal complexes are larger than those in the polymeric Cp analogs, illustrating the lower coordinating ability of the 1,2-diaza-3,5-diborolyl ligands.

These ligands exhibit remarkable variability in their coordination behaviour and due to

the difference in the radii of the ring atoms the assignment of the coordination mode is

not immediately obvious. A view of the perpendicular projection onto the CB2N2 planes

provided a good indication for determining the hapticity of the 1,2-diaza-3,5-diborolyl

ligands. The η1 (C), η2 (BC), η3 (NBC), and η4 (BNNB) coordination modes have been

assigned to the alkali metal complexes incorporating these novel π ligands. By

comparison, the cyclopentadienyl ligand displays almost exclusively an η5-coordination

185

in alkali metal sandwich complexes. No trends regarding the preference of the ligand for

a specific coordination mode could be established at this point, but it appears obvious that

the energetic differences between the different hapticity modes are small.

Group 14 metallocenes (3.2 – 3.3) incorporating the 1,2-diaza-3,5-diborolyl ligands were synthesized by the metathesis reaction of the alkali metal salts with the corresponding metal halides, and thoroughly characterized. The crystal structures of the heterogermanocene (3.2) and stannocenes (3.3a and 3.3b) feature typical bent sandwich structures with dihedral angles ranging from 50.1(1) to 53.1(1)º, and are wider than those measured for cyclopentadienyl germanium and tin complexes. Analogous to the alkali metal complexes, the 1,2-diaza-3,5-diborolyl ligands also displayed more diverse coordination behaviour toward the metals than cyclopentadienyl does. In contrast to the

η5-coordination mode of the cyclopentadienyl analogs, the phenyl derivatives 3.2 and

3.3b reveal η3,η3 ligation of the 1,2-diaza-3,5-diborolyl ligands to the metal centers, and

the methyl derivative 3.3a is best described as an η4,η4-coordination mode. In these

structures, the metals avoid contact with the nitrogen atoms and hence the heterocyclic

ligands coordinate preferentially to the metal ions through the skeletal carbon and boron

atoms. A similar observation was observed for the bis(1,2-azaborolyl)tin complex.

The attempt to synthesize a silicocene containing the 1,2-diaza-3,5-diborolyl

ligands led instead to the isolation of the trichloro(diazadiborolylsilyl) complexes (3.1a,

3.1b). The structural analysis of 3.1b revealed a pyramidal SiCl3 moiety that is σ-

coordinated by the 1,2-diisopropyl-1,2-diaza-3,5-diphenyl-3,5-diborolyl ligand. The

multinuclear NMR and MS analyses of these silicon complexes were consistent with the

structural data confirming the identity of these complexes.

186

The tin complexes 3.3a and 3.3b possess similar chemical properties as the

stannocenes featuring carbon-based ligands. Reactions of 3.3 with SnCl2 and

[H(Et2O)2]B(C6F5)4 led to the formation of the neutral 1,2-diaza-3,5-diborolyltin(II)

chloride (3.5) and cationic 1,2-diaza-3,5-diborolyltin borate complexes (3.6a and 3.6b,

respectively). The identity of these tin complexes was confirmed by multinuclear NMR

and MS analysis. Moreover, a tetravalent tin(IV) complex 3.4 was prepared by the

oxidative addition of 3.3a with dichloromethane via the insertion of the C-Cl bond into

the tin center. The spectroscopic and structural characterization of 3.4 provided

conclusive information for the identity of the complex. The crystal structure of 3.4

1 exhibited the expected structure with tetracoordinated tin and the two CB2N2 rings η -

coordinated to the Sn(CH2Cl)Cl moiety.

A series of novel zinc, cadmium, and mercury sandwich complexes of the 1,2-

diaza-3,5-diborolyls (4.1 – 4.3) were synthesized and characterized. Structural

determination of these heterocyclic group 12 complexes showed that the ring carbon

plays a dominant role in the binding of the ligand to the metal and the geometry of these

complexes is fairly symmetric. The molecular structure of four base-free complexes

4.1a, 4.1b, 4.2b and 4.3 features a linear geometry at the metal, with C-M-C angles

ranging from 174 to 180°. For compound 4.2a, the molecule crystallized with one unit of

BrLi(thf)3 that coordinated to the cadmium center and as a result a more acute C-Cd-C angle of 159° was observed. The coordination of the ligand in these group 12 metallocenes is typically η1(π),η1(π). Except for zincocene 4.1a, where the Zn-C bond is

noticeably tilted towards the ligand and the coordination mode of the ligand is better

3 3 described as η ,η . The M-C bond is nearly perpendicular onto the CB2N2 planes in

187

compounds 4.1b, 4.2a · BrLi(thf)3, 4.2b and 4.3 (85.0 – 88.9°), while the C-C bond

involving the ring carbon forms angles of 12.2 to 21.7° with the same ring plane. The σ

character of the η1 bonding in group 12 cyclopentadienyl derivatives is more pronounced

than in these group 12 metallocenes featuring heterocyclic ligands. This could be

explained by the more effective stabilization of a planar configuration for the negatively

charged carbon, through π donation to the electron-deficient boron atoms in 1,2-diaza-

3,5-diborolyl as compared to the stabilization through delocalization over the ring

structure in cyclopentadienyl. The metal-carbon bond lengths in all complexes are

comparable to those observed in the cyclopentadienyl analogs. The multinuclear NMR

and MS analysis of these group 12 metallocenes revealed the presence of the sandwich

structure in solution. This is in agreement with the strong covalent character of the

metal-ligand interaction shown in the solid-state structures.

The coordination chemistry of the 1,2-diaza-3,5-diborolyl ligands with early and

late transition metals was investigated. The rhodium (5.1b and 5.1e), iron (5.2),

ruthenium (5.3) and zirconium (5.4) complexes containing the 1,2-diisopropyl-diaza-3,5- diphenyl-diborolyl and 1,2-cyclopropyl-diaza-3,5-diphenyl-diborolyl ligands were synthesized and characterized. The identity of these complexes was confirmed by multinuclear NMR and MS studies as well as elemental analysis. The solid-state structure of these metal complexes features the sandwich geometry for their cyclopentadienyl analogs. In all complexes, the 1,2-diaza-3,5-diborolyl ligands are η5- coordinated to the metal centers. The displacement of the metal with respect to the center of the CB2N2 ring is dependent on the electronic properties of the metals. For the electron-rich transition metals such as the rhodocenes 5.1b and 5.1e, ferrocene 5.2 and

188 ruthenocene 5.3, the metals are positioned nearly above the center of the ligands. It is worth noting that in 5.2, the Fe atom is slightly shifted towards the N atoms of the ligand.

For the electron-deficient zirconocene 5.4, the Zr atom is noticeably shifted from the central position and is situated closer to the more electron-rich N atoms. The heterocyclic ligand is nearly planar with the exception of the zirconium complex, where the ring structure is folded considerably along the B(1)⋅⋅⋅B(2) axis with the skeletal carbon atom pointing towards the Zr atom resulting in unusually long Zr-B bonds. The cyclovoltammogram of 5.2 and 5.3 revealed one reversible oxidation process, which is associated to the metal center (M2+/M3+), and the oxidation potentials are comparable to those observed in analogous Cp*2Fe and Cp*2Ru.

These novel heterocyclic ligands display an intriguing reactivity towards metal carbonyl complexes. Reactions of the alkali metal salts (2.4a – e) with various metal carbonyls yielded a mixture of the 1,2-diaza-4-oxa-3,5-diborolidine (5.6) and alkynyl metal carbonyl complexes (5.5). Compounds 5.6 and 5.5 are generated though the insertion of the CO group of the metal carbonyl into the C−B bond of the ligand, followed by a substitution of the CCH3 fragment with an O atom of the carbonyl group.

The identity of these compounds was confirmed by the NMR and MS analysis, and the molecular structure of 1,2,3,5-tetraphenyl-1,2-diaza-4-oxa-3,5-diborolidine 5.6c featuring a planar five-membered OB2N2 framework was determined.

In addition to the coordination studies of the 1,2-diaza-3,5-diborolyl ligands, the synthesis of a heterobicyclic 1,5-diaza-2,4,6,8-tetraborolidine (5.7) featuring a planar

C2B4N2 ring framework was described. The NMR characterization of the isolated product revealed a mixture of the cis and trans-isomers that are separable by

189

crystallization. The crystal structure of 5.7 features a trans-conformation with a

relatively planar skeletal framework. Stepwise deprotonation of 5.7 was successfully achieved using one or two equivalents of K[N(SiMe3)2] to yield the mono and

dipotassium salts (5.8 and 5.9). The crystal structure of 5.9(tmeda)2 featured the

dianionic N2B4C2 ligand bifacially coordinated by the potassium ions that are chelated by

a TMEDA molecule. The reaction of 5.9 with [Cp*RuCl]4 prompted the cleavage of the

N−N bond of the bicyclic ligand and yielded a triple-decker sandwich complex 5.10. The

structural determination for 5.10 revealed a triple-decker ruthenocene structure

containing an unusual heterocyclic eight-membered middle deck with a B4N2C2 skeleton.

The ruthenium centers in this derivative are closer to each other than in any other reported multi-decker sandwich complex. The electrochemical studies of 5.10 using cyclic voltammetry revealed three reversible, well separated one-electron transfer steps indicating efficient electron delocalization over the framework.

7.2. Coordination Chemistry of 1,2,4-Triaza-3,5-diborolyl Ligand

Alkali metal complexes containing 1-methyl-3,5-diphenyl-4-methylamino-1,2,4- triaza-3,5-diborolyl ligand were synthesized by the selective deprotonation of 1,2,4- triaza-3,5-diborole 1.63b at the ring nitrogen. The structural characterization of these alkali metal complexes revealed 2D polymeric structures for the lithium (6.1a) and sodium salts (6.1b) and a 1D polymeric arrangement in the potassium salt (6.1c). The structures are dominated by the M−N σ interactions between the 1,2,4-triaza-3,5- diborolyl ligands and the alkali metals. The polymeric structure of the potassium derivative also displayed K−N and K−B π interactions involving the B2N3 ring. A

190

dimeric rhodium complex 6.2 consisting of Rh(cod) fragments bridged by triazadiborolyl

rings σ-coordinating through the amido ring nitrogen atoms was prepared by the reaction

of 6.1a with [Rh(cod)Cl]2. Spectroscopic and structural characterization of 6.2

unambiguously provided conclusive proof for the identity of this dimeric rhodium

complex.

The σ-coordination site at the amido ring nitrogen was blocked using

triphenylborane in order to explore the π-coordination properties of the ligand. The

borane adducts 6.3a – c however did not show any π-coordination between the B2N3 ring and alkali metal ions, as illustrated by the crystal structure of the lithium and potassium derivatives. Instead, the lithium ion is bonded to the pendant methylamino group of the ligand and three acetonitrile molecules in 6.3a(CH3CN)3, while the potassium ion is

coordinated by the pendant methylamino group of the ligand and three phenyl groups in

6.3c. Attempts to synthesize metal complexes by reactions of these boron adducts with

various transition metal reagents were unsuccessful.

The more symmetrical 1,2-dimethyl-3,5-diphenyl-4-methylamino-1,2,4-triaza-

3,5-diborolidine 6.4 was prepared by the reaction of the potassium salt 6.1c with

iodomethane. The subsequent deprotonation of the pendant methylamino group in 6.4,

resulted in the formation of a tricyclic dilithioborate complex 6.5 with B4N8 framework.

The structural investigation of 6.5 revealed a tricyclic structure consisting of two B2N3 linked by one B2N4 ring. The lithium ions are coordinated to the three nitrogen atoms of

the tricyclic ligand and one THF molecule. The oxidation of 6.5 with FeCl2(thf)2 resulted

in the formation of the neutral tricyclic tetrahydrazidotetraborane 6.6. The crystal

191 structure of 6.6 revealed a planar structure with a tricyclic B4N8 ring backbone consisting of one 6- and two 5-membered rings, isostructural with s-indacene. Although the tricyclic B4N8 framework is essentially planar, the geometry around all methyl substituted nitrogen atoms is pyramidal. This is indicative for the localization of the π-electrons on the nitrogen atoms, which is supported by the theoretical data. The electrochemical study of 6.6 showed two irreversible oxidation steps, but no reduction of 6.6 was observed.

The coordination behavior of the investigated 1,2,4-triaza-3,5-diborolyl ligand resembles closely the coordination behavior of its organic analogs pyrrolyl, indolyl and carbazolyl. The latter ligands exhibited σ-coordination modes in their alkali metal salts and π-coordination modes in many of their transition metal complexes. It can be therefore concluded that the 1,2,4-triaza-3,5-diborolyl ligands are promising candidate for the formation of “inorganic” sandwich compounds. According to our investigations, the replacement of the active hydrogen of the pendant methylamino group with a less reactive organic group is a mandatory condition for the successful use of heterocycles as

π-ligands in transition metal chemistry.

7.3. Suggested Future Research Directions

The rich coordination chemistry of the heterocyclic 1,2-diaza-3,5- diborolyl and 1,2,4-triaza-3,5-diborolyl ligands opens a number of appealing research directions. The unusual reactivity of the 1,2-diaza-3,5-diborolyl ligands towards metal carbonyls will be investigated further in order to gain a better understanding of the mechanism of the reaction. The reactivity studies of these anionic diazadiborolyl ligands

192 could be extended by the investigation of the reactions of these alkali metal complexes with carbon monoxide and anhydrous oxygen instead of the metal carbonyls (Scheme

7.1). The viability and generality of this reaction as a tool for carbon-carbon bond formation and activation of oxygen-containing substances will be investigated. The reaction of cyclic or acyclic bis(organoboranes) with various oxygen-based reagents such

” as CO, O2, NO, SO and R 2CO should be explored. The expected products of these reactions are depicted in Scheme 7.2.

Scheme 7.1.

R O R 1) CO B B + Me C C H 2) HX, -MX M Me N N R R R' R' BB

NN O 1) O R R R' R' 2 B B O 2) HX, -MX + 2.4 - 2.6 NN Me H R' R'

Scheme 7.2.

O R'(H2C)C CH + O R2B BR2 + R'H2CC N R2B BR2 NO CO O O C + O2 R2B BR R'H2C H 2 H CH2R' 2 R2BH R' CCH R2B BR2 SO2 S O C + R2B BR2 R'H2C H

R"2CO

R'H2C R" O CC + R2B BR2 H R'

193

Scheme 7.3.

Ph Ph Me B N B Me H B N B Ph Ph Me B N B Ph Ph Me M Ph Ph Ph Ph H B N B Me B N B B N B Ph Ph Ph K Ph Me Me 1) KHMDS Me M H B N B N B B B B Ph Ph Me N MX2 Ph Ph THF Ph Ph B N B Me M M Me Me H N Ph Ph Ph Ph B N B B B THF B N B 2) MX2 B N B Ph Me Me Me Ph n Me Ph Ph N M B N B H B B Ph Ph Ph Ph Ph Ph B N B Me Me 5.8 M Ph Ph B N B H B N B Me Me Ph Ph B N B H Ph Ph

Boron-containing heterocyclic ligands have played an important role in the

chemistry of multidecker sandwich complexes.64 The preliminary studies of the

heterobicyclic 1,5-diaza-2,4,6,8-tetraborolyl ligand showed that the mono and dianionic ligands (5.8 and 5.9) are promising building blocks for the construction of polydecker sandwiches. The synthesis of multidecker sandwich complexes utilizing the heterobicyclic 1,5-diaza-2,4,6,8-tetraborolyl ligands would be appealing. The synthetic approach for the assembly of the polydecker sandwich complexes is depicted in Scheme

7.3. The synthetic route involves the successive stacking of two smaller sandwich complexes to a metal center. The electronic properties of these polydecker sandwich complexes will be examined.

Me Me R R R R Al Al BB

NN PP R' R' R' R'

In addition, the development of new heterocyclic analogs by the substitution of

the skeletal boron and nitrogen atoms with aluminium and phosphorus atoms,

respectively, will also be an exciting extension of the research project. The synthesis of

194 these ligands will be attempted using the same ring closing approach described in

Chapter Two of this thesis, followed by an investigation and comparison of their coordination properties to both heterocyclic and organic cyclopentadienyl ligands.

195

CHAPTER EIGHT

Experimental Methods

8.1. General Considerations

All reactions and manipulations of products were performed under an argon atmosphere using standard Schlenk line and inert-atmosphere glove box techniques.

Solvents were dried and deoxygenated prior to use. Starting materials were prepared according to reported procedures or purchased from commercial suppliers.

8.1.1. Solvents

Diethyl ether (Et2O) and tetrahydrofuran (THF) were dried over sodium/benzophenone ketyl and stored in evacuated 500 mL glass bombs over sodium/benzophenone ketyl. Toluene, benzene and pentane were dried over sodium metal and hexane was dried under sodium/potassium alloy. Dichloromethane (DMC), triethylamine (NEt3) and tetramethylpiperidine (TMP) were dried over calcium hydride.

The deuterated solvents were purchased from Cambridge Isotope Laboratories, Inc. and dried prior to usage: tetrahydrofuran-d8 (THF-d8) (potassium metal), benzene-d6 (C6D6)

196

(sodium/benzophenone ketyl), toluene-d8 (Tol-d8) (potassium metal) and

dichloromethane-d2 (CD2Cl2) (99% anhydrous, calcium hydride).

8.1.2. Reagents

The compounds SnMe4, SnPh4, Me3SiNMe2, Na[N(SiMe3)2], K[N(SiMe3)2], KH,

t BCl3, BPh3, HNMe2, H2N-NH2⋅2HCl, [ BuOC=ONH]2, nBuLi (1.6 M solution in

hexane), HCl (1.0 M solution in diethyl ether and 4.0 M solution in dioxane), Cp*H,

anhydrous ZnCl2, CdBr2, HgCl2, GeCl2 · dioxane, SnCl2, [Rh(cod)Cl]2, RuCl3⋅4H2O,

ZrCl4, were commercial reagents purchased from Aldrich and used as received. The

reagent [PhNH]2 (Aldrich) was purified by washing with hexane, and SiCl4 was distilled

prior to use. H2N-NMeH was dried over CaH2 and vacuum transferred into a glass bomb.

PhB(NMe2)2 was prepared by addition of a PhBCl2 solution to HNMe2 in hexane and

distilled (b.p. 75 °C, atmospheric pressure). FeCl2(thf)2 was prepared by the Soxhlet

182 extraction of anhydrous FeCl2 with THF. [Cp*RuCl]4 and ZrCp*Cl3 were prepared

according to literature procedures and the purity was checked by 1H NMR spectroscopy

183 (δ in THF-d8: 1.56 ppm (s, C5(CH3)5 in [Cp*RuCl]4) and 1.58 ppm (s, C5(CH3)5 in

184 Cp*ZrCl3). Compounds B(C6F5)3 and LiB(C6F5)4 were prepared according to reported

185 procedures. [H(Et2O)2]B(C6F5)4 was prepared by addition of a HCl solution to

186 LiB(C6F5)4 in THF solution.

197

8.1.3. Instrumentations

The NMR spectra were collected on a Bruker Advance DRX-400 instrument (1H,

400 MHz; 7Li, 155.3 MHz; 11B, 128.2 MHz; 13C, 100.5 MHz; 29Si, 79.4 MHz; 113Cd, 88.6

MHz; 119Sn, 148.7 MHz; 199Hg, 71.4 MHz) spectrometers. The 1H and 13C NMR spectra

1 1 were calibrated with respect to C6D5H ( H, 7.15 ppm), THF-d7 ( H, 3.58 ppm), CDHCl2

1 1 13 13 ( H, 5.32 ppm), toluene-d7 ( H, 2.09 ppm), C6D6 ( C, 128.39 ppm), THF-d8 ( C, 67.57

13 13 ppm), CD2Cl2 ( C, 54.00 ppm) and toluene-d8 ( C, 138.39 ppm). The other NMR

7 spectra were referenced to the following external standards: LiCl in D2O ( Li, 0 ppm),

11 113 BF3⋅Et2O in C6D6 ( B, 0 ppm), Cd(ClO4)2 in THF-d8 ( Cd, 0 ppm), Me4Sn in C6D6

119 199 ( Sn, 0 ppm) and HgCl2 in THF-d8 ( Hg, 1495 ppm).

The electron impact (EI), electrospray ionization (ESI), and high-resolution mass

spectrometry (HRMS) and the elemental analyses were performed by the Analytical

Instrumentation Laboratory of the Department of Chemistry, University of Calgary. The

EPR spectroscopy was performed on a Bruker EMX 113 spectrometer with the help of

Dorothy Fox (Department of Chemistry, University of Calgary), Tracey Roemmele and

Dr. René Boeré (University of Lethbridge). Cyclovoltammograms were recorded on a

PARstat 2273 potentiostat at scan rates between 50 – 1000 mVs-1. The cell had a

platinum disk working electrode, a platinum wire auxiliary electrode, and a silver wire

0/+1 0 pseudo reference electrode. The internal standard [Cp2Co] was used with E = 1.36 V vs. ferrocene and -0.82 V vs. standard calomel electrode in THF.

198

8.1.4. X-ray Crystallography

X-ray structures of 2.3c, 2.3d, 2.4a, 2.4c(thf)3, 2.5c(thf)3, 2.6b(thf), 2.6b(thf)2,

3.1b · 0.25C5H12, 3.2, 3.3b, 3.4, 4.1a, 4.1b, 4.2a · BrLi(thf)3, 4.2b, 5.1b, 5.1e, 5.2, 5.3,

5.4, 5.6c, 5.7, 5.9(tmeda)2, 5.10, 6.1b, 6.1c, 6.2, 6.3a(CH3CN)3, 6.3c, 6.5, and 6.6 were

collected, solved and refined by Dr. M. Parvez, 2.5b(thf)3 by Dr. D. Eisler and Andrea

Corrente, 2.5b, and 2.6c(thf) by Dr. J. Konu, and 3.3a, 4.3, and 6.1a by Dr. R. McDonald

(University of Alberta). Single crystals were coated with Paratone 8277 oil (Exxon) and mounted on a glass fiber. All measurements were made on a Nonius KappaCCD diffractometer, excepted for 3.3a, 4.3, and 6.1a that were measured on a Bruker SMART

1000 CCD/PLATFORM diffractometer, with graphite monochromated MoKα radiation (λ

= 0.71073 Å). The crystal data, data collection and refinement parameters for the

complexes are tabulated and summarized in Appendix One. The data were collected188 at a temperature of 173(2) K using ω and ϕ scans and corrected for Lorentz and polarization

effects and for absorption using multi-scan method.189

The structures were solved by the direct methods188 and expanded using Fourier

techniques.190 The non-hydrogen atoms were refined anisotropically. Hydrogen atoms

were included at geometrically idealized positions and were not refined. The final cycle

of full-matrix least-squares refinement using SHELXL97191 converged with unweighted

and weighted agreement factors, R and wR (all data), respectively, and goodness of fit,

F2. The weighting scheme was based on counting statistics and the final difference map

was essentially featureless. The figures were plotted with the aid of Diamond.192

199

8.1.5. Computational Calculations

The computational studies of the model structures 5.10a, [5.10a]2-, 6.6, [6.6]2+ and [6.6]2- were performed by Dr. H. Tuononen at the University of Jyväskylä. DFT calculations were performed for complexes 5.10a and [5.10a]2- using the Turbomole

5.9.1 program package.181 The molecular structures were optimized by using a

combination of the PBE1PBE exchange-correlation functional179 with the Ahlrichs'

triple-zeta valence basis sets augmented by one set of polarization functions (def-

TZVP)193a; corresponding ECP basis set was used for the Ru nuclei.193b Program ADF

2006.01194 was used to perform an energy decomposition195 and Mayer bond order196 analyses for optimized structure of 5.10a by employing fragment molecular orbitals

+ 2- calculated for [CpRu] and [B4N2C2H6] subunits. The analyses utilized the PBEPBE

GGA-functional179 in combination with STO-type all-electron basis sets of TZP quality;197 scalar relativistic ZORA Hamiltonian was applied in all ADF calculations.198

The MO calculations of 6.6, [6.6]2+ and [6.6]2- were done with the Turbomole 5.9.1 and

Gaussian 03 program packages.181 The molecular structures were optimized with the

PBEO hybrid functional179 using the Aldrich’s TZVP basis sets.180

200

8.2. Experimental Details for Chapter 2

Preparation of HMeC(BRCl)2 and substituted hydrazine compounds

1,1-Bis(chloromethylboryl)ethane (2.1a):68

SnMe4 (5.73 g, 0.030 mol) was added dropwise to 1,1-bis(dichloroboryl)ethane

(2.824 g, 0.015 mol) at 0 °C over 20 minutes. The neat mixture was subsequently heated to 40 °C for 2 h to ensure the completion of the reaction. The desired product was used without isolation due to the significant loss of product resulting from the separation of the

1 SnMe3Cl by-product. H NMR (400 MHz, C6D6, 25 ºC): δ = 0.79 (s, 6H, BCH3), 1.88 (d,

3 11 3H, JHH = 6.3 Hz, HCCH3), 2.11 (s, br, 1H, HCCH3); B NMR (128 MHz, C6D6, 25

ºC): δ = 74.8 (s, br).

1,1-Bis(chlorophenylboryl)ethane (2.1b):68

1,1-Bis(chloroboryl)ethane (5.31 g, 27.7 mmol) was added to Ph4Sn (7.90 g, 18.5

mmol) and the neat mixture was stirred at ambient temperature for 2 h. Subsequently, the

colorless suspension was heated and maintained at 80 °C overnight. Vacuum distillation

yielded PhSnCl3 at 60 - 80 °C / 0.1 torr and the desired product at 100 - 115 °C / 0.10

1 torr, as a colorless liquid (6.50 g, 85%). H NMR (400 MHz, C6D6, 25 ºC): δ = 1.51 (d,

3 3 3H, JHH = 6.6 Hz, HCCH3), 3.22 (q, 1H, JHH = 6.6 Hz, HCCH3), 7.01 – 7.16 (m, 6H, m-,

11 p-C6H5), 7.98 (d, 4H, o-C6H5); B NMR (128 MHz, C6D6, 25 ºC): δ = 66.9 (s, br).

1,1-Bis(dimethylaminochloroboryl)ethane (2.1c):69

A solution of Me3Si-NMe2 (3.67 g, 0.0313 mol) in pentane (20 mL) was added dropwise to a solution of 1,1-bis(dichloroboryl)ethane (3.00 g, 0.0156 mol) in pentane

201

(20 mL). The mixture was heated under reflux for 2 h and then distilled at 42 °C / 0.1

torr. The desired product was obtained as colorless liquid (2.44 g, 73%). 1H NMR (400

MHz, C6D6, 25 ºC): δ = 1.32 (s, 4H, HCCH3 & HCCH3), 2.36 (s, 6H, N(CH3)2), 2.60 (s,

11 6H, N(CH3)2); B NMR (128 MHz, C6D6, 25 ºC): δ = 38.7 (s, br).

Me2C=N-N=CMe2:

Hydrazine (10.0 g, 0.312 mol) was slowly mixed with acetone (100 mL) under

external cooling, in a flask containing molecular sieves (3 Å, 100 g). The mixture was

left sitting at ambient temperature for 24 h and then the solution was decanted and

distilled at atmospheric pressure. The product was collected as a clear colorless liquid at

1 134 °C (26.7 g, 74%). H NMR (400 MHz, C6D6, 25 °C): δ = 1.80 (d, 12H, CCH3).

1,2-Diisopropylhydrazine (2.2a):70

A Parr hydrogenation apparatus was loaded with Me2C=N-N=CMe2 (26.7 g,

0.238 mol), ethanol (36 mL), acetic acid (5.35 mL) and the Pt/C catalyst (10%, 100 mg).

The apparatus was deoxygenated and stirred under 5 bar of H2 for 24 h. The resulting

mixture was filtered and the filtrate was decomposed with a hydrochloride solution (12.1

M, 50 mL). Volatiles were removed in vacuum and the residue was dissolved in water

(60 mL) and treated with solid KOH (34.0 g, 0.606 mol). The product was extracted with

diethyl ether (3 × 50 mL), the combined extracts were dried over amalgamated aluminum

foil and distilled at atmospheric pressure, yielding an air sensitive colorless liquid boiling

1 3 at 124 °C (16.4 g, 59%). H NMR (400 MHz, C6D6, 25 °C): δ = 0.97 (d, 12H, JHH = 6.2

3 13 Hz, CH(CH3)2), 2.34 (s, br, 2H, NH), 2.74 (sep, 2H, JHH = 6.2 Hz, CH(CH3)2); C

NMR (400 MHz, C6D6, 25 °C): δ = 21.6 (s, CH(CH3)2), 50.9 (s, CH(CH3)2).

202

1,2-Bis-tert-butoxycarbonylpyrazolidine (Pz-boc):90

A solution of di-tert-butyl hydrazodiformate (5.23 g, 22.5 mmol) in

dimethylformamide, DMF, (50 mL) was added dropwise to a stirred suspension of

sodium hydride (1.14 g, 47.3 mmol) in DMF (30 mL) at room temperature, which

resulted in the formation of hydrogen gas. The mixture was stirred for 30 min and neat

1,3-dibromopropane (2.29 mL, 22.5 mmol) was added dropwise to mixture afforded a

pale yellow suspension. The reaction mixture was stirred overnight. The

dimethylformamide was removed by distillation at reduce pressure (90 °C / 10 torr) and

the crude product was extracted with diethyl ether (50 mL) yielded a pale yellow

solution. Volatiles were removed under full vacuum and the desired product was isolated

1 as a colorless oil (5.82 g, 95%). H NMR (400 MHz, CD2Cl2, 25 °C): δ = 1.43 (s, 18H,

t 3 O=CO Bu), 1.98 (quint, 2H, JHH = 7.2 Hz, N(CH2)3), 3.15 (m, 2H, N(CH2)3), 3.82 (m,

13 2H, N(CH2)3); C NMR (400 MHz, CD2Cl2, 25 °C): δ = 26.2 (s, N(CH2)3), 28.5 (s,

t t O=CO Bu), 46.9 (s, N(CH2)3), 81.3 (s, O=CO Bu).

Pyrazolidine Hydrochloride (2.2c·HCl): 90

A 4 M hydrogen chloride / dioxane solution (32 mL, 0.128 mmol) was added

dropwise to a stirring solution of Pz-boc (5.82 g, 21.4 mmol) in diethyl ether (20 mL).

The mixture was stirred for 3 h at room temperature and the desired product precipitated

out of solution and was collected by filtration under argon as a colorless hygroscopic

1 3 solid (2.88 g, 93%). H NMR (400 MHz, DMSO-d6, 25 °C): δ = 1.96 (quint, 2H, JHH =

3 13 7.1 Hz, N(CH2)3), 3.04 (t, 2H, JHH = 7.3 Hz, N(CH2)3), 9.53 (s, br, NH); C NMR (400

MHz, DMSO-d6, 25 °C): δ = 26.2 (s, N(CH2)3), 46.6 (s, N(CH2)3).

203

Preparation of 1,2-diaza-3,5-diborolidines

1,2-Diisopropyl-4-methyl-3,5-dimethyl-1,2-diaza-3,5-diborolidine (2.3a):

A solution of 2.1a (1.20 g, 7.96 mmol) in hexane (20 mL) was added slowly to a

solution containing 2.2a (0.90 g, 7.96 mmol) and triethylamine (1.61 g, 15.9 mmol) in

hexane (15 mL). The suspension was stirred at ambient temperature for 30 min and

subsequently filtered. The filtrate was concentrated in vacuum and the residual liquid

was distilled under reduced pressure. The product was distilled at 78 oC / 5 torr as a clear

1 colorless liquid (0.93 g, 60%). H NMR (400 MHz, C6D6, 25 ºC): δ = 0.25 (q, br, 1H,

3 3 JHH = 8.2 Hz, HCCH3), 0.64 (s, 6H, BCH3), 1.17 (d, 12H, JHH = 6.8 Hz, CH(CH3)2),

3 3 13 1.19 (d, 3H, JHH = 8.2 Hz, HCCH3,), 3.67 (sep, 2H, JHH = 6.8 Hz, CH(CH3)2); C NMR

(100 MHz, C6D6, 25 ºC): δ = 0.17 (s, br, BCH3), 9.01 (s, HCCH3), 22.9 (s, CH(CH3)2),

11 + 47.2 (s, CH(CH3)2); B NMR (128 MHz, C6D6, 25 ºC): δ = 46.9 (s, br); MS (EI , 70 eV):

+ + 11 m/z (%): 194.2 (59) [M] , 179.2 (100) [M–Me] ; HRMS for H24C10N2 B2 calcd.

194.2126, found 194.2131.

1,2-Diisopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolidine (2.3b):

A solution of 2.2a (1.27 g, 10.9 mmol) and triethylamine (2.21 g, 21.8 mmol) in hexane (30 mL) was slowly added to a stirred solution of 2.1b (3.00 g, 10.9 mmol) in hexane (20 mL). The suspension was stirred at ambient temperature for 12 h and the white precipitate was filtered off. Volatiles were removed under full vacuum, leaving

1 behind the product as a colorless oil (2.46 g, 71%). H NMR (400 MHz, C6D6, 25 °C): δ

3 3 = 1.12 (d, 3H, JHH = 6.8 Hz, HCCH3), 1.13 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.14 (d,

3 3 6H, JHH = 6.9 Hz, CH(CH3)2), 3.91 (sep, 2H, JHH = 6.9 Hz, CH(CH3)2), 7.20 – 7.31 (m,

204

13 6H, m-, p-C6H5), 7.46 (d, 4H, o-C6H5); C NMR (100 MHz, C6D6, 25 °C): δ = 9.32 (s,

HCCH3), 24.0, 24.2 (s, CH(CH3)2), 26.8 (s, br, HCCH3), 49.4 (s, CH(CH3)2), 127.9 (s, p-

11 C6H5), 128.2 (s, m-C6H5), 133.0 (s, o-C6H5), 141.8 (s, br, i-C6H5); B NMR (128 MHz,

+ + C6D6, 25 °C): δ = 46.0 (s, br); MS (EI , 70 eV): m/z (%): 318 (100) [M] , 303 (88) [M–

+ + Me] , 190 (48) [M–(iPrN)2Me+H] .

1,2-Diphenyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolidine (2.3c):

A solution of 1,1-diphenylhydrazine (737 mg, 4.00 mmol) and triethylamine (810 mg, 8.00 mmol) in benzene (50 mL) was added to a stirring solution of 2.1b (1.10 g, 4.00 mmol) in hexane (20 mL) producing a pale yellow suspension. The mixture was stirred at ambient temperature for 12 h and the white precipitate was filtered off. Volatiles were removed under full vacuum and pale yellow solid was obtained. The solid was washed twice with hexane (25 mL) and the product was collected as a colorless solid (0.98 g,

63%). Colorless thin plate crystals were obtained by cooling a concentrated solution of

1 2.3c in benzene and hexane at -35 °C. H NMR (400 MHz, C6D6, 25 °C): δ = 1.45 (d, 3H,

3 3 3 JHH = 8.2 Hz, HCCH3), 1.65 (q, 1H, JHH = 8.2 Hz, HCCH3), 6.65 (t, 2H, JHH = 7.3 Hz,

3 3 p-NC6H5), 6.75 (t, 4H, JHH = 7.5 Hz, m-NC6H5), 6.87 (d, 4H, JHH = 7.5 Hz, o-NC6H5),

13 7.11 (m, 6H, m- & p-BC6H5), 7.44 (d, 4H, o-BC6H5); C NMR (100 MHz, THF-d8, 25

°C): δ = 11.0 (s, HCCH3), 23.0 (s, br, HCCH3), 127.1 (s, p-NC6H5), 128.2 (s, m-NC6H5),

129.1 (s, o-NC6H5), 129.3 (s, p-BC6H5), 130.2 (s, m-BC6H5), 134.9 (s, o-BC6H5), 137.5

11 (s, br, i-BC6H5), 142.1 (s, i-NC6H5); B NMR (128 MHz, THF-d8, 25 °C): δ = 44.0 (s, br); MS (EI+, 70 eV): m/z (%): 386 (76) [M]+, 371 (21) [M - Me]+.

205

1,2-Diphenyl-4-methyl-3,5-dimethylamino-1,2-diaza-3,5-diborolidine (2.3d):

A solution of 1,1-diphenylhydrazine (175 mg, 0.950 mmol) and triethylamine

(192 mg, 1.90 mmol) in benzene (20 mL) was slowly added to a stirring solution of 2.1c

(200 g, 0.950 mmol) in hexane (20 mL) producing an orange-yellow suspension. The mixture was stirred at ambient temperature for 3 h and the white precipitate was filtered off. Volatiles were removed under full vacuum and a colorless solid was obtained (154 mg, 51%). Colorless thin plate crystals were obtained by cooling a concentrated solution

1 of 2.3d in toluene and hexane at -35 °C. H NMR (400 MHz, toluene-d8, 25 °C): δ = 0.65

3 3 (q, 1H, JHH = 8.1 Hz, HCCH3), 1.01 (d, 3H, JHH = 8.1 Hz, HCCH3), 2.57 (s, 12H,

13 N(CH3)2), 6.82 (t, 2H, p-C6H5), 6.96 (m, 4H, m-C6H5), 7.19 (m, 4H, o-C6H5); C NMR

(100 MHz, toluene-d8, 25 °C): δ = 8.9 (s, br, HCCH3), 10.2 (s, HCCH3), 39.3 (s, br,

N(CH3)3), 40.3 (s, br, N(CH3)3), 116.6 (s, m-C6H5), 119.8 (s, p-C6H5), 129.0 (s, o-C6H5),

11 + 149.5 (s, i-C6H5); B NMR (128 MHz, toluene-d8, 25 °C): δ = 37.4 (s, br); MS (EI , 70

+ 11 eV): m/z (%): 320 (26) [M] ; HRMS for H26C18N4 B2 calcd. 320.23436, found

320.23559.

1,2-Cyclopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolidine (2.3e):

A colorless solution of potassium hexamethyldisilazane, KHMDS, (2.987 g, 14.97 mmol) in THF (20 mL) was slowly added to a stirring white suspension of 2.2c·HCl

(1.086 g, 7.488 mmol) in THF (30 mL). The mixture was stirred at room temperature

overnight afforded a light brown suspension. Triethylamine (1.515 g, 14.97 mmol) was

added to the mixture and a solution of 2.1b (2.058 g, 7.488 mmol) in hexane (20 mL) was

slowly added to the suspension mixture, which immediately yielded a yellow suspension.

206

The mixture was stirred at room temperature for 5 h and the white precipitate was filtered

off. Volatiles were removed under full vacuum and a pale yellow residue was obtained.

The desired product was washed a few times with pentane (25 mL) and colorless solid of

1 2.3e was isolated (1.605 g, 71%). H NMR (400 MHz, C6D6, 25 °C): δ = 1.44 (d, 3H,

3 3 JHH = 8.4 Hz, HCCH3), 1.55 (m, 2H, NCH2CH2), 1.73 (q, 1H, JHH = 8.4 Hz, HCCH3),

3 3 3.21 (m, 4H, NCH2CH2), 7.31 (t, 2H, JHH = 7.5 Hz, p-C6H5), 7.40 (t, 4H, JHH = 7.5 Hz,

3 13 m-C6H5), 7.77 (d, 4H, JHH = 7.5 Hz, o-C6H5); C NMR (100 MHz, C6D6, 25 °C): δ =

11.7 (s, HCCH3), 22.5 (s, br, HCCH3), 30.9 (s, NCH2CH2), 44.2 (s, NCH2CH2), 128.6 (s,

11 m-C6H5), 129.3 (s, p-C6H5), 134.9 (s, o-C6H5), 137.6 (s, br, i-C6H5); B NMR (128 MHz,

+ + THF-d8, 25 °C): δ = 39.3 (s, br); MS (EI , 70 eV): m/z (%): 274 (45) [M] ; HRMS for

11 H20C17N2 B2 calcd. 274.18126, found 274.18256.

Preparation of alkali metal salts containing 1,2-diaza-3,5-diborolyl ligands

1,2-Diisopropyl-4-methyl-3,5-dimethyl-1,2-diaza-3,5-diborolyllithium (2.4a):

A yellow solution containing lithium 2,2,6,6-tetramethylpiperidide, LiTMP, was

formed in situ by the reaction of 1.6 M n-butyllithium, nBuLi, in hexane (11.0 mL, 17.7 mmol) and 2,2,6,6-tetramethylpiperidine, TMP, (2.50 g, 17.7 mmol) in THF (3 mL). The solutions of 2.3a (1.14 g, 5.90 mmol) in THF (3 mL) and LiTMP were pre-cooled to -35

ºC, mixed and kept at this temperature for another 2 h. The yellow solution was allowed

to warm to room temperature overnight. The solvent was removed under vacuum

producing an orange solid, which was washed a few times with hexane (30 mL) and dried

under vacuum. The product was obtained as a colorless powder (1.07 g, 91 %). Colorless

crystals were obtained by slow evaporation of a solution of 2.4a in a mixture of THF and

207

1 3 benzene. H NMR (400 MHz, THF-d8, 25 °C): δ = 0.33 (s, 6H, BCH3), 1.16 (d, 12H, JHH

3 13 = 6.8 Hz, CH(CH3)2), 1.66 (s, 3H, CCH3), 3.76 (sep, 2H, JHH = 6.8 Hz, CH(CH3)2); C

NMR (100 MHz, THF-d8, 25 °C): δ = 0.17 (s, br, BCH3), 13.6 (s, CCH3), 23.0 (s,

13 CH(CH3)2), 48.2 (s, CH(CH3)2); C NMR (100 MHz, THF-d8, -50 °C): δ = 1.4 (s,

BCH3), 13.1 (s, CCH3), 22.9 (s, CH(CH3)2), 48.4 (s, CH(CH3)2), 86.7 (s, br, B2CCH3);

11 7 B NMR (128 MHz, THF-d8, 25 °C): δ = 38.3 (s, br); Li NMR (155 MHz, THF-d8, 25

°C): δ = -2.34 (s); MS (EI+, 70 eV): m/z (%): [M–Li+H]+ 194.2 (36), 179.2 (100) [M–Li–

Me+H]+.

1,2-Diisopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolyllithium (2.4b):

A yellow solution of LiTMP was formed in situ by the reaction of 1.6 M nBuLi in

hexane (3.20 ml, 5.12 mmol) and TMP (0.724 g, 5.12 mmol) in THF (3 mL) and the

solution of 2.3b (1.63 g, 5.12 mmol) in THF (20 mL) were pre-cooled to -35 °C for 1 h.

The pre-cooled solutions of 2.3b and LiTMP were mixed and kept at -35 °C for 2 h. The

solvent was removed under vacuum producing a pale yellow residue, which was washed

several times with hexane (30 mL) and dried under vacuum. The product was obtained

1 as an off-white powder (680 mg, 41%). H NMR (400 MHz, THF-d8, 25 °C): δ = 1.07 (d,

3 3 12H, JHH = 6.8 Hz, CH(CH3)2), 1.72 (s, 3H, CCH3), 3.80 (sep, 2H, JHH = 6.8 Hz,

3 3 CH(CH3)2), 6.90 (t, 2H, JHH = 7.3 Hz, p-C6H5), 7.06 (t, 4H, JHH = 7.3 Hz, m-C6H5), 7.45

3 13 (d, 4H, JHH = 7.3 Hz, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 14.7 (s, CCH3),

23.9 (s, CH(CH3)2), 50.7 (s, CH(CH3)2), 94.8 (s, br, B2CCH3), 124.1 (s, p-C6H5), 126.6

11 (s, m-C6H5), 135.1 (s, o-C6H5), 151.2 (s, br, i-C6H5); B NMR (128 MHz, THF-d8, 25

7 °C): δ = 37.9 (s, br); Li NMR (155 MHz, THF-d8, 25 °C): δ = -2.8 (s).

208

1,2-Diisopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolylsodium (2.5b):

A solution of 2.3b (200 mg, 0.629 mmol) in THF (20 mL) was added to a stirred

solution of sodium hexamethyldisilazane, NaHMDS, (115 mg, 0.629 mmol) in THF (15

mL) forming a yellow mixture, which was stirred at ambient temperature for one day to

ensure the reaction had completed. The volatiles were removed under vacuum producing

a colorless solid, which was washed twice with hexane (30 mL) and dried under vacuum.

The product was obtained as a colorless powder (159 mg, 75%). Colorless thin plates of

2.5b were obtained by slow evaporation of a solution of 2.5b in benzene and THF at

room temperature and colorless block crystals of 2.5b(thf)3 were obtained by storing a

concentrated THF and hexane solution of 2.5b at -35 °C for few days. 1H NMR (400

3 MHz, THF-d8, 25 °C): δ = 1.12 (d, 12H, JHH = 6.8 Hz, CH(CH3)2), 1.75 (s, 3H, CCH3),

3 3 3.90 (sep, 2H, JHH = 6.8 Hz, CH(CH3)2), 6.98 (t, 2H, JHH = 7.3 Hz, p-C6H5), 7.13 (t, 4H,

3 3 13 JHH = 7.3 Hz, m-C6H5), 7.46 (d, 4H, JHH = 7.3 Hz, o-C6H5); C NMR (100 MHz, THF-

d8, 25 °C): δ = 13.9 (s, CCH3), 23.9 (s, CH(CH3)2), 50.5 (s, CH(CH3)2), 91.5 (s, br,

B2CCH3), 125.0 (s, p-C6H5), 127.1 (s, m-C6H5), 135.0 (s, o-C6H5), 149.2 (s, br, i-C6H5);

11 B NMR (128 MHz, THF-d8, 25 °C): δ = 39.1 (s, br).

1,2-Diisopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolylpotassium (2.6b):

A solution of 2.3b (1.35 g, 4.24 mmol) in THF (20 mL) was added to a stirred solution of KHMDS (0.950 g, 4.24 mmol) in THF (15 mL) forming a yellow mixture.

The mixture was stirred at ambient temperature for 12 h and the volatiles were removed under vacuum producing a colorless solid, which was washed twice with hexane (30 mL) and dried under vacuum. The product was obtained as a colorless powder (1.41 g, 93%).

Colorless crystals of 2.6b(thf) and 2.6b(thf)2 were obtained by slow diffusion of hexane

209

into a THF solution of 2.6b and cooling a concentrated solution of 2.6b in THF and

1 hexane at -35 °C, respectively. H NMR (400 MHz, THF-d8, 25 °C): δ = 1.12 (d, 12H,

3 3 JHH = 6.8 Hz, CH(CH3)2), 1.73 (s, 3H, CCH3), 3.89 (sep, 2H, JHH = 6.8 Hz, CH(CH3)2),

3 3 6.97 (t, 2H, JHH = 7.3 Hz, p-C6H5), 7.13 (t, 4H, JHH = 7.3 Hz, m-C6H5), 7.43 (d, 4H,

3 13 JHH = 7.3 Hz, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 14.3 (s, CCH3), 23.9

(s, CH(CH3)2), 50.3 (s, CH(CH3)2), 96.8 (s, br, B2CCH3), 124.7 (s, p-C6H5), 127.1 (s, m-

11 C6H5), 134.6 (s, o-C6H5), 149.9 (s, br, i-C6H5); B NMR (128 MHz, THF-d8, 25 °C): δ =

38.6 (s, br).

1,2-Diphenyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolyllithium (2.4c):

A yellow solution of LiTMP, formed in situ by mixing the 1.6 M nBuLi in hexane

(0.33 mL, 0.53 mmol) with TMP (75 mg, 0.53 mmol) in THF (3 mL), and the solution of

2.3c (0.20 g, 0.53 mmol) in THF (10 mL) were pre-cooled to -35 °C for 1 h. The pre- cooled solutions of 2.3c and LiTMP were mixed and kept at -35 °C for 2 h. The solvent was removed under vacuum producing a pale yellow solid, which was washed twice with hexane (30 mL) and dried under vacuum. The product was obtained as a creamy white powder (0.26 g, 90%). Colorless prismatic crystals of 2.4c(thf)3 were obtained by slow

1 diffusion of hexane into a THF solution of 2.4c. H NMR (400 MHz, THF-d8, 25 °C): δ =

3 3 1.73 (s, 3H, CCH3), 6.35 (t, 2H, JHH = 7.2 Hz, p-NC6H5), 6.70 (t, 4H, JHH = 7.3 Hz, m-

3 3 NC6H5), 6.84 (d, 4H, JHH = 7.5 Hz, o-NC6H5), 6.94 (t, 2H, JHH = 6.1 Hz, p-BC6H5),

3 3 13 7.06 (t, 4H, JHH = 7.2 Hz, m-BC6H5), 7.34 (d, 4H, JHH = 7.2 Hz, o-BC6H5); C NMR

(100 MHz, THF-d8, 25 °C, 10 Hz LB): δ = 14.8 (s, CCH3), 98.1 (s, br, B2CCH3), 118.1

(s, p-NC6H5), 122.5 (s, m-NC6H5), 124.5 (s, p-BC6H5), 126.7 (s, o-NC6H5), 127.5 (s, m-

11 BC6H5), 135.1 (s, o-BC6H5), 148.8 (s, br, i-BC6H5), 149.6 (s, i-NC6H5); B NMR (128

210

7 MHz, THF-d8, 25 °C): δ = 40.5 (s, br). Li NMR (155 MHz, THF-d8, 25 °C): δ = -0.82

(s).

1,2-Diphenyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolylsodium (2.5c):

A solution of 2.3c (0.358 g, 0.927 mmol) in THF (10 mL) was added to a stirred solution of NaHMDS (0.170 g, 0.927 mmol) in THF (15 mL) forming a yellow mixture.

The mixture was stirred at ambient temperature for 3 h to ensure the reaction was complete. The volatiles were removed under vacuum producing a colorless solid, which was washed twice with hexane (30 mL) and dried under vacuum. The product was obtained as a colorless powder (0.40 g, 97%). Colorless block crystals of 2.5c(thf)3 that were obtained by cooling a concentrated solution of 2.5c in THF and pentane at -35 °C.

1 3 H NMR (400 MHz, THF-d8, 25 °C): δ = 1.82 (s, 3H, CCH3), 6.45 (t, 2H, JHH = 7.1 Hz,

3 3 p-NC6H5), 6.76 (t, 4H, JHH = 7.5 Hz, m-NC6H5), 6.87 (d, 4H, JHH = 8.4 Hz, o-NC6H5),

3 3 7.02 (t, 2H, JHH = 7.4 Hz, p-BC6H5), 7.12 (t, 4H, JHH = 7.8 Hz, m-BC6H5), 7.27 (d, 4H,

3 13 JHH = 7.8 Hz, o-BC6H5); C NMR (100 MHz, THF-d8, 25 °C, 10 Hz LB): δ = 13.8 (s,

CCH3), 93.6 (s, br, B2CCH3), 119.5 (s, p-NC6H5), 123.0 (s, m-NC6H5), 125.4 (s, p-

BC6H5), 127.1 (s, o-NC6H5), 127.8 (s, m-BC6H5), 134.9 (s, o-BC6H5), 146.7 (s, br, i-

11 BC6H5), 148.5 (s, i-NC6H5); B NMR (128 MHz, THF-d8, 25 °C): δ = 40.1 (s, br).

1,2-Diphenyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolylpotassium (2.6c):

The solution of 2.3c (0.200 g, 0.518 mmol) in THF (10 mL) was added to a stirred solution of KHMDS (0.116 g, 0.518 mmol) in THF (15 mL) affording a yellow mixture.

The mixture was stirred at ambient temperature for 2 h and the volatiles were removed under vacuum yielded a pale yellow solid, which was washed twice with hexane (30 mL)

211

and dried under vacuum. The product was obtained as a light brown powder (0.20 g,

89%).). Colorless prismatic crystals of 2.6c(thf) were obtained by slow diffusion of

1 hexane into a THF solution of 2.6c. H NMR (400 MHz, THF-d8, 25 °C): δ = 1.78 (s, 3H,

3 3 CCH3), 6.46 (t, 2H, JHH = 7.2 Hz, p-NC6H5), 6.77 (t, 4H, JHH = 7.4 Hz, m-NC6H5), 6.84

3 3 3 (d, 4H, JHH = 8.0 Hz, o-NC6H5), 7.01 (t, 2H, JHH = 7.4 Hz, p-BC6H5), 7.12 (t, 4H, JHH

3 13 = 7.3 Hz, m-BC6H5), 7.34 (d, 4H, JHH = 6.9 Hz, o-BC6H5); C NMR (100 MHz, THF-

d8, 25 °C, 10 Hz LB): δ = 14.4 (s, CCH3), 97.7 (s, br, B2CCH3), 119.4 (s, p-NC6H5),

122.8 (s, m-NC6H5), 125.3 (s, p-BC6H5), 127.2 (s, o-NC6H5), 128.0 (s, m-BC6H5), 134.8

11 (s, o-BC6H5), 147.2 (s, br, i-BC6H5), 148.5 (s, i-NC6H5); B NMR (128 MHz, THF-d8,

25 °C): δ = 40.1 (s, br).

1,2-Cyclopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolyllithium (2.4e):

A yellow solution of LiTMP, formed in situ by mixing the 1.6 M nBuLi in hexane

(2.52 mL, 4.03 mmol) with TMP (570 mg, 4.03 mmol) in THF (3 mL) and the solution of

2.3e (1.10 g, 4.03 mmol) in THF (10 mL) were pre-cooled to -35 °C for 1 h. The pre- cooled solutions of 2.3e and LiTMP were mixed, affording an orange mixture, which was stirred at room temperature for 2 h. The solvent was removed under vacuum producing a pale yellow residue, which was washed with hexane (30 mL) and dried under vacuum.

The product was obtained as a colorless powder (1.02 g, 91%). 1H NMR (400 MHz,

3 THF-d8, 25 °C): δ = 2.18 (quint, 2H, JHH = 6.4 Hz, NCH2CH2), 2.23 (s, 3H, CCH3),

3 3 3.63 (t, 4H, JHH = 5.9 Hz, NCH2CH2), 6.99 (t, 2H, JHH = 7.1 Hz, p-C6H5), 7.13 (t, 4H,

3 3 13 JHH = 7.3 Hz, m-C6H5), 7.58 (d, 4H, JHH = 7.3 Hz, o-C6H5); C NMR (100 MHz, THF-

d8, 25 °C): δ = 15.7 (s, CCH3), 31.5 (s, NCH2CH2), 46.0 (s, NCH2CH2), 91.6 (s, br,

B2CCH3), 125.1 (s, p-C6H5), 127.4 (s, m-C6H5), 134.7 (s, o-C6H5), 145.5 (s, br, i-C6H5);

212

11 7 B NMR (128 MHz, THF-d8, 25 °C): δ = 31.5 (s, br); Li NMR (155 MHz, THF-d8, 25

°C): δ = -2.16 (s); MS (EI+, 70 eV): m/z (%): 274 (8) [M-Li+H]+; HRMS for

11 H20C17N2 B2 calcd. 274.18126, found 274.18037.

1,2-Cyclopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolylsodium (2.5e):

A solution of 2.3e (0.300 g, 1.09 mmol) in THF (10 mL) was added to a stirred

solution of NaHMDS (0.200 g, 1.09 mmol) in THF (15 mL) forming an orange mixture.

The mixture was stirred at ambient temperature for 3 h and the volatiles were removed

under vacuum producing an orange residue, which was washed twice with hexane (30

mL) and dried under vacuum. The product was obtained as an off-white powder (0.232

1 3 g, 72%). H NMR (400 MHz, THF-d8, 25 °C): δ = 2.25 (quint, 2H, JHH = 6.5 Hz,

3 NCH2CH2), 2.29 (s, 3H, CCH3), 3.74 (t, 4H, JHH = 6.5 Hz, NCH2CH2), 6.99 (t, 2H,

3 3 3 JHH = 7.1 Hz, p-C6H5), 7.17 (t, 4H, JHH = 7.3 Hz, m-C6H5), 7.66 (d, 4H, JHH = 7.3 Hz,

13 o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 16.2 (s, CCH3), 32.1 (s, NCH2CH2),

46.5 (s, NCH2CH2), 92.4 (s, br, B2CCH3), 125.0 (s, p-C6H5), 127.5 (s, m-C6H5), 134.7 (s,

11 o-C6H5), 146.2 (s, br, i-C6H5); B NMR (128 MHz, THF-d8, 25 °C): δ = 32.1 (s, br).

1,2-Cyclopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolylpotassium (2.6e):

A solution of 2.3e (0.315 g, 1.15 mmol) in THF (10 mL) was added to a stirred solution of KHMDS (0.230 g, 1.15 mmol) in THF (15 mL) affording an orange solution.

The mixture was stirred at ambient temperature for 2 h and the volatiles were removed under vacuum yielded a creamy yellow residue, which was washed twice with hexane (30 mL) and dried under vacuum. The product was obtained as a colorless powder (0.33 g,

92%). Thin needles of colorless crystals were obtained by recrystallization of 2.6e in

213

1 THF solution. H NMR (400 MHz, THF-d8, 25 °C): δ = 2.25 (s, br, 2H, NCH2CH2), 2.25

3 (s, 3H, CCH3), 3.70 (s, br, 4H, NCH2CH2), 6.96 (t, br, 2H, p-C6H5), 7.15 (t, 4H, JHH =

3 13 7.1 Hz, m-C6H5), 7.64 (d, 4H, JHH = 6.7 Hz, o-C6H5); C NMR (100 MHz, THF-d8, 25

°C): δ = 16.5 (s, CCH3), 32.2 (s, NCH2CH2), 46.5 (s, NCH2CH2), 91.1 (s, br, B2CCH3),

11 124.8 (s, p-C6H5), 127.5 (s, m-C6H5), 134.6 (s, o-C6H5), 145.2 (s, br, i-C6H5); B NMR

(128 MHz, THF-d8, 25 °C): δ = 31.7 (s, br).

8.3. Experimental Details for Chapter 3

Preparation of the trichloro(1,2-diaza-3,5-diborolylsliyl) complexes

1,2-Diisopropyl-3,5-dimethyl-1,2-diaza-3,5-diborolyl-trichlorosilane (3.1a):

Silicon tetrachloride (97 μL, 0.85 mmol) was added to a colorless solution of 2.4a

(0.17 g, 0.85 mmol) in THF (15 mL). The clear mixture was stirred at ambient temperature for 2 h and the volatiles were subsequently removed under vacuum leaving behind a thick residue. The product was extracted with hexane and isolated upon removal of the solvent in form of a colorless crystalline solid (0.21 g, 75%). 1H NMR (400 MHz,

3 C6D6, 25 °C): δ = 0.64 (s, 6H, BCH3), 1.08 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.13 (d,

3 3 3 6H, JHH = 6.9 Hz, CH(CH3)2), 1.29 (s, 3H, CCH3, JSiH = 13.0 Hz), 3.68 (sep, 2H, JHH =

1 6.9 Hz, CH(CH3)2); H NMR (400 MHz, THF-d8, 25 °C): δ = 0.55 (s, 6H, BCH3), 1.26

3 3 (s, 3H, CCH3), 1.34 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.37 (d, 6H, JHH = 6.9 Hz,

3 13 1 CH(CH3)2), 4.07 (sep, 2H, JHH 6.9 Hz, CH(CH3)2); C{ H} NMR (100 MHz, THF-d8,

25 °C): δ = 0.6 (s, br, BCH3), 10.7 (s, CCH3), 22.3 (s, CH(CH3)2), 22.9 (s, CH(CH3)2),

214

11 1 35.4 (s, br, B2CCH3), 47.6 (s, CH(CH3)2); B{ H} NMR (128 MHz, C6D6, 25 °C): δ =

29 1 + 43.9 (s, br); Si{ H} NMR (79 MHz, THF-d8, 25 °C): δ = 5.7 (s); MS (EI , 70 eV): m/z(%): 327(13) [M]+, 312(40) [M-Me]+, 269(14) [M-Me-iPr]+.

1,2-Diisopropyl-3,5-diphenyl-1,2-diaza-3,5-diborolyl-trichlorosilane (3.1b):

Silicon tetrachloride (32 μL, 0.28 mmol) was added to a colorless solution of 2.6b

(0.10 g, 0.28 mmol) in THF (15 mL). The pale yellow mixture was stirred at ambient temperature for 2 h and the volatiles were subsequently removed under vacuum leaving behind a thick residue. The product was extracted with hexane and isolated upon removal of the solvent in the form of a colorless crystalline (0.10 g, 79%). Colorless crystals were obtained by cooling a concentrated pentane solution of 3.1b to -35 °C. 1H NMR (400

3 3 MHz, THF-d8, 25 °C): δ = 1.20 (d, br, 6H, JHH = 6.3 Hz, CH(CH3)2), 1.31 (d, 6H, JHH =

3 3 6.9 Hz, CH(CH3)2), 1.43 (s, 3H, JSiH = 12.9 Hz, CCH3), 4.19 (sep, 2H, JHH 6.9 Hz,

13 1 CH(CH3)2), 7.25 – 7.31 (m, 6H, m- and p-C6H5), 7.35 (d, 4H, o-C6H5); C{ H} NMR

(100 MHz, THF-d8, 25 °C): δ = 10.2 (s, CCH3), 23.7 (s, CH(CH3)2), 24.3 (s, CH(CH3)2),

36.2 (s, br, B2CCH3), 49.9 (s, CH(CH3)2), 128.0 (s, m-C6H5), 128.3 (s, p-C6H5), 132.7 (s,

11 1 o-C6H5), 140.0 (s, br, i-C6H5); B{ H} NMR (128 MHz, THF-d8, 25 °C): δ = 41.8 (s,

29 1 + br); Si{ H} NMR (79 MHz, THF-d8, 25 °C): δ = 7.0 (s); MS (EI , 70 eV): m/z(%):

450(100) [M]+, 435(96) [M - Me]+, 407(10) [M - iPr]+, 393(20) [M – Me – iPr + H]+;

11 35 HRMS for H27C20N2 B2 Cl3Si calcd. 450.1195, found 450.1199, for

11 35 37 H27C20N2 B2 Cl2 ClSi calcd. 452.1166, found 452.1157.

215

Preparation of the germanium and tin complexes

Bis(1,2-diisopropyl-3,5-diphenyl-1,2-diaza-3,5-diborolyl)germanium (3.2):

A solution of 2.6b (100 mg, 0.280 mmol) and GeCl2·dioxan (32 mg, 0.14 mmol)

in Et2O (20 mL) was stirred for 2 h at ambient temperature. Volatiles were removed

under vacuum and the product was extracted with pentane (20 mL). KCl was filtered off

and the red filtrate was dried under vacuum yielding 3.2 as an orange-red solid (90 mg,

91%). Red crystals were obtained by cooling a concentrated pentane solution of 3.2 to -

1 35 ºC for days. H NMR (400 MHz, C6D6, 25 °C): δ = 1.30 (s, br, 6H, CH(CH3)2), 1.32

(s, br, 6H, CH(CH3)2), 2.10 (s, 3H, CCH3), 4.37 (s, br, 2H, CH(CH3)2), 7.15-7.20 (m, 6H,

13 1 m- and p-C6H5), 7.34 (d, br, 4H, o-C6H5); C{ H} NMR (100 MHz, C6D6, 25 °C): δ =

12.9 (s, CCH3), 24.5 (s, br, CH(CH3)2), 25.8 (s, br, CH(CH3)2), 50.5 (s, CH(CH3)2), 110.9

(s, br, B2CCH3), 127.9 (s, p-C6H5), 128.0 (s, m-C6H5), 133.8 (s, o-C6H5), 141.1 (s, br, i-

11 1 C6H5); B{ H} NMR (128 MHz, C6D6, 25 °C): δ = 36.4 (s, br).

Bis(1,2-diisopropyl-3,5-dimethyl-1,2-diaza-3,5-diborolyl)tin (3.3a):

A solution of 2.4a (0.105 g, 0.525 mmol) and anhydrous SnCl2 (0.050 g, 0.262

mmol) in THF (30 mL) was stirred for 3 h at ambient temperature. Volatiles were

removed under vacuum and the red oily residue was extracted with hexane (30 mL).

Concentration and cooling of the hexane solution to -35 ºC yielded red crystals of 3.3a

1 (108 mg, 81.8%). H NMR (400 MHz, C6D6, 25 °C): δ = 0.77 (s, 6H, BCH3), 1.22 (d, 6H,

3 3 3 JHH = 6.9 Hz, CH(CH3)2), 1.31 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 2.71 (s, 3H, J119SnH =

1 11.3 Hz, CCH3), 3.87 (s, br, 2H, CH(CH3)2); H NMR (400 MHz, THF-d8, 25 °C): δ =

3 3 0.42 (s, 6H, BCH3), 1.27 (d, 6H, JHH = 6.8 Hz, CH(CH3)2), 1.34 (d, 6H, JHH = 6.8 Hz,

216

3 3 CH(CH3)2), 2.00 (s, 3H, J119SnH = 9.4 Hz, CCH3), 3.97 (s, 2H, JHH = 6.9 Hz, CH(CH3)2);

13 1 C{ H} NMR (100 MHz, C6D6, 25 °C): δ = -0.34 (s, br, BCH3), 11.7 (s, CCH3), 23.9 (s, br, CH(CH3)2), 26.4 (s, br, CH(CH3)2), 48.3 (s, CH(CH3)2), 104.4 (s, br, B2CCH3);

13 1 C{ H} NMR (100 MHz, THF-d8, -20 °C): δ = 2.39 (s, br, BCH3), 12.8 (s, CCH3), 23.3

11 1 (s, br, CH(CH3)2), 48.3 (s, br, CH(CH3)2), 78.8 (s, br, B2CCH3); B{ H} NMR (128

119 1 MHz, C6D6, 25 °C): δ = 33.8 (s, br); Sn{ H} NMR (148 MHz, THF-d8, 25°C): δ = -

1975.9 (s); MS (EI+, 70 eV): m/z(%): 310.2 (10) [SnL]+, 194.3 (38) [L]+; HRMS for

11 120 H23C10N2 B2 Sn calcd. 313.1069, found 313.1063; Anal. Calcd for H46C20N4B4Sn: C,

47.61; H, 9.19; N, 11.10. Found: C, 46.74; H, 9.10; N, 11.11.

Bis(1,2-diisopropyl-3,5-diphenyl-1,2-diaza-3,5-diborolyl)tin (3.3b):

A solution of 2.6b (0.30 g, 0.84 mmol) and anhydrous SnCl2 (0.080 g, 0.42

mmol) in THF (30 mL) was stirred for 2 h at ambient temperature. Volatiles were

removed under vacuum and the red oily residue was extracted with pentane (30 mL).

The red filtrate was dried under vacuum yielding 3.3b as a red crystalline solid (0.30 g,

95%). Red crystals were obtained by cooling a concentrated pentane solution of 3.3b to -

1 35 ºC. H NMR (400 MHz, C6D6, 25 °C): δ = 1.26 (s, br, 6H, CH(CH3)2), 1.32 (s, br, 6H,

CH(CH3)2), 2.44 (s, 3H, CCH3), 4.33 (s, br, 2H, CH(CH3)2), 7.12-7.19 (m, 6H, m- and p-

3 13 1 C6H5), 7.38 (d, 4H, JHH = 6.1 Hz, o-C6H5); C{ H} NMR (100 MHz, C6D6, 25 °C): δ =

11.5 (s, CCH3), 25.0 (s, br, CH(CH3)2), 26.2 (s, br, CH(CH3)2), 50.3 (s, CH(CH3)2), 112.0

(s, br, B2CCH3), 127.5 (s, p-C6H5), 128.6 (s, m-C6H5), 134.1 (s, o-C6H5), 141.0 (s, br, i-

11 1 C6H5); B{ H} NMR (128 MHz, C6D6, 25 °C): δ = 30.0 (s, br); HRMS for

11 120 H27C20N2 B2 Sn calcd. 437.1382, found 437.1419; Anal. Calcd for H54C40N4B4Sn: C,

63.82; H, 7.23; N, 7.44. Found: C, 63.93; H, 7.52; N, 7.49.

217

Bis(1,2-diisopropyl-diaza-3,5-dimethyl-diborolyl)tin(chloromethyl)chloride (3.4):

Red crystals of 3.3a (69 mg, 0.14 mmol) were dissolved in CH2Cl2 (3 mL) and the

dark red solution slowly turned to light yellow solution upon stirring at ambient

temperature overnight. Removal of the volatiles under vacuum yielded a thick light

yellow residue of the desired product (36 mg, 45%). Colorless crystals were obtained by

slow evaporation of a pentane / THF solution of 3.4 at ambient temperature. 1H NMR

3 (400 MHz, C6D6, 25 °C): δ = 0.78 (s, 6H, BCH3), 0.83 (s, 6H, BCH3), 1.16 (d, 6H, JHH =

3 3 6.9 Hz, CH(CH3)2), 1.19 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.26 (d, 6H, JHH = 6.9 Hz,

3 3 CH(CH3)2), 1.29 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.68 (s, 6H, J117SnH = 111 Hz,

3 2 3 J119SnH = 116 Hz, CCH3), 3.25 (s, 2H, JSnH = 11 Hz, SnCH2Cl), 3.79 (sep, 4H, JHH 6.9

13 1 Hz, CH(CH3)2); C{ H} NMR (100 MHz, C6D6, 25 °C): δ = 0.89 (s, br, BCH3), 12.5 (s,

2 JSnC = 48 Hz, CCH3), 22.6 (s, CH(CH3)2), 23.1 (s, CH(CH3)2), 23.3 (s, CH(CH3)2), 30.7

11 1 (s, SnCH2Cl), 47.0 (s, CH(CH3)2; B{ H} NMR (128 MHz, C6D6, 25 °C): δ = 42.6 (s,

119 1 + br); Sn{ H} NMR (148 MHz, C6D6, 25 °C): δ = 45.0 (s, br); MS (EI , 70 eV): m/z

+ + + + (%): 590(6) [M] , 575 (6) [M-Me] , 347 (26) [M-L-CH2Cl] , 312 (9) [M-L-CH2Cl2] ,

192 (50) [L-H]+.

1,2-Diisopropyl-3,5-diphenyl-1,2-diaza-3,5-diborolyltin chloride (3.5):

A dark red solution of 3.3b (100 mg, 0.133 mmol) and anhydrous SnCl2 (25 mg,

0.133 mmol) in THF (15 mL) was stirred for 24 h at ambient temperature. Volatiles were removed under vacuum forming a thick orange residue, which was washed twice with hexane (10 mL) and dried under vacuum affording 3.5 as a yellow solid (80 mg, 64%).

1 3 H NMR (400 MHz, C6D6, 25 °C): δ = 1.05 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.14 (s,

218

3 6H, JHH = 6.9 Hz, CH(CH3)2), 2.87 (s, 3H, CCH3), 4.06 (s, br, 2H, CH(CH3)2), 7.19 (t,

3 3 3 2H, JHH = 7.3 Hz, p-C6H5), 7.30 (t, 4H, JHH = 7.3 Hz, m-C6H5), 7.60 (d, 4H, JHH = 6.8

13 1 Hz, o-C6H5); C{ H} NMR (100 MHz, C6D6, 25 °C): δ = 11.5 (s, CCH3), 24.6 (s, br,

CH(CH3)2), 26.6 (s, br, CH(CH3)2), 49.5 (s, CH(CH3)2), 95.1 (s, br, B2CCH3), 127.3 (s, p-

11 1 C6H5), 127.4 (s, m-C6H5), 133.7 (s, o-C6H5), 138.9 (s, br, i-C6H5); B{ H} NMR (128

+ + MHz, C6D6,, 25 °C): δ = 36.0 (s, br); MS (EI , 70 eV): m/z (%): 472 (7) [M] .

1,2-Diisopropyl-4-methyl-3,5-dimethyl-1,2-diaza-3,5-diborolyltin tetrakis-

pentafluorophenylborate (3.6a):

A colorless solution of [H(Et2O)2]B(C6F5)4 (99 mg, 0.12 mmol) in CH2Cl2 (10 mL) was added to the dark red crystals of 3.3a (60 mg, 0.12 mmol). The mixture turned bright yellow upon stirring at ambient temperature for 1 h. Volatiles were removed under vacuum yielding a thick yellow residue, which was washed twice with pentane. The product was dried under full vacuum and isolated as a yellow solid (115 mg, 94.2%). 1H

3 NMR (400 MHz, CD2Cl2, 25 °C): δ = 0.81 (s, 6H, BCH3), 1.57 (d, 6H, JHH = 6.8 Hz,

3 CH(CH3)2), 1.63 (d, 6H, JHH = 6.8 Hz, CH(CH3)2), 2.05 (s, 3H, CCH3), 4.32 (sep, 2H,

3 13 1 JHH = 6.8 Hz, CH(CH3)2); C{ H} NMR (100 MHz, CD2Cl2, 25 °C): δ = -1.73 (s, br,

BCH3), 9.94 (s, CCH3), 23.7 (s, br, CH(CH3)2), 28.1 (s, br, CH(CH3)2), 51.1 (s,

1 1 CH(CH3)2), 124.5 (s, br, i-B(C6F5)4), 136.9 (d, JCF = 247 Hz, m-B(C6F5)4), 138.8 (d, JCF

1 11 1 = 246 Hz, p-B(C6F5)4), 148.8 (d, JCF = 240 Hz, o-B(C6F5)4); B{ H} NMR (128 MHz,

CD2Cl2, 25 °C): δ = 36.9 (s, br, BCH3), -17.4 (s, B(C6F5)4); ESI-MS (m/z, rel. intens.):

+ + - 312.7 (100) [SnL] , 193.0 (83) [L] , 678.4 (100) [B(C6F5)4] .

219

1,2-Diisopropyl-4-methyl-3,5-diphenyl-1,2-diaza-3,5-diborolyltin tetrakis-

pentafluorophenylborate (3.6b):

A colorless solution of [H(Et2O)2]B(C6F5)4 (198 mg, 0.239 mmol) in CH2Cl2 (10 mL) was added to the dark red crystals of 3.3b (180 mg, 0.239 mmol), and stirred for 18 h. Volatiles were removed under vacuum leaving behind a thick yellow residue, which was washed twice with hexane. The product was dried under full vacuum and isolated as

1 a yellow solid (230 mg, 86.5%). H NMR (400 MHz, CD2Cl2, 25 °C): δ = 1.53 (d, br, 6H,

3 CH(CH3)2), 1.61 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 2.11 (s, 3H, CCH3), 4.63 (s, br, 2H,

13 1 CH(CH3)2), 7.28-7.47 (m, 10H, o-, m- and p-C6H5); C{ H} NMR (100 MHz, CD2Cl2,

25 °C): δ = 10.5 (s, CCH3), 25.9 (s, br, CH(CH3)2), 53.0 (s, CH(CH3)2), 124.4 (s, br, i-

1 B(C6F5)4), 129.1 (s, m-C6H5), 129.9 (s, p-C6H5), 133.7 (s, o-C6H5), 136.9 (d, JCF = 246

1 Hz, m-B(C6F5)4), 138.8 (d, JCF = 240 Hz, p-B(C6F5)4), 142.3 (s, br, i-C6H5), 148.8 (d,

1 11 1 JCF = 241 Hz, o-B(C6F5)4); B{ H} NMR (128 MHz, CD2Cl2, 25 °C): δ = 33.5 (s, br,

+ BCH3), -17.4 (s, B(C6F5)4); ESI-MS (m/z, rel. intens.): 437.1 (35) [SnL] , 317.1 (100)

+ - [L] , 678.8 (100) [B(C6F5)4] .

220

8.4. Experimental Details for Chapter 4

Preparation of the zinc, cadmium, and mercury complexes

Bis(1,2-diisopropyl-3,5-dimethyl-1,2-diaza-3,5-diborolyl)zinc (4.1a):

A solution of 2.4a (0.150 g, 0.750 mmol) and anhydrous ZnCl2 (0.051 g, 0.375 mmol) in THF (30 mL) was stirred for 20 h at ambient temperature. The solvent was removed under vacuum and the residue was extracted with hexane (30 mL).

Concentration and cooling of the solution to -35 ºC yielded colorless crystals of 4.1a (110

1 3 mg, 65.1%). H NMR (400 MHz, C6D6, 25 ºC): δ = 0.73 (s, 6H, BCH3), 1.23 (d, 6H, JHH

3 = 6.9 Hz, CH(CH3)2), 1.30 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.76 (s, 3H, CCH3), 3.88

3 13 (sep, 2H, JHH 6.9 Hz, CH(CH3)2); C NMR (100 MHz, C6D6, 25 ºC): δ = -0.3 (s, br,

13 BCH3), 11.8 (s, CCH3), 23.5 (s, CH(CH3)2), 24.2 (s, CH(CH3)2), 47.9 (s, CH(CH3)2); C

NMR (100 MHz, THF-d8, -50 ºC): δ = 2.1 (s, br, BCH3), 13.5 (s, CCH3), 22.9 (s,

11 CH(CH3)2), 23.8 (s, CH(CH3)2), 48.0 (s, CH(CH3)2), 63.7 (s, br, B2CCH3); B NMR

o + + (128 MHz, C6D6, 25 C): δ = 37.6 (s, br); MS (EI , 70 eV): m/z(%): 450.3 (28) [ZnL2] ,

+ + 11 64 257.1 (23) [ZnL] , 194.2 (39) [L] ; HRMS for H46C20N4 B4 Zn calcd. 450.3386, found

450.3412.

Bis(1,2-diisopropyl-3,5-diphenyl-1,2-diaza-3,5-diborolyl)zinc (4.1b):

A solution of 2.6b (0.100 g, 0.280 mmol) and anhydrous ZnCl2 (0.019 g, 0.140

mmol) in THF (30 mL) was stirred for 3 h at ambient temperature forming a fine

suspension. Volatiles were removed in vacuum and the residue was extracted with

hexane (20 mL). Subsequent solvent removal left behind the product as a colorless

221

powder. Colorless block crystals of 4.1b were obtained by cooling a concentrated pentane

1 solution to -35 °C (57 mg, 58 %). H NMR (400 MHz, THF-d8, 25 °C): δ = 1.14 (d, 6H,

3 3 JHH = 6.8 Hz, CH(CH3)2), 1.27 (d, 6H, JHH = 6.8 Hz, CH(CH3)2), 1.53 (s, 3H, CCH3),

3 4.18 (sep, 2H, JHH = 6.8 Hz, CH(CH3)2), 7.22 - 7.36 (m, 6H, m- and p-C6H5), 7.36 - 7.38

13 1 (m, 4H, o-C6H5); C{ H} NMR (100 MHz, THF-d8, 25 °C): δ = 11.7 (s, CCH3), 23.8 (s,

CH(CH3)2), 24.6 (s, CH(CH3)2), 50.6 (s, CH(CH3)2), 69.3 (s, br, B2CCH3), 127.5 (s, p-

11 1 C6H5), 128.3 (s, m-C6H5), 134.0 (s, o-C6H5), 142.0 (s, br, i-C6H5); B{ H} NMR (128

MHz, C6D6, 25 °C): δ = 39.0 (s, br); ESI-MS (m/z, rel. intens., CH3CN): m/z(%):

+ + + 698.2(5) [ZnL2] , 381.0(100) [ZnL] , 319.1(50) [L + H] .

Bis(1,2-diisopropyl-3,5-dimethyl-1,2-diaza-3,5-diborolyl)cadmium (4.2a):

A solution of 2.4a (0.100 g, 0.500 mmol) and anhydrous CdBr2 (0.068 g, 0.250

mmol) in THF (30 mL) was sonicated for 30 min, forming a fine grey suspension. The

solvent was subsequently removed in vacuum and the oily residue was extracted with

hexane (20 mL). Solvent removal left behind 4.2a as a colorless crystalline solid (138

mg, 94 %). Colorless needles of 4.2a were obtained by cooling a concentrated pentane

1 4 solution to -35 °C. H NMR (400 MHz, C6D6, 25 °C): δ = 0.74 (s, 6H, JCdH = 8.7 Hz,

3 3 BCH3), 1.24 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.26 (d, 6H, JHH = 6.9 Hz, CH(CH3)2),

3 3 1 1.84 (s, 3H, JCdH = 59.6 Hz, CCH3), 3.85 (sep, 2H, JHH 6.9 Hz, CH(CH3)2); H NMR

4 3 (400 MHz, THF-d8, 25 °C): δ = 0.47 (s, 6H, JCdH = 7.7 Hz, BCH3), 1.18 (d, 6H, JHH =

3 3 6.8 Hz, CH(CH3)2), 1.27 (d, 6H, JHH = 6.8 Hz, CH(CH3)2), 1.50 (s, 3H, JCdH = 46.4 Hz,

3 13 1 CCH3), 3.77 (sep, 2H, JHH 6.8 Hz, CH(CH3)2); C{ H} NMR (100 MHz, C6D6, 25 °C):

2 δ = -0.26 (s, br, BCH3), 13.2 (s, JCdC = 39.7 Hz, CCH3), 23.4 (s, CH(CH3)2), 23.7 (s,

13 1 CH(CH3)2), 47.8 (s, CH(CH3)2), 77.6 (s, br, B2CCH3); C{ H} NMR (100 MHz, THF-d8,

222

2 D1 = 10 sec, LB = 3 Hz, 25 °C): δ = 1.44 (s, br, BCH3), 14.3 (s, JCdC = 43.1 Hz, CCH3),

23.2 (s, CH(CH3)2), 23.8 (s, CH(CH3)2), 48.3 (s, CH(CH3)2), 70.6 (s, br, B2CCH3);

11 1 11 1 B{ H} NMR (128 MHz, C6D6, 25 °C): δ = 37.9 (s, br); B{ H} NMR (128 MHz, THF-

113 1 d8, 25 °C): δ = 40.9 (s, br, LW1/2 = 486 Hz); Cd{ H} NMR (88 MHz, C6D6, 25 °C): δ =

113 1 + 256.2 (s); Cd{ H} NMR (66 MHz, THF-d8, 25 °C): δ = 397.9 (s); MS (EI , 70 eV):

+ + + m/z(%): 498.7(60) [CdL2] , 305.3(26) [CdL] , 192.3(63) [L] ; HRMS for

11 112 H46C20N4 B4 Cd calcd. 498.3122, found 498.3124.

Bis(1,2-diisopropyl-3,5-diphenyl-1,2-diaza-3,5-diborolyl)cadmium (4.2b):

A mixture of 2.6b (0.100 g, 0.280 mmol) and anhydrous CdBr2 (0.038 g, 0.140

mmol) in THF (30 mL) was sonicated for 30 min, forming a grey suspension. The

solvent was subsequently removed in vacuum and the residue was extracted with hexane

(20 mL). Solvent removal left behind 4.2b as a colorless powder. Pale yellow prismatic

crystals (84 mg, 80 %) of 4.2b were obtained by cooling a concentrated pentane solution

1 3 to -35 °C. H NMR (400 MHz, C6D6, 25 °C): δ = 1.13 (d, 6H, JHH = 6.8 Hz, CH(CH3)2),

3 3 1.17 (d, 6H, JHH = 6.8 Hz, CH(CH3)2), 1.87 (s, 3H, JCdH = 58.7 Hz, CCH3), 4.06 (sep,

3 2H, JHH 6.8 Hz, CH(CH3)2), 7.22 (t, 2H, p-C6H5), 7.31 (t, 4H, m-C6H5), 7.62 (d, 4H, o-

1 3 C6H5); H NMR (400 MHz, THF-d8, 25 °C): δ = 1.16 (d, 6H, JHH = 6.8 Hz, CH(CH3)2),

3 3 1.24 (d, 6H, JHH = 6.8 Hz, CH(CH3)2), 1.58 (s, 3H, JCdH = 58.7 Hz, CCH3), 4.16 (sep,

3 2H, JHH 6.8 Hz, CH(CH3)2), 7.23 - 7.29 (m, 6H, m- and p-C6H5), 7.40 - 7.43 (m, 4H, o-

13 1 2 C6H5); C{ H} NMR (100 MHz, THF-d8, 25 °C): δ = 13.2 (s, JCdC = 42.2 Hz, CCH3),

23.9 (s, CH(CH3)2), 24.3 (s, CH(CH3)2), 50.4 (s, CH(CH3)2), 77.4 (s, br, B2CCH3), 127.7

11 1 (s, p-C6H5), 128.4 (s, m-C6H5), 134.2 (s, o-C6H5), 142.2 (s, br, i-C6H5); B{ H} NMR

223

113 (128 MHz, THF-d8, 25 °C): δ = 36.6 (s, br); Cd NMR (88 MHz, C6D6, 25 °C): δ =

3 286.1 (sep, JCdH = 60.5 Hz); ESI-MS (m/z, rel. intens., CH3CN): m/z(%): 745.1(6) [CdL2

- H]+, 430.1(25) [CdL]+, 317.1(100) [L]+.

Bis(1,2-diisopropyl-3,5-dimethyl-1,2-diaza-3,5-diborolyl)mercury (4.3):

A solution of 2.4a (0.100 g, 0.500 mmol) and anhydrous HgCl2 (0.068 g, 0.250

mmol) in THF (30 mL) was stirred for 3 h at ambient temperature, forming a grey

suspension. The solvent was removed in vacuum and the residue was extracted with

hexane (30 mL). Removal of the solvent in vacuum left behind colorless, crystalline 4.3

(138 mg, 94 %). X-ray quality single crystals of 4.3 were obtained by slow evaporation

1 4 of a pentane solution at -35 °C. H NMR (400 MHz, C6D6, 25 °C): δ = 0.71 (s, 6H, JHgH

3 3 = 20.2 Hz, BCH3), 1.26 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.29 (d, 6H, JHH = 6.9 Hz,

3 3 CH(CH3)2), 1.84 (s, 3H, JHgH = 142.5 Hz, CCH3), 3.82 (sep, 2H, JHH = 6.9 Hz,

1 4 CH(CH3)2); H NMR (400 MHz, THF-d8, 25 °C): δ = 0.45 (s, 6H, JHgH = 19.8 Hz,

3 3 BCH3), 1.29 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.32 (d, 6H, JHH = 6.9 Hz, CH(CH3)2),

3 3 13 1 1.59 (s, 3H, JHgH = 143.4 Hz, CCH3), 3.96 (sep, 2H, JHH = 6.9 Hz, CH(CH3)2); C{ H}

2 NMR (100 MHz, C6D6, 25 °C): δ = -0.13 (s, br, BCH3), 13.6 (s, JHgC = 64.5 Hz, CCH3),

23.4 (s, CH(CH3)2), 23.8 (s, CH(CH3)2), 47.9 (s, CH(CH3)2), 77.8 (s, br, B2CCH3);

13 1 C{ H} NMR (100 MHz, THF-d8, -20 °C): δ = -0.35 (s, br, BCH3), 13.5 (s, CCH3), 23.4

11 1 (s, CH(CH3)2), 23.8 (s, CH(CH3)2), 48.2 (s, br, CH(CH3)2), 77.3 (s, br, B2CCH3); B{ H}

11 1 NMR (128 MHz, C6D6, 25 °C): δ = 39.0 (s, br); B{ H} NMR (128 MHz, THF-d8, 25

199 1 + °C): δ = 38.9 (s, br); Hg{ H} NMR (71 MHz, C6D6, 25 °C): δ = -2063.3 (s); MS (EI ,

+ + + 70 eV): m/z(%): 586.5(21) [HgL2] , 393.3(1) [HgL] , 193.3(20) [L] .

224

8.5. Experimental Details for Chapter 5

Preparation of the rhodium, iron, ruthenium and zirconium complexes

1,2-Diisopropyl-diaza-3,5-dimethyl-diborolyl rhodium cylcooctadiene (5.1a):

A solution of 2.4a (100 mg, 0.500 mmol) and [Rh(cod)Cl]2 (123 mg, 0.250 mmol) in THF (20 mL) was stirred for 4 h at ambient temperature producing a dark brown solution. Volatiles were removed under vacuum and the residue was extracted with hexane (20 mL). The LiCl salt was filtered off and the product was dried under vacuum

1 yielding a brown solid of 5.1a (120 mg, 57%). H NMR (400 MHz, THF-d8, 25 °C): δ =

3 0.51 (s, 6H, BCH3), 0.89 (s, 3H, CCH3), 1.29 (d, 6H, JHH = 6.7 Hz, CH(CH3)2), 1.64 (d,

3 6H, JHH = 6.7 Hz, CH(CH3)2), 1.84 (q, br, 4H, CH2), 2.18 (m, br, 4H, CH2), 3.53 (s, br,

3 13 1 4H, CH), 3.80 (sep, 2H, JHH 6.7 Hz, CH(CH3)2); C{ H} NMR (100 MHz, THF-d8, 25

°C): δ = -0.26 (s, br, BCH3), 10.5 (s, CCH3), 23.4 (s, CH(CH3)2), 26.2 (s, CH(CH3)2),

11 1 32.4 (s, CH2), 48.7 (s, CH(CH3)2), 72.7 (s, br, CH), 81.3 (s, br, B2CCH3); B{ H} NMR

(128 MHz, THF-d8, 25 °C): δ = 26.5 (s, br).

1,2-Diisopropyl-diaza-3,5-diphenyl-diborolyl rhodium cylcooctadiene (5.1b):

A solution of 2.6b (217 mg, 0.608 mmol) and [Rh(cod)Cl]2 (150 mg, 0.304

mmol) in THF (20 mL) was stirred for 3 h at ambient temperature forming a dark brown

solution. Volatiles were removed under vacuum and the residue was extracted with hexane (20 mL). The KCl salt was filtered off and the product was dried under vacuum and recrystallized from hexane yielding an orange-brown solid of 5.1b (199 mg, 62%).

Thin plate yellow crystals were obtained by cooling the concentrated pentane solution of

225

1 5.1b at -35 ºC. H NMR (400 MHz, C6D6, 25 ºC): δ = 1.10 (s, 3H, CCH3), 1.15 (d, 6H,

3 3 JHH = 6.3 Hz, CH(CH3)2), 1.24 (d, 6H, JHH = 6.9 Hz, CH(CH3)2), 1.93 (d, br, 4H, CH2),

3 2.43 (s, br, 4H, CH2), 3.23 (sep, 2H, JHH = 6.8 Hz, CH(CH3)2), 3.86 (s, br, 4H, CH), 7.24

13 1 (t, 2H, p-C6H5), 7.37 (t, 4H, m-C6H5), 7.85 (d, 4H, o-C6H5); C{ H} NMR (100 MHz,

CD2Cl2 25 ºC): δ = 9.8 (s, CCH3), 25.3 (s, CH(CH3)2), 26.8 (s, CH(CH3)2), 32.1 (s, br,

CH), 32.2 (s, CH2), 49.5 (s, CH(CH3)2), 82.6 (s, br, B2CCH3), 126.8 (s, p-C6H5), 127.7 (s,

11 1 m-C6H5), 134.2 (s, o-C6H5), 141.7 (s, br, i-C6H5); B{ H} NMR (128 MHz, C6D6, 25

ºC): δ = 28.9 (s, br); MS (EI+, 70 eV): m/z(%): 528.3(94) [Rh(cod)L]+, 418.1(85) [RhL]+,

+ 11 317.2(57) [L] ; HRMS for H39C28N2Rh B2 calcd. 528.2354, found 528.2387; Anal.

Calcd for H39C28N2RhB2: C, 63.67; H, 7.44; N, 5.30. Found: C, 63.76; H, 7.27; N, 5.39.

1,2-Cyclopropyl-diaza-3,5-diphenyl-diborolyl rhodium cyclooctadiene (5.1e):

A solution of 2.4e (0.10 g, 0.36 mmol) and [Rh(cod)Cl]2 (0.088g, 0.18 mmol) in

THF (20 mL) was stirred for 2 h at ambient temperature forming a dark red solution.

Volatiles were removed under vacuum and the dark red residue was extracted with

pentane (20 mL) and diethyl ether (5 mL). The LiCl salt was filtered off and the product

was dried under vacuum and recrystallization in hexane yielding an orange-yellow solid

of 5.1e (77 mg, 45%). Yellow prismatic crystals of 5.1e were obtained by slow

evaporation of hexane and THF mixture of 5.1e at ambient temperature. 1H NMR (400

MHz, C6D6, 25 ºC): δ = 1.36 (m, 2H, NCH2CH2), 1.71 (s, 3H, CCH3), 1.80 (m, 4H, CH2),

2.18 (m, 4H, CH2), 2.85 (m, 2H, NCH2CH2), 3.19 (m, 2H, NCH2CH2), 3.52 (s, br, 4H,

13 1 CH), 7.30 (t, 2H, p-C6H5), 7.43 (t, 4H, m-C6H5), 7.88 (d, 4H, o-C6H5); C{ H} NMR

(100 MHz, C6D6, 25 ºC): δ = 12.0 (s, CCH3), 29.9 (s, NCH2CH2), 32.3 (s, CH2), 44.9 (s,

1 NCH2CH2), 73.7 (d, JRhC = 13.6 Hz, CH), 82.6 (s, br, B2CCH3), 128.2 (s, p-C6H5), 128.4

226

11 1 (s, m-C6H5), 134.2 (s, o-C6H5), 139.0 (s, br, i-C6H5); B{ H} NMR (128 MHz, C6D6, 25

ºC): δ = 23.8 (s, br); MS (EI+, 70 eV): m/z (%): 484 (100) [Rh(cod)L]+, 376 (11) [RhL]+,

+ 11 274 (24) [L+H] ; HRMS for H31C25N2Rh B2 calcd. 484.17284, found 484.17196; Anal.

Calcd for H31C25N2RhB2: C, 62.03; H, 6.46; N, 5.79. Found: C, 62.09; H, 6.39; N, 5.77.

Bis(1,2-cyclopropyl-diaza-3,5-diphenyl-diborolyl)iron (5.2):

THF (15 mL) was added to a mixture of 2.4e (0.10 g, 0.36 mmol) and FeCl2(thf)2

(0.048 g, 0.18 mmol) at -78 ºC. The red mixture was slowly warmed to ambient temperature and was stirred for 2 h affording a dark brown solution. Volatiles were removed under full vacuum yielding dark brown residue, which was extracted with hexane (15 mL) and THF (5 mL) forming a suspension. The LiCl was filtered off and the dark brown filtrate was dried under vacuum. The product was obtained as a reddish- brown solid (0.095 g, 89%). Red thin plate crystals of 5.2 were obtained by slow evaporation of a hexane solution of 5.2 at ambient temperature. 1H NMR (400 MHz,

CD2Cl2, 25 ºC): δ = 1.55 (s, 3H, CCH3), 2.20 (m, 1H, NCH2CH2), 2.36 (m, 1H,

NCH2CH2), 2.94 (q, br, 2H, NCH2CH2), 3.75 (t, br, 2H, NCH2CH2), 7.33 – 7.37 (m, 6H,

13 1 m- and p-C6H5), 7.68 (d, br, 4H, o-C6H5); C{ H} NMR (100 MHz, CD2Cl2, 25 ºC): δ =

15.5 (s, CCH3), 27.7 (s, NCH2CH2), 46.2 (s, NCH2CH2), 65.5 (s, br, B2CCH3), 128.0 (s,

11 1 m- and p-C6H5), 134.9 (s, o-C6H5), 137.9 (s, br, i-C6H5); B{ H} NMR (128 MHz,

+ + CD2Cl2, 25 ºC): δ = 13.6 (s, br); MS (EI , 70 eV): m/z (%): 602 (28) [FeL2] , 328 (10)

+ + 11 [FeL] , 274 (100) [L] ; HRMS for H38C34N4 B4Fe calcd. 602.28181, found 602.28500;

Anal. Calcd for H38C34N4FeB4: C, 67.86; H, 6.36; N, 9.31. Found: C, 67.48; H, 6.48; N,

9.26.

227

1,2-Cyclopropyl-diaza-3,5-diphenyl-diborolyl pentamethylcyclopentadienyl

ruthenium (5.3):

A solution of 2.4e (0.10 g, 0.36 mmol) and [Cp*RuCl]4 (0.093 g, 0.089 mmol) in

THF (20 mL) was stirred for 2 h at -78 °C for 1 h. The mixture was slowly warmed to

ambient temperature and stirred for an additional 30 min affording a dark brown solution.

Volatiles were removed under vacuum and the dark brown solid was extracted with

hexane (20 mL). The LiCl salt was filtered off and the product was dried under vacuum.

The product was obtained as a light yellow solid (0.15 g, 80%). Yellow block crystals of

5.3 were obtained by slow evaporation of a hexane solution of 5.3 at ambient

1 temperature. H NMR (400 MHz, C6D6, 25 ºC): δ = 1.34 (m, 1H, NCH2CH2), 1.51 (s,

15H, C5(CH3)5), 1.91 (m, 1H, NCH2CH2), 2.36 (s, 3H, CCH3), 2.55 (m, 2H, NCH2CH2),

3.20 (m, 2H, NCH2CH2), 7.29 (t, 2H, p-C6H5), 7.43 (t, 4H, m-C6H5), 7.78 (d, 4H, o-

13 1 C6H5); C{ H} NMR (100 MHz, C6D6, 25 ºC): δ = 11.0 (s, C5(CH3)5), 15.9 (s, CCH3),

26.5 (s, NCH2CH2), 47.0 (s, NCH2CH2), 80.0 (s, br, B2CCH3), 81.6 (s, C5(CH3)5), 127.6

11 1 (s, p-C6H5), 128.1 (s, m-C6H5), 134.7 (s, o-C6H5), 139.1 (s, br, i-C6H5); B{ H} NMR

+ (128 MHz, C6D6, 25 ºC): δ = 14.7 (s, br); MS (EI , 70 eV): m/z (%): 509 (100)

+ 11 [Cp*RuL] ; HRMS for H34C27N2Ru B2 calcd. 510.19516, found 510.19853; Anal. Calcd

for H34C27N2RuB2: C, 63.68; H, 6.73; N, 5.50. Found: C, 63.81; H, 6.80; N, 5.50.

1,2-Cyclopropyl-diaza-3,5-diphenyl-diborolyl pentamethylcyclopentadienyl

zirconium dichloride (5.4):

A solution of 2.4e (0.10 g, 0.36 mmol) and Cp*ZrCl3 (0.12 g, 0.36 mmol) in THF

(20 mL) was stirred for 2 h at -78 °C affording an orange solution. The mixture was

228

slowly warmed to room temperature and stirred for an additional 30 min. Volatiles were

removed under vacuum and the orange residue was extracted with benzene (15 mL). The

LiCl salt was filtered off and the product was dried under vacuum yielded a thick orange-

yellow residue of 5.4 (0.13 g, 32%). Colorless block crystals of 5.4 were obtained by

slow evaporation of the hexane and benzene solution of 5.4 at ambient temperature. 1H

NMR (400 MHz, C6D6, 25 ºC): δ = 1.42 (m, 2H, NCH2CH2), 1.88 (s, 15H, C5(CH3)5),

2.30 (s, 3H, CCH3), 3.22 (m, 2H, NCH2CH2), 3.76 (m, 2H, NCH2CH2), 7.27 (t, 2H, p-

13 1 C6H5), 7.33 (t, 4H, m-C6H5), 7.82 (d, 4H, o-C6H5); C{ H} NMR (100 MHz, C6D6, 25

ºC): δ = 12.8 (s, C5(CH3)5), 16.6 (s, CCH3), 31.9 (s, NCH2CH2), 46.7 (s, NCH2CH2),

108.5 (s, br, B2CCH3), 125.6 (s, C5(CH3)5), 128.2 (s, m-C6H5), 129.8 (s, p-C6H5), 135.9

11 1 (s, o-C6H5), 136.6 (s, br, i-C6H5); B{ H} NMR (128 MHz, C6D6, 25 ºC): δ = 42.6 (s,

+ + + br); MS (EI , 70 eV): m/z (%): 568 (0.2) [LCp*ZCl2] , 274 (100) [L+H] ; HRMS for

11 35 90 H34C27N2 B2 Cl2 Zr calcd. 568.13322, found 568.13032.

Preparation of 1,2-diaza-4-oxa-3,5-diborolidines and lithium tungsten propynyl

pentacarbonyl complexes

1,2-Cyclopropyl-3,5-diphenyl-4-oxa-1,2-diaza--3,5-diborolidine (5.6e) and

lithium tungsten propynyl pentacarbonyl (5.5):

THF (10 mL) was added to a mixture of 2.4e (100 mg, 0.357 mmol) and W(CO)6

(125 mg, 0.357 mmol) at -78 C. The mixture was stirred at ambient temperature for 2 h affording an orange solution. Volatiles were removed under full vacuum and hexane (20 mL) was added to the orange residue, upon sonication of mixture for 30 min a yellow- orange suspension was obtained. The yellow solid of 5.5 was filtered and dried under

229

vacuum (99 mg, 75%). The orange filtrate was dried under vacuum yielding a creamy

1 tan solid of 5.6e (67 mg, 72%). Analysis of 5.6e: H NMR (400 MHz, C6D6, 25 ºC): δ =

3 3 1.62 (quint, 2H, JHH = 6.8 Hz, NCH2CH2), 3.16 (t, 4H, JHH = 6.8 Hz, NCH2CH2), 7.31 –

13 1 7.39 (m, 6H, m- and p-C6H5), 8.03 (d, 4H, o-C6H5); C{ H} NMR (100 MHz, C6D6, 25

ºC): δ = 32.0 (s, NCH2CH2), 43.0 (s, NCH2CH2), 128.2 (s, m-C6H5), 130.1 (s, p-C6H5),

11 1 134.3 (s, o-C6H5), 134.9 (s, br, i-C6H5); B{ H} NMR (128 MHz, C6D6, 25 ºC): δ = 26.3

(s, br); MS (EI+, 70 eV): m/z (%): 262 (100) [M]+. Analysis of 5.5: 1H NMR (400 MHz,

13 1 THF-d8, 25 ºC): δ = 1.65 (s, 3H, CH3); C{ H} NMR (100 MHz, THF-d8, 25 ºC): δ =

1 2 7.08 (s, WC≡CCH3), 92.1 (s, JWC = 93.6 Hz, WC≡CCH3), 102.6 ppm (s, JWC = 23.9 Hz,

1 1 WC≡CCH3), 201.6 (s, JWC = 123.6 Hz, WC≡O), 203.2 (s, JWC = 136.7 Hz, WC≡O

trans); Anal. Calcd for H20C27LiO10W: C, 38.86; H, 4.40. Found: C, 39.32; H, 4.77.

1,2,3,5-Tetraphenyl-4-oxa-1,2-diaza--3,5-diborolidine (5.6c):

THF (10 mL) was added to a mixture of 2.4c (0.10 g, 0.26 mmol) and W(CO)6

(0.092g, 0.26 mmol) at -78 C. The mixture was stirred at ambient temperature for 2 h affording an orange-yellow solution. Volatiles were removed under full vacuum and hexane (20 mL) was added to the orange residue; upon sonication of mixture for 30 min a yellow-orange suspension was obtained. The yellow solid of 5.5 was filtered off. The yellow filtrate was dried under vacuum yielding a light brown solid of 5.6c (65 mg,

67%). Colorless crystals were obtained by cooling a concentrated hexane solution of 5.6c

1 at -35 °C. H NMR (400 MHz, THF-d8, 25 ºC): δ = 7.12 – 7.33 (m, 6H, m- and p-C6H5),

13 1 7.86 (d, 4H, o-C6H5); C{ H} NMR (100 MHz, THF-d8, 25 ºC): δ = 128.0 (s, p-NC6H5),

128.4 (s, m-NC6H5), 129.6, (s, o-NC6H5), 130.1 (s, m-BC6H5), 130.9 (s, p-BC6H5), 135.12

230

11 1 (s, o-BC6H5), 140.2 (s, i-NC6H5); B{ H} NMR (128 MHz, THF-d8, 25 ºC): δ = 29.0 (s,

+ + 11 br); MS (EI , 70 eV): m/z (%): 374 (100) [M] ; HRMS for H20C24N2 B2O calcd.

374.17617, found 374.17584.

Preparation of the heterobicyclic 1,5-diaza-2,4,6,8-tetraborolidine, its mono-

and dipotassium complexes and the triple-decker ruthenocene with C2B4N2 ring framework

2,4,6,8-Tetraphenyl-3,7-dimethyl-1,5-diaza-2,4,6,8-tetraborolidine (5.7):

A colorless solution of KHMDS (1.17 g, 5.86 mmol) in THF (20 mL) was slowly added to a stirred white suspension of hydrazine dihydrochloride salt (0.308 g, 2.93 mmol) in THF (20 mL). The mixture was refluxed until hydrazine dihydrochloride salt dissolved in solution affording a light yellow suspension. Triethylamine (1.19 g, 11.7 mmol) was added to the mixture and stirred for 30 min. A solution of 2.1b (1.61 g, 5.86 mmol) in hexane (30 mL) was slowly added to the suspension, which immediately yielded a yellow suspension. The mixture was stirred at room temperature for 3 h and the white precipitate was filtered off. Volatiles were removed under full vacuum and a colorless solid was obtained. The desired product was washed a few times with pentane

(25 mL) and a colorless solid of 5.7 was isolated as mixture of cis and trans-isomers

(0.660 g, 52%). Colorless crystals of the trans-isomer of 5.7 were obtained by the recrystallization of the solid in a pentane and THF mixture at -35 °C. trans-5.7: 1H NMR

3 3 (400 MHz, C6D6, 25 °C): δ = 1.19 (d, 3H, JHH = 7.8 Hz, HCCH3), 1.80 (q, 1H, JHH = 7.6

3 Hz, HCCH3), 7.03 – 7.09 (m, 6H, m- and p-C6H5), 7.30 (d, 4H, JHH = 6.6 Hz, o-C6H5);

13 C NMR (100 MHz, C6D6, 25 °C): δ = 10.2 (s, HCCH3), 31.8 (s, br, HCCH3), 127.6 (s,

231

1 p-C6H5), 127.7 (s, m-C6H5), 134.0 (s, o-C6H5), 137.1 (s, br, i-C6H5); cis-5.7: H NMR

(400 MHz, C6D6, 25 °C): δ = 1.38 (s, 3H, HCCH3), 1.92 (s, br, 1H, HCCH3), 7.03 – 7.09

3 13 (m, 6H, m- and p-C6H5), 7.25 (d, 4H, JHH = 6.6 Hz, o-C6H5); C NMR (100 MHz, C6D6,

25 °C): δ = 9.5 (s, HCCH3), 34.3 (s, br, HCCH3), 129.3 (s, p-C6H5), 129.7 (s, m-C6H5),

11 134.9 (s, o-C6H5), 137.5 (s, br, i-C6H5); 5.7: B NMR (128 MHz, THF-d8, 25 °C): δ =

+ + 11 54.1 (s, br); MS (EI , 70 eV): m/z (%): 435 (100) [M] ; HRMS for H28C28N2 B4 calcd.

436.26247, found 436.26376; Anal. Calcd for H28C28N2B4: C, 77.17; H, 6.48; N, 6.43.

Found: C, 76.22; H, 6.93; N, 6.45.

Monopotassium salt of 1,5-diaza-2,4,6,8-tetraborolyl (5.8):

A solution of 5.7 (0.20 g, 0.46 mmol) in THF (10 mL) was added to a stirred solution of KHMDS (0.091 g, 0.46 mmol) in THF (15 mL) afforded a bright yellow solution. The mixture was stirred at ambient temperature for 18 h and the volatiles were removed under vacuum yielding a yellow solid, which was washed with hexane (30 mL) and dried under vacuum. The product was obtained as a light yellow solid (0.21 g, 97%).

1 3 H NMR (400 MHz, THF-d8, 25 °C): δ = 0.91 (d, 3H, JHH = 8.2 Hz, HCCH3), 1.69 (q,

3 1H, JHH = 8.4 Hz, HCCH3), 1.91 (s, 3H, CCH3), 6.84 – 6.93 (m, 12H, m- and p-C6H5),

13 7.17 – 7.22 (m, 8H, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 12.0 (s, HCCH3),

14.6 (s, CCH3), 26.3 (s, br, HCCH3), 108.2 (s, br, B2CCH3), 125.3, 127.2 (s, p-C6H5),

11 126.8, 126.9 (s, m-C6H5), 135.2, 135.3 (s, o-C6H5), 141.9, 146.0 (s, br, i-C6H5); B NMR

(128 MHz, THF-d8, 25 °C): δ = 41.3 (s, br).

232

Dipotassium salt of 1,5-diaza-2,4,6,8-tetraborolyl (5.9):

THF (10 mL) was added to a mixture of 5.7 (0.125 g, 0.287 mmol) and KHMDS

(0.114 g, 0.574 mmol) at -78 ºC affording a bright orange solution. The mixture was

stirred at ambient temperature for 18 h and the volatiles were removed under vacuum

yielding an orange solid, which was washed with hexane (30 mL) and then with toluene

(10 mL) and dried under vacuum. The product was obtained as a light orange powder

(0.120 g, 82%). Orange prismatic crystals of 5.9 that were suitable for X-ray diffraction analysis were obtained by slow evaporation of a benzene and THF solution of 5.9 at

1 ambient temperature. H NMR (400 MHz, THF-d8, 25 °C): δ = 1.91 (s, 3H, CCH3), 6.75

13 – 6.78 (m, 12H, m- & p-C6H5), 7.20 – 7.22 (m, 8H, o-C6H5); C NMR (100 MHz, THF- d8, 25 °C): δ = 15.5 (s, CCH3), 101.5 (s, br, B2CCH3), 124.5 (s, p-C6H5), 126.2 (s, m-

11 C6H5), 135.9 (s, o-C6H5), 147.0 (s, br, i-C6H5); B NMR (128 MHz, THF-d8, 25 °C): δ =

39.8 (s, br, LW1/2 = 948 Hz).

Triple-decker ruthenocene, Cp*Ru(Me2Ph4C2B4N2)RuCp* (5.10):

THF (15 mL) was added to a mixture of 5.9 (0.24 g, 0.46 mmol) and [Cp*RuCl]4

(0.24 g, 0.23 mmol) at -78 ºC. The dark red mixture was stirred for 2 h at -78 °C and then slowly warmed to ambient temperature affording a dark brown solution. Volatiles were removed in vacuum yielding a dark brown residue, which was extracted with hexane (10 mL) and benzene (10 mL) forming a suspension. The KCl was filtered off and the dark brown filtrate was dried under vacuum. The product was obtained as a brown solid. Dark brown prismatic crystals of 5.10 (52 mg, 25 %) was obtained by slow evaporation of a hexane solution of 5.10 at ambient temperature. 1H NMR (400 MHz,

3 C6D6, 25 °C): δ = 1.19 (s, 15H, C5(CH3)5), 2.65 (s, 3H, CCH3), 7.39 (t, 2H, JHH = 7.3 Hz,

233

3 3 1 p-C6H5), 7.51 (t, 4H, JHH = 7.3 Hz, m-C6H5), 8.24 (d, 4H, JHH = 7.3 Hz, o-C6H5); H

NMR (400 MHz, CD2Cl2, 25 °C): δ = 1.24 (s, 15H, C5(CH3)5), 2.49 (s, 3H, CCH3), 7.37

13 (m, 6H, m- and p-C6H5), 7.98 (d, 4H, o-C6H5); C NMR (100 MHz, C6D6, 25 °C): δ =

10.5 (s, C5(CH3)5), 23.6 (s, CCH3), 88.4 (s, C5(CH3)5), 127.6 (s, p-C6H5), 128.7 (s, m-

13 C6H5), 137.6 (s, o-C6H5), 141.2 (s, br, i-C6H5); C NMR (100 MHz, CD2Cl2, 25 °C): δ =

10.4 (s, C5(CH3)5), 23.1 (s, CCH3), 88.4 (s, C5(CH3)5), 127.2 (s, m-C6H5), 128.2 (s, p-

11 C6H5), 137.3 (s, o-C6H5), 142.2 (s, br, i-C6H5); B NMR (128 MHz, C6D6, 25 °C): δ =

11 + 23.2 (s, br); B NMR (128 MHz, CD2Cl2, 25 °C): δ = 24.1 (s, br); MS (EI , 70 eV): m/z

+ 11 102 (%): 907 (68) [M] ; HRMS for H48C55N2 B4 Ru2 calcd. 907.28244, found 907.28371;

Anal. Calcd for H51C64N2B4Ru2: C, 64.45; H, 6.79; N, 2.95. Found: C, 64.63; H, 6.56; N,

3.26.

Cyclic voltammetry of 5.2, 5.3 and 5.10:

The measurements were performed in the absence of oxygen, in anhydrous

tetrahydrofuran, dichloromethane and dimethoxymethane, at an analyte concentration of

1 mM, and in the presence of 0.1 M [nBu4N]PF6 as a supporting electrolyte. A PARstat

2273 Revision 2273 potentiostat was used at scan rates between 50 and 1000 mVs-1. The

three-electrode cell had a platinum disk working electrode, a platinum wire auxiliary

0/+1 0 electrode, and a silver wire pseudo-reference electrode. [Cp2Co] with E = -1.36 V verse ferrocene and -0.80 V versus a saturated calomel electrode in THF (E0 = -1.33 and -

0.87 V in CH2Cl2, -1.31 and -0.80 V in DME, respectively) was used as an internal

standard.

234

8.6. Experimental Details for Chapter 6

Preparation of the alkali metal, rhodium and borane adducts of the 1,2,4-triaza-

3,5-diborolyl ligand

1-Methyl-3,5-diphenyl-4-methylamino-1,2,4-triaza-3,5-diborolidine (1.63b):58b

1,2,4-Triaza-3,5-diborolidine 1.63b was prepared according to the literature procedure. The product was obtained as a colorless solid (75 % yield). 1H NMR (400

3 MHz, THF-d8, 25 °C): δ = 2.45 (d, 3H, JHH = 6.2 Hz, HNCH3), 3.21 (s, 3H, BNCH3),

3 4.14 (q, 1H, JHH = 6.2 Hz, HNCH3), 7.27-7.36 (m, 6H, C6H5), 7.49 (s, br, HNB), 7.60 (d,

13 2H, o-C6H5), 7.93 (d, 2H, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 35.6 (s,

BNCH3), 43.3 (s, HNCH3), 128.3 (s, m-C6H5), 128.7 (s, m-C6H5), 129.0 (s, p-C6H5),

11 129.3 (s, p-C6H5), 134.3 (s, o-C6H5); B NMR (128 MHz, THF-d8, 25 °C): δ = 27.5 (s,

58b br, LW1/2 = 330 Hz). The NMR data was in good agreement with the reported values.

1-Methyl-3,5-diphenyl-4-methylamino-1,2,4-triaza-3,5-diborolyllithium (6.1a):

A yellow solution of LiTMP was prepared from nBuLi in hexane (3.11 mL, 1.6

M, 4.99 mmol) and TMP (0.704 g, 4.99 mmol) in THF (3 mL). Solutions of 1.63b (1.32 g, 4.99 mmol) in THF (3 mL) and LiTMP were pre-cooled to –35 °C, mixed and kept at this temperature for 2 h. The resulting yellow solution was allowed to warm to room temperature overnight, and the solvent was removed under vacuum, leaving behind a yellow residue that was washed twice with hexane (30 mL) and dried under vacuum. The product was obtained as a colorless powder (1.24 g, 92.1 %). Colorless crystals were obtained by slow diffusion of hexane into a THF solution of 6.1a. 1H NMR (400 MHz,

235

3 THF-d8, 25 °C): δ = 2.36 (d, 3H, JHH = 6.4 Hz, HNCH3), 3.22 (s, 3H, BNCH3), 3.75 (q,

3 1H, JHH = 6.4 Hz, HNCH3), 7.05-7.23 (m, 6H, m- and p-C6H5), 7.60 (d, 2H, o-C6H5),

13 7.77 (d, 2H, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 39.0 (s, BNCH3), 44.0 (s,

HNCH3), 125.9 (s, p-C6H5), 126.7 (s, p-C6H5), 127.5 (s, m-C6H5), 127.9 (s, m-C6H5),

11 134.5 (s, o-C6H5), 134.6 (s, o-C6H5), 139.9 (s, br, i-C6H5), 144.0 (s, br, i-C6H5); B NMR

7 (128 MHz, THF-d8, 25 °C): δ = 24.2 (s, br), 27.8 (s, br); Li NMR (155 MHz, THF-d8, 25

°C): δ = -1.24 (s); MS (EI+, 70 eV): m/z(%): 264(36) [M–Li+H]+, 179.2(100) [M–Li–

Me+H]+.

1-Methyl-3,5-diphenyl-4-methylamino-1,2,4-triaza-3,5-diboroly sodium (6.1b):

A solution of 1.63b (340 mg, 1.28 mmol) in THF (20 mL) was added to a stirred solution of NaHMDS (236 mg, 1.28 mmol) in THF (15 mL). The resulting yellow mixture was stirred at ambient temperature for 3 h to ensure completion of the reaction.

The solvent was removed under vacuum leaving behind a light yellow residue that was washed twice with hexane (30 mL) and dried under vacuum. The product was obtained as a colorless powder (340 mg, 92.4 %). Colorless crystals were obtained by slow

1 diffusion of hexane into a THF solution of 6.1b. H NMR (400 MHz, THF-d8, 25 °C): δ

3 3 = 2.35 (d, 3H, JHH = 6.4 Hz, HNCH3), 3.25 (s, 3H, BNCH3), 3.80 (q, 1H, JHH = 6.4 Hz,

3 HNCH3), 7.08-7.28 (m, 6H, m- and p-C6H5), 7.56 (d, 2H, JHH = 6.6 Hz, o-C6H5), 7.83 (d,

3 13 2H, JHH = 6.6 Hz, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 39.5 (s, BNCH3),

44.0 (s, HNCH3), 126.6 (s, p-C6H5), 127.1 (s, p-C6H5), 128.1 (s, m-C6H5), 128.2 (s, m-

C6H5), 133.9 (s, o-C6H5), 134.4 (s, o-C6H5), 139.3 (s, br, i-C6H5), 142.7 (s, br, i-C6H5);

11 B NMR (128 MHz, THF-d8, 25 °C): δ = 24.7 (s, br), 27.4 (s, br).

236

1-Methyl-3,5-diphenyl-4-methylamino-1,2,4-triaza-3,5-diborolylpotassium (6.1c):

A solution of 1.63b (1.50 g, 5.68 mmol) in THF (30 mL) was added slowly under stirring to a suspension of potassium hydride (0.228 g, 5.68 mmol) in THF (20 mL). The hydride dissolved with concomitant evolution of a gas, yielding a yellow solution. The mixture was stirred at ambient temperature for 3 h to ensure completion of the reaction and the solvent was subsequently removed under vacuum. The yellow residue was washed twice with hexane (30 mL) and dried under vacuum, yielding a light yellow powder (1.68 g, 98 %). Yellow crystals were obtained by slow diffusion of hexane into a

1 3 THF solution of 6.1c. H NMR (400 MHz, THF-d8, 25 °C): δ = 2.40 (d, 3H, JHH = 6.4

3 Hz, HNCH3), 3.27 (s, 3H, BNCH3), 3.87 (q, 1H, JHH = 6.4 Hz, HNCH3), 7.06-7.27 (m,

3 3 6H, m- and p-C6H5), 7.56 (d, 2H, JHH = 6.6 Hz, o-C6H5), 7.89 (d, 2H, JHH = 6.6 Hz, o-

13 C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 39.4 (s, BNCH3), 44.0 (s, HNCH3),

126.3 (s, p-C6H5), 127.1 (s, p-C6H5), 128.2 (s, m-C6H5), 128.3 (s, m-C6H5), 133.5 (s, o-

11 C6H5), 134.4 (s, o-C6H5), 139.0 (s, br, i-C6H5), 143.1 (s, br, i-C6H5); B NMR (128 MHz,

THF-d8, 25 °C): δ = 26.6 (s, br).

1-Methyl-3,5-diphenyl-4-methylamino-1,2,3-triaza-3,5-diborolyl cyclooctadiene

rhodium dimer (6.2):

A solution of the 6.1a (100 mg, 0.370 mmol) and [Rh(cod)Cl]2 (91 mg, 0.185 mmol) in THF (10 mL) was stirred for 2 h at ambient temperature. The bright red solution was concentrated in vacuum to 3 mL and subsequently cooled to -35 °C for 2 days to afford a yellow crystalline solid that was separated by decantation, washed with pentane and dried under vacuum (42 %). Yellow, thin prismatic crystals were obtained by

237

cooling a concentrated solution of 6.2 in a mixture of hexane and THF to -35 ºC. 1H

NMR (400 MHz, CD2Cl2, 25 °C): δ = 1.30 (m, 2H, CH2), 1.89 (m, 2H, CH2), 2.20 (s, 3H,

HNCH3), 2.23 (m, 2H, CH2), 2.77 (m, 2H, CH2), 2.90 (s, 3H, BNCH3), 3.30 (q, br, 2H,

CH), 3.46 (t, br, 2H, CH), 3.63 (s, br, 1H, HNCH3), 7.24 – 7.35 (m, 5H, BC6H5), 7.55 (t,

13 1H, p-BC6H5), 7.67 (t, 2H, m-BC6H5), 8.86 (d, 2H, o-BC6H5); C NMR (100 MHz,

CD2Cl2, 25 °C): δ = 28.9 (2, CH2), 32.8 (s, CH2), 39.3 (s, BNCH3), 43.3 (s, HNCH3),

1 1 76.6 (d, JRhC = 13.0 Hz, CH), 84.3 (d, JRhC = 13.0 Hz, CH), 127.9 (s, m-BC6H5), 128.0

(s, m-BC6H5), 128.5 (s, p-BC6H5), 128.8 (s, p-BC6H5), 133.8 (s, o-BC6H5), 135.2 (s, o-

11 + BC6H5); B NMR (128 MHz, CD2Cl2, 25 °C): δ = 30.6 (s, br); MS (EI , 70 eV): m/z(%):

474 (10) [(cod)RhL]+, 364 (5) [RhL–2H]+, 264 (100) [L+H]+.

1-Methyl-3,5-diphenyl-4-methylamino-2-triphenylboryl-1,2,3-triaza-3,5-diborolyl

lithium (6.3a):

A solution of BPh3 (475 mg, 1.96 mmol) in THF (15 mL) was added dropwise to

a stirred yellow solution of 6.1a (530 mg, 1.96 mmol) in THF (20 mL) and the pale

yellow mixture was stirred at ambient temperature for 12 h. Subsequent removal of the

volatiles in vacuum left behind a colorless solid that was washed with hexane and dried

under vacuum, yielding a colorless powder (930 mg, 93 %). Colorless crystals of

6.3a(CH3CN)3 were obtained by cooling a concentrated solution of 6.3a in a mixture of

1 acetonitrile and hexane to -35 °C. H NMR (400 MHz, THF-d8, 25 °C): δ = 2.22 (d, 3H,

3 3 JHH = 6.4 Hz, HNCH3), 2.69 (s, 3H, BNCH3), 3.56 (q, 1H, JHH = 6.4 Hz, HNCH3), 6.62

(t, 3H, p-B(C6H5)3), 6.75 (t, 6H, m-B(C6H5)3), 7.39 (d, 6H, o-B(C6H5)3), 6.71 (m, 3H, p-

and m-BC6H5), 6.96 (m, 2H, o-BC6H5), 7.13 (t, 1H, p-BC6H5), 7.21 (t, 2H, m-BC6H5),

238

13 7.66 (d, 2H, o-BC6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 37.1 (s, BNCH3), 43.6

(s, HNCH3), 122.6 (s, p-B(C6H5)3), 125.7 (s, m-B(C6H5)3), 124.8 (s, p-BC6H5), 126.6 (s, m-BC6H5), 127.1 (s, p-BC6H5), 127.6 (s, m-BC6H5), 134.1 (s, o-BC6H5), 135.1 (s, o-

BC6H5), 136.6 (s, o-B(C6H5)3), 138.9 (s, br, i-BC6H5), 142.1 (s, br, i-BC6H5), 162.9 (s, br,

11 i-B(C6H5)3); B NMR (128 MHz, THF-d8, 25 °C): δ = -4.60 (s, br, B(C6H5)3), 27.3 (s, br,

BC6H5).

1-Methyl-3,5-diphenyl-4-methylamino-2-triphenylboryl-1,2,3-triaza-3,5-diborolyl

sodium (6.3b):

A colorless solution of BPh3 (150 mg, 0.619 mmol) in THF (15 mL) was added

dropwise to a stirred yellow solution of 6.1b (177 mg, 0.619 mmol) in THF (20 mL). The

pale yellow mixture was stirred at ambient temperature for 12 h and the volatiles were

subsequently removed in vacuum. The residue was washed twice with hexane and dried

under vacuum, yielding the desired product as a colorless powder (280 mg, 85.9 %). 1H

3 NMR (400 MHz, THF-d8, 25 °C): δ = 2.19 (d, 3H, JHH = 6.4 Hz, HNCH3), 2.71 (s, 3H,

3 BNCH3), 3.54 (q, 1H, JHH = 6.4 Hz, HNCH3), 6.62 (t, 3H, p-B(C6H5)3), 6.74 (t, 6H, m-

B(C6H5)3), 7.39 (d, 6H, o-B(C6H5)3), 6.70 (m, 3H, p- and m-BC6H5), 6.96 (m, 2H, o-

13 BC6H5), 7.13 (t, 1H, p-BC6H5), 7.21 (t, 2H, m-BC6H5), 7.66 (d, 2H, o-BC6H5); C NMR

(100 MHz, THF-d8, 25 °C): δ = 37.1 (s, BNCH3), 43.7 (s, HNCH3), 122.6 (s, p-

B(C6H5)3), 125.7 (s, m-B(C6H5)3), 124.7 (s, p-BC6H5), 127.1 (s, p-BC6H5), 126.6 (s, m-

BC6H5), 127.7 (s, m-BC6H5), 134.1 (s, o-BC6H5), 135.1 (s, o-BC6H5), 136.7 (s, o-

11 B(C6H5)3), 138.9 (s, br, i-BC6H5), 142.5 (s, br, i-BC6H5), 162.9 (s, br, i-B(C6H5)3); B

NMR (128 MHz, THF-d8, 25 °C): δ = -6.45 (s, br, B(C6H5)3), 25.0 (s, br, BC6H5).

239

1-Methyl-3,5-diphenyl-4-methylamino-2-triphenylboryl-1,2,3-triaza-3,5-diborolyl

potassium (6.3c):

A colorless solution of BPh3 (160 mg, 0.662 mmol) in THF (10 mL) was added

dropwise to a stirred yellow solution of 6.1c (200 mg, 0.662 mmol) in THF (15 mL). The pale yellow mixture was stirred at ambient temperature for 12 h and the volatiles were

subsequently removed in vacuum. The colorless solid residue was washed with hexane

and dried under vacuum yielding a colorless powder (300 mg, 83 %). Colorless crystals

were obtained by slow evaporation of a solution of 6.3c in a mixture of THF and Et2O.

1 3 H NMR (400 MHz, THF-d8, 25 °C): δ = 2.24 (d, 3H, JHH = 6.4 Hz, HNCH3), 2.69 (s,

3 3H, BNCH3), 3.56 (q, 1H, JHH = 6.4 Hz, HNCH3), 6.65 (t, 3H, p-B(C6H5)3), 6.78 (t, 6H,

m-B(C6H5)3), 7.43 (d, 6H, o-B(C6H5)3), 6.72 (m, 3H, p- and m-BC6H5), 6.70 (m, 2H, o-

13 BC6H5), 7.15 (t, 1H, p-BC6H5), 7.23 (t, 2H, m-BC6H5), 7.65 (d, 2H, o-BC6H5); C NMR

(100 MHz, THF-d8, 25 °C): δ = 37.1 (s, BNCH3), 43.6 (s, HNCH3), 122.6 (s, p-

B(C6H5)3), 125.7 (s, m-B(C6H5)3), 124.8 (s, p-BC6H5), 127.1 (s, p-BC6H5), 126.6 (s, m-

BC6H5), 127.6 (s, m-BC6H5), 134.1 (s, o-BC6H5), 135.1 (s, o-BC6H5), 136.6 (s, o-

11 B(C6H5)3), 138.9 (s, br, i-BC6H5), 142.1 (s, br, i-BC6H5), 162.9 (s, br, i-B(C6H5)3); B

NMR (128 MHz, THF-d8, 25 °C): δ = -6.58 (s, br, B(C6H5)3), 25.3 (s, br, BC6H5).

240

Preparation of tricyclic dilithiodiborate and tetrahydrazidotetraborane with

B4N8 frameworks

1, 2-Dimethyl-3,5-diphenyl-4-methylamino-1,2,4-triaza-3,5-diborolidine (6.4):

A solution of 1.63b (0.700 g, 2.65 mmol) in THF (20 mL) was slowly added to a suspension of KH (0.106 g, 2.65 mmol) in THF (15 mL). KH dissolved with evolution of hydrogen producing a yellow solution that was stirred at ambient temperature for another

2 h. MeI (0.165 mL, 2.65 mmol) was added to the mixture and the colorless suspension was stirred for an additional hour and then concentrated under vacuum to ca. 3 mL.

Hexane (30 mL) was added and the KI by-product was filtered off. The solvent was subsequently removed in vacuum, leaving behind the product as a colorless powder (638

1 3 mg, 88 %). H NMR (400 MHz, C6D6, 25 °C): δ = 2.32 (d, 3H, JHH = 6.4 Hz, HNCH3),

3 2.88 (s, 6H, (NCH3)2), 3.76 (q, 1H, JHH = 6.4 Hz, HNCH3), 7.25 - 7.34 (m, 6H, m- and p-

3 13 C6H5), 7.74 (d, 4H, JHH = 6.6 Hz, o-C6H5); C NMR (100 MHz, C6D6, 25 °C): δ = 31.7

(s, (NCH3)2), 43.5 (s, HNCH3), 128.8 (s, m-C6H5), 129.1 (s, p-C6H5), 134.4 (s, o-C6H5),

11 135.0 (s, br, i-C6H5); B NMR (128 MHz, THF-d8, 25 °C): δ = 28.6 (s, br).

Tricyclic B4N8 dilithiodiborate (6.5):

A solution of LiTMP was prepared from 1.6 M nBuLi in hexane (1.12 ml, 1.80

mmol) and TMP (0.254 g, 1.80 mmol) in THF (3 mL). The solutions of 6.4 (0.500 g,

1.80 mmol) in THF (3 mL) and LiTMP were pre-cooled to -35 °C and mixed, yielding an

orange solution that was stored at -35 °C for a day. Subsequently, it was allowed to

warm to room temperature and volatiles were removed in vacuum, leaving behind a

yellow residue that was washed twice with hexane (30 mL) and dried under vacuum. The

product was obtained as a colorless powder (388 mg, 60.5 %). Needle-like crystals were

241

1 obtained by recrystallization of 6.5 from THF at -35 ºC. H NMR (400 MHz, THF-d8, 25

°C): δ = 2.01 (s, 6H, NCH3), 2.28 (s, 6H, NCH3), 2.71 (s, 6H, NCH3), 7.12 - 7.31 (m,

3 3 12H, m- and p-C6H5), 7.53 (d, 4H, JHH = 6.8 Hz, o-C6H5), 7.77 (d, 4H, JHH = 6.8 Hz, o-

13 C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 34.2 (s, NCH3), 39.1 (s, NCH3), 41.3 (s,

11 NCH3), 126.3, 127.2 (s, p-C6H5), 127.5, 128.0 (s, m-C6H5), 134.0, 135.9 (s, o-C6H5); B

7 NMR (128 MHz, THF-d8, 25 °C): δ = 2.7 (s) and 28.5 (s, br); Li NMR (155 MHz, THF-

d8, 25 °C): δ = -2.18 (s).

Tricycle B4N8 tetrahydrazidotetraborane (6.6):

A solution of 6.5 (80.0 mg, 0.112 mmol) in THF (3 mL) was added to solid

FeCl2(thf)2 (38.0 mg, 0.141 mmol). The dark mixture was allow to stand at room temperature for a day, and was then concentrated to 1 mL and cooled to -35 °C for several days until colorless crystals of 6.6 formed and were separated by decantation (10

1 mg, 22 %). H NMR (400 MHz, THF-d8, 25 °C): δ = 2.66 (s, 6H, NCH3), 2.83 (s, 6H,

3 NCH3), 2.87 (s, 6H, NCH3), 7.31 - 7.33 (m, 6H, m- and p-C6H5), 7.55 (d, 4H, JHH = 5.6

13 Hz, o-C6H5); C NMR (100 MHz, THF-d8, 25 °C): δ = 32.9 (s, NCH3), 33.7 (s, NCH3),

11 42.4 (s, NCH3), 128.4 (s, m-C6H5), 128.9 (s, p-C6H5), 134.2 (s, o-C6H5); B NMR (128

+ + MHz, THF-d8, 25 °C): δ = 25.1 (s, br); MS (EI , 70 eV): m/z (%): 400(51) [M] , 385(13)

+ + 11 [M–Me] , 278(40) [M–Ph–3Me] ; HRMS for H28C18N8 B4: calcd. 400.2809, found

400.2838.

242

Cyclic voltammetry of 6.6:

The measurements were performed in the absence of oxygen, in anhydrous tetrahydrofuran, at an analyte concentration of 1 mM, and in the presence of 0.1 M

[nBu4N]PF6 as a supporting electrolyte. A PARstat 2273 Revision 2273 potentiostat was used at scan rates between 50 and 1000 mVs-1. The three-electrode cell had a platinum disk working electrode, a platinum wire auxiliary electrode, and a silver wire pseudo-

0/+1 0 reference electrode. [Cp2Co] with E = 1.36 V vs. ferrocene and -0.80 V vs. a saturated calomel electrode was used as an internal standard.

243

References

[1] Kealy, T. J.; Pauson, P. L. Nature 1951, 168, 1039.

[2] (a) Togni, A.; Halterman, R. L. (Eds.) Metallocenes, Wiley-VCH, Weinheim,

1998. (b) McKnight, A. L.; Waymouth, R. M. Chem. Rev. 1998, 98, 2587.

[3] Archer, R. D. Inorganic and Organometallic Polymers, Wiley-VCH, New York

2001.

[4] Arrayás, R. G.; Adrio, J.; Carretero, C. Angew. Chem. Int. Ed. 2006, 45, 7674.

[5] (a) Ashe III, A. J.; Al-Ahmad, S.; Fang, X. J. Organomet. Chem. 1999, 581, 92.

(b) Lee, B. Y.; Bazan, G. C. J. Organomet. Chem. 2002, 642, 275.

[6] (a) Herberich, G. E.; Greiss, G.; Heil, H. F. Angew. Chem. Int. Ed. Engl. 1970, 9,

805. (b) Herberich, G. E.; Becker, H. J.; Carsten, K.; Engelke, C.; Koch, W.

Chem. Ber. 1976, 109, 2382. (c) Ashe III, A. J.; Shu, P. J. Am. Chem. Soc. 1971,

93, 1804.

[7] (a) Scherer, O. J. Angew. Chem. Int. Ed. Engl. 1990, 29, 1104. (b) Scherer, O. J.

Acc. Chem. Res. 1999, 32, 751.

[8] (a) Kuhn, N.; Horn, E.-M.; Boese, R.; Augart, N. Angew. Chem. Int. Ed. Engl.

1988, 27, 1368. (b) Kuhn, N.; Horn, E.-M.; Boese, R.; Augart, N. Angew. Chem.

Int. Ed. Engl. 1989, 28, 342. (c) Kuhn, N.; Jendral, K.; Boese, R.; Bläser, D.

Chem. Ber. 1991, 124, 89.

[9] de Lauzon, G.; Deschamps, B.; Fischer, J.; Mathey, F.; Mitschler, A. J. Am.

Chem. Soc. 1980, 102, 994.

244

[10] (a) Ashe III, A. J.; Mahmoud, S.; Elschenbroich, C.; Wünsch, M. Angew. Chem.,

Int. Ed. Engl. 1987, 26, 229. (b) Ashe III, A. J.; Kampf, J. W.; Pilotek, S.;

Rousseau, R. Organometallics 1994, 13, 4067.

[11] (a) Ashe III, A. J.; Diephouse, T. R. J. Organomet. Chem. 1980, 202, C95. (b)

Ashe III, A. J.; Diephouse, T. R.; Kampf, J. W.; Al-Taweel, S. M.

Organometallics 1991, 10, 2068. (c) Ashe III, A. J.; Kampf, J. W.; Al-Taweel, S.

M. Organometallics 1992, 11, 1491.

[12] (a) Ashe III, A. J.; Kampf, J. W.; Al-Taweel, S. M. J. Am. Chem. Soc. 1992, 114,

372. (b) Ashe III, A. J.; Kampf, J. W.; Puranik, D. B.; Al-Taweel, S. M.

Organometallics 1992, 11, 2743.

[13] Fagan, P. J.; Nugent, W. A. J. Am. Chem. Soc. 1988, 110, 2310.

[14] Ashe III, A. J.; Kampf, J. W. Organometallics 1992, 11, 1491.

[15] Douglas, T.; Theopold, K. H. Angew. Chem. Int. Ed. Engl. 1989, 28, 1367.

[16] Sendlinger, S. C.; Haggerty, B. S.; Rheingold, A. L.; Theopold, K. H. Chem. Ber.

1991, 124, 2453.

[17] McComas, C. C.; Ziller, J. W.; Van Vranken, D. L. Organometallics 2000, 19,

2853.

[18] Ogasawara, M.; Nagano, T.; Yoshida, K.; Hayashi, T. Organometallics 2002, 21,

3062.

[19] Freeman, W. P.; Tilley, T. D.; Rheingold, A. L., Ostrander, R. L. Angew. Chem.

Int. Ed. Engl. 1993, 32, 1744.

[20] Freeman, W. P.; Tilley, T. D.; Rheingold, A. L. J. Am. Chem. Soc. 1994, 116,

8428.

245

[21] (a) Dysard, J. M.; Tilley, T. D. J. Am. Chem. Soc. 1998, 120, 8245. (b) Dysard, J.

M.; Tilley, T. D. J. Am. Chem. Soc. 2000, 122, 3097.

[22] Lui, S.-Y.; Hills, I. D.; Fu, G. C. Organometallics 2002, 21, 4323.

[23] Chen, J.; Fang, X.; Bajko, Z.; Kampf, J. W.; Ashe III, A. J. Organometallics

2004, 23, 5088.

[24] (a) Ashe, A. J., III; Kampf, J. W.; Schiesher, M. Organometallics 2000, 19, 4681.

(b) Ashe III, A. J.; Fang, X.; Kampf, J. W. Organometallics 2000, 19, 4935.

[25] Liu, S.-Y.; Lo, M. M.-C.; Fu, G. C. Angew. Chem. Int. Ed. Engl. 2002, 41, 174.

[26] Liu, S.-Y.; Hills, I. D.; Fu, G. C. J. Am. Chem. Soc. 2005, 127, 15352.

[27] Ashe III, A. J.; Fang, X. Org. Lett. 2000, 2, 2089.

[28] Hänssgen, D.; Odenhausen, E. Chem. Ber. 1979, 112, 2389.

[29] (a) Ashe III, A. J.; Fang, X.; Kampf, J. W. Organometallics 1998, 17, 2379. (b)

Ashe III, A. J.; Fang, X.; Kampf, J. W. Organometallics 1999, 18, 1821.

[30] (a) Schmid, G.; Schulze, J. Angew. Chem. Int. Ed. Engl. 1980, 19, 54. (b)

Amirkhalili, S.; Höhner, U.; Schmid, G. Angew. Chem. Int. Ed. Engl. 1982, 21,

68. (c) Schmid, G.; Zaika, D.; Lehr, J.; Augart, N.; Boese, R. Chem. Ber. 1988,

121, 1873. (d) Schmid, G. Comments Inorg. Chem. 1985, 4, 17.

[31] Schmid, G.; Kampmann, D.; Höhner, U.; Bläser, D.; Boese, R. Chem. Ber. 1984,

117, 1052.

[32] Schmid, G.; Boltsch, O.; Bläser, D.; Boese, R. Z. Naturforsch. 1984, B39, 1082.

[33] Herberich, G. E.; Ohst, H. Adv. Organomet. Chem. 1986, 25, 199.

[34] Ashe III, A. J.; Bajko, Z.; Fang, X.; Kampf, J. W.; Yang, H. Phosphorus, Sulfur,

and Silicon 2004, 179, 711.

246

[35] Yang, H.; Fang, X.; Kampf, J. W.; Ashe III, A. J. Polyhedron 2005, 24, 1280.

[36] Bazan, G. C.; Rodriguez, G.; Ashe III, A. J.; Al-Ahmad, S.; Müller, C. J. Am.

Chem. Soc. 1996, 118, 2291.

[37] Ashe III, A. J.; Al-Ahmad, S.; Fang, X.; Kampf, J. W. Organometallics 1998, 17,

3883.

[38] For reviews see: (a) Cosgriff, J. E.; Deacon, G. B. Angew. Chem. Int. Ed. 1998,

37, 286. (b) Sadimenko, A. P.; Basson, S. S. Coord. Chem. Rev. 1996, 147, 247.

[39] (a) Guzei, I. A.; Baboul, A. G.; Yap, G. P. A.; Rheingold, A. L.; Schlegel, H. B.;

Winter, C. H. J. Am. Chem. Soc. 1997, 119, 3387. (b) Guzei, I. A.; Yap, G. P. A.;

Winter, C. H. Inorg. Chem. 1997, 36, 1738.

[40] Perera, J. R.; Heeg, M. J.; Schlegel, H. B.; Winter, C. H. J. Am. Chem. Soc. 1999,

121, 4536.

[41] Becker, G.; Becker, W.; Knebl, R.; Schmidt, H.; Weeber, U.; Westerhausen, M.

Nova Acta Leopold 1985, 59, 55.

[42] Bartsch, R.; Hitchcock, P. B.; Nixon, J. F. J. Chem. Soc., Chem. Commun. 1987,

1146.

[43] Bartsch, R.; Hitchcock, P. B.; Nixon, J. F. J. Organomet. Chem. 1988, 356, C1.

[44] Bartsch, R.; Hitchcock, P. B.; Nixon, J. F. J. Chem. Soc., Chem. Commun. 1988,

819.

[45] Bartsch, R.; Hitchcock, P. B.; Nixon, J. F. J. Organomet. Chem. 1989, 373, C17.

[46] Bartsch, R.; Gelessus, A.; Hitchcock, P. B.; Nixon, J. F. J. Organomet. Chem.

1992, 430, C10.

247

[47] Scheer, M.; Deng, S.; Scherer, O. J.; Sierka, M. Angew. Chem. Int. Ed. 2005, 44,

3755.

[48] Szarvas, L.; Bajko, Z.; Fusz, S.; Burck, S.; Daniels, J.; Nieger, M.; Gudat, D. Z.

Anorg. Allg. Chem. 2002, 628, 2303.

[49] Zheng, W.; Zhang, G.; Fan, K. Organometallics 2006, 25, 1548.

[50] Schmidpeter, A.; Willhalm, A. Angew. Chem. Int. Ed. Engl. 1984, 23, 903.

[51] Lee, V. Y.; Kato, R.; Ichinohe, M.; Sekiguchi, A. J. Am. Chem. Soc. 2005, 127,

13142.

[52] Lee, V. Y.; Kato, R.; Sekiguchi, A.; Krapp, A.; Frenking, G. J. Am. Chem. Soc.

2007, 129, 10340.

[53] Lee, V. Y.; Ichinohe, M.; Sekiguchi, A. J. Am. Chem. Soc. 2000, 122, 12604.

[54] (a) Paine, R. T. and Narula, C. K. Chem. Rev. 1990, 90, 73. (b) Schubert, U. and

Hüsing, N. Synthesis of Inorganic Materials, Viley-VCH, Weinheim, 2000. (c)

Corso, M.; Auwärter, W.; Muntwiler, M.; Tamai, A.; Greber, T.; Osterwalder, J.

Science 2004, 303, 217.

[55] (a) Wakamiya, A.; Ide T.; Yamaguchi, S. J. Am. Chem. Soc. 2005, 127, 14859. (b)

Sham, I. H. T.; Kwok, C.-C.; Che C.-M.; Zhu, N. Chem. Commun. 2005, 3547.

[56] (a) Niedenzu, K.; Fritz, P.; Jenne, H. Angew. Chem., Int. Ed. Engl. 1964, 3, 506.

(b) Nölle, D.; Nöth, H. Angew. Chem., Int. Ed. Engl. 1971, 10, 126. (c) Nölle, D.;

Nöth, H. Z. Naturforsch. 1972, 27 B, 1425.

[57] (a) Nöth, H.; Regnet, W. Z. Anorg. Allg. Chem. 1967, 352, 1. (b) Nölle, D.; Nöth,

H.; Winterstein, W. Z. Anorg. Allg. Chem. 1974, 406, 235.

248

[58] (a) Nölle, D.; Nöth, H. Chem. Ber. 1978, 111, 469. (b) Engelhardt, U.; Park, S. S.

Acta Cryst. 1996, C 52, 3248.

[59] (a) Su, M.-D.; Chu, S.-Y. J. Phys. Chem. 1989, 93, 6043. (b) Bridgeman, A. J.;

Rothery, J. Inorg. Chim. Acta 1999, 288, 17.

[60] Urnėžius, E.; Brennessel, W. W.; Cramer, C. J.; Ellis, J. E.; Schleyer, P. v. R.

Science 2002, 295, 832.

[61] Scherer. O. J.; Schwalb, J.; Wolmershauser, G.; Kaim, W.; Gross, R. Angew.

Chem. Int. Ed. Engl. 1986, 25, 363.

[62] Reingold, A. L.; Foley, M. J.; Sullivan, P. J. J Am. Chem. Soc. 1982, 104, 4727.

[63] Breunig, H. J.; Rosler, R.; Burford, N. Angew. Chem. Int. Ed. 2000, 39, 4148.

[64] (a) Grimes, R. N. Chem. Rev. 1992, 92, 251. (b) Siebert, W. Pure & Appl. Chem.

1988, 60, 1345.

[65] Kuhlmann, T.; Roth, S.; Rozière, J.; Siebert, W. Angew. Chem. Int. Ed. Engl.

1986, 25, 105.

[66] (a) Harder, S. Coord. Chem. Rev. 1998, 176, 17, and erratum, Coord. Chem. Rev.

2000, 199, 331. (b) Jutzi, P.; Burford, N. Chem. Rev. 1999, 99, 969.

[67] (a) Deforth, T.; Kaschke, M.; Stock, H.; Pritzkow, H.; Siebert, W. Z. Naturforsch.

1997, 52B, 823. (b) Enders, M.; Gangnus, B.; Hettrich, R.; Magos-Martin, Z.;

Stephan, M.; Pritzkow, H.; Siebert, W.; Zenneck, U. Chem. Ber. 1993, 126, 2197.

[68] Ederer, B.; Metzler, N.; Nöth, H. Chem. Ber. 1993, 126, 2003.

[69] Fisch, H.; Pritzkow, H.; Siebert, W. Angew. Chem. Int. Ed. Engl. 1984, 23, 608.

[70] Lochte, H. L.; Noyes, W. A.; Bailey, J. R. J. Am. Chem. Soc. 1922, 44, 2556.

249

[71] (a) Boese, R.; Maulitz, A. H.; Stellberg, P. Chem. Ber. 1994, 127, 1887. (b)

Jäschke, T.; Jansen, M. Z. Anorg. Allg. Chem. 2004, 630, 239.

[72] Kohata, K.; Fukuyama, T.; Kuchitsu, K. J. Phys. Chem. 1982, 86, 602.

[73] Krahulic, K. E.; Enright, G. D.; Parvez, M.; Roesler, R. J. Am. Chem. Soc. 2005,

127, 4142.

[74] Boyle, T. J.; Pedrotty, D. M.; Alam, T. M.; Vick, S. C.; Rodriguez, M. A. Inorg.

Chem. 2000, 39, 5133.

[75] (a) Jubb, J.; Gambarotta, S.; Duchateau, R.; Teuben, J. H. J. Chem. Soc., Chem.

Commun. 1994, 2641. (b) Bonomo, L.; Solari, E.; Scopelliti, R.; Latronico, M.;

Floriani, C. Chem. Commun. 1999, 2227.

[76] (a) Evans, W. J.; Boyle T. J.; Ziller, J. W. Organometallics 1992, 11, 3903. (b)

Dinnebier, R. E.; Behrens U.; Olbrich, F. Organometallics 1997, 16, 3855. (c)

Nie, W.; Qian, C.; Chen Y.; Sun, J. Organometallics 2001, 20, 5780.

[77] (a) Herberich, G. E.; Hostalek, M.; Laven, R.; Boese, R. Angew. Chem. Int. Ed.

Engl. 1990, 29, 317. (b) Herberich, G. E.; Eigendorf, U.; Englert, U. Chem. Ber.

1993, 126, 1397. (c) Grigsby, W. J.; Power, P. P. J. Am. Chem. Soc. 1996, 118,

7981. (d) Scheschkewitz, D.; Menzel, M.; Hofmann, M.; von Ragué Schleyer, P.;

Geiseler, G.; Massa, W.; Harms, K.; Berndt, A. Angew. Chem. Int. Ed. 1999, 38,

2936.

[78] (a) Hoic, D. A.; Davis, W. M.; Fu, G. C. J. Am. Chem. Soc. 1995, 117, 8480. (b)

Herberich, G. E.; Englert, U.; Schmidt, M. U.; Standt, R. Organometallics 1996,

15, 2707. (c) Herberich, G. E.; Ganter, B.; Pons, M. Organometallics 1998, 17,

250

1254. (d) Ghesner, I; Piers, W. E.; Parvez, M.; McDonald, R. Organometallics

2004, 23, 3085.

[79] Cradock, S.; Purves, C.; Rankin, D. W. H. J. Mol. Struct. 1990, 220, 193.

[80] Fu, G. C. Adv. Organomet. Chem. 2001, 47, 101.

[81] Hitzbleck, J.; Deacon, G. B.; Ruhlandt-Senge, K. Angew. Chem. Int. Ed. 2004, 43,

5218.

[82] (a) Aoyagi, T.; Shearer, H. M. M.; Wade, K.; Whitehead, G. J. Organomet. Chem.

1979, 175, 21. (b) Rogers, R. D.; Atwood, J. L.; Rausch, M. D.; Macomber, D.

W.; Hart W. P. J. Organomet. Chem. 1982, 238, 79. (c) Rabe, G.; Roesky, H. W.;

Stalke, D.; Pauer, F.; Sheldrick, G. M. J. Organomet. Chem. 1991, 403, 11. (d)

Herberich, G. E.; Fischer, A. Organometallics 1996, 15, 58. (e) Kähler, T.;

Behrens, U.; Neander, S.; Olbrich, F. J. Organomet. Chem. 2002, 649, 50. (f)

Dinnebier, R. B.; Schneider, M.; van Smaalen, S.; Olbrich F.; Behrens, U. Acta

Cryst. 1999, B 55, 35. (g) Cole, M. L; Jones, C.; Junk, P. C. J. Chem. Soc., Dalton

Trans. 2002, 896.

[83] (a) Lee, B. Y.; Wang, S.; Putzer, M.; Bartholomew, G. P.; Bu, X.; Bazan, G. C. J.

Am. Chem. Soc. 2000, 122, 3969. (b) Lee, B. Y.; Bazan, G. C. J. Am. Chem. Soc.

2000, 122, 8577.

[84] Deacon, G. B.; Delbridge, E. E.; Evans, D. J.; Harika, R.; Junk, P. C.; Skelton, B.

W.; White, A. H. Chem. Eur. J. 2004, 10, 1193.

[85] Kuhn, N.; Henkel, G.; Kreutzberg, J. Angew. Chem. Int. Ed. Engl. 1990, 29, 1143.

[86] Harder, S.; Prosenc, M. H.; Rief, U. Organometallics 1996, 15, 118.

251

[87] Giesbrecht, G. R.; Gebauer, A.; Shafir, A.; Arnold, J. J. Chem. Soc., Dalton

Trans. 2000, 4018.

[88] (a) Jordan, V.; Behrens, U.; Olbrich, F.; Weiss, E. J. Organomet. Chem. 1996,

517, 81. (b) Karsch, H. H.; Graf, V. W.; Reisky, M. Chem. Commun. 1999, 1695.

Organomet. Chem. 2002, 649, 252. (d) Evans, W. J.; Giarikos, D. G.; Ziller, J. W.

J. Organomet. Chem. 2003, 688, 200.

[89] (a) Hoic, D. A.; DiMare, M.; Fu, G. C. J. Am. Chem. Soc. 1997, 119, 7155. (b)

Zheng, X.; Herberich, G. E. Eur. J. Inorg. Chem. 2003, 2175.

[90] Boros, E. E.; Bouvier, F.; Randhawa, S.; Rabinowitz, M. H. J. Heterocyclic

Chem. 2001, 38, 613.

[91] Jutzi, P.; Reumann, G. J. Chem. Soc., Dalton Trans. 2000, 2237.

[92] Rhodes, B.; Chien, J. C. W.; Rausch, M. D. Organometallics 1998, 17, 1931.

[93] (a) Jutzi, P.; Kanne, D.; Krüger, C. Angew. Chem. Int. Ed. Engl. 1986, 25, 164.

(b) Jutzi, P.; Holtmann, U.; Kanne, D.; Blom, R.; Gleiter, R.; Hyla-Kryspin, I.

Chem. Ber. 1989, 122, 1629.

[94] (a) Jutzi, P.; Mix, A.; Rummel, B.; Schoeller, W. W.; Neumann, B.; Stammler,

H.-G. Science 2004, 305, 849. (b) Bertrand, G. Science 2004, 305, 783.

[95] (a) Scibelli, J. V.; Curtis, M. D. J. Am. Chem. Soc. 1973, 95, 924. (b) Grenz, M.;

Hahn, E.; du Mont, W.-W.; Pickardt, J. Angew. Chem. Ed. Int. Engl. 1984, 23, 61.

[96] Schumann, H.; Janiak, C.; Hahn, E.; Loebel, J.; Zuckerman, J. L. Angew. Chem.

Int. Ed. Engl. 1985, 24, 773.

252

[97] Jutzi, P.; Schlüter, E.; Hursthouse, M. B.; Arif, A. M.; Short, R. L. J. Organomet.

Chem. 1986, 299, 285.

[98] (a) Fernholt, L.; Haaland, A.; Jutzi, P.; Kohl, F. X.; Seip, R. Acta Chem. Scand.

Ser. A 1984, 38, 211. (b) Schöpper, A.; Saak, W.; Weidenbruch, M. J.

Organomet. Chem. 2006, 691, 809.

[99] Winter, J. G.; Portius, P.; Kociok-Köhn, G.; Steck, R.; Filippou, A. C.

Organometallics 1998, 17, 4176.

[100] (a) Fischer, E. O.; Grubert, H. Z. Naturforsch., Teil B 1956, 11, 423. (b) Atwood,

J. L.; Hunter, W. E. J. Chem. Soc. Chem. Comm. 1981, 925.

[101] (a) Jutzi, P.; Kohl, F. X.; Hoffmann, P.; Kruger, C.; Tsay, Y. Chem. Ber. 1980,

113, 757. (b) Cowley, A. H.; Lasch, J. G.; Norman, N. C.; Stewart, C. A.; Wright,

T. C. Organometallics 1983, 2, 1691. (c) Schumann, H.; Janiak, C.; Hahn, E.;

Kolaxa, C.; Loebela, J.; Rauschb, M. D.; Zuckermann, J. J.; Heeg, M. J. Chem.

Ber. 1986, 119, 2656. (d) Cowley, A. H.; Jutzi, P.; Kohl, F. X.; Lasch, J. G.;

Norman, N. C.; Schlüter, E. Angew. Chem. Int. Ed. Engl. 1984, 23, 616. (e)

Burkey, D. J.; Hanusa, T. P. Organometallics 1995, 14, 11. (f) Heeg, M. J.;

Janiak, C.; Zuckerman, J. J. J. Am. Chem. Soc. 1984, 106, 4259. (g) Sitzmann, H.;

Boese, R.; Stellberg, P. Z. Anorg. Allg. Chem. 1996, 622, 751.

[102] (a) Constantine, S. P.; Hitchcock, P. B.; Lawless, G. A.; De Lima, G. M. Chem.

Commun. 1996, 1101. (b) Constantine, S. P.; Cox, H.; Hitchcock, P. B.; Lawless,

G. A. Organometallics 2000, 19, 317.

253

[103] Armstrong, D. R.; Beswick, M. A.; Cromhout, N. L.; Harmer, C. N.; Moncrieff,

D.; Russell, C. A.; Raithby, P. R.; Steiner, A.; Wheatley, A. E. H.; Wright, D. S.

Organometallics 1998, 17, 3176.

[104] (a) Davidson, M. G.; Stalke, D.; Wright, D. S. Angew. Chem. Int. Ed. Engl. 1992,

31, 1226. (b) Armstrong, D. R.; Duer, M. J.; Davidson, M. G.; Moncrieff, D.;

Russell, C. A.; Stourton, C.; Steiner, A.; Stalke, D.; Wright, D. S.

Organometallics 1997, 16, 3340.

[105] (a) Bos, K. D.; Bulten, E. J.; Noltes, J. G.; Spek, A. L. J. Organomet. Chem. 1975,

99, 71. (b) Constantine, S. P.; De Lima, G. M.; Hitchcock, P. B.; Keates, J. M.;

Lawless, G. A.; Marziano, I. Organometallics 1997, 16, 793. (c) Jutzi, P.; Kohl,

F.; Krüger, C. Angew. Chem. Int. Ed. Engl. 1979, 18, 59. (d) Jutzi, P.; Kohl, F.;

Krüger, C.; Wolmershäuser, G.; Hoffmann, P.; Stauffert, P. Angew. Chem. Int.

Ed. Engl. 1982, 94, 66. (e) Hani, R.; Geanangel, R. A. J. Organomet. Chem. 1985,

293, 197.

[106] Schmid, G.; Zaika, D.; Boese, R. Angew. Chem. Int. Ed. Engl. 1985, 24, 602.

[107] Wadepohl, H.; Pritzkow, H.; Siebert, W. Organometallics 1983, 2, 1899.

[108] Kuhn, N.; Henkel, G.; Stubenrauch, S. J. Chem. Soc. Chem. Commun. 1992, 760.

[109] Mitzel, N. W.; Riede, J.; Schmidbaur, H. Acta Cryst. 1997, C53, 1335.

[110] Cowley, A. H.; Ebsworth, E. A. V.; Mehrotra, S. K.; Rankin, D. W. H.;

Walkinshaw, M. D. J. Chem. Soc., Chem. Commun. 1982, 1099.

[111] (a) Hosmane, N. S.; de Meester, P. J. Am. Chem. Soc. 1986, 108, 6050. (b)

Hosmane, N. S.; Islam, M. S.; Pinkston, B. S.; Upali Siriwardane, U.; Banewicz,

J. J.; Maguire, J. A. Organometallics 1988, 7, 2340. (c) Hosmane, N. S.; Yang, J.;

254

Lu, K.-J.; Zhang, H.; Siriwardane, U.; Islam, M. S.; Thomas, J. L. C.; Maguire, J.

A. Organometallics 1998, 17, 2784.

[112] (a) Wrackmeyer, B. Annu. Rep. NMR Spectrosc. 1985, 16, 73. (b) Richter, F.;

Dargatz, M.; Hartung, H.; Schollmeyer, D.; Weichmann, H. J. Organomet. Chem.

1996, 514, 233. (c) Růžička, A.; Jambor, R.; Císařová, I.; Holeček, J. Chem. Eur.

J. 2003, 9, 2411.

[113] Casas, J. S.; Castellano, E. E.; Ellena, J.; García-Tasende, M. S.; Sánchez, A.;

Sordo, J.; Taboada, C.; Vidarte, M. J. Appl. Organometal. Chem. 2003, 17, 940.

[114] Jones, J. N.; Moore, J. A.; Cowley, A. H.; MacDonald, C. L. B. Dalton Trans.

2005, 3846.

[115] Budzelaar, P. H. M.; Boersma, J.; van der Kerk, G. J. M.; Spek, A. L.;

Duisenberg, A. J. M. J. Organomet. Chem. 1985, 281, 123.

[116] (a) Fischer, B.; Wijkens, P.; Boersma, J.; van Koten, G. Smeets, W. J. J.; Spek, A.

L.; Budzelaar, P. H. M. J. Organomet. Chem. 1989, 376, 223. (b) Burkey D. J.;

Hanusa, T. P. J. Organomet. Chem. 1996, 512, 165. (c) Fernández, R.; Resa, I.;

del Río, D.; Carmona, E. Organometallics 2003, 22, 381.

[117] (a) Resa, I.; Carmona, E.; Gutierrez-Puebla, E.; Monge, A. Science 2004, 305,

1136. (b) Alvarez, E.; Grirrane, A.; Resa, I.; del Río, D.; Rodríguez, A.; Carmona,

E. Angew. Chem., Int. Ed. 2007, 46, 1296.

[118] (a) Fischer, B.; van Mier, G. P. M.; Boersma, J.; Smeets, W. J. J.; Spek, A. L. J.

Organomet. Chem. 1987, 322, C37. (b) Smeets, W. J. J.; Spek, A. L.; Fischer, B.;

van Mier, G. P. M.; Boersma, J. Acta Cryst. 1987, C43, 893. (c) Barr, D.;

255

Edwards, A. J.; Raithby, P. R.; Rennie, M.-A.; Verhorevoort, K. L.; Wright, D. S.

J. Organomet. Chem. 1995, 493, 175.

[119] Bentz, D.; Wolmershäuser, G.; Sitzmann, H. Organometallics 2006, 25, 3175.

[120] Cummins, C. C.; Schrock, R. R.; Davis, W. M. Organometallics 1991, 10, 3781.

[121] (a) Fischer, B.; van Mier, G. P. M.; Boersma, J.; van Koten, G.; Smeets, W. J. J.;

Spek, A. L. Recl. Trav. Chim. Pays-Bas 1988, 107, 259. (b) Sitzmann, H.;

Wolmershäuser, G. Z. Anorg. Allg. Chem. 1995, 621, 109. (c) Hitchcock, P. B.;

Keates, J. M.; Lawless, G. A. J. Am. Chem. Soc. 1998, 120, 599.

[122] (a) Davies, A. G.; Goddard, J. P.; Hursthouse, M. B.; Walker, N. P. C. J. Chem.

Soc., Dalton Trans. 1985, 471. (b) Lorberth, J.; Berlitz, T. F.; Massa, W. Angew.

Chem., Int. Ed. Engl. 1989, 28, 611.

[123] (a) Ly, H. V.; Forster, T. D.; Maley, D.; Parvez, M.; Roesler, R. Chem. Commun.

2005, 4468. (b) Ly, H. V.; Forster, T. D.; Corrente, A. M.; Eisler, D. J.; Konu, J.;

Parvez, M.; Roesler, R. Organometallics 2007, 26, 1750.

[124] (a) Greenwood, N. N.; McGinnety, J. A.; Owen, J. D. J. Chem. Soc. A 1971, 809.

(b) Allmann, R.; Bätzel, V.; Pfeil, R.; Schmid, G. Z. Naturforsch. 1976, 31B,

1329. (c) Goeta, A. E.; Howard, J. A. K.; Hughes, A. K.; Johnson, A. L.; Wade,

K. Chem. Commun. 1998, 1713.

[125] Hasselgren, C.; Dean, P. A. W.; Scudder, M. L.; Craig, D. C.; Dance I. G. J.

Chem. Soc., Dalton Trans. 1997, 2019.

[126] (a) Greenwood, N. N.; McGinnety, J. A.; Owen, J. D. J. Chem. Soc., Dalton

Trans. 1972, 989. (b) Schaper, T.; Preetz, W. Inorg. Chem. 1998, 37, 363. (c)

Littger, R.; Englich, U.; Spencer, J. T. Inorg. Chem. 1997, 36, 6434.

256

[127] (a) Colquhoun, H. M.; Greenhough, T. J.; Wallbridge, M. G. H. J. Chem. Soc.,

Dalton Trans., 1979, 619. (b) Lewis, Z. G.; Welch, A. J. Acta Cryst. 1993, C49,

715. (c) Teixidor, F.; Ayllòn, J. A.; Viñas, C.; Kivekäs, R.; Sillanpää, R.; Casabò,

J. J. Organomet. Chem. 1994, 483, 153.

[128] (a) Kulzick, M. A.; Price, R. T.; Andersen, R. A.; Muetterties, E. L. J.

Organomet. Chem. 1987, 333, 105. (b) Duan, Z.; Hampden-Smith, M. J.;

Sylwester, A. P. Chem. Mater. 1992, 4, 1146.

[129] Huck, S.; Ginsberg, A.; Pritzkow, H.; Siebert, W. J. Organomet. Chem. 1998,

571, 107.

[130] Ginsberg, A.; Pritzkow, H.; Siebert, W. J. Organomet. Chem. 2001, 619, 7.

[131] Loginov, D. A.; Muratov, D. V.; Starikova, Z. A.; Petrovskii, P. V.; Kudinov, A.

R. Russ. Chem. Bull., Ed., 2006, 55, 1581.

[132] Adams, H.; Bailey, N. A.; Mann, B. E.; Taylor, B. F.; White, C.; Yavari, P. J.

Chem. Soc. Dalton Trans. 1987, 1947.

[133] (a) Day, V. W.; Reimer, K. J.; Shaver, A. J. Chem. Soc. Chem. Comm. 1975, 403.

(b) Sünkel, K.; Stramm, C.; Lang, M.; Kempinger, W.; Hofmann, J. Inorganica

Chimica Acta 1998, 269, 111.

[134] Siebert, W.; Hettrich, R.; Pritzkow, H. Angew. Chem. Int. Ed. Engl. 1994, 33, 203.

[135] Hettrich, R.; Kaschke, M.; Wadepohl, H.; Weinmann, W.; Stephan, M.; Pritzkow,

H.; Siebert, W.; Hyla-Kryspin, I.; Gleiter, R. Chem. Eur. J. 1996, 2, 487.

[136] (a) Schmid, G.; Amirkhalili, S.; Höhner, U.; Kampmann, D.; Boese, R. Chem.

Ber. 1982, 115, 3830. (b) Schmid, G.; Höhner, U.; Kampmann, D.; Zaika, D.;

257

Boese, R. Chem. Ber. 1983, 116, 951. (c) Schmid, G.; Haske, S.; Zaika, D.;

Boese, R.; Bläser, D. Chem. Ber. 1994, 127, 73.

[137] Wallace, C. E.; Selegue, J. P.; Carrillo, A. Organometallics 1998, 17, 3390.

[138] Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877.

[139] (a) Kölle, U.; Grub, J. J. Organomet. Chem. 1985, 289, 133. (b) Hashidzume, K.;

Tobita, H.; Ogino, H. Organometallics 1995, 14, 1187. (c) Gassman, P. G.;

Winter, C. H. J. Am. Chem. Soc. 1988, 110, 6130.

[140] Ashe III, A. J.; Al-Ahmad, S.; Pilotek, S.; Puranik, D. B. Organometallics 1995,

14, 2689.

[141] Fang, X.; Yang, H.; Kampf, J. W.; Holl, M. M. B.; Ashe III, A. J.

Organometallics 2006, 25, 513.

[142] Corey, J. Y.; Zhu, X.-H.; Brammer, L.; Rath, N. P.; Acta Cryst. 1995, C51, 565.

[143] Bajgur, C. S.; Tikkanen, W.; Petersen, J. L. Inorg. Chem. 1985, 24, 2539.

[144] (a) Schmid, G.; Höhner, U.; Kampmann, D.; Schmidt, F.; Bläser, D.; Boese, R.

Chem. Ber. 1984, 117, 672. (b) Schmid, G.; Schmidt, F.; Boese, R. Chem. Ber.

1985, 118, 1949. (c) Kershner, D. L.; Rheingold, A. L.; Basolo, F.

Organometallics 1987, 6, 196.

[145] (a) Thiele, B.; Paetzold, P.; Englert, U. Chem. Ber. 1992, 125, 2681. (b) Greiwe,

P.; Beez, V.; Pritzkow, H.; Siebert, W. Eur. J. Inorg. Chem. 2001, 381.

[146] Hillman, M. E. D. J. Am. Chem. Soc. 1963, 85, 1626.

[147] Edwin, J.; Siebert, W. J. Organomet. Chem. 1981, 215, 255.

[148] Heldt, I.; Behrens, U. Z. Anorg. Allg. Chem. 2005, 631, 749.

[149] Evans, W. J.; Brady, J. C.; Ziller J. W. Inorg. Chem. 2002, 41, 3340.

258

[150] Cloke, F. G. N.; Kuchta, M. C.; Harker, R. M.; Hitchcock, P. B.; Parry, J. S.

Organometallics 2000, 19, 5795.

[151] (a) Oelckers, B.; Chávez, I.; Manríquez, J. M.; Román, E. Organometallics 1993,

12, 3396. (b) Summerscales, O. T.; Cloke, F. G. N. Coord. Chem. Rev. 2006, 250,

1122.

[152] Geiger, W. E.; Salzer, A.; Edwin, J.; von Philipsborn, W.; Piantini, U.; Rheingold,

A. J. Am. Chem. Soc. 1990, 112, 7113.

[153] Kudinov, A. R.; Loginov, D. A.; Starikova, Z. A.; Petrovskii, P. V.; Corsini, M.;

Zanello, P. Eur. J. Inorg. Chem. 2002, 3018.

[154] Herberich, G. E.; Englert, U.; Pubanz, D. J. Organomet. Chem. 1993, 459, 1.

[155] O'Hare, D.; Brookes, J.; Watkin, D. J. J. Mater. Chem. 1991, 1, 691.

[156] Herberich, G. E.; Englert, U.; Marken, F.; Hofmann, P. Organometallics 1993,

12, 4039.

[157] Ghosh, S.; Noll, B. C.; Fehlner, T. P. Angew. Chem. Int. Ed. Engl. 2005, 44, 6568.

[158] Hoover, J. M; DiPasquale, A.; Mayer, J. M.; Michael, F. E. Organometallics

2007, 26, 3297.

[159] Manríquez, J. M.; Ward, M. D.; Reiff, W. M.; Calabrese, J. C.; Jones, N. L.;

Carroll, P. J.; Bunel, E. E.; Miller, J. S. J. Am. Chem. Soc. 1995, 117, 6182.

[160] Westerhausen, M.; Wieneke, M.; Nöth, H.; Seifert, T.; Pfitzner, A.; Schwarz, W.;

Schwarz, O.; Weidlein J. Eur. J. Inorg. Chem. 1998, 1175.

[161] Gregory, K.; Bremer, M.; Bauer, W.; Schleyer, P. v. R.; Lorenzen, N. P.; Kopf, J.;

Weiss, E. Organometallics 1990, 9, 1485.

259

[162] Hacker, R.; Kaufmann, E.; Schleyer, P. v. R.; Mahdi, W.; Dietrich H. Chem. Ber.

1987, 120, 1533.

[163] (a) Bock, H.; Arad, C.; Näther, C.; Havlas, Z. Helv. Chim. Acta 1997, 80, 606. (b)

Bock, H.; Arad, C.; Näther, C.; Havlas, Z. J. Organomet. Chem. 1997, 548, 115.

[164] (a) Nöth, H.; Rojas-Lima, S.; Troll, A. Eur. J. Inorg. Chem. 2005, 1895. (b) Nöth,

H.; Troll, A. Eur. J. Inorg. Chem. 2005, 3524.

[165] Athimoolam, A.; Gambarotta, S.; Korobkov, I. Can. J. Chem. 2005, 83, 832.

[166] (a) Esbak, H.; Behrens, U. Z. Anorg. Allg. Chem. 2005, 631, 1581. (b) Gregory,

K.; Bremer, M.; Schleyer, P. v. R.; Klusener, P. A. A.; Brandsma, L. Angew.

Chem., Int. Ed. Engl. 1989, 28, 1224.

[167] (a) Raja, R.; Thomas, J. M.; Jones, M. D.; Johnson, B. F. G.; Vaughan, D. E. W.

J. Am. Chem. Soc. 2003, 125, 14982. (b) Rouzaud, J.; Jones, M. D.; Raja, R.;

Thomas, J. M.; Johnson, B. F. G.; Thomas, J. M.; Duer, M. J. Helv. Chim. Acta

2003, 86, 1753.

[168] de Bruin, B.; Kicken, R. J. N. A. M.; Suos, N. F. A.; Donners, M. P. J.; den

Reijer, C. J.; Sandee, A. J.; de Gelder, R.; Smits, J. M. M.; Gal, A. W.; Spek A. L.

Eur. J. Inorg. Chem. 1999, 1581.

[169] Kuhn, N.; Schulten, M.; Zauder, E.; Augart, N.; Boese, R. Chem. Ber. 1989, 122,

1891.

[170] (a) Wittig, G.; Keicher, G.; Rückert, A.; Raff, P. Liebigs Ann. 1949, 563, 110. (b)

Wittig, G. Angew. Chem. 1950, 62, 231. (c) Cassaretto, F. P.; McLafferty, J. J.;

Moore, C. E. Anal. Chim. Acta 1965, 32, 376.

[171] Hoffmann, K.; Weiss, E. J. Organomet. Chem. 1974, 67, 221.

260

[172] Alaviuhkola, T.; Bobacka, J.; Nissinen, M.; Rissanen, K.; Ivaska, A.; Pursiainen,

J. Chem. Eur. J. 2005, 11, 2071.

[173] Maruyama, K.; Ito T.; Yamamoto, A. Transition Met. Chem. 1980, 5, 14.

[174] (a) Turner W. R.; Elving, P. J. J. Phys. Chem. 1965, 69, 1067. (b) Turner W. R.;

Elving, P. J. Anal. Chem. 1965, 37, 207.

[175] Soriano Jartín, R.; Ligabue, A.; Soncini A.; Lazzeretti, P. J. Phys. Chem. A. 2002,

106, 11806.

[176] Hafner, K.; Stowasser, B.; Krimmer, H.-P.; Fischer, S.; Böhm M. C.; Lindner, H.

J. Angew. Chem., Int. Ed. Engl. 1986, 25, 630.

[177] Cary, D. R.; Green J. C.; O'Hare, D. Angew. Chem., Int. Ed. Engl. 1997, 36, 2618.

[178] The structures of 6.6, [6.6]2+, and [6.6]2- were optimized with the PBE0 hybrid

functional179 using the Ahlrichs' TZVP basis sets.180 All calculations were done

with the Turbomole 5.9.1 and Gaussian 03 program packages.181

[179] (a) Perdew, J. P.; Burke K.; Ernzerhof, M. Phys. Rev. Lett. 1996, 77, 3865. (b)

Perdew, J. P.; Burke K.; Ernzerhof, M. Phys. Rev. Lett. 1997, 78, 1396. (c)

Perdew, J. P.; Ernzerhof M.; Burke, K. J. Chem. Phys. 1996, 105, 9982. (d)

Ernzerhof, M.; Scuseria, G. E. J. Chem. Phys. 1999, 110, 5029.

[180] Schäfer, A.; Huber C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829.

[181] (a) TURBOMOLE, Program Package for ab initio Electronic Structure

Calculations, Version 5.9.1 Theoretical Chemistry Group, University of

Karlsruhe, Karlsruhe, Germany, 2007. (b) R. Ahlrichs, M. Bär, M. Häser, H. Horn

and C. Kölmel, Chem. Phys. Lett. 1989, 162, 165. (c) Gaussian 03, Revision D.02,

Frisch, M. J. et al. Gaussian, Inc., Wallingford CT, 2004.

261

[182] Ittel, S. D.; English, A. D.; Tolman, C. A.; Jesson, J. P. Inorg. Chim. Acta. 1979,

33, 101.

[183] Fagan, P. J.; Ward, M. D.; Calabrese, J. C. J. Am. Chem. Soc. 1989, 111, 1698.

[184] Wolczanski, P. T.; Bercaw, J. E. Organometallics 1982, 1, 793.

[185] Massey, A. G.; Park, J. J. Organomet. Chem. 1964, 2, 245.

[186] Jutzi, P.; Müller, C.; Stammler, A.; Stammler, H.-G. Organometallics 2000, 19,

1442.

[187] Otwinowski, Z.; Minor, W. Methods Enzymol. 1997, 276, 307.

[188] R. Hooft, COLLECT. Nonius BV, 1998, Delft, The Netherlands.

[189] Altomare, A.; Cascarano, M.; Giacovazzo, C.; Guagliardi, A. J. Appl. Cryst.

1993, 26, 343.

[190] Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W. P.; de Gelder, R.;

Israel, R.; Smits, J. M. M. The DIRDIF-94 program system, Technical Report of

the Crystallography Laboratory, 1994, University of Nijmegen, The Netherlands.

[191] Sheldrick, G. M. SHELXL97, 1997, University of Göttingen, Germany.

[192] DIAMOND – Visual Crystal Structure Information System, CRYSTAL

IMPACT, Postfach 1251, D-53002 Bonn.

[193] (a) The TZVP basis sets were used as they are referenced in the Turbomole 5.9.1

internal basis set library. See ftp://ftp.chemie.uni-karlsruhe.de/pub/ for explicit

basis set listings. (b) Bergner, A.; Dolg, M.; Kuechle, W.; Stoll, H.; Preuss, H.

Mol. Phys. 1993, 80, 1431.

[194] ADF2006.01, SCM, Theoretical Chemistry, Vrije Universiteit, Amsterdam,

Netherlands, http://www.scm.com.

262

[195] (a) Ziegler, T.; Rauk, A. Theor. Chem. Acta 1977, 46, 1. (b) Ziegler, T.; Rauk, A.

Inorg. Chem. 1979, 18, 1558. (c) Ziegler, T.; Rauk, A. Inorg. Chem. 1979, 18,

1755. (d) Bickelhaupt, F. M.; Baerends, E. J. Reviews in Computational

Chemistry; Eds. Lipkowitz, K. B.; Boyd, D. B.; Wiley: New York, 2000, 15, 1–

86.

[196] (a) Mayer, I. Chem. Phys. Lett. 1983, 97, 270. (b) Mayer, I. Chem. Phys. Lett.

1985, 117, 396. (c) Mayer, I. Theor. Chim. Acta 1985, 67, 315. (d) Mayer, I. Int.

J. Quant. Chem. 1986, 73, 477.

[197] (a) van Lenthe, E.; Baerends, E. J. J. Comp. Chem. 2003, 24, 1142. (b) Chong, D.

P.; Lenthe, E.; van Gisbergen, S. J. A.; Baerends, E. J. J. Comp. Chem. 2004, 25,

1030.

[198] (a) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1993, 99, 4597.

(b) van Lenthe, E.; Baerends, E. J.; Snijders, J. G. J. Chem. Phys. 1994, 101,

9783. (c) van Lenthe, E.; Ehlers, A. E.; Baerends, E. J. J. Chem. Phys. 1999, 110,

8943.

263

APPENDIX ONE

Tabulated Selected Crystal Data Collection Parameters and Selected Bond Lengths

and Bond Angles

264

Table 2.1. Selected Data and Structure Refinement Details for 2.3c, 2.3d and 2.4a.

2.3c 2.3d 2.4a

Empirical formula C26H24B2N2 C18H26B2N4 C10H23B2LiN2 Formula weight 386.09 320.05 199.86 Crystal system Triclinic Triclinic Orthorhombic

Space group P-1 P-1 P212121 a (Å) 9.274(5) 9.751(3) 8.573(5) b (Å) 10.117(7) 10.426(3) 9.925(6) c (Å) 12.837(8) 10.457(3) 14.892(9) α (deg) 106.43(3) 117.06(2) 90 β (deg) 91.33(3) 90.45(2) 90 γ (deg) 110.43(3) 101.59(2) 90 V (Å3) 1072.6(12) 921.4(5) 1267(1) Z 2 2 4 -3 dcalc (g cm ) 1.195 1.154 1.048 μ(Mo Kα) (mm-1) 0.068 0.068 0.058 Theta range (deg) 2.60 to 25.27 3.82 to 25.31 7.2 to 25.0

Independent reflections 3854(Rint = 0.0268) 3333(Rint = 0.0233) 1266(Rint = 0.032) Data/restraints/parameters 3854 / 0 / 272 3333 / 0 / 217 1266 / 0 / 142 GOF on F2 1.021 1.052 1.05

R1(F) [I > 2σ(I)] 0.0605 0.0436 0.044 2 wR2(F ) [all data] 0.1591 0.1035 0.098

265

Table 2.2. Selected Data and Structure Refinement Parameters for 2.5b, 2.5b(thf)3, 2.6b(thf) and 2.6b(thf)2.

2.5b 2.5b(thf)3 2.6b(thf) 2.6b(thf)2

Empirical formula C20H27B2N2Na C32H51B2N2NaO3 C24H35B2KN2O C28H43B2KN2O2 Formula weight 340.05 556.36 428.26 500.36 Crystal system Monoclinic Monoclinic Monoclinic Monoclinic Space group P21/n P21/c Cc Pn a (Å) 11.711(2) 17.4664) 6.055(1) 11.852(7) b (Å) 10.170(2) 9.358(2) 17.628(5) 6.165(2) c (Å) 17.149(3) 20.269(4) 22.867(6) 20.203(1) β (deg) 106.34(3) 90.48(3) 96.50(2) 96.06(2) V (Å3) 1959.8(7) 3313(1) 2425(1) 1468(1) Z 4 4 4 2 -3 dcalc (g cm ) 1.152 1.116 1.173 1.132 μ(Mo Kα) (mm-1) 0.085 0.080 0.236 0.207 Theta range (deg) 2.35 to 25.03 3.19 to 27.49 3.5 to 27.5 3.5 to 27.5

Independent reflections 3418(Rint = 0.027) 7579(Rint = 0.061) 2759(Rint = 0.023) 3335(Rint = 0.043) Data/restraints/parameters 3431 / 0 / 230 7579 / 0 / 366 2759 / 2 / 271 3335 / 2 / 316 GOF on F2 1.070 1.016 1.05 1.03

R1(F) [I > 2σ(I)] 0.0528 0.0593 0.036 0.053 2 wR2(F ) [all data] 0.1274 0.138 0.090 0.122

266

Table 2.3. Selected Data and Structure Refinement Parameters for 2.4c(thf)3, 2.5c(thf)3

and 2.6c(thf).

2.4c(thf)3 2.5c(thf)3 2.6c(thf)

Empirical formula C38H47B2LiN2O3 C38H47B2N2NaO3 C30H31B2KN2O Formula weight 608.34 624.39 496.29 Crystal system monoclinic monoclinic monoclinic

Space group C2/c P21 P21/n a (Å) 22.732(7) 9.154(5) 11.393(2) b (Å) 8.931(2) 14.153(5) 9.782(2) c (Å) 35.350(1) 13.912(7) 24.168(5) α (deg) 90 90 90 β (deg) 104.18(1) 100.47(2) 99.73(3) γ (deg) 90 90 90 V (Å3) 6958(3) 1772(2) 2654.8(9) Z 8 2 4 -3 dcalc (g cm ) 1.161 1.170 1.242 μ(Mo Kα) (mm-1) 0.071 0.083 0.226 Theta range (deg) 3.26 to 27.47 2.98 to 27.47 3.30 to 25.02

Independent reflections 7828(Rint = 0.0367) 4189(Rint = 0.047) 4623(Rint = 0.0356) Data/restraints/parameters 7828 / 0 / 415 4189 / 0 / 416 4623 / 0 / 325 GOF on F2 1.006 1.020 1.028

R1(F) [I > 2σ(I)] 0.0557 0.0555 0.0479 2 wR2(F ) [all data] 0.1295 0.1180 0.1205

267

Table 2.4. Selected Bond Lengths (Å) and Bond Angles (°) for 2.3c.

C(1)-B(1) 1.572(4) C(1)-B(2) 1.581(4) B(1)-N(1) 1.412(3) B(2)-N(2) 1.416(3) N(1)-N(2) 1.452(3) C(1)-C(2) 1.462(4) N(1)-C(3) 1.429(3) N(2)-C(9) 1.438(3) B(1)-C(15) 1.568(4) B(2)-C(21) 1.565(4) C(1)-B(1)-N(1) 108.4(2) C(1)-B(2)-N(2) 107.6(2) C(1)-B(1)-C(15) 126.8(2) C(1)-B(2)-C(21) 127.9(2) C(15)-B(1)-N(1) 124.7(2) C(21)-B(2)-N(2) 124.2(2) B(1)-N(1)-N(2) 110.4(2) B(2)-N(2)-N(1) 111.1(2) B(1)-N(1)-C(3) 129.7(2) B(2)-N(2)-C(9) 128.5(2) C(3)-N(1)-N(2) 118.1(2) C(9)-N(2)-N(1) 117.6(2) B(1)-C(1)-B(2) 102.3(2)

C(2)-C(1)-B(1) 121.0(2) Σpentagon angles 539.9 C(2)-C(1)-B(2) 125.3(2)

Table 2.5. Selected Bond Lengths (Å) and Bond Angles (°) for 2.3d.

C(1)-B(1) 1.578(2) C(1)-B(2) 1.592(2) B(1)-N(1) 1.497(2) B(2)-N(2) 1.467(2) N(1)-N(2) 1.457(2) C(1)-C(2) 1.550(2) N(1)-C(3) 1.415(2) N(2)-C(9) 1.409(2) B(1)-N(3) 1.389(2) B(2)-N(4) 1.399(2) C(1)-B(1)-N(1) 109.0(1) C(1)-B(2)-N(2) 109.5(1) C(1)-B(1)-N(3) 128.9(1) C(1)-B(2)-N(4) 126.8(1) N(3)-B(1)-N(1) 122.1(1) N(4)-B(2)-N(2) 123.5(1) B(1)-N(1)-N(2) 104.5(1) B(2)-N(2)-N(1) 109.1(1) B(1)-N(1)-C(3) 122.5(1) B(2)-N(2)-C(9) 129.6(1) C(3)-N(1)-N(2) 114.4(1) C(9)-N(2)-N(1) 114.5(1) B(1)-C(1)-B(2) 98.9(1)

C(2)-C(1)-B(1) 111.2(1) Σpentagon angles 531.0 C(2)-C(1)-B(2) 111.391)

268

Table 2.6. Selected Bond Lengths (Å) and Bond Angles (°) for 2.4a.

Li(1)-C(1) 2.480(6) Li(1’)-C(1) 2.225(6) Li(1)-B(1) 2.354(6) Li(1)-B(2) 2.370(6) Li(1)-N(1) 2.075(5) Li(1)-N(2) 2.112(5) Li(1’)···B(1) 2.543(6) Li(1’)···B(2) 2.750(6) Li(1’)···N(1) 3.313(6) Li(1’)···N(2) 3.179(6) Li(1)···Li(1’) 4.381(8) C(1)-B(1) 1.513(4) C(1)-B(2) 1.498(4) B(1)-N(1) 1.457(4) B(2)-N(2) 1.479(4) N(1)-N(2) 1.466(3) C(1)-C(2) 1.519(4) B(1)-C(3) 1.589(4) B(2)-C(4) 1.593(4) N(1)-C(5) 1.481(3) N(2)-C(8) 1.484(3)

C(1)-B(1)-N(1) 107.5(3) C(1)-B(2)-N(2) 107.6(2) C(1)-B(1)-C(3) 128.3(3) C(1)-B(2)-C(4) 128.8(3) C(3)-B(1)-N(1) 124.2(3) C(4)-B(2)-N(2) 123.4(3) B(1)-N(1)-N(2) 109.6(2) B(2)-N(2)-N(1) 108.3(2) B(1)-N(1)-C(5) 123.4(3) B(2)-N(2)-C(8) 125.1(2) C(5)-N(1)-N(2) 119.0(2) C(8)-N(2)-N(1) 118.7(2) B(1)-C(1)-B(2) 106.9(3)

C(2)-C(1)-B(1) 126.4(3) Σpentagon angles 539.9 C(2)-C(1)-B(2) 126.6(3)

Symmetry transformations used to generate equivalent atoms: x-1/2, -y+1/2, -z+1 and x+1/2, -y+1/2, -z+1

269

Table 2.7. Selected Bond Lengths (Å) and Bond Angles (°) for 2.5b, 2.5b(thf)3,

2.6b(thf) and 2.6b(thf)2.

a 2.5b (M = Na) 2.5b(thf)3 2.6b(thf) 2.6b(thf)2 (M = Na) (M = K) (M = K) C(1)-B(1) 1.514(3) 1.494(3) 1.503(4) 1.491(7) C(1)-B(2) 1.496(3) 1.498(3) 1.506(3) 1.486(6) B(1)-N(1) 1.452(3) 1.460(3) 1.458(4) 1.465(6) B(2)-N(2) 1.481(3) 1.453(3) 1.459(3) 1.465(6) N(1)-N(2) 1.449(2) 1.460(3) 1.444(3) 1.456(5) C(1)-C(2) 1.526(3) 1.521(3) 1.516(3) 1.530(6) B(1)-C(9) 1.581(3) 1.589(3) 1.595(4) 1.579(7) B(2)-C(15) 1.590(3) 1.586(3) 1.589(4) 1.584(6) N(1)-C(3) 1.482(3) 1.479(3) 1.481(3) 1.476(5) N(2)-C(6) 1.489(3) 1.471(3) 1.473(3) 1.472(5) M(1)-C(1) 2.801(2) 2.616(2) 3.058(2) 3.127(4) M(1’)a-C(1) 2.624(5) – 2.755(5) - 3.033(2) 3.156(4) M(1)-B(1) 2.884(2) 2.874(2) 3.211(3) 3.802(4) M(1)-B(2) 2.629(2) 3.101(3) 3.752(3) 3.425(4) M(1’)-B(1) 2.517(5) – 2.654(5) - 3.477(3) 3.364(4) M(1’)-B(2) 3.971(5) – 3.399(5) - 3.315(3) 3.895(4) M(1)-N(1) 2.872(2) 3.351(2) 3.898(3) 4.419(4) M(1)-N(2) 2.610(2) 3.540(2) 4.256(2) 4.143(4) M(1’)-N(1) 2.766(5) – 3.144(5) - 3.992(3) 4.131(4) M(1’)-N(2) 3.077(5) – 3.695(6) - 3.832(3) 4.483(4) M(1)-O(1) - 2.333(2) – 2.733(2) 2.726(4), 2.344(2) 2.816(4) M(1)···M(1’) 5.01(1) – 5.21(1) - 6.055(3) 6.165(4) B(1)-C(1)-B(2) 104.9(2) 103.7(2) 103.3(2) 104.2(4) C(2)-C(1)-B(1) 128.5(2) 129.7(2) 129.6(2) 128.8(4) C(2)-C(1)-B(2) 126.6(2) 126.5(2) 127.1(2) 126.9(4) C(1)-B(1)-N(1) 108.3(2) 110.0(2) 109.8(2) 110.1(4) C(1)-B(1)-C(9) 129.2(2) 128.0(2) 126.7(2) 127.4(4)

270

C(9)-B(1)-N(1) 122.5(2) 121.9(2) 123.5(2) 122.5(4) C(1)-B(2)-N(2) 109.2(2) 109.8(2) 110.1(2) 110.0(4) C(1)-B(2)-C(15) 126.3(2) 126.2(2) 126.0(2) 127.6(4) C(15)-B(2)-N(2) 123.5(2) 124.0(2) 123.8(2) 122.4(4) B(1)-N(1)-N(2) 109.8(2) 107.5(2) 108.5(2) 107.4(3) B(1)-N(1)-C(3) 125.7(2) 121.6(2) 126.0(2) 125.3(3) C(3)-N(1)-N(2) 120.3(2) 116.3(2) 112.8(2) 112.2(3) B(2)-N(2)-N(1) 107.1(2) 108.5(2) 107.9(2) 107.7(3) B(2)-N(2)-C(6) 124.4(2) 126.9(2) 124.6(2) 123.9(3) C(6)-N(2)-N(1) 114.7(2) 113.9(2) 111.3(2) 112.0(3)

Σpentagon angles 539.2 539.5 539.6 539.4 aIn complex 2.5b, M(1’) is Na(2) as shown in Figure 2.6.

Symmetry transformations used to generate equivalent atoms for 2.6b(thf): x+1, y, z and x-1, y, z

Symmetry transformations used to generate equivalent atoms for 2.6b(thf)2: x, y-1, z and x, y+1,z

271

Table 2.8. Selected Bond Lengths (Å) and Bond Angles (°) for 2.4c(thf)3, 2.5c(thf)3, and

2.6c(thf).

2.4c(thf)3 (M = Li) 2.5c(thf)3 (M = Na) 2.6c(thf) (M = K)

C(1)-B(1) 1.501(3) 1.482(6) 1.499(3) C(1)-B(2) 1.490(3) 1.499(6) 1.479(3) B(1)-N(1) 1.464(2) 1.466(5) 1.458(3) B(2)-N(2) 1.481(3) 1.486(5) 1.494(3) N(1)-N(2) 1.437(2) 1.438(4) 1.443(2) C(1)-C(2) 1.527(3) 1.522(5) 1.524(3) B(1)-C(15) 1.586(3) 1.589(3) 1.583(3) B(2)-C(21) 1.579(3) 1.586(3) 1.585(3) N(1)-C(3) 1.407(2) 1.415(4) 1.404(3) N(2)-C(9) 1.431(2) 1.418(5) 1.427(3) M(1)-C(1) 2.351(4) 2.680(4) 3.021(2) M(1’)-C(26) - - 3.052(2) M(1)-O 1.947(4) – 2.009(3) 2.325(3) – 2.360(4) 2.644(2) M(1)-B(1) 2.908(4) 3.189(5) 3.249(3) M(1)-B(2) 2.983(4) 2.939(5) 3.466(3) M(1)···N(1) 3.747(4) 3.572(3) 3.848(2) M(1)···N(2) 3.755(4) 3.540(3) 3.896(2) M(1)···M(1’) - - 7.363(1) B(1)-C(1)-B(2) 104.6(2) 105.7(3) 105.0(2) C(2)-C(1)-B(1) 128.6(2) 125.6(3) 126.8(2) C(2)-C(1)-B(2) 126.5(2) 128.2(3) 128.2(2) C(1)-B(1)-N(1) 108.6(2) 108.7(3) 108.9(2) C(1)-B(1)-C(15) 129.5(2) 129.4(3) 127.4(2) C(15)-B(1)-N(1) 121.5(2) 121.8(3) 123.2(2) C(1)-B(2)-N(2) 109.7(2) 108.3(3) 109.5(2) C(1)-B(2)-C(21) 131.0(2) 128.1(4) 129.7(2) C(21)-B(2)-N(2) 119.2(2) 123.3(3) 120.7(2) B(1)-N(1)-N(2) 109.5(2) 109.2(3) 109.3(2)

272

B(1)-N(1)-C(3) 131.6(2) 128.5(2) 131.2(2) C(3)-N(1)-N(2) 115.3(2) 116.0(3) 118.3(2) B(2)-N(2)-N(1) 107.1(1) 107.6(3) 106.7(2) B(2)-N(2)-C(9) 120.5(1) 125.0(3) 119.9(2) C(9)-N(2)-N(1) 113.5(1) 115.5(3) 114.2(2) O-M(1)-O 99.3(2) – 103.2(2) 89.1(1) – 110.3(1) - C(1)-M(1)-O 112.6(2) -120.(2) 105.6(1) – 143.1(1) 121.2(6)

Σpentagon angles 539.5 539.5 539.4

273

Table 3.1. Selected Data and Structure Refinement Parameters for 3.1b · 0.25 C5H12, 3.2, 3.3a, 3.3b and 3.4.

3.1b · 0.25 C5H12 3.2 3.3a 3.3b 3.4 Empirical formula C20H27B2Cl3N2Si C42.5H60B4GeN4 C20H46B4N4Sn C40H54B4N4Sn C21H48B4Cl2N4Sn Formula weight 469.53 742.78 504.54 752.80 589.46 Crystal system Monoclinic Orthorhombic Orthorhombic Monoclinic Monoclinic Space group P21/n P212121 P212121 P21/c P21/c a (Å) 6.755(6) 9.456(2) 10.835(1) 21.502(6) 17.337(4) b (Å) 19.491(9) 21.440(5) 14.179(1) 9.181(2) 9.619(3) c (Å) 20.603(10) 22.1033) 17.720(2) 20.537(7) 19.596(5) α (deg) 90 90 90 90 90 β (deg) 93.47(4) 90 90 97.32(1) 113.78(2) γ (deg) 90 90 90 90 90 V (Å3) 2708(3) 4481(2) 2722.5(4) 4021(2) 2990(1) Z 4 4 4 4 4 -3 dcalc (g cm ) 1.152 1.101 1.231 1.243 1.309

2θmax (deg) 50 25.0 52.8 27.5 27.5 μ(Mo Kα) (mm-1) 0.393 0.715 0.951 0.667 1.049

Independent reflections 4738 (Rint = 0.049) 7886 (Rint = 0.000) 5580 (Rint = 0.027) 9178 (Rint = 0.040) 6829 (Rint = 0.034) Data/restraints/parameters 4738 / 0 / 267 7886 / 0 / 462 5880 / 0 / 268 9178 / 0 / 442 6829 / 0 / 303 GOF on F2 1.11 1.109 1.030 1.02 1.008 0.0366 R1(F) [I > 2σ(I)] 0.066 0.0430 0.0221 0.042 2 0.0996 wR2(F ) [all data] 0.189 0.1251 0.0524 0.102

274

Table 3.2. Selected Bond Lengths (Å) and Bond Angles (°) for 3.1b · 0.25 C5H12.

Si(1)-C(1) 1.829(5) Si(1)-Cl(2) 2.038(2) Si(1)-Cl(1) 2.040(2) Si(1)-Cl(3) 2.039(2) C(1)-B(1) 1.592(7) C(1)-B(2) 1.584(7) B(1)-N(1) 1.382(7) B(2)-N(2) 1.405(7) N(1)-N(2) 1.454(5) C(1)-C(2) 1.568(6) B(1)-C(9) 1.574(7) B(2)-C(15) 1.560(7) N(1)-C(3) 1.477(5) N(2)-C(6) 1.473(6) C(1)-B(1)-N(1) 107.5(3) C(1)-B(2)-N(2) 107.6(2) C(1)-B(1)-C(9) 128.3(3) C(1)-B(2)-C(15) 128.8(3) C(9)-B(1)-N(1) 124.2(3) C(15)-B(2)-N(2) 123.4(3) B(1)-N(1)-N(2) 109.6(2) B(2)-N(2)-N(1) 108.3(2) B(1)-N(1)-C(3) 123.4(3) B(2)-N(2)-C(6) 125.1(2) C(3)-N(1)-N(2) 119.0(2) C(6)-N(2)-N(1) 118.7(2) B(1)-C(1)-B(2) 106.9(3) C(1)-Si(1)-Cl 104.0(8) - 107.3(8) C(2)-C(1)-B(1) 126.4(3) Cl-Si(1)-Cl 111.4(1) - 113.3(1)

C(2)-C(1)-B(2) 126.6(3) Si(1)-C(1) / CB2N2 64.1

C(2)-C(1)-Si(1) 106.0(3) C(2)-C(1) / CB2N2 48.1

Σpentagon angles 539.8

275

Table 3.3. Selected Bond Lengths (Å) and Bond Angles (°) for 3.2.

Ge(1)-C(1) 2.186(4) Ge(1)-C(21) 2.199(3) Ge(1)-B(1) 2.553(4) Ge(1)-B(3) 2.600(4) Ge(1)-B(2) 2.592(4) Ge(1)-B(4) 2.520(4) Ge(1)···N(1) 2.946(3) Ge(1)···N(3) 2.929(3) Ge(1)···N(2) 2.921(3) Ge(1)···N(4) 2.794(3) C(1)-B(1) 1.543(5) C(21)-B(3) 1.532(5) C(1)-B(2) 1.534(5) C(21)-B(4) 1.515(5) N(1)-B(1) 1.433(5) N(3)-B(3) 1.429(5) N(2)-B(2) 1.430(5) N(4)-B(4) 1.448(5) N(1)-N(2) 1.429(4) N(3)-N(4) 1.432(4) C(1)-C(2) 1.527(5) C(21)-C(22) 1.518(5)

B-C(Ph) 1.580(5) - 1.600(5) N-C(iPr) 1.469(5) - 1.486(4) Ge(1)-plane(1) 2.147(4) Ge(1)-plane(2) 2.119(4) C(1)-Ge(1)-C(21) 105.4(1) B(2)-C(1)-B(1) 103.2(3) B(4)-C(21)-B(3) 103.8(3) C(2)-C(1)-B(1) 125.5(3) B(3)-C(21)-C(22) 124.3(3) C(2)-C(1)-B(2) 124.2(3) B(4)-C(21)-C(22) 127.4(3) N(1)-B(1)-C(1) 108.0(3) N(3)-B(3)-C(21) 108.3(3) N(1)-B(1)-C(9) 124.0(3) N(3)-B(3)-C(29) 125.5(3) C(1)-B(1)-C(9) 127.8(3) C(21)-B(3)-C(29) 126.2(3) N(2)-B(2)-C(1) 108.1(3) N(4)-B(4)-C(21) 107.6(3) N(2)-B(2)-C(15) 124.6(3) N(4)-B(4)-C(35) 123.9(3) C(1)-B(2)-C(15) 127.2(3) C(21)-B(4)-C(35) 128.5(3) N(2)-N(1)-B(1) 109.9(3) B(3)-N(3)-N(4) 109.5(3) B(1)-N(1)-C(3) 127.0(3) B(3)-N(3)-C(23) 129.6(3) N(2)-N(1)-C(3) 122.8(3) N(4)-N(3)-C(23) 120.3(3) N(1)-N(2)-B(2) 110.1(3) N(3)-N(4)-B(4) 109.3(3) B(2)-N(2)-C(6) 129.4(3) B(4)-N(4)-C(26) 125.6(3) N(1)-N(2)-C(6) 119.8(3) N(3)-N(4)-C(26) 123.5(3)

CB2N2/CB2N2 53.1(1)

ΣC(1)-B(1)-N(1)-B(2)-N(2) 539.3 ΣC(21)-B(3)-N(3)-B(4)-N(4) 538.5

276

Table 3.4. Selected Bond Lengths (Å) and Bond Angles (°) for 3.3b.

Sn(1)-C(1) 2.435(3) Sn(1)-C(21) 2.451(3) Sn(1)-B(1) 2.706(3) Sn(1)-B(3) 2.749(4) Sn(1)-B(2) 2.715(3) Sn(1)-B(4) 2.666(4) Sn(1)···N(1) 2.947(3) Sn(1)···N(3) 2.970(2) Sn(1)···N(2) 2.981(3) Sn(1)···N(4) 2.846(3) C(1)-B(1) 1.517(4) C(21)-B(3) 1.533(4) C(1)-B(2) 1.530(4) C(21)-B(4) 1.527(5) N(1)-B(1) 1.450(4) N(3)-B(3) 1.438(4) N(2)-B(2) 1.437(4) N(4)-B(4) 1.451(4) N(1)-N(2) 1.426(3) N(3)-N(4) 1.425(3) C(1)-C(2) 1.533(4) C(21)-C(22) 1.524(4)

B-C(Ph) 1.571(4) - 1.588(4) N-C(iPr) 1.473(4) - 1.482(4) Sn(1)-plane(1) 2.355(3) Sn(1)-plane(2) 2.330(3) C(1)-Sn(1)-C(21) 103.92(9) B(2)-C(1)-B(1) 104.9(2) B(4)-C(21)-B(3) 104.4(2) C(2)-C(1)-B(1) 125.8(3) B(3)-C(21)-C(22) 125.3(3) C(2)-C(1)-B(2) 126.2(3) B(4)-C(21)-C(22) 127.2(3) N(1)-B(1)-C(1) 107.3(3) N(3)-B(3)-C(21) 107.7(3) N(1)-B(1)-C(9) 123.2(3) N(3)-B(3)-C(29) 125.1(3) C(1)-B(1)-C(9) 129.5(3) C(21)-B(3)-C(29) 127.2(3) N(2)-B(2)-C(1) 107.4(2) N(4)-B(4)-C(21) 106.9(3) N(2)-B(2)-C(15) 123.0(3) N(4)-B(4)-C(35) 123.5(3) C(1)-B(2)-C(15) 129.6(3) C(21)-B(4)-C(35) 129.6(3) N(2)-N(1)-B(1) 109.8(2) B(3)-N(3)-N(4) 109.9(2) B(1)-N(1)-C(3) 120.3(2) B(3)-N(3)-C(23) 129.2(2) N(2)-N(1)-C(3) 129.7(2) N(4)-N(3)-C(23) 120.0(2) N(1)-N(2)-B(2) 110.3(2) N(3)-N(4)-B(4) 110.2(2) B(2)-N(2)-C(6) 126.4(2) B(4)-N(4)-C(26) 121.6(2) N(1)-N(2)-C(6) 123.3(2) N(3)-N(4)-C(26) 124.5(2)

CB2N2/CB2N2 50.2(1)

ΣC(1)-B(1)-N(1)-B(2)-N(2) 539.7 ΣC(21)-B(3)-N(3)-B(4)-N(4) 539.1

277

Table 3.5. Selected Bond Lengths (Å) and Bond Angles (°) for 3.3a.

Sn(1)-C(1) 2.428(2) Sn(1)-C(2) 2.432(2) Sn(1)-B(1) 2.766(2) Sn(1)-B(3) 2.774(3) Sn(1)-B(2) 2.672(3) Sn(1)-B(4) 2.663(2) Sn(1)-N(2) 2.834(2) Sn(1)-N(4) 2.7437(2) Sn(1)···N(1) 2.998(2) Sn(1)···N(3) 2.9360(2) C(1)-B(1) 1.524(3) C(2)-B(3) 1.523(4) C(1)-B(2) 1.524(4) C(2)-B(4) 1.520(3) B(1)-N(1) 1.433(3) B(3)-N(3) 1.439(4) B(2)-N(2) 1.445(3) B(4)-N(4) 1.453(4) N(1)-N(2) 1.443(2) N(3)-N(4) 1.441(3) C(1)-C(17) 1.517(3) C(2)-C(27) 1.526(4)

B-C(Me) 1.578(4) - 1.596(4) N-C(iPr) 1.464(3) - 1.479(3) Sn(1)-plane(1) 2.347(1) Sn(1)-plane(2) 2.366(2) C(1)-Sn(1)-C(2) 102.53(8) B(1)-C(1)-B(2) 104.9(2) B(3)-C(2)-B(4) 105.0(2) C(17)-C(1)-B(1) 125.2(2) C(27)-C(2)-B(3) 124.1(2) C(17)-C(1)-B(2) 125.9(2) C(27)-C(2)-B(4) 126.5(2) N(1)-B(1)-C(1) 107.0(2) N(3)-B(3)-C(2) 107.4(2) N(1)-B(1)-C(18) 125.4(2) N(3)-B(3)-C(28) 125.5(3) C(1)-B(1)-C(18) 126.7(2) C(2)-B(3)-C(28) 127.1(3) N(2)-B(2)-C(1) 106.4(2) N(4)-B(4)-C(2) 106.1(2) N(2)-B(2)-C(19) 125.9(2) N(4)-B(4)-C(29) 125.5(2) C(1)-B(2)-C(19) 127.6(2) C(2)-B(4)-C(29) 128.4(3) N(2)-N(1)-B(1) 109.3(2) N(4)-N(3)-B(3) 109.5(2) N(2)-N(1)-C(11) 119.5(2) N(4)-N(3)-C(21) 121.0(2) C(11)-N(1)-B(1) 130.6(2) C(21)-N(3)-B(3) 128.9(2) N(1)-N(2)-B(2) 109.9(2) N(3)-N(4)-B(4) 109.3(2) N(1)-N(2)-C(14) 118.8(2) N(3)-N(4)-C(24) 118.6(2) C(14)-N(2)-B(2) 129.8(2) C(24)-N(4)-B(4) 130.5(2)

CB2N2/CB2N2 50.0(1)

ΣC(1)-B(1)-N(1)-B(2)-N(2) 537.5 ΣC(2)-B(3)-N(3)-B(4)-N(4) 537.3

278

Table 3.6. Selected Bond Lengths (Å) and Bond Angles (°) for 3.4.

Si(1)-C(1) 2.184(3) Si(1)-Cl(1) 2.385(1) Si(1)-C(2) 2.195(3) Si(1)-C(12) 2.167(3) C(2)-B(1) 1.564(5) C(12)-B(3) 1.570(6) C(2)-B(2) 1.554(5) C(12)-B(4) 1.572(6) B(1)-N(1) 1.404(4) B(3)-N(3) 1.410(5) B(2)-N(2) 1.414(5) B(4)-N(4) 1.408(5) N(1)-N(2) 1.447(4) N(3)-N(4) 1.457(4) C(2)-C(3) 1.547(5) C(12)-C(13) 1.552(5)

B-C(Me) 1.573(6) - 1.585(5) N-C(iPr) 1.463(5) - 1.482(4) C(1)-Cl(2) 1.769(3) C(12)-Sn(1)-C(1) 109.7(1) C(12)-Sn(1)-Cl(1) 104.0(1) C(12)-Sn(1)-C(2) 120.8(1) C(1)-Sn(1)-Cl(1) 101.2(1) C(1)-Sn(1)-C(2) 111.5(1) C(2)-Sn(1)-Cl(1) 107.4(1) B(1)-C(2)-B(2) 101.9(3) B(3)-C(12)-B(4) 101.9(3) C(3)-C(2)-B(1) 120.3(3) C(13)-C(12)-B(3) 115.8(3) C(3)-C(2)-B(2) 122.4(3) C(13)-C(12)-B(4) 117.4(3) C(2)-B(1)-N(1) 108.2(3) C(12)-B(3)-N(3) 108.7(3) C(2)-B(1)-C(4) 125.0(3) C(12)-B(3)-C(14) 125.3(3) C(4)-B(1)-N(1) 126.7(3) C(14)-B(3)-N(3) 125.8(3) C(2)-B(2)-N(2) 108.0(3) C(12)-B(4)-N(4) 108.4(3) C(2)-B(2)-C(5) 126.5(3) C(12)-B(4)-C(15) 126.8(3) C(5)-B(2)-N(2) 125.5(3) C(15)-B(4)-N(4) 124.8(4) B(1)-N(1)-N(2) 110.3(2) B(3)-N(3)-N(4) 110.1(3) B(1)-N(1)-C(6) 128.8(3) B(3)-N(3)-C(16) 128.2(3) C(6)-N(1)-N(2) 120.2(3) C(16)-N(3)-N(4) 119.3(3) B(2)-N(2)-N(1) 110.1(3) B(4)-N(4)-N(3) 110.8(3) B(2)-N(2)-C(9) 130.5(3) B(4)-N(4)-C(19) 126.7(3) C(9)-N(2)-N(1) 118.4(3) C(19)-N(4)-N(3) 122.2(3)

Sn(1)-C(2)/CB2 73.6 Sn(1)-C(12)/CB2N2 59.7

C(2)-C(31)/CB2 34.2 C(12)-C(13)/CB2N2 44.7

ΣC(2)-B(1)-N(1)-B(2)-N(2) 538.5 ΣC(12)-B(3)-N(3)-B(4)-N(4) 539.9

279

Table 4.1. Selected Data and Structure Refinement Details for 4.1a, 4.1b, 4.2a · BrLi(thf)3, 4.2b and 4.3.

4.1a 4.1b 4.2a·BrLi(thf)3 4.2b 4.3 Empirical formula C20H46B4N4Zn C40H54B4N4Zn C32H68B4BrCdLiN4O3 C40H54B4CdN4 C20H46B4HgN4 Formula weight 549.34 699.48 799.40 746.51 586.44 Crystal system Monoclinic triclinic Monoclinic triclinic Monoclinic Space group P21/n P1¯ P21/n P1¯ I2/m a (Å) 9.396(2) 10.681(3) 9.396(2) 10.940(3) 16.095(2) b (Å) 21.241(7) 12.141(4) 21.241(7) 12.001(4) 15.295(2) c (Å) 21.107(7) 17.176(4) 21.107(7) 16.987(5) 11.350(1) α (deg) 90 75.483(14) 90 76.011(16) 90 β (deg) 96.85(2) 80.656(15) 96.85(2) 80.947(16) 101.249(1) γ (deg) 90 66.124(13) 90 66.866(16) 90 3 V (Å ) 4182(2) 1967(1) 4182(2) 1985(1) 2740.6(5) Z 4 2 4 2 4 -3 dcalc (g cm ) 1.270 1.181 1.270 1.249 1.421

2θmax (deg) 50.2 55.0 50.2 55.0 52.8 μ (Mo Kα) (mm-1) 1.511 0.657 1.511 0.582 5.629

Independent reflections 7345(Rint = 0.024) 8921(Rint = 0.029) 7345 (Rint = 0.024) 9039 (Rint = 0.038) 2918(Rint = 0.031) Data/restraints/parameters 7345 / 0 / 425 8921 / 0 / 451 7345 / 0 / 425 9039 / 0 / 442 2918 / 0 / 240 GOF on F2 1.04 1.02 1.04 1.12 1.08 0.020 R1(F) [I > 2σ(I)] 0.038 0.042 0.038 0.061 2 0.050 wR2(F ) [all data] 0.097 0.109 0.097 0.189

280

Table 4.2. Selected Bond Lengths (Å) and Bond Angles (°) for 4.1a.

Molecule 1 Zn(1)-C(1) 2.002(3) Zn(1)-C(11) 2.001(3) Zn(1)-B(1) 2.537(4) Zn(1)-B(3) 2.528(4) Zn(1)-B(2) 2.443(4) Zn(1)-B(4) 2.478(4) C(1)-B(1) 1.532(5) C(11)-B(3) 1.525(5) C(1)-B(2) 1.539(5) C(11)-B(4) 1.537(5) N(1)-B(1) 1.432(5) N(3)-B(3) 1.429(6) N(2)-B(2) 1.442(5) N(4)-B(4) 1.412(5) N(1)-N(2) 1.443(4) N(3)-N(4) 1.433(5) C(1)-C(2) 1.534(5) C(11)-C(12) 1.527(5)

B-C(Me) 1.578(6) – 1.589(6) N-C(iPr) 1.451(8) – 1.476(4) C(1)-Zn(1)-C(11) 105.4(1) B(2)-C(1)-B(1) 104.1(3) B(4)-C(11)-B(3) 103.8(3) C(2)-C(1)-B(1) 126.1(3) B(3)-C(11)-C(12) 126.6(3) C(2)-C(1)-B(2) 124.1(3) B(4)-C(11)-C(12) 123.3(3) N(1)-B(1)-C(1) 107.7(3) N(3)-B(3)-C(11) 107.5(3) N(1)-B(1)-C(3) 123.3(3) N(3)-B(3)-C(13) 124.6(4) C(1)-B(1)-C(3) 129.0(3) C(21)-B(3)-C(13) 127.9(4) N(2)-B(2)-C(1) 107.4(3) N(4)-B(4)-C(11) 107.7(3) N(2)-B(2)-C(4) 125.6(3) N(4)-B(4)-C(14) 125.4(4) C(1)-B(2)-C(4) 127.0(3) C(21)-B(4)-C(14) 126.9(4) N(2)-N(1)-B(1) 110.4(3) B(3)-N(3)-N(4) 110.1(3) B(1)-N(1)-C(5) 124.5(3) B(3)-N(3)-C(15) 143.4(5) N(2)-N(1)-C(5) 121.9(3) N(4)-N(3)-C(15) 105.1(4) N(1)-N(2)-B(2) 109.3(2) N(3)-N(4)-B(4) 110.0(3) B(2)-N(2)-C(8) 128.8(3) B(4)-N(4)-C(18) 145.7(5) N(1)-N(2)-C(8) 119.0(2) N(3)-N(4)-C(18) 102.3(4)

Zn(1)-C(1) / CB2N2 81.1 Zn(1)-C(11) / CB2N2 82.6

C(1)-C(2) / CB2N2 12.2 C(11)-C(12) / CB2N2 14.4

ΣC(1)-B(1)-N(1)-B(2)-N(2) 538.9 ΣC(11)-B(3)-N(3)-B(4)-N(4) 539.1

281

Molecule 2 C(21)-B(5) 1.541(7) Zn(2)-C(21) 2.004(4) C(21)-B(6) 1.507(6) Zn(2)-B(5) 2.445(5) N(5)-B(5) 1.420(6) Zn(2)-B(6) 2.546(5)

N(6)-B(6) 1.406(5) B-C(Me) 1.586(7) – 1.59(2)

N(5)-N(6) 1.424(4) N-C(iPr) 1.477(5) – 1.484(5)

C(21)-C(22) 1.56(3) C(21)-Zn(2)-C(21’) N(5)-B(5)-C(21) 107.7(3) N(6)-B(6)-C(21) 108.4(4) C(21)-B(5)-C(23) 113.8(7) C(21)-B(6)-C(24) 127.7(4) N(5)-B(5)-C(23) 138.4(7) N(6)-B(6)-C(24) 123.8(4) B(5)-N(5)-N(6) 108.8(3) B(6)-N(6)-N(5) 110.9(3) B(5)-N(5)-C(25) 132.9(4) B(6)-N(6)-C(28) 122.4(4) N(6)-N(5)-C(25) 116.2(3) N(5)-N(6)-C(28) 124.5(3)

B(6)-C(21)-B(5) 103.1(3) Zn(2)-C(21) / CB2N2 80.3

B(5)-C(21)-C(22) 137.4(8) C(21)-C(22) / CB2N2 16.2

B(6)-C(21)-C(22) 114.2(11) ∑C(21)-B(5)-N(5)-B(6)-N(6) 538.9

Symmetry transformations used to generate equivalent atoms: -x, -y+2, -z

282

Table 4.3. Selected Bond Lengths (Å) and Bond Angles (°) for 4.1b.

Zn(1)-C(1) 1.999(2) Zn(1)-C(21) 2.004(2) Zn(1)···B(1) 2.626(2) Zn(1)···B(3) 2.564(2) Zn(1)···B(2) 2.603(2) Zn(1)···B(4) 2.570(2) C(1)-B(1) 1.550(3) C(21)-B(3) 1.556(3) C(1)-B(2) 1.542(3) C(21)-B(4) 1.537(3) N(1)-B(1) 1.427(3) N(3)-B(3) 1.424(3) N(2)-B(2) 1.429(3) N(4)-B(4) 1.433(3) N(1)-N(2) 1.446(2) N(3)-N(4) 1.443(2) C(1)-C(2) 1.534(3) C(21)-C(22) 1.527(3)

B-C(Ph) 1.574(3) – 1.581(3) N-C(iPr) 1.470(3) – 1.479(2)

C-Caromatic 1.359(4) - 1.398(3) C-Caliphatic 1.508(4) - 1.549(4) C(1)-Zn(1)-C(21) 174.71(8) B(2)-C(1)-B(1) 102.77(16) B(4)-C(21)-B(3) 102.68(17) C(2)-C(1)-B(1) 125.78(17) B(3)-C(21)-C(22) 126.73(18) C(2)-C(1)-B(2) 123.89(17) B(4)-C(21)-C(22) 122.59(18) N(1)-B(1)-C(1) 108.21(18) N(3)-B(3)-C(21) 108.48(17) N(1)-B(1)-C(3) 123.79(19) N(3)-B(3)-C(23) 125.87(18) C(1)-B(1)-C(3) 127.96(18) C(21)-B(3)-C(23) 125.66(18) N(2)-B(2)-C(1) 108.99(17) N(4)-B(4)-C(21) 108.52(17) N(2)-B(2)-C(9) 126.04(18) N(4)-B(4)-C(29) 122.7(2) C(1)-B(2)-C(9) 124.97(18) C(21)-B(4)-C(29) 128.7(2) N(2)-N(1)-B(1) 110.24(16) B(3)-N(3)-N(4) 109.77(16) B(1)-N(1)-C(15) 124.6(2) B(3)-N(3)-C(35) 127.83(16) N(2)-N(1)-C(15) 120.19(17) N(4)-N(3)-C(35) 117.22(15) N(1)-N(2)-B(2) 109.38(15) N(3)-N(4)-B(4) 110.02(17) B(2)-N(2)-C(18) 128.08(18) B(4)-N(4)-C(38) 126.09(18) N(1)-N(2)-C(18) 117.05(16) N(3)-N(4)-C(38) 120.24(18)

Zn(1)-C(1) / CB2N2 87.8 Zn(1)-C(21) / CB2N2 86.9

C(1)-C(2) / CB2N2 20.2 C(21)-C(22) / CB2N2 19.7

ΣC(1)-B(1)-N(1)-B(2)-N(2) 539.6 ΣC(21)-B(3)-N(3)-B(4)-N(4) 539.5

283

Table 4.4. Selected Bond Lengths (Å) and Bond Angles (°) for 4.2a · LiBr(thf)3.

Cd(1)-C(1) 2.248(3) Cd(1)-C(11) 2.255(3) Cd(1)···B(1) 2.872(4) Cd(1)···B(3) 2.796(4) Cd(1)···B(2) 2.831(4) Cd(1)···B(4) 2.828(4) C(1)-B(1) 1.529(6) C(11)-B(3) 1.522(6) C(1)-B(2) 1.519(5) C(11)-B(4) 1.526(6) N(1)-B(1) 1.433(6) N(3)-B(3) 1.438(5) N(2)-B(2) 1.426(5) N(4)-B(4) 1.442(5) N(1)-N(2) 1.438(5) N(3)-N(4) 1.457(4) C(1)-C(2) 1.533(5) C(11)-C(12) 1.531(5)

B-C(Me) 1.589(6) – 1.591(6) Cd(1)···Br(1) 2.731(8)

N-C(iPr) 1.40(1) – 1.47(1) Br(1)-Li(1) 2.229(2)

Li(1)-O(thf) 1.920(7) – 1.941(7) C(1)-Cd(1)-C(11) 159.3(1) B(2)-C(1)-B(1) 103.1(3) B(4)-C(11)-B(3) 103.6(3) C(2)-C(1)-B(1) 124.4(3) B(3)-C(11)-C(12) 125.3(3) C(2)-C(1)-B(2) 124.5(3) B(4)-C(11)-C(12) 124.6(3) N(1)-B(1)-C(1) 108.2(3) N(3)-B(3)-C(11) 108.7(3) N(1)-B(1)-C(3) 125.0(4) N(3)-B(3)-C(13) 125.5(4) C(1)-B(1)-C(3) 126.5(4) C(21)-B(3)-C(13) 125.4(3) N(2)-B(2)-C(1) 109.1(3) N(4)-B(4)-C(11) 108.5(3) N(2)-B(2)-C(4) 122.3(4) N(4)-B(4)-C(14) 124.8(4) C(1)-B(2)-C(4) 128.6(4) C(11)-B(4)-C(14) 126.7(4) N(2)-N(1)-B(1) 109.3(3) B(3)-N(3)-N(4) 108.6(3) B(1)-N(1)-C(5) 141.3(5) B(3)-N(3)-C(15) 130.0(3) N(2)-N(1)-C(5) 103.2(5) N(4)-N(3)-C(15) 115.8(3) N(1)-N(2)-B(2) 109.3(3) N(3)-N(4)-B(4) 109.2(3) B(2)-N(2)-C(8) 141.4(5) B(4)-N(4)-C(18) 126.1(3) N(1)-N(2)-C(8) 103.8(5) N(3)-N(4)-C(18) 116.0(3)

Cd(1)-C(1) / CB2N2 85.0 Cd(1)-C(11) / CB2N2 87.7

C(1)-C(2) / CB2N2 15.6 C(11)-C(12) / CB2N2 19.1

ΣC(1)-B(1)-N(1)-B(2)-N(2) 539.0 ΣC(11)-B(3)-N(3)-B(4)-N(4) 538.6

284

Table 4.5. Selected Bond Lengths (Å) and Bond Angles (°) for 4.2b.

Cd(1)-C(1) 2.201(5) Cd(1)-C(21) 2.206(5) Cd(1)···B(1) 2.706(6) Cd(1)···B(3) 2.734(5) Cd(1)···B(2) 2.711(5) Cd(1)···B(4) 2.758(6) C(1)-B(1) 1.555(7) C(21)-B(3) 1.543(8) C(1)-B(2) 1.535(8) C(21)-B(4) 1.542(7) N(1)-B(1) 1.423(7) N(3)-B(3) 1.432(7) N(2)-B(2) 1.444(7) N(4)-B(4) 1.429(7) N(1)-N(2) 1.450(6) N(3)-N(4) 1.445(6) C(1)-C(2) 1.517(7) C(21)-C(22) 1.535(7)

B-C(Ph) 1.570(7) – 1.588(8) N-C(iPr) 1.472(6) – 1488(6)

C-Caromatic 1.365(9) - 1.407(8) C-Caliphatic 1.504(9) - 1.534(8) C(1)-Cd(1)-C(21) 175.3(2) B(2)-C(1)-B(1) 102.8(4) B(4)-C(21)-B(3) 103.5(4) C(2)-C(1)-B(1) 122.8(4) B(3)-C(21)-C(22) 124.8(4) C(2)-C(1)-B(2) 127.8(5) B(4)-C(21)-C(22) 126.3(4) N(1)-B(1)-C(1) 108.6(4) N(3)-B(3)-C(21) 108.2(4) N(1)-B(1)-C(3) 125.5(5) N(3)-B(3)-C(23) 127.4(5) C(1)-B(1)-C(3) 125.9(5) C(21)-B(3)-C(23) 124.4(4) N(2)-B(2)-C(1) 109.0(4) N(4)-B(4)-C(21) 108.0(4) N(2)-B(2)-C(9) 121.3(5) N(4)-B(4)-C(29) 123.1(4) C(1)-B(2)-C(9) 129.6(5) C(21)-B(4)-C(29) 128.6(5) N(2)-N(1)-B(1) 110.2(4) B(3)-N(3)-N(4) 109.8(4) B(1)-N(1)-C(15) 128.2(4) B(3)-N(3)-C(35) 127.0(4) N(2)-N(1)-C(15) 116.3(4) N(4)-N(3)-C(35) 116.6(4) N(1)-N(2)-B(2) 109.1(4) N(3)-N(4)-B(4) 110.2(4) B(2)-N(2)-C(18) 125.7(4) B(4)-N(4)-C(38) 124.6(4) N(1)-N(2)-C(18) 119.2(4) N(3)-N(4)-C(38) 120.7(4)

Cd(1)-C(1) / CB2N2 87.6 Cd(1)-C(21) / CB2N2 87.9

C(1)-C(2) / CB2N2 17.6 C(21)-C(22) / CB2N2 18.4

ΣC(1)-B(1)-N(1)-B(2)-N(2) 539.7 ΣC(21)-B(3)-N(3)-B(4)-N(4) 539.7

285

Table 4.6. Selected Bond Lengths (Å) and Bond Angles (°) for 4.3.

Hg(1)-C(10) 2.168(6) Hg(1)-C(20) 2.177(6) Hg(1)-B(11) 2.780(6) Hg(1)-B(21) 2.778(7) Hg(1)-B(12) 2.759(7) Hg(1)-B(22) 2.729(7) C(10)-B(11) 1.557(9) C(20)-B(21) 1.553(9) C(10)-B(12) 1.546(10) C(20)-B(22) 1.550(10) N(11)-B(11) 1.407(9) N(21)-B(21) 1.390(12) N(12)-B(12) 1.410(10) N(22)-B(22) 1.416(11) N(11)-N(12) 1.454(7) N(21)-N(22) 1.486(12) C(10)-C(17) 1.536(11) C(20)-C(27) 1.524(9)

B-C(Me) 1.57(1) – 1.60(1) N-C(iPr) 1.41(2) – 1.61(2) C(10)-Hg(1)-C(10’) 175.6(3) C(20)-Hg(1)-C(20’) 176.1(3) B(12)-C(10)-B(11) 102.2(5) B(22)-C(20)-B(21) 103.0(6) C(17)-C(10)-B(11) 124.8(6) B(21)-C(20)-C(27) 123.8(6) C(17)-C(10)-B(12) 123.3(6) B(22)-C(20)-C(27) 124.7(6) N(11)-B(11)-C(10) 108.4(5) N(21)-B(21)-C(20) 109.3(7) N(11)-B(11)-C(18) 126.5(6) N(21)-B(21)-C(28) 124.3(8) C(10)-B(11)-C(18) 125.0(6) C(20)-B(21)-C(28) 126.4(7) N(12)-B(12)-C(10) 108.8(6) N(22)-B(22)-C(20) 107.6(6) N(12)-B(12)-C(19) 124.0(6) N(22)-B(22)-C(29) 126.0(7) C(10)-B(12)-C(19) 127.2(6) C(20)-B(22)-C(29) 126.4(7) N(12)-N(11)-B(11) 109.8(5) B(21)-N(21)-N(22) 108.8(7) B(11)-N(11)-C(11) 136.7(8) B(21)-N(21)-C(21) 115.2(9) B(11)-N(11)-C(11’) 121.0(7) B(21)-N(21)-C(21’) 147.6(10) N(12)-N(11)-C(11) 108.8(7) N(22)-N(21)-C(21) 131.2(8) N(12)-N(11)-C(11’) 122.8(7) N(22)-N(21)-C(21’) 100.1(8) N(11)-N(12)-B(12) 109.9(5) N(21)-N(22)-B(22) 110.3(6) B(12)-N(12)-C(14) 113.3(8) B(22)-N(22)-C(24) 140.2(9) B(12)-N(12)-C(14’) 134.7(8) B(22)-N(22)-C(24’) 111.1(8) N(11)-N(12)-C(14) 130.0(7) N(21)-N(22)-C(24) 104.4(9) N(11)-N(12)-C(14’) 109.9(7) N(21)-N(22)-C(24’) 132.4(8)

Hg(1)-C(10) / CB2N2 87.6 Hg(1)-C(20) / CB2N2 88.9

C(10)-C(17) / CB2N2 20.2 C(20)-C(27) / CB2N2 21.7

ΣC(10)-B(11)-N(11)-B(12)-N(12) 539.1 ΣC(20)-B(21)-N(21)-B(22)-N(22) 539.0

Symmetry transformations used to generate equivalent atoms via twofold axis: 1/2, y, 1/2

286

Table 5.2. Selected Data and Structure Refinement Parameters for 5.1b, 5.1e, 5.2, 5.3 and 5.4.

5.1b 5.1e 5.2 5.3 5.4 Empirical formula C31H46B2N2Rh C25H31B2N2Rh C34H38B4FeN4 C27H34B2N2Ru C27H34B2Cl2N2Zr Formula weight 571.23 484.05 601.77 509.25 570.30 Crystal system Monoclinic Orthorhombic Monoclinic Monoclinic Monoclinic Space group P21/c F d d 2 P21/c P 21/c P 21/c a (Å) 19.266(6) 27.714(2) 11.770(5) 12.180(3) 8.452(3) b (Å) 10.158(2) 38.709(5) 15.036(3) 8.6670(11) 18.184(4) c (Å) 15.259(7) 8.2410(15) 17.927(7) 24.155(6) 17.094(7) α (deg) 90 90 90 90 90 β (deg) 103.398(14) 90 105.487(14) 101.776(11) 95.829(14) γ (deg) 90 90 90 90 90 V (Å3) 2905(2) 8841(2) 3057.4(19) 2496.2(9) 2613.6(15) Z 4 16 4 4 4 -3 dcalc (g cm ) 1.306 1.455 1.307 1.355 1.449 μ(Mo Kα) (mm-1) 0.610 0.787 0.525 0.645 0.644

2θmax (deg) 27.5 27.46 27.46 27.48 27.38.

Independent reflections 6607(Rint= 0.027) 2699(Rint = 0.0262) 6954(Rint = 0.0437) 5685(Rint = 0.0322) 5893(Rint = 0.0320) Data/restraints/parameters 6607 / 0 / 325 2699 / 1 / 272 6954 / 0 / 390 5685 / 0 / 294 5893 / 0 / 313 GOF on F2 1.06 1.067 1.031 1.039 1.027 0.037 0.0280 0.0477 0.0342 0.0386 R1(F) [I > 2σ(I)] 2 0.097 0.0628 0.1176 0.0884 0.0936 wR2(F ) [all data]

287

Table 5.3. Selected Data and Structure Refinement Parameters for 5.6c, 5.7, 5.9(tmeda)2 and 5.10.

5.6c 5.7 5.9(tmeda)2 5.10 Empirical formula C24H20B2N2O C28H28B4N2 C20 H29 B2 K N3 C48H56B4N2Ru2 · 1.5 C6H14 Formula weight 374.04 435.76 372.18 1035.59 Crystal system Monoclinic Monoclinic Triclinic Monoclinic Space group C2/c P21/n P -1 C2 a (Å) 23.629(2) 14.1990(10) 7.4770(5) 24.955(8) b (Å) 10.785(4) 6.0800(2) 11.9330(9) 19.629(5) c (Å) 8.325(5) 14.6470(11) 12.3420(10) 10.626(3) α (deg) 90 90 94.030(3) 90 β (deg) 109.72(2) 103.012(3) 97.951(4) 114.964(15) γ (deg) 90 90 101.459(5) 90 V (Å3) 1997(1) 1232.01(13) 1063.43(14) 4719(2) Z 4 2 2 4 -3 dcalc (g cm ) 1.244 1.175 1.162 1.458 μ(Mo Kα) (mm-1) 0.075 0.066 0.258 0.682

2θmax (deg) 27.47 27.42 24.9 27.48

Independent reflections 2262(Rint = 0.0355) 2777(Rint = 0.0305) 3649(Rint = 0.051) 5560(Rint = 0.0333) Data/restraints/parameters 2262 / 0 / 132 2777 / 0 / 154 3649 / 0 / 236 5560 / 1 / 517 GOF on F2 0.995 1.004 1.04 1.061 0.0401 0.0502 0.051 0.0387 R1(F) [I > 2σ(I)] 2 0.1057 0.1296 0.138 0.1125 wR2(F ) [all data]

288

Table 5.4. Selected Bond Lengths (Å) and Bond Angles (°) for 5.1b.

C(1)-B(1) 1.516(4) Rh(1)-C(1) 2.218(3) C(1)-B(2) 1.512(4) Rh(1)-B(1) 2.377(3) B(1)-N(1) 1.457(4) Rh(1)-B(2) 2.335(3) B(2)-N(2) 1.472(4) Rh(1)-N(1) 2.334(2) N(1)-N(2) 1.481(3) Rh(1)-N(2) 2.284(2) C(1)-C(2) 1.510(4) Rh(1)-C(21) 2.166(3) B(1)-C(9) 1.585(4) Rh(1)-C(22) 2.185(3) B(2)-C(15) 1.579(4) Rh(1)-C(25) 2.085(3) N(1)-C(3) 1.478(3) Rh(1)-C(26) 2.076(3)

N(2)-C(6) 1.485(3) Rh(1)-CB2N2 plane 1.921(3)

C(1)-B(1)-N(1) 107.1(2) C(1)-B(2)-N(2) 106.7(2) C(1)-B(1)-C(9) 127.2(2) C(1)-B(2)-C(15) 128.6(2) C(9)-B(1)-N(1) 125.7(2) C(15)-B(2)-N(2) 124.6(2) B(1)-N(1)-N(2) 109.3(2) B(2)-N(2)-N(1) 108.6(2) B(1)-N(1)-C(3) 131.5(2) B(2)-N(2)-C(6) 127.0(2) C(3)-N(1)-N(2) 119.3(2) C(6)-N(2)-N(1) 119.6(2)

B(1)-C(1)-B(2) 106.8(2) Σpentagon angles 538.5 C(2)-C(1)-B(2) 126.0(2) C(2)-C(1)-B(1) 125.9(2)

289

Table 5.5. Selected Bond Lengths (Å) and Bond Angles (°) for 5.1e.

C(1)-B(1) 1.528(6) Rh(1)-C(1) 2.217(3) C(1)-B(2) 1.540(6) Rh(1)-B(1) 2.385(4) B(1)-N(1) 1.452(5) Rh(1)-B(2) 2.324(4) B(2)-N(2) 1.454(5) Rh(1)-N(1) 2.309(3) N(1)-N(2) 1.449(4) Rh(1)-N(2) 2.249(3) C(1)-C(2) 1.510(5) Rh(1)-C(6) 2.146(4) B(1)-C(14) 1.582(6) Rh(1)-C(7) 2.182(4) B(2)-C(20) 1.575(6) Rh(1)-C(10) 2.076(4) N(1)-C(3) 1.460(5) Rh(1)-C(11) 2.081(4)

N(2)-C(5) 1.461(5) Rh(1)-CB2N2 plane 1.908(3) C(3)-C(4) 1.526(6) C(4)-C(5) 1.531(6) C(1)-B(1)-N(1) 105.3(3) C(1)-B(2)-N(2) 104.5(3) C(1)-B(1)-C(14) 134.3(4) C(1)-B(2)-C(20) 132.7(4) C(14)-B(1)-N(1) 120.4(4) C(20)-B(2)-N(2) 122.8(4) B(1)-N(1)-N(2) 110.4(3) B(2)-N(2)-N(1) 110.9(3) B(1)-N(1)-C(3) 139.8(3) B(2)-N(2)-C(5) 139.4(4) C(3)-N(1)-N(2) 109.2(3) C(5)-N(2)-N(1) 108.0(3)

B(1)-C(1)-B(2) 107.6(3) Σpentagon angles 538.5 C(2)-C(1)-B(2) 125.2(3) C(2)-C(1)-B(1) 126.2(3)

290

Table 5.6. Selected Bond Lengths (Å) and Bond Angles (°) for 5.2.

C(1)-B(1) 1.511(4) C(18)-B(3) 1.521(4) C(1)-B(2) 1.517(4) C(18)-B(4) 1.510(4) B(1)-N(1) 1.472(3) B(3)-N(3) 1.488(3) B(2)-N(2) 1.489(3) B(4)-N(4) 1.489(4) N(1)-N(2) 1.436(3) N(3)-N(4) 1.439(3) C(1)-C(2) 1.513(3) C(18)-C(19) 1.520(3) B(1)-C(6) 1.571(4) B(3)-C(23) 1.571(4) B(2)-C(12) 1.568(4) B(4)-C(29) 1.572(4) N(1)-C(3) 1.469(3) N(3)-C(20) 1.472(3) N(2)-C(5) 1.478(3) N(4)-C(22) 1.473(3) C(3)-C(4) 1.521(4) C(20)-C(21) 1.519(4) C(4)-C(5) 1.526(4) C(21)-C(22) 1.528(4) Fe(1)-C(1) 2.192(2) Fe(1)-C(18) 2.177(2) Fe(1)-B(1) 2.203(3) Fe(1)-B(3) 2.217(3) Fe(1)-B(2) 2.198(3) Fe(1)-B(4) 2.213(3) Fe(1)-N(1) 1.971(2) Fe(1)-N(3) 1.973(2) Fe(1)-N(2) 1.997(2) Fe(1)-N(4) 1.986(2)

Fe(1)-CB2N2 plane(1) 1.676(3) Fe(1)-CB2N2 plane(2) 1.674(3)

C(1)-B(1)-N(1) 104.6(2) C(18)-B(3)-N(3) 103.8(2) C(1)-B(1)-C(6) 133.7(2) C(18)-B(3)-C(23) 133.9(2) C(6)-B(1)-N(1) 121.7(2) C(23)-B(3)-N(3) 122.2(2) C(1)-B(2)-N(2) 104.3(2) C(18)-B(4)-N(4) 104.6(2) C(1)-B(2)-C(12) 133.7(2) C(18)-B(4)-C(29) 132.6(2) C(12)-B(2)-N(2) 121.6(2) C(29)-B(4)-N(4) 122.8(2) B(1)-N(1)-N(2) 111.0(2) B(3)-N(3)-N(4) 110.8(2) B(1)-N(1)-C(3) 137.0(2) B(3)-N(3)-C(20) 138.8(2) C(3)-N(1)-N(2) 109.7(2) C(20)-N(3)-N(4) 108.9(2) B(2)-N(2)-N(1) 110.0(2) B(4)-N(4)-N(3) 110.0(2) B(2)-N(2)-C(5) 138.9(2) B(4)-N(4)-C(22) 138.8(2) C(5)-N(2)-N(1) 108.7(2) C(22)-N(4)-N(3) 109.0(2) B(1)-C(1)-B(2) 109.5(2) B(3)-C(18)-B(4) 109.6(2)

291

C(2)-C(1)-B(2) 127.5(2) C(19)-C(18)-B(4) 125.1(1) C(2)-C(1)-B(1) 122.9(2) C(19)-C(18)-B(3) 125.2(2)

Σpentagon angles 539.4 Σpentagon angles 538.8 Dihedral angles 1.26(9)

Table 5.7. Selected Bond Lengths (Å) and Bond Angles (°) for 5.3.

C(1)-B(1) 1.522(4) Ru(1)-C(1) 2.296(2) C(1)-B(1) 1.511(4) Ru(1)-B(1) 2.322(3) B(1)-N(1) 1.479(3) Ru(1)-B(2) 2.323(3) B(2)-N(2) 1.487(3) Ru(1)-N(1) 2.127(2) N(1)-N(2) 1.431(3) Ru(1)-N(2) 2.126(2) C(1)-C(2) 1.516(3) Ru(1)-C(6) 2.194(2) B(1)-C(16) 1.573(4) Ru(1)-C(7) 2.185(2) B(2)-C(22) 1.578(4) Ru(1)-C(8) 2.162(2) N(1)-C(3) 1.468(3) Ru(1)-C(9) 2.154(2) N(2)-C(5) 1.469(3) Ru(1)-C(10) 2.162(3)

C(3)-C(4) 1.521(4) Ru(1)-CB2N2 plane 1.836(2) C(4)-C(5) 1.527(3) Ru(1)-Cp* plane 1.797(2)

C(1)-B(1)-N(1) 104.5(2) C(1)-B(2)-N(2) 104.5(2) C(1)-B(1)-C(16) 133.8(2) C(1)-B(2)-C(22) 133.6(2) C(16)-B(1)-N(1) 121.6(2) C(22)-B(2)-N(2) 121.9(2) B(1)-N(1)-N(2) 110.7(2) B(2)-N(2)-N(1) 110.5(2) B(1)-N(1)-C(3) 139.2(2) B(2)-N(2)-C(5) 138.9(2) C(3)-N(1)-N(2) 109.0(2) C(5)-N(2)-N(1) 108.9(2)

B(1)-C(1)-B(2) 109.2(2) Σpentagon angles 539.4

C(2)-C(1)-B(2) 125.1(2) CB2N2-Ru(1)-Cp* 174.4 C(2)-C(1)-B(1) 125.5(2)

292

Table 5.8. Selected Bond Lengths (Å) and Bond Angles (°) for 5.4.

C(1)-B(1) 1.509(4) Zr(1)-C(1) 2.597(3) C(1)-B(2) 1.517(4) Zr(1)-B(1) 2.715(3) B(1)-N(1) 1.467(4) Zr(1)-B(2) 2.757(3) B(2)-N(2) 1.489(4) Zr(1)-N(1) 2.372(2) N(1)-N(2) 1.467(3) Zr(1)-N(2) 2.357(2) C(1)-C(2) 1.518(4) Zr(1)-Cl(1) 2.430(1) B(1)-C(16) 1.572(4) Zr(1)-Cl(2) 2.423(1) B(2)-C(22) 1.569(4) Zr(1)-C(6) 2.525(3) N(1)-C(3) 1.473(3) Zr(1)-C(7) 2.543(3) N(2)-C(5) 1.481(3) Zr(1)-C(8) 2.531(3) C(3)-C(4) 1.510(4) Zr(1)-C(9) 2.561(3) C(4)-C(5) 1.508(4) Zr(1)-C(10) 2.524(3)

Zr(1)-CB2N2 plane 2.156(3) Zr(1)-Cp* plane 2.231(3)

C(1)-B(1)-N(1) 105.8(2) C(1)-B(2)-N(2) 104.6(2) C(1)-B(1)-C(16) 132.8(2) C(1)-B(2)-C(22) 132.6(2) C(16)-B(1)-N(1) 121.3(2) C(22)-B(2)-N(2) 122.8(2) B(1)-N(1)-N(2) 108.8(2) B(2)-N(2)-N(1) 108.0(2) B(1)-N(1)-C(3) 132.3(2) B(2)-N(2)-C(5) 132.5(2) C(3)-N(1)-N(2) 109.3(2) C(5)-N(2)-N(1) 107.5(2) B(1)-C(1)-B(2) 104.9(2) Cl(2)-Zr(1)-Cl(1) 97.51(4)

C(2)-C(1)-B(2) 128.9(2) CB2N2-Zr(1)-Cp* 141.0

C(2)-C(1)-B(1) 122.3(2) Σpentagon angles 532.3

293

Table 5.9. Selected Bond Lengths (Å) and Bond Angles (°) for 5.6c.

O(1)-B(1) 1.396(1) B(1)-O(1)-B(1’) 108.7(1) B(1)-N(1) 1.414(2) O(1)-B(1)-N(1) 108.7(1) N(1)-N(1’) 1.447(2) O(1)-B(1)-C(8) 122.5(1) B(1)-C(8) 1.560(2) C(8)-B(1)-N(1) 128.8(1) N(1)-C(2) 1.424(1) B(1)-N(1)-N(1’) 106.8(6)

C-CBPh 1.383(2) - 1.401(2) B(1)-N(1)-C(2) 132.1(9)

C-CNPh 1.380(2) - 1.390(2) C(2)-N(1)-N(1’) 119.2(6)

Σpentagon angles 539.7

Symmetry transformations used to generate equivalent atoms: -x, y, -z+1/2

Table 5.10. Selected Bond Lengths (Å) and Bond Angles (°) for trans-5.7.

C(1)-B(1) 1.574(2) B(1)-C(1)-B(2) 101.9(1) C(1)-B(2) 1.571(2) C(2)-C(1)-B(2) 119.9(1) B(1)-N(1) 1.437(2) B(1)-C(1)-C(2) 120.7(1) B(2)-N(1’) 1.436(2) C(1)-B(1)-N(1) 108.9(1) N(1)-N(1’) 1.480(2) C(1)-B(1)-C(3) 126.9(1) C(1)-C(2) 1.529(2) C(3)-B(1)-N(1) 124.2(2) B(1)-C(3) 1.565(2) C(1)-B(2)-N(1’) 109.2(1) B(2)-C(9) 1.560(2) C(1)-B(2)-C(9) 126.9(1)

C-CBPh 1.375(3) - 1.401(2) C(9)-B(2)-N(1’) 124.1(1) B(1)-N(1)-B(2’) 140.9(1)

Σpentagon angles 539.1 B(1)-N(1)-N(1’) 109.5(1) B(2)-N(1’)-N(1) 109.6(1)

Symmetry transformations used to generate equivalent atoms: -x+1, -y, -z

294

Table 5.11. Selected Bond Lengths (Å) and Bond Angles (°) for 5.9(tmeda)2.

C(1)-B(1) 1.512(4) K(1)-N(1) 2.889(2) C(1)-B(2) 1.494(4) K(1)-N(1’) 2.820(2) B(1)-N(1) 1.482(4) K(1)-N(2) 2.809(3) B(2)-N(1’) 1.474(4) K(1)-N(3) 3.051(2) N(1)-N(1’) 1.442(3) K(1)-B(1) 3.376(3) C(1)-C(2) 1.516(4) K(1)-B(2) 3.268(3) B(1)-C(3) 1.585(4) K(1)-B(1’) 3.268(3) B(2)-C(9) 1.585(4) K(1)-B(2’) 3.357(3) N(2)-C(15) 1.456(4) K(1)-C(1) 3.691(3) N(2)-C(17) 1.445(4) K(1)-C(1’) 3.514(3) N(2)-C(18) 1.464(4) K(1)···K(1’) 5.525(1) N(3)-C(16) 1.465(3) N(3)-C(19) 1.460(4) C(15)-C(16) 1.496(5)

N(3)-C(20) 1.465(5) C-CPh 1.370(5) – 1.405(5) B(1)-N(1)-N(1’) 108.8(2) B(1)-C(1)-B(2) 104.7(2) B(2)-N(1’)-N(1) 107.9(2) C(2)-C(1)-B(2) 128.4(2) B(1)-N(1)-B(2’) 143.3(2) B(1)-C(1)-C(2) 126.2(2) C(1)-B(1)-N(1) 108.5(2) C(1)-B(2)-N(1’) 109.6(2) C(1)-B(1)-C(3) 129.1(2) C(1)-B(2)-C(9) 129.4(2) C(3)-B(1)-N(1) 122.1(1) C(9)-B(2)-N(1’) 120.5(2) K(1)-N(1)-K(1’) 150.78(8) N(1)-K(1)-N(1’) 29.22(6)

Σpentagon angles 539.5(2) N(2)-K(1)-N(3) 61.71(7)

Symmetry transformations used to generate equivalent atoms: -x+2, -y+1, -z

295

Table 5.12. Selected Bond Lengths (Å) and Bond Angles (°) for 5.10.

C(1)-B(1) 1.542(10) C(2)-B(3) 1.582(9) C(1)-B(2) 1.565(9) C(2)-B(4) 1.558(9) B(1)-N(1) 1.437(9) B(3)-N(1) 1.400(9) B(2)-N(2) 1.440(9) B(4)-N(2) 1.462(9) C(1)-C(3) 1.549(9) C(2)-C(4) 1.526(8) B(1)-C(25) 1.578(9) B(3)-C(37) 1.591(9) B(2)-C(31) 1.608(9) B(4)-C(43) 1.594(9) N(1)⋅⋅⋅N(2) 2.548(7) Ru(1)⋅⋅⋅Ru(2) 3.249(1) Ru(1)-N(1) 2.128(5) Ru(1)-C(5) 2.216(6) Ru(1)-N(2) 2.118(5) Ru(1)-C(6) 2.243(6) Ru(1)-B(1) 2.520(7) Ru(1)-C(7) 2.218(6) Ru(1)-B(2) 2.546(7) Ru(1)-C(8) 2.211(7) Ru(1)-B(3) 2.500(8) Ru(1)-C(9) 2.224(6) Ru(1)-B(4) 2.488(7) Ru(2)-C(15) 2.198(7) Ru(1)-C(1) 2.402(5) Ru(2)-C(16) 2.179(7) Ru(1)-C(2) 2.548(7) Ru(2)-C(17) 2.160(8) Ru(2)-N(1) 2.129(5) Ru(2)-C(18) 2.184(6) Ru(2)-N(2) 2.132(6) Ru(2)-C(19) 2.189(6)

Ru(2)-B(3) 2.378(6) C-CPh 1.361(11) - 1.414(10)

Ru(2)-B(4) 2.351(7) C-CCp* 1.389(11) - 1.451(10)

Ru(2)-C(2) 2.257(6) C-CCp*Me 1.479(11) - 1.535(11)

Ru(1)-plane CB4N2 1.578(4) Ru(1)-plane Cp* 1.862(4)

Ru(2)-plane CB2N2 1.592(4) Ru(2)-plane Cp* 1.816(4) B(1)-C(1)-B(2) 123.8(6) B(3)-C(2)-B(4) 130.1(6) C(3)-C(1)-B(2) 117.9(6) C(4)-C(2)-B(4) 114.4(6) B(1)-C(1)-C(3) 117.7(6) B(3)-C(2)-C(4) 115.4(6) C(1)-B(1)-N(1) 113.1(6) C(2)-B(3)-N(1) 110.7(6) C(1)-B(1)-C(25) 124.8(6) C(2)-B(3)-C(37) 126.6(6) C(25)-B(1)-N(1) 122.1(6) C(37)-B(3)-N(1) 122.7(6) C(1)-B(2)-N(2) 114.1(6) C(2)-B(4)-N(2) 112.1(6) C(1)-B(2)-C(31) 123.4(6) C(2)-B(4)-C(43) 127.0(6) C(31)-B(2)-N(2) 122.0(5) C(43)-B(4)-N(2) 120.8(6) B(1)-N(1)-B(3) 166.4(6) B(2)-N(2)-B(4) 172.1(6)

CB4N2-Ru(1)-Cp* 2.7(2) CB2N2-Ru(2)-Cp* 4.7(2)

296

Table 6.1. Selected Data and Structure Refinement Parameters for 6.1a, 6.1b, 6.1c, and 6.2 · thf.

6.1a 6.1b 6.1c 6.2 · thf

Empirical formula C14H17B2LiN4 C14H17B2N4Na C14H17B2KN4 C26H37B2N4ORh Formula weight 269.88 285.93 302.04 546.13 Crystal system monoclinic monoclinic orthorhombic Triclinic

Space group P21/c P21/c P212121 P -1 a (Å) 13.464 (2) 13.419(7) 7.746(4) 9.818(2) b (Å) 10.0183 (15) 10.143(5)Å 11.973(4) 10.130(3) c (Å) 11.2470 (17) 99.82(3) 16.742(9) 13.284(3) α (deg) 90 90 90 77.906(13) β (deg) 100.283 (3) 11.526(8) 90 77.701(13) γ (deg) 90 90 90 87.761(15) V (A3) 1492.7 (4) 1545.8(15) 1552.7(13) 1262.2(5) Z 4 4 4 2 3 dcalcd (g cm ) 1.201 1.229 1.292 1.437

2θmax (deg) 52.94 27.5 25.03 27.6 μ (mm-1) 0.071 0.098 0.338 0.703

Independent reflections 3072 (Rint = 0.0922) 3520 (Rint = 0.076) 2711 (Rint = 0.0256) 5750 (Rint = 0.037) Data/restraints/parameters 3072 / 1 / 239 3520 / 0 / 193 2711 / 0 / 193 5750 / 0 / 307 GOF on F2 1.038 1.01 1.056 1.03

R1(F) [I > 2σ(I)] 0.0621 0.060 0.0379 0.038 2 wR2(F ) [all data] 0.1995 0.182 0.1014 0.098

297

Table 6.2. Selected Data and Structure Refinement Parameters for 6.3a(CH3CN)3, 6.3c, 6.5(thf) and 6.6.

6.3a(CH3CN)3 6.3c 6.5(thf) 6.6

Empirical formula C38H41B3LiN7 C32H32B3KN4 C19H27B2LiN4O C18H28B4N8 Formula weight 635.15 544.15 356.01 399.72 Crystal system monoclinic Monoclinic Monoclinic Triclinic

Space group C2/c P21/c P21/n P -1 a (Å) 32.087(12) 9.111(4) 13.149(5) 7.240(3) b (Å) 11.022(5) 17.582(4) 9.299(4) 9.157(6) c (Å) 23.883(7) 18.059(6) 16.145(6) 16.983(12) α (deg) 90 90 90 75.88(3) β (deg) 118.10(2) 91.344(14) 100.17(3) 83.19(4) γ (deg) 90 90 90 76.55(4) V (A3) 7451(5) 2892.1(17) 1943.1(13) 1059.7(11) Z 8 4 4 2 3 dcalcd (g cm ) 1.132 1.250 1.217 1.253

2θmax (deg) 0.067 27.5 27.5 25.0 μ (mm-1) 25.0 0.212 0.074 0.077

Independent reflections 6573 (Rint = 0.046) 6566 (Rint = 0.034) 4398 (Rint = 0.027) 3693 (Rint = 0.031) Data/restraints/parameters 6573 / 0 / 447 6566 / 0 / 366 4398 / 0 / 247 3693 / 0 / 277 GOF on F2 1.02 1.00 1.02 1.01

R1(F) [I > 2σ(I)] 0.062 0.046 0.047 0.046 2 wR2(F ) [all data] 0.177 0.126 0.126 0.146

298

Table 6.3. Selected Bond Lengths (Å) and Bond Angles (°) for 6.1a.

N(1)-N(2) 1.435(3) N(3)-N(4) 1.439(3) B(1)-N(1) 1.400(4) B(2)-N(2) 1.407(4) B(1)-N(3) 1.443(4) B(2)-N(3) 1.446(4) B(1)-C(3) 1.572(4) B(2)-C(9) 1.571(4) N(1)-C(1) 1.448(3) N(4)-C(2) 1.465(4) N(2)-Li(1) 2.068(5) N(4)-Li(1”) 2.100(5)

N(2)-Li(1’) 2.013(5) C-Caryl 1.29(3) - 1.44(3) N(3)-B(1)-N(1) 105.2(2) N(3)-B(2)-N(2) 109.3(2) N(3)-B(1)-C(3) 125.9(2) N(3)-B(2)-C(9) 125.3(2) C(3)-B(1)-N(1) 128.7(3) C(9)-B(2)-N(2) 125.4(2) B(1)-N(1)-N(2) 112.5(2) B(1)-N(3)-B(2) 108.0(2) B(1)-N(1)-C(1) 130.4(2) B(1)-N(3)-N(4) 126.5(2) C(1)-N(1)-N(2) 116.6(2) B(2)-N(3)-N(4) 124.8(2) N(1)-N(2)-B(2) 105.1(2) N(3)-N(4)-C(2) 108.4(2) Li(1)-N(2)-Li(1’) 78.2(2) N(2)-Li(1)-N(2”) 101.8(2) N(1)-N(2)-Li(1’) 129.6(2) N(1)-N(2)-Li(1) 116.2(2) B(2)-N(2)-Li(1’) 111.0(2) B(2)-N(2)-Li(1) 115.5(2) N(3)-N(4)-Li(1”) 130.1(2) C2-N4-Li(1”) 106.0(2) N(2”)-Li(1”)-N(4) 151.2(3) N(2*)-Li(1”)-N(4) 100.8(2)

Σpentagon angle 540.0

Primed atoms are related to unprimed ones via the crystallographic inversion center: 0, 0, 0

Double-primed atoms are related to unprimed ones via the symmetry operation: -x, 1/2+y, 1/2–z

Starred atoms are related to unprimed ones via the symmetry operation: -x, -1/2+y, 1/2–z

299

Table 6.4. Selected Bond Lengths (Å) and Bond Angles (°) for 6.1b.

N(1)-N(2) 1.433(3) N(3)-N(4) 1.437(3) B(1)-N(1) 1.411(4) B(2)-N(2) 1.400(4) B(1)-N(3) 1.446(4) B(2)-N(3) 1.448(4) B(1)-C(3) 1.561(4) B(2)-C(9) 1.587(4) N(1)-C(1) 1.444(3) N(4)-C(2) 1.458(4) N(2)-Na(1) 2.350(3) N(4)-Na(1”) 2.449(3)

N(2)-Na(1’) 2.416(3) C-Caryl 1.370(4) - 1.409(4) N(3)-B(1)-N(1) 104.7(2) N(3)-B(2)-N(2) 109.9(2) N(3)-B(1)-C(3) 127.1(3) N(3)-B(2)-C(9) 126.0(2) C(3)-B(1)-N(1) 128.2(3) C(9)-B(2)-N(2) 124.1(2) B(1)-N(1)-N(2) 112.7(2) B(1)-N(3)-B(2) 127.4(2) B(1)-N(1)-C(1) 131.1(2) B(1)-N(3)-N(4) 124.1(2) C(1)-N(1)-N(2) 115.9(2) B(2)-N(3)-N(4) 110.8(2) N(1)-N(2)-B(2) 104.9(2) N(3)-N(4)-C(2) 107.8(2) Na(1)-N(2)-Na(1’) 84.32(8) N(2)-Na(1)-N(2”) 95.68(8) N(1)-N(2)-Na(1’) 122.3(2) N(1)-N(2)-Na(1) 124.1(2) B(2)-N(2)-Na(1’) 110.8(2) B(2)-N(2)-Na(1) 109.0(2) N(3)-N(4)-Na(1”) 120.9(2) C(2)-N(4)-Na(1”) 107.5(2) N(2*)-Na(1”)-N(4) 89.32(9) N(2”)-Na(1”)-N(4) 151.6(1)

Σpentagon angle 540.0

Primed atoms are related to unprimed ones via the crystallographic inversion center: 0, 0, 0

Double-primed atoms are related to unprimed ones via the symmetry operation: -x, 1/2+y, 1/2–z

Starred atoms are related to unprimed ones via the symmetry operation: -x, -1/2+y, 1/2–z

300

Table 6.5. Selected Bond Lengths (Å) and Bond Angles (°) 6.1c.

N(1)-N(2) 1.430(3) N(3)-N(4) 1.440(3) B(1)-N(1) 1.410(4) B(2)-N(2) 1.401(4) B(1)-N(3) 1.454(3) B(2)-N(3) 1.470(4) B(1)-C(3) 1.574(4) B(2)-C(9) 1.579(4) N(1)-C(1) 1.455(4) N(4)-C(2) 1.468(4) N(2’)-K(1) 2.681(3) N(1)-K(1) 3.009(2) N(4)-K(1’) 2.889(3) N(2)-K(1) 2.800(2) C(3)-K(1)’ 3.127(3) B(2)-K(1) 3.149(3)

N(3)-B(1)-N(1) 104.6(2) N(3)-B(2)-N(2) 109.7(2) N(3)-B(1)-C(3) 129.7(2) N(3)-B(2)-C(9) 126.3(2) C(3)-B(1)-N(1) 125.6(2) C(9)-B(2)-N(2) 124.0(2) B(1)-N(1)-N(2) 113.4(2) B(1)-N(3)-B(2) 107.1(2) B(1)-N(1)-C(1) 131.1(2) B(1)-N(3)-N(4) 125.4(2) C(1)-N(1)-N(2) 114.6(2) B(2)-N(3)-N(4) 127.3(2) N(1)-N(2)-B(2) 104.9(2) N(3)-N(4)-C(2) 110.5(2) K(1’)-N(2’)-K(1) 108.6(1) N(2’)-K(1)-N(2) 122.86(7) N(1’)-N(2’)-K(1) 126.6(2) N(2’)-K(1)-N(1) 104.38(7) B(2’)-N(2’)-K(1) 125.8(2) N(2’)-K(1)-B(2) 104.32(8) N(3)-N(4)-K(1’) 91.4(1) N(2’)-K(1)-N(4”) 125.59(7) C(3)-K(1)-N(4) 66.26(7) N(2’)-K(1)-C(3”) 122.73(7)

Σpentagon angle 539.7

Primed atoms are related to unprimed ones via the symmetry operation: x-1/2, -y-1/2, -z-1 Double-primed atoms are related to unprimed ones via the symmetry operation: x+1/2, -y-1/2,-z-1

301

Table 6.6. Selected Bond Lengths (Å) and Bond Angles (°) for 6.2 · thf.

N(1)-N(2) 1.447(3) N(3)-N(4) 1.435(3) B(1)-N(1) 1.406(4) B(2)-N(2) 1.460(4) B(1)-N(3) 1.446(4) B(2)-N(3) 1.419(4) B(1)-C(3) 1.566(4) B(2)-C(9) 1.584(4) N(1)-C(1) 1.444(4) N(4)-C(2) 1.453(4) N(2)-Rh(1) 2.141(2) N(2)-Rh(1’) 2.168(2) C(15)-Rh(1) 2.133(2) C(19)-Rh(1) 2.153(3) C(16)-Rh(1) 2.115(3) C(20)-Rh(1) 2.115(3)

N(3)-B(1)-N(1) 105.7(2) N(3)-B(2)-N(2) 108.7(2) N(3)-B(1)-C(3) 129.2(3) N(3)-B(2)-C(9) 123.3(3) C(3)-B(1)-N(1) 125.0(3) C(9)-B(2)-N(2) 128.0(3) B(1)-N(1)-N(2) 111.9(2) B(1)-N(3)-B(2) 109.3(2) B(1)-N(1)-C(1) 131.5(2) B(1)-N(3)-N(4) 124.3(2) C(1)-N(1)-N(2) 116.1(2) B(2)-N(3)-N(4) 126.4(2) N(1)-N(2)-B(2) 104.3(2) N(3)-N(4)-C(2) 111.0(3) Rh(1)-N(2)-Rh(1’) 98.33(9) N(2)-Rh(1)-N(2’) 81.67(8) Rh(1)-N(2)-N(1) 118.2(1) C(15)-Rh(1)-C(20) 82.2(1) Rh(1)-N(2)-B(1) 112.8(2) C(16)-Rh(1)-C(19) 81.4(1) C(15)-Rh(1)-N(2) 93.6(1) C(15)-Rh(1)-N(2’) 164.5(1) C(16)-Rh(1)-N(2) 95.0(1) C(16)-Rh(1)-N(2’) 156.1(1) C(19)-Rh(1)-N(2) 168.8(1) C(19)-Rh(1)-N(2’) 97.2(1) C(20)-Rh(1)-N(2) 152.9(1) C(20)-Rh(1)-N(2’) 95.3(1)

Σpentagon angle 539.9

Primed atoms are related to unprimed ones via the symmetry operation: -x+1, -y, -z+1

302

Table 6.7. Selected Bond Lengths (Å) and Bond Angles (°) for 6.3a(CH3CN)3.

N(1)-N(2) 1.448(3) N(3)-N(4) 1.440(3) B(1)-N(1) 1.412(4) B(2)-N(2) 1.420(4) B(1)-N(3) 1.422(4) B(2)-N(3) 1.452(4) B(1)-C(3) 1.573(4) B(2)-C(9) 1.581(4) N(1)-C(1) 1.450(3) N(4)-C(2) 1.465(3) N(2)-B(3) 1.579(4) N(4)-Li(1) 2.013(7) B(3)-C(15) 1.634(4) N(5)-Li(1) 1.979(5) B(3)-C(21) 1.646(4) N(6)-Li(1) 2.073(6) B(3)-C(27) 1.653(3) N(7)-Li(1) 2.028(7)

N(3)-B(1)-N(1) 105.9(2) N(3)-B(2)-N(2) 107.6(2) N(3)-B(1)-C(3) 129.5(2) N(3)-B(2)-C(9) 120.4(2) C(3)-B(1)-N(1) 124.5(3) C(9)-B(2)-N(2) 131.9(2) B(1)-N(1)-N(2) 110.9(2) B(1)-N(3)-B(2) 109.4(2) B(1)-N(1)-C(1) 126.3(2) B(1)-N(3)-N(4) 125.5(2) C(1)-N(1)-N(2) 120.5(2) B(2)-N(3)-N(4) 124.8(2) N(1)-N(2)-B(2) 106.1(2) N(3)-N(4)-C(2) 111.2(2) B(3)-N(2)-N(1) 134.3(2) N(3)-N(4)-Li(1) 121.2(2) B(3)-N(2)-B(2) 119.3(2) C(2)-N(4)-Li(1) 116.6(3) N(2)-B(3)-C(15) 108.0(2) N(4)-Li(1)-N(5) 110.8(3) N(2)-B(3)-C(21) 111.9(2) N(4)-Li(1)-N(6) 95.6(2) N(2)-B(3)-C(27) 107.9(2) N(4)-Li(1)-N(7) 119.9(3)

Σpentagon angle 539.9

303

Table 6.8. Selected Bond Lengths (Å) and Bond Angles (°) for 6.3c.

N(1)-N(2) 1.447(2) N(3)-N(4) 1.439(2) B(1)-N(1) 1.405(2) B(2)-N(2) 1.414(2) B(1)-N(3) 1.437(3) B(2)-N(3) 1.452(2) B(1)-C(3) 1.575(3) B(2)-C(9) 1.580(3) N(1)-C(1) 1.451(2) N(4)-C(2) 1.466(3) N(2)-B(3) 1.576(3) N(4)-K(1) 2.833(2) B(3)-C(15) 1.658(3) K(1)-C(9) 3.042(2) B(3)-C(21) 1.659(3) K(1)-C(20’) 2.993(2) B(3)-C(27) 1.640(3) K(1)-C(21’) 3.108(2)

N(3)-B(1)-N(1) 105.87(16) N(3)-B(2)-N(2) 108.06(16) N(3)-B(1)-C(3) 128.53(16) N(3)-B(2)-C(9) 121.82(16) C(3)-B(1)-N(1) 125.49(17) C(9)-B(2)-N(2) 129.88(16) B(1)-N(1)-N(2) 111.09(14) B(1)-N(3)-B(2) 108.72(14) B(1)-N(1)-C(1) 127.53(15) B(1)-N(3)-N(4) 126.61(15) C(1)-N(1)-N(2) 120.78(13) B(2)-N(3)-N(4) 124.59(15) N(1)-N(2)-B(2) 106.21(13) N(3)-N(4)-C(2) 110.80(15) B(3)-N(2)-N(1) 120.9(1) N(3)-N(4)-Li(1) 109.8(1) B(3)-N(2)-B(2) 132.7(1) C(2)-N(4)-Li(1) 123.3(1) N(2)-B(3)-C(15) 108.3(1) N(4)-K(1)-C(9) 63.75(4) N(2)-B(3)-C(21) 108.9(1) N(4)-K(1)-C(20’) 134.56(5) N(2)-B(3)-C(27) 110.8(1) N(4)-K(1)-C(21’) 98.36(4)

Σpentagon angle 539.9

Primed atoms are related to unprimed ones via the symmetry operation: x, -y+1/2, z-1/2 Double-primed atoms are related to unprimed ones via the symmetry operation: x, -y+1/2, z+1/2

304

Table 6.9. Selected Bond Lengths (Å) and Bond Angles (°) for 6.5(thf).

N(1)-N(2) 1.475(2) N(1)-C(1) 1.462(2) N(3)-N(4) 1.445(2) N(2)-C(2) 1.445(2) B(1)-N(2) 1.441(2) N(4)-C(3) 1.451(2) B(1)-N(3) 1.408(2) B(1)-C(4) 1.579(3) B(2)-N(1) 1.562(2) B(2)-C(10) 1.636(2) B(2)-N(3) 1.556(2) N(1)-Li(1) 2.189(3) B(2)-N(4’) 1.566(2) N(3)-Li(1) 2.326(3) O(1)-Li(1) 1.901(3) N(4)’-Li(1) 2.034(3) C(10’)···Li(1) 2.775(3)

N(2)-N(1)-B(2) 104.4(1) N(3)-B(1)-N(2) 107.9(1) C(1)-N(1)-B(2) 116.1(1) N(2)-B(1)-C(4) 120.6(1) C(1)-N(1)-N(2) 109.7(1) N(3)-B(1)-C(4) 131.5(1) B(1)-N(2)-N(1) 108.7(1) N(1)-B(2)-N(4’) 107.7(1) N(1)-N(2)-C(2) 114.3(1) N(3)-B(2)-N(4’) 106.0(1) B(1)-N(2)-C(2) 125.2(1) N(4’)-B(2)-C(10) 116.2(1) B(1)-N(3)-B(2) 110.0(1) N(3)-B(2)-N(1) 98.4(1) B(1)-N(3)-N(4) 133.0(1) N(1)-B(2)-C(10) 115.8(1) N(4)-N(3)-B(2) 112.7(1) N(3)-B(2)-C(10) 110.8(1) N(3)-N(4)-B(2’) 107.7(1) N(1)-Li(1)-N(3) 63.0(1) N(3)-N(4)-C(3) 114.8(1) N(4’)-Li(1)-N(3) 69.4(1) B(2’)-N(4)-C(3) 124.1(1) N(4’)-Li(1’)-N(1) 73.3(1)

Primed atoms are related to unprimed ones via the symmetry operation: -x, -y, -z+1

305

Table 6.10. Selected Bond Lengths (Å) and Bond Angles (°) for 6.6.

Molecule One Molecule Two N(1)-N(2) 1.436(2) N(5)-N(6) 1.441(2) N(3)-N(4) 1.438(3) N(7)-N(8) 1.438(3) B(1)-N(2) 1.428(4) B(3)-N(6) 1.423(4) B(1)-N(3) 1.434(3) B(3)-N(7) 1.434(3) B(2)-N(1) 1.429(3) B(4)-N(5) 1.427(3) B(2)-N(3) 1.434(2) B(4)-N(7) 1.440(3) B(2)-N(4’) 1.436(3) B(4)-N(8’) 1.427(3) N(1)-C(1) 1.446(3) N(5)-C(10) 1.449(3) N(2)-C(2) 1.455(3) N(6)-C(11) 1.453(3) N(4)-C(3) 1.463(3) N(8)-C(12) 1.449(3) B(1)-C(4) 1.570(3) B(4)-C(13) 1.566(3)

N(3)-B(1)-N(2) 106.5(2) N(7)-B(3)-N(6) 106.8(2) N(2)-B(1)-C(4) 123.7(2) N(6)-B(3)-C(13) 122.8(2) N(3)-B(1)-C(4) 129.5(2) N(7)-B(3)-C(13) 130.4(2) N(3)-B(2)-N(1) 107.9(2) N(7)-B(4)-N(5) 108.4(2) N(1)-B(2)-N(4’) 126.4(2) N(5)-B(4)-N(8’) 127.5(2) N(3)-B(2)-N(4’) 125.7(2) N(3)-B(4)-N(8’) 124.2(2) N(2)-N(1)-B(2) 107.2(2) N(6)-N(5)-B(4) 106.7(2) C(1)-N(1)-B(2) 125.9(2) C(10)-N(5)-B(4) 125.8(2) C(1)-N(1)-N(2) 115.7(2) C(10)-N(5)-N(6) 114.3(2) B(1)-N(2)-N(1) 109.2(2) B(3)-N(6)-N(5) 109.5(2) B(1)-N(2)-C(2) 127.8(2) B(3)-N(6)-C(11) 126.7(2) N(1)-N(2)-C(2) 114.7(2) N(5)-N(6)-C(11) 114.8(2) B(1)-N(3)-B(2) 108.7(2) B(3)-N(7)-B(4) 108.3(2) B(1)-N(3)-N(4) 128.8(2) B(3)-N(7)-N(8) 129.6(2) N(4)-N(3)-B(2) 122.4(2) N(8)-N(7)-B(4) 122.1(2) N(3)-N(4)-B(2’) 111.9(2) N(7)-N(8)-B(4’) 113.7(2) B(2’)-N(4)-C(3) 118.5(2) B(4’)-N(8)-C(12) 121.8(2) N(3)-N(4)-C(3) 112.1(2) N(7)-N(8)-C(12) 114.6(2)

Σpentagon angle 539.5 Σpentagon angle 539.5

Σhexagon angle 720 Σhexagon angle 720

Primed atoms are related to unprimed ones via the symmetry operation: -x+1, -y+1, -z+1